Physiological Aspects of Crop Productivity

Physiological Aspects
of Crop Productivity

Proceedings of the 15th Colloquium of the International Potash Institute held in

Wageningen/The Netherlands 1980
Copyright: International Potash Institute
P.O. Box 41
CH-3048 Worblaufen-Bern/Switzerland
Phone (031) 58 53 73, Telex 33430 ipibe ch

Printed by: 'Der Bund' AG, Bern/Switzerland

 Contents

Page
Opening Session

van Diest, A. Soils and crops in the Netherlands - Poten-

tialities and constraints 9

Bijloo, J.D. Organization of the agricultural research in

the Netherlands 17

Ist Session Breeding for higher yields

Lapton, F.G.H. Breeding for higher yields 27

Woolhouse, H. W. The possibilities for modification of' the

pattern of photosynthetic assimilation of
CO, in relation to the growth and yield of
crops 37

Mengel, K. Assimilate transport through phloem tissue 51

Stoy, V. Grain filling and the properties of the sink 65

Arnon. 1. Co-ordinator's report on the Ist session 77

2nd Session Hormonal regulation of yield formation

Michael, G. and Beringer, H. The role of hormones in yield formation 85

Bruinsma, J. The endogenous hormonal pattern and its

interference by exogenous plant growth
regulators 117

5
Mizrahi, Y. The role of plant hormones in plant adapta-
tion to stress conditions 125

Tromp, J. and Wertheim, S.J. Synthetic growth regulators: mode of action

and application in fruit production 137

Cooke, G.W. Co-ordinator's report on the 2nd session 151

3rd Session The role of nutrients in yield formation -

Beringer, H. Nutritional and environmental effects on

yield formation 155

Krauss, A. Influence of nitrogen nutrition on tuber

initiation of potatoes 175

Haeder, H.E. Effect of potassium nutrition on sink inten-

sity and duration 185

Walker, D.A. Regulation of starch synthesis in leaves-The

role of orthophosphate 195

Mengel, K. Co-ordinator's Report on the 3rd session 209

4th Session The use of models in crop physiology

Penning de Vries, F.W.T. Simulation models of growth of crops,

particularly under nutrient stress 213

Malet, Ph. Modelling in Mediterranean area: adjust-

ment problems of crop models 227

Hansen, G.K. A model for temperate European vegetative

crops 239

Wood, J. The mathematical expression of crop re-

sponse to inputs 263

Laudelout, H. Co-ordinator's report on the 4th session 273

6
 Chairman of the Colloquium:
Prof. Dr. A. van Diest, Dept. of Soils and Fertil-
izers, Agricultural University, Wageningen/The
Netherlands; Member of the Scientific Board of
the International Potash Institute

Opening Session
Soils and Crops in the Netherlands -
Potentialities and Constraints
A. van Diest, Department of Soils and Fertilizers, Agricultural University, Wageningen/The
Netherlands; Member of the Scientific Board of the International Potash Institute; Chairman
of the 15th [.P.I.-Colloquium-

The surface area of the Netherlands is 3.3 x 10 ha, of which in 1975 a little over
2 x 10f6 ha were in agricultural use. The area under cultivation has been decreasing
steadily since the second world war. Some data are shown in Table 1.

Table I. Land use in the Netherlands, ha x 10 -

Type of use 1950 1975
G rassland ..................................... .............. . 13 17 1286
A rable land ................................................... 898 675
Horticultural land .............................................. 122 114
T otal
.. . .................. .............................. . 2337 20 75

Percentagewise, the grassland area has decreased little in this period. There are two
main reasons for this absence of decline. First, a fairly large percentage of our soils
can hardly be used for anything but grass production. The groundwater tables are
too high for any crop demanding a deep soil for normal root development. These
high groundwater tables are often mandatory for the maintenance of our lowmoor
peat soils. Without them, these soils would oxidize away even faster than they already
do now.
The second reason has to do with the guaranteed prices obtained by our dairy farmers
for their milk. Many Dutch farmers are not adventurous enough to accept the risk of
making a fortune one year and going broke the next year, for instance by growing
potatoes. They prefer a rather steady income afforded them by the guaranteed milk
price system, even when it means making long days seven days a week with hourly
wages not far above and often below our minimum wage level, which, by the way, is
said to be the highest in the world.
At the present time, the system of guaranteed milk prices within the European Common
Market is under heavy fire, and it seems unlikely that in the future Dutch dairy farmers
can count on receiving a fixed price for every litre of milk they deliver.
In the past, small farmers in this country grew fodder crops to be fed to a small dairy
herd consisting, in many cases, of not more than ten head of cattle. This system of
mixed farming was practiced on our sandy soils which lie to the east and south of a
NE-SW diagonal over the map of the Netherlands. To the north and west of it, we
find most of our young marine loams and lowmoor peat soils.
* Prof. Dr. A. van Diest, Dept. of Soils and Fertilizers. Agricultural University, De Dreijen 3,
NL-6703 BC Wageningen/The Netherlands

9
Originally, many of these sandy soils were unfit for even the lowliest type of subsistence
farming, since they lack acceptable levels of natural fertility and water-holding capac-
ity. On the other hand, these relatively high and dry soils offered the necessary pro-
tection against the destructive forces of sea- and riverwater, and for that reason they
were still in relatively high demand to live on.
For many centuries, the sparse quantities of nutrients present in these soils were
systematically concentrated on small stretches of land around the settlements. In this
way, the soils near the villages were continually enriched at the expense of the soils
lying farther away from these villages. Sheep and pigs served in the transfer of the
nutrients. After these animals had destroyed the originally woody vegetation on these
soils, the sheep grazed the heather that had replaced the forest vegetation.
Overnight the sheep were kept in barns, and the manure they produced in these barns,
mixed with the sod of the heathfields used as litter, was never returned to the heath-
fields but always to the arable land lying around the villages.
In some cases, this system of nutrient transfer lasted for ages, but in other instances
the impoverished soils failed to sustain even a heather vegetation, and became bare.
The winds then started to shift the exposed sand around, and in many cases, the
arable soils surrounding the villages were covered by the sand, forcing the population
to abandon their homesteads and to move on to new lands and to an uncertain future.
The reason I dwell upon this phenomenon has to do with the fact that so many of
these small farms on which mixed farming was practiced, were located on or near
these old arable soils. These farmers had never been rich, but after the Second World
War, the economic conditions forced them either to stop farming or to switch to
another type of farming.
Nowadays, many of these farmers are engaged in meat and egg production. They
raise large numbers of hogs, calves or poultry in long, windowless sheds. The few
hectares of land they have is used more as dumping ground for liquid manure than as
meadow or arable field. Compared with the past, the livelihood of these farmers is
much improved, but they have caused problems in that their farm buildings constitute
countless eyesores in an originally attractive landscape, and their liquid manure forms
a serious threat to the maintenance of a clean environment.
About 20 years ago, these farmers converted their arable land on which traditionally
rye, potatoes, and mangolds were grown, into permanent pasture. When the excessive
use of liquid manure made it increasingly difficult to maintain acceptable swards,
they switched to growing silage maize. The remarkable feature about this maize is
that it can stand excessively high quantities of liquid manure without showing much
of a yield reduction. The lack of any alternative crop which could be grown in rotation
however, forms a severe hazard to the continued growing of maize on these soils.
As you probably know, feed conversion by hogs is greatly improved by the addition
of copper to the feed. This practice, however, has led in certain regions to soil Cu
contents high enough to endanger the lives of such Cu-susceptible animals as
sheep, and the health of certain crops. It is to be expected that measures will soon have
to be taken to restrict the use of copper in hog feed. However, when such measures are
contemplated, one should also see to it that they are enforceable in the entire European
Common Market. Anybody paying attention to discussions on such issues in Brussels,
knows how difficult it is to come to any sort of an agreement.
So far, I may have raised the impression that agriculture in the Netherlands is faced
with nothing but problems. It is still my intention to present a balanced view of our

10
Dutch agriculture, and with some of the less favorable aspects off my chest, I can now
justifiably proceed to some more pleasant features of Dutch agriculture.
Our country lies mainly between 51 0 and 530 30' Northern latitude. The climate is
temperate and maritime, which means that rainfall, with an annual total of 70 cm,
is rather evenly distributed through the year, with a low of 40 mm in March and a
high of 75 mm in August. which by the way is sometimes unfavorable for harvesting
operations during that latter month. In the period September-May, precipitation
exceeds evapotranspiration, and in the remaining months the situation is reversed.
The average monthly temperatures range from a low of 2°C in January to a high of
19'C in July.
Taking into account the average percentages of sunshine in the various months, and
assuming respiration losses to be V, of gross photosynthesis, one arrives at the follow-
ing values of potential dry-matter production of a standard crop in the Netherlands
(Table 2):

Table 2. Potential dry-matter production of a standard crop in the Netherlands (kg per ha)
for the various months of the growing season,
April ....................... 5500 August .................... 6,400
M ay ........................ 7,100 Septem ber ................ 4,700
Jtne ........ ................. 7,600 O ctober ..................... 3,200
July ..... ............. 7,300 Total ...................... 41,800
Source: ran ifeelst H.D.J. et al.: Verslagen van Landbouwk undige Onderzockingen No. 879
(1978)

The various reasons why actual production levels are still so far removed from the
potential production are well known to most of you. I name the more important
limiting factors:
I. incomplete light interception.
2. low temperatures.
3. insufficient water supply.
4. insufficient nutrient supply.
5. incidence of pests and diseases.
6. losses during and after harvest.
Out of these 6, only the number 4 is relatively unimportant in the Netherlands. The
factor no. 2 is often responsible for the dominant influence of factor no. I : due to low
temperatures in springtime, it is often June before leaf area indices reach values high
enough to guarantee nearly complete light interception.
For the three important arable crops winter wheat, potatoes, and sugar beet, data on
potential production, light interception and calculated growth rates are presented in
Figure I. It is seen that it is only in the months of June and July that winter wheat
intercepts a high percentage of incident light and, consequently, shows calculated
growth rates that approach the potential rates. Because of the relatively low heat
requirement of this crop, it forms a closed green canopy much earlier than do potato
and sugar beet. The relatively early ripening of wheat, however, limits its yield.
As can further be seen from Figure I, the weak point of potato is the large discrepancy
shown between actual and potential growth rates in the spring. Again here, low spring
temperatures dictate farming practices, that do not allow the canopy to close until
mid June.

I
 Pot.pro kg dm ho doy - Light
Gross t interception
400 300 . .. ... 100
23s
300 225 --

a 60
200 150

POTATOES
too 7S20

O 0-A
llv 0
April May June July Aug Sept Oct,

Pot prod.,kg.dm ho'doy 1 ' Light

Gross Net interception
tO0• 300 100

- "so
300 225 3--

b N; 60
200 150
40
SUGAR BEETS
10075 20
2Av0
April May June July Aug Sept, Oct.

Pot.prod.c,kg dm ho"day "' *A Light

Gross Net interception
O0 300 100

330 225 - - '-

z .2 60

200 150 * 40
WHEAT " £
100 75 '
20

0 April May June July Aug. Sept. Oct.

Fig./. Potential production rate (line I), percentage of light intercepted by the crop (line 2)
and calculated crop growth rate (line 3) for three crops in the Netherlands. Source: Sibma L.,
Neth. J.Agric. Sci. 25. 25, 278-287 (1977).

12
Likewise, in the case of sugar beet, we can observe that in the period of almost com-
plete light interception, the potential growth rate has already passed its mid-June peak.
Weather conditions delay closure of the canopy to until the end of June.
During the extremely dry summer of 1976, many dairy farmers in the Netherlands
saw the advantages of having an irrigation system. Hence, in dry periods, one can
now see sprinkler irrigation systems in operation on pastures everywhere in the country.
Irrigation of arable land is still rare, partly because the jets of water delivered by the
sprinklers cause damage to young potato and sugar beet plants. The main reason for the

tuber yield
tons/ha

70 _ _150 200 mm
trn m

50-
100 mm
50O 50 mm

40-

30-

20-

10-

00 200 300 N taiten up

kg/ha

sandy soil clay loam

100-

200
N applied
kg/ha

Fig.2. The relationship between N applied, N taken up and calculated potato tuber yields
on two soil types and at four levels of available soil moisture. Source: ran Heemst H. D.J
ef al.: Verst. Landbouwk. Onderz. No. 879 (1978).

13
absence of artificial irrigation systems in arable crop production has to do with the
fact that crops which root more deeply than permanent grass may need no supplemen-
tal irrigation when grown on soils with an adequate water-holding capacity. Dutch
soils are at field capacity at the end of most winters. On soils with a good water-
holding capacity, this means that 200 mm of plant-available water is stored in the
profile, enough to sustain normal crop growth in drier than normal springs.
Many of our marine loam soils have adequate water-holding capacities, but the sandy
soils I spoke of earlier are deficient in this respect. This also explains why our rural
forebears were so interested in raising the water-holding capacities of these soils by
applying annually large quantities of sheep manure mixed with heath sod.
Potatoes especially are sensitive to water shortage, as can be seen from the information
presented in Figure 2. In this type of presentation, the efficiency of the fertilizer used
and the response of the crop to varying quantities of N absorbed is given for different
water-holding capacities of the soils. The graphs show that
1. clay soils have a higher natural N-fertility than sandy soils.
2. the efficiency of added fertilizer N is 60% on sandy soils, and 50% on clay soils.
3. 70 kg of grain are produced for every kg of N absorbed when N is a yield-limiting
factor,
4. a difference of 20 tons of potatoes is to be expected between soils with low- and
soils with high water-holding capacities.
The yields presented in the Figures 1 and 2 are calculated yields. Experiments con-
ducted during the last 10 years have shown that these calculated yields can indeed be
obtained, and in some cases surpassed.
With adequate moisture and nutrient supplies in an average year, the yield levels of
our crops are considerably affected by the extent of crop-protection measures. Some
information on this point is presented in Figure 3 for winter wheat. Taking into
account the discrepancies between the potential and actual growth rates of wheat, as
shown in Figure 1, and assuming that the yield index for winter wheat is 40%, it can
be calculated that annual yields of approximately 10 tons of grain per ha must be
attainable. In experiments in which particular attention is given to pest and disease
control, the calculated goals can be approached very closely. On other experimental
fields with normal crop-protective measures, the differences between potential and
actual yields vary from year to year, but on the average amount to 2 tons of grain. If
differences between potential yields and actual yields obtained with normal crop-
protective measures become smaller than 1 ton per ha, it becomes questionable
whether the additional protective measures needed to approach potential yields will
pay for themselves.
Most striking, however, are the discrepancies between the average farmer's yield and
the potential yield. Of course, on the average farm, there may be water- and, occa-
sionally, nutrient deficiencies.
For potatoes, yields closely approaching 100 tons per ha have already been obtained.
It is rather tempting to cite more such figures obtained with other crops. However,
some sobering remarks must be made here also.
Considering the most important crops grown in the Netherlands, namely grass, cereals,
potatoes and sugar beet, it is questionable whether the European Common Market
will be ready in the coming years to absorb any large increases in such crops or in
commodities derived from grass, such as meat and dairy products. If the answer to
the question is in the negative, it must be realized that any further increase in the

14
 grain yield
kg/ha

0
10,000 0 potential yields o B

0~~

9000- a - ,blue print farming- yields

8000
7000- -
highest yields on experimental fields
6000

5000- ..
4000 o national average yields

L I I I tII

1965 1971
Fig.3. Potential and actual winter wheat yields in the Netherlands. 1965-[971. Source:
Goucmian,, J.and van Keulen U.: Hohenheimer Arbeiten, Heft 97, 40-52 (1978).
production of such crops is bound to result in a price decline. Clever farmers may
succeed in raising their yields enough to offset such price declines, but it must be
realized that not all farmers in the Netherlands may be smart enough to stay in
business under such circumstances. However, for the maintenance of our landscape,
we also need these not-so-smart farmers, who in this respect may be even more
valuable than their clever colleagues, who are more prone to change their rural
homesteads into factory-like enterprises.
However, there are more markets in the world than the European Common Market,
and with two-thirds of the world population still undernourished, there remain enough
reasons to pay continued attention to the improvement of the productivity levels of
our crops.
To that effect, in the Netherlands we have a well developed, but rather complicated
system of agricultural teaching, research, and extension. Since the organizers of this
Colloquium considered it worthwhile to let you hear something about the organization
of agricultural research in the Netherlands, we asked Mr. Byiloo, one of the secretaries
of our National 4griculturalResearch CowMVil, to tell you something about the activities
of this Council and about its place in relation to all agricultural organizations in the
Netherlands.

15
Organization of the Agricultural Research
in the Netherlands
J. D. Bijloo, National Council for Agricultural Research. Wageningen/The Netherlands-

History

It will be hiell)1i1, before discussing the organization and structure of the agricultural
research in the Netherlands today. to give a bird's eye view on the history and growth
of research.
In 1821. a veterinary school was Founded at Utrecht with the main purpose of teaching
on animal diseases. In line with the liberal philosophy of that period of 'laisser faire,
laisser aller'. agriculture had to look after itself. The first school of agriculture and
forestry was only opened at Wageningen about 25 years later, in 1876.
The first industrial revolution enhanced the possibilities of transport within the United
States of America and between the USA and Europe. As a consequence. enormous
quantities of food, tmainly cereals flooded the European markets, reSLIIting in the
first agricuItural crisis.
A National Committee in the Netherlands concluded in 1886, that, instead of giving
direct financial aid to the farmers, tie government should adopt a more positive
approach by slimulating education, research and extension in agriculture.
Ten years before, in 1876 the first Testing Station for seeds, animal feeds and fertilizers
had been founded at Wageningen. Then, as a result of the report of the National
Committee, between 1877 and 1900, seven regional experimental stations with testing
facilities were opened, and these also had an extension function.
in the following years up to World War 1[, a number of specialized Research Institutes
and Experimental Stations were founded first by the Iinistry of Economic Affairs and
later by the Ministry of Agriculture and Fisheries. the Netherlands Organization for
Applied Science (TNO) and private (agricultural) industries.
In the meantime, education both in agriculture and veterinary science had been given
to a higher status by the foundation of the ligrictlnral Unitersit,' at Wageningen
(1918) and the Facu/t of Veterinar , Science of the University at Utrecht (1925).
Later on, and especially after World War Ii, more Experimental Stations and a large
number of Research Institutes for basic and applied research were founded and existing
institutions enlarged. These developments resulted in 'Wageningen' becoming inter-
nationally recognized centre for agricultural research which it is today. The Faculty of
Veterinary Science (14 sub-faculIties) and the University of Wageningen (70 sub-facul-
ties) and showed considerable growth.
* r. J.D. Bijl,,. National Council for AgricuLItural Research, P.O. Box 297, NL-2501 BD
Den Haag/The Netherlands

17
In order that research should function properly as a policy instrument, it is necessary
that research programmes should be closely co-ordinated. The National Council for
Agricultural Research (NRLO) was founded to improve communication between
government, research institutes of the Ministry of Agriculture and the users ofresearch.
This was achieved by:
- co-ordination of research programmes (of government research institutes)
- policy development
- furthering the relations between industry and science
- implementing and efficient execution of research programmes.
Most of the government institutes are foundations. This legal form has been chosen
in order to establish a system of co-directorship embracing agricultural industry,
farmer's organizations, consumer organizations and other bodies.
The Ministry has in general two types of research institutions, namely:
- institutes and experimental stations belonging to one of its divisions with policy
responsibilities
- institutes grouped under the division of the Central Directorate of Agricultural
Research (DLO).
The Division of Agricultural Research (DLO), however, co-ordinates the research
activities of all policy divisions of the Ministry which use research as a policy instru-
ment. Research co-ordinators of all divisions constitute the staff which prepares the
proposals for allocation of material means and personnel to the (34) institutes. These
proposals are discussed and decisions are made in a ministerial committee (COLO)
presided over the two Directors-General of the Ministry of Agriculture. The institutes
and experimental stations of the Ministry of Agriculture account for about 2/3 of the
research budget of the Ministry.

National council for agricultural research (NRLO)

The National Councilfor Agricultural Research (NRLO) was reorganised in 1970 to

coordinate the aforementioned activities with the research efforts of other ministerial
bodies (as far as concerns agricultural research) and with agricultural research by
private industry.
The NRLO, initially charged only with the coordination of research by the Ministry
of Agriculture, developed into an organisation in which the participants are either
producers or users of agricultural research. The main producers of research are the
research organizations of
- the Ministry of Agriculture and Fisheries
- the Agricultural University
- the Dutch Organization for Applied Science (TNO)
- the Veterinary Faculty of the State University of Utrecht
- research under the auspices of:
- the Jsselmeerpolders Development Authority
- the Netherlands Institute for Sugar Production
- the Netherlands Institute for Dairy Research
The main users of research include a wide range of bodies. They fall into three groups:
- ministries and other governmental bodies

18
- organizations in the food and agricultural sector
- interest groups within society.
In principle, any new body can join NRLO, provided that its participation promotes
the aims of NRLO and that it accepts the terms of operation of NRLO.

Tasks and objectives

Agriculture is described as:
- use of living organisms to fix solar energy as organic matter
- conversion of these raw materials into consumer goods and products for industry
- management of the physical environment and flora and fauna.
Agricultural research includes such activities as:
- conservation, development and use of soil, water, forests, etc.
- protection against harmful biotic and non-biotic factors
- efficient production and quality improvement
- product development and processing
- efficient marketing, including pricing and quality
- development of human resources, economics of communities, areas, etc.
- aid to developing countries.
The main tasks of NRLO are in the field of
- planning and
- co-ordination of research
with the following objectives:
- tuning of agricultural research programmes to the needs of society
- ensuring that research and development related to agriculture and land resources
(including research facilities) are used efficiently
- providing solicited and unsolicited advice
- providing a channel for links between agricultural research and other branches of
research
- assuming responsibility for systematic recording of all research projects concerned
with agriculture and land resources
- preparing periodical surveys of agricultural science and of needs for research in
agriculture and land resources.
Two aspects of the task and status of NRLO need further clarification: the relation
of NRLO to the participating research organizations and the advisory function of
NRLO.

Relations to participating research organizations

Policy and administration of agricultural research are under the control of the partners
namely the Ministry of Agriculture and Fisheries, the Agricultural University, the
Veterinary Faculty and TNO, The Netherlands Institute for Sugar Production and
the Netherlands Institute for Dairy Research. Such control includes the research
programme, collaborative research, establishment of services to research investment
in buildings and staffing policy. NRLO does not interfere with the policy and adminis-
tration of research organizations.

19
The statutes imply that, in establishing NRLO, the persons and bodies administering
and carrying out research considered it useful to consult each other and to consult the
users of agricultural science as to how research should best be organized and used.
Hence, NRLO does not compete with or attempt to administer participating bodies,
but is a voluntary group in which the partners try to harmonize research policy and
research programmes. The heart of the matter is that NRLO is an organization of
partners or even just a secretariat or postbox. Collaboration is voluntary but carries
responsibility. The NRLO acts like a friendly policeman.
Advisory function of NRLO
From time to time, advice on various matters has been offered to various governmental
bodies, but this is certainly not a major function. The main way in which NRLO
achieves its goals is consultation. NRLO has not the authority to lay down the law
for the partners. It can offer advice, solicited or unsolicited, on aspects of research
policy. Advice is limited to the efficiency or effectiveness or research programmes. It
can bring into consideration matters that indirectly affect agricultural research, for
instance national science policy.
Constituent bodies and manner of operation
Besides the Council, NRLO has a secretariat and a system of constituent bodies (or
'Substructure'). Most of NRLO's work is in the Substructure. The variety of constituent
bodies reflects the wide spectrum of agricultural research with its diversity of dis-
ciplines, scientists and interest groups. The government's agricultural research budget
in 1978 was aboutf350 million for about 100 institutes, research stations and university
departments, about 1000 research workers, 4000 research projects. In addition, about
f]130 million was spent on agricultural research in industry. Several hundred of groups
attempt to bring together participants involved in similar work to discuss their
problems.
The Substructure has four types of groups (Figure 1).
Divisions are groups in which research policy makers and research managers arrange
the main lines of research for a broad field. NRLO has divided its field of operations
into five divisions, namely:

16%

Div. 111 39% ment of nature and environment

Div. 11 Plant production
21% Div. Il Animal production
Div. IV Processing and marketing
Div. V Agriculture and society

Div. IV
20%/ Div.V

20
 Council

CentralProject
Administration

Documentation

Program Advisory Cordntg

eCommittees

Special lCmite
Study Groups

Working
Parties

Fig. I. National Council for AgricIltural Research (NRLO).

Program Advisory Conimitlees are bodies in which users of agricultural research

convene. Their main function is to feed viable requests for research into the planning
process. They are organized according to subjecl-interests, for instance a large crop
or a group of crops, groups of animals, marketing problems etc.
Coordiating Comnittees (COOC) comprising contact committees, working parties
and study committees. These are groups in which research workers and research
managers consult with each other, and plan and co-ordinate research in their subject.
They are organized according to research fields, for instance water quality, phyto-
pathology and crop protection, animal housing and maintenance research, processing,
etc.

21
The following structure shows how research managers of the participating bodies are
brought together according to a particular discipline.

Res. inst. Min. Agric and Fish. Org. Ap. Fac. Vet Industry
DLO Univ. FId. St. Res. TNO Univ.
COOC A xx X X x X
COOC B x x x x x
COOC C x x x x x

Special Study Groups: parties with the task to adequately prepare policies for special
subjects by means of study and consultation.
The following is characteristic of the way NRLO tries to operate: Decision-making
is spread as widely as possible. Contact committees and working groups make their
own decisions and only pass matters beyond their competence to a higher level, the
co-ordinating committees. In turn, the co-ordinating committees regulate the main
lines of research and decide how to implement the research programme within their
fields. Only matters beyond their competence are referred to the division. Only matters
which the divisions cannot control are put before the National Council. There is wide,
rapid and open communication of views and information between different groups
and levels in the organization. Exchange is not only between scientists, research
administrators and policy-makers, but also between producers and users of research.
Free exchange is necessary if largely autonomous groups are to work in harmony.
Communication must be not only up the hierarchy, but from group to group, and
within the group. Rapid and free communication allows for assessment, adjustment
and regulation both by superiors but and within the groups themselves. NRLO makes
use of selective exchange of written information and of liaison officers between bodies
with a common interest. In principle, information is public and freely available.
The organisation of NRLO is continually under review. It is necessary from time to
time to examine whether or not particular groups are still serving a useful purpose;
whether or not they are still needed. There may be overlap among groups within NRLO
or with outside groups. It must be possible to set up new groups quickly with clear
information on the task to be performed and with the proper composition. Clear
arrangements are needed for procedures and terms of operation of the various groups.
The work must be assessed frequently and alterations made in the functioning of
groups as necessary. Groups which have completed their task or ceased to serve their
purpose must not linger on but must be dissolved.
Thus, for the programming, co-ordinating and planning of agricultural research the
Council establishes specific platforms at different levels for consultation and decision-
makings between research management sectors, research workers and potential users
of scientific results. This participation model fairly well meets the requirements of
modern management, namely a balance between top--down and down-- top move-
ments, with respect to planning and execution of research.

22
Research on subjects falling within the scope of this Colloquium on Physiological
Aspects of Crop Productivity is organized in the
- Co-ordinating Committee on Biological Aspects of Plant Growth and Development
which is made up as follows:
Members 7 Research Institutes
6 Experimental Stations
8 Sub-Faculties of the University of Wageningen
Contact-Cotnmittee on Plant growth regulators
Sub-committee on Fundamental Studies
Sub-committee on Applied Research
Working Group on Senescence
Contact-Conunitteeon Vase-life of cut-flowers
Contact-Comnittee on Fruit-physiology
Committee on Tissue-cultures
Committee an Supplementary irradiation in protected cultivation

23
 Chairman of the 1st Session
Prof. Dr. 1. Arnon, Settlement Study Center,
Rehovot/Israel: Member of the Scientific Board
of the International Votash Institute

Breeding for Higher Yields

Breeding for Higher Yields
F.G.H.Lupton, Plant Breeding Institute, Trumpington, CanibridgelEngland*

Summary

About half of the rapid increase in yield of cereal crops during the past thirty years is due to
the introduction of new varieties. Comparison of old and new varieties shows that there has
been little change in total crop biomass, but that most of the extra yield is due to increased
harvest index. This is often associated with shorter straw, but there is a positive association
between yield and height in many crops. Dwarfing genes in wheat are pleiotropically associated
with characters likely to increase yield, so that further increases may involve the selection of
tall progenies which carry these genes.
Increased grain yield has been associated with lower protein content but little change in
protein yield per hectare. Most grain protein has been assimilated by anthesis when crop
protein content is proportional to biomass. Grain protein content may therefore be increased
by increasing biomass or nitrogen harvest index.
Greater reliability of performance may be obtained by breeding for durable disease resistance
or by breeding crops adapted to speciic environments, such as those where yield may be
limited by drought.

1. Introduction

1.1 Yields of cereal crops have increased more rapidly during the past thirty years than
at any other time during the history of mankind. During this time the national average
yield of wheat in the United Kingdom has increased from 2.4 t/ha in 1950 to 5.2 t/ha
in 1979 and some farmers are obtaining yields of I I t/ha and more. At the same time
the average yield of barley has increased from 2.3 to 4.1 t/ha. Similar results have been
obtained from other countries in Western Europe and equally spectacular figures are
recorded from some of the developing countries, though the levels of productivity are
lower. In India, for example, average wheat yields increased from 8.3 q/ha in 1966 to
14.1 q/ha in 1976, and at the same time increased availability of irrigation increased
the area available for wheat cultivation, so that national production increased from
10.4 to 29.1 million tonnes during these ten years (Rao [27]).

* F.G.H. Lupton MA, PhD. FIBiol; Plant Breeding Institute, Mars Lane, Trumpington,
Cambridge CR2 21-Q/England

27
2. Comparison of old and modern varieties

2.1 Much of this increase is due to improved husbandry and heavier use of fertilisers,
but these advances would have been impossible without the introduction of new
varieties, and in particular the introduction of shorter strawed varieties, especially
of wheat. There is necessarily considerable interaction between these two components
of yield improvement, but Silvey [31] has calculated that new varieties have accounted
for approximately half the yield increases in wheat in the UK between 1947 and 1977.

2.2 Austin et al. [2] report an interesting comparison of old and new winter wheat
varieties grown under conditions in which disease attack was prevented by prophy-
lactic sprays and lodging prevented by supporting the crops with netting. They showed
that under these conditions the yields of the modern varieties were about 50% greater
than those of the older ones, and that the straw length of the modern varieties was
little more than half that of the older ones. However the total biomass of the aerial
parts, that is yield of grain and straw, had changed very little. An analysis of the
components of yield showed that it was not possible to identify any single component
as the chief cause of yield improvement, the number of ears per plant being greatest in
one variety, the number of grains per ear in a second and the 1000 kernel weight in
a third. The yield increase had however been associated with a progressive change in
harvest index which had increased from 34% in the oldest varieties to 50% in the
highest yielding modern selections. This result, which agrees with that found by
others (e.g. van Dobben [12]), suggests that future yield advances are likely to be
associated with further increases in harvest index. But, as Austin et al. [2] point out,
there must be a limit to the extent by which this process can continue, because of the
agronomic requirements of the crop, so that breeders will have to exploit other
attributes, such as total dry matter production, if long term yield advances are to be
obtained.

3. Future trends in yield improvement

3.1 Total dry matter production can be increased by selecting varieties which either
make more efficient use of incident solar energy by more efficient photosynthesis per
unit of leaf surface or which allow the light to penetrate more uniformly through the
crop canopy, so that lower leaves can make a greater contribution to total photo-
synthesis. Alternatively, the breeder can select varieties in which the photosynthetic
period is extended, or in which the rate of photosynthesis of older leaves falls off more
slowly.

3.2 Genotype differences in rate of photosynthesis per unit of leaf area have been
reported by a number of workers. For example, Dautnma [lO] demonstrated a 22%
variation in the maximum rate of photosynthesis of the flag leaves of a range of spring
wheat varieties, and similar results in spring barleys. Dantian found that the maximum
rate of photosynthesis took place when the leaves first reached their fully expanded
areas. The subsequent decline in photosynthetic activity may also differ with variety,
as was shown by Aslat and Hunt [1 in a comparison of the spring wheat varieties
Opal and Kolibri. in this case, the leaves remained green for about the same time,

28
and although the initial fall in photosynthetic rate was much greater in Kolibri. there
was little varietal differences in rate of photosynthesis 42 days after anthesis. But other
workers have reported wide differences in leaf longevity, particularly in the flag and
second leaves. and these have been shown to be anl important factor in determining
yield variation (Spierlz, ftel Ha' and Kupers [33]; Lupton, Oliver and Ruckenhauer
[2/]).

3.3 The measurements of leaf photosynthesis so far discussed have been made on single
leaves of plants growing in containers, and have not in general been confirmed by
measurements made on plants growing in field conditions. The discrepancies between
fiekl and pot experiments may partly arise because of differences in leaf angle and other
aspects of crop geometry which may cause differences in light penetration through the
crop canopy. The effects of such differences were studied by Austin et al. [4] who
compared two similar but not isogenic lines, one with erect and the other with drooping
foliage The rates of photosynthesis of" the genotypes were determined by measuring
the carbon dioxide uptake of small area of field crop sealed under a plastic cuvette
and connected to an infia red gas analyser. Results showed that the rate of photo-
synthesis of the genotype with the erect leaves was consistently greater than that of the
genotype with the drooping leaves. Surprisingly, however, the grain yields of the two
genotypes were not significantly different, and Austin and his colleagues showed that
this was dttc to the translocation of more reserve carbohydrate from the stem to the
ear by the genotype with the drooping leaves.

4. Interaction of 'Source' and 'Sink'

4.1 Austin's observations illustrate the importance of the capacity of the ear to store
carbohydrate in determining varietal differences ingrain yield. The interaction of
'source' and 'sink' has been studied by many workers, and it has been sutggcsted by
Bit,hani [8] and others that the high yields obtained from short strawed wheat
varieties may be partly due to reduced competition for carbohydrate between the stem
and ear of such varieties at the time of rapid ear development shortly before anthesis.
This might result in more photosynthate being available for ear development, so that
larger ears, and in particular, ears with more grains per spikelet, can develop. There
is, however, little experimental evidence to support this attractive concept, and several
workers (e.g. Rawson, Giford and Brcner [28]) have shown that removal of the
ear. or reduction in its size. has little direct effect on the rate of photosynthesis of the
leaves.

4.2 A more direct analysis of the interaction of source and sink was made by LupltOn,
Oliver and Ruckenbauer [21]. They made detailed measurements of the pattern of
development of the ears and vegetative parts of a range of selections from two crosses
between tall and shorter strawed wheat varieties. These were then used to calculate
Multiple regression equations relating yield to the components of crop development.
Thcir results showed that approximately 50% of the yield variation could be aceCotnted
for by consideration of flag leaf area duration and measurements concerned with
photosynthetic capacity. bilt thai simtltaneotts consideration of these components
and measurements of the growth rate of the ear privnordia accounted for 92% of the

29
total variation. These experiments thus show that crop development can only be
satisfactorily explained by simultaneous consideration of components concerned with
source and sink.

4.3 A further illustration of the extent to which yield may be limited by the capacity of
the ear to store photosynthate is given by an analysis of the physiological basis of
of heterosis in wheat (Lupton [19]). In these experiments, rates of photosynthesis and
patterns of dry weight increase were studied in two sets of parents with their F1 hybrids.
In both cases, the F, hybrids showed marked heterosis for dry weight accumulation
and rate of photosynthesis until two weeks before maturity. There was then a marked
fall in the dry weights of the F1 hybrids but a continued increase in the dry weights of
the parents. This suggests that, possibly as a result of incomplete restoration of fertility,
the sink capacity of the F1 ears was not sufficient to utilise the extra carbohydrate which
had been assimilated.

5. Exploitation of dwarfing genes in wheat

5.1 There has been a consistent trend to breed successively shorter strawed varieties of
cereal crops during the past thirty years or more. However, consideration of random
populations, such as the progenies of randomly selected F2 plants, shows a positive
correlation between height and yield, as has been demonstrated by Snape, Law and
'orland [32] for wheat, Riggs and Hayter /29] for barley and Sampson [30] for
oats, In making agronomically acceptable short strawed selections, breeders have
therefore been restricting their future progress, and as has been mentioned previously,
this trend must eventually be reversed. Most of the progress in oat and barley breeding,
and the earlier work with wheat, has involved selection for recombinations of minor
genes, but the spectacular successes with semi-dwarf wheat varieties have involved the
use of reduced height (rht) genes derived from the Japanese variety Norin 10. The
mode of action of these genes is associated with disturbance of the gibberellic acid
(GA) balance such that plants with the dwarfing genes have a very high GA content,
the expression of which is prevented by an endogenous anti-gibberellin. This phenom-
enon has been studied in detail by Gale and Law [15] who suggest that the high yields
of many of the semi-dwarf cultivars might arise because of other effects of the GA
which were not blocked by the endogenous anti-gibberellin. The GA may, for example,
disturb the apical dominance so as to increase tillering or at a later stage of crop devel-
opment, to increase spikelet fertility.

5.2 Wheat plants with the rht genes can be identified by spraying seedlings with GA.
This causes marked internode extension in plants lacking these genes, but has no
effect on plants which carry them (Gale and Gregory [14]). The effect of the rht
genes on yield was demonstrated by Gale and Law [15] using segregating populations
classified in this way. They demonstrated the expected positive correlations between
height and yield in plants homozygous for the rht genes and in those lacking them. But
they also found that, over a given range of heights, plants with the rht genes signifi-
cantly outyielded those without them. They therefore suggest that breeders should
exploit the GA insensitivity test to select for 'tall dwarfs' - that is for taller plants
which carry the rht dwarfing genes.

30
6. Breeding for increased protein yield

6.1 When selecting for increased grain yield, it is important that the plant breeder
should at the same time consider the protein content of his new varieties. This is
obviously important in developing coutries, where wheat comprises a large part of
the human diet. But it is essential to view the problem in its proper perspective and
in particular to appreciate that protein content is of secondary importance when the
total energy value of the diet is limiting. Under these circumstances proteins consumed
will be used as an additional source of energy and will not therefore be available for
growth or tissue replacement. Nevertheless, programmes to breed wheat and barley
varieties with increased protein percentages, and with improved amino acid balance
of these proteins have been started in India and other developing countries, At the
same time, increased areas of wheat cultivation in these countries following the intro-
duction of high yielding short strawed varieties has led to a reduction in the area
devoted to grain legumes such as chick pea (Cicer arietammo ). This has resulted in a
reduction in the protein content of the diet in some areas, and in particular in a fall
in the content of lysine and other limiting amino acids.

6.2 Even in Western Europe, cereals provide an important part of the protein in the
human diet, and also in the diet of animals, which consume 47% of the wheat and
650X, of the barley grown in the United Kingdom (HGCA [16]). So far as human diet
is concerned, protein content is most important in relation to bread making quality
in wheat and to malting quality in barley, though in the latter case, the maltster
requires a product with low protein content.

6.3 When a range of wheat or barley varieties are grown under the same conditions,
there is a strong negative association between protein percentage and grain yield,
though the protein content of any variety may be increased by appropriate appli-
cations of nitrogenous fertilizer. This situation was demonstrated by Pushman and
Binghain [25] who compared the yields of a range of winter wheat varieties grown
under three contrasting regimes. They also showed that despite the negative regression
of protein percentage on yield observed under each regime, certain varieties consistently
had higher protein contents than would have been expected from the overall regres-
sions, suggesting that it should be possible for the plant breeder to select for high
yielding varieties with satisfactory protein content. They also showed that the protein
yield per hectare of modern varieties was up to 10% greater than that of older varieties
they had replaced.
6.4 Despite the encouraging results obtained by Ddbereiner, Day and ion Bulow f/ I],
there is no immediate likelihood of developing cereal plants capable of synthesising
proteins from atmospheric nitrogen. Increase in protein percentage can therefore
only be achieved by selection for varieties which either assimilate soil nitrogen more
effectively or which achieve a better distribution of protein within the plant. It is
however important to appreciate that any increase in grain protein, whether by
reduction of nitrate, or in the longer term by fixation of atmospheric nitrogen, can
only be achieved by the expenditure of photosynthetic energy which might otherwise
have been used in the synthesis of carbohydrates. Furthermore, the energy reqtired
to synthesise a gramme of protein is about double that required to synthesise a gramme
of carbohydrate.

31
6.5 It was suggested by filrich and Hageman [13] that nitrate reductase activity was
rate limiting in the nitrogen nutrition of wheat, and that it should therefore be possible
to select genotypes capable of synthesising large amounts of protein by measuring
differences in nitrogen reductase activity. This possibility was investigated by Austin,
Rossi and Blackwell [6] who were unable to find evidence that nitrate reductase
activity was rate limiting, but rather that the rate of nitrogen uptake was closely
associated with overall crop growth rate. This was confirmed by Austin et al. [3] who
demonstrated a close relationship between crop dry weight and total crop nitrogen
content at anthesis. This study also showed that most of the nitrogen present in the
crop at maturity had been assimilated by anthesis, though some post anthesis nitrogen
uptake took place in varieties in which stem dry weight also increased during this
period.
6.6 There are thus four ways in which selection for increased grain protein content
may be attempted:
(i) The breeder may select for maximum dry weight at anthesis. As discussed earlier,
such selection may also be necessary in order to achieve increased yield, though
it may be difficult to combine with selection for shorter strawed varieties.
(ii) He may select for varieties which break the close correlation of total dry matter
and protein content. Such varieties have been identified by Vogel, Johnson and
Mattern [34] and have been widely used in breeding programmes in many
countries.
(iii) He may select for varieties in which a high proportion of total plant nitrogen is
translocated to the grain. Austin et al. [3] report significant differences in this
ratio, defined as nitrogen harvest index, though high values may be associated
with varieties in which photosynthetic activity of the leaves and ears falls off more
rapidly after anthesis, leaving more time for the translocation of nitrogen from
those organs to the grain.
(iv) Finally he may select for genotypes in which nitrogen uptake continues after
anthesis, but the metabolic energy required for this process may again be as-
sociated with genotypes with long continued photosynthetic activity, and hence
with lower nitrogen harvest index.
6.7 Selection for bread making quality requires simultaneous selection for quality and
quantity of protein. Problems associated with selection for protein quality do not lie
within the scope of the present review, but reference may be made to the recently
evolved sodium dodecyl sulphate (SDS) test (Axford, McDermott and Redinan [7]).
In this test, flour samples are shaken up in solutions of SDS, which causes the long
protein molecules of the flour gluten to uncoil to form flocculent particles. Comparison
of the rate of sedimentation of these particles provides a rapid assay of the suitability
of the flours for bread making. The test has the advantage of being nearly independent
of the protein content of the samples concerned and has greatly increased the efficiency
of early generation tests of baking quality in wheat (Blackman and Gill [9]).

7. Breeding for reliability of performance

7,1 Any analysis of the prospects of breeding for yield must consider also problems
of breeding for reliability of performance and the need for more widespread use of

32
improved husbandry techniques so as to narrow the gap between national average
yields and those obtained by the best farmers. Detailed consideration of problems of
disease and pest control lie outside the scope of the present review, though losses due
to these organisms are one of the principal causes of year to year variations in yield.
Mention should however be made of the recent work on breeding for durable resistance
to leaf diseases (Johnson [18]), and to the use of mixtures and multilines in order to
limit the risk of losses due to newly arising virulent races, especially of airborne path-
ogens attacking leaves and stems (Wolfe and Barrett [36]).

7.2 It has been suggested that work on breeding for disease resistance is no longer
necessary because the diseases can now be readily and cheaply controlled by fungicides.
But I consider this a dangerous concept, partly because the cost of fungicides and their
application is likely to increase, but more particularly because of the very real danger
that strains of the pathogens may develop with virulence against sprayed crops. Such
virulence has already been reported on several occasions, as in the case of ethirinool
tolerance of powdery mildew in barley (Wolfe [35]). [ therefore suggest that although
fungicides have a very valuable role to play, their action should be monitored in
variety trials with untreated controls. This should lead to an integrated control
system in which losses are restricted by the use of low rates of fungicide on varieties
with a moderate level of durable disease resistance.

8. Breeding for specific adaptation

8.1 The breeding and trialling procedures used in most countries are designed to
select varieties which perform well over a wide range of environments. Although most
systems, such as that used in Britain (Patterson and Silvey [24]), include trials grown
at contrasting sites and continued over a number of years, the varieties chosen for
commercial exploitation are normally those which perform well over a wide range of
environments and have been selected for wide adaptation from an early stage in the
breeding programme. Furthermore the trial sites used rarely include areas of low
fertility or severe contamination by soil borne diseases, largely because such sites are
not usually available on experimental farms, and trials grown on them frequently have
large coefficients of variability.

8.2 It has been suggested, however, that varieties selected for overall good performance
over a range of sites may yield less well under specified conditions than varieties
selected specifically for those conditions. A project to test this hypothesis in relation
to winter wheat breeding has recently been started at Cambridge (Hughes and Oliver
[17]). This involves comparisons of the performance of selections made following a
normal autumn sowing at the Plant Breeding Institute with that of selections from a
very late sowing and from sowings on a sandy site at which water deficit is likely to
occur. It is planned to multiply early generation selections made at each of the three
sites so that they can be compared in yield trials under each environment. There are
obvious logistic limitations to the economic value of varieties bred to meet narrowly
defined ecological requirements. It is hoped, however, that these experiments will show
whether it is desirable to breed for specific environments, and that the environments
considered are of sufficiently widespread occurrence to warrant the development of
varieties specifically adapted to them.

33
9. Breeding for water deficit

9.1 It is in particular felt that there is considerable scope in breeding varieties which
are tolerant of water deficit. Austin and Quarrie [5]) considered a range of experiments
with wheat and barley grown on light soils in Eastern England and found that correc-
tion of water deficit had led on average to an 18% increase in crop yield. It is, of course,
unlikely that a yield advance of this size could be obtained by breeding, but there is
considerable genetic variability available in characters associated with water economy.

9.2 Lupton et al. [20] demonstrated varietal differences in root development in a range
of wheat varieties. They showed that short strawed varieties had more roots at depths
of 50 cm or more below the surface than taller ones, and that this was associated with
a more effective uptake of phosphorus. But a more extensive root system may not be
associated with better drought tolerance. Indeed, an extensive root system in the early
stages of crop development may lead to an early exhaustion of soil water, and thus to
a water deficit during the critical period of grain formation (Passioura [23]; Lupton
and Pushtnan [22j).

9.3 Varietal difference in drought tolerance may therefore be better studied by con-
sideration of the water relations of the aerial parts. Quarrie [26] studied the relation-
ship of abscisic acid and proline accumulation in droughted plants with the develop-
ment of water stress and with changes in stomatal conductance. He compared eight
spring wheat cultivars grown under severely droughted conditions in pots and found
the expected close correlations of abscisic acid and proline contents with chang-
ing leaf water potential. But he also found varietal differences in the rate of abscisic
acid accumulation as leaf water potential decreased, with a suggestion that the more
drought tolerant varieties accumulated abscisic acid more slowly. Assessment of the
rate of abscisic acid accumulation may therefore provide a method of assaying drought
tolerance, but it is unlikely that such a relationship has a general application, and it
is possible that simple morphological characters such as stomatal distribution or even
leaf waxiness may be equally useful in assessing varietal differences in drought toler-
ance.

10. Longer term prospects

10.1 Although there appears to be no immediate prospect that the steady increases in
cereal yields which have been so pronounced a feature of the past thirty years will
cease, it is clear that little further increase in harvest index can be expected. Further
yield improvements must therefore be associated with increases in total crop biomass
and probably with some increase in crop height. Greater biomass will facilitate greater
uptake of soil nitrogen, so that new varieties may also have a higher protein content.

10.2 But long term prospects for yield improvement depend on the identification of
genotypes which are either more efficient photosynthetically or in which photosynthesis
continues longer. There is at present little known variation in photosynthetic efficiency
within the temperate grass species and little prospect of extending the duration of
photosynthesis without an undue disturbance of modern crop husbandry procedures.

34
But there is considerable variation between genera and in particular between the
so called C3 and C4 species. This suggests the possibility of exploiting the techniques
now being developed in genetic engineering to develop wheat or barley varieties with
the capacity to utilise the C4 photosynthetic pathway. Although clearly a very long
term objective, it is hard to appreciate the economic, political and sociological ad-
vantages such varieties would offer.

10.3 Meanwhile we must remain realistic in our outlook, and [ would suggest that in
the shorter term increases in yield will depend more and more on the breeding of
varieties which are more consistent in their performance, suffer less from the vagaries
of climate and disease, and are better adapted to the specific environments in which
they are grown.

II. References

I. Aslant A. and hunt L.A.: Photosynthesis and transpiration of the flag leaf in four spring
wheat cultivars. Planta 141, pp. 23-28 (1978)
2. Austin R. B., Binthan J.,Blackwell R.D., Evans L. T., Ford MI A., Morgan C. L. and
Taylor M.: Genetic improvements in winter wheat yields since 1900 and associated
physiological changes. J. agric. Sci., Camb. (in press) (1980)
3. A ustin R. B., Ford Al,A., Edrich J.A. and Blackwell R. D.:The nitrogen economy of wheat.
.1.agric. Sci.. Camb. 88, pp. 159-167 (1977)
4. Autstin R.B., Fo,rd At. A., Edrich J.A. and Hooper B.E.: Some effects of leaf posture on
photosynthesis and yield in wheat. Anti. appl. Biol. 83. pp. 425-446 (1976)
5. Austin R.B. and Quarrie S.A.: Crop characteristics and the water economy of cereal
plants. Crop Phys. and Cereal Breed. (Ed. J,H.J.Spierz and T. Kramer) Pudoc, Wage-
ningen. pp. 122-125 (1979)
6. Austin R.B., Rossi L. and Blackwell R. D.: Relationships between nitrate reductase
activity, plant weight and nitrogen content in seedlings of Triticum and Aegilops. Ann.
Bot. 42, pp. 429-438 (1978)
7. A.v/rd D. W. E., McDermott E. E. and Redman D. G.: Small scale tests of bread making
quality. Mill. Feed Fert. 161. pp. 18-20 (1978)
8. Bing/tam J.: Physiological objectives in breeding for grain yield in wheat. The Way Ahead
in Plant Breeding: Proc. 6th Congr. Eucarpia, Cambridge, 1971 (Ed. F.G.H.Lupton,
G.Jenkins and R.Johnson) pp. 15-29 (1972)
9. Blackinann J.A. and Gill A. A.: A comparison of some small-scale tests for bread making
quality used in wheat breeding. J. agric Sci., Camb.(in press) (1980)
10. Dantiona G.: Rates of photosynthesis in leaves of wheat and barley varieties. Neth. J.
agric. Sci. 21, pp. 181-187 (1973)
II. Dobereiner J.,Day .. M. and von Bu/ow J.F.W.: Associations of nitrogen fixing bacteria
with roots of forage grass and grain species. Proc. Int. Winter Wheat Conf., Zagreb,
June 1975. pp. 221-237. 1975
12. van Dobber, B. H.: Influence of temperature and light conditions on dry matter distri-
bution, development rate and yield in arable crops. Neth. J. agric. Sci. 10, pp. 377-289
(1962)
13. Eilrich G.L. and Hageman R.H.: Nitrate reductase activity and its relationship to accu-
mulation of vegetative and grain nitrogen in wheat (Triticun aeslirian L.). Crop Sci. 13,
pp. 56-66 (1973)
14. Gale M.D. and Gregory R.S.: A rapid method for early generation selection of dwarf
genotypes in wheat. Euphytica 26, pp. 733-738 (1977)
15. Gale M. D. and Law C.V.: The identification and exploitation of Norn 10 sei-dwarfing
genes. Ann. Rep. Pl. Breed. Inst. Camb. 1976, pp. 21-35 (1977)
16. HGCA: Cumulative monthly statistics. Bull. Home Grown Cereals Auth. 14, No.25
(1980)

35
17. Hughes W.G. and Oliver R.H.: Yield variation in wheat. Ann. Rep., PI. Breed. Inst.
Camb. 1979 (in press) (1980)
18. Johnson R.: Practical breeding for durable resistance to rust diseases in self-pollinating
cereals. Euphytica 27, pp. 529-540 (1978)
19. Lupton F.G.H.: The physiologic basis for heterosis in wheat. Heterosisin plant breedingi
Proc. 7th Congr. Eucarpia, Budapest, 1974 (Ed. A.Janossy and F.G. H. Lupton) pp. 71-79
(1976)
20. Lupton F.G.H., Oliver R.H., Ellis F.B., Barnes B.T., Howse K.R., Welbank P.J. and
Taylor P.J.: Root and shoot growth of senti-dwarf and taller winter wheats. Ann. appl.
Biol. 77, pp. 129-44 (1974)
21. Lupton F.G.H., Oliver R.H. and Ruckenbauer P.: An analysis of the factors determining
yield in crosses between semi dwarf and taller wheat varieties. J. agric. Sei., Camb. 82,
pp. 483-496 (1974)
22. Lupton F.G.H. and PuslunanF.M.: The crop, the environment and the genotype. Bread
(Ed. A.Spicer) pp. 67-89, London, Applied Science Publishers, 1975
23. PassiouraJ.B.: The effect of root geometry on the yield of wheat growing on stored water.
Aust. J. agric. Res. 23, pp. 745-752 (1972)
24. Patterson H. D. and Silvey V.: Statutory and reconmended list trials of crop varieties in
the United Kingdom. J. Roy. Stat. Soc. 143 (in press) (1980)
25. Puslnan F.M. and Bingham J.: The effects of granular nitrogen fertilizer and a foliar
spray of urea on the yield and bread making quality of ten winter wheats. J. agric. Sci.,
Camb. 87, pp. 281-292 (1976)
26. Quarrie S. A.: Genotypic differences in leaf water potential, abscisic acid and proline con-
centrations in spring wheat during stress. Ann. Bot. (in press) (1980)
27. Rao M. V.: Wheat in India: an overview. Wheat Research in India. (Ed. P.L.Jaisival)
Indian Council of Agricultural Research, New Delhi, 1978
28. Rawson H.M., GifobrdR.M. and Breniner P.M.: Carbon dioxide exchange in relation to
sink demand in wheat. Planta 132, pp. 19-23 (1976)
29. Riggs T.J. and Hayer A.M.: A study of the inheritance and interrelationships 9 f some
agronomically important characters in spring barley. Theor. appl. Genet. 46, pp. 257-264
(1975)
30. Sampson D.R.: Additive and non additive genetic variances and genotypic correlations
for yield and other traits in oats. Can. J. Genet. Cyt. 13, pp. 864-872 (1971)
31. Silvey V.: The contribution of new varieties to increasing cereal yield in England and
Wales. J. nat. Inst. agric. Bot. 14, pp. 367-384 (1979)
32. Snape J. W., Law C. N. and Worland A.J.: Whole chromosome analysis of height in wheat.
Heredity 38, pp. 25-36 (1977)
33. Spierizi. H. J., ten Hag B. A. and Kupers L.J. P.: Relationship between green area duration
and grain yield in some varieties of spring wheat. Neth. J.agric. Sci. 19, pp. 211-222 (197 1)
34. Vogel K.P., Johnson V.A. and Mattern P.J.: Re-evaluation of common wheats from the
USDA world wheat collection for protein and lysine content. Res. Bull. 272, Agric. exp.
St., Univ. Nebraska, p. 36 (1975)
35. Wolfe M.S.: Pathogen response to fungicide use. Proc. 8th Brit. Insect. Fung. Conf.,
pp. 813-822 (1975)
36. Wolfe M.S. and Barrett J.A.: Population genetics of powdery mildew epidemics. Ann.
New York Acad. Sci. 287, pp. 151-163 (1977)

36
The Possibilities for Modification of the
Pattern of Photosynthetic Assimilation of
C02 in Relation to the Growth and Yield of
Crops
. W. Woolhouse, Department of Plant Sciences University of Leeds/England-

Sioninary

Photosynthetic CO, fixation involves the reductive pentose phosphate cycle with ribulose
1,5-bisphosphate carboxylase (RuBPcase) as the carboxylating enzyme. Loss of COz in
photorespiration involves an oxidative reaction sequence based on phosphoglycollic acid
generated by an oxygenase reaction catalysed by RuBPC-ase. It is argued that the oxygenase
reaction is not a biochemical vestige but an important fail-safe reaction which prevents
damaging accumulation of superoxide radicals under conditions of CO, depletion. Mecha-
nisms for avoiding photorespiratorv loss of CO, are briefly considered and it is concluded
that the C, mechanism offers the only feasible route to such economies at the present time.
The ecological distribution of C, and C4 species is briefly considered with particular respect to
the temperature requirements for growth of C4 species It is concluded that even the better-
known C, species of temperate regions are imperfectly adapted with respect to the prevailing
temperatures when compared to C, species from the same habitats, although this incomplete
adaptation does not appear to be due primarily to the photosynthetic apparatus. The possi-
bilities for introducing attributes of C4 photosynthesis into C, species showing better adap-
tation to low temperatures is considered and the possibility of using intermediate C,/C,
species for an approach to this problem is explored.

it is now firmly established that the basic mechanism of CO, assimilation in photo-
synthesis is mediated by the carboxylation of RuBP coupled to the reductive pentose
phosphate cycle. The enzyme which catalyses the primary carboxylation is ribulose
1,5-bisphosphate carboxylase, detailed studies of the reaction mechanism suggest the
sequence shown in Figure 1. It is also generally agreed that the inhibition of photo-
synthesis by oxygen, known originally as the Warburg effect, arises from a competitive
reaction between 02 and CO, at the site of carboxylation on the ribulose bisphosphate
carboxylase enzyme (Bosves et al. [1971j: Badger and Andr-ews [1974]: Laing et al.
[1974]). This alternative oxygenase function of the enzyme yields phosphoglycollate
which is further metabolised according to the reactions shown in Figure 2. Some
aspects of this process have been the source of controversy (Zelitch [1975] Robinson
and Gibbs [1974]) but for the present author the burden of evidence is in favour of
this scheme. It accords well with many physiological observations such as the in-
creased rate of CO, fixation at reduced pO,, lowered rates of net photosynthesis at
high temperatures at which it may be shown that the efflux of CO. in the light is

* Prof. Dr. ff. W. Woolhouse, Professor of Botany. Department of Plant Sciences, The Uni-
versity, Leeds LS2 9 J"; from 1. 10. 80 Professor of Biology, The University of East Anglia
and Director. John Innes Institute. Colney Lane, Norwich. NR4 7UH/England

37
 R R R Glcut
CabrIn O OH I Glycollote
ot (2)--C'zlOR' --- C-OR c,

I I - -
C=O C=O 0 O
I2 I I
R2
PE
R

Co2

R 1-OR'

IC _ O_ 2PGA

0 0
C=0 Oy

R2

Fig I. Overall reaction sequence for the carboxylase and oxygenase reactions of the enzyme
ribulose- 1,5-bisphosphatc carboxylase/oxygenase.

CH20PO 3

o C, HtOO C' coo-

O0

PHOTOSYNTHETIC C-O PHOTORESPIRA TORY

CARBON C-O. CARSON coo-

REDUCTION
CYCLE
CH2 0P0 OXIDATION
CYCLE coo-
I
CYCLYCLE

\C.,o1, -7/
(!, C.2O0"3O
Coo- C... JC N3 ,..

,.,-- C coo
CO0-

Fig.2. The proposed photorespiratory carbon oxidation cycle whereby phosphoglycollate

generated in the oxygenase reaction of RuBP carboxylase is further metabolised through
glycine to yield CO2, NH, and glyceric acid-3-phosphate.

increased and there is a parallel increase in the oxygenase in preference to the carbo-
xylase reactions arising partly from intrinsic changes in the enzyme and partly from
the relatively greater solubility of 0, over CO2 at higher temperatures. Other physio-
logical observations which support this mechanism include, the effect of high tempera-
tures and high pO2 in increasing the carbon dioxide compensation point. Compelling

38
evidence for the basic biochemical aspects of the scheme coies from the finding by
Summerville and Ogren [1979] that mutants of Arabidopsis thalliana lacking the
enzyme phosphoglycollate phosphatase die at ordinary ambient CO 2 concentrations,
having their tissues choked with phosphoglycollate, but will survive at elevated pCO,
when the carboxylation reaction yielding PGA is Favoured.

The raison d'6tre for photorespiration

Since the net effect of the oxygenase reaction and the subsequent photorespiratory
cycle is to depress the rate of CO, fixation, plant physiologists have been much puzzled
to know why it should take place. Many different carboxylating enzymes have been
discovered in animals and plants, but RuBP carboxylase is the only one of these which
is known to possess an alternative oxygenase function. [t is difficult to escape the con-
clusion that this oxygenase function of the enzyme would not have persisted in green
plants if there was not some very clear selective pressure favouring its retention. It may
be urged that apparently vestigeal structures are known to occur, as for example the
vermiform appendix in man, but a biochemical vestige which led to a depressed net
rate of photosynthesis would surely be at a selective disadvantage. The most plausible
explanation for this phenomenon to date, is that it functions as a safety mechanism
without which the photosynthetic apparatus would prove fatally unstable. On land and
in water it may frequently happen that the supply of CO2 for photosynthesis becomes
limiting when other conditions, particularly the level of irradiance are favourable, as
for example in dry summery weather on land when water deficits cause stomatal
closure leading to decreased uptake of CO,. or warn conditions amongst dense vege-
tation in water when the solubility and rate of diffusion of CO, become limiting. Under
conditions such as these one may frequently encounter circumstances in which the
electron transport system of photosynthesis gets out of step with the carbon assimilat-
ing system, so that the pools of ATP and N ADPH are fully charged and the reducing
traps in the electron transport chain are filled. It is in these circumstances that the
chloroplast appears particularly unstable; for example algal mutants lacking carote-
noids, in which the photosynthetic membranes are very labile, may be cultured in low
light or under reduced pO, but are photobleached under normal conditions of illumi-
nation and pO 2 . It is suggested that when electron traps such as the primary acceptors

I SOD~ P.O,

0. ) O-4 (Potential mrnmblano damage)

* Pr.m
or y accepto of PSI - ,edo, int,e,mediales- -NADPH.-4 (CO, rduc0ion)

h, , PSI PSI'

Fi,.3. Suggested reaction sequence for generation of free radicals causing damage to the
thynakoid membranes. The reaction marked (1) ) represents the suggested flow of elec-
trons from a primary acceptor for PSI to 0. when the acceptor is fUlly reduced. SOD: denotes
superoxide dismutase.

39
for PSI are filled other compounds, most notably oxygen, may accept electrons leading
to the formation of superoxide radicles and H 2 0 2 as shown in Figure 3.
The presence of high levels of the enzyme superoxide dismutase (SOD) in chloroplasts
may be taken as further circumstantial evidence of the propensity for formation of
superoxide radicles in these organelles. It would be instructive to explore the physio-
logical responses of mutants lacking the chloroplast SOD, under conditions of high
light and limiting CO2, if indeed such mutants could survive at all. In any event the
essence of the proposal which we are considering here is that under conditions of
limiting CO 2 the tendency to saturated traps and possible overburdening of the SOD
system may be obviated by continued consumption of reductants through the oxyge-
nase - photorespiratory cycle (Figure 4); the consequence is a wasteful 'burning off'
of carbon but this will be a low price to pay if the alternative is an oxidative disruption
of the thylakoid membrane system.

NADPH.
{ CO. + NH,

Reductive Phosphoglycollate
seposphat Oxidation cycle
cycle
(RPC)

41
sPhosphoglycollate
Sucrose
AT A

Starch
CO, Os

Fig.4. Diagram to illustrate the potential role of photorespiration (phosphoglycollate oxi-

dation cycle) in a 'safety' valve mechanism of NADPH, oxidation. When photosynthesis is
proceeding normally electron flow away from potential free radical formation (Figure 3)
is maintained by NADPH2 consumption in the reductive pentose phosphate cycle (RPC)
R indicates a variable resistance to uptake of CO,; when stornatal closure or other factors
increase R and thereby limit CO,fixation then the oxygenase reaction:is favoured and NADPH,
oxidation is sustained by coupling of the phosphoglycollate oxidation cycle to the RPC as
shown. Under these conditions a reversal of carbon flow from starch, broken arrows (---- -+)
may be-required to sustain the RPC cycle.

40
Consequences of the 'Oxygenase safety-valve' system

I have laboured the foregoing speculations concerning oxygenase reactions and photo-
respiration in order to emphasise that if these reactions in fact possess the selective
importance which has been suggested, it may be impracticable, or indeed a fools
errand, to try and breed out or otherwise eliminate them. in nature there appear
to have been two avenues to the circumventing of photorespiratory loss of carbon,
both of which, Crassulacean acid metabolism (CA M) and the C, syndrome (Figures 5a
and 5b). involve the introduction of a cytoplasmic carboxylation reaction which is not
susceptible to oxygen and can trap CO, which would otherwise be lost. It is significant
that the C, and CAM mechanisms, both of which are polyphyletic occurring in many
plant families, predominate and appear to have had their origins in environments
where the high radiant energy loads and limited water supplies have conspired to
constrain the stornatal responses in the direction of reduced conductance to CO thus
2
raising the potential hazards of superoxide generation noted in the preceding section.
Comparison of Figures 5a and 5b shows that a central feature of the fully developed
C, syndrome is the spatial separation of the supplementary PEP carboxylation and
the main RuBP carboxylation, between the mesophyll and bundle sheath respectively.
In CAM on the ofherhand both carboxylases are located in the same cell but differen-
tial regulatory mechanisms for the two enzymes ensure that generally speaking their
activities are separated in time ( Woolhouse [1978]).
Crassulacean acid metabolism generally occurs in succulents which are distributed
over many of the climatic zones of the world often occupying niches which may

hHexose
......... s .6rc.

PGA
Rea.
. .DA..Reductive
.. tion
Gluc;ePp en to se
phosphate
cycle
RuD
co, Pyruvate

all theMalate

Reactions occurring in DARK

m Reactions occurring in LIGHT

Fige. 5a. Diagrammatic representation of Crassulacean Acid Metabolism. The reacion sequence
indicated in the shaded portion of the diagram indicates events which take place in the dark
period; the reactions which occur in the light are shown in the unshaded portion.

41
 COI PEP - OPLASTID
0. oloocelo'e MesOphyll'I

I Cell

Mo oltpy

. hnIle

CALVIN OCYCLE4- RIWuOSe 1.5bisph osphate

Fig.Sb. Diagrammatic representation of C4 photosynthesis showing the separation of the two

carboxylation reactions between different cell types.
experience occasional severe droughts in otherwise mesic environments. Thus for
example CAM species may be found on cliffs and dry stone walls in Britain and other
parts of the temperate zones and as epiphytes on rainforest trees as well as in the
ground flora of arid regions.
The C4 mechanism of photosynthesis is found predominantly in plants of warm
regions in habitats where direct insolation is high. The distribution of C4 species with
respect to latitude in North America has been studied for monocotyledons (Teeri
and Stowe [1976]) and dicotyledons (Stowe and Teeri [1978]). In the case of the
monocotyledons, a stepwise multiple regression analysis to examine the distribution
of C4 grasses in relation to a number of climatic variables in 27 regions of North
America showed a significant correlation with only one parameter, the normal July
minimum temperature (Figure 6). Distribution of the C4 dicotyledons correlated not
with temperature but with rates of summer pan evaporation and dryness ratio of the
habitat.

42
 0,*
0

30 40 50 60 70 80
Normal July minimum temperature, *F

Fie.6. Relationship between the distribution of C, grasses in North America and the mean
July minimum temperature. (From Tee,i and Siowe [1976]).

In a study of the distribution of C4 species with altitude in Kenya Tieszell et al. [1979]
showed a correlation between the decreasing numbers of C4 species and increasing
numbers of C3 species with increased altitude. Exceptions to the correlation could be
related to features of' the microhabitat and local rainfall factors.
An interesting example of the responses of C3 and C4 species in the same environment
is provided by Ceratoides lanata (C 3) and Atriplex conferta (C.,) in the Great Basin of
Utah. These two species occur together in populations which appear to be in equili-
brium: when their growth. productivity, carbon balance and water consumption are
compared on an annual basis the two species are seen to perform similarly. Only when
these parameters are investigated in more detail at monthly intervals do differences
become apparent, Atriplex performing less well than Ceratoides in the cooler spring
and autumn months but sustaining a higher growth rate in the hot dry months of July
and August. Thus we Mist conclude that although the two species occur naturally
together the adaptation of the C, Atriplex has not proceeded far enough to allow the
same growth rates on the C3 Ceat1ides under cool conditions.
Our studies with the C4 species Sparlina x townsendii reveal a similar situation. Figure 7
shows the rate of photosynthesis as a function temperature in S. x townsendit compared
with that of C, grasses which occur naturally in the same region (Long and Woolhouse
1 1979]). In contrast to other C grasses S. x townsendii photosynthesised at similar
rates to the C3 species at 5°C and 1OOC and at up to twice the rate of the C , species
at higher temperatures. In order to investigate the photosynthetic response of S. x
iownsendii in more detail measurements were made of rates of photosynthesis inder
controlled conditions in the laboratory, simulating the range of variables likely to be
encountered in the field (Long and Woolhouse /1978a and 1978b]) from which rates
of photosynthesis in the field could be predicted. The validity of the predictions is
dependent on the relative importance of other factors, probably acting indirectly,
which might modify the photosynthetic rate under field conditions; they were tested
by means of measurements of the rate of photosynthesis in the field (Long and Wool-
house [1979]; Longq and Incoll [1979]). Table I compares rates of photosynthesis in
the field with those predicted for the same leaves using the laboratory data. There is a
substantial measure of agreement between the two except in the early part of the year
(see below).

43
 200
C

1 C
50

O,

0 0-

30 40
0

Leaf temperature ('C)

rate in Sprlina ownsendii (Q4 () com-
Fig. 7. Temperature response of photosynthetic () and Sesleria albeans (C).
pared to that of another C species Pennisetunpurpureta
(From Long, ncoll and Woolhouse [1975]).

from laboratory
Table 1. Comparison of the mean values of leaf photosynthetic rate predicted
(Fc
)
with the actual values recorded in the field (Fc) (after Long and Incoll [1980]).
studies
F F .'
Date
0.09 0.12 -
12 A pril 1974 ................................. 0.39 0.58 -
21 A pril 1974 ................................. 0.36*
0.29
12 M ay 1974 ................................. 0.59 0.67
1 June 1974 .................................. 0.50 0.53
15 June 1974 ................................. 0.49 0.53
11 July 1974 .................... : .............
0.76 0.85
25 July 1974 .................................. 0.78 0.80
23 August 1974 ............................... 0.86
1974 ............................. 0.90
18 September 0.66 0.73
29 September 1974 .............................
0.51 0.54
23 October 1974 .............................. 0.39 0.41
17 N ovem ber 1974 .............................
rate, in ng (CO) a(leaf area)-2
F is the mean of all measurements of leaf photosyntheticFC is the mean leaf photosynthetic
each date.
s made between 11.00 and 13.00 I G MT on
- ,
photon flux densities and leaf temperatures
rate predicted from the laboratory data using the
time that the field measurements were made. An asterisk indicates that the
recorded at the
and F' is significant at the 95% confidence interval of Student's t-test.
difference between Fc

44
Temperature adaptation in C, and C4 species

Although the success of Spartina x townsendri in an ecological sense cannot be denied

(since its origin as an allopolyploid in the late nineteenth century ithas come to
occupy some 25,000 hectares around the coasts of N.W. Europe); the limitations
imposed by the temperature requirements of growth noted for other C4 species still
apply. At the time of writing (12 May 1980) the author has just visited a salt marsh on
the East coast of Yorkshire at which location the two dominant species are Pu,ccinelfia
,narithna (C,) and S. x townsendii (C). At this date the Puccinellfie is in active and
vigorous growth whilst the first shoots of Sparlina are only just emerging. It is interest-
ing to note that the model for photosynthesis in Sparlina under field conditions,
(Long and Incoll [1980]), Table I,conspicuously breaks down in the early spring
period, suggesting that photosynthesis pet- se may have little to do with the inability of
the species to grow in this early spring period.
Another interesting example of adaptation of growth in relation to temperature, in
this case on a shorter time scale, is shown in Figure 8,which compares leaf extension
rates for Dactylis glonerata (C3 ) and Paspahnn dilatatin (C4 ) during days in the month
of August 1978. Over the period in question the weather was warm and dry and the
diurnal increment of leaf extension was similar for the two species. When however

1278--24- -

o &20
B06
8 12
.,

E
800

E 400

8 10 12 14 16 16 20 22

Time (GNIT)

Fig.8. Upper trace: Auxanonleter traces of the rate of leaf extension in Paspahon dilatatin
and DactklisgIo/nerataduring a day in August 1978 at the University Field Station at Leeds.
M/idle trace: ThernocouplC measurements of the temperature of tIle apical Tneristenl of
Pas,lhne (P) over the same period.
Lower troace: Incident radiation measured over the same p-riod using a Kipp Solarimeter.
(Siotn and If'oolouse, unpublished resuIts). For explanation see text.

45
the extension growth was investigated in some detail using an auxanometer, it was
found that the growth of Dactylis was taking place almost entirely at night, the water
potentials by day being too low to provide the required turgor for growth; whereas
the growth of Paspalum was confined to the daytime, the night temperatures being too
low to permit growth. Clearly in this instance temperature requirements for processes
other than photosynthesis were limiting growth of the Paspalum.

Temperature limitations to growth in C3 and C4 species

It is widely recognised that low temperatures are the primary limitation to plant
growth and yield in the agricultural systems of the temperate zones ( Waring [1971]).
A great deal of effort has been put into the evaluation and selection of genotypes of
many crop plants in order to obtain forms which are able to survive, grow and develop
at low temperatures. Significant success has been achieved in selection of frost resistant
strains in many species but the ability to actually grow at low temperatures appears to
be a more deep-seated characteristic which is less prone to intraspecific variation.
At the present time there is interest in the development of maize (C4) as a forage crop
in Britain. Significant progress has been made with the breeding of varieties with
desirable foliage characteristics and the ability to make rapid growth during the summer
months. The really substantial limitation to the advancement of the crop is the slow
rate of growth under the cool conditions of an average English spring. It is of interest
to note that the problem appears to reside primarily with the capacity for cell division
and expansion at low temperatures rather than with the slower rate of photosynthesis
(Monteith and Elsion [1971]). Attempts to find varieties with better low temperature
characteristics have been made by searching for stocks grown at high altitude in the
Andes; closer investigation of these genotypes in their natural surroundings however,
reveals that they take much longer to grow - often from 12-15 months from sowing to
harvest; i.e. they survive but do not grow significantly in the cooler parts of the day and
so have little to offer with respect to the contracted seasons of the temperate zones.
Some degree of intraspecific variation in ability to grow at low temperatures has been
found in the C3 forage grasses, particularly Lolium perenne, Festuca arundinacea and
Dactylis glotnerata (Cooper [1964]; MacColl and Cooper [1967]; Robson [1967]).
These results however were obtained purely by selection and the forms obtained show
only limited growth responses at lower temperatures; the intrinsic factors which limit
growth at low temperatures were not investigated. Deeper insight into these problems
is afforded by the work of Raison and colleagues suggesting that the physical state of
membrane lipids may be a major factor determining the lower temperature limit for
growth of a given species (Raison [1973]; Pike and Berry [1979]). Table 2 shows the
phase separation temperatures for phospholipids of a group of cool temperate C3
grasses and a group of C4 species from warmer climates. These values were obtained
using the polyene fatty acid, trans parinaric acid which dissolves in the membrane
lipids and changes the intensity and polarisation of its fluorescence with phase transi-
tions in the surrounding lipid molecules. Comparative studies of winter-active and
summer-active species from the same environment have shown close correlations
between these phospholipid phase separation temperatures and the prevailing tempera-
ture during the period of active growth (Pike and Berry [1979]). The problem which
we now face is whether fundamental attributes such as the composition of these

46
Table 2. Phospholipid phase separation temperatures for annual grasses

Species' Separation Temperature2 ('C)

Cool Climate
A venafa an ............... .................... - 9
Avena sativa .................................... - II
Bronius rigidas ............. ................... . 10
fordel n vu&rare .................................
.. 6
Warm Climate
Chloris virgaia ................................... 4
Digitaria saiquin,als ............................ 8
Pan ict...texantan . ..................... .......... 7
Zea m ays .. .................................... 9

All plants were grown at constant 27°C and a 16 hr pholoperiod.

2 Determnined as the average value from trans-parinaric acid fluorescence intensity and
polarization ratio plots of data from 2 samples.

membrane phospholipids are so inimical to a species as to be not readily modifiable

by selection, and if so, are we always going to find that when evolution has proceeded
to the level of differentiation of the C, syndrome it has always done so in environments
wherein the species possess temperature requirements for growth which cannot be
readily modified. If this temperature limitation should prove to be rigid then will this
preclude the significant exploitation of the C4 process for recycling of carbon under
temperate conditions? Recent studies of intermediate C,/C, species suggest that this
may be an over-pessimistic view.

The relative yields of C3 and C4 crops and the question of intermediate forms

Before there is any embarkation on a crusade to introduce the C4 mechanism for

carbon conservation into a wider range of crops including temperate species it is
important to recognise that the view that C4 species are capable of higher yields under
field conditions has been challenged (Gijford [1974]). However, Table 3 modified
after Monteith (1978), shows that when a careful assessment of published results is
made, the maximum short term growth rates of C, species are consistently 20% or more
in excess of those of C, species, Figure 9 shows that this advantage is also reflected in
the standing dry weight of the crops at harvest. These results point strongly to the

Table 3. Maximum short-term growth rates

Group Species C (g r- d-) Reference
C, Penniseu, ityphoides ...... 54 Begg [1965j
Zen mays ................ 52 Aurata and Togari [1975]
52 Williams et al. [1965]
S.r./h,m viagare .......... 51 Loomis and Williams (1963]
C, Agrosienma githago 39 Newton [1968]
Sozant,n tuberasan ........ 37 Lorenz [1944]
Orvz, satliva .............. 36 Mhrata and Togari [1975]
Typha latifolia ............ 34 Penfoiund [ 1956]

47
 80

03

2: A
40

r) 0

l00 200 300

Length of growing season (days)

Fig.9. Standing dry weight of crops at harvest in relation to length of growing season. (From
Monteith [1978]).
C3 kale C ,4bulrush millet +
potatoes 7 maize y
sugar beet 0 sorghum A
rice 0 sugar cane 0
cassava 0 napier grass U
oil palm &

potential for increased yields if net rates of photosynthesis can be increased; such
arguments would apply even in temperate climates in the warm summer months,
provided that such benefits were not offset by the adverse effects of temperature on
growth which, as we have seen, applies for most C4 crops which,are currently available,
in the early and later parts of the season.
A possible new approach to this problem is suggested by the discovery of species
intermediate between C3 and C4 in their physiological responses and biochemical
make-up (Brown and Brown [1975]; Kennedy and Laetsch [1974]) and of intraspecific,

48
ecotypic, variation in the extent of development of C, photosynthesis in the species
Mollugo verticillata (Sayre and Kennedy [1977]; Sayre et al. [1979j).
In Mo//ago chloroplasts are present in both mesophyll and bundle sheath and RuBP
carboxylase is present in all of the chlorogenous cells. Some of the "'CO,enters meta-
bolism by the immediate carboxylation of RuBP and some by the preliminary car-
boxylation of PEP to form malate. If, as seems probable, malate can enter and be
decarboxylated in plastids of the mesophyll or the bundle sheath, we have the poten-
tial for a partial re-cycling of CO, which is not dependent upon a fully developed
Krantz anatomy. Sayre et al. (1979) have shown ecotypic variation in the levels of
a number of enzymes associated with the C3 and C4 pathways in different ecotypes of
Xl. verlicullaia. The quantitative aspects of Soyre et al.'s results leave much to be
desired, as assessed from a reading of their published methods. This need not however
detract from the exciting possibilities raised by their work, including such matters as
the factors regulating the relative levels of the key enzymes of C3 and C4 metabolism
and the impact of changes in these levels in C, species with respect to their effects on
recycling of CO,, the net rate of photosynthesis and the growth rates of the plants.

Conclusion

The biochemistry of photorespiration and of the C4 pump which is favoured tinder

conditions of high photorespiration are becoming understood. The ecological aspects
of the advantages of the C, syndrome tinder a range of conditions are less clear, being
complicated by temperature responses affecting other aspects of growth than the
photosynthetic input. There is a continuing challenge to understand the genetic
control of the C., syndrome in order that its introduction into species having lower
temperature requirements for growth may be explored.

References

I. Badger, l. R. and Andrews, T.J.: Effects of CO,, 0, and temperature on a high affinity
forin or ribulosC diphosphate carboxylase-oxygenase from Spinach. Biochem. Biophys.
Res. Commun. 60, 204-210 (1974)
2. Bowes. G.. Ogren, WL. I.and Hageman. R. If.: Phosphoglycollate production catalysed
by ribulose diphosphate carboxylasc. Biochem. Biophys, Res. Common. 45, 716-722(1971)
3. Brown, R.H. and Brown, W.V.: Photosynthetic characteristics of Panicton iniloides,a
species with reduced photorespiration. Crop Sci.. 15, 681-685 (1975)
4. Cooper, J. P.: Climatic variation in forag- grasses. 1. Leaf development in climatic races
of Lolium and Daetylis. J. Appl. Ecol. /, 45-62 (1964)
5. Gilford, R.M.: A comparison of potential photosynthesis. productivity and yield of
plant species with differing photosynthetic metabolism. Aust. J. Plant Physiol. 1, 107-117
(1974)
6. Kennedy, R. A. and Laeisch, W. M.: Plant species intermediate for C,/C, photosynthesis.
Science, 184, 1087-1089 (1974)
7. Laink. iW.A.. Ogren, W. L. and Hageman, R. H.: Regulation of soybean net photosynthetic
CO, fixation by the interaction of CO2, 0. and ribulose-1,5 diphosphate carboxylase.
Plant Physiol. 54, 678-685 (1974)
8. Lo,kg, S.Y, and Woulhou.e. If. It.: The responses of net photosynthesis to vapour pressure
deficit and CO, concentration in Spartina x townsendii (sense lato). J. Exp. Bot. 29,
567-577 (1978a)

49
 9. Long, S.P. and Woolhouse, H. WV.: The responses of net photosynthesis to light and
temperature in Spartina x townsendii (sensu Into). J. Exp. Bot. 29, 803-814 (1978b)
10. Long, S. P. and Woothouse, H. W.: Primary production of Spartina marshes. Proceedings
of the 1st European Ecological Symposium. Blackwells (1979)
11. Long, S.P. and Incoll, L.D.: The prediction and measurement of photosynthetic rate of
Spartina x townsendii (sensu Iota) in the field. J. Applied. Ecol., 16, 879-891 (1979)
12. MacColl, D. and Cooper, J.P.: Climatic variation in forage grasses. 111. Seasonal changes
in growth and assimilation in climatic races of Loliun, Dactylis and Festuca. J. Appl.
Ecol. 4, 113-127 (1967)
13. Monteth, J.L.: Reassessment of maximum growth rates for C 3 and C4 crops. Exp.
Agric. 14, 1-5 (1978)
14. Monteith, J.L. and Elston, J.F.: Microclimatology and crop production. In: Potential
Crop Production. Ed.: P.F.Wareing and J..P. Cooper. 23-42, 1971
15. Pike, C.S. and Berry, J.A.: Phase separation temperatures of phospholipids from warm
and cool climate plants. Carnegie Inst. Year Book. 163-168 (1979)
16. Raison, J.K.: Temperature-induced phase changes in membrane lipids and their influence
on metabolic regulation. Symp. Soc. Exp. Biol. 27, 485-512 (1973)
17. Robinson, J.M. and Gibbs, M.: Photosynthetic intermediates, the Warburg effect, and
glycollate synthesis in isolated spinach chloroplasts. Plant Physiol. 53, 790-797 (1974)
18. Robson, M.J.: A comparison of British and North African varieties of tall Fescue (Festuca
arundinacea).Leaf growth during winter and the effects on it of temperature and daylight.
J. Appl. Ecol. 4, 475-484 (1967)
19. Sayre, R. T_ and Kennedy, R.A.: Ecolypic differences in C, and C, photosynthetic activity
in Mollago verticillata; a C.-C 4 intermediate plant. Planta 134, 257-262 (1977)
20. Sayre, R.T., Kennedy, R.A. and Pr/ngnitz, D.J.: Photosynthetic enzyme activities and
localisation in Mollugo verticillata populations differing in the levels of C, and C, cycle
operation. Plant Physiol., 64, 293-299 (1979)
21. Stowe, L.G. and Teeri, J.A.: The geographic distribution of C4 species of the Dicotyle-
donac in relation to climate. American Naturalist, 112, 609-623 (1978)
22. Somnerville, C.R. and Ogren, W.L.: A phosphoglycollate phosphatase-deficient mutant
of Arabidopsis. Nature, 280, 833-836 (1979)
23. Teeri, J.A. and Stowe, L.G.: Climatic patterns and the distribution of C4 grasses in
North America. Oecologia (Berl.) 23, 1-12 (1976)
24. Tieszen, L.L., Seryhnba, M.M., hnbanba, S.K. and Troughton, J.H.: The distribution
of C, and C4 grasses and carbon isotope discrimination along an altitudinal-and moisture
gradient in Kenya. Oecologia, 37, 337-350 (1979)
25. Waring, P.F.: Potential crop production in Britain - some conclusions. In: Potential
Crop Production. Ed. P.F_ Waring and J.P,Cooper.362, 1971
26. Woolhouse, ff. W.: Light gathering and carbon assimilation processes in photosynthesis,
their adaptive modifications and significance for agriculture. Endeavour New Series, 2,
35-40 (1978)
27. Zeliteh, L: Improving the efficiency of photosynthesis, Science. 188, 626-633 (1975)

50
Assimilate Transport through Phloem Tissue
K. Mengel, Institute of Plant Nutrition, Justus-Liebig-Universiiy Giessen/Federal Republic
of Germany*

Summary

Phloem strands represent the highway on which photosynthates are transported from the
physiological source to the physiological sink. For most crop species the major organic
solutes translocated in the phloem are sucrose and amino acids. It is suggested that these
compounds arc released from the mesophyll cells into the apoplast, from where they are
actively absorved by the sieve cell companion cell complex. This active transport is mediated
by an ATPase, which pumps H, into the apoplast and drives the active sucrose uptake in
form of a H '-sucrose cotransport.
Phloem unloading is a less well understood process. It is suggested that it is mainly a passive
leakage of solutes out of the phloem cells. The solutes follow a concentration gradient which
is maintained by the metabolic activity of the physiological sink.
The phloen transport is basically a mass flow driven by the osmotic uptake of water in the
physiological source and by the water release in the physiological sink. The transport process
itself is passive, but phloem loading as well as the processes in the physiological sink inducing
a concentration gradient are active processes.
Potassium has a specific effect t,n phloem transport, probably by promoting the phloem load-
ing.

I. Introduction

Crop plants are highly developed organisms comparable in some respect with modern
economic systems in which production and consumption centres are located at various
places and connected by transport ways such as water channels, railways or highways.
The same is true for crop plants which produce photosynthates mainly in the leaves.
A major proportion of these photosynthates is transported to other plant parts such
as stems, roots and fruits in which organs some of the assimilates are stored and some
of them respired. The production centres of crop plants are termed source, the centres
for storage and consumption sink. The most important 'highway' by which both are
connected is the phloem tissue. Xylem tissue represents also a major 'highway' in
crop plants, but it is mainly used for the transportation of water and inorganic solutes,
whereas the phloem tissue is the main pathway for the transportation of organic solutes,
predominantly sucrose and amino acids. Transportation of these compounds from

* Prof.Dr. K. Mengel, Institut fur Pflanzenernihrutng, Justus-Liebig-Universitit, Sodanlage 6,

D-63 Giessen/Federal Republic of Germany

51
source to sink may be a limited process in crop and yield production and it is for this
reason that phloem transport merits attention for agronomists, too [11].
In this paper only a brief outline on the topic 'phloem transport' can be presented.
Mainly physiological as well as nutritional and to a lesser extent morphological and
anatomical questions will be considered. In many cases deductions and derivations
from plant species other than crops will be made as there are still few experimental
results for crop species.

2. Phloem loading

It is now generally accepted that the site of phloem loading is the sieve cell companion
cell complex [14]. Thus experiments carried out with C-14 labelled CO2 showed that
newly synthesised assimilates are mainly accumulated in the companion cell [49].
These are connected with the sieve cells by numerous plasmodesmata and both are
considered as a functional unit for collecting photosynthates in the physiological sink.
The sieve cell companion cell complexes form a narrow network (minor vein system)
of which an autoradiograph is shown in Plate ]. The mean distance of the mesophyll
cells from these fine veins is only about 70 ltm or the diameter of about two mesophyll
cells. This may show that the network is rather narrow and thus well suited for
collecting the photosynthates produced in the mesophyll cells.
The transport pathway from the mesophyll cell towards the minor veins is not yet
completely clear. The pathway may be a symplasmic one following the endoplasma
retiCulum extending from the mesophyll cells through the plasmodesmata to the sieve
elements or companion cells [14]. Experimental data of Geiger et al. [16] obtained
with sugar beet leaves, however, provide convincing evidence, that the free space
(apoplast) of the mesophyll tissue is an essential part of the pathway photosynthates
follow when transported from the mesophyll cells towards the sieve cell companion
cell complex. It is assumed, that sucrose is synthesized in the cytoplasm of the nieso-
phyll cells from triosephosphates [29]. Sucrose leaks into the apoplast and from
here it is actively secreted into the sieve cell companion cell complex. At this point
sucrose must be transported against a steep concentration gradient. The range of
sucrose concentrations in the various tissues of sugar beet is shown in Table 1 [30],
giving evidence thai the sucrose concentration in the veins is higher by a factor of 10
than the sucrose concentration in the mesophyll.
The uphill transport of sucrose from the apoplast into the sieve cell requires energy
which according to experimental results of Sovonick et al. [45] is provided in the form
of ATP. This ATP dependence is supported by the fact that additions of ATP promoted,
whereas the inhibitor DNP inhibited the sucrose phloem loading process. Giaquinta
[17] has proposed a model for this sucrose loading of sieve tubes and recently substan-
tiated his hypothesis with experimental results [18].
Such a model is illutrated in Figure 1. Basically it consists of an ATPase which builds
up a proton gradient across the membrane by splitting ATP into ADP + Pj and excreting
the resulting H+ into the apoplast. It is assumed that the sucrose uptake is a sucrose -
proton cotransport mediated by a carrier of which the affinity depends on its proto-
nation.Thus according to Tanner [48] high sucrose uptake rates are observed in a pH
range of 5 to 6, whereas at pH 8 in the apoplast the sucrose uptake is virtually nil.
Analogous results were obtained by Malek and Baker [32] when studying the sucrose

52
Plate I: ALitoradtograph showing the minor vein systen of a sugar beet leaf (magnification
x 10). The aUtoradiograph was obtained after treating the leaf with C-14 labelled sucrose and
finally removing the free space sucrose by rinsing. (Photo: Geiger)

Table I. Sucrose concentration in various tissues of sugar beet 30]

Tissue Concentration, NI 10-

M esophyll .................... 3- 3.5
M inor veins .... ..................... .... ...... .... 20-25
M ajor veins .... ......................... .......... 50-80
Phloem sap ........................................ 200-300
Storage tissue of the root ............... 400-600

53
 Free space Sieve element
Sucrose
H ATP

ATP- A DP®D.P
/K®
ase

Membrane
+ +
Fig.]. Scheme of ATPase activity, H efflux, K influx and 1-1-sucrose cotransport

uptake by Ricinus petioles. Tanner [48] stresses the fact that the Km of sugar uptake
equaled to the K, of H + disappearence, a finding which provides indirect evidence that
processes are closely related. Sucrose uptake by the phloem tissue is specifically
inhibited by p-chloro-mercuribenzene sulfonic acid (= PCMBS), which also inhibits
the excretion of H + into the apoplast. Fusicoccin has the reverse effect, it stimulates
H+ production and sucrose uptake [18]. In this process K + may also be involved, a
question which will be treated more thoroughly below. It is of interest that fusicoccin,
which is a mycotoxin, activates membrane ATPase [33]. Thus the assumption, that
ATPase activity, sucrose uptake, and H+ excretion into the apoplast are closely
associated, is justified- The hypothesis that ATPase drives the active sucrose uptake of
phloem tissue is consistent with experimental results of Gilder and Cronshaw [21], who
found high ATPase activity at the surface of sieve element cells.
Besides sucrose, amino acids are abundant in phloem sap [23]. Consequently amino
acid uptake by sieve cells should also be an active process. Baker [3] reported, that
fusicoccin promoted the uptake of amino acids, too. Thus the suggestion that the
proton gradient across the membrane also drives the active uptake of amino acids
is justified.
It appears that in plant tissues the H+ gradient functions in an analogous way to the
Na+ gradient in animal tissue, the latter gradient driving the active uptake of sugars
and amino acids [26].
The rate of phloem loading depends on the provision of photosynthates which are
partially respired in order to produce ATP required for the ATPase activity. Comor [6]
suggests that for the active uptake of I sucrose, 1.5 ATP are required. The rate of
sucrose uptake depends also on the sucrose concentration in the sieve tube. High inter-
nal sucrose concentrations reduce the net uptake rate [19]. Obviously the sucrose
carrier functions like a conveyer belt of which the deloading is effected by a high

54
internal level of the transported material. Further mechanisms controlling phloem
loading are feasible, but until now not yet elucidated. Geiger [15] suggests that the
turgor pressure in the sieve tube controls the uptake of photosynthates, low Iurgor
pressure promoting and high pressure reducing the uptake. Loading depends also
on the water potential of the source leaf and too low a potential may impair the loading
process [50], A further factor of phloem loading is the physiological age of the source
leaf and the capacity of loading aquired when the leaf has reached a certain stage of
development.

3. Phloen unloading

Not much is known about the mechanism of phloem unloading. The process occurs
in the physiological sink and as this may differ depending on the kind of plant organ,
such as young leaf tissue, storage tissue of stems, roots or fruits, the mechanism of
unloading could also differ for the various organs. Some authors assume that un-
loading is merely a leakage, which means that anatomy and membrane permeability
of the phloem endings should allow a major leakage of sieve tube solutes. Jenner [28]
suggested that sucrose enters cereal grains mainly via the vascular tissue from where
it diffuses into the endosperm cavity and from here into the free space of the endo-
sperm. The transport of sucrose from the free space into the endosperm cell is also
considered to be a passive process This assumption is supported by the sucrose con-
centrations found by Jenner [28] in the various parts of a wheat grain: Vascular
bundle 220. liquid of the endosperm cavity 78, free space of the endosperm 32, and
endosrerm cells 19 mM sucrose. The sucrose leves demonstrate that there is a steep
sucrose concentration gradient allowing sucrose diffusion fjom the sieve tubes into
the endosperm cells. The diffusive flux is driven by starch synthesis, which 'consumes'
sucrose and thus maintains a sucrose concentration gradient. In all cases in which the
photosynthates are consumed by respiration or used by the synthesis of macromolecules
(polysaccharides. proteins nucleic acids) the model outlined may work, provided that
the phloem tissue does not actively reasorb the solutes which have been leaked from
the sieve cells. This question is not yet clarified. Cronshaw [7J demonstrated that
unloading tissue also showed ATPase activity.
Reabsorption of sucrose may be prevented by invertase present in the free space
splitting sucrose into glucose and fructose [10]. According to Gayler and Glasiou [13]
invertase plays a role in sucrose accumulation in the storage parenchyma cells of
sugar cane. Sucrose hydrolysis occurs in the free space and the resulting hexoses are
transported into the storage cells, in which they are again used for the resynthesis
of sucrose. According to Kirsanrov [30] the storage tissue of sugar beet lacks invertase,
and thus the sucrose of the phloem tissue is directly stored. Stein and Willenbrink [47]
suggested that sucrose is actively transported into the parenchyma cells of the sugar
beet storage tissue. The barrier, which must be overcome, is the plasmalemma and the
energy required for this active process is supposed to be ATP. Whether at this point
of sucrose accumulation an ATPasc is functioning in an inverse direction is still a
matter of speculation. The finding that plasmalemma does contain ATPase supports
such a possibility [7]. The sucrose, which is finally stored in the vacuoles, has to
permeate the tonoplast, too. According to recent results of Willenbrink and Doll [52]
the sucrose tonoplast transport shows a carrier mediated kinetic in the lower sucrose

55
concentration range (<20 mM), whereas in the higher concentration range (30-250
mM) diffusion dominates.
Phloem unloading is strictly controlled by sink metabolism and this in return has a
remarkable impact on phloem loading. Thus sink metabolism controls phloem loading
and even photosynthesis to some degree in a kind o feed back relation. This is the
reason why the photosynthetic rates adjust very closely to the demand of photo-
synthates, and in many cases it is not the assimilation in the source leaves but the meta-
bolization of photosynthates in the physiological sink which is the limiting process in
crop production. This has been shown for starch formation in wheat grains by Mengel
and Haeder [35] and for protein synthesis in barley grains by Beringer and Koch [4].
It is supposed that the sink metabolism is closely controlled by photohormones. Thus
Anfhannner and Solansky [2] showed, that application of kinetin to theears of spring
barley resulted in an increase of grain weight and in improved photosynthesis by the
flag leaf. Pbloem unloading in root nodules of legumes depends much on the strength
of the physiological sink, which means on the number of nodules and on the activity
of the Rhizobium bacteroids [43].
The flexibility in CO2 assimilation in the source leaves is not infinite. Egli el al. [9]
removed 2/3 of the leaf area of soybeans and found that such a rigorous measure
reduced the single seed weight and the number of seeds per plant considerably.

4. Mechanism of phloem transport

The mechanism of phloem transport has for years been a matter of research and of
speculation. Miinch's [40] 'Druckstrbmungstheorie' has been much discussed and
critisized but today it becomes more and more evident that the principle of the Md7nch
system seems be to correct. More than 10 years ago Biddulph [5] in an article about
'Mechanisms of Translocation of Plant Metabolites' wrote: '... it should be stressed
that there is, in reality, only one understandable mechanism of propulsion for trans-
location: This is the MUnch system. All other systems (activated diffusion, streaming
cyclosis and surface active movement) are based on unknown or inadequate mechanisms
as we known them and therefore require a belief that future undisclosed evidence will
support them. This could occur, but for the present the pragmatic person must favour
the workable hypothesis.'
The basic process of this Munch system is mass flow, which is driven by metabolic
processes of the source and of the sink. Phloem loading by the source tissue exerts a
'push' and unloading by the sink tissue a 'pull' both driving the mass flow in the
phloem [14]. There exists a logical sequence of physiological steps in this system
(Figure 2): Active phloem loading results in a decrease of the water potential of the
phloem sap; hence water moves from the surrounding tissue into the sieve cell and
thus increases its turgor. In the physiological sink the reverse process is true, the
solutes are removed from the sieve cells, thus the water potential is increased and
water moves out of the sieve cells into the surrounding tissue. Both processes, osmotic
water uptake in the sieve cells of the source and water loss from the sieve cells of the
sink, result in a water flow from the source to the sink [51, 53]. A main problem in
this system was the question whether the sieve pores of the sieve plates are blocked by
callose or not. Fisher [12] showed that in the petiole tissue of soya bean about 70%
of the sieve plate pores were essentially free from obstruction (see Plate 11). It is now

56
 Photo-
synth ate
Ph to-

Source

EI

r o synthate

' Sink

------ Water transport

.- Transport of photosynthates

Fig.2. Schematic presentation showing phlocn loading with photosynthate at the source and
unloading at the sink. Loading and unloading induce water movement from the source to

57
Plate 11: Sieve plate in a soya bean petiole which was quickly frozen in sit". The functional
condition of this sieve tube was verified by microauto-radiography. (Photo: Fisher)

generally accepted that the synthesis of callose and the blocking of sieve pores by
callose plugs occurs as a consequence of wounding phloem tissue. By this mechanism
the plant protects itself from a major loss of photosynthates [53]. In non affected
phloem tissue, however, the sieve pores are open and thus allow a mass flow from one
sieve tube to the other.

58
From this 'push and pull' system, as described above and shown in Figure 2, it can
easily be deduced that the metabolic activity of the source and especially of the sink
influences the rate of mass flow in the phlocm. A high consumption of photosynthates
in the sink will result in a steeper concentration gradient and thus promote the release
of photosynthates from the sieve cells.
Partition of the photosynthates in the entire plant also has to be considered under the
aspect of the strength of the physiological sink, but depends also on the distance
between source and sink. Generally lower leaves supply roots and lower plant parts
with photosynthates, whereas the upper leaves mainly provide younger leaves and
fruits with assimilates [31]. This pattern of assimilate distribution is consistent with
the assumption that the phloem flow Follows a pressure gradient

d',p/da
= Turgor pressure
a - Distance

According to this system phloem transport is a passive process, driven by loading and
unloading of the sieve cells. ATP consumed by the pholem is supposed to maintain
the phloem cells but not to drive the phloem flow [51]. The effect of temperature on
phloem transport is, according to Giaquinta and Geiger (20], related to the viscosity
dependence of the phloem sap on temperature and not a consequence of a low meta-
bolic activity.
The composition of the phloem sap may vary from species to species and also depends
much on metabolic condiions. As the aquirement of phloem sap meets with difficulties
in many crop species often only indirect information is available about the phloetn sap
composition. [t is suggested that in most plant species sucrose is the most important
organic solute (200-300 mM). followed by amino acids (30-20X) mM) and by organic
acids [23, 36]. Of the amino acids glutamine was most abundant in the phloem sap
of Ricinus commul.i. According to Secer (44] of all soluble amino acids in developing
wheat grains glutarnine showed the highest N turnover This finding may be an indirect
proof that glutamine plays a major role as phloem sap constituent in cereals also.
K + by far dominates the inorganic solutes. Whether it is directly involved in phloem
transport. will be considered in the following section.

5. Effect of potassium on phloern transport

The importance of K+ for phloem transport has been recently treated by Mengel [34]
and Mengel and Kirkhv 37]. For this reason only a brief comment will be given here.
There exists a remarkable amount of data proving the specific and favourable effect
of K + on the translocation of photosynthates from the physiological source towards
the sink. As this K:' effect has also been observed in cases in which photosynthesis was
not influenced by K+, the promoting K+ etfect cannot be regarded as an indirect one
related to the higher supply with photosynthates [24, 38]. Meikgel and Haeder [36] in
studying the pholem flux rate and pholem sap composition of Ricinus co,mmunis
found that K - scarcely influenced the phlocm sap composition but increased the flux
rate significantly. Table 2 shows the phloem sap composition of the two K± treat-

59
 Table 2. Effect of K' supply on the solute concentration in the phlocni sap of Ricinus conumnis
[36]

mM
K, K2
Sucrose ...................................................... 228 238
Raffi nose .................................................... 1.7 2.5-
G lucose 6 phosphate .......................................... 0.56 0.82**-
Fructose 6 phosphate ......................................... 0.17 0.25**
A m ino acids ................................................. 192 192
A T P ...................................... ............... 0.73 0.72
U T P ........................................................ 0.14 0.15
UD P-glucose ................................................ 0.45 0.50
M alate .......................... ........................... 0 .83 1.32*
K ........................................................... 47 66**
Osmotic potential (in bar) ..................................... -12.5 -14.5**
Significant difference at * 5%, ** 1% and *** 0.1 % level

ments. It is obvious that the concentration of the main solutes, sucrose and amino
acids, did not differ significantly, although the flux rate in the K, treatment was nearly
twice as high as in the K,-treatment. This means that the flux rate of the phloem solutes
was also nearly doubled. Mengel and Haeder [36] suggested that the higher K +
treatmant resultet in a higher provision of ATP, wich in return favoured the pholem
loading process.
Because of the particular experimental set up in the Ricinus experiment of Mengel and
Haeder [36] the sink metabolism was eliminated and thus could not be responsible for
the higher flux rate.
Protein- and polysaccharide synthesis are in many cases main processes of the physio-
logical sink. Both are promoted by K+, as K + activates starch synthases and also
enzymes responsible for protein synthesis. In vitro experiments have shown that
maximum K + activation is obtained within a concentration range of between 40 to
80 mM K. This is a level which is frequently exceeded in plant tissues. Thus Pierce and
tfighibotham [42] reported a K + concentration of 140-215 mM in the cytoplasm of
Avena coleoptile cells. Even in the press sap of K + deficient tissues of bean leaves
Arneke [1] found a K + concentration of 50-70 mM. From this evidence it would
appear that under conditions of suboptimum K + supply, the K + concentration for
optimum enzyme activation may still be adequate. This statement is supported by
recent experimental data of Mengel et al. [39], who found that K+ promoted the
amino acid turnover in developing wheat grains. This favourable effect, however, was
not related to the K + content of the grains, but resulted exclusively from the import
rate of amino acids into the grains.
An analogous situation is true for the starch synthesis in potato tubers. Hawker et al.
[25] found in in vitro experiments that maximum K + activation of starch synthesizing
enzymes was found in a concentration range of 50-100 mM K+. in a treatment
suffering heavily from K + deficiency Haeder et al. [22] found a K+ tuber content of
2.4% K in the dry matter which is about 100 mM K + in the press sap. These examples
may demonstrate that it is unlikely that the favourable effect K+ has on the long
distance transport of photosynthates is related to the influence of K+ on metabolic
processes in the sink.

60
It is an intriguing question whether K1 has a direct effect on the ATPase responsible
for the phloem loading process. Tanner [48] reported that the ATPase produced H +
was equivalent to K;' uptake. This finding would suggest that Ky balances the
clectropotentials at either side of the membrane, a K+ function analogous to that in
the thylakoids [41]. Tanner's observation is consistent with the fact that fusicoccin,
a specific ATPasc activator, stimulated Ff+ efflux as well as K + influx [18]. Malek and
Baker [32] in studying phloem loading of hollow Rictas petioles found that K, and
low pH levels stimulated sucrose uptake by the phloem. The authors proposed a
scheme in which sugar transport is coupled to an ATPase driven H+ efflux and K+
+
influx (see Figure I). According to this scheme a certain Km level in the free space of
source leaves is required in order to drive the H+ efiLux pump. These results contrast
with experimental data of Hatchings [27], who found that K + inhibited rather than
promoted the sugar uptake of Ricinus cotyledons. This author assumes that sucrose
and H+I are absorbed in a kind of a coransport which can be affected competitively
by K 1 .
Doman and Geiger [8] have some doubts about a direct effect of K+ on phloem
loading. They found that externally supplied K + promoted the release of photo-
synthates from the mesophyll cells and they claim that the favourable effect of K + on
phloem transport results from an enhanced release of photosynthates from mesophyll
cells into the free space of the leaf tissue. Maximum release rates were obtained, when
KCI concentrations of 15 mM were applied. It is questionable whether under normal
conditions Kt concentrations of this order of magnitude prevail in the leaf apoplast.
Thus it appears that the question of the effect of Km on phloem loading is not yet
clarified and requires further research.
PotassiuL is the most abundant cation species in the phloem sap, where it balances
the anion equivalents. Whether the K + in the phloem sap has still other more specific
functions is an open question. Spanner [46] proposed an electroosmotic flow through
the sieve plates in which K + plays a major role. One assumption of his rather compli-
cated concept is that the pores of the sieve plates are occluded with P proteins. This
assumption has been disproved by Fisher's [12] investigations which showed that
most of the sieve pores are open and not blocked by callose or P-proteins. In this
context it should be emphasized that the K + concentration of the phloem sap is not
particularly high but in the range of concentration found in ocher tissues and generally
even lower than the K- concentration in meristematic tissues. Under this aspect
it appears that one of the main function of K+ in the phloem sap is to provide the
physiological sink with K+. This is at least true for sink tissues with a higher water
content.

6. Conclusions

Phloem transport is an important process in crop production. Phloem loading depends

on the photosynthetic activity, on the water potential and on the K- status of the
source leaf. In most cases the provision of photosynthates is not a limiting factor in
crop growth, whereas insufficient K. status or too low a water potential may impair
phloem loading and thus affect crop production.
Phloem unloading depends much on the metabolic activity in the physiological sink.
In vegetative growth phloem unloading and the metabolic consumption of photo-

61
synthates is unlikely to limit growth. Filling of storage tissue and especially the growth
of fruits comprise biochemical processes which may limit the consumption of photo-
synthates in the sink and thus crop growth. This is particularly true for starch synthesis
in cereal grains. The step of starch formation is often a limiting factor of grain yield.
The photosynthetic potential of a cereal by far exceeds its sink capacity. In order to
exploit this photosynthetic potential to a greater extent the limiting process of starch
synthesis in the grains should be better understood. Recent still unpublished results of
our laboratory have shown that grain growth is not primarily a question of starch
synthesizing enzymes of which the activity is a consequence and not a cause of the
growth rate. The latter seems to be controlled by phytohormones.
Future research should aim at identifying this grain growth limiting factor, in order
to overcome it.

7. References

I. Arneke, W.: Der Einfluss des Kaliums auf die Komponenten des Wasserpotentials und
auf die Wachstumsrate von Phaseolus vulgaris L. Diss. FB 19, Justus Liebig University
Giessen, 1980
2. Aufhannner, WV.and Solansky, S.: Beeinflussung der Assimilatspeicherungsprozesse in der
Sommergerstenahre durch Kinetinbehandlungen. Z. Pflanzenernfdhr. Bodenk. Heft 4,
503-515 (1976)
3. Baker, D. A., Malek, F. and Dehvar, F. D.: Phloem loading of amino acids from the
petioles of Ricinus leaves. Ber. Deutsch. Bot, Ges. 93, 203-209 (1980)
4. Beringer, H. and Koch, K.: Stickstoffmelabolismus in einer normalen und in einer lysin-
reichen Gerste bei unterschiedlicher Kaliumernthrung. Landw. Forsch. Sonderh_ 34111,
36-44, Kongressband (1977)
5. Biddulph, 0.: Mechanisms of translocation of plant metabolites. In: 'Physiological Aspects
of Crop Yield' (R.C.Dinauer, ed.) pp. 143-164. Am. Soc. Agron. Madison USA, 1969
6. Contor, E., Rotter, M., Waldhauser, E. M. and Bong Heuy Clto: Sucrose-proton symport
in the Ricinus seedling. Ber. Deutsch. Bot. Ges. 93, 211-219 (1980)
7. Cronshaw, J.: Histochemical localization of enzymes in the phloem. Ber. Deutsch. Dot.
Ges. 93, 123-139 (1980)
8. Da,n, D.C. and Geiger, D.R.: Effect of exogenously supplied foliar potassium on
phloem loading in Beta vulgaris L. Plant Physiol. 64, 528-533 (1979)
9. Egli, D.B., Gossett, D.R. and Leggett, J.E.: Effect of leaf and pod removal on the distri-
bution of C-14 labeled assimilate in soybeans. Crop Sci. 16, 791-794 (1976)
10. Eschrich, W.: Free space invertase, its possible role in phloem unloading. .Ber. Deutsch.
Bot. Ges. 93, 363-378 (1980)
11. Evans. L. T.: The physiological basis of crop yield. In: Crop Physiology (L. T.Evans, ed.)
pp. 327-355. Cambridge University Press, 1975
12. Fisher, D.B. : Structure of functional soybean sieve elements. Plant Physiol. 56, 555-569
(1975)
13. Gayler, K.R. and Glasziou, K. T.: Sugar accumulation in sugarcane (Carrier-mediated
active transport of glucose). Plant Physiol. 49, 563-568 (1972)
14. Geiger, D.R.: Phloem loading. In: Transport in Plants 1,Phloem Transport (M.H.Zn-
mer,nann and J. A. Milburn, eds.) pp. 396-431. Springer-Verlag Berlin, Heidelberg, New
York, 1975
15. Geiger, D.R.: Control of partitioning and export of carbon in leaves of higher plants.
Bot. Gaz. 140, 241-248 (1979)
16. Geiger, D.R., Sovonick, S.A., Shock, T.L. and Fellows, R.J.: Role of free space in trans-
location in sugar beet. Plant Physiol. 54, 892-898 (1974)
17. Giaquinta, R.: Possible role of pH gradient and membrane ATPase in the loading of
sucrose into the sieve tubes. Nature 267, 369-370 (1977)
18. Giaquinta,R.: Phloem loading of sucrose. Involvement of membrane ATPase and proton
transport. Plant Physiol. 63, 744-748 (1979)

62
19. Giaquinta, R.: Mechanism and control of phloern loading of sucrose. Ber, Deutsch. Bot.
Ges. 93, 187 201 (1980)
20. Giaquinto, R. and Gekqer, D. R.: Mechanism of inhibition of translocation by localized
chilling. Plant Physiol. 51, 372-377 (1973)
21, Gilder, J. and Cronsham, J.: The distribution of adenosine-triphosphatase activity in
differentiating and mature phloem cells of Nicoiana tabactopt and its relationship to
phloem transport. J. Ultrastr. Res. 44, 388-404 (1973)
22. flaeder, H.E., iWengel, K. and Forster, H.: The effect of potassium on translocation of
photosynthates and yield pattern of potato plants. J. Sci. Fd Agric. 24, 1479-1487 (1973)
23. Ill. S. Ml. and Baker, D. A.: The chemical composition of Rieincus phloem exudate.
Planta 106, 131-140 (1972)
24. Hart. C.E.: Effect of potassium deficiency upon translocation of C-14 in detached blades
of sugarcane. Plant Physiol. 45. 183-187 (1970)
25, Hawker, . S., Nwrschner. H. and Krouss, A.: Starch synthesis in developing potato tubers.
Physiol. Plant. 46, 25-30 (1979)
26. [leinz, E.: Na+-linked Transport of Organic SOlutes. Springer-Verlag, Berlin, Heidelberg.
New York, 1972
27. Itachbigs. AT: Sucrose and proton cotransport in Ricinuscotyledons. Planta 138, 237-241
(1978)
28. lAwner. C. F.: Factors in the grain regulating the accumulation of starch. In: Mechanism of
Regulation of Plant Growth (R. L. Bielewski, A. R. Ferguson, M. A. Cresswell, eds.)
pp. 901-908, The Royal Soc. of New Zealand, Wellington 1974
29, Kelly. G.j., latzko, E. and Gibbs, M.: Regulatory aspects of photosynthetic carbon
metabolism Ann. Rev. Plant 'hysiol, 27. 181-205 (1976)
30, Kursanol, A.L.: Assimilattransport und Zuckerspeicherung in der Zuckerrdbe. Z.
Zuckerind. 24. 478-487 (1974)
31. Molo., D.J. and Cho,melski, W.A.: Distribution of C-14 labeled assimilates in rape
plants. Crop Sci. 16, 530-532 (1976)
32. Malek. F. and Baker. D.A.:. Proton co-transport of sugars in phloem loading. Planta 135,
297-299 (1977)
33, Marr&, E.: Eftects of fusicoccin and hormones on plant cell membrane activities: obser-
vations and hvpothesis. It: Regulation of Cell Membrane Activities in Plants (E.Nlarrti
and 0. Ciffrri, eds.) pp. 185-2 0 2. Elesvier North Holland, Amsterdam, 1977
34. Mengel, K.: Effect of potassium on the assimilate conduction to storage tissue. Her.
Deutsch. Rot. Ges. 93, 353-362 (1980)
35. Meige, K. and Hader. ]I. E.. The effect of potassium and light intensity on the grain
yield production of spring wheat. 4th Intern. Coll. on the Control of Plant Nutrition,
pp. 463-475, Gent (1976)
36. Mengel, K. and Haeder, H.E.: Effect of pOtassitim supply onl the rate of phloem sap
exudation and tie composition of phloem sap of Riims communis. Plant Physiol. 59,
282-284 (1977)
37. Menqel K. and Kirk, E. A.: Pot assium in crop production. Adv. Agron. 1980 (in press)
38. Mentwel. K. and Viro, l.. Effect of potassium supply on the transport of photosynthates
to the fruits of tomatoes (L.Lcopersicoor esc,lewuim). Physiol. Plant. 30, 295-300 (1974)
39. leot,el, K., Se(er, M. and Koch. K.: The effect of potassito inon protein formation and
amino acid turnover in developing wheat grain. Agron. J. 1980 (in press)
40. M/nch, E.: Die Stoffbewegungen in der Ptlanze. G. Fischer-Verlag. Jena. 1930
41. Pfifger, R.: Investigations on light-induced cation fluxes of isolated chloroplasts. Her.
Deutsch. Hot. Ges. 87, 383-388 (1974)
42. Pierce, W.S. and Higinhothan, N.: Compartments and fluxes of K-, Na' and Cl- in
Arena coleoptile cells. Plant Physiol. 46, 666-673 (1970)
43. Ruassell. W.J. and Johnson. D.R.. Carbon-14 assimilate translocation in nodulated and
nonnodulated soybeans. Crop Sci. 15, 159-161 (1975)
44. Se,er. X1.: Der Einfluss des Kaliums auf die Stickstoffl'netabolisierung und die Korn-
proteinbildung bei Sommerweizen. Kali-Briefe (Bintehof) 14 (6) (1978)
45. Sovonick, S.A., GeOrer, D. R. and Fellows, R.J.: Evidence tor active phloem loading in
the minor veins of sugar beet. Plant Physiol. 54, 886-891 (1974)
46. Spanner. D. C.: Elect roosmotic flow. ln: Transport in Plants. Phlloem Transport (Af. tt.
Zimmermann and J.A.Xilh rn, eds.) pp. 301-327. Springer-Verlag, Berlin, Heidelberg.
New York, 1975

63
47. Stein, M. and Willenbrink, J.: Zur Speicherung der Saccharose in der wachsenden Zucker-
r0be. Z. Pflanzenphysiol. 79, 310-322 (1976)
48. Tanner, W.: Proton-sugar cotransport in lower and higher plants- Ber. Deutsch. Bot.
Ges. 93, 167-176 (1980)
49. Trip, P.: Sugar transport in conducting elements of sugar beet leaves. Plant Physiol. 44,
717-725 (1969)
50. Wardlaw, J.F.: The effect of water stress on translocation in relation to photosynthesis
and growth. 11. Effect during leaf development in Lolain tenulentuin. Austr. J. Biol. Sci.
22, 1-16 (1969)
51. Wardlaw, J.F.: Phloem transport: Physical chemical or impossible. Ann. Rev. Plant
Physiol. 25, 519-539 (1974)
52. Willenbrink, J. and Doll, S.: Charcteristics of the sucrose uptake system of vacuoles
isolated from red beet tissue. Kinetics and specifity of the sucrose uptake system. Planta
147, 159-162 (1979)
53. Zi,ntnermann, M.: Translocation of nutrients. In: Physiology of Plant Growth and Devel-
opment (M.B. Wilkins, ed.) pp. 383-417, 1969

64
Grain Filling and the Properties of the Sink
V. Stoy. Biological Department. Sval6f AB. Sval6v/Sweden-

Sunlinarv
Grain yield is determined by both grain number and grain weight, grain number being the
most variable determinant and grain weight the most stable one, although the weight of an
individual grain depends very much on its position within the ear. Grain growth follows an
overall pattern, which is more or less identical for all cereal crops and which is essentially
characterized by its rate and duration. Carbon assimilates to sustain this growth are produced
both before and after anthesis and may be stored temporarily at a larger scale than commonly
assumed. Grain growth can in principle be subject to source, transport or sink limitation and
different opinions concerning this question are briefly discussed. particular importance is
attached to processes which can be imagined to terminate grain growth. At the end a more
detailed presentation is given of some recent investigations on the importance of grain position
within the car on final grain weight.

1. Introduction
The physical process of grain filling starts after anthesis when fertilization has taken
place and the grains begin to develop. The foundations for this process are laid much
earlier in plant life, however, and it is essential to realize that grain filling therefore
cannot be considered as an. isolated phenomenon but has to be related to various
aspects of plant development before anthesis.
Most important among these pre-anthesis events is the process of floral initiation since
grain number is largely determined during it course and hence to a large extent the
capacity of the plant to store assimilates in the grains. Grain number is presumably
the most important yield determinant in wheat, both per unit ground area (Evans [9]
and per single culm (Ledcnt [301).and deficiencies in grain number can only be com-
pensated by increases in grain weight to a certain extent. It is true that experiments in
which the number of grains per ear was reduced artificially, have shown that the mean
weight of the remaining grains may increase considerably beyond that found in an
intact ear growing in the same environment (Stoy [45]. Bingham [4], Bremner and
Rawso,, [7/), but other experiments, in which grain number was altered by applying
different temperatUres during ear formation instead of removing grains artificially
(Warrington, Dunstone and Green [51],Rawson and Batgga [37]), indicate that in intact
plants grain weight is a much more independent character than in plants with
artificially reduced grain numbers (Figure I).
* Dr. V. Stoy, Biological Department, Sval6f AR, S-26800 Svaldv/Sweden

65
 60

50

40

0.

20 i 20 0.

0
ID 0

20 V 40 50
Weigh,i pcr grain tmgs

Fig. 1. The poor relationship between grain number and weight per grain in the temperature
transfer experiments by Rawson and Bagga [37]. The symbols represent different cultivars and
different temperature treatments.

lt is clear, on the other hand, that the final weight of an individual grain depends very
much on its position within the ear. Investigations on wheat, particularly by Australian
workers (Rawson and Evans [38], Evans, Blinghain and Roskains [11], Brentner [4],
Brener and Rawson [7]), have shown that grains nornmally differ very much in weight
both between and within the various spikelets of an ear. The reasons for the observed
variations in grain development in different grain positions within the ear are not
quite clear. There are good indications, however, that both growth potential of the
individual grains and resistance to movement of assimilates towards these grains may
play an important role (Breinner and .Rawson [7]). Before discussing these problems
in more detail it may be appropriate to mention some general features of grain growth
and of assimilate supply to the grains.

2. The growth of the cereal grain

Grain growth follows an overall pattern which is more or less identical for all cereal
crops (Figure 2). During the first two weeks following anthesis the weight of the grain
increases slowly whereas its volume expands more rapidly owing to pericarp and grain
coat expansion and intense cell division activity in the endlosperm (Evers [.14]). After
this initial lag period grain weight increases much faster and in many cases at a
remarkably constant rate during most of the grain filling period. Sometimes, parti-
cularly at high temperatures and for short grain filling periods, the growth rate starts
to decrease relatively soon, however. During the phase of rapid growth most of the
starch is synthesized and incorporated into the previously formed endosperm cells.
Towards the end of the grain filling period the dry weight increase slows down more
and more and frequently even a slight decline in weight may occur shortly before
maturity. The rate of grain filling is normally defined as the slope of the apparently
linear part of the growth curve and the duration of the grain filling period as the
extrapolated value of linear growth from floret weight at anthesis to maximum grain
weight (Sofield, Evans and Wardlaw [40]) (see also Figure 2).

66
 160
q Final
140 yield

j120 -0
4

so T-
60 -
40
4-

0 l0 20 30 40 50 60
Days after anthcsis
fPg. 2. Grain growth in maize. T is the effective filling period (after Duncan [81 from Golds-
worthy).

Both rate and duration of the grain filling process are significantly affected by various
environmental factors. ThUs high temperatures after anthesis increase the rate of grain
growth whereas the length of the grain filling period is shortened. Inversely, low

6 (a) 15/10c 1.8 (b)

21/16 *C 15/10 OC
. 50 - 1.5 21/16 C

u40, 1.2
00 -30125 'C
30 /I 0.9C / 2

S20 / 0.6 /

10- 003 /

0 0O.
20 40 60 80 10 0 20 40 60 80 100
Days after anthesis

Fie. 3. The effect of temperature on the accumulation of (a) dry weight and (b) nitrogen in
grains of the basal florets of the middle spikes of Late Mexico 120' wheat (after Evans,
W,dlaw and Fis(her [12] from Sofield, Evans and Wardlaw [40]).

67
temperatures decrease the growth rate but lengthen the filling period (Figure 3)
(Sofield, Evans and Wardlaw [40], Spiertz [42]). In most cases the increase in grain
filling duration more than compensates the decrease in grain filling rate, hence higher
final grain weights are obtained at low temperatures than at high ones (Softeld, Evans
and Wardlaw [40], Kolderup [29]). The temperature effect on grain weight is also
obtained if only the ear itself is treated as was shown by Ford, Pearman and Thorne[16].
This indicates that the increased grain growth rate and shortened grain filling period
is caused by metabolic and developmental events within the grain itself and not by
effects on the carbohydrate supply from the rest of the plant.

3. The transfer of assimilates into the grains

A constant grain growth rate can only be maintained if there is a constant flow of
assimilates into the grains. Until recently it was generally agreed that 'current'
photosynthesis, i.e. photosynthesis occurring more or less simultaneously with the
grain filling process, is the most dominating source of carbon assimilates for grain
growth (Porter, Pal and Martin [36], Buttrose and May [8], Thorne [49]). However,
there is good evidence now that normally at least part - and under special circumstances
quite a substantial fraction - of the carbon found in the mature grains must have been
fixed before anthesis (Gallagher, Biscoe and Scott [18], Gallagher, Biscoe and Hunter
[17], Bidinger, Musgrave and Fischer [3], Austin et al. [1], Stoy [48]). Moreover,
there are good reasons now to question whether photosynthetic production after
anthesis may actually be considered as 'current'.

SKANDIA nlB8 0-0 grain growth rote

eo-
Fi.4maiae ulted Tte of fixation

'12 I FJ¥
200

an 12- I 0us

I
It\

IoI

_S -'/'o -" ' 1; o .o

I&y friinhoi
Fig 4.Coparso
bewen gai Irwhrt n acltdrt ffxto fgancro
asiiae ifil-owmaeilothintrwetvreySada11B(etudo
nubise
an4S-, rslt)

68
Asa matter of fact there seems to be no difference in principle between the contributions
made by these two sources of assimilates since the products of 'current' photosynthesis
are also to a large extent.stored lemporarily before actually entering the grains. This
is very clearly demonstrated in an experiment done at our institute last summer. Small
field plots of rye, wheat and triticale were pulse-labelled with 4CO at various stages
of development, beginning about 35 days before anthesis and ending at about leaf
senescence(= ca. 45days after anthesis). Plant samples were harvested both immediately
after treatment and at full maturity and analysed for weight and radioactivity in various
plant parts. With the aid of these data the amounts of assimilate, which were produced
at the various growth stages and finally recovered from grains, were calulated. As
shown in Figure 4 fixation of carbon recovered from the grains in the winter wheat

SOLID - Rcovn,-dot
m.,tut y
40

0o 0eRoov,,ed onc
wee
rduto oft
300 r

E
20.

.3 -2 1 0 t
12 3 4 5 0 7
weeks from anthes1s

Fi,. 5. Translocation of carbon assimilates, fixed at various stages of development and

recovered from the grains in the main shoot ear one week after anthesis and at maturity.
Material: field-grown winter wheat Solid (Bertholdvron and Sfo.l, unpublished results).

cUltivar Skandia IllB had already begun 40 days before anthesis. The fixation intensity
reached at peak 12 days after anthesis but declined again almost immediately. 35 days
after anthesis there was no longer any net production of carbon assimilates but in fact
some carbon recovered from the grains was fixed even 45 days after anthesis, this
uptake however being more than compensated for by losses through respiration. In
contrast, grain growth was stable for a much longer time and did not start to decline
until carbon fixation had virtually ceased.
It is thus quite obvious that a substantial relocation of carbon assimilates must have
taken place, not only of substances formed before anthesis but also of those produced
later. Direct proof of the occurence of such relocation is given by an experiment,
conducted at Svaldv in 1978. Whole plants were pulse-labelled with 14CO, as in the
1979 experiment. One third of these plants was harvested immediately after treatment,
another third one week after treatment and the remaining third at full maturity. The
results of this experiment (Figure 5) demonstrate very clearly that a large proportion
of the grain carbon, fixed in the weeks following anthesis, was not translocated directly
to the grains (i.e. within a week) but stored tenporarily somewhere else. These surplus
assimilates. together with those formed before anthesis. thus represent the source of

69
carbon used in the late phase of grain filling when 'current' photosynthesis has more
or less ceased completely. Very similar results have been reported by Austin et al. [1]
and by Spiertz and van de Haar [43].

4. Limitation and termination of grain growth

The question now arises as to what mechanism that regulates the flow of assimilates
into the grain. There are three major processes which basically can be imagined to
exert a control function (see e.g. Evans, Wardlaw and Fischer [12]. These are 1) the
photosynthetic production of carbon assimilates (source limitation), 2)the translocation
of these assimilates into the grains (transport limitation) and 3) the metabolic activity
of the grain itself (sink limitation)-
Source limitation is not likely to control grain growth during its earlier stages since,
as we have seen, there exists a definite surplus production of assimilates during the
first weeks after anthesis. This surplus is used to supplement the rapidly diminishing
flow of assimilates from 'current' photosynthesis during late stages of grain filling but
it is still an open question whether the combined contribution of assimilates from these
two sources may limit grain yield or not.
Jenner [20, 21, 22, 23] and Jenner and Rathjen [24, 25, 26, 27, 28] in a series of
experiments with wheat studied the question whether the concentration of sucrose in
grains, and particularly in the endosperm cells, changes with grain development in such
a way as to indicate a limitation in grain growth due to restricted assimilate supply.
They could not find support for such a hypothesis, on the contrary sucrose concen-
tration frequently increased somewhat towards the end of the starch accumulation
period, and they therefore concluded that the cessation of starch accumulation is
attributed to loss of specific metabolic activities within the endosperm itself rather
than to a response to signals from other parts of the plants (Jennerand.Rathjen[27]).
The concept of sink limitation, at least under certain conditions has also been accepted
by many other workers (Evans and Rawson [10], Bremner and Rawson [6], Sofield,
Evans and Wardlaw [40]) ; although causes for cessation of grain growth other than
impaired metabolic activity in the endosperm have been suggested by some of them.
Thus Sofieldet al.[41] found that accumulation of dry matter, nitrogen and phosphorus
by the grain, and the entry of water into it, all ceased at the time that lipids are deposited
in the pigment strand situated between the vascular bundle in the grain furrow and the
endosperm cavity. These authors therefore suggest that this deposition is actually the
event that terminates grain growth.
These considerations do not exclude, however, that source limitation may nevertheless
exist under certain circumstances. It has been pointed out earlier (Stay [46, 47]) that
a proper balance between sink and source strength is essential if maximum yields are
to be obtained in cereal crops. Similarly Austin et al. [2] stated recently that in order
to achieve further genetic gain in wheat yields total dry matter production, and hence
source potential, has to be increased in the new varieties. The reason is that harvest
index cannot possibly be raised beyond a certain ceiling value and apparently this is
already closely approached in many varieties of to-day.
The possibility of a transport limitation has also been considered by many workers.
Thus Evans, Wardlaw and Fischer [12] in reviewing previous work on this subject came
to the conclusion that 'since there are a number of conditions in which the rate of

70
assimilate production exceeds the rate of storage, the limitations to carbohydrate
transport to the ear, within the ear, within spikelets, and within grains, merits further
analysis'. Evans and co-workers (Evans er al. [13]) have measured the phloem area in
the peduncle of wheats with widely differing grain production and found a close
proportionality between these areas and the maximum rates of assimilate import by the
ears. They pointed out, however, that this does not prove that the vascular capacity
is actually limiting ear or grain growth. particularly not since phloem development
may be controlled by the ear through some kind of feed-back mechanism. Later
experiments by Wardlaw and Moncur [50] moreover demonstrated that half of the
vascular system in the peduncle may be severed without any reduction in the rate of
accumulation of shoot assimilate in the ear.

5. Grain growth as related to position within the ear

Whereas the capacity of the vascular system in the peduncle thus seems to be more
than sufficient to ensure an adequate supply of assimilate to the ear as a whole the
situation appears to be less clear with respect to the supply of the individual grains.
As already mentioned in the introduction the location of the grain within the ear is
of great importance for its final weight and several reasons can be suggested to account
for this position effect.
Experiments by Rawson and Evans [38] and Evans, Bit4g.ham and Roskin.sv[/], in
which they sterilizcd various combinations of florets in wheat cars around anthesis,
demonstrated that such sterilizations lead to compensatory grain setting in florets
which normally would have been empty. It is therefore obvious that normally there
are restrictions on grain setting in intact ears and it was concluded by Evans, Bingham
and Roskams [//] that these restrictions are correlative and hormonal in nature
rather than due to competition for assimilates. A position effect is therefore re-
cognizable already at the stage of grain setting and this effect is presumably initiated
as early as in the first phases of ear development. However, floret position is of great
importance also for the further development of the grain, once it has been set.
In a recent paper, Bremnet and Rawson [7] discuss this problem thoroughly. Figure 6,
taken from their paper, shows the variation in grain weight in an intact, mature ear
with respect to position of the spikelcts within the ear and of the florets within the
spikelets. The grains in the lower, central spikelets clearly have an advantage of some
kind over the grains in the basal and particularly in the apical spikelets of the ear.
Within each spikelet the second and third grains from the base are with few exceptions
considerably heavier than the other ones. The question is now how to explain these
effects.
According to the authors three factors in particular have to be considered in this
connection: I) the availability of assimilate. 2) the intrinsic capacity of the grain to
accumulate dry matter - 'growth potential', and 3) the resistance within the phloem
to the movement of assimilate to the grain. Changes in the relative availability of
assimilate can be introduced either by removing part of the grains from the ear or by
shading the plants. Grain removal increases assimilate availability to the remaining
ones and therefore ought to yield information about the growth potential of the grains
whereas shading reduces assimilate availability and thereby should reveal differences
in resistance to assimilate transport into the grains.

71
 (c)
3
4

67

0
35

17

isA

L3 9 5O 0 70
Weight per grain (iS)

Fig. 6. Intact wheat ear profiles showing weight per mature grain for grains a, b, c, and d in
each spikelet position (numbering from the ear apex). (After Bremner and Rawson [7j.)

Figure 7 demonstrates the results from a grain removal experiment (Brenmer and
Rawson [7]). It is obvious that the grains in most positions had a potential to grow
larger than they normally do in intact ears but it was also found that some of the
grains in fact got smaller when their surrounding neighbours in the same spikelet were
removed. This surprising effect may be explained in various ways but since it was
evident particularly when the apical grains were removed some beneficial influence on
assimilate transport into the spikelet might be attributed to the apical grains. One can
only speculate about the nature of this influence at the moment.
The grain removal experiments also gave interesting results with respect to differences
in grain development between spikelets. Thus the differences in weight between grains
from different spikelets were of the same order as in intact ears, even if only one grain
was left per spikelet. This shows that the growth potential of the grains is to a certain
extent fixed according to the position of the spikelet, and this is presumably the result
of the differences in morphological development within the spike, which are visible
already at or even before anthesis. A more synchronous development of the ear
structure might be a worthwhile goal in future plant breeding, and it has in fact been
suggested by Fisher [15] that the high grain production of wheat varieties having
Norin-10 in their ancestry depends partly on a better synchronization of ear develop-
ment in these types. Another possible way of achieving a high degree of synchronization
would be to use branched ears with a great number of simultaneously developing small
spikelets as has been suggested by Rawson and Ruwali [39].

72
 . .i L

0 :00 .0 0 0 7 1 .0 . oO 70

Or o L

F4,. 7. Wheat eat promiles ol ,eight per grain at maturity fo)r grains a, b.e,. dd forintact cars
(solid lines and circles) and for ears in which all but the grains designated bv open circles verc
removed at ) days after aniiesis (after Bretn, and Raison [7]).

The shading experiments of Bremir and Rawvsmi f[7 showed that a severe shortage of
assimilates resulted in a nearly uniform redLiCtionl in grain weight in all spikelets.
These were thus linked 6in parallel' to the SOUrCC of assimilate and there was no sign
that there were any differences in resistance to assimilate transport between them. In
contrast, there was a clear indication that there was at least a partial in series' - linkage
within the individual spikelets, that is. apical grains were affected more than basal ones.
These findings are consistent with known facts about the vascular connections within
the spikelet (Zee and O'Brie'n [53j, Hanif and Ltutge, [19]). According to these
authors the three basal grains in a wheat spikelet are connected with the rachis by
independent vascular strands, (hey are thus linked 'in parallel'. However, the lengths
of these strands diffet and may therefore offer diff'erent resistances to assimilate
movement, thereby Superimposing anl'in series'-effwc over tile 6 in par-allel'-finkage.
This 'in scries7-position is still more evident with the more distal grains in the spikelet
since these are not directly connected with the iachis.
IInConclusion1 Bremnt/er and RawonVql r7J therefore state that resistance Io miovemntn
of assimilate within the car, partiCUlarly within spikelets. may be anlimpoitant factor
in regulating the growth of individual grains and that the limitations to growth of
distal grains, as well as of the car as a whole, probably lies more in the maintenance
roeS gradin s than in the physical capacity ol the Vascular syster Itremains
aLiCit
to be found, however, how these gradents o and withinh etar are maintained and
controlled and what can be done to increase thenr magnitude and dUration. Hormonal

73
influence may be of vital importance in this case as in so many others of growth and
development (see e.g. Michael, Allinger and Wilberg [31], Michael and Seiler-
Kelbitsch [32], Monla and Michael [34], Mounla [33], Moun/a, Bangerth and Stoy
[35], Wheeler [52], Stamp [44]) and it is therefore quite appropriate that ihe whole
of the next session of this Colloquium is devoted to this intriguing aspect of plant life.

6. References

I. Austin R. B., Edrich, J. A., Ford, M. A. and Blackwell, R. D.: The fate of the dry matter
4
carbohydrates and ' C lost from the leaves and stems of wheat during grain filling. Ann
Bot. 41, 1309-1321 (1977)
2. Austin, R. B., Binghan, J., Blackwell, R. D., Evans, L. T., Ford, M. A., Morgan, C. L. and
Taylor, M.: Genetic improvements in winter wheat yields since 1900 and associated
physiological changes. J. agric. Sci., Camb. (in press) (1980)
3. Bidinger, F, Musgrave, R. B. and Fischer, R. A.: Contribution of stored pre-anthesis
assimilate to grain yield in wheat and barley. Nature 270, 431-433 (1977)
4. Bingham, J.: Investigations on the physiology of yield in winter wheat by comparison of
varieties and by artificial variation in grain number per ear, J. agric. Sci., Cam. 68, 411-
422 (1967)
5. Bremner, P. M.: Accumulation of dry matter and nitrogen by grains in different positions
of the wheat ear as influenced by shading and defoliation. Aust. J. biol. Sci. 25, 657-668
(1972)
6. Bremner, P. M. and Rawson, Ht. M.: Fixation of 14CO, by flowering and non-flowering
glumes of the wheat ear, and the pattern of transport of label to individual grains. Aust.
J. biol. Sci. 25, 921-930 (1972)
7. Bremner, P. M, and Rawson, H. M.: The weights of individual grains of the wheat ear in
relation to their growth potential', the supply of assimilate and interaction between grains.
Aust. J. Plant Physiol. 5, 61-72 (1978)
8. Buttrose, Al. S. and May, L. H.: Physiology of cereal grain. 1. The source of carbon for
the developing barley kernel. Aust. J. biol. Sci. 12, 40-52 (1959)
9. Evans, L. T.: The influence of irradiance before and after anthesis on grain yield and its
components in microcrops of wheat grown in a constant daylength and temperature
regime, Field Crops Res. 1, 5-19 (1978)
10. Evans, L. T. and Rawson, H. M.: Photosynthesis and respiration by the flag leaf and
components of the ear during grain development in wheat. Aust. J. biol. Sci. 23, 245-254
(1970)
11. Evans, L. T., Bingham, J. and Roskanzs, M. A.: The pattern of grain set within ears of
wheat. Aust. J. biol. Sci. 25, 1-8 (1972)
12. Evans, L. T., Wardlaw, I. F. and Fischer, E. A.: Wheat. fi: Evans L. T. (ed.): Crop
physiology, some case histories. pp. 101-149, Cambridge, 1975
13. Evans, L. T., Dunstone, R. L., Rawson, H. M. and Williams, R. F.: The phloem of the
wheat stem in relation to requiremnts for assimilate by the ear. Aust. J. biol. Sci. 23,
743-752 (1970)
14. Evers A. D.: Development of the endosperm of wheat. Ann. Bot. 34, 547-555 (1970)
15. Fisher, J.E.: Developmental morphology of the inflorescence in hexaploid wheat
cultivars with and without the cultivar Norm 10 in their ancestry. Can. J. Plant Sci. 53,
7-15 (1973)
16. Ford, M. A., Pearnan,J. and Thorne, G. N.: Effects of variation in ear temperature on
growth and yield of spring wheat. Ann. appl. Biol. 82, 317-333 (1976)
17. Gallagher, J. N., Biscoe, P. V. and Hunter, B.. Effects of drought on grain growth. Nature
264, 541-542 (1976)
18. Gallagher, J. M., Biscoe, P. V. and Scott, R. K.: Barley and its environment. V. Stability
of grain weight. J. appl. Ecol. 12, 319-336 (1975)
19. Hanif, M. and Longer, R. H. M.: The vascular system of the spikelet in wheat (Triticun
aestium). Ann. Bot. 36, 721-727 (1972)

74
20. Jenner, C.F.: Relationship between levels of soluble carbohydrate and starch synthesis
in detached ears of wheat. Aust. J.biol. Sci. 23, 991-1003 (1970)
21. Jeaner.C.F.: The uptake of sucrose and its conversion to starch in detached ears of wheat.
J, exptl. Bot. 24, 295-306 (1977)
22. Jeaner, C.F.: Factors in the grain regulating the accumulation of starch. In: Bieleski. i.L.,
Fere,uson, A. R. and Cresswell, M.M.(eds.): Mechsnisms of regulation of plant growth.
Roy. Soc. New Zealand, Bull. 12, pp. 90t-908, Wellington (1974)
23. Jenner, C.F.: Physiological investigations on restrictions to transport of sucrose in wheat.
Aust. J. Plant Physiol. 3, 337-347 (1976)
24. Jenner, C.F.and Rathjen. A.J.: Factors limiting the supply of sucrose to the developing
wheat grain, Ann. Bot. 36. 729-741 (1972)
25. Jlener, C.F.and Rathjen,A.J.: Limitations to the accumulation of starch in the developing
wheat grain. Ann. Bot. 36, 743-754 (1972)
26. Jener, C.F.and Rathjen, A.J.: Factors regulating the accumulation of starch in ripening
wheat grain, Aust. J. Plant Physiol. 2. 311-322 (1975)
27. Jenner. C.F.and Ratijen, A.J.: Supply of sucrose and its metabolism in developing grains
of wheat. Aust. J. Plant Physiol. 4, 691-701 (1977)
28. Jener, C.F.and Rathjen. A.J.: Physiological basis of genetic differences in the growth
of grains of six varieties of wheat. Aust. J. Plant Physiol 5,249-262 (1978)
29. Kolderup, F.: Application of different temperatures in three growth phases of wheat.
1,Effects on grain and straw yields. Acta Agr. Scand. 29, 6-10 (1979)
30. Ledent, J.F.. Relationships between cuhn yield and morphological characters in winter
wheat (Triticum aestivum.. L) genotypes. Cereal Res, Commun. 5. 89-99 (1977)
31. Michael, G and Sei/er-Kelhisch. H.: Cytokinin content and kernel size of barley grain as
affected by environmental and genetic factors. Crop Sci. 12, 162-165 (1972)
32. Michael, G., Afinget, P. and Wilherg, E.: Einige Aspekte zur hornionalen Regulation der
Korngr6sse bei Getreide. Z. Pflanzenernihr. Bodenkde. 125. 24-35 (1970)
33. Moun/a. M.A. Kh.: Gibberellin-like substance in parts of developing barley grain. Physiol.
Plant. 44, 268-272 (1978)
34. Mlouna, M. A. Kh. and Michael, G.: Gibberellin-like substances in developing barley
grain and their relation to dry weight increase. Physiol. Plant. 29. 274-276 (1973)
35. Mamla, .,.A.Kh. Ratr,qerth, F. and Stmv, V..Gibberellin-like substances and indole
type auxins indeveloping grains of normal- and high-lysine genotypes of barley. Physiol.
Plant. 48, 568-573 (1980)
36. Porter. H. K.. Pal. N. and Martin. R. V.: Physiological studies in plant nutrition. XV.
Assimilation of carbon by the ear of barley and its relation to the accumulation of dry
matter in the grain. Ann, Bot. 14, 55-68 (1950)
37. Rawson, I/. M. and 8igga. A. K.: Influence of temperature between floral initiation and
flag-leaf emergence on grain number in wheat. Aust. J. Plant Physio. 6, 391-400 (1979)
38. Rawson, H. Al. and Evats. L T. : The pattern of grain growth within the ear of wheat.
A tiSt. j. biol Sci. 23, 753-764 (19701
39. Rawson. 1. Wl.and Ratali, K. N.: Ear branching as a means of increasing grain uniformity
in wheat. Aust. J. agric. Res. 23. 551-559 (1972)
40. Soied.J.-Evas, L. T. and Wardlaau, ./. F.: The effects of temperature and light on grain
filling in wheat. In: Bieleski, R. L. Fergus'on, A. R.and Cresswell M.M. (eds.) Mechanisms
of regulation of plant growth. Roy. Soc. New Zealand. Bull. 12. pp. 909-915. Wellington
(1974)
41. Sofield. J.,Evans. L. T.. Cook. M. G. and Wardlat, I. F.:Factors influencing the rate and
duration of grain filling in wheat. Aust. .J. Plant Physiol. 4, 785-797 (1977)
42. Spiert, J.HI.J.: Grain growth and distribution of dry matter in the wheat plant as
influenced by temperature, light energy and ear size, Neth. J. agric. Sci, 22, 207-220 (1974)
43. Spiertz, J.H. J. and van de Har, H.: Cultivar differences and nitrogen effects on grain
yield, crop photosynthesis and distribution of assimilates in winter wheat. Neth. . agric.
Sci. 26. 233-249 (1978)
44. Stamp, P. H.: Untersuchungen an verschiedenen Weizensorten zur Assimilat- und Stick-
stoffeinlagerung in das Korn unter der besonderen Btricksichtigung des Einflusses Von
Wirkstoffien. Diss.. Kiel, 98 pp., 1975
45. Sto. 11.: Photosynthesis. respiration, and carbohydrate accomulation in spring wheat in
relation to yield. Physiol. Plant, Suppl. IK, 1-125 (1965)

75
46. Stoy, K4: Assimilatbildung und -verteilung als Komponenten der Ertragsbildung beim
Getreide. Angew. Bot. 47, 17-26 (1973)
47. Stoy, V.: Trockensubstanzproduktion und Assimilateinlagerung in das Getreidekorn.
2. Pflanzenernihr. Bodenkde. 140, 35-50 (1977)
48. Stoy, V.: The storage and re-mobilization of carbohydrates in cereals. In: Spiertz, J. H. J.
and Kramer, T. (eds): Crop physiology and cereal breeding, Wageningen, pp. 55-59,1979
49. Thorne, G. N.: Photosynthesis of ears and flag leaves of wheat and barley. Ann. Bot. 29,
317-329 (1965)
50. Wardlaw, 1. F. and Moncur, L.: Source, sink and hormonal control of translocation in
wheat. Planta 128, 93-100 (1976)
51. Warrington, J. J., Dunstone, R. L. and Green, L. M.: Temperature effects at three deve-
lopment stages on the yield of the wheat ear. Aust. J. agric. Res. 28, 11-27 (1977)
52. Wheeler, A. W.: Changes in growth-substance contents during growth of wheat grains.
Ann. appl. Biol. 72, 327-334 (1972)
53. Zee, S. Y. and O'Brien, r. P.: Vascular transfer cells in the wheat spikelet. Aust. J. biol.
Sci. 24, 35-49 (1971)

76
Co-ordinator's Report on the Ist Session
. irnon, Settlement Study Center, Rehovot/[srael; Member of the Scientific Board of the
International Potash Institute-

Breeding for Higher Yields

The traditional role of the coordinator is to present a balanced summary of the papers
given at his session and the following discussion. It has been agreed with the chairman
that I should deviate front this routine. My presentation will be concerned with
certain aspects of plant breeding for higher yields, that were not covered in Dr. Lupton's
stimulating paper, and with points raised by him that are controversial. Also, Dr.
Lupton has very wisely restricted himself to treating the problems of breeding for high
yields in two crops only - wheat and barley - though the title of his paper 'Breeding
for Fligher Yields' might give rise to wider expectations.
Amongst the various points raised by Dr. Lupton, was one statement on which Iwould
like to take a stand. I quote: 'Ithas been suggested that varieties selected for overall
good performance over a range of sites may yield less under specified conditions than
varieties selected specifically for those conditions'.
Breeding objectives appear to follow a cyclical pattern like ladies fashions! When
was a student, breeding for specific adaptation was the basic tenet of any breeding
programme. As a young plant breeder, I wasted a few years breeding varieties of
wheat for specific adaptation to a number of different ecological niches in what was
then known as Palestine - which within the conditions of a province-sized country,
has as many ecological niches as a respectable continent. Imagine my chagrin, when
after a few years of efforts, I found that an introduced variety, Florence-Aurore, bred
in Rabat/Morocco, consistently outyielded all my varieties itall the locations for
which they had been bred.
'ro add insult to injury - one very early selection, bred specifically for the conditions
prevailing in the Jordan valley, was shown by Vavilov, with whom we had made an
exchange of breeding material, to be the outstanding variety in the Polar Circle!
Variability within a given ecological niche, can be greater from year to year, than
between niches in a given year. Only varieties with wide adaptability can cope with
such situations.
The one lesson we have learnt in the last 30 years is that outstanding varieties in ad
crops are those with wide adaptability. Ilow otherwise explain that in our country
the outstanding varieties of sugar beets come from Holland, wheat from Mexico (with
Japanese parentage), barley from Montana and from Germany, maize from Illinois,
cotton from California, roses from France. peas from England. etc.

* Prof. Dr. I. Arnon, Settlement Study Center, P.O.B. 555, Rehovot/Israel

77
 What are the miracle wheats of the Green Revolution if not varieties with an extremely
broad spectrum of adaptability? Once you have such varieties, you can tack on one or
more locally desirable characteristics.
Another point made by Dr. Lupton is that further increases in yield potential are likely
to be associated with increases in harvest index or increases in total dry matter pro-
duction. Dr. Lupton qualifies this statement by adding that increases in the harvest
index are limited by the agronomic requirements of the crop - and increases in biomass
produced are limited too, because the harvest index is concurrently reduced.
The inevitable conclusion from this would be that we have practically almost reached
a dead end in our breeding efforts for higher yields.
And yet, there are alternatives. One is to combine in the same plant the high levels of
biological yield already achieved, with the high harvest indices available. And the
means to achieve this aim are F1 hybrids. We have been able to show that the highest
yielding dwarf hybrid sorghum grown in Israel Rs 610 is a cross between two mediocre
parents - the restorer strain has a high biological yield; the male-sterile strain has a
high harvest index. The hybrid combines the two favourable characteristics of its
parents. It does not often happen that way, but it can be done.
By analogy with wheat and barley, sorghums are also selfpollinated, and semi-dwarf
types were developed some 20 years earlier than in wheat, not in order to improve
yields, but because the engineers were unable to solve the problems of mechanical
harvesting of tall sorghum. They placed an order with plant breeders for tailor-made,
80 cm high plants.
This brings us to the question of commercial F1 hybrids of wheat and barley. The
original great success in F, hybrids - in maize - was limited to a single cross-pollinated
crop, in which manual emasculation was economically feasible. The great break-
through came with the chance discovery of a male-sterile plant in sorghum - followed
by some of the most sophisticated genetic manipulation (before genetic engineering
became fashionable) leading to the very successful F, hybrids in a number of important
crops. Progress in wheat and barley is, to the best of my knowledge, still lagging
behind - but hybrid F1 wheats and barleys are definitely a possibility which cannot be
ignored.
Breeding for drought-resistance has been a consistent theme for as long as I remember
and probably the greatest source of wasted breeding efforts in the whole field of plant
breeding.
Breeding for drought-resistance means different things to different workers: Breeding
for production under conditions of insufficient moisture supply over the growing
season, or breeding for production under overall favourable moisture conditions when
short periods of drought are encountered during the growing period.
In the first case the crop is under water stress during most of the time, with occasional
periods of favourable moisture supply, and in the second case, the plant has to over-
come occasional short periods of stress only.
The criteria used for adaptation to water stress are different. According to the above
mentioned breeding objectives:

a) ability of the plants to survive under drought conditions - this is the criterium used
by Mizrahi in his paper;
b) to endure drought with relative small loss in yield, this is the most generally used
criterium;

78
c) to be efficient in the use of water, and able to recover rapidly after short periods
of stress.

a) Survival under dioilaI conditions is based on breeding for xerophytic characters

which is generally achieved at the expense of productivity. The most frequent result
is survival of the crop - but not survival of the farmer. In Kenya. a plant breeder
told me proudly that he had been successful in breeding a drought-resistant maize
for a low-rainfall area and showed a little runt of a cob with 4-6 kernels. [ told
him that his 'success' explained why Government drought-relief was necessary in
9 years Ou of ten in order to prevent starvation of the people growing the 'drought-
resistant' maize.

b) Applying the criteriUnm of 'ability to endure drought with relative small loss in
yield leads to the absurd situation whereby a variety that produces say 1000 kg/ha
under favourable moisture conditions and 500 kg/ha under drought is by definition
more drought-resistant than a variety that produces 6 tons under favourable
conditions and 600 kg Under drought conditions.

We iltus/ accept the fact that below a certain minimum of water supply, you caMot
obtain economically acceptable yields of grain. The only alternatives are: a) to improve
available moisture in the soil by agronomic means; b) if this is not possible - not to
grow cereals. Breeding efforts areffitile for crops that are almost continuously under
water stress!
At and above this minimunm moistUre requirement, the use of varieties with wide
adaptability is the logical answer to the considerable variability characteristic of the
drought-prone areas.
Another characteristic of these areas is, that even in good years, there occur periods of
water stress. Ability to recover rapidly after such stress periods should be part of the
genetic make-up of varieties with wide adaptability for these regions.
These are the varieties with high water use efficiency (WUE). where

WUE= Yield (Y)

Evapotranspiration (ET)
ET should be as near optimum as possible; the major factor to achieve high WUE
that is amenable to manipulation is to breed for high yields under the wide variety of
conditions likely to be encountered in a given region in the course of several years.
I Would like to refer to one more objective of breeding for specific conditions which
was being advocated recently in some of the developing countries, in particular in
India. Because of the energy crisis, plant-breeders have been challenged to produce
varieties that are capable of giving high yields without the need for high levels of
fertilizers. Wherever t have had personal contacts with those responsible for breeding
policy. I attempted to dissuade them from investing money and effort in following
this latest example of wishful thinking. Lcan do no better than quote Borlng on this
subject: 'We will succeed in producing such varieties of crop plants about six months
after utopian political leaders, sociologists and economists produce a new race of
men who needs no food in order to grow strong bodies, maintain health, will effectively
enjoy life and speak eloquently'.

79
Amongst the possibilities mentioned by Dr. Lopion is that by genetical engineering
it will be possible to transfer to economic cereals the ability to fix atmospheric nitrogen,
at present the monopoly of the leguminous plants. He mentioned in passing that
N-fixation was an energy-consuming process. I think that very few people realize how
great is the price in terms of carbohydrate energy that legumes pay for the energy
they apparently receive free from the atmosphere. The bacterial fixation of nitrogen
requires the same amount of energy per unit N as the chemical manufacturing process
of the fertilizer. I n soybeans, for every 100 kg of N fixed, 400 kg of carbohydrates are
expended; even at 1976 oil prices, the cost of the fuel amounted to no more than 10%
of the value of the carbohydrates sacrificed, It is plausible to assume that the main
reason that plant breeders have not been able to raise the level of yields of legumes
such as soya, despite considerable breeding efforts, anywhere near the achievements in
cereals, is due to the privilege of symbiotic N-fixation enjoyed by the legumes.
And now efforts are being devoted to transfer the same privilege to the economic
cereals!
One of the characteristics of much of the subsistence agriculture in the tropics is
mixed cropping. Dr. Lupton mentioned how a mixture of varieties inhibits the spread
of diseases. A mixture of crops should be- and is-, still more effective in this respect.
it has been shown that mixed cropping also has many other advantages: mainly
greater efficiency in the utilization of incident light, available nutrients and water, etc.
Wisely, the International Research Centres, instead of ignoring mixed cropping as a
primitive method, are devising modern versions of mixed cropping in alternate rows,
relay cropping, etc. What is needed to complement these activities is the breeding of
varieties of the different crops involved in mixed-cropping that are tailor-made to
grow in combination with other crops.
Who knows, maybe this will also be the next step forward in the agriculture of the
developed countries!
Two other breeding activities that need to be mentioned are mutation breeding and
interspecific crosses. Who has forgotten the high hopes raised by the use of radiation
and chemical mutagens a couple of decades ago? In particular, it would be interesting
to hear what has come out of the considerable efforts invested by our Swedish collea-
gues in mutation breeding and what are the conclusions they have reached.
Work on Wheat x Rye crosses has been going on for many years - Trisecale has
become a reality. What are its potentials - present and future, is another question
that has not been mentioned hitherto.
In the course of the lively discussion that followed the coordinator's report, Dr. Lupton
and Dr. Stoy made the following comments:
Dr. Lupton agreed that widely adapted varieties have great economic value, and have
in many cases been successfully cultivated over large areas of the world. But because
it has often proved difficult or impossible to explain some of the wide variations in
yield observed from site to site, the possibility of breeding varieties with specific
adaptation is being investigated as a component part of a wider programme at the
Cambridge Plant Institute in selection for more general adaptability.
Dr. Lupton stated that further increases in yield of wheat and barley were to be
expected by selecting for higher total biomass amongst plants with rht genes. In his
opinion, there was no advantage in using F, hybrids for this purpose.
As to the Triticale programme, this is relatively new at Cambridge, and is aimed at the
production of a winter crop for grain for animal feed. It is still too early to assess the

80
value of this programme, but Dr. Luptoi expressed enthusiasm as to its future pros-
pects.
Dr. Stoy objected to the term 'mutation breeding' as implying a special and unique
method of plant breeding. Like other valuable gene sources, artificial induced muta-
tions should be incorporated into ordinary breeding programmes. Many of these
mutations can already be found in a wide array of commercial varieties. Theoretical
investigations on experimental mutagenesis and on radiation biology have important
implications for our understanding of vital biological processes and are thus indirectly
of considerable value to agriculture.

In conclusion
We have come a long way from the time when individual selection and hybridization
were the only tools available to plant breeders. With an array of ingenious methods
available to them, they have achieved remarkable quantum jumps in the productivity
of those crops in which major efforts have been invested.
However, I have the feeling, confirmed by what Dr. Lupton stated, that we are not far
from reaching the ceiling of what can be achieved with the existing breeding tools,
A parallel development is the enormous progress made in the understanding of the
processes involved in the physiology of yiehl, a concept that did not exist as a distinct
discipline a couple of decades ago.
What has been achieved in this field has been amply demonstrated at this colloquium
in a number of remarkable papers. If studies on the physiology of yield are not to
remain an intellectual exercise, and plant breeding is not to come to a dead end - a
meeting of minds of plant breeders and physiologists is essential in order to define
new plant breeding objectives and to plan the strategy that needs to be adopted.
The minds should not only meet but should also work together in interdisciplinary
teams to achieve these aims.
I sincerely hope that this ColloqUium will have made a contribution towards the
realization of this desirable marriage of convenience.

81
 Chairman of the 2nd Session
Dr. G. W. Cooke, Chief Scientific Officer, Agri-
cultural Research Council, London/England;
Member of the Scientific Board of the Inter-
national Potash Institute

Hormonal Regulation
of Yield Formation
The Role of Hormones in Yield Formation
G. Michael, Institute of Plant Nutrition, University of Flohenheim/Federal Republic of
Germany- and H.Beringer, Agricultural Research Station Bilutehof, Hannover/Federal
Republic of Gerniany**

Summiary

The example Of cereals is used to it is demonstrate how phytohorriones play a role in the
regilation processes leading to yield forImation: they are involved (i)in the control of lateral
bud growth and thus the number of tillers and ears per plant and per hectare. (ii) in the
regulation of the number of grains per ear and the interactions within the ear which are
responsible for the great differences in grain size in the ear, and (iii) in the control of storage
processes, in particular the duration of grain filling and hence the control of the formation of
the individual grain. Many influences of environmental factors as well as of crop husbandry
and crop breeding can be interpreted uinder the aspect of changes in the hormonal status in
the plant. These aspects justifV the request that with regard to yield formation more attention
shotld be given to the phytohorniones than it is done today.

I. Introduction

Yield formation of crops is a complex process. Its elucidation has always been one of
the main concerns in crop science in the widest sense. The most important prere-
quisite of yield formation is of course photosynthesis and the availability of assimilates
('source'). It has. however, been observed that the storage organs ('sink') themselves
which are of interest to us in terms or yield also exert an important regulative function
in yield formation and that phytohorniones participate in this control function in a
more or less specific way. About 10 years ago phytohormones were only casually
mentioned e.g. in a meeting tinder the title of 'Physiological aspects of crop yield'
(Easth,t ti t. [1969]). Meanwhile there are more and more indications that hormonal
control is involved in the processes leading to crop production. Taking cereals as an
example it will be attempted to elucidate some steps of yield formation under the
aspect of hormonal control. The picture as it presents itself today is incomplete and
to a large extent hypothetical, as on the one hand biochemical knowledge of the
formation and function of plant hormones is rather inadequate and as on the other
hand many results which can be found scattered in the literature come from experi-
ments which were mostly conducted and evaluated with practical agronomic aims
and not under the aspect of hormonal regulation. In this review it is therefore intended
* Irof. em. Dr. Dr.h.c.C. Aichael. InsitiOt for Pflanzencrnahrung der Universitat 1-folien-
helm P.O.Box 106, D-700) Stuttgart 70/Federal Republic of Germany
** Prof. Dr.,U. Beringer, Agricultural Research Station Bntehof, 1.0,.Box 3209. D-3000 Han-
nover/Federal Republic of Germany

85
to show how such relationships can be recognized and if they may provide a basis for
improvement in the breeding and husbandry of crops.
For the discussion of yield physiology cereals are a suitable subject, as the storage
capacity is set with the development and the appearance of the inflorescence. Inde-
terminate crops such as potatoes and sugar beet, however, are always able to produce
new sink capacities (tubers, storage cells).
The grain yield per plant and per hectare is determined by
A. the number of ears per plant or per hectare,
B. the number of grains per ear and
C. the single grain weight.

It will be demonstrated to what extent hormonal control participates or is assumed

to participate in the formation of these yield components.

2. Number of ears per plant

The number of ears per plant is determined by the rate of lateral bud growth. It
depends (Brouwer [1972]; Langer et al. [1973]; Scheffer and Bruder [1979]) on the
species, the variety, the size and vigour of the seed grain, the depth of sowing and
various environmental factors such as light (planting density), photoperiod, tempera-
ture and availability of water and mineral nutrients during the period of tiller
formation. It could be shown (Heyland [1961]) that a 2-week interruption in the
supply of mineral nutrients at the 2nd-3rd leaf stage of barley resulted in the highest
reduction of the numbers of ears per plant (to 55%) and also of yield (to 61%),
whereas later interruptions in the nutrient supply influenced growth and yield of the
plants less severely. The same was observed for deficiency or discontinuation in the
supply of potassium (Mengel and Forster [1968]), magnesium (Forster,pers. comm-)
and phosphate (Black [1970]; Beringer, this volume). The conclusion seems to be
justified that the nutrient supply and hence photosynthesis and the availability of
assimilates in the early stages of plant development participate in regulating the
number of tillers per plant, particularly as defoliation at this stage reduced tiller
development by almost 50 percent (Langer et al. [1973]).
There is also another factor which proved to be rather important at least for tiller
formation in the advanced stages of plant development. That is the number of
developing grains or ears: The number of tillers for instance increased to 180 and 290
percent, respectively, when 50 percent of the ears or all of them were removed (Aspnall
[1963]). It is therefore not only the amount of nutrients itself which promotes tiller
formation, but also a specific effect associated with developing spikelets and grains
which inhibits growth of additional tillers through a system of apical dominance.
Auxin was found to be one of the factors inducing - directly or indirectly - such apical
dominance (AspinaIl [1963]; Leopold [1949]) in gramineae and other species.
Numerous findings corroborate the existence of such a control mechanism on tillering
by auxins:
- Apical dominance is particularly high in phases of high meristematic and metabolic
activity (hence probably also a high auxin production) (Langer et al. [1973]).

86
- Under long-day conditions less tillers are produced than in short days (Scheibe and
Ellermann [1967]). Accordingly, the auxin content in barley plants was 78% higher
tinder long-day conditions than in short days (Leopold [1949]) which may be con-
sidered as one of the causes for the lower number of tillers.
- Foliar applied auxin decreased the number of tillers in grasses from relatively 100
to 40 (Leopold [1949]).
--The number of tillers increased when the apex of grass plants was damaged or when
the effect of endogenous auxin was inhibited by X-rays (from 100 to 375) or by the
application of more or less inhibitory substances such as coumarin (to 200%) or
triiodobenzoic acid (TIBA) (even to 600%, Leopold [1949]). In other cases, too,
treatment with TIBA increased the number of tillers (Saint et al. [1975]; Ackerson
and Chilcote [1978]).
- Of two wheat cultivars one produced in general only 2-3 shoots, the other 5-8 shoots.
A higher content of total and bound IAA, which may have led to higher apical
dominance, was already observed in the seed grain of the first cultivar as compared
to the second (Rademacher [1978]).
- Where ears were removed before emergence, axillary buds normally dormant
developed into ear bearing tillers (Thorne [1962]).
- A uniculm variety of barley (Donald [1968]; Gale [1974]; Stay [1973]) showed
a significant increase in IAA content in all parts (also at the culm base) at the time of
shooting as compared to the variety 'Europa' (Table 1).
All these examples are strong evidence for the participation of auxin in the control
of tiller formation, which requires further investigation.
Participation in the regulation of tiller formation can also be expected from cyto-
kinins (Phillips [1975]). It is known that cytokinins generally favoLr the growth of
bUds and tillers (Lethain [1967]; Sachs and Thiman [1967]; BruinsIna [1979]). Higher
contents in zeatin, a cytokinin, were paralleled by a higher spur formation in yong
apple trees (Buban et al. [1979]). Accordingly participation of cytokinins in the
control of the number of tillers can also be expected in cereals, as the application of
kinetin sometimes increases the number of tillers (Langer et al. [1973]). Endogenous
cytokinins are mainly produced in apical meristems (see Bruinsnta, this volume) and
with regard to tillering the growing root apices may play an important role in cereal
plants. The cytokinins are transported from the root into the shoot (Weiss and Vaadia
[1965]; Kende [1965]). A close relationship exists therefore between root growth and
the cytokinin content of the whole plant (reviewed by Torrey [1976] and Bruins,na
[1979]).
Table I. IAA content of uni-culm barley plant parts in % of those of the control ('Europa')
(Sya, Hangerih, Schroeder. pers. communication)
Time of sampling Roots Basal Remain.
Culml parts culm
A t tillering ................................... . 150 196 160
13 days after sowing ........................... 117 140 128
21 days after sowing .......... ............... 134 186 128
28 days after sowing ............ .............. 147 141 101

All factors favouring growth and branching of roots (e.g. Bergmann [1954]: R.
Brouwer [1977]) such as the supply of carbohydrates, nitrogen and other mineral

87
nutrients (Wagner and Michael [1971] (Table 2), Saltehnacher and Marschner[1977],
G6ring and Mardanow [1976]) lead to an increase in the production of cytokinins in
the root and their exportation into the shoot. On the other hand cytokinin activity in
the shoot decreases e.g. due to deficiency of roots in carbohydrates, in water (Itai and
Vaadia [1965]) or oxygen supply (Burrows and Carr [1969]). These findings indicate
why the planting density and light intensity, in particular the incidence of light on the
lower leaves which mainly supply the roots (Lupton [1966]), and also fertilizer
application, soil cultivation and aeration as well as shading and partial defoliation
have such a strong influence on tiller development, namely through stimulation or
inhibition of root growth and hence also of the cytokinin activity in the plant.
Table 2. Influence of different N-supply on cytokinin activity in roots and exudates of sun-
flower plahts (Wagner and Michael [1971])
Treatment Root-DM Root growth Exudate/pl./day Root extract
g/pl. mm/day ml Kin.** Kin. **/pl.
a)*+N ................. 0.9 not determ. 2.8 0.33 0.7
-N ................. 0.9 not determ. 1.4 0.05 0.3
b) -N, -N ............. 1.2 0 4.1 0.05 0.3
-N, +N ............. 1.4 4.5 4.2 0.29 3.5
Sunflowers were grown 5 weeks in nutrient solutions
a) 7 days in solutions with and without N respectively, then plants were decapitated and
bleeding sap was collected over 4 days.
b) Plants for 5 days in N-free solution, then I day with and without N resp., bleeding
for I day.
* Kinetin-equivalents x 10-9 M.
All these observations agree with the opinion expressed by Langer et al. [1973]) that
at least three factors, auxin, cytokinin and assimilate availability (possibly also
giberellins and abscisic acid; Gale [1974]; Phillips [1975]) are involved in controlling
lateral bud growth. This requires however more elucidation as to how far the avail-
ability of assimilates or the 'general level of energy substrate' in the plant (Mitchel
[1953]) control tillering directly or indirectly, i.e. through changes in the hormonal
balance. If it is acknowledged that apart from auxins cytokinins also have an important
function in the control of tillering, the increase in the tiller number after partial removal
of spikelets of main culms will not merely be the consequence of a decreased auxin
flow, but may be interpreted as an improved supply in cytokinins to tiller buds. For,
when removing the most important sink of the cereal plant, i.e. developing grains, then
the roots will be better supplied with carbohydrates. Root growth will be stimulated
(Zink, personal communication) and hence cytokinin production will be higher. The
increased growth of lateral buds after removing of grains is then the consequence not
only of the diminished basipetal auxin flow, but also of the higher cytokinin supply
from the roots. Control of tillering is therefore not achieved in most cases by a single
phytohormone but by a specific balance.
Such a balance of phytohormones probably plays also a role when a farmer attempts
to increase tillering by N fertilization. N fertilization increases the production of cyto-
kinins by promoting root growth. At the same time, of course, shoot growth is
intensified and with it apical dominance as well. The number of tillers is therefore
affected by the balance of these two phytohormones (and others) at the site of tiller
growth.

88
Hormonal balances are probably also effective in the formation of ears on the tillers.
It is known that only a reduced number of' tillers survive to reach the stage of ear
formation. This severe reduction of tillers during ear emergence is often of greater
importance with respect to final yield than the number of originally initiated tiller buds.
Tiller reduction occurs at a period of high growth rates of the plants and of high
demands in assimilates. It seems, however (Thorne [1962]: Laude ef (it, [1967]), that
the cause of this reduction might be less due to competition for assimilates but more
due to competition for phytohormones from the roots or due to changes in the hor-
monal balance in general. It is therefore highly recommended that phytohormones
should be further investigated in order to clarify this agriculturally important problem
of ear formation on the lateral tillers (Aafhammer [1980]).

3. Number of grains per ear

On one hand, a high number of grains is caused genetically. It has long been an
important aim of plant breeders (Stov [1977]). On the other hand it is known
(Brouwer [1972]) that a high number of grains is attained for instance in short days
by good light conditions and temperatures which should not be too high in order to
ensure long-term growth of the apex and good differentiation. It is not only the number
of floral primordia which is decisive for the number of developing grains but also -
as in tiller formation - the number of florets which continue to develop until polli-
nation. As induction and development of flower primordia coincide with a period of
considerable increase in vegetative growth of the plant (shooting, ear emergence) with
high demands in carbohydrates, it may be assumed that the vegetative and generative
organs compete for the available nutrients. But this competition is mostly without
major importance, on the contrary:
- The grain number per ear is often higher in semi-dwarf varieties than in standard
height wheat varieties (mostly due to a lower reduction in florts) although a lower
amount of available nutrients could be expec.ed in semi-dwarf plants. Dwarf cultivars
are characterized by a lower total dry matter yield per plant, a lower rate of dry matter
increase at anthesis and by a lower I'C supply to the ear at that time (Bremner and
Davidson [1978]), Accordingly, the control of the grain number per ear seems to be
largely independant of the respective assimilate supply.
- Short term nutrient deficiency, e.g. a 2-week discontinuation in the nutrient supply
during the initation of the inflorescence did not lead to a decrease bitt to an increase
in the grain number by 25-34% (HeYland [ 1961]),
- Short term shading before anthesis also led sometimes to an increase in the number
of grains per ear (Gifford et tl. [1973]; see also Spier/z [1974]).
- Even drastic defoliation of maize in the 5-leaf stage increased yield level due to an
increase in the number of grains per ear (Crookstone and Hicks [1978]).
Although these results were only obtained under specific conditions, they show that
the carbohydrate supply in the plant should not generally be considered as the limiting
factor in the formation of generative organs and that a certain deficiency may even
favour the formation of these organs.
This assumption is corroborated by observations concerning the regulative processes
within the spikelet and the ear. It is known that the first grain in the spikelets and the
grains in the centre of the ear are the oldest and heaviest, whereas the distal ones
within the spikelet as well as the basal and apical grains in the spike are generally

89
 grain weight
mg/spike let
150-

100-

50

4 8 12 16
spikelet position
Fig. 1. Weight of individual spikelets at various positions (1=topmost spikelet) for intact
ears (o) and ears in which the first floret of the eight central spikelets (No. 5 to 12) was
sterilized before anthesis (0). (Data of Rawson and Evans [1970]).

smaller or do not develop at all. The different development of the grains within the
spikelet may have several reasons. Regulation might, for instance, be due to anatomi-
cal reasons, as the first florets of the spikelets which are supplied directly by the
principal vascular bundles of the rachilla are better provided with carbohydrates than
the terminal florets. It is, however, not very likely that the different carbohydrate
supply is an essential reason for insufficient setting and development of the more distal
grains (Hanif and Longer [1972]). Instead it is quite reasonable to assume that
hormonally controlled interactions exist within both, the spikelet and the ear which
lead to a 'suppression' of the younger buds and grains by the older ones - once again
like in the control of tiller development - through a system of apical dominance
(similar to the hierarchy and pecking order in a poultry yard!). Such interactions are
already known in some plants: Removal of older fruits increased fruit set of sub-
sequently opening flowers of Lupinus plants (van Steveninck [1957]) and reduced the
abortion of young fruits in the racemes of Phaseolus vulgaris (Tamas el al. [1979]).
The following findings support the assumption that such interactions may also occur
in the spikelets and ears of cereals and that phytohormones are involved in these
interactions which participate in controlling the number of grains per ear:
- When the older florets of the central spikelets were sterilized (Rawson and Evans
[1970]; Evans [1972]) (Figure 1), the younger florets of the spikelet as well as the
grains occupying basal and apical positions in the ear respectively, were better
developed and the weight of the whole ears was even increased by 20%, although the

90
heavier first grains of some spikelets were missing. Itwas concluded for normal ears
that the setting of grains in the distal florets and consequently the grain yield per ear
may be reduced by the rapid development of grains from fiorets which reached
anthesis first (Rawson and Eva,s [1970]).
There is obviously a strong interaction within the spikelet based on a system of apical
dominance (or a 'medial dominance' within the car!). The increase in the weight of
the manipulated ear by 20% incidentally shows that the supply of assimilates is - at
least in this case - not the limiting factor, but is sufficient to produce higher yields. It
is therefore not only the source but predominantly the sink which is responsible for
regulation.
- Similar results were obtained by Walpole and Morgan [1973]. They showed that
florets which do not normally carry grains in wheat ears will do so if some of those
floiets which carry grains are sterilized. They suggest that 'the developing grain may
produce phytohormones which either directly or indirectly inhibit the development
of other, younger fiorets'.
- Radley [1978] confirmed these results. When spikelets were deprived of the two
lower grains, the weight of the 3rd grain increased (e.g, by 30%). When spraying
TIHA oil the ears instead of removing the grains, the weight of the 3rd grain signifi-
cantly increased as well,in this case by 19%. As TIBA inhibits auxin transport (prob-
ably also the transport from older to younger grains), this result indicates that this
phylohormone is involved in the regulation of grain growth. Such regulation probably
also decides whether the distant grains of a spikelet will ever be developed or'reduced'.
- Higher auxin contents in older and more developed grains ( Wheelcr [1976]; Radley
[1978]) as well as high export rates of auxins also allow the assumption that auxin
is involved in the interactions within the spikelet (first results with the agar block
method confirm auxin movement in the ear; Sjittand Bangerth, personal com-
munication).
- Number of grains per ear increases with decreasing day-length (Rawson [1971]).
If. like in the aforementioned experiments by Leopold (chapter 2), a lower auxin
content in the plant tinder shoriday conditions is suggested, then a smaller 'apical'
dominance may be considered as the cause of the higher grain number.
These examples will of course have to be corroborated and completed by exact data on
phytohormone activity. Only then will it be possible to make significant statements
on the hormonal control of the grain number per spikelet and per ear. Should it turn
out that such hormonal regulation determines the development within the ear and
that more advanced grains inhibit those which develop later, it would be desirable 10
achieve a more synchronized development within the car (Sioy [1976]). This seems
to be the case in high-yielding semidwarf *Mexican' wheat derived from Norin-10.
In the standard wheat the fastest growing spikelet primordia occur about one-third
of the way from the base and not at the base of the developing spike. This leads to an
inflorescence with a fusiform or spindle-shaped appearence (Figure 2). whereas the
developing inflorescence of Norin-10 and its derivatives, elongated with high number
of spikelet primordia, have a cylindrical, spathelike appearance indicating that spikelet
development along the spike is much more synchronous than in common wheat
(Fisher [1973]). The different development may be regarded as a result of a hormonal
control possibly realized by gibberellins (Stov and Hagherg [1967]: Hohnes [1973a];
Stoy [1977j), but also by auxins. It may be that Norin-10 derivatives are characterized
by a pattern of dominance phenomena deviating from the one in standard cultivars.

91
Weaker dominance would mean less suppression of bud formation and of bud develop-
ment within the spikelets and ears. A stronger dominance in the primary and secondary
axis (due to a higher auxin activity?), however, could also be suggested, allowing a
'greater number of spikelets to form' and preventing 'precocious development of the
basal florets, thereby allowing more fertile florets to form in each spikelet' (Fisher
[1973]).

Fig.2. Developing inflorescences of a standard wheat (left side) and of a Norin-10 derivative
(right side). (Schematic, after J.E.Fisher [1973]).

This aspect of eventual participation of phytohormones in the regulation of the

number of grains per ear requires further investigation. The use of auxin inhibitors
such as TI:BA or chlor-fluorenol (Schneider [1970]) or 'antiauxins' (e.g. PCIB;
Popp [1979]) as well as hormone analyses should provide further informations. ifthe
hypothesis of hormonal control is confirmed, it will be a valuable basis for crop
breeding and crop husbandry. Both disciplines aim at a more uniform development
of basal, medial and apical spikelets and of the more distal grains in the spikelet. Yield
increases were also obtained in other plants, e.g. in Arachis hypogeae, by eliminating
the inhibitory effect of the earlier basal-nodal pods and by better synchronisation in
the development of the fruits (Amir [1969]; see also Bruinsina, this volume).
Similar to the hormonal regulation of tiller growth, cytokinins also participate in the
regulation of grain number per ear. In general, cytokinins promote flower development
(see Bruinsma, this volume). It has to be expected that they have a favourable effect on
grain number per ear as well. This can already happen in the process of differentiation
in the shoot apex and also later during the process of grain setting itself. Cytokinins
are known to delay senescence and exert a 'retention effect' (Mothes [1960]; Allinger

92
ef al. [1969]). In the latter capacity they could, similarly to TIBA, inhibit the export of
auxins from the older grains. In this way they could overcome 'apical' dominance and
thereby contribute to an increase of number of grains per ear (Ruckenbauer and
Kirby [1973]).
Some other results also indicate the participation of cytokinins in the control of the
grain number per ear. Reducing tiller density by thinning 19 days before anthesis in-
creased the number of wheat grains in comparable ears e.g. from 51 to 65 (Martinez-
Carrasco and Thorne [1979]) possibly by promoting cytokinin supply to the remaining
shoots. Seier-Kelhitsch et al. [1974] could indeed show that after removal of lateral
shoots the hormonal status of plants was changed (Table 3): The remaining main
shoots grew taller (higher gibberellin activity?) and their exudates yielded more cyto-
kinins than those from control plants. Even cytokinin activity in the grains was higher
in the treated plants probably due to the higher root/shoot ratio and a higher phyto-
hormone import from the roots. This example clearly demonstrates that a removal of
lateral tillers, which is sometimes done in sink-source studies, might drastically change
the hormonal status of the remaining plant. Higher grain numbers per ear (Martinez-
Carrasoa and Thorne [1979]) and higher single grain weights (Table 3) can be the
result of the changed hormonal status.

Table 3. Effect of removing lateral tillers of barley plants (Seiler-Kelbitsch et al. [1974])

Control Thinned
plants plants
Length of main culi s, cm .................................. 93 10M
Single grain weight. m g .... ................................ 57 62
Kin. eq x 10- NI/1000 grains 14 days after pollination .......... 16 41
Root dry weight, g/pot .................................... 17 17
Root dry weight, g/100 culm s ............................... 41 109

When taking the effect of such hormonal balances into account, the following expla-
nation may be given for the grain increase in maize cobs due to drastic removal of
leaf blades in the 5-leaf stage (Crookstone and Hicks, see above, this chapter). By
reducing the leaf area less foliar auxins but many cytokinins from the roots reach the
apex because the leaves as receiving organs for cytokinins are missing (or as they do
not transpire according to shading experiments by Giffbrd et al. [1973] and are
therefore hardly able to take up cytokinins from the root). The temporary existing
new hormonal balance (increase of cytokinin/auxin ratio in the apex) favours the
initation and further development of spikelet and flower primordia. If there is any
competition between vegetative and generative organs then it is presumably not
for carbohydrates in the first place, but for phytohormones ( Wheeler [1976)).

4. Single grain weight

The setting of grains within spikelets and ears seems to be primarily determined by
hormonal interactions within the ear. Last not least, these also play also a role in the
growth of the individual grain. Apart from that, of course, rate and duration of
assimilate supply are responsible for the determination of final single grain weight.

93
Therefore much attention has been paid in the past to leaf growth, canopy develod-
ment and leaf area index. Since Archbold [1945] demonstrated grain growth to be
largely based on current assimilation, particular attention has been given to crop
photosynthesis during the grain filling period. More recently, however, evidence has
been accumulated that grain yield is influenced not only by the capacity of the crop
to provide assimilates (source) but also by the capacity of the grains to store these
assimilates (sink capacity) (Evans and Wardlaw [1976]). Manipulations at the
assimilating organ (source), at the transport system and the site of incorporation
(sink) indicate that the leaves normally operate at levels below their maximal capacity
(reviewed by Neales and Incole [1968]) and that they possess enough 'photosynthetic
reserves' (Roemer [1971]). This high photosynthetic capacity enables the plant to
'compensate' when individual photosynthetic organs get damaged (by insects, hail
injury etc.) and is mostly sufficient to produce higher grain yields. The assimilating
organs and photosynthesis are therefore in many cases not the yield limiting factor ('no
lack of assimilates', Rawson and Evans [1971]) as often assumed, at least not in
regions with maritime climate (see chapter 5). In this case a high photosynthetic rate
is unlikely to be the cause, but rather the consequence of a high sink activity.
In accordance with the subject of this review the discussion of the development of a
single grain will emphasize the storage processes and the control performed by the
grains themselves (concerning their 'sink capacity') as well as focus on the question
how far phytohormones are involved. It is conceivable that phytohormones participate

4.1) in the regulation of the storage capacity of the single grain, e.g. of the number and
size of endosperm cells which store the assimilates,
4.2) in the intensity of the storage processes and finally
4.3) in the duration of the grain filling period.

4.1 Storage capacity of the single grain

The determination of grain size and grain weight begins already before storage
commences. This could be demonstrated by shading experiments. Shading during the
actual grain filling period decreases the grain yield sometimes to a lesser extent than
shading before this period (Saghir et al. [1968]; Gifford et al. [1973]; Stoy [1973,
1977]). This happens when shading is done at the time of intensive cell division
shortly after anthesis when the number of cells and the storage capacity are mainly
determined (Wardlaw [1970]; Evans and Wardlaw [1976]; Spiertz [1974]). This
early determination of cell numbers and storage capacity of a single grain is one of the
reasons why cultivar differences in final grain size become obvious in higher rates of
grain growth during the 4th an 5th week after anthesis (Stamp and Geister [1976]).
Unfortunately only few data are available on the relationship between the number of
cells (in particular of endosperm cells) and grain filling processes or grain size. These
data demonstrate that grains of large-size mutants possess more and larger aleurone
cells than those of cultivars with small grains (Seiler-Kelbitsch el al. [1975]). The
number of endosperm cells is of course more important. They store the largest portion
of assimilates and determine grain size ( Wardlaw [1970]). Brocklehurst [1977] investi-
gated in two wheat cultivars the relationship between the weight of mature grains and

94
 I Cyt GA IAA ABA
100

% fr.w.

j %%ABA

anthesis maturation
Fig.3. Phytohormone activity per grain during grain development (schematic rel. values)
max. = 100)
fresh weight/grain
-- cytokinin (Cyt)
gibberellin (GA)
0-0-- auxin (IAA)
. abscisic acid (ABA) from GoIdbach and Michael [1976]

Ack •50
5

80- - 40

CD
O 60" 0" 308

C 0

c oDLJ. 1970.20 ',-

C J# ' Ack >

. - 1 %% 10 a"
Or -

6/11 6/15 6/19 6/23 7/1

Fig.4. Cytokinin activity per 1000 kernels (*-) and dry weight of 1000 kernels (o-o
during grain development ofdiflerent barley varieties. (Ack=Ackcrmanns MCZ; Or=Oriol).
(Michael and Seider-Kelbitsch [1972]).

95
the number of endosperm cells 2 weeks after anthesis, when cell division is practically
terminated. The two cultivars differed in the single grain weight by 21% and in the
numberof endosperm cells by 19%. When the number of grains per spikelet was reduced
at anthesis, the weight of the remaining grains could be increased by 47% in the one
cultivar and by 13% in the other. The number of endosperm cells increased in a similar
way (by 30% and 8%, respectively). Radley[1978] arrived at a similar result: Removing
the first two grains in the spikelet increased the weight of the 3rd grain and also the
number of endosperm cells. These results obtained by Wardlaw, by Brocklehurst and
by Radley indicate a clear relationship between number of cells which were formed
after anthesis and the grain size.
In the light of these findings the regulation mechanism for cell division and cell
number gains importance. As far as phytohormones are concerned, it seems most
likely that primarily cytokinins may be involved in these processes, for they are known
to stimulate cell division within a certain concentration range (which is reflected in
their name!). There is evidence for their participation in early grain development.
Cytokinin activity shortly after fertilization is particularly high (Figure 3) and its
maximum seems to coincide with the maximum of mitosis (Tollenaar [1977]). It
could moreover be demonstrated that grains of large-grain varieties or mutants show
a significantly higher cytokinin activity than those of small grain ones (Figure 4).
It is noteworthy that this can already be observed at a time when the difference in grain
size of both cultivars is not yet detectable (Michael and Seiler-Kelbitsch [1972];
Michael et al. [1973]; Seiler-Kelbitsch et al. [1975]). This result, now being confirmed
for other varieties (Herzog and Geisler [1977]) indicates that cytokinins are involved
in the growth of grains during the first phase of their development, probably through
the regulation of cell number and storage capacity. It may be assumed that breeding
of high yielding varieties on the basis of high single grain weight unconsciously
selected plants with high cytokinin activity leading to a high cell number and thus
providing an important- prerequisite for high dry matter weight. The results cited all
indicate that greater attention should be given to the cell number of developing grains
and that it is necessary to recognize the importance of phytohormones which play a
decisive role in the regulation of grain size. These aspects urgently require further
elucidation.
It is not yet possible to make definitive statements on the origin of cytokinins in the
grains. As they are already found in the ear before anthesis ( Witham and Miller [1963];
Wheeler [1972]; Herzog and Geisler [1977]) synthesis in the ear itself is likely.
But a supply from other vegetative plant parts which are still capable of cell division
cannot be excluded, e.g. from roots with still active root apices. Results are available
(Michael and Seiler-Kelbitsch [1972]) that grains of awned ears as compared to
de-awned ears show higher cytokinin contents probably due to an increased trans-
port of substances, also of cytokinins, from the roots to the ears and as a consequence of
transpiring awns (see chapter 4.3). Thorough investigation is however necessary to
determine how far these (presumably relatively small) amounts of cytokinin im-
ported from the roots at the beginning of grain development are of importance in
comparison with the large amounts produced in the ear. The question of how far
other phytohormones such as gibberellins and auxins are effective besides cytokinins
has also to be investigated. As their maxima in the grain only occur at later stages of
development (Figure 3), the target of their function is probably to be found in the
control of other processes (see later).

96
The storage capacity of the endosperm depends also on the size of the endosperm
cells (and number of amyloplasts per cell?). According to the findings by Radley[1978]
again phyrohormones are involved in the regulation of cell size in the grain: Removing
the first two grains in the spikelet favoured - as already mentioned - the development
of the third grain. The number of cells increased, but the cells were smaller.
If phytohormones can be assumed to control cell division in the grain, the question
arises if other tissues in the grain e.g. the embryo can be influenced too. Already Ashby
[1932] considered a high cell number of the embryo as one of the causes for the
'phenomenon of hybrid vigor'. Recent publications on maize also confirm a close
relationship between embryo growth and final grain size (Saulesku and Popa [1965]:
Bjarnason and Polltner [1972]). This is possibly due to a supply of phytohornlones
(Khan et al. [1973]) to the endosperm by the embryo which has in general a longer
growth phase than the endosperm.
Finally yield can be influenced by a more physical control of (hormonal regulated?)
cell growth. Rice grains cannot grow larger than the hull permits; and the size of
outer and inner glumes is determined as early as 5 days before anthesis (Mirata and
ilaisashima [1975]). Also in maize the spatial conditions of the cob may influence
grain growth (Zink. personal communication): By manipulated pollination, cobs were
obtained (i) which showed a normal grain density only in their lower part and others (ii)
which had barren spots all over the cob. The grain number was in both cases reduced
to approx. 35% of normal cobs. [lence the carbohydrate demand of the ears was equal
in both cases. Yet the grain size was not the same: while it hardly changed in (i) with
not more than 5% increase over the control, the single grain weight in (ii) was signifi-
cantly (about 350,,) above the control.

4.2 intensity of grailn filling

Extent and intensity of the hormonally controlled flow of assimilates into the storage
organs not only depend on the storage capacity. They can also have other causes
inherent in the grain establishing a certain 'attraction'.
The view that phytohormones exert an 'attraction' originates from investigations on
long distance and on cell to cell transport e.g. by Nitsch [1950] and Afolhes et al.
[1959, 1961]. Since that time numerous examples (see Alliger et al. [1969]) have de-
monstrated that endogenous or applied phytohormones such as cytokinins, gibberellins
and auxins increase the 'attractive power' of a tissue. They establish metabolic sinks,
for instance at the point of application (Seth and Wareing [1967]). But the physio-
logical and biochemical reasons for such an attraction have not yet been definitely
clarified.
A contribution of phytohormoncs to the attraction and promotion of storage pro-
cesses in cereal grains would also be expected if we recall the path of the carbohydrates
within the grain: Sucrose normally enters the grain through the vascular bundle
running along the furrow. It then travels across the tissue of the pigment strand and
the remains of the nucellus into the endospern cavity. From this point sugar is
assumed to move radially outward across the endosperm probably in the intercellUlar
apparent free space. From here it is taken up by the cell which converts the sugar into
starch (Jelner [.1970]; Jenner and Ratijen [1975]; Sokri and Sharnon [1975]; Jenner

97
[1976]). In this chain of individual steps hormonal controls can be suggested at several
places. For instance already at the first step, the unloading of the sieve tubes is of
particular importance for the transport of carbohydrates within the grain. The rate of
phloem unloading determines the sucrose concentration in the cavity and accordingly
the rate of sucrose diffusion along the concentration gradient in the free space of the
grain. It may often be regarded as a limiting factor in the transport of carbohydrates.
A participation of phytohormones in phloem unloading has been suggested (Zimmer-
inann [1958]; Nakata and Leopold [1967]) and in cereal grains the size of the cavity,
constituting an intermediate depot for sucrose, seems to be influenced by phytohor-
mones (Radley [1978]). Furthermore it has to be expected that gibberellins and auxins
are involved when the carbohydrates pass through the cell membranes (Wood and
Paleg[1972]). Finally they also participate in the promotion of synthesis and activities
of enzymes thereby influencing the storage process and its intensity.
The following observations advocate a participation of phytohormones in the intensity
of grain filling:
- The major proportion of gibberellins (Mounla [1978]) and auxins (Radentacher
[1978]) in the grain is located in the accumulating endosperm.
- IAA treatment of wheat ears increased the translocation of assimilates from the flag
leaf to the car (Prochazka [1978]).
- Grains of a variety ('Kolibri') with distinctly higher single grain weights had
significantly higher IAA contents (per grain and per g fresh weight) than those of a
small grain variety ('Chinese Spring') (Rademacher [1978]).
- A fairly close relationship exists between maximum growth rate of seeds and the
levels of auxin and gibberellin (Wheeler [1972]; Mounla and Michael [1973J;
Euwens and Schwabe [1975]; Radenacher [1978]).
Although correlations like these do not, of course, prove a causal involvement of
hormones in seed growth, they provide consistent evidence for the concept of hormonal
control (Bewly and Black [1978]).
The opposite case, that a phytohormone will not exert an attraction, but a repulsion of
assimilates can also be found. Such a phytohormone is abscisic acid - at least at a
certain concentration (Mc Who and Jackson [1976]; Doerffiing [.1978]; Dewdney and
Mc Wha [.1979]; Fenner and Doerffling [1979]). Wagner [1974] studied the assimilation
of 4 CO2 by excised barley shoots and the transport of 14 C into the ears which had
previously been treated with phytohormones. The 14 C uptake by ears was 145% for
GA 3, 120% for kinetin but only 20% for ABA-treated ears compared with the controls.

4.3 Duration of grain filling

The duration of grain filling is the subject of many investigations (Stoy [1965];
Lupton [1966b]; Binghain [1967]; Daynardel al. [.1971]; Evans and Wardlaw [1976];
Hageman [1977]; Tollenuar [1977] etc.). Environmental conditions which extend the
duration of grain filling, such as cool climate (Beringer [1966]) or specific agronomic
measures, are a priori important prerequisites for the formation of high grain weights
and grain yields. The extension of grain filling is also an important target in plant
breeding (Rawson and Evans [1970]; Daynard el al. [1971]; Tollenaar [1977]).
High yielding varieties with high single grain weights are not generally characterized

98
by a long total growth period but mostly by a long period between anthesis and
maturity, and yields of hybrids are highly correlated with the effective duration of
filling more than with its rate (Evans and Wardlaw [1976]). A long filling period is
especially advantageous if assimilation happens to become the limiting factor at the
time of highest grain growth rate (Sloy [1965]). Late senescence of the flag leaf and
ofawns is therefore an important aim of plant breeding (Iaensel[1965]; Stoy [1973];
Olgheimi e al, [1976]). But it has to be pointed out here that delayed flag leaf
senescence which could be checked by chlorophyll determinations, is only an indicator
not a guarantee for large grains. Nevertheless in comparing different wheat cultivars,
Stamp and I-erzog [ 1976] found correlations (r = - +0.7) between final grain weight
and flag leaf chlorophyll content 4 weeks after ear emergence.
There are numerous examples which show that senescence is controlled to a large
extent by phytohormones. They can retard (cytokinins, gibberellins, auxins) or
accelerate senescence (e.g. abscisic acid). Cytokinins will be discussed first. Their
retarding effect on senescence has already been investigated by Richmond and Lang
[1957j and Mothes [1960] and has often since been confirmed.
Cytokinins may affect yield formation primarily during the first phase of grain growth
through the promotion of cell division and the build up of storage capacity. At this
time cytokinins reach their maximum in the grain (Figure 3). Later, when their activity
has already decreased, their main function is presumably the retardation of senescence
of the whole plant, keeping the assimilating organs green and active for a longer time.
Cytokinins are mainly produced in the neristematic tissues of the plant. During grain
filling the root apices might be the major sites of cylokinin synthesis (Weiss and
Vaadia [1965]: Siaton et al. [1967j; Kleinen and Kliti [1974]) provided that the
roots arc still growing at that time. The nodes of the culm could be another site of
cytokinin synthesis. Leaves excised with the meristematic nodal tissue remain green
for a longer time than those without nodes (P. MNartin. personal communication). But
in general the root system is of greatest importance in the cytokinin supply to the culms,
because shoots depend on the import of root-produced cytokinins (Vonk [1979]).
This import into the main shoots is especially high if the lateral shoots are removed.
By such manipulation the root dry weight per 100 shoots rose from 41 to 109 g and
the remaining shoots were indeed green for a longer time. The cytokinin activities of
the exudates and even of the grains ware significantly higher than those of the exudates
and grains of intact plants (Seiler-Kelhitsch et al. [1974], Table 3).
The cytokinin production of the roots might explain that all those husbandry techni-
ques which promote root growth even during the period of grain filling enable the
plants to remain green and active for a longer time. Such an extension of the active
phase of the plant can be achieved by good soil strUcture (soil cultivation and aeration)
and by fertilizer application, in particular by adequate nitrogen dressings. it can be
demonstrated (Table 2) that roots immediately stop growth and cytokinin export into
the shoot when temporary nitrogen deficiency occurs or inhibitors are added, whereas
these functions increase with adequate N supply (Wag,ner and Michael [1971];
Sattelmacher and Niarschner [1977]: Ktauss 11978]).
Not only the quantity, but also the form of N application affected the growth rate of
sunflower plants differently and parallel to that the cytokinin content (Salama and
Wareiig [/979]). Even N top-dressings at flowering stimulated cytokinin production
again in the root of maize followed by a considerable increase of cyokinins in the
root exudate (e.g. 1.22 kinetin-equiv. x 10-"II in exudates per plant per day com-

99
pared with 0.29 equiv. in exudates of plants without topdressing; Wagner and Michael
[1971]). This may sometimes lead to additional tiller formation in cereals. But, more
important, cytokinin production is certainly involved when, after late N dressings,
not only protein content in grains and the proportion of prolamin (zein in maize)
but also grain size and grain yield are enhanced (Gilzel [1968]; Orlorius and
Hoefner [1976]; Zink [1979]). A positive correlation between zein content, kernel
size and grain yield of maize (Tsai et al. [1978]) has presumably to be interpreted
in this way because all three parameters may be regarded as the consequence (i)
of the plant being still well supplied with N during grain development, (ii) of the
root system being thereby maintained physiologically active and (iii) of the shoots
being well supplied with cytokinins. This example, too, shows that N fertilization
has a positive effect on yield not only due to the availability of the nutrient itself but
also through the hormonal status of the plant (Michael [1974, 1979]). This effect
was also be demonstrated by F.Zink (pers. comm.): Nitrogen application to the
soil or to the husks, respectively, during the phase of grain formation of maize plants
increased the N content of the grains (by 20-30%). However, only the N application
via the soil (and the roots) had the additional effects of increasing the number of
grains by 11-12% and the single grain weight by 13-14%, probably by promoting root
growth and phytohormone activity.
A similar effect has to be attributed to K fertilisation. Haeder (see this volume) could
demonstrate in barley that the incorporation of 14C sucrose into the grains 42 days
after ear emergence was significantly higher in plants well supplied with K (K2 and
K, plants) than in K, plants. Concomitant with increased K contents chlorophyll
content was two and three times higher than in the K, plants. The good K nutrition
had also retarded senescence and extended the grain filling period surely due to
increased cytokinin production in the roots. This can be assumed because the roots
of high K plants still looked white and very healthy at this advanced stage of ripening.
Accordingly, K fertilization can have a similar positive effect on the cytokinin level
in the plant like N fertilization. The reverse is the case when the reduction of root
growth (by excessively high temperatures or flooding) leads to the inhibition of
cytokinin production and accelerated cytokinin degradation and hence to premature
senescence (Sition et al. [1967.]; Mizrahi and Richnond [1972]) and lower grain yield
(Beringer [1966]; Spiertz [1974]).
These considerations apply in a similar way to gibberellins which can also retard
senescence (Allinger et al. [1969]; Fletcher and Adedipe [1970]; Davies et al. [1977]).
GA is synthesized in leaves as well as in root apices (Groebe and Ropers [1978]).
Root-originating GA is transported along the xylem into the shoot (Carr et al.
[1964]; Jones and Phillips [1966]). A detrimental effect on root growth (anaero-
biosis, stress) leads to the inhibition of GA export (till new adventitious roots are
produced) (Reid and Crozier [1971]). This fact underlines the importance of an active
root system in providing the shoot with phytohormones, which retard senescence.
The supply of the shoot with phytohormones from the root is however not without
complication. A competition within the shoot for phytohormones is often observed
(Varga and Bruinsma [1974]) and this competition is presumably the reason
why the removal of ears mostly retards senescence of leaves and culms and why
the removal of leaves causes the ears to remain green for a longer time (Hohnes
[1973b]) combined with higher chlorophyll and cytokinin contents in the ear (Wheeler
[1976]). Moreover, the transpiration of leaves and ears affects the distribution of

100
hormones within the shoot. Although the presence of cytokinin nucleotides in the
phloem exudate could be detected in some cases ([Vonk [1979]), transport from the
root along the xylem by the transpiration stream has to be generally assumed (Weiss
and Vaacia [1965j; Kende [1965]). Hence the awns must be considered in a new
light. The main function of the (drought resistant) awns (Grundbacher [1963]) in
grain formation is undoubtedly the supply of assimilates in particular in climates
with frequent dry spells during grain filling. Removal of awns (Sagronisky [1954])
can then lead to a decrease in grain weight. This decrease in grain weight is generally
attributed to a lack of assimilation towards the end of grain formation. The results,
however, also allow the assumption that awns most likely affect grain yield by their
Iranspiration and consequently their attraction of phytohormones from the root:
- Removal of awns caused accelerated yellowing and ripening of ears.
- Removal of awns decreased the supply of the ears and grains with substances which
are translocated via the xylem: Ca- + and kinetin (Michael et al. [1969]).
- Removal of awns led to a significant reduction in the cytokinin content of the grains
(Michael and Seiler-Kelhitsch [1972]).
- Yield depressions due to removal of awns were greater al 45 percent relative
humidity than at 95 percent when the transpiration of the plants is practically nil
(Michael ct al. [1970]).
- In de-awned ears the application of kinetin to the ears had in most cases a positive
effect on grain weight, whereas the effect of exogenous kinetin always remained
doubtful in awned ears (Michael et al. 19701).
These observations suggest that awns, besides their function as suppliers of carbo-
hydrates have an influence on the phytoiormonal status of the plant due to their
transpiration. A similar aspect has also be taken into account when evaluating the
results of experiments with shading. Shading inhibits not only photosynthesis (source)
but also transpiration and possibly the transport of substances such as phytohormones
from the roots to the ears. Consequently shading can effect sink capacity as well.

A retarding effect on senescence is also to be expected from gibberellins and auxins

whereas abscisic acid (ADA) often shows the opposite effect. This effect is rather
negative with regard to grain filling processes: ABA decreases - depending of its
concentration - the size of the sink and promotes senescence (Doerffliag [971, 1977,
1978]: Jackson and Osborne [1972]; McWha and Jackson [1976]; Dewdney and
McWha [1978]; Fenner and Doerfflinq (1979]). Hence it reduces the duration of
assimilation of the green parts as well as the duration of grain filling. The consequence
is often premature ripening (Mengel and Kirkhyr [1978]).
ABA activity in the cereal grain increases during grain growth (as well as in other
fruits: Doerffling [1970]: Alleweldt et al. [ 1975]) more or less continously (Milborrow
[1974]; Mc Wta [ 1975]; Golbach and Michael / 1976]; King [1976]: Radley [1976]).
It is the phytohormone reaching its maximum in the grain at the latest (Figure 3) i.e.
when the increase in grain dry weight is terminated and fresh weight decreases. Many
results, also those from experiments with applied abseisic acid (King [1976]; Tanlas
el al. [1979]) suggest that ABA (directly or in connection with ethylene: Jackson and
Osborne [1972]) plays an essential role in the control of the duration of grain filling.
ABA is assumed to induce changes in membrane structure. permeability (e.g. also of
the pericarp of the grains: RadIy [1976]) and enzyme activities (Pai and Benzioni

l1
[1973]; Evans [1974]; Ho and Varner [1976]) before the occurence of visible grain
ripening (Chin and Beevers [1970]). These changes initiate water loss from the grain
and the end of grain filling (Radley [1976]) even if sugar content in stem and
grains may still be high (Tollenaar [1977]; Breidert and Schoen [1979]) and the
grain (in particular the young ones) could still grow due to their storage capacity.
According to our present knowledge the question: 'What causes the grain suddenly
to stop growth?' (Evans and Wardlaw [1976]) can be explained to a large extent by
ABA. In considering the hormonal balance, the influence of ABA becomes particularly
great during the later stage of grain development. At this time the activity of the other
phytohormones with a positive effect on grain filling has already decreased considerably
(Figure 3).
This is supported by results from two wheat cultivars (Rademacher [1978]). In
contrast to the grains of 'Kolibri' the ABA content per kg grain fresh weight of
'Chinese Spring' was 50 percent higher at the time of ABA maximum and final
maturity. Grains of 'Chinese Spring' ripened faster and achieved a single grain weight
of 25 mg as compared to 38 mg in the other cultivar. Accordingly, a close relationship
seems to exist between ABA content, duration of grain filling and grain weight.
Opinions, however, still differ as far as the origin of ABA in the grain is concerned.
Besides production in the ear (Milborrow and Robinson [1973]; Radley [1978];
Dewduey and MeWha [1978]), an additional import from the leaves may be ex-
pected. ABA is transported from the mature leaves to the shoot tips (Zeevart [1977])
or to young leaves (and the apical parts of the plants) which are temporarily richer in
ABA than old ones (Goldbach et al. [1975]). It is known that in situations of water
stress ABA is very quickly synthesized in leaves (Wright and Hiron [1970]; Mizrahi
and Richmond [1972] and others). The sites of ABA synthesis are the chloroplasts,
from where, after an increase in the permeability of the chloroplast membrane, ABA
leaks into the cytosol (Milborrow [1979]). High amounts of ABA are accumulated
also in ageing as well as in N deficient leaves (Goldbach et al. [1975]; Krauss [1978]).
Fully expanded leaves of sunflower plants grown in complete nutrient solution had on
the average 6.8 ig ABA/kg fr.w., those of plants grown in N-deficient solution
contained 21[ig ABA. In old leaves the differences were 8.1 against 29.8 l.g ABA
(Goldbach et al.[1975]). As ABA moves easily and rapidly in the phloem (Doerffling
and BOttger [1968]; Eschrich [1970]; Bellandi and Doerffling [1974]) and in specific
cases also in the xylem (Lenton et al. [1968]; King [.1979]) it may be assumed that it
migrates also from ageing or stressed leaves into young growing leaves (see above) and
that it is transported also into the ears. This assumption is supported by the following:

- Removal of ageing leaves at later stages of the grain development often keeps the
ears green for a longer time.
- After application of 14C labeled ABA to the flag leaf of spring wheat, up to 21
percent of the exported ABA could be found in the grains after 12-16 hours (Goldbach
and Goldbach [1977]; Goldbach et al. [1977]).
- When leaves of these plants were exposed to warm air (36*C) while the ears were
protected, ABA contents increased in the wilting leaves (from 50 to 78 lg ABA/100 g
dry matter). But also ABA in grains rose (from 44 to 83 i'g ABA/100 g dry matter)
as well as in glumes and awns from 22 to 82 jig ABA/100 g dry matter (Goldbach et al.
[1977]). These results indicate that transport of ABA from lower plant parts to the
ear and grains has to be expected.

102
This ABA import by grains may possibly explain why decreasing the number of
spikelets per ear (for instance from 14 to 4; Lupton and Ali [1966]) did not always
increase but sometimes even decreased single grain weight (for instance from 57 to
46 mg). perhaps because the imported ABA was distributed among fewer grains.
It has to be admitted that ABA can be easily metabolized (Zeevart [1977]), in the
grain predominantly in the aleUron layer (Dashek et al. [1979]). [ts metabolites
phaseic acid and dihydrophaseic acid are much less effective than ABA (Dashek et al.
[1979]). Investigations on the metabolisation of ABA have however mostly been
conducted with applied ABA and it is (doubtful whether endogenous ABA is as
quickly metabolised (Milborrow [1974]; Dewdnev and McWha [1979]). There are
also results indicating that roetabolisation of ABA in the cereal grain is relatively
slow as compared with other fruits (King [1979]) and is particularly slow in older
grains as compared to younger ones (Goldbach and Goldhbch [1977]). It has therefore
1o be expected that in the last stages of ripening grains are effectively influenced by
ABA, which could have been imported from the leaves. Adverse growth conditions

._ 16-

0) 44: I
4-4
Mn 26 C" 8

S fresh weight %
60 26 C%1%

S40- ;0 -. 26C 18 C
AM .rdry weight

20 27 35 47
days after pollination
Fig. 5. ABA equivalent content. fresh weight and dry weight of 1,000 barley grains (Oriol)
during ripening at day temperatures nf IS C and 26 C. Standard error is indicated by vertical
bars. (Goldbach and Mi,hfel [1976])

103
which have a detrimental effect on the vegetative plant parts could then also affect
grains via hormones and decrease yields due to premature ripening and shrivelled
grains.
It is noteworthy that ABA also has other functions in the grain which are of particular
interest to agriculture. ABA, being primarily located in the embryo (Goldbach and
Michael [1976]; King [1976]) is one of the controlling factors of seed dormancy
(Wareing and Saunders [1971]; King [1976]; Walton [1977]). Higher ABA contents
in the grain inhibit sprouting with its known negative effect on grain quality. The
risk of precocious germination of the grains within the ears is particularly high when
grain ripening is accelerated by dry heat which is followed by a humid period (Brouwer
[1972]; Grahl and Schroedter [1975]; Johansson et al. [1976]). One reason for
sprouting could be that ABA, after having induced ripening, is more quickly meta-
bolised in a hot period than at lower temperatures (Goldhach and Michael [1976])
(Figure 5). This would lead to a faster break of seed dormancy. It could be demon-
strated with a specific example (although not in all cases, see King [1979]) that the
grains of the Swedish variety 'Kristina' which is inclined to sprouting, had a lower
ABA content during grain development and particularly towards the end of ripening
than the grains of a more resistant variety 'Oriol' (Goldbach and Michael [1976]).
Should the participation of ABA in the control of these processes be confirmed (see
King [1976, 1979]), this additional example would support the necessity for further
detailed research on the phytohormonal status of the grain for a better understanding
of grain development and yield formation.

5. Summary and prospects of future development

The aim of this paper is to demonstrate the importance of phytohormones and their
stimulating effect on crop production which has not received due attention in the
past. The function of phytohormones in the regulation of cereal yield is demonstrated
by several examples.
Phytohormones are involved in the control of tiller formation, probably in the form
of apical dominance. The strong modification of the number of tillers and ears per
plant by environment as well as by breeding must be interpreted in relation to the
phytohormonal status of the plant. Also the second yield component, the number of
grains per ear, can be studied under this aspect. There are many indications that the
formation of spikelets in the ear and of florets in the spikelets are hormonally controlled
and that the interactions between individual grains within the ear arc also based on a
system of apical dominance. Finally the third yield component, the single grain weight,
has again to be seen under the aspect of hormonal regulation. In this case hormonal
control concerns the storage capacity of the single grain (e.g. the number of accumu-
lating cells) as well as the intensity and duration of grain filling.
The numerous examples about the participation of phytohormones in yield formation
offer many valuable aspects to crop husbandry and crop breeding:
a) Many husbandry techniques, long recommended and approved, can be better
interpreted if phytohormones are taken into consideration. The promotion of the
growth and duration of an active root system by adequate soil tillage and fertilizer
supply not only improves water and nutrient uptake, but also favours a hormonal
balance, which leads to a better harvest index and higher economic yield. In this
connection, mycorrhiza, could be mentioned too. This is generally assumed to

104
stimulate plant growth through a better availability of mineral nutrients. But mycor-
rhiza could also be effective as a supplier of cytokinins to the host plant (Schweers and
Meyer [1970]).
Furthermore the hormonal aspect throws a new light on fertilizer application (Michael
[1974, 1978]). The effect of nitrogen fertilization is not fully understood if only the
supplyof N as an essential nutrient and its direct influence on metabolism areconsidered.
It has to be broadened by the indirect influences through the hormonal status of the
plant (see Beringer, this volume). The application of nitrogen increases the cytokinin
content and thus enhances the number of ears per plant, the number of grains per ear
and the filling processes of single grains by retarding senescence. The different effects
of N H 4-N and NO,-N have to be considered as well ( Wakloo [1970]). Under certain
conditions they can cause changes in the redox potential and pH value of cells resulting
in different rates of hormone synthesis or particularly degradation ( Whity and Hall
[1974]). For instance NH,-fertilization compared to NO,-application caused a
higher zeatin content of young apple trees which was paralleled by an increase in spur
formation (Buban et al. [1979]). Unfortunately, there is a lack of data on hormonal
analysis in other cases. It is therefore not yet possible to interpret on a hormonal
basis the frequently observed differences in growth pattern and grain yields after
NH 4 - and N0,-fertilization (McKee [1949]; Klem, [ 1966]; Steineck [1966];Sommer
and Poleischny [1972]; a.o.). (For further details about possible effects of fertilizer
application on yield formation via phytohorniones see the papers by Beringer, Kraus
and Haeder in this volume and Rajagopal and Rao [1972]: Michael [1974], [1978];
Men gel and Kirkbv [1978]).
The situation is, by the way, similar for the application of pesticides or synthetic growth
regulators, such as CCC. They can result in yield increases although the plants are
healthy or without any risk of lodging (AfJhamner [1980]). Possibly these substances
exhibit hormone-like activities or affect the hormonal status of the plant. In the case
of CCC the initial inhibition of root growth is followed by an increase of growth and
hormone production, a favourable effect lasting till grain and yield formation (Julg
and Sturn [1964]; Hamus [1967]; Ruckeabauer [1968j, Dressel and Kihn [1971]).

b) Consideration of the hormonal status also facilitates understanding of the effect of

environmental stress situations (see Mizrahi, this volume) which may determine grain
yield not only by reducing photosynthesis but also through phytohormones. In most
cases probably the latter aspect is the more important one, and changes in the rate of
photosynthesis should not be defined as the cause. but instead as a consequence of
reduced sink capacity. When, for instance, shading during vegetative growth causes
depressions in final grain yield, then the sequence of events in many cases is likely to be:
Shading -- changes in the phytohormonal balance - less sink capacity - less
photosynthesis.
These ideas touch on a basic problem in crop cultivation. This is the question of the
limiting factor in yield formation and the difficulty of deciding where to influence
growth and yield : on the side of photosynthesis (source) or on the side of sink capacity.
Ample literature is available on this subject (see e.g. Stoy [1977], Evans and WVardlaiv
[1976]: Hagemann [1977]) yet it is not possible to give a universal answer. Sometimes
the source and sometimes the sink seems to be limiting (Gifford et al,[1973]). In
modern crops and under most growing conditions the sink is generally a limiting
factor (see chapter 4 and e.g. Rawson and Etans [1971]). On the other hand there

105
are specific cases in which the supply with carbohydrates may be regarded as the
limiting factor in grain growth e.g. during the last days of grain filling or after injury
and premature senescence of leaves (Stay [1973]; Tollenaar [1977]; Aufliamner
[1980]). But even in such cases it would have to be investigated how far a hormonal
control of the sink (via ABA?) is involved. For, 'source' should not only be regarded
as 'source of carbohydrates' or other 'nutrients' but also, at least sometimes, as
'source of phytohormones'! Hence, in all cases, phytohormones which control the
sink, play an important role because they regulate the distribution of assimilates
within the plant (Thorne [1971]) and help to establish better 'energy storage'
(Mengel [1977]).
When discussing source-sink relationships one methodical aspect should always be
taken into account: In many experiments the source is manipulated by removing or
shading of individual leaves. But such manipulations do not only affect the source.
They often also change the hormonal status and hence sink capacity and both intensity
and duration of grain filling. It is therefore advisable to reconsider these investigations
and their interpretation under the aspect of hormonal control. The same remarks
also apply to the elaboration of mathematical models on crop productivity in which
the source is frequently overestimated as the limiting factor.

c) A phytohormonal consideration of yield formation also offers new aspects inplant

breeding. Improvement of yield potential by breeding is mainly based on improvement
of the harvest index. Unconsciously breeders selected for yield increase through
changes in the hormonal status of the plant (Linser et at. [1958]; Schwanitz (1960];
(see also chapter 4.1). Some symptoms of certain hormonal balances, for instance
duration of flag leaves, ears and awns are already selection criteria in plant breeding
(Haensel [1965]; Stay [1973]; Olugbemi et al. [1976]). An additional criterion for
bigger grains could be the number of endosperm cells (see chapter 4.1) whereas a
selection for a more synchronized development within the inflorescence (Figure 2)
may improve number of grains/ear. Though incomplete, our present knowledge on the
control of sink capacity and the filling processes gives rise to expectations that the
number of grains per ear and single grain weight can be further improved by systematic
breeding. It is noteworthy in this respect that these two yield components have a
different hormonal control. It should therefore be possible to improve each component
independently.
Although a negative correlation has frequently been observed between the two yield
components one should bear in mind that an increase in grain number per ear occurs
after repression of hormonal inhibition within the ear, for instance, by the formation
of grains from apical grain primordia in the spikelet and of apical and basal grains
in the ear. In most cases these grains are set later and will therefore not reach the
same size as the older ones due to a medial dominance (IAA?) or a hormonal con-
trolled termination of the grain filling processes (by ABA?).
The use of growth analysis and the interpretation of yield differences on the basis of
the different yield components therefore represents a great progress in plant breeding
and yield physiology (Bruinsma [1966]). It would be advisable to differentiate even
more in further research and to measure for instance single grain weight not only as a
summation effect but also to determine grain size distribution in order to identify
causes of eventual yield differences.
Although an interdependence of grain number and single grain weight would not be

106
expected from a physiological standpoint, an interdependence of other components
must be taken into account. Iffor instance the varietal difference in ABA activity
with its decisive influence on the termination of the grain filling process as well as
on the inhibition of sprouting could be confirmed in further experiments, then a
highly important consequence would arise: Breeding efforts in achieving higher
single grain weight are probably often linked with allincrease e.g. in the cytokinin
content of the grains and possibly with a decrease in ABA content. The frequently
observed low resistance to sprouting in large grain cuLhivars (as a consequence of lower
ABA content) however would then he an inevitable consequence of these hormonal
changes effected by breeding. Similarly if plant breeders would try to improve stomatal
regulation and drought resistance via ABA (Luplan; Mizrahi, this volume) then an
accelerated leaf senescence, a shorter grain filling period and a lower crop yield could
be the consequence. These examples show how important the knowledge of the hor-
monal status of the plants and its consideration are for the understanding of yield
Formation and for further evolution of crops through breeding and cultivation.
Two more examples will be given for the same reason. Breeding succeeded in increasing
the resistance to lodging in maize (Josephson and Kincer [1977]) through changing
the position of the cob in relation to lower sections of the stem, while the number of
leaves per plant remained the same. This morphological change wts however accom-
panied by earlier maturity and lower yield. Physiologically such a correlation is not
surprising. A lower insertion of the cob with its high sink capacity brought the cob
closer to the root. Consequently root growth was probably affected by a reduced
assimilate supply resulting in a lower cytokinin export and in accelerated leaf senes-
cence. A yield physiologist would therefore rather recommend a higher insertion of
the cob!
Finally the much discussed problem of leaf orientation (Donald /1963]: Pendelton
et al. [1968]; Eians [1975]) and the strong association between high yields and
erected leaves (Tanner at el. [1966]) should be considered. This problem is pre-
dominantly discussed under the aspect of a more Un iform light distribution within the
canopy, a high photosynthetic rate and a better supply of the storage tissues with
carbohydrates, This problem touches also on the hormonal status of the plant: (i)
Better light interception by lower leaves is associated with more active and prolonged
root growth and therefore with more prolonged water and nutrient uptake and export
of cytokinins (Evans [1975j) which again contribute to the prolongation of the
filling period and to an increase in the single grain weight (AriYanayagam etal. [1974];
Froehlich el al. [1977]. (it) As the auxin activity is closely associated with leaf
orientalion (erect leaf orientation associated with low auxin activity; Paler and
Phillips [1963]: Phillip [1964, 1975]) an influence on the formation of grains (see
chapter 3) would quite be possible. (iii) Leaf orientation is possibly only a link in the
Ixeromorphic syndrome' (characterized by a stronger root system, higher leaf blade
and midrib thickness, shorter plant structure. erect leaves, higher photosynthetic
rate in dry periods and other characteristics). Accordingly. leaf orientation as changed
by breeding would only be a symptom for fundamental changes of tihe metabolism
inClUding the phytohormonal status.
All these examples show that plant breeders should not only rely on statistical data
but also be aware of physiological processes (Adams /1967]) in order to comply with
the old demand not only to select for yield but breed for yield in selecting for yield
prerequisites and those parameters that buid yield (7anaer and lones / 1965]).

107
d) From the agricultural point of view it would last, but not least, be interesting to
know whether specific application of phytohormones would lead to yield increases.
This problem is discussed in detail by Bruinsina (this volume) and by Bangerth [1980].
Of the numerous investigations which have been carried out with cereals some
positive findings are available (Barnsley [1964]; Herzog and Geister [1977J). These
results are however difficult to reproduce (Kursanow [1966]; Michael et al. [1970];
Aufhanmer and Solansky [1972]). This is due to the complexity of biological systems
as well as to the insufficient knowledge about the type of phytohormones needed and
their time of application (Aufliammer [1980]). A systematic and well understood
application of synthetic growth regulators will at the best be a perspective for the
future (Marschner[1978]) and many screening tests for potential growth regulators
are required. It should however always be kept in mind that the interactions between
phytohormones are highly complex (Popp [1979]) so that the slightest changes can
have enormous effects. One of our major attempts should therefore be to influence
by breeding and by husbandry the endogenous hormones to attain their optimal
synthesis and effects (Michael [1978]). The most important step to reach this goal is
however the promotion of the idea that phytohormones play a decisive role in the
regulation of yield formation, to propagate this line of thinking and to encourage
further research in this direction. This was also the objective of this paper!

Acknowledgements

The authors would like to extend their special thanks to all their collegues, in particular
to Prof. Dr. H. Marschner, Prof. Dr. P. Martin, Dr. F. Zink whose friendly advice
contributed to the preparation of the present paper, including Mrs. M_ Labrenz for her
translation work and the 'Deutsche Forschungsgerneinschaft'which financially supported
the experiments by the authors and their coworkers.

6. References

Ackerson R.C. and Chilcote D.O.: Effect of defoliation and TIBA (triiod bencoic acid) on
tillering, dry matter production, and carbohydrate reserves of two cultivars of Kentucky
Bluegrass. Crop Sci. 18, 705-708 (1978)
Adams M. T.: Basis of yield component compensation in crop plants with special reference
to the field bean Phaseolus vidg. Crop Sci. 7, 505 (1967)
Alleweldt G., Dfiring H. and Waitz G.: Untersuchungen zum Mechanismus der Zuckerein-
lagerung in wachsende Weinbeeren. Angew. Bot. 49, 65-73 (1975)
Allinger P., Michael G. and Mart,, P.: Einfluss von Cytokininen und anderen Wuchsstoffen
auf die Stoffverlagerung in abgeschnittenen Bdttern, Flora A 160, 538-551 (1969)
Amir J.: A study on the reproduction stage of the groundnut Arachis hypog. L. Ann. Bot. 33,
333-338 (1969)
Archbold H. K.: Some factors concerned in the process of starch storage in the barley grain.
Nature 156, 70-73 (1945)
Ariyanayagain R.P., Moore C.L. and Carangal V.R.: Selection for leaf angle in maize and its
effect on grain yield and other characters. Crop Sci. 14, 551-556 (1974)
Ashby E_: Studies on the inheritance of physiolog. characters. ]1. Further experiments upon
the basis of hybrid vigor. Ann. Bot. 46, 1007-10032 (1932)
Aspinall D.: The control of tillering in the barley plant. ii. The control of tiller-bud growth
during ear development. Aust. J. biol. Sci. 16, 285-304 (1963)

108
Aufliaumer W.: Phytohormone und Wachstumsregler bei Getreide Ond Mais. It: Anwen-
dungsmdglichkciten von Phytohormonen und Wachstumsregulatoren in der Pflanzen-
produktion (F. Bargerih, Ed.) U[lmer Verlag Stuttgart 1980
Aufliamnier W. and Solanski, S.: Ein Beitrag zur Kontrolle des Assimilatetransportcs tinter
dem Einfluss von Wirkstoffen. Z. Acker- u. Pflanzenbau 135, 143-155 (1972)
Bwnaerth F. (Ed.): Anwendiingsmogl ch kei ten von Phytormonen und WaclstumsregUlatoren
in der Pflanzenproduktion. Hohenheinner Arbeiten, Ulmer Verlag Stuttgart 1980
Barnslev G.E.: N-6-Benzyladenin as a senescence inhibitor. Shell Res. Ltd. Brit. Patent 961,
092. Chem. Abstr. 61. 13816 (1964)
Bellandi D.M.. Doe,jjfing K.: Transport of abscisic acid-2-C-14 in intact pea seedlings.
Physiol. Plant. 32. 365-368 (1974)
Bertger H.: Einfluss von Reifegrad, N-Dtingung auf Fettbildung uni Fettsaiurezusammen-
sctzLng in Haferkornern. Z. Pflanzcncrnihr. Bodenkde 114, 117-127 (1966)
Berh4re, //.: EinflUss der "I'emperat1Jr alof Ertrag kind Fetlbildung in -laferkodrnern. Z. Ptl.
Ernihr. Bodenk. 116, 45-53 (1967)
B r,mamr IV.: Wurzelwachstum und Ernteertrag. Z. Aeker- and Pflanzenbau 97. 337-368
(1954)
Bet,li J.D. and Blak MI: Physiology and biochemistry of seeds in relation to germination.
Vol. /. Development, germination and growth. Springer Berlin, Heidelberg, New York
1978
Rhicham J.: Investigation on the physiology of yield in winier wheat by comparison of
varieties and artificial variation in grain-number per ear. J. agric. Sci., 68, 41 t-422 (1967)
Bjeatron AM.ani Poffiner [Y.G.: The maize germ: Its role as a contributing factor to protein
quantity and quality. Z. Pflanzenzucht. 68, 83-89 (1972)
Black A. L.: AdventitoUs roots, tillers and grain yield of spring wheat as influenccd by
N-P-lertilization. Agr. J. 62, 32-36 (1970)
Breidert D. and Sc/hen W. J. : Die Bildung knd Speicherung von Kohlcnhydraten and Proteinen
in reicenden Gerslenkaryopsen. Angew. Bot. 53, 65-81 (1979)
Brenner P.Mf.: Accumulation of dry iatter and nitrogen by grains in different positions of
the wheat ear as influenced by shading and defoliation. Aust. J. biol. Sci. 25. 657-668
(1972)
Br'emner P. M4.and I)avidson J. L.: A sttidy of grain number in two contrasting wheat cultivars.
Aust. J. agric. Res. 28, 431-441 (1978)
Brocklehtusi P. A.: Factors controlling grain weight in wheat. Nature 266, 348-349 (1977)
Biouiner R. Root fUnctiOning. In: Lcn,drberg J.J., Cutting C. V (Eds.) Environmental effects
on crop physiology, Proc. Synmp. 1975 Long Ashion (1977).
B,-ouier W.: Handb. d. Spez. Pflanzenbates. Parey Berlin. Hamburg 1972
Bru,insma J.: Analysis of growlh, development and yield in a spacing cxperimnt with winter
rye (Sec. cer.). Neth. J. agric. Sci. 14. 198-214 (1966)
Birtansma .: Rolle der Cytokinine bzi Bliten- Utd Frucliten twickhImg. Z. Pflanzenernihr.
Rodenkde 140, 15-23 (1977)
Bm-i-nsma J.: Root hormones and overground development. In: Plant Regul. atid World
Agriculi. (Ed. T. K. Scott) (Plenum Publ. Corp.). 35-47 (1979)
Btbon T., Varga A., Tromp J.. Kneg E. and Brminsmoa J.: Effect of anmonifn and nitrate
nutrition on the levels of zeatin and amino nitrogen in xylem sap of apple roots. Z.
Pflanzenphysiol, 89, 289-295 (197))
Hruoow [V.J. and Carr D.J.: Effect of flooding the root system of sunflower plants on the
cytokinin content in the xylem sap. Physiol. Plant. 22. 1105-1112 (1969)
Carr D.J.. Reid D, M. and Skene KGXM: The supply of gibberellin from the root to the shoot.
Planta 63. 382-392 (1964)
Chin, T Y. and Beevers L.: Changes in endogenous growth regulators it) NOsMxiO1111 ( Tropao-
him) leaves during senescence. Planta 92. 178-188 (1970)
Crookston R. K. and Icks D. R.: Early defoliation affccts corn yield. Crop Sci. /8, 485-489
(1978)
)ashek W. V., Shigh B.N. and WahtontD.C.: Abscisic acid localization and metabolism in
barley aleuron layers. Plant Physiol. 64. 43-48 (1979)
Davies P.J., Probsting W.M. and Gianjfrg,na T.J.: -formonal relationships in whole plant
senescence. Proc. 9. Int. Conf. Plant Growth substances (Pilet P. . Ed.) Lausanne,
273-280 (1977)
Daimrd T. B., Taimer J.PW. and )unnn W.G." Crop Sci. I, 45-48 (1971)

109
Dewdney S.J. and McWha J.A.: Abscisic acid and the movement of photosynthetic assi-
milates towards developing wheat grain. Z. Pflanzenphysiol. 92, 183-186 (1978)
Deutdney S.J. and McWha J.A.: The metabolism and transport of abscisic acid during grain
filling in wheat. J. exp. Bot. 29, 1299-1308 (1979)
Doerffling K.: Quantitative V&randerungen des Abscisins5uregehaltes wihrend der Frucht-
entwicklung von So/nnan lycopers. Planta 93, 233-242 (1970)
Doerfffing K.: Das Phytohormon Abszisins5ure. Biolog. Rundschau 9, 129-143 (1971)
Doerffling K.: Speicherungsprozesse - die Rolle der Phytohormone. Z. Pfianzenerniihr.
Bodenkde 140, 3-14 (1977)
Doerffling K.: Growth. Fortschr. Bot. 40, 150-167 (1978)
Doerffling K. and Bdtlger M.: Transport von Abscisinsaure in Explantaten, Blattstiel und
Internodialsegmenten. Planta 80, 299 (1968)
Donald C.M.: Competition among crop and pasture plants. Adv. Agr. 15, 1-118 (1963)
Donald C.M.: The breeding of crop idiotypes. Euphytica 17, 385-403 (1968)
Dressel L.J. and Kuehn H.: Obur die unterschiedliche Reaktion von in CCC-haltiger Ndhr-
I6sung gewachscnen Roggenpflanzen. Z. Pflanzencrnihr. Bodenkde 129, 236-245 (1971)
Eahin J.D., Haskins F.A., Sullivan C. Y. and van Bavet C. H. M.: Physiolog. aspects of crop
yield. Proc. Symp. Nebrasca. Wisconsin 1969
ESetowens C.J. and Schwabe W. W.: Seed and pod development in Pisul sat. in relation to
extracted and applied hormones. J. exp. Bot. 26, 1-14 (1975)
Eschrich W.: Biochemistry and fine structure of phloem. Anna. Rev. Plant Physiol. 21, 193-214
(1970)
Evans L.T.: Storage capacity as a limitation on grain yield. Int. Rice Inst., Los Banos,
Philippines, 499-511 (1972)
Evans L. T.: Rapid responses to plant hormones. Annu. Rev. Plant Physiol. 25, 195-223
(1974)
Evans L. T.: The physiological basis of crop yield. h,: Evans (Ed.) Physiology - some case
histories, Cambridge 327-355, 1975
Evans L. T., Bingham J., Jackson P. and Sutherland J.: Effect of awns and drought on the
supply of photosynthates and its distribution within wheat ears. Ann. appl. Biol. 70,
67-76 (.1972)
Evans L. T. and Wardlaw 1 F.: Comparative physiology of grain yields in cereals. Adv, Agr.
28, 301-359 (1976)
Fanner R. and Doerffling K.: Effect of kinetin and abscisic acid on the distribution of 2-amino
(l-14C) isobutyric acid in leaves of Zea mays. Plant Physiol. 48, 131-133 (1980)
Fisher J.E.: Developmental morphology of the inflorescence in hexaploid wheat cultivars
with and without the cultivar Norin-10 in their ancestry. Can. J. Plant Sci. 53, 7-15 (1973)
Fletcher R.A. and Adedipe N.O.: Hormonal regulation of leaf senescence in intact plants.
Plant growth substances 1970 (Carr ). J. Ed.), 571-580
Froehlich W.G., Pol,ner W.G. and Klein D.: Performance of upright-leaved liguleless-2
and liguleless-3 maize hybrids adapted to central European climatic conditions. Z.
Pflanzenzichtung 79, 134-144 (1977)
Gale M.: Towards super cereals. New Scientist 61, 248-251 (1974)
Giflbrd R.M., Breinner P.B. and Jones D.B.: Assessing photosynthetic limitation to grain
yield in a field crop. Aust. J. agric. Res. 24, 297-307 (1973)
G6ring H. and Mardanow A.A.: Einfluss von Stickstoffmang2l auf das K/Ca-Verhaltnis und
den Zytokining-halt jung.r Kirbispilanzen. Biochem., Physiol. Pflanzen (BPP) 170,
261-264 (1976)
Go/dbach H. and Goldbach E.: Abseisic acid translocation and influence of water stress on
grain abscisic acid cantent. 1 exp. Bot. 28, 1342-1350 (1977)
Goldbach H., Goldbach E. and Michael G.: Transport of abscisic acid from leaves to grains
in wheat and barley plants. Naturwissenschaften 64, 488-489 (1977)
Goldbach E., Goldcbach H., Wagner H. and Michael G.: Influence of N-deficiency on th2 abscisic
acid content of sunflower plants. Physiol. Plant. 34, 138-140 (1975)
Goldbach H. and Michael, G.: Abscisic acid content of barley grains during ripening as
affected by temperatur and variety. Crop Sci. 16, 797-799 (1976)
Graebe I.E. and Ropers H.J.: Gibberellins. in: Phytohorniones and related compotnds;
Lethain, Goodwin, Higgins (Edts). Elsevier Amsterdam, 107-203, 1978
Grahl A. and Schroedter H.: Witterung vor der Reife und Keimruhe von Weizen unter b.-
sonderer Bericksichtigung der Auswuchsvorhersage. Seed Technol. 3, 815-826 (1975)

J10
Gfinzel G.: Die Wirkung extrem hoher und spiter N-Gaben auf die QUalitatsmerkmale aus-
gewaihller Winter- und Sommerweizensorten. Z. Acker- u. Pflanzenbau 128, 93-116
(1968)
Grundbacher F.J.: The physiological function of cereal awns. Bot. Rev. 29. 366-381 (1963)
Haensel. H.: Physiologic der Ertragsbildung und die Zichtung auf Ertrag bci Getreide.
Z. Planzenziichtung 54. 97-1 10 (1965)
Hagemtn R. H.: Factors affecting yield of cereal grains via physiological processes. Syrp.
Plant Growth Regulators Working Group. Arcansas, 14-42, 1977
[Hnif M. and Lmncer R. H, M.: The vascular system of the spikelet in wheal. Ann. Bot. 36,
721-727 (1972)
anitus /1.: Die Becinflussung des Wurzelsvsterns von Weizen durch CCC. Z. Acker- u.
Pflanzenbau 125, 40-46 (1967)
HerzOg H. and Geisler G.: Der Einfluss der Cytokininapplikation auf die Assimilatecin-
lagerng und die endogene Cytokininaktivitit der Karyopsen bei zwei Somnuerweizen-
sorten. Z. Acker- u. Pflanzenbau 144. 230-242 (1977)
Hevland K. U.: 01her die Bedeutung der Enahrung in verschiedenen Entwickingsstadien for
den Ertrag der Sommergerste. Z. Acker- u. I'llanzenbau 113, 41-65 (1961)
Ho D.7-H, and Vainer i.E.: Response of barley aleuron layers to abscisic acid. Plant Physiol.
57. 175-178 (1976)
Hol...es D.P.: Inflorescence development of semidwarf and standard wheat cultivars in
different photoperiod and nitrogen treatments. Can. J. Bot. 51, 941-956 (1973a)
Holme.s D. P.: Effect of defoliation on chlorophyll loss in seneseing wheat inflorescences and
on grain maturation. Can. J. Plant Sci. 53. 499-500 (1973b)
Itai C. II. and Henzioni t.: Regulation of the senescence and related metabolic processes.
Proc. 8. Int. Conf. Plant Growth Substances, Tokyo. 1022-1026 (1973)
I1ai C.H.. Ben,zioni .. and Oriin L.: Correlations changes in endogenous hormon level and
shoot growth induced by short heat treatment to the root. Physiol. Plant. 29. 355-360
(1973)
Itai C. 11.and Vaadia Y.: Kinetin-like activity in root exudate of water stress sunflower plants.
Physiol. Plant. 18, 941 .944 (1965)
Ja,ckso M.B. and Osborne D.J,: Abscisic acid, auxin and ethylene in explant abscision.
J. exp. Bot. 23, 849-862 (1972)
Jenner C. F.: Relationship between level of soluble carbohydrate and starch synthesis in
detached ears of wheat. Aust. i1biol. Sci. 23. 991-1003 (1970)
lener C. F.: Wheat grains and spinach leaves as accumlators of starch. Proc. Symp. Trans-
port, transfer processes in plant. Wardlaw and Passioura Edts. Australia, 73-83 (1976)
Jenner C.F. and Ratlijen A.J.: Factors regulating the accumTlation of starch in ripening
wheat grains. Aust. J. Plant Physiol. 2, 311-322 (1975)
.lolansson N.- Joren,ssm,, I. Maltiss m B.. Olsson G. and Srensson G.: The heat reqUirereont of
dillerent wheat cultivars to attain ipeness to germinate and the prospects of developing
a warning system for farmers from this. Sver. Utsiades for. Tidskr. 86, 233-241 (1976)
Jotes R.L. and Phil/ips I,D.J.: Organs of gibberellin synthesis in light grown sunflower
plants, Plant Physiol. 41. [381-1386 (1966)
Josephson L.I. and Kincer I/. C.: Selection tor lower ear placement in two synthetic poptI-
lations of maize. Crop Sci. 17, 499-502 (1977)
lonr .1. und Srm, H.: Wachstumsreguliirende WirkUng vol Chlorcholinchlorid. Landw.
Forschung 17, 1-9 (1964)
Kende II.: Kinetin-like factors in root exudate of sunflowers. Proc. Nat. Alkad. Sci. 53,
1302-1307 (1965)
Khan Ai.A. terheck R., Wcoers E.C. and ian, On,kelen, HA.: Finbryoless wheat grain. Plant
Physiol. 51. 641-645 (1973)
Khe P. W.: Abscisic acid in developing wheat grains and its relation to grain growth and
maturity. Planta 132, 43-51 (1976)
Kin,g R, W.: Abscisic acid synthesis and metabolism in wheat ears. AtOst. J. Plant Physiol. 6,
99-108 (1979)
Kleme, F. and Klmobt D.." Half-live of sRNA from primary roots of zea mays. A contribution
to the cytokinin production. Physiol. Plant, 3/. 186-188 (1974)
Klemm K.: Der Einfluss der N-Form auf die Ertragsbildung verschiedener KtltUrpflanzen.
Bodenkultur 17, 265-284 (1966)

Il
Krauss A.: Tuberization and abscisic-acid content in Solanum tuberosum as affected by
nitrogen nutrition. Potato Res. 21, 183-193 (1978)
Kursanow A.L., Kulaeva O.N. and Konovalov J.B.: Ober die M6glichkeit der Anwendung
von Kininen far die Aktivierung des Reifens und Keimens der Samen. Agrochimija
107-114 (1966)
Langer R. H.,, Prasad P. C. and Laude H.M.: Effect of kinetin on tiller bud elongation in wheat
(Trit. aest. L.) Ann. Bot. 37, 565-571 (1973)
Laude H.M., Ridley J.R. and Suneson C.A.: Tiller senescence and grain development in
barley. Crop Sci. 7, 261-263 (1967)
Lenton J.R., Boven M.R. and Saunders P.F.: Detection of abscisic acid in the xylem sap of
willow (Salix vin.) by gas-liquid chromatography. Nature 220, 86-87 (1968)
LeopodA. C.: The control of tilering in grasses by auxin. Am. J. Bot. 36, 437-440 (1949)
Le/hanz D. S.: Chemistry and physiology of k inetin-like compounds Annu. Rev. Plant Physiol.
18, 349-364 (1967)
Linser H., Kiermayer 0. und Youssef E.: Der Wuchsstoffgehalt verschiedener Brassica-
Pflanzen. Planta 52, 173-186 (1958)
Lupton F. G. H.: Translocation of photosynthetic assimilates in wheat. Ann. appl. Biol. 57,
355 (1966a)
Lupton F.G.H.: The use of physiological characters in breeding for yield in wheat. Acta
Agric. Scand., Eucarpia Suppl. 16, Stockholm, 174-176 (1966b)
Lupton F.G.H. and Ali M.A.M.: Studies on photosynthesis in the ear of wheat. Ann. appl.
Biol. 57, 281-286 (1966)
Marschner H.: Erndhrungs- und ertragsphysiologische Aspekte der Pflanzenerndhrung.
Angew. Bot. 52, 71-87 (1978)
Martinez-CarrascoR. and Thorne G.N.: Physiological factors limiting grain size in wheat.
J. cxp. Bot. 30, 669-679 (1979)
McKee H.S.: Review of recent work on nitrogen metabolism. New Phytologist 48, 1-83
(1949)
McWha JA.: Changes in abscisic acid levels in developing grain of wheat (Trit. aest.)
J. exp. Bot. 26, 823-827 (1975)
McWha J.A. and Jackson D.L.: Some growth promotive effects of abscisic acid. J. exp.
Bot. 27, 1004-1007 (1976)
Mengel K.: Energy absorption, energy conversion, and energy storage of crops in relation to
endogenous and exogenous factors. hi: Fertilizer use and production of carbohydrates.
Proc. 13. Coil. Internat. Potash lnst., Bern, 61-82 (1977)
Mengel K. and Forster H.: Der Einfluss einer zeitlich variierten unterbrochenen K-Erndhrung
auf Ertrags- und Qualitdltsmerkmale von Gerste. Z. Acker- u. Pflanzenbau 127, 317-326
(1968)
Mengel K. and Kirkby E.A.: Principles of plant nutrition. Intern. Potash Inst., Bern 1978
Michael G.: Nitrogen fertilizer application, hormon activity and crop production. Proc. 7.
Intern. Coll. Plant Analysis and Fertilizer Probl., Hannover, 307-316 (1974)
Michael G.: Stickstoffernihrung, Phytohormonaktivitlit und Stoffbildung bei Kulturpflanzen.
Landw. Forsch. 32, 110-118 (1979)
Michael G., Allinger P. and Wilberg E.: Einige Aspekte zur homonalen Regulation der Korn-
grosse bei Getreide. Z. Pflanzenern&hr. Bodenkde 125, 24-35 (1970)
Michael G. and Seiler-Kelbitsch H.: Cytokinin content and kernel size of barley grains as
affected by environmental and genetic factors. Crop Sci. 12, 162-165 (1972)
Michael G., Seiler-Kelbitsch. H. and Wagner H.: Some aspects on hormonal regulation of
grain size of barley. Proc. Internat. Sympos. Breeding and productivity of barley. Kro-
meriz. 397-412, 1973
Michael G., Schultz R. and Wagner H.: Modeilversuche zur Bedeutung der Granne far die
Versorgung des Gerstenkornes mit Cytokininen. Flora A 160, 306-316 (1969)
Milborrow B. V.: The chemistry and physiology of abscisic acid. Annu. Rev. Plant Physiol. 85,
259-307 (.1974)
Milborrow B. V.: Antitranspirants and its regulation of abscisic acid. Aust. J. Plant Physiol.6,
249-254 (1979)
Milborrow B. V. and Robinson D. R.: Factors affecting the biosynthesis of abscisic acid. J.
exp. Bot. 24, 537-548 (1973)
Mitchel K.J.: Influence of light and temperature on the growth on rye grass 11. Control of
lateral bud development. Physiol. Plant. 6, 425-443 (1953)

112
Mizrahi Y. and Richmond A.E.: Abscisic acid in relation to mineral deprivation. Plant
Physiol. 50, 667-670 (1972)
Mothes K.: Uber das Altern der Bl5tter Uod die M6glichkeit der Wiederverjungung. Natur-
wiss. 47. 337-351 (1960)
Mothes K. and Enge/brecht L.: On the activity of a kinetin-like root factor. Life S.t 2,852-857
(1963)
Mothes K., 1nietbrecht L. und Kulajeva 0.: DIbcr die Wirkung des Kinetins auf Sticksot-
verteilung und Eiweisssyntheie in isolierten Blittern. Flora 147, 445-465 (1959)
V4othe.v K., EngelbreclaI L. und Schfitte H. R.: fiber die AkkUnulation von Aminoisobutter-
saire im Blatgewebe tinter dem Einlluss von Kinetin. Physiol. Plant. 14, 72-75 (1961)
Mounia Nl. A. Ki,: Gibberellin-like substances in parts of developing barley grain. Physiol.
Plant. 44. 268-272 (1978)
Mounla A.A. Kt. and Michael G.: Gibberellin-like substances in developing barley grain and
their relation to dry weight increase. Physiol. Plant. 29, 274-276 (1973)
Mustrata Y and Mas utshima S.: Rice. /i: Crop physiology, soie case histories (f.vans L. T.
Ed.). Cambridge 73-99 (1975)
Nakara S. and Leopold A. C.: Radioautographic study of translocation in bean leaves. Amer.
..Pot. 54. 769-772 (1967)
Neales "- F. and Incole I. D.: The control of leaf photosynthesis rate by the level of assimilate
concentration in the leaf. A review of the hypothesis. Bot. Rev. 34, 107-125 (1968)
Nitsh ., P.: Growth and morphogenesis of the strawberries as related to auxin. Amer. J. Bt.
37, 211-215 (1950)
01iteeni L. B., A ustin R. B. and lingham J.: Effect of awrs on the photosynthesis and yield
of wheat. Trit. aest. Ann. appl. Biol. 84, 241-250 (1976)
Orlorhs K. und Hoefier W.: Eirlluss einer variierten Stickstoffdbngung auf Assimilations-
leisitUng Und Ertragsbildtng Von Sommerweizen. Z. Pflanzenernhr. Bodenkde 139,
63 I-(40 (1976)
PalmerJ. H. and Phillips L. J.: The effect of the terminal bud, indolylacetic acid. and nitrogen
supply on the growth and orientation of the petiole of Helianahus ann. Physiol. Plant. 16,
572-584 (1963)
Pen,dleton .1.W. Smith G.E., Winter S. R. and john.vton T.J.: Field investigations on the
relationship of leaf angle in corn to grain yield and apparent photosynthesis. Agron. J. 60,
422-424 (1968)
Phillips 1. D.J.: Root-shoot hormone relations. [. The importance of an aerated root system
in the regulation of growth hormone levels in the shoot of HIelianthus ann. Ann. Bt. 28,
17-35 (1964)
Phillips 1. O.J.: Apical dominance. Ann. Rev. Plant Physiol. 26, 341-367 (1975)
Popp M.: Probleme der Intcrakion zwischen Phvtobornionen. Ber. Deutsch. hot. (Ges.92,
323-339 (1979)
Proctsazka S.: Effect of indol-3-acetic acid on the translocation of assimilates in winter wheat
in the period of kernel formation. Acta Univ. Agr. Brne XXTV/. 99-104 (1978)
Radenacher W.: Gaschroniatographische Analyse der Verainderungen in Hormongehalt
des wachsenden Weizenkornes. Diss. Gottingen 1978
Radle, M.: The development of wheat grains in relation to endogenous growth substances.
J. exp. Hot. 27, 1009-1021 (1976)
Rodl, Xf.: Factors affecting grain enlargement in wheat. J. exp. Bet. 29, 918-934 (1978)
Ratri,opal 1/. and Rao J.M., Growth response of tomato plants to exogenous applications
of IAA and GA, under nitrogen stress, Ind. J. Plant Physio. XV. 95-102 (1972)
Raissn H. .: An upper limit for spikelet number per ear in wvheat as controlled by photo-
period. Aust. J. Agr. Res. 22. 537-546 (1971)
Ravson H.,VI. and Evans L. T.: The pattern of grain growth within the ear of wheat. Aust.
J. biol. Sci. 23, 753-764 (1970)
Rawvoni H. AV.and Evans L. T.." The contribution of stem reserves to grain development in a
range of wheat ctultivars of different height. Aust. J. Agr. Res. 22, 851 (1971)
Ranson ll. and Hafitr G.: Translocation and remobilisation of 14-C assimilated at
difl"erent stages by each leaf of the wheat plant. Aust. J. biol. Sci. 22. 321-331 (1969)
Reid D.V1. and Crozier A.: The effect of water logging on the gibbereltin content and growth
of tonato plants. .. exp. Bot. 22. 39-48 (1971)
Rielan..tid E. A. and l.ang A.: Effect of kinetin on protein content and survival of detached
xanthium-leaves. Science 125. 650-651 (1957)

113
Roemer W.: Untersuchungen fiber die Auslastung des Photosyntheseapparates bei Gerste und
weissem Serif in Abhingigkeit von den Umweltbedingungen. Arch. Bodenfruchtbark.
Pflanzenprodukt. 15, 415-423 (1971)
Ruckenbauer P.: CCC-Versuche mit Winterweizenpflanzen in Ndhrlbsungen. Bodenkultur 19,
306-319 (1968)
.Ruckenbauer P. und Kirby E. J. M.: Effect of kinetin on the growth and development of barley
and its interactions with root size. J. agric. Sci. 80, 2l1-217 (1973)
Sachs T. and Thiman K.J.: The role of auxins and cytokinins in the release of buds from dor-
mance. Am. J. Bot. 54, 136-144 (1967)
Saeulesku N. und Popa T.: Der Keimling als Ertragsfiihigkeitsanzeiger bei Mais. Z. Pflanzen-
z6chtung 54, 1-5 (1965)
Saghir A.R., Khan A.R. and Worzella W.: Effects of plant parts on the grain yield, kernel
weight, and plant height of wheat and barley. Agr. J. 60, 95-97 (1968)
Sagronisky H.: Zur Bedeutung der Gerstengranne fur die Kornentwicklung. Z. Pflanzen-
zibchtung 33, 267-284 (1954)
Saini A., Sud A. and Nanda R.: Growth regulators in relation to tillering in wheat. Ind. ..
Plant Physiol. 18, 140-146 (1975)
Sakri F.A.K. and Shannon J.C.: Movement of 14-C-labeled sugars into the kernels of wheat
(Trit. aest. L.) Plant Physiol. 55, 881-889 (1975)
Salama, A.M.S. and Wareing P.F.: Effects of mineral nutrition on endogenous cytokinins
in plants of sunflower (Helianth. ann.) J. exp. Bot. 30, 971-981 (1979)
Sattelnacher 13. and Marschner H.: Nitrogen nutrition and cytokinin activity in solanum
tuberosum. Physiol. Plant, 42, 185-189 (1977)
Scheffer, K. and Bruder W.: Untersuchungen fiber das Bestockungsverm6gen und scin Ein-
fluss auf die Ertragsbildung von Sommerweizen- und Sommergerstensorten. Kalibriefe
(Biintehof) 14, (8), 545-554 (1979)
Scheibe A. und Ellermann M.: Ober den Einfluss von Licht und Temperatur auf den Be-
stockungsvorgang somnierannueller Getreidearten. Z. Acker- u. Pflanzenbau 126, 197-228
(1967)
Schneider G.: Morphactins: Physiology and performance. Annu. Rev. Plant Physiol. 21,
499-536 (.1970)
Schwanitz F.: Das Ertragsproblem in entwicklungsphysiologischer Sicht. Zilchter 30, 45-56
(1960)
Schweers W. und Meyer F. H.: Einfluss der Mycorrhiza auf den Transport von Assimilaten.
Ber. Deutsch. Bot. Ges. 83, 103-119 (1970)
Seiler-Kelbitsch, H., Michael G., Hdiuser H. and Fischbeck G.: Cytokiningehalt und Korn-
entwicklung von Gerstenmutanten mit unterschiedlicher Korngr6sse. Z. Pflanzen-
zfichtung 75, 311-316 (1975)
Seiler-Kelbitsch H., Michael G. Und Wilberg E.: Beziehungen zwischen Korngewicht und
CytokininaktivitAt bei Sommergerste, untersucht am Beispiel des Abschneidens von
Bestockungstrieben. Angew. Botan. 48, 299-307 (1974)
Seth A. K. and Wareing P. F.: Hormone directed transport of metabolites and its possible role
in plant senescence. J. exp. Bot. 18, 65-77 (1967)
Sition D.A., Itai C. and Kende H.: Decreased cytokinin production in the roots as a factor in
shoot senescence. Planta 73, 296-300 (1967)
Sommer K. und Poletschny H.: N-serve und Siniazin bei Ammonium- und Nitraternahrung
von Weizen. Landw. Forschung 25, 203-215 (1972)
Spiertz J. H.J.: Grain growth and distribution of dry matter in the wheat plant as influenced
by temperature, light energy and ear size. Neth. J. agr. Sci. 22, 207-220 (1974)
Stamp P.H. and Geisler G.: Der Verlauf des Kornwachstums in Abh5ngigkeit von der Korn-
position bei zwei Sommerweizensorten. Z. Acker- u. Pflanzenbau 142, 264-274 (1976)
Stamp P.H. und Herzog H.: Untersuchungen zur Fahnenblattalterung und zum Kornwachs-
turn einiger deutscher Sommerweizensorten. Z. PflanzenzOchtung 77, 330-338 (1976)
Sieveninck, van R. .M.: Factors affecting the abscission of reproductive organs in yellow
lupins (Lup. lut.). The effect of different patterns of flower removal. J exp. Bot. 8,
373-381 (1957)
Steineck 0.: Die Wirkung der N-Form auf die Stoffbildung von Hafer bei konstantem Niihr-
stoffangebot in Nihrl6sungskultur. Bodenkultur 17, 248-264 (1966)
Stay V.: Photosynthesis, respiration and carbohydrate
/
accumulation in spring wheat in
relation to yield. Physiol. Plant Suppl. I Lund 1965

114
Stoy V.: Photosyntheseleistung und Assimilateverlagerung als ertragsbegrenzende Faktoren
im Getreidebau. Vortrzige F.Pflanzenz6chter 13, 34-51 (1973)
Siov V.: Giber die Assimilatebildung bei Cetreide und Ackerbohnen. Getreide, Mchl ind
Brot. 30, 85-89 (1976)
Sunt' V/: Trockensubstanzproduktion und Assimilateverlagerung in das Getreidekorn. Z.
Pflanzenerndhr. Bodenkunde 140, 35-50 (1977)
SioY V. and Hagberg A.: Effects of growth regUlators on ear density mUtanis in barley.
Hereditas 58. 359-384 (1967)
Tuains J.A.. Wa/lace D. H., LudJbrd P.M.and Oburn J. L.: Effect of older fruits on abortion
and abscisic acid concentration of younger frints in Phaseolus v'ui. L. Plant Physiol. 64,
620-622 (1979)
7tannerJ. IV., Gardener C.J., Stoskopf N.C. and Reinbergo E.: Some observations on upright-
leaf-typ small grains. Can. J.Plant. Sci. 46, 690 (1966)
Tanner J.IV. and lones G.E.: ProdUction physiology. In: Genes to GCenus (Greet F.A. and
Armn T.J. Edts.) Symp. Plant Growth. Internat. Min. a.Chein. Corporat. Illinois, 1965
Thorne G.N.: Survival of tillers and distribution of dry inatter between ear and shoot of
barley varieties. Ann. Bot. 26. 37-54 (1962)
Thurne G.N., PIysiolog. factors limiting the yield of arable crops, In: Potential crop pro-
duction (Weartig and Cooper. Edts.) London, 143-158 (1971)
Tietz A. and Doerj/lig K.: Veriinderungen in Gehalt von Abscisinsliure und Indolcssigsidure
der Chloroplastenfarbstoffe in Pisunrkeimilingcn durch GA,. Planta 85, 118-125 (1969)
laWIenuar M,: Sink-source relationship during reprodUctive development in maize. A review.
Maydica XXII, 49-75 (1977)
TonreiJ.G.: Root hormones and plant growth. Anno. Rev. Plant Physiol. 27, 435-459 (1976)
Tvai C. Y.. Hi,be,r. A,.and Warren 1I, L.. Relationship of the kernel sink for N to maize
productivity. Crop Sci. 18. 399-404 (1978)
'a g A. and Brttbutna ,.: rhe growth and ripening of tomato fruits at different levels of
endogenous cytokiTlins I. Hortic. Sci. 49, 135-142 (1974)
l,),tk C. R.: Origin of cytukinins transported in the phloeni. Physiol. Plant. 46, 235-240 (1979)
Wag,tner //.. WichstotfgesteuerteAssinmilateverlagClrungbei Gerste. Angew. Bolan. 48, 331-338
(1974)
ILcagner H. Und Michael G.: Der Einfluss unterschiedlicher Stickstoflversorgung auf die
Cytokininbildung in WurMln. Biochem. Physiol. Pflanzen (BPP) 162, 147-158 (1971)
Waklo J,L.: Role of mineral nutrients and growth regulators in the apical dominance in
So/ainim sismbrifoi, Planta 91, 190- 194 (1970)
[Valpole P.R. and M organ D.G.: The effect of floiet sterilization oilgrain number and grain
weight in wheat ears. Ann. Hot. 37, 1041-1048 (1973)
ia/lton D.C.: Abscisic acid and seed germination. It: Physiology and biochemistry of seed
dormancy and germination (Khan A,A. Ed.) Amsterdam. 145-156, 1977
lVardllo J.F.: The early stages of grain development in wheat: Response to light and tempera-
ture in a single variety. Aust. J.bio. Sci. 23. 765-776 (1970)
friceing P. F. and Sounders P.F.: Hormones and dormancy. Anini. Rev. Plant Physiol. 22,
261-288 (1971)
Weiss C. and Vaadia Y.: Kinctin-like activity in root apices of sunflower plants. Life Sci. 4,
1323-1326 (1965)
Wheeler .4.14/: Changes in growth-subsiance content during growth of wheat. J.appl. Biol.
72. 327-334 (1972)
Wheeler A. JV.: Some treatments affecting growth substances in developing wheat ears. Ann
appl. Hot. 83, 455-462 (1976)
W,ity C. ).and flall R. W.: A cytokinin oxidase inZea mays, Can. .j. Biochcn. 52, 789-799
(1974)
Withan, F. t. and M/iller C.O.: Level of a kinctin-like factor in the developing maize plant.
Plant Physiol. 38. XXXVI1 (1963)
iood 4.and Pale, L.G.: Gibberellic acid ol the permeability of model mernbran systems.
Plant Physiol. 50, 103-108 (1972)
tright S,T. C. and Hiron R. IV. P.: The acCUotmlation of abscisic acid inplants dutiring Wilting
and Under water stress. In: Plant Growth Substalnces (Carr D.1.. Ed.) Berlin, 291-298,
1970
Zeeiart J.A.D . Sites of abscisic acid synthesis and metabolistm in RicinOs comm. Plant
Physiol. 59. 788-791 (1977)

115
Zimnernann M.H.: Translocation of organic substances in trees. 111. The removal of sugars
from sieve tubes in White Ash. Plant Physiol. 33, 213-217 (1958)
Zink F.: Effect of late application of nitrogen on the grain protein of three high-yielding maize
hybrids with different protein or lysin content. In: Seed protein improvement in cereals
and grain legumes. Proc. Symp. Neuherberg, Internat. Atom En. Ag. Vienna, 1979

116
The Endogenous Hormonal Pattern
and its Interference
by Exogenous Plant Growth Regulators
J. Bruinsma, Department of Plant Physiology, Agricultural University Wageningen/
The Netherlands*

Sumniari

The activities of sinks and sources in plants are regUlated by the endogenous hormonal
patterns that result from the biosynthesis, translocation, and metabolism of the various
components of these patterns.
The application of exogenous plant growth regulators may interfere with these processes or
supplement components at suboptinal concentrations. Examples of such effects are
presented and a number of complications are discussed: the relationship between hormone
level and effect need not be positive, exogenous substances may arrive at other sites than the
corresponding endogenous ones and, accordingly, exert different effects.
It is concluded that the fate of exogenous regulators and their effects upon the amounts and
ttirnover rates of the endogenous components of the hormonal pattern have to be further
elucidated in order to arrive at better and new possibilities for their application.

1. Introduction

The metabolic activities of' sources and sinks in plants not only depend oil their
nutritional state, they are also regulated by the hormonal patterns prevailing in these
organs. The various hormonal substances locally influence the expression of the genetic
information from the nuclei of the tissue cells and also affect the functioning of enzymes
and other proteins in the cytoplasm and its membranes. In these actions tile hormones
mutually interfere, they act together, additively or even synergistically, or they coun-
teract each other. The entire hormonal pattern determines the resulting activity and.
thereby, the growth and development of the organ.
The hormonal pattern within a given organ is the result of the biosynthetic rates of
the various hormones, of their translocation towards and from the organ, and of their
inactivation by metabolic conjugation or breakdown. The application of exogenous
plant growth regulators in order to direct crop productivity exerts its cfl'cct by inter-
ference with these processes. Plant growth regulators can simulate the mode of action
of plant hormones or interact with the synthesis. translocation. or inactivation of
these substances.
In the following, a few examples will illustrate the biosynthesis and long distance
transport of phytohoiniones. We shall then discuss how far the exogenous application
* Prof. Dr. j1. Bruinsipta. Department of Plant Physiology of the Agricultural University,
Arboretumlaan 4. NL-6703 BD Wageningen/The Netherlands

117
of plant growth regulators can inform us about the endogenous hormonal pattern.
The effect of exogenous regulators even with determination of the endogenous
components of the pattern may be insufficient to give us a reliable insight. Deter-
minations of turnover rates of hormones and exogenous regulators, and of their
localization within the tissue and the cells are required to understand the regulation
of sinks and sources in crop plants by application of growth regulators.

2. Biosynthesis of phytohormones

Many of the plant hormones are isoprenoid compounds, formed partially or entirely
from one or more isoprene units (Figure I). They are therefore chemically related to
such substances as the carotenoids, steroids, essential oils, resins, and rubber. Growth-
regulating effects of e.g. steroids, incidentally mentioned in the literature (Geuis
[11]), might be due to end-product inhibition of the synthesis of these substances so
that the synthesis of hormonal isoprenoids can be stimulated. The cytokinins derive
their side chain from a single isoprene unit. The abscisins, like abscisic acid and
xanthoxin, are trimers, the gibberellins tetramers of these units.
Inhibitors of the biosynthesis of the gibberellins have been screened as growth retar-
dants by the phytopharmaceutical industry. Such substances as chlormequat and
ancymidol specifically reduce effects of endogenous gibberellins without adverse side
effects so that they can safely be used to retard stem growth, e.g. in cereals and in
ornamentals, or to promote flower formation, e.g. in pome fruits (Jung [14], Skytt
Andersen [21]).

Oc CHt3 ,CH 2 3ATP 3 OP+P

HO 2 OH ,20
CH CH

mevalonic acid isopentonyl- and dimothylallyl-

pyrophosphals (isoprene units)

CYTOKININS ABSCISINS GIBBERELLINS

N N P ga,.n,yla,rnyl-

e~}1) . pyrophoSphale
ro w
dimet.pn.grohylallyw tth retardants

F20P j O 11 Qk .,en .
ab.cisic acid OH
(ABA)

HN CHH2 0OH 0H

0 0 C<
H HO C' HO 601
H .oOH
trans-zeatin xanthoxin gibber.llic ac,d (GA 3 )

Fig. 1. Phytohormones derived from isoprene units

118
Other hormones are derived from the metabolism of amino acids (Figure 2). The
auxins, indoleacetic acid and phenylacetic acid, probably originate from the amino
acids tryptophan and phenylalanine. respectively. Ethylene is derived from homoserine
with a recycling of methionine, while at least in fungi glultamic acid can also play a role
as precursor (Lieherman [17.). The biosynthesis through methionine has recently
been elucidated (Adams and Yatg [I, Lfurssen et ed. [18]) and occurs in three steps.
File second reaction is the most interesting one, since it is generally rate-limiting. It can
be greatly enhanced by auxins and numerous effects of the application of, e.g.,
synthetic auxins in plant growth and development are based on the acceleration of
ethylene formation through this pathway. The same step can also very effectively be
inhibited by mcthionine analogs, e.g. aminoethoxyvinyl glycine (AVG). and this
opens highly interesting perspectives to counteract ethylene effects, especially its
promotion of such senescence phenomena as leaf yellowing, flower wilting, and fruit

The product of this second reaction, the cyclic aminoacid I-amninocyclopropane-l-

carboxylic acid (ACC), requires peroxidation to be decomposed into ethylene and
other products. In the roots of inundated plants, insufficient oxygen is present for this
conversion and ACC is translocated as such by the transpiration stream into the
foliage. Within the leaves ethylene is generated from its precursor and causes the
epinasty that is so characteristic for inundated plants. ACC is much easier to apply to
plants than the gaseous ethylene and has therefore possibilities for use as a plant
growth regulator to induce etfbyille effects.

AUXINS ETHYLENE

NH2
CH-C-C' H0H -C&NH2 CHrS-CHCHrC-C,
H H O F I
NHz ' OH I N H OH

iPI ha 0sr0I methion,ne PPI, P ,

HH
-
t.dlo,,n,t,1

' pCH-C -C' C-C--Caeoh, NH,

AVcH
~~A0.J 1,J
en,Ip ruvic acid
-- S-denosine--
S-adnuy[meih,~on,ne (SAM)
d ,vhc acd A
H
Ioy AVO

CH2 e 0

CH, NH,
4C0

O-- yip acidIACC)

ICarb

?
CH?CH.I H-( K NHI'CO2

Feg. 2. Phytohorniones derived run amino acids

119
3. Translocation of phytohormones

The precursor of ethylene, ACC, offers an example already of the long-distance trans-
port of hormones. As a gas, ethylene travels through the tissue mainly via the inter-
cellular space but its precursor can be translo&ated throughout the plant in the
vascular bundles, in our example through the xylem vessels from the roots to the
overground parts.
Since many agricultural measures, such as the supply of water, mineral nutrients, and
man-made chemicals, are applied to the soil, the signals of the root system to the
aerial parts are an important link in the chain from such a measure to a response in
crop productivity. These signals are more often than not of a hormonal nature
(Bruinstna [5]). For example, drought causes stomata closure which, in turn,
decreases the entry into the foliage of carbon dioxide and its conversion into dry matter.
This stomatal closure occurs even before the water potential of the leaf tissue drops
appreciably. It is the roots which first become aware of the lack of water. As an
immediate response, they reduce their biosynthesis of cytokinins, so that the tran-
spiration stream to the leaves contains less of these hormones. This signal causes the
chloroplasts in the mesophyll of the leaves to release large quantities of abscisic acid
which acts in two ways. On the one hand, it reduces the loading of sucrose into the leaf
veins. The retention of assimilates in the leaves allow a lowering of the osmotic
potential of the leaf cells so that they can more powerfully compete for the available
water in the plant. On the other hand, the abscisic acid moves form the mesophyll into
the guard cells of the leaf epiderm and causes the stomata to close by interference with
the cation exchange of these cells. In this way an early hormonal signal from the roots
exerts a hormonal response within the leaves to minimize the effect of an adverse soil
factor (Mansfield and Wellubrn [19]). Cytokinins, and also such other hormones as
gibberellins and abscisins, and even a precursor like ACC, can act as messengers from
root to shoot (Bruinsina [5]).
However, a complication is that the site of biosynthesis may not be restricted to the
roots. it is well established that gibberellins and abscisins are synthesized in aerial
parts, maybe even predominantly. But even for the cytokinins it is not quite sure that
they are not produced in overground plant parts, at least under particular conditions.
There is no evidence for Ihe production of cytokinins in vegetative or reproductive
buds, in flowers, fruits, or seeds (Bruinsmna [4], Van Staden and Brown [23]), How-
ever, recent experiments with rootless plants show that they can at least maintain their
cytokinin content and their ability to perform lateral shoot growth (e.g. Wang and
Wareing [26]). Rootless sprouts cultured from tobacco pith callus, are able to convert
14C-adenine into isopentenyladenine, zeatin, and their ribosides, which proves that
cytokinin synthesis can occur outside the roots (Chert and Petschow [9]). The cut
flower stalk of Yucca produces phloem sap containing the nucleotides of isopentenyl-
adenine and zeatin. Removal of the rhizome causes a decrease in the nucleotide content
which lasts only a day, but removal of leaves results in a decreased cytokinin content
for at least three days (Vonk [25]). This strongly indicates supply from the leaves
that are known sites of conversion of cytokinins into their nucleolides and glucosides.
Cuttings of woody plants, such as poplar, apple, and grape, often show increased
cytokinin levels during overwintering or cold storage. It is still unknown whether
this cytokinin is synthesized de nova, e.g. in the cambium cells, or released from

120
inactive conjugates still present from their earlier formation from root-sUpplied
cytokinins (Van Stacten and Davey [24]).
It is these problems of distribution, storage, and redistribution of hormones that
complicate many attempts to locate sites of biosynthesis and to elucidate the connected
source-sink relationships.
A nice recent example of a hormone-directed source-sink relationship is the explana-
tion offered by Brun,, et al. [8] for the observation that the photosynthetic activity of
soybean leaves is considerably reduced upon removal of the developing fruits. Pod
removal was also accompanied by closure of stomata and increased abscisic acid levels
in the leaves. It is indicated that the fruits stimulate the photosynthetic activity of the
leaves by their auxin production. The auxin produced by the pods checks the abscisic
acid level in the leaves, thereby allowing the stomata to remain open and the niesophyll
to have a good supply of carbon dioxide for their photosynthetic activity.

4. Exogenous regulators as supplements to the endogenous hormonal pattern

A practical perspective of plant hormone studies is that they hopefully lead to indi-
cations how to direct the growth and nutrient distribution in crop plants by external
applications of growth regulators. One possibility is the treatment with regulators
that interfere with the biosynthesis or translocation of endogenous hormonal sub-
stances, another is the application of regulators to simulate the action of hormones or
to increase their endogenous levels.
An example oft he latter possibility, again with soybean, is the addition of the synthetic
cytokinin, benzyladenine (BA), to plants that have started fruiting (Crosby et aef1. /0]).
Flower buds occuring above the fertilized flowers which are forming pods tend to
abort. This abortion can be prevented by BA application, so that the numbers of
pods per plant can be considerably increased. Because the number of seeds per pod
and the weight per seed are also enhanced by the treatment, it was suggested that the
exogenous BA replenishes an endogenous shortage of cytokinin. caused by the con-
siderable cytokinin intake of the developing pods. Also experiments at our laboratory
with plants of the spider flower (C/eome spinosa) demonstrated that the auxin pro-
duction of developing ovules produces a sink activity of the fruits which by its cyto-
kinin attraction deprives the flower buds above of sulficient cytokinin for their
normal development. This leads to the occurrence of sterile zones on the fruiting
flower stalk (De Jong and Brtsma [13]). In the case of soybean, the endogenous
zeatin content of the flower ovaries varied between different varieties from t50 to
600 ng/g, the varieties with the least cytokinin being the most responsive to the BA
application.
With flower buds of Begonia, too. exogenous cytokinins directly promote flower
development. Experiments with Culture in vitro revealed considerable differences in
effectiveness of the different cytokinins. as is frequently met with in the application of
this class of regulators, BA being the most effective (Berghoef and Brinsma [3j).
Other groups of regulators, however, affected the flower development only indirectly.
Auxins and gibberellins, for example, influenced the number and sex of the flowers
through a change in the relative sink activities of. on the one hand, the inflorescence
primordia and, on the other hand. the apices of main stem and side branches. The
stimulation of vegetative development by these hormones decreased the nutritional

121
 status of the inflorescence primordia in such a way that fewer female flowers could be
formed (Berghoef and Bruinsma [2]).
Application of growth regulators to replenish endogenous hormone contents has been
used successfully to promote flowering of conifers. In sylviculture, selection is directed
towards increased wood production and, since flowering may considerably reduce the
formation of wood, such selection unintentionally involves a reduced fertility. As a
consequence, the seed production of many woody species is very unsatisfactory. In
1958, Kato in Japan discovered that treatment with gibberellic acid, GA 3, strongly
promoted cone production in Cryptomeria japonica. Since then, this gibberellin has
been successfully applied for the stimulation of flowering in many coniferous species.
Trees that take up to seven decades to become fertile can be induced to flower already
in their seedling stage. Originally, these results could only be obtained within the order
of the Cupressales, attempts in the order of the Pinales were unsuccessful, and most
of the economically important species belong to the Pinaceae: Pius, Picea, Abies,
Pseudotsuga, etc. Later studies revealed that these species will respond to other
gibberellins (e.g. GA,,), which are less polar i.e., contain less hydroxyl groups.
Conditions that promote cone formation, e.g. water stress or nitrate fertilization, also
enhance the occurrence of these less polar gibberellins in the needles (Pharis [20]).
It therefore appears that endogenous gibberellins are required first for vegetative
growth and are not available for the process of cone formation, and that the exogenous
addition of certain gibberellins adds to endogenous levels of specific gibberellins to
such an extent that flowering can be induced.
However, the endogenous level of a hormone is not always a reliable indicator of its
physiological importance. This, too, can be illustrated with gibberellins. The main
gibberellin in rice plants is GA, next to which a small amount of GA, also occurs.
Nevertheless it is GA, that stimulates the growth of the rice plant, the large amount of
GA,9 probably being the pool from which the active substance is formed in limited
amounts only (Kurogochi et al. [16]). In genetically dwarfed pea plants, or under
conditions restricting the growth of the pea plant, gibberellin GA 20 accumulates. The
high gibberellin level coincides here, too, with a low growth rate, Under growth-
promoting conditions the GA, 0 level decreases so that there is apparently an inverse
relationship between the amount of gibberellin and the growth rate. Under these
conditions GA,, is converted into GA, which is the active gibberellin in pea stem
elongation (Koehler [15], Sponsel et al. [22]). These examples demonstrate that the
amount of hormone need not correspond with its physiological significance and,
moreover, need not be related to its effect, the relationship may even be inverse.
Turnover rates of these substances may be more relevant than their absolute amounts.

5. Effects of compartmentalization

Turnover rates of hormones can be determined by the addition of the substance

involved containing a radioactive label and following the fate of the label. In these
very advanced studies, however, another problem arises, viz. that of compartmen-
talization. When radioactive gibberellin GA.9 is added to pea stem tissue, it is quite
stable indicating a low rate of turnover. However, endogenous GA, is readily
metabolized into an inactive catabolite. If this different fate of exogenous and endogen-
ous molecules of the same substance is not due to an isotope effect, then it points to

122
a different localization in the tissue, probably within the cells. The endogenous GA,9
being located within a compartment in which it is enzymatically converted and which
is inaccessible for the exogenously applied molecules (Sponsel et al. [22]).
A similar case was studied at our laboratory with the auxin. IAA. Whereas radioactive
IAA, applied to hypocotyl segments of light-grown sunflower seedlings, is basipetally
translocated at a rate of 7 mm h- 1, the endogenous IAA is hardly transported at all,
neither basipetally nor laterally upon unilateral illumination (Brainsma et al. [7]).
It is apparently pooled in a compartment such as the vacuole from which it is hardly
released, Therefore, exogenously applied regutlators may arrive at other sites than
where the corresponding endogenous hormones reside and accordingly their fate and
function may be very different. For example, exogenoLisly applied IAA may move
laterally upon unilateral illunination, suggesting a role of endogenous IAA in photo-
tropic curvature. In the light-grown sunflower hypocotyl, however, the endogenois
IAA is not affected by a phototropic stimulus but the abscisin, xanthoxin, accUtimulates
at the illuminated side causing bending by its growth-inhibiting effect (ruminsma et al.
[6]). Such observations underline the necessity not to rely only oil experiments with
exogenous applications but to perform exact determinations of the endogenous
compounds involved.

6. Conclusion

In summary, the application of plant growth regulators, among other measures, has a
function in the regulation of source and sink activities to direct crop growth and
development. On the one hand, these substances can interfere with the biosynthesis,
translocation. and metabolism of plant hormones. On the other hand, they can replace
or supplement phytohormones when the level of the latter is suboptimal. Accounting
for the effects of these substances requtires the determination of their fate within the
plant and their effect on the level and turnover of the endogenous substances that
constitute the hormonal pattern, Increasing knowledge of these factors may offer
better and new opportunities for the applications of plant growth regulators in the
production of food. flower, fibre and fuel crops.

7. References

I. Adams. D. 0. and Y..te,. S. F.: Ethylene biosynthesis: identification of I-alminocyclo-

propane-I -carboxylic acid as an intermediate in the conversion of methionine to ethvlene.
Proc. Natl. Acad. Sci. U.S.A. 76, pp. 170-174 (1979)
2. Berghoef J. and Bruinsma. J.: Flower development of Begonia f-amconis Liebtnl I. Effects
of growth-regulating substances and environmental conditions on the composition of the
inflorescence. Z. Pflanzenphysiol. 93, pp. 303-315 (1979)
3. Berghoef, J. and Bruinsma, J.: Flower development of Be .o.nafizamoni., Liebin. II. Effects
of nutrition and growth-regulating substances on the growth of Ilower buids in tiro.
Z. Pflanzenphysiol. 93, 345-357 (1979)
4. Bruinsma. J.: Rolle der Cytokinine bei Bliten- und Fruchteentwicklung Z. Vflanzen-
erniihrung Lnd BodenkLnde 140. pp. 15-23 (1977)
5. Bruinsmca. J.: .oot hormones and overground developIent. I: Sc-at, T. K. (ed.), Plant
regulation and world agriculture, pp. 35-47. New York 1979
6. Brtnsma, J., Ira,msseil,J i. and Kneet, I.. Phototropisn as a phenomenon of irhihition,
In: Sk..og. F. (ed.), Plant growth substances. Berlin, Heidelberg. New York, 1980

123
 7. Bruinsina, J., Karssen, C. M., Bensehop, M. and Dori, J. B. van: Hormonal regulation of
phototropism in the ligth-grown sunflower seedling, Helianthus annts L.: immobility
of endogenous indoleacetic acid and inhibition of hypocotyl growth by illuminated
cotyledons. J. Exp. Bot. 26, pp. 411-418 (1975)
8. Brun, W.A., Brenner, M. L., Setter, T. L. and Heitz, M. B.: Hormonal control of source/
sink activity in fruiting soybeans. In: Skoog, F. (ed.), Plant growth substances. Berlin,
Heidelberg, New York, 1980
9. Chen, C.-M, and Petschow, B.: Cytokinin biosynthesis in cultured rootless tobacco plants.
Plant Physiol. 62, pp. 861-865 (1978)
10. Crosby K. E., Aung, L. H. and Buss, G. R.: Hormonal effects on soyabean reproductive
development. In: Skoog, F. (ed.), Plant growth substances. Berlin, Heidelberg, New York,
1980
II. Geuns, J. M. C.: Steroid hormones and plant growth and development. Phytochem. 17,
pp. 1-14 (1978)
12. Goodwin, T. W.: Biosynthesis of terpenoids. Ann. Rev. Plant Physiol. 30, pp. 369-404
(1979)
13. Jong, A. W. de and Bruinsma, J.: Pistil development in Cleome flowers. IV. Effects of
growth-regulating substances on female abortion in Cleonze spinosa Jacq. Z. Pflanzen-
physiol. 73, pp. 152-159 (1974)
14. Jung, J.: Possibilities for optimalization of plant nutrition by new agrochemical substances
- especially in cereals. In: Scott, T. K. (ed.), Plant regulation and world agriculture.
pp. 279-307, New York 1979
15. Koehler, D.: Zur Gibberellinakkumulation bei im Wachstum gehermnten Erbsenkeim-
lingen (Pistan sativun). 2. Pflanzenphysiol. 65, pp. 404-409 (1971)
16. Kurogochi, S., Murofutshi, N., Oia, Y. and Takahashi, N.: Identification of gibberellins
in the rice plant and quantitative changes of gibberellin A, throughout its life cycle.
Planta 146, pp. 185-191 (1979)
17. Lieberinann, M.: Biosynthesis and action of ethylene. Ann. Rev. Plant Physiol. 30, pp.
533-591 (1979)
18. Ltirssen, K., Naumann, K. and Schrlder, R.: -Aminocyclo-propane-l-carboxylic acid -
an intermediate of the ethylene biosynthesis in higher plants. Z. Pflanzenphysiol. 92,
pp. 285-294 (1979)
19. Mansfield, T. A. and Wellburn, A. R.: Abscisic acid and the functioning of stomata. In:
Skoog, F. (ed.), Plant growth substances. Berlin, Heidelberg, New York, 1980
20. Pharis,R, P.: Manipulation of flowering in conifers through the use of plant hormones.
In: Mikshe, J. P. (ed.): Modern methods in forest genetics. pp. 265-282, Berlin-Heidel-
berg 1976
21. Skytt Andersen, A.: Plant growth retardants: present and future use in food production.
hi: Scott, T. K. (ed.), Plant regulation and world agriculture. pp. 251-277, New York, 1979
22. Sponsel, V. M., Kirkwood, P. S. and MacMillan, J.: Metabolism of gibberellins in im-
mature seeds of Pisani sativioan. it: Skoog, F. (ed.), Plant growth substances. Berlin, Hei-
delberg, New York, 1980
23. Staden, J. van and Brown, N. A. C.: Investigations into the possibility that potato buds
synthesize cytokinins. J. exp. Bot. 30, pp. 391-397 (1979)
24. Staden, J. van and Davey, J. E.: The synthesis, transport and metabolism of endogenous
cytokinins. Plant, Cell & Environment 2, pp. 93-106 (1979)
25. Vonk, C, R.: Origin of cytokinins transported in the phloem. Physiol. Plant. 46, pp. 235-
240 (1979)
26. Wang, T. L. and Wareing, P. F.: Cytokinins and apical dominance in Solanum andigena:
lateral shoot growth and endogenous cytokinin levels in the absence of roots. New Phytol.
82, pp. 19-28 (1979)

124
The Role of Plant Hormones in Plant
Adaptation to Stress Conditions
Y. Mizrahi,. Department of Biology, Research and Development Authority,
Ben-G urion University of the Negev, Beer-Sheva/Isracl-

Suti'mmary

It was hypothesized that tunder stress conditions which impair plant growth, the cytokinin
activity decreases, while the activity of growth inhibitors, such as ABA, increases. The hor-
monal changes enable the plant to survive under the new environment conditions.
The concentration of growth inhibitors and cytokinins in plants under various types of stress
and during recovery from them was found to be in agreement with the above hypothesis.
Accordingly, when applied externally prior to the water stress, cytokinin impairs plants'
adaptability, whereas ABA improves it.
This thesis, which is supported by additional physiological and genetic evidence from the
literature. can also explain phenomena which cannot be accounted for by the other adaptation
hypotheses.
Possible agricultural applications of these findings will be discussed.

Stress is defined as an environmental change which induces reduction of plant growth.

Adaptation is defined as changes which occur in the plant in response to stress condi-
tions and which enable the plant to survive and maintain limited growth.
The most important lactor which reduced plant growth in many areas in the world is
water stress. Water stress develops when the water efliux from the plant is greater than
the water influx into the plant. This phenomenon led many scientists to describe the
plant as an aqueduct between the soil and the atmosphere. By defining the plant as an
aqueduct, it became possible to consider conductivities and resistances to water flow.
The driving force of water is considered to be the differences in water potentials
between various plant parts minus all the resistances in this aqueduct.
Water potential is defined as:

,/w=r+ t+ p

where osmotic potential

matrix potential
p- hydrostatic potential= turgor pressure.

* Dr. Y. Xizrahi, Department of Biology, Research and Development Authority, Ben-Gurion

University of the Negev. Beer-Sheva/Israel

125
Scientists used to consider a decrease in turgor pressure as the main cause of the
reduction in growth under water stress conditions. However, under saline drought
conditions, following an adaptive period of time, plant turgor can be equal to or even
greater than that of non-stressed plants, and still plant growth is inibited. These and
other unexplained phenomena led Itai and Vaadia [3] in the early sixties to hypothesize
that under root stress the amount of shoot cytokinins declines. Later, we developed
the hypothesis that under stress conditions which impair plant growth, cytokinin
activity decreases and the activity of growth inhibitors such as ABA increases. This
hormonal change is part of the adaptation process which enables the plant to survive
under the new environmental conditions.

The evidence to support this thesis is as follows:

1. EndogenoUs cytokinin levels decline under saline stress conditions. The amount of
cytokinins decreases with increasing salinity of the root medium [3].

Hi.Flaccid leaves from osmotically stressed plants, when removed to distilled water,
regain their turgor. The Iranspiration rate of such leaves is lower than that of control
leaves which have been removed from non-stressed plants. Pre-treatment with ABA
and kinetin decreases and increases, respectively, the transpiration rate of such leaves
([6j; Figure 1).

Il1. Endogenous analysis of ABA in stressed leaves shows that under stress conditions
ABA concentration increases and that after recovery it declines to pre-stress levels
([8]; Figure 2). The ABA concentration increases initially and after an adaptation
period declines, but does not reach the level of control leaves (Figure 3).
When analyses were made with bioassay systems it was found that growth inhibition
under stress conditions can not be explained by ABA increase alone. This is because
the extraction equivalent of 50 mg of leaves did not contain half of the ABA in the
extraction equivalent of 100 mg of leaves when the basis for calculation was an ABA
response curve. This means that at least one other ABA-like compound with a different
response curve increases in leaves upon exposure to stress conditions (Table 1). The
results of Loveys and Kreidnian [5] confirm these findings.

IV. Genetic evidence was published [2, 4] which demonstrated the involvement of
ABA in the adaptation of tomatoes to water stress. These authors are working on a
wilty mutant of tomato which is unable to close its stomata even when the cells are
in plasmolysis. Leaves of this mutant contain only one-tenth of the ABA which exists
in the normal leaf, JIber and Tal [2, 14] showed that application of ABA can reverse
the phenotype of the wilty mutant to a normal one with normal stomatal closure.

V. It has been found that NaCl-stressed plants can withstand water stress induced by
eliminating aeration from the nutrient solution ([8]; Figure 4). A precise experiment
was done by measuring the water saturation deficit (W.S.D.) as a parameter of the
plant's water status ([8]; Figure 5). Tobacco plants were exposed to 6 g/I NaCI at
successive time intervals, so that on the day of the experiment we had plants which
had been exposed to salinity for various times. Water stress was induced by removing

126
 .4 NON-STRESS STRESS
t.2

0
10
w
E
108

o.6

S0A
2-

02

0 20 40 60 80 100 0 20 40 60 80 100
Tme, minutes

Fig. I. The effect of1osmotic root stress, ABA and kinetin on the transpiral ion rate of detached
tobacco leave (from ref. [6 3)

140 S

120,
0 0
1' -
" 80 "" "
812
0 -o \SR
7 6 1 NS
A. 40 " "

' 20
0
0 48 96
Time hours
kig.2. The effect Of NaCl in the root niediUnrl on leaf ABA content of tobacco plants. S,
0.0 g/I NaCI NS. non-stressed planis; SR. stress for 48 hours followed by recovery in Hoag-
land solution (from ref. :8.)

127
 500

t 40 0 , ,
- #

E /
N 300 /
E
200 - S-
C)Ca

100- NS

m 100
I 2 4 6 9 12

DAYS
Fig.3. Change in ABA content of tobacco leaves on exposure to NaCl. S, 6.0 g/Il NaCI; NS,
non-stressed plants (from ref. :8

Fig.4. Appearance of tobacco plant after 3 hours under water stress induced by lack of root
aeration. S, 6 g/l NaCI in the root medium; NS non-stressed plants (from ref. 8)

128
 24

20

-------- -----------------

S12
._o

0 4

Days in 4.5 otm NoCI

Fig.5. Effect of lack of root aeration and li me of exposure to salinity (6 g/l NaCI) oil the water
saturation deficit of tobacco leaves (from ref.[8j), Broken line, control value after 3 h under
water stress circles, values after 3 h under water stress: black dots. values before the exposure
to water stress, Water stress was induced by lack of root aeration.

Table I. Endognous ABA concentrations from stressed and non-stressed leaves as calculated
from the wheat coleoptile elongation test, using ABA (in nanograms per tube) for the response
curve
Treatment Pubes containing Tubes containing
extract equivalent to extract equivalent to
100 mg leaf dry weight 50 mg leaf dry weight
C ontrol ....... .................. 1.3 0.026
Stress ................. ..... 20.0 2.6

the air which had been bubbling through the nutrient solution. Three hours later
the W.S.D. of the leaves was measured. As shown in Figure 5. the W.S.D. values
declined with increasing length of exposure to NaCI. reaching a plateau after 4 days,
which in our opinion is the time required by the plant to adapt to NaCI stress. Plant
hardiness to water stress according to these measurements is in accord with endogenous
ABA (Figure 3). which reached a peak on the fourth day of exposure to salinity.

VI. If ABA increase and cytokinin decrease are adaptive responses to water stress, one
would expect ABA to serve as an anti-stress agent while cytokinins would alleviate
stress symptoms. Figure 6 demonstrates this very clearly. Tobacco plants treated with
ABA and kinetin and then exposed to water stress by eliminating aeration from the
nutrient solution, showed that ABA was better and kinetin was worse than the control.

129
VII. The question arises whether these hormonal changes are associated only with
water stress. In an attempt to answer this question, tobacco plants were exposed to
6 g/l NaCl in the root medium under high (95-100%) and low (50-75%) relative
humidity. Under high relative humidity there was no change in leaf ABA content,
even though the plant was exposed to NaCI ([7]; Figure 7). Plants were exposed to
mineral deprivation by transferring them from half-strength Hoagland solution to
distilled water [9]. Figure 8 shows that leaf ABA content increases with time in
mineral-deprived plants and declines when plants recover in Hoagland solution. The
mineral-deprived plants withstood water stress better than controls, as a result of their
elevated internal ABA content. Figure 9 presents the results of leaf W.S.D. of mineral-
deprived plants after their exposure to water stress induced by lack of root aeration,
as done with NaCI-stressed plants.
This means that plants with totally different turgor values which have not been exposed
to water stress will react the same way as water-stressed plants; the explanation for
their behavior is the ABA content. Other stresses, such as boron toxicity and chilling,
will increase the ABA content of the leaves and reduce transpiration ([1]; Table 2).
These stresses also induce hardening against water stress as well as against chilling
injury, as measured by relative leakage. This is in correlation with endogenous ABA
levels ([I]: Table 3). It was also found that growth reduction caused in cucumber
seedlings by chilling stress can be antagonized by ABA treatment or other stress which
increases the ABA concentration in the leaves ([12, 13]; Figure 10). Our scheme is
summarized in Figure 11.

NP
Fig.6. Effect of pretreatment with hormones on tobacco plant turgor under water stress
induced by lack of root aeration. Left, control plant; middle, ABA treatment; right, kinetin
treatment.

130
 NS "HIGH" S

90
80
70
60 69 71
50
W 40
t o30
20

to
100

50
so I

70-
20
50

,
875 8.75
TIME (ini)
Fig. 7. Effect of salinity and air relative humidity on the ABA concentration of tobacco leaves.
The arrows designate the ABA peak in gas chrotnaiograph according to a 3H electron-capture
detector. The numbers are picograms of ABA per tng leaf dry weight. '-igh', high relative
humidity (98%); 'Low', low relative humidity: NS, no stress applied (without NaCI); S,
stress applied (6 g/Il NaCI): * diluted 1/10 (from ref. [7])

140 I

120

E l00
E 80"

4 0
in
< H
20 -

0D 2 4 7

DAYS
Fig,8. Changes in the amount of cis-trans ABA extracted from the leaves of plants grown in
distilled water (W), plants grown in half-strength Hoagland solution (H); and plants grown
in distilled water for 4 days and then transe rred [or 3 days to half-strength Hoagland solttion
(W-H), the recovery treatment (frorn rcf. [9J)

131
 0

1C

I 2 47
DAYS

Fig.9. Effect of 3-hour period of cessation in root aeration, on the water saturation deficit of
leaves of plants kept for increasing periods in distilled water. W, H, and W-H as for Figure 8
(from ref. /91)

0 .l>o 0~

0-10 Id'le
A BA (MI

Fig. 10. The appearance of cucumber cotyledons 6 days after seedlings were exposed to chilling,
seedlings were exposed for 24 h to ABA or NaCI in the root medium. Chilling stress consisted
of 20C, 80% R.H., in the dark for 24 h.
A, cotyledons of non-chilled seedlings; B, cotyledons of chilled seedlings (from ref. J32)

132
 ABA
EROGENOUS

WAEERSTRESS WATERS?cRTAGC
OSMOTPC STRESSES INCREASED ENOOGENOiS i,proved iNCREASED OSMOTICUM
LACKOF ROOTAEAIATION
- CONTENT OF LACKOF ROOTAERATION
MINERAL OEPRRATION- A B A LOW ROOT TEMP.
EXPOSURE TO LOW TEMP - SUBFREEZING TEMP

Fie. II. Scheme of the involvement of ABA in various stresses.

Tahle 2. Abscisic acid content and extent of transpiration (as percentages of control) of
tobacco leaves exposed to various stresses

Treatment Initial stress ABA content* Transpiration*

I Moderate leaf dehydration 360 46
2 Severe leaf dehydration 660 29
3 Short mineral deprivation 260 90
4 Long mineral deprivation 400 53
5 BO33 - toxicity 200 63
6 Salination 450 56
* percent of the appropriate control. Values of controls were 49, 59, 65 x 10- 2 g mg-I dry
weight foitreatments 1,2,and 6, respectively; and 20, 25, 35 x 10- -g mg- dry weight for
treatments 3,4. and 5,respectively (from Ref. ,'.)
** Percent of the appropriate control. Values of the controls ranged from 9 to 10.5 x 10-3 ml
hr-1 cnI -

Table 3. Response ol tobacco leaves to bCtow-frCzing temperature and reduced oxygen in

the root medimn after prior exposure to various initial stresses
Treatments I and 2: detached leaves of tobacco plants water saturation deficit (WSD) levels
of 15 and 35% (moderate and severe leaf-dehydration, respectively); treatments 3 and 4:
leaves of tobacco plants deprived of minerals for 3 and 7 days (short and long mineral depri-
vation, respectively); teatment 5: leaves of tobacco plants grown in half-strength Hoagland
nutrient solution to which was added 290 ug/I of HJ1O,; treatment 6: leaves of tobacco plants
grown in half-strength Hoagland nutrient solution to which was added 6 g NaCI/I (from
Ref/I),

Treatment No. Initial stress Resistance to subsequent stress of

below-freezing reduced 0, in
temperature root meditm
Relative leakage- Leaf WSD**
I Moderate leaf dehydration 80
2 Severe leaf dehydration 47 -
3 Short mineral deprivation 85 75
4 Long mineral deprivation 60 50
5 BO-toxicity 74 60
6 Salination 42 53
Percent of the appropriate control. Relative leakage of the controls ranged from a ratio
of 0,50 to 0.63.
** Percent of the appropriate control i.e.. plants fron which root aeration was withheld but
which were not pre-exposed to an initial stress. The WSD of the control treatments ranged
from 17 to 23%.

133
Possible agricultural uses of ABA and cytokinins

ABA can be used as an antitranspirant if the price becomes reasonable. Figure 12 shows
that under drought conditions, ABA-treated barley plants grew better than controls,
and cytokinin-treated plants were the worst. Under severe drought conditions only
ABA-treated plants survived and produced grains ([10]; Figure 13).
ABA can be used to prevent damage due to transfer of plants from the nursery to the
field. In cases where such transfer is followed by water stress, ABA application can
help the plants. In Indiana, it was the practice to grow tomato seedlings in greenhouses
free from frost danger and then transplant them in the field. Some damage caused by
water stress was a major problem. Plants treated with ABA withstood transplanting
much better than untreated ones, and produced uniform stands in the field with
undamaged tomato seedlings. Since ABA is so expensive, one can increase its endogen-
ous concentration by exposing plants (in the nursery, cuttings under mist propagation,
plantlets produced by tissue culture, etc.) to mild stress-like mineral deprivation, low
temperature, etc., which will be of benefit when transferring these plants to the field.
Some good farmers who run nurseries used to say that it is good practice to avoid
fertilization and irrigation of beds before transferring the plants, in order to produce
uniform and good stands in the field.
Larque-Saavedra and Wain [4] found that corn and sorghum varieties which can
withstand water stress increased their ABA content under water shortage even more
than regular varieties. They suggested measurement of ABA concentration as a tool
for selecting drought-tolerant lines.

Fig. 12. Effect of kinetin (kin) and ABA on barley plants under drought conditions.
(Cont = control plant) (from ref. [10])

134
Fig.13. Effect of ADA-sprays on non-watered wheat seedlings when direct loss of water from
the soil was prevented. C, control (water spray without ABA): A 20, 20 ppm ABA; A 100,
100 ppm ABA. Photograph taken 70 days after sowing (from ref. [10])

Cytokinins can be used for speeding up the water loss of hay. This may be of value
in areas where early rain might spoil the hay if drying is not rapid.
In closed greenhouses, where the relative humidity is very high, plants could be grown
with saline water which can not be used for outdoor irrigation. O'Leary in Arizona
has shown that it is possible to overcome lethal doses of NaCI by increasing the
relative humidity [11].
An inexpensive analog of ABA, if available, could be used to prevent injury caused
by chilling or below-freezing temperatures. This would require sufficient warning before
temperatures decline to the dangerous range.

References

1. Boussiba, S., Rikin, A. and Richmond, A.E.: The role of abscisic acid in cross adaptation
of tobacco plants. Pl. Physiol. 56, 337-339 (1975)
2. Imber, D. and Tal, M.: Phenotypic reversion of flacca, a wilty mutant of tomato, by
abscisic acid. Science, N.Y. 169, 592-593 (1970)
3. hai, C. and Vdadia, Y.: Kinetin-like activity in root exudate of water-stressed sunflower
plants. Physiologia Pl. 18: 941-949 (1965)
4. Larque-Saavedra and Wain, R.L.: Studies on plant growth-regulating substances, XLI I.
Abscisic acid as a genetic character related to drought tolerance. Ann. appl. Biol. 83,
291 297 (1976)
5. Loveys, B.R. and Kreidman, P.E.: Internal control of stomatal physiology and photo-
synthesis. 1. Stomatal regulation and associated changes in endogenous levels of abscisic
acid and phaseic acid. Aust. J. Pl, Physiol. 1, 407415 (1974)

135
 6. Mizrahi, Y., Blumenfeld, A. and Richmond, A.E.: Abscisic acid and transpiration in leaves
in relation to osmotic root stress. Pl. Physiol. 46, 169-171 (1970)
7. Mizrahi, Y., Blumenfeld, A. and Richmond, A.E.: Abscisic acid and cytokinin contents of
leaves in relation to salinity and relative humidity. P1. Physiol. 48, 752-755 (1971)
8. Mizrahi, Y., Blumenfeld, A. and Richinond, A. E.: The role of abscisic acid and salination
in the adaptive response of plant, to reduced root aeration, Pl. Cell Physiol. 13, 15-21
(1972)
9. Mizrahi, Y. and Richmond, A.E.: Abscisic acid and mineral deprivation. Pl. Physiol. 50,
667-670 (1972)
10. Mizrahi, Y., Scherings, S.G., Arad, S. and Richmond, A.E.: Aspects of the effect of ABA
on the water status of barley and wheat seedlings. Physiologia P1. 31, 44-50 (1973)
II. O'Leary, J. W.: High humidity overcomes lethal levels of salinity in hydroponically
grown salt-sensitive plants. PI. Soil 42, 717-721 (1975)
12. Rikin, A., Blumenfeld, A. and Richmond, A.E,: Chilling resistance as affected by stressing
environments and abscisic acid. Bot. Gaz. 137, 307-312 (1976)
13. Rikin, A. and Richmond, A.E.: Amelioration of chilling injuries in cucumber seedlings by
abscisic acid. Physiologia Pl. 38, 95-97 (1976)
14. Tal, M. and Imber, D.: Abnormal stomatal behaviour and hormonal imbalance inflacca,
a wilty mutant of tomato. II. Auxin and abscisic-like activity. Pl. Physiol. 46, 373-376
(1970)

136
Synthetic Growth Regulators: Mode of
Action and Application in Fruit Production
J. Tro.p and S. J, Wertheim. Research Station for Fruit GJrowing,
Wilhelminadorp/'Fhe Nethcrlands*

Summal- Y

This communication deals with the application of growth regulators in the modern production
of apples and pears and discusses, in general terms, the mode of action of these agents.
The use of newly developed chemicals such as M&B 25,105, which induce lateral shoot
formation in ntursery trees may advance cropping one year.
In view of the necessity to keep a balance between lower-bud formation, shoot growth, and
fruit set. special attention is given to: the reduction of flowering by gibberellins, the promotion
of flowering by growth retardants such as SADH and CCC, flower and fruit thinning by the
ethylene-releasing agent CEPA. the auxin-like substances NAA and NAAm. and the insecticide
carbaryl, the promotion of partlicriocarpic fruit set in pear by gibberellins. and the promotion
of fruit set by growth retardants. via reduction of shoot-growth competition.
Finally. the prevention of pre-harvest fruit drop by NAA, the acceleration of ripening by
CEPA. and the reductive effect of GA,+, on skin russeting in apple fruits are discussed.

1. Introduction

In modern intensive fruit-growing, small, closely planted trees have replaced the
large, widely spaced, trees used formerly. At the same time the economic life span of an
orchard has been reduced from more than 50 years to barely 10-15 years for apple and
15-20 years for pear. In these intensive systems early fruit-bearing is essential, espe-
cially to balance the natural habit of vigorous growth in young fruit trees and also to
give an early return on the high financial investmends that fruit-growing necessitates
today.
The maintenance of a balance is also of crucial importance in later years of orchard
life. To prevent biennial bearing, the delicate balance between fruit set, flower-bud
formation and shoot growth must be controlled adequately. Especially with respect to
economic considerations the alternation of light crops oF usually large fruits with poor
storage quality and heavy crops with too many undersized fruits is highly undesirable.
To control fruit cropping and vegetative growth, a number of cUhural techniques are
available sttch as prLining, tying clown of branches to reduce growth and favott
flowering, fruit thinning by hand and. of course, the use of dwarfing rootstocks.
Nitrogen fertilization may stimulate either growth or flowering, the decisive factor

* Dr. J. Totp, and Dr. S.J.IVertheim. Research Station for Fruit Growing. Brugstraa 51.
NL-4475 AN Wilhelminadorpifhe Netherlands

137
being the time of application. Furthermore, supply of water during the first spring after
planting - the critical phase in tree life - has been found to favour tree development
and, consequently, early bearing.
Since the discovery of plant hormones in the late twenties it has become clear that
these substances play a key role in the integration and coordination of growth and
differentiation in the various plant parts. Furthermore, because plant response to
environmental conditions and cultural measures is mediated via small amounts of these
chemical substances, it is not surprising that attempts have long been made to regulate
plant behaviour by chemical means other than fertilizers. Use has been made of
natural hormones, e.g. IAA (indoleacetic acid) and various gibberellins (mostly GA 3
or GA4 +7 ), man-made substances that behave like hormones, e.g. NAA (l-naphta-
leneacetic acid) and CEPA [(2-chloroethyl) phosphonic acid], and man-made chemicals
capable of affecting the production or translocation of hormones, e.g. the growth
retardants MH (maleic hydrazide), TIBA (2,3,5-triiodobenzoic acid), SADH (succinic
acid 2,2-dimethylhydrazide, Alar), and chlormequat [(2-chloroethyl) trimethylam-
moniumchloride, CCC].
Although much research has been done on the effect of growth regulators on the
growth and development of fruit trees, relatively few applications have found extensive
use in commercial fruit-growing. Nevertheless, growth regulators have proven to be
very helpful in the solution of specific problems. In the present communication a
number of applications will be discussed, mainly on the basis of data obtained in work
done on apple and pear in our Research Station. For a recent review on the use of
growth regulators in citriculture, the reader is referred to Monselise[28]. Forpomology,
a short historical review has recently been published by Martin [26].

2. Early cropping

2.1 Branching in nursery trees

For modern fruit-growing, nursery trees having an adequate number of lateral shoots
(feathers') that are not inserted too low and make a wide angle with the main stem
are greatly preferred to unfeathered trees. Feathers form the basis for the primary
branch system, and they provide room for early production of flower buds. Properly
feathered trees may start to bear fruit a year earlier. Van Oosten [42] found a clear
positive relationship between the number of feathers at planting time and total yield
from the second and third years. The natural ability for branching differs between
cultivars. Some cultivars produce lateral shoots in the nursery readily, but others
have to be stimulated to do so.
The failure to form lateral shoots is an example of correlative inhibition mediated by
auxin produced by the young growing leaves in the apical bud (Phillips [30]). Thus,
removal of the site of auxin production by shoot tipping (Werthein [46], see also
Tables I and 2) or blocking of the translocation of auxin by TIBA stimulates feathering.
However, besides being too laborious for practical use, shoot-tip removal usually
induces the outgrowth of relatively few shoots that grow too steeply. In contrast, as
Table 1 shows, the removal of young expanding leaves, with the actual growing point
left intact, results in a sufficient number of wide-angled shoots. Obviously, apical
dominance should only be reduced temporarily; the resumption of auxin supply after

138
recovery of the still growing shoot tip is needed to stimulate widening of the crotch
angles, as Jankiewicz and Kozvra [13] have shown for apple.
In attemps made in apple and pear to induce branching chemically, tests were per-
formed with a variety of systemic growth regulators all having the ability, depending
on the concentration used, to damage young tissues without killing the apex. They
include Off-Shoot-O (Quinlant [31]), a mixture of methyl esters of fatty acids used
in a number of plant species to promote branching (Cathey el al. [4]), MH (Jankie-
wicz and Kozyra [13], Baldini et al, [3]), and diphenylurea (Baldini et al. [3]). In the
last few years a few newly developed agents have become available, i.e.. M&B 25,105
(propyl 3-t-butylphenoxyacetaie), NC 9634 [(3-phenyl-l,2,4-thiadiazole-5-yl) thioacetic
acid], and PP 528 (active ingredient not yet released) (Quinlan [31, 32], Wertheim
[45, 46]).
A few data are given in Tables I and 2. 'Fable t clearly shows that leaf removal
(especially when done twice) and M&B 25,105 stimulated the formation of widely
branched feathers, whereas tipping, i.e., complete removal of the apex, resulted in
very steep lateral shoots. In all treatments an appreciable number of shoots arose lower
than 30-40 cm above the graft union. For practical reasons such shoots are not
desired, and since they may also inhibit outgrowth of buds inserted higher up, another

Table 1. Formation of lateral shoots (> 10 cm) in nursery trees of apple cv. Red Boskoop and
Benoni in 1977 as affected by shoot tipping, manual removal of young leaves, and application
of M&B 25.105. Dates of treatment: 24 June (R.B) and 5 July (B). Second leaf-removal 5 July
(R.B) and I I July (B). Concentration of M&B 25,105: 1500 ppm active ingredient. From
Wertleun [46].
Treatment Red Boskoop Bcnoni
Total > 40 cm Crotch Total > 40 cm Crotch
above angles above angles
union too small union too small
Untreated 4.7 3.2 0.0 2.A 1.3 0.0
Shoot-tipping 6.8 3.7 1.2 5.0 4.1 2.3
Leaf-removal (once) 6.2 4.0 0.1 3.8 2.9 0.2
Leaf-removal (twice) 7.6 5.5 0.2 4.5 3.8 0.4
M&13 25,105 11.0 6.1 .3 6.6 5.6 1.7

Table 2. Formation of lateral shoots (> 10 cm) in nursery trees of apple cv. Winston in 1978
as affected by various treatments. Concentration of Amex-A820: 14(00 ppm active ingredient
and of M&B 25,105: 1500 ppm active ingredient.

Treatment Total > 30 cm above union

L U ntreated ...................... ....... 7.0 0.7
2. Manual removal of low shoots (21 June) . 2.7 2.3
3. Amcx-A820 (21 June) .................... 2.9 1.9
4. Manual leaf removal (17 and 27 July) ... 5.6 0.6
5. M & B 25,105 (17 July) ................... 6.2 0.7
2+4 (21 June and 17 July resp.) ............ 2.8 2.4
2 1-5 (21 June and 17 July resp.) ............ 4.3 3.6
3 1-4 (21 June and 17 July resp.) ............ 3.6 1.8
3 + 5 (21 June and 17 July resp.) ......... 5.4 3.6

139
experiment was performed in which the unwanted feathers were removed by hand or
their growth inhibited by the growth-controlling agent Amex-A820 (N-sec-butyl-4-tert-
butyl-2,6-dinitro anilin). In some cases these treatments were combined with removal of
young expanding leaves by hand or chemically. As Table 2 shows, prevention of
feathering in the lower part of the stem stimulated bud outgrowth in higher regions,
especially where lateral shoot production was stimulated by M&B 25,105.
These two experiments were done with different cultivars, which might explain the
marked difference in response to treatments between them. For instance, M&B 25,105
and leaf removal even failed completely to induce feathers in cv. Winston in 1978.
Another relevant factor could be growth vigour, which even for the same cultivar can
vary appreciably in successive years since it is dependent on cultural and environ-
mental conditions. Quinlan [31] found that the apple cv. Spartan budded on the
dwarfing roolstock M.9 formed distinctly fewer 'natural' lateral shoots in the nursery
than it did when budded on the vigorous rootstock MM.106. Although both responded
positively to various branching agents, the difference between rootstocks persisted.
The question of whether such agents act by lowering the auxin level in the shoot tip
was only studied for NC 9634. For about two weeks after the treatment, the shoot
tips contained markedly less auxin-like substances than did the untreated controls
(Abbas [1]).
It is conceivable that the level of endogenous cytokinins determines the response to
the various chemicals that influence branching. Deficiency of cytokinin is believed to
be a factor in the release of buds from apical dominance (Phillips [30]). Multiple
sprays of the cytokinin 6-benzylaminopurine stimulated bud-outgrowth in apple
(Kender and Carpenter [15]).

2.2 Control of growth vigour

In sonic cultivars of apple and pear, especially when not budded on dwarfing root-
stocks, vegetative growth is so pronounced in the first few years after planting that
cropping is delayed. Although tying down of branches may help to restrict growth
and stimulate flowering, the use of growth retardants, especially SADH and chlor-
mequat, has proven to be a reasonably good alternative. Both of these retardants
inhibit internode extension and, to a somewhat lower degree, node formation, thus
giving the tree a compact shape. The SADH treatment leads to a typical thickening of
the growing shoot. SADH is usually used in apple, but in pear chlorequat is preferred.
Due to the undesirable side-effect of reduced fruit size and poor keeping quality of the
fruits of cultivars such as Cox's Orange Pippin, the use of retardants is avoided in
more mature orchards where growth vigour is not usually a major problem.
There is a wealth of information available on the more practical aspects of the use of
growth retardants in fruit-growing, including not only SADH and chlormequat but
also TIBA and the ethylene releasing agent CEPA (Williams [48], Li7dders and Debor
[24, 25]). However, the action of these compounds is not fully understood. Since
gibberellins are known to stimulate growth and inhibit flowering, there is a strong
feeling that the various agents interfere at some point in the synthesis or translocation
of gibberellins. The relationship between growth and flowering will be discussed in
more detail in the next section.

140
3. Flower-bud formation, fruit set, and shoot growth

In fruit trees these three phenomena are closely interrelated, and any attempt to affect
one of them without affecting the others has hardly any chance of success. However,
for convenience the following considerations are dealt with in separate sections.

3.1 Flower-ud finrmation and shoot growth

The concept of an antagonism between vegetative growth and flowering is widely

accepted not only for fruit trees (Davis [6]) but also in a more general sense in plants
(Lang [17]). Since young developing leaves in the shoot tip are rich sources of
gibberellins and the inhibiting effect of applied gibberellins has been demonstrated
in many experiments, it seems obvious to ascribe this antagonism to the action of
gibberellins.
As Lwcktill / 22] pointed out, the first irreversible steps leading to flower initiation in
an apple bud are bound to a critical number of about 20 nodes per bud. in addition
there is a requirement for a certain minimum duration of the plastochron (the interval
between the formation of successive primordia) amounting to about 7 days (FidJdrd
[8]). and the presence of bracts (Fijbird [9]). This means that if a bud contains say
6 nodes at the beginning of the season (a normal situa ion), the actual flower initiation
will not start before the first half of August. "The inhibitory effect of growth on
flowering is considered to be an apical dominance phenomenon, the gibberellins
produced in the young shoot-top leaves inducing a prolongation of the plastochron.
It this occurs, the 20-node stage is reached later in the season when the internal and
environmental conditions may be less appropriate for flower initiation.
In terms of this conception it is not surprising that application of gibberellins to reduce
flowering is most effective early in the season, i.e.. long before the actual flower
initiation occurs. This was clearly shown in an experiment on two-year-old potted
apple trees cv. Golden Delicious that were sprayed 3 times with either GA., or GA 4 + 7.
GA, sprayed at full bloom or 2 weeks later reduced flowering on spurs and shoots
(Table 3), but had little efftet when sprayed 4 weeks alter fullbloom. The effect of
GA 4., 7 was more marked, and even the response to the latest treatnent was appre-

Table 3. Shoot growth (shoots I cm) and flowering on spurs (expressed as percentage of
10
flower Clusters arising from the total number of floral and vegetative buds) in two-year-old
apple trees cv, Golden Delicious as atfected by the time of application of GA, and GA 1 4 7
(concentration: 800 ppm active ingredient). Pot experiment. Froim Tromp/ 41 1.

Treatment Mcan shoot Mean / of flower clusters/tree

glowth/trce irrespective if no. o well-developed
(cm) of qUalitV Jlowers/clusters >.2
U ntreated ....................... 55.9 74.5 44.9
GA, at fullbloom ................ 964 44.2 31.8
GA 3 3 weeks after full bloom ...... 85.9 50.6 30.6
GA 3 4 weeks after full bloom ...... 70.9 64.8 46.7
GA,-, at full bloom .............. 98.6 29.5 25.7
GA,-- 2 weeks after full bloom 85.7 26.5 20.0
GA,17 4 weeks aflter full bloom ..... 102.1 52.8 35.2

141
ciable. Cluster quality, here defined according to the number of well-developed flowers
per cluster, was rather poor but seemed to be improved by the gibberellin treatments.
Obviously, after such treatments flowers are initiated only in the best buds. The shoot
growth data suggest a stimulative effect of the gibberellin sprays. In work done by
Tromp [40], GA, applied just after shoot-growth cessation did not affect flowering in
young apple trees. Therefore, the conclusion that gibberellins are hardly or not at
all involved in the actual formation of flowers seems justified (see also Table 5).
In section 2.2 reference is made to the chemicals SADH, chlormequat, TIBA, and
CEPA as agents that inhibit growth and promote flowering. Data obtained in two
simple pot experiments with apple are given in Tables 4 and 5, from which it is clear
that when SADH is applied 'early' (about 3 weeks after full bloom) flowering abund-
ancy is increased in the following year and growth vigour is reduced. Table 4 also
shows that when the early SADH spray is combined with an application of GA 4 +7, the
promotive effect of the former is completely nullified. A 'late' (shortly after shoot
growth cessation, around the end of July) GA, 7 spray was much less effective in this
respect. As comparison of an 'early' with a 'late' SADH sprays shows (Table 5), SADH
affects flowering irrespective of time of application, which suggests that it may affect
flower-bud formation directly, and also that the question of whether SADH is an
anti-gibberellin is not relevant for its effect on flowering. Nevertheless, SADH and
other growth retardants may interfere with the synthesis of gibberellins. For instance.

Table 4. Shoot growth (>5 cm) and flowering on shoots (expressed as percentage of flower
clusters airising from the total number of buds) in one-year-old apple trees cv. Schone von
Boskoop as affected by SADH and GA 4 . 7 (concentration: 1800 and 800 ppm active ingredient,
respectively) applied either 'early' (end of May) or 'late' (end of July, shortly after shoot
growth cessation). Pot experiment.
Treatment Mean shoot Mean % of flower
growth/tree (cm) clusters/tree
U ntreated ................................... 86.2 56.9
No SAD14; GA 'early' ........................ 107.2 32.9
SADH 'early'; no GA ........................ 58.8 84.8
SADH 'early'; GA 'early' . .................... 84.5 37.3
SADH 'early'; GA 'late' ...................... 72.5 60.5

Table 5. Shoot growth (shoots >5 cm) and flowering on shoots (expressed as percentage of
flower clusters arising from the total number of buds) in one-year-old apple trees cv. Cox's
Orange Pippin as affected by SADH (concentration: 1800 ppm active ingredient) applied
either 'early' (beginning of June) or 'late' (mid-July, shortly after shoot growth cessation). Pot
experiment. From Tromp [41].
Treatment Mean shoot Mean % of flower clusters/tree
growth/tree irrespective if no. of well-developed
(cm) of quality flowers/cluster >2
Untreated ......................... 140.3 57.9 34.1
SADI4 'early' ...................... 78.6 71.8 59.1
SADH 'late' . ....................... 131.0 70.9 53.1

142
Rya o and Sansavini [35] found a reduction of diffusible gibberellin in apices of
upright shoots of sweet cherry sprayed with SADH. Similarly, Hoad and Monselise
[12] showed a reduction of gibberellin-like substances in the shoot tips of SADH-
treated apple trees. As has been found for a variety of plant systems (Ninnemann et al.
[29], Reid and Carr [34], Zeevaart [49]), chlormequat is also able to lower ihe
internal gibberellin level. Seen in relation to the role of cytokinins in flower-bud
formation (Luckwill [21]), the finding that chlormequat raises the cytokinin level in
the bleeding sap of Vitis vinifera is particularly interesting (Skete [38]).

3.2 Flower-bud formation and fruit set

If no corrective steps are taken, high fruit set may lead to biennial bearing, this
phenomenon being characterized by an alternation of light crops and superfluous
flower-bud formation with heavy crops and poor production of flower buds in
successive years. Therefore, correction of fruit load either by improving fruit set or by
fruit thinning is essential, especially in cultivars showing a strong natural tendency for
biennial bearing. With respect to fruit thinning, it should be remarked that in more
regularly bearing cultivars such asGolden Delicious (at least in cool northern regions),
fruit thinning can also be strongly recommended to increase fruit size.
Due to hormonal stimuli arising mainly from the fertilization process the levels of
auxin, gibberellin, and cytokinin in the young fruit show a distinct increase, although
not all to the same extent. The hormones are thought to create 'sinks', i.e., are able
to attract nutrients needed for fruit growth. tf, due to poor fertilization, there is a
lag in the rise of the hormone level, competition for nutrients with other sinks such
as fruits and growing shoots is unsuccessful, and fruit abscission occurs. With respect
to the inhibitory effect of the presence of fruits on flower-bud formation, there is
ample evidence that gibberellin produced in the developing fruits or. more accurately,
in the developing seeds (Chan and Cain [5]), is the causal factor. Luckwill [21] found
for the strongly biennial apple cv. Emneth Early that the time when the fruits become
inhibitory to flower-bud formation (4-5 weeks after full bloom) corresponds with the
time when the gibberellin level starts to increase. Furthermore, [load [11] recently
found in difusates collected from apple fruitlets prior to the period in which the
actual flower initiation normally occurs, that there were more hormones, including
gibberellin-like substances, in the biennial cultivars than in those fruiting more
regularly. He also found that the amount of gibberellin in the diflusates increased with
the number of seeds. This may explain why in certain strongly biennial cuItivars,
which are sell-fertile and consequently have fruits with many seeds, the thinning of
young fruilets is too late to prevent inhibition of flower-bud formation. Such Cutivars
require thinning of flowers. However. even when the seed content is not parlicularly
high, to overcome biennial bearing fruit thinning should be clone within about 4 weeks
alter fill bloom (see Table 3).
Thinning of fruits or flowers with chemicals is a method widely used in commercial
fruit production. Numerous chemicals have been tested in countless experiments done
in all fruit-growing areas of the world. However, most of these agents were either
ineffective or showed unwanted side-effects, and at least for commercial use in Western
Europe only a few are left, i.e., CEPA for flower thinning and for fruit thinning NAA
or its amide (NAAm) and especially the insecticide carbaryl (1-naphlalene-N-mcthyl-
carbaniate).

143
In practice the flower-thinning effect of CEPA is used in, for instance, the strictly
biennial, self-fertile apple cv. Benoni- As Table 6 shows, CEPA applied just before
flowers open (10% open flowers stage) clearly stimulates flower abscission. Due to the
serious risk of overthinning and reduction of fruit growth, the concentration should
be kept as low as possible. Usually, hand thinning remains necessary but is much less
laborious than without application of CEPA. For many apple cultivars CEPA is
totally useless, because the risk of overthinning is too great. The thinning action of
CEPA is attributed to its release of ethylene after being absorbed by the plant. The
abscission-promoting effect of ethylene is well documented (Abeles [2]).

Table 6. Results of hand thinning and chemical thinning with CEPA (concentration; 1440 ppm
active ingredient) in four-year-old (1974) apple trees of cv. Benoni. Time of spraying: balloon
stage of flowers. Hand thinning done shortly after flowering. Field experiment.
1974 1975 1976
Hand CEPA Hand CEPA Hand CEPA
+ + +
hand hand hand
Flower clusters/tree before thinning . 210 209 165 196 425 354
Fruits/100 flower clusters .......... 43 40 52 45 54 36
Yield/tree (kg) ................... . 11.8 10.7 15.2 16.9 23.2 14.0
Hand-thinned fruits/tree .......... 934 205 946 445 1183 177

The auxin-like substances NAA and NAAm were formerly rather popular as thinning
agents in our region but due to adverse side-effects on fruit size and leaf condition
they have been largely superseded bycarbaryl. Since auxin inside the fruits is supposed
to inhibit fruitlet abscission, the thinning effect of NAA and NAAm is surprising. This
effect is probably due to the fact that auxin stimulates ethylene synthesis, which might
predominate after the application of auxin-like substances. Alternatively, a high
concentration of auxin is reported to induce seed abortion resulting in less hormone
production and ultimately in fruit abscission. Recently, Schneider [36] suggested that
the reduced movement of metabolites to the fruit he observed in NAA-treated trees
initiated the abscission process; ethylene would not play a primary role. This concept
would also explain why NAA and NAAm reduce fruit size.
The mechanism underlying the thinning action of carbaryl is not known, but since
it is chemically related to the naphtyl-auxins it may act via the same mechanism
(which is not completely known, as we have seen). Although carbaryl, being a broad-
spectrum insecticide, is undesirable in modern fruit-growing, especially for the main
cultivar Golden Delicious a belier thinning agent is not available. This cultivar gives
no biennial-bearing problems under our conditions, and fruit thinning is only needed
to improve fruit size. As e.g. Table 7 shows, the share of off-grade fruit can be reduced
appreciably by carbaryl, although an additional hand thinning is sometimes needed.
Thinning almost always reduces total yield but enhances the yield of fruits of good
size.
Promotion of fruit set is especially needed in years when, due to adverse weather
conditions at blossoming time, flowers are dhmaged and the activity of pollinating

144
Table 7. Examples of the effect of hand and chemical thinning with carbaryl (concentration:
750 ppm active ingredient) on grading in full-grown apple trees cv. Golden Delicious. Time
of spraying: 25 days after ftll bloom. Hand thinning done after June drop. Field experiment.

Orchard I Orchard 2
Un- land Chem. Un- Hand Chem.
treated thinning thinning treated thinniing thinning
Yield/tree (kg >70mm .. 8.8 23.3 21.8 9.1 23.6 20.1
Yield/tree (kg) <70ma ... 29.0 3.6 12.8 35.0 12.0 18.6
Total yield/tree (kg) ....... 37.8 26.9 34.6 44.1 35.6 38.7

insects is restricted. Since the receptacle is less sensitive to low temperatures, attempts
to stimulate parthenocarpic fruit set by application of gibberellins may be very
successful ( Varga [43]). Especially in pear cultivars having a strong tendency to
natural parthenocarpy (e.g. Triomphe de Vienne) either GA,, or GA 4 + 7 sprayed at
early bloom is widely used to promote frutit set. An example is given in Table 8; here
the improvement of frIit set was so marked that for fruit-size considerations, fruit
thinning had to be done. There is some risk that fewer flowers for the following year
will be initiated, but this effect is minimal when the concentration is kept low and any
application after ftll bloom is avoided. With respect to the market quality. partheno-
carpic fruits are usually slightly more elongated and smaller than seeded fruits, and
in some cultivars severe malformations may occur. Therefore, the use of these sprays
is discoUraged for some cultivars.

Table S. Examples of the effect of GA 3 (concentration: 15 ppm active ingtedient) on fruit-set

(expressed as fruits per 100 flower clusters) before June drop, and after thinning by hand, at
harvest, and on yield in full-grown pear trees cv. Trioniphe de Vienne. Time of spraying;
early bloom. Field experinent.

Experiment I Experiment 2
Untreated GA, Untreated GA,
5 June. before June drop ................... II 71 16 67
8 September, harvest date .................... 7 19 8 19
Y ield/tree (kg) .. ........... ................ 22.5 36.1 23.3 40.3

In apple the natural tendency to parthenocarpic fruit set is low and only a multi-
hormonal stimulus may have success. Kot(,b and Schwtabe [16] reported fruit set of
emasculated flowers of cv. Cox's Orange Pippin treated with a mixitUre of CiA, an
auxin, and a cytokinin, but more recent data given by Jonk,ers [14.] and obtained in
trials in Wilhelrninadorp show so much variation that possibilities for large-scale use
are unlikely.

3.3 Fruit set and shoot growth

It is normal for many fruit species that, due to unsUccessful competition for metabolites
between fruitlets and growing shoots, a large proportion of the initially set fruits drop

145
prematurely. This early drop, the so-called June drop, is particularly pronounced at
high temperatures (Westphal-Stevels [47], Grauslund [10]), undoubtedly due to
stimulation of shoot-tip activity. Shoot-fruit competition and consequently fruit drop
can be reduced by removing the shoot tips by hand (Quinlan and Preston [33],
Grauslund[10]), which, of course, is unsuitable for commercial orchards. Fortunately,
comparable results can be obtained by the application of growth retardants early in
the season, as has been shown in numerous experiments (Luckwill and Child [23],
Link [18]). However, the results are often inconclusive and, in addition, adverse
effects on fruit size and, especially in apple, on storage behaviour, have prevented
general use of this method in The Netherlands so far, at least to obtain a larger crop.
Finally, it should be remarked that in years when fruit set is extremely poor, for
example due to early frost damage, shoot growth may be very vigorous. This kind
of situation has been discussed in section 2.2.

4. Miscellaneous

4.1 Prevention of pre-harvest fruit drop

Especially in early cultivars of apple and pear, severe fruit drop may occur shortly
before the anticipated picking time, probably due to a decrease in the auxin content
of the seeds and, as ripening proceeds, to an increase in the production of ethylene.
Sprays of auxin-like substances such as NAA, NAAm, and 2,4,5-trichlorophenoxy-
propionic acid (2,4,5-TP) can prevent this pre-harvest drop satisfactorily, and this
application of growth regulators has been practised widely in fruit-growing for a long
time (Martin [26]). With respect to the mode of action of these auxins as applied,
the stimulatory effect of auxin on ethylene synthesis is obviously overshadowed by its
abscission-retarding effect.

4.2 Acceleration of the ripening process

In many types of fruit, ripening is characterized by an increased production of ethylene.
Therefore, it is not surprising that the ethylene-releasing agent CEPA can advance
ripening in, amongst others, apple and pear (Looney [20], Sharples [37], Link [19]).
Application of CEPA one week or a few weeks before the anticipated harvest time may
advance early marketing of fruits whose red colour is well developed and whose taste
may be better than that of untreated fruits picked at the same time (McBride and
Faragher [27]).
Table 9 gives some data for the summer cultivar of apple James Grieve Lired. Since
CEPA used alone may induce severe pre-harvest drop of fruits, it was applied in
combination with NAA. It is evident from Table 9 that CEPA improved colour
development and that a delay in picking increased the proportion of dropped fruits
dramatically, despite the addition of NAA. The later cultivars, e.g. Cox's Orange
Pippin and Golden Delicious, also responded to CEPA, although less strongly, with
respect to improvement of fruit finish, but due to the more advanced ripening, storage
life may be markedly shortened (data not given).

146
Table 9. Combined effect of CEPA (concentration: 480 ppm active ingredient) and NAA
(concentration: 10 ppm active ingredient) on pre-harvest fruit drop and red colour development
at two picking dates in full-grown apple trees cv. James Grieve Lired. Time of spraying:
12 August. Field experiment.

Picked on 26/8 Picked on 9/9

Untreated CEPA + Untreated CEPA +
NAA NAA
Dropped fruits (M )........................... 5 4 10 78
Fruits (% ), skin <1/3 red ................. 71 22 27 0
Fruits (% ), skin 1/- red ................ 23 56 35 6
Fruits (M ),skin > 2/, red ................. 6 22 38 94

4.3 Decrease in fruit-skin russeting

In certain apple cultivars, particularly Golden Delicious, the occurrence of skin

russefing may appreciably decrease the market value of the fruit. However, a few
years ago Taylor [39] observed that two or three sprays of GA 4 .,., applied early in the
season (at petal fall and two or three weeks later) could reduce skin russeting in
Golden Delicious fruits. This beneficial effect of GA,+, was confirmed in later
experiments by Fecher [7]. Wertheim [44] reported that in fruits of cv. Cox's Orange
Pippin, GA,+« may reduce cracking induced by drought. There is general agreement
that GA 3 is ineffective, which may be related to the fact that GA,, unlike GA 4 +,, does
not occur naturally in apple.
Table 10 shows recent spray results obtained for two apple cultivars in The Nether-
lands. In all cases skin smoothness was improved markedly. [n the local cultivar
Karmijn de Sonnaville, where fruit russeting and cracking are of serious concern to the
growers, GA.,17 may prove to be very useful. Preliminary results indicate that if the
applied concentration is kept low, there arc no adverse side-effects on fruit size and
flower-bud formation.
Skin russeting is initialed by uncontrolled cell division and enlargement in the epider-
mal layer. Seen in relation to the role of gibberellins in cell extension, it is plausible
that Eccher [7] found that application of GA,+ 7 favours normal development of the
fruit epidermis. He also found much higher concentrations of GA 4 +7 in fruits of a
normal clone of Golden Delicious as compared with a clone usually giving russeting
fruits, which suggests that russeting is due to a hormonal deficiency or unbalance.

Table 10. Examples of the effect of GA ,- (concentration 10 ppm active ingredient., 4 sprays)
on skin smoothness (weight percentage of smooth or slightly russeted fruits) in full-grown
trees of two apple cultivars. Time of spraying: late bloom and 10, 20. and 30 days after that.
Field experiment.
Treatment Orchard
2 3 4 5 6
Karmijn de Sonnaville Golden Delicious
Untreated ... ................. 18 50 21 42 82 60
GA4 .7 ........ 59 90 32 58 96 83

147
5. Concluding remarks

In spite of the extensive research done on plant growth regulators in the last few
decades, at least in the fruit-growing field, the number of compounds that are of real
general use to the grower is still rather small. The usually very variable results, some-
times leading to severe financial losses, are discouraging, and a basis for more reliable
prediction of what will happen after a certain treatment is badly needed.
With respect to this objective it must be kept in mind that the effect of each regulator
spray used, for example, to improve regularity in fruit bearing, is dependent not only
on environmental conditions throughout the season but also on the nutrient and
hormone status of the tree at the time of application, which in turn are determined by
such. factors as (he 'history' of the tree (e.g., fruit load in the preceding year), the
climatic 'history', and the soil conditions. Estimation of the nutrient status of a tree
is relatively easy and, for example, leaf analysis to determine mineral levels has proven
to be a useful tool to assess fertilizer needs. However, hormone analyses of plant
tissue are extremely complicated and the analytical data are often difficult to interpret.
It is nevertheless felt that study of the more fundamental aspects in the field of hormone
research in trees should be encouraged in order especially to increase the effectiveness
of the more 'applied' type of research.

6. References

I. Abbas, M. F: Association between branching in maiden apple trees and levels of endo-
genous auxins. Acta Hort. 80, 59-61 (1978)
2. Abeles, F. B.: Ethylene in Plant Biology, pp. 178-190. Academic Press, New York - London
1973
3. Baldini, E., Sansavini, S. and Zoeca, A.: Induction of feathers by growth regulators on
maiden trees of apple and pear. J. hort. Sci. 48, 327-337 (1973)
4. Cathey, H. M., Steffens, G. L., Start, N. W. and Zinunertnann, R. H.. Chemical pruning of
plants, Science 153, 1382-[383 (1966)
5. Chan, B. G. and Cran,, J. C.. The effect of seed formation on subsequent flowering in apple.
Proc. Am. Soc. hort. Sci. 91, 63-67 (1967)
6. Davis, L. D.: Flowering and alternate bearing. Proc. Am. Soc. hort. Sci. 70, 545-556 (1957)
7. Eccher, T.: Russeting of Golden Delicious apples as related to endogenous and exogenous
gibberellins. Acta Hort. 80, 381-391 (1978)
8. Firford,R. M.: The morphogenesis of apple buds. 1.Tire activity of the apical meristem.
Ann. Bot. N. S. 29, 167-180 (1965)
9. Fulford, R. M.: The morphogenesis of apple buds. Ill. The inception of flowers. Ann. Bot.
N.S. 30, 207-219 (1966)
10. Graushtnd, J.: Effects of temperature, shoot-tipping, and carbaryl on fruit set of apple
trees. Acta Hort. 80, 207-217 (1978)
II. 11oad, G. V.: The role of seed derived hormones in the control of fiowering in apple. Acta
Hort. 80, 93-103 (1978)
12. Hoad, G. V and Monselise, S. P.: Effects of succinic acid 2,2-dimethylhydrazide (SADH)
on the gibberellin and abscisic acid levels in stem tips of M 26 apple rootstocks. Scientia
Hort. 4, 41-47 (1976)
13. Jankiewicz, L. S. and Kozyra, J.: Use of maleic hydrazide (MH) to induce branches in
maiden fruit trees. Acta Agrobot. 26, 203-207 (1973)
14. Jonkers, H.: Induction of parthenocarpy in apple and pear with mixtures of growth
regulators. Acta Hort. 80, 123-127 (1978)
15. Kender, W. J. and Carpenter, S.: Stimulation of lateral bud growth of apple trees by
6-benzylaminopurine. J. Am. Soc. hort. Sci. 97, 377-380 (1972)

148
16. Kotob, M. A. and S(l(,he, n/. W.: Induction of parthenocarpic fruit ili Cox's Orange
Pippin apples. J. hort. Sci. 46, 89-93 (1971)
17. Ltwi, A.: Auxins in flowering. Encyclopedia of Plant Physiology 14, 909-950. Springer-
Verlag, Berlin - G.ttingen - Heidelberg 1961
18. Lin, /I.: Uber del Ejnfluss von Alar au f \Vachsti tm. Er trag und FrlchtqUalillit bei cinigen
Aptelsorten. Erwerbsobstbau 17. 152-155 (1975)
19. Lwtk. .:Einsatzin6glichkeiten von \VachlstumsrCgulaioren im Apfelanbait. Erwerbsobst-
batL 18 19-122 (1976)
20. Loone, A'.E.: Control of fruil maturation and ripening with growth regulators. Acta
Hort. 34, 397 406 (1973)
21. Luckuill, L C.: The control of growth and friitfulIlCss of apple trees. Physiology of Tree
Crops pp 237-254. Academic Press, London - New York 1970
22. /tw,,il, .( .:A new look at the process of rit bud formation in apple. Proc. XIXth
int. hoti. Congr. 1974, 3, 237-245 (1974)
23. Lmit0l L.C. and Child, R. D.: Growth retardants in apples: summary of experiments
1964-1966. Ann. Rep. Long Ashnon Res. Sm. 1966. 74-85 (1967)
24. Lit,dem P.and Debor, 11.It'.: Application and effect of Ajar on apples. Aktuellc Literatur
Informationen aus den Obstbau 62. pp. 41. Berlin 1977
25. Lftidc,s, P.and Debmr. I/ t: Application and effect of CCC on frtuit plants and grape
vines. Aktttelle Infornationen ats dent Obstbaut 70, pp. 34, Berlin 1978
26, main, G. C. Growth regUlators in poimology: the cnd of the beginning. HortSci. 14,
326-328 (1979)
27. Mb Th/d. R. L. and Faai('her, .1,D.: The sensory eval4tatioc of Jonathan and Delicious
apples treated with the growth regIator Ethephon. J.Sci. Fd Agric. 29. 465-470 (1978)
28. Almiselise, S. P.: The use of growth regulators in citricUItUre. a review. Scicntia Hort. 11,
151-162 (1979)
29. Minnemon. I/,. Z,ecirt, I A. 1.)., Kende, II. and Lang. 4.:The plant growth retardant
CCC as inhibitor of gibberellin biosynthesis inFiesa,tim, mtonilift)rme, Planta 61, 229-235
(1964)
30. Phillips, I. D. I: Apical domninance. Ann. Rev. Plant Physiol. 26, 341-367 (1975)
31.Quhm,.. D.: Chenical induction ol lateral branches (feathers). Acta iort. 65. 129-138
(1978)
32. Quitnln, .1. D.. The use of growth regulators for shaping young fruit tiees. Acta Hor, 80,
39-48 (1978)
33. Quiml.. . D.and Presoru. i, P.. The inflocrce of shoot competition on fr uilretention and
cropping of apple Ices. J. hort. Sui. 46. 525-534 (1971)
34. Reid,D. M.;.and Car. D.J Effects of a dwarting compound. CCC. on the production and
export of gibberellin-like substances by root systems. Planta 73, 1-11 (1967)
35. Ringo, K. and Sanmmin, S.: Effects of sUicCiic acid 2.2-dimethyllhydrazide ott flowering
and gibberellic acid conteit of sweet cherty (IPimius aim,L.). J. Tort. Sci. 47, 173-178
(1972)
36. Scneiider, G. W.: the mode of action of apple thinning agents. Acta Ilort. 80,225-231
(1978)
37, Shatles, R.0.: Effects of growth regulators on the storage life of apples and pears. Acta
Hort, 34, 421-428 (1973)
38. Skene, K.G. A,.: Tite telationship between the etfects of CCC oilroot growth and cytokinmin
levels in the bleeding sap of Tins linifru L.I. exp. Hot. 21. 418-431 (1970)
39. Ty'.B. K.: ReduCtin of apple skin russeting by gibberellin A\ 1 , hort. Sci. 50, 169-
172 (1975)
40. I.:Ellects of growthregultig substances and tree orientationt tilt growth and
1ip
lowet-bud formation in apple. J. hort. Sci. 47, 55 533(1972)
41. I7) ' J.. The interactonn ofgrwth reguttla orsand tree orientation On fruit-bld forinatin.
Acta Hort. 34, 185-188 (1973)
42. Van Osen, [. J.: Effect of initial tree qualit on yield. Acta Hcrt. 65. 123 127 (1978)
43. V'arga, A.: TuinbottwkUndige toepassingen van gibberellazuIr. Dissertatie Wageningen,
pp. 84. Veenman & Zone,. Wageningen 1963
44. Wertheim, S.I. The drop of flowers and fruits in apple, with special reference to the Jtine
drop of Cox's Orange I'ippin and its control with growth regutlaors. Meded. Landb.
l-logesch Wageningen 71 (17). 1-73 (1971)

149
45. Werthein, S.J.: Induction of side-shoot formation in the fruit-tree nursery. Acta Hort.
80, 49-54 (1978)
46. Wertheim, S.J.: Manual and chemical induction of side-shoot formation in apple trees
in the nursery. Scientia Hort. 9, 337-345 (1978)
47. Westphal-Stevels, J. M. C.: Bestuiving en vruchtzetting van Cox'Orange Pippin en
Jonathan in een fytotron (11). Fruitteelt 60, 874-875 (1970)
48. Williams, M. W.: Chemical control of vegetative growth and flowering of apple trees.
Acta Hort. 34, 167-173 (1973)
49. Zeevaart, J.A. D.: Reduction of the gibberellin content of Pharbilis seeds by CCCand
after-efects in the progeny. Plant Physiol. 41, 856-862 (1966)

150
Co-ordinator's Report on the 2nd Session
G.W. Cooke, Agricultural Research Council, London/England; Member of the Scientific
Board of the International Potash Institute-

Hormonal Regulation of Yield Formation

It became clear in the first session that a full understanding of the factors affecting
the formation of yield requires a knowledge of the roles of the substances which
regulate the processes of nutrient uptake, growth and transport of tile products of
photosynthesis to the organ which is to be harvested. This was stressed by Professor
Stoy.
In the first paper Professors Michael and Beringer reviewed our present knowledge
of the various roles of plant hormones in building yield, taking cereals as an example.
They pointed out that the 'picture' is incomplete and to a large extent hypothetical.
Biochemical knowledge of the formation and function of hormones is inadequate and
there has been too little research directly planned to elucidate their roles. The authors
discussed what is known about control of the number of ears per plant, numbers of
grains per ear and weight of a single grain. They gave detailed attention to the processes
involved in grain filling.
There are implications of present knowledge for crop husbandry practices including
cultivations, fertilizer use. and the application of growth control substances. The
effects of stress are now better understood as are the relationships between 'source'
and 'sink' in building yields. Improved knowledge of hormone regulation of physio-
logical processes should also assist plant breeders in defining their objectives. They
conclude 'One of our major attempts should therefore be to influence by breeding and
by husbandry the endogenous hormones to attain their optimal synthesis and efiects.'
The endogenous hormones of plants interact: their net effect may be altered when
exogenous plant growth regulators are applied. These effects were discussed by
Professor Bruinsina. lie pointed out that further hormone studies should show how
to direct the growth and nutrient distribution in crops by applying growth regulators.
Sonic regulators may alter either biosynthesis or translocation of endogenous hor-
mones; other substances may increase the concentrations of hormones or enhance
their action when the natural amounts are too sinall for maximIm effect. Better
knowledge of the interactions between hormones and regulators will offer new oppor-
tunities for altering growth patterns in the ways we desire.
When a crop yield is less than the potential set by its genetic makeup, the reason is
that it has been subject to stress. There are many causes of stress, lack of nutrients,

* Dr. G.W. Cooke, Chief Scientific Officer, Agricultural Research Council, 160 Great Portland
Street, London, WI N 6 DT/England

151
water, or the attack of a pest or disease are the obvious ones. Dr. Mizrahi discussed
the ways in which plants may alter hormone balances so that they may survive in spite
of the stress factor or factors. He dealt in particular with effects of water stress occur-
ringwhen plants are grown in saline solutions. He showed that under such stress the
activity of cytokinin decreased while the activity of abscisic acid (ABA) increased. This
suggested that ABA might be used to protect plants from excessive loss of water by
transpiration if it became cheap enough.
In the last paper Drs. Tromp and Wertheint dealt specifically with apples and pears.
They showed how regulators can be used to modify the growth patterns of young
trees before and after planting. They can also be used to control the factors that are
important in the production of fruit-flower bud formation, fruit set, thinning, and
preventing preharvest drop, The ripening process can be accelerated, and some quality
aspects (russetting for example) improved.
The gains in knowledge in recent years of the nature and functions of plant hormones
are impressive. This the Session showed clearly. The papers and the discussion also
showed how much more research is needed before we can be confident of manipulating
the processes within the plant to secure more of the product we desire (called yield).
We need more knowledge of cause and effect in work on plant hormoies. It also seems
that we need to know much more about the concentrations of plant hormones present
in the parts of plants. Where high concentrations are found there may be a temptation
to say that the substances were formed by the plant to achieve certain growth or
transport objectives. But could they not be sometimes the result of disordered metabo-
lism? - there is the very old example of putrescine found in potassium-deficient barley.
One question that was not discussed in the session was the analysis or assay of plant
hormones. As one who is ignorant of this branch of biochemistry may I ask if analytical
methods are well established and accurate? Another question I would like to ask is
whether we are likely to discover more hormones that regulate plant processes. Our
knowledge of recognised growth substances is so recent that I am tempted to forecast
that others will be discovered. If this is so would it not be worthwhile to coordinate
biochemical work on known processes to identify new compounds involved in their
regulation. A final question arises - do we have an adequate chemical effort to
synthesize analogues of known plant hormones (and of those to be discovered) with
the intention of introducing substances with more of the desired activity than is
possessed by the compounds that nature provides?
Whatever may be the answers to my questions I am convinced that we will continue
research that will increase our ability to control the processes of plant production to
benefit mankind. As with other aspects of agricultural science our objective must be
so to understand all components of production processes that we can completely
control the system to provide output at levels decided by need, available inputs, and
by economic circumstances. When we reach this stage we can claim to be capable
managers of our environment; the purpose of management is to avoid surprises - such
as those we often experience now when yields are far below accepted potentials.

152
 Chairman of the 3rd Session
Prof. Dr. K. Vegel.
Instituteof Plant Nutrition. Justus-Liebig-Univer-
sity, Giessen/Federal Republic of Germany

The Role of Nutrients

in Yield Formation
Nutritional and Environmental Effects
on Yield Formation
U. Beringer, Agricultural Research Station Boutehof,
Hannover/Federal Republic of Germany-

SuInfiary

Environnental factors and plant nutrients influence yield formation by being catalysts and
substrates of plant growth. But they may also have regulating functions at a cellular level. The
paper emphasises the latter point in elaborating how light, temperature, ion gradients, etc.,
can be considered as stimuli for the oriented and coordinated development of a crop. Morpho-
genesis, photo- and thermoperiodism are dealt with, source-sink relationships are discussed
and the role of nutrients for the endogeneous level and balance of phytohormones as regulators
of yield formation is stressed.

1. Introduction

Innumerable glasshouse and field trials have proved the essentiality and benefit of
fertilizers for crop yield. They have also demonstrated the effects which light intensity
and photoperiod, temperature, physical and chemical soil factors, water supply and
irrigation. biotic and abiotic stress situations, husbandry techniques and finally the
genetic predisposition of a plant might have on growth and development of a crop
canopy. A review about such a broad topic might run into the risk of enumerating
yield data obtained under different conditions of growth. This should be avoided,
because it will hardly reveal the causal events which have led to the final economic
yield mostly measured only gravimetrically. It should rather emphasize which experi-
ments allow conclusions about certain metabolic processes decisive for the ontogeny of
the crop and its physiology of yield formation.
With this aim in mind it is necessary to refer first to our present knowledge about basic
principles of morphogenesis. Plant nutrients and environmental factors are not only
substrates for and catalysts of plant growth. They can also be considered as stimuli for
certain steps of plant development, to which the plant responds in the sequence:
stinulus reception -- stimulus transduction -i physiological response (Bentup [3]).
Perception of environmental stimuli like light or temperature occurs through special
sensor molecules (for instance phytochrome (Zeevaart [60/), or changes in membrane
permeability (phase transitions after chilling, Ljons [36]). Transduction of stimuli
can happen through phytohorniones moving to a target tissue, which finally responds
to the stimulus by oriented growth (Lethain et al. [35]).
Prof. Dr. H. Beringer, Director, Agricultural Research Station Bintehof, P.O. Box 3209,
D-3000 Hannover/Fed. Rep. of Germany

155
Oriented growth in the context of this colloquium primarily means the development
of yield parameters i.e. of sink capacity and its filling with photosynthates. For these
processes the root as the absorbing organ for water and nutrients and the photo-
synthesizing shoot must be considered and special attention has to be paid to source-
sink relationships.
Finally examples will be given of the effects which water and mineral nutrients might
have on the regulation of yield formation and interpretations will be attempted. This
will, unavoidably, be rather speculative, because in biological research, especially in
such a plastic process as yield formation, causal analysis is impossible and even
factorial analysis insufficient for the design of a general 'system'.

2. Basic principles of morphogenesis

In the ontogeny of an organism cell division and cell expansion are accompanied by
the differentiation of cells according to their future functions. Already in the fertilized
egg cell an unequal distribution of the cytoplasm i.e. a certain differentiation or
polarity takes place. Such unequal cell divisions occur repeatedly during growth.

2.1 Polarity a basic phenomenon of the cell

A well-known example for polarity in plants is Pfeffer's experiment on the polarity in

the willow stem. Placing a stem cutting in moist air, roots will develop at the originally
basal end, while shoot buds grow at the morphologically apical end regardless of the
orientation of the cutting ( Wareing and Phillips [54]).
Polarity can be induced by light, by external electric fields and other signals. As an
example the light-induced orientation of outgrowth of a spore of the Fungus Botrytis,
known to farmers as the cause of grey mould in grapes is given (Figure I). If a spore
of Botrytis is partially illuminated and shaded (hatched area) respectively. the out-
growth of the rhizoid will take place at the site exposed to the light. Application of
plane-polarized light will induce an outgrowth in the plane of the incident light. From
this response to light the existence of a photoreceptor molecule must be postulated.
Another mechanism for cell polarity is known from the marine brown alga Fucus.
The zygote secretes a signalizing substance, called rhizin, whose chemical nature is
still unknown. This signal is uniformly distributed around the zygote. But as soon as
the zygote comes into the vicinity of a substratum, for instance a rock to which it will
later be attached, this substratum is a diffusion barrier. Consequently a concentration
gradient of the signal is built up, leading to the development of the rhizoid.
Polarity in Fucus zygotes can also be induced by imposed gradients of H+, K±, Ca++.
They generate a bioelectric field in such a way that Ca ions enter the zygote at the
rhizoidal pole and that K ions leave the cell at the apical side (Mohr and Schopfer
[40], Weisenseel [55]).
Bioclectric fields and surface potentials have also been measured in roots of onions,
broad beans, lupins and maize (Jaffe and Nuticelli [32]). A recent investigation on
roots of barley seedlings by Weisenseel et al. [56] has shown, that current enters
both the main elongation zone of the root as well as the growing tips of root hairs.
The current leaves again at non-growing sections of the root. Primarily protons are

156
 Botrytis

Focus
diflus,on b.rr.e,
stagnant ftr U.lrt. in the
Setwater seawae seawater

F4,. / Induction of poIaritV in a spore of Bo/wr/i by Unidirectional (a) or polarized (b) light
anid development of the rhizoid in a Ficim zygote by concentration gradients of a stimulus
(after Hemtrup, 1971 ' and Wleienseel 1979 )

responsible for this current. They seem to leak into growing cells or cell parts and to
be piumped OLIt of nol-growling ones.
Though the meaning of root surface potentials for the development of lateral roots
and root hairs under soil conditions is yet unknown, their existence reveals the
biophysical participation of nutrient ions in root growth.

2.2 Geo- and phototropism

Plant nutrients are also involved in the response mechanism of plants to geo- and
photoiropic stimuli. This is shown in Figure 2. Unilateral illumination of the maize
coleoptile causes its bending towards the light source. Bending starts after a lag phase
and the angle of bending increases with time. Before exposure to light the concentra-
tions of Ca, K, and1P respectively are uniform in both longitudinal halves of the
colcoptilc. But in response to the light stimulus Ca ions accumulate on the concave
side, whereas K and 1 accumIulate on the faster growing convex side suggesting a
causal link between such ionic gradients and the curvature of the coleoptile. The
light Stifnulus. however. can be replaced by exogencous application of indote acetic
acid (LAA). This indicates that an auxin gradient between both celeoptile sides is the
cause of the observed ion gradients; this the more so because auxin inhibitors prevent
the establishment of these ion gradients. Nevertheless comparison of the curves for
the individual ions and lor tie angle of coleoptile bending reveals that redistribution
of ions within the colcoptile precedes bending. which indicates that they are not a
consequence of the different growth intensities within the concave and convex hall.
In terms of growlh regulation by phytohormones and plant nutrients it can be assumed
from this and other examples given by Goswani and Audus [28]. that auxins may

157
 2t40'

- 10o 20'

-20 ' _

0 I 2 3 4

40*

10, • 20*

0 1 2 3 4
Tme (h)

Fig. 2. Influence ofa phototropic stimulus on curvature (v) and ion concentration (o) in maize
coleoptiles (after Gosw,and and A4udus [1976])

change membrane behaviour by activating ion pumps located in the membrane. On

the other hand the question may also be raised: how does ion composition and ion
concentration influence the effect of auxins and other growth regulators during plant
growth and yield formation?

2.3 Flower induction

The economic yield of most crops is seed or fruit. Flower 'induction and flower fer-
tility are therefore indispensible prerequisites for yield formation. Agronomists have
long been aware that some crops yield better if grown at a certain day-length or in a
more favourable climate and breeders have been successful in selecting lines less
dependent on such environmental conditions.
2.3.1 Pholoperiodism

During the transition from vegetative to reproductive growth shoot apices cease
developing additional leaf primordia and start to form of flower primordia. The
physiology of this ontogenetic step is interpreted by the hypothesis that certain genes,
which were suppressed and non-operational during vegetative development, are
activated. This activation can be either an autonomous process within the apical
meristern i.e. a genetically determined step (for instance early and late flowering
cultivars) or a transition of the apex induced by yet unidentified flowering hormones

158
(Florigen). Grafting experiments have proved that the flowering hormone is synthe-
sized within the leaves (Figure 3), from where it is translocated in the phloem to the
stem apex, being the target tissue. After reaching a threshold level there, it stimulates
RNA - and protein synthesis as well as DNA replication and mitosis. Within a few
hours the vegetative apex is then irreversibly changed into a generative apex (evocation).
Consecutively flower primordia are formed and the flower develops.
The physiology of flower formation in itself is a very complex process, described in a
comprehensive review by Zeevaart [60]. Again the nutritional status of the plant is
involved and probably modifies the photoperiodic response. Diomainto-Bonnand
(cit. in Zeevaurt [60]) observed for instance that Nicotiana gluthinosa flowered earliest
under long-day conditions at a high level of nutrition, while at low nutritional level
it responded as a quantitative short-day plant and at an intermediate level of nutrition
it was indifferent to the photoperiod. Taking such findings into consideration as well
as the participation of other exogeneous and endogeneous factors in flower induction,
photoperiodism may be a regulatory mechanism. But its quantitative importance for
economic yield of crops is didicult to assess. In addition it has been observed repeatedly
in the production of cereals and especially of grain legumes that the number of flowers
by far exceeds the final number of seeds (Beringer a. Tekete [67, Damnbroth [12],
Et.s [15]). Flower induction, therefore, seems very rarely to be a yield limiting
factor.

site of synthesis (leaves)

-,'

transport (phloem, all directions)

target tissue (shoot apices)

flower induction
(evocation)

morohogenesis of
flower orimordia

flower develowment

He,. 3. Path of the floral stituaIs (after Mohr and Schopfe, [1978])

159
2.3.2 Vernalisation and thertnoperiodisn
Some crops are known to need low temperature treatment (vernalisation) to accel-
erate or even to initiate flower formation. The range of the vernalising stimulus is
between 0 and < + 10'C, its duration required by most crops between 10 to 40 days
(Geisler [22]).
Winter cereals can be disposed for the perception of a low temperature stimulus when
grains are imbibed and swelling of the caryopsis has started. The target tissue is the
shoot meristem of the embryo, which is obviously influenced in its metabolism and its
differentiation by exposure to vernalizing temperatures. Although the molecular details
are yet unknown and 'Vernalin' is still as hypothetical as 'Florigen', the effect of low
temperature treatment in the seedling stage is remembered by all succeeding cells so
that finally formation of ears and flowers is accelerated.
In comparison to winter cereals rape and sugar-beet need a vernalizing stimulus when
in the rosette stage of growth. Only then are the apical meristems competent to perceive
the stimulus for which they have an absolute requirement for shoot elongation and
formation of inflorescences.
Thermoperiodism i.e. the fluctuation of temperature between day and night could
also regulate yield formation. Went [57] found higher growth-rates of tomato shoot
axes if temperatures at night were lower than by day in comparison to constant
day/night temperatures. This might be due to less dark respiration at lower temperature
and consequently more assimilates could be utilised for growth. Experiments with
sugar-beet support this view (Ulrich [50]) though the regulating metabolic steps are
definitely more complex.
In spite of the plausible argument that thermoperiodism with cooler nights and smaller
respiratory losses stimulates growth and dry matter production, one should be careful
with generalizations. In experiments with different temperature treatments metabolic
intensity and duration of yield formation are changed. Beringer [4] studied the effect
of day/night temperatures of 12°/12°C and of 300/12'C respectively during grain
development on the yield of oats. Twenty-three days after anthesis grains were bigger
and nearly mature in the warm treatment and would have indicated a clear thermo-
periodic response. Grains in the cool treatment, however, needed 72 days to maturity
and outyielded the warm treatment by nearly 30%.
In summary all environmental factors, as demonstrated here for light and temperature,
can act as stimuli and accordingly as regulating factors for certain stages of ontogeny.
As stimuli they can affect the endogeneous balance of phytohormones i.e. act bio-
chemically. They can also influence cell differentiation and plant development bio-
physically. As the hormonal aspects of yield formation are discussed elsewhere in this
volume (Michael and Beringer, p. 35) the biophysics of growth and yield regulation
will briefly be considered.

2.4 Biophysical aspects of stimuli action

Stimuli are perceived by substances, which are able to quickly alter their structure in
response to the stimuli. Besides chlorophylls and anthocyanins as major pigments,
higher plants synthesize sensor pigments in small quantities. Such a sensor pigment
is phytochrome, a chromoprotein in which light only changes the position of one
double bond in the chromophore moiety (Figure 4). Irradiation with red light (665 nm)

160
 Protein

Pr (blue} H

H H H

photoconversions 11
Protein

PP
Pfr (blue-green) H H -

0N N N N 0
HAR H

Fig. 4. Silggested structure of the chroniophore of phytochrome (fronm Mohr and Schopfer
! 1978])

leads to the formation of the blue green phytochrome Prr and this molecule is con-
verted back into the blue phytochrome (P,) if dark red light (- 720 nm) is applied.
These conversions are not fully complete, instead different equilibria are established
between the two (Mohr and Schopfer [40]).
The phytochrome system has multiple functions in plants. In photomorphism the
induction or repression of enzymes by phytochrome has been found and in photo-
periodism a high level of Per induces the formation of the flower hormone in long-day
plants (LDP) while in short-day plants (SDP) it represses it. Thus the response to the
Prr signal is different in LDP and SDP.
The effect of a vernalising stimulus on plant development might probably be perceived
by temperature sensitive membranes. This speculation is based on the identity of
vernalising and chilling temperaires. As a result of chilling injury in sensitive species
Lyons [36] described a phase transition of membranes from liquid-crystalline into
Ihe solid-gel phase combined with increased membrane permeability. Lyons' model
night therefore be a useful working hypothesis for molecular studies on vernalisation.
Plant nutrients can also participate in biophysical mechanisms of growth control. Not
only are ion gradients and electric fields built up in response to external stimuli as
shown above, but also the ion gradients themselves could induce certain steps of cell
differentiation. Evidence for interactions between phylohormones and nutrients is
increasing (see chapter 4) the role of nutrients for osmotic potential and osmoregula-
tion in cells has been more emphasized recently (Zinmernianu [62]).

161
3. Establishment and utilisation of yield sinks

Besides the initiation of reproductive plant parts, their number and further develop-
ment into seeds is of the utmost importance for final economic yield. That proportion
of total dry matter production which is deposited in storage tissues is generally ex-
pressed by the harvest index of the crop. Its magnitude depends on leaf area and photo-
synthesis (source) on the one hand and storage or sink capacity on the other. Both are
interrelated by many means, but the regulation of harvest-index is widely unknown.
Many studies on source-sink relationships have been conducted in the past. The
question, however, if photosynthesis or storage capacity is most likely the yield-limiting
factor is and will remain a controversal point in crop physiology (Gorasky et al. [27],
King et al. [33], Silbereisen and Buchloh [48]) further references in Michael and
Beringer p. 35 and in the comprehensive reviews by Evans [15] and Nasyrov [42].

3.1 Root system and shoot/root ratio

A close correlation between source and sink has been inferred from studies on the
shoot/root ratio of plants. If the environment is constant, there is usually a logarithmic
linear relationship between the weights of shoots and roots during vegetative growth
(Russel [44]), but on the other hand shoot/root ratio is modified by many factors
like light intensity, nitrogen supply, soil compaction, etc. Though the facts are known,
the physiological processes involved in the regulation of the shoot/root ratio are
hardly elucidated.
Very likely phytohormones are involved in the coordinated growth of root and shoot.
In comprehensive reviews Goodwin et al. [26] and Torrey [49] cite several examples
for roots being the major site of cytokinin and giberellic acid synthesis. But the data
collected so far are still insufficient for general conclusions. This is because differences
between species, cultivars, stage of plant development have to be taken into account as
well as the effects stress situations might have. Accordingly the suggestion by Sachs
[45] that a positive feed-back control of shoot and root growth would exist by
assuming root originating cytokinins to stimulate shoot apical growth and, vice versa,
shoot originating auxins to promote root development has to be questioned (Goodwin
et al. [26]).
How does nutrient supply influence root growth and what might be the regtlating
events? In an experiment with barley seedlings Drew [14] applied nutrient solution
with low and high N, P, K concentrations respectively to different root zones. If a
solution with high P or high N concentration was applied to the section 4-8 cm,
stimulated root growth occured in this zone. A localized application of the high K
solution, however, resulted in a root pattern visually comparable to the control plant,
receiving high nutrient supply to the total root. While in the P-treatment root and
shoot growth and consequently the shoot/root ratio was the same as in the control
plants (table 1), localized application of N and K resulted in growth reduction of
shoots and roots and in a decrease of the shoot/root ratio. The mineral contents of the
shoot do not suggest a nutrient deficiency. Consequently the changes in the growth
pattern could be due to changes in the phytohormonal balance (see chapter 4).

162
Table I. Influence of a high N, P, K supply to root section 4-8 cm on shoot and root growth
or barley seedlings (Drew [14])
root shoot shoot/ N P K
mg mg root % in shoot - DM
Control ( H-I) .- .................. 174 410 2.4 4.8 0.76 6.7
N (L-Fl-L)* . ...................... 150 266 1.8 4.8 0.84 7.3
P (L-H-L) ........................ 170 402 2.4 4.5 0.63 4.0
K (L-H-L) ........................ 151 316 2.1 4,9 0.76 5.6
* low and high nutrient supply respectively to root sections 0-4; 4-8 and >.8 cm.

3.2 Initiation and filling of storage tissues

The regulation of yield formation in cereals has been reviewed in detail by Michael and
Berhi,er (p. 35) who emphasized the regulative role of phytohormones in tillering,
grain setting and grain filling and considered sink capacity to be the main yield-
limiting factor.
This conclusion is partly supported, partly attenuated by an experiment of Forster and
Beringer (1980). As shown in Table 2. intact barley plants responded to increasing
levels of K-fertilizer with increasing grain yield. This increase was due to all yield
components and especially to increasing grain size. Removal of the flag and adjacent
leaf at anthesis caused insignificant reductions in number of ears and of grains. But
grain size was reduced by 6 to 21% as compared to grain size in control plants, SUr-
prisingly it was not the K 4-plants which were assumed to have the highest compensating
ability, btll the K 3-plants which showed the smallest depression in grain size. The
explanation for the significant 15% reduction of grain size in the K 4 treatment might
be the higher numbers of ears and grains in the K 4-plants. Due 10 the good K supply
since planting these plants had the highest tillering and grain set i.e.they developed a
higher storage capacity. After removal of the two uppermost leaves the remaining leaf
area was obviously unable to supply enough assimilates for normal grain development.
Accordingly it can be concluded, that in this experiment the equilibrium between
source and sink was rather static, though at different absolute yield plateau.

Table 2. Effect of K-nutrition and leaf area reduction on yield and yield components of barley
(Forster and Beri, er [1980])

Control Reduced leaf area

% change against
control
grains ears/ grains/ sgw ears/ grains/ sgw
g/pl. plant ear mg/gr. plant ear
K ............................. 1.4 2.4 32.7 18.6 0 -I -15
K ........................... 4.3 3.4 41.0 30.1 -3 -6 -21"
K .............. .. ....I.. . 5.3 3,4 41.1 37.3 -7 0 - 6
K I ............... .......... 7.6 3.7 43.8 46.8 -5 -2 1l5*
G D 5% ....................... 1.4 0.6 2.3 6.3

163
 The build-up of storage capacity in crops of which vegetative parts like tubers or
roots are harvested, i.e. in indeterminate crops, is more complex than in cereals.
In potato, tubers are initiated over a very wide range of developmental stages and in
general the older the plant the more prone it is to form tubers (Moorby and Mih/horpe
[41]). Superimposed on the effect of age, environmental conditions like short days,
low temperature, high daily irradition and low nitrogen supply hasten tuber initiation
(see Krauss, p. 123).
The storage root of sugar-beet is histologically characterized by several concentric
rings of meristematic and vascular tissue between which the sugar storing parenchy-
matic tissue is located. The factors determining and regulating the development and
activity of these secondary vascular meristems are unclear, but auxins and cytokinins
are probably involved (Fick et at. [16]). Bouilliene et al. [9] have shown that these
concentric, peripheral meristems are already formed by the time the young plant has
developed 6-12 leaves. According to Milford and Watson [39] six peripheral meristems
are formed at the 10-14 leaves stage. Three or four additional outer rings are produced
later, but they occupy less than 20% of the total cross-sectional area of the root. The
outer rings seem to be of less importance for the actual storage capacity of the root,
because root weights obtained in the field are generally far below the potential root
weight (Fick et al. [16]). Milford and Watson [39] observed that increased nitrogen
supply did not alter the number of cells and of pheripheral rings, but increased cell
volumes and the areas of both parenchymatic and meristematic tissues. This suggests
nitrogen that acted as a 'motor' of general growth, so that larger proportions of
translocated assimilates were utilized for root growth than for sucrose accumulation.
Although the mechanisms of sugar transport and storage in the beet are known
(Giaquinta [24, 25], Wyse [58]), knowledge on the regulation of yield is scarce and
concerns mostly gravimetric source-sink investigations. By reducing the number of
leaves to 6 or 8 per sugar-beet plant Dainbroth and Branun [13] noticed a reduction
in total growth, but the rate of photosynthesis per cm 2 leaf rose by 23 (8 leaves/plant)
and 39% (6 leaves/plant) respectively which indicates that in normal plants the leaf
apparatus possesses photosynthetic reserves and is unlikely to be the yield limiting
factor. If a sugar-beet plant can grow with less leaf area, then probably higher plant
populations/ha could be envisaged coupled with higher beet yields. First tests gave
the results presented in Figure 5. Whereas increasing plant population reduced leaf
area of control plants from 5500 cm 2 to 3350 cm 2, it decreased it only from 2120 to
1760 cm2 in plants which had 6 leaves left. Single beet weight decreased sharply
with increasing plant population, but the rate of depression was less in plants with
lower leaf numbers. On a yield/ha basis (Figure 6) higher plant densities of course
led to more leaf mass and due to competition between individual plants reduced single
beet weight, which was compensated by higher beet numbers. Consequently beet
yield in the leafy controls was quite stable, but in the treatments with plants having
only 6-8 leaves beet yields were raised to 550 dt/ha at the highest plant density. Sugar
yields/ha increased parellel to beet yields.
In terms of yield physiology these data suggest that sugar-beet breeding could prob-
ably focus on the selection of less leafy cultivars, which would allow higher plant
populations. But as the sugar percentages, inserted in Figure 6, show, the sugar
industry might not be satisfied with beets having only 12% sucrose. Therefore the
emphasis must also be to improve sugar accumulation in the root. Here the question
still remains open, if the number and activity of the 10-12 concentric secondary

164
 800
P,5 %

6001 - 4000 . Control

a 6 Leaves/PL
500 8 Leaves/PL.
3000

400j
] "Cootrol
300 / • 6 Leaves/P2
6 Leaves/P.

6 8 10 12 14 16 RantS/i 6 8 10 12 14 16 Plants/rn

F,. 5. Effect of rednced leaf number and plant density on growth or a single sugar beet plant
(Damhrolh and Bramini "1979])

P,5%
P'5% 700
550T
600 70 9S
500 2 o500 212.2)
S400 601 02At (123
450

.40020 e"
*Gcntr- *Coltrol (li,t *Control
01 46fath96
Leaves/F.
:.8 Leaves/Pt *6 8 Lea.vs/Pl. *6
&8
LeavesPl.
Leaves/Ft
350 100 Leavs/fl 40

6 0 4168 ol12l41 Rzrmsft 6 810 12 14 16Planta4T?

F, 6. Eflect of reduced ICar nUiber and plant density on yield Of sugar beet/ha (Dambroth
and Bramn 19797)

meristems are possibly not necessary for better sucrose accumulation. The grouping
of sugar-beet cultivars into 'yield' (large root with moderate sucrose concentration)
or 'sugar' types (smaller roots, but with greater sucrose concentration) indicates that
there are substances or mechanisms in the plant deciding about the utilisation of
photosynthates for root growth and sucrose accumulation respectively.

165
4. Role of mineral nutrients in the regulation of yield formation

Studies which reveal some regulatory functions of water and nutrients in growth and
yield formation have mostly been done by inducing a water or nutrient stress during
certain growth stages. The response of the plant has then been followed up morpho-
logically or chemically. What a stress situation might mean on a molecular and there-
fore regulatory level has been emphasized in studies on osmoregulation (Zimmermann
[62]) as well as on the synthesis and balance of phytohormones (see the papers by
Michael and Beringer, Bruinsina, Mizrahi, Kraus, Haeder in this volume).

4.1 Water status of the plant

Normal plant growth requires turgidity of cells. Turgor potential is the driving force
for cell extension after which cell differentiation starts. When a certain turgor pressure
is achieved membrane resistance reaches its maximum and net fluxes of ions and water
stop. Very likely growth hormones change this pressure-resistance balance in favour
of higher influx rates as was demonstrated by auxin applications to Valonia cells
(Zimmermann [61]).
If better nitrogen supply to sugar-beets resulted in larger cell volumes (Milford and
Watson [39]), it may be suggested that at better nitrogen supply more osmotically
active substances were in the root and caused higher water uptake by cells, i.e. in-
creased turgor potential and consequently cell volume. Similarly better growth of
Phaseolus vulgarisseedlings at higher K concentrations was paralleled by higher turgor
potentials in leaves, associated with larger cells and leaf area. This is plausible,
because K ions are a major contributing factor to osmotic and turgor potential (Arneke
[2]).

4.2 Nitrogen supply

Yield increases by nitrogen fertilizers are obtained all over the world. They are mostly
interpreted by nitrogen being a substrate for the synthesis of organic N compounds.
In addition the nutritional status in N can also be looked at from a hormonal view
insofar as nitrogen stimulates meristematic growth and cytokinin synthesis as shown
in sunflower seedlings by Wagner and Michael [51]. This result agrees with the finding,
that late N-dressing to rice increased both N and cytokinin content in root exudates
(Yoshida and Oritani [59]). Also the observation that nitrogen deficiency leads to the
initiation of new potato tubers by a relative promotion of abscisic acid content points
into the same direction (Saftelmacher and Marschner [46], see Krauss, page 123).
Interesting in this connection are also data by Trolldenier (unpublished) on the inter-
action between mineral nutrition and yield formation of wheat infected by Gduman-
nonyces graininis (Table 3). Although straw and grain yields of infected plants were
significantly lower than those of the control plants, improved nutrition with nitrogen
raised straw yields or infected plants by 22% (P,K 1) and 84% (P,K,) respectively
and increased grain yield by 34% (P1 K1) and 85% (P2K 2 ). These increases, which
also reveal distinct N and PK interactions, are interpreted by Trolldenier in accordance
to Garret [21]. This author reported that at higher N rates, in spite of more lesions,
root regeneration is faster than the infection by the pathogen, thereby compensating

166
Table 3. Influence of NPK nutrition on yield of wheat (g/pot) infected with Gil'uniannom ices

graminis (Troldenier. unpublished)

straw grains
cont r. infected contr. infected
N1',K, ................ ....... 57.5 31.9 (100) 76.4 23.4 (100)
N _,P,K ................ ........ . 52.5 38.9 (122) 56.3 31.3 (134)
N ,P,K ............................. 68.0 37.3 (100) 94.6 26.6 (100)
N,P_,K ........................ 80.9 68.5 (184) 116.3 49.1 (185)

the infection better. In the context of results cited above the higher rates of root
growth could also have led to higher cytokinin production.

4.3 Phosphorus supply

Like nitrogen stress phosphorus deficiency also causes lower levels of cytokintins in
root exudates of tomatoes, which was associated with a lower number of flowers on the
first truss. Applications of kinetin increased the number of flowers in P-deficient
seedlings (Mcnai , and van Staden [37.). This suggests that V-deficiency affected
flower formation via the hormonal balance rather than through lack of assimilates or
of P-containing substances.

4.4 Potassium supply

Many enzymes are activated by potassilurn and potassium is involved in photo- and
oxidalive phosphorylation. Better energy status and intensification of anabolic
reactions in the plant by adequate K nutrition as well as the importance of K for
osmoregulation, i.e. for stress tolerance are generally considered to be the yield
stimulating effects of potassium (Beringer [5], Beringer and Trolie,ier [7]).
When a suboptimal K supply to maize shortens the filling period, reduces single grain
weight and also decreases the rate of grain filling (Peaslee [43]), or when on the other
side adequate K nutrition promotes grain yield of cereals primarily via increased grain
size accompanied by retarded senescence of leaves and roots, then, it might be suggest-
ed to suggest, an effect of K on the level and balance of phytohormones,
Morphogenetic effects of varying K nutrition were investigated in So/anum sisymbri-
folidiu by Wak/iloo [52. 53]. Plants grown at a low, but not deficient, K level grew
larger, set more flowers of which 36% were sterile. Exogeneous applications of
giberellic acid increased sterility. On the contrary plants with higher K supply had
shorter internodes. less flowers, but only a 501 sterility. Kinetin application to low K
plants had the same effect as better K nutrition, i.e. reduction of plant height and of
sterility. These adverse effects of GA and kinetin simulating the differences between
low and high K nutrition indicate indeed a role o1 potassium nutrition in the hormonal
balance and accordingly in the regulation of yield formation. This is Supported by data
given in Table 4. In this biotest for cytokinins 40 mM K in the medium had the same
promotional effect on growth and chlorophytl content as 10 ng kinetin/l. Additionally
K and kinetin acted synergistically as indicated by a nearly trebled cotyledon weight
and a doubled chlorophyll content.

167
Table 4. Influence of 40 mM KCI and 10 mg/1 kinetin on growth and chlorophyll content of
cucumber cotyledons (Green and Muir [1978])
K Kin. mg/cotyl. chlorophyll
absorbance
- . . ....................... ............... 27.4 0.19
40 . ........................................ 37.4 0.40
- 10 .................................... 38.4 0.46
40 10 ........................................ 108.0 0.99

These results showing interaction between potassium supply and cytokinins might
be an explanation for the findings, that an interruption in the nutrient supply to
cereals during tillering has the most serious consequences for final grain yield (Forster
[17], Heyland [31], Mengel and Forster [38]).
Table 5 shows yield depressions by 20 to 40%. As far as ear number is concerned, they
are understandable, because no K was offered during tillering. Surprisingly, however,
also single grain-weight was reduced, although after tillering K was supplied again in
the nutrient solution and photosynthetic activity could have been similar to that of
control plants. A temporary K-deficiency during tillering could have resulted in in-
creased uptake of other cations, in reduced nitrate reduction and protein synthesis
(Forster and Mengel [19]). But what the real regulating factors were, remains specu-
lation.

Table 5. Effect of an interrupted K supply during tillering on relative grain yield of cereals
(controls= 100) (Mengel and Forster [1968]; Forster [1973])
Barley Wheat Oat
grain yield .......................................... 58.8*** 83.3** 81.5-**
ears/plant .......................................... 84.1** 90.4 " 90.9*
grains/ear .......................................... 92.5* 97.2 91.3*
grain-weight ........................................ 75.2*** 94.2** 96.9
* P<5%
* <1%
• <0.1%

4.5 Magnesium supply

The importance of an adequate nutrient supply during early vegetative growth of

cereals also applies to magnesium. Under the aspect of yield formation in cereals this
nutrient seems to be especially needed during tillering and grain filling. In nutrient
solutions containing 0.2 me Mg/I Forster (pers. comm.) found 5.4 tillers/barley plant
as compared to 5.8 tillers in the treatment containing 2.0 me Mg/l. Grain yields/plant
increased from 2.3 to 3.6 g. This was partly due to better tillering, but primarily to
promotion of single grain weight from 25.8 to 32.1 mg. These results are fully supported
by pot experiments with several soils. Highly significant correlations between rate of
Mg-uptake and rate of dry matter production were only found during early vegetative
growth and during grain development (Grinme, pers. comm.).

168
Magnesium is probably directly involved in the process of assimilate storage. Firstly
because Mg contents in grains are generally higher than in straw and secondly because
there is a highly significant positive correlation between Mg content and single grain
weight (Figure 7). Starch formation in chloroplasts is favoured by a high ratio of
triose-P: inorganic P (see Walker p. 143). In the amyloplast of a cereal grain starch
deposition could probably be promoted by magnesiUm keeping the concentration of
inorganic P low through a continuous synthesis of Mg-phytate.

*mg
*
-~ r48-
/%

5 , /o0
/ . 36

0 -r -. I --- r
5 10 15 5 to
Mg-content of straw ree/og) Mg- content of grain (me/100g)

Fie. 7. Relationship between Ig-contnt or straw and grains (left) and between Mg-content
and single grain weight in barley (Gimnt [pers. comm.])

4.6 Microntlricnt supply

Increasing rates of Nf[K fertilizers in countries with high agricultural productivity/ha

and the occurence of deficiency symptoms in plants grown on acid or calcareous
virgin soils have led to an enormous awareness of the role of micronutrients in yield
formation. The recent literature on yield increases in many crops by micronutrients,
has been reviewed by Bussler [10].
Micronutrients are known to be enzymatic cofactors. But what does the activity of a
certain enzyme or a micronutrient stress mean in the process of yield formation?
Microscopic B-deficiency symptoms are primarily detectable in cambial tissues, which
become broader by an accuMulation of non-differentiating cells. Kouchiand Kimazow
[34] observed in tonatoes that cell division and cell elongation is inhibited in root tips
and that formation of lateral roots is initiated. The latter seems to be due to higher
levels ofendogeneous auxins (Bohnsack and Albert [8]). Cytokinin contents, however.
were low in roots and root exudates of B-deficient sunflower seedlings ( Wagner and
Atichael [5/]). Boron might be considered as a regulating factor of cell division
(Cohen and Lepper 11]), but it may not directly influence cytokinin synthesis in the
root. Instead adequate B supply may indirectly promote root development by better
shoot growth and assimilate supply.
Copper deficiency has been reported to cause male sterility in wheat (Grahani [29j).
The critical period in copper-deficient plants is the early boot stage. when the meiotic
divisions of pollen mother cells exert a localised demand for copper. An impaired

169
translocation of nutrients, water and assimilates to the reproductive parts could also
be assumed, because in copper-deficient plants cell wall lignification is inhibited, less
xylem-vessels are built-up and phloem cells remain small (Bussier [10]).
Fe-chlorosis as well as Zn deficiency are nowadays encountered increasingly and world-
wide. In grapes a lack of both micronutrients caused the formation of numerous,
thin sideshoots (G(7rtel [20]). This suggests a break of apical dominance or at least
a shift in the hormonal balance, which is supported by lower levels of giberellic acid
found in Zn-deficient bean plants (Shkolnik et aL [47]) or stunted growth combined
with higher tiller formation of maize ( Wradzilo, cit. in Bussler [10]). Also the pref-
erential localisation of Zn in the cytoplasm of meristematic root cells could be looked
at from a hormonal point of view (Gerebtzoff and Ranaut [23]).

5. Concluding remarks

Fertilizers and other environmental growth factors affect plant growth and yield
formation in a most complex way which is best intergrated by the plant itself through
its plasticity. They can be considered as substrates and catalysts of plant production.
But they are more than this. They are also stimuli and regulators of plant development.
The perception of gravity, of photo- and thermoperiods, the build-up of ionic gradients
within cells and tissues, the synthesis as well as the balance of phytohormones are
only a few examples of how environment and nutrients are involved in the differentia-
tion of cells and in the realization of the genetic yield potential of crops.
Future agronomic research will have to pay even more attention than in the past to the
uptake of nutrients during the growth cycle, to the balance of macro- and micro-
nutrients, to the mutual physiological and morphological interactions in the growth of
roots, shoots and storage tissues. Only a detailed knowledge of these parameters will
lead to an understanding of the cybernetics of plant production and to a scientifically
based straight-forward improvement of plant productivity.

6. References

1. Allsopp, A.: The significance for development of water supply, osmotic relations and
nutrition. In: Ruhland, W. (ed.): Handbuch der Pflanzenphysiologie, Vol. 15/1 pp, 504-
552; Springer, Berlin-Heidelberg-New York, 1965
2. Arneke, WV.: Einfluss des Kaliums auf die Komponenten des Wasserpotentials und auf
die Wachstumsrate von Phaseolus vulgaris, Dissertation, Univ. Giessen, 1980
3. Bentrup, F. W.: Reception and transduction of electrical and mechanical stimuli. In:
Haupt, W.and Feinlelb, M.E. (eds.) Encyclopedia of Plant Physiology New Series Vol. 7,
pp. 42-70, Springer, Berlin-Heidelberg-New York, 1979
4. Beringer, H.: Einfluss der Temperatur auf Ertrag und Fettbildung in Haferkbrnern.
Z. Pflanzenernhr. Bodenk. 116, 45-53 (1967)
5. Beringer, H.: Functions of potassium in plant metabolism with particular reference to
[ yield. In: Sekhon, G.S. Potassium in soils and crops pp. 185-202. New Delhi, Pot. Res-
Inst. India, 1978
6. Beringer, H. and Tekete, 4.: Influence of mineral nutrition on seed development of two
groundnut cultivars. Comm. Soil Sci. Plant Analysis 10, pp. 491-501 (1979)
7. Beringer, 11. and Trolldenier, G.: Influence of K nutrition on response to environmental
stress. Proc. 10 Congr. Int. Pot. Inst., Bern, pp. 189-222 (1978)

170
 8. Bohnsack, C. W. and Albert. L.S.: Early effects of boron deficiency on indoleacetic acid
oxidase levels of squash root tips. Plant Physiol. 59 pp. 1047-1050 (1977)
9. Bouilhkne, R,. Kronacher, P. and Roubaix, .J. de: Etapes morphologiques et chimiques
dans le cycle v6g6tatif de Ia betterave sucriere Pubis. Inst. beige Amelior. Better. 8.
pp. 339-357 (1940)
10. Bussler, WV.:Mangelerseheinungen an hheren Pflanzen. IIISptrenn'hrstoffe. Z. PM.
Krankh. u. PI.-Schutz. 86, pp. 205-235 (1979)
11.Cohen. M.S.and Lepper, R.: Effect of boron on cell elongation and division in squash
roots. Plant Physiol. 59. 884-887 (1977)
12. Danbroh. M.: Ertragsphysiologisehe DCttungen for die Diskrepanz zwischen Bltenzahl
und Fruchtansatz bei Ackerbohnen. Gottinger Pflanzenzuchter Seminar 2. pp. 26-31,
1974
13. Datbroth, If. and Bratim, A,: ientical improvements for a better ratio between beets
and leaves in sugar beet. Paper pres. at ECE Symp. Economical and technical aspects
of sugar beet production, Kiew, Sept. 79, 1979
14. Drei,, M.C.: Comparisons of the effects of a localized supply of phosphate, nitrate,
anmmonium and potassium on the growth of the seminal root system and the shoot in
barley. New Phytol. 75, pp. 479-490 (1975)
15. Evtr. L. T. (cd.): Crop Physiology pp. 374 Cambridge University press, 1975
16. Fick, G. W., Loomis. R.S. and Williats. WA.: Sugar beet. In: Evans. L. T. (ed.) Crop
Physiology pp. 259-295 Cambridge University Press, 1975
17. Forster, H1.: Der Einfluss ter Kalium- und Stickstoffanlieferung an die Pfianze auf Er-
tragsmerkmale tind Ertragsbildung bei Getrcide. Landwirtsch. Forsch. 26, pp. 221-227
(1973)
18. Forster. I1.and Bebiger. U.: KornwachstL]Tn von Weizen bei redizierter BlattflNche und
variierter Kalinernlahrung. Die Bodenkultur (1980, in press)
19. Foro/er, H. and Mengel, K.: Der IinflUtss einer im ,Ugendstadium kttrzfristig unterbroche-
nen K-Ernihrulng atof Ertiragsbildtng Und ehalt an MineralstofTen und loslichen Anlino-
suren. Z. Ackel- tnd Pflanzenbau 130, pp. 203-213 (1969)
20. Gdrtel. W.: Dic Mikroniihrstoffic, ihre BedcLltung fir die Ikebenernalhrung unter beson-
derer Bericksictifigung der Mangel- und lberschusserscheirulngen. Weinberg u. Keller
21. pp. 435-508 (1974)
21. Gur,rel, S.D.: Pathogenic root infecting fungi. pp. 294 Cambridge University Press, t970
22. Gel-er, G.: Pflanzenbau. em Lehrbuch. Botanischc GrUndlagen tnd Technik der Pflan-
zenprodulktion pp. 474. Parey, Berlin-Hamburg. 1979
23. Ge-ebtzo.ff' A. and Rankat, I. L.: Essai de localisation histochimique du zinc chez Horde ... i
ildgnre et toxicite. Physiologia Pl. 23, 574-582 (1970)
24. Giaqidoilu. R.: Phloem loading of Sucrose. Plant Physiol. 63, pp. 744-748 (1979)
25. Giaquinta, R.: Sucrose translocation and storage in the sugar beet. Plant Physiol. 63.
pp. 828-832 (1979)
26. Goodi'n, P. B., Golbow. B. I. and Let/am,, D. S.: Phytohorniones and growth correlations.
In: Le-ham, Guodwtin and H/,rmins (eds.): Phytohormones amd related compounds Vol. 2,
pp. 215-249: Elsevier. Amsterdarn-Oxford-New York, 1978
27. Gorasky.S. R.. Pendleton, J. IW., Peters. D. B.. Borer. J.S. and Bener/in. J.E,: Internal
water stress and apparent photosynthesis with soy-beans differing in pubescence. Agron.
J. 63. 674-676 (1971)
28. Goseami. K. K. A. and Audits, Li.,. Distribution of Ca. K and P in Heliantlhus annmus
h/pocoflls and Zen tluvs coleoptiles in relation to tropic stinltii and curvature, Ann. Bot.
40. pp, 49-64 (1976)
29. Grahinm, R.D.: Male sterility in wheat plants deficient in copper. Nalure 254, pp. 514-515
(1976)
30. Green, .1.F. and Mur, R. ,.: The effect of potassium on cotyledon expansion Induced
by cylokinins. Physiol. Plant. 43, pp, 213-218 (1978)
31. Heilan,t K. U.. fiber die Bedeutung der Ertiaihrttng in verschiedenen Entwicklungsstadien
ir den Ertrag der Somnergerste. Z. Acker- u. Pflanzenbau, 113, pp. 41-65 (1961)
32. Jaffe, L. F. and Nutice/i. R.: Electrical controls of development. Ann, Rev. Biophys.
Bioeng. 6. pp. 445-476 (1977)
33. Kia,, R. I-'.. Wardiao. J.F. and Et-ans, L. T.: Effect of assimilate utilization on photosyn-
thetic rate in wheat. Planta 77. pp. 261-276 (1976)

171
34. Kouchi, H. and Kutnazawa, K.: Anatomical responses of root tips to boron deficiency.
Soil Sci. Plant Nutr. (Tokyo) 21, pp. 21-28 and 137-150 (1975)
35. Lethamn, D.S., Higgins, T.J. V., Goodwin, P.B. and Jacobson, J.F.: Phytohormones in
retrospect. In: Lethai, Goodwin and Higgins (eds.): Phytohormones and related com-
pounds Vol. 1, pp. 1-18, Amsterdam-Oxford-New York, 1978
36. Lyons, J.M.: Chilling injury in plants. Ann. Rev. Plant Physiol. 24, pp. 445-466 (1973)
37. Menary, R.C. and van Staden, I.: Effect of phosphorus nutrition and cytokinins on
flowering in the tomato, Lycopersicon esculentunt. Austr. J. Plant Physiol. 3, pp. 201-205
(1976)
38. Mengel, K. and Forster, H1.: Der Einfluss einer zeitlich variierten, unterbrochenen K-Er-
nfihrung auf Ertrags- und Qualit-tsmerkmale von Gerste. Z. Acker- u Pflanzenbau,
127, pp. 317-326 (1968)
39. Milford, G.F. and Watson, D.J.: The effect of nitrogen on the growth and sugar content
of sugar beet. Ann. Bot. 35, pp. 287-300 (1971)
40. Mohr, H. and Schopfer, P.: Lehrbuch der Pflanzenphysiologie, pp. 608. Springer, Berlin-
Heidelberg-New York, 1978
41. Moorby, . and Mifthorpe, F.L.: Potato. In: Evans, L. T. (ed.) Crop Physiology, pp.
225-257, Cambridge University Press, 1975
42. Nasyrov, Y.S.: Genetic control of photosynthesis and improving of crop productivity.
Ann. Rev. Plant Physiol. 29, pp. 215-237 (1978)
43. Peaslee, D. .: Effects of N, P and K on yield, rates of kernel growth and grain filling
periods of two corn hybrids. Soil Sci. Plant Anal. 8, pp. 373-389 (1977)
44. Russel, R.Scorl: Plant root systems, pp. 298, McGraw-Hill, London, 1977
45. Sachs, 7.: A possible basis for apical organization in plants. J. theoret. fiol. 37, pp. 353-
361 (1972)
46. Sattelnacher, B. and Marschner, B.: Tuberization in potato plants as affected by appli-
cations of nitrogen to the roots and leaves. Potato Res. 22, 49-57 (1979)
47. Shkolik, M. Y., Davidova, IKN. and Mochenyat, K.H.: Effect of Zn on the content of
giberellin like substances in bean leaves. Soviet Plant Physiol. 22, pp. 894-897 (1975)
48. Silbereisen, R. and Buchloh, G.: Uber die Bedeutung der Blitter am Neuwuchs for die
Fruchtausbildung beim Apfel. Gartenbauwissensch. 35, pp. 55-90 (1970)
49. Torrey, J. G.: Root hormones and plant growth. Ann. Rev. Plant Physiol. 27, pp. 435-459
(1976)
50. Ulrich, A.: The influence of temperature and light factors on the growth and development
of sugar beets in controlled climatic environments. Agron. J. 44, pp. 66-73 (1952)
51. Wagner, H. and Michael, G.: Der Einfluss unterschiedlicher Stickstoffversorgung auf
die Cytokininbildung in Wurzeln von Sonnenblumenpflanzen. Biochern. Physiol. Pflanzen
162, pp. 147-158 (1971)
52. Wakhloo, J.L.: Studies on the growth, flowering and production of female sterile flowers
as affected by different levels of foliar potassium in Solanun sisyinbrffolium. 1.Effect of
K content of the plant on vegetative growth and flowering. J. Exp. Bot. 26, pp. 425-432
(1975)
53. Wakhloo, J.L.: I. Interaction between foliar potassium and applied giberellic acid and
6-furfurylaminopurine. J. Exp. Bot. 26, pp. 433-440 (1975)
54. Wareing, P.F. and Phillips, I.D.J.: The control of growth and differentiation in plants.
Pergamon Press, Oxford, pp. 303, 1971
55. Weisen,seel, M.H.: Induction of polarity. In: Haupt, W. and Feinteib, M.E. (eds.): En-
cyclopedia of Plant Physiology New Series, Vol. 7, pp. 485-505, Springer, Berlin-Beidel-
berg-New York, 1979
56. Weisenseel, M. H., Dorn, A. and Jaffe, L. F.: Natural H' currents traverse growing roots
and root hairs of barley. Plant Physiol. 64, pp. 512-518 (1979)
57. Went, F.W.: Plant growth under controlled conditions. Amer. J. Bot. 31, pp. 597-618
(1944)
58. Wyse, R.: Sucrose uptake by sugar beet tap root tissue. Plant Physiol. 64, pp. 837-841
(1979)
59. Yoshida, R. and Oritani, T.: Studies on nitrogen metabolisms in crop plants. 1. Effects of
nitrogen topdressing on cytokinin content in the root exudate of rice plants. Proc. Crop
Sci. Soc. (Japan) 43, pp. 47-51 (1974)
60. Zeevaart, J.A.D.: Physiology of flower formation. Ann. Rev. Plant Physiol. 27, pp. 321-
348 (1976)

172
61. Zin mermann, U.: Cell tLirgor pressure regulation and turgor JNessure-mediaitd transport
processes. In: Jening, D. H. (ed,): Integration of activity in the higher plant. pp. 117-154,
Cambridge University Press. 1977
62. Zimmermann. U.: Physics of turgot- and osmoregulation. Ann. Rev. Plant Physiol. 29.
pp. 121- 148 (1978)

173
Influence of Nitrogen Nutrition
on Tuber Initiation of Potatoes
A. Krmss. Institute of Plant Nutrition. University of Hohenheir. Stuttgart/Federal Republic
of Germany / The International Potato Center, CiP, Lima/Peru*

Suilanry

Tuber initiation in potato plants is exogenously controlled by daylength, temperature and N

nutrition. High or continuous N supply delays or inhibits and low or discontinuous N supply
favours tuber initiation. The fitlience of N is indirect and based on alterations in the endo-
genous phytohornione balance. At continuous N supply the content of the 'inhibiting' hor-
muone gibberellin (GA) is relatively high and the content of the assumed 'promotor' abscisic
acid (ABA) is low. The corresponding narrow ABA/GA-ratio promotes shoot growth and
delays tuber initiation. Interruption of the N supply (discontinuous N) inverts the ABA/GA-
ratio by simultaneous increase in the ABA and decrease in the GA content which shifts the
phytohormone balance in favour of promotion'. At the same time the cytoki n in (CY ) con-
tent of the shoot of the induced plants is high even though CYT export from the roots has
ceased. Restoring the N supply after onset of tuber initiation again alters the ABA/GA-ratio
into an 'inhibitory' constellation similar to that of continuous N supply. This can lead to
regrowth, i.e. cessation of tuber growth and stolon formation. In the interaction between N
supply phytohormone level a nd tuber initiation the root system, as a source of phytohormones,
obviously has a key function.

The time of tuber initiation is an important stage in the yield fornation of potatoes.
Together with the length of the vegetation period the time of tuber initiation determines
the potential duration of bulking. For a given vegetation period the later tuber
initiation occurs the shorter is the time actually available for assimilate storage and
tuber growth. When tuber initiation is delayed the end of the vegetation period can
restrict further tuber growth.
The time of tuber initiation correlates with the haulrn growth. Factors which favour
haulm growth usually delay tuberization and vice versa. This inverse relationship has
been clearly demonstrated by Ivins and Bremner [15] as regards the effect of N
nutrition on shoot and tuber growth (Figure I). At a low rate of N fertilizer tuber
initiation occurs relatively early. [n contrast (he shoot growth is restricted and the
leaf area index (LAI,) remains small. With respect to yield formation, the relatively
short leaf area duration limits the duration of assimilate influx due to early senescence.
The bulking rate (Y,) declines althotIgh the vegetation period would still allow a
longer phase of tuber growth.

Dr. A. Krauss, Institute of Plant Nutrition, University of Hohenheim, P.O. Box 106, D-7000

Stuttgart/Federal Republic of Germany

175
 - - LA!2

) : LAI 1

(U I

- time
Fig.l Interactions between shoot and tuber growth in potatoes as affected by N nutrition
(after lvins and Bremner [15])

Increasing the N supply promotes shoot growth and increases the leaf area index
(LAI,). Simultaneously tuber initiation is delayed. If such plants are harvested early
the yield will be much smaller than would be the case with earlier tuber initiation.
High N supply gives longer leaf area duration and the phase of assimilate storage and
tuber growth (Y2) is extended.
However, this benefit from high N supply can only be realised if the weather remains
favourable for tuber growth so that high applications of N fertilizer to potatoes will
be utilized only in years with a long vegetation period (Chutterbuck and Simpson [4]).
The time of tuber initiation is also controlled by climatic factors such as daylength
and temperature. In respect to tuber initiation the potato plant is a 'short day plant',
i.e. to initiate tubers the daylength must be shorter than the 'critical photoperiod'.
Temperature obviously modifies the length of the critical photoperiod. Low tempera-
lure hastens (Burt [3]; Slater [39]) and high temperature delays tuber initiation
presumably by shortening of the critical photoperiod with increasing temperatures
(Bodlaender [2]).
While we have considerable knowledge about the effect of environmental factors on
the time of tuber initiation, less is known about the endogenous regulation mechanism
controlling the process of tuber initiation.
There is little doubt that phytohormones are closely involved in this regulation
mechanism despite contradictory views as to which hormone is the controlling factor.
Whilst Okazawa and Chapman [28] consider the balance of'promoting' to 'inhibiting'
phytohormones to be an important control mechanism, Ewing [8] and Kumar and
Wareing [22] prefer a specific 'stimulus' and Mauk and Langille [26] assume that
the cytokinins act as triggers in tuber initiation.
Most of the studies on the effect of exogenous factors on the hormonal control of
tuber initiation have concerned the photoperiodic reaction of the potato plants. Both
non-induced longday plants and plants grown at constant high temperature are
characterized by a relative by high content of gibberellins (GA) (Railton and Wareing
[32]; Pont Lezica [30]; Krauss and Marschner [20]). In contrast, induced plants
have a relatively low GA content. The inhibitory effect of GA is also revealed if the

176
GA content of the plants is increased by applying GA.,. On the other hand, if the GA
content is reduced by CCC, tuber initiation occurs even tinder non-inducing conditions
(I[amnes and Nel [14]).
'Promoters' of tuber initiation are obviously the endogenous inhibitors, in particular
abscisic acid (ABA). The promoting effect of ABA applications to the shoot (El
Antablv et al. [I]) or stolons (Krauss und Marschner [19]) of non-induced plants
provides some evidence for this assumption.
The 'stimulus* theory (Ewing and Wareing [9]) is based on the observation that
cuttings from induced plants will form tubers even after transfer into non-inducing
daylengths. Forthermore the induced cuttings can transfer the impulse when grafted
onto non-induced stocks (Ioc. cit.). The shorter the daylength to which the mother
plant is exposed the higher is the content (activity?) of the 'stimulus' in the cuttings.
With increasing reduction of the daylength, and thus with increasing induction of the
mother plant, the dormancy of the basal bud will first be broken and subsequently
stolons and linally tubers are formed.
The cytokinins (CYT) are also considered to be 'tuberizing hormones' because the
CYT content is higher in induced shortday plants than in plants grown under long
daylength (Forsine and Langille [10]: Motk and Langille [26]). However, even if
CYTs are essential to induce in-Viro tub-rization (Palvner and Smith [29]) it is still
an open question whether the increase in the CYT content following transfer of the
plants to inducing daylength is the precondition for or the consequence of tuber
initiation (see also Satteinacher and AIarschner [36]). Results obtained by Tizio and
Biait [40) and the effect of applying CYT to intact growing stolons which either
showed no response (Krau." and Marsc/ner [19/) or converted stolons into vertical
growing leafy shoots (Krnar and Wareing [21]) support a critical view.
While the pholoperiodic effect has been much investigated there is little information
on the effect of N nutrition on the endogenous regulation mechanism in potato plants.
The inhibitory effect of a high N supply could be interpreted as the result of an increased
sink strength of the growing shoot tips which will divert the assimilate distribution
mainly acropetally and restrict the basipetal translocation. The switch in assimilate dis-
tribution causing a relative shortage of assimilatles in the stolon tip would be respon-
sible for inhibition of tuber initiation. The results of Lovell and Booth [24], which
actually show that the assimilate distribution can be diverted by increase in the sink
activity of the shoot tip after treating the leaves with GA,. support this assumption.
An argument against a substrate-dependent tuber initiation is that stolon growth is
not restricted by high N supply (Krauss and Morschner [18]) although the amount of
dry matter produced by a growing stolon is comparable to that accumulated by
growing tubers immediately after initiation.
It is more likely that in analogy to the photoperiodic effect N nutrition also controls
tuber initiation by affecting the phytohormonc level in the plant. Interactions between
N nutrition and the endogenous phytohormones are known from other plant species
such as sunflower, tomato or tobacco (loc. cit.). however it is difficult to adopt these
results for the specific process 01 tuber initiation in potatoes.
The effect of N on the hormonal regulation mechanism in potatoes was studied in
water culture experiments. Using water culture it is possible to control precisely the
N supply and thus tile N uptake as well as the start of tuber initiation.
N was supplied either continuously, i.e. the N concentration of the nutrient solution
was kept above a I mil limit, or, in comparison, discontinuously. In the latter treat-

177
ment, a 3-day period of N supply alternated with a 6-day period without N (N
withdrawal).
At continuous N supply and uptake tuber initiation was inhibited even though the
plants were subjected to inducing climatic conditions (12 hrs daylength and constant
20'C). The shoot growth was not restricted and continued at a constant rate (Figure 2).
In contrast tuber initiation occured when N was supplied discontinuously, the first
visible swelling of the stolon tips occuring in the last third of a N withdrawal period.
Shoot growth ceased at the same time.

C) cont N

0o discont N

oo
/
0-

z
+N -N +N
to t6 t 9 days

Fig. 2. Effect of continuous and discontinuous N.supply on shoot growth and tuber initiation
(arrow = tuber initiation)

Providing the plants with a continuous N supply immediately after onset of tuber
growth can lead to an inversion of tuber initiation, i.e. 'regrowth' of the tubers
(Figure 3), tuber growth ceasing and one or more stolons being formed on the tuber
apex.
Reversible cessation of tuber growth following abundant N supply can also occur at
a later stage of tuber growth (Xrauss and Marschner [18]). However, when tuber
growth resumes after decrease in the N supply this is not seen as 'normal' enlargement
of the original shape but growth is mainly restricted to certain areas of the tuber
resulting in malformation (knobby tubers) and secondary growth (Figure 4).
Presumably the appearance of malformation and secondary growth in the field after
distinct fluctuations in the weather conditions is also the result of fluctuations in the
availability and uptake of N, e.g. when, after a dry period during tuber growth,
following rain N again becomes available.
The effect of N on tuber initiation and tuber growth is sensitive enough to produce
chain-like tubers (Figure 5). This will happen if, following 'regrowth' the N supply is
again interrupted inducing a second phase of tuber initiation. It is important to note
that in case of secondary tuber formation growth is restricted to the youngest tuber.

178
Fig.3. Regrowth of tubers following abundant N supply

Fig.4. Malformations of tubers as a result of disturbances in tuber growth caused by N

Fig.5. Secondary growth after alterating N supply

179
The contents of CYT, ABA and GA were determined in plants under different N
treatments (Figure 6). At 6 ontinuous N supply the ABA and GA content in individual
parts of the plant remained fairly constant. The CYT content tended to decrease in
the roots and stolons but to show a strong increase in the root exudate.

cont N discont N cont N discont N

-N :+N -N +N
roots 0 exu&ate 6-0
S/
--
2-
o 1- * . ..... o *..
e

0o shoots stolons/tubers: 7

0 6 9 6 90 6 9 69dy

Fig.6. Effect of N supply on the phytohormone content (relative values) in different parts of
potato plants -- CYT; -a- ABA ; - - GA (after Sajgel,acler and Marsclnr [36]; Krauss
[17]; Krouss and Marschner [20])

interruption of the N supply, i.e. providing inducing conditions, is characerized

partly by a steep increase in the ABA content and a simultaneous decrease in the GA.
content of the shoot. The GYT content in the root exudate is conspicuously reduced
whilst the CYT content of shoot and stolons show the opposite trend.
After restoring the N supply the hormone levels return to values similar to those at
the beginning of the experiment, i.e. the ABA and CYT content of the shoot de-
creases and the CYT content of the root exudate as well as the GA content of the
shoot increases. Only in stolons and developing tubers is there a steady accumulation
of hormone.
The change in the hormone level with the mode of N supply agrees with results
obtained on other plant species. N deficiency also caused an increase in ABA content
in sunflower (Go/ldhach et al. [12]) and tobacco (Mizrahi and Richmond [27]), the
decrease in the CYT content of the root exudate is comparable to results with sun-
flowers (Salama and Wareing [35]: Wagner and Michael [41]) and a decrease in the
GA content in N deficiency was also found in tomato (Rajagopal and Rao [33]) and
cranberries (Luke and Eck [25]).
The change in the ABA content of the shoot may reflect steep alterations of the ABA
content in roots and root extidate. The exact reason for this close negative relationship
between N and ABA is not known. However, it is conceivable that N withdrawal

I80
causes an onset of senescence of the roots. The cessation in CYT export in the root
exudate might support this assumption.
Lack of CYT usually promotes senescence accompanied by an increase in the ABA
content as shown by Even-Chen and Itai [7] with detached leaves. In agreement with
this, application of CYT could delay both senescence and increase in ABA. Therefore,
it is possible that the decreased ABA content of the roots after restoring the N supply
is also linked with the resumed CYT production of the roots.
Presumably the decrease in the GA content of the shoot during N withdrawal is also
connected with the physiological status of the roots. In the roots GA is produced
(Jones and Phillips [16]) and, at least in part, metabolized as well (Crozier and Reid
[5]). Stress situations such as periodical N deficiency could therefore interfere with
these processes and affect the GA export as has been shown in waterlogged tomato
plants (Reid el aL 134]).
A second way by which the GA content of the shoot might be changed is by altering
GA production or GA turnover in the shoot. It may be assumed that the GA pro-
duction in the shoot tips is linked with their meristematic activity. Interruption of the
N supply caused cessation of the shoot growth (Figure 2) and this is certainly corre-
lated with reduced meristematic activity and consequently lower GA production of
the apical tips. tn addition the increased CYT content could stimulate a higher GA
turnover rate as was found in peas (Railton [31]).
The increase in the CYT content of the shoot after interruption of the N supply
strongly contrasts with the reduced CYT transport in the root exudate. This could
indicate that during tile phase of tuber initiation the CYT content of the shoot is no
longer controlled by the roots but rather by the tuber initiation itself. Accumulation
of CYT in the shoot despite a lack of CYT supply from the roots could result from
activation of biologically inactive forms such as CYT-ribotides. The change in the
CYT spectrum observed during the phase of tuber initiation (Satteinacher and
Mwrschter [36]) makes this assumption realistic.
Support for a direct tuberization/CYT interaction is also found in other results from
Sattehnacler and Mcrschner [37] in which they demonstrated that following inter-
ruption fo the N supply to juvenile plants, i.e. non-induced plants, the CYT content
of the shoot reflected very well the reduced CYT transport from the roots.
Assuming that the increased cY'r content of the shoot is the result of and not the
reason for tuber initiation, the simultaneous change in the ABA and GA content of
the shoot, i.c. the ABA/GA-ratio, can be considered as a precondition for tuber
initiation. Independently of whether the change in the ABA/GA-ratio is the result of
exogenously induced alterations in the endogenous phytohornone level (N supply.
daylength, temperature) or affected directly by applications of hormones or growth
regulators (CCC) a wide ABA/GA-ratio promotes and a narrow ABA/GA-ratio
inhibits tuber initiation.
During the phase of interrupted N supply the ABA/GA-ratio becomes wider as a
result oft'he simultaneous increase in ABA and decrease in GA contents. Consequently
tubers are induced. When N is supplied again the ABA/ratio becomes narrower, i.e.
it is shifted in favour of 'inhibition'. This can lead to regrowth. Regrowth of the
tubers can also be initiated by application of GA 3 to the shoot (Lippert et at [23])
or the tuber (Kriauss and illarse/hter [19]), i.e. by an artifical increase of the GA
content which decreases the ABA/GA-ratio.
Non-induced plants with a continuous N supply have a fairly constant ABA content

181
and the GA content shows only a slow gradual decrease. In consequence the ABA/GA-
ratio does not show a drastic change.
It is important to note that the regulatory effect of N is obviously linked exclusively
to the root system. Continuous N supply will inhibit tuber initiation only when N is
supplied to the roots but not when given to the leaves (Sattehnacher and Marschner
[38]).
The pronounced effect of N on tuber initiation and tuber growth as reported has
consequences for estimating the N fertilizer requirement of potatoes in the field.
If early tuber initiation is desired, e.g. for the early crop, for seed potatoes, or in
areas with a short growing season the N supply should be low because early depletion
of the available N in the soil allows an early change in the ABA/GA-ratio. However,
with early tuber initiation, the tubers become the dominant sink much earlier and
consequently shoot growth is restricted. This source limitation restricts the final yield.
Increasing the amount of N fertilizer extends N availability in the soil. A longer phase
of N uptake delays the change in the ABA/GA-ratio. Shoot growth is enhanced and
tuber initiation inhibited. Although better shoot development (higher LAI and longer
LAD) increases the yield potential, late tuber initiation carries with it the risk that an
early end to the growing season, e.g. low temperature in autumn, may prevent the full
utilization of the yield potential. Therefore the length of the season determines how
far the shoot growth should be promoted and tuber initiation delayed by N nutrition.
A deliberate delay in tuber initiation by using high rates of N could however, be a
method of increasing tuber yield for instance at low latitudes where the natural short
days promote tuber initiation at the expense of shoot development.
Unfortunately it is still not possible to control the time of tuber initiation in the field
independently of shoot growth. Even when application trials with CCC show pro-
mising results (Gifford and Moorby [11]; Gunasena and Harris [13]) the long
persistence of CCC in potato plants and tubers (Dekhuijzen and Bodlaender [6])
affects the development of the daughter plants and thus prohibits the wide use of such
a method in commercial potato growing.

References

I. El-Antably, 11.M.M., Wareing, P.F. and Hillman, J.: Some physiological responses to
d'l abscisin (dormin). Planta 73, pp. 74-90 (1967)
2. Bodlaender, K.B.A.: Influence of temperature, radiation and photoperiod on development
and yield. The growth of the potato. Proc. 10th Easter School agric. Sci. Nottingham
p. 19 9, 1963
3. Bart, R. L.: Influence of short periods of low temperature on tuber initiation in the potalo.
Eur. Potato J. 7, pp. 197-208 (1964)
4. Clunerbuck, B.J. and Simpson, K.: The interactions of water and fertilizer nitrogen in
effects on growth pattern and yield of potatoes. J. agric. Sci. 91, pp. 161-172 (1978)
5. Crozier, A. and Reid, D.M.: Do roots synthesize Gibberellins? Can. J. Bot. 49, pp. 967-
975 (1971)
6. Dekhuijzen, H.M. and Bodlaender, K.B.A.: Distribution and persistence of chlormequat
in potato plants. Pestic. Sci. 4, pp. 619-627 (1973)
7. Even-Chen, Z. and ltai, C.: The role of abscisic acid.in senescence of detached tobacco
leaves. Physiol. Plant. 34, pp. 97-100 (1975)
8. Ewing, E.E.: Critical photoperiods for tuberization: a screening technique with potato
cuttings. Amer. Potato 1.55, pp. 43-53 (1978)

182
 9. Ewhg, E. E. and Wareing, P. F.: Shoot, stolon, and tuber formation on potato (Solan''l,
tuberostan L.) cuttings in response to photoperiod. Plant Physiol. 61, pp. 348-353 (1978)
10. Forsite, P. L. and Langille, A.R.: Endogenous Cytokinins in Solanum tuberosum Is
influenced by photoperiod and temperature. Physiol. Plant. 34. pp. 75-77 (1975)
1i. Qi//brd, R.M. and Moorhy, f.: The effect of CCC on the initiation of potato tubers.
Eur. Potato J. 10. pp. 235-238 (1967)
12. Golelbach. E., Golbah, It., Wagner, H. and Michael. G.: Influence of N deficiency on
the abscisic acid content of sunflower plants. Physiol. Plant. 34. pp. 138-140 (1975)
13. Gu,nasena. if. P. ,. and Harris. P. M.: The effect of CCC. nitrogen and potassium on the
growth and yield of two varieties of potatoes. J. agric. Sci. 76, pp. 3 3-52 ([971)
14. famnes, P.S. and Nel, P.C.: Control mechanisms in the tuberization process. Potato
Res. 18, pp. 262-272 (1975)
15, I ns, J.D. and Brenwer, P.M.: Growth. development and yield in the potato. Out.
Agric. 4, pp. 211-217 (1965)
16. Jones, R.L. and Phillips. I.D.J.: Organs of gibberellin synthesis in light-grown sunflower
plants. Plant Physiol. 41, pp. 1381-1386 (1966)
17. Krarass, A.: Tuberization and abscisic acid content in Solnutan a,herosiot as affected by
nitrogen notrition. Potato Res. 21, pp. 183-193 (1978)
18. Krauss, ,4. and larvch,ter. H.: Einfluss der Sticksloffernihrung der Kartoffeln auf [ndik-
lion und Wachstumsrate det Knolle. Z. Pfianzenernahr.. Bodenktinde 128, pp. 153-168
(1971)
19. Krau,', A. und Marvchner, H.: Einfluss vol Stickstoffernahrung Lind Wuclsstoffappli-
kation aut die Knolleninduktion bei Kartollelpflanzen. Z. Pflanzenerniihr., Bodenkunde
139, pp. 143-155 (1976)
20. Kra,,ss. A. and Mat,scher. I.: InfLience of nitrogen nutrition. daylength and temperature
on contents of gibberellic and abscisic acid and on tuberization of potato plants. Potato
Res. (submittee 1980)
21. Kimar, 1). and Wareitg, P.F.: Factors controlling stolon development in the potato
plant. New Phytol. 71, pp. 639-648 (1972)
22. Kutmtr, D. and Wanreh, P. -,:Studies on tuherization in So/amtma andigena, I. Evidence
for the existence and movement of a specific tuberization stimulus. New Phytol. 72.
pp. 283-287 (1973)
23. Lipperi, L.F., Rappport. L. and Timm, f: Systematic induction of sprouting in white
potatoes by foliar application of gibberellin. Plant Physiol. 33. pp. 132-133 (1958)
24. Lovell. P.H. and Booth. A.: Effects of gibberellic acid on growth, tuber formation and
carbohydrate distribution in Solanum tuberosum. New Phytol. 66, pp. 525-537 (1967)
25. Lu,ke. N.C. and Eek, P.: Endogenous gibberellin-like activity in cranberry at different
stages of development as influenced by nitrogen and daminozide. J. Amer. Soec. hort.
Sci. 103, pp. 250-252 (1978)
26. Malihk, C.S. and Latnille, A.R.: Physiology of tuberization in Solrm,,, triberosim L.
Cis-zcatin riboside in the potato plant: its identification and changes in endogenous levels
as influenced by temperature and photoperiod. Plant Physiol. 62, pp. 438-442 (1978)
27, Mizrahi, Y and Richmond. A. E.: Abscisic acid in relation to mnineral d.privation. Plant
Physiol. 50. pp. 667-670 (1972)
28. Okazawa. Y. and Chapman, Hf.W.: Regulation of tuber formation in the potato plant.
Physiol. Plant, 15, pp. 413-419 (1962)
29. Pc/ner, C.E. and Stmlh. O.E.: Cytokinins and tuber initiation in the potato Solanio,n
iuberosum L. Nature 221, pp. 279-280 (1969)
30. Poni Lezica, R.F.: Evolution des substances de type gibberellines chez In pomime de terre
pendant la tub5risation, en relation avec Ia 1ongIeur du jour et In tWmprattre. Potato
Res. 13, pp. 323-331 (1970)
31. Railon. J.D.: Eflects of N'-benzvladeninc on the rate of turnover of 3H-GA, by shoots
of dwarf Pisto swmirny. Planta 120. pp. 197-200 (1974)
32. Railon, J.D. and tiare/t. P. F..Effects of daylength on gibberellins in leaves of Solann,.
andigena. I. Changes in levels of free, acidic gibbcrellin-iike substances. Pliysiol. Plant. 28,
pp. 88-94 (1973)
33. Rijak,opa/, V/. and Rao, J.l.: Changes in the endogenous level of atxins and gibberellin-
like substances in the shoot apices of nitrogen-deficient tomato plants (L,copersico
esci/entm Mill.). Atst. J. Bot. 22. pp. 429-435 (1974)

183
34. Reid, D. M., Crozier, A. and Harvey, B.M. R.: The effects of flooding on the export of
gibberellin from the root to the shoot. Planta 89, pp. 376-379 (1969)
35. Salanta El-, D.A. and Wareing, P.F.: Effects of mineral nutrition on endogenous cyto-
kinins in plants of sunflower. J. exp. Bot. 30, pp. 971-981 (1979)
36. SaIrel,nacher, B. and Marschner, H.: Relation between nitrogen nutrition, cytokinin
activity and tuberization in Solantnn tuberossn. Physiol. Plant. 44, pp. 65-68 (1978)
37. Sanenacher, B. and Marschner, H.: Nitrogen nutrition and cytokinin activity in Solanum
tuberosum. Physiol. Plant. 42, pp. 185-189 (1978)
38. Sattehnacher, B. and Marschner, H.: Tuberization in potato plants as affected by appli-
cations of nitrogen to the roots and leaves. Potato Res. 22, pp. 49-57 (1979)
39. Slater, J. W.: The effect of night temperature on tuber initiation of the potato. Eur.
Potato J. 11, pp. 14-22 (1968)
40. Tizio, R. and Biain, M.M.: Are cytokinins the specific factors for tuber formation in the
potato plant? Phyton 31, pp. 3-13 (1973)
41. Wagner, H. und Michael, G.: Cytokinin-Bildung in Wurzeln von Sonnenblumen bei
unterschiedlicher Stickstoffernihrung und Chloramphenicol-Zusatz, Naturwiss. 56,
p. 3 7 9 (1969)

184
Effect of Potassium Nutrition
on Sink Intensity and Duration
l/.E. Haeder. Agricultural Research Station BOntehof.
-lannover/Federal Republic of Germany*

Sionniari

The filling rate i.e.the sink intensity of cereal grains can be enhanced by K nutrition of the
plants. This is demonstrated in short teirmn experiments where detached barley cars from plants
of different K status were supplied with equal amounts of 14C sucrose at diffcrent stages of
grain development. Ears Irom plants with a high K status accumulated more C/24 irs in their
grains. Within the grains 1'C was mainly incorporated into starch, but also into amino acids
and organic acids. Although K enhanced absolutely 11C incorporation into all these fractions,
no specific promotion of starch synthesis by K could be found. Accordingly K has a stimulating
effect on sink intensity a general way. An extension of sink duration. i.e. a prolonged grain
filling period, by increasinig K nutrition was also observed in pot experiments with wheat.
This could have been due to retarded leaf senescence and consequently better supply with
assimilates. Btll analysis of abscisic acid (ABA) during grain development suggests an involve-
ment of this phytnlhormone. Increasing K nutrition delayed the occurence of the ABA
maximum and decreased its absolute level in the grains. Very likely other phytohormones
and changed source-sink relationships also have to be taken into consideration. The pro-
motional effects of K nutrition on sink intensity and sink dutration, i.e. on grain size need
further elucidation.

1.Introduction

In most crops the plant parts producing photosynthates (source) are different from
the tissues which store these photosynthates (sink). The central problem of yield
physiology is the discussion about the importance of source and sink for yield forma-
tion (Gifford el al. [4j, SloY [23]). In some cases the sink is considered to be more
important for yield formation than source (Ponnner [18]). But recently it was suggest-
ed. that both are similarly important (Soy [25], Tol[enaar [26j).
The effect of K nutrition on source i.e. on photosynthesis in leaves as well as on the
translocation of photosynthates towards grains and tubers has been studied in several
experiments (IHarder et al. [7]. Mengel and Haler [12]). From these results it was
slated, that K enhances the photosynthetic rate in leaves (Mengel and ilaeder [13])
and the flow rate of assimilates in the phloem vessels (Me,twel and f/ueder [14]) to
grains and tubers (Haeder et al. /7j). More recently the influence of K on several

* Dr. II.E. Hoeder, Agricultural Research Station Buntehof. P.O. Box 3209. D-3000 Hanno-
ver/Federal Republic of Germany

185
components of the sink has been examined (Forster [2]). The sink can be subdivided
into three components:
1. sink potential i.e. the number of grains per ear or tubers per plant, which are
available for filling with assimilates.
2. sink intensity i.e. the demand of the storage tissues for photosynthates or the
filling rate.
3. sink duration i.e. the extension of filling period.
The number of grains per ear is largely fixed genetically. Higher nutrient supply
to barley for instance increased the number of tillers per plant, when applied in early
stages of growth, and single grain weight, when applied later (Stoy [24]). But in
barley the grain number per ear was hardly affected (Fuchs [3]) ; in contrast to
wheat (Forster [2]).

2. Effect of K on sink intensity

In studying the influence of K on sink intensity it is generally difficult to distinguish

between effects on source and on sink. This is explained by the fact that different K
concentrations in the nutrient solution affect vegetative growth and the transport of
photosynthates from the leaves to the storage organs. To investigate sink intensity,
however, it is necessary to supply the same amounts of photosynthates i.e. sucrose,
to all ears. For our experiment therefore detached ears with the uppermost culm
internode (15 cm length) were used.
The ears were cut from plants with different K nutrition, as indicated by the different
K contents of the straw:
K,=7.0; K,=21.7; K 3 = 36.8 mg K/g dry matter.
K contents in grains did not differ very much.
Sucrose uptake by the culm was for 24 hours. Then the ears were dried, separated into
grains, rachis, awns, and culms. In all these organs 14 C labeling was measured.
At different stages of grain filling (14, 21,28 and 35 days after ear emergence) eight ears
per treatment were put in 10 ml of a solution containing 50 mg ' 4C sucrose per ml,
according to a technique described by Jenner and Rathjen [10]. Incorporation of
supplied 14C sucrose into the grains during the whole grain filling period was enhanced
according to the K level of plants (Figure 1). The translocation of 14C into the grains
generally declined with increasing age of the grains and correlated negatively with dry
matter content in the ear. When dry matter content of the ear reached about 60%, the
accumulation of photosynthates in grains dropped (Spiertz [21]).
In our experiment this dry matter content was reached at day 35 after ear emergence
(Table 1). It is noteworthy that at 21 and 28 days after ear emergence grains of K,
and K 3 treatments already had a higher dry matter content. Nevertheless these grains
also incorporated more 14C sucrose during 24 hours which indicates the promoting
effect of K nutrition on sink intensity. At day 35 after car emergence dry matter
content of K3 ears was only 61.7% as compared to 66% in K, and K 1 respectively,
This lower dry matter content or in other words the higher water content of ears of
,(3plants at this age reveals a slight extension of grain filling (sink duration) and a

186
 30

20

10

* Awns
* I Rachis

0.. Grains

14 21 28 35
Days after ear emergence

Fig. /. ''C distribution in ears of barley plants. Detached ears were supplied with "C sucrose
for 24 hours and then harvested. Values were obtained by combustion of dry plant naterial
in an oxidizer and measuring I'CO,. Each value represents an average of four measurements.
The vertical bars at the right side of the colunis represent LSD 5% for three K treatments
each.

Table 1. Dry matter content in ears of barley plants (%)

Days after ear emergence K, K K,

14 ........................ 33.6 33.8 32.7
21 . .. .. .. .. .. ... .. .. .. .. ... .. .. . ........ .. .. .. .. .. ... . .. 39 .8 4 0 .7 4 2 .6
28 ......... ..................................... ......... 44 ,5 49 .5 49.5
35 .. .. ... .. . .. .. .. .. .. .. .. .. .. .. .. ... .. .. .. ... .. .. . ... . .. . 6 6 .8 6 5.8 6 1.7

positive effect on sink intensity. The promotion of the rate and duration of sucrose
uptake by grains in the K.- and K.,-treatrents could have several reasons. Better K-
flUtrition of plants could have resulted in better assimilate supply to the ears during
anthesis and shortly after, when the number of endosperni cells is determined ( Tolle-
naar [26]). Consequently a higher sink capacity could have been established, being
then the cause for the higher "'C sucrose uptake by detached ears. Also K- and K,-

187
 plants could have had a balance of phytohormones favouring sink intensity. This
assumption is supported by determinations of abscisic acid (ABA) content (see later).
Enhanced sink intensity could also be due to morphological alteration. In comparisons
of high-yielding wheat varieties with old cultivars Pommer [18] noticed increased
cross sections of conducting vessels in the ear rachis.
A higher demand of grains for photosynthates is correlated with a higher rate of starch
synthesis in the grains (Jenner and Ratlijen [11]). In our experiments 14C in the
grains was fractionated into starch, sugar, organic acids and amino acids. In general,
K enhanced the rate of 14 C incorporation into the starch fraction (Figure 2). But the
differences between treatments were only significant at day 21 after ear emergence
and at day 35, when grain filling generally ceased. K also stimulated 14C incorporation
into organic acids and amino acids. Accordingly it can be assumed, that grain metab-
olism in total had been stimulated by K.
The decrease of 14C incorporation into starch during grain development was not
4
caused by a decreasing sugar supply to grains. Until day 28 after ear emergence ' C
labeling of the sugar fractions stayed at the same level while labeling of the starch frac-

900

800

700

60G Amino acids -

Organic a

a 500 Sugar

coc
F400

300 Starch

200
K1, K K3
100 -
.

14 21 28 35
Days after ear emergence

Fig. 2. " C distribution in barley grains. Ears were treated as described in the legend of
Figure 1. Values were obtained by extraction of grains with ethanol and subsequently with
HC. ETOH extract was separated by column chromatography. Each value represents an
average of four measurements. The vertical bars represent LSD 5% for three K treatments
each.

188
tion decreased very rapidly. Jenner and Ra/hjen [ II] suggested that grain filling in ears
is not limited by sucrose supply but by the conversion rate of sucrose into starch, This
suggestion might be confined to earlier stages of grain filling. At the final stage, when
most of the leaves are becoming senescent, the remaining green leaf area could become
yield-limiting (Spieriz [22]). In our experiment this occurred at day 35 after ear
emergence (Figure 2). At that time the supply of "C sucrose into the grains was
generally very small. But in comparison to K, plants, grains of K, and K3 plants
accumulated about three and five times more I C.
It was proved by several experiments, that K stimulates the activity of starch synthe-
sizing enzymes (Evans and Wilde [I], Hawker et al. [8], Hawker et al. [9]). Murata
and Akazawa [15] could enhance the activity of starch synthase in barley and rice
seeds for about 30' by increasing the K concentration in the incubation medium
from zero to 100 milli molar (Table 2). The activity of starch synhase in grains of
wheat plants with different K status was studied by us over the whole grain filling
period. There was no clear effect of K nutrition on the activity of starch synthase
extracted from grains. Despite different K contents in the straw at anthesis (K, = 3.0;
K2 - 10.3 and K,= 23.0 mg K per g dry matter) the content in the grains did not differ
very much (8.1 : 8.4 and 8.8 mg K per g dry matter respectively). The small differences
in K content of grains could be the reason for the lack of response in enzyme activity
to K-treatments. But parallel to the in vitro determination of starch synthase starch
content in the grains was measured. Here a definite K effect on starch accumulation
was observed (Table 3). From the beginning of starch accumulation in grains differences
in starch content per grain were visible between the K-treatments. At full maturity of
K, plants (August 2nd) the grains of K, and K., plants had accumulated about three
times as much starch than grains of K, plants. This result supports the suggestion
mentioned above, that increasing K supply enhances sink intensity and sink duration.

Table 2. Effect of K concentration in the incubation medium on starch synthase activity in

barley and rice seeds (Miraw. and Akazawa f1 /969])

KCI applied AIP formed relative

(M) (n moles) (%)
Barley ....................................... 0 53.4 1t0
01 72.3 135
R ice ..................... ................... 0 37.5 100
0.1 51.1 136

Table 3. AmoLnt of starch in wheat grains. The grains were collected from plants with different

K status
Dale of harvest K, K, K,

ing starch/grain
June 22 ... ...................................... 1.0 I9 3.2
Ju nc 26 . .......... ....... ........ . ................... 2.1 4 .7 4 .5
July 7 .......... ............... ... .................... . 8.4 14 .5 15.2
Ju ly 20 ................................................... 9.8 20 .6 23.2
July 27 .................... ............................. [1.5 25.4 28.7
A ug ust 2 ................................................. 1.2 29 .6 32.1

189
3. K-effect on sink duration

Duration of grain filling seems to be more important for final grain size than sink
intensity (Spiertz [21], Stcy [24]). The effect of K nutrition upon duration of grain
filling has been observed in several field and pot experiments (Peaslee [17], Forster
[2]). ]n our pot experiment with wheat the duration of grain filling was extended by
K nutrition for 20 and 27 days respectively in comparison to K, (Table 4). The pro-
longed filling resulted from both earlier anthesis (4 days) and retarded maturity
(16 days and 23 days). Accordingly single grain weight was doubled. The question
arises, if the prolonged filling period was caused by a direct effect of K on sink or
primarily an effect on source, which influenced sink. An answer can only be given by
measuring metabolic compounds, which are involved in the regulation of senescence
in the grain itself. Abscisic acid (ABA) is involved in the regulation of senescence in
grains (Goldbach and Michael [6]). The initiation of senescence in grains is accom-
panied by a sharp increase in ABA content for a short time. Then ABA content drops
again very rapidly (Radley [19]. The ABA peak seems therefore to be a suitable
indicator for the onset of ripening. The amounts of ABA in grains of the same wheat
experiment were measured during the whole filling period.

Table 4. Duration of grain filling period in wheat plants and single grain weight influenced b)
K nutrition

K, K, K3
Date of anthesis ......................... June 16 June 12 June 12
Date of maturity ......................... July 30 August 15 August 22
Filling period (days) ...................... 44 64 71
Single grain weight (ng DM) .............. 19.0 40.3 43.2

In the first three weeks after anthesis the amount of ABA in the grains increased slightly
and similarly in all K treatments (Figure 3). But then the K status of the plants had a
significant influence on further increase in ABA content. The better the K supply of
the plants, the later the grains reached their ABA maximum. Additionally this ABA
maximum was lower, the better the K-nutrition. This is clear evidence that K pro-
longed sink duration in the grains. In all K treatments the appearance of the ABA peak
nearly coincides with the decline in linear grain growth (Figure 4), indicating the end
of grain filling or the beginning of water loss of grains respectively.
The question remains open as to what might be the link between K nutrition and ABA
content in grains. At first a connection between ABA content and dry matter content
in the grains could be assumed, because a correlation between decreasing water content
and increasing ABA content in plant tissues is known (Goldbach and Goldbach [5]).
But during the whole grain filling period dry matter content in grains of K, plants was
always lower than dry matter content in grains of K, plants (Figure 5). With ABA
content it was the reverse. Accordingly it might not be only dry matter content of
grains which affects ABA content. Probably products of chlorophyll catabolism in
leaves and culms have contributed to the increasing ABA content in grains (A'aqvi
and Engvild [16], Rasehke et al. [20]). Chlorophyll content of the upper internodes

190
 K,

14, :1
12 K2

10 K
"

C K3

20 30: 4,"2
0~

2 .~K

20
June July
F J Lne
at 3.
4. Conrtents
2: Full ofnritry
absisic atacid
Julyin 30
g (K),
rains
ofAugust
wheat during
15 (K 2)grain filling period.
and August First
22 (K,). anthesis
Values are
averages of three i easiuretents. Te vertical bars r resent standard errors of the values.

40 / ..... o K 9
,
I .

C -
30,

a,3 .....

0 0. . . .. 828N
JuneJu-
F .LI u . c r as o f s i*l g r ifeg t 'n
g a n [ w e t d r n r i i l n e i d i e e d
oK
n ce n of
ut ritio n F u rIh e r c o m n le nd o Fi re .

V9
 60 K3
K2
50 50.
.. q
-Ka
40 ... -

30...

20

10

0
20 30- 4 18 28
June July

Fig. 5. Increase of dry matter content in grains of wheat during grain filling period. Grains
were obtained from plants with different K nutrition. Further comment in legend of Figure 3.

for instance, declined in the same period, when ABA content in grains increased
(Table 5). This correlation suggests, that sink duration might also be affected by the
metabolism in the leaves. It is premature to assign a sink controlling function exclu-
sively to ABA. More likely the balance of phytohormones has to be considered. It is
known from nutrient solution experiments, that K nutrition keeps roots more vigorous,

Table 5. Decline of chlorophyll content in the upper internode of wheat plants during the
final stage of grain filling
Date of harvest K, K, K,
mg Chl./g FW
July 20 ................................................... 0.7 2.4 2.5
July 25 ................................................... 1.2 1.3 1.3
July 27 ................................................... - 0.8 1.2
July 31 ................................................... - 0.5 0.6
A ugust 2 .................................................. - 0.1 0.4

even during the final stages of ripening (Mengel and Haeder [13]). During this
period root tips are the only meristematic tissues in the plant and probably the major
site of cytokinin synthesis, This phytohormone is assumed to stimulate grain filling
(see Michael, p. 35).
In summary the results reported demonstrate favourable effects of K on important
processes in yield formation.

192
4. References

I. .vans, H.J. and Wi/ldes, R. A.: Potassium and its role in enzyme activation. Pro:. 8th
Coll. Internal. Potash Inst., pp. 13-39 (1971)
2. Forster, H.: Einfliuss der Kali umernfi hru ng allf Ausbildung und Chlorophylg halt des
Fahnenblattes und auf die Kornertragskomponenten von Somn rweizen. Z. Acker- und
Pflanzenbau 143. pp. 169-178 (1976)
3, Fuchs. PP.: UnicrsuchUngen zum Einfluss der Stickstoffdingtng aLlf(die Anlage und A.iS-
bildung des Ertragsmerkmals Ahrchenzahl je Ahre be Winterroggen. Winterweizen und
zweizeiliger Gerste. Arch. f. Acker- u. Pffanzenbau U. Bodenkunde 19, pp, 227-286 (1975)
4, Gifford. R. )W.. Brener. P. Xf. and dones. D. B.: Assessing photosynthetic limitation to
grain yield in a field crop. Aust. J. agric. Res. 24. pp. 297-307 (1973)
5. Goodhath, H. and Goldbct. E.: Abscisic acid translocation and influence of water stress
on grain abscisic acid content. J. exp. Bot. 28, pp. 1342-1350 11977)
6. Goldhach, H. and Michael. G.: Abscisic acid content of barley grains during ripening as
affected by temperature and variety. Crop. Sci. 16. pp. 797-79o) (1976)
7. Haeder. Hl. E., Meinel. K. and Fmrsier, I-.: The effect of potassifflm on translocation of
photosynthates and yield pattern of potato plants. J. Sci. Fd. Agric. 24, pp. 1479-1487
(1973)
8. Hawker. J. S.. arschner 1. and Downln, PP J. S.: Effects of sodiurn and potassium on
starch synthesis in leaves. Aust. J. Plant Physiol. /, pp. 49t-501 (1974)
9. Ha,ker, J. S.. Wa,rschner, H. and Kraiess. A.: Starch synthesis in developing potato
tubers. Physiol. Plant. 46. pp. 25-30 (1979)
10. lenner, C. F. and Ratljel, A.J.: Factors limiting the suIpply of sucrose to the developing
wheat grain. Ann. Bot. 36, pp, 729-741 (1972)
II. Je,n,er, C. /- and Rathen, A.J,: Supply of sucrose and its metabolism in developing
grains of wheat. Aust. J. Plant Physiol. 4. pp. 691-701 (1977)
I 2. Menge/. K. and Hceder. H. E.: Photosynthese und Assimilat-transport bci WeiLen wahrend
der Kornausbildung bei Unterschiedlicher Kalitimernahrung. Z. Acker- und Pflanzenbau
140. pp. 206-213 1974)
13. VIe,n,'el. K. and Haeder. H. E.: The efect of potassitutm and light intensity onl the grain
yield production of spring wheat. Proc. 4th Iniernat. Coll. oil Control of Plant Nutr,,
pp. 463-475 (1976)
14. Mlea eo. K. and Ilecder, H. E.: Effect of potassiuml supply on the rate of phloem sap
eXudation and the Comllposition of phloein. saip of Ri(inus cmmO nono, Plant Physiol. 59,
pp. 282-284 (1977)
15, Aurat, T. ani Aku,,,u. T.. StimUlation effect of potassitin ion on starch synlhetase
of different plant origins. Plant Cell Physiol. /0. 457-460 (1969)
16. Naqi, S. M. and Eigvi/. K. C.: Action of abscisic acid on auxin transport and its relation
to phototropisn. Physiol. Plant. 30, pp. 283-287 (1974)
17. Peaslee, 1. E.: Effects of nitrogen. phosph orus and potassiLnl nutrition on yield, rates
of kernel growd and grain filling periods of two corn hybrids. Soil Sci. Plant Anal. 8,
pp. 373-389 (1977)
18. Pommet, G.: Ertragsphysiologie und ZuchIung. Bayer. Landw. Jahrb. 54, pp. 290-302
(1977)
19. Rade', X1.: The development of wheat grain in relation to endogenous growth sufbstances.
.J. exp. Bot. 27, pp. 1009-1021 (1976)
20, Rasehke, K., Firm,. R. 1). and Pierce, NI.: Stomatal closure in cesponse to xanthoxin and
abscisic acid. Vlanta 125, 149-160 (1975)
21. Spiert. J. H.J.: Grain growth and distribution of dry matter in the wheat plant as
inflUenced by temperature, light energy and car size. Neth. J. agric. Sci. 22. pp. 207-220
(1974)
22. Spiertz. J. iJ.: "The influence of temperature and light intensity on grain growth in
relation to the carbohydrate and nitrogen economy of the wheat plant. Neth. J. agric. Sci.
25, pp. 182-197 (1977)
23. Stat'. V,: Source and sink properties as related to yield in dillerent barley genotypes.
Proc. 3th Internal. Barley Genet. Synp. pp. 641-648 (1975)
24. Sto, V,: Illanzenphysiologische Merkmale as Selektionskritcrien in der Zuchtung auf
Ertrag bei Getreide. Ber. Arbeitstag. Saatichti pp. I 1-29. Gtimpenstein. 1976

193
25. Stoy, V.: Trockensubstanzproduktion und Assimilateinlagerung in das Getreidekorn.
Z. Pflanzcnernihr. Bodenkunde 140, pp. 35-40 (1977)
26. Tollet,aar, M.: Sink-source relationships during reproductive development in maize.
A review. Maydica 22, pp. 4 9 - 7 5 (1977)

194
Regulation of Starch Synthesis in Leaves
The Role of Orthophosphate
D.A. Walker, ARC Research Group on Photosynthesis. Department of Botany, University
of Sheffield/England*

Sw loI.y

I) The isolated chloroplast is a Pi-consuming, triose phosphate exporting organeile.

2) The rate of Pi recycling within the stronia occasioned by starch synthesis appears to be
too slow to support maximl,l photosynthesis.
3) When isolated chloroplasts are illuminated in Pi-deficient mediurn, build-up of triose
phosphate and PGA in the stronia coincides with allosteric activation of ADIlglucose
pyrophosphorylase and starch synthesis is favoured.
4) When isolated chloroplasts are iluaninated in high external [Pi]. triose phosphate export
is enforced. ADPglucose pyrophosphorylase activity is diminished and starch synthesis is
depressed.
5) Manipulation of Pi concentrations within the leaf by mannose feeding suggests that Pi
concentration may play a similarly important role in determining the extent of starch
svnt hesis ill
iwo.

I. Introduction

Chloroplasts capable of rapid rates of CO, fixation (Walker [29]: Backe et at. [2]:
Jenson and Basshant [18]) and CO-dependent 0, evolttion (Walker and Hill [33])
were first isolated from C3 plants in the 1960's. Their availability has led to the
development of two important and entirely unforeseen concepts. The first is that the
chloroplast is not a self-sufficient entity capable of carbon assimilation according to
the classic overall equation
CO + H,20- CH_,O + 0, .. (I).
Instead, it is a Pi-consuming, triose phosphate exporting organelle ( Walker and Herold
[35]: Walker and Robisxon [36]) and both its ability to assimilate CO 2 and transport
elaborated carbon to the cytoplasm (Figure I) are better expressed by the relationship

3CO, + 2H,O + Pi -- triose phosphate + 302... (2).

The second, related, concept is that Pi does not enter the chloroplast in a free and
Uncontrolled fashion bit only via the Pi translocator (Hetdh [12], Walker and Crafts
r34]) which mediates an obligatory stoichionietric exchange between Pi and certain

D.A. Walker, Professor of Biology. Deparanicnt of Botany, University of Sheffield, Sheffield

SIO 2TN/England

195
 Pi

02 0C02

Triose phosphate

Fig.]. The principal function of the chloroplast is to import CO, and orthophosphate and
export triose phosphate (TP). Some triose phosphate is converted within the stroma to hexose
monophosphates and starch

phosphorylated metabolites. Of these, the import of Pi and the associated export of

triose phosphate is most important in this context. It has also become clear that this
exchange has a regulatory function (Herold and Walker [16]; Walker and Herold
[35]; Walker [32]) and can affect both the rate of photosynthesis and the nature of
the photosynthetic products. In this regard there is a particularly clear correlation
between [Pi] and starch synthesis. This paper seeks to outlinethe salient features of this
relationship and, in the interest of brevity, refers the reader to a number of relevant
reviews rather than to the original literature. The entire subject is treated at length by
rLdwards and Walker [6].

2. The relationship between [Pil, and the movement of metabolites between stroma and
cytosol

In order to arrive at a fuller understanding of the manner in which orthophosphate can

influence starch synthesis (Section 4) and the distribution of fixed carbon between
starch and soluble products it is useful to examine the interrelationship between [Pi],
metabolite movement and induction.
When chloroplasts, protoplasts or intact leaves are brightly illuminated after a
relatively long period in the dark, photosynthetic carbon assimilation and its asso-
ciated 02 evolution does not start immediately at its maximum rate (for a review,
see Walker [30]). The reason for this initial lag or induction period is still a matter for

196
debate although the Osterhout and Haas hypothesis (Osterihou and Haas [23])
remains as valid a basis for discussion and experiment as it did when it was first pro-
posed more than half a century ago. This hypothesis proposes two mechanisms:
(a) the build-up of metabolites depleted in the dark. (b) the re-activation of catalysts
inactivated in the dark. In our own work with isolated chloroplasts we have been led
to believe that the former is the more important. Many enzymes of the photosynthetic
carbon cycle can be light activated (for reviews, see Laizko and Kelly [20./ and
Walker [32]) but only FBPase (of those so far studied) is present in such low activity
in the lark and displays such marked increased activity in the light that it could
obviously impose a major limitation. Even this enzyme, however, can be activated so
quickly in the light (in a matter of seconds) that it is difficult to see how it can account
for induction lags of several minutes (Leeg,ood and Walker [21]). Moreover. FBPase
in isolated chloroplasts can be pre-activated in the dark with dithiothreitol without
abolishing the lag. Conversely there is no doubt that, in isolated chloroplasts, the
lag may be shortened by small quantities of PGA or triose phosphates or that it may
be lengthened by super-optimal concentration of Pi ( Walker [31], Herol and Walker
[16]). These facts together with the reversal of the Pi-inhibition by certain cycle
intermediates (such as PGA, iriose phosphates and pentose monophosphates) led to
the notion of enforced export of cycle intermediates from the stroma in the presence
of high [Pi] ( Walker and Crolts [34]) and to the expei inents by Held! and his col-
leagues which defined the Pi translocator (Heldl [13]). These showed that there is
obligate stoichiometric exchange of Pi. DHAP and PGA across the inner of the two
chloroplast envelopes (Douce and Jovard 15]; Heldl 113]). The outer envelope is
freely permeable to metabolites but Pi will only enter the chloroplast in exchange for
an internal phosphorylated metabolite such as PGA or triose phosphate (Fliege el al.
:71). This, indeed, is regarded as the principal function of the Pi-translocator although
any of the metabolites translocated by this mechanism can, it is believed, exchange
with any other in any direction. For example, external PGA can enter the chloroplas
in exchange for internal CHAP and can. in this way, promote a shuttle which accom-
plishes indirect export of ATP to the cytoplasm (Heber [/0], and Walker and Crofts
[34/, Walker [31]). It may be noted that PGA is exported as the divalent anion
PG A '- ) and that it is for this reason that the alkaline pH of the illuminated stroma
allows tie retention of a relatively large pool of PGA' - which would otherwise be
rapidly dissipated by exchange for external Pi down a favourable concentration
gradient (Fliege el al. [7]). As already indicated, the relative contributions made to
induction by light activation of enzymes and autocatalytic build-Up of intermediates
is still disputed. For the present purpose, however, it will be assumed (a) that the
inIuction lag represents the time taken for cycle intermediates to reach their steady
state level and (b) that, even if there are circumstances invito in which light activation
takes so long or the induction period is so short, that the one is governed by the other,
the interrelationship between [Pi], the lag, and metabolite transport would still not be
materially affected. In this regard, it is important to recognise that the total level of
photosynthetic cycle metabolites within the chloroplast is less important than the
relative or partial concenrations. ThtUs RBP carboxylase is already limited by CO,
concentration in most circumstances and the total catalytic capacity of this enzyme is
not so high that the rate of carboxylation could be maintained in the presence of low
RBP concentration. Similarly, for reasons that wvill be given below (Section 6) the
cycle cannot operate tnless the ratio of PGA/pentose monophosphate is high.

197
 It will be argued, therefore, that the extension of the induction lag which is observed
when isolated chloroplasts are incubated in relatively high Pi does reflect autocatalysis
on the one hand and enforced export of triose phosphate on the other. These con-
flicting processes could in theory, come into balance. Indeed, relatively small increases
in Pi extend the lag without diminishing the final rate of photosynthesis. However,
larger quantities of Pi also depress the final rate and it might then be supposed that,
while the build-up of intermediates is fast enough to allow photosynthesis to start,
accelerating loss finally comes into balance with production at levels too low to permit
maximal photosynthesis. Secondary effects may also play a role. Thus, Antimycin A
can stimulate under such conditions (Walker [30]), possibly by relieving a back-
pressure of protons within the thylakoid compartment caused by excess cyclic electron
flow (Slovacek and Hind [27]; Slovacek et al. [28]). It is evident that non-cyclic
electron transport to NADP must ultimately be limited by re-oxidation of NADPH
and that this in turn will be related to the turn-over of the photosynthetic carbon cycle.

3. The chloroplast as a Pi-consuning organelle

As already noted in the introduction, the chloroplast is a Pi consuming organelle.
This was first demonstrated by Cockburn et al. in 1967 [4] who showed that photo-
synthesis by isolated chloroplasts soon ceased in Pi-deficient medium and that, in the
short term, there was then an evolution of three molecules of 02 for each molecule of
Pi added. This was consistent with equation 2 rather than equation 1 and, as techniques
improved, it became possible to detect the decreased rate of 02 evolution associated
with starch synthesis (Herold and Walker [16]). The latter reaction sequence does not
involve Pi consumption and data of this sort clearly indicates that a prime function of
the chloroplast is Pi import and triose phosphate export. Internal synthesis of free poly-
chloroplast is Pi import and triose phosphate export. Internal synthesis of free poly-
saccharide or free sugars is, in itself, too slow to permit photosynthesis to proceed at
maximal rates, Similarly the rate of starch synthesis in spinach and, conceivably, in
many species, is too slow to re-cycle Pi at a rate adequate to support maximal.photo-
synthesis (Herold and Walker [16]). This is not to deny that some species can
incorporate a very high proportion of total carbon assimilated into starch under
certain conditions. What it does question is whether such starch formation is ever
materially greater than about 30% of maximal photosynthesis.

4. The relationship between Pi concentration and starch synthesis

It follows from the preceding sections that if [Pi] is suboptimal theproportion of newly
fixed carbon which is incorporated into starch is likely to increase because starch and
free sugars are the only photosynthetic products which do no involve net uptake of Pi.
[It is now known, with certainty, that sucrose (which is often the major 'end-product'
of photosynthesis) is synthesised in the cytoplasm from exported triose phosphate
(Whittingham et al. [38]; Robinson and Walker [26]) so that the formation of this
sugar, although it does not contain Pi, still involves Pi import into the stroma]. Clearly,
in these circumstances, starch synthesis could occur by default because the formation
of other compounds was precluded. However, this is far from being the whole story.
Two other factors are involved. Firstly the export of triose phosphate from the

198
 Envelopes

Starch --- HP -,-- TP

Pi

G Cytoplasm

Envelopes

Starch -HP - TP
0\Mannose

Pi Mannose - P
b Cytoplasm
Fg.2. Normal pathway of photosynthetically fixed carbon and proposed diversion of this
by niannose. (a) The normal pathway in which some newly synthesized triosephosphate (TP)
is converted to starch via hexose diphosphate and monophosphate (14P) but in which there is
a considerable export of triose phosphate to the cytoplasm via the phosphate translocator.
This export is linked to orthophosphate (Pi) import. (h) In tie presence of exogenous mannose
cytoplasmic orthophosphate becomes depleted through the action of hexokinase. The resulting
nannose phosphate is not further metabolized and is itself (like other hexosephosphates and
free mannose) unable to move through the inner envelope of the chloroplast at a significant
rate. As the concentration of cytoplasnuc orthophosphate falls, the rate of exchange of ortho-
phosphate and triose phosphate is diminished and newly fixed CO. is diverted to starch
synthesis

chloroplast is diminished in the presence of low Pi (Figure 2) so that more starch

precursors are retained within the stroma than usual (Iferold and Walker [16],
Walker [32]), Secondly, ADPglucose pyrophosphorylase, a key enzyme in starch
synthesis, is allosterically regulated by Pi and PGA (Preiss and Levi [24]). The
absolute concentration of these compounds is less important in regulation than their
relative concentrations so that when te PGA/Pi is high. as it is in these circumstances.
the enzyme is 'switched-on'. In this way enzyme activation coincides with increased
availability of substrate and although photosynthesis itself decreased in low [Pi] the
absolute, as well as the relative, amount of starch synthesis can increase.

199
5. Direct and indirect evidence of Pi modulation of starch synthesis

1) Preiss and his colleagues (see e.g. Preiss and Levi [24]) have carried out extensive
measurements of ADPglucose pyrophosphorylase activity and have clearly and
elegantly demonstrated, for example, that the enzyme is activated when the PGA/
Pi ratio is high.
2) Heldt et al [14] have shown that triose phosphate export is a function of external
[Pi] and that starch synthesis is increased in isolated chloroplasts kept in low Pi
medium (Figure 3).
3) When leaf discs of certain species are floated on mannose there is a sequestration
of cytoplasmic Pi as mannose phosphate (Chen-She, el aL [3]). This results from
the combined action of phosphorylation and hexokinase
Pi 7 ADP 7K mannose phosphate
H,0 .iL ATP -IL mannose

C
.0

.total carbon fixation

E

0 '

I-

E
E
0
incorporation into starch

1 2 3
mMP
Fig.3. Starch synthesis and photosynthesis in isolated spinach chloroplasts as a function of
orthophosphate concentration (after Heldt et al. [14])

200
This sequestration of Pi depresses photosynthesis but simultaneously increases the
proportion of fixed carbon which enters starch. This effect depends on the extent to
which mannose phosphate is subsequently metabolised. In some species, in which
this further metabolism is presumably limited, the increase in starch is very marked
(Table 1). In other species, feeding with compounds such as glucosamine (which is also
phosphorylated in the presence of hexokinase) also results in substantial increases in
starch. [it may be noted that whereas iannose stimulated starch synthesis in all C3
species investigated it was inhibitory in all CA species tested (Herold el al. [ 15]). It
was concluded that the inhibitory action of Pi-sequestration on the reaction catalysed
by phosphate pyruvate dikinase was the overriding factor in C4 species].

Table I. Stimulation of starch syntlhcsis in leaves of spinach beet (Beta iucais iullivar) by
mannose
Conjira/ Nfaninose
Leaf I .................... .................... . 2.9 58.9
4.3 54.1
L eaf 2 .. ...............
... ....... ............... 5.2 52.8
4.7 58.7
L eaf 3 .... ....... ....... ........................... 5.5 55.3
2.8 74.7
Leaf 4 ................................................ 3.7 83 .4
4.1 81.5
Lea f 5 . . . .... ......... ................................ 5.1 63.3
5.1 67.0
M ean ............. ....................... ...... . 4 .3 65.0
Standard deviation ........... ......................... 0.9 11.3
Pairs of leaf discs were floated on a buffer or buffer +0.2 M mannose as indicated (after
Chen-She et al. !3_i)

Feeding with 14C labelled mannose showed that the increased starch synthesis did not
derive from direct incorporation of mannose into starch. Instead, when 321P was used,
radioactivity accumulated in mannose phosphate and it could also be demonstrated
that there was a decrease in the total quantity of free Pi which was present in the leaf
(Chen-She et ci. [3]: Herald et ci. [15]). Recent work with isolated protoplasts by
Herol and McNeil (personal conillniCation) has shown that starch synthesis is
increased and photosynthesis decreased by mannose, as in the experiments with leaf
discs, but that the chloroplasts released from the proloplasts by passage through a
nylon mesh showed unimpaired photosynthesis. All of these results are consistent with
the notion that the mannose effect is indirect and that it increases starch by lowering
cytoplasmic Pi.

6. Source sink relationships and Pi modulation

The internal factors which limit photosynthetic carbon assimilation are not known
with any certainty. RBP carboxylase continues to be regarded as one of the most
likely limiting factors although its maxilnum potential (in high CO,) in spinach

201
 exceeds the average rate of photosynthesis in air by an order of magnitude and its
characteristics (Akazawa [1]) no longer seem inadequate in relation to the CO, con-
centration which is likely to obtain in the chloroplast, (given a Vmax of 600 and a
Km(CO2 ) of i I M only 52 ppm CO, would be needed to support a rate of 100 tvmoles
CO2 fixed mg - 1 chlorophyll hr-'). Conversely the maximum rate of coupled non-cyclic
electron flow often brings about the evolution of 02 at little more than 250 [imoles.
mg - 1 chlorophyll hr - 1, a rate which is not often exceeded by the average C3 leaf even
in augmented CO,. Facts such as these suggest that the upper limits of photosynthesis
may be fixed by electron transport capacity and certainly there is little doubt that at
normal temperatures photosynthesis is often limited by the available light intensity.
If this sets the upper limit what dictates actual performance within the limit? That
distant sinks can affect photosynthesis now seems beyond reasonable doubt (Geiger
[9]; Habeshaw [11]; Humphries and Thorne [J7]; King et al. [19]). The fact that the
starch content of leaves responds to all manner of factors, including sink activity,
excision etc. is a matter of everyday observation (see e.g. Herold and Walker [16]).
How might such distant events make their occurrence felt within the chloroplast?
Again, Pi concentration, has been postulated as a controlling mechanism. According
to this hypothesis photosynthetic carbon assimilation can only reach the maximum
rate allowed by light intensity when Pi import and triose phosphate export are also
maximal. Certainly, if internal recycling of Pi is entirely dependent on starch synthesis
and if the rate of starch synthesis itself cannot equal the maximum rate of photo-
synthesis then carbon assimilation must be limited by the transport requirements
implied by equation 2. This indicates that photosynthesis will be maximal when the
availability ofPi (supplied by internal and external recycling of Pi) is exactly in balance
with triose phosphate production. If the external [Pi] became too great, optimal
regeneration of the CO2 acceptor would be slowed by enforced export of triose
phosphate. Evidence exists which suggest that triose phosphate is probably exported
against a concentration gradient (Giersch et al. [8]) and this, in turn implies a need
to maintain a high, but not too high, downhill gradient of Pi from cytosol to stroma.
If Pi could always enter the cytosol from the vacuole in an uncontrolled fashion this
control would be lost and it is hard to see how high rates of photosynthetic carbon
assimilation could be maintained. If, however, the [Pi] in the cytoplasm is maintained,
within reasonable limits, the photosynthetic rate will be a function of the [triose
phosphate]/[Pi] ratio within the cytosol. (The cytoplasmic concentrations of triose-
phosphate and PGA may well be as important in this regard as that of Pi because
all of these compounds act as competitive inhibitors of one another in respect to
the Pi translocator)- In this way cytoplasmic recycling of Pi is probably very important.
This is usually achieved by sucrose synthesis (Figure 4) which consumes triose
phosphate and, in so doing, releases Pi which can re-enter the stroma (Herold and
Walker [16]; Walker and Robinson [36]; Walker and Robinson [37]). If sucrose
phosphate synthase and sucrose phosphate phosphatase are inhibited by free sucrose,
as has been suggested (see e.g. Whittinghain et al. [38]) sink activity could make itself
felt by ensuring optim3l [triose phosphate]/[Pi] ratios. Active sinks would promote
optimal ratios. Inactive sinks would slow photosynthesis (and favour starch synthesis)
by incurring high ratios of [triose phosphate]/[Pi].
As before, it would be reasonable to enquire how the RPP pathway (as distinct from
ADPglucose pyrophosphorylase - Section 4) 'recognises' the fact that it is attempting
to operate in a Pi deficient situation. Orthophosphate enters the RPP pathway via

202
 Chlorop/ost
en.elopes
Stromo

-p CO,
GIP -.- GP.F6P
ATPA
JC P

A DPG FBP-_-TP P~
2 1p(
t s - DP- ATPP
Pu P+
uDP

PI
cyoplsm UTP ,' ADP
CyrplsmS.ctose-P + UDP> : ATP,~Y

Pi - - Feedback
Sco. -
. ihhbifion

4
TRANSLOCA77ON

F,.4. Pathways of starch and sucrose synthesis in leaves indicating their role in re-cycling of
Pi and tihe coosequent possibility of distant sinks exerting a controlling On the rate of photo-
synthesis and partitioning between starch and slcrose.

photosynthetic phosphorylation but this process has such a high affinity for Pi that
it is indifferent to [Pi] until this reaches very low levels. Conversely, the reduction of
PGA to triose phosphate is extremely sensitive to [ATP]/[A[DPI ratios and, therefore,
readily inhibited by A DP (Lillcy and IWalker [22], Robinson and Walker /25]), Clearly
once the [Pi] concentration falls below the level needed to maintain an appropriate
portion of the total adenylate as ATP the reduction of PGA (and therefore the operation
of the entire cycle) coUld be affected (Robinson and Walker [25]; Walker and Robinson
[36]).
When spinach is actively growing under favourablc conditions the starch that accu-
inUlates by day is more or less completely remobilisect by night (Hero/d and Walker
[16]). Viewed in this context, transitory starch is seen as excess of production over
day-light consumption. If. for whatever reason, sink actively is depressed there wvill be
some tendency for photosvnthesis to fall bit also accumulation will tend to exceed
total consUmption and leaf starch will not be fully remobilised by night. This appears
to happen in spinach, for example, after floral initiation or when growth is ctArtailed
by deficiency or disease.

7. Starch degradation

In general, starch in storage tisstIs is believed to be degraded by the combined action

of' several enzymes (Preiss and Levi 124]). Starch phosphorylase catalyses the
phosphorolysis of a-I,4 glucosyl chains yielding glucose-I-phosphate
a-I ,4 (glucosyl) + Ph = x-I,4 (g,LOSYI), + glucose-l-phosphate

203
m-amylase attacks a-1,4 linkages of amylose or amylopectin. P-amylase attacks only
the non-reducing ends of chains, splitting of pairs of glucosyl units as maltose. R-enzyme
attacks the 1,6 branching points of amylopectin making this component of starch
susceptible to further attack by P-amylase of phosphorylase. Disproportionating
enzyme can transfer groups of glucosyl units between short-chain dextrins producing
a mixture of longer and shorter chain dextrins and glucose.
In leaves, it seems likely that P-amylase is cytoplasmic and may only be concerned with
starch degradation of senescent chloroplasts or those which have taken on the
characteristics of amyloplasts. A stromal endoamylase which has somewhat different
characteristics to other plant a-amylases (having no requirement for Ca++ and lacking
heat stability) is believed to initiate starch degradation in chloroplasts (Priess and
Levi [24]). With the help of R-enzymes and D-enzymes (Figure 5) this produces
substrates which readily undergo phosphorolysis. Glucose-I-phosphate is then con-
verted to triose phosphates via the following sequence which demands the availability
of phosphofructokinase activity.

'/ Ia'

-amOt=e a-l,4-(Gluc). Enye a-1,4 :a-1,6-(luc),

04
-( lu)

00 ' G-6P

GIP
qI u C +GP

FbP g - G3 P

Fig.5. Some aspects of starch degradation in leaves and the nature of the degradation products
released from the chloroplast (after Preiss and Levi [24])

glucose-I -phosphate -glucose-6-phosphate - fructose-6-phosphate -

/ glyceraldehyde-3-phosphate
fructose-I,6-bisphosphate \ dihydroxyacetonephosphate

204
These triose phosphates can, of course, be exported fromt the chloroplast via the Pi
translocator. High Ii favours degradation in three ways. (a) by initiating phosphor-
olysis, (b) by inactivating ADPglucose pyrophosphorylase (by providing a low [PGA]/
[Pi] ratio) and (c) by accelerating triose phosphate export. Kraminer (personal corn-
munication) has, in tact, shown that an increase in orthophosphate concentration will
switch synthesis to degradation in illuminated chloroplasts.

8. Concluding comment

The relationship between starch synthesis in isolated chloroplast and external [Pi]
seems remarkably clear. Low Pt favours retention of triose phosphate within the
stroma (Figure 6) and concomitant allosteric activation of ADPglucose pyrophos-
phorylase. Thus the availability of substrate coincides with activation of the enzyme
and starch synthesis is favoured.

/ndue/on Low external P1"

Envelopes
neady toHigh external P

FiI6. Sunmry of regulatory etfects of ti.

"To, hlf" durng induction molest of1the triose phosphate product is fed back into thle cycle or
expor'ted itnto the cytosol -
Hoytn,,t let!/t in the stead.'' state. unlder optimal conditions thle chloroplast tmainly imports Pi
and exports trnose phosphatce-
To1p ti,: when the ex tern at (or cytosol ic)t Pi is low. nore of"the carbon t raflic flows in to starch
and this is favoured by the concooxit ant activator of AD P glucose pyrophosphorylase. Low
Pi slows ilhotosynt hsis by aticti ng the AT iA D rat o (section 6).
ouottlt: external [Pi] inhibits High photosynthesis and starch synthesis because of
enforced export of trnose phosphate via thle i t rans] ocator and deactivat ion of A DP glucose
prophosplho ryla se

205
Currently, attempts are being made to relate these observations to the infinitely more
complicated situation which obtains within the intact leaf and the intact plant.
Manipulation of cytoplasmic [Pi] by mannose feeding suggests that the concentration
of Pi in the cytoplasm may be as important (in its effect on the rate of photosynthesis
and the proportion of carbon which enters starch) as that in the medium surrounding
isolated chloroplasts.

9. References

1. Akazawa, T.: Ribulose-1,5-Bisphosphate Carboxylase. In: Photosynthesis II. Encyclopedia

of Plant Physiology. New Series, Vol. 6. (Eds. M.Gibbs and E.Latzko), pp. 2 08-229.
Springer, Berlin, Heidelberg, New York, 1979
2. Burcke, C., Walker, D.A. and Baldry, C. W.: Some effects of sugars and sugar phosphates
on carbon dioxide fixation by isolated chloroplasts. Biochem. J. 101, pp. 636-641 (1966)
3. Chen-She, S.H., Lewis, D.H. and Walker, D.A.: Stimulation of photosynthetic starch
formation by sequestration of cytoplasmic orthophosphate. New Phytol. 74, pp. 383-392
(1975)
4. Cockburn, W., Baldry, C. W. and Walker, D.A.: Oxygen evolution by isolated chloroplasts
with carbon dioxide as the hydrogen acceptor. A requirement for orthophosphate or
pyrophosphate. Biochini. Biophys. Acta. 131, pp. 594-596 (1967)
5. Douce, R. and Joyard, J.: Structure and function of the plastid envelope. In: Advances in
Botanical Research. Vol. 7. (Ed: H. Woolhouse), pp. 1-116. Academic Press, London,
1978
6. Edwards, G.E. and Walker, D.A.: C3, C4. Some Aspects of Photosynthetic Carbon
Assimilation. Packard Publishing Ltd., Chichester, 1980 in press.
7. Fliege, R., Flugge, U.I., Werdan, K. and Heldt, H. W.: Specific transport of inorganic
phosphate, 3-phosphoglycerate and triose phosphates across the inner membrane of the
envelope in spinach chloroplasts. Biochim. Biophys. Acta. 502, pp. 232-247 (1978).
8. Giersch, Ch., Heber, U., Kaiser, G., Walker, D.A. and Robinson, S.P.: Intracellular
metabolite gradients and flow of carbon during photosynthesis of leaf protoplasts. Archives
of Biochem. & Biophys. Submitted (1980).
9. Geiger, D.R.: Effect of translocation and assimilate demand on photosynthesis. Canadian
Journal of Botany, 54, pp. 2337-2345 (1976)
10. Heber, U.: Metabolite exchange between chloroplasts and cytoplasm. Ann. Rev. Plant
Physiol. 25, pp. 393-421 (1974)
II. Habeshaw, D.: Translocation and the control of photosynthesis in sugar beet, Planta,
Berlin, 110, pp. 213-226 (1973)
12. Heidi, H. W.: Metabolite carriers of chloroplasts. In: Encyclopedia of Plant Physiology,
Vol. 3: Intracellular Transport and Exchange Processes. (Eds. U. Heber and R.Slocking),
pp. 137-143. Springer, Heidelberg (1976)
13. Heldt, H. W.: Metabolite transport in intact spinach chloroplasts. In: Topics in Photo-
synthesis, Vol. . The Intact Chloroplast. (Ed. J. Barber), pp. 215-234. Elsevier, Amster-
dam, 1976
14. Heldi, H. W., Chon, C.J., Maronde, D., Herold, A., Stankovic, Z.A., Walker, D.A.,
Kraininer, A., Kirk, M.A. and Heber, U.: The role of orthophosphate and other factors
in the regulation of starch formation in leaves and isolated chloroplasts. Plant Physiol.
59, pp. 1145-1155 (1977)
15. Herold, A., Lewis, D. H. and Walker, D.A.: Sequestration of cytoplasmic orthophosphate
by mannose and its differential effect on photosynthetic starch synthesis in C3 and C4
species. New Phytol. 76, pp. 397-407 (1976)
16. Herold, A. and Walker, D.A.: Transport across chloroplast envelopes - the role of phos-
phate. In: Membrane Transport in Biology, Vol. 1. (Eds. G.Giebisch, D. C. Tosteson and
H. H. Ussing), pp. 411-439. Springer, Heidelberg, New York, 1979
17. Humphries, E. C. and Thorne, G.N.: The effect of root formation on photosynthesis of
detached leaves. Annals of Botany, London, 28, pp. 391-400 (1964)

206
18. Jenson, R.G. and Bassham, J.A.: Photosynthesis by isolated chloroplasts. Proc. Nat.
Acad. Sci. US 56, Pp. 1095-1 l01 (1966)
19. King, R. W., Wardlw., 1. F. and Evans. L. T: Effect of assimilate utilization on photo-
synthetic rate in wheat. Planta, Berlin, 77, pp. 261-276 (1967)
20. Lalzko, E. and Kelly. G.J.: Enzymes of the Redtctive Pentosc Phosphate Cycle. fi:
Photosynthesis It. Encyclopedia of Plant Physiology. New Series, Vol. 6. (Eds. M. Gihbs
and E.Latzko). pp. 239-250. Springer, Berlin, Heidelberg, New York, 1979
21. Leegood, R.C. and Walker, D.A.: Autocatalysis and light activation of enzymes in relation
to photosynthetic induction in wheat chloroplasts. Archives of Biochern & Biophys.,
200,pp. 575-582 (1980)
22. Lillet, R. A4cC. and Walker. D.A.: The reduction of 3-phosphoglyccrate by reconstituted
chloroplasts and by chloroplast extracts. Biochim. Biophys. Acta, 368, pp. 269-278 (1974)
23. Os/erbot. W.J. V. and 1Iaas,A. R.C.: On the dynamics of photosynthesis. J. Gen. Physiol. 1,
pp. 1-16 (1919)
24. Pries. J.and Levi. C.: Metabolism of Primary Products of Photosynthesis. Metabolism
of Starch in ILeaves, In: Photosynthesis I. Encyclopedia of Plant Physiology. New Series,
Vol. 6. (Eds. M. Gibbs and E,Latzko). pp. 282-312. Springer, Heidelberg, New York. 1979
25. Robinson, S. P. and Walker, D. A.: "Thccontrol of 3-phosphoglycerate reduction in isolated
chloroplasts by the concentrations of ATP, ADP and 3-phosphoglycerate. Biochim.
Biophys. Acta. 545, pp. 528-536 (1979)
26. Robinson ,S.P. and Walker. D.A.: The site of sucrose synthesis in isolated leaf protoplasts.
FEBS Letters 107. 295-299 (1979)
27. Slovacek. R.E. and Hind. G.: Intluence of Antimycin A and uncoutplers on anaerobic
photosynthesis in isolated chloroplasts. Plant Physiol. 60. pp. 538-542 (1971)
28. Slovacek. R.E., Mills, J.D. and Hind. G.: The function of cycle electron transport in
photosynthesis. FEBS ILetcrs 87, pp. 73-76 (1978)
29. Walker, D.A.: Correlation between photosynthetic activity and metnbrane integrity in
isolated pea chloroplasts. Plant Physiol 40. pp. 1157-1161 (1965)
30. Walker, D.A.: C' 2 fixation by intact chloroplasts: photosynthetic induction and its
relation to transport phenomena and control meclanistns. In: The Intact Chloroplast.
Chapter 7, (ed. J. Barber) pp. 235-278. Elsevier, Amsterdam, 1976
31. WalkAer, D.A.: Plastids and intracepllar transport. tt: Transport in Plants III. Encyclo-
pedia of Plant Physiology. New Series. Vol. 3. (Eds. C. R.Stocking and U. ieer).
pp. 85-136. Springer, Berlin. Heidelberg. New York, 1976
32. Walker, D.A.: Regulatory mechanisms in photosynthetic carbon netabolism. hf: Current
Topics in Cellular Regulation. Vo;lti1e II(eds. B.L.Horecker and E.Stadanan)n , pp. 203-
241. Academic Press, New York, 1976
33. Walker, D.A. and Hll. R.: The relation of oxygen evolution to carbon assimilation with
isolated chloroplasis. Biochim. Biophys. Acia, 131, pp. 330-338 (1967)
34. Walker, D,A.and CrofT., A.R.: Photosynthesis. Ann. Rev. Biochen. 39. pp. 389-428
(1970)
35. Walker, 1). A. and Herold. A,: Can the chloroplast support photosynthesis unaided? In:
Photosynthetic Organclfcs: StructuLre and Function. (Eds. Y. Flji/t. S.Karoli, K. S/hina
and S. Mioachi). Special IssuC of Plant and Cell Physiology, pp. 295-310 (1977)
36. Walker, D.A. and Robinson., S.P.: Regulation of Photosynthetic Carbon Assimilation.
In: Photosynthetic Carbon Assimilation. Basic Life Sciences. Vol. 1/. Proceedings of a
Symposium held at Brookhaven National Laboratory. Upton. New York. May 31-
June 2. 1978. (Eds. H. W.Siegehnan and G. Himd). pp. 43-59. Plenun Press, New York.
1978
37. Walker,. D.A. and Robinson, S. P.: Ch loruplast and Cll. A contemporary view of photo-
synthetic carbon assimilation. (presented at the 'Botaniker-lagung' in Marbtirg, Sept. 13.
1978). Ber. Deutsch. Bot. Gcs. Bd. 91 pp. 513-526 (1978)
38. Whittingluim. C.P., Key, AJ. and Bird. I. F.: The Enzvmology of Sucrose Synthesis in
Leaves. hi: Photosynthesis [I. Encyclopedia of Plant Physiology, New Series. Vol. 6 (Eds.
M.Gibbs and F. Latzko), pp. 313-326. Springer. Berlin. Heidelberg. New York, 1979

207
Co-ordinator's Report on the 3rd Session
K. Mengel, Institute of Plant Nutrition. Justus-Liebig-University, Giessen/Federal Republic
of Germany

The Role of Nutrients in Yield Formation

This is very a broad topic and could not be completely covered by the speakers who
concentrated on some recent aspects. Prof Beringer commented on the role of
nutrients in forming yield and also referred to the effects of other endogenous factors
such as light intensity, day length and temperature. emphasizing that these endogenous
factors may affect plant development and yield production via phytohormones and
phytochromes. The causal relationship between such endogenous factors and bio-
chemical and biophysical processes is still obscure, but there is evidence that endoge-
notis factors may establish ionic gradients, affect membrane permeability and induce
the synthesis of phytohormones. Prof. Beringcr especially stressed the finding that
some of the effects of plant nutrients on plant development and crop production may
be indirect by influencing phytohormone synthesis and turnover.
This important aspect of the effect of plant nutrients was also emphasized by the
following speakers. Dr. Haeder presented research data showing that suboptimurm K+
supply resulted in increased ABA levels in wheat plants. To my knowledge this is the
first experimental evidence of a relationship between K nutrition and ABA. Dr. Haeder
suggested that the poor grain filling of the crop with a suboplimum K supply could be
related to the raised ABA levels in these plants. Further research, however, is needed
to support this suggestion.
Dr. Krauss provided evidence that there is a close relationship between nitrogen
nutrition and the GA-and ABA levels in potato tuber. Both phy(ohormones and the
ratio between them are crucial for tuber initiation and tuber growth. The direct
response of the phytohormone level to a change in nitrogen supply is a strong argument
for a direct causal relationship between nitrogen metabolism and phytohormonc turn
over. A possible sequence of factors and processes could be as follows:
External factor--phytohormone synthesis-,
gene activation--enzyme synthesis- netabolism
The experimental verification of such relationships is difficult to achieve, due to the
Fact that in working with entire plants, alteration of one factor does not only affect one
particular metabolic process but may also affect other processes. This dilemma was
raised by Dr. Woolhouse in the discussion.

* Prof. Dr. K. Menti'e/, Institut farlPlfanzenernah rung,J ustus-Liebig-Uni versitAt,Sudanlage 6,

D-63 Giessen/Federal Republic of GernlanV

209
Dr. Walker's paper dealt with the importance of phosphate for the export of photo-
synthates out of the chloroplast. He showed that a low P level in the medium may delay
the export of triosephosphate. This is a very basic process, as both energy and organic
carbon required for crop production depend on this export. It is suggested that a
number of observations associated with phosphate deficiency could be related to this
chloroplast export. In the chloroplast, phosphate may influence enzyme activity by
an allosteric effect on the ADP glucose-pyrophosphorylase. This example shows that
plant nutrients may directly control enzyme activity.
Whether the import of photosynthates into amyloplasts is also controlled by a
mechanism analogous to that concerning the export of triosephosphate out of the
chloroplast, is an important question in relation to starch formation in grains and
tubers. This remains an open question and it is not yet clear whether the phosphate
level may play a crucial role in starch production of storage organs.

210
 Chairman of the 4th Session
Prof. Dr. It. Laudelout. Department of Soil
Science, University of Louvain-la-Neuve/Bel-
gium; Member of the Scientific Board of the
International Potash Insti te

The Use of Models

in Crop Physiology
Simulation Models of Growth of Crops,
Particularly under Nutrient Stress
Dr. F. W. T. Penning de Vries, Department of Theoretical Production Ecology, Agricultural
University. Wageningen/The Netherlands-

S11111marY
This paper considers the development of crop growth modeling, and analyzes some current
models of crop growth under non optimal nutrient supply.
Modeling of growt and production in agricultural situations without any nutrient stress
received most attention till now. and models have been developed that predict growth and
transpiration for those conditions fairly reliably. Some of their aspects, though, still need
mt ch improvement, such as plant morphology and assimilate distribution.
As more knowledge about availability of nitrogen and minerals in the soil becomes operational,
modeling in this field is becoming more relevant for growth and production studies. Some
aspects of crop production in these conditions are discussed. It appears that models of growth
and production still need considerable strengthening by experimentation and modeling before
they will be fairly predictive.

1. Introduction

This paper presents a brief survey of the modeling of crop growth, with an emphasis
on growth and production under nutrient stress. Two viewpoints will be taken: a
historical one to follow the development of models and modeling, and an analytical
one to describe contemporary models at different levels of crop production.
Before these viewpoints are given, it is useful to repeat briefly the definition of some
terms, and to structure the broad field of modeling by distinguishing 4 levels of crop
production and by characterization of 3 levels of development of models.
A 'model' is defined as a schematic representation of a 'system', the latter being a
'coherent part of the real world'. 'Simulation' is the 'building and utilisation of models'.
Many types of models can be distinguished. Only the most important group of
tangible models, that of explanatory, dynamic simulation models, will be considered
in this paper. In such models, processes are described by mathematical equations.
With the proper combination of those equations growth can be calctlated. Stich
calculations are performed for relatively short time intervals, e.g. of I day, after which
the computed increase in dry matter is added to the biomass already present. Changes
in the lcaf surface area and in stocks of soil water and soil nutrients can be computed
similarly. For the next day, all calculations are repeated. accounting for the changes
in biomass, leaf surface and stocks that took place and for changes in the plant environ-
ment (e.g. radiation) that may have taken place. Cycles of such calculations are repeat-
* Dr. F. IV. T. Peim,ting de Vries, Department of Theoretical Production Ecology, Agricultural
University, Bornsestceg 65, NL-6708 PD Wageningen/The Netherlands

213
ed until the growing season is completed. Such models are explanatory in the sense
that the simulated growth is calculated from, and hence-based on, underlying physio-
logical, physical and biochemical knowledge. Models are then used to integrate
knowledge of processes fundamental to crop growth in order to compute the increase
in crop biomass, and thus help to bridge levels of knowledge and fields of science
(De Wit [60]). The models are called 'dynamic' because the growth rate throughout
the growing season can be simulated with them, and not only the final production,
and because the final production of the crop is not a complex, though static regression
of yield on weather and soil variables. It is the resultant of interacting processes and
of environmental conditions whose effects depend on their timing with respect to the
current state of the crop and of the soil. The above terms and concepts are extensively
discussed and elaborated by De Wit [60], De Wit and Goudritan [62], Brockington
[8] and Penning de Vries [47].
For agricultural production, an elegant and practical delimitation of system was
proposed by De Wit [63]. He distinguishes 4 levels of plant production, ignoring
diseases, weeds and pests. The systems of plant growth and crop productivity at each
of these levels can be considered as belonging to one broad class. Those levels are:

Production level 1: Growth in conditions with ample plant nutrients and soil water.
The crop growth rate is then determined by weather conditions and amounts to
100-350 kg dry matter ha- 1 day- ' ; crop production depends on the growth rate and the
duration of the growing season. This situation is sometimes realized in field experi-
ments and in glasshouses. Major elements of this type of system are the dry weight
of leaves, stems, reproductive or storage organs; major processes are photosynthesis,
growth and maintenance, biomass distribution and leaf area development.

Production level 2: Plant growth is limited by water shortage part of the time and the
duration of the growing season may also depend on soil water availability. This occurs
on heavily fertilized soils in semi arid regions and in temperate climates. The situation
is neither very common in agriculture, nor in natural ecosystems. The extra elements
in this class of systems are the plant and soil water balances; crucial processes are
transpiration and other processes of loss or gain of water from the soil.

Production level 3: Plant growth is limited by shortage of nitrogen (N) most of the time
and sometimes by water shortage. This is quite a common situation in agricultural
systems and is also normal in nature. Important elements of this class of systems are
the N in the soil and in the plant; important processes are the transformation of
nitrogenous compounds in the soil and other processes of the N-balance, absorption
by roots, growth-availability interactions and redistribution within the plant.

Production level 4: Plant growth is mainly limited by the availability of other elements
of which shortage of phosphorus (P) is most common. Growth rates are 10-50 kg dry
matter ha- 1 day-' over a growing season of about 100 days. This situation occurs in
heavily exploited areas where no fertilizer is used, such as in the poorest areas of the
world. Important elements of this class of systems are the P and N contents of the soil
and plants, and important processes are their transformations in the soil, absorption
by roots and the response of plant growth to their absolute and relative availabilities.
Though it is rare to find cases that fit exactly into any of these production levels, it

214
is a very practical simplification in the beginning of a study to reduce specilic cases
to one of them. Only in more detailed studies, the complex situation in which different
limitations intertwine during growth, earns consideration.
Three stages of development of models can be distinguished in each of these situations
(Penning dc Vries [47): preliminary, comprehensive and summary models. During its
development, a model moves gradually from one stage into the next, All of these
3 stages do exist at the I ° and 2' level of crop production, but only preliminary models
are developed at the levels 3 and 4.
Preliminary models have a structure and contain data that reflect current scientific
knowledge about the modeled system. but they are simple because insight into the
explanatory level is still vague and imprecise. As a consequence, the value of such
models for prediction is quite limited. However. as modeling provides a means to make
explicit and to quantify hypotheses about processes in a system, preliminary models
can be useful for the development of science in showing the consistency of such hypoth-
eses with other information. Moreover, preliminary models are a first step towards
comprehensive models, which reflect systems of which essential elements arethoroughly
understood and in which much knowledge is incorporated. Comprehensive models are
often fairly good for predictive purposes, though they should be used with utmost
precaution in situations that differ considerably from those in which the models were
evaluated. Comprehensive models are typically intricate and little accessible for its
potential users. Summary models should therefore be made of satisfactory comprehen-
sive models. A summary model is a model of a comprehensive model: essential aspects
are reproduced in a simple way. and aspects that are only marginally important are
shed predecessor. Its predictive value is about the same as that of its, but its simulated
results contain much less detail and arc thus more useful to non-specialists.

2. A brief history of modeling growth and production

Since a little over a decade, modeling crop production and plant growth receives a fair
amount of attention and had its share of publicity. At any time, models have been
published that differ enormously in stage of evolution: some are well developed while
others are still preliminary. This brief history emphasizes the most sophisticated
nodels on a subject at any moment in time.

2.1 1953-1968: preliminary models

Regrcssion models of crop productivity have been built for a long time to provide
predictions of yields, and wvere usually based on rainfall. Scientists. seeking an explana-
tion tor such relationships, started to calculate potential and actual canopy photo-
synthesis on the basis of increasing operational knowledge of leaf photosynthesis.
The first study was as early as 1953 by Moisi and Saeki. Preliminary growth models
considered photosynthesis itl detail, but calculated growth. very simply, by substracting
daily respiration from daily photosynthesis. These models were suitable for well
developed leaf canopies and under good growth conditions, as leaf photosynthesis was
measured n similar circumstances. Biochemical and physiological research had
revealed in detail i mechanisins of respiratory processes, but there was little understand-
ing of how their rates were geared to processes in whole plants (Beevers [6, 7]. As

215
 a result, crop respiration was modeled simply as a constant rate per unit plant weight
or leaf area, care being taken to arrive at an intuitively fair cumulative fraction of 20%
(De Wit [57]), 33% (Loomis and Williams [32]) or 30% (De Wit [58]) of gross
photosynthesis over the whole growing period. The limited understanding of metabolic
activity and respiration restricted the predictive value of such models considerably.
Those early models did not simulate the distribution of new biomass over leaves,
stems, reproductive organs and roots. Although its importance was well realized
(Ross [48]), almost no mechanistic concept was available to simulate the patterns
of growth of organs. Formation of plant organs was specified according to plant age
or development stage. An exception was the distribution of new biomass over roots
and shoots under mild water stress, according to a 'functional balance'. This concept
was included in Brouwer and De Wit's model [9] and its principle was adopted later
in many other growth models.
Effects of moderate or severe water shortage on growth and production were not
included in these models. Neither were effects of nutrient shortage. The leaf surface
area of the crop, an important determinant of canopy photosynthesis, was not
simulated but based on direct measurements. Finally lack of appropriate field obser-
vations made evaluation of these models, and the appreciation of their results, quite
subjective.
Writing simulation models required considerable skill in using computers. The com-
puter language FORTRAN was used predominantly for programming. Yet, simulation
of micro-economic systems had already led to the development of the simulation
language DYNAMO (Forrester[16]), that was remarkably suitable to crop physiol-
ogists.

2.2 1969-1976: comprehensive models at the production levels I and 2

2.2.1 Models
The models developed in agricultural research described canopy photosynthesis,
respiration and growth, and mimicked the distribution of biomass according to
experimental results or attempted to simulate it. Transpiration was simulated in some
of the models. Annual crops or vegetations were modeled almost exclusively, with
particular emphasis on vegetative crops. Some models were set up to simulate types
or crops (ELCROS, by De Wit et al. [59]; SPAM, by Sinclair et al. [54]; ARID
CROP by Van Keulen [26]), but most were specific for certain species (such as: SIM-
COTfor cotton [1]; SIMED and ALSIM for alfalfa by Holt et al. [22] and Fick [14],
resp.; SO YMOD for soybeans by Curry et al. [JO]; SUBGRO by Fick et al. [13] for
sugar beet.) Obviously, these models had much of their scientific content in common.
ELCROS, SUBGRO and SIMED simulate plant growth under optimal soil nutrient
and soil water conditions. ARID CROP and SPAM simulate growth at production
level 2. Most other agricultural models were meant to simulate plant growth under
actual field conditions, i.e. at production level 3 and in situations where diseases and
pests may also play an important role. The results of such models seem of direct
interest for application, but they require treatment of the whole range of interactions
of growth with nutrient availability, water shortage and diseases. These aspects were
not well understood, so that such models were necessarily inaccurate and hence
difficult to evaluate. They could be extrapolated to conditions only slightly different
from those in which they originated.

216
In biological sciences, ambitious attempts were made to model parts of ecosystems,
particularly in the Biome studies in the IBP programs in the United States of America.
Examples are studies of desert ecosystems (Goodall [17]), tundra ecosystems (Miller
and Tieszen [38]) and grassland ecosystems (hins [25]). The Biorne models had a
wider scope than agricultural models, and conglomerate aspects as different as soil
water and nutrients, weather, native plant species with different physiological and
ecological characteristics, and species of herbivores and their predators were included.
Consequently, plant growth received less attention, and its simulation lagged some
years behind its formulation in agricuItUral models.
An important development in modeling crop growth was the quantification of energy
requirements for growth and maintenance processes, to both of which respiration is
related. McCree [35] quantified experimentally respiration coefficients for both
processes in white clover plants, and determined subsequently such coefficients for
other species and for other temperaturcs (McCree [36]). Penning de Vries and co-
workers, prompted by inadequacies of the ELCROS model, published a study of plant
metabolism [41, 42, 43] that showed how respiration coefficients for growth pro-
cesses can be derived by straightforward stoichiometry from the biochemical compo-
sition of the biomass. Insight into maintenance processes was improved, but its
quantification remained essentially experimental. This approach to metabolism and
respiration has been adopted since in most crop growth models (Penning f/e Vries
[47]).
The model ARID CROP ( Van Kenlen [26]) successfully linked the soil water balance
with growth of the vegetation ont well Fertilized soils in semi arid regions. This compre-
hensive model simulated in detail water transport in the soil, and used a summary of
C-balance processes and of potential canopy transpiration from ELCROS to calculate
the transpiration coefficient of the canopy. Division of the crop transpiration rate by
this coefficient gives crop growth. This principle has been applied since in other models
at this production level.
2.2.2 Modeling
The Trebon Conf,rence in 1969, organized by the Intenational Biological Program
(Set/ik [53]). marked the beginning of development of comprehensive growth
models. This development resulted from advances in the knowledge about the subject
and by advances in modeling techniques.
A major paper in Trcbon was given by De Wit /60] abouL concepts in modeling, and
was followed by an excellent discussion about merits of simulation (Wagtgoner
[56]. The main points of those papers are the distinction between demonstrative and
explanatory models, the idea that models can serve to integrate knowledge from difte-
rent levels of biological organization, and a discussion of limitations to modeling of
physiological processes. The conference, and many following ones. showed a large
interest in modeling and considerable optimism about its future. The latter was based,
among other things, on the large success of canopy photosynthesis models in the pre-
ceding years (Fmqruhiar 12]). in addition, suitable coMpiter ilcilities appeared to
be ready for use in this field, at least in sonic situations. As a result, this period
witnessed the proliferation of extensive and complex models.
The frequent comparison of simulation results and field data upgraded the quality of
model prediction considerably. Disagreement between experimental and simulated
restlts sometimes led to experimental results being discarded, bit LisuaIlly caused

217
reconsideration of model parameters and design of new experiments. The danger that
adjusting parameters to a particular situation results in a curve fitting procedure of
little value is clearly recognized (e.g. De Wit [60]) but not always resisted.
As a result, not all scientists were impressed by these developments. Hesketh and
Jones [21] express concern about the superficiality with which, they claim, many
models were built. Passioura [40], reflecting on a developing fashion of modeling,
stated that 'there is a much quicker way of getting... a framework in which to hang
ones research... than spending a year to create a comprehensive simulation model'.
Seligman [51] comparing grassland models, saw the future value of such models
rather than their actual performance as a justification for modeling, and the lack of
an alternative integrated approach strengthened his conclusion.
On the technical level, many modelers continued to write their simUlafion programs
in FORTRAN. it has the important advantage of being widely available but large
programs are often difficult to read. Others preferred user oriented simulation lan-
guages like DYNAMO, or C.S.M.P. (IBM[24]) that require less programming skill.
]n the U.S. Grassland Biame, the simulation language SIMCOMP (Gustafson [19])
was developed and intensively used.

2.3 1977-?: Comprehensive and summary models at the production levels I and 2, and prelimi-
nary models at the level 3

2.3.1 Modeling
Evaluation is an important issue in this period. Some important papers were written
around 1977 (Baker and Curry [2], Penning de Vries [44], Innis [25]) about concepts,
terminology and procedures of evaluation. It is this point in time that has been taken
as the benchmark for the beginning of this period. Its end cannot yet be indicated.
In theory, evaluation is dealt with very seriously indeed. In practice, under the pressure
for quick results and due to sloppiness of modelers, only a few models are thoroughly
evaluated. In fact, experiments that are only used for evaluating model behaviour and
not for derivation of one or more parameters are still rare. Still, the emphasis on
evaluation is quite a healthy one.
Some models grow considerably in size and complexity. Particularly models that
integrate many disciplines increase in size and maintain all known detail. Their growth
suggests that their users are satisfied with their behaviour, at least to a certain extent.
One should keep in mind, however, that if the scope of the model increases but its
transparency decreases, the advantage of modeling can turn into a disadvantage by
disorienting and misleading its user. The fact that this occurred quite a few times made
some modelers and many non-modelers shy away from large simulation models. This
leads to an important question: how to deal with very large models? Is increasing
complexity inevitable in the process of model development? Technical limitations are
sometimes also a reason for this concern, but they are generally offset by technical
developments. One route leading away from the dilemma of further development that
brings unmanageable complexity can be the use of summary models of well studied
aspects. Summary models can be a suitable replacement for comprehensive models
if less detail is needed than the large model can provide. But a summary can only be
made when the subject is really well understood, and as yet, few summary models of
crop growth have been published (Penning de Vries [47]).

218
Many comprehensive models are published in this period. However. it is alarming that
these publications are usually too brief to get a good impression of the model. Fairly
complete descriptions are sometimes given in little available internal reports or
tincitable literature. The series of Simulation Monographs 165] is a good exception.
2.3.2 Models
A few generalities of the developments of models are presented here; some examples
of the most sophisticated models will be discussed in the next chapter. Not much pro-
gress has occurred in the last few years in modeling the C-balance processes, which are
of principal interest at production level I.Further developments arc still needed in the
understanding of the dry matter distribution over the organs of plants, and that of the
development of the surface area of the leaves. Such improvements are necessary to
make models more versatile and to better predict the yields of economic products
rather than total dry matter production. The reader is referred to Loomis et el. j33]
and Pennin.g de Vries [47] for further detail.
Transpiration and soil water balance models were reviewed by Hall [20]. Hsiao ettit.
123] and Fischer and Tinner [15]. Basic concepts changed little recently, and fair
simulation models for growth at production level 2 are available. Models to simulate
the daily course of carbon and water balances produce the effect of water stress always
through stomatal closure. Some growth models that use time steps of I day reduce
photosynthesis by a factor that depends on the average water stress of the last day.
Other models with I-day time steps calculate growth by dividing daily transpiration
by a transpiration coefficient (g water g- 1 dry matter). This approach seems indeed
superior due to the constancy and predictability of this coefficient, particularly when
transpiration is expressed relative to potential evapotranspiration (Doorenbos and
Kassala ilj). This constancy arises since 1. gross photosynthesis and transpi-
ration are almost proportional throughout the day. because stomata offer the largest
diffusion resistance to both, particularly under water stress: 2. the efficiency of the
growth process is unaffected by water stress; 3. the rate of maintenance respiration
diminishes probably slightly as a result of the lower overall metabolic activity.
It has recently become evident that the transpiration coefficient of plants without
water stress often. hut not always, shows a similar constancy. It is constant if stonlata
open and close in correspondence with the rate of photosynthesis, as the result of a
regulating mechanism that maintains a constant CO,-concentration within stomata
over a wide range of photosynthetic rates (Raschke [48], Gouchiaan and Van Laar
: 18 . lVong et al. /64]:). It has as a consequence that the rate of canopy photosyn-
thesis determines the rate of canopy transpiration under fixed atmospheric conditions,
which leads to a transpiration rate in tie field that is much lower than is often thought.
This was shown with the model BACROS, that simulates such a mechanism (Van
Keulen el ell. [29], and confirmed experimentally.
The effect of a sub-optimal N-supply to crops is studied increasingly with simulation
models. Four of such models will be discussed in 3.3. Such models concentrate on the
rate of absorption of N by the crop and on its concentration of N. In what way low
or high N-concentrations modify the distribution of dry matter over organs, and how
low concentrations reduce photosynthesis and growth is poorly understood. The
problem of how the growth rate is related to the rate of uptake of N is confounded by
the observation that tissues have maximum and minimunLm concentrations of N that
diminish with age of the plant (e.g. Peinig de Vries el al. 145/) and by the tact that

219
a part of the absorbed N remains mobile and is translocated from old parts to growing
tissues if the N-uptake by the root system becomes too low. Both aspects are not yet
well-known, and the existing models of uptake, redistribution and growth are largely
descriptive in nature, rather than explanatory. Examples of such models are given
elsewhere [47].
Uptake of N by the crop under conditions of N-limitation is relatively fast and effi-
cient, so that the question of uptake under conditions of N-limited growth becomes
largely one of the availability of N in the soil. Reviews by Beek and Frissel [5], Reuss
and Innis (in: [25j) and Van Veen [55] and further research by Seligman and Van
Keulen [52] and Kral (in: [46]) showed considerable progress in the modeling of
mineralization of N from soil organic matter and of immobilization back into it,
and of the recovery of fertilizer N.
Modeling crop growth at production level 4 is still almost absent. The effect of P-
shortage on growth and also the uptake of P by roots are even less known than those
for N. One simulation model has been published for growth of a natural vegetation
(ELM, Sauer, in: [25]) and it will be discussed in 3.3. It regards absolute concentra-
tions of P in tissues. Penning de Vries et al. [45] suggested that the ratio of P to N in
the plant may often be a more relevant variable than the absolute concentration of
either of them, as these elements are functionally related in the metabolism of cells.
They showed that the P/N ratio varies over a narrow range of 0.04 g P g-1 N in
P-starved plants to 0.15 g P g-l N in N-starved plants. This implies that the effect of
P-shortage can be seen as reduced availability of N to the crop. Cole et al. (in: [25])
presented an interesting preliminary model on transformations of organic P in the soil
and on uptake by roots. It was coupled to the growth model of Sauer mentioned earlier.
Beck [4] discussed transformations of forms of inorganic P in fertilized soils.

3. Some current crop growth models

Some examples of relatively advanced crop growth models at 3 of the 4 levels of

production will be discussed. No models are published about growth limited by
P-availability. For more detail, the reader is referred to the original papers and to
recent reviews [33, 47].

3.1 Crop growth with ample nutrients and water

De Wit's [61] model BACROS may serve as an example of a comprehensive model.
It simulates vegetative growth of crops at non limiting levels of soil water and soil
nutrients (production level I) on the basis of standard meteorological observations. It
considers neither germination nor the reproductive growth phase. It has been develop-
ped over more than a decade by De Wit and a team of co-workers, and was called
ELCROS at an earlier stage. It comprises a thorough treatment of the processes of the
C-balance and of transpiration. Laboratory research, literature study and frequent
evaluations led to a model that simulates growth, yield and water use quite reliably over
a wide range of environmental conditions for C3 and C4 crops. Its structure reflects
cereal and grass crops, and small but specific sets of parameters and functional re-
lationships specify the actual species under consideration. The model is adaptable to
other types of species, but no further work has not yet been done. Like all models in

220
this group De Vit's model is particularly weak in the section ofl regulation of distli-
bution of biomass, and in the development of leaf surface area. The latter limitation
is a serious handicap in simulation of the early stages of growth.
An example of a summary model at this production level is that by Van Kei/en [27]
for potential rice production. It is based on structure, data and concepts of BACROS.
Its basic growth equation can be given as:

growth = (G P × 0.68 - MC x DW) x CE

GP stands for gross photosynthesis (in kg CO, m -2 day-'; 0.68 converts it to kg

glucose M- 2 day-), DW for total dry weight (kg dry matter n-), MC for the mainte-
nance coefficient (kg. kg-' day-1) and CE for the conversion efficiency of the growth
process (kg . kg-). Van Keuien distributes biomass formed over roots, leaves, stems
and. after flowering, to inflorescences plus seeds in predetermined proportions and
related to the physiological age of the crop. GP is calculated from standardized data.
The leat surface area, required in the photosynthesis calculation, is found by dividing
the leaf weight by 0.1 kg. ni - -. MC is 0.02-0.015; the effect of temperature on MC
could be neglected as this model was applied in a fairly constant environment. CE
depends only on the chemical composition of the biomass formed, and a value of
about 0.7 is common. Final yield is calculated by proceeding with time steps of 10 days.
The structure of this summary model is correct for many other crops with a determinate
growth pattern; parameters and functions might need adjustment, The model is also
principally correct at lower levels of productivity. However, the equation focusses on
photosynthesis as the limiting factor for growth, which is then not correct. Thus the
above summary growth model nlay be applied to other productivity levels with great
caution.

3.2 Crop growth with ample nutrients

An example of a comprehensive model at production level 2 is ARID CROP by

Van Kei/en [26.] and Van Keulen et al. [28]. Ii simulates the growth of annual vege-
tations of natural pastures in semi-arid regions with winter rains, and has also been
applied to wheat. The model SORGFfor simulation of growth of sorghum (Maas and
Arkin [34]), designed for optimization of growth and water use during the season, has
much in common with ARID CROP, scientifically speaking. Both simulate the water
balance of a number of soil layers by accounting for rain and drainage from one layer
into the next, for evaporation from the surface and for extraction of water from the
rooted layers, due to transpiration. Runoff of water from surfaces with a slope under
intensive rain and interception were added to ARID CROP for its use in a summer
rainfall area (Stroosijder, in: [46]). Canopy transpiration is close to the potential
evapotranspiration when the soil is well covered. The transpiration rate is reduced if
soil water runs low in one or more rooted layers. Daily growth equals daily transpira-
tion divided by the transpiration coefficient. The latter is equal to tile ratio of the
transpiration rate and the growth rate with ample water but otherwise in similar
conditions. The transpiration coefficient is lower when photosynthesis determines the
rate of transpiration through the regulation of the internal CO,-concentration in
stomata [2.3.2) than when the regulation is absent, Effects of severe water stress on
development, growth and death of tissues are included in ARID CROP, although
little physiological basis exists for their actual formulation. ARID CROP simulates

221
growth and the soil water balance in winter and in summer rainfall areas fairly re-
liably [28, 46].
A summary model of a soil water balance and the growth of a vegetation of annuals
with ample nutrients in semi-arid regions has been described by Stroosnijder (in:
[46]). It is based on concepts and data of ARID CROP. It presents a calculation of
the amount of soil water that is available to the plants over time intervals of 10 days,
and computes growth by dividing the transpired water by the transpiration coefficient.

3.3 Crop growth under conditions of N- or P-limitation

Given the large amount of field experimentation and also the numerous measurements
of ion uptake by plant roots, the number of simulation models that try to combine, or
even to make use of all this information is still amazingly small. Only a few models for
growth at this level of production were found: GOSSYM+RHIZOS (Baker et al.
[3], Lambert et al. [31]) for cotton, PAPRAN (Seligman et al. [50], Seligman and
Van Keulen [52] and ELM (Sauer, in: [25]) for natural pastures in semi-arid regions,
and SO YMOD/OARDC (Meyer et al. [37]) for soybeans. The approaches in these
models differ considerably. However, there are many parameters and even processes
that are poorly known in such systems so that there is sufficient freedom in the choice
of parameters to let all models show a fairly realistic behaviour. This remark is not to
criticize these models, as for each choice good arguments are presented, but it under-
lines the preliminary stage of these models. As these models are much larger than the
growth-nutrients interactions discussed, their value should not be inferred from the
description below.
The ELMmodel harbours the simpliest description of the reduction of the growth rate
by N-shortage: a reduction factor between I and 0 by which the potential growth
rate is multiplied. The factor depends on the current N-concentration in the plants,
but young plants are more reduced at any concentration than old plants are. N in tops
and roots and the translocation of N from one to the other are simulated. The sup-
porting sub-models of N-transformation (Reuss and Innis, in: [25]) and of P-trans-
formations (Cole et al., in: [25]) in the soil and of N and P uptake are sophisticated
and thoughtfully developed, The rate of uptake of nitrate ha-' is a function of root
biomass, of nitrate concentration in the soil solution, and of the soil water potential.
The uptake of P depends on similar factors, and further on the concentration of N in
the roots. Too high concentrations in the plants are avoided by reducing the rates of
uptake.
The uptake of N in RHIZOS equals the amount of nitrate dissolved in the transpira-
tion stream. This view neglects the contribution of diffusion to the N-supply of roots,
which can be substantial (Van Keulen [30]). GOSSYM is coupled to RHIZOS and
calculates the potential growth rates of the organs of the cotton plant and the corre-
sponding demands for C and for N. The actual growth rates are found by multipli-
cation of these maximum rates by reduction factors due to C-stress and to N-stress.
These factors are numerically equal to the current reserves of carbohydrates and of N
divided by the demands for them, a formulation that resembles the transpiration coeffi-
cient concept. The pool of labile N is filled through uptake by roots and through
'mining' for N in older organs. In addition to growth rates, the C-stress and N-stress
factors modify the morphogenetic development of the plants and control the rate of
abortion of flowers.
PAPRAN also computes the growth rate by multiplication of the potential rate,

222
derived from ARID CROP, by a reduction factor. This factor depends on the difference
between the actual N-concentration in the tissues and the minimum concentration.
The latter diminishes from about 2.5% to 1%, as the plant ages. At any moment. its
optimum concentration is about 2 x its minimum value, and above this there may still
be some luxury consumption of N. The daily uptake of N by the crop is essentially
equal to half its demand for N (the current weight of its biomass times the difference
between its current maximum and actual N-concentration), limited only by a maxi-
nmum rate of N-absorption by the crop and not exceeding the amount of N available
in the soil. Redistribution of N from vegetative parts to grain occurs up to a maximum
seed yield or until the minimum N-concentration of the straw is attained, and as long
as weather conditions permit. The amount of N available in the soil is computed as
the resultant of mineralization and immobilization of N in 10 soil layers, in which
inorganic N. and N in fresh and 'stable' organic matter are considered (Seliq,,,an and
Va,n Keiden [52)). The rates, efficiencies and regulations of these processes are simu-
lated according to current knowledge. From comparisons of simulations and experi-
ments the authors conclude that the model provides a good framework for further
investigations but also that more physiological and soil micro-biological information
is required before the model can be applied more widely.
SO YMOD/OARDC represents a particular case: that of a legume which only assimi-
lates N as N, in its root nodules. The growth of N-free material and of proteins are
basically simulated independently, but their ratios are not allowed to exceed the limits
of 2% N and of 6% (8% in the seed). The rate of N-assimilation is controlled by the
average level of N in the plant. Photosynthesis and growth are simulated for each node
separately; both rates become reduced at all N-levels below the maximum concentra-
tion. and are 0.0 at 2% N. Thus, the basic limitation to crop growth remains to be
canopy photosynthesis.
The direct effect of water shortage is simulated as the reduction of photosynthesis
and thus of growth in all models, except in ELM where a direct relationship between
the growth rate and the soil water potential is programmed. An indirect effect of water
stress results from a reduced uptake of N. and is related to the soil water potential in
ELI and to the transpiration rate in RHIZOS. This indirect effect is small in PAP-
RAN as diffusion of nitrate towards the roots is fast and plays an important role in
the N-supply ( Van Ketien [30]). (There is no N-uptake in SO YMOD.) Both effects
of water shortage reduced the final yield most when the crop was relatively well sup-
plied with N. The reduction is only small in case the final yield was limited by the small
amount of N absorbed by the crop and the minimum concentration of N was attained.
It is interesting to note that the transpiration coefficient of vegetation of annuals of
natural range lands in the Sahel, where production is limited by N or P, was observed
to be as low as that of crops under optimal growth conditions. This is ascribed to the
functioning of the stnomatal internal CO,-concentration regulating mechanism (2.3.2.
Penning (it Vries and Dfiteve [46]). Annual grasses of range lands in the mediterranean
zone do not regulate their stomata in this way ( Van Kei/en, pers. comm.). which leads
to less efficient water use. This featlAre of stomatal regulation has not yet been included
in models at this production level.
The differences between those models indicate that modeling at the levels of crop
production 3 and 4 still needs considerable strengthening by research involving experi-
mentation and modeling before such models will be fairly predictive and more widely
applicable.

223
Acknowledgements

This paper is heavily based on work from colleagues in the Department of Theoretical
Production Ecology of the Agricultural University, Wageningen, The Netherlands.
The author's familiarity with their work may have caused too large an emphasis on it.
Mrs. Merkelijn helped to trace many publications on models.

4. Literature cited

1. Anonymus: Computer simulation of a cotton system. ARS-S-52. Agric. Res. Service,

USDA, POB 5465, Mississippi State, Miss., 39762, USA, 1975
2. Baker, C.H. and Curry, R.B.: Structure of agricultural simulators: a philosophic view.
Agricultural Systems 1, 201-218 (1976)
3. Baker, D.N., Lambert, J.R., Phene, C.. and McKinion, J.M.: GOSSYM: a simulator
of cotton crop dynamics. In: Computers applied to large scale agricultural enterprises.
Proc. U.S.-U.S.S.R. seminar, Moscow, Riga, Kishinev, pp. 100-133, 1976
4. Beek, J.: Phosphate retention by soil in relation to waste disposal, Ph. D thesis Agricul-
tural Univ., Wageningen, The Netherlands, 1979
5. Beek, J. and Frissel, M.J.: Simulation of nitrogen behaviour in soils, PUDOC, Wagenin-
gen, The Netherlands, 1973
6. Beevers, H.: Respiratory metabolism in plants. Row, Petersen and Company, New York,
1961
7. Beevers, H.: Respiration in plants and its regulation. In: Setlik, 1970
8. Brockington, N. R.: Computer mcdeling in agriculture. Oxford Univ. Press, London, 1979
9. Brouwer, R. and Wit, C. T. de: A simulation model of plant growth with special attention
to root growth and its consequences. Proe. 15th Easter School Agric. Sci., Univ. of
Nottingham, 224-242, 1969
10. Curry, R.B., Baker, C.11. and Streeter, J.G.: SOYMOD I, a dynamic simulator of
soybean growth and development. Trans. ASAE 18 (5), 963-974 (1975)
I1. Doorenbos, J. and Kassain, A. K,: Yield response to water, FAO irrigation and drainage
paper 33, FAO, Rome, 1979
12. Farquhar, G.D.: Modeling of photosynthetic reponse to environmental conditions. Vol.
14, chapter 16, New Encyclopedia Plant Physiology (in press). Springer Verlag, 1981
13. Fick, G. W., Williams, W.A. and Loomis, R.S.: Computer simulation of dry matter
distribution during sugar beet growth. Crop Sci. 13, 413-417 (1973)
14. Fick, G. W.: The mechanism of alfalfa regrowth, a computer simulation approach.
Search, Agriculture 7 (3), 1-26, 1977. Cornell Univ. New York
15. Fischer, R.A. and Turner, N.C.: Plant productivity in the arid and semiarid zones. Ann.
Rev. Plant Physiol. 29, 277-317 (1978)
16. Forrester, J. W.: Industrial Dynamics. MIT Press, Boston, USA, 1961
17. Goodall, D. W.: Ecosystem simulation in the US/IBP Desert Biome. Summer Computer
Simulation Conference, Montreal, 777-780, 1973
18. Goudriaan, J. and Van Lar, H. H.: Relations between leaf resistance, CO,-concentrations
and CO-assimilation in maize, beans, Lalang grass and sunflower. Photosynthetica, 12
(3), 241-249 (1978)
19, Gustafson, J.D.: SIMCOMP 3.0. In: Innis, 1978 (1978)
20. Hall, A.E.: Mathematical models of water loss and plant water relations. Vol. 14, chapter
7, New Encyclopedia Plant Physiology (in press). Springer Verlag, 1981
21. Hesketh, J.D. and Jones, J. W.: Some comments on computer simulators for plant growth.
1975. Ecol. Model. 2 (3), 235-247, 1976
22, Holt, D.A., Bula, R.I., Miles, G.E., Schreiber, G.E. and Peart, R.M.: Environmental
physiology, modeling of alfalfa growth. I. Conceptual development of SIMED. Res.
Bull. 907, Purdue Univ., Indiana, USA, 1975
23. Hsiao, *T.C., Fereres, E., Acevedo, E. and Henderson, D. W.: Water stress and dynamics
of growth and yield of crop plants. In: Ecological Studies vol. 19, Lange, Kappen and
Schulze, eds. Springer Verlag, Berlin. pp. 281-305, 1976

224
24. I. B. M.: Continuous System Modeling Program Ill, General system information manual
(GHI 9-7000) and Users manual (SH19-7001-2). IBM data processing div., 1133 West-
chester Avenue, White Plains, NY 10604. USA. 1975
25. Innis, G.S.: ed. Grassland Simulation Model. Springer Verlag, New York, 1978
26. Kettlen, H. van: Siiulaltion of water use and herbage growth in arid regions. Simulation
Monograph, PUDOC, Wageningen. The Netherlands, 1975
27. Keten, H. an: A calculation method for potential rice production. Contr. Centr. Res.
Inst. Agric. Bogor 21, Central Research Institute for Agriculture, Bogor. Indonesia. 1976
28, Kei/et. H. van, Selik , mn. N.G. and BenI.jin, R. IV.:Simulation of water use and
herbage growth in arid regions: a reevaluatitn and further development of the model
ARID CROP. Agric. Systems (1980)
29. Keiden, H. van, Van Laar. H. H.. Lounerse, IV.and Gourian,, J.: Physiological aspects
of increased CO_-conccntration. Experientia (1980)
30. Ketlen,H. ran: Modeling the interaction of water and nitrogen. Proe. ICARDA sym-
posiu 'Inreasing the effectiveness of water and nitrogen in rainfed farming systems in
mediterranean type environments., Webb. ed, (1980)
31. Lombert. J.P.. aket,. D.N.and Phene. C.J.: Dynamic simulation of processes in the soil
under growing row crops: RIIZOS. hi: ComptIters applied to the management of large
scale agricultural enterprises. Proc. U.S.-U.S.S.R. seminar. Moscow, Riga, Kishinev, 1976
32. Lo ./v, R.S. and Wi/aIS, W. .: Maximam cr01) production, an estimate. Crop Sci., 3,
67-72 (1963)
33, Loomis. R.S.. Rahbi,tge. R. and A' I: Explanatory models in crop physiology. Ann.
Rev. Plant Physiol 30, 339-367 (1979)
34. Maas, S.J. and 'kin. G.F.: Users guide to SORGF,a dynamic sorghum growth model
with feed-back capacity. Program and model documentation 78-1. Texas Agric. Exp.
Station, Texas A+ M Univ. Texas, USA, 1978
35, MCie, K.J._ Ai equation for tie rate of respiration for white clover plants under
control led conditions. In: Setlik. pp. 221-231, 1970
36. MCree. K.: Equations for the rate ef dark respiration of whitc clover and grain sorghtim
as functitns of dry weight, photosynthetic rate and temperatUrc Crop Sci. 14 (4),
509-514 (1974)
37. Meer, G.E.. Cury, R.B., Sneeter, J.G. and klederski, I/,/. SOYMOD/OARDC,
a dynamic simulator of soybean growth, development and seed yield, I. Res. ult. I 113,
Ohio Agric. Res. Devel. Center, Wooster, Ohio, USA. 1979
38. Miller, P.C. and Tieszen. L. L.: A preliminary model of factors aliecting primary pro-
duction in the arctic tundra. Arct. Alp. Res. 4, 1-18 (1972)
39. Monsi, Nt. and Saeki. T.: Uber den Lichtfaktor in den Pflanzengesellschaften tnd seine
Bcdeutung fur die Stoffproduklion. Jap. J. BOt., /4. 22-52 (1953)
40. Passiom,ra. J.B.: Sense and nonsense in crop si11ulation. .I. Aust. Inst. Agric. Sci., 39,
181-183 (1973)
41. Penaiag de Vi/es. FI I/. T., Rrinsii,. A.B. and fun Lrta. LI. : Products. requi reme nts
and etliciency of biological synihesis, a quantitative approach. J.Theor. Biol. 45, 339-377
(1974)
42. Penning rk'Vries, F. II. T.: The cost of maintenance processes in plant cells. Ann. Bot. 39,
77-92 (1975)
43. Pennat, cl lre. F. W. 7.: Use of assimilates in higher plants, It:Photosyntlhesis and
Productivity in Different Environments. (ooper, cd. Cambridge Univ. Press, Ca)mbridge,
UK. pp. 459-480. 1975
44, Penning de Fries. F. I,.T.: Evaluation of simulation models in agriculture and biology:
conclusions of a workshop. Agricultural Systenis 2, 99-107 (1977)
45. Penng de Fries, r. F.I T., Krt, J. M\.and Vait Keide, II.: Productivity of Salhclian grass-
lands and the availability of nitrogen and phosphorus roI l the soil. in:Nitrogen cycling
in West African ecosystems. Proc. SCOPE-UNEP Nitrogen Unit Workshop. Ibadan.
Rosswall ed. Roy. Swed. Acad. Sci. 1980
47. Penntiat, le I,-ies. F. IK T. Modeling growth and production. \1ol. 14. chapter 17, New
Encyclopedia Plant Physiology (in press). Springer Vcrlag.
48. Ramehke. K.: Stomatal action. Ann. Rev,. Plant Physiol. 26. 309-340 (1975)
49. Ross, Y.R.: Mathcmtical description of plant growth. Doklady Akadenii Nauk. SSSR
17. 481-483 (1966)

225
50. Seligman, N.G., Keulen, H. van and Goudriaan, J.: An elementary model of nitrogen
uptake and redistribution by annual plant species. Oecologia 21, 243-261 (1975)
51. Seligman, N.G.: A critical appraisal of some grassland models. In: Critical evaluation of
systems analysis for research and management. Arnold and De Wit, eds.: Simulation
Monograph, PUDOC, Wageningen, The Netherlands, 1976
52. Seligman, N.G. and Van Keulen, H.: PAPRAN, a simulation model for annual pasture
production limited by rainfall and nitrogen supply. In: Models for behaviour of nitrogen
in soils and uptake by plants. Frissel and van Veen eds., PUJDOC, Wageningen, The
Netherlands (in press)
53. Setlik, L: Prediction and measurement of photosynthetic productivity. Proc. TBP/PP
Tech. meeting Trebon, PUDOC, Wageningen, The Netherlands, 1970
54. Sinclair, T. R., Allen, L.H. and Stewart, D. W.: A simulation model for improving crop
productivity. Summer Computer Simulation Conference, Boston, 784-794, 1971
55. Veen, J.A. van: The behaviour of nitrogen in soil. Ph. D. thesis Agricultural Univ.,
Wageningen, The Netherlands, 1977
56. Waggoner, P.E.: Consultation on how models are made, how they are tested and what
they tell us of experiments to be done. In: Setlik, 1970, pp. 575-579
57. Wit, C. T. de: Potential photosynthesis of crop surfaces. Neth. J. Agric. Sci. 7, 141-149
(1959)
58. Wit, C.T. de: Photosynthesis of leaf canopies. Agr. Res. Rep. 663, PUDOC, Wageningen,
The Netherlands, 1965
59. Wit, C. T. de, Broinver, R. and Penningde Vries, F. W. T.: The simulation of photosynthetic
systems. In: Setlik, 1970, pp. 47-70
60. Wit, C.T. de: Dynamic concepts in biology. In: Setlik, 1970, pp. 17-23
61. Wit, C.T. de: Simulation of assimilation, respiration and transpiration of crops. Simu-
lation Monograph. PUDOC, Wageningen, The Netherlands, 1978
62. Wit, C. T. tie and Goudriaan, J.: Simulation of ecological processes. Simulation Mono-
graph, PUDOC, Wageningen, The Netherlands. (Japanese edition from Kodansha Ltd.,
Tokyo), 1978
63. Wit, C. T. de: In: Production primaire au Sahel. Penning de Vries and Djiteye, eds.,
Agric. Res. Rep. (in prep.), PUDOC, Wageningen, The Netherlands, 1980
64. Wong, S.C., Cowan, I.R. and Farquhar, G.D.: Stomatal conductance correlate with
photosynthetic capacity. Nature 282, 424-426 (1979)
65. Simulation Monograph Series, PUDOC, Wageningen, The Netherlands

226
Modelling in Mediterranean Area:
Adjustment Problems of Crop Models
Ph.,Malet, Station de Hioclimatologie, I.N.R.A. d'Avignon, Montfavet/France*

Summ,ari,

Despite the specific characters of the Mediterranean areas, there is no need for a different
approach to the building of analytical crop models. But the problems of adjusting a model
to a new area lead to focussing investigations on the organogenesis processes.
As an example a new scheme of organogenesis is presented for wheat; and by means of a
statistical model, the relationship between organogenesis and growth is shown to be the key
process in plant adjustment to change in the environment.
Such co-ordinated research in the Mediterranean and in more northern and even in more
southerly areas can undoubtedly widely increase the generality of crop models.

I. Introduction

A model is a simplified representation of a phenomenon. The picture depends not

only on our knowledge of this phenomenon but also on the intended use. Therefore,
a model represents a reality distorted for the purpose of achieving a particular aim
and this obviously restricts its generality.
As a consequence, the goals to be reached must be clearly determined before building a
model. Now, the aims obviously depend first on the diversity of cropping systems in
the considered area. In Mediterranean areas, there are numerous cropping systems.
The most intensive (e,g. greenhouse vegetables and flowers) can be found surrounded
by or adjacent to an extensive pastoral area. Likewise, typical production systems, for
instance of olives or rice, are often found near cereal farming which is very similar to
that of temperate areas.
In the first place a choice must be made between systems typical of the area and others
which are also 'ound in the temperate climate. The latter have been chosen up to now,
on the one hand because of their nutritional and economic importance, and on the
other because similar research has been done in other areas. Inmaking this choice we
are faced with the problem of adjusting a given model to different environments.
Demonstrating that no statistical model can be used outside its building area is easy
(M,alet [5j) ;neither its coefficients nor even its variables can be transferred elsewhere.
This is the more embarrassing as yield forecasting models have also been proved to be

* Ph. Malet, Station de Bioclinatologie, Domaine St-Paul, F-84140 Montravet/France

227
necessarily statistical. That is the reason why, in order to solve this dilemma, I have
suggested (Malet [5] and [6]) a scheme which takes into account both analytical
and yield forecasting models.
The adjustment to different environment conditions might be thought to be less
difficult for analytical models based on plant physiology. Unfortunately, such is not
the case.
As a consequence this lecture will be focussed on the main transposition problems
and the necessary steps to be taken to solve them.

2. The main adjustment problems of models

2.1 Nature of data and frequency of introducing data

If, as we said before, a model is closely dependent on its goals, it is closely so on the
nature of data and the frequency of introducing inputs.
Now, a model only uses the variables which are not permanently at their optima: for
instance (Figure 1), in a area where the factor X 1 never varies outside the range A,
this factor will never be taken into account in a model for this area considering the
dR
response -- of plants. A typical example of a factor generally at its optimum in a

dt

dR
Fig.]. Theoretical response d- of a crop to one factor X

model is the cultivar used, for it is generally well adapted to the area concerned. But
in the areas where such a factor X, varies outside the range A, it must be taken into
account. On the contrary, another factor may reach its optimum range and so become
meaningless.
de Wit and Goudriaan [16] very adequately remark that 'a system must be simulated
with a time interval of integration less than its time constant' (i.e. its time of response).
As nearly all the biological processes concern enzyme reactions and as it is well
established that the response time of this kind of chemical reaction is related to the

228
level of external factors (e.g. temperature), it may be necessary to Tnodify the time-
intervals of integration for another area.
To sum up. adjusting a model for use in other places is impossible without knowing:
a) the variation intervals of all the variables in the area where the model was buiIt
and also in the new area,
b) the growth rate laws and the response times of plants in the range of variation
found in both areas.
Moreover. there must be the same possibility of collecting meaningful data in the two
areas and this is not often the case.

2.2 Relationships helween variables

Each region not only has its own variation intervals of variables but also its own
relationships between variables.
For instance, according to Pemopn's formula*, the potential evapotranspiration
(I.E.f.) is the sort of a radiative term and an advective term. Now the weight of the
lattcr varies according to the area: on an average, advection has a more important
weight in the Mediterranean area than in the temperate climate (Seguin [//]). So the
relationship between P.E.T. and sun radiation varies from one region to another. And
it is the same for most relationships between any climatic factors.
This change in relationships is also evident for agronomic and technical factors: for
instaice in France there is a close relationship between the winter wheat cultivar used
by the farmers and the amount of fertilizers (Striziks and Pots [12j :. but it has been
shown to differ according to the area (011iiier [10] ;,
As a consequence in order to adapt a model to a different area it is necessary that the
interaction between two factors in the behaviour of a crop can be taken into account
independently of the relationships between these two factors in the area for which
the model was built. This cannot be asserted till all the parameters which are assumed
to be constant are proved to be so and not to be in fact depending on some other
factors. It is hardly possible to guarantee this.

2.3 I)eelopmenl las

in lost growth niodels. development stages are state variables considered as external
records. This information can be given by a development model, but unfortunately
such a development model also lacks generality.
For instance, the number of leaves (n) appearing on tie main stems of Maris Htnt sman
winter wheat is

- in England (Gallagher [3]) n = 0.9 10 2-O.8

- in France (Brittany) (GraiLer [4]): n= 1.2 10-2'0-3
where Xv0 are the accumulaied iemperatures above zero since sowing time. Thus. while
the change in climate is quite small, the laws are very different. On the other hand
Vincent /14], Delcolle and Gurnade /I. (4f Figure 2) find laws very similar to

* Which can be simplified as follows: P.E.T,-a.Rn I-b Ea. where Rn is the net radiation
and Ea the atmospheric drying power (advection)

229
 U)W
(n U

Eto 0
LZL

N 3D tnO
NN N

o) * 04 Cr)

o4 0,1

0 tnci
o

0
* 444 t

0 44 '
0
230
00 4 O

- * to
'4. .4.

- 1 D 01

Fig. 2. Leaf number of a winter wheat (Talent) as a function of accumulated temperatures

(from Delecolle and Gurniade [1])

230
Gallagher's but for different culiivars and above all under quite different climates. The
differences between (he :0 coefficients are quite reasonable and can easily be explained
by the change of temperature ranges (Dwand [2]): so. an accumulated temperature
law has but a local significance.
On the contrary, the difference between the Goiger's constant term (-3) on one hand
and De/colle and Gurnade's (-I) or Gallagher's (-0.8) on the other can only be ex-
plained by the difference in methods of observing leaf appearance.
Anyway, the presence of this constant term is disturbing: the logic of the accumulated
temperature law would rather lead us to assume relationships such as: n -k Z0 with
n=o if E0=o.
Thus there is a delay in leaf appearance and in order to explain it DeleCcolle and
Gurnale have to suppose an 'endogenous' cause. In fact, the explanation probably
lies in the statistical nature of the accumulated temperature law which makes no
attempt at analysing the morphogenesis processes, de Witt and Goudriaan [16] and
ML [9] emphasize this lack,

3. New approaches and new model schemes

3.1 Building and using areas of analytical models

The preceding paragraphs 2.1 and 2.2 lead to the conclusion that real generality of an
analytical model cannot be achieved without covering the widest possible variation
intervals of factors. As a consequence, building models specific to Mediterranean
areas seems absurd.
On the other hand, analytical investigations in these areas coordinated with studies in
the temperate climate can undoubtedly widely increase the generality of models. In
France such a CoUse was chosen for winter wheat in 1975. However wheat is not
economically important in the French Mediterranean area but we thought that wider
ranges of factors were absolutely necessary, even though the results expected would
be more useful inthe temperate areas.

3.2 Difference betneen grosth and development

Four different phenomena make a plant organ (e.g. a leaf) visible to the naked eye:
1,The initiation of a priniordim at a date which depends on the initiation rate of
the preceding organs,
2. A period of delay (latency), which is in fact a more or less complete blocking of
cellular divisions. For some organs, this period is hardly ever appreciable.
3. The period of cell division.
4. The cellular elongation, necessary to make the organ visible.
Ifwe wish to distinguish clearly between development and growth. we I ust only call
development the two first phases (initiation and dclay), because obviously the other
two are growth phases. Moreover the difference between development and growth is
only meaningful at tlie level of each organ. So. it is then better to use the word 'organo-
genesis

231
Let us remark that an organ is sometimes initiated in several steps, and that the growth
of each of its primordia follows the same chronological order: such is the case of the
internodes of the wheat stem (Vincent [14]). Therefore, 'elementary organogenesis
cntities' must be considered every time to see a clear boundary between growth and
development.

3.3 Working system of organogenesis: case of wheat

This level of investigation leads into modelling the biochemical processes concerned.
That is the way chosen in France on Vincent's suggestion ( Vincent [14], Vincent and
Malet [15]). Obviously this method of investigation is much slower than the simple
use of thermal time (accumulated temperatures) because we need to observe and
analyse organogenesis step by step, both under controlled conditions (Malvoisin and
Vincent [8]) and in the field (DeNcolle and Gurnade [1]).
According to these researches, each organogenesis step of wheat (for leaf initiation as
well as for a development stage of the ear) depends on the working of a control centre.
Wheat organogenesis can then be described as a sequence of centres which start
working each in its turn without any break up to the last (i.e. to that which starts
pollen meYosis).
Each of these centres (in the medullary axis) has a dual activity:
- An initiation activity by ordering impulses towards the territories of organogenesis
(at the bottom of the apex) due to a stock of activators.
- A control of cellular division in the previously initiated organs by a group of
inhibitors. The rate of consumption of these inhibitors determines the working span
of each centre and, as a consequence, the duration of each organogenesis'phase.
During the vegetative.period (Figure 3) each control centre initiates (at the bottom
of the apex), first a leaf primordium (with its bud which may become a tiller stem) and
also a future control centre. So, during the whole leaf initiation period, the number of
future control centre above the working one remains constant (five for wheat).
The activator stock progressively increases from one control centre to the next, whereas
the inhibitor stock remains nearly constant. So, when the ratio between the former
and the latter reaches a genetic threshold, the control centre which starts working at
this moment orders the initiation of the last leaf and of the last control centre. The
number of control centres necessary to reach this stage can obviously vary a little from
one year to another in field conditions. As the activator stock of the five waiting control
centres above is at its highest level, the inhibition stock of the working centre is, so to
speak, cancelled. Then the tiller buds are set free and they start their own growth: that
is tillering. The beginning (At) of this important period is marked by the appearance
of some alternately white and green stripes under the apex. The apex is set free also and
starts flower initiation by the double ridge stage.
The next five control centres which are the five nodes of the stem will not order the
initiation of any new leaf or control centre, but will order the five last stages of the
ear organogenesis. Moreover, they will make the five internodes grow each in its
turn: that is the shooting period (Figure 4).
Thus, this organogenesis can be represented as a sequence of tanks (Figure 5). The
maximum storage capacity of each is fully determined by the genetic programme. The
rate at which each tank is used up determines the timing of organogenesis: it depends

232
on the temperature and the red and infrared light penetrating cellular tissues of leaves
and even of the medullary axis. That means that tie external factors which influence
organogenesis are: air temperature. light and any factors which affect the transmission
of temperature and light inside the cellular tissues: particularly the plant structure,
hence the indirect influence of mineral and water supply and of net photosynthesis.

EAR
PRIMORDIUM

I +
I
* a I
WAITING -
CONTROL * L EAVES
CENTRES . . PRIMORDIA

a a
n + I

n LEA I
-------------------------------.
CONTROL

CEN TR E
DRAINED CFR'A V-LLVEM LEAF
CONTROL
CENTRE - - - --
n-I
CENTRE a I a n-I
CONTRO I

TEMP. - PHOTOPERIOD
t
WATER STRESS

-WHEAT VEGETATIVE DEVELOPMENT-

[ig.3. Vegetative organogenesis of wheat

233
 >

S_- _DEVELOPING
. .EAR

I .... 4--- -- 1

FLOW RECORDER
I S I

- aI///A m 511
V1IIIIII11111111111111
9 '1LE S I

SI PRIMORDA

GROWING li
.INTERNODE

NODE r ------ INDEPENDENr

L rl mIn LEAF

PHO TOPERIOD

-WHEAT FLORAL DEVELOPMENT.

f g.4. Apex organogenesis of wheat

234
 PHOTOPERIOD ----
TEMPERATURE <---I

WA TER STRESS
----
L LEAVES

Ii I
I I

TI TO
ACVA INITO WOKN

INHIBITOR WORKING
ACTIVATOR TANK
TANK CONTR

ACIATRINHI BI TOR PRECEDING

ACTI AKO TA NK CON TROL

1! 6LEVEL CONTACT
,.ol CENTRE

-WHEAT DEVELOPMENT MODEL - (DETAIL)_

Fi.5 . Ana logic scheme of working control centres diring ogailnogenesis o I wheat

235
Such a model obviously has a very wide range of use: it will allow us to analyse the
consequences not only of climatic variation but also of differences in nutrient supply
and above all, of variations in structure linked to genetic variations. It must be
emphasized that this hope is mainly based on the logical scheme and not on mathe-
matical equations and this is very important. Besides, these equations have not yet
all been determined, as this requires refined and unusual observations of organogenesis
in great number. That is an obvious drawback.
The necessary field observations for winter wheat were started in France in 1975 at
four different sites in each year - two in the North and two in the Mediterranean area.
About two more years will be needed to gather all the equations.

3.4 Relationships between growth and development

Thanks to the preceding model very close links between growth and organogenesis
have been brought to light: for instance, synchronism between the growth of the stem
internodes and ear organogenesis.
It is also possible to show a statistical link between growth and development of plants
within a crop (Malet [7)). Set ti the end of the development phase (including initiation
and delay period) for a plant (i) within a crop. Let us suppose that the variability of
these starting times has no consequence in growth rates of the elementary organo-
genesis entities which have just appeared (see figure 6a); we can then express growth
Yi (t) as follows:
yi (t)= F (t-ti)+ ei (t) (1)

where the function (F) stands for the growth law common to all the plants and ej (t)
the residual error. The mean-square and the skewness coefficient (g,= m 3/mru/2 ) then
develop in time as shown in the Figure 7.
As a consequence, even if the organogenesis variability (i.e. of starting times ti) has
no influence on the individual growth laws, it nevertheless affects their variability- But
the most frequent case is represented in Figure 6b, where a difference in ti changes the
growth law. So, organogenesis must be taken into account in growth models by means
of feedback.
As an alternative, we can choose a development model according to a feedforward
scheme, i.e. 'the utilisation of a predictive model to modify the present behaviour of
the system on the basis of an anticipated future effects' (Rosen [13]). Now, the model
(1) is of the predictive kind when growth conditions are at their best (i.e. without the
interference of limiting factors). Because in that case (Malet [7]), the first order
and second
an seo orderL
od derivatives of the mean growth (y according to time can be

determined at every moment from the mean-square and the third moment of the
growths and of the starting times (ti). This very interesting predictive property points
out the well known priority of organogenesis for the growth of living creatures.
Under less propitious environmental conditions, as in the Mediterranean climate, this
property can vanish: the mean-square and the third moment no longer run their
theoretical course; but by analysing the different evolutions, the periods when one or
several environmental factors become limiting can be determined precisely.

236
 Growth Growth

ti m

fg,6 Influences of starting tirme ((,) on growth: a) case of a delay as expressed by the relation
(I) in text. b) the geieral case

Growth Mean.square Skewness

coefficient

ime time time

Its ts

Fig. 7. Evolution of statistical parameters of growth according 1o the relation (1)in text
(from Malet [7J)

4. Conclusion

Despite the specific characters of the Mediterranean areas as regards climate, crop
systems and agronomic techniques, Ihere is no need for a diiflerent approach to the
building of analytical and physiological plant models.
The change From one area to another only leads us into studying organogenesis
processes and their links with growth more closely. Moreover a statistical model points
oUt that the relationship between organogenesis and growth is the key process in plant
adjustmenl to change in the environment.
Obviously this analytical approach takes time and is expensive; but it has very impor-
tant and useful issues: for instance, simulating adjustment capacilies of cUtivars
selected for a given area to other environments would obviate the need for long and
expensive experiments and study of behaviour in every area whenever a new cultivar
is introduced.
This aim demands the building of models which clearly inClUde all the plant structre
parameters which can vary from one Cultivar to another. This leads us first into

237
observing and minutely describing each stage of organogenesis, then into considering
as closely as possible the biochemical processes which rule over it. This approach is
the best possible way for building general models which take into account both growth
and organogenesis.

5. References

1. DelNcolle R. and Gurnade J.C.: Liaisons entre le ddveloppement et ]a morphologie du

bl tendre d'hiver. 1.Stades de d6veloppement de ]'apex, apparition des feuilles et crois-
sance de ]a tige. 20 p. Preprint Station Bioclimatologic INRA 84140 Montfavet/France,
1980
2. Durand R.: Action de ]a tempdrature et du rayonnement sur la croissance. Ann. Physiol.
Veg. 9, No.1, 5-27 (1967)
3. Gallagher J.N.: Field studies of cereal leaf growth. 1.Initiation and expression in relation
to temperature and ontogeny. J. Exp, Bot. 30, No. 117, 625-636 (1979)
4. Granger Y.: Relations spatiales en culture de bW6 tendre d'hiver sous climat atlantique.
146 p.+Annex. Thesis E.N.S.A. Rennes/France, 1979
5. Malet Ph.: Perspectives et intdrat de d6veloppement pour ]a Communauti Europ&enne
des moddles agromdt6orologiques de prevision de rdcoltes susceptibles d'intdgrer les
diverses donn6es acquises par t61ddtection. 284 p. Rapport C.E.E. Station de Bio-
climatologie INRA, 84140 Montfavet/France, 1978
6. Malet Ph.: Future prospects of crop-weather models and use of remote sensing data.
15 p. Preprint ISPRA COURSES, Applications of remote sensing to agricultural pro-
duction forecasting, Ispra Italy. Oct. 15-19 1979. (in press) (1979)
7. Malet Ph.: Liaison statistique entre te d6veloppement et la croissance chez les plantes.
Application dun mod6le simpifi6 6 f'analyse de ]a croissance. Ann. Agron. 30, No.5
415430 (1979)
8. Malvoisin P. and Vincent A.: Croissance et organogdn6se du bld: Analyse de deux situa-
tions. 19 p. Preprint. Le Tocsin du Radiateur (in press) (1979)
9. Nix H.: Simplified simulation models based on specified minimum data sets: the cropval
concept (11). 35 p. Preprint ISPRA COURSES. Application of remote sensing to agri-
cultural production forecasting Ispra, Italy. Oct. 15-19 1979 (in press) (1979)
10. Offivier M.: Utilisation des facteurs agronomiques dans I'6laboration de moddles agro-
md(torologiques r6gionaux de prdvision des rendements du bld d'hiver. 75 p. Thesis Univ.
Sciences et Techniques du Languedoc. Montpellier/France, 1979
I. Seguin B.: Estimation de I'E.T.P. en climat m6ditdrran&en du sud-est de ]a France:
Comparaisons des m6thodes A l'6chelle locale et probl&mes pos6s par 1'extension Z
rNchelle r6gionale. La Meteorologic 6th Ser., 11, 33-40 (1977)
12. Strizycks S. and Pous B.: Etude des facteurs influant sur les rendements en bld tendre.
Cahiers de Statistiques Agricoles No. 22, 41-57 (1975)
13. Rosen R.: Feed forwards and global system failure: a general mechanism for senescence.
J.theor. Biol. 74, 579-590 (1979)
14. Vincent A.: Sur la logique dv Bid, p. 34+fig., Note interne Laboratoire Croissance et
Ddveloppement INRA BP 1540, 21034 Dijon/France, 1977
15. Vincent A. and Malet Ph.: Principes de fonctionnement intdgr6 du bId. Le Tocsin du
Radiateur 5, 282-288 (1977)
16. de Wiit C. T. and Goudraan J.: Simulation of ecological processes, p. 175, C.A.P.D.,
Wageningen, 1978

238
A Model for Temperate European
Vegetative Crops
G.K. Hmansen,. Royal Veterinary and Agricultural Universitiv. Department of Crop Hus-
bandry and Plant Breeding, Taastrup/Denmark*

Sun,mary

A model was developed from experimentally estimatcd parameters, and used in planning
new experiments, By alternating experimental and modelling phases, the model is continually
revised.
Since the availability of water to crops is a most important determinant of yield within a
region, a continuous dynamic model describing water state and fluxes in soil and plant to
atmosphere "'as constructed. The model was based on soil physical and plant physiological
parameters deternined in laboratories, and on climatic parameters. Some important input
parameters are: soil water characteristic curve, soil hydraulic conductivity, root growth
and distribution, plant resistance to water uptake. plant water characteristic curves, leaf area
ndex, plant resistance to water-vapour flux, plant growth (i.e. photosyntlhesis+respiration).
Climatic input parameters are: global and net radiation, temperature, air humidity, wind
velocity, rate and amount of precipitation. The soil profile is divided into 10 cn layers within
which root and soil characteristics are assumed to be evenly distributed. Time intervals are
0.002 days. The output potentials are many, the most important perhaps being evapotranspi-
ration, soil and plant water status, and plant growth. Field tests were made over two growing
seasons. Changes in soil water content were measured by a neutron probe, evapotranspiration
and photosynthesis by tie Bouen ratio method, and plant water status by the pressure boob
technique, Inputs were tie actual climatic paratneters measured. Stimulated and measured
results of evapotranspiration agreed were for both wet and dry periods whether expressed
as cumulative or 30 minute averages.
The model has been further developed to consider distribution and utilization of assimilates.
'his part of the model is based on first order rate constants and on the yield coefficients for
conversion of assimilates into strUctural plant materials. The parameters involved were
determined separately for above and below ground plant organs.

1. Introduction

[n connection with a co-project, Sinuated Plat Production', a dynamic continuous

simulalion model of water state and transpiration in the soil-plant atmosphere system
was developed (Hasen [1975]). in this model crop growth was especially related to
transpiration rate and dry matter already acCUIulated. The model has been further
extended for photosynthesis, respiration and growth. The whole model complex is
named Hejidal' (in Nordic mythology: Hejmdal guarded the Rainbow Bridge, and

* G.K. Ilasen. Royal Veterinary and Agricultural University. Department of Crop Hus-
bandry and Plant Breeding, Agrovej 10. DK-2630 Taastrtup/Dcnniark

239
was able to hear grass growing). Hejmdal is a state variable model based on the
assumption that the state of the system can be quantitatively characterized at any
moment and that the changes in the system can be described by equations (Figure 1).
The model was originally written in Dynamo - a continuous dynamic simulation lan-
guage, where the basic tool is integration by Euler's method. The time step used is
0.002 day. Hejmdal is based on physiological parameters of Italian ryegrass, and
considers the water relationships, the carbon balance and vegetative growth. The model
is made up of two parts, one of which treats the water state and flow in the soil-plant
atmosphere system, the other treats photosynthesis, respiration, utilization of assi-
milales and growth.

PRECIP ITA ITON

A ERCETED U RADIAION D

SPHERE

CONTNT RAT IL ORDE ROO

KU I ISAIB
.N
I-'NI EOTNTlONR

POTN
TMO DRAI TETAT

I
ER AC
TEWATPR k VAPORAT
frATER OI.~ tI
WR TOH

OOTO
1AFA- PWATRPL
GIRN SOLWTR OETAL Qn_
XFROT ATIHNUWTERC- N
Q. ATER CONTENTT
B AALECO
T TERCEPTED
IA I
MEI TEI ANT R

A. SOILT .... METERAanl

F,g. I. Flow diagram or Hejmdal. Rectangles represent integration equations, circles auxiliary
equations. Solid lines indicate flow of materials, broken lines flow of information. Clouds
represent driving variables outside the system.

240
2. Description of the model
2.1 Soil-plant-water relationships
A detailed description and programming has been given earlier (Hansen [1975]).
Figure I using Forrester's 1 1968] symbols illustrates the model of water relationships
in the soil-plant-atmosphere continuum. Briefly, state variable or levels arc shown
by rectangles, rates of flow are indicated by rectangles. rates of flow are indicated by
'valv' symbols, and auxiliary variables by circles. Material flows are indicated by solid
lines and information flows by circles. 'Clouds' indicate soUrces and sinks - the driving
variables outside the system.
The driving variables are the macromcteorological data: Total global radiation, net
radiation. vapour pressure, wind speed, and temperature.
The soil profile is divided into a number of 10 cm thick compartments, within each of
which distribution of water and root is assumed to be uniform. Two processes of water
flow in soil are simulated:
- vertical flow between adjacent compartments and
- radial flow towards the root surfaces within the compartlments.
Vertical flow between the compartments is simulated under the influence of developing
-
gradients in matric water potential. The hydraulic conductivity, K, given by A,N N
at soil water content smaller than the air entrance value, above which K is taken
as saturated hydraulic conductivity. The paranienters A and N were found by the
method of' Millington and Qdrk [1961]. Mean hydraulic conductivity between two
adjacent compartments is taken as the arithmetic mean of the two conductivitics.
For heterogeneous soils, water characteristic curves and parameters of hydraulic
conductivities are required for each soil layer. Precipitation not intercepted by the
crop is distributed homogeneously on the surface layer, from which also evaporation
may occur. Soil water content beyond the deepest compartment is assuImed to be
constant over the simulation period. The number of compartments may be selected
in such a way that this requirement is fulfilled.
Radial flows towards the roots are based on the cylindrical model of Garcbwr [1960]
adapted to steady rate according to ,lakobsen [1974J. In this model oisture movement
takes place from a soil cylinder surrounding each single root towards the root located
in the central axis. All water taken up is assumHed to come from the soil cylinder and
no water enters the cylinder from neighbouring cylinders. The radius of the cylinders
depends on average root density of the compartment. At steady rate conditions, equal
amounts of water are taken along the radial pathway. The numerical water potential
at the root surface b,T at steady rate is:

' N(I-N) q[ln(r,r,-') - 0.5] (l-N)1l1l -N)

- 2-. A
where L,s is soil water potential (numerical), q is water uptake per unit length of root,
r, is half the distance between neighbouring roots. r, is root radius. A and N arc
parameters of soil hydraulic conductivity. Change in flow rate will cause disturbance
in steady rate. A new steady rate will be approached in accordance with an exponential
delay function. The delay time is calculated by:

(I. /(Y- q '2

D) = L,Z- (r,-r,) (2)
(d ,Jd G)q

241
where Lv is root density, Z is thickness ofs oil compartment, and d'pJdS is the reciprocal
differential water capacity. For many water characteristic curves d,/d6 can be
described by:

= ae-{3e (3)
d6

where a and l3are parameters; 0 is volumetric water content.

By means of the delay time, actual water potential at the root surface Fra is calculated.
Water uptake by roots is given by

-
p
-qrp
(4)
U
RR

where U is rate of uptake, 4p is numerical plant water potential and RR is the

hydraulic resistanc of the roots. RR has been estimated in Italian ryegrass plants
(Hansen [1974]), and these results are used in the model. RR was found to vary
with flow rate, being smallest at the highest rate of flow. To get values of RR for
each soil compartment, the whole plant value is distributed between compartments
inversely related to relative amount of roots present.
Total growth of tops and roots in this submodel is related to transpiration rate by a
state variable
d (Dm) = aT - bDm (5)
dt

where Dm is accumulated dry matter, T transpiration rate, t is time, a and b are

parameters. This empirical equation takes into account the close relation between
photosynthesis and transpiration, and will therefore also reflect the influence of water
stress on growth. The b parameter is an expression for maintenance respiration.
Total growth of roots is related to growth of tops assuming a constant top-root
relationship. When the tops are reduced by cutting, root growth is decreased according
to an experimentally obtained function until the original top-root relation is again
obtained. Root growth is divided into two parts, of which one is for horizontal
extension, the other for extension in depth. The horizontal part is divided between
the soil compartments, relative to the amount of roots already present (exponential
growth). The part for depth extension penetrates the lowest compartment at a given
rate. As the lower boundary is crossed, root growth in this compartment becomes
horizontal.
Total water content of the aerial part of the crop is a state variable, where the rate
of change is given by the amount of water taken up, minus the amount lost by trans-
piration.
Water uptake by leaves from intercepted water is not considered. Water potential is
given by an experimentally determined relation to water content (Hansen [1974]).
The model of Monteith [1965a] is used to determine transpiration. The external
aerodynamic transfer resistance is calculated according to Pennian and Long [1960]
with wind speed measured 2 m above ground surface. The roughness length and the
zeroplane displacement is related to plant height, and assumed not to depend on
wind speed.

242
The plant stomatal resistance is given by
r,= f(S) + 1.6/L + g(%p) (6)
where Si is global radiation and L is leaf area index. The parameters involved are
obtained by applying a detailed light distribution model with experimentally obtained
results of single leaves on stoma(al resistance relationships to water potential oil
light (S. E.Jensen. unpubl.).
Net radiation absorbed by the plant is calculated by
R. = R,, (1-c -1"4) (7)

where R, is net radiation above the plant.

If the soil water content of the surface compartment is higher than 0.32, maximum
evaporation friom the soil surface is calculated from:
e- 4
0.9 (IZ, 0 L)
H (A + y)

where H is latent heat of evaporation. A is the slope of the saturated vapour pressure
- temperature curve and y is the psychrometer constant. The factor 0.9 is given to
correct for soil heat flux. As the surface compartmen( dries up, evaporation is reduced,
and the actual soil evaporation is given by
E,, = Esm (BE - 0.23) (9)
where B is a constant, found from Olesen [1970].
Interception of precipitation by the plants is assumed to depend on leaf area index,
L, and the amount of water already intercepted. Itis given by
SKL
I- S-KL (10)

where S is precipitation accumulated during periods, when intercepted water is present

and K is a parameter found from Rijtetma [1965]. The rate of change in interception
is derived from the rate of precipitation. It is assumed that evaporation only occurs
as long as leaves are wet, and that transpiration first takes place, when the intercepted
water has disappeared. Itmeans that evapotranspiration is calculated according to
Moweilh [1965a] with negligible stomatal resistance dLUring interception.

2.2 Photovnhesis, respirationand growth

Liqht iterception by j/lage. For simplicity the model of At onteith [1965b] was used.
In this model, the foliage is regarded as consisting of discrete layers and the leaves
are assumed to be distributed Uni forrnly in a horizontal plane. Patefield and Atustin
[1971] modified these assumptions, as they may be true for canopies where leaf area
index. L. is large, but are not true for young seedlings where L is small. Thus when L
is less than one the assumption that all leaves are sTn11it is not strictly true. According
to Moneith [1965b], the following equations give the areas of' sunlit and once shaded
leaves in a canopy with leaf area index, L:

243
 (I-..sL)
Sunlit, A 0 - -(11)

Once shaded, A, = (J-sL-(1-s) LsL-)/(I-s) (12)

where A is the radiation not intercepted by a unit leaf area. As pointed out by Patefield
and Austin [1967] these equations give rise to inconsistencies for non-interger values
of L. To avoid this they used the following model for light interception:
L
Ao{1s11 -S _ for L>1

L for L < I

I-sL-(I-s) LSL-1 for L>2 (13)

1-S
A, = L-A for L<I L <2
0 for L < I
A2 = L-Ao-A
where A,= sunlit leaf area index
A,= once shaded leaf area index
A2 = twice or more shaded leaf area index.
The extinction coefficient, K, of ryegrass varies with L (Jensen [1977]). As K= 1-s'
s can for ryegrass approximately be given as:
s=0.75 - 0.025 L. (14)
The higher s is, the higher is A.. When s varies with L, then A0 shows a maximum at
L=4.5. Increment of L above six, is equal to increment in A2 (Figure 2).
Photosynthesis. Rate of photosynthesis F for a single leaf can approximately be
described by
F= RL + CI CO/rb (15)
el + C,/rb
where F =gross photosynthesis
RL= respiration in the light
e= photochemical efficiency
I =absorbed photosynthetic active radiation
C, = CO,-concentration at leaf surface

For sunlit leaves photosynthesis is:

FAo = F (l-s) A0 (16)
for once shaded leaves:
FA1 = F (l-s) n A1 (17)
and for twice shaded leaves
FA, = F (l-s) t' A2 (18)

244
 A0 A, orA 2 - index

ILA

10

5= 0.750 S= 0.75-026A

LA LAI

0 5A0 o
5

Fig. 2. Variation of sunlit (A,), once shaded (A;). and twice or more shaded (A,) leaf area
indices together wvith total leafUarea index (LA]I) at di fferent values ofrs.

where is the proportion of the light transmitted by the leaf (= 0.04). The lolal GO 2-
uptake Pr, by the crop is given by summation:

P= FA 0 +FA + FA_, (19)

Whereas respiration in light can not be measured directly, the light at the light corn-
pensat ion point, Ic, and the CO.,- cuncentIration at the CO. cornpensatlion point, C,
can : both of them depend on Rh. Therefore they were used in place of R as suggested
L
by AAock l. [1971]. At
S(L-I) (Ca-C=)/(ra -r'b)
F-(I-1¢) +I (Ca-C )/(r, ± rb) (0

where C2 = CO 2-concentration above the crop (2 m)

r =the aerodynamic resistance

245
 C. varies with temperature (Hansen [1977]) and with leaf water potential
(Sheehy et al. [1975]). In the model, the highest value for C.at any time is
selected. 1. varies with temperature according to:
I. = 14.9+0.4 TL (21)
According to Acock et al. [1971] is 15.10-16 g CO 2 J-'. The CO2 -concentration
was assumed to be 325 ppm two meters above the crop. The crop resistance to CO, -
diffusion is given by:
r'b = 1.57 rb (22)
The aerodynamic resistance for CO, is equal to that for water vapor. Crop temper-
ature, TL, influences C. and respiration and is given by:

TL= HLra+T
Cp pa 2 (23)

where HL is sensible heat flux from the crop:

HL = RNa - LEL (24)
CP and p. are heat capacity of air and density of air, respectively.
Growth and maintenance depend on the supply of assimilates. For model purposes
plants or crops are often considered as consisting of two carbon components, structural
and storage material (Warren Wilson [1972], Thornley [1976]). In a model of
Thornley [1977], structural material was further divided into degradable and non-
degradable material, to analyse the processes behind maintenance respiration, defined
as the respiration necessary to maintain status quo of the plant organ i.e. structural
and storage material. Maintenance respiration, one of the two respiratory components,
has been estimated experimentally from the intercept of specific respiration plotted
against relative growth rate (Thornley and Hesketh [1972]), and as the respiration
after 48 h in darkness [McCree [1974]).
It has also been estimated from two dimensional regression analyses with respira-
tion as the dependent variable, and the diurnal increment in structural material
as one and total accumulated structural material as another independent variable
[Hansen [1978]). The parameter of the latter gives the maintenance coefficient.
Defined in this way, maintenance respiration is the respiratory requirement for
maintaining the structural plant material including also maintenance of ion gradients.
Thornley's [1977] model for vegetative plant growth is based on the concept that
respiration is directly proportional to the amount of storage material or its conversion
into structural material, a part of which is degradable, and recycled through the pool
of storage material. Although Thornley on basis of this model doubted the general
concept of regarding respiration as consisting of two components, Barnes and Hole
[1978] showed that the Thornley model did not conflict with this view. The parameter
of the model can be estimated from pulse labeling plants with 14CO2, and measurements
of the respiratory loss of 14CO2 in a subsequent dark period (Thornley [1977]).
As the initial values of storage and structural material are not known, it is more diffi-
cult to estimate the parameters from measurements of 12 CO2 -effiux in darkness. Anal-
ysis has been carried out to estimate the kinetic parameters of the Thornley model for
tops and roots and to find parameters for partitioning assimilates between tops and

246
roots (Hansen, in prep.). Based on these estimates the model can be applied as a whole
plant or crop growth model, which can also stimulate the within day variation. As
models of plant or crop photosynthesis have already been established, e.g. de Wit
et al. [19701, the present model uses photosynthesis as the driving variable. A visual
display of the whole plant model is shown in Figure 3, where the parameters of tops

K, *
KT.W'WR'W
K DW
R
_

Fg.3. The model of assimilate utilization in whole ptanis. Wi0 is dry weight of recently assi-
milated carbohydrates. Ws.r and WVSR are storage pools of tops and roots, respectively
\Vor and WGR are structurat mlaterials of tops and roots, respectively KT, KR, KGT, KGR,
K OT and KDr, are fi rst order rate constants for tops (T) and roots (R). YGT and Y0GR are
the e,°iciencies of conversion of storage materials in/to structural materials.

and roots are subscribed T and R. respectively. The model has five state variables:
the pool of recently assimilated carbohydrates, Wp. the storage pool for tops and
roots, Wsr and WsR, and the structural nntcrial We;T and W01 1 . The latter were
divided into degradable and non-degradable materials in Thornier's model, a com-
partmentation, which could not be used in the present modification. Total dry weights
of tops and roots are obtained by:
WT_ WST-I- WOT-WP (25a)
WT= WsG K+WGR (25b)

Supply of assimilate to W is gross photosynthesis, P . It is assumed that the uni-

directional transports from Wp to top and root storage pools are obtained by first
order kinetics, the rate constants being KT and Kr, respectively.

PT- KTWj} (26a)

P= KR' W1 (26b)
The remaining parts of the model are analogous to the Thorite model, but without
conepartientat ion of structural material into degradable and non-degradable parts.
The formulations are equal for tops and roots. Thus, omitting the subscriptions and:
R the following first order difierential equations are valid:

247
dWs
dt- =P - K,-Ws+KD'WG (27)

dWG YG "KG 'WS - KD'WG (28)

dt
Where t is time, K G and KD are first order rate constants and YG is the efficiency of
conversion of storage material into structural material (Table 1).
Table 1. Values of applied parameters for the assimilate utilization part of Helmdal
at 200C.
Parameters References
KT ................... 1.0 Day-' Hansen [1980] (in prep.)
K t ................... 0.36 Day -- ' Hansen [1980] (in press)
KGT .................. 2.20 Day ' Hansen [1980] (in press)
KGR ....... ...... 3.20 Day-- Hansen [1980] (in press)
KDT .................. 0.151 Day-' Hansen [1980, (in press)
KDR .................. 0.079 Day- 1 Hansen [1980] (in press)
YGT ................. 0.83 G CH,O/CH2 O Hansen [1978J
YGR .................... 0.54 G CIt0/CH2 0 Hansen [1978]

3. Validation
Experiments to collect validation data were carried out at the Hoeibakkegaard
Experimental Farin outside Copenhagen, in 1973 and 1974 with Italian ryegrass for
which the model was developed. The experiments were described by Jensen [1977]
and the field trial by Hansen and Jensen [1977]. Transpiration was measured by the
Bowen ratio method (Jensen [1977]) and soil water content was followed with the
neutron probe method (Jakobsen [1977]).
Leaf water potential was measured occasionally during 24 h periods by the pressure
chamber method. Meteorological data used in the model were taken from the weather
station adjacent to the experiment. Both dry and wet periods occured in the two
seasons giving a wide range of conditions for the test.
Figure 4 shows accumulated measured and simulated actual evapotranspiration for
two periods in 1973 and 1974. The simulated values agree well with the Bowen ratio
values, but not with the neutron moderation method. The divergence of neutron
moderation value in the 2nd period in 1973 was due to a calibration change.
A more critical test is daily totals. Comparison between the two estimations of this is
shown in Figures 5a-b. The greatest divergences occur on rainy days. At the start of
the 1974 period the crop was poorly developed, which might have caused the disagree-
ment but despite this, there were periods with very close agreement.
The profile measurements for estimation of actual evapotranspiration by the Bowen
ratio method were made only from 06,00 to 18.00 hrs so that hence evaporation and
dewfall are not included in the evapotranspiration sums. However, results from a
weighable lysimeter enabled these to be estimated at 7 and 4 mm in 1973 and 1974,
respectively.
1974 measurements were made throughout the 24 hours over a 15 day period. Ac-
cumulated 24 hours values and accumulated 06.00-18.00 values for this period were

248
 mm
140
120
120/ (a) 1973 1" -

100 -
/

80 0

60 9

40 Cd- SimuLated
20-- Bowen ratio
SO---o Neutron

11 20 30 10 20
May June

mm
140
(b) 1973
120
100 -i

80-
- SimuLated
Bowen ratio

40 O---O Neutron / -/

20 -- --

V
f I I

22 30 10 20 31
June July

fig.4a-h. Measured and simulated actual evapotranspiration accumulated during two seasons.

249
 mm

160
140 (c) 1974
120 -
100 "'""

80
60
40 - Simulated
20- Bowen ratio
2 . . ---- o Neutron
10 20 30 10 20 30 9

May June July

mm

160
140 (d) 1974
120
100
80
60 .
1.0 .ro" - Simulated

20 Bowen ratio
o---o Neutron
10 20 30 10 19
July August

Fig.4c-d. Measured and simulated actual evapotranspiration accumulated during two seasons.

250
Actual evcpotrinspirotion, mm Precipitation, mm
5 50
Simulated 19 73
4 -Bowen ratio 40
1. Precipitat ion
3 30
2 .-- ' .x 20
2V to
5 10 15 20 25 31
5 May

4 K \ 40

3 30

S5 3 10 15 20 25 30 30
5J 'I une- - \ - so
2 -#II*20

-* ' . .. - 10
0

1 5 10 15 20 25 31
July

Fik.5a. Comparisons of measured daily totals of evapotranspiration with simulated resulIts.

respectively 37.3 and 36.3 mni indicating tfiat dewfull and night evapotranspiration
were substantially equal over the period.
Another critical test is comparison of half-hour values obtained by the Bowen ratio
method, with the instantaneous simulated values (Figure 6). The reSUlts shown were
obtained during a period of developing water stress, commencing at high soil water
content and ending at severe water deficit. On the first day (May 29) there was a phase
displacement in the diUrnal cycle between the measured and the Simulated values. the
lattet being delayed. On June 13 the trough in the middle of the day was (lie to
clouded sky. On the last days (.1-une 19 and 20), both Simulated and measured values
increased steeply at Sunrise due to intercepted dfew. When the dew disappeared, sinu-

251
o

o
0 >

\C

.j/ -V

-6 ~ ~ ----

Fig. 5b. Comparisons of measured daily totals of evapotranspiration with simulated results.

lated values decreased more than measured ones. These different courses were due
to the assumption in the model that no transpiration through stomata occurs as
long as intercepted water is present.
Comparison of moisture profiles obtaine by the neutron moderation method with
those by the simulation model on selected days during the growth season are shown
in Figure 7. The measured results were higher in the uppermost layer, where the
neutron moderation method presents difficulty. In spite or these discrepancies there

252
 8
MAY 29th 1973 ,'- -

MAY 30th 1973

8
2

0 12 18 24
4-

2 N\ I \
6-

JUNE 13th 1973

2

2
E
o I2 ' 24
JUNE 19th 1973

2-

0 61 21 18 2T4
6- JUNE 20th 1973
__ -- Simuloted

4 ," Bowen
2-

...
0 6 12 18 24
Hours
F/e. 6. Comparisons of measured half hour values of evapotranspiration rates with instan-
taneous simulated results.

253
 0

May 24th June 20th +

so - 1974 1974 +

/4

120. E+

0 1+

- 0 I I t I I I

40 "-.

July 16th "July 30t

80 1974 I 1974
5 10 15 2t5 3 1 5 2 5 3

I -- -
120 - -- +--, Measured
iuoe- neutron I1

I I I I I I I
5 10 15 20 25 30 5 10 15 20 25 30

Volumetric soil water content

Fig.7. Soil moisture profiles as measured by the neutron moderation probe and obtained
from simulation.

was close agreement. The sharp change'of water content at a depth of 40-50 cm
obtained by both methods, shows the lower boundary of the root profile.
On selected days, leaf water potential was measured by the pressure chamber method.
Experimental values for single leaves are compared with the result of simulation in
Figure 8; it should be pointed out that the latter represents average plant water
potential.
Considering all the parameters and variables determining plant water potential, the
agreement obtained was regarded as being satisfactory.
To validate the model for photosynthesis, CO,-fluxes measured above the crop
were used. These were integrated half hour values, while the simulation outputs were

254
Bar Bar
- 1s May 29-30 th 1973 t % / r30 - 30 - June 21 th 1973 *

¢
i
+

/
/

-0
1 -15
18 0 5 12 i18 5 12 is 24
Hours Hours
Bar
-5 Bar
-2 +
May 30 th 1974,
-20 June 25 th 1971.
I VV

+t
1X -20-
-10 +%"

1- 0
-5t 6 12 124

Bar Hours
0 12 18
August 13th 1974
-10 Hours
A

0 24* 5 12 18 24
Hours
Fig.8. Measured values of single leaf water potentials
compared with simulated values of
crop water potentials.

255
instantaneous values, but the driving radiation was based on hourly integration. Thus
the simulated results were also smoothed, especially on days with fluctuating radialion.
The CO2 -fluxes measured from 6 a.m. to 6 p.m. are compared with simulated and
measured values in Figure 9 which also shows global radiation. On May 25th 1973
with a small crop and fluctuating radiation the CO,-flux showed much fluctation, but
agreement was close. As the crop grew e.g. June 6th and 7th, simulated values were
higher than the measured ones. On these days radiation was high and there was no
water stress. The highest photosynthetic rates were obtained at 10 a.m. for both
measured and simulated results, although radiation was highest at 12 o'clock. From
13th to 20th June water stress developed and photosynthesis was small but agreement
was close. During the period of water stress, photosynthetic rates at the start of the
day were high as on days when water supply was sufficient. Dew, intercepted during
the night caused full turgidity, until removed by evaporation. The crop was cut
June 21st.
The resuJts show that the model can mimic the diurnal trend of photosynthesis
especially on days when photosynthesis is limited by low radiation or water stress.

May 29th 1973

120 o 1,2
May 25 th 1973

80: 800 ' "LAI=1,38-- "LAI= 2,46 0,8

E

40 A L, o
0 II I 0
0
LAI= 6,00
June 6th
U.120 1973
o' May 30th 1973 / Fa
12<

04
t f
80 .. LAI=2,96- - 0,80
40 ,,0,4,

0 Lj I I u I n

6 12 18 6 12 18
Hours Hours

Fig.9a. Simulated and measured CO-fluxes and global radiation.

256
When these are not limiting, the model overestimates photosynthesis. This is due to
the assumption that CO,-concentration remains constant at 325 ppm, two meters
above the crop.
This may not be true as CO,-concentration at this height is influenced by the rate of
photosynthesis, especially when wind velocities are small. Better agreement might be
obtained by assuming constant CO, concentration at a greater height.

June 7th 1973 June 13th 1973

120 1.2
LAI= 6.96 LAI =10,46
80
80 0,8

44

0 \
IoI _ _ _I _ _ _60 _ _ _ _ a
120 June 17th 1973 June 19th 1973 1.2
0 0 E
E_ LAI I.8

eoo g 00- U'? .O

X LAI 12.3 z
40 9. < 04"
7 .40 U0 4
0._

U- /,,

0 I*' 4-iII0
120 June 20th 1973 -June 29th 1973 o 1.2
0

so 0.
80 g ' 0.8

0
Y LAI=12.3
40 o * 0,

K ,
6 12 18 6 12 18
Hours Hours

Fig.9b. Simulated and measured CO,-fluxes and global radiation.

257
Small plots were harvested weekly to measure dry matter production. The harvested
yield agreed closely with integrated values of photosynthesis and respiration for the
first month but later simulated values were higher (Figure 10). One reason for this is
the above mentioned discrepancies of photosynthesis. Another reason is that the model
made no assumption of scenecence and death of leaves.
The assimilate utilization part of the model has not been tested in the field. Assuming a
Q 0 of 2 for the temperature variation in the kinetic parameters the behaviour of this
part of the model has been tested at 10, 20 and 30'C (Figure 11). The stored materials
and hence also respiration show diurnal variation with a single peak about 16 hours
after commencement of the light period and most pronounced at high photosynthesic
rate and high temperature. Two diurnal peaks of root respiration have been found
experimentally one 6 hours the other 16 hours after the start of the photoperiod
(Hansen [1977, 1980]). Thus the model can at present partly mimic this diurnal
pattern. To get the other peak, further compartmentation of the storage pools may
be necessary.
The simulated top/root :ratio is smallest at the highest temperature (Figure 12). This
pattern is also known from literature.

4
'0

Z
0
-3
W

-
2
.C

e Simulted
o-- - Harvested

I
11 15 20 25 30 1 5 10 15 20
MAY JUNE
Fig. I0. Simulated and harvested dry matter yield of tops.
258
 T=300c I

0.5

04 G 2
PF PG=:1
I /
I', ,\I I
0.3

0.30 I \

020I

0.15
0.20 I 0

o,-I -----
-- - -
.
I_- =---.--.
I.
. . .....--
-
"

0.10
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
DAYS
Fig. II. Simulated variation of respiration in tops (- ) and in roots (- ---- ) at three
different temperatures and at two different diurnal photosynthetic supply of assimilates.

259
 6.5
0N

5.0

C3
2c)3.5

2.0
0.0 0.5 1.0 1.5 2.0 25 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Days
Fig. 12. Simulated trends in top/root ratio at three different temperatures.

4. Conclusion

Hejindal should not be considered as a complete model; it is still under development.

It is used as a tool by which consequencies of experimental work may be obtained and
parameters can be investigated, and it is a guide for planning experiments. The next
step in the development should be to include the uptake and assimilation of nitrate.

References

Acock, B., Thornley, J. H.M. and Warren Wilson, J.: Photosynthesis and energy conversion.
In: Potential Crop Production, pp. 43-75 (P. F. Wareing and J. P. Cooper, eds.) Heine-
mann Educational Books, 1971
Barnes, A. and Hole, C.C.: A theoretical basis of growth and maintenance respiration.
Ann. Bot. 1217-1221 (1978)
Forrester, J. W.: Principles of systems. Wright Allen Press. Cambridge, Mass., 183 pp., 1968
Gardner, V. R.: Dynamic aspect of water availability to plants. Soil Sci, 89, 63-73 (1960)
Hansen, G.K.: Resistance to water flow in soil and plants, plant water status, stomatal
resistance, and transpiration of Lolium multflorum as influenced by transpiration demand
and soil water depletion. Acta Agric. Scand. 24, 83-92 (1974)
H.nsen, G.K.: A dynamic continuous simulation model of water state and transportation in
the soil-plant-atmosphere system. 1. The model and its sensitivity. Arta Agric. Scand.
25, 129-149 (1975)
Hansen, G.K.: Adaption to photosynthesis and diurnal oscillation of root respiration rates
for Lolium niultiflorum. Physiol. Plant 39, 275-279 (1977)
Hansen, G.K.: Utilization of photosynthates for growth, respiration and storage in tops and
roots of Lolium multiflorum. Physiol. Plant 42, 5-13 (1978)
Hansen, G.K.: Diurnal variation of root respiration rates and nitrate uptake as influenced
by nitrogen supply. Physiol. Plant 48, 421-427 (1980)

260
Jakobsen, B.F.L.: Water and phosphate transport to plant roots. Acta Agric. Scand. 24,
55-60 (1974)
Jensen, S.E.: Milling i marken for testning af modellcr. In: Simuleret planteproduktion
(Hansen G.K.. Jensen, S.E. and Jakobsen, B.F.L., eds.), D.S.R. The Royal Vet. and
Agric. Univ., Copenhagen, 1977
McCree, K.J.: Equations for the rate of dark respiration or white clower and grain sorghum,
as functions of dry weight, photosynthetic rate and temperature. Crop Sci. 14, 509-514
(1974)
Milington, R.J. and Quirk, .. P.: Permeability of porous solids. Trans. Faraday Soc. 57 (7),
1-8 (1961)
Monteith, J. L.: Evaporation and environment. In: The state and movement of water in living
organism (Fogg G.E., ed.), pp. 205-234, Acad. Press, London, 1965a
Monfeit, J.L.: Light distribution and photosynthesis in field crops. Ann. Bot. N.S. 29,
17-37 (1965b)
Olesen, S.E.: Evaporation og vandtransport i relation til jordens overfladestruktur og
vandindhold. Ph. D. Thesis, The Royal Vet. and Agric. Univ. Copenhagen, 1970
Penman, I. L. and Long, L. F.: Weather in wheat. Quart. J. Meteor. Soc. 86, 16-50 (1960)
Reitena, P.E.: An analysis of actual evapotranspiration, Wageningen Agric. Res. Rep 659
(1965)
Sheehy, J.E., Green, R.M. and Robson, M.J.: The influence of water stress on the photo-
synthesis of a simulated sward of perennial ryegrass. Ann. Bot. 39. 387-401 (1975)
Thornley, J, H.M.: Mathematical Models in Plant Physiology. Acad. Press. London, 318 pp,
1976
Thornley, J. H. M.: Growth and maintenance respiration: A re-interpretation. Ann. Bot. 41.
1191-1203 (1977)
de Wit, C. T., Brotwer, R. and Penning de Vries, F. W. T.: The simulation of photosynthetic
systems. In: Prediction and measurement of photosynthetic activity. Proc. Int. Biol.
Programme. Te:hn. Meeting Trebon, pp. 47-70, PUDOC, Wageningen, 1970

261
The Mathematical Expression
of Crop Response to Inputs
J. Wood, Statistics Department, Rothaisted Experimental Station, Harpenden, Hertfordshire/
England-

Sunmary

Some of the curves commonly used to express crop response arc listed and described. Possible
motivations for fitling models are discussed and shown to have important implications both
for selecting a mordel on general grounds and for matching a model to a set of data. Whilst no
single function is proposed as being the best, the advantages or' the split-line, in terms of its
ease of interpretation, and the quadratic-over-linear, in terms of its flexibility, are mentioned.
In the end though, the investigator must be guided by his own set of data, his aims. and any
a-priori knowledge about the likely form of the response. Experimental design is identified
as being of crucial importance in response work.

1. tntroduction

The mathematical expression of crop response to inputs covers a large area. 'Response'
itself can be anything from yield to nitrogen content and 'inputs' could refer to. among
other things, fertilizers, seedrate or the joint effect of several factors. The data available
to the would-be modeller could come from experiments performed either in the field
or the laboratory, or have been collected in a survey.
[n order to simplify matters, the situation considered here is the investigation of the
yield response to a single nutrient through field trials testing several levels of the
fertilizer under consideration, and the subsequent description of the data by simple
mathematical functions. Thus, whilst many of the following observations are quite
general, others will not apply or be incomplete when the situation is actually very
different or more complex than that just outlined,

2. Why fit models?

Before embarking on the technical problems of fitting a model to a set of data, it is

important for the investigator to consider why he is fitting a model at all. A clear

* J. Wood, M. Sc., Statistics Department, Rothamsted Experimental Station. Harpenden,

Herts. AL5 2JQ/England

263
identification of aims at this stage should help in deciding which classes of functions
are worth considering and how best to discriminate between particular curves. Four
possible motives are discussed below: the testing of a theoretical model, the compact
summarisation of a set of data, the estimation of various statistics and as an aid in the
interpretation of series of experiments. Naturally these reasons for fitting models need
not be mutually exclusive. The main point is that some kind of reason is essential;
needless to say a vague feeling that a curve of some sort would be nice is insufficient
justification for fitting a model.

2.1 The testing of a theoretical model

If we have at our disposal a model based on theory, it is natural to wish to fit it to data
in order to assess its performance in practice. In addition, a theoretical model will
contain parameters which have a clear biological interpretation, e.g. the amount of a
nutrient already present in the soil. Fitting the model will provide estimates of these
quantities which, of interest in themselves, may also in some cases be compared with
independent derivations, e.g. the results of soil analysis.
Unfortunately this is not usually the situation in agronomy in general and in fertilizer
response work in particular. Crop response is a function of many complex and inter-
acting mechanisms which are not fully understood, whereas there is only sufficient
information in the average field experiment to justify the fitting of a relatively simple
mathematical expression. Thus the theoretical support for any potentially useful
model will, of necessity, be limited. Although it can be argued that any theoretical
justification is better than none, this particular aspect of the matter will not be given
much emphasis here but will be covered rather by consideration of the extent to which
the behaviour of a model is biologically plausible. In other words, modelling is seen
as being a means to an end; the models themselves are not regarded as being necessarily
'true'.

2.2 The summarisation of data

It is quite feasible that an investigator may wish to fit a model simply to obtain a
compact summary of his data. This is a perfectly reasonable, if limited, aim; a response
curve along with an indication of its goodness-of-fit is a more parsimonious description
of a set of data than a list of the mean yields at each level of fertilizer, and allows for
interpolation where required. If this is all that is desired, it is only necessary that the
model should be an adequate representation of the data. It should fit well, faithfully
mirror identifiable features of the data and it should not produce features of its own
that are not supported by the data. Naturally it will sometimes happen by chance alone
that a set of data indicates a pattern that previous experience will lead us to dismiss.
As a trivial example, an experiment might show the rate of response to fertilizer as
being slightly greater at high than at low rates of application. Here it would make little
sense to fit a curve that reflected this apparent effect - a straight line would be the most
appropriate model.
Obviously, if the size of the response observed in a set of data is of the same order of
magnitude as the standard error of the mean yields, then the results are better sum-
marised by a simple statement of this fact rather than by fitting a curve. All that the
latter is likely to achieve in this case is a spurious impression of accuracy.

264
2.3 The use of models in estimation

[n addition to requiring a model as a simple stinmary of his data, an investigator will

usually Wish to use it to estimate various statistics of interest. Examples of such
statistics are the economic optimum dressing, the average rale of response up to the
optimum, and the maximum yield. They can be classified into two groups: a) those
that can be estimated within the range of dressings tested in the experiment and b)
those whose estimation involves extrapolation. An example of the latter is the esti-
mation of the amount of nutrient in the soil available to the plant by back-projection
of the response curve to the point of zero yield. Since estimation by extrapolation
requires a greater faith in the overall validity of the model than is in general being
assumed here, attention will be restricted to the first case,
In addition to the points made in Section 2.2 about the general fit of a curve, tile
performance of a model in estimation is now of crucial importance. Itshould not
introduce bias into the estimate being considered and, if a particular region of the
data is of special interest, tile fit of tile
model in this area becomes relatively more
critical. For example, in attempting to estimate maximum yield, one may be prepared
to some extent to sacrifice overall goodness-of-lit in the interests of a better fit in the
region of the maximum.
The fitting of a curve principally for estimation when there is at best a limited faith
inthe 'truth' of the model raises a general question: why approach the problem in
such a roundabout fashion? For instance, in order to estimate the economic optimum
dressing, one could, simply using tile raw data, successively compare the yield at each
level of fertilizer with that from the lowest level included in the experiment and choose
the dressing that maxiluised Ihe net profit. However, the use of a curve offers two
advantages here. Firstly, a model allows the estimated optimum to lie between dress-
ings actually tested. Secondly, and more important, a fitted curve Utilizes all the data
in a trial and hence should be more accurate for estimating optima than the somewhat
piecemeal method outlined above. Implicit in (he second reason given is the assumption
that all the points on the curve contain some useful information concerning the statistic
of' interest. This should be the case with, for example, a centrally located optimum,
but the assumption becomes less valid as the optimum gets close to either end of the
range of dressings tested. Inextreme cases fitting a curve will be of little benefit, For
instance, where a crop responds strongly up to the highest rate of fertilizer included
inan experiment, the most that can be said about the optimum dressing is that it is
greater than, or equal to, this rate - a conclusion that could be reached without re-
course to model fitting.

2.4 Fhe summarisation of series of experiments

It is in the summarisation of series of experiments that modelling is likely to be espe-

cially worthwhile. A fitted curve provides a simpler description of the response in a
trial than a list of yields and so facilitates comparisons between siles. Various statistics
of interest can readily be calculated from a set of curves, similarities in these between
sites noted, and an attempt made to relate differences to site characteristics.
As might be expected, the pitfalls associated with this use of modelling are also large.
In the interests of simplicity, other things being equal, it would be nice to fit a model
of the same mathematical form to all the data. This has the additional benefit of allow-

265
ing one to test whether the overall model can be further simplified by constraining
certain of the parameters to take the same values at all sites.
If a single function really does fit well to all the results of a series of trials, taking into
consideration the points made in Sections 2.2 and 2.3, then things could hardly be
better. However this is usually too much to expect, which leaves the investigator with
the problem of how many different models he is going to fit, and which of these is
most appropriate to the results of each experiment. If possible, it is convenient to
choose functions that are related to one another. For instance, a single flexible model
capable of representing a range of responses and collapsing readily into one or two
simpler forms when required is worth consideration.
It may be possible to group the sites in a series by factors likely to affect response,
such as rainfall and soil-type, in the hope that the form of response will be more
consistent within the groups, which can then be treated separately for the purpose
of model fitting. This approach should certainly be tried if feasible.
To sum up, the main emphasis in this sort of work is on the overall performance of a
model. For instance, consideration of whether a particular curve leads to bias in
estimation over the sites as a whole becomes critical. The final aim is to find the impor-
tant patterns in a large set of data and the investigator will be prepared to sacrifice
goodness-of-fit at individual sites to some extent in the interests of fitting as few
different types of model as possible. This is fair enough - so long as real and important
inconsistencies are not ironed out in the process and, since there are no general rules
for deciding whether or not this is the case, model fitting in this context remains
something of an art.

3. Some common response curves

This section contains brief descriptions of some of the models commonly used to
express crop response to fertilizer, followed by a discussion of the relationships
between them and their various strengths and weaknesses. Where mathematical
formulae are given, y represents the crop yield, x the amount of fertilizer applied, and
a, b, c and d the parameters to be estimated. These expressions are given in a general
form. In addition, certain constraints on the parameters are necessary in order that
the curves assume a 'sensible' shape over the range of interest. For example, some will
take only positive and others only negative values. Whilst it would be inappropriate
to explicitly state all of these constraints here, it is assumed throughout that they
apply; naturally only the 'sensible' shapes are of interest in the present context.

3.1.1 Polynomials (Figure ]a)

This class of functions includes:
the straight line y=a+bx
the quadratic y=a+bx+cx2
and the cubic y=a+bx+cx 2 +dx 3 .

The straight line is the simplest useful function available for describing crop response.
The quadratic may be used when there is evidence that the response is curved, but has

266
 ,. Polynomials
sir aight lin t .0lra C

h Exponentials
lLit(NX niitsab 'rih0 h
iiOI

Fig. I. Examples of reponse curves

the disadvantage of being symmetrical about its maximum. If the shape of the true
response is obviously not symmetrical, a transformation of the x axis may help.
For instance, taking square roots of the amounts of fertilizer tested in an experiment
before fitting a quadratic to the results leads to the square root curve, y=a+b0' +
cx, which shows yield rising to a maximum before falling off more gradually. Alterna-
tively, an extra term could be added to the quadratic to allow for asymmetry, which
leads to the cubic. This, however, is rarely a satisfactory model.

267
 ,.Inverse Polynomials
Y Iiear-over -Ijnear y linear-over-quadratic

quadratic -over- I near

Split Lines
,i,
two linear segments three linear sgments

3.1.2 Exponential curves (Figure Ib)

The simple exponential takes the form
y = a - bcx
It describes the situation where yield rises ever more gradually towards a (theoreti-
cally unattainable) maximum as more fertilizer is applied. It was used by Mitscherlich -
it is often referred to by his name - and later by Crowther and Yates [2].
This curve can be easily modified by the addition of a linear term to have a descending
asymptote and thus represent the situation where yield rises to a maximum and then
declines.
The modified form is y=a - bex - dx.

268
3.1.3 Inverse polynomials (Figure tc)
This class of curves is described in detail by Nelder [4]. The particular models that
will be considered here can be written:

a+ bx
y = --
I +cex- the linear-over-linear or L/L

a+±bx
andy = a dbx the linear-over-quadratic or L/Q

Also considered is a modification of the L/L proposed by Greenwood ei af. [3] which
can be written:
a+bx+cx2
y =-- 1 - the quadratic-over-linear or Q/L.

The L/L shows yield rising always towards a limiting value, whilst both the L/Q and
the Q/L describe situations where yield attains a maximum value and then declines
with increasing fertilizer application.

3.1.4 The split lite (Figure Id)

Two intersecting straight lines, or the split line, has been used extensively by Boyd
(see, e.g., Doyd et al. [I]) as a model of crop response to fertilizer.
This model is easy to visualize - yield is seen as rising linearly up to a certain level at
which there is an abrupt turning point. Beyond this point yield either continues to
rise, though more slowly, stabilizes, or declines again in a linear fashion.

3.2 Relationships between the models

The models listed above can be grouped together by their general shape and the
ntmber of parameters that require estimation.
The Mitscherlich and L/L. which are three-parameter models, both describe situations
of diminishing returns where the addition of extra fertilizer becomes less and less
beneficial, though without ever causing yield to decline. The other two three-para-
meter cUrves, the quadratic and the square root, show yield rising to a maximtum and
then declining, as do the modified Mischerlich. the L/Q and the Q/L - all four-
parameter models. However. whereas the quadratic is symnletrical about its maximum
value and the shape of the square root curve past the olaximUm determined by the
shape of the rising part of the Curve, the foL1r-patameter models are more flexible and
better able on the whole to cope with asymmetrical responses.
Naturally, the Mitscherlich curve and the L/L can be regarded as special cases of their
modified forms, the modified Maseherlich and the Q/L. The L/Q can also reduce to
the L/L and, in addition, both the Q/L and ihe modified iitscherlich can reduce to the
simple quadratic, the latter by letting the parameter c tend to unity.
The split-line, with its abrupt turning point, stands somewhat apart from the smooth
Curves though, interestingly enough, both the exponential and the inverse polynomial
models can be extended to forms approximating to intersecting straight lines (see,
e.g., Boyd et al, [1]).

269
3.3 Choice of model

Naturally, yield does not increase indefinitely with the application of fertilizer. It will
either reach a plateau, from which it may eventually decline, or it will start to decline
shortly after attaining a maximum. In the latter case, there is no reason to suppose
that its decline will be a mirror-image of its rise, a characteristic of the quadratic,
and indeed the automatic fitting of quadratic curves can lead to serious misrepresen-
tation of data. Where however an experiment provides little information on the form
of response past the turning point, or where yield rises up to the highest dressing
tested, a quadratic (or simpler still in the second instance, a straight line) can perform
perfectly well within the range of the data. When there is more information on the
shape of the yield decline, it is natural to consider one of the more flexible models
- the L/Q, the Q/L and the modified Mitscherlich. When looking for a model to
describe the results from a series of experiments, it is relevant to note that the L/Q
is unsuitable for representing a set of parallel responses and that the Q/L has the
added attraction of collapsing naturally into the simpler forms of the L/L and the
quadratic where appropriate.
With a plateau-type response, the L/L and Mitscherlich curves are obvious possi-
bilities for choice of model.
Whatever the general form of the response, the split-line can perform best of all if
the data indicates that turning points are too abrupt to be adequately described by a
smooth curve. (Where yield reaches a plateau and then eventually decreases, the
split-line may be extended by adding a third linear segment if desired.) Providing of
course that this model really does fit, the split-line has the advantage of being parti-
cularly easy to interpret. The parameters - the initial rate of response, the turning
point and the rate of response past the turning point - are of direct practical importance
and estimates of the economic optimum dressing derived from the model are stable to
price changes.

4. Practical problems

Ideally, using a measure of common-sense and being guided by the data as to the likely
form of the response, we should be able to pick a curve of the same general form and
fit it, bearing in mind that we are after a model that is as simple as possible, consistent
with being an adequate representation of the data.
However, practical problems arise. Which curve should be chosen if there are several
that fit equally well, and what can be done if the message of the data is ambiguous?
It is certainly not uncommon for a number of different models to fit equally well to a
set of data - perhaps here it will not matter too much which is chosen, though it
is worth checking whether the choice of model has a serious effect on the estimation
of any statistics of interest. However, ambiguities in a set of data will necessarily
weaken the faith we have in any curve fitted to the results.
If we are restricted to data from a single experiment, there is little more that can be
said. However, with a series of experiments we can do rather better. Averaging the
results over the sites may help to clarify the true form of response. However, this can
only work if it is reasonable to expect that the sites will behave similarly, or if they
can be readily partitioned into relatively homogeneous groups within which the form

270
of the response is likely to be stable. Finally, since with an unbiased model differences
between the observed and fitted yields should be random, it is possible to test whether a
model is imposing its own form on a set of data by examining these residuals for
pattern.

5. Experimental design

Models, as described in the previous sections, are tools for the interpretation of the
reqults of experiments and are Far less important in themselves than the thought that
goes into the design of the experiments and the care that goes into their execution.
The experimenter investigating fertilizer response must make the following decisions:
1. How many different rates of fertilizer will be tested'?
2. What range of dressings will they span?
3. How will they be spaced?
4. What other factors should be included in the experiment?
5. How Much replication is necessary?
And, if there arc to be a series of trials:
6. low many sites are needed, and how should they be selected?
7. How many years should the experiment run?
Many of these questions are obviously related. The answers to most of them will
depend on the resources available to the experimenter and the constraints imposed
by the necessity of restricting individual experiments to a reasonable size.
To diScuss all these questions here would be impossible, and only a couple of obser-
vations will be made. Firstly, at least five, and preferably more, different rates of
fertilizer must be tested if the aim is to investigate the form of the response curve.
Secondly, there must be sufficient replication, possibly achieved by averaging over
other factors, to ensure that the form of the response is not obscured by experimental
error. What constitutes 'sufficient' will vary with the likely size of the errors associated
with the particular crop being grown and the expected size of the response, but. as a
guide, I would suggest a ninimum of four for cereals where one might expect a coeffi-
cient of variation of about 5,,.
It only remains to emphasise that no amount of curve fitting can substitute for having
a good experiment in the first place.

6. References
I. Beid D..A.. Yucn L. T. K. and Needh,am P.: Nitrogen requirement of cereals 1. Response
curves. J.agric. Sci.,Camb.87. pp. 149-162 (1976)
2. Cro,ther E.M. and Yates F.: Fertilizer policy in war time. The fertilizer requirements
of arable crops. Emp. J. exp. Agiic. 9, pp. 77-97 (1941)
3. Greenwood D.J., Wood J. T., Cleaver f.J. and Htt .1.: A theory for fertilizer response.
J. agric. Sci., Camb. 77. pp. 511-523 (1971)
4. Nelder J.A.: Inverse polynomials, a useful group of m nit i-factor response functions.
Biometrics 22, pp. 128-141 (1966)

271
Co-ordinator's Report on the 4th Session
H. Laudelout, Department of Soil Science, University of I-oLvain-la-Neuve/Belgium; member
of the Scientific Board of the International Potash Institute*

The Use of Models in Crop Physiology

Though the lime available for this session meant that the sLibject could not be covered
completely, it was our aim that the methods described by the speakers should, as far
as possible, be representative of the whole field.
The methods described by Wood in the first paper may be regarded as transitional
between the classical methods of biometry and modelling proper. These methods set
out to express data on crop response to given factors in as concise a manner as possible.
These methods have been used by agronomists for the past fifty years or so and, though
they may strike some as being, intellectually, not very exciting, they are of great
practical value; the computation involved is relatively simple and they are well
adapted to econometric research. Various possible methods are available and the
research worker selects those which are best suited to his practical needs.
The other extreme was represented by the paper given by Petming de Vries. The
approach here is essentially deterministic, the aim being to cover as many ecological
factors as possible and to explain the results obtained. Whenever such methods are
discussed, two questions are invariably raised: the first about the information which
the agronomist or physiologist should give to the modeller, and the second concerns
the extent to which the exercise can assist the experimenter in explaining his results
and in planning further work. Such methods have ambitious aims and demand a
large input of information without which they cannot be effective.
A rather different type of modelling was described by Hansen. This has more limited
aims and is easier to use since, in the words of the author. 'Experimentation and
modelling phases alternate and thus the model is under continuous dynamic revision'.
1here are also models, which were not described, with more limited application,
the aim of which is only to utse existing information in such as way as to improve its
predictive value. Their value is restricted to the extent to which the quantitative
information already available is limited. Examples can be found in the fields of soil
physical chemistry, bio-climatology etc.,
An interesting kind of model was described by Mede. This is an analogue model whose
predictive value can be assessed only by a sufficiently large number of confirmatory
trials.

* Prof. Dr. 1-1.Laudelout. Department of Soil Science, University of Louvin-la-Neuve, Place

Croix du Sud 2, B-1348 Louvin-la-Neuve/B3elgiun)

273
There will always be objections to the pretensions of models to simulate natural
processes and these arise either because their predictive value is limited or because the
information fed into the model may be imprecise, the one being the consequence of the
other. One objection which is rarely mentioned is that of the high cost of developing
and using models. Development requires many man hours even though advanced
programming languages are used and consequently the use of numerical methods
pose no problem. Even so, the execution time on the computer of some models may be
such that the cost of using them exceeds that of a field experiment. When finance for
agricultural research is limited such considerations must carry considerable weight,
Ihough the modellers, as a rule, remain silent on this point.

274