Professional Documents
Culture Documents
(C.L.E.AN.)
Developments in Hydrobiology 117
Series editor
H. J. Dumont
Coastal Lagoon Eutrophication
and
ANaerobic Processes (C.L.E.AN.)
Nitrogen and Sulfur Cycles and Population Dynamics
in Coastal Lagoons
Edited by
Contents
Growth of the seaweed Ulva rigida C. Agardh in relation to biomass densities, internal nutrient
pools and external nutrient supply in the Sacca di Goro lagoon (Northern Italy)
by P. Viaroli, M. Naldi, C. Bondavalli & S. Bencivelli ... . . . .. . .. ..... ..... .. .. ... . 93-103
Macrophyte communities and their impact of benthic fluxes of oxygen, sulphide and nutrients
in shallow eutrophic environments
by P. Viaroli, M. Bartoli, C. Bondavalli, RR Christian, G. Giordani & M. Naldi .. ... 105-119
Differential anaerobic decomposition of seagrass (Zostera noltii) and (Monostroma obscurum)
biomass from Arcachon Bay (France)
by S. Bourgues, 1. Auby, R de Wit & PJ. Labourg ................................ 121-131
Nitrification, denitrification, and nitrate ammonification in sediments of two coastal lagoons
in Southern France
by S. Rysgaard, N. Risgaard-Petersen & N.P. Sloth .......... , ... ..... ... . . ... .... 133-141
Benthic oxygen respiration, ammonium and phosphorus regeneration in surficial sediments
of the Sacca di Goro (Northern Italy) and two French coastal lagoons: a comparative study
by M. Bartoli, M. Cattadori, G. Giordani & P. Viaroli ............................. 143-159
Seasonal variation in rates of heterotrophic nitrogen fixation (acetylene reduction) in Zostera
noltii meadows and uncolonised sediments of the Bassin d' Arcachon, south-west France
by D.T. Welsh, S. Bourgues, R de Wit & RA. Herbert............................ 161-174
Relationships between porewater organic carbon content, sulphate reduction and nitrogen
fixation (acetylene reduction) in the rhizosphere of Zostera noltii
by D.T. Welsh, P. Wellsbury, S. Bourgues, R de Wit & RA. Herbert ............... 175-183
The biogeochemistry of two eutrophic marine lagoons and its effect on microphytobenthic
communities
by LJ. Stal, S.B. Behrens, M. Villbrandt, S. van Bergeijk & F. Kruyning ............ 185-198
Sulfur bacteria in sediments of two coastal ecosystems: the Bassin d' Arcachon and the Etang
du Prevost, France
by B.E.M. Schaub & H. van Gemerden .......................................... 199-210
Sulphide release from anoxic sediments in relation to iron availability and organic matter
recalcitrance and its effects on inorganic phosphorus recycling
by G. Giordani, M. Bartoli, M. Cattadori & P. Viaroli ............................. 211-222
Preface
The Coastal Lagoon Eutrophication and ANaerobic processes research programme (CLEAN) was one of the
projects funded under the European Community Environment Research Programme 1991-1994. Coastal lagoons
constitute an important transition zone between the terrestrial environment and the open ocean. They comprise
13% of the world's coastline from the tropics to the polar regions. Their productivity is derived from interactions
between the land, oceans and atmosphere. They are enriched by the oceanic and continental inputs and are amongst
the most productive ecosystems in the biosphere.
The effect of human activity on these coastal ecosystems, whether direct or indirect, has accelerated considerably
in recent years and the impact of such changes on these sensitive and economically important environments is now
a cause of major concern. Along the European coasts, many shallow coastal lagoons receive large amounts of
organic and mineral nutrients which has led to a rapid increase in eutrophication over the two last decades.
The objective of this project was to study different lagoon systems subject to increasing levels of eutrophication
in order to identify the key processes controlling these changes and how they impact on the species composition
and population dynamics of the indigenous biological communities.
The papers in this special volume provide a comprehensive set of data on the effects of eutrophication on the
population dynamics of the indigenous biota, biodiversity, trophic relationships, biogeochemical cycling, nutrient
inputs and fluxes in two constrasting lagoon systems, the highly eutrophic Etang du Prevost and the moderately
eutrophic Bassin d'Arcachon. By adopting a multidisciplinary approach it has been possible to identify the key
processes regulating eutrophication in these lagoons and their impact on the community structure of the primary
and secondary producers as well as the nutrient exchanges which occur at the sediment-water interface. These data
provide a valuable source of new information for scientists working on related problems in similar ecosystems as
well as those involved in the management of these important natural resources.
The editors wish to acknowledge the assistance given by the numerous referees who carefully reviewed the
manuscripts and to thank Professor Henri Dumont, editor-in-chief of Hydrobiologia for his invaluable assistance
in the production of this special volume. We also wish to acknowledge the support and encouragement given by Dr
Canice Nolan of the European Commission Directorate DG XII to publish the data obtained during the course of
this project as a single special volume of Hydrobiologia.
We are grateful to the Conservatoire du Littoral (Mrs Musson) and to the Conseil General of Gironde (Mr
Boutet) for field work facilities and permission to enter and work on the site of Certes in Arcachon Bay and the
Company of Maguelone, and their director Mr Sallas for help during sampling work in the Prevost lagoon.
Finally, it is a pleasure to thank the staff of the Centre d'Oceanographie et de Biologie Marine, Arcachon, for
the unstinting support that they have given to all the participants during the course of this project. Particular thanks
are due to Fran~oise Truong who provided the administrative support for the project.
PARTICIPATING LABORATORIES
FRENCH GROUP Centre d'Oceanographie et de Biologie
marine
Pierre Caumette
Coordinator 2 rue Jolyet
Jacques Castel 33120 Arcachon
Universite Bordeaux I
Key words: Coastal lagoons, eutrophication, macrophytes, microorganisms, nitrogen cycle, sulphur cycle
Abstract
The conditions of eutrophication are described in three lagoon systems differing by their structure, their catchment
area and their connection with the sea: the Bassin d' Arcachon on the Atlantic coast, SW France, the semi-artificial
fish ponds of the Bassin d' Arcachon, and the Etang du Prevost on the Mediterranean. The Bassin d' Arcachon is a
shallow semi-enclosed bay, strongly influenced by climatic factors and tidal currents. The Bassin receives significant
inflow of freshwater and the waters are only partially renewed. The greatest part of the primary production is due to
the seagrass Zostera noltii. Although the ecosystem remains on the whole in steady state, some evidence of potential
eutrophication are visible. For instance, the flux of nitrogen into the Bassin d' Arcachon has increased by more than
50% during the last 25 years. The most significant change among primary producers is the massive development
since 1988 of the green alga Monostroma obscurum. The fish 'reservoirs' of the Bassin d' Arcachon are man-made
enclosures designed for extensive aquaculture and where the water renewal is only possible during certain periods
of time. Thus, because of the shallowness and the confined nature of these fish ponds, acute eutrophication is
sometimes observed in summer. The Etang du Prevost is extremely eutrophic due to agricultural and urban run-off.
Red waters occur periodically during the warm summer months as a consequence of ecological events beginning in
the early spring with a bloom of green macroalgae (Ulva sp.). In summer, the algal biomass is degraded by aerobic
heterotrophic bacteria; the oxygen demand encompasses the oxygen production, leading to the predominence of
anaerobic processes and dystrophic crisis. From the comparison of the selected sites, three stages of eutrophication
are recognized according to the conceptual model of Nienhuis (1992) describing the relation between the relative
dominance of primary producers connected to the availability of nutrients. Such macroscopic observations should,
now, be explained by the study of microbiological processes including meiofauna, protozoa, bacteria and all the
components of the microbial loop.
forests, marshes
-l..
0
LAGOON
agriculture
internal
recycling
systems
beach processes
reinforcement structures water diversion
tide gates irrigation
fishing
Figure 1. The lagoon as concentrating mechanism and controllable interface between land and sea (redrawn from Lasserre, 1979a).
Microbial communities in coastal lagoons ton, filamentous algae and Ulva sp. benefit from the
increased levels of nutrients and will grow in volume
Many shallow coastal lagoons exhibit well defined and dominance at the expense of flowering plants and
salinity gradients from seawater salinity near the con- other algae. They commonly dominate in many coastal
nection with the sea to low salinity in the continental lagoons.
areas of the lagoon receiving river inputs, or to brines In the sediment, the microbial community is more
in solar evaporation lagoons. Such variable conditions complex depending on the gradients of oxygen pene-
of salinity have a very important impact on the dis- trating the sediment and subsequently on the interface
tribution and the selection of the components of the between oxic and anoxic phases. This interface is gen-
biological communities of coastal lagoons. erally found within the first 2 mm of the sediment.
In the water column bacterial communities are Therefore many different kinds of semi-aerobic and
dominated by aerobic heterotrophic bacteria, either anaerobic bacteria coexist, living by fermentation or
with strictly oxidative metabolism or facultatively fer- anaerobic respiration. In most coastal lagoons, it is
mentative metabolism. These bacteria are halotolerant known that the anaerobic bacterial community is dom-
organisms and are mainly of terrestrial origin (Car- inated by sulphate-reducing bacteria that can transform
mouze & Caumette, 1985). The distribution of marine up to 50% of the organic material (J~rgensen, 1983)
bacteria depends on the influence of tides (Erchen- but produce H2S, a toxic compound that accumulates as
brecker & Stevensen, 1975). Although many authors FeS2 (pyrite) in the anoxic environment. If light reach-
have reported on their distribution and physiology (see es the anoxic layers, phototrophic sulphide-oxidizing
references in Caumette, 1989a), little is known about bacteria can grow and sometimes populate these envi-
their activity and role in coastal lagoons. They miner- ronments by forming dense purple to brown colored
alize organic material either in free living conditions, masses (Caumette, 1989b; Van Gemerden, 1983). In
or associated with clays or particulate organic matter. addition to sulphur bacteria, anoxic layers of coastal
They contribute to the release of inorganic nutrients. It lagoons are densely populated by nitrogen bacteria and
is known that fast growing plants such as phytoplank- methanogenic bacteria (Caumette, 1989a). However,
Xll
7.
7. In coastal lagoons, the trophic structure is charac-
80
terized by a number of different primary food types
70
SO JANUARY JULY and a highly connected food web of generalist feeders.
60
.0
Much of the emergent and submergent plant materi-
30 SO
al enters the food web as detritus. Because of the net
20 40 I
nutrient uptake which occurs during the fermentation
10
30 of detritus, this material may have a higher nutritional
20 value than the original plant material. For both detri-
I'm
10 tus and epiphytes, it is not just plant material, but an
entire community which is consumed, including bac-
7. I'm
•0
teria, fungi, microalgae and protozoans .
30
The increased pool of autochthonous particulate
OCTOBER
'0 matter results in intense microbial activity in the
20
30 surface sediment. This might be further enhanced
10
20 by benthic-pelagic coupling in the topmost layer of
60 I'm 10 the sediment. Due to the good light conditions, the
high primary production resulting from eutrophication
7. 40 SO I'm
60 60
7- exceeds the capacity of the heterotrophic decomposers
SO SO
to completely mineralize the biomass. As a conse-
MAY NOVEMBER
.0 40
quence, the pool of particulate organic matter in the
sediment is increasing due to the imbalance between
30 30
production and mineralisation.
20 20
In coastal lagoons, the excess detritus is accumu-
10 10
lated from late autumn through winter. The eutrophic
S 10 20 30 40 SO 60 I'm conditions prevailing in spring and summer contribute
Figure 2. Seasonal variation of the particle size (first mm of the to a rapid decomposition of detritus. As a result decom-
sediment) in a semi-enclosed lagoon ecosystem: the fish ponds of poser production and the microbial loop are also stimu-
the Bassin d' Arcachon (Castel, unpublished). lated. This provides an abundant food source on which
luxurious development of the meiofauna can occur. In
this case, the ratio of detritus to biomass results in
although many papers describe these bacteria and their an imbalance if the energy released from the detri-
role in marine or coastal estuarine environments (Her- tus by the meiofauna is dissipated too rapidly. These
bert, 1975; McFarlane & Herbert, 1984; Marty et aI., high decomposition rates result in an increased oxygen
1990), little is known on their distribution and role in demand and may lead to a temporary disappearance of
coastal lagoons. dissolved oxygen in the water column (Las serre et aI.,
Lagoon environments display very high densities of 1976). This phenomenon causes mass mortality of the
meiofauna composed mainly of nematodes, harpacti- benthic macrofauna and fish.
coid copepod, oligochaetes, ostracods, turbellarians, These summer dystrophic crises lead to the for-
and polychaetes (Castel, 1992). These meiofauna com- mation of high bacterial biomasses in the lagoons as
ponents may consume much more energy in the form a consequence of high bacterial activity involved in
of organic matter (detritus + microftora + bacteria) mineralization (both aerobic and anaerobic bacteria).
than do the macrofauna (Las serre et aI., 1976). Fur- Such biomass can further be used as a food source by
thermore respiration rates when related to biomass are the meiofauna. In many coastal lagoons, particularly
very high, indicating the importance of the micro- and at the sediment surface, bacteria play an important role
meiofaunal communities in coastal lagoons (Table 2). as a food source for higher organisms via the detrit-
One gram of meiofauna respires a rate at least ten ic food chain (Coull, 1973). However the quantitative
times greater than an equivalent macrofaunal biomass. importance of bacteria was not appreciated until rela-
In shallow eutrophic areas, the ratio macrofauna, meio- tively recently (Fenchel & l¢rgensen, 1977) although
fauna, microfauna, in term of biomass is approximately some early studies (see Zobell, 1946) emphasized the
100, 10, 1; whereas from a metabolic point of view it importance of bacteria in marine food chains.
is 4, 2, 1.
xiii
Table 2. Total in situ oxygen consumption and relative uptake by micro- and meiofauna (in
percent) in man-made lagoons (fish ponds) and salt marshes in the Bassin d' Arcachon (from
Lasserre et al., 1976 and Lasserre, 1979b).
Recent studies have shown very high bacterial pro- redox state of the sediments and overlaying water col-
ductivity in coastal lagoons (references in Torreton, umn. It is well established that their oxidation is mainly
1991) but few describe accurately the interrelationships mediated by chemotrophic prokaryotes that use oxy-
between such biomass and the consumers in the water gen as terminal electron acceptor.
column and benthos (Gophen et aI., 1974; Caumette Such organisms live around the redoxcline of the
et aI., 1983; Rieper, 1982; Carman & Thistle, 1985; lagoon ecosystem, depending on reduced compounds
Decho & Castenholz, 1986; Souza-Santos et aI., 1996). from the anoxic layer and oxygen from the overly-
Fenchel & J0rgensen (1977) calculated that bacteria in ing oxic layer. Thus reduced nitrogen compounds are
the detritic food chain provided between 60-80% of the generally oxidized via the activity of ammonia- or
diet of the meiofauna. On average, meiofauna graze at nitrite-oxidizing bacteria. In contrast, reduced sulphur
a rate of 1% of the standing stock of both heterotrophic compounds can be oxidized under anoxic conditions.
bacteria and autrotrophic microalgae per hour (Mon- Chemotrophic bacteria that perform anaerobic respi-
tagna, 1995). Meiofauna grazing is therefore broadly ration such as Thiobacillus denitrificans generate a
in equilibrium with microbial production. However, proton electrochemical gradient as electron flow from
new more accurate data are now required to clarify our reduced sulphur compounds to nitrate, nitric or nitrous
understanding of the functioning of coastal lagoons, in oxides. Moreover, in the upper part of the anoxic layers
terms of carbon turnover, organic matter fluxes via the reached by light, phototrophic sulphur oxidizing bac-
biogeochemical cycles, and the mineralization and the teria are able to oxidize reduced sulphur compounds
production of organic material through the food chain. to sulphate as a consequence of their an oxygenic pho-
tosynthesis. Thus, in coastal lagoons, both cycles are
Biogeochemical nitrogen and sulphur cycles in functioning in similar conditions: reduction of nitro-
coastal lagoons gen or sulphur compounds occurs under anoxic con-
ditions whereas oxidation of the reduced compounds
The nitrogen and sulphur cycles include a diverse range mostly occur at the interface between oxic and anoxic
of oxidation and reduction reactions (Figures 3 & 4). zones, i.e. the redoxcline. Almost all eutrophic shallow
These can be divided into two categories: dissimi- water coastal lagoons have a redoxcline occurring at
latory reactions that are found principally amongst the sediment surface. Therefore most of these metabol-
prokaryotes and assimilatory reactions that occur in ic processes coexist within the first millimeters of the
both prokaryotes and eucaryotes (Ferguson, 1988). sediment. In eutrophic coastal lagoons, ammonia is the
In both cycles, the production of reduced com- major nitrogen compound produced in anoxic layers
pounds such as NH 3 , NO;, H2S, SO are either the (Koike & Hattori, 1978; Nishio et aI., 1982, 1983; Mar-
end products of dissimilatory anaerobic respiration ty et aI., 1990). However, production of reduced nitro-
or derive from aerobic and anaerobic degradation of gen compounds is lower compared to that of reduced
organic material (Herbert, 1982; Caumette, 1986; Her- sulphur compounds which are derived from sulphate
bert & Nedwell, 1990). These compounds accumulate reduction which is the most important anaerobic respi-
in the anoxic layers mainly in the sediment of eutroph- ratory process occurring in coastal lagoons (Caumette,
ic coastal lagoons. Their oxidation is dependent on the 1986) and coastal marine environments (J0rgensen,
XIV
IOxic conditions I
Aerobic
sulfur Assimilatory
oxidation sulfate reduction
Dissimilatory
sulfate reduction Anaerobic
degradation of
organic matter
I Anoxic conditions I
Figure 4. Sulphur cycle.
xv
" .
III
~;
interface between the oxic and anoxic zones. However, 5km
Origin Years
1970 1980 1990
12
i"
~ 8
10
•
E2
0
Outer bay
Inner bay
Eyre river
M
0
Z 6
Z \:':::::::::::::::) Zostera no/tii
N
+
0 4 _ Zostera marina
:;;:
.I
Z
~
2
0 .... ....co
.... ....enen
.... co '" '" co en C>
'" Figure 8. Location of the seagrass beds of Zostera noltii and Z.
C> ~ N ~ II) N~
co co co co co co co co co en en en en
en en en en en en en en en en en en en en en en marina in the Bassin d' Arcachon (redrawn from Auby, 1993).
~ ~ ~ ~ ~ ~ ~ ~ ~ ~
Table 4. Estimated annual production ofthe different primary producers in the Bassin d' Arcachon
(from Auby et al., 1993).
reported in the bassin in 1962. These are example of and outflow are regulated by sluice gates (Figure 11).
species having extended their area of distribution. Oth- The fish ponds or fish reservoirs have a characteristic
er species have appeared since 1968 as a consequence shape: channels or ditches, 1.5 to 2 m in depth, feed
of the importation of the Pacific oyster Crassostrea large expanses known as 'itats', each of which covers
gigas Thunberg, including the tunicate Styella clava an area ranging from 1000 to 10 000 m:::. In the canals
Herdman, some Bryozoans and Annelids. More recent- the fish are 'penned' during the winter months to coun-
ly, the large brown alga Sargassum muticum (Yen- teract the colder temperatures prevailing in the shallow
do) Fensholt accidentally introduced in Great Britain areas. These flats, 20-50 cm deep, are separated from
with the Pacific oyster, invaded the coast of Britan- each other by ridges formed from the mud removed
ny and appeared some years ago in Arcachon. Even during the digging of the ditches. All inflows and out-
small organisms « 1 mm) such as the Harpacticoid flows can be regulated by sluices located at intervals
Copepod Stenhelia latioperculata Ito (originally from along these embankments.
Japan) are likely to have been introduced with oys- At high tide, the sluices are manoeuvred in such a
ters. Conversely, some species have disappeared due way that a current is created and the fish, both imma-
to either modifications of hydrological conditions (e.g. ture and adults, are drawn into the reservoirs. The fish
the clam Venus verrucosa Linne) or to overexploitation are prevented from returning to the sea by a system of
(e.g. the bivalve Chlamys varia Linne). mesh frames. These operations are usually carried out
during the cooler seasons, from March to November.
Fish ponds of the Bassin d'Arcachon, a moderately Every two weeks, when high tides occur at the time of
eutrophic lagoon system the full or new moon, water from the sea is allowed to
flow into the reservoir through the sluices. This oper-
General description ation lets in the young fish and natural mineral salts,
The 'fish reservoirs' of the Bassin d'Arcachon (Fig- and replenishes the reservoir by exchanging the water
ure 10) are man-made enclosures created in the lagoon- held in the ponds and fresh seawater. This operation
al marshes (wetlands) and where a number of eury- is carried out at low tide. The young fish are free to
haline fish are farmed: grey mullet (Chelon labrosus move and to grow inside the complicated network of
Risso, Liza ramada Risso), sea bass (Dicentrarchus basins, in which the salinity may vary from very dilute
labrax Linne), eels (Anguilla anguilla Linne) and gilt brackish water to full seawater.
head bream (Sparus aurata Linne). Such fish reservoirs Because ofthe shallowness and the confined nature
are also known from the Charente and Vendee along of the fish ponds, the salinity regime is extremely vari-
the Atlantic coast. These structures, designed for tradi- able both in time and space. The salinity ranges from
tional extensive (without food supply) aquaculture, are almost freshwater to hypersaline (60%0). The salinity
comparable to the 'valli', situated along the Adriatic regime of the fish reservoirs is strongly linked to (I) the
coast, and also to the tropical 'tambaks' in Indonesia. relative location of the ponds between the sea and the
They constitute mixohaline and shallow (0.2-1.5 m continental freshwater inputs, and (2) the renewal of
depth) environments where the rich input of detritus the water through the sluices. The man-induced renew-
plays a prominent role in the food chain. Originally al of water is supposed to maintain salinity compatible
they were salt-pans exploited since the end of the Mid- with biological activities. However, the maintenance
dle Ages. Progressively, during the late eighteenth cen- of a given salinity requires a careful and periodic oper-
tury the exploitation of the salt decreased and the salt- ating of the sluices since, due to the shallowness of
pans were converted into fish ponds. These fish reser- the ponds, the salinity tends to vary widely (Figure 12)
voirs schemes flourished all through the nineteenth between periods of renewals. The same situation can
century up to the end of the Second World War but they also be observed for nutrients (Escaravage, 1990). In
have since gone into decline because of the labour costs such ponds, the greatest part of the regulation of nutri-
and the relatively low yield (50 kg ha- 1) of this type of ent concentrations is controlled by in situ biochemi-
extensive aquaculture. These semi-enclosed lagoonal cal processes, especially at the benthic level. These
systems cover a surface area of approximately 940 ha mechanisms are hardly affected by the renewal of the
in the Bassin d' Arcachon. The old salt-evaporation overlaying water.
areas and the salt-pan runoff ditches are separated by
embankments from the sea and through which inflow
xx
500 m
Figure 10. Map of the fish ponds of the Bassin d' Arcachon (Certes reservoirs).
Figure 11. Schematic diagram of the fish ponds of the Bassin d' Arcachon. Left: salt marsh comprising the 'slikke' (SI) = mudflat, the 'schorre'
(Sch) = high marsh and 'estey' (es) = small channel. The ponds are separated from the salt marsh by a dyke (d) along which a sluice (ec) is
established. Right: typical structure of the reservoirs, Pr = channel (width: 3-4 m, depth: 0.8-1.5 m), PI = shallow basins (width: 10-40 m,
length: 100-800 m, depth: 0.2--0.5 m). Redrawn from Lasserre (1979a) and Castel (1989).
From eutrophy to dystrophy Cladophora vadorum (Aresch.) Klitz at the water sur-
As in many inshore areas, the rich input of detritus, face, and mats of green algae (Lamprothamnium papu-
organic matter, bacteria and benthic microtlora play losum J. Groves, Chaetomorpha aerea (Dillwyn) Klitz,
a prominent role in the food chain of the fish ponds. or cyanobacteria at the sediment surface. The biomass
The most important primary producers are sea grasses of Ruppia cirrhosa has been estimated to 126 g W W
(Ruppia cirrhosa Petag), and filamentous green algae: m- 2 for the above-ground and to 51 gWWm- 2 for the
XXI
30
29 °c a
28
25
27
r:i'
o 20 25
!' 15
c:
22
~ 10 21
5 19
18
O~~~~--~~~~~r-~Lr~~r---~
April May June July August Sept. Oct. 20 1 10 20 1 10 20 1 10
Figure 12. Temporal variation of salinity in a fish pond of the Bassin JUNE I JULY I AUGUST J SEPT.
_ntpellier
l'Arnel
r
N
du Prevost
2 km
Mediterranean Sea
Figure 14. Location of the principal lagoons along the Languedocian Mediterranean coast.
water exchanges. The Etang du Prevost is situated in 'Le Lez', and seawater through the only connection
a lagoonal complex delimited to the South-West by with the sea 'Le Grau du Prevost'. Although the lagoon
the Etang de Thau and the Sete promontory and to the is situated in a nontidal zone, the water level can vary
North-East by the wetlands of the Camargue and Rhone from -0.30 m to +0.40 m relative to the mean level
delta (Figure 14). These ponds are of recent geological because of the strong winds. In the meantime, the
formation. The sandy and detrital zones in this region current speed can reach 0.40 m s-1 in the lagoon and
explain their continental formation. The creation by 0.80 m s-1 in the 'grau'. When the wind is blowing,
the currents of offshore bars (,cordons littoraux') has 20 to 25% of the water can be renewed in one day.
allowed the transformation of the coastline and the for- In contrast, during calm periods, the water remains
mation and evolution of these coastal lagoons since the stagnant. This is particularly true in summer. The wind
quaternary period. action may induce very strong and rapid variations of
The water in these lagoons is slowly renewed, the salinity (Figure 15). In summer, because of the
via small channels ('graus') communicating with the evaporation, the salinity can reach 40%0.
Mediterranean sea. These shallow bodies of water are The system is extremely eutrophic due to agricul-
subjected to continental inputs. They are directly influ- tural and urban run-off. For the Etang du Prevost, 85-
enced by urban or industrial effluents. 90% of the input of Nand P are of domestic origin
In contrast to the Bassin d' Arcachon, the Etang du (Table 5). The input per volume unit of lagoonal water
Prevost is a shallow lagoon with an average depth of is approximately 24.6 t 106 m- 3 for the total nitrogen
about 0.8 m and a surface area of380 ha (Figure 14). It and 4.3 t 106 m- 3 for the total phosphorus (Anony-
receives fresh water and sewage carried by a small river
xxiii
water
sediment
Lux (~02
,/ i /
water
B red water
white water
Figure 15. Salinity fields (%0) in the Prevost lagoon during two
different climatic conditions: wind blowing from the Northeast
(22.02.1973) and wind blowing from the Southwest (10.03.1973) Etang du Prevost. Its flora is essentially composed of
(redrawn from Guelorget & Perthuisot, 1992). green algae (Ulva and Enteromorpha spp.). The bio-
mass of these algae may reach 5 kg W W m- 2 and
Table 5. Annual input and stocks of nitrogen and phos- sometimes accumulates at densities up to 30 kg W
phorus, expressed in tonnes, in the Etang du Prevost
(anonymous, 1991).
W m- 2 . In summer, this biomass is rapidly degraded
by aerobic heterotrophic bacteria whose numbers and
Origin Nitrogen Phosphorus activities increase rapidly. During this period the water
turns anoxic and becomes rich in sulphide, which leads
Domestic input 60.2 11.1
to severe dystrophic crisis.
Agriculture 5.5 0.5
Industry 1.4 0.2
Sediment stock (10 cm) 475 140
The dystrophic crisis
Annual flux in sediment 22 2.4
The Etang du Prevost is regularly subject to dystrophic
Viva stock 20.6 1.8 crises, locally called the 'malai·gues'. Such dramat-
ic events appear every year during the warm summer
months for a period of about 15 days. They have a dra-
matic consequence for aquaculture, particularly oyster
mous, 1991). As a comparison, the input of nitrogen culture that takes place in the lagoon. Macroscopical-
in the Bassin d' Arcachon is only 3 t 106 m- 3 . ly, the dystrophic crises are characterized by coloured
Such large inputs of nutrients lead to the devel- waters (white, red or black waters). They originate
opment of high biomass of primary producers in the from a perturbation of the sulphur cycle.
xxiv
a
a .... a
U~
N a
0
O
0 ....
a
\D .... U 0
(")
0
........
.... 0
a
....l!)0N a
(")
N
l!)00
.... l!)0
0
....
....0 0
a N
(") 0 0"1 .... a
a a \0
....
(")1'- l!)
°8~ (")
0
a
co 0
....
N 0 0"1 .... 0
(")
l!)
~
0..,
0
l!) 0 ..... (") ..... a~
0
....a
N a 1'- .... a
~
011 N p:: ~
U)U) O~ U)
011 N p::
U)U) oro U)
U•
,...,... ,...,...
,...
I
I
CO
,
I
0 (0
T"" T""
0
a .....
a U co a
a .... ..... 0 1'- ..... 0
UN a NO 0 a
N ..........
..... aco
l!) N
0
a N a N ...... a
l!)0
..... (")
N
0 .... ......0
......... a a
(") 0 oO~
N
a 00
(") ....
a .... (") ....... O"l""'i .......
N
0 ~ a
......... 0 0 ~
a 0"1 a
.... ~ a .....
...... N (")0 ...... N
a
.........
~O 0 II
~
011 N p::
II U) U) O~ U)
~
011 N p::
U)U) O~ U)
,...
--
,...
, ....
....
Q) ,...,...
Ci:I
,...
Q)
Ci:I 3:
(0 I
I
It) 3: :=:
Q)
T"" "C .s::: 0
~ 3: N
~I:J
Figure 17. The Prevost lagoon during a dystrophic crisis (June-August 1977). I: before the dystrophic crisis, II: appearance of anoxic water
with presence of sulfide, III: appearance of red and white waters, IV: after the dystrophic crisis. J, 2, 3 and C refer to sampling stations. S04 '
S- and O2 are expressed in mg I-I. B =phototrophic bacteria, SR =sulfate reducers. Bacterial counts are expressed in numbers ml- I .
xxv
Table 6. Dissolved nitrate and ammonia in the water of the Bassin d' Arcachon
(central and inner parts), fish reservoirs of Certes and Etang du Prevost. Values
are given for the summer 1993 (Sloth et aI., 1993) except for the fish reservoirs
(summer 1985, Castel, unpublished).
Bassin d' Arcachon, central part (station A) 0.32 - 2.0 2.0 - 2.4
Bassin d' Arcachon, inner part (station B) 0.32 - 1.3 1.5 - 2.5
Fish reservoirs (station Cl) 0.7 - 1.01 4.6 - 8.3
Etang du Prevost (station 1 l) 1.2 - 4.9 5.7 - 10.5
dans l't~tang du Prevost a Palavas, Herault. Vie Milieu 25B: 175- Caumette, P. & B. Baleux, 1980. Etude des eaux rouges dues a
204. la proliferation des bacteries photosynthetiques sulfo-oxydantes
Anonymous, 1991. Efficacite de la reduction de la masse des nutri- dans I' etang du Prevost, lagune saumatre mediterraneenne. Mar.
ments dans la prevention des malillgues. Application aux etangs BioI. 56: 183-194.
Palavasiens. Summary Report, Region Languedoc-Roussillon, Caumette, P., M. Pagano & L. Saint-Jean, 1983. Repartition verticale
Agence de I'Eau Rhone-Mediterranee-Corse, 7 pp. des bacteries, du phytoplancton et du zooplancton dans une partie
Auby, I., 1991. Contribution a I'etude des herbiers de Zostera stratifi6e d'une lagune tropicale (lagune Ebrie, Cote d'Ivoire).
no/tii dans Ie Bassin d' Arcachon: dynamique, production et Hydrobiologia 106: 135-148.
degradation, macrofaune associee. Doct. Thesis, Univ. Bordeaux Coull, B.C., 1973. Estuarine meiofauna: A review: trophic rela-
I, 162 pp. tionships and microbial interactions. In L. H. Stevenson & R. R.
Auby, I., 1993. Evolution de la richesse biologique du Bassin d' Ar- Colwell (eds), Estuarine Microbial Ecology. University of South
cachon. Scientific Report 91 5 527 019, IFREMER, Soc. Sci. Carolina Press, Columbia: 499-512.
Arcachon, 222 pp. Decho, A. W. & R. W. Castenholz, 1986. Spatial patterns and feed-
Auby, I., G. Bachelet & P. J. Labourg, 1993. Biomass and ing of meiobenthic harpacticoid copepods in relation to resident
species composition of macrophytes in Arcachon Bay and Prevost microbial flora. Hydrobiologia 131: 87-96.
lagoon, with a compilation of data on primary production in Arca- Erckenbrecher, C. & L. H. Stevenson, 1975. The influence of tidal
chon Bay. In P. Caumette (ed.), CLEAN Progress report, part II, flux on microbial biomass in salt marsh creeks. Limnol. Oceanogr.
EU Environment Programme DG XII, Brussels: 65-72. 20: 618-625.
Auby, I., F. Manaud, D. Maurer & G. Trut, 1994. Etude de la pro- Escaravage, V, 1990. Daily cycles of dissolved oxygen and nutrient
liferation des algues vertes dans Ie Bassin d' Arcachon. Scientific content in a shallow fishpond: The impact of water renewal.
Report, IFREMER, Arcachon, 163 pp. + annex. Hydrobiologia 207: 131-136.
Bachelet, G., X. de Montaudouin, I. Auby & P. J. Labourg, 1994. Fenchel, T. & B. B. J0rgensen, 1977. Detritus food chain in aquatic
A comparative study of the seasonal changes in macrophytes and ecosystems: the role of bacteria. Adv. Microb. Ecol. 1: 1-55.
macrozoobenthos assemblages in three coastal lagoons under Ferguson, S. J., 1988. The redox reactions of the nitrogen and sulphur
varying degrees of eutrophication. In P. Caumette (Coord.) cycles. In J. A. Cole & S. J. Ferguson (eds), The nitrogen and
C.L.E.A.N. Progress Report 1994. EU Environment Programme sulphur cycles. Cambridge University Press: 1-29.
DG XII, Brussels: 353-367. Gophen, M., B. Z. Cavari & T. Berman, 1974. Zooplankton feeding
Baleux, B., P. Caumette & M. Troussellier, 1979. Numeration et on differentially labelled algae and bacteria. Nature (London)
approche qualitative des populations bacteriennes des lagunes 247: 393-394.
amenagees de Certes (Arcachon). II. Bacteries sulfo-oxydantes Guelorget, O. & J. P. Perthuisot, 1992. Paralic ecosystems. Biologi-
photosynthetiques. Publ. Sci. Tech. CNEXO, Brest, Actes collo- cal organization and functioning. Vie Milieu 42: 215-251.
ques 7: 529-542. Herbert, R. A., 1975. Heterotrophic nitrogen fixation in shallow
Bouchet, J. M., 1968. Etude oceanographique des chenaux du Bassin estuarine sediments. J. expo mar. BioI. Ecol. 18: 215-225.
d' Arcachon. Doct. Thesis, Univ. Bordeaux, 200 pp. Herbert, R. A., 1982. Nitrate dissimilation in marine and estuarine
Carman, K. R. & D. Thistle, 1985. Microbial food partitioning by sediments. In D. B. Nedwell & c. M. Brown (eds), Sediment
three species of benthic copepods. Mar. BioI. 88: 143-148. microbiology. Academic Press, New York: 53-71.
Carmouze, J. P. & P. Caumette, 1985. Les effets de la pollution Herbert, R. A. & D. B. Nedwell, 1990. Role of environmental fac-
organique sur les biomasses et activites du phytoplancton et des tors in regulating nitrate respiration in intertidal sediments. In
bacteries heterotrophes dans la lagune Ebrie (Cote d' Ivoire). Rev. N. P. Revsbech & J. S0rensen (eds), Denitrification in Soil and
Hydrobiol. trop. 18: 183-212. Sediment. Plenum Press, New York: 77-90.
Castel, J., 1978. Plancton estival dans les etangs saumatres du Bassin J0rgensen, B. B., 1982. Ecology of the bacteria of the sulphur cycle
d'Arcachon. Bull. Off. natn. pech. Tunisie 2: 303-319. with special reference to anoxic-oxic interface environments. Phi-
Castel, J., 1989. Structure et evolution d'un ecosysteme lagunaire los. Trans. r. Soc. London, Ser. B 298: 543-561.
amenage: les 'reservoirs a poissons' du Bassin d' Arcachon. In J0rgensen, B. B., 1983. The microbial sulphur cycle. In W. Krumbein
Connaissance et gestion de la frange littorale et du proche plateau (ed.), Microbial Geochemistry. Blackwell Scientific Publications,
continental. Conseil de l'Europe, Strasbourg: 417-428. Oxford: 91-124.
Castel, J., 1992. The meiofauna of coastal lagoons and their impor- Koike, I. & R. Hattori, 1978. Denitrification and ammonia formation
tance in the food web. Vie Milieu 42: 125-135. in anaerobic coastal sediments. Appl. envir. Microbiol. 35: 278-
Castel, J., P. J. Labourg, V Escaravage, I. Auby & M. E. Garcia, 282.
1989. Influence of seagrass beds and oyster parks on the abun- Labourg, P. J., 1975. Contribution a I'hydrologie des etangs
dance and biomass patterns of meio- and macrobenthos in tidal saumatres de la region d' Arcachon: description des phenomenes
flats. Estuar. coast. Shelf Sci. 28: 71-85. d'eaux blanches. Bull. Soc. Linn. Bordeaux 5: 3-8.
Caumette, P., 1986. Phototrophic sulfur bacteria and sulfate-reducing Labourg, P. J., 1979. Structure et evolution de la macrofaune
bacteria causing red waters in a shallow brackish lagoon. FEMS invertebree d'un ecosysteme lagunaire amenage (reservoirs de
Microbiol. Ecol. 38: 113-124. Certes). Publ. Sci. Tech. CNEXO, Brest, Actes colloques 7: 575-
Caumette, P., 1989a. Les lagunes et les marais maritimes. In M. 590.
Bianchi, D. Marty, J. C. Bertand, P. Caumette & M. Gauthier Labourg, P. J., 1985. Ecologie et utilisation des zones humides du
(eds), Micro-organismes dans les ecosystemes oceaniques. Mas- Bassin d'Arcachon. In Actes du colloque sur les zones humides
son editions, Paris: 249-282. du littoral Aquitain, D.R.A.E. Aquitaine, Bordeaux: 5-28.
Caumette, P., 1989b. Ecology and general physiology of phototroph- Lasserre, P., 1979a. Coastal lagoons. Sanctuary ecosystems, cradles
ic bacteria in benthic environments. In Y. Cohen & E. Rosenberg of culture, targets for economic growth, Nature & Resources,
(eds), Microbial mats, Ecological physiology of benthic micro- Unesco 15: 2-21.
bial communities. ASM Publications, Washington DC, USA: Lasserre, P., 1979b. Programme coordonne Ecotron sur Ie site d' Ar-
283-304. cachon (Aquitaine, France): controle de la production biologique
XXVlll
marine dans un ecosysteme lagunaire amenage (reservoirs 11 pois- pods for various species of bacteria. Mar. Eco!. Prog. Ser. 7:
sons). Bilan, synthese et perspective. Pub!. Sci. Tech. CNEXO, 303-307.
Brest, Actes colloques 7: 463-502. Robert, R., N. Guillocheau & Y. Collos, 1987. Hydrological para-
Lasserre, P. & H. Postma (eds), 1982. Proceedings of the interna- meters during an annual cycle in the Arcachon Basin. Mar. Bio!.
tional symposium on coastal lagoons, Bordeaux, France, 8-14 95: 631-640.
September 1981. Oceano!. Acta, No. sp., 461 pp. Sloth, N. P., N. Risgaard-Petersen, S. Rysgaard & S. P. Pelegri, 1993.
Lasserre, P., J. Renaud-Mornant & 1. Castel, 1976. Metabolic activ- Nitrification, denitrification and nitrate ammonification in sedi-
ities of meiofaunal communities in a semi-enclosed lagoon. Pos- ments of two coastal lagoons in Southern France. In P. Caumette
sibilities of trophic competition between meiofauna and mugilid (Coord.) C.L.E.A.N. Progress Report 1994. EU Environment Pro-
fish. In G. Persoone & E. Jaspers (eds), Proc. 10th Europ. Symp. gramme DG XII, Brussels: 159-185.
Mar. Bio!., vo!' 2. Population dynamics, Universa Press, Wet- Soriano-Sierra, E., 1988. Contribution 11 l'ecologie des etangs
teren: 393-414. saumfitres du Bassin d' Arcachon: structure, dynamique et
Marty, D., G. Esnault, P. Caumette, E. Ranaivoson-Rambeloarisoa productivite d'une phytocenose dominee par Ruppia cirrhosa
& J. C. Bertrand, 1990. Denitrification, sulfato-reduction (Petag.). D.E.A., Univ. Bordeaux I, 18 pp. + annex.
et methanogenese dans les sediments superficiels d'un etang Souza Santos, L. P., J. Castel & P. 1. P. Santos, 1996. The role
saumfitre mediterraneen. Oceano!. Acta 13: 199-210. of phototrophic bacteria as food for meiobenthic harpacticoid
McFarlane, G. T. & R. A. Herbert, 1984. Dissimilatory nitrate reduc- copepods inhabiting eutrophic coastal lagoons. Hydrobiologia
tion and nitrification in estuarine sediments. J. Gen. Microbio!. (this volume).
130: 2301-2308. Torreton, J. P., 1991. Importance des bacteries heterotrophes aerobies
Montagna, P. A., 1995. Rates of metazoan meiofaunal microbivory: dans une lagune eutrophe tropicale (Lagune Ebrie, Cote d'Ivoire).
a review. Vie Milieu 45: 1-9. Biomasse, production, exportations. These Doctoral. Faculte des
Nienhuis, P. H., 1992. Ecology of coastal lagoons in the Netherlands Sciences de Luminy, 246 pp.
(Veerse Meer and Grevelingen). Vie Milieu 42: 59-72. Van Gemerden, H., 1983. Physiological ecology of purple and green
Nishio, T., I. Koike & A. Hattori, 1982. Denitrification, nitrate reduc- bacteria. Ann. Inst. Pasteur Microbio!. 134: 73-92.
tion and oxygen consumption in coastal and estuarine sediments. Visscher, P. T., P. Quist & H. van Gemerden, 1991. Methylated
App!. envir. Microbio!. 43: 648-653. sulfur compounds in microbial mats: in situ concentrations and
Nishio, T., I. Koike & A. Hattori, 1983. Estimates of denitrification metabolism by a colorless sulfur bacterium. App!. Envir. Micro-
and nitrification in coastal and estuarine sediments. App!. envir. bio!. 57: 1758-1763.
Microbio!. 45: 444-450. Zobell, C. E., 1946. Marine Microbiology. Chronica Botanica Co.
Rieper, M., 1982. Feeding preferences of marine harpacticoid cope- Waltham, Mass., 240 pp.
PART I
Eutrophication effects on population dynamics,
biodiversity and trophic relationships
in coastal lagoons
Hydrobioiogia 329: 3-17, 1996. 3
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (CLE.AN.).
©1996 Kluwer Academic Publishers.
Abstract
Activity and numbers of heterotrophic bacteria have indicated that, as expected, Prevost Lagoon is more eutrophic
than Arcachon Bay. Amplification and sequence analysis of the 16S rRNA genes from DNA samples extracted
directly from the environment allow the determination of phylogenetic relationships among members of microbial
communities in natural ecosystems without the need for cultivation. Analysis of partial 16S rRNA gene sequences
obtained from Stations A and 11 revealed that, in both environments, a relatively large number of clones related
to CytophagalFlexibacterlBacteroides as well as to a-Proteobacteria were found. One hundred percent similarity
with the sequences of the data bases were not found for any of the more than a hundred clones studied. In fact for
most clones maximum similarity was below 95% for the nucleotide series sequenced. Similarity was not higher
with any of the sequences found for the 14 isolates (pure cultures) obtained from the same samples. Redundancy,
i.e. number of identical sequences, was higher in the samples from Arcachon. In addition, sequences related to
representatives of ten major phylogenetic branches of Eubacteria were obtained from Prevost Lagoon, however
only five branches were represented by the data from Arcachon. These findings indicate a higher bacterial diversity
in Prevost Lagoon.
One such problem is the need to use a conversion fac- depth sample was related to Cyanobacteria, Syne-
tor to transform thymidine incorporation into bacterial chococcus and group SAR7 (Giovannoni et aI., 1990).
production (Fuhrman & Azam, 1980). These authors also described a novel group, the so-
The bacterial diversity present is obviously a key called Marine GroupA, unrelated to any subdivisions
factor in actually carrying out the fast degradation of described by Woese (1987). They also found clones
the organic matter accumulated by primary production. related to Gram-positives, a- and ,-Proteobacteria
The diversity of marine prokaryotes is not well known, and CytophagaiBacteroideslFlexibacter.
most of them are not yet cultured (Jannasch & Jones, DeLong (1992) described, in the Pacific Ocean, two
1959), showing the inability of standard microbiolog- novel groups related to the domain Archaea, group I
ical techniques to cultivate most microorganisms. As within the group Crenarchaeota and group II within
a powerful alternative, a molecular approach based the group Euryarchaeota and related to methanogenic
on the analysis of rRNA sequences has been used to microorganisms. In cold Antarctic waters, DeLong
determine the species composition of microbial com- et aI. (1993) found several clones related to group I,
munities (Pace et aI., 1986). The sequences of rRNAs and a group of sequences appeared close to group II.
directly extracted from naturally occurring organisms In the study by Fuhrman et aI. (1993) in the Pacific
are determined using molecular techniques and com- Ocean, the fourth group of clones appeared to be with-
pared with known rRNA sequences, using molecu- in the archaeal group described by DeLong (1992).
lar phylogeny techniques. The differences in sequence Studies have been made to estimate the diversity of
are the basis of the phylogenetic analysis to identify free-living organisms and aggregate-attached organ-
organisms in the environment by hybridization with isms (DeLong et aI., 1993). Clones of free-living
organism-specific probes and may help to determine microorganisms appeared to be relatd to the SARlI
the community structure (Amann et aI., 1995). group (Giovannoni et aI., 1990), ,-Proteobacteria
Previous 16S rRNA studies in the oligotrophic and the CytophagaiFlexibacterlBacteroides group.
waters ot'the Pacific and Atlantic Oceans (Fuhrman Those of aggregate-attached microorganisms were
et aI., 1993; Giovannoni et aI., 1990; Schmidt phylogenetically close to Cyanobacteria, Cytopha-
et aI., 1991) have disclosed phylogenetic lines not gaiFlexibacter; Planctomyces and ,-Proteobacteria.
described by classical microbiological methods. No The most abundant phylogenetic types detected in
sequence directly recovered from an environmental macroaggregate-associated bacterial populations fell
sample had 100% similarity with a previously cul- within the CytophagaiFlexibacter group (DeLong et
tured and described microorganism. In the Sargasso aI., 1993).
Sea, Giovannoni et aI. (1990) determined that their In the present work two coastal lagoons that repre-
clones formed two different groups represented by sent two model systems have been studied. The Arca-
clone SAR7 and SARlI. SAR7 grouped with pho- chon lagoon on the French Atlantic coast is a typical
totrophic oxygenic organisms (Cyanobacteria, chloro- tidal system in which the water in the lagoon is flushed
plasts and prochlorophytes) and SARlI grouped with by strong currents keeping a relatively low nutrient
the a-Proteobacteria group. In the North Pacif- concentration and a strong connection with the neigh-
ic Ocean, Schmidt et al. (1991) recovered clones bouring Atlantic waters. The Prevost Lagoon, on the
related to Cyanobacteria and to a new phylogenet- other hand, is located on the Mediterranean coast. With
ic group described by Giovannoni et aI. (1990) in a relatively narrow communication with the Mediter-
the Sargasso Sea. Other clones were close to Pro- ranean Sea that is virtually free oftides, the connection
teobacteria, but two clones (AL030 and AL033) between this lagoon and the marine habitat is compar-
did not belong to any phylogenetic lines previous- atively restricted. The Prevost lagoon is much more
ly described (Schmidt et aI., 1991). No archaeal eutrophic than that of Arcachon and affected to a far
clones were recovered in this sample. In the Pacif- greater degree by adjacent urban developments, with a
ic Ocean again, Fuhrman et aI. (1993) took samples history of dystrophic crises (Caumette, 1986).
at 100 and 500 m to observe the difference due to We have studied the activity and numbers of het-
the varying depth of water. In both sample depths erotrophic bacteria by classical techniques (thymidine
they found clones with 100% similarity, not with cul- incorporation and glucose utilization/incorporation
tured microorganisms, but with clones belonging to the and plating on marine nutrient agar). Sediment sam-
phylogenetic line SARlI described by Giovannoni et ples were also obtained for viable counts and isolation.
aI. (1990). Another group of clones from the 100 m Besides, water samples from two sampling sites (A
5
and 11) were processed to obtain DNA from the bac- tive medium mFC (Difco) (APHA-AWWA-WPCF,
terioplankton, and then libraries of 16S rDNAs direct- 1975).
ly amplified from both environments were construct-
ed, to study phylogenetic diversity. The 16S rDNA Classification of isolates
from some isolates obtained from the two environ-
ments were also partially sequenced and analyzed. Some colonies were classified to the genus level per-
forming biochemical tests by the standard methods
(Bradshaw, 1992): oxidase, catalase, 0fI<~ sensitivity
Materials and methods to vibriostatic agent 01129 (all the media were supple-
mented with 2% w/v artificial sea water salts). Mor-
Sampling sites and periodicity phology, Gram reaction and motility were determined
using a microscope.
Sampling was always performed at high tide to obtain
water samples and at low tide to get sediment cores. Biomass collection
Station A was reached by boat and water samples
were collected in sterilized II glass bottles. Stations A Different volumes of water samples, ranging from 1-31
and B (in Arcachon Bay), Stations C, and C 2 (Certes of Station 11 to 10-15 I of Station A were obtained in
Domain) and Stations X and 11 (Prevost Lagoon) were plastic containers previously treated with 0.1 N HCl to
selected. In station 11 samples were taken about 30 m eliminate contaminant DNA. They were processed to
from the shore with a water depth of about 1 m. Sam- obtain bacterial biomass and DNA. Debris and eukary-
ples were always obtained from superficial water by otic cells were removed by using glass microfibre filters
opening the flask just below the surface. Four sampling (Whatman GFID). Microorganisms were collected by
campaigns were carried out in May and September positive pressure on a 9 cm f2f 0.22 /Lm pore-size filter
1993 and May and September 1994. Sediment sam- (Durapore; Millipore Corp., Bedford, MA). Water was
ples were obtained in 5 ml syringes and the two upper pumped till filter clogging decreased flow rate to a few
ml separated from the lower three and processed inde- ml per min. Filters were stored at - 80°C until DNA
pendently. Sampling sites are described in Castel et al. extraction.
(1995).
Determination of microbial heterotrophic activities
Heterotrophic bacteria enumeration and isolation
Tritiated thymidine incorporation: Water samples were
Total heterotrophic aerobes were counted by plating on incubated at in situ temperature in the dark, on a rotary
nutrient agar supplemented with 2% w/v artificial sea shaker for 40 min with a final concentration of 10 nM of
water salts and incubated at 20 and 30°C for 24 or 48 tritiated thymidine (Amersham Life Science), labelled
hours. For sediment samples dilutions in sterile artifi- in the methyl position with a specific activity of70-85
cial sea water were performed. Colonies were isolated Ci/mmol. A blank was obtained by adding formalde-
at random and subcultured twice. 1 ml of bacterial cul- hyde solution (final concentration 2%). Five mI from
tures were harvested in sterile 2% marine salt water by each sample were taken in propylene tubes pre-chilled
centrifuging at 10000 rpm for 2 min and pellets were on ice for I min and an equal volume of cold TCA 10%
stored at -20°C. was added. After 5 min the mixture was filtered and
Some samples were plated onto selective media for rinsed three times with 5 ml of ice-cold 5% TCA. The
enumeration of specific physiological groups. Sulphite filtration was carried out in an ice bath. The filters were
reducing clostridia were counted from every sample by dried at 40 to 50°C for 15 min, chilled, placed in scin-
the most probable number (MPN) method using TSC tillation vials and radio-assayed by liquid scintillation
selective medium (Merck) with cycloserine to inhibit counting (Fuhrman & Azam, 1980).
the vegetative forms (APHA-AWWA-WPCF, 1975). Glucose metabolism: Water samples were incubat-
ed at in situ temperature in the dark for 1 hour with
Faecal coliforms 3 /LCi of D-U'4C Glucose (250 mCi/mmol), obtained
from Amersham Life Science, in a bottle plugged with
These pollution indicators were determined by the Whatman No. 1 paper. Another sample (blank) was
standard membrane filtration method, using the selec- fixed by adding formaldehyde solution for 1 hour. After
6
incubation 0.10 ml of H 2 S04 was added to the sample. (Promega), 0.2 JiM of each oligonucleotide primer
The Whatman No. I paper was soaked with 0.2 ml of and 100 ng of template DNA in a total volume of
,B-phenethylamine. In the dark, under the same condi- 100 ILl. The sequence of the forward primers was
tions, the sample was incubated again for 1 hour. The 5'-AGAGTTTGATCATGGCTCAG-3' (Lane et aI.,
sample was filtered through 0.2 ILm pore size (Celotate, 1985) at nucleotide positions 8-27 of the E. coli num-
Millipore Corp., Bedford, MA.) filters then rinsed with bering (Brosius et aI., 1981). Reverse primer was 5'-
25 ml of sterile distilled water. The filters were dried GGTTACCTTGTTACGACTT-3' (Lane et aI., 1985)
at 40 to 50°C for 15 min and counted. The Whatman at positions 1509-1491 (E. coli numbering). Reac-
No. 1 paper was placed in scintillation vials with 4 ml tion mixtures were overlaid with mineral oil (Light
of scintillation liquid and counted (Wright & Hobbie, White Oil; Sigma) prior to a thermal cycling regime
1966). (35 cycles on a Perkin Elmer Cetus 480) of 94°C for
Thymidine incorporation estimates were measured I min, 50°C for 1 min, and 72 °C for 2 min. After the
in pmoles of thymidine per unit volume and time to final cycle was completed, the amplicons were allowed
allow comparisons between studies. In order to trans- to extend at 72 °C for 10 min. Negative controls were
form thymidine incorporation into bacterial produc- included with no addition of template DNA. Five ILl of
tion an empirical conversion factor, 2 x 10 18 cells the PCR products were analyszed on 1% (w/v) agarose
mol- I (Fuhrman & Azam, 1980), was used. The ratio gels.
between dividing cells and bacterial plate counts gave
the specific growth rate (IL). We assumed a value of Purification and cloning of the PCR products
20 fg C cell-I (Moriarty, 1990), to represent the pro-
ductivity estimate in carbon units. We also estimated PCR products were purified using GeneClean II kit and
the doubling time from the growth rate (ln2/IL) (Mori- cloned into a T-tailed-vector prepared as follows: Blue-
arty, 1990). script plasmid (Stratagene, La Jolla, CA) was digested
with EcoRV and incubated with Taq I DNA polymerase
DNA extraction and purification (l U/ILg plasmid120 ILl volume) using standard buffer
conditions in the presence of 2 mM dTTP for two
Thawed filters were cut with a sterile razor blade into hours at 72 °c (Marchuk et aI., 1990) and ligated at
small pieces and vortexed in 2 ml of STE buffer 16°C overnight. XLBlue Escherichia coli strain was
(10 mM Tris-HCl [pH 8], 1 mM EDTA, 100 mM transformed with a standard protocol (Maniatis et aI.,
NaCl) in 50 ml conical based polypropylene centrifuge 1982). Plasmidic DNA was extracted using the Wizard
tubes. A 0.1 volume of 10% w/v sodium dodecyl sul- Minipreps DNA Purification System (Promega).
phate (SDS) was added and the tubes placed in a boil-
ing water bath for 2-5 min (Fuhrman et aI., 1988). Primers and sequencing reactions
After spinning at 5000 rpm for 10 min, the super-
natant fluids were transferred into 1.5 ml Eppendorf Primers with 5'-TTACCGCGGCTGCTGGCACG-3'
tubes and the DNA was purified with the GeneClean II sequence (Martinez-Murcia, 1993) at positions 533-
kit (BIO 101, Inc.). DNA from bacterial cultures was 514 (E. coli numbering) were chosen to sequence ca.
extracted as previously described (Martinez-Murcia 200 bp of a relatively variable region of each select-
& Rodriguez-Valera, 1994). Quantification and puri- ed clone. Primers were synthesized by the phospho-
ty (referred to protein content) of the DNA obtained ramidite method (Beaucage & Caruthers, 1981) on a
were estimated and 5 ILl aliquots of DNA dilutions Model 392 RNA/DNA Applied Biosystems synthe-
were electrophoresed (Martinez-Murcia & Rodriguez- sizer. Clones randomly selected were sequenced by
Valera, 1994). the dideoxy-chain termination method (Sanger et aI.,
1977) with a Sequenase Version 2.0 Kit (United States
PCR Amplification of J 6S rRNA genes Biochemicals) following the manufacturer's protocol.
Cl May 93 70 1.7 x 102 6.0 X 104 1.5 X 105 4.0 X 104 6.0 X 104
Sept 93 1.8 x 103 2.0 X 103 1.7 X 105 2.5 X 105 1.2 X 105 1.7 X 105
May 94 9.0 x 102 1.4 X 103 1.9 X 104 2.5 X 104 7.4 X 104 8.8 X 105
Sept 94 1.6 x 103 1.6 X 103 1.7 X 104 1.2 X 104 2.0 X 104 2.0 X 104
C2 May 93 3.7 x 102 1.0 X 103 2.0 X 104 3.0 X 104 5.0 X 103 2.0 X 104
Sept 93 1.8 x 103 4.2 X 103 9.0 X 105 1.0 X 106 1.4 X 105 2.5 X 105
May 94 1.8 x 103 1.8 X 103 6.0 X 104 7.6 X 104 9.0 X 104 6.7 X 104
Sept 94 1.3 x 103 1.9 X 103 9.0 X 104 7.0 X 104 6.0 X 103 4.0 X 103
sin Molecular Biology Package Version 8. Phylogenet- About 100 isolates were obtained from different
ic analysis were carried out with DNA STAR Package samples of Stations A and 11 during the September
(Lasergene) in a Quadra 840 Macintosh Computer. 1993 campaign. They were characterized to provide a
primary classification at the genus level. As shown in
Table 3, in both cases the predominant heterotrophic
Results microflora in the water column were related to the gen-
era Vibrionaceae as expected, and Pseudomonaceae.
Heterotrophic bacterial numbers and activities in the In the case of the Prevost Lagoon there also appeared
two lagoons to be a significant number of Enterobacteriaceae that
could be allochthonous microorganisms due to previ-
The study of the role ofbacterioplankton starts with the ously described waste water pollution detected in this
estimation of microbial biomass and activities in water environment. In the sediment there was also a signif-
samples. The results of the viability counts (average of icant proportion of Gram Positive bacteria, both rods
the three sampling campaigns) at the sampling sites, and cocci.
faecal coliforms and sulphite reducing clostridia are Thymidine incorporation estimates, assimilation
shown in Table 1 and Table 2. and respiration of glucose values, growth rates and
8
Table 2. Faecal coliforms and clostridia numbers (cels/lOO ml) in samples
from stations in Arcachon Bay (A and B), Certes Domain (CI and C2) and
Prevost lagoon (X and 11). ND =Non-determined data.
A May 93 ND 0 60 80
Sept 93 2 0 170 20
May 94 0 2 1560 3316
Sept 94 2 4 43 250
B May 93 ND ND 0 0
Sept 93 ND ND 10 20
May 94 ND 4 4137 3320
Sept 94 ND 9 9 4
X May 93 ND 0 130 ND
Sept 93 4 10 110 110
May 94 0 30 1080 630
Sept 94 33 4 ND ND
II May 93 ND 0 350 ND
Sept 93 20 50 130 130
May 94 0 60 1115 3036
Sept 94 14 9 610 620
doubling times of the heterotrophic bacteria are shown referred to the numbers of bacterial plate counts as the
in Table 4. In Station A values of bacterial production heterotrophic activities are related with: (i) bacterial
ranged from 0.014 to 0.52 J-Lg C L -Ih- I , in Stations numbers and (ii) distinct activity of different bacterial
C 1 and C2 (Certes Domain) from 1.70 to 4.65, and 0.27 populations. We observed the highest values of assimi-
to 2.5 J-Lg C L -I h -I respectively. Estimates in Prevost lation of glucose in early summer 1993-1994 in all the
were in the range of 0.11 to 3.55 J-Lg C L -Ih- I (Sta- Stations. The highest heterotrophic activity estimated
tion X) and 0.80 to 8.0 J-Lg C L -I h -I (Station 11). The by glucose assimilation was in both the Stations in
highest values at most sampling stations were found at Certes Domain (C l and C2). Maximal incorporation of
the end of summer in 1993. both thymidine and glucose was measured in Station
In order to compare estimates of bacterial produc- 11 (Prevost Lagoon) rather than Station A (Arcachon
tion in terms of thymidine incorporation and assimila- Bay), showing that the microbial community of Station
tion of organic compounds (glucose) those values were 11 presented more heterotrophic activity per cell. In
9
Table 3. Taxonomic status of the isolates from Stations A
The sequences were compared with all available 16S
(Arcachon Bay) and II (Prevost lagoon) during the Septem-
ber 1993 campaign. The total number of isolates from water, rDNA sequences of the GenBank and EMBL Data
superficial sediment and deep sediment from both stations and Library. Pairwise sequence comparisons with selected
the percentage of each taxonomic group are shown. numbers of several phylogenetic branches were per-
formed, and two dendrograms of sequence similari-
Station A
Water Superficial Deep
ty were constructed from two matrices of calculated
sediment sediment
percentages, each containing clones from Stations A
and 11 respectively. Some of these results have been
Total number 24 10 23 reported elsewhere (Benlloch et aI., 1995).
Vibrio, Plesiomonas, No identical sequences were found on comparing
Aeromonas 67% 60% 57% the clones with those of the data banks. The high-
Pseudomonas 25% 30% 26% est similarity values found were those corresponding
Enterobacteriaceae
to clone 38a (Station A) with a representative of an
Acinetobacter 8%
undescribed marine eubacterial group (Fuhrman et aI.,
Bacillus 13%
1993) (98.9%), and to clone 5011 (Station 11) with Tro-
Lactobacillus 4%
pherima whippelli, a High G+C content Gram-positive
Planococcus 10%
(99%). Overall similarity values with sequences of
Station II the data banks were higher for clones from Station A
Water Superficial Deep than from Station 11. In Station A clones appear to be
sediment sediment related to different phylogenetic branches. A total of
Total number 15 17 24 eighteen clones (Figure 1) were relatively close to the
Vibrio, Plesiomonas, CytophagaiFlexibacterlBacteroides group with simi-
Aeromonas 40% 41% 54% larity values ranging from 88.4 to 94.6%. Five clones
Pseudomonas 40% 29% 25% were related to the a-Proteobacteria branch (93.6 to
Enterobacteriaceae 20% 6% 12.5% 96.8%). Four clones appeared to be phylogenetically
Acinetobacter close, albeit at a very low similarity (ranging from 72.6
Bacillus 8.5% to 85.6%), to a group of 16S rDNA fragments gener-
Lactobacillus 18% ated by PCR from a soil sample (Liesack & Stacke-
Streptococcus 6% brandt, 1992). Two clones clustered with a marine
group of bacteria (Fuhrman et aI., 1993) and anoth-
er two with a representative of SAR 11 group from a
Sargasso Sea sample (Giovannoni et aI., 1990), both
groups described only by 16S rRNA sequencing.
In Station 11, twenty-two clones were related to the
summary, all data point to a major difference between CytophagaiFlexibacterlBacteroides group with simi-
the heterotrophic bacterial biomass of each of the two lar values ranging from 79.1 to 94.7% (Figure 2). Eight
lagoons. As had been expected from the higher level of clones clustered with the a-Proteobacteria (sequence
eutrophication (Caumette, 1986), it was much greater similarities between 74.7 and 95.2%) and seven clones
in the Prevost lagoon. were related to High G+C Gram-positive Bacteria
(85.6-99.6%). Two clones appeared to be phylo-
Analysis of bacterial biodiversity by direct genetically close to ,,(-Proteobacteria with similar-
amplification and sequencing of the small subunit ity values ranging from 89.7 to 90%. One clone
rRNA genes related to an unknown Actinomycete (85.6%), anoth-
er to Cyanobacteria (98.5%) and another to 8-
Amplification products of the 16S rRNA genes, using Proteobacteria. Although in Station 11, only one clone
DNA obtained from Stations A and 11 (samples did not cluster with the known bacterial phylogenet-
obtained in September 1993) as templates, were cloned ic branches of Figure 2 (viz: Bacterium soil clone
in E. coli. About 50 clones from each environment group, CytophagaiFlexibacterlBacteroides group, a-
were randomly selected and ca. 210 bp determined (3-8-Proteobacteria, Marine group, High G+C Gram-
of the 16S rDNA sequences (ranging from positions positive, Cyanobacteria, Actinomycetes), the sequence
283 to 499; E. coli numbering) (Brosius et aI., 1981). analysis showed a relationship with some members
......
0
Table 4. Tritiated thymidine incorporation and estimates of bacterial secondary production, growth rates, assimilation and respiration of i4C-Glucose due to bacteria in water samples from
stations in Arcachon Bay (A and B), Certes Domain (C i and C2) and Prevost lagoon (X and 11).
Station Date Incorporation Specific Bacterial Growth Doubling Assimilation Respiration Specific Specific
estimate incorporation production rate time mol Gluc mol Glue assimilation respiration
mol tim L-ih- i estimate ILgCL-ih- i lL(h- i ) t.J (h) L-ih- i L-ih- i estimate estimate
xl0- i2 pmol time eel-ih- i X 10- 9 X 10- 9 mol Glue eel-ih- i mol Glue cel-ih- i
X 10- 9 X 10- 9
A May 1993 5.20 1.04 x 10- 5 0.26 0.026 26.65 1.78 (84.7%) 0.32 (15.3%) 3.56 x 10- 6 0.64 X 10- 6
Sept 1993 10.40 0.86 x 10- 5 0.52 0.029 23.9 0.95 (86.2%) 0.15 (13.8%) 1.12 x 10- 6 0.178 X 10- 6
May 1994 1.51 0.75 x 10- 5 0.07 0.014 49.5 0.47 (88.1 %) 0.06 (11.9%) 2.36 x 10- 6 0.318 X 10- 6
Sept 1994 9.16 3.27 x 10- 5 0.46 0.081 8.55 0.42 (91.3%) 0.04 (8.7%) 1.48 x 10- 6 0.14x 10- 6
Ci May 1993 69.0 40.6 x 10- 5 3.45 1.006 0.69 2.71 (73.3%) 0.98 (26.7%) 15.9 x 10- 6 5.8 X 10- 6
Sept 1993 93.0 4.65 x 10- 5 4.65 0.110 6.30 6.58 (71.1 %) 2.67 (28.9%) 3.29 x 10- 6 1.33 X 10- 6
May 1994 34.0 2.42 x 10- 5 1.70 0.060 11.55 20.8 (85.1 %) 3.62 (14.9%) 6.7 x 10- 6 2.6 X 10- 6
Sept 1994 47.0 2.93 x 10- 5 2.35 0.073 9.50 5.21 (99.9%) 0.99 (1.9%) 3.25 x 10- 6 0.62 X 10- 6
C2 May 1993 50.3 5.03 x 10- 5 2.51 0.100 6.93 8.74 (79.1 %) 2.31 (20.9) 8.74 x 10- 6 2.31 X 10- 6
Sept 1993 27.0 0.64 x 10- 5 1.35 0.005 128.4 9.55 (74.6%) 3.24 (25.4%) 2.27 x 10- 6 0.77 X 10- 6
May 1994 ND ND ND ND ND 28.1 (84.3%) 5.24 (15.7%) 15.6 x 10- 6 2.91 X 10- 6
Sept 1994 5.44 0.28 x 10- 5 0.27 0.007 97.6 0.77 (79.3%) 0.20 (20.7%) 0.41 x 10- 6 0.11 X 10- 6
X May 1993 71.0 4.7 x 10- 5 3.55 0.174 3.94 51.3 (93.1 %) 3.82 (6.9%) 34.2 x 10- 6 2.54 X 10- 6
Sept 1993 23.0 1.64 x 10- 5 1.15 0.041 16.9 1.90 (76.9%) 0.57(23.1%) 1.35 x 10- 6 0.41 X 10- 6
May 1994 2.20 0.23 x 10- 5 0.11 0.006 123.7 0.52 (90.4%) 0.05 (9.6%) 0.53 x 10- 6 0.06 X 10- 6
Sept 1994 4.40 0.34 x 10- 5 0.22 0.008 82.5 0.12 (80.3%) 0.03 (19.7%) 0.09 x 10- 6 0.02 X 10- 6
within these groups. Clone 1611 showed the highest to Vibrio vulnificus). One showed 90.7% similarity
similarity value with Lactobacillus rimae (80.4%). to Aeromonas schubertii. Two isolates showed about
Some clones were found more than once. From the 90% similarity to Bacillus cereus. Finally one showed
analysis of Station A clone I3a (93.5% with Flexibac- 90.7% similarity to Yersinia ruckeri.
ter columnaris) and clone lOa (93.8% with Rhodobac- Four strains, representatives of phototrophic bac-
ter adriaticus) were found 5 times; clone 41a (85.6% teria isolated by using specific media and phenotypi-
with Bacterium soil clone) and clone 33a (93.9% cally characterized in another laboratory (Laboratoire
Cytophaga latercula) four times; clone 48a (95.9% d'Oceanographie Biologique, Bordeaux University)
with Clavibacter xyli) and clone 60a (94.4% Flexibac- were also included in our study. Isolate 5811, identified
ter maritimus) three times; clones 37a (9l.7% with as Thiocapsa sp. showed the highest similarity value
Cytophaga aquatilis), 14a (93.4% with Antarcticum with Chromatium vinosum (91 %), Rhodopseudomonas
vesiculatum) and 26a (93.4% with Flexibacter mar- sp. 2105 showed 95.7% similarity with Rhodopseu-
itimus) twice each. In Station 11 the following clones domonas marina, Rhodobacter sp. 10 2102 showed
appeared twice: clone 1411 (91.6% with Flavobacte- 100% similarity with Rhodobacter sufjidophilus and
ria gondwanense), clone 1111 (91.9% with Cytopha- Chromatium sp. 2203 showed 87.2% with Codakia
ga succinicans), clone 3111 (88.9% with Arthrobac- orbicularis gill symbiont (probably an artifact due to
ter globijormis), clone 4511 (79.5% with Alcaligenes the relatively small number of phototrophic bacteria
eutrophus), clone 7011 (79.l % with Cytophaga lat- 16S rRNA sequences deposited in the data banks).
ercula) and clone 911 (98.5% with Prochlorococcus Results from isolates representative of the Nitrogen
marinus). Clone 2011 (95.2% with Erytrobacter sp.) Cycle Group (Dept. of Biological Sciences, Dundee
appeared three times. Only one sequence (a 200 bp University) were that isolate AAN 013 (ammonifi-
fragment; post. 283-499) was found to be identi- er, Gram-positive, sporulated rod) showed an iden-
cal in both Stations, showing 92.5% similarity with tical sequence to that of isolate AAN 036 (ammoni-
Rhodobacter adriaticus. An almost identical second fier, Gram-positive, sporulated rod), similarity value
sequence (only one nucleotide difference) was also of92.3% with Clostridium bijermentans. Isolate ANO
found in Stations A and II and its closest relationship 011 (denitrifier, Gram-positive cocci) revealed a simi-
was with Cy/ophaga marinoflava (94.6%). larity percentage of 100% with Staphylococcus cohnii
Fourteen randomly taken colonies (seven from (all three from Station A). Isolate PNO 019 (anaer-
each lagoon) isolated in rich medium from the same obic, Gram-negative cocci) showed 92.7% similari-
water samples used for direct DNA extraction were ty with Carnobacterium funditum, and isolate PAN
subjected to 16S rDNA analysis (partial 16S rDNA 031 (amonifier, Gram-positive, sporulated rod), 98.1 %
sequencing of the same variable region, 283-499; E. with Eubacterium tenue, both from Station 1l.
coli). All the isolate sequences were different from
the sequences obtained by direct PCR amplification.
Although again no perfect matching to sequences of Discussion
the data bases was found here, in both environments
most of the isolates had relatively high similarity Heterotrophic bacterial numbers and activities in the
to the typical heterotrophic aquatic bacteria (Vibrio, sampling sites
Aeromonas, Alteromonas). Of the seven isolates from
Station A, three showed more than 95% similarity Out of the five sampling sites studied, station A is
to sequences of known species of this group (95.8% clearly different from the rest, that are all located in
Alteromonas migrifaciens, 96% Vibrio marin us and stagnant waters. Station A contains much less bacteri-
94.9% Alteromonas haloplanktis). One showed 89.6% al biomass and has less heterotrophic bacterial activity.
similarity to Alteromonas haloplanktis. Two showed This is to be expected given the effective tide flushing
over 94% similarity to Bacillus species. Finally one of the waters located at this site. Station A is very much
isolate showed a 93.8% similarity to the Gram-negative influenced by the coastal Atlantic waters. In the water
'j-Proteobacteria, Moraxella phenylpyruvica. Of the column (Table 1), the main difference occurs between
seven isolates from Station 11 three also showed Stations A and 11, Stations C 1, C2 and X being some-
relatively high similarity to be mentioned group of what intermediate. The numbers obtained were slightly
well known aquatic bacteria (96.6% to Vibrio mari- higher at 30°C in Stations C 2 , X and 11, which are
nus and 93.1 % to Alteromonas haloplanktis and 92.7 stagnant waters and probably reach higher tempera-
12
.-----22a
Microscil/a aggregans
1--_ _ _
Flavobacterium breve
L.._ _ _ _ _
L_______C===-Halobacterium salinarium
Haloferax volcanii
37.9 ___.----~--~---~__- - . - - - , r - - - , - -__-,
35 30 20 15 10 5 o
Figure J. Similarity tree showing the relationships of partial l6S rONA sequences from Arcachon Bay (Station A) and representative sequences
of the data base. The scale bar indicates sequence divergence (%).
13
1. a Proteobacteria
2. r Proteobacteria
3. Actinomycete
4. Cyanobacteria
5. f3 Proteobacteria
6. LowGC Gram-positive bacteria
7. High GC Gram-positive bacteria
8. 8 Proteobacteria
9. Bacteria soil group
10. CytophagaiFlexibacterlBacteroides
11. Spirochaete
12. Archaea
Figure 2. Similarity tree showing the relationships of partial 16S rDNA sequences from Prevost Lagoon (Station II) and representative
sequences of the data base. The scale indicates sequence divergence (%).
14
tures than Station A which is much more similar to teria and Chi suggest that bacteria should respond to
open ocean waters known to contain important num- increases in phytoplankton nutrients. If bacterial pop-
bers of psychrophilic microbes. In any case the water ulations were controlled by forces acting via bacte-
in the Prevost Lagoon, particularly in the areas distant riovore ingestion, bacterioplankton abundance should
from the Mediterranean Sea Inlet (Station 11) is charac- be inversely correlated with bacteriovore abundance or
terized by considerably higher viable counts, which are biomass, and there is no evidence of this.
in agreement with the known higher productivity of this Heterotrophic bacterial production values in open
lagoon (Caumette, 1986). The presence of significant oceans range from 0.08 JLg C L -I h- I in the Warm Core
faecal contamination, as indicated by faecal coliforms Ring Gulf to 0.45 JLg C L -Ih- I in the Coral Sea, Aus-
(Table 2) and sulphite-reducing clostridia (Table 3) in tralia. In coastal ocean waters, from 2.2 JLg C L -Ih- I
Station 11 (14-20 FC and 50-60 clostridia/IOO ml) is at Scripps Pier to 91 JLg C L -Ih- I in Bietri, Ivory
a clear indication that this lagoon is subject to consid- Coast (Ducklow and Carlson, 1992). Values are higher
erable urban pollution from the surrounding popula- in coastal waters than in open ocean waters. Thymi-
tion centres. In contrast, Arcachon Bay appears to be dine incorporation in Arcachon Bay ranged from 0.014
virtually free of significant faecal contamination. The to 0.52 JLg C L -I h -I, with values similar to those
presence of clostridia in the sediment is not significant for open oceans. In the Certes Domain the average
since the spores can be accumulated in this environ- value was 2.8 JLg C L -Ih- I and in Prevost Lagoon
ment for very long periods, nor is it an appropriate 4.55 JLg C L -Ih- I. These data support our original
pollution indicator in this environment. Even so, the hypothesis that both systems represent very different
increase in clostridia counts at the end of the summer situations that can be considered as different models.
in Arcachon Bay is significant, probably as an indi- One (Arcachon) with a low level of eutrophication and
cation ofthe cumulative pollution suffered throughout urban pollution, relatively low productivity and strong
the summer season. influence of the surrounding ocean waters; the other
The sediments varied considerably from the above (Prevost) highly eutrophic, with significant levels of
pattern. In the top two centimetres of the sediment urban pollution and very different (isolated) from the
(Table 1) the numbers of heterotrophs seem to be the neighbouring Mediterranean Sea waters.
equivalent of the water numbers, i.e. Station A showed
the highest counts, followed by C 2 , Band C I , while Analysis of bacterial biodiversity by direct
the Prevost Lagoon stations showed lower numbers amplification and sequencing of the small subunit
in the superficial sediment. This apparent discrepancy rRNA genes
can be explained if we assume that while in Prevost
Lagoon the larger (or a very significant) part of the pri- In oligotrophic ocean waters the studies on micro-
mary productivity takes place in the water column due bial diversity of naturally occurring organisms by
to the low light penetration in these eutrophic waters molecular approaches have shown that clones were
while in the more oligotrophic waters of Arcachon mostly phylogenetically related with Cyanobacteria
Bay most of the primary productivity takes place in (Fuhrman et aI., 1993; Giovannoni etal., 1990) SARll
the surficial sediment, richer in nutrients and directly group (a-Proteobacteria) (Giovannoni et aI., 1990),
exposed to sunlight at low tide. In the deep sediment I'-Proteobacteria (Fuhrman et aI., 1993), Marine
(Table 1) sample results revert to the water pattern Group A (Fuhrman et aI., 1993) (a new phylogenet-
showing higher numbers in the Prevost Lagoon, prob- ic branch) and novel groups of the Archaea domain
ably because this deep anaerobic sediment is not relat- (DeLong, 1992; Fuhrman et aI., 1992; DeLong
ed to the superficial sediment or water productivity. et aI., 1994). Minor numbers were related with
In any case the deep sediment samples are probably Gram-positives (Fuhrman et aI., 1993), Cytopha-
dominated by strict anaerobes that are not detected by gaiFlexibacte rlBacte roides (Fuhrman et aI., 1993) and
present culture techniques used. Planctomyces (Fuhrman et aI., 1993). Other molec-
Ducklow & Carlson (1992), in an extensive study ular approaches to compare coastal waters with open
of oceanic bacterial production, concluded that bac- ocean waters have been used. For example, an oligonu-
terial biomass is controlled by resources rather than cleotide probe derived from the SARlI (Giovannoni
by predation. Bacterial abundance and bacterial pro- et aI., 1990) sequence retrieved from the Sargasso
duction are statistically correlated with phytoplankton Sea was hybridized against total RNA extracted from
biomass (as Chi). Positive correlations between bac- picoplankton samples obtained from both origins (Gio-
15
vannoni et aI., 1990). The results showed that the rel- cellulose and probably playa major role in the turnover
ative importance of these groups of Proteobacteria of matter in nature. Their main habitats are soils,
was lower in several coastal samples, suggesting that decaying plant material and animal dung (Reichen-
the organisms contributing these sequences may be bach, 1991). It seems that at the end of summer, after
more abundant in relatively oligotrophic open ocean proliferation of macroalgae (Ulva sp., Enteromorpha
waters. Another approach compared different samples sp.) decomposition of the macroalgae material would
by hybridization of total DNA obtained from different stimulate development of Cytophagales. They could be
sampling sites (Lee & Fuhrman, 1991), the results of even more important in the aggregate-associated biota
this work also suggested that the bacterial community that is probably retained by the glass-fibre filter.
of coastal waters may be different from that of open In Station A, four clones were phylogenetically
ocean water. related to a novel eubacterial soil group described by
The bacterial biodiversity revealed by our 16S Liesack and Stackebrandt (1992). This phylogenetic
rDNA study in the two coastal lagoons showed some line was not represented in oligotrophic ocean waters,
differences from any of those previous studies carried it could reflect the continental influence in coastal
out in open ocean waters. Two clones from Station lagoons. We also found clones related with members of
A (8a and 39a) and one from Station 11 (6511) were Gram-positives (low and high G+C content). Fuhrman
closely related to the widely distributed SARlI group (1993) found 9 clones belonging to a new phyloge-
(94.2%, 96.8% and 88% respectively). But these are netic group apparently related to the high G+C Gram-
very few representatives in comparison with the larger positive Bacteria and represented by clone BDAI-5, a
number of clones found in ocean waters (Fuhrman et single isolate from the Pacific was identical to this one
aI., 1993). In both stations, we also found clones relat- and the rest varied only at I or 2 positions. However,
ed with the Marine Group A described by Fuhrman the Gram-positive type sequences that we found are not
et ai. (1993): clone 38a and 17a (98.9% and 96.1%) related to those found by Fuhrman (1993). They could
and clones 1911,211,6911 and 5511 (86.7%,70%, also be of terrestrial origin. Remarkable also is the rel-
89.8% and 82.0% respectively). So the novel phylo- atively low representation found for ,-Proteobacteria
genetic lines described as being predominant in olig- (only 2 clones in Station 11 and none in Station A).
otrophic ocean waters are present in coastal lagoons This group has been found in open ocean waters by the
although the proportion of clones belonging to these 16S rDNA technique (Fuhrman et aI., 1993) but they
groups retrieved was much lower, suggesting that they were also in a minority, which is in contrast with previ-
could be less abundant in coastal lagoons. Another ous studies about oligotrophic ocean waters. However,
group that was relatively little represented in our study most of the heterotrophic bacteria that are easily iso-
is the Cyanobacteria that was found as an important lated from the ocean belong to this phylogenetic clus-
component of the picoplankton of open ocean waters ter. Furthermore, our own isolates from the lagoons
(Giovannoni et aI., 1990). Only a clone in Station 11 studied, although none showed a perfect match with
was similar to a Cyanobacteria. The above does not any sequence deposited in data banks, all correspond-
necessarily mean that this group of prokaryotes is not ed to organisms belonging to the ,-Proteobacteria.
present in the environments studied. The large cellu- These results support the view that some groups of the
lar size of most members of the well known groups ,-Proteobacteria (Vibrio for example) are particular-
of Cyanobacteria excludes them from our study since ly well fitted for growth in laboratory cultures, and
they would be retained by the glass fibre filter which although abundant in marine habitats, they are prob-
nominal pore size is 2.7 /Lm. ably in a minority or at least not predominant in this
On the other hand, in the coastal lagoons stud- kind of environment.
ied here we have observed a great number of repre- Regarding both coastal aquatic systems, the
sentatives of the CytophagalFlexibacteriBacteroides study of their diversity reflects some shared simi-
group. There is evidence that members of the Cytopha- larities. The major group represented is Cytopha-
galFlexibacterlBacteroides group are present in ocean galFlexibacterlBacteroides in both Stations and (Y-
waters but in macroaggregate-associated bacterial pop- Proteobacteria. However, they also differ in some
ulations (DeLong et aI., 1993). These microorganisms aspects. An analysis in detail revealed that only one
are Gram-negative bacteria, gliding, aerobic, organ- identical sequence was found in both environments
otrophs. Many of them are able to degrade biomacro- (with a 92.5% similarity to Rhodobacter adriaticus).
molecules like proteins, chitin, pectin, agar, starch or The Prevost Lagoon showed higher diversity by using
16
the 16S rRNA gene amplification technique. The num- Caumette, P., 1986. Phototrophic sulfur bacteria and sulfate/reducing
ber of different sequences retrieved was larger: 43 dif- bacteria causing red waters in a shallow brackish coastal lagoon
(Prevost Lagoon, France). FEMS Microbiol. Ecol. 38: 113-124.
ferent sequences in the 52 clones studied, while from DeLong, E. F., 1992. Archaea in coastal marine environments. Proc.
Arcachon out of the same number of clones only 31 Natl. Acad. Sci. USA 89: 5685-5689.
different sequences were found i.e. redundancy was DeLong, E. F., D. G. Franks & A. L. Alldredge, 1993. Phylogenetic
higher. Furthermore, in Arcachon we found represen- diversity of aggregate-attached vs free-living marine bacterial
assemblages. Limnol. Oceanogr. 38: 924-934.
tatives of only five of the major phylogenetic branch- DeLong, E. F., K. Y. Wu, B. B. Prezelin & R. V. M. Jovine, 1994.
es of eubacteria (Woese, 1987) while in the Prevost High abundance of Archaea in Antarctic marine picoplankton.
lagoon we found representatives of ten different phy- Nature 371: 695-697.
Ducklow, H. w. & c. Carlson, 1992. Oceanic Bacterial Production.
logenetic branches including one Actinomycete, one
Adv. Microb. Ecol. 12: 113-181.
Cyanobacteria and a relatively large number of low Fuhrman, 1. A. & F. Azam, 1980. Bacterioplankton secondary pro-
G+C Gram-positives. These results suggest that bac- duction estimates for coastal waters of British Columbia, Antarc-
terial biodiversity is higher in the eutrophic Mediter- tica and California. Appl. en vir. Microbiol. 39: 1085-1095.
Fuhrman, 1. A., D. E. Comeau, A. Hagstron & A. Chan, 1988.
ranean lagoon that in the relatively oligotrophic and
Extraction from natural planktonic microorganisms of DNA suit-
colder Atlantic one and that there are also some differ- able for molecular biological studies. Appl. envir. Microbiol. 54:
ences regarding the groups represented in both envi- 1426-1429.
ronments. Fuhrman, 1. A., K. McCallum & A. A. Davis, 1992. Novel major
archaebacterial group from marine plankton. Nature 356: 148-
149.
Fuhrman, 1. A., K. McCallum & A. A. Davis. 1993. Phylogenetic
Acknowledgements diversity of subsurface marine microbial communities from the
Atlantic and Pacific Oceans. Appl. envir. Microbiol. 59: 1294-
1302.
This work was supported by grants PB 93/0930 of Giovannoni, S. 1., T. B. Britschgi, C. L. Moyer & K G. Field, 1990.
the DGICYT, BIO 93/0750 of the CICYT, and the Genetic diversity in Sargasso Sea bacterioplankton. Nature 345:
C.L.E.A.N. project contract No. EV5V-CT92-0080 of 60-63.
the European Commission. A.1.M.M. was recipient of Jannasch, H. W. & G. E. lones, 1959. Bacterial population in sea
water as determined by different methods of enumeration. Lim-
a post-doctoral contract of the Spanish Ministry of Edu- nol. Oceanogr. 4: 128-140.
cation and Science. Thanks are offered to Prof. Waltre Lane, D. J., B. Pace, G. 1. Olsen, D. A. Stahl, M. L. Sogin & N.
Sttihmer for technical support to carry out the phyloge- R. Pace, 1985. Rapid determination of 16S rDNA sequences for
netic analysis. Secretarial assistance by K. Hernandez phylogenetic analysis. Proc. natn. Acad. Sci. USA 82: 6955-
6959.
is gratefully acknowledged. Lee, S. & 1. A. Fuhrman, 1991. Spatial & temporal variation of
Natural Bacterioplankton assemblages studied by total genomic
DBA cross-hybridization. LimnolOceanogr. 36: 1277-1287.
Liesack, W. & E. Stackebrandt, 1992. Occurrence of novel groups of
References
the Domain Bacteria as revealed by analysis of genetic material
isolated from an Australian Terrestrial Environment. J. Bacteriol.
Amann, R. I., W. Ludwig & K-H. Schleifer, 1995. Phylogenetic 174: 5072-5078.
identification and in situ detection of individual microbial cells Maniatis, T.. E. F. Fritsch & 1. Sambrook, 1982. Molecular cloning:
without cultivation. Microbiol. Rev. 59: 143-169. a laboratory manual. Cold Spring Harbor Laboratory Press, Cold
APHA-AWWA-WPCF, 1975. Standard methods for the examination Spling Harbor, N.Y.
of the water and wastewater. 14th ed., (Rand, M. c., A. E. Green- Marchuk, D., M. Drumm, A. Saulino & F. S. Collins, 1990. Con-
berg & M. J. Taras, Eds), American Public Health Association, struction of T-vectors, a rapid and general system for direct
Washington. cloning of unmodified PCR products. Nucl. Acids. Res. 19: 1154.
Beaucage, S. L. & M. H. Caruthers, 1981. Dideoxynucleotide phos- Martinez-Murcia, A. 1., 1993. Doctoral Thesis. Department of
phoramidites - a new class of key intermediates for deoxypolynu- Microbiology, University of Reading, Reading, U.K
cleotide synthesis. Tetrahedrom Lett. 22: 1859-1862. Martinez-Murcia, A. 1. & F. Rodriguez-Valera, 1994. The use of
Benlloch, S., F. Rodriguez-Valera & A. 1. Martinez-Murcia, 1995. artibrarily primed PCR (AP-PCR) to develop taxa specific DNA
Bacterial diversity in two coastal lagoons deduced from 16S probes of known sequence. FEMS Microbiol. Lett. 124: 265-270.
rDNA PCR amplification and partial sequencing. FEMS Micro- Moriarty, D. 1. w., 1990. Techniques for estimating bacterial growth
bioI. Ecol. 18: 267-280. rates and production of biomass in aquatic environments. Meth-
Bradshaw, L. 1., 1992. Laboratory Microbiology. Fourth Edition. ods in Microbiology. Vol. 22. Pag. 211-234.
Saunders College Publishing. Pace, N. R., D. A. Stahl, D. J. Lane & G. J. Olsen, 1986. The analysis
Brosius, J. P., T. J. Dull, D. D. Sleeter & H. F. Noyer, 1981. Gene of natural microbial populations by ribosomal RNA sequences.
organization and primary structure of a ribosomal RNA operon Adv. Microb. Ecol. 9: I-55.
from E. coli. 1. Mol. BioI. 148: 107-.127. Reichenbach, H., 1991. The Order Cytophagales. In: The Prokary-
Castel, 1., P. Caumette & R. Herbert, 1996. Eutrophication gradients otes, Second Edition (Balows, A., H. G. Triiper, M. Dworkin, W.
in coastal lagoons. Hydrobiologia 329: xi-·xxx.
17
Harder & K.-H. Schleifer, Eds), pp. 3631-3675, Springer Verlag, Schmidt, T. M., E. F. DeLong & N. R. Pace, 1991. Analysis of a
New York. marine picoplankton community by 16S rRNA gene cloning and
Sanger, F., S. Nicklen & A. R. Coulson, 1977. DNA sequencing sequencing. 1. Bacteriol. 173: 4371-4378.
with chain-terminating inhibitors. Proc. natll. Acad. Sci. USA Woese, C. R., 1987. Bacterial Evolution. Microbiol. Rev. 51: 221-
74: 5463-5467. 271.
Wright. R. T. & 1. E. Hobbie, 1966. Use of glucose and acetate by
bacteria and algae in aquatic ecosystems. Ecology 47: 447-464.
Hydrobiologia 329: 19-31, 1996. 19
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996. Kluwer Academic Publishers.
Key words: Biodiversity, 16S rRNA, halophilic Bacteria, halophilic Archaea, hypersaline environments
Abstract
New methods based on PCR amplification of 16S rRNA genes from DNA samples extracted directly from the
environment allow the description of microbial diversity in natural ecosystems without the need for cultivation.
We have applied this technique to an extreme environment presumed to have very low diversity: the crystallizer
ponds of a marine saltern with salinity over NaCl saturation. The molecular methodology has shown that indeed
very low diversity can be found here. Prokaryotes belonging to the Bacteria domain are a minor component and
only members of a closely related cluster of sequences were found, all relatives of the a-Proteobacteria (ca. 83%
to Rhodopseudomonas marina). Halophilic Archaea were as expected the largest component of biomass in this
environment. All the clones sequenced corresponded again to a highly homologous cluster (probably members
of the same genus). However, all the sequences diverged considerably from the ones of the described genera of
halophilic Archaea, in fact the data are consistent with the idea that the 16S rRNA genes directly amplified from the
saltern correspond to members of an undescribed genus. This is remarkable since many collection strains sequenced
come specifically from this saltern. Furthermore, 16S rDNA obtained from archaeal cultures isolated from the same
sample had no homology to the sequences obtained by PCR amplification, instead they appear to be members of
the well known genus Haloarcula. However, this concurs with the findings of other authors who obtained different
organisms by culture from those detected by the sequences retrieved directly by PCR. A possible explanation is
that culturability, in standard media, is the exception rather than the rule. To study the biodiversity gradient present
along the salinity gradient found in a multi-pond solar saltern we have also applied a novel molecular strategy. This
method is based on the restriction digestion of a population of 16S rDNA sequences directly amplified from an
environmental sample. Digested fragments separated by polyacrylamide gel electrophoresis generate characteristic
profile data for estimation of diversity and overall similarities between the organisms of different environments. The
methodology has been applied to a set of five ponds covering the salinity gradient from about twice that of seawater
(6.4%) to NaCl precipitation (30.8%). Bacterial (eubacterial) diversity estimated from the complexity of the banding
pattern obtained by restriction of the amplicons from the different ponds decreased with increasing salinity while
for Archaea (archaebacteria) the reverse was true i.e. the higher the salinity the higher the number of bands. The
similarities in taxonomic composition of the prokaryotic populations present in those ponds were evaluated from
the number of restriction bands shared by the different samples. The relationships found among the different
environments were independent of the enzyme used for digestion and were consistent with previous descriptions
obtained by the study of isolates from the different environments. This technique appears to be promising as a rapid
method for microbial biodiversity fingerprinting useful to compare several environments and detect major shifts in
species composition of the microbial population.
on marine agar plates (Giovannoni et aI., 1990b), while found in the Atlantic and Pacific Oceans (Fuhrman
it is known that about half are metabolically active. et aI., 1993). A new archaeal phylogenetic branch
DNA reassociation techniques have shown that only 1 diverging deeply from the thermophilic branch or Cre-
in about 200 genomes present in a Norwegian soil was narchaeota has also been found in ocean waters includ-
recovered by culture (Torsvik et aI., 1990). This major ing deep ocean samples (500 m) and permanently cold
drawback has recently been bypassed - at least to some Antarctic waters (DeLong et aI., 1994; Fuhrman et aI.,
extent - by the use of molecular techniques (Pace et aI., 1992). The low temperature of the environments from
1986). The most widely used and, up to the moment, which these sequences were recovered, as well as the
most fruitful approach has been the PCR amplification low G + C of the 16S rDNA lead the authors to sug-
(or cloning) of 16S rDNA directly from the sample gest that the phenotype of the corresponding organism
(DeLong, 1992; DeLong et aI., 1994; Fuhrman et aI., would not be thermophilic at all and therefore very
1993; Giovannoni et aI., 1990a; Giovannoni et aI., different from other Crenarchaeota. A novel line of
1990b; Liesack & Stackebrandt et aI., 1992; Schmidt Bacteria was also discovered in soil with less than 75%
et aI., 1991; Stackebrandt et aI., 1993; Torsvik et aI., homology to any known organism (Liesack & Stacke-
1990). The sequences of the rDNA genes retrieved brandt et aI., 1992). The fact that no perfect matching
from the environment give an estimate of the micro- to any known sequence has been found for any of the
bial diversity present. The comparison of the sequences 16S rDNAs directly retrieved from the environment,
with data banks allows the phylogenetic placement of not even in soil, an environment lengthily studied by
the microorganism. If the amplified DNA were to show classical microbial ecology approaches, is also remark-
enough homology to a cultivated taxonomic group it able. There are also a number of papers reporting the
would even be possible to predict the physiology of the results of the technique applied to thermal environ-
microorganism detected and concomitantly its puta- ments. Since extreme environments are characterized
tive ecological role. However, the number of samples by low biological diversity, results could be simpler
and sequences per sample is limited by the sequencing to interpret. In fact the number of different sequences
capabilities of the laboratory and therefore most studies found was in general lower. For example, from a pink
that have been carried out investigate only about 100 filament community in Octopus Spring (84-88 QC,
sequences at the most. To be able to evaluate the biodi- located in Yellowstone National Park) of 35 clones
versity present in a sample rapidly, and also to compare studied 26 were identical (Reysenbach et aI., 1994). In
several samples, for example when time or spatial gra- the cyanobacterial mat also at Octopus Spring (waters
dients are studied, a more expedient methodology is at55 QC), 5 identical sequences were found in 14 clones
required. Restriction fragment length polymorphism studied (Weller et aI., 1991). By restriction analysis of
of 16S rDNA has been shown to be highly informa- 16S rRNA genes the clones obtained in a deep-sea
tive (Avaniss-Aghajani et aI., 1994; Moyer et aI., 1994; hydrothermal vent, after application of statistical tests,
Ralph etaI., 1993) and could therefore be used to obtain showed that only about 10 OTU (Operational Taxo-
a fingerprint of microbial biodiversity in the environ- nomic Units) were present in the environment (Moyer
ment under study with a relatively small investment et aI., 1994). In a study carried out with the sedi-
of time and cost, so allowing several samples to be ment in another thermal environment, Jim's Black Pool
analyzed simultaneously. (74 QC), again in Yellowstone, of 98 clones obtained
The cloning and PCR amplification of 16S rDNA with archaeal primers, three sequences were repeat-
technique has already been applied to a number of ed 23, 8 and 7 times respectively while the remainder
marine, soil and thermal (extreme) natural environ- were represented 1-4 times in the collection (Barns
ments (Barns et aI., 1994; DeLong, 1992; DeLong et aI., 1994). However, in this work a considerable
et aI., 1994; Fuhrman et aI., 1993; Giovannoni et aI., diversity of archaeal 16S rRNA sequences was found
J990a, 1990b; Liesack & Stackebrandt et aI., 1992; and although some clones had high sequence similarity
Schmidtetal., 1991; Stackebrandtetal., 1993; Torsvik to cultivated thermophiles (about 97%) others showed
et aI., 1990; Ward et aI., 1990; Weller et aI., 1991). This great evolutionary distance from any known organism.
has lead to the discovery of a remarkable unknown Nevertheless, a perfect matching between clones and
diversity of both Eubacteria and Archaea. A new 16S rRNA sequences of cultivated microorganisms has
eubacterial group distantly related to a:-Proteobacteria never been described, even when isolated from specif-
represented by clone SAR 11 first described by Gio- ically the same environment and location.
vanonni et al. (1990a) is widely distributed, being
21
Hypersaline environments represent a very inter- Haloferax volcanii ATCC 29605 T , Haloferax gib-
esting model, particularly the multipond saltern in bonsii ATCC 33959 T , Haloferax denitrificans ATCC
which a gradient of salinities develops spatially, dif- 35960T , Haloferax 'larsenii' ATCC 49116, Haloar-
ferent ponds representing different salinity values that cula hispanica ATCC 33960T , Haloarcula califor-
range from sea water (ca. 3.5% w/v) to NaCI precipita- niae ATCC 33799 T , Haloarcula marismortui ATCC
tion (over 35% w/v) (Javor, 1983a; Rodrfguez-Valera 43049 T , Halobacterium salinarium ATCC 33171 T ,
et aI., 1985). Along this gradient biodiversity decreas- Halobacterium lacusprojundii ATCC 49239 T , Halo-
es as the environment becomes more hypersaline i.e. coccus morrhuae CCM 537 T , Halococcus saccha-
more extreme. In the crystallizer ponds, where NaCl rolyticus ATCC 49257 T , were cultivated on 25%
is harvested, only members of the Archaea domain are (w/v) salt water agar supplemented (Rodrfguez-Valera
recovered by culture and they belong to only 3 or 4 et aI., 1981) with 0.5% yeast extract, 1% starch and
species of haloarchaea (formerly halobacteria) (Oren, 1% glucose at 37°C for approximately one week.
1990; Rodrfguez-Valera et aI., 1985). Microscopically, Escherichia coli ATCC 25922, Pasteurella multo-
samples of the crystallizer water also show very little cida ATCC 43137 T , Listeria monocytogenes ATCC
diversity, long slender rods predominating, often gas 15313 T , Proteus mirabilis ATCC 29906 T , Salmonella
vacuolated similar to the cells of some species of the sp. ATCC 35664 and Aeromonas salmonicida NCIMB
genus Halobacterium and some square bacteria like 1102 T were cultivated as recommended by the Amer-
those described by Walsby (Javor, 1983b; Oren, 1990; ican Type Culture Collection.
Rodrfguez-Valera et aI., 1985; Walsby, 1980).
In the present work direct 16S rDNA amplifica- Biomass collection
tion and sequencing were applied to the DNA obtained
from a crystallizer pond. Bacterial and archaeal popu- Approximately 5 I of superficial water (10 cm below
lations were also characterized by the analysis of the the surface) were collected from ponds with differ-
16S rRNA genes directly amplified from the envi- ent salinity of a solar saltern located in Santa Pola
ronment and digested with frequent cutting restriction (Alicante, Spain). Debris and eukaryotic cells were
endonucleases. The technique has indeed allowed us removed by using glass microfiber filters (Whatman
to characterize bacterial biodiversity in five selected GF/D). Microorganisms were collected by positive
ponds with different salinities (6.4-30.8%) and to com- pressure filtration on a 9 cm, 0.22 11m pore-size fil-
pare the similarity of prokaryotic populations present ter (Durapore; Millipore Corp., Bedford, MA). Water
in those ponds. was pumped until filter clogging decreased the flow
rate to a few ml min-I. Filters were stored at -80°C
until DNA extraction.
Materials and methods
DNA extraction and purification
Salinity
Thawed filters were cut with a sterile razor blade into
The total salt concentrations were determined by the small pieces and vortexed in 5 ml of 1% w/v SDS, TE
ash content of 10 ml of water dried at 110 °C for 24 h buffer (100 mM Tris-HCl; pH 8; 10 mM EDTA), in
and afterwards at 400°C for 4 h. 50 ml conical based polypropylene centrifuge tubes.
Acid-washed glass beads (Sigma) were added and vor-
Media and strains texed until resuspension (1-2 min). The sample was
incubated at 37°C for 1 h after addition of 50 111 of
To isolate pure cultures from the crystallizer water, lysozyme (Boehringer) in TE buffer (1 % w/v). After
100 /1,1 were inoculated onto plates containing 25% boiling for 10 min and vortexing for 2 min, 1 ml of
(WN) salt water agar supplemented with 0.5% yeast lysis buffer (4% w/v SDS; 100 mM EDTA; 50 mM
extract (Rodrfguez-Valera et aI., 1981). Colonies Tris-HCI; pH 8), and 30 111 proteinase K (14.7 mg
were isolated at random and subcultured twice; 1 ml ml- 1 stock solution) was added. The samples were
of bacterial cultures was harvested in sterile 25% vortexed for 1 min, incubated at 56°C for 1 hand
salt water by centrifuging at 10000 rpm for 2 min boiled for 5 min. As no complete lysis was observed
and pellets were stored at -20°C. Culture collec- at this step, a thermal shock (5 min at -80°C and
tion strains, Haloferax mediterranei ATCC 33500T , 5 min at boiling) was applied. The DNA was extract-
22
ed once with a volume of phenol-Tris (pH 8) and a with EcoRV and incubated with Taq I DNA polymerase
volume of chloroform-H 20; 0.1 volume of 2M NaAc (1 UIJLg plasmid/20 JLI volume) using standard buffer
(pH 4.8) and 3 volumes of absolute ethanol were added conditions in the presence of 2 mM dTTP for two
and kept overnight at -20°C. After centrifuging at hours at 72 °C (Marchuk et aI., 1990) and ligated at
12000 rpm for 10 min, the pellet was dried and DNA 16°C overnight. XLBlue Escherichia coli strain was
purified using the GeneClean II Kit. DNA from bac- transformed with a standard protocol (Maniatis et aI.,
terial cultures was extracted as previously described 1982). Plasmidic DNA was extracted using the Wizard
(Martinez-Murcia & Rodriguez-Valera, 1994). Quan- Minipreps DNA Purification System (Promega).
tification and purity (referred to protein content) of
the DNA obtained were estimated and 5 JLI aliquots Primers and sequencing reactions
of DNA dilutions were analyzed (Martinez-Murcia &
Rodriguez-Valera, 1994). Primers with 5'-TTACCGCGGCTGCTGGCACG-3'
sequence (Martinez-Murcia, 1993) at positions 533-
PCR amplification of 16S rDNA 514 (E. coli numbering) for the Bacterial Domain and
5'-CCCCGTAGGGCCCGG-3'(Barns et aI., 1994) at
Reaction mixtures contained 50 mM KCl, 10 mM positions 330-316 (H. salinarium numbering) for the
Tris-HCI (pH 9), 1.5 mM MgCI 2 , 0.1 % Tri- Archaeal Domain were chosen to sequence ca. 200 bp
ton X-I00, 200 JLM each of deoxyribonucleotide of a relatively variable region of each selected clone.
triphosphate (dATP, dCTP, dGTP, dTTP; Pharma- Sequencing primers 5' -GGATTTCACTCCTACCCC-
cia LKB Biotechnology), 2U of Taq I DNA poly- 3' and 5'-CCATTGTAGCCCGCGTGT-3' at posi-
merase (Promega), 0.2 pM each of oligonucleotide tions 646-629 and 1203-1186 respectively (H. sali-
primer and 100 ng of template DNA in a total narium numbering) were designed in the present
volume of 50 JLI. The sequences of the forward study from a 16S rRNA sequence alignment
primers were 5'-TTCCGGTTGATCCTGCCGGA-3' of members of Halobacteriaceae (Kamekura &
(DeLong, 1992) ranging from nucleotide posi- Seno, 1993). These together with the primers 5'-
tions 2 to 21 of the Halobacterium salinari- ACCGGCGTTGACTCCAATT-3' (DeLong, 1992)
um (Larsen & Grant, 1989) numbering system and 5'-GACGGGCGGTGTGTGCA-3' (Barns et aI.,
(Mankin et aI., 1985) for the Archaeal Domain and 1994) at positions 929-912 and 1368-1352 respec-
5'-AGAGTTTGATCATGGCTCAG-3' (Lane et aI., tively (H. salinarium numbering) and the Bluescript
1985) at nucleotide positions 8 to 27 of the E. coli sequencing primers, were used when determination
numbering (Brosius et aI., 1981) for the Eubacterial of the complete archaeal 16S rDNA sequences were
Domain. Reverse primer for both Domains was 5'- required. Primers were synthesized by the phospho-
GGTTACCTTGTTACGACTT-3' (Lane et aI., 1985) ramidite method (Beaucage & Caruthers 1981) on a
at positions 1509-1491 (E. coli numbering). Reac- Model 392 RNAIDNA Applied Biosystems synthe-
tion mixtures were overlaid with mineral oil (Light sizer. Clones randomly selected were sequenced by
White Oil; Sigma) prior to a thermal cycling regime the dideoxy-chain termination method (Sanger et aI.,
(35 cycles on a Perkin Elmer Cetus 480) of 94°C for 1977) with a Sequenase Version 2.0 Kit (United States
1 min, 55°C for 1 min, and 72 °C for 2 min. After the Biochemicals) following the manufacturer's protocol.
final cycle was completed, the amplicons were allowed
to extend at 72 °C for 10 min. Negative controls were Southern hybridization
included with no addition of template DNA. Five JLI of
the PCR products were analyzed on 1% (w/v) agarose Different amounts of DNA extracted from the crystal-
gels to check for purity. PCR products were precipitat- lizer were dot blotted onto a Hybond-N-Nylon mem-
ed, dried and resuspended in 25 JLI of sterile distilled brane (Amersham). Total 16S rDNA amplified with
water. either eubacterial or archaeal primers were digoxi-
genin labelled by random primed synthesis accord-
Purification and cloning of the PCR products ing to the procedure recommended by Boerhinger
Mannhein (DIG-System, Boerhinger Mannheim) and
PCR products were purified using GeneClean II kit and then separately hybridized to the total DNA blots.
cloned into a T-tailed-vector prepared as follows: Blue- Hybridization was carried out overnight at high strin-
script plasmid (Stratagene, La Jolla, CA) was digested gency conditions (68-70 0C). Filters were washed
23
twice for 5 min at room temperature with 2 x SSC, data was computed using the NTSYS program version
0.1 % SDS and twice for 15 min at 68°C with 0.1 x in IBM Pc.
SSC and 0.1 % SDS. Chemoluminescent detection was
performed using Lumiphos™ as a substrate for the Nucleotide sequence accession numbers
alkaline phosphate linked to the anti-digoxigenin anti-
body (Boerhinger Mannheim) following the recom- Amplified products were cloned and the sequences
mendations of the manufacturer. After two washes for determined in the present study were deposited in the
15 min in 0.2 N NaOH, 0.1 % w/v SDS at 37°C, and EMBL data base with accession numbers: X84084,
equilibration in 2 x SSC, blots were stored at -20°C X84330, X84331, X84322, X84323, X84324.
and reused up to six times. Probe solutions were also
maintained at - 20 ° C, and reused following denatura-
tion (10-15 min at 90-100 0C). Results
Restriction endonuclease digestions The mUlti-pond solar saltern selected for this study
has been studied previously by other techniques and
Enzymatic digestions were performed by incubating the gradient of physicochemical conditions as well as
25 III of the amplified products with 5U of each endonu- the microbial populations that develop along this gra-
clease and the corresponding enzyme buffer. Diges- dient have been described (Rodrfguez-Valera et aI.,
tions were continued overnight at 37°C for Alul, HinjI, 1985; Rodrfguez-Valera et aI., 1981; Rodrfguez-
and Mbo!. Digested products were precipitated with Valera, 1986). The water of the five ponds sampled
ethanol, dried and resuspended in 5 III of dye loading had salinities representing the increase gradient from
buffer. about twice the seawater salinity (Preparation Pond
PR3, 6.4%) located in the so-called carbonate domain
Electrophoresis (Barbe et aI., 1990) to saturation for sodium chloride,
halite domain (Crystallizer Pond CR30, 30.8%), the
The digested products were heated for 2 min at 96 ° C concentrator ponds have intermediate salinities (gyp-
and chilled on ice before loading on 55 cm wedge- sum domain) (as shown in Table 1). Bacterial and
shaped (0.2-0.6 mm) 6% (w/v) polyacrylamide dena- archaeal primers gave good amplification from all the
turing (7 M urea)-TBE gels. Electrophoreses were per- samples except the two ponds oflower salinity PR3 and
formed in TBE buffer at 55°C using a LKB Macrophor PR6 that gave no amplification when archaeal primers
2010 sequencing unit operated at 50 W per gel. The were used, indicating a low presence or non-existence
bromophenol blue was allowed to migrate to the bot- of archaeal representatives at low salinity. Archaeal
tom; the gel was washed in 10% v/v glacial acetic groups have been found in marine samples of differ-
acid for 20 min, and silver stained by using the Silver ent origins by 16S rDNA direct PCR amplification and
Sequence TM Kit (Promega) following the manufac- sequencing but the samples were always obtained from
turer's protocol. The gel was photographed using a open ocean, relatively oligotrophic waters (DeLong,
14/1 x 17/1 EDF film (Promega). 1992; DeLong et aI., 1994; Fuhrman et aI., 1992).
Apparently the much more eutrophic, as well as more
Analysis of the sequence data saline waters of these two ponds are devoid of a sig-
nificant proportion of this phylogenetic group.
The determined sequences were compared with 16S Although the 16S rRNA genes were amplified
rRNA sequences obtained from the GenBank and the from the crystallizer with both eubacterial and archaeal
EMBL Data Library by using programs of the Wis- primers, the relative abundance of both domains was
consin Molecular Biology Package Version 8. Com- very different as shown by hybridization of the prod-
plete 16S rDNA sequences of archaeal clones HACI ucts from both amplifications to the total DNA extract-
and HAC14 were manually aligned to those from ed and blotted on membranes (see Figure 1). The
other halophilic Archaea (Kamekura & Seno 1993). amplified archaeal rDNA gave much stronger signal,
A matrix of calculated similarity percentages was the weak hybridization shown by the bacterial rDNA
analyzed by UPGMA. Pairwise comparisons of the indicates a relatively small representation, particularly
band patterns were manually performed, and a matrix considering that a part could be due to hybridization
(presence-absence for each character) constructed. The of highly conserved 16S rDNA regions to the archaeal
24
Table 1. Number of digested fragments of total 16S rONA PCR-amplified from environmental
samples using bacterial and archaeal primer sets. Clone HAC I was obtained from the Crystallizer
Pond 30 in a previous study.
Table 2. Similarity values for 16S rRNA gene sequences of halophilic Archaea.
2 3 4 5 6 7 8 9 10 II 12 13 14
I. CLONE HACI 100 99.0 91.1 90.4 87.3 87.7 87.1 87.1 86.5 86.3 87.6 86.7 87.5 86.7
2. CLONE HAC14 100 90.4 90.0 87.1 87.4 87.2 87.0 90.7 85.4 87.4 86.7 87.1 86.5
3. H. volcanii 100 98.5 87.8 88.6 90.1 91.2 88.8 89.6 88.4 88.1 88.9 89.7
4. H. mediterranei 100 89.3 88.3 88.2 89.5 88.6 87.3 87.8 87.5 90.1 88.2
5. H. sinaiiensis M 100 97.7 91.4 88.8 87.0 86.7 87.0 87.2 86.1 86.1
6. H. marismortui rrnb 100 94.9 89.4 87.6 88.7 87.7 88.2 87.1 87.0
7. H. marismortui rrna 100 90.1 90.0 90.0 90.1 88.8 87.6 87.4
8. N. magadii 100 90.1 89.4 90.5 87.6 93.0 88.0
9. H. cutirubrum 100 88.3 99.8 87.8 90.0 88.5
10. H. morrhuae 100 88.4 86.3 90.1 88.6
II. H. salinarium 100 87.4 90.1 88.6
12. H. sodomense 100 87.4 94.7
13. N. occultus 100 87.9
14. H. saccharovorum 100
DNA. In any case considering the salt concentration bases were carried out and the highest sequence sim-
in the pond (30,8%) not many Bacteria would be ilarity value was found to be that with Rhodopseu-
physiologically active (Oren, 1990; Rodrfguez-Valera, dornonas marina (82.6% and 83.6%, clones HBCl and
1986). Previous studies using differential inhibitors HBC8 respectively). Summarizing, the bacterial DNA
for halophilic Archaea showed that in the crystallizer present in the crystallizer belonged mainly to a single
ponds most (if not all) the active biomass was archaeal group (possibly a single genus) distantly related to the
(Oren, 1990). a-Proteobacteria.
Five clones of the amplified eubacterial16S rDNA A total of 12 archaeal 16S rDNA clones from
from the crystallizer pond were randomly selected and the crystallizer were randomly selected and sequences
sequence analysis of ca. 200 bp (ranging from 283 of ca. 200 bp (positions 100-300; H. salinari-
to 487; E. coli numbering), revealed they belong to urn numbering) were determined. Analysis of data
a highly homogeneous phylogenetic group. Clones showed the twelve clones have identical sequence
HBCI, HBC5 and HBC7 showed identical sequences at these positions. The above cloned-fragments were
(HBC for Halophilic Bacterial Clone). Clones HBC2 obtained from several amplification reactions and
and HBC8 showed differences with HBCl in only cloning experiments. Some of these selected archaeal
2 and 14 nucleotide positions respectively. Sequence clones were subjected to analysis of another variable
comparisons with all16S rRNAs available in the data region of the 16S rDNA sequence (positions 384-
25
Total DNA from: HACI and HAC14 and other haloarchaea (as shown in
crvstallizer 1 Iwl2S CRJO Table 2) ranged from ca. 86% (Halococcus morrhuae)
to 91 % (viz. Haloferax volcanii, Haloferax mediter-
100ng 50ng 25 ng 50ng 50ng
ranei) . The dendrogram of Figure 2 shows that both
PROBE clones are, although relatively related, as distant from
members of the genus Haloferax as the other genera
Archaea are from one another. These results revealed that clones
HACI and HAC14 clearly correspond to a group dis-
tinct from all known haloarchaea, at least at the genus
level. Signature sequences for the different genera
are found in the region between positions 151-200
(see Figure 3). In this region 16 positions differentiate
Haloarcula from Haloferax, 22 discriminate Halobac-
Bacteria terium from Haloferax, 11 separate Haloarcula from
Halococcus, and 17 Halococcus from Haloferax. This
region is also highly divergent in all the crystallizer
clones with at least 20 positions different to any of the
Figure 1. Southern blot showing hybridizations of different amounts other genera. An inspection in detail of the complete
of total DNA from a crystallizer pond using archaeal and eubacte- 16S rDNA sequences of clones HACI and HAC14
rial amplified 16S rDNA products as probes. DNA extracted from
archaeal and eubacterial isolates (CRIO and lIWI25 respectively) revealed different nucleotide composition at 21 posi-
were used as controls (Benlloch et aI., 1995). tions which are conserved in all haloarchaea examined.
An example is illustrated in Figure 3; at position 196
all sequences exhibited thymine, but all selected clones
600; H. salinarium numbering). Clones HAC4, HAC8, showed cytosine. The considerably long branch in the
HACl1, HACI2, HAC14 and HAC16 had an identical dendrogram (see Figure 2), the genus-specific signa-
sequence at this stretch indicating that they may belong ture, and changes in relatively conserved positions of
to a single species (HAC for Halophilic Archaeal the sequenced molecule all strengthen the view that
Clone), However, this sequence showed 12 and 11 they belong to a new, undescribed genus of haloar-
nucleotide differences from those of HAC 1 and HAC6 chaea. In our study a high frequency of cloning 16S
respectively. Moreover, the complete 16S rDNA insert rDNA fragments with highly similar sequence (i.e.
of clones HACI and HAC14 was sequenced (1436 and with the same sequence at 100-300 region) has been
1438 nucleotides respectively ranging from positions 1 taken as a suspicion that members of this unknown
to 1458 of the H. salinarium 16S rRNA sequence genus are apparently numerous in the environment ana-
(Mankin et aI., 1985). The two sequences were highly lyzed. This is particularly noteworthy considering that
similar, although 15 nucleotide differences were found, three of the species and culture collection strains of
indicating that they correspond to related species of the haloarchaea (Haloferax mediterranei, Haloferax gib-
same genus. Similarities from the sequence compar- bonsii, Haloarcula hispanica) have been isolated from
isons with those of other Halobacteriaceae are shown the same saltern studied here. The results of the present
in Table 2 and Figure 2. As described by Kameku- study are reminiscent of that of Ward et al. (1990)
ra and Seno (1993) it has been shown that in gen- where a well known community, the Octopus Spring
eral the accepted taxonomy of haloarchaea fits well cyanobacterial mat was studied by 16S rDNA direct
with the phylogenetic relationships derived from the amplification. Like our results, 16S rDNA clones were
known 16S rDNA sequences. Sequence similarities markedly different from the 16S rDNA sequences of
between genera range from 86 to 91 % while species the organisms commonly isolated from this spring and
inside each genus are normally over 98% homology. considered to be the main components of the popu-
An exception is the species Halobacterium saccha- lation - Chloroflexus auranticus and Synechococcus
rovorum which should be included in a new genus, lividus (highest similarity around 90%).
as has already been suggested (Grant & Ross, 1986). DNA of four strains of haloarchaea isolated from
As shown in Figure 2, H. sodomense does not belong the same sample was extracted and a part of their 16S
to the genus Halobacterium, but is instead related to rDNA sequenced. All the strains isolated had the same
H. saccharovorum. The similarities between the clones sequence in the region 130-270 (H. salinarium num-
26
84 88 92 96 100
%
CLONE 1
I CLONE 14
L
-------C===
Haloferax volcanii
Haloferax mediterranei
r Haloarcula sinaiiensis
r-----[===========
r---------I Haloarcliia marismortui rrnb
l ___________ Haloarcliia marismortlli rrna
......
Natronobacterillm magadii
Natronococclis OCCllitliS
Halobacterium cutirubnlm
-[========
Halobacterium salinarillm
' - - - - - - - - - - - - - - - - - - - Halococcus morrhllae
L___________ Halobacterillm sodomense
Halobacterillm saccharovorum
Figure 2. Dendrogram of 16S rONA sequence similarities from clones HACI (Clone I), HACI4 (Clone 14) and members of Halobacteriaceae
(8enlloch et al., 1995).
151 200
Natronococcus occultus
Natronobacterium magadii
Halobacterium salinarium
Y12
Halobacterium cutirubrum
Haloarcula sinaiiensis M
Haloarcula marismortui rrnB
Ha/oarcula sinaiiensis m
Haloarcula marismortui rrnA
Halococcus morrhuae-16008
Halococcus morrhuae
Halobacterium sodomense
Haloferax gibbonsii
Haloferax denitrificans
Haloferax volcanii
Haloferax mediterranei
Clone 1 HA
Clone 14 HA CGCCCCGG
•
Figure 3. Alignment of a 16S rONA region from members of Halobacteriaceae and cloned fragments. The boxed areas are nucleotide
signatures that differentiate at the genus level. The start showed a nucleotide position where all selected HAC clones have changed a conserved
nucleotide. Numbers indicate positions following Hb. salinarium numbering (19). Clone I, identical sequence for all selected HAC clones.
YI2, Halobacterium sp. (14) (8enlloch et aI., 1996).
bering) and showed the highest homology (97%) to frequently. The specific discriminatory power of the
Haloarcula marismortui belonging very probably to a digestion enzymes was tested by computing against
closely related species. the data from all species of the genus Aeromonas, the
Enzymes Alul, Hinjl and Mbol were selected after 16S rDNAs of which show only 1 to 33 nucleotide
a computer analysis of many restriction endonucleas- differences (Martinez-Murcia et aI., 1992). As a
es that are known to cleave the 16S rDNA sequences conclusion of this test, enzymes such as Acil cut-
27
- -
fied from cultures of Escherichia coli ATCC 25922,
Pasteurella multocida, ATCC 43137 T , Listeria mono-
cytogenes ATCC 15313 T , Proteus mirabilis ATCC
29906T , Salmonella sp. ATCC 35664, Aeromonas
salmonicida NCIMB 1102 T , then a mixture of all the
amplified 16S rDNAs were digested with HinjI. All
the bands detected for every individual reference strain
could be detected as well in the digested mixture (data
•
not shown).
Bacterial and archaeal16S rDNA fragments ampli-
fied from the saltern waters were separately digest-
ed using three different enzyme combinations: HinjI,
Mbol + HinjI and Alul + HinjI. The products were sep-
arated by polyacrylamide gel electrophoresis (see Fig-
ure 4), the numbers of detected bands for each sample
are listed in Table I. In the case of Bacteria it is clear
that the total number of bands decreases at increasing
salinities in all the enzyme/enzyme combinations used.
For this phylogenetic group the environment becomes
more extreme at increasing salinities (over 1M NaCl),
therefore a general decrease in biodiversity is to be
expected (Rodrfguez-Valera et aI., 1981). On the oth-
er hand, the archaeal16S rDNA gave more restriction
fragments at higher salinities. Halophilic Archaea have
generally very high growth optima (over 20% NaCl)
and most of the species do not grow below 10% NaCl
(Larsen & Grant, 1989), therefore for this phylogenet-
ic group the low salinity conditions can be considered
an environmental extreme, and thus their decrease in
diversity.
The patterns of each sample were compared by
identifying fragments of identical size in the same
digestion from different samples. In this way a similar- Figure 4. Polyacrylamide gel showing Hinfl + MboI digestions of
(A) archaeal 16S rONA amplified from: lane I, Concentrator pond
ity coefficient was obtained for every pair of samples (CO)I08; 2, COlli; 3, Crystallizer Pond (CR)30; 4, clone HACI
and a similarity dendrogram generated. The dendro- (see reference 31); 5, isolate 9 from CR30; 6, isolate 10 from CR30;
gram shows the similarity level regarding archaeal or 7, molecular marker V (pBR 322 ONA.Hae III; Boehringer Minion);
bacterial populations in the systems studied. Analy- and (B) bacterial 16S rONA from: lane I, CR30; 2, COlli; 3,
CO 108; 4, Preparation pond (PR) 6; 5, PR3 (Martinez-Murcia et al.,
sis of bacterial patterns using HinjI (see Figure 5A) 1995).
revealed that 16S rDNA sequences from two concen-
28
-y
A A
Hinfl
Hinfl
CR30 r-----------lr- CR30
_____---; L - COIll
[ COIl!
~
' - - - - - - - - - - COIOS
COlli
CO! OS e
PRG Mho! + Hinfl , - - - - - - - - CR30
PR3
.---------i'----_ _ _ _ _ COIJ 1
e
Mhol + Hinfl ' - - - - - - - - - - - - - - co\Os
CR30
~
COlli
COIOS
i
PIl6 Figure 6. Dendrogram of archaeal 16S rDNA-RFLP similari-
I'R3 ties obtained by digestion with HinfI (A), AluI + HinfI (8), and
Mbo! + HinfJ (e) (Martinez-Murcia et aI., 1995).
,
0.0
,
0.2 0;4 0;6
,
0." I.",
Figure 5. Dendrogram of bacterial 16S rDNA-RFLP similari-
ties obtained by digestion with Hinf! (A), AluI + Hinjl (8), and
MboI+HinfI (e) (Martinez-Murcia et aI., 1995). seems to be present at COlD8 (13.3%) that is close to
the lower salinity tolerated by most halophilic Archaea
(about 10%). Probably a restricted and physiologically
trator samples (COllI and COlD8) were more similar specialized group of halophilic Archaea is present here.
to each other than to those from other water samples. Our results illustrate the different salt tolerance that
In the same way, 16S rDNA RFLP from the prepara- characterizes halophilic Bacteria and Archaea.
tion ponds were relatively very similar to each other. One possible application of the technique described
The sample from the crystallizer pond (CR30) not only here is to check the presence of an organisms charac-
showed lower diversity but was also less related to the terized only by sequence, i.e. non-cultivated, in a given
16S rDNA restriction patterns from the concentrators environment. The clone designated HACl was subject-
and preparation ponds. The same relationships were ed to l6S rDNA restriction digestion, together with the
found in the diagram of Figure 5B (using enzymes amplicon of the crystallizer and with some halophilic
Alul + HinjI) and Figure 5C (Mbol + HinjI). The rela- archaeal collection strains of the genera Haloferax,
tionships of archaeal 16S rDNA RFLP patterns were Haloarcula, Halobacterium, Halococcus, and two iso-
different from those of Bacteria. The sequences ampli- lates from the crystallizer CR30 as a reference (results
fied from the concentrator COllI were more relat- not shown). The high intensity restriction bands of
ed to those from the crystallizer CR30 than to those clone HAC I were of the same size as the major bands of
from COlD8 (Figure 6A; HinjI). The same relation- the crystallizer. However, profiles from clone HACl,
ships were found when Alul + HinjI or MboI + HinjI and reference strains were different from each oth-
were used (Figures 6B and 6C). er. The sample collected in which clone HACl was
With regard to Archaea the situation is somewhat obtained (Benlloch et a!., 1995) came from the same
reversed. The crystallizer and pond COllI (2l.6% crystallizer pond but it was almost a year before the
salinity) had similar archaeal populations. Again this sample was analyzed by digestion of 16SrRNA genes.
range (20-30%) corresponds to the normal range of Therefore, our results support the hypothesis (Benl-
growth of halophilic Archaea. A different population loch et aI., 1995) that the group of clones characterized
29
by 16S rDNA sequencing are predominant in the crys- et aI., 1990a). Those findings seem to indicate a more
tallizer and, so far, have not been isolated. moderate diversity.
A concise method for assessing microbiological
diversity, based on the digestion of the 16S rDNA
Discussion sequences directly amplified from a complex popu-
lation of microorganisms, is also reported (Martfnez-
The extreme environments that have been studied Murcia et aI., 1996). This methodology allows rapid
by the PCR technique show, as might be expected, comparison of the microbial populations present in
less diversity than sea water or soil, frequently hav- several samples since more than 30 samples can be
ing a higher redundancy in the sequenced 16S rDNA run in a single gel. The technique may have sever-
clones. However, as in high diversity environments al applications, for example, scanning serial samples
high homologies to cultured organisms are very sel- in a gradient in order to identify major boundaries
dom found. The extremely hypersaline waters of the in terms of microbial biodiversity, or to estimate the
crystallizer studied here show indeed even less diversi- contribution of some isolates or 16S rDNA clones to
ty by the direct rDNA amplification methodology than the total microbial popUlation. The application of this
in a previous work by culture isolation (Rodrfguez- method to a well understood environmental gradient,
Valera et aI., 1985). However, the organisms retrieved that found in the mUlti-pond saltern, illustrates its reli-
by one technique are markedly different from those ability since previous knowledge obtained from more
obtained by the other, even when both techniques are classical microbial ecology approaches has been large-
applied to the same sample. Our data can be interpreted ly confirmed, independently of the enzyme or combi-
if we consider that the culture methods strongly bias nation used.
the results towards microorganisms that are a minority Although the techniques based on direct detection
in the environment. If this hypothesis is correct and of the 5S rRNA present in the sample are more direct
applicable to other environments it has a major impact (obviating the need for PCR amplification and restric-
on our understanding of microbial diversity. tion enzyme digest) (Hofle, 1992) they also have some
The results reported to date from soil, ocean and disadvantages when compared with the methodology
thermal environments have shown a remarkably low described here. First of all, the extraction and manip-
level of homology to 16S rRNA sequences from cul- ulation of RNA is time-consuming and the sensitivity
tured microorganisms present in the data bases. In fact, to degradation also introduces the risk of not retrieving
homologies over 92% are rarely found and very often the less represented species. Also 5S rRNAs are very
values around 85% are the highest scores for a given small and the size differences do not always reflect
sequence comparison, this is within the range of differ- the variety of organisms present. We consider that the
ent phylum homology (Liesack& Stackebrandt, 1992). number of bands alone is an acceptable estimate of the
This has generally been interpreted as a reflection of total biodiversity present, particularly when different
the extremely high real bacterial diversity, orders of enzyme combinations are used. Of course, we cannot
magnitude higher than that shown by culture and iso- disprove that more than one sequence could be con-
lation, i.e. barely 3000 cultured bacterial species that tained in a single band or that incidentally one single
are known and have 16S rRNA genes sequenced. If organism could give a disproportionately large or small
the real number of species is several orders of mag- number of bands leading to an inaccurate evaluation
nitude greater, the probability of retrieving any of the of the real diversity present. However, both events are
known bacteria is relatively small unless a very large sufficiently infrequent to allow statistically significant
number of sequences from the natural environment evaluations by means of the method described here.
is screened. On the other hand, some data indicate Another possible source of error may be incomplete
less diversity. Most studies have detected consider- digestion. To avoid this problem it is essential to adjust
able redundancy in the sequences retrieved from the the ratio DNA/enzyme accurately and always digest a
environment, very often two representatives of each DNA control of which the restriction pattern is known;
sequence are found and sometimes five or more rep- incubation times should always be very long to reduce
etitions have been recorded. Identical sequences have this possibility.
been obtained from samples from different oceans and It is remarkable that the topology of the den-
by different authors (Fuhrman et aI., 1993; Giovannoni drograms is similar in all cases irrespective of the
enzyme/enzyme combination used. This is a strong
30
Javor, B. J., 1983b. Nutrients and ecology of the Western Salt and from a microbial mat at an active, hydrothermal vent system,
Exportadora de Sal saltern brines. 6th International Symposium Loihi Seamount, Hawaii. Appl. Envir. Microbiol. 60: 871-879.
on Salt. Vol. I. Salt Institute. Muyzer, G., E. C. De Waal & A. G. Uitterlinden, 1993. Profiling of
Kamekura, M. & Y. Seno, 1993. Partial sequence of the gene for a complex microbial populations by denaturing gradient gel elec-
serine protease from a halophilic archaeum Hal(}terax mediter- trophoresis analysis of Polymerase Chain Reaction-Amplified
ranei R4, and nucleotide sequences of 16S rRNA encoding genes genes coding for 16S rRNA. Appl. Envir. Microbiol. 59: 695-
from several halophilic archaea. Experientia 49: 503-513. 700.
Kushner, D. J. & M. Kamekura, 1988. Physiology of halophilic Oren, A., 1990. Estimation of the contribution of halobacteria to the
eubacteria. In: Rodriguez-Valera, F (ed.), Halophilic Bacteria, bacterial biomass and activity in solar salterns by the use of bile
Volume I. CRC Press, Inc. Boca Raton, Florida: 109-138. salts. FEMS Microbiol. Ecol. 73: 41-48.
Lane, D. J., B. Pace, G. J. Olsen, D. A. Stahl, M. L. Sogin & Pace, N. R., D. A. Stahl, D. J. Lane & G. J. Olsen, 1986. The analysis
N. R. Pace, 1985. Rapid determination of 16S rRNA sequences of natural microbial populations by ribosomal RNA sequences.
for phylogenetic analysis. Proc. Natl. Acad. Sci. USA. 82: 6955- Adv. Microb. Ecol. 9: I-55.
6959. Ralph, D., M. McClelland, J. Welsh, G. Baranton & P. Perolat, 1993.
Larsen, H. & W. D. Grant, 1989. Genus I: Halobacterium. Elazari- Leptospira species categorized by arbitrarily primed polymerase
Volcani 1957, 207 AL , p. 2219. In J. T. Staley, M. P. Bryant, chain reaction (PCR) and by mapped restriction polymorphisms
N. Pfennig & J. G. Holt (eds), Bergey's Manual of Systematic in PCR-amplified rRNA genes. J. Bacteriol. 175: 973-981.
Bacteriology, Vol. 3. The Williams & Wilkins Co., Baltimore. Reysenbach, A-L., G. S. Wickham & N. R. Pace, 1994. Phylogenet-
Liesack, W. & E. Stackebrandt, 1992. Occurrence of novel groups of ic analysis of the hyperthermophilic pink filament community in
the Domain Bacteria as revealed by analysis of genetic material Octopus Spring, Yellowstone National Park. Appl. Envir. Micro-
isolated from an Australian Terrestrial Environment. J. Bacteriol. bioI. 60: 2113-2119.
174: 5072-5078. Rodriguez-Valera, F, F Ruiz-Berraquero & A. Ramos-Cormenzana,
Maniatis, T., E. F Fritsch & J. Sambrook, 1982. Molecular cloning: 1981. Characteristics of the heterotrophic bacterial populations
a laboratory manual. Cold Spring Harbor Laboratory Press, Cold in hypersaline environments of different salt concentrations.
Spring Harbor, N.Y. Microb. Ecol. 7: 235-243.
Mankin, A. S., V. K. Kagramanova, N. L. Teterina, P. M. Rubtsov, Rodriguez-Valera, F, 1986. The ecology and taxonomy of aero-
E. N. Belova, A. M. Kopylov, L. A. Baratova & A. A. Gogdanov, bic chemoorganotrophic halophilic eubacteria. FEMS Microbiol.
1985. The nucleotide sequence of the gene coding for the 16S Rev. 39: 17-22.
rRNA from the archaebacterium Halobacterium halobium. Gene Rodriguez-Valera, F, A. Ventosa, G. Juez & J. F Imhoff, 1985.
37: 181-189. Variation of environmental features and microbial populations
Marchuk, D., M. Drumm, A. Saulino & F S. Collins, 1990. Con- with salt concentrations in a mUlti-pond saltern. Microb. Ecol.
struction of T-vectors, a rapid and general system for direct II: 107-115.
cloning of unmodified PCR products. Nucl. Acids. Res. 19: 1154. Sanger, F, S. Nicklen & A. R. Coulson., 1977. DNA sequencing
Martinez-Murcia, A. J., S. Benlloch & M. D. Collins, 1992. Phylo- with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA.
genetic interrelationships of members of the genera Aeromonas 74: 5463-5467.
and Plesiomonas as determined by 16S rDNA sequencing: lack Schmidt, T. M., E. F DeLong & N. R. Pace, 1991. Analysis of a
of congruence with results of DNA-DNA hybridization. Int. J. marine picoplankton community by 16S rRNA gene cloning and
Syst. Bact. 42: 412-421. sequencing. J. Bacteriol. 173: 4371-4378.
Martinez-Murcia, A. J., 1993. Doctoral Thesis. Department of Stackebrandt, E., W. Liesack & B. M. Goebel, 1993. Bacterial diver-
Microbiology, University of Reading, Reading, U.K. sity in a soil sample from a subtropical Australian environment
Martinez-Murcia, A. J. & F Rodriguez-Valera, 1994. The use of as determined by 16S rDNA analysis. The FASEB J. 7: 232-236.
arbitrarily primed PCR (APPCR) to develop taxa specific DNA Torsvik, v., S. Ktire, R. S0rheim & J. Goks0yr, 1990. Comparison
probes of known sequence. FEMS Microbiol. Lett. 124: 265-270. of phenotypic diversity and DNA heterogeneity in a population
Martinez-Murcia, A. J., S. G. Acinas & F Rodriguez-Valera, 1996. of soil bacteria. Appl. Envir. Microbiol. 56: 776-781.
Evaluation of prokaryotic diversity by restrictase digestion of 16S Walsby, A. E., 1980. A square bacterium. Nature 283: 69.
rDNA directly amplified from hypersaline environments. FEMS Ward, D. M., R. Weller & M. M. Bateson, 1990. 16S rRNA
Microbiol. Ecol. (in press). sequences reveal numerous uncultured microorganisms in a nat-
Mellado Duran, M. E., 1994. Bacterias halofilas moderadas: filoge- ural community. Nature 345: 63-65.
nia y construccion de vectores de clonacion. PhD Thesis, Uni- Weller, R .. J. W. Weller & D. M. Ward, 1991. 16S rRNA sequences of
versity of Sevilla. uncultivated hot spring cyanobacterial mat inhabitants retrieved
Moyer, C. L., F C. Dobbs & D. M. Karl, 1994. Estimation of diver- as randomly primed cDNA. Appl. Envir. Microbiol. 57: 1146-
sity and community structure through restriction fragment length liS!.
polymorphism distribution analysis of bacterial 16S rRN A genes
Hydrobio[ogia 329: 33-43, 1996. 33
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Key words: Coastal lagoons, eutrophication, phototrophic bacteria, purple nonsulfur bacteria, purple sulfur bacteria,
bacterial diversity
Abstract
Sediment samples collected from coastal lagoons on the French Mediterranean (Prevost Lagoon) and Atlantic
coasts (Arcachon Bay and Certes fishponds) have been studied in order to determine the population densities and
the species diversity of the different groups of anoxygenic phototrophic bacteria (purple sulfur bacteria, purple
nonsulfur bacteria and green sulfur bacteria) present in these ecosystems. Several strains of each group were isolated
in pure culture and characterized by their physiological properties. The occurrence of purple nonsulfur bacteria in
organic rich sediments of the Arcachon Bay and the dominance of purple sulfur bacteria in the Prevost lagoon and
Certes fishponds are discussed with respect to their community structure and abundance. The diversity differences
of the phototrophic bacterial strains isolated from both environments are also discussed.
The phototrophic bacterial population determined as as A. V. S. and amounts of total organic carbon were
colony forming units (CFU) in the 0 -2 cm layer of the much lower (Figures IA and 2).
sediment samples collected from each sampling site are Analyses of colonies of phototrophic bacteria
presented in Figure IA. Cell numbers increased from which grew in the agar shake tubes showed differences
March to June; from June to September they were sta- in the community structure between the sampling sites
ble in most of the stations, with the exception of station (Figure IB). Purple sulfur bacteria of the Chromati-
B where populations decreased to numbers similar to aceae were always dominant at the non-tidal stations
those recorded in March. The highest numbers, rang- (X, 11 and C), representing between 70 and 85% of the
ing from 5 x lO6 to 5 x 107 CFU cm -3, were found in total phototrophic community. Nevertheless, at these
the Certes Fishponds (station C) and the inner stations sampling sites, the population of purple nonsulfur bac-
of the Prevost Lagoon and Arcachon Bay (stations 11 teria increased from March to September, increasing
and B). These stations are also characterized by hight from 2-5% to 20-30%. In contrast, at station B, sub-
amounts of acid-volatile sulfide (A. V. S.) and total jected to tidal influence in the Arcachon Bay, purple
organic carbon as shown in Figure 2. In contrast, at the sulfur bacteria were less dominant and purple nonsul-
two sites directly subjected to tidal influences (stations fur bacteria represented a significant component of the
A and X), the numbers of phototrophic bacteria as well phototrophic bacterial community through the duration
of this investigation. At station B purple nonsulfur bac-
36
ro
U r--
20-30. J x
1
11
u 30-50. 3- +
'cro 2
~
'-"
....
0>
(A)
0 '"
I-<
(1)
0-5.
ro
......
0
E.....
5-10.
10-20. ~.
I- Q
20-30. +1
4 6 8 10 12 .....S
"0
30-50. +1
A B
(1)
tI)
AVS. (pmol.cm-3)
0-5. 01
Figure 2. Total organic carbon and acid volatile sulfide in the 0-2 cm
horizon of the sediment at the sampling stations in March 1993.
5-10.
10-20.
+
20-30. ~
30-50. c
teria composed up to 40% of the phototrophic bacterial Figure 3. Vertical distribution of phototrophic bacteria in the top
community. five em of the sampling sites in September 1993. Error bars indicate
standard error calculated from three samples.
In September 1993 the phototrophic bacterial po-
pulations were determined from zero to five cm depth
in the sediments (Figure 3). In the non-tidal stations,
the highest numbers were found in the upper 5 mm
of the sediment. Numbers decreased below this hori- by the genera Rhodobacter and/or Rhodovulum and
zon however phototrophic bacteria were still present Rhodopseudomonas and/or Rhodobium at station B.
at 5 cm depth, ranging from 105 (station X) to 107 From the three different ecosystems several strains
CFU cm- 3 (station C). At the tidal stations A and B, were isolated and identified according to their morpho-
the highest numbers were found in the 5-10 mm depth logical and physiological properties, in order to assess
horizon of the sediment. their taxonomic position and their metabolic capacities.
Data presented in Table 2 shows the major characteris-
Diversity o/phototrophic bacteria tics of some isolates. On the basis of their morphology
and carotenoids composition, several strains could be
Analyses of the data on the diversity of the phototro- classified with the genera Chromatium (Chr.), Thiocap-
phic bacteria in the samples collected in Septem- sa (Tea.), Thiocystis (Tcs.), Rhodobacter (Rba.) and/or
ber 1993 show that there were significant differences Rhodovulum (Rdv.), Rhodopseudomonas (Rps.) and/or
between the sampling sites (Table 1). Among the Chro- Rhodobium (Rbi.), Rhodospirillum (Rsp.) and Prosthe-
matiaceae, bacteria belonging to the genus Thiocapsa cochloris (Prs.).
were dominant (40-45%) at station 11 in the Prevost Data from the physiological tests (Tables 3 and 4)
Lagoon whereas in the Arcachon Bay and Certes Fish- enable several strains to be identified to the species
ponds, the motile purple sulfur bacteria of the genera level. Thus it was found that bacteria resembling Chr.
Chromatium and Thiocystis were dominant. Green sul- gracile and Tea. roseopersicina as well as bacteria
fur bacteria were detected only in the non-tidal stations, resembling the genera Rhodobacter and/or Rhodovu-
and were a minor component of the phototrophic bac- lum were found in all the lagoons. Tidal station A in the
terial community (0,5 to 4%). Among them, bacteria Arcachon Bay was characterized by isolates belonging
of the genus Prosthecochloris were dominant. A sig- to the species Chr. purpuratum and Rdv. suljidophilum,
nificant feature was the presence of substential num- whereas at station B the last species was the dominant
bers of purple nonsulfur bacteria, mainly represented member of the purple nonsuifur bacteria isolated.
37
Table 1. Composition of phototrophic bacterial communities obtained from the different sam-
pling sites in September. Water: Phototrophic bacterial numbers (CPU ml- 1) and percentages
are calculated from a single enumeration, sediments: Community composition are calculated
from three samples and from the average percentages of the different sediments depth (percent-
ages ± SE).
A B C II X
Water
Sediments
Community
(%)
PNSB 4.1 ± 1.9 42.7 ± 2.9 21.4 ± 9.1 25.7 ± 13.1 37.6±4.2
PSB 95.9 ± 1.9 57.3 ± 2.9 75.0 ± 10.9 72.5 ± 13.8 61.7 ± 3.9
Chr.fTes. 63.9 ± 11.2 26.7 ± 3.7 39.5 ± 11.6 29.1 ± 6.9 29.8 ± 3.5
Chr. (P) 10.7 ± 7.2 1204 ± lOA 10.2 ± 0.9 8.5 ± 704 3A± 2.3
Tea. 21.3 ± 12.9 18.2 ± 6.9 25.3 ± 3.2 34.9 ± 8.1 28.5 ± 3.2
Community: PNSB, purple nonsulfur bacteria; PSB, purple sulfur bacteria (Chr.fTcs., red
Chromatium sp. and/or Thillcystis sp.; Chr. (P), purple Chromatiaceae; Tea .• Thillcapsa sp.);
GSB, green sulfur bacteria; N.D., not detected.
Prevost lagoon
Arcachon Bay
Certes lagoon
and allow the phototrophic development of anoxygenic When the results obtained in this study at the five
phototrophic bacteria. Thus, the availability of organic different sampling sites are compared, some impor-
matter, anoxic conditions, reduced sulfur compounds tant features are evident. The inner stations of these
and light will control the numbers and activities of the ecosystems (stations 11, Band C) appear to have more
phototrophic bacteria (Caumette, 1992). favorable conditions for the development and main-
tenance of phototrophic bacteria. These stations are
39
Table 3. Physiological characteristics of some phototrophic sulfur bacteria isolated from Etang du
Prevost, Bassin d' Arcachon and Certes fishponds
Requirements
vitamins
NaCl (% w/v) 0.5 0 0.5 0.5 0.5 ND ND
N2-fixation + + + + + + + +
S04-red + + + + +
Oxygen m m an m m m an m
Optimum pH 7.1 7.2 7.2 7.2 7.0 7.15 7.15 7.2
Electron donors
Sulfide + + + + + + (+) +
Sulfite + + + (+)
Thiosulfate + + + + + + +
Sulfur + + + + + + + +
Formate +
Acetate + + + + + + + +
Propionate + + + + + + +
Butyrate + + + +
Valerate + + + +
Caprylate +
Lactate + + + + + +
Glycolate + + +
Pyruvate + + + + + + + +
Malate + + + + + (+) (+) +
Fumarate + + + + + + (+) +
Succinate + + + + + + (+) +
Citrate
Glucose
Fructose + + +
Methanol
Ethanol +
Propanol +
Glycerol + +
Yeast extract + + + + + + + +
Casamino- (+) + (+)
acids
Requirements
NaCI (% w/v) 0.5 0.5 0.5 0.5 0.5 NO NO NO
N2-fixation + + + + + + + +
S04-red + + + + + + + +
Oxygen ae m m ae ae ae ae ae
Optimum pH 6.6 6.8 6.7 6.8 6.7 6.8 6.8 6.8
Electron donors
Sulfide + + + + + + +
Thiosulfate + + + + +
Sulfur +
characterized by high levels of organic matter and have tions A and B) were less eutrophic and characterized
high acid volatile sulfide concentrations, indicating the by lower populations of phototrophic bacteria.
strong activity of sulfate-reducing bacteria (Bourgues Whilst similarities in the total population of pho-
et aI., 1994). In contrast, the two tidal stations (sta- to trophic bacteria were observed in all three lagoon
systems, major differences were recorded in the com-
41
munity structure. For example, green sulfur bacteria In the Certes Fishponds, the inner station C was
were only found in the sediments at the sampling sta- characterized by high organic matter and acid-volatile
tions C and 11. These bacteria do not tolerate oxygen sulfide levels as was station 11 in the Prevost Lagoon.
as the purple bacteria do, and therefore have never The main difference between these two sampling
been found in the eutrophic but tidal station B in the sites was their macrophyte composition (Auby et ai.,
Arcachon Bay. Phototrophic purple bacteria that are 1993) and the abundance of meiobenthos (Buffan &
commonly found in coastal marine sediments (Pfennig, Castel, 1993). At station C, Ruppia cirrhosa (Pet.)
1989) were also widely distributed at all the sampling Grande was the dominant component of the macro-
stations investigated in this study. Other phototrophic phyte biomass and maintained a relatively stable stand-
bacteria such as members of the Chlorofiexaceae or ing crop throughout the year. At station 11, as well as
the Heliobacteriaceae have not been detected in these station X, the principal primary producer was the alga
lagoon systems, even when enrichment cultures were Ulva sp. which could achieve biomass levels of 370 g
applied. m- 2 of dry mass. Station 11 was characterized by a
Among the purple bacteria, the purple nonsulfur sharp decrease in meiobenthos abundance during the
bacteria were well represented at all the sampling summer, whereas in the Certes Fishponds meioben-
sites, especially at station B. These bacteria are widely thos biomass increased significantly in summer due to
distributed in organic rich aquatic environments such the presence of transient meiofauna populations (poly-
as sewage contaminated waters (Siefert et ai., 1978; chaete larvae) which could reach densities greater than
Kobayashi, 1977) or lagoons subjected to excessive 3000 individuals cm- 2 (Buffan & Castel, 1993). This
pollution (Deveze & Fauvel, 1966; Caumette, 1987). meiofauna may play a significant role in the biotur-
At station B the sediments receive large quantities bation of the sediments, leading to the resuspension
of organic matter. The total biomass of the macro- of reduced sulfur compounds trapped in the deeper
phytes Monostroma obscurum and Zostera sp. at this sediment layers. These ecological considerations may
site may reach 260 g m- 2 of dry mass (Auby et ai., account for the observed differences in community
1993). Decomposition of this organic matter favor structure of phototrophic bacteria at the two sites.
high sulfate-reduction rates at this station (Finster et In the Certes Fishponds, numbers of phototrophic
ai., 1994) and the concomitant release of low molecu- bacteria were higher than those at the Prevost Lagoon
lar weight organic compounds which are metabolized and the community structure was more diverse, par-
by purple nonsulfur bacteria. The Rhodovulum and ticularly among the dominant purple sulfur bacte-
Rhodobium strains isolated from the Arcachon Bay ria which included Chromatium gracile, Chromatium
were found to be able to utilise a wide range of sim- buderi, Thiocystis sp. and Thiocapsa roseopersicina.
ple organic compounds (organic acids, sugars, sugar- In contrast, the community of purple sulfur bacteria in
alcohols, amino acids) but also sulfide. These isolates the Prevost Lagoon was strongly dominated by bacte-
are less sensitive to oxygen than purple sulfur bacte- ria resembling Thiocapsa roseopersicina, as was previ-
ria, and therefore dominated the phototrophic bacterial ously reported by Caumette (1986). Thiocapsa roseop-
community at station B because oxygen diffused down ersicina is a purple sulfur bacterium well adapted to the
into the sediments at low tide periods. fluctuating conditions of oxygen and sulfide (De Wit
At sampling station A, a tidal site dominated by the & van Gemerden, 1987, 1990; Schaub & Van Gemer-
macrophyte Zostera noltii Hornem. that forms beds den, 1994). Such fluctuating conditions occur period-
without decaying algae (Auby et ai., 1993), purple ically in the Prevost Lagoon. Macroalgae are rapidly
nonsulfur bacteria only composed a minor compo- degraded with a concomitant rapid consumption of dis-
nent of the phototrophic bacterial community. At this solved oxygen. Further degradation is carried out by
site, purple sulfur bacteria dominated the community. sulfate-reducing bacteria that form hydrogen sulfide in
Among them, several strains resembling Chromatium the entire water column, leading to dystrophic crises
purpuratum were isolated. This is a typical marine and (Amanieu et ai., 1975). During these crises, Thiocapsa
strictly anaerobic species isolated from marine sponges roseopersicina is the dominant phototrophic bacterium
(Imhoff & Trtiper, 1980) and marine sediments (Ma- that develops in the water column, producing red water
theron, 1972) (Table 3). This bacterium also formed and oxidizing the toxic sulfide (Caumette, 1986). In
purple layers on sandy sediments adjacent to station contrast, at station B which is exposed to air period-
A. ically, as well as at the non-tidal station C, the high
levels of iron in the sediments playa significant role
42
in trapping the hydrogen sulfide generated by sulfate Cerruti, A., 1938. Le condizioni oceanographiche et biologiche del
reducing bacteria (Stal et aI., 1994). mar piccolo di Taranto durante I' Agosto del 1938. Boll. Pesca.
Piscicolt. Idrobiol. 14: 711-751.
Cline, J. D., 1969. Spectrophotometric determination of hydrogen
sulfide in natural waters. Limnol. Oceanogr. 14: 454-458.
Acknowledgements Deveze, L. & Y. Fauvel, 1966. Un phenomene bacterien d'eaux
rouges dans l'etang d'Ingril (Herault). Rev. Travaux de l'Institut
des Peches Maritimes 30: 365-374.
This study was financed by a joint ED project on de Wit, R. & H. van Gemerden, 1987. Chemolithotrophic growth of
Coastal Lagoon Eutrophication and Anaerobic pro- the phototrophic purple sulfur bacterium Thiocapsa roseopersic-
cesses (CLEAN), Contract number EV5V - CT92 - ina. FEMS Microbiol. Ecol. 45: 117-126.
0080. de Wit, R. & H. van Gemerden, 1990. Growth and metabolism
of the purple sulfur bacterium Thiocapsa roseopersicina under
combined light/dark and oxic/anoxic regiments. Arch. Microbiol.
154: 459-464.
References Eichler, B. & N. Pfennig, 1986. Characterization of a new platelet-
forming purple sulfur bacterium, Amoebobacter pedil!/()rmis sp.
nov. Arch. Microbiol. 146: 295-300.
Amanieu, M., B. Baleux, O. Guelorget & P. Michel, 1975.
Finster, K., A. Gammelgaard & L. S. Hansen, 1994. Sulfate reduc-
Etude biologique et hydrobiologique d'une crise dystrophique
tion rates, inorganic sulfur pools and the accumulation of volatile
(Malaigue) dans l'etang du Prevost it Palavas (Herault). Vie
fatty acids in the Ba~sin of Arcachon and the Prevost lagoon,
Milieu, serie B 25: 175-204.
southern France: a comparative study. In Caumette, P. (coord.),
Auby, I., G. Bachelet & P. J. Labourg, 1993. Biomass and species
Proceedings of the CLEAN Meeting, part II, Ferrara, Italy: 223-
composition of macrophytes in Arcachon Bay and Prevost
252. European Community Environmental Programme, DG XII,
Lagoon, with a compilation of data on primary production in
Brussels.
Arcachon Bay. In Caumette, P. (coord.), Proceedings of the
Forti, A., 1933. II fenomeno del lago di sangue nello stagno di
CLEAN Meeting, part II, Alicante, Spain: 65-72. European
Pergusa in Sicilia. Nuovo Giorno Bot. Ital. 40: 76.
Community Environmental Programme, DG XII, Brussels.
Guerrero, R., C Pedros-Alio, I. Esteve & J. Ma~, 1987. Com-
Bourgues, S., O. Pringault & R. de Wit, 1994. Spatial and temporal
munities of phototrophic sulfur bacteria in lakes of the Spanish
variations of oxygen uptake, sulfide production and population
Mediterranean region. Acta Academiae Aboensis 47: 125-151.
of sulfur cycle bacteria at the CLEAN stations. In Caumette,
Guyoneaud, R., A. Hirschler, R. Matheron & P. Caumette, 1993.
P. (coord.), Proceedings of the CLEAN Meeting, part II, Fer-
Phototrophic bacterial communities in Arcachon bay and Prevost
rara, Italy: 109-132. European Community Environmental Pro-
lagoon. In Caumette, P. (coord.), Proceedings of the CLEAN
gramme, DG XII, Brussels.
Meeting, part II, Alicante, Spain: 341-355. European Communi-
Buffan, E. & J. Castel, 1993. Distribution of benthic microbionta
ty Environmental Programme, DG XII, Brussels.
in Arcachon Bay and Prevost Lagoon. In Caumette, P. (coord.),
Heldt, J. H., 1952. Les eaux rouges. Soc. Sci. Nat. Tunisie 5: 103-
Proceedings of the CLEAN Meeting, part II, Alicante, Spain: 73-
106.
89. European Community Environmental Programme, DG XII,
Imhoff, J. H. & H. G. Triiper, 1980. Chromatium purpuratum, sp.
Brussels.
nov., a new species ofthe Chromatiaceae. Zbl. Bakt., l. Abt. Orig.
Cahet, G., 1965. Contribution it I' etude des eaux et des sediments de
CI: 61-69.
Bages-Sigean (Aude). III. Reduction des composes soufres. Vie
Jlirgensen, B. B., 1977. The sulfur cycle of a coastal marine sediment
Milieu 16: 917-981.
(Limfjorden, Denmark). Limnol. Oceanogr. 22: 814-832.
Castel, J., P. Caumette & R. A. Herbert, 1996. Eutrophication gradi-
Jlirgensen, B. B. & T. Fenchel, 1974. The sulfur cycle of a marine
ents in coastal lagoons. Hydrobiologia 329: xi-xxx.
sediment model system. Mar. BioI. 24: 189-201.
Caumette, P., 1986. Phototrophic sulfur bacteria and sulfate-
Jlirgensen, B. B., N. P. Revsbech & Y. Cohen, 1983. Photosynthesis
reducing bacteria causing red waters in a shallow brackish coa~tal
and structure of benthic microbial mats: Microelectrode and SEM
lagoon (Prevost lagoon, France). FEMS Microbiol. Ecol. 38:
studies of four cyanobacterial communities. Limnol. Oceanogr.
113-124.
28: 1075-1093.
Caumette, P., 1987. Role des bacteries phototrophes et des bacteries
Kampf, C & N. Pfennig, 1980. Capacity of Chromatiaceae for
sulfato-reductrices dans les milieux lagunaires. O.R.S.T.O.M.
(ed.), Paris, 304 pp. chemotrophic growth. Specific respiration rates of Thiocystis vio-
Caumette, P., 1992. Bacterial communities in coastal lagoons. An
lacea and Chromatium vino.wm. Arch. Microbiol. 127: 125-135.
overview. Vie milieu 42: 111-123. Kobayashi, M., 1977. Utilization and disposal of wa~tes by photo-
synthetic bacteria. In H. G. Schlegel & J. Barnea (eds), Microbial
Caumette, P. & B. Baleux, 1980. Etude des eaux rouges dues it
energy conversion. Pergamon Press, Oxford: 443-453.
la proliferation des bacteries photosynthetiques sulfo-oxydantes
Lassen, C, R. de Wit, J. Lorentzen & N. P. Revsbech, 1993. Distri-
dans I' etang du Prevost, lagune saumiitre mediterraneenne. Mar.
bution of light, 02 and oxygenic photosynthesis. In Caumette, P.
BioI. 56: 183-194.
(coord.), Proceedings of the CLEAN Meeting, part II, Alicante,
Caumette, P., M. Pagano & L. Saint-Jean, 1983. Repartition verti-
Spain: 23-40. European Community Environmental Programme,
cale du phytoplancton des bacteries et du zooplancton dans un
DG XII, Brussels.
milieu stratifie en baie de Bietri (Lagune Ebrie, Cote d'Ivoire).
Matheron, R. & R. Baulaigue, 1972. Bacteries photosynthetiques
Hydrologia 106: 135-148.
sulfureuses marines. Assimilation des substances organiques et
Caumette, P., K. Schmidt, H. Biebl & N. Pfennig, 1985. Char-
mim!rales, et influence de la teneur en chlorure de sodium du
acterization of a Thiocapsa strain containing okenone as major
carotenoid. System. Appl. Microbiol. 6: 132-136. milieu de culture sur leur developpement. Arch. Mikrobiol. 86:
291-304.
43
Pfennig, N., 1989. Ecology of phototrophic purple and green sulfur Siefert, E., R. L. Irgens & N. Pfennig, 1978. Phototrophic purple and
bacteria. In H. G. Schlegel & B. Bowien (eds), Autotrophic Bac- green bacteria in a sewage treatment plant. Appl. Envir. Micro-
teria. Science Technique Publisher, Madison and Springer Verlag, bioI. 35: 38-44.
Berlin: 97-116. Stal, L., F. Kruying, S. B. Behrens & M. Villbrandt, 1994. Spatial
Pfennig, N. & S. Wagener, 1986. An improved method of prepar- and seasonal variations of iron, phosphate and sulfide in the sedi-
ing wet mounts for the photomicrography of microorganisms. J. ments of two eutrophic coastal lagoons. In Caumette, P. (coord.),
Microbiol. Meth. 4: 303-306. Proceedings of the CLEAN Meeting, part II, Ferrara, Italy: 167-
Pfennig, N. & H. G. Triiper, 1992. The family Chromatiaceae. In A. 186. European Community Environmental Programme, DG XII,
Balows, H. G. Triiper, M. Dworkin, W. Harder & K. H. Schleifer Brussels.
(eds), The Prokaryotes. Springer-Verlag, New-York: 3200-3221. Stirn, 1., 1971. Ecological consequences of marine pollution. Rev.
Pierson, B., A. Oesterle & G.L. Murphy, 1987. Pigments, light pen- Int. Oceanogr. medit. 24: 13-46.
etration, and photosynthetic activity in the multi-layered micro- van Gemerden, H., R. de Wit, C. S. Tughan & R. A. Herbert, 1989a.
bial mats of Great Sippewisset salt marsh, Massachussets. FEMS Development of mass blooms of Thiocapsa roseopersicina on
Microbiol. Ecol. 45: 365-376. sheltered beaches on the Orkney Islands. FEMS Microbiol. Ecol.
Schaub, B. E. M. & H. van Gemerden, 1994. Simultaneous pho- 62: Ill-liS.
totrophic and chemotrophic growth in the purple sulfur bacteri- van Gemerden, H., C. S. Tughan, R. de Wit & R. A. Herbert, 1989b.
um Thiocapsa I'Oseopersicina MI. FEMS Microbiol. Ecol. 13: Laminated microbial ecosystems on sheltered beaches in Scapa
185-196. Flow, Orkney Islands. FEMS Microbiol. Eco!. 62: 87-102.
Hydrobi%gia 329: 45-55, 1996. 45
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Abstract
Zooplankton variability in a lagoon of the Po Delta, the Sacca di Goro, was studied in relation to changes in
its hydrodynamic and hydrological features. From 1987 to 1992 the lagoon was affected each summer by severe
anoxia due to the decomposition of dense and widespread beds of Viva rigida. In August 1992 a canal was opened
through the sand bank closing off the lagoon from the sea in order to improve the water circulation. This hydraulic
intervention led to significant environmental changes in the lagoon: during the two subsequent years the summer
dystrophic crises were brief and less severe, due to a sharp decrease in the Viva cover. At the same time a clear
increase in phytoplankton biomass values was observed. Following the opening of the channel, the most remarkable
change in zooplankton was a significant density increase of calanoids, particularly of Acartia tonsa.
.' . ~
SACCA DI GORO
,
./ I
I
"
.... "
,',
........ ,'
I
. o 2 Km
I 1
... • '0,' • ,/
'
,
... I
i,
... "\
i
i
!
I
I
I
...
.. ,
FiRure 1, Sacca di Goro: sampling stations are indicated; the arrow shows the position of the sand bank cut.
Materials and methods identification was, when possible, made to the species
level for the copepods and cladocerans. For data pro-
Zooplankton samples were collected at two stations, cessing all identified taxonomic units were subdivided
St. 4 in the centre of the lagoon and St. 8 in the Valle di into 18 major groups: Noctiluca, tintinnids, Synchaeta,
Gorino (Figure 1), at monthly intervals from Novem- other rotifer species, polychaete larvae, gastropod lar-
ber 1987 to October 1989 and then from March to vae, bivalve larvae, cirriped larvae, decapod larvae,
December 1992, from April to November 1993 and cladocerans, copepod nauplii, copepod copepodids
from February to December 1994. Tidal phase vari- and adults (subdivided into five categories: freshwa-
ations were not the same for the different sampling ter copepods, brackish water copepods, euplankton-
dates. ic calanoids, euplanktonic cyclopoids, euplanktonic
Sampling was carried out with a 15 I Patalas trap: harpacticoids), tunicates and 'other groups'.
at least 150 I of water were taken from the entire water Canonical Correspondence Analysis (CCA) (Ter
column (maximum depth was about 2 m) for each Braak, 1986) was applied in data analysis. CCA allows
sample and then filtered through a 50 JIm mesh net. ordination of a set of samples, each defined by a list
A buffered formaldehyde solution (4%) was used as of species abundances, under the constraint that ordi-
fixative. nation axes are a linear combination of environmental
Simultaneously with zooplankton sampling, water variables measured for each sample. This multivariate
temperature, salinity, dissolved oxygen (Winkler technique assumes that species abundance response to
method, APHA, 1975) and phytoplankton chlorophyll- variables follows a unimodal, bell-shaped curve. CCA
a content (trichromatic method, APHA, 1975) were provides ordination of both samples and species points,
measured at both the surface and near bottom layers highlighting the importance of species in determining
of the water column (for 1994 data are available only sample composition.
from March to August). A simple graphical representation (biplot) can be
Zooplankton counting was carried out on subsam- drawn, where environmental variables are represented
pIes taken by a Hansen-Stempel pipette. Taxonomic by arrows in the ordination space. Then sample (and
47
species) point projections on the arrows are estimates ic episodes, following improvised emergency inter-
of the 'score' of each point with respect to environ- ventions on the sand bank aimed to favour a temporary
mental variables. In the case of species, this score can inflow of seawater into the lagoon (Ferrari et aI., 1993).
be thought of as species performance estimates along The increase of phytoplankton chlorophyll-a coincided
the corresponding environmental gradient. with the reduction of Viva biomass. Other phenome-
All ordination techniques are by their own nature na closely related to the decreased spreading of this
explorative, thus a significance test of the results has nitrophilous macro alga have been detailed by Viaroli
a purely indicative value. In the case of CCA, a cru- et al. (1993b) and Bartoli & Viaroli (pers. comm.): dur-
cial issue is that the species data set is fully or part- ing 1993 and 1994 a shorter phase of spring depletion
ly explained by the environmental variables consid- of dissolved inorganic nitrogen and a shorter duration
ered. To assess the significance of ordination axes of summer anoxia in a narrow easternmost area of the
for explorative purposes, a Monte Carlo permutation Valle di Gorino were observed. In addition, experi-
test is usually performed: in each one of a number of ments carried out in 1993 and 1994 with the dark and
CCAs, the order of environmental variable values is light bottle method and dark core incubation showed,
randomly permuted. From the distribution of the ordi- in comparison with the previous years, a significant
nation statistic chosen (usually the eigenvalue), one reduction of the metabolic acti vity (production and res-
can estimate the probability that environmental vari- piration) of the macroalgal compartment compared to
ables would have no significant effect on the species the planktonic compartment as well as a decrease in
data set, by computing the percentage p of runs yield- the sediment oxygen demand.
ing statistics equal to or greater than that observed. Zooplankton abundance and composition also
A CCA of the Sacca di Goro data was performed underwent significant changes. Until 1992 total zoo-
using abundances of the 18 taxonomic groups, after plankton density at St. 8 was characterized by sudden,
a log( 1 + x) transformation. Environmental variables very high peaks, generally linked to phytoplankton
were: temperature, salinity, dissolved oxygen and phy- chlorophyll-a maxima. In the subsequent years total
toplankton chlorophyll-a content: data recorded at the density had more reduced fluctuations; furthermore,
surface layer of the water column were used. The its seasonal trend showed less pronounced differences
excessive sensitivity of CCA to rare species was com- from St. 4 (Figure 2).
pensated by using down weighting as suggested by The copepod nauplii were, on the average, the dom-
Hill (1979). CCA was performed by means of the inant group, with the highest abundance peak in the
CANOCO program (Ter Braak, 1988), running on a Valle di Gorino (1299 ind 1-1 in July 1992); they fol-
IBM-compatible Pc. Separate CCAs were conducted lowed the same general trend observed for total zoo-
for the two sampling stations, with Monte Carlo signif- plankton (Figure 2). The species belonging to the five
icance tests using first eigenvalues extracted and 1000 categories of copepodids and adults of copepods are
permutations each time. The biplot diagrams of envi- listed in Table 1. Freshwater copepods occurred with a
ronmental variables and taxa scores are represented very low frequency. The brackish water fraction, main-
separately for the periods before and after the cut of ly represented by epibenthic and phytal harpacticoids,
the sand-bank; in other diagrams sample scores for the showed larger abundances at St. 8 (16 and 17 ind 1-1
entire period of investigation are given separately for in July 1988 and August 1992, respectively) during
each year. the decomposition phase of macroalgal beds; at the
same station, a decreasing trend was observed after
August 1992: maximum density was 9 ind 1-1 in 1993
Results (May) and 2 ind I-I in 1994 (December). Abundances
of neritic euplanktonic cyclopoids (chiefly belonging
The change in the lagoon hydrological conditions after to the genus Oithona) and harpacticoids did not vary
the hydraulic intervention in August 1992 was reflect- appreciably: density maxima were 7-8 and 5 ind 1-1,
ed, first of all, by higher values of phytoplankton respectively, both before and after August 1992.
chlorophyll-a content at St. 8 (Figure 2). In previ- From 1987 to 1989 calanoids were characterized
ous years, particularly in 1992, chlorophyll-a con- by relatively high abundance values during winter
tent of the water ranged around values less than 5 Ji,g months, mainly sustained by neritic forms, in particu-
1-1 through the entire growing season: isolated phyto- lar Acartia clausi Giesbrecht. In 1993 and 1994 they
plankton blooms occurred only after the summer anox- showed sharply higher densities than in previous years
48
ChIoropb)'U. II
f
~
60
-. --
51. • ~
SI. • .
II
1\
I, 40
I'
I ,
I I 20
Total zooplallktoa
2000
1500
1000
'00
Copepod aaupJII
2000
1500 1500
1000 1000
500
Figure 2. Sacca di Goro: phytoplankton chlorophyll-a concentration at the surface layer, density of total zooplankton and copepod nauplii at
St. 4 and St. 8 from 1987 to 1989 and from 1992 to 1994.
at both stations (Figure 3): this increased abundance peaks of calanoid density in September 1989 (75 ind
was due almost exclusively to Acartia tansa Dana, a I-I at StA) and then in July (189 ind I-I at St. 4)
species typical of highly eutrophic coastal areas and and August 1994 (198 ind 1-1 at St. 8) were mostly
harbour environments (Brylinski, 1981). A. tansa was supported by A. tansa. This led to the progressive dis-
found for the first time in the Mediterranean Sea by appearance of the congeneric species Acartia margalefi
Gaudy & Vinas (1985) in a great brackish water pond Alcaraz and Acartia latisetasa Kriczaguin, which are
near Marseille and for the first time in the Adriatic Sea typical of the most confined areas of the lagoons of the
by Farabegoli et al. (1989) in the lagoons of the Po Adriatic Sea: after 1989 no individuals belonging to
River Delta. Over the period under investigation, A. these two species were found. The congeneric marine
tansa gradually came to be the dominant species of the species A. clausi was found in most samples, especial-
plankton copepod component of the Sacca di Goro: the ly at St. 4, but its abundance gradually decreased both
49
Synchaela
ind'I- 1 51.4 -0--
SOO SOO 5L. - -. --
400 400
300 300
200 200
100 100
0 0
Gastropod larvae
ind·I- 1
90 90
'II
60 60
,
"
I
I I
I
I I
30 30 I I
,, 1,
I I
Calanoids
200
~
150
'I
II
I
100 I
I
I
SO I
~
~\ 1\ .... ~\
0 r/!:i!-~,~ I ,~,k I I I I ,~,
....
NPI 'MAMI IAIONPI'MANI IAIO
>0
MANIlA .ONPI 'MAMI IA 10 NPI .NANI I ASONP
00
00
Figure 3. Sacca di Goro: density of Synchaeta, gastropod larvae and calanoids (copepodids and adults) at St. 4 and St. 8 from 1987 to 1989 and
from 1992 to 1994.
in the Valle di Gorino and in the centre of the lagoon, Among meroplankton, decapod larvae exceeded
until it represented only a slight fraction of the Acartia 2 ind 1-1 only at St. 8 from May to July 1993 and
complex density in 1994. in June 1994; from 1987 to 1994 they were signif-
A clear increase in the Valle di Gorino was observed icantly increasing at St. 4 too. Polychaete, bivalve
for marine cladocerans as well: from 1987 to 1992 at and cirriped larvae showed conspicuous densities at
St. 8 they never exceeded 1 ind 1-1, whereas they both stations, generally from March to September,
showed a peak of 11 ind 1-1 in both September 1993 throughout the research period. Density maxima for
and September 1994. The highest density value (37 ind these three taxa were found at St. 4: 417 ind 1-1 in
1-1) was recorded in June 1994 at St. 4. The most August 1989 for polychaete larvae, 196 ind 1-1 in July
important species was Podon polyphemoides Leuck. 1988 for bivalve larvae and 35 ind 1-1 in July 1989 for
cirriped larvae. The high densities of bivalve larvae
50
Table 1. Species of copepods occurring in the zooplankton
samples collected in the Sacca di Goro.
can be related to the presence in the southern area of
the lagoon of extensive banks of the cultivated clam
Freshwaters cope pods Tapes philippinarum Adams & Reeve. The consider-
Acanthocyclops sp. able cirriped larvae abundance seems to be dependent
Brackish copepods on the large number of wood and cement artificial sub-
Calanipeda aquaedulcis Kritz strata widespread in all the lagoon. Gastropod larvae,
Halicyclops rotundipes Kiefer probably belonging to the genus Hydrobia, a small
Halicyciops sp. snail closely associated with the Ulva thalli, at St. 8
Cyclopina gracilis Claus decreased sharply from the peak of84 ind 1-1 in August
Cyciopina sp. 1992 to maximum densities of 8 to 9 ind 1-1 in 1993
Ergasilidae
and 1994 (Figure 3).
Parategastes sphaericus Claus
A decreasing trend for Synchaeta was evidenced at
Canuella perplexa Scott
both the stations: the peaks observed in 1988 (222 ind
Ecfinosol1U1 dentatum Steuer
1-1 at St. 4 and 165 ind I-I at St. 8 in July) and
Ecfinosoma sp.
1989 (485 ind 1-1 at St. 4 and 141 ind 1-1 at St. 8
Horsiella sp.
Microarthridion ./allax Perkins
in March) progressively decreased; in 1994 a remark-
Harpacticus fiexulosus Ceccherelli ably lower density was noted (Figure 3). Other rotifer
Harpacficus fiexus Brady & Robertson species rarely occurred, although in August 1992 iso-
Tisbe holoturiae Humes lated peaks (255 ind 1-1 at St. 4 and 43 ind 1-1 at St. 8),
Tisbe sp. mainly sustained by Brachionus plicatilis O.E Milller,
Amphiascus parvus Sars were recorded.
Amonardia normani Brady Noctiluca and tintinnids were generally found in
Schizopera compacta De Lint low density, except for a tintinnid abundance peak of
Diosaccopsis rubeus Blian 260 ind I-I at St. 8 in August 1992.
Ameira parvula Claus Both the occurrence frequency and abundance of
Nitocra typica Boeck tunicates (appendicularians), a typical marine group
Mesochra pygmaea Claus generally associated with neritic species of copepods,
Asellopsis sarmantica lakubisiak were quite low throughout the research period; at St.
Robertgurneya simi/is A. Scott
4, however, their frequency and abundance (with a
Euplanktonic calanoids
maximum of 6 ind I-I in September 1992) appeared
Calanus helgolandicus Claus
more relevant than at St. 8, where tunicates were found
Paracalus parvus Claus
only in three samples with a maximum of 1 ind 1-1 in
Pseudocalanus elongatus Boeck
September 1993.
Clausocalanus sp.
Temora longicornis Milller
The main results of CCA are presented in the
Temora stylitera Dana
biplot diagrams of the environmental variables and
Centropages ponticus Karavaev taxa scores for St. 4 and St. 8, separately for the pre- and
Centropages typicus Kr6yer post-cut period (Figure 4). The statistical significance
Acartia clausi Giesbrecht of the effects of the environmental variables, tested
Acartia latisetosa Kriczaguin by the Monte Carlo permutation procedure, should be
Acartia margaleft Alcaraz noted. Before the channel was cut, at both stations
Acartia tonsa Dana the community composition could be accounted for by
Euplanktonic cyclopoids environmental variables: the p value is < 0.01 for St.
Oithona nana Giesbrecht 4 and 0.01 for St. 8. In contrast, after August 1992 the
Oithona similis Claus environmental variables do not significantly explain
Oithona sp. the zooplanktonic community structure: p values are
Oncaea subtilis Giesbrecht 0.20 and 0.28 for St. 4 and St. 8, respectively.
Oncaea media Giesbrecht CCA ordination before August 1992 at St. 4 is
Oncaea sp.
characterized by the importance of salinity which con-
Corycaeus sp.
tributes particularly to the first axis in contrast both to
Euplanktonic harpacticoids
chlorophyll concentration and temperature. The typ-
Microsetella norvegica Boeck
ically marine taxa, tunicates, cyclopoids, harpacti-
Euterpina acutifi,(Jns Dana
51
St. 4
dl
(a) 22%
(b)
25%
ti
ro
cl ro
cd C dU be
58% 52%
T no
dl
sy
ha
fc
p< om p = 0.20
St. 8
(a) dl ! 22%
I (b) be 28%
ti gl
C
fc
rt:
ro
50% 53%
T-
ed
0
o
p =0.01 p =0.28
coids, and especially calanoids whose relative abun- on the upper right quadrant where salinity and chloro-
dance peaked during the winter months, are associated phyll content are at their maxima: indeed, in 1988 and
with the highest salinity values. At St. 8 ordination 1989 cyclopoids and calanoids occurred with isolated
space shows the larger part of taxa aggregated close to summer peaks after the dystrophic crisis.
the origin; tunicates, cyclopoids and calanoids, how- After August 1992 ordination at both stations still
ever, are separated from the other groups and located points to a close relationship between salinity and the
52
two typically marine taxa represented by tunicates and phytes to the dominance of phytoplankton within the
cyclopoids, whereas calanoids, which were more and primary producer community.
more dominated by Acartia tansa, are associated with The response of zooplankton to the environmental
autochthonous meroplankton taxa (gastropod, cirriped changes was analysed on an incomplete set of sam-
and bivalve larvae at St. 4, polychaete larvae at both ples. Moreover, data analysis was done on mixed
St. 4 and St. 8). Calanoids, in contrast to what was groups identified according to taxonomic and ecologi-
observed before the channel was cut, appear to be cal analogies among species, and statistical processing
inversely related to chlorophyll content. was carried out for principally explorative purposes.
Environmental variable plots calculated over the Nevertheless, these results seem promising.
entire investigation period, for an easier graph interpre- CCA ordination of both taxa and samples sug-
tation, are presented only for the first year (Figure 5). gests two lines of systemic interpretation of the evo-
At St. 8 the first axis is clearly defined by the opposition lution of the lagoon system we have investigated.
between dissolved oxygen and temperature; such an First, the significance of the differences in zooplank-
ordination underscores the central importance of sum- ton structure before and after the channel was cut is
mertime anoxia for this station. After 1992 a trend of demonstrated: a biocoenosis strongly influenced by
samples (summer ones included) to move towards the the forcing of environmental variables is replaced,
left quadrants, where dissolved oxygen is at its maxi- after 1992, by a biocoenosis more resilient to exter-
mum, is quite evident. At St. 4, on the contrary, the key nal factors. Secondly, the main ecological differences
variable appears to be salinity, which is in opposition between the two sampling areas are represented syn-
to chlorophyll content, thus confirming what had been thetically: in the centre of the lagoon zooplankton
observed in biplots of taxa scores for the same station. variability appears to be dependent on the action of
two factors, autochthonous production (phytoplankton
chlorophyll-a) and the influence of the sea (salinity);
Discussion the Valle di Gorino is characterized chiefly by summer
anoxia, but a tendency to attenuation of dystrophic
The case of the Sacca di Goro is representative of a con- events after 1992 is pointed out.
dition of eutrophy common to many coastal lagoons Analysis of the spatial-temporal patterns of the
and not only those in the Northern Adriatic (Viaroli et individual zooplankton taxa shows regularities and
ai., 1992; Ceccherelli et ai., 1994). The eutrophication consistent trends, above all in the Valle di Gorino:
process accelerated rapidly during the 1980's, leading here, after August 1992, a decrease in taxa strictly
to the enormous growth of Viva biomass and the conse- associated with Viva thalli (e.g., brackish water cope-
quent severe summer anoxia due to its decomposition. pods and gastropod larvae) and, more generally, in
The first dystrophic episode in the lagoon took place the minute zooplankton forms, in particular rotifers,
in 1987, and every summer up to 1992 was charac- was observed. Concurrently, the grazing chain appears
terized by increasingly more severe anoxic crises. The to have been strengthened by the increased density
hydraulic intervention in August 1992 did not result in of larger-sized forms of filter feeders: cladocerans,
a substantial reduction of water column trophism, inas- decapod larvae and above all calanoids represented
much as it remained elevated due to the heavy nutrient by Acartia tansa. The latter species has most proba-
loading from both the catchment area and the sediment. bly only recently colonized the lagoon (Farabegoli et
It did, however, attenuate environmental stress which ai., 1989), but within a period of a few years it has
was particularly severe in the Valle di Gorino and by become the dominant planktonic copepod (resulting
then also evident in the centre of the lagoon (Ferrari et in a reduction in the diversity of the Acartia complex
ai., 1993; Viaroli et ai., 1993a, 1993b). typically confined to the lagoon) and probably a key
The series of data examined in this study shows the species of the entire plankton community. A short-term
transition from an extreme condition towards an envi- zooplankton sampling campaign, we carried out in the
ronment which allows the recovery of a highly resilient Sacca di Goro in July 1994 (unpublished data), con-
ecosystem. The hydrodynamic and hydrological con- firmed the primary importance of A. tansa: in the Valle
ditions created by cutting of the channel resulted first of di Gorino, at ebb tide, copepodids and adults of this
all in the reduction of the area covered by the Viva beds species reached density peaks of more than 500 ind
and a shift from the dominance of macroalgae and epi- I-i.
53
(a)
1988 1989
4 37
47% 9
s 5 10
c 1112 8
1992 1993
4 6 8
5 4 5"
9
7
10
1994
6
5
78
Figure 5. CCA of the zooplankton in the Sacca di Goro from 1988 to 1994: scores of environmental variables and samples in the plane of
the first two axes of the ordination for St. 4 (a) and St. 8 (b). The axes in the yearly subplots have the same scaling; arrows of environmental
variables are shown only in the first subplot. The numbers correspond to sampling months; 5' and 5" indicate two different samples collected
in May 1993.
54
(b)
1988 26% 1989
6 4 5
2
o 7 10 5 6 7
8 2
4
1992 1993
6 5'
3 9 5"
5 8 7
10
7 49
10
8
1994
6
3
8'
Figure5b.
55
Abstract
A recolonization field experiment of two different artificially disturbed sediments (both defaunated sand and
defaunatedlreduced sand through organic enrichment) was carried out in the Sacca di Goro (Adriatic sea, Po
river Delta, Italy). Copepods showed themselves better colonizers than nematodes. In particular, copepods, in the
defaunated sand, were able to reach the same densities as the control site after only seven days from the beginning
of the experiment. In the reduced-sand, copepod recolonization occurred more slowly but reached the densities
found in both azoic and control sediments at the end of the experiment (15 days), when the values of total carbon
content decreased. The recovery evolution of the community structures was mostly dependent on the different
behaviour of the active epibenthic species of the harpacticoids (e.g. Canuella perplexa T. & A. Scott, 1893, Ameira
parvula (Claus, 1866), Robertgurneya similis (A. Scott, 1896)) and of the passively transported endobenthic ones
(e.g. Asellopsis sarmatica lakubisiak, 1938, Ectinosoma dentatum Steuer, 1940).
man induced disturbances. Recently they are more and The experimental site was located on the subtidal
more affected by increased eutrophication which had side of a sand bank facing the lagoon. From this sta-
caused, above all, a huge growth of macroalgal beds tion (CR, Figure 1) the sand (Md = 0.207 mm) was
(Ulvaceae) whose decomposition resulted in organic gathered which was made azoic for the experiment
enrichment of the sediments. The latter may cause by repeated freezing (-20°C) and thawing (40°C)
complete anoxia and reduced conditions in the sedi- and dried in the air for about one month. In order to
mentary habitat, which represents the main factor of obtain reduced conditions a part of this azoic sand was
acute disturbance for the biota, especially for benthic enriched with powdered dry Ulva thalli. The amount
organisms (Viaroli et aI., 1992). of this addition corresponded to a total carbon content
There are very few studies in the literature on the of about 100 g C m- 2 in the uppermost 3 cm of sedi-
recovery rates of sublittoral meiofauna communities ment. Thirty-six plastic containers (each with a surface
with respect to experimental organic enrichment of area of 49 cm 2 ) were filled up to the rim: 18 with the
sediment (see Coull & Chandler, 1992). simply defaunated sand (DF treatment) and 18 with
The aim of the present study was to investigate, the organically enriched sand (RD treatment). Before
in one of these lagoons (Sacca di Goro), whether being submerged in situ, all of those containers were
the recolonization process and the recovery rate of a incubated for 5 days in filtered sea water until com-
meiobenthic community in sandy patches were differ- plete reduced conditions were reached in those with
ently affected by two kinds of artificially induced dis- organically enriched sand. Containers were randomly
turbance. We compared the effects of either a merely arranged in a steel tray which was placed directly on
defaunated sand or a similarly defaunated but com- the sandy bottom of the experimental station (TR, Fi-
pletely reduced sand by organic enrichment with pow- gure 1). Only the sixteen internal containers were sam-
derded Ulva thalli. The latter tried to simulate one of pled in order to avoid the edge effect. Sampling was
the most severe disturbance event which periodically scheduled on 1-3-7-15 days from the beginning ofthe
occurs in these environments. In particular, copepod experiment because a previous study, carried out only
species compositions and community structure differ- with defaunated sand in the same site (Colangelo et aI.,
ences among the different sediment treatments and the 1994), had shown that the main changes in the reco-
natural environment were analyzed to determine if dif- very of the copepod assemblage occurred during the
ferent species behaved differently. The meiobenthic first two weeks. On each sampling occasion, 4 replicate
assemblage studied was that of a fine sand shoal in the containers were randomly sampled: two with defau-
investigated lagoon. nated sand and two with reduced sand. From each con-
tainer, 2 pseudoreplicate cores (3.14 cm 2 each) were
directly collected by hand (a total of 4 cores from DF
Materials and methods treatment and 4 from RD treatment). At the control site
(CR, Figure 1) 2 pairs of pseudoreplicate cores were
The experiment was carried out in the Sacca di Goro sampled from the sandy bottom by means of a remo-
(Figure 1), a subtidal shallow water embayment, one tely operated multiple corer. The uppermost 3 cm of
of the largest (26 km2 ) in the Po river Delta (Northwest sediment were retained, stained with rose bengal and
Adriatic coast) in June 1993. This lagoon shows wide preserved in 8% buffered formalin. In the laboratory,
yearly variations in salinity (10.8-32.9%0) (Colombo et sampled were sieved through both 1.0 and 0.03 mm
aI., 1991). In spite ofthe limited tidal range (typical of mesh size. The fraction on the 0.03 mm sieve was cen-
the Mediterranean sea) tidal effects are not negligible, trifuged three times with a LUDOX-AM water solution
mainly in the West-central part, owing to the wide sea (60%) in order to separate organisms from sediments.
mouth (2.5 km). Sediments are fine-grained (from silty Meiobenthic organisms were counted to major taxa.
clay to clayey silt) in the inner part and become coarser Copepods were identified to the species, staged and
toward the sea, where relict and recent sand bars are sexed.
located (Dal Cin & Pambianchi, 1991). Owing to the At the start of the experiment and on each sam-
eutrophication conditions of this lagoon (Pugnetti & pling date, two samples of surface sediment from CR
Zaccaria, 1991), severe anoxic crises, mainly due to and four random marginal containers (not sampled
decomposition of Ulva rigida Agardh beds, frequently for meiofauna), two with DF sand and two with RD
occur during the summer. sand, were analyzed for the following abiotic parame-
ters: Eh, with a 'Radiometer' platinum combined elec-
59
SACCA 01 GORO
o 2 Km
L..J I
Figure 1. Map of the Sacca di Goro lagoon. Locations of control (CR) and experimental (TR) sites are shown.
Start 3 7 15
Source of df F-value
variation Total Nematodes Copepods
meiofauna
Densities of nematodes were always statistically process: two epibenthic harpacticoids, the muddy-sand
lower in both treatment containers than in the control dweller Canuella perplexa T. & A. Scott, 1893 and the
(Table 2). Their constantly low densities in the treat- eurytopous Ameira parvula (Claus, 1866), and two
ment containers resulted in no significant interaction endobenthic species, the sand dwellers: Ectinosoma
with respect to time. The recolonization behaviour of dentatum Steuer, 1940 and Asellopsis sarmatica Jaku-
copepods appeared to be different. In the defaunated bisiak, 1938. The behaviour of each of them appeared
sand, they reached the same densities as the control to be different in relation to their functional charac-
after seven days from the beginning of the experiment. teristics (Appendix Table 1). The ANOVA F-values
In the reduced sediments the copepod recolonization for the two epibenthic species are reported in Table 3.
occurred more slowly. Density values of this taxon Their average densities and the SNK test results are
were not significantly different among ambient sedi- shown in Figure 3. Canuella perplexa seems to be a
ment and treatments only at the end of the experimental slow colonizer, its abundance values were significantly
period. lower in the treatment containers, particularly in those
Within the copepod assemblage, four species of with reduced sediments, until day 15. By contrast,
harpacticoid played a major role in the recolonization Ameira parvula confirmed itself as an opportunistic,
61
Table 3. Two-way ANOVA results for treatments (control; defaunated; reduced) and times (days from
the beginning of the experiment) relative to the major copepod species (for details of experimental
design see 'Material and Methods ').
-
141313
NEMATODA Canuella perplexa
.... EZ?J CR ....
121313 EZ?J CR
*-
60 a
* 0 DF 0 DF
ieee RD RD
see 40
.A.
a
61313 ....
ab ....
41313 20 .... b
ru b
~~
21313
!. * ~*i
I
E
u 13 l 11'1,,- ..!. ru
I
0
;i**a I,..., a
...
v 1 3 7 15
E
U
1 3 7 15
M
1l ...v
-
C
-
'-i 1413 M
S0 .... 40 a
a a b
60 ........ a
b b
a b
C
....
40 b 20
•
*
lUi In ! I. !
b
*a -
11
20
0 I,1, .! I~; 0
a
1 :3 7 15 1 3 7 15
DSbi S Da\Js
Figure 2. Average densities (± SO) per core (3.14 cm 2 )
of nematodes Figure 3. Average densities (± SO) per core (3.14 cm 2 ) of Canuella
and copepods in each treatment on each sampling date. Averages perplexa and Ameira parvula in each treatment on each sampling
with the same symbols among treatments on the same sampling date date. Averages with the same symbols among treatments on the same
are not significantly different. Symbols referring only to the legend sampling date are not significantly different. The same letters mean
labels in the Nematoda graph are relative to significant differences no significant differences among sampling date average values within
between treatment factor averages. The same letters mean no signif- each treatment. CR: control; OF: defaunated sand; RO: reduced sand
icant differences among sampling date average values within each (see text as for detail).
treatment. CR: control; OF: defaunated sand; RO: reduced sand (see
text as for detail).
*-
*CJ OF jectories than clusters. The trend of the development in
RO community structure in the azoic sand appeared to be a
28
constant progression. On the other hand, the date point
trend of the reduced sand community has a different
trajectory because the recolonization started later and
18
its pioneer phase showed marked differences from the
~In~
defaunated sand assemblage. However, the structure of
OJ
I
8 ml ,l, rI, the two communities converged toward a very similar
community in both the azoic and reduced sand by day
1 3 7 15
E
U 15.
'<t The ANOVA results of the diversity index values
.-I
(')
are shown in Table 4, trends of their averages in Fig-
38
Asellopsis sarmatica ure 6 and SNK test comparisons in Table 5. It is easy
"D ... EZZ/ CR to see a faster increase of the diversity values in the
....
C
*CJ OF
defaunated sand than in the reduced sand. However,
Z 28 *- RO
the values of all diversity indexes became not signifi-
cantly different between control and treatments toward
the end of the experimental period. The univariate mea-
18 sures of diversity take into account the distribution of
individuals amongst species but do not utilize species
8 ~~ in In~ identity. The fact that diversity was the same for the
communities in the three types of sediments by day 15
1 3 7 15
means that, in general, the copepod assemblages in the
Oau s two treatments reached an organized structure similar
Figure 4. Average densities (± SD) per core (3.14 cm 2 ) of Ecti- to that of the control. Instead the multivariate species-
nosoma dentatum and Asellopsis sarmatica in each treatment on
each sampling date. Treatment legend labels with the same symbols
dependent MDS method was much more sensitive in
indicate no significant difference between treatment factor averages. discriminating the three communities. This means that
CR: control; DF: defaunated sand; RD: reduced sand (see text as for the dominance relationships among species changed
detail). in the defaunated and reduced sand assemblages com-
pared to the natural environment. In particular the dom-
inance of the epibenthic species increased in the exper-
process was confirmed by the results for Robertgur- imental sand containers (Appendix Table 1).
neya similis (A. Scott, 1896) (Table 3). This species
was rare in the CR but it showed no abundance differ-
ences between CR and RD treatment and it exceeded Discussion
significantly the environment densities in the DF treat-
ment by day 7 (Appendix Table 1). The endobenthic The present investigation shows meiobenthic organ-
species, Asellopsis sarmatica and Ectinosoma denta- isms rapidly colonized the experimental containers.
tum, were present in the treatments from day 1 (Fig- In particular, the recovery of meiobenthic commu-
ure 4), but their average densities in the experimental nity structures seems to occur after just two weeks,
sand containers were always significantly lower than irrespective of the experimental conditions which sim-
in the control site (Table 3). ulated two different kinds (and levels) of disturbance
The overall pattern of community changes of the impact on the sedimentary environment. Fast colo-
copepod assemblage is shown by the results of the nization rates have been reported also in other stud-
MDS ordination (Figure 5). The date points of the ies on recolonization of disturbed habitats. After a
control (CR) form a separate and homogeneous clus- mechanically induced disturbance, Sherman & Coull
ter which indicates the relatively constant copepod (1980) observed that densities of most meiobenthic
assemblage in the environment. On the contrary, the taxa reached the same levels as those at the control
date points of both DF and RD treatment communities site after just 12 hours. Le Guellec (1988), working
63
GORO
COPEPODA
DF15
,
+
RD15
\' ... _ _ _ RD7
.- CR15
-- - \
Y...
I \ I ""
\ DF7 "",
;
t'I \ '1\
\
\
\
\
"
CR7 I 'R03
a \ "
I
I
DF3
\
~
R9 1
CR3CR1
'------~
,
'DF1
"
DItJENS 1
(stress=O.06)
Figure 5. MDS ordination plot of date points for the three treatments. CR: control; DF: defaunated sand; RD: reduced sand. 1-3-7-15: days
from the beginning of the experiment.
Source of df F-value
variation No species Hill's NI Evenness
with exogenous azoic sand, reported a similar result As for the effect of physicochemical alterations of
after only two tidal cycles. Thistle (1980) also found sediments induced by organic enrichment, the present
that the recolonization of defaunated enteropneust fae- investigation showed that, in general, colonization by
cal mounds by subtidal harpacticoids occurred in less most meiofaunal populations and recovery of meioben-
than 24 hours. Sun & Fleeger (1994), even though thic assemblage structure seemed to be delayed in
they were mainly interested in investigating meiofau- reduced sand compared to the defaunated sand. The
nal recolonization of azoic sediment mimicking bot- recovery rates of different meiobenthic taxa were relat-
tom depressions, remarked how the abundances of the ed both to their different movement mechanism, and
dominant copepods showed no significant differences consequent ability to recruit to sediment, and to the
between experimental and ambient sediments after 24- degree of disturbance of the experimental sediments.
48 hours. Nematodes were the least active colonizers, but this
0'1
.j::o.
Appendix 1. Average densities (± SD) per core (3.14 cm 2 ) of meiobenthic copepod species recorded in each treatment on each sampling date from the beginning of the experiment. Functional
habits: M =mud; S =sand; Ph =phyta1; E =eurytopous; epi =epibenthic; end =endobenthic.
CONTROL DEFAUNATED REDUCED
3 7 IS 3 7 IS 3 7 IS
Canuella perplexa (MS-epi) 33.8 (±22.7) 26.0 (± 4.8) 8.S (± 4.7) 16.S (± 6.2) 0.8 (± I.S) 1.8 (±O.S) 1.3 (±1.3) 6.3 (± 2.1) 0.3 (±O.S) 0.3 (±O.S) 11.0(± 8.1)
Ectinosoma dentatum (S-end) s.o (± 4.2) 10.0 (± 4.8) 8.0 (± 4.6) 6.8 (± 4.1) 0.8 (± 1.0) I.S (± 1.3) 2.8 (± 2.4) O.S (± 1.0) 0.3 (±O.S) 0.3 (±O.S) O.S (± 0.6)
Microarthridion fallax (MS-end) 2.S (± 1.0) 2.3 (± 1.3) O.S (± 0.6) 0.8 (± 1.0)
Harpacticus fiexus (S-epi) 1.8 (± 1.0) 1.0 (± 0.8) 0.3 (± O.S) O.S (±0.6)
Tisbe holothuriae (Ph) 0.8 (± I.S) 3.0 (± 4.8) 0.3 (±O.S) O.S (± 1.0) 0.8 (±1.0) 2.0 (± 2.7)
Dactylopusia tisboides (E-epi) 0.3 (± O.S) O.S (± 1.0) 0.3 (±O.S) 0.3 (±O.S)
Amphiascus parvus (Ph) 0.3 (± O.S) 0.3 (± O.S)
Robertgurneya similis (E-epi) 0.8 (± 1.0) 0.8 (± 1.0) O.S (± 0.6) 0.8 (± 1.0) 0.8 (± I.S) 2.S (±0.6) 4.8 (± 3.9) 0.3 (±O.S) 0.3 (±0.5) 1.0 (± 1.2)
Schizopera compacta (E-epi) 0.8 (± 1.5)
Ameira parvula (E-epi) 17.5 (± 9.6) 27.8 (± 79) 12.8 (± 9.6) 16.8 (± 16.8) 1.3 (±2.5) 5.8 (±4.1) 19.8 (±4.3) 19.8 (±12.1) 2.3 (± 1.7) 14.5 (± 11.9)
Mesochra pontica (E-epi) 0.3 (± 0.5) 0.3 (±0.5) 0.5 (± 1.0)
Enhydrosoma gariene (M-end) 0.3 (± 0.5)
Paralaophonte congenera (E-epi) 0.3 (± 0.5) 0.8 (± 1.0) 1.3 (± 1.5) 0.3 (±0.5) 0.3 (±O.S) O.S (±0.6) 0.5 (±0.6) 0.5 (± 0.6)
Asellopsis sarmatica (S-end) 13.8 (± 6.6) 15.8 (± 6.1) 13.0 (± 5.3) 16.8 (± 10.2) 2.0 (±2.5) 3.3 (±1.0) 3.3 (±2.9) 0.8 (± O.S) 0.3 (±0.5) 0.8 (±1.0) 0.3 (± O.S)
Cyclopina gracilis (E-epi) 0.3 (± 1.0) 0.3 (±O.S) O.S (±0.6) O.S (± 0.6)
TOTAL 75.S (±35.5) 84.8 (±13.4) 46.S (±19.8) 61.3 (±31.4) 45 (±4.4) 13.0 (±3.9) 29.0 (±6.3) 3S.3 (±13.3) I.S (±1.7) 1.0 (±1.4) 6.0 (±3.5) 31.0 (±21.4)
65
Table 5. Studentized-Newman-Keuls post hoc test for mUltiple
comparisons of diversity index values for the copepod assem-
manence of reduced conditions in deeper layers of sed-
blage, relative to the significant interaction. Average values with iments (Table 1) prevents those species living beneath
the same symbols among treatments on the same sampling date the surface of the sediment from colonizing success-
are not significantly different. The same letters indicate no statisti- fully (Gee et aI., 1985). Thus endobenthic species may
cally significant differences among sampling date average values
within each single treatment. CR: control; DF: defaunated sand;
take much longer time to recover. Probably, the short
RD: reduced sand (see text as for detail) duration of this experiment might have been critical
in masking the time settlement potential of these slow
Treatments No species Hill's NI Evenness colonizing species (Sun & Fleeger, 1994).
CRxTimes V' 6.75 a V' 4.08 a 59.41 a Epibenthic copepods (e.g. Canuella perplexa,
3 V' 7.00 a V' 4.47 a 61.59 a Ameira parvula and Robertgurneya similis) seem to
7 V' 6.75 a V' 4.93 a 65.90 a be very good colonizers. This depends on their abil-
15 V' 5.50 a V' 4.03 a 65.50 a ity to enter actively the water column to be trans-
ported (Kern, 1990). In fact, their densities in both
DFxTimes * 1.50 a * 1.39 a 16.82 a experimental treatments fully reached ambient sedi-
3 V' 4.50 b V' 3.80 b 69.98 b ment levels toward the end of the experimental period
7 *V' 5.25 b *2.8Iab 52.30 ab (Figure 3; Table 3). However, notwithstanding their
15 V' 5.50 b V' 3.51 b 58.89 ab dispersal capability, the recolonization of the reduced
sand appeared to be delayed with respect to the azoic
RDxTimes * 1.25 a * 1.46 a 19.15 a sand. Probably, the epibenthic species show a selective
3 * 1.00 a * 1.50 a 22.50 a settlement behaviour in relation to the chemical status
7 * 3.50 b *V' 3.07 b 72.73b
of the sediments. Although they are able to settle rapid-
15 V' 5.50 b V' 3.23 b 56.77 ab
ly in the merely defaunated sand, they might postpone
their settlement in the reduced sand perhaps waiting
for more oxygenated conditions of surface sediment.
result was probably due to our experimental set-up. Gee et al. (1985) carried out a mesocosm exper-
This did not allow lateral invasion by crawling through iment on the effects of organic enrichment on meio-
sediment interstices which is the main route by which fauna by adding powdered macroalgae (Aseophyllum)
nematodes recolonize azoic sediment patches (Chan- to experimental containers of mud in two different
dler & Fleeger, 1983). Nematodes are known to occur quantities (50 and 200 g C m- 2 respectively). In their
deeper in sediments (Jacobs, 1984) and to be passive- high dose treatments, after 56 days, all the burrowing
ly suspended in the water column only when shear- species had disappeared and the fauna was dominated
currents exceed a certain threshold (Sherman & Coull, by only five species of the Tisbe guild, even though
1980; Palmer, 1988). So, in our experiment, they could no specimens of this genus were found in the field
arrive in the containers only by passive transport via samples. Since members of the genus Tisbe are oppor-
water column. That is probably the reason why we tunist organisms this may explain their presence in
observed a poor recolonization of experimental sand these organically stressed environments. In the present
by those organisms. The same thing probably applied, study, at intermediate quantity of powdered macroal-
to a certain extent, to endobenthic copepod species, gae addition (100 g C m- 2 ), the opportunistic role
Ectinosoma dentatum and Asellopsis sarmatica. Their seemed to be taken by Ameira parvula (Figure 3).
abundances in the treatments never reached those of The adults of this species have an active migratory
the control site, not even at the end of the experimental behaviour (Kern, 1990) and are able to settle massively
period. These endobenthic species seem to be less good (mainly with ovigerous females) during the first phase
colonizer but only because, probably, they are able to of recolonization of defaunated sediments (Colange-
burrow into the sediment as flow increases (e.g. at high lo et aI., 1994). Also Warwick et al. (1990) found
water) and thus avoid transport (Palmer, 1988). Con- an Ameira sp. which seemed to swim actively at high
sequently, few of them randomly encounter the con- water slack and to be capable or redistributing itself in
tainers. On the other hand, even if no statistical differ- the sandy habitat, in order either to avoid predation or
ences were found between densities in defaunated and to search out sediment patches more attractive in terms
reduced sand, in the latter the abundances of these two of available food resources.
species always seemed to be lower (Figure 4). At high The ordination plot of the copepod assemblage
levels of organic enrichment (RD treatment) the per- (Figure 5) and the trends of the diversity index values
66
L-+-l~
8
OJ
.-1
U
It is clear that, besides the granulometric charac-
OJ 6 teristic of sediments, settlement of meiobenthic organ-
. --
0..
... ' ....
UI
4
1 . ~. t·········
.",-
isms (that of harpacticoid at least) seems to be con-
"-0 . .V" trolled by the chemical status of the sedimentary habi-
0 2 )//1 tat. We have still to understand the complicated dynam-
Z ics in the sediment strata. Notwithstanding the perma-
-··-··-·v'
9 i nent negative redox potentials, the oxidative processes
1 3 "7 15 of the sediment carbon content seem to have been set up
19
quickly in the investigated reduced sand (Table 1). This
probably means that oxic/anoxic and sulphidic chemo-
8 clines become established in the sediment (Revsbech
..-j et aI., 1980). Oxic microenvironments could exist amid
Z
6 reduced sediment and enable co-occurring aerobic and
UI
anaerobic processes on a millimetre scale (Fenchel,
0-1
0-1 4 1992). For meiobenthos, the eco-physiological adap-
'-1
J: tive mechanism involved in resistance to reduced con-
2
ditions are still largely unknown. However, the present
results seem to emphasise that meiobenthic popula-
9
1 3 "7 15 tions, living in the complicated systems around the
above-said chemoclines, are probably adapated to
exploit a set of chemical and ecological microniches.
1.9
,...
'J 9.8
Acknowledgements
UI
III 9.6
OJ
C The authors wish to thank Mr Vittorio Gaiani for his
C e.4
OJ precious collaboration in field work. The research was
)
UJ 9.2
~'-"-l supported by a grant from the ECC-CLEAN Project
(contract No. EV5V CT92 0080).
e.9
1 3 "7 15
Days References
0-0 CR 0'·0 OF v-··-v RD
Ceccherelli, V. U., G. Caramori, M. A. Colangelo, O. Fornasari, V.
Figure 6. Trends of diversity indices average values (± SO) for the Gaiani, C. Mantovani & G. Reggiani, 1992. La successione eco-
copepod assemblage in each treatment. CR: control; OF: defaunated
logica nelle comunita bentoniche di ambienti lagunari in relazione
sand; RO: reduced sand.
ai fenomeni di disturbo. S.It.E.lATTI 15: 135-153.
Chandler, G. T. & J. W. Fleeger, 1983. Meiofaunal colonisation of
azoic estuarine sediment in Louisiana: mechanisms of dispersal.
J. expo mar. BioI. Ecol. 69: 175-188.
(Figure 6) demonstrated the timing and pattern of the Colangelo, M. A., V. Gaiani & V. U. Ceccherelli, 1994. Evoluzione
community recovery. During the first recolonization del popolamento meiobentonico su sabbia artificialmente defau-
phase (up to day 3), when higher values of both car- nata. BioI. Mar. Medit. I: 243-247.
bon content and negative redox potential were record- Colombo, G., C. Camerota, R. Bisceglia, V. Zaccaria, V. Gaiani &
A. Carrieri, 1991. Variazioni spaziali e temporali delle caratteris-
ed in the reduced sediments (Table 1), the greatest tiche fisico-chimiche delle acque e della biomassa fitoplanctonica
differences in community composition were observed. della Sacca di Goro. In S. Bencivelli & N. Castaldi (eds), Stu-
Afterwards, when the levels of the organic enrichment dio integrato sUll'ecologia della Sacca di Goro. Franco Angeli,
decreased, the biotic structuring forces of the commu- Milano: 9-38.
Coull, B. C. & M. A. Palmer, 1984. Field experimentation in meio- Pearson, T. H. & R. Rosenberg, 1978. Macrobenthic succession
faunal ecology. Hydrobiologia 118: 1-19. in relation to organic enrichment and pollution of the marine
Coull, B. C. & G. T. Chandler, 1992. Pollution and meiofauna: field, environment. Oceanogr. Mar. Bio!. ann. Rev. 16: 229-311.
laboratory, and mesocosm studies. Oceanogr. Mar. Bio!. ann. Rev. Pugnetti, A. & V. Zaccaria, 1991. Ricerche sui fitoplancton della
30: 191-271. Sacca di Goro. In S. Bencivelli & N. Castaldi (eds), Studio inte-
Dal Cin, R. & P. Pambianchi, 1991. I sedimenti della Sacca di Goro grato sull'ecologia della Sacca di Goro. Franco Angeli, Milano:
(Delta del Po). In S. Bencivelli & N. Castaldi (eds), Studio inte- 59-71.
grato sull'ecologia della Sacca di Goro. Franco Angeli, Milano: Revsbech, N. P., J. S\lrensen, T. H. Blackburn & J. P. Lomholt,
253-263. 1980. Distribution of oxygen in marine sediments measured with
Fenchel, T., 1992. What can ecologists learn from microbes: life microelectrodes. Limno!. Oceanogr. 25: 403-411.
beneath a square centimetre of sediment surface. Funct. Eco!. 6: Rofes, G., 1980. Etudes des sediments. Methodes de prelevement et
499-507. d'analyse pratiquees au laboratoire de sedimentologie. CTGREF,
Field, J. G., K. R. Clarke & R. M. Warwick, 1982. A practical strategy DIVISION QEPP, Etude 47, 50 pp.
for analysing multispecies distribution patterns. Mar. Eco!. Prog. Sherman, K. M. & B. C. Coull, 1980. The response of meiofauna to
Ser. 8: 37-52. sediment disturbance. J. expo mar. Bio!. Eco!. 46: 59-71.
Gee, J. M., R. M. Warwick, M. Schaanning, J. A. Berge & w. G. Sun, B. & J. W. Fleeger, 1994. Field experiments of the colonisation
Ambrose Jr, 1985. Effects of organic enrichment of meiofaunal of meiofauna into sediment depressions. Mar. Eco!. Prog. Ser.
abundance and community structure in sublittoral soft sediments. 110: 167-175.
J. expo mar. Bio!. Eco!. 91: 247-262. Thistle, D., 1980. The response of a harpacticoid copepod communi-
Heip, C., R. M. Warwick, M. R. Carr, P. M. J. Herman, R. Huys, N. ty to a small scale natural disturbance. J. mar. Res. 38: 381-395.
Smol & K. van Holsbeke, 1988. Analysis of community attribut- Underwood, A. J., 1981. Techniques of analysis of variance inexper-
es of the benthic meiofauna of FrierfjordlLangesundfjord. Mar. imental marine biology and ecology. Oceanogr. Mar. Bio!. ann.
Eco!. Prog. Ser. 46: 171-180. Rev. 19: 513--605.
Jacobs, L. J., 1984. The free-living inland aquatic nematodes of Viaroli, P., A. Pugnetti & I. Ferrari, 1992. VIva rigida C. AGARDH
Africa - a review. Hydrobiologia 113: 259-291. growth and decomposition processes and related effects on nitro-
Kern, J. C., 1990. Active and passive aspects of meiobenthic cope- gen and phosphorus cycles in a coastal lagoon (Sacca di Goro,
pods dispersal at two sites near Mustang Island, Texas. Mar. Eco!. Po River Delta). In: G. Colombo, l. Ferrari, V. U. CecchereIIi,
Prog. Ser. 60: 211-223. R. Rossi (eds), Marine eutrophication and population dynamics.
Le Guellec, c., 1988. Colonisation d'un sable azo·ique exogene par Olsen & Olsen, Freedenborg: 77-84.
la meiofaune. Cah. Bio!. Mar. 29: 469-481. Warwick, R. M., K. R. Clarke & J. M. Gee, 1990. The effect of dis-
Palmer, M. A., 1988. Dispersal of marine meiofauna: a review and turbance by soldier crabs Mictyris platycheles H. Milne Edwards
conceptual model explaining passive transport and active emer- on meiobenthic community structure. J. expo mar. Bio!. Eco!. 135:
gence with implications for recruitment. Mar. Eco!. Prog. Ser. 48: 19-33.
81-91.
Hydrobio[ogia 329: 69-78,1996. 69
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Key words: Meiobenthos, harpacticoid copepods, vertical distribution, emergence, Canuella perplexa, Amphiascus
parvus
Abstract
The vertical distribution of meiobenthic copepods was investigated within muddy sediments of a eutrophic lagoon
(fish ponds of Arcachon Bay, France). The aim of the study was to determine if in muddy sediments, as previously
established in sandy sediments, meiobenthic copepods migrate vertically according to the seasons or diel periods.
Two experimental approaches were used, viz: a three-season comparison was made of the diel vertical distribution
of the harpacticoid Canuella perplexa T. & A. Scott (1893) and secondly the depth distribution of a meiobenthic
copepod assemblage was followed for a 24 h period, in shallow water subtidal locations. The harpacticoid C. per-
plexa vertically migrated through the top three centimeters of the sediment, showing diel and seasonal variations in
depth distribution. The differential vertical distributions shown by the dominant meiobenthic populations suggest
that emergence into the water column may mainly concern surface dwelling copepods. The physical and biological
factors affecting seasonal and die I changes in the copepod assemblage of the fish ponds are discussed.
67% of all meiofauna from shallow (50 cm depth) blage over a diel period. Seasonal changes were inves-
muddy areas were restricted to the top first cm. Yingst tigated by comparing diel depth distribution patterns of
(1978) counted 88% ofharpacticoids within the upper- C. perplexa within 3 diel samplings carried out during
most cm of muddy sediments from a 14 m deep bottom. different seasons.
Similarly Jensen (1983) found 98% of the total number
of harpacticoids in the top first cm of a sublittoral soft
bottom. Smith & Coull (1987) determined that mud Materials and methods
flats contain in the top first cm approximately twice
as many meiofauna as the first 10 cm of sandy sedi- Study site
ments. Furthermore, investigations using flume exper-
iments could not find evidence that meiobenthic cope- Sampling sites (stations Cl and G) were located in the
pods vertically migrate in muddy sediments (Palmer & fish ponds of Arcachon Bay which are semi-enclosed
Molloy, 1986; Coull et aI., 1989). However, prelimi- and shallow lagoonal impoundments (see description
nary field routine investigation carried out in 1993 in by Castel et aI., 1996). Sediments of the fish ponds
Arcachon Bay showed a surprising vertical distribution which are partly covered by canopies of the seagrass
of meiobenthic copepods (Buffan & Castel, 1993). In Ruppia cirrhosa, are rich in organic detritus and silt
the muddy sediment sampled in the fish ponds, 74% (36-58.4%), (Escaravage & Castel, 1989). In the sam-
of the copepod assemblage were recorded in the 2- pling area POC content generally varies between 4.5%
5 cm depth horizon, in spring, when the assemblage to 6.5% (Etc heber, pers. comm.). The water depth at the
was numerically dominated by Canuella perplexa T. sampling stations is 0.4 to 0.6 m. Station G was locat-
& A. Scott, 1893, a common harpacticoid of coastal ed near a deeper channel allowing renewal of water
environments (Heip, 1973; Castel, 1992; Ceccherelli whereas station Cl was more isolated.
et aI., 1994). These data suggested that in these muddy
sediments, migrations of meiobenthic copepods from Sample collection
the surface to the deeper layers may be significant.
A number of studies have shown that meiofauna Sediment was hand-collected using plastic core tubes
and especially harpacticoids display upward migra- (3 cm internal diameter and 5 cm length) and imme-
tions from sediment into the overlaying water column diately preserved in liquid nitrogen. In the laboratory,
(Hicks, 1986; Walters, 1988; Armonies, 1989). Such frozen cores were sliced every 1 cm from 0 to 5 cm
movements are related to a number of factors, includ- depth (in June, this was made immediately after the
ing light levels (Armonies, 1988), diel periods (Wai- sampling of sediment cores). Slices were then washed
ters, 1988; 1991; Walters & Bell, 1994), tidal periods through a 63 f.J,m sieve to collect the total copepod
(Bell et aI., 1988), copepod densities (Service & Bell, assemblage or through a 200 f.J,m sieve to selectively
1987; Walters, 1991), and species behaviour (Walters, collect Canuella perplexa. Samples were preserved in
1991). Therefore in the present study, we have investi- 5% formalin and stained with Rose Bengal. The fauna
gated the vertical distribution of meiobenthic copepods were extracted from the sediment by flotation in Ludox
in a eutrophic lagoon using both an estimate of the HS-40 (De Jonge & Bouwman, 1977). Meiobenthic
density distribution according to depth in the sediment copepods were sorted under a dissection microscope
and a quantitative estimate of emerging copepods in and identified by microscopic observations. Physico-
the water column. chemical variables (temperature of the sediment sur-
Field investigations were conducted to answer the face, dissolved oxygen, and salinity) were measured
following 3 questions concerning the vertical distri- during the sampling periods.
bution patterns of meiobenthic copepods in subtidal Copepods emerging from the sediment were col-
muddy sediments. lected in the water column using emergence traps
1. Does the pattern vary over diel periods and are made as previously described by Walters & Bell
these variations linked to the night-day cycles? (1986). Emergence traps (Figure 1) collect any organ-
2. Can the diel patterns be related to the emerging ism actively moving at least 12 cm above the sedi-
behaviour of harpacticoids? ment surface. Each trap was closed using an alumini-
3. Does the diel pattern vary according to seasons? um cap, filled with filtered seawater (63 f.J,m), care-
Diurnal changes were investigated by comparing fully positioned on a cylindrical support, and the cap
depth distribution of a meiobenthic copepod assem- was removed. The supports (plexiglass cylinders of
71
o 6 em
o
o
o
Figure 1. Emergence trap positioned between Ruppia patches in the fish ponds (station G, October). Modified from Walters & Bell (1986).
23.8 cm 2 surface) were carefully inserted into the sed- sunset and sunrise which may be the periods of max-
iment to a depth of 4 cm, at least 1 hour before trap imum emergence over a diel period (Armonies, 1988;
placement. At the end of the field incubation, the trap Walters, 1991). Four replicate cores were collected
contents were washed through a 63 p,m sieve and pre- on the 15 th October at 17:00, 18:00 (sunset), 19:00,
served in 5% v/v formalin until sorting of copepods. 24:00, and on the 16th October at 01 :00, 06:00, 07:00
Only adult copepods were identified to the species lev- (sunrise), 08:00, 13:00, 14:00, 17:00, 18:00 (sunset),
el. 19:00. Two cm of overlaying water were sampled for
each core. These water aliquots were then separat-
Sampling designs ed from sediment and considered as a slice in the
results. From 19:00 on the 15 th October to 19:00 on
Station Cl was investigated for 21 h in June 1993 (03- the 16 th October, three replicate emergence traps were
04 June) and for 45 hin March 1994 (27-28-29 March). placed between Ruppia patches at each sampling time
The aim of both samplings was to specifically study the to be collected at the following sampling time.
vertical distribution patterns of C. perplexa over diel
periods under different seasonal conditions. Every 3 Statistical analysis
hours, 3 (in June) to 8 (in March) replicate cores were
collected over the diel periods. Friedman two-way analysis of variance was used to
Station G was investigated for 26 hours in October test for significant differences between averages of
1994 (15-16 October). Preliminary samplings have copepod densities between sampling times over the
indicated both that C. perplexa was abundant at sta- diel periods, or according to depth. The Kolmogorov-
tion G and that the copepod assemblage was more Smirnov two-sample test was used to compare average
diversified than those from station C 1. Thus, we decid- densities of C. perplexa, at a depth horizon, between
ed to sample station G to study a possible difference two sampling periods or between two depth horizons.
in vertical distribution pattern between co-occurring The significance level of probability (P) was 0.05.
species. Furthermore, the sampling was conducted
to evaluate the emerging fraction of copepods and
to examine the possible relationship between emerg-
ing behaviour and the vertical distribution patterns of
copepods. Sampling points were concentrated around
72
-
>-
'iii
cQ)
80%
60%
0 0-1 em
diurnal peaks of surface sediment temperature respec-
tively (Figure 3). In contrast, during the June sampling
copepods were mostly found in the 2-3 cm depth hori-
---
"tI
iii o 1-2 em zon at higher temperatures (43-61 % between 16:00
0 40% and 19:00), whereas at lower temperatures most indi-
• 2-3em
0 viduals were found in the 1-2 cm depth horizon (41-
:.!!.
0 20% 67% at 01:00-04:00) (Figure 3). In October the tem-
--"-+-
0% .......
March June October
perature did not show a significant diurnal fluctuation,
however, copepods tended to migrate down, from the
first to the second and the third centimeters below the
Samplings surface, between 06:00 and 08:00 and between 17:00
Figure 2. Mean vertical distribution of the harpacticoid C. perplexa in
and 19:00, (Figure 3). These results indicate that fac-
the top 3 cm of the sediment in March (station Cl), June (station Cl), tors other than surface sediment temperature influence
and October (station G). the depth distribution of C. perplexa.
During the sampling programme the salinity varied On average harpacticoids represented 87% of the
between 10%0 (in March), 17%0 (in June), and 29%0 (in meiobenthic copepod community, with copepodites
October). Dissolved oxygen similarly varied between and adults representing 30% and 57% of the communi-
the 3 diel periods, with, on average, a minimum of ty respectively. The most abundant species, Canuella
2.3 mg 1-1 (early morning) and a maximum of 13.8 mg perplexa and Amphiascus parvus Sars 1906 comprised
I-I (sunset). Generally, the copepods (79.4% to 98%) 72% of the harpacticoid community. The remain-
were restricted to the top 3 cm of the sediment. Conse- der of the assemblage comprised 8 species: Schiz-
quently the vertical distribution of copepods was stud- opera sp. Sars 1905 (7.3%); Paralaophonte brevi-
ied in the 3 cm depth column. rostris Claus 1863 (7.06%); Mesochra pygmea Claus
1863 (S.8%) ; Robertgumeyasp. (2.7%); Cletocamptus
Vertical distribution ofe. perplexa over diel periods confluens Schmeil 1894 (1.7%); unidentified species
and between seasons (June 1993, March 1994, and (1.3%); Amonardia normani Brady 1872 (0.82%); and
October 1994, samplings) Paradactylopodia sp. (0.68%). No ovigerous females
were observed among the population of C. perplexa.
Between sampling periods (Figure 2), the mean tem- In contrast, A. parvus ovigerous females were found.
perature of the sediment surface varied from 16.3 °C The abundances of copepods in the emergence traps
(March) to 19.4 °C (October) and 21.4 °C (June). On differed over the 24 h period, (Figure 4). Friedman
average 60% of the copepods were concentrated in the analysis of variance was performed on data sets con-
top first cm in March (P<0.05, Friedman test), where- cerning traps incubated in the field for equivalent dura-
as in June 80% of the population was found in the tions (i.e. a set pooling abundances found in traps incu-
1-3 cm depth horizon (P = 0.05, Friedman test). In bated in the field for 1 h, and a set pooling abundances
October 89.7% of individuals were concentrated in the found in traps incubated in the field for > 1 h dura-
uppermost 2 cm (P<O.OOI, Friedman test) and equally tion). Average abundances were significantly higher
distributed in the 0-1 cm and 1-2 cm depth horizons. (P<O.OS) in emergence traps incubated between 19:00
Over the diel periods (Figure 3), the temperature and 24:00 (291 ± lOS ind trap-lor 23.8 cm 2 sedi-
variation at the sediment surface was greater in March ment surface) and between 06:00 and 07:00 (P=O.OS,
and June (around 10 0c) than in October (1°C). In 161 ± 87 ind trap -I) than in other traps incubated for
March most individuals (74-87%) were found in the equivalent durations respectively. Considering that the
top first cm at higher temperatures whereas at lower total copepod abundance in the sediment reached on
temperatures most individuals (51-59 %) were found average 120S ± 712 ind 23.8 cm- 2 over the sampling
in the 1-2 cm depth horizon. The percentage of the campaign, the emerging fraction can be estimated to
73
~I ~_
07:00 respectively. C. perplexa was not found in trap
contents. The species composition was thus different
n _____ --- - -
between the emergence periods, with the second emer-
o gence phase mainly attributable to a single species,
100% ~-------------=--,
A. parvus.
80%
60% Vertical distribution patterns oj meiobenthic
40% copepodsjrom station G (October 1994 sampling)
20%
Most copepods (79.4%) were distributed in the top
3 cm of the sediment, 4% were found deeper, and
16.6% were found in the overlaying water (i.e. in
October. station G
the 2 cm water column above the sediment sur-
~I --- - - - - - - - - ------ - - - - - -
face). On average, the density of copepods which
showed an emerging behaviour (i.e. density of the
non-emerging copepod C. perplexa was removed), sig-
nificantly decreased with depth (P<O.OOI, Friedman
100% test) with 104.5 ± 24 ind 7 cm-3, 50 ± 15 ind 7 cm- 3 ,
80% 26.5 ± 7 ind 7 cm-3, at 1 cm, 2 cm, and 3 cm below
60%
the surface respectively.
40%
Over the diel period (Figure 5) the vertical distribu-
tion pattern of copepods (c. perplexa excepted) tended
Ull to change around sunset. Before sunset (between 13 :00
StS
Sampling time
and 18:00), densities in the overlaying water and in the
first cm of the sediment showed an increase averag-
• 2·3 em IlilJ 1·2em D 0·1 em
ing 14% and 13% of the total assemblage respectively,
Figure 3. Vertical distribution of the harpacticoid C. perplexa in the whilst in the deeper sediment strata densities tended to
top 3 cm over diel periods: for 45 hours in March, 2l hours in June,
and 26 hours in October. decrease by 13.5% and 12.5% of the total assemblage
in the 1-2 cm and the 2-3 cm depth horizons respec-
tively. These results indicate that before sunset cope-
be 15-33% and 6-21 % of the total copepod assem- pods tended to migrate upwards and to regroup at the
blage for the 19:00-24:00 and the 06:00-07:00 emer- sediment water interface. At sunset (between 17:00 and
gence periods respectively. However, since our sam- 19:00), densities in the overlaying water (0-2 cm above
pling method does not allow the collecting of emerg- sediment) tended to decrease. This decrease averaged
ing copepods which were swimming between 2 cm between 7 to 37% of the total assemblage. This change
and 12 cm above the sediment surface, these percent- occurred at the beginning of the emergence period sug-
74
400 s
s u
350 u n u
n
'f 300 s I
n
s
u e S
~ 250 e
M t e
N
'-'
..
0.
...CI
; 150
200
c.
Z 100
50
0
0 0 0 0 0 0 0 0
~ ~ 19:00-24:00 '? 01 :00-06:00 ~ ~ 08:00-13:00 ~ 14:00-17:0f.J '? q
co en
;:; ...... co
~ ~
0
co
0 "'"
~ ~
0">
~
I
60
I
0
0
0
0
6
0
0
0
I
0
0
I
0
0
I
0
I
,.:..: co V
N
t:ri ,.:..: M ,.:..: ~
co
0 c::::l
gesting that a number of copepods began to swim up in 1973; Jensen, 1983; Coull et aI., 1989). Vertical distri-
the water column (above the overlaying water). During bution of metazoans is generally related to the oxygen
the night, copepods were mainly found in the first cm concentration in pore water. In the muddy sediments
of the sediment. of the fish ponds, the maximum depth of oxygen pen-
During both emergence periods (19:00-24:00 and etration was 5 mm (Schaub & van Gemerden, 1996).
06:00-07:00), the vertical distribution patterns of the However, infaunal activity (e.g. tube building by poly-
emerging copepod A. parvus and the non-emerging chaetes and crustaceans) provides micro-oxygenated
copepod C. perplexa were compared (Figure 6). Dur- halos (Meyers et aI., 1987) and phanerogam roots
ing both periods the depth distribution of C. perplexa may enhance the oxygen penetration at depth. In addi-
was consistent with the average distribution pattern tion, potential food sources for copepods such as pho-
observed over the whole die 1 period (Figure 2). In totrophic microorganisms are also present through the
contrast, the depth distribution of A. parvus tended top 3 cm of the sediment (Buffan et aI., 1993 and
to change during the second emergence phase (06:00- unpublished data), indicating that copepods may find
07:00). Density of the copepod in the first cm tended to food sources below the sediment surface.
decrease, inducing a more uniform depth distribution
in the top 3 cm (Figure 6). Diel and seasonal variations o/the vertical
distribution pattern o/the harpacticoid C. perplexa
100%
80%
60%
40%
20%
0%
888 888
~ =~ 8 (:i 8
.2-3 em !IE 1-2 em 0 0-1 em UillI Overlaying
water
Figure 5. Vertical distribution of the copepod assemblage at station G in October; C. perplexa population was removed. Samples were taken at
different time intervals over a 26 hour time period.
!1
c:
centages of individuals were found in the top cm in
March and October when average sediment surface
~ temperatures approached 17 -19°C. Furthermore, the
'j5 2 uniform distribution of copepods in the uppermost 2 cm
.c r-----------------------~ observed in October was tied to a vertical homogeniza-
.c
g. 3
C
tion of temperatures in the sediment due to the stabil-
ity of the diurnal temperature. However, it should be
taken into account that the change of sampling sta-
o 10 20 30 40 50
tion, from station C1 (March and June) to station G
Mean density (n° 7 em-3 ) (October) may also have influenced the observed dif-
ferences in depth distribution between seasons. These
Figure 6. Mean vertical distribution of the most abundant harpacti-
coids, C. perplexa and A. parvus. in the top 3 cm of the sediment, vertical distribution patterns are not totally consistent
during both emergence periods (October, station G). with those of Huys et al. (1986) who studied a subtidal
sandy station located in the Southern Bight of the North
76
Sea. These authors showed that C. perplexa migrated Emergence into the water column and vertical
towards the sediment surface in March and stayed close distribution patterns ofmeiobenthic copepods
to the surface throughout spring and summer (90% of
the population in the upper 2 cm), and migrated deep- During the October sampling, the emerging fraction
er in autumn. In the present study, the mean vertical was estimated to be 15 to 33% (postsunsetemergence)
distribution patterns observed in March and in Octo- and 6 to 21 % (pre-sunrise emergence) of the copepod
ber (Figure 2) support these results whereas the June assemblage. These results are consistent with those of
vertical distribution does not. This was probably due Walters & Bell (1994) who observed that 7 to 31 %
to the difference of summer temperatures between the of the copepod assemblage emerged during the post-
Southern Bight of the North Sea and Arcachon Bay. sunset period.
The authors related vertical movements of C. perplexa As expected, active emergence of meiobenthic
with breeding periods, indicating that during extensive copepods into the water column was associated with
reproductive activity (spring and summer) the cope- light changes over the diel period (Figure 4). Emer-
pod was mainly found in the upper millimeters of the gence throughout the night and particularly around
sediment. Considering that C. perplexa shows 2 or sunset has previously been well established (Walters,
3 reproductive periods occurring in spring, summer 1988, 1991; Walters & Bell, 1994). In the present study
and late summer to early autumn (Castel & Lasserre, the time of emergence was species-specific. Indeed,
1977; Ceccherelli & Mistri, 1991), and considering the migrating species A. normani and Paradactylopo-
that breeding activity is related to temperature (Harris, dia sp. were found in traps sampled only after sunset,
1972c), it is likely that in the present study, reproduc- whereas A. parvus was found in traps sampled during
tive activity also influenced the vertical migration of both emergence phases. It may be deduced that A. nor-
C. perplexa through the sediment. This is supported mani and Paradactylopodia emerged only during the
by the fact that the nauplius of C. perplexa is plank- first part of the night. Another possibility was that both
tonic (Vincx & Heip, 1979), suggesting that ovigerous copepods swam for a long period (from around 19:00-
females migrate to the sediment surface for laying. 24:00 to 06:00) in the water column. Traps placed
Oxygen concentrations also showed diel variations above the sediment at 06:00 would not capture them
which were linked to temperature variations which if the copepods were already swimming through the
both responded to the diel cycle of light intensity (Fig- water column. Such long excursions of meiobenthic
ure 3). Indeed, oxygen concentrations measured in the copepods in the water column have been observed. For
water column reflected the photosynthetic activity of example, Bell et a!. (1989) determined that an emerg-
microphytes which was obviously more intensive dur- ing copepod (Zausodes arenicolus) remained in the
ing days than during nights. There were no marked water column of field aquaria for more than 9 hours.
changes in mean oxygen concentrations between sea- Using experimental field aquaria, these authors also
sons (March: 4.4-8.3 mg 1-1, June: 3.4-1l.9 mg 1-1, showed that the composition of the migrating species
October: 5.4-9.7 mg 1-1), thus, seasonal changes in assemblage determines the time, the duration and/or
the vertical distribution of C. perplexa could not be the number of emergence periods occurring during the
attributed to this factor. Over diel periods we found no night.
evidence for an effect of diel variations in oxygen con- We found evidence which related the emerging
centrations on the vertical distribution of this copepod behaviour with the vertical distribution pattern of cope-
even if the copepod tended to move up during (June) or pods in the sediment. At first, emerging copepods
after (March and October) the decrease of oxygen con- (Figure 5) and non-emerging copepods (population of
centration at night. However, considering that oxygen C. perplexa) (Figure 2) showed different vertical dis-
is a key factor for the depth distribution of meiofauna tribution patterns over the diel period (October, sta-
(McLachlan, 1978; Meyers et a!., 1987), it is possi- tion G). The emerging assemblage was in general
ble that the diel vertical distribution of C. perplexa found mainly in the first cm of the sediment, whereas
responded to the combined effects of diel variations of C. perplexa showed a deeper vertical distribution. This
temperature and oxygen concentrations. suggests that emergence into the water column may
mainly concern surface dwelling copepods. Secondly,
the depth distribution of A. parvus only (the dominant
species emerging at sunrise) tended to change during
emergence phases. Density of this copepod tended to
77
decrease in the top first cm while it was emerging into gan, P. O. Ceidigh & P. J. S. Boaden (eds), Biology of benthic
the water column (Figure 6, second emergence phase, organisms. Pergamon Press: 129-146.
Castel, J., P. Caumette & R. Herbert, 1996. Eutrophication gradients
06:00-07:00). Consequently, it seems that both the ver- in coastal lagoons as exemplified by the Bassin d' Arcachon and
tical occupation of space and diel changes of the ver- the Etang du Prevost. Hydrobiologia 329: xi-xxx.
tical distribution, particularly between day and night, Ceccherelli, V. U. & M. Mistri, 1991. Production of the meiobenthic
may be related in part to the swimming behaviour of harpacticoid copepod Canuella perplexa. Mar. Ecol. Prog. Ser.
68: 225-234.
harpacticoids. Ceccherelli, V. U., M. Mistri & P. Franzoi, 1994. Predation impact
on the meiobenthic harpacticoid Canuella perplexa in a lagoon
of the Po River Delta, Italy. Estuaries 17: 283-287.
Acknowledgments Coull, B. c., M. A. Palmer & P. E. Myers, 1989. Controls on
the vertical distribution of meiobenthos in mud: field and flume
studies with juvenile fish. Mar. Ecol. Prog. Ser. 55: 133-139.
We are grateful to Dr I. Auby, L. P. Souza-Santos, Escaravage, V. & J. Castel, 1989. Application de la notion de con-
E. Antajan, and Y. Dubau for assistance collecting finement aux peuplements meiobenthiques des lagunes endiguees
du Bassin d' Arcachon (Cote atlantique). Acta Oecol. Gener. 10:
the samples throughout mosquito-infested nights. To
1-17.
M. Parra for making emergence traps and to A. M. Cas- Escaravage, v., M. E. Garcia & J. Castel, 1989. The distribution of
tel for assistance in sorting the trap contents. To meiofauna and its contribution to detritic pathways in tidal flats
S. Bourgues, L. P. Souza-Santos and P. 1. P. Santos (Arcachon Bay, France). In J. D. Ros. (ed.), Topics in marine
Biology, Scient. Mar. 53: 551-559.
for their advice and discussions concerning sampling,
Harris, R. P., I 972a. The distribution and ecology of the interstitial
laboratory processing, and earlier draft of the manu- meiofauna of a sandy beach at Whitsand Bay, east Cornwall. J.
script. We thank Dr W. Arrnonies and Dr I. Ferrari for mar. bioI. Ass. U.K. 52: 1-18.
critical comments which were helpful to improve the Harris, R. P., 1972b. Seasonal changes in the meiofauna population
of an intertidal sand beach. J. mar. bioI. Ass. U.K. 52: 389-403.
quality of the manuscript. We are also very grateful
Harris, R. P., I 972c. Reproductive activity of the interstitial copepods
to G. Hello and Dr D. T. Welsh for their assistance of a sandy beach. J. mar. bioI. Ass. U.K. 52: 507-524.
with the English version. This work was supported Heip, c., 1973. Partitioning of the brackish water habitat by copepod
by a E.U. Environment programm (CLEAN contract species. Hydrobiologia 41: 189-198.
Hicks, G. R. F., 1986. Distribution and behaviour of meiofaunal
EV5V-CT92-0080). copepods inside and outside seagrass beds. Mar. Ecol. Prog. Ser.
31: 159-170.
Huys, R., R. L. Hennan & c. Heip, 1986. Seasonal fluctuations in
References vertical distribution and breeding activity of a subtidal harpacti-
coid community in the southern bight, North Sea. Neth. J. Sea
Res. 20: 375-383.
ArIt, G., 1973. Vertical and horizontal microdistribution of the meio-
Jensen, P., 1983. Meiofauna abundance and vertical zonation in a
fauna in the Greifswalder Bodden. Oikos Suppl. 15: 105-111.
sublittoral soft bottom, with a test of the Haps corer. Mar. BioI.
Annonies, W., 1988. Hydrodynamic factors affecting behaviour of
74: 319-326.
intertidal meiobenthos. Ophelia 28: 183-193.
Joint, I. R., J. M. Gee & R. M. Warwick, 1982. Detennination of
Annonies, W., 1989. Meiofaunal emergence from intertidal sedi-
fine-scale vertical distribution of microbes and meiofauna in an
ment measured in the field: significant contribution to nocturnal
intertidal sediment. Mar. BioI. 72: 157-164.
planktonic biomass in shallow waters. HelgoHinder wiss. Meere-
Jonge, V. N. De & L. A. Bouwman, 1977. A simple density sepa-
sunters. 43: 29-43.
ration technique for quantitative isolation of meiobenthos using
Bell, S. S., G. R. F. Hicks & K. Walters, 1988. Active swimming in
the colloidal silica LUDOX-TM. Mar. BioI. 42: 143-148.
meiobenthic copepods of seagrass beds: species patterns and role
McIntyre, A. D. & D. J. Murison, 1973. The meiofauna of a flatfish
in reproductive behavior. Mar. BioI. 98: 351-358.
nursery ground. J. mar. bioI. Ass. U.K. 53: 93-118.
Bell, S. S., G. R. F. Hicks & K. Walters, 1989. Experimental investi-
McLachlan, A., 1978. A quantitative analysis of the meiofauna and
gations of benthic re-entry by migrating meiobenthic copepods.
the chemistry of the' redox potential discontinuity zone in a shel-
J. expo mar. BioI. Ecol. 130: 291-303.
tered sandy beach. Estuar. coast. mar. Sci. 7: 275-290.
Buffan, E. & J. Castel, 1993. Distribution of benthic microbiota in
McLachlan, A., P. E. D. Winter & L. Botha, 1977. Vertical and
Arcachon Bay and Prevost lagoon. C.L.E.A.N. progress report
horizontal distribution of sublittoral meiofauna in Algoa Bay,
part II: 73-89.
South Africa. Mar. BioI. 40: 355-364.
Buffan, E., L. P. Souza-Santos & J. Castel, 1993. Daily variations of
Meyers, M. B., H. Fossing & E. N. Powell, 1987. Microdistribution
vertical distribution and pigment gut content of the meiobenthic
of interstitial meiofauna, oxygen and sulfide gradients, and the
copepod Canuella perplexa in a brackish lagoon. C.L.E.A.N.
tubes of macro-infauna. Mar. Ecol. Prog. Ser. 35: 223-241.
progress report part II: 133-148.
Montagna, P. A., J. E. Bauer, D. Hardin & R. B. Spies, 1989. Vertical
Castel, J., 1992. The meiofauna of coastal lagoon ecosystems and
distribution of microbial and meiofaunal populations in sediments
their importance in the food weeb. Vie Milieu 42: 125-135.
of a natural coastal hydrocarbon seep. J. mar. Res. 47: 657-680.
Castel, J. & P. Lasserre, 1977. Colonisation et distribution spatiale
Palmer, M. A. & R. M. Molloy, 1986. Water flow and the vertical
des copepodes dans des lagunes semi-artificielles. In B. F. Kea-
distribution of meiofauna: a flume experiment. Estuaries 9: 225-
228.
78
Schaub, B. & H. van Gemerden, 1996. Sulfur bacteria in sediments Walters, K., 1988. Diel vertical migration of sediment-associated
of two coastal ecosystems: the Bassin d'Arcachon and the Etang meiofauna in subtropical sand and seagrass habitats. J. expo mar.
du Prevost, France. Hydrobiologia, this volume. Bio!. Bco!. 117: 169-186.
Service, S. K. & S. S. Bell, 1987. Density-influenced active dispersal Walters, K., 1991. Influences of abundances, behavior, species com-
ofharpacticoid copepods. J. expo mar. Bio!. Bco!. 114: 49-62. position, and ontogenic stage on active emergence of meiobenthic
Smith, L. D. & B. C. Coull, 1987. Juvenile spot (Pisces) and grass copepods in subtropical habitats. Mar. Bio!. 108: 207-215.
shrimp predation on meiobenthos in muddy and sandy substrata. Walters, K. & S. S. Bell, 1986. Diel patterns of active vertical
J. expo mar. Bio!. Bco!. 105: 123-136. migration in seagrass meiofauna. Mar. Bco!. Prog. Ser. 61: 111-
Thielemans, L. K. T. & c. Heip, 1984. The response of a harpacticoid 124.
copepod community to sediment disturbance in a semi-enclosed Walters, K. & S. S. Bell, 1994. Significance of copepod emergence
lagoon. Hydrobiologia 118: 127-133. to benthic, pelagic, and phytallinkages in a subtidal seagrass bed.
Vincx, M. & C. Heip, 1979. Larval development and biology of Mar. Bco!. Prog. Ser. 108: 237-249.
Canuella perplexa T. and A. Scott, 1893 (copepoda harpacticoi- Yingst, J. Y., 1978. Patterns of micro and meiofauna abundance
da). Cah. Bio!. Mar. 20: 281-299. in marine sediments, measured with the adenosine triphosphate
assay. Mar. BioI. 47: 41-54.
Hydrobio[ogia 329: 79-89, 1996. 79
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Key words: Feeding ecology, meiofauna, harpacticoid copepods, coastal lagoons, phototrophic bacteria, microalgae
Abstract
Laboratory experiments were undertaken using Amonardia normani and Schizopera cf compacta, two meiobenthic
harpacticoid copepods commonly found in coastal lagoons. The first experiments were designed to determine if the
phototrophic sulfur bacteria Chromatium gracile can be ingested by these copepods and at what concentrations.
Egestion rate was used as an index of feeding rate. The response of the egestion rate, expressed in numbers of
faecal pellets produced by copepod per day, as a function of bacterial concentration followed the functional model.
A. normani attained constant feeding rates from the bacterial concentration of 1 x 107 cells ml- I (5 f..Lg C ml- I )
onwards, S. cf compacta attained constant feeding rates from 2.6 x 107 cells ml- I (13 f..Lg C ml- I ) onwards.
The faecal pellet volume changed significantly (p<0.05) between food concentrations for A. normani but not for
S. cf compacta (p>0.05). In order to investigate the effect of the phototrophic bacterial diet on the population
dynamics of A. normani three groups of nauplii were maintained at 2 x 107 cells ml- I and observed every day. The
mortality of these nauplii was very high compared to those maintained on a diatom diet (Nitzschia constricta); only
in one of the groups did some copepodites develop but no adults were ever observed. Adults fed on bacteria did
not have different (p>0.05) survival rates compared to those fed on diatoms, nevertheless, the number of nauplii
produced was significantly less (p<0.05) on the bacterial diet. These results lead us to suggest that although the
phototrophic sulfur bacteria (Chromatium gracile) can be ingested by both copepod species it cannot sustain the
full development of the A. normani population. Thus, a bloom of phototrophic sulfur bacteria does not seem to be
a favourable situation for opportunistic benthic copepods to colonize eutrophic coastal lagoons after a dystrophic
crisis.
Relationship between bacterial concentration and laboratory were separated from the maintenance cul-
egestion rate of copepods ture. Each group received a bacterial or a diatom sus-
pension at concentrations which resulted in maximum
A single cope pod female was placed in a 23.7 cm 2 glass egestion rates, i e., 2 x 107 cells ml- I for C. gracile
dish containing 10 ml of bacterial cell suspension of (see Results) and 0.13 p,g chlorophyll a ml- I (about
known concentration. Five replicate dishes were used 7 x 105 cells ml- I ) for N. constricta (Souza-Santos, in
for each bacterial concentration and for controls (con- preparation). One day later, the hatched nauplii were
sidered here as copepods in sterile, 0.22 Ilm filtered placed individually in the cavities of sterile micro-
seawater, or without food). These dishes were incubat- plates, containing about 0.3 ml of the food suspension.
ed for 24 h, accounting for the possibility of endoge- To reduce evaporation, the cavities situated at the bor-
nous diel cycles of feeding rate (Souza-Santos et aI., der of the microplates were filled with distilled water
1995), at the same conditions as copepod cultures. At and the micro-plates were placed inside aquaria with
the end of the incubation period and after checking for a humid atmosphere. Two groups of 45 and 43 nauplii
copepod survival, the content of each dish was pre- were fed on N. constricta and three groups of 76, 49
served in 4% v/v formalin for further analysis. and 50 nauplii were fed on C. gracile.
The number of faecal pellets produced by each All copepod stages were observed every day until
copepod during 24 h was counted under a stereo- they died. The criterion used to identify the dead ani-
microscope. To estimate the total volume of faecal mals was the absence of mobility after mechanical
pellets produced, the length and width of 20 to 30 pel- stimulation. Once and if the copepodite stage was
lets for each food concentration were measured using attained, the animals were put together in a sterile
a camera lucida. The volume of pellets was estimated glass dish containing 10 ml of the food suspension.
assuming a cylindrical shape. After laying the first egg sac, each female was isolated
Experiments were carried out twice with A. nor- with a male in a sterile plastic flask containing 10 ml of
mani. At the first experiment the feeding rates the food suspension. As soon as a new brood hatched,
increased continually with food concentrations. the adult couple was transferred to a new flask and the
Regarding the functional model, it means that the nauplii released were preserved in 4% v/v formalin and
food concentrations were lower than the 'critical' stained with Rose Bengal to be counted.
food concentration, defined here as the concentration Food in microplates (nauplius development) was
from which no significant increase in feeding rates is not renewed. During the copepodite stages the food
observed. In the second experiment higher food con- suspension was renewed every two days and the living
centrations were tested and the feeding rates attained animals counted. During the reproductive period, the
the expected plateau. This resulted in different ranges food suspension was changed every time a new brood
of tested bacterial concentrations between the two was produced (between two and three days) or each
copepod species (Table 1). five days for the post-reproductive period.
old suspensions were preserved in 4% v/v formalin significantly up to a concentration of 1 x 107 cells
and stained with Rose Bengal for further counting of ml- I or 5.2 /Lg C ml- I (Table 1, Figure 1),
hatched nauplii. from this concentration onwards the number of fae-
cal pellets did not change significantly (ANOVA,
F=2.2, d.f.=4119, p=0.1095). The pellet volume,
Results although, decreased significantly (ANOVA, F=4.43,
d.f.=41104, p=0.0024) from 4 x 107 cells ml- I to
Relationship between bacterial concentration and 10 x 107 cell ml-I. Only from 2.6 x 107 cells ml- I
egestion rate of copepods and above did the number of faecal pellets produced
by S. cf compacta differ from control values (Table 1).
The carbon content of C. gracile strain CE2201 cell The number of faecal pellets produced from 2.6 x 107
was estimated as 5.19 x 10- 1 pg from the measure cells ml- I or 13.6 /Lg C ml- I onwards did not change
of cell volume (4.72 /Lm 3 ) and as 5.27 x 10- 1 pg by significantly (Figure 2) (ANOVA, F= 1.0, dJ. = 2112,
weighing the bacterial culture. The mean value of p = 0.3961). The mean pellet volume did not change
0.52 pg C cell- I was thus used for the transformation significantly (ANOVA, F = 0.54, d.f. = 3/98, p = 0.66)
of cell concentration in biomass. between bacterial concentrations at least between
The tested food items sedimented on the bottom of 2.6 x 107 cells ml- I and 26 x 107 cells ml- I .
the flasks; therefore the copepods could feed on them
without another kind of substrate. Effect of bacte rial diet on the population dynamics of
Under the control conditions, without food, cope- Amonardia normani
pods normally produce from 0 to 3 typical faecal pel-
lets per copepod per day (Table 1) or seldom many First experiment
(about 20) very small, transparent pellets (not consid- The nauplii of A. norrnani are strictly benthic and could
ered here). This faecal pellet production should be a feed on both C. gracile and N. constricta, based on
result of the feeding activity in the maintenance cul- the presence of many faecal pellets in cavities of the
tures. micro-plate.
A. normani egestion rates were different from The survival of A. normani was very low in the
the control values from 3 x 104 cells ml- I onwards three groups of nauplii fed on C. gracile (Figure 3)
(Table 1). The number of faecal pellets produced compared with those fed on the diatom N. constricta
and the mean pellet volume for A. normani increased (Figure 4). The lethal time for 50% of the population
83
5
7 (xlO )
150
0
6 •
,-...
120 C'?
.....
5 S
I
>,
'"
"C
::t
--
('.)
..... S
-
90 4 ;::l
I
Q, 0
;;-
I -
0
('.) ....,
....,rn
-
3 III
Q.) 60 III
0-
!
Q.)
2 .::
Q,
z
0
S
'"
III
30
1
O~~--~~--~~--J-~--~--~~~O
o 2 4 6 8 10 12 14 16 18 20 22
7
concentration (cells.ml-l) (xlO )
Figure I. Variation of the number of faecal pellets produced (No. pellets. cop-I day-I) and mean faecal pellet volume (ILm 3 ) of Amonardia
normani in relation to Chromatium gracile concentration (cells ml- I ). Vertical bars denote the confidence intervals of the means (p = 0.05).
ranged from 7 to 10 days for the bacterial diet and the diatom diet and the bacterial diet: 13.8 and 0.41
was about 37 days for the diatom diet. Only in one nauplii . female- t . day-t, respectively.
group fed on bacteria, nauplii developed to copepodites
(C 5 stage) but no adults were observed. A. normani
presented a complete life-cycle on the diatom diet with Discussion
an intrinsic rate of increase (ro) of 0.3 (Souza-Santos,
in preparation). Data on the effect of food concentration on the phys-
iological rates of meiobenthic copepods are scarce.
Second experiment The results of Heinle et al. (1977) indicated that Coul-
The survival of adult A. normani fed during one week lana canadensis produced more eggs at higher algal
on N. constricta (95%) or on C. gracile (87%) was concentration. Rieper (1978,1985) and Rieper & Flo-
not significantly different (Student's t-test, t= 1.86, tow (1981) showed that the feeding rates of both Tisbe
d.f. =4, P =0.14) at the end of the experiment (Fig- holothuriae and Paramphiascella vararensis increased
ure 5). On the other hand, fecundity showed different with food (bacteria, algae and ciliate) concentrations.
variations through time. In the N. constricta diet the Heteropsyllus pseudonunni had higher egg production
fecundity increased significantly with time (regression at increasing algal concentration but not at increas-
analysis - R 2 =76%, p=O.OI). On the C. gracile diet ing detritus concentration (Ustach, 1982). However,
the fecundity decreased significantly with time (regres- these studies did not test a large range of food con-
sion analysis - R 2 = 60%, P =0.04). At the end of the centration and did not discuss the existence of a 'criti-
experiment the fecundity was significantly different cal' food concentration above which the physiological
(Student's t-test, t= 22.4, dJ. =4, p<O.OOOI) between rates (such as feeding and egg production) are inde- .
pendent of the food concentration. Only for Coullana
84
5
(x10 )
35 3.5
-----
0 30 3.0 •
.......,
C'?
..... S
2.5 .......,
::l.
>. 25
I
CIl
C!J
-
'0
..... S
/
I 20 2.0 ;j
0.. 0
>
-
0
....,
-
C)
-
....,00 15 1.5 C!J
C!J C!J
0..
C!J
!
0.. ~
1.0
Z
0 10 ! CIl
C!J
S
5 0.5
0 0.0
0 5 10 15 20 25 30
concentration (cells.ml-1) (xlO 7 )
Figure 2. Variation of the number of faecal pellets produced (No. pellets. cop-l day-I) and mean faecal pellet volume (/Lm 3 ) of Schizopera
ct: compacta in relation to Chromatium gracile concentration (cells ml- 1 ). Vertical bars denote the confidence intervals of the means (p =0.05).
110 20
100 a
0
0 100
80 15
....... 90
'ct. 60
'-'
~
.3
tij ~
80 10~
..
.~ > ;:l
.f;
...'"
(.)
~ 40
V)
;:l
Ul
70
5
20
60 t---9~\ ~
..
0
12 34 45 56 67 78 50
Days 0 2 3 4 5 6 7 8
f
concentrations of 1 x 107 cells ml- l (5.2 ftg C ml- l ) ~ 10]
80
and 2.6 x 107 cells ml- l (13.6 ftg C ml--l), for each
..
>
.f; ;:l
a
(.)
species respectively.
70
'"
r..
As proposed by Frost (1972) and Mullin et al.
(1975) three models were adjusted to the results f ¢/ ·1
5
80
60 ...
observed
-e-
-1.ge-7(C-1.4e5)
40 Ivlev F=88 (I-e
~
F= 93 (C-9.6e4)
Michaelis-Menten
20 -£- 4e6+(C-9.6e4)
rectilinear F=2 + 8e-6 C
F=IOI - 2e-7 C
0
O.OE+OO 5.0E-;-07 l.OE+08 1.5E+08 2.0E+08
Bacterial cone. (cells.ml-l)
...
40,-----------------------------------------~
observed
....... -e-
~
~ 30 Ivlev
....... -fr-
I
ci. Michaelis-Menten
o(,) -£-
vi
rectilinear
] 20
Q)
0-
~
(,)
(,)
<S 10
o
Z
o .-~---.-------.------.------.------,-------~
O.OE+OO S.OE+OS 1.0E+06 l.SE+06 2.0E+06 2.SE+06 3.0E+06
Bacterial cone. (cells.ml-l)
Figure 6. Feeding rates of Amonardia normani in relation to Chromatium gracile concentrations. observed data and expected curves estimated
by the Ivlev model, Michaelis-Menten model and the rectilinear model. a) all data, b) bacterial concentrations below the 'critical' one.
species and a functional response to these concentra- able to express the feeding rate by individual or by
tions. However, Montagna plotted a relative algal feed- body biomass per unit of time. Thus, we have replot-
ing rate (x 10- 4 h- 1) versus the total carbon added (see ted the results of Vanden Berghe & Bergmans (1981),
his Figure 1). This relative rate, although interesting now with the algal feeding rate expressed in /-lg C.
to measure meiofauna grazing impact on food stocks, ind- 1 day-l versus the total carbon added or the algal
is worthless in physiological studies where it is prefer- carbon added (Figure 7a and b). The linear regression
87
0.4 i ·--'1
j • a bacterial densities in the sediments of the fish ponds of
>-
I
I
T. ho/othuriae
'V • Certes were higherthan 107 cells ml- J during the three
~ 0.3 1I sampling periods (March, June and September 1993)
"0
.§ j
T. battagliai
• •
(Guyoneaud et aI., 1996). Caumette (1986) observed
t
()
Cl
II T. furcata
'V 'V phototrophic bacteria densities of 107 cells ml- J in the
;0.2
•
'V
~ I
sediment of Prevost Lagoon during the summer and
Cl
c::: 1 • 'V iii • in the water column the densities attained 108 cells
~Q) 0.1
u.
1.
i
'V
• ml- J during 'red waters'. These results indicate that
1 'V
the 'critical' phototrophic bacterial concentrations esti-
•
0 I mated in the present study can be found by meiobenthic
copepods in the field, making possible the exploitation
20 40 60 80 100 120 of this food at maximum rates in natural conditions.
Total carbon added (~g)
Variation of the faecal pellet volume as a function
0.4 of bacterial concentration differed for each copepod
• b species. S. cf compacta faecal pellet volume did not
>:
T. ho/othuriae
'V • change significantly between the tested concentrations
~ 0.3
"0
T. battagliai
• • (Figure 2), for A. normani the faecal pellet volume
r1r
c:::
(3 T. furcata increased up to the 'critical' concentration, did not
Cl
'V 'V change significantly up to 5.3 X 107 cells ml- J and
•
'V
were observed (though not counted) every day until impossible the population growth and colonization of
they died, but development to copepodites was not this lagoon, unless another food resource is available.
observed. Other meiobenthic copepods, such as P. vararensis
The adults that fed on C. gracile had similar sur- (Rieper, 1984), can grow using phototrophic bacteria
vival rates to those fed on diatoms and were able to as sole food source and may colonize lagoons during
produce nauplii, although in smaller number than in 'red waters'. On the other hand, the utilization of the
the diatom diet (Figure 5). The increase of fecundi- phototrophic sulfur bacteria as a food complement for
ty observed in the diatom diet during the experiment both tested copepods cannot be discarded.
can be a result of the lower density of copepods in the
experimental dishes as compared to the maintenance
dishes (Fava & Crotti, 1979) or can be due to a higher Acknowledgments
food supply (in the maintenance dishes the food con-
centration was not controlled and was renewed only The authors are grateful to P. Caumette and R. Guy-
once a week). The decrease of fecundity observed in oneaud for providing the original culture of Chro-
the bacterial diet can represent a delay of the response matium gracile strain CE2201 isolated from Certes
to diet change, meaning that the first broods could still sediments and their suggestions during all the work.
be related to the diatom diet in the maintenance cul- Sincere thanks to all staff of Prof. P. Caumette for prac-
tures. tical help during the preparation of bacterial culture
Caumette et al. (1983) demonstrated, by means of media. The authors thank G. Bachelet, M. Kirchner and
in situ gut-content analyses, that Acartia clausi fed V. U. Ceccherelli for critical reading ofthe manuscript.
on phototrophic bacteria (Rhodopseudomonas sp. and LPSS and PJPS acknowledge a CAPES postgraduate
Chromatium sp.) present in the water column of a studentship (Brazil). This study was partly supported
tropical lagoon. Laboratory experiments also showed by a CEC Environment programme (CLEAN contract
that this copepod could lay eggs when feeding on EV5V - CT92 - 0080).
three different strains of phototrophic bacteria, indi-
cating assimilation. Nevertheless, egg production was
lower than when the copepods were fed on algae References
(Caumette, 1985). Rieper (1984) tested Chromatium
vinosum and Rhodopseudomonas palustris as food Bougis. P., 1974. Ecologie du plancton marin. Tome II - Le zoo-
plancton. Masson, Paris, 200 pp.
for Tisbe holothuriae and Paramphiascella vararensis. Castel, J., 1979. Adaptation and reproductive cycle of the harpacti-
For P. vararensis both phototrophic bacteria represent- coid copepod Amonardia normani (Brady, 1872) in semi-
ed adequate food resource for its complete life-cycle. enclosed lagoons of Arcachon Bay, France. In E. Naylor and
In the case of T. holothuriae, a species known to eat G. Hartnoll (eds), Cyclic phenomena in marine plants and ani-
mals. Pergamon Press, Oxford and New York: 131-138.
on several bacteria strains (Rieper, 1978; Guerin & Castel, J., 1992. The meiofauna of coastal lagoon ecosystems and
Rieper-Kirchner, 1991), the C. vinosum diet result- their importance in the food web. Vie Milieu 42: 125-135.
ed in no egg production and the R. palustris diet in Castel, J. & P. Lasserre, 1977. Colonisation et distribution spatiale
an incomplete life-cycle. Decho & Castenholz (1986) des copepodes dans des lagunes semi-artificielles. In B. F. Kee-
gan, P. O. Ceidigh & P. J. S. Boaden (eds), Biology of benthic
proposed that Thiocapsa sp (even with sulfur globules) organisms. Pergamon Press Oxford and New York: 129-146.
is an important food resource for Mesochra lilljebor- Caumette, P., 1985. Developpement des bacteries phototrophes et
gi based on distributional data and laboratory feeding sulfato-redutrices dans des lagunes peu profondes et des lagunes
experiments. stratifiees. Etude de leur rOle dans Ie cycle du soufre et dans
la production de biomasse. Doctoral Thesis. University of Aix-
In conclusion, the meiobenthic copepods Amonar- Marseille III. France, 325 pp.
dia normani and Schizopera cf compacta can ingest Caumette, P., 1986. Phototrophic sulfur bacteria and sulfate-
the phototrophic bacteria Chromatium gracile. Their reducing bacteria causing red waters in a shallow brackish coastal
lagoon (Prevost Lagoon, France). FEMS Microbiol. Ecol. 38:
higher feeding rates were observed from the bacterial
113-124.
concentration of 107 cells ml- 1 onwards. Neverthe- Caumette, P., 1992. Bacterial communities in coastal lagoons. An
less, it can be suggested that adults of the opportunistic overview. Vie Milieu 42: 111-123.
species Amonardia normani entering shallow coastal Caumette, P., M. Pagano & L. Saint-Jean, 1983. Repartition verticale
du phytoplancton, des bacteries et du zooplancton dans un milieu
lagoon from the sea during blooms of phototrophic stratifie en Baie de Bietri (Lagune Ebrie, Cote d'Ivoire). Relations
sulfur bacteria will produce a small number of nau- trophiques. Hydrobiologia 106: 135-148.
plii that will not develop, making it difficult or even
89
Conover, R. 1., 1966. Factors affecting the assimilation of organic Pfennig, N. & H. G. Triiper, 1981. Isolation of members of the fam-
matter by zooplankton and the question of superfluous feeding. ilies Chromatiaceae and Chlorobiaceae. In Starr M. P., Stolp H ..
Limno!. Oceanogr II: 1066--1071. Triiper H. G., Balows A., Schlegel H. G. (eds), The Prokaryotes.
Decho, A. W. & R. W. Castenholz, 1986. Spatial patterns and feed- Springer Verlag, Berlin: 279-289.
ing of meiobenthic harpacticoid cope pods in relation to resident Rieper, M .. 1978. Bacteria as food for marine harpacticoid copepods.
microbial flora. Hydrobiologia 131: 87-96. Mar. Bio!. 45: 337-345.
Fava, G. & E. Crotti, 1979. Effect of crowding on nauplii produc- Rieper, M., 1982. Feeding preferences of marine harpacticoid cope-
tion during mating time in Tisbe clodiensis and T. holothuriae pods for various species of bacteria. Mar. Eco!. Prog. Ser. 7:
(Copepoda, Harpacticoida). Helgoltinder wiss. Meeresunters. 32: 303-307.
466-475. Rieper, M., 1984. Relationships between bacteria and marine cope-
Frost, B. W., 1972. Effects of size and concentration of food parti- pods. In: Bacteriologie Marine. CNRS. Paris: 169-172.
cles on the feeding behavior of the marine planktonic copepod Rieper, M., 1985. Some lower food web organisms in the nutrition
Calanus pacijicus Limno!. Oceanogr. 17: 805-815. of marine harpacticoid copepods: an experimental study. Hel-
Giere, 0., 1993. Meiobenthology. The microscope fauna in aquatic goltinder. wiss. Meeresunters. 39: 357-366.
sediments. Springer-Verlag, Berlin, 328 pp. Rieper, M. & C. Flotow, 1981. Feeding experiments with bacteria,
Guerin, J.-P. & M. Rieper-Kirchner, 1991. Influence of three bacteria ciliates and harpacticoid copepods. Kieler Meeresforsch., Son-
strains on the population dynamics of Tisbe holothuriae (Cope- derh. 5: 370-375.
poda, Harpacticoida). Helgoltinder. wiss. Meeresunters. 45: 493- Schiemer, F., 1982. Food dependence and energetics of free liv-
511. ing nematodes. I. Respiration, growth and reproduction of
Guyoneaud, R., R. Matheron, R. Baulaigue, K. Podeur, A. Hirschler Caenorhabditis briggsae (Nematoda) at different levels of food
& P. Caumette, 1996. Anoxygenic phototrophic bacteria in supply. Oecologia (Berl.) 54: 108-121.
eutrophic coastal lagoons of the french mediterranean and atlantic Schiemer, F., A. Duncan & R. Z. Klekowski, 1980. A bioenergetic
coasts (Prevost Lagoon, Arcachon Bay). Hydrobiologia 329 study of a benthic nematode, Plectus palustris de Man 1881.
(Dev. Hydrobio!. 117): 33-43. throughout its life cycle. II. Growth, fecundity and energy budgets
Harris, R. P., 1977. Some aspects of the biology of the Harpacti- at different densities of bacterial food and general ecological
coid Copepod, Scottolana canadensis (Willey), maintained in considerations. Oecologia (Berl.) 44: 205-212.
laboratory culture. Chesapeake Sci. 18: 245-252. Shaw, B. A., P. J. Harrison & R. J. Andersen, 1994. Evaluation of
Heinle, D. R., R. P. Harris, J. F. Ustach & D. A. Flemer, 1977. the copepod Tigriopus calit(Jrnicus as a bioassay organism for
Detritus as food for estuarine copepods. Mar. Bio!. 40: 341-353. the detection of chemical feeding deterrents produced by marine
Karl, D. M., 1986. Determination of in situ microbial biomass, via- phytoplankton. Mar. Bio!. 121: 89-95.
bility, metabolism, and growth. In J. S. Pointdexter & E. R. Lead- Souza-Santos, L. P., J. Castel & P. J. P. dos Santos, 1995. Feeding
better (eds), Bacteria in nature (Vo!. 2) Methods and special appli- rate cycle of the epibenthic harpacticoid copepod Harpacticus
cations in bacterial ecology. Plenum Pre,s, New York - London: fiexus: Laboratory experiments using faecal pellets counts. Vie
85-176. Milieu 45: 75-83.
Lonsdale, D. J. & J. S. Levinton, 1989. Energy budgets oflatitudinal- Ustach, J. F., 1982. Algae, bacteria and detritus as food for the
Iy separated Scottolana canadensis (Copepoda: Harpacticoida). harpacticoid copepod, Heteropsyllus pseudonunni Coull and
Limno!. Oceanogr. 34: 324-331. Palmer. J. expo mar. Bio!. Eco!. 64: 203-214.
Montagna, P. A., 1995. Rates of metazoan meiofaunal microbivory: Vanden Berghe, W. & M. Bergmans, 1981. Differential food prefer-
a review. Vie Milieu 45: 1-9. ences in three co-occurring species of Tisbe (Copepoda, Harpacti-
Montagna, P. A. & G. F. Blanchard & A. Dinet, 1995. Effect of pro- coida). Mar. Eco!. Prog. Ser. 4: 213-219.
duction and biomass of intertidal microphytobenthos on meio- Williams, T. D. & M. B. Jones, 1994. Effects of temperature and
faunal grazing rates. 1. expo mar. BioI. Eco!. 185: 149-165. food quantity on postembryonic development of Tisbe battagliai
Mullin, M. M., 1963. Some factors affecting the feeding of marine (Copepoda: Harpacticoida). J. expo mar. Bio!. Eco!. 183: 283-
copepods of the genus Calanus. Limno!. Oceanogr. 8: 239-250. 298.
Mullin, M. M., E. F. Stewart & F. J. Fuglister, 1975. Ingestion by
planktonic grazers as a function of concentration offood. Limno!.
Oceanogr. 20: 259-262.
Parsons, T. R., R. J. LeBrasseur & J. D. Fulton, 1967. Some observa-
tions on the dependence of zooplankton grazing on the cell size
and concentration of phytoplankton blooms. J. Oceanogr. Soc.
Jap. 23: 10-17.
PART II
Eutrophication effects on biogeochemistry
of nitrogen and sulfur
in coastal lagoons
Hydrobi%gia 329: 93-103, 1996. 93
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (CLE.AN.).
©1996 Kluwer Academic Publishers.
Abstract
Growth of the seaweed Viva rigida C. Agardh was investigated in relation to biomass densities, internal nutrient
pools and external nutrient supply. Research was carried out from 23 March to 5 July 1994 in the Sacca di Goro
(Po Delta, Northern Italy), whose south-eastern part was covered by extensive mats of Viva rigida. Two types of
field experiments were conducted by incubating VIva thalli inside large cages. In the first experiment, beginning
on 23 March, 100 g of wet thalli were placed into the cages, allowed to grow for two weeks, then collected and
replaced. This procedure was repeated 8 times over the study period. In the second experiment, VIva thalli were
left inside the cages and collected at selected time intervals (14, 27, 41, 64 and 76 days) in order to simulate the
effects of increased density on growth and nutrient storage.
We recorded specific growth rates (NGR) ranging from 0.025 to 0.081 d- 1 for a period up to two months in the
repeated short-term experiments performed at relatively low initial algal densities (300-500 g AFDW m- 3 ). These
NGR resulted significantly related to dissolved inorganic nitrogen (DIN) in the water column. Tissue concentrations
of total Kjeldahl nitrogen (TN) were almost constant, while extractable nitrate decreased in a similar manner to
DIN in the water column. Total phosphorus showed considerable variation, probably linked to pulsed freshwater
inflow.
In the long-term incubation experiment, NGR of Viva was inversely related to density. Internal concentrations of
both total P and TN reached maximum values after one month; thereafter P concentration remained almost constant,
while TN decreased below 2% w/w (by dry weight). The TN decrease was also accompanied by an abrupt decrease in
nitrate tissue concentration. The biomass incubated over the two month period suffered a progressive N limitation as
shown by a decreasing N:P ratio (49.4 to 14.6). The reciprocal control of Viva against biogeochemical environment
and viceversa is a key factor in explaining both resource competition and successional stages in primary producer
communities dominated by Viva. However, when the biomass exceeds a critical threshold level, approximately
1 kg AFDW m- 3 , the macroalgal community switches from active production to rapid decomposition, probably as
a result of selfshading, biomass density and development of anaerobic conditions within the macroalgal beds.
of maximum growth appears to be determined most- line, while sandy-mud prevails in the eastern area (Dal
ly by nitrogen availability (Lapointe & Tenore, 1981; Cin & Pambianchi, 1991). Freshwater flows into the
Duke et aI., 1989a,b; Fujita et aI., 1989; Sand-Jensen, northeastern part of the lagoon from the Po di Goro,
1991; Couthino & Zingmark, 1993). Compared with while the western and central zones are affected mainly
ammonium, little work has focused on the effects of by the Po di Volano canal. In 1992, the Po di Volano
nitrate upon growth and biochemical composition of canal discharged about 3100 tons of nitrate, 560 tons
Chlorophyceans (Lapointe & Tenore, 1981; O'Brien of ammonium, and 46 tons of total phosphorus into the
& Wheeler, 1987; Corzo & Niell, 1992). However, it Sacca di Goro (Alvisi et aI., 1993).
has been proposed that nitrate is the main macronu-
trient affecting the growth cycle of other macroal-
gae (Chapman & Craigie, 1977; Da Costa Braga & Materials and methods
Yoneshigue-Valentin, 1994). Several Phaeophyceans
may accumulate nitrate in vacuoles to very high con- Research was carried out from 23rd March to 5th July
centrations thereby gaining an additional nitrogen pool 1994 in the southeastern part of the lagoon where the
in addition to amino acids and proteins (Chapman & mean water depth is l.2 m (station M, Figure 1). Two
Craigie, 1977; DeBoer, 1981). types of experiments were conducted incubating VIva
The excessive growth of green macroalgae (green thalli inside cy lindrical enclosures positioned at a depth
tides) is causing severe dystrophic crises in a number of 0.5 - 0.7 m in the lagoon. Enclosures were built using
of lagoons located along the northern Mediterranean a plastic fencing net (mesh =10 mm) to obtain cages
coast (Amanieu et aI., 1975; Izzo & Hull, 1991; Sfriso with no limits to water exchanges.
et aI., 1992). Amongst these, the Sacca di Goro has
been experiencing summer dystrophic episodes due Field observations
to VIva rigida C. Ag., whose excessive growth and
decomposition has been related to nitrate availability In addition to the experiments, three to six macroalgal
(Viaroli et aI., 1992; 1993; 1995). samples were collected around the experimental site
The data presented in this paper are part of a long- within a 0.5 km 2 area. The macroalgal biomass was
term research project on the eutrophication ofthe Sacca harvested with a rake-like device over an area ranging
di Goro. We report field experiments which simulate from 2 to 10 m 2 .
natural growth conditions of VIva. The experiments Commencing on the 23rd March, water samples
aimed (1) to analyze biomass parameters of VIva as were collected approximately twice a week at site M
related to biomass densities, internal nutrient pools at a depth of about 0.5 m with a Ruttner bottle. Nitrite
and external nutrient supply and (2) to evaluate the was determined spectrophotometrically after diazota-
ability of VIva to store nutrients over short and long tion; nitrate was measured as nitrite after cadmium
time periods. The latter objective is especially relevant reduction (APHA, 1975); ammonium was determined
in order to assess to what degree VIva may act as a by the indophenol-blue method (Koroleff, 1970) and
nutrient sink. soluble reactive phosphorus (SRP) by the ascorbic acid
method (Valderrama, 1977).
Water temperature, salinity, and dissolved oxygen
Study area were recorded approximately every four hours from an
in situ buoy equipped with specific probes (Hydronaut
A typical estuarine environment of the Po Delta is rep- 601). Total daily irradiance was measured by hemi-
resented by shallow-water embayments locally called spherical pyroheliometer at the meteorological station
'sacca'. The Sacca di Goro lies south of the Po di in Volano, near to the mouth of the Po di Volano (cour-
Goro, the southernmost deltaic branch, at approxi- tesy of the Ufficio Meteorologico Regionale, Regione
mately 44°47'-44°50' Nand 12 °15'-12°20' E. The Emilia Romagna).
lagoon has an area of 26 km 2 and an average depth of
about 1.5 m; it is connected to the sea by a 1.5 km- Long-term experiment
wide mouth (Figure 1). The bottom is flat and the
sediment is composed of typical alluvial mud with a From 23rd March to 7th June macroalgal thalli were
high clay and silt content in the northern and central incubated in cylindrical enclosures in the field and col-
zones; sand is more abundant near the southern shore- lected at selected time intervals to simulate the effects
95
, .~. ... :
" ~. .. .. to .. ••
ADRIATIC SEA
o 2K",
. , .....
~ <J
Figure 1. General layout of the Sacca di Goro, and position of the sampling site (st. M.). Shaded area indicates the maximum spreading of VIva
in June 1994.
of increased density upon growth and nutrient stor- March to 5th July, eight subsequent experiments were
age. On 23rd March, 25 cylindrical enclosures (mean carried out incubating the macroalga inside cylindrical
~ = 30 cm; height = 50 cm, mesh = 10 mm) were posi- plastic enclosures (mean ~ =30 cm; height =35 cm,
tioned at station M. Viva thalli were collected around mesh = 10 mm). On 23rd March, five enclosures were
the experimental site, washed with sea water, chopped positioned at the experimental site (station M, Fig-
into small pieces (approximately 20 x 20 cm) and ure 1). Each cage was filled with 100 g of wet Viva
weighed. Five subsamples of the biomass were col- thalli, as described for the long-term experiment. From
lected and stored in an ice box and transferred to the then onwards, the same five cages were sampled and
laboratory. Each cage was then filled with 100 g of wet refilled approximately every two weeks: sampling
Viva thalli, corresponding to 10.8 ± 1.1 g dry weight. dates (end of an incubation and beginning of the next
Afterwards, the cages were sampled on the 6th (the one) were 6th and 19th April, 3rd, 10th and 26th May,
14th day of incubation) and 19th April (the 27th day), 7th and 23rd June, and 5th July. Sampling of the cages
on 3rd (the 41st day) and 26th (the 64th day) May and and refilling were performed in a similar manner to
7th June (the 76th day). Each time, five cages were that described for the long-term experiment.
sampled and not replaced. The harvested biomass was
washed with sea water and kept in an ice box until Analytical procedure for macroalgal samples
transferred to the laboratory.
In the laboratory VIva thalli were cleaned to remove
Short-term experiments the epiphytes, rinsed with tap water to remove salt, and
oven- dried at 70°C to determine the dry weight (DW).
Viva thalli were incubated over the short-term peri- Subsamples ranging from 20 to 50 mg were ana-
ods (approximately two weeks) in order to simulate lyzed for the determination of total nitrogen content
the ephemeral behaviour of this seaweed. From 23rd
96
~§
0"" 160
'"O~
a>.3
Results ~:Jl 80
~~ d
Ci~
o.
Field observations
100 DIN
";"
80
In the experimental area, Viva growth started in z nitrate
(5 60
early spring when temperature and light conditions E 40
became favourable (Figure 2). Afterwards, the bio- :1-
20
mass increased rapidly until June, reaching the highest e
0
net growth rate at the end of May (Figure 3). The 60 90 120 150 180 210
growth period ceased at the end of June when a nearly Julian Day
continuous bed of Viva extended as far as the centre Figure 2. Total irradiance measured at the meteorological station in
of the lagoon. In the more sheltered areas, macroalgal Volano (a), water temperature (b) and salinity (c), dissolved oxygen
(d), dissolved inorganic nitrogen (DIN) and nitrate (e) concentrations
biomass reached values of around 500 g AFDW m- 2 ,
in the water column at station M (23 March-15 July 1994).
corresponding to a wet weight greater than 5 kg m - 2• If
the actual thickness of the macroalgal bed (50-70 cm)
is taken into account the density of Viva is approxi-
mately 700-1000 g AFDW m- 3 . Total nitrogen content in the Viva thalli ranged from
At the end of June, the Viva biomass underwent 2.5 to 3.0% DW, and was similar to the levels measured
a rapid decay which was coincident with a progres- in previous years; minima of about 2.0% DW were also
sive decrease in dissolved oxygen concentration in the typical of the biomass immediately prior the decompo-
water column (Figure 2). sition phase (Viaroli et aI., 1992, 1993). The increase
In the water column at station M, the most important in tissue-N which accompanied the decomposition of
form of dissolved inorganic nitrogen was nitrate, espe- Viva is probably dependent on the nitrogen enrich-
cially in April and to a lesser extent in May. Ammoni- ment of the decomposing biomass due to leaching of
um was almost always undetectable (Figure 2). Nitrate nitrogen-free compounds (Pugnetti et aI., 1992) as well
was in excess in April, and at the end of May. Soluble as growth of the decomposers. Total phosphorus con-
reactive phosphorus was undetectable at all sampling tent in Viva biomass ranged from 0.11 to 0.24% DW
dates, as is usual in the spring phase of macroalgal and showed no well defined seasonal trend. However,
growth (Viaroli et aI., 1993). the atomic N:P ratio in the macroalgal tissue (30.9 to
97
300 3
200 2
100
O+-~~~~-r----.----'----~ O+-----~----~----~----~----,
60 90 120 150 180 210 60 90 120 150 180 210
45
30 0.2
15
O+-----r---~~--~----~----~ O.O+-----r----~----r_--_.,_---...,
60 90 120 150 180 210 60 90 120 150 180 210
Julian days Julian days
Figure 3. Free-Hoating Ulva biomass at site M and total Kjeldahl nitrogen, total phosphorus and N:P ratios (by atoms) in macroalgal tissue
(23 March-15 July 1994). Bars indicate one standard deviation.
49.4) suggested that free-floating Uiva had a nutrition- physiological stress. Subsequently, Ulva decomposed
al status which tended toward P-limitation rather than rapidly losing approximately 50% biomass within two
N-limitation (Atkinson & Smith, 1983; Lapointe et aI., weeks.
1992). The data suggest that the observed growth pat-
tern may be related to the elemental composition of
Long-term experiment the Ulva tissue. From 23rd March to 19th April,
total phosphorus content in the Ulva thalli increased
Figure 4 shows the main results ofthe long-term exper- from 0.12 ± 0.02 to 0.31 ± 0.01 % Ow. Concurrent-
iment which simulated the natural growth of Ulva sub- ly, the total nitrogen reserve grew from 2.59 ± 0.14
jected to increased macroalgal density. From the 23rd to 3.04 ± 0.13% ow. Afterwards, tissue-P continued
March, the biomass incubated inside the enclosures to remain almost constant in the range 0.2-0.3% OW,
underwent an almost linear increase that lasted approx- while tissue-N decreased to 1.76 ± 0.39% ow. The
imately 60 days. Ouring the growth phase the macroal- latter value is significantly lower than that required for
gal biomass stored a considerable quantity of nitrogen active growth and might indicate that throughout the
(from 8.0 ± 0.8 to 68.2 ± 15.3 g N m- 3 ) and phospho- incubation period the macroalga was nitrogen limited
rus (from 0.4 to 7.3 ± 3.2 g P m- 3 ), thus functioning (Fujita et aI., 1989). A further indication that nitrogen
as a nutrient sink. was limiting was that atomic N:P ratio was less than
The maximum biomass value (1351 ± 212 g 30, well below the threshold proposed by Atkinson &
AFDW m -3) reached on 26th May was about six times Smith (1983) for balanced growth.
higher than that recorded at the commencement of the The dilution of total nitrogen due to the biomass
experiment (231 ± 23 g AFOW m- 3). Similarly, such increase coincided with an almost linear decrease of
a major increase of biomass within a single enclosure nitrate in the macroalgal tissue. Nevertheless, during
caused a large proportion of the Ulva tissue to suffer the early incubation phase Ulva took up nitrate active-
98
~
45
1200
30
800
400 15
O+---~----~----r----r--~ 0
60 90 120 150 180 210 60 90 120 150 180 210
25
2
~
0
60 20 0 -~--y-~
90 ~ 150 180 210 60 90 120 150 180 210
1500
1000
500
O+---~~--~----~~--r----'
o+---~----~--~~--~--~
60 90 120 150 180 210
60 90 120 150 '80 210
8
02
4
O+-~~----~----r---~--~ O_O~--~-----'------r---~'------'
60 90 120 150 180 210 60 90 120 150 180 210
Julian days Julian days
Figure 4. Long-term experiment (23 March-7 June). Changes of Viva biomass, atomic N:P ratio, and nitrogen and phosphorus content in
macroalgal tissue during the long-term incubation at station M. Lett: units as g m- 3 ; right: units as % dry weight. Bars indicate one standard
deviation.
99
-
imum from 10th to 26th May (from 488 to 548 ± 72 g
0.10
AFDW m- 3 ). Macroalgal growth was accompanied
:g. 0.05 by nutrient uptake and luxury storage in the biomass.
Q)
The maximum nitrogen uptake was observed from 19th
"@
0.00 April to 3rd May (from 15.2 to 36.4 ± 3.0 g N m- 3 ).
~
-3
1200 Ulva biomass (g AFDW m ) Atomic NIP ratio
60
45
800
400 v ~ V 7
V V
30
r V V ,/
V
~
l~
15
~
,/
V V V ,/ V V
o -
,/
o
30
20
10
1500
1000
500
-3
Total Phosphorus (gP m ) Total Phosphorus (% dry weight)
4 0.4
Figure 6. Short-term experiments (23 March-5 July). Changes of Viva biomass, atomic N:P ratio, and nitrogen and phosphorus content in
macroalgal tissue during the long-term incubation at station M. Left: units as g m- 3 ; right: units as % dry weight. For each experiment initial
(light) and final (dark) values of each parameter considered are shown. Dates indicated in the lower panels represent the end of an incubation
period and the beginning of the subsequent one. Bars indicate one standard deviation.
101
::s!- 0.05
mated on 19th April as well as the 26th May, the
remaining growth rate estimates correlated significant-
CD
"§
.r= 0.00
•
19 April
ly with mean DIN concentration in the water column 26 May
~
(r = 0.966, p<0.007), with the linear model explaining 2 -0.05
Cl
93.2% of the total variance. The lower values observed Qi
on 19th April and 26th May coincided, to a large extent, c: -0.10
with a decrease in total light irradiance and water tem- 0 20 40 60 80 100
perature (see Figure 2). mean DIN concentration (p.mol N 1. 1 )
Figure 7. Short-tenn experiments (23 March-5 July). Upper pan-
el: net growth rates of Ulva incubated over the short-tenn periods.
Discussion Lower panel: relationship between net growth rates and the corre-
sponding mean dissolved inorganic nitrogen (DIN) concentration in
the water column. Bars indicate one standard deviation.
Light, temperature, and nutrient availability in the
water are thought to exert direct control on macroal-
gal productivity, although the seasonal growth cycle
can also be conditioned by internal factors, such as transient nitrate availability by means of active trans-
circannual rhythms (Luning, 1993). Moreover, many port system which leads to a luxury uptake saturating
chlorophytes have been found to integrate short-term the nitrate and nitrite reductase systems and leading to
environmental variability, by virtue of either light or nitrate accumulation within vacuoles (DeBoer, 1981).
nutrient saturation characteristics. Thus, at a relatively Our experimental site is a similarly variable envi-
low growth rate they can control a fluctuating environ- ronment, with nitrate concentrations varying on both a
ment (Ramus & Venable, 1987; Duke et aI., 1989a). tidal and seasonal basis. Nitrate can be also supplied
However, when macroalgal densities exceed a certain by episodic freshwater discharge from the Po di Volano
critical threshold, growth becomes more controlled by canal. Thus the control of nitrate availability by nitrate
internal factors rather than by the biogeochemical envi- storage in macroalgal tissue is of paramount impor-
ronment. tance in regulating the growth of Viva. In the long-term
Despite a fluactuating and sometimes very stress- incubation experiments, Viva appears to control nitrate
ful environment, Ulvaceans are capable of utilizing availability for approximately 40 days. This is coinci-
ammonium when available in transiently large quanti- dent with macroalgal densities less than 850 g AFDW
ties and convert it rapidly into new biomass; further- m -3, as well as with high nitrate concentrations in the
more, it seems that these macro algae have evolved water column. Thereafter, the progressive decrease of
mechanisms of coping with variable supply (Duke tissue nitrate can be considered as an indication of the
et aI., 1989a). In fluctuating environment dominated onset of nitrogen limitation. As a matter of fact, the
by nitrate inputs, many seaweeds appear to control the decrease in internal nitrate reserve may indicate that
102
the nutritional requirements has outstripped the exter- The ability of Viva to store P and N may have rel-
nal nitrate availability. The net result is that the storage evant effects not only upon the macroalgal growth, but
capacity allows this macro alga to maintain an active also on nutrient cycling in the lagoon ecosystem. In
growth for at least three weeks after nitrate depletion. the long-term experiments nutrients were accumulated
We would also consider that Viva is a ruderal within macroalgal biomass for about two months, with
species specialised in rapid colonisation, with large mean rates of 3.3% d- 1 for Nand 4.6% d- 1 for P. In
individuals suffering frequent extinctions due to the the short-term experiments, beside a greater variabil-
removal by waves and tidal currents (Littler & Lit- ity, we recorded accumulation rates of approximately
tler, 1980). In the short-term experiments, resembling 8% d- 1 for Nand 10% d- l for P. Thus Viva can be
this ephemeral behaviour of Viva, net growth rates considered as a nutrient sink over very long periods of
were directly related to the dissolved nitrate concentra- time. Furthermore, high uptake rates coupled with the
tions in the water column since there was a significant removal of floating macroalgae by currents and storms
decrease in net growth rates when nitrate became limit- would likely favour a considerable nutrient loss.
ing in the water. Thus we can postulate that in the Sacca The reciprocal control of Viva against biogeo-
di Goro nitrate is one of the key factors determining chemical environment and viceversa is a key factor in
Viva biomass production, and it is in agreement with explaining both resource competition and successional
previous studies by Viaroli et aI. (1992; 1993). stages in primary producer communities dominated by
Temporary luxury uptake and storage of nutri- Viva (Smith & Horne, 1988; Pugnetti et aI., 1992).
ents may confer selective advantages to Ulvaceans by However, it is evident that when biomass exceeds
buffering against transient nutrient availability. Fur- certain critical tresho1ds, approximately a thousand g
thermore, several Ulvaceans are able to control pro- AFDW m-\ the macroa1gal community shifts from
longed nitrogen depletion via storage of nitrogen in active production to rapid decomposition, probably due
photosynthetic pigments and rubisco molecules (Duke to strong selfshading, biomass density and anaerobic
et aI., 1989a). Phosphorus is a different matter due to its processes starting within the macroalga1 beds. This is
rapid turn-over. However, in both long and short-term likely the case for Viva rigida which accumulates in the
experiments we recorded considerable accumulation sheltered lagoons of the northern Adriatic coast: at low
of P within the Viva biomass. In the short-term experi- biomass, growth appears controlled by nutrient avail-
ments, the potential importance ofN relative to P in the ability, whereas when biomass exceeds 2-3 kg wet
incubated biomass is difficult to assess due to the high weigth, the density-dependent factors become domi-
degree of variability. On 23rd March, at the beginning nant in driving the entire system towards dystrophic
of the experiments, the N:P ratio was 49.4: 1 indicating cnSlS.
a strong P-limitation, whereas in the subsequent period
Viva tended to become N-limited (N:P ratio = 19.6:1).
In early May, after the increase of nitrate concentration Acknowledgments
in the water column, availability of P, rather than N,
seemed of paramount importance in limiting macroal- This research was supported by the joint EU project
gal growth (N:P ratio=47.0:1). Considering that N:P 'CLEAN': Coastal Lagoon Eutrophication and ANaer-
ratios less than 10 indicate N-limitation, while N:P obic processes (Contract No EV5V-CT92-0080). Part
ratios greater than 30 indicate P-limitation (Atkinson of our work was made possible thanks to Mr Vitto-
& Smith, 1983; Smith, 1984), we can therefore con- rio Gaiani (Dipartimento di Biologia, Universita di
clude that the nutritional status of macroalgae incubat- Ferrara), Mr Tony Zappata (Battello Hydra, Amminis-
ed over short periods would seem to depend mainly trazione Provinciale di Ferrara), Mr Mimmo Cavalca
on nutrients supplied exogenously from the surround- (lstituto di Ecologia, Universita di Parma) and Dr Mat-
ing water. In contrast, macroalgae incubated in the teo Cattadori (lTIS 'G. Gali1ei', San Secondo, Parma).
long-term experiment underwent a clear decrease in We are indebted with Prof. Paolo Menozzi (Istituto di
N:P ratios from 49.4:1 to 14.6:1 indicating a progres- Ecologia, Universita di Parma) and Prof. R. A. Her-
sion to N-limitation. In this case, the nutritional status bert (Department of Biology, Dundee University) who
was controlled by nutrient dilution due to the increased revised the manuscript.
macroalgal density rather than by the biogeochemical
environment.
103
Abstract
The impact of macrophyte communities on benthic fluxes has been analyzed in three shallow coastal environments:
Etang du Prevost (Mediterranean coast of France), characterized by the large floating macro-alga Ulva rigida;
Certes fishponds (Bassin d' Arcachon), covered by Ruppia cirrhosa; and the inner intertidal mud-flat in the
Arcachon Bay (French Atlantic coast), which has extensive Zostera noltii meadows. In these bodies of water,
primary production is dependent primarily on the dominant seagrasses and macroalgae that are also responsible for
the large quantity of organic matter deposited on the sediment surface. In 1993 and 1994, fluxes of oxygen, sulphide
and nutrients were measured in early and late summer, which were selected in order to represent the production and
decomposition phases of the dominant macrophytes. Experimental work was undertaken to measure: (1) standing
crop of dominant macroalgae and rooted phanerogams and the elemental and macromolecular composition of plant
biomass; (2) benthic fluxes of oxygen, sulphide, nitrogen and phosphorus using incubation of multiple dark and
light benthic chambers; (3) water-sediment profiles of free-sulphide in sediment cores with rooted phanerogams
(Ruppia) as well as with floating seaweeds (Ulva).
At the selected sampling sites, in addition to external (tides) and/or internal (sediment reactivity) factors, we
observed differences in benthic fluxes which were clearly related to growth patterns and structure of the macrophyte
communities. The Z. noltii meadows were stable and characterized by slow growth and almost constant biomass.
In the more sheltered sampling station in the Certes fishponds, R. cirrhosa beds showed a summer decrease due to
extensive epiphyte growth. During the decomposition phase, significant fluxes of free-sulphide occurred inside the
dark benthic chambers, probably due to the metabolism of the epiphytic layer. In the Etang du Prevost, U. rigida
achieved high biomass levels, even though the macroalgal beds exhibited a patchy distribution due to wind action
and the hydrodynamics of the lagoon. In the decomposition phase, which was coincident with the annual dystrophic
crisis the rapid decomposition of Ulva led to high fluxes of free sulphide.
The shift from the production to decomposition phase resulted in substantial changes in nutrient recycling only in
the macro-algal-dominated system. During the growth period dissol ved inorganic nitrogen and phosphorus were kept
at low levels due to macrophyte uptake. In contrast during the decomposition phase when the macroalgal biomass
was mineralised, nitrogen and phosphorus were rapidly recycled. The same processes did not occur in the Certes
fishponds probably because of the greater internal buffering capacity linked either to plant morphology/physiology
or to the properties of the sediment.
growth, which in turn is accompanied by a decline of oxygenated or even supersaturated layer, and the lower
attached phanerogam communities due to decreased one, which is anoxic and highly reduced.
water transparency, build up of epiphytes and occur- Hence, growth patterns (intensity, duration and
rence of reducing conditions in the bottom water and rate) and distribution of macroalgae and seagrasses
surface sediments (Twilley et ai., 1985; Sand-Jensen may explain the changes which occur in other pro-
& Borum, 1991; Nienhuis, 1992; Williams & Ruck- cesses within the coastal ecosystems (Sand-Jensen &
elshaus, 1993). Aquatic environments with extensive Borum, 1991). For example, benthic processes are
phytoplankton blooms frequently progress to systems particularly sensitive and can undergo catastrophic
dominated by opportunistic, free-floating macroalgae changes as result of sudden non-linear responses to
belonging to the Ulvales and Cladophorales. the external stresses. The structure or form of plant
The development of such large and fast grow- communities may mitigate against the drastic changes
ing primary producers plays a key role in oxygen of external forcing, since aquatic ecosystems with root-
production and consumption and regulation of the ed phanerogams are more resilient than those with
entire benthic metabolism (for an up-dated review see green macroalgae. Such mitigation can contribute to
Sand-Jensen & Borum, 1991). Usually, fast growing the resistance and resilience of ecosystems, acting
macroalgae exibit nutrient uptake and storage greatly as buffers against external perturbations (Sherman,
in excess of their physiological needs. This is the case 1994).
for Viva rigida C. Agardh, whose luxury uptake cou-
pled with high growth rates results in nutrient deple-
tion in the water column, which in turn results in the Aims and objectives
pulsing trends seen especially in the dissolved inorgan-
ic nitrogen pooi. As a consequence, Viva outcompetes In the three ecosystems investigated the principal pri-
other primary producers due to its ability to store nitro- mary producers are seagrasses and macroalgae which
gen, bringing about a shortened food chain (Smith & are also responsible for the bulk ofthe vegetative detri-
Horne, 1985; Pugnetti et ai., 1992; Fong et ai., 1994). tus deposited on the sediment surface (Auby et ai.,
There is also experimental evidence of a strict rela- 1994; Castel et ai., 1996). It would therefore be
tionship between growth and decomposition rates: fast expected that benthic fluxes are dependent primari-
growing plants tend to have high nutrient concentra- lyon growth patterns and distribution of the primary
tions and low fibre content and also decompose more producers and decomposition rates of the plant detri-
rapidly because their high NIC and PIC ratios stimu- tus. We postulated that the intensity of water-sediment
late microbial growth (Buchsbaum et ai., 1991; Duarte, exchanges in lagoons dominated by slow growing root-
1992; Enriquez et ai., 1993). Therefore, the replace- ed angiosperms, such as Ruppia cirrhosa (Pet.) Grande
ment of seagrass meadows by macro algae can effect and Zostera noltii Hornem, would show less variation
changes not only in the timing and rate of detritus pro- than those occurring in lagoons dominated by oppor-
duction but also in decomposition rate, which in turn tunistic faster- growing macroalgae, such as Viva rigi-
leads to a decline in water quality. For example, fast da. In addition to studying benthic fluxes in relation
macroalgal growth followed by decomposition of the to growth patterns, we also investigated the effects
accumulated biomass is considered to be one of the of plant morphology on water stratification. It was
most important factors in the occurrence of dystrophic anticipated that the development of pelagic macroal-
crises in several Italian lagoons (Izzo & Hull, 1991; gae would lead to laminar stratification which limits
Sfriso et ai., 1992; Viaroli et ai., 1995). In addition to water mixing and gas exchange, while the growth of
the above-mentioned effects, changes also occur in the rooted phanerogams would enhance vertical exchange
composition of the primary producer community as a and gas transport throughout the water column and
consequence of eutrophication. Rooted phanerogams surficial sediment.
provide gas transport throughout the surficial sedi- In 1993 and 1994, fluxes of oxygen, sulphide and
ment contributing to its oxygenation, while floating nutrients were measured in early and late summer.
macroalgae can promote physical stratification of the These sampling periods were selected to represent pro-
water (Crawford, 1992). For example, the large thalli duction and decomposition phases of the dominant
of Viva act as physical barriers separating the water macrophytes (Christian et ai., 1996). In this paper we
column into two distinct layers: the uppermost well- present data from selected experiments in which the
107
Study area
Our work addressed the effects of macrophyte commu-
nities on benthic fluxes in three shallow coastal envi-
ronments: Etang du Prevost (mediterranean coast of
France) characterized by the large floating macroalga
Ulva rigida; fishponds of Certes (Bassin d' Arcachon), Figure 1. General layout of the Arcachon Bay and the Etang du
covered by Ruppia cirrhosa; and the inner intertidal Prevost, and positions of the experimental sites.
flat in the Bassin d' Arcachon (west Atlantic coast of
France), which is dominated by Zostera noltii (Fig-
ure 1). Detailed descriptions of experimental sites have
been reported by Auby et ai. (1994). Here we refer mined by spectrophotometry after acid extraction of
to the Etang du Prevost as station 11, fishponds of the ashes (Aspila et aI., 1976). Approximately 2-5 g
Certes as station Cl and the inner site in the Bassin were also analyzed for hemicellulose, cellulose and
d' Arcachon as station B. lignin (Goering & Van Soest, 1970).
Biomass, and elemental and macromolecular Estimation of benthic fluxes by incubation of multiple
composition of dominant macrophytes dark and light benthic chambers
The biomass of macroalgae and macrophytes was Fluxes of nutrients, oxygen and free-sulphide were
determined by direct harvesting inside circular cores determined using dark and light cylindrical benthic
(30 cm i.d.) The harvested plants were sorted, washed chambers and bell jars with volume of about 14 litres
with both lagoon and tap water, weighed and oven dried and cross-sectional area of approximately 700 cm 2 .
at 70 ° C until they reached a constant dry weight (DW). On 20 May and 31 August 1994, at Station B three
During the sampling of August 1994 the epiphytes cov- dark and three light benthic chambers were incubated
ering Ruppia (station C 1) were carefully removed from over the Zostera beds. Each experiment was carried out
the host plants by means of forceps and oven dried at around midday and because of the tidal cycle incubated
70°C. At station Cl sestonic chlorophyll-a was also for only 3--4 hours.
measured by conventional techniques (APHA, 1975). At stations C 1 and 11, each experiment began in
Dried subsamples were used to determine the ele- the early afternoon and lasted approximately 50 hours.
mental and macromolecular composition of the har- From 23 until 25 May 1994, at station C 1 we compared
vested biomasses. Subsamples in the range 100- three different conditions: bare sediment, low (70 g
200 mg were analyzed with a LECO 600 CHN ele- DW m- 2 ) and high (110 g DW m- 2 ) Ruppia biomass.
mental analyzer to determine the total C and N con- From 23 to 25 August 1994, we studied the impact
tent. About one gram of dried biomass was ignited at of the growth of the epiphytic community by analyz-
550°C in a muffle furnace to determine ash and ash ing the following conditions: bare sediment, Ruppia
free dry weight (AFDW). Total phosphorus was deter- (86 ± 11 g DW m- 2 ) in excess of epiphytes (52 ± 14 g
108
Table 2. Major characteristics of the opportunistic macroalga Viva rigida and the rooted
phanerogams Zostera noltii and Ruppia cirrhosa and epiphytes growing on Ruppia at
the experimental sites. Units: % dry weight; ud.: not detectable; n.d.: not determined.
Oxygen fluxes estimated by benthic chamber experiments 250 Biomass (g OW m -2) 20 May 1994
E:'J 20 May o 31 August 200
20 150
15 100
'J::
10
'"'E 5
0 (\1
0
"0
E -5
E
-10 r:J Ruppia o Epiphytes
-15 250 -2
NP DR GP Biomass (g OW m ) 26 August 1994
N.~~ON~W~ON~~~O~~
- . . - . - . . - ...... ('\I(\IN<'I!INt")("')M
(Figure 3).
At station C 1, from 23 to 25 May 1994 we com-
pared three different levels of macrophyte cover using N~~~ON ~~~ ON~~~ON~
-..--..-..-NNN NNC")("')("')
the benthic chambers: bare sediment, low (70 g DW cross section (m)
m- 2 ) and high (110 g DW m- 2 ) Ruppia biomass.
As expected, changes in oxygen production and con- Figure 3. Variations in biomass of Ruppia cirrhosa and its epiphytes
(top and intermediate panels) and relationship between epiphytes and
sumption were closely related to the standing stock sestonic chlorophyll-a (bottom panel) across a section at station C I
of Ruppia, while the role played by the plankton and in 1994.
microphytobenthic communities was less significant
(Figure 4). Analysis of the results shows that the oxy-
gen decrease in the sediment plus plankton chambers trations ranging between 50 and 240 f.1,moll- 1 inside
was initially almost as high as the other chambers, all benthic chambers (data not shown).
while afterwards consumption was minimal due to the From 23 to 25 August 1994, we studied the impact
lack of production. Variations in oxygen concentra- of the growth of the epiphytic community by analyz-
tion in the lagoon fully conformed to those observed ing bare sediment, and two different assemblages of
inside the benthic chambers in the light which con- macrophytes and epiphytes: Ruppia in excess of epi-
tained the highest biomass and followed a clear dai- phytes (R>E) and vice-versa (R<E). In the light ben-
ly trend consistent with that reported by Thimel & thic chambers the oxygen concentration followed a
Labourg (1992). Under all the experimental condi- clear daily trend with night anoxia, which was coupled
tions nitrate, nitrite, and soluble reactive phospho- with free-sulphide production (Figure 5). Inside the
rus (SRP) remained below the level of dectection. R>E chamber sulphide increased to 12.3 mmol m- 2
The ammonium concentration remained constant in over about 8 hours and in R <E chamber it reached
the light benthic chambers (around 5 f.1,moll- I ), while 25.9 mmol m- 2 after about 8 hours. Inside the dark
it showed significant increases principally in the dark benthic chambers, oxygen was depleted within a few
chambers without macrophyte cover and containing a hours. Afterwards an almost linear increase of free-
high biomass (Figure 4). However, most of the dis- sulphide was observed. After 49 hours, 58.3 (bare sed-
solved nitrogen was recycled and stored in the organic iment), 86.1 (R>E) and 118.1 (R<E) mmol S2- m- 2
pool, as shown by dissolved organic nitrogen concen- were released to the water column. Significant sul-
111
sediment+plankton
150
sediment+plankton
6000
~
E 100 E
ON z 4000 light
o 0
E
E
50
dark
light
E
::l..
15C
Ruppia (LB)+sediment+plankton Ruppia (LB)+sediment+plankton
6000
~
E 100 E
z 4000
)l
light
0
E dark
::::..
50
2000
.•..• 0 ••.....
0
12 24 36 48 0 12 24 36 48
150
Ruppia (HB)+sediment+plankton Ruppia (HB)+sediment+plankton
6000
a..... ~
E 100 \) o
E
z 4000
1
dark o light 0
E
50 ::~,,/ ::::..
2000
/ ·'0,
: "
O+---~~~~~.-~--~-.-.-
0. ..... .".. 0 .
\
o+-----,-----~----_r-----r--
o 12 24 36 48 o 12 24 36 48
23 May incubation time (hours) 23 May incubation time (hours)
12a.m. 12 a.m.
Figure 4. Main results of the experiment with benthic chambers carried out at station CI from 23 until 25 May 1994. Ruppia biomass inside
benthic chambers: 70 g DW m- 2 (LB) and 110 g DW m-- 2 (HB).
112
phide fluxes occurred in spite of the expected buffer- p<O.OOOl) with bare sediment, and in the dark benthic
ing capacity due to the high iron content (Bartoli et ai., chamber with Viva, where SRP followed a nearly lin-
1996). If we consider that sulphide production was ear increase (27.1 ± 11.7 tlmol P m- 2 h- i , r= 0.933,
related to the Ruppia/epiphyte ratio we can postulate P<O.OOOl). Inside the light and dark bell jars and in
that sulphide production occurred in close proximi- the light benthic chamber with Viva, most of the SRP
ty to epiphyte layer. Under these conditions phosphate was present in the macroalgal biomass and from these
release might have been expected but SRP release from data we can infer that most of SRP was obtained from
the sediment was negligible. All the inorganic forms of the sediment. Over the same period, a significant linear
both nitrogen and phosphorus in the water were below increase in ammonium was observed inside the dark
the level of detection, while dissolved organic nitrogen benthic chamber containing Viva (34.7 ± 8.2 tlmol N
concentrations were similar to those observed in May m- 2 h- i , r=0.978, p<O.OOOl). After about 12 hours
(data not shown). ammonium produced in the dark benthic chamber with
Viva was almost three times higher than in the dark
Etang du Prevost, station 11 benthic chamber with bare sediment and the dark bell
In May 1994, the lagoon of Prevost was covered by jar containing the macroalga (Figure 6). These results
floating mats of Viva rigida with a patchy distribution highlight the potential activity of Viva and its epiphytes
of biomass (from 76 to 631 g DW m- 2 ) and net growth of immobilising available phosphorus and nitrogen and
rates of 0.078 ± 0.022 d- i (Viaroli et ai., 1994). Subse- slowing their recycling, even under the highly reduc-
quently, the Viva standing crop decreased dramatically ing conditions of the dark bell jar. In contrast, such
due to harvesting, which started in early summer. In conditions appear to enhance SRP release from anoxic
August, at station 11 the sediment was almost denuded, sediment leading to an imbalance between phosphorus
while along the southern and western shoreline there and nitrogen recycling.
were large accumulations of filamentous green algae. Benthic fluxes during the Viva growth phase were
From 16 to 18 May 1994, we studied three different compared to those measured during the decomposition
experimental conditions: benthic chambers with bare phase in 1993. From 24 to 26 August 1993, the incuba-
sediment, benthic chambers with sediment covered by tion of about 500 g DW m- 2 of decaying macroalgae
Viva and bell jars with Viva plus the plankton commu- resulted in anoxia accompanied by the production of
nity but without sediment. Macroalgal biomass inside large quantities of free-sulphide in both the light and
the incubation chambers was approximately 500 g DW dark benthic chambers (Figure 7). In the dark chamber
m- 2 . Changes in oxygen fluxes were to a large extent without macroalgae, the sulphide flux through the bare
influenced by macroalgal activity as indicated by the sediment was several times lower and was delayed by
maximum oxygen increase in the light benthic cham- almost 24 hours. The presence of free sulphide in the
ber (9.69 mmol O 2 m- 2 h- i ) and bell jar (30.75 mmol water of the Prevost lagoon is not unusual probably
O 2 m- 2 h- i ) containing Viva (Figure 6). In the light due to the large amount of organic matter undergoing
chamber with bare sediment, the oxygen concentration rapid degradation and the low buffering capacity of the
decreased showing that consumption prevailed over system (Caumette, 1986). Although the low buffering
production processes. A similar trend was observed capacity of the sediments may explain the high sul-
in the corresponding dark chamber, where oxygen phide fluxes recorded, the great difference in fluxes
uptake was almost constant at 4.68 mmol 02 m- 2 between the dark chamber containing bare sediment
h- i (Figure 6). These data indicate that Viva was and those containing Viva infers that sulphide produc-
responsible for the bulk of the oxygen production, tion was initiated within the decomposing macroalgal
while benthic metabolism accounted for most of the mats.
oxygen consumption. The association of metabolic The dramatic changes in redox conditions which
activities of both macroalgae and sediment resulted occurred in the dark benthic chambers were accompa-
in major changes in the physico-chemical properties nied by the substantial release of both ammonium and
of the water in the light and dark benthic chambers and SRP. Dark recycling of P and N was clearly related to
bell jars. For example the pH increased from 7.9 to 9.7. Viva decay, even though significant ammonium flux-
Significant SRP fluxes were observed inside the light es were also observed in the bare sediment chamber.
(8.4 ± 1.5 tlmol P m- 2 h- i , r= 0.988, p<O.OOOl) and However, in the latter case the ammonium to phospho-
dark chambers (15.6 ± 3.1 tlmol P m- 2 h- i , r= 0.984, rus ratios were close to the conventional Redfield ratio
(N:P= 16:1), whereas in the chamber with Viva SRP
113
LIGHT BENTHIC CHAMBERS DARK BENTHIC CHAMBERS
bare sediment bare sediment
100~--------------------1 150 100
sulphide not detectable
~
~ E 75 E 75 ~
(5 (5 100 sulphide E
E E (5
-.S
-.S
~.......•
50 50 E
c: <D
-0 -.S
<))
c:
~ gs,
~
o 25 :sen 50 ,,
25 S<-
~
<D
,,
0------- -0---0
0
0
12 24 36 48 12 24 36 48
J\
<D
:s
en 25
<D
0>
S<-
Cl.
-0' -'" 25
0>
S<-
.
.0-
<D 0 <D p'--- 0
,.-
~ ~
0 0
12 24 36 48 0 12 24 36 48
1lI/
E E
(5 ,, (5
-.S E -.S 50 E
II "
50
<D
:'Q -.S <D
-0 .0/.0' -.S
..c: c: E c:
10
:sen 50
Cl. <D Cl. <))
0> 0>
~ 25 S<- "Il' 25 S<-
<D <D _-..0 0
0
~ ~ ,/0--
0 0 0
12 24 36 48 0 12 24 36 48
23 Aug incubation time (hours) 23 Aug incubation time (hours)
12a.m. 12 a.m.
Figure 5. Main results of the experiment with benthic chambers carried out at station CI from 23 until 25 August 1994 (explanation given in
the text).
-""
DISSOLVED OXYGEN AMMONIA NITROGEN SOLUBLE REACTIVE PHOSPHORUS
a.····· O
~ SO····"J.....
01 • •• pi 0+1----,-----r----r----~
0 12 24 36 48 o 12 24 36 48
')I ')I
E 150 '"E 1500 dark
E 2000 1
z Cl.
100 \ light ! 1000 ~
$gE \ 0 "-
1000
50
r \ ...., ... /p....Il......o.••o.••••.•• /f 500
./
/ light 1- / light
oI 4 II '" .. .., II ,. 0
0 12 24 36 48 0 12 24 36 48 0 12 24 36 48
16 May Incubation time (hours) 16 May Incubation lime (hours) 16 May incubation time (hours)
12 a.m. 12 a.m. 12 a.m.
Figure 6. Main results of the experiment with benthic chambers carried out at station 11 from 16 until 18 May 1994.
115
50 E
1
",,'
>,,4
sulphide
25 c:
<I>
~
o
O~o--r~~~--~--~~O
12 24 36 48 o 12 24 36 48
E sulphide /0 E 750
"0100
~
~ /,,'
,,
50 <")' E
~
1 500
50
,,
<I>
~ , .sc: 32
:a
~
SO ,, 25
~
.c
.Q.
~ 250
25
,'.tS $< <I>
<I> o jg I
jg
0~4--r--~----~--~~0
O~/*-\-r--~~--~--~~o
o 12 24 36 48 o 12 24 36 48
4 4 4
E E E
- 20
]
~
3 ammonium 3 3
.§. ".0. ............
E 2 ", .................. 2 E 2
::::>
'2
'2
~ , '
"
-
~ 10
E
co ~ phosphorus ~co
o+----r---,----~---r~0 O+---~----~--~----r-~O
o 12 24 36 48 o 12 24 36 48
sediment+Ulva sediment+Ulva
5~------------------~5 40~--------------------~~
,,1>_----
,
E
4
phosphorus
4 <")'
E E 30 , ,, <")'
E
,,
1
20
E
1 (/)
"0
120
E I
I
1 ammonrum
,
phosphorus
VI
::::> 2 ::::> I 2
'2 o '2 , 10 o
.c
~ fil-
o
~ 10 : ~
o
E .c E .c
co a. co a.
04---~----...---~----r-~0
o 12 24 36 48
24 Aug incubation time (hours) 24 Aug incubation time (hours)
12a.m. 12a.m.
Figure 7. Main results of the experiment with benthic chambers carried out at station 11 from 24 until 26 August 1993.
116
flux was at least three times greater. This trend allow reflects the extent to which each site can support pro-
us to conclude that Ulva decomposition may result cesses associated with oxygen availability, i.e. benthic
in an imbalance between ammonium and SRP flux, autotrophy. This index becomes very useful when com-
probably because ammonium was primarily recycled paring sites on a seasonal basis (Figure 8). For exam-
from the decomposing biomass, while SRP was also ple, station 11 which was autotrophic during the spring
released from anoxic sediment (Figure 7). In contrast, growth phase clearly became heterotrophic in the sum-
net ammonium release was not recorded in the light, mer decomposition phase. Station Cl underwent slight
indicating that macroalgae may function as a nitrogen changes probably linked to shifts in community struc-
sink, even under the extreme conditions which devel- ture, while station B was highly autotrophic at all times.
oped after several hours of incubation. In the latter case, the relationship between oxygen
production and consumption was presumably overes-
Intersystem comparison timated due to the short-term incubation. Moreover,
high oxygen availability would be likely maintained
The macroscopic differences observed at the selected by tidal flushing. Results from the potential sulphide
sites are to a large extent dependent on hydrodynam- fluxes induced by incubating macrophytes inside the
ic factors: station B is well flushed by tidal currents, benthic chambers show that they were proportional
station 11 is slowly flushed due to the bathymetry of to oxygen consumption (Figure 9). This pattern high-
the Etang du Prevost, while station Cl is in a man- lights the relationship between macrophyte community
made lagoon which undergoes limited water exchange and ecosystem stability. As previously discussed, data
(Escaravage, 1990). However, we can reasonably pos- from station B should be considered carefully being
tulate that internal factors linked to macrophyte com- sulphide fluxes affected by tidal flushing.
munity structure also contribute to the maintenance of If we consider oxygen and sulphide fluxes as indi-
such diversity (Odum, 1990; Brinson, 1993). cators of the stability and resilience of these lagoon
Standing crop data provide basic information on systems, then a clear progression in macrophyte suc-
the structure of the primary producer communities. At cession is evident viz.:
intertidal station B, the Z. noltii meadows are character- Zostera 2: Ruppia > Ruppia + epiphytes » Ulva
ized by slow growth and an almost constant biomass.
In the more sheltered station Cl, R. Cirrhosa, which with the Zostera system more buffered than the Ulva
has a morphology similar to that of Z. noitii, under- one. This explains why the Etang du Prevost is subject
went a period of decomposition in the summer due to to annual dystrophic crises (Caumette. 1986), while
extensive growth of epiphytes. At station 11, in the the fishponds of Certes experience a seasonal decay
Etang du Prevost, U. rigida reached very high biomass as well as summer anoxia mainly due to the massive
levels. However, ephemeral macroalgae such as Ulva development of epiphytes on the Ruppia (Escaravage.
have a patchy distribution because they are affected 1990; Thimel & Labourg, 1992).
by prevailing hydrodynamic conditions, such as those A quite similar pattern can be also recognized when
induced by wind and currents. the fibre content of the detritus is considered, which is
The differences in structure and metabolism of the an index of organic matter biodegradability or recalci-
primary producers can be related to the wide range of trance (Godshalk & Wetzel, 1978; Mann, 1988):
fluxes of oxygen, nutrients, and free-sulphide which Zostera2:healthy Ruppia (refractory»
were recorded at the different sampling sites. Initially,
we used a first ordination of such differences by using Ruppia epiphytes (intermediate» Ulva (labile)
the metabolism based trophic index (BTSI) modified
from that proposed by Rizzo et al. (in press). BTSI The large differences between the rooted and float-
is based on the relative magnitude of hourly rates of ing systems have also been demonstrated by means of
maximum net oxygen fluxes measured under both light vertical profiles taken along the water-sediment hori-
and dark conditions (Table 3). In this scheme maxi- zon (Figure 10). In the system with Ulva high sul-
mum oxygen production in the light (X-axis) is plot- phide concentrations were detected in the water layer
ted against oxygen uptake in the dark (Y-axis). Rela- beneath the floating thalli. These values were almost
tive positions are indicative of degrees of autotrophy twice as high as those recorded either in the cores
and heterotrophy. We used this approach to assign to without macroalgae or with Ruppia. In the latter appre-
each benthic system a categorical value which broadly ciable amounts of sulphide were detected only a few
·2 .,
Net maximum oxygen productivity (mmol m h )
·5 0 5 10 15
0
--;- highly
.s= autotrophic
~
E
"0
E ·10
.s.
OJ
-'"
.!!!
0.
::> net
<::
OJ
C) ·20 heterotrophic net
i:<' autotrophic C1·HB
0
-'" c
to
Cl
·30
·2 .,
Net maximum oxygen productivity (mmol m h )
·2 .,
0 10 20 Dark oxygen uptake (mmol m h
--;-
• dark benthic chambers C light benthic chambers
s:: highly
':' autotrophic
E B
"0
Figure 9. Relationships between dark oxygen uptake and maximum
E -10 potential sulphide fluxes in dark and light benthic chambers. Sym-
.s. net
10
11·S
C1·R<E
bols as in Figure 8.
~ heterotrophic
.!!!
0.
::>
<::
OJ
~
·20 net
~ autotrophic almost independently from that in the sediment and
~
11·U may thus determine internal fluxes which overlap the
'"
Cl c
·30 - ' - - - - - - - ' - - - - - - - - - - - - - ' benthic ones. These results are consistent with those
of Thimel & Labourg (1992), who recognised the key
FiRure 8. Ordination of the sites considered according to the BTSI
role of detritus in determining internal carbon flows.
(metabolism based trophic state index). Top panel: growth phase,
bottom panel: decay phase. Symbols - station B: with Zostera To some extent, the epiphytic subsystem at sta-
(B); station ll: with Viva (ll-V) and with bare sediment (II-S); tion CI resembles the subsystem which developed at
station Cl: with low (Cl-LB) and high (Cl-HB) biomass of Ruppia, station 11 in the Etang du Prevost due to accumula-
Ruppia in excess of epiphytes (R> E), epiphytes in excess of Ruppia
(R<E), and bare sediment (Cl-S). Further details are given in the
tion of Viva. Both appear to result in significant direct
text. effects on water quality, even though station 11 is less
protected against dystrophic crisis. These differences
between the two sites, which are undergoing similar
centimeters below the sediment surface and can be slow water reneval, can be explained primarily by the
explained by the deposition of a thick bed of epiphytes different macrophyte communities.
with an organic matter content in the range 20-30%
w/w. Such organic matter consists mainly of mucilagi-
nous aggregates which are readily degradeable and can Concluding remarks
sustain high levels of microbial activity which exceed
the buffering capacity of the available iron (Bartoli Viva rig ida , Ruppia cirrhosa and Zostera noltii show
et aI., 1996). Likewise, the significant fluxes of sul- very different growth patterns as well as production
phide observed in the dark and light benthic chambers rates (Figure 11). In a number of dystrophic lagoons,
can be explained by the production of free-sulphide Viva rigida reaches huge biomasses (up to 103 g DW
by the epiphytes growing on the leaves of the Rup- m- 2 ) in few weeks (Sfriso et aI., 1992). Unlike, root-
pia or by the epiphytes covering the sediment surface. ed phanerogams they have annual cycles and their
These results demonstrate the importance of epiphytes biomasses never exceed a few hundred g DW m- 2
in establishing very active localised systems at the sed- (Verhoeven, 1980, Philippart, 1995).
iment - water interface and around the Ruppia leaves. The fast growth of Viva and the very high biomass
In the short-term, the epiphytic pathway may operate produced does not enable the top down mechanisms
118
station 11 station C1
ing the amplitude of such disturbance. When this kind
mm r - - - - - mm ...----:-:---::----,-, of biogeochemical control is ineffective, the effects of
w"'ith-ou-t7:
U,--lva' without Ruppia
70 20
dystrophic events would presumably lower the resis-
50
15 tance as well as the resilience of the lagoon community
10 leading to a positive feedback.
30
10 0~------------4
Acknowledgments
·10 +--.----.---.---1 ·5 +'-'-"'-'-'',"-''-''-"-.,'.-
...=----1
o 50 100 150 200 o 2
~M S 2· IlM S 2· This research was supported by the joint EU project
'CLEAN': Coastal Lagoon Eutrophication and ANaer-
mm , - - - - - - - - - , mm,-----w~it~hR~u-pp~·a' obic processes (Contract No EV5V-CT92-0080) We
70 20
gratefully acknowledge the support and facilities pro-
15
50~-7U~lv-a~th-a~IIi---,1 vided in Arcachon by Professor Pierre Caumette, Dr
10 Jacques Castel, Dr P. 1. Labourg, Dr R. de Wit and
30
their colleagues during the course of this project. We
10 ••.••.
-"co_~ oi---------+ are indebted with Dr M. Cattadori (Istituto di Ecologia,
1o~··~···---- ____ o __ Universita di Parma) who contributed to the field and
·10+-.--.----.---.-..,.--1 .5 +--r--.---=;R=.....q
o 6 8 10 12 o 50 100 150 200 laboratory work and useful information on the refrac-
mM S 2· ~M S 2· tory compounds.
Figure 10. Sulphide profiles determined along a water·sediment hori·
zon in sediment cores collected at station CI (right) and II (left) in
August 1994. References
oxidized A.P.H.A., A.WWA., WP.c.F., 1975. Standard methods for the
macrophyte
________ community examination of water and wastewaters. 14th ed. A.P.H.A. Wash-
controls ington, 1193 pp.
.,::!2
Aspila, K. I., H. Ageminian & A. S. Y. Chau, 1976. A semiautomat-
~---,/
'0 .,
..; ">.
-,.,
a.C)
ed method for the determination of inorganic, organic and total
phosphate in sediments. Analyst 101: 187- 197.
'E" .,-
.,c: Auby, I., F. Manaud, D. Maurrer & G. Trut, 1994. Etude de
oc: ., la proliferation des algues vertes dans Ie Bassin d' Arcachon.
i
".;: >
0-'::: Rapport IFREMER·CEMAGREF-SSA·SABARC. IFREMER,
'" l1S Arcachon, 163 pp.
TI~ reduced
" - Bartoli, M., M. Cattadori, G. Giordani & P. Viaroli, 1996. Benthic
t
;;:: 0
oxygen respiration, ammonium and phosphorus regeneration in
surficial sediments of the Sacca di Goro (Northern Italy) and two
biogeochemical controls french coastal lagoons: a comparative study Hydrobiologia 329
(Dev. Hydrobiol. 117): 143-159.
Figure II. Schematic representation of intrasystem variablity in Brinson, M. M., 1993. Changes in the functioning of wetlands along
macroalgal (M) and seagrass (S) communities. Macrophyte con· environmental gradients. Wetlands 13: 65-74.
trois: density, self shading, nutrient limitation, timing of detritus Buchsbaum, R., I. Valiela, T. Swain, M. Dzierzesky & S. Allen,
production, molecular recalcitrance, etc. 1991. Available and refractory nitrogen in detritus of coastal
vascular plants and macroalgae. Mar. Ecol. Prog. Ser. 72: 131-
143.
Castel, J., P. Caumette & R. Herbert, 1996. Eutrophication gradi·
to control primary production and can lead the system ents as exemplified by the Bassin d' Arcachon and the Etang du
organization to a shortened food chain (Smith & Horne, Prevost. Hydrobiologia 329 (Dev. Hydrobiol. 117): xi-xxx.
1985, Pugnetti et aI., 1992). Thus, energy flow is char- Caumette, P., 1986. Phototrophic sulphur bacteria and sulphate
reducing bacteria causing red waters in a shallow brackish coastal
acterized by wide seasonal pulses with great organic lagoon (Prevost Lagoon, France). FEMS Microbiol.-Ecol. 38:
matter accumulation and then great instability. In this 113-124.
manner, a change in the community structure can affect Christian, R. R., E. Fores, F. A. Comin, P. Viaroli, M. Naldi &
the organic matter conversion and determine a distur- I. Ferrari, 1996. Nitrogen cycling networks of coastal ecosys·
terns: influence of trophic status and primary producer form.
bance in the ecosystem functioning (e.g. dystrophy). Ecol. Model. 87 (in press).
Moreover, biogeochemical reactions (e.g. precipitation Cline, J. D., 1969. Spectrophotometric determination of hydrogen
of metal sulphide) can exert a bottom-up control lower- sulphide in natural waters. Limnol. Oceanogr. 14: 454-459.
119
Crawford, R. M. M., 1992. Oxygen availability as an ecological Rizzo, W. M., S. K. Dailey, G. 1. Lackey, R. R. Christian, B. E. Berry
limit to plant distribution. Adv. eeo!. Res. 23: 93-185. & R. L. Wetzel, 1995. A metabolism-based trophic state index
Duarte, C. M., 1992. Nutrient concentration of aquatic plants: pat- for comparing the ecological values of shallow water sediment
terns across species. Limno!. Oceangr. 37: 882- 889. habitats. Estuaries (in press).
Enriquez, S., C. M. Duarte & K. Sand-1ensen, 1993. Patterns in Sand-Jensen, K. & J. Borum, 1991. Interactions among phytoplank-
decomposition rates among photosynthetic organisms: the impor- ton, periphyton, and macrophytes in temperate freshwater and
tance of detritus C:N:P content. Oecologia 94: 457-471. estuaries. Aquat. Bot. 41: 137-175.
Escaravage, v., 1990. Daily cycles of dissolved oxygen and nutri- Sfriso, A., B. Pavoni, A. Marcomini & A. A. Orio, 1992. Macroal-
ent content in a shallow fishpond: the impact of water renewa!. gae, nutrient cycles, and pollutants in the lagoon of Venice. Estu-
Hydrobiologia 207: 131-136. aries IS: 517-528.
Fong, P., T. C. Foin & 1. B. Zelder, 1994. A simulation model of Sherman. K., 1994. Sustainability, biomass yields, and health of
lagoon algae based on nitrogen competition and internal storage. coastal ecosystems: an ecological perspective. Mar. Eco!. Progr.
Eeo!. Monogr. 64: 225-247. Ser. 112: 277-301.
Giesen, W. B. 1. T., M. M. Van Katwijk & c. Den Hartog, 1990. Smith, D. W. & A. 1. Horne, 1985. Experimental measurement
Temperature, salinity, insolation and wasting disease of eelgrass of resource competition between planktonic microalgae and
(Zostera marina L.) in the Dutch Wadden Sea in the 1930's. Neth. macroalgae (seaweeds) in mesocosms simulating the San Fran-
1. Sea Res. 25: 395-404. cisco Bay-Estuary, California. Hydrobiologia 159: 259-268.
Godshalk, G. L. & R. G. Wetzel, 1978. Decomposition of aquatic Stevenson, 1. c., 1988. Comparative ecology of submersed grass
angiosperms. III. Zostera marina L. and a conceptual model of beds in freshwater, estuarine, and marine environments. Limno!.
decomposition. Aquat. Bot. 5: 329-354. Oceanogr. 33: 867-893.
Goering, H. K. & P. 1. Van Soest, 1970. Forage fibre analysis (appa- Thimel, A. & P. 1. Labourg, 1992. In situ metabolism of a benthic
ratus, reagents, procedure and some applications). U.S. Dept. community in a shallow brackish lagoon (Fish-impoundments of
Agric., Agric. Res. Ser., Agric. Hanab., 379 pp. Arcachon Bay, Atlantic coast, France). Vie Milieu 42: 185-192.
Izzo, G. & V. Hull, 1991. The anoxic crises in dystrophic pro- Twilley, R. R., W. M. Kemp, K. W. Staver, J. C. Stevenson &
cesses of coastal lagoons: an energetic explanation. In C. Rossi W. R. Boynton, 1985. Nutrient enrichment of estuarine sub-
& E. Tiezzi (Eds), Ecological Physical Chemistry. Elsevier Sci. mersed vascular plant communities. Algal growth and effects
Pbls. Amsterdam: 559-572. on production of plants and associated communities. Mar. ECD!.
Koroleff, F, 1970. Direct determination of ammonia in natural Prog. Ser. 23: 179-191.
waters as indophenol blue. Information on techniques and meth- Valderrama, 1. c., 1977. Methods used by the Hydrographic Depart-
ods for seawater analysis. I.C.E.S. Interlaboratory Rep. No.3: ment of National Board of Fisheries, Sweden. In K. Grasshoff
19-22. (ed.), Report of the Baltic Intercalibration Workshop. Interim
Kuhl, M. & B. B. J"srgensen, 1992. Microsensor measurements of Commission for the Protection of the Environment of the Baltic
sulfate reduction and sulfide oxidation in compact microbial com- Sea: 14-34.
munities of aerobic biofilms. App!. envir. Microbio!. 58: 1164- Van Gemerden, H., C. S. Tughan, R. de Wit & R. A. Herbert, 1989.
1174. Laminated microbial ecosystems on sheltered beaches in Scapa
Mann, K. H., 1988. Production and use of detritus in various freshwa- Flow, Orkney Islands. FEMS Microbio!. Eco!. 62: 87-102.
ter, estuarine, and coastal marine ecosystems. Limno!. Oceanogr. Verhoeven, 1. T. A., 1980. The ecology of Ruppia- dominated com-
33: 910-930. munities in Western Europe. II. Structure and dynamics of the
Nienhuis, P. H., 1992. Eutrophication, water management, and the macroflora and macrofauna communities. Aquat. Bot. 8: 1-85.
functioning of Dutch estuaries and coastal lagoons. Estuaries IS: Viaroli, P., M. Bartoli, C. Bondavalli, M. Cattadori, G. Giordani &
538-548. M. Naldi, 1994. Oxygen, sulphide and nutrient fluxes in shallow
Odum, W., 1990. Internal processes influencing the maintenance of eutrophic lagoons with different primary producer communities.
ecotones: do they exist? In R. 1. Naiman & H. Decamps (eds), I. Macrophyte communities and their impact on benthic fluxes.
The Ecology and Management of Aquatic- terrestrial Ecotones. In P. Caumette (Coord.), C.L.E.A.N. Progress Report, European
MAB 4. The Parthenon Publishing Group, Paris: 91-102. Commission: 327-352. European Community Environment Pro-
Philippart, C. 1. M., 1995. Seasonal variation in growth and biomass gramme, DG XII, Brussels.
of an intertidal Zostera noltii stand in the Dutch Wadden Sea. Viaroli, P., M. Bartoli, C. Bondavalli & M. Naldi, 1995. Oxygen
Neth. 1. Sea Res. 33: 205-218. fluxes and dystrophy in a coastal lagoon colonized by Viva rigida
Pugnetti, A., P. Viaroli & I. Ferrari, 1992. Processes leading (Sacca di Goro, Po River Delta, Northern Italy). Fresenius Envir.
to dystrophy in a Po River Delta lagoon (Sacca di Goro): Bull. 4: 381-386.
phytoplankton-macroalgae interactions. Sci. tot. Envir. supp!. Williams, S. L. & M. H. Ruckelshaus, 1993. Effects of nitrogen
1992: 445-456. availability and herbivory on eelgrass (Zostera marina) and epi-
phytes. Ecology 74: 904-918.
Hydrobi%gia 329: 121-131, 1996. 121
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Key words: Zostera noltii, Monostroma obscurum, anaerobic decomposition, nutrient recycling
Abstract
Arcachon Bay is characterized by extensive meadows of the seagrass Zostera noltii. Moreover, as a consequence of
eutrophication, massive proliferations of the macroalga (Monostroma obscurum) have occurred since the beginning
of 1990s.
This paper describes the anaerobic decomposition of biomass of both species under experimental conditions
by two methods. Firstly, the dynamics of decomposition were studied in situ using litter bags. The remaining
biomass and the elemental composition of the decomposing macrophytes were monitored. Secondly, degradation
was studied in experimental containers under anoxic conditions in which the release of inorganic nutrients and the
development of fermentative and sulfate-reducing bacterial populations were followed.
The decomposition rate of total biomass was faster for macroalgae than for the vascular plants, thus corroborating
previous observations. However, both in situ and laboratory experiments showed that the anaerobic decomposition
of the seagrass Z. noltii resulted in rapid release of inorganic Nand P, and increasing CIN and CIP ratios of the
residual biomass. As a result, the recycling of inorganic nitrogen and phosphorus compounds was slightly more
efficient for Z. noltii than for M. obscurum. Recycling of inorganic nutrients appears to be of a great importance to
the whole ecosystem, because of the extensive spreading of Z. noltii in the bay.
Stat ion Z
Figure 1. a. Distribution of Zostera nO/Iii (pointed) and Zostera marina (black) meadows in Arcachon Bay and location of the study site Z.
b. Repartition of Monostroma obscurum biomass in Arcachon Bay (July 1993). pointed: 0-10 g fresh weight m- 2 (intertidal) or m- 3 (subtidal)
hatched: 10-500 g fresh weight m- 2 (intertidal) or m- 3 (subtidal) black: > 500 g fresh weight m- 2 (intertidal) or m- 3 (subtidal).
Grazing by herbivores on Z. noltii is quantitatively et aI., 1994). In addition, the export of this macroal-
of little importance, as comparable to other vascular ga by currents is presumably of minor importance,
plants. The major herbivores are popUlations of the because its production takes place predominantly in
dark-bellied Brent geese (Branta bernicla bernicla) the confined parts of the bay. As a result, a large
which overwinter in the lagoon. Although numerous fraction enters the detritus food chain and is min-
(up to 40000 birds), their consumption represents at eralized in the sediments. Macroalgae are therefore
most 2% of the above-ground annual production of important in nutrient cycles in coastal lagoons, because
Z. noltii (Auby, 1991). A small part of the biomass is they act as both nutrient sinks and sources, during
exported out of the bay by the currents. However, the growth and decay phases, respectively (Lavery &
importance of this biomass export remains unquanti- McComb, 1991). The relatively fast decomposition
fied. Hence, a large part of the Z. noltii detritus decom- rate of macroalgae provides also a rapid recycling of
poses at the sediment surface or in the sediment itself, nutrients which may be an important factor to maintain
supplying large quantities of organic matter to the sys- high productivity (Hanisak, 1993).
tem. The slow decomposition rates characteristic of This paper reports the results of a comparative
different seagrass species have been described exten- study of the decomposition of Z. noltii and M. obscu-
sively in the literature (Harrisson & Mann, 1975; Thay- rum detritus under anoxic conditions. The research
er et aI., 1977; Godschalk & Wetzel, 1978; Pellikaan, was conducted by both in situ and laboratory incu-
1984). This feature is often attributed to its relatively bation experiments. The decay rates were compared
high content of refractory constituents, such as struc- and attention was paid to the importance of inorganic
tural carbohydrates, lignin and cellulose, as well as to nutrient recycling.
the C:N:P ratios (Enriquez et aI., 1993).
The macro alga M. obscurum constitutes a food
source for gastropods and amphipods. Nevertheless,
it appears that the grazing rate is extremely low (Auby
123
Sites of sampling and in situ experiments Dry weight of macrophyte detritus was determined
after drying at 60°C for 48 hours. Analysis of Car-
In 1992-1993, Z. noltii leaves and M. obscurum thalli bon, Nitrogen and Phosphorus content were performed
were sampled at station Z which was located at mean at the Institut Europeen de l'Environnement de Bor-
tide level (M.T.L.) on an intertidal mudflat, in Arca- deaux (IEEB). Organic carbon content was measured
chon Bay (Figure 1). The in situ decomposition experi- according to Le Corre (1983), nitrogen content with
ments were performed adjacent to the sampling station. the Kjeldahl method (Allen, 1970) and phosphorus
The freshly collected macrophytes were transport- content by a colorimetric method after acidic digestion
ed to the laboratory in cool boxes and immediately with HN03-H2S04 (APHA, AWWA, WPCF, 1980).
sorted and rinsed with sea water. During spring 1992, Water samples were filtered (Whatman membrane,
decomposition experiments were conducted both with 0.45 /.lm pore size) and stored frozen in polyethylene
leaves of Z. noltii and thalli of M. obscurum; addition- vials until analysed for NHt, NO;- +N03 and PO~
al experiments were performed in other seasons with using a Technicon autoanalyser at the Laboratory of
M. obscurum thalli only. At the start of the experiments, CREMA (l'Houmeau), according to Treguer & Le
replicate litter bags of 1 mm nylon mesh (21 x 21 cm) Corre (1975). Sulfide concentrations were determined
were filled with the same amount of freshly harvested colorimetrically according to Cline (1969).
and rinsed leaves or thalli (30 to 50 g wet weight) and
buried 1 cm deep in the sediment. Minielectrode mea- Bacterial enumerations
surements showed that permanently anoxic conditions
occurred at these sediment depths. The entire in situ Sulfate-reducing bacteria (SRB) were enumerated
experiments covered the periods listed in Table I. At using the Most Probable Number technique in liq-
selected time intervals, one or two bags were harvested uid medium with 10 mM sodium lactate and 10 mM
for subsequent analyses of remaining dry weight and sodium acetate (Pfennig et aI., 1981). The medium
elemental composition (i.e., C, Nand P contents). Pri- was dispensed aseptically in Hungate's tubes and the
or to these analyses, the harvested detritus was gently growth of SRB was determined by the formation of a
rinsed with sea water to remove sediment particles and brown precipitate after addition of CuS04/HCI solu-
subsequently with fresh water to remove salt. tion (5 mM/50 mM) (Caumette, 1986).
Fermentative bacteria were enumerated by count-
Laboratory experiments ing Colony Forming Units (C.F.U.) on agar plates.
The growth medium consisted of filtered (0.22 /.lm)
Laboratory experiments were performed to investigate sea water, 5 g peptone, 5 g yeast extract, 15 g agar-
inorganic nutrient release. To simulate environmen- agar. The pH was adjusted to 7.2-7.4. The medium
tal anoxic dark conditions, aquaria (21 x 21 x 22 cm) was autoclaved and dispensed into Petri dishes, which
were wrapped in aluminium foil and sealed with sili- were spread using a portion from aseptically prepared
cone to prevent oxygen ingress. A volume of 30 cm 3 decimal dilution series and incubated in anaerobic jars.
of sediment from station Z was added, the aquaria
completely filled with seawater (9.7 I) and amend- Curve-fitting
ed as follows: (1) control treatment containing sedi-
ment alone. (2) Sediment plus 100 g wet weight of The time courses of in situ biomass decomposition
M. obscurum (equivalent to 10.78 g dry weight), and were modelled according to first order kinetics using
(3) Sediment plus 100 g wet weight of Z. noltii (equiv- the exponential function:
alent to 14.23 g dry weight). Treatments 2 and 3 were
(Eq.l )
established in duplicate. Incubation was performed at
ambient laboratory temperature (20-25 0c) for about or according to the logistic function:
2 months. During incubation, water samples for chem-
ical analyses and bacterial enumerations were with- Wo
(Eq.2)
drawn through a sample port with syringes (50 ml) to
Wt = I +bx e ki '
WINTER
(18-42) 1993 44% in 30 days not significant 9.3°C 5.17
M.obscurum
is the specific degradation rate (per day) and b is a M. obscurum biomass was low, but increased substan-
coefficient specific to the logistic function only, which tially after a lag period of 15 days. The initial dry
is related to a time delay. Curve fitting was achieved by weight of 5.40 g decreased by 76% to 1.30 g over the
an iterative procedure to minimize the sum of squares 25 days incubation period. For Z. noltii, the initial dry
of the residuals (Sokal & Rohlf, 1981). weight of 7.11 g decreased to 1.64 g after 66 days,
corresponding to a loss of77%. Thus, the phanerogam
biomass decomposed to a similar extent in a period
Results which was double that found for M. obscurum.
The decomposition processes of macrophytes can
In situ experiments generally be modelled by first order kinetics (exponen-
tial function) or if a significant time lag is observed,
In situ time course experiments of decomposition of by a logistic function according to equation 1 and
Z. noltii and M. obscurum biomass differed consid- 2, respectively (see Method section). We found that
erably (Figure 2). Initially the decomposition rate of first order kinetics provided a good description for
125
100~-----------------------------------,
100
SUMMER
o~ 60
0>
c:
'2
'~ 40
£
;f!. 20
o,
o 10 20 30 40 50 60 70
o 5 10 15 20 25 30 35 40
Day!.
Days
Figure 2, Loss of dry weight of Monostroma obscurum (open trian-
gles) and Zostera no/tii (solid circles) in litter bags, as percentage of i..-, _ _ _ _ _ .:D.=__ D.
remaining dry weight. Spring experiment (March 1992), 100 A
AUTUMN
:g, 80
~
the decomposition of Z. noltii but not for M. obscu- "& 60
rum which was best modelled by the logistic equation. g'
C
Accordingly, the calculated times required for 50% loss .~ 40
<D
of dry weight were 20 and 37 days for M. obscurum CI:
20.--------------------------------,
50~------------------------------~
Z. no/Iii
~--.------
15
~40
~;,g o
:§
Z. noltii
::- 30 o 10
c 5
c
CD
z
8 20 M. obscurum o M. obscurum
c 5
o
.0
810
O~~~~~~~~~~~~~~~~~
O~~~~~~~~~~~~~~TT~rl
o 10 20 30 40 50 60 70
Days
o 10 20 30 40 50 60 70
Days Figure 5. CIN atomic ratios during the spring decomposition of
Monostroma obscurum (open triangles) and Zostera noltii (solid
circles) in litter bags.
8
~'
~8 during the first two weeks while total biomass remained
;,g
~ virtually constant (lag phase). Afterwards, whilst a
cCD rapid loss of remaining biomass occurred, N-content
c4
0
<.> increased again to 6% (after 21 days) (Figure 4 and
c Table 1). In contrast, a minor but continuing decrease
CD
g2
Cl
of nitrogen content was found for Z. noltii.
Z Initial phosphorus contents were 0.10% and 1.10%
for M. obscurum and Z. noltii, respectively. Never-
0 theless, P-content of the latter appeared to be excep-
0 10 20 30 40 50 60 70
Days
tionally high in March 1992. Comparably, Viaroli
(pers. comm.) found values ranging from 0.2 to 0.4%
in May and August 1993 at station B in Arcachon
1.2.,-------------------------------__---,
Bay, which is a more sheltered than station Z. After
a week of incubation, P-content was 0.25% and then
decreased further to 0.087%. In contrast, phosphorus
content of M. obscurum detritic material increased dur-
ing the decay process.
At the beginning of the spring experiment, the
CIN ratios were 8.6 and 12.6 for M. obscurum and
Z. noltii, respectively (Figure 5). Algal detritus CIN
ratio increased during the first two weeks of the exper-
imental period and then decreased (CIN = 5.39 after 25
days). On the other hand, the CIN ratio of seagrass
10 20 30 40 50 60 70 leaves detritus gradually increased during the incuba-
Days tion up to 15.75.
The similarity of the decomposition process of
Figure 4. Carbon. nitrogen and phosphorus content as percentage M. obscurum in spring and autumn was also confirmed
(weight/weight) of dry weight during the spring decomposition of by the time course of the CIN ratio, which followed the
Monostroma obscurum (open triangles) and Zostera no/tii (solid
circles) in litter bags.
same trend. In fact, CIN, in autumn, equaled to 4.43 at
the beginning of the experiment. This ratio increased
up to 5.18 during the two first weeks and then decreased
(CIN =3.82 after 30 days).
127
4000 4000
:?3000 :?3000
a a
c c
0 0
.;; 2000 .~ 2000
E
<l>
C
Q)
"c:
81000
"
c:
81000
0
0 10 20 30 40 50 60 70 10 20 30 40 50 60 70
Days Days
40 40
:?30
a
/ :? 30
ac:
c:
~ 20
// g20
0
E
<l>
/ ai
"c:
810
"
c:
810
nilrite+nitrate
10 20 30 40 50 60 70 10 20 30 40 50 60 70
Days Days
200 200
:?150 :?150
a a
c: c:
.!2 .!2
~
100 e 100
Q) ~
"0
c:
50
"c:
0
50
() phosphate () phosphate
D.... 0.
0 0
0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70
Days Days
20 20
:? 15 :? 15
.s.c .s.c:
0
.~ 10 ~ 10
C C
""c:
Q)
"c:
0 5 0 5
() ()
0
0 10 20 30 40 50 60 70 10 20 30 40 50 60 70
Days Days
Figure 7. Variation of ammonium (solid and open squares), nitrite + nitrate (solid and open triangles), phosphate (solid and open circles) and
sulfide (solid and open lozenges) concentrations in the water of aquaria containing M. obscurum and Z. no/tii.
129
be explained by efficient P-release and the higher not be explained by this factor. Phosphorus appears to
P-content. Maximal phosphate concentration reached be present in labile form in Z. noltii and its leach-
16.6 11M in the M. obscurum containers after the two ing is very efficient, confirming the results of Auby
first weeks of incubation but this value decreased with (1991). Another factor influencing in situ decomposi-
time to 0 after 55 days. A concentration of 87.4 J-tM tion rates may be detritus fragmentation. In fact, Chi-
was observed in the Z. noltii aquarium water at the end ronomus larvae were detected from the first week in
of incubation. Conceivably, the 'mass-balance' calcu- the litter bags containing decaying M. obscurum but
lations for phosphorus are more difficult than for nitro- were never found in the Z. noltii bags. The mechanical
gen compounds, because of the relative importance of activity, facilitating the detritus breakdown in parti-
P release or binding to sediment particles. cles of small size, is likely an important factor for
In both macrophyte incubations, H2S was detected the colonization by microorganisms (Pellikaan, 1982).
from the first day of incubation. H 2S concentrations Z. noltii leaves are composed of more than 40% of
increased gradually with time up to 10.4 mM in the cellulose, hemicellulose and lignin (relative to total
M. obscurum container and to 16.3 mM in the case of dry weight: 14.6%, 19.1% and lO.2%, respectively;
Z. noltii. Auby, 1991). Macroalgae are characterized by low-
The number of sulfate-reducing bacteria (SRB) er contents of these components which are between
increased with time in all treatments (Table 2). Howev- lO% and 34% for Viva lactuca Lin. (Brouard, 1983;
er, the increase was higher when macrophyte material Faucille, 1984). This difference in refractory structural
was present. After 43 days of incubation SRB num- constituents can explain the slow decay of phanerogam
bers reached 5.5 x 107 and l.9 x 106 bacteria/ml for relative to macroalgal decomposition. In addition, vas-
M. obscurum and Z. noltii containers, respectively. In cular plants are known to produce toxic substances,
contrast, trends in the density of fermentative bacteria which are liberated early during the decomposition
were different Fermentative bacteria were present in process, thus causing a delay of microbial coloniza-
high numbers at the start but did not increase signifi- tion resulting in a slower degradation rate. It has been
cantly in the Z. noltii combination, whereas a signif- reported thatZ. marina contained bioactive compounds
icant increase was detected in the M. obscurum con- (phenolic acid and flavonoids) shown to inhibit micro-
tainer. bial hydrolytic enzymes (Horner et aI., 1988; Swain,
1979).
Three phases are generally recognized in seagrass
Discussiou decomposition processes based on changes in decay
rate with time (Rice & Tenore, 1981; Pellikaan, 1982).
In situ, the decomposition rate of M. obscurum was During the first phase a rapid loss of dry weight does
about twice as fast as for Z. noltii. Even slower degra- occur, presumably corresponding to the leaching of
dation rates have been found for Z. noltii (Hemminga soluble compounds. The second phase is characterized
& Nieuwenhuize, 1991). M. obscurum decomposition by microbial colonization, reflected by an increase of
rates were however similar to those reported for the nitrogen content. Finally, the third one corresponds to
green alga Caulerpa cupressoides (Williams, 1984). the loss of refractory material. These 3 phases should
Our findings thus corroborate previous studies that imply corresponding changes in C/N ratio. Neverthe-
decomposition of marine vascular plants is a relative- less we were not able to detect these stages during in
ly slow process (Fenchel, 1977; Harrisson & Mann, situ Z. noltii decay, as it followed first order kinetics
1975; Thayer et aI., 1977; Godschalk & Wetzel, 1978; and C/N ratios increased throughout the experiment.
Pellikaan, 1984). Unambiguous interpretations of the time course of C/N
According to Rice & Tenore (1981), the rate of ratios are, however, difficult, because the leaching of
decomposition is largely dependent on the chemical soluble compounds, element enrichment of the detri-
composition of the macrophyte. Macroalgal tissues tus by microbial colonization and protein binding reac-
are commonly richer in nitrogen than vascular plant tions may counteract whilst they occur simultaneously
material (Duarte, 1992). The nitrogen content mea- (Pellikaan, 1984).
sured in spring for both macrophytes were however Surprisingly, degradation of M. obscurum followed
rather similar (4.45% and 3.90% for M. obscurum first order kinetics only in summer and a lag peri-
and Z. noltii, respectively). The difference between od of two weeks has been found both in spring and
M. obscurum and Z. noltii decomposition rates can thus autumn. During the lag phase a decrease in N-content
130
took place, whilst this parameter increased again after centrations released from both decaying macrophytes
degradation had started at a high rate. were also very high. The release of inorganic phos-
In contrast to the in situ observations, the labo- phorus from the sediment, favoured by the reducing
ratory decomposition experiments showed an almost conditions prevailing in the containers, could partially
equal loss of dry weight for both macrophytes (50.8% explain the quantities measured during our experiment.
and 46.8% for M. obscurum and Z. noltii, respective- The regeneration, due to the in situ decomposition of
ly) at the end of incubation period (55 days). Thus, Viva rigida under anaerobic conditions, accounted for
for M. obscurum, the decay process studied in the lab- more than 50% of the increase of phosphorus concen-
oratory was slower than observed in situ. Firstly, the trations observed in the Sacca di Ooro lagoon during
high sulfide concentrations measured in the overlying dystrophic crises, when particulate + dissolved phos-
water in the experimental containers partially explain phorus reached values of 14.5 flM in the water column.
this difference. H 2 S is toxic for many bacterial species, However, nitrogen regeneration was less efficient in
but in the field this compound diffuses away and can be the latter system, thus causing an imbalance between
trapped by sedimentary iron. The iron content in Arca- Nand P release (Viaroli et ai., 1992).
chon Bay is particularly high (Stal et ai., 1996; Viaroli In Arcachon Bay, the release of inorganic nutrients
et ai., 1996) and minielectrode measurements show from decaying detritus is of a great importance for both
that free H 2 S is rarely present or if present at a con- macrophytes. According to Iizumi et ai. (1982), the
centration below 0.5 mM (data not shown). Second- mechanism of ammonium regeneration from organ-
ly, Chironomus larvae were never found in laboratory ic nitrogen in sediments with eelgrass beds consti-
containers, whereas they were present in M. obscu- tutes the main pathway for supplying nitrogen to the
rum litter bags (see above), thus suggesting that the whole ecosystem. Our study demonstrated the poten-
degradation of this macrophyte is indeed enhanced by tial importance of the decomposition of the domi-
detritus fragmentation caused by these animals. nant macrophytes of Arcachon Bay, mainly Z. noltii
In the laboratory experiments, populations of phanerogam, as an important nitrogen and phospho-
sulfate-reducing bacteria were well developed and rus source in the ecosystem. Nitrogen recycling may
actively participated in the decomposition. Assuming become less efficient due to losses caused by cou-
that two moles of sulfide are produced for each mole pled nitrification-denitrification processes. However,
of carbon mineralized (J0rgensen, 1977), we have esti- coupled nitrification denitrification were inhibited in
mated that sulfate reduction accounted for around 75% our experiments, but were also very low at compa-
of the mineralization of macrophyte carbon. rable intertidal stations in the bay (Rysgaard et ai.,
The laboratory experiments demonstrated that 1996). Thus, this coupled process is likely not impor-
anaerobic decomposition of both M. obscurum and tant in affecting N-regeneration in the field. Seagrasses
Z. noltii results in a high inorganic nutrient release. appear to be the most important factor in determining
Water column concentration of dissolved Nand P nutrient availability because of their huge biomass in
species reflect the net result of sediment flux and this ecosystem. We demonstrated that the release of
macrophyte nutrient release minus the uptake due to nutrients from Zostera noltii was a relatively fast pro-
bacterial processes. We found that the concentrations cess. Monostroma obscurum, characterized by a faster
of inorganic N-compounds in the water column in decay enhances this nutrient recycling even though it
containers amended with M. obscurum or Z. noltii accounts only for the 20% of the total plant biomass.
provide a good measure of the N-recycling during
the decomposition process. A more complicated pat-
tern was observed for dissolved phosphate, because Acknowledgments
this compound reached a maximum and subsequently
decreased. The authors highly acknowledge Miss M. Durin for her
Ammonium was the dominant form of dissolved contribution to the Zostera noltii degradation experi-
nitrogen released from decaying macrophyte detritus ment and Dr A. Herbland, director of the laboratory of
in marine environments. Both Williams (1984) and CREMA (L'Houmeau, La Rochelle, France), for the
Hanisak (1993) reported the same conclusions. Due use of the Technicon autoanalyser. We are also grate-
to the anoxic conditions in the containers, bacterial ful to the staff of IFREMER Arcachon, especially to
nitrification was arrested and ammonium was therefore O. Trut for his kindness and assistance during nutri-
not converted into nitrite and nitrate. Phosphate con- ents measurements and his valuable help for drawing
131
Figure 1. Finally, we thank E. Buffan, Y. Dubau for !izumi, H., A. Hattori & C. P. McRoy, 1982. Ammonium regene-
helping us in the data presentation and D. Welsh for ration and assimilation in eelgrass (Zostera marina) beds. Mar.
BioI. 66: 59-65.
his help with the English presentation. This work was J~rgensen, B. B., 1977. The sulfur cycle of a coastal marine sediment
supported by a E. U. Environment programm (CLEAN (Limfjorden, Denmark). Limnol. Oceanogr. 22: 814-832.
contract EV5V-CT92-0080). Lavery, P. S. & A. J. McComb, 1991. Macroalgal-sediment nutrient
interactions and their importance to macroalgal nutrition in a
eutrophic estuary. Estuar. coast. shelf Sci. 32: 281-295.
Le Corre, P., 1983. Dosage du carbone organique particulaire. In
References CNEXO (ed.), Manuel des analyses chimiques en milieu marin.
BNDOlDocumentation, Brest (France): 203-208.
Allen, S. E., 1970. Chemical analysis of ecological materials. Black- Pellikaan, G. c., 1982. Decomposition processes of eelgrass, Zostera
well Scientific Publications, Oxford: 185-186. marina L. Hydrobiol. Bull. 16: 83-92.
APHA, AWWA, WPCF. 1980. Standard methods forthe examination Pellikaan, G. c., 1984. Laboratory experiments on eelgrass (Zostera
of water and wa~tewater. 15th edn. 1134 pp. marina L.) decomposition. Neth. J. Sea Res. 18: 360-383.
Auby, I., 1991. Contribution 11 I' etude des herbiers de Zostera noltii Pfennig, N., F. Widdel & H. G. Triiper, 1981. The dissimilatory
dans Ie Bassin d' Arcachon. These Doctorat, Universite Bordeaux sulfate-reducing bacteria. In M. P. Starr, H. Stolp, H. G. Triiper,
1,234 pp. A. Balows & H. G. Schlegel (eds), The prokaryotes, a handbook
Auby, I. & P. J. Labourg, 1996. Seasonal dynamics of Zostera noltii on habitats, isolation and identification of bacteria. Springer-
Hornem. in the Bay of Arcachon (France). J. Sea Res. (in press). Verlag, Berlin, vol. I: 926-940.
Auby, I., F. Manaud, D. Maurer & G. Trut, 1994. Etude de la pro- Ribes, E., 1988. Contribution 11 ['etude de la proliferation des
liferation des algues vertes dans Ie Bassin d' Arcachon. Rapport algues vertes dans Ie Ba~sin d' Arcachon. Contrat I.F.R.E. MER
SIBA: 163 pp. 875527053. 31 pp.
Brouard, F., 1983. Digestion aerobie de la biomasse vegetale aqua- Rice, D. L. & K. R. Tenore, 1981. Dynamics of carbon and nitro-
tique. These de Docteur Ingenieur, INSA Toulouse, 175 pp. gen during the decomposition of detritus derived from estuarine
Caumette, P., 1986. Phototrophic sulfur bacteria and sulfate- macrophytes. Estuar. coa~t. shelf Sci. 13: 681-690.
reducing bacteria causing red waters in a shallow brackish coastal Rysgaard, S., N. Risgaard-Petersen & N. P. Sloth, 1996. Nitrification,
lagoon. FEMS Microbiol. Ecol. 38: 113-124. denitrification and nitrate ammonification in sediments of two
Cline, J. D., 1969. Spectrophotometric determination of hydrogen coastal lagoons in Southern France. Hydrobiologia 000 (Dev.
sulfide in natural waters. Limnol. Oceanogr. 14: 454-458. Hydrobiol. 000): 000-000.
Duarte, C. M., 1992. Nutrient concentration of aquatic plants: pat- Sokal, R. R. & F. J. Rohlf, 1981. Biometry. W. H. Freeman &
terns across species. Limnol. Oceanogr. 37: 882-889. Company, San Fransisco, 859 pp.
Durieu de Maisonneuve, M., 1855. Notes detachees sur quelques Stal, L. J., S. B. Behrens, M. Villbrandt, S. van Bergeijk & F. Kruyn-
plantes de la flore de la Gironde, et description d'une nouvelle ing, 1996. The biogeochemistry of two eutrophic marine lagoons
espece d'Avena. Actes Soc. linn. Bordeaux. 20: 1-83. and its effect on microphytobenthic communities. Hydrobiologia
Enriquez, S., C. M. Duarte & K. Sand-Jensen, 1993. Patterns in (this volume).
decomposition rates among photosynthetic organisms: the impor- Swain, T., 1979. Tannins and lignins. In G. E. Rosenthal &
tance of detritus C:N:P content. Oecologia 94: 457-471. D. H. Janzen (eds), Herbivores: their interactions with secondary
Faucille, S., 1984. Digestion anaerobie de vegetaux aquatiques. plant metabolites. Academic Press, New York: 657-{582.
Vegetaux marins et d'eau douce. These Institut national Poly- Thayer, G. w., D. W. Engel & M. W. Lacroix, 1977. Sea~onal
technique de Lorraine, 100 pp. distribution and changes in the nutritive quality of living, dead
Fenchel, T., 1977. Aspects of the decomposition of seagrasses. In and detrital fractions of Zostera marina L. J. expo mar. BioI. Ecol.
C. P. McRoy & C. Helfferich (eds), Seagrasses ecosystems: a 30: 109-127.
scientific perspective. Marcel Dekker, New York: 123-145. Treguer, P. & P. Le Corre, 1975. Manuel d'analyse des sels nutritifs
Godschalk, G. L. & R. G. Wetzel, 1978. Decomposition of aquatic dans J'eau de mer (utilisation de l'autoanalyseur II Technicon
angiosperms. c. Zostera marina L. and a conceptual model of R). 2e ed. Laboratoire d'Oceanologie Chimique, Universite de
decomposition. Aquat. Bot. 5: 329-354. Bretagne occidentale. 110 pp.
Hanisak, M. D., 1993. Nitrogen relea~e from decomposing sea- Viaroli, P., M. Bartoli, C. Bondavalli, R. R. Christian, G. Giordani
weeds: species and temperature effects. J. appl. Phycol. 5: 175- & M. Naldi, 1996. Macrophyte communities and their impact
181. on benthic fluxes of oxygen, sulphide and nutrients in shal-
Harrisson, P. G. & K. H. Mann, 1975. Detritus formation from eel- low eutrophic environments. Hydrobiologia 329 (Dev. Hydro-
grass (Zostera marina L.): the relative effects of fragmentation, bioI. 117): 105-119.
leaching and decay. Limnol. Oceanogr. 20: 924-934. Viaroli, P., A. Pugnetti & I. Ferrari, 1992. Viva rigida growth and
Hemminga, M. A. & J. Nieuwenhuize, 1991. Transport, deposition decomposition processes and related effects on nitrogen and phos-
and in situ decay of seagrasses in a tropical mudflat area (Banc phorus cycles in a coa~tallagoon (Sacca di Goro, Po River Delta).
d'Arguin, Mauritania). Neth. J. Sea Res. 27: 183-190. In G. Colombo, I. Ferrari, V. U. Ceccherelli & R. Rossi (eds),
Horner, J. D., J. R. Gosz & R. G. Cates, 1988. The role of carbon- Marine eutrophication and population dynamics. Olsen & Olsen,
based secondary metabolites in decomposition. Am. Nat. 132: Fredensborg (Denmark): 77-84.
869-883. Williams, S. L., 1984. Decomposition of the macroalga Caulerpa
cupressoides (West) c. Agardh: field and laboratory studies. J.
expo mar. BioI. Ecol. 80: 109-124.
Hydrobiologia 329: 133-141, 1996. 133
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Key words: Nitrification, denitrification, nitrate, ammonification, sediments, isotope, 15N, ammonium, flux
In temperal coastal areas, nitrogen typically limits periods of dystrophic crisis during summer, when most
primary productivity, which makes processes reduc- of the water column goes anoxic. The sampling site
ing nitrogen availability of primary concern in revers- (Station 11) was located at 60 cm water depth in the
ing the effects of eutrophication. This can be accom- western and very sheltered part of the lagoon.
plished by either lowering the input of combined nitro- Undisturbed sediment cores were collected at each
gen, increasing water exchange, or managing the area locality in June and September 1993 and in January
in a way where nitrate removal by denitrification is 1994. In the Etang du Prevost, 30 cm long Plexiglas
enhanced. However, it is important to gain detailed core tubes with an i.d. of 5.2 cm were used for sediment
information about nitrogen cycling in different ecosys- sampling and incubation, whereas 40 cm long core
tems, suffering from different eutrophication stress, if tubes with an i.d. of 10 cm were used for sampling and
a successful management is to be obtained. incubation of sediment cores in the Bassin d' Arcachon
In this paper we describe the benthic nitrogen trans- to contain intact eelgrass plants.
formations in two coastal lagoons of southern France:
Bassin d' Arcachon on the Atlantic coast and Etang Flux and denitrification measurements
du Prevost on the Mediterranean coast. The biolog-
ical structure of the two lagoons was very different Within a few hours of sampling, the sediment cores
due to different nutrient loading and hydrodynamics. were transported to the laboratory and placed in an
Bassin d' Arcachon was dominated by benthic prima- open tank filled with air-saturated in situ water and
ry producers, thus representing less eutrophic con- maintained at in situ temperatures. The sediment in
ditions whereas Etang du Prevost was dominated by all cores was adjusted to an equal height of 8 cm,
pelagic primary producers and represented a dystroph- and a 3-cm long Teflon-coated magnet was suspended
ic ecosystem. 5 em above the sediment surface. When seagrass was
present, the magnet was placed 5 em above the leaves.
Momentum for rotation of the small magnets was pro-
Materials and methods vided by a large external magnet rotating at 60 r.p.m.
to ensure a homogenous mixing of the water column.
Experimental sites and sediment sampling Five cores were illuminated at a light irradiance of
400 fLmol photons m- 2 S-l in the 400-700 nm range,
Bassin d' Arcachon has an area of 155 km 2, with a 3-km and 5 cores were incubated in darkness.
wide connection to the sea. The tidal range in the area is Measurements of oxygen and combined nitrogen
2-3 m, and the bay is dominated by large tidal mudflats fluxes across the sediment-water interface were initi-
occupying about 70% of the total area. A thorough ated by closing the sediment cores with floating glass
description of the hydrology of the basin was published lids. The lids could be removed during the incubation
by Robert et al. (1987) who reported salinities varying period and water samples collected with a syringe.
from 28-35%0 in the outermost zone to 20-35%0 in Samples were collected 3 times at regular time inter-
the inner zone. The water temperature varies between vals during the incubation period. The total incubation
5 °C during winter and 25°C during summer. Most time ranged from 1 to 8 h, resulting in a 10-20% change
of the tidal flats are covered by the widgeon grass from the initial O 2 concentration. Samples for O 2
Zostera noltii Hornem. The sampling site (Station A) determinations were collected in glass vials (Exetain-
was located on a tidal flat in the central part of the basin er, Labco, High Wycombe, UK) and Winkler reagents
about 1 km north of the city of Arcachon. were added immediately (Strickland & Parsons, 1972).
The lagoon Etang du Prevost has an area of 4 km 2 Samples collected for ammonia and nitrate determina-
and is linked to the Mediterranean by a single 12 m tions were filtered and frozen in 20 ml plastic vials.
wide channel. The tidal range is about 20 cm, and The 15N experiment for determining anaerobic
most of the lagoon is permanently water-covered with nitrate reduction rates was initiated by adding K I5 N03"
an average depth of 150 cm. The salinity is uniform (99.9 atom%) to the water in the open incubation tank
throughout the lagoon, but varies between 25%0 during to a final concentration of 30 fLM 15N03". A syringe
winter and 42%0 during dry periods in the summer. The was used to exchange the water in each sediment core
temperature ranges from 5 °C during winter to 27°C with 15N03" -rich water from the reservoir in order to
during summer. The lagoon is characterized by large obtain a uniform mixing of the added isotope in all
spring blooms of green macroalgae (Viva sp.) and by cores. Initial samples from the water column were
135
taken after the water had equilibrated with the sedi- rates of 15N isotopes were calculated as
ment for 15 min. The core-tubes were then closed and 15N _ (cse~ - Cini)(q,~e~ + V2) + (Cwater - Cini)(VI - V2)
incubated as described above. Samples for analysis of p x - tA '
concentrations and 15N enrichment of NO} were col- (2)
lected from the water column in the core tubes with where p 15 N x is the production rate of the relevant iso-
a syringe and frozen in 20 ml plastic vials. Samples tope ('5NHt, 14NI5N or 15NI5N), Csed and CwateT are
for 15N analysis of N2 and NHt were collected from the concentrations of the isotope in the sediment-water
both the water column and the porewater. The latter suspension and the water column, respectively, Cini
was accomplished by carefully mixing the sediment is the initial concentration of the isotope, Vsed is the
and water column with a metal stick after the addition volume of the sediment core, <I> is the sediment poros-
of 250 p,1 7M ZnCh to the sediment surface .. Water ity, VI and V2 are the volumes of the water column
samples and sediment-water suspension samples for before and after water-samples are collected, respec-
N2 isotope analysis were preserved in gas-tight vials tively, t is the incubation time, and A is the surface
(Exetainer, Labco, High Wycombe, UK.) with 2% v/v area. Rates of denitrification per m2 were estimated
ZnCh solution, and samples for NHt concentrations from the equations derived by Nielsen (1992):
and isotope analysis were immediately frozen in 50 ml
plastic vials. D I5 = p('4N 15 N) + 2p('s N I5 N),
p('4N I5 N)
Analysis of concentrations and isotopic composition
of nitrogen species
DI4 = 2p(ISNI5N) *D I5 , (3)
1993, a high abundance of Vlva sp. was observed ry production. Most of this primary production will
in the water column and the surface sediments were be deposited within the lagoon because only a rela-
well-oxidized and inhabited by dense populations of tively small amount will be lost by tidal exchange.
benthic infauna. In late June-early July 1993, howev- Thus, degradation of this organic material increases
er, the lagoon underwent dystrophic crisis with anoxia the internal nitrogen cycling between water and sedi-
and occurrence of sulfide in the water column (Viaroli ment. The nitrification capacity in the sediment made
et al., 1996), resulting in a substantial mortality of it possible, however, that a minor fraction of nitro-
benthic infauna. In the months following the Sep- gen was lost from the system via coupled nitrification-
tember 1993 campaign, the system stabilized and by denitrification during each mineralization/assimilation
January 1994 the lagoon was colonized by a dense cycle. About 22% of the NHt produced by miner-
community of benthic microphytes. These changes alization (NHt flux + nitrification) was nitrified and
in the overall ecosystem structure were reflected in 27% of the NO.1 formed was subsequently denitri-
both the exchange rates of combined nitrogen across fied. Thus, 8% ~f the NHt liberated by mineraliza-
the sediment-water interface and in the processes of tion underwent coupled nitrification-denitrification. If
nitrification and nitrate reduction within the sediment nitrification-denitrification activity occurs continuous-
(Tables 2 and 3). ly, approximately 50 kg N would be removed from the
The highly abundant infauna present in June 1993 lagoon each day, which represents an internal capacity
ventilated the uppermost few centimetres of the sedi- for recovering eutrophication stress if external loading
ment and may therefore have been responsible for the is reduced.
high sediment 02 uptake rates recorded in the dark The dystrophic crisis in the Etang du Prevost in the
(Table 2). Although low concentrations of NO.1 in the early summer that resulted in sulfide in the water col-
overlying water during the June campaign (Table 1) umn and mass mortality of benthic fauna, complete-
could only support low rates of denitrification (Dw, ly inhibited sediment nitrification and thus the basis
Table 3), high rates of nitrification within the sedi- for coupled nitrification- denitrification. Nitrate from
ment created suitable conditions for significant cou- the overlying water was therefore the only source for
pled nitrification-denitrification activity (Dn' Table 3). denitrification during the September field campaign
A minimum estimate of the nitrification rate can be (Table 3). Since the rate of denitrification based on
calculated as the sum of the rates of NO.1 release and water column NO.1 (Dw) is proportional to the NO.1
coupled nitrification-denitrification (Rysgaard et al., concentration in the water column (Nielsen et al.,
1993). The estimated nitrification rate during darkness 1990), the'" 4 x higher NO.1 concentration in the over-
was 3.3 mmol m- 2 d- 1 , a high rate compared to nitri- lying water in September compared with June caused
fication activities measured at other marine shallow- the rate of Dw to increase also by '" 4 x (Table 3).
water locations (e.g., Henriksen et al., 1981; Nishio However, the total rate of denitrification in September
et al., 1983; Henriksen & Kemp, 1988; Kemp et al., was only about half of that measured in June because
1990; Sloth et al., 1992; Rysgaard et al., 1993). nitrification did not occur.
This high nitrification activity was probably caused Nitrate ammonification was a significant process in
by a combination of the presence of a dense benth- September and accounted for approximately 33% of
ic infaunal population and high rates of NHt regen- the anaerobic nitrate reduction activity in the sediment
eration (Table 2) originating from decomposition of (Table 3). Although there was a net removal of NO.1
organic matter. Stimulation of nitrification and coupled from the water column to the sediment, a substantial
nitrification-denitrification by the presence of ventilat- part of this NO:;- was then returned to the water column
ing animals in the sediment has also been described as NHt through the dissimilatory reduction pathway.
by other investigators (Aller, 1988; Kristensen, 1988; We did not measure nitrate ammonification rates in
Henriksen & Kemp, 1988; Binnerup et al., 1992; Peli- June (i.e. before the dystrophic crisis) and thus it is
gri et al., 1994), and the mechanism responsible for this difficult to evaluate whether the occurrence of nitrate
stimulation has been attributed to an increase in sedi- ammonification was a consequence of the dystrophic
ment surface area caused by animal irrigation (Aller, crisis. However, determinations of the most probable
1988; Kristensen, 1988). numbers of nitrate ammonifiers and denitrifiers (Welsh
Fluxes of DIN (NO.1 + NHt) were directed from & Herbert, pers. comm.) in the sediments showed that
the sediment towards the water column, indicating that the former group was dominant in the uppermost sed-
the sediment was a nitrogen source for pelagic prima- iment layers in both June and September.
137
Table 1. Concentrations of NO; and NHt, and temperature of the water column at the sampling stations.
Table 2. Net sediment uptake and release rates of 02 NO; , and NHt in light and dark incubated cores. Positive and negative values
represent net release and net uptake rates, respectively. Standard errors are shown in parenthesis (n '" 5). Nitrate uptake rates were
not determined at Bassin d' Arcachon in January due to analytical problems.
O2 NO- NHt
3
(mmolm- 2 d- l ) (mmol m- 2 d- I ) (mmol m- 2 d- I )
Locality June September January June September January June September January
Etang du Prevost, -83 (24) -33 (23) 172 (12) 2.3 (0.4) -1.1 (0.2) -6.3 (0.2) 7.6 (1.1) 2.2 (2.5) -8.1 (0.5)
Light
Etang du Prevost, -222 (46) -Ill (5) -32 (0.4) 2.3 (0.3) -1.2 (0.3) -6.5 (0.3) 15.6 (4.2) 10.3 (2.8) -2.3 (0.6)
Dark
Bassin d'Arcachon 52 (9) 99 (25) 28 (4) 0.1 (0.1) -2.0 (0.5) nd -2.6 (1.4) -2.0 (0.5) -1.0 (0.1)
Light
Bassin d' Arcachon -44 (6) -84 (6) -35 (5) 0(0) -1.4 (0.2) nd -1.6 (0.7) -1.4 (0.2) 0.8 (0.1)
Dark
Since the release of NHt from the sediment of water column. Denitrification was based on both NO;-
Etang du Prevost in September 1993 was much high- diffusing from the water column and on NO;- pro-
er than the NO;- uptake from the water column, a duced by nitrification within the sediment (Table 3). If
net efflux of DIN from the sediment to the water no benthic micro algal assimilation of NO;- occurred,
column supplied the pelagic primary production with it would be expected that the rate of denitrification
nitrogen. As in June, a significant internal N-cycling based on water column NO;- (Dw) would be high-
between the sediment and the water column at the er than the rate obtained during the September cam-
Etang du Prevost was possible. The lack of nitrification paign because the water column NO;- concentration
within the sediment, however, disconnected the link was considerably higher during winter (Table 1 and
between N-mineralisation and N-removal via coupled 3). However, the measured rate of Dw in January was
nitrification-denitrification that was present in June. 25% of the rate during September and this account-
Consequently, nitrogen could only be lost from the ed for <2% of the net NO;- uptake by the benthic
lagoon through denitrification based on water column microalgae. The rate of coupled nitrification- denitrifi-
NO;- , through export of organic and inorganic nitrogen cation (Dn) in January was only 5% of the activity mea-
by tidal exchange to the sea, or by permanent burial of sured in June, despite a much lower sediment uptake
refractory N- compounds within the sediment. of 02 and higher NHt concentrations in the water col-
The dominance of benthic microalgae in January umn in January (Table 1). This suggests that benthic
had a strong influence on both 02 exchange rates microalgae suppressed both Dw and Dn activity. Other
across the sediment- water interface and the inorgan- studies have shown that benthic microphytes may sig-
ic N-metabolism within the system, as has also been nificantly reduce both the nitrification activity and the
shown in other shallow marine systems, e.g. Ander- rate of coupled nitrification-denitrification in marine
sen et aI., 1984; Nielsen et aI., 1990; Sundbiick et aI., sediments (Henriksen & Kemp, 1988; N. Risgaard-
1991. In contrast to the situation in June and Sep- Petersen, in prep; Dalsgaard et aI., in prep). The mech-
tember, the sediment was net oxygen producing and anisms responsible for this reduction are not satisfacto-
net NO;- and NHt consuming (Table 2), indicating rily explained, but it has been suggested that the release
that benthic micro algae assimilated nitrogen from the of bacteriostatic compounds by the benthic microalgae,
138
Table 3. Rates of denitrification of NO;- supplied from the water column (D w ), coupled nitrification-denitrification (Dn ), and nitrate ammonification
of NO;- supplied from the water column DNRAw, in light and dark incubated cores. Standard errors are shown in parenthesis (n =5).
Dw Dn DNRA w
(mmol m- 2 d- 1) (mmolm- 2 d- 1 ) (mmol m- 2 d- 1)
Locality June September January June September January September
Etang du Prevost, 0.09 (0.02) 0.44 (0.16) 0.05 (0.03) l.01 (0.18) 0(0) 0.01 (0.01) 0.13
Light
Etang du Prevost, 0.08 (0.01) 0.44 (0.13) 0.17 (0.01) 0.79 (0.2) 0(0) 0.09 (0.06) 0.31
Dark
Bassin d' Arcachon 0.001 (0.001) 0.00 I (0.00 I) 0.082 (0.015) 0.03 (0.02) 0(0) 0.14 (0.03) 0.07
Light
Bassin d' Arcachon 0.00 I (0.00 I) 0.009 (0.002) 0.277 (0.051) 0.02 (0.02) 0.01 (0.01) 0.19 (0.01) 0.13
Dark
fluctuating pH and 02 conditions, and competition for N-cycling in sediments at the Bassin d'Arcachon
NHt may be important factors (Henriksen & Kemp,
1988). The intertidal sediments at the bay of Arcachon were
Nitrate ammonification was not quantified during much more stable than the sediments at the Etang du
January, but an estimate of the maximum rate of anaer- Prevost throughout all three sampling periods. The sed-
obic NO;;- reduction based on the actual NO;;- and 02 iment was inhabited by productive eelgras (Zostera
concentrations in the water column, and the sediment noltii) populations throughout all three field cam-
02 uptake in the dark (Christensen et aI., 1990), sug- paigns, as shown by the net oxygen production during
gested that nitrate ammonification accounted for less light incubations of intact plants (Table 2). Concentra-
than 10% of the net uptake of N01 . tions of NO l and NHt were low in the water column
Fluxes of DIN (NO;;- and NHf) were directed from during June and September (2-5 J.lM), during January
the water column towards the sediment surface in Jan- the NO;;- concentration was elevated (25 J.lM) due to
uary, leading to a net removal of inorganic nitrogen increased riverine input.
from the water above the sediments, and most of this Rates of denitrification of NO;;- supplied from the
DIN uptake was attributed to algal assimilation. In water column (Dw) and rates of coupled nitrification-
contrast to the June and September field campaigns, denitrification (Dn) were highest in January (Table 3).
a close and efficient coupling between NHt regen- The lowest rate of Dw occurred in June, while the low-
eration and algal N-assimilation within the sediment est rate of Dn occurred in September. The observed
was established during the January field campaign. seasonal pattern of Dw may be explained by differ-
The close coupling between NHt regeneration and ences in the NO;;- load to the system, as Dw is pro-
algal N-assimilation in the sediment, that prevented portional to the NO;;- concentration in the overlying
DIN release to the overlying water and suppressed water (Nielsen et aI., 1990). The observed pattern of
the coupled nitrification- denitrification activity may Dn may be linked directly to the seasonal variation in
contribute to a prolonged retention of nitrogen in the the nitrification activity. The highest rate of potential
sediment during the winter months. This may then be nitrification in sediments is usually observed during
liberated to the water column during the spring months winter when oxygen penetrates deep into the sediment
when floating macroalgae recolonize the lagoon and and when competition for NHt from benthic prima-
shade out the benthic microalagae. Consequently, the ry producers is weak (Hansen et aI., 1981). Oxygen
loss of benthic microalgae will increase the efflux of consumption in Arcachon Bay, was lowest in January,
nutrients from the sediment to the water column and indicating a deeper 02 penetration depth and thus better
thus further stimulate the growth of VIva sp .. conditions for nitrifying bacteria (Jensen et aI., 1993).
NHt release from the sediment was only observed
during January suggesting that the NHt availability
for nitrifiers was higher because uptake by seagrass
was lower during the winter period.
139
In the results presented thus far we have focused NO;- assimilation by Z. noltii, and extremely low abun-
solely on nitrification-denitrification processes associ- dance of nitrifiers (Welsh & Herbert, pers. comm.) may
ated with the sediment surface. Rooted benthic macro- contribute to the low denitrification activity relative to
phytes may also influence these processes in deep- nitrate ammonification activity.
er sediment strata surrounding the rhizosphere. The During June and September, there was a net uptake
release of 02 via the roots (Sand-Jensen et aI., 1982) of inorganic nitrogen from the water column during
to the sediment may stimulate nitrification and thus light and dark conditions, although there was a small
provide an additional NO;- source for denitrification net release ofNO;- in June. In January, NHt was only
(Iizumi et aI., 1980; Christensen & S¢rensen, 1986; released to the overlying water in the dark (Table 2).
Reddy et aI., 1989; Caffrey & Kemp, 1990). However, In September, when the net fluxes of NO-j and NHt
our measurements of coupled nitrification- denitrifica- were directed towards the sediment, less than 1% of
tion activity associated with the rhizosphere of Z. noltii the NO-j uptake from the water column was due to
showed no evidence for 15N enrichment of the sediment denitrification (Table 2 and 3). The low denitrifica-
N2 pool even when sediment cores containing intact tion capacity of the bay of Arcachon suggests that
plants were incubated in the light (400 J.lmol photones only a minor fraction of the NO;- input to the basin
m- 2 s-l) for 7-24 h and enriched with 15NHt (final is removed from the system via denitrification as N2
sediment concentration of 15NHt was 1 mM). Cou- gas, and that most NO;- is assimilated by Z. noltii
pled nitrification- denitrification activity was therefore directly either from the water column or from the sedi-
exclusively associated with the sediment surface in the ment via anaerobic nitrate-ammonification. The small
Bassin d' Arcachon. nitrification capacity within the sediment prevented a
Overall, the sediment in the bay of Arcachon had significant N removal from decomposing plant or algal
a low N- removal capacity via denitrification. Rates material via coupled nitrification-denitrification. Since
of both denitrification based on water column NO;- inorganic combined nitrogen was not released from
and coupled nitrification-denitrification were extreme- the sediment during summer, an efficient coupling
ly low during summer and winter (Table 3) compared between sediment NHt regeneration and assimilation
with rates measured at other localities (e.g. Seitzinger, by Z. noltii retained nitrogen within the sediment.
1988; Kemp et aI., 1990; Rysgaard et aI., 1993; Rys-
gaard et aI., 1995; Nielsen et aI., 1995). One fac-
tor contributing to the low denitrification activity was Conclusions
the dominance of bacteria with a nitrate ammonifica-
tion pathway (Welsh & Herbert, pers. comm.) com- In the present study the nitrogen cycle was described
pared to denitrifiers. In September we investigated the for two very different lagoons, the seagrass-dominated
importance of nitrate ammonification, and in accor- Bassin d' Arcachon and the algal-dominated dystrophic
dance with the quantitative dominance of ammonifiers Etang du Prevost. Denitrification rates in both lagoons
over denitrifiers observed by Welsh & Herbert (pers. were low and played a minor role as a nitrogen sink.
comm.), nitrate ammonification accounted for 95% of Throughout the year, the seagrass in Bassin d' Arca-
the total dissimilatory reduction ofNO;- supplied from chon controlled the nitrogen cycle due to its high nitro-
the water column (Table 3). It has been suggested gen uptake rates. Consequently, nitrogen fluxes were
that nitrate ammonification is a secondary metabol- directed into the sediment from the water column, and
ic pathway for various groups of bacteria including the sediment in this lagoon acted as a significant nitro-
Clostridia sp. and sulphate reducing bacteria (Tiedje, gen sink. In contrast, the sediment of the hypereu-
1988). In Bassin d' Arcachon, sulphate reduction rates trophic lagoon Etang du Prevost acted as a significant
were very high in the sediment (Isaksen & Finster, Sub- N source during the summer months, due to extensive
mitted) and sulphate reducing bacteria with a capacity mineralization and lack of benthic primary production.
for nitrate ammonification were highly abundant (Fin- During winter, benthic microalgae colonized the sedi-
ster, pers. comm). Recently, it has been shown that ment surface and changed the sediment in the lagoon
sulphate reducing bacteria may have a higher affinity from being a nitrogen source to the overlying water
for NO;- at lower NO;- concentrations than denitrifiers to being a sink due to their high assimilation rates.
(Dalsgaard & Bak, 1994). Thus, low NO;- input to the It is likely, however, that this assimilated nitrogen is
anoxic sediment strata due to relatively low NO;- con- liberated to the water column at the onset of summer
centrations in the water column (Table 1), competing thereby fueling the extensive growth of the floating
140
macroalgae, Ulva sp .. Thus, nitrogen cycling in Etang Henriksen, K., J. I. Hansen & T H. Blackburn, 1981. Rates of
du Prevost is much more unstable and unpredictable as nitrification, distribution of nitrifying bacteria and nitrate fluxes
in different types of sediment from different types of sediment
compared with Bassin d' Arcachon. from danish waters. Mar. BioI. 61: 299-304.
Iizumi, H., A. Hattori & C. P. McRoy, 1980. Nitrate and nitrite in
interstitial waters of eelgrass beds in relation to the rhizosphere.
Acknowledgments J. Exp. Mar. BioI. Ecol.: 191-201.
Isaksen, M. E & K. Finster, 1995. Sulphate Reduction in the Root
Zone of a Seagrass Bed (Zostera no/tii) in a Tidal Area, The Basin
We thank N. P. Revsbeck and Silvia Pelegri for helpful of Arcachon, France. Mar. Ecol. Prog. Ser. (in press).
discussions concerning this manuscript. This work was Jensen, K., N. P. Revsbech & L. P. Nielsen, 1993. Microscale dis-
tribution of nitrification activity in sediment determined with
financially supported by the ED Programme EV5V-
a shielded microsensor for nitrate. Appl. Envir. Microbiol. 59:
CT93-0245, the Danish Research Academy and the 3287-3296.
Danish National Science Research Council. Kemp, W. M., P. Sampou, J. Caffrey &M. Mayer, 1990. Ammonium
recycling versus denitrification in Chesapeake Bay sediments.
Limnol. Oceanogr. 35: 1545-1563.
Kristensen, E., 1988. Benthic fauna and biogeochemical processes in
References marine sediments: microbial activities and fluxes. In T. H. Black-
burn & J. S!ilrensen reds], Nitrogen cycling in coastal marine
Aller, R. C., 1988. Benthic fauna and biogeochemical processes in environments. Wiley: 275-299.
marine sediments: The role of burrow structures. In T H. Black- Nielsen, K., Nielsen, L. P. & P. Rasmussen, 1995. Estuarine nitrogen
burn & 1. S!ilrensen [eds], Nitrogen cycling in coastal marine retention independently estimated by the denitrification rate and
environments. Wiley: 301-338. mass balance methods: a study of Norsminde Fjord, Denmark.
Andersen, T K., M. H. Jensen &J. S!ilrensen, 1984. Diurnal variation Mar. Ecol. Prog. Ser. 119: 275-283.
of nitrogen cycling in coastal, marine sediments. Mar. BioI. 83: Nielsen, L. P., 1992. Denitrification in sediment determined from
171-176. isotope pairing. FEMS Microb. Ecol. 86: 357- 362.
Binnerup, S. J., K. Jensen, N. P. Revsbech, M. H. Jensen & Nielsen, L. P., P. B. Christensen, N. P. Revsbech & J. S!ilrensen, 1990.
J. S!ilrensen, 1992. Denitrification, dissimilative reduction of Denitrification and photosynthesis in stream sediments studied
nitrate to ammonium, and nitrification in a bioturbated estuar- with microsensor and whole core techniques. Limnol. Oceanogr.
ine sediment as measured with 15N and microsensor techniques. 35: 1135-1140.
Appl. Envir. Microbiol. 58: 303-313. Nishio, T, I. Koike & A. Hattori, 1983. Estimates of denitrification
Blackburn, T H., 1993. Turnover of 15NHt tracer in sediments. In and nitrification in coastal and estuarine sediments. Appl. Envir.
Kemp, P. E, Sherr, B. E, Sherr, E. B. & Cole, J. J. (eds) Handbook Microbiol. 45: 444--450.
of methods in aquatic microbial ecology. Lewies Publishers, Boca Peligri, S. P., L. P. Nielsen & T H. Blacknurn, 1994. Effect of
Raton, USA: 643- 648. Corophium vo/utator (Pallas) on denitrification in artificial micro-
Bower, C. E. & T Holm-Hansen, 1980. A salicylate- hypochlorite cosms, measured by 15N isotope pairing technique. Mar. Ecol.
method for determining ammonia in seawater. Can. J. Fish. Sci. Prog. Ser.: 285-290.
37: 794--798. Reddy, K. R., W. H. Patrick Jr. & C. W. Lindau, 1989. Nitrification-
Caffrey, J. & W. M. Kemp, 1990. Nitrogen cycling in sediments with denitrification at the plant root-sediment interface in wetlands.
estuarine populations of Potamogeton peljilliatus and Zostera Limnol. Oceanogr. 34: 1004--1013.
marina. Mar. Ecol. Prog. Ser. 66: 147-160. Risgaard-Petersen, N., S. Rysgaard & N. P. Revsbech, 1993. A
Christensen, P. B., L. P. Nielsen, J. S!ilrensen & N. P. Revsbech, 1990. sensitive assay for determination of 14 N/ 15 N isotope distribution
Denitrification in nitrate-rich streams: Diurnal and seasonal vari- in N03'. J. Microbiol. Meth. 17: 155-164.
ation related to benthic ocygen metabolism. Limnol. Oceano gr. Risgaard-Petersen, N. & S. Rysgaard, 1995. Nitrate reduction in
35: 640--651. sediments and waterlogged soil measured by 15N techniques. In
Christensen, P. B. & J. S!ilrensen, 1986. Temporal variation of deni- Alef & Nannipieri (eds) Methods in Applied Soil Microbiology
trification in plant-covered, littoral sediment from Lake Hampen, and Biochemistry Academic Press. : 277-288.
Denmark. Appl. Environ. Microbiol. 51: 1174--1179. Robert, R., N. Guillocheau & Y. Collos, 1987. Hydrobiological
Dalsgaard, T & E Bak, 1994. Nitrate reduction in a sulfate-reducing parameters during an annual cycle in the Arcachon Basin. Mar.
bacterium, Desu(fovibrio desu(turicans, isolated from rice paddy BioI. 95: 631-640.
soil: Sulfide inhibition, kinetics, and regulation. Appl. Envir. Rysgaard, S., P. B. Christensen & L. P. Nielsen, 1995. Seasonal
Microbiol. 60: 291-297. variation in nitrification and denitrification in estuarine sediment
Duarte, C. M., 1995. Submerged aquatic vegetation in relation to colonized by benthic microalgae and bioturbating infauna. Mar.
different nutrient regimes. Ophelia 41: 87-112. Ecol. Prog. Ser. 126: 111-121.
Grashoff, K., M. Erhardt & K. Kremling, 1983. Methods of seawater Rysgaard, S., N. Risgaard-Petersen, L. P. Nielsen & N. P. Revsbech,
analysis, 2nd edn. Verlag Chemie, Weinheim, Germany. 1993. Nitrification and denitrification in lake and estuarine sed-
Hansen, J. I., K. Henriksen & T H. Blackburn, 1981. Seasonal dis- iment measured by 15N dilution technique and isotope pairing.
tribution of nitrifying bacteria and rates of nitrification in coastal Appl. Environ. Microbiol. 59: 2093-2098.
marine sediments. Microb. Ecol. 7: 297-304. Sand-Jensen, K., C. Prahl & H. Stockholm, 1982. Oxygen release
Henriksen, K. & W. M. Kemp, 1988. Nitrification in estuarine and from roots of submerged aquatic macrophytes. Oikos 38: 349-
coastal marine sediments. In T H. Blackburn, and J. S!ilrensen 354.
reds], Nitrogen cycling in coastal marine environments. Wiley:
201-249.
141
Sloth, N. P., L. P. Nielsen & T. H. Blackburn, 1992. Nitrification nitrate reduction to ammonium. In A. 1. B. Zender (ed.), Biology
in sediment cores measured with acetylene inhibition. Limno!. of anaerobic microorganisms Wiley and Sons: 174-244.
Oceanogr. 37: 1108-1112. Viaroli, P, M. Bartoli, C. Bondavalli, R. R. Christian, G. Giordani
Seitzinger, S. P., 1988. Denitrification in freshwater and coastal & M. Naldi, 1996. Macrophyte communities and their impact
marine ecosystems: Ecological and geochemical significance. on benthic fluxes of oxygen, sulphide and nutrients in shal-
Limno!. Oceanogr. 33: 702-724. low eutrophic environments. Hydrobiologia 329 (Dev. Hydrobio!.
Strickland, 1. D. H. & T. R. Parsons, 1972. A practical handbook of 117): 105-ll9.
seawater analysis. Bull. Fish. Res. Bd Can. 167: 311 pp. Welsh, D. T., S. Bourgues, R. de Wit & R. A. Herbert, 1996. Seasonal
Sundback, K., V. Enoksen, W. Graneli & K. Petterson, 1991. Influ- variation in rates of heterotrophic nitrogen fixation (acetylene
ence of sublittoral microphytobenthos on the oxygen and nutrient reduction) in Zostera no/tii meadows and uncolonized sediments
flux between sediment and water: A laboratory continuous-flow of the Bassin d' Arcachon, South-West France. Hydrobiologia
study. Mar. Eco!. Prog. Ser. 74: 263-279. 329 (Dev. Hydrobio!. ll7): 161-174.
Tiedje, 1. M., 1988. Ecology of denitrification and dissimilatory
Hydrobio[ogia 329: 143-159,1996. 143
p. Caumette, 1. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Abstract
During 1994 net sediment-water fluxes of oxygen, ammonium and inorganic phosphorus as well as sediment
profiles of organic matter, nitrogen, phosphorus and iron were determined in three shallow eutrophic environments.
Investigations were conducted monthly from March to December at five stations in the Saccadi Goro (Po River Delta,
Italy). In the late summer, samples were collected from a single site in the Prevost lagoon (French Mediterranean
coast) and three stations in the Bassin d' Arcachon (French Atlantic coast). In the Sacca di Goro, water-sediment
exchanges of O 2 , NHt and PO!- were estimated by means of core incubation in the dark. Benthic fluxes for the
French lagoons were in part determined experimentally using benthic chambers and in part from the literature.
In general in the Sacca di Goro the highest oxygen uptake and nutrient release rates were found at the central sites,
affected by macroalgal growth. At the sampling site adjacent to the freshwater inlet, sediment-water exchanges
were principally influenced by tidal activity. In terms of organic matter and nutrient levels, sediments from the
Sacca di Goro and from the Prevost lagoon, both colonised by the floating macroalga Ulva rigida C. Agardh, were
similar. Sediments from the inner sheltered site in the Bassin d' Arcachon, invaded by the rooted macrophyte Ruppia
cirrhosa (Pet.) Grande, showed the highest total Nand P content (363 ± 157 j.lmol N cm- 3 and 15 ±2 j.lmol P cm- 3
as average values in the top 10 cm of sediment), but were low in pore water ammonium and orthophosphate probably
due to the high sequestering capacity of the system and/or efficient coupling between bacterial nutrient regeneration
and assimilation by the plant roots. In addition the outer tidal stations in the Bassin d' Arcachon, invaded by rooted
macrophytes, were low in pore water nutrients. A different trend was evident in the Prevost lagoon where the
concentrations of exchangeable inorganic phosphorus and ammonium were appreciable (0.28 ± 0.07 j.lmol P cm- 3
and 2.4 ± 1.4 j.lmol N cm- 3 as average values in the top 10 cm of sediment). High amounts of dissolved organic
nitrogen were found in the pore water at all the sites investigated showing the key role of the organic nitrogen in
the recycling of nitrogen in these systems.
The hypothesis that iron is a key factor in controlling phosphorus release is discussed since the Sacca di Goro,
which is subject to dystrophic crises, is richer in iron than the Bassin d' Arcachon, which is a more buffered system.
Bassin d'Arcachon
+ Sacca di Gam
N
+
+
N
1 km
Etang du Prevost
Figure 1. Map of the considered lagoons.
Material and methods porosity, organic matter content, total Kjeldhal nitro-
gen, inorganic and organic P, exchangeable NHt and
Samples were collected monthly from March to PO!-, dissolved organic N, total ferrous and ferric iron
December 1994 at five stations in the Saccadi Goro (Po were determined in sediment cores sliced into 2 cm
River Delta, Italy). All the investigated sites (stations sections from the surface to 10 cm depth.
G, 1,4,5/8,8) are mesotidal and the water depth ranges In the Sacca di Goro sediments were collected by
between 0.7 and 1.5 m. In August/September 1994, means of Plexiglass cores (height 30 cm, i.d. 5 cm);
samples were also collected at station 11 in the Prevost bottom water samples were collected in a 11 Ruttner
lagoon (French Mediterranean coast) and at three sta- bottle. Samples were collected on 16 March, 12 April,
tions in the Bassin d' Arcachon (French Atlantic coast). 5 May, 7 June, IS July, 8 August, 9 September, 13
In the Sacca di Goro, exchanges of oxygen, ammo- October and 14 December at the five sites shown in
nium and soluble reactive phosphorus (SRP) across Figure 1. The sampling dates were chosen so that
the water-sediment interface were measured by means they would express a seasonal representativeness. Sed-
of core incubations in the dark. Benthic fluxes for the iments were not collected at station 5/8 in July, at
French lagoons were in part determined experimen- station 4 in August, at station 1 and at station 8 in
tally in benthic chambers and in part from the litera- December. In the French lagoons sediment cores were
ture. Along with sediment-water exchanges, density, collected by hand at low tide: on 19 August at sta-
146
tion 11 (Prevost lagoon), on 23 August at station Cl dissolved organic nitrogen and dissolved organic phos-
(Certes fishponds), on 2 September at station A and on phorus. Extractable ammonium and extractable reac-
4 September at station B (Bassin d' Arcachon). tive phosphorus were kept in solution washing 2 or 3
times 1 g about of wet sediment respectively with 10 ml
Sacca di Gora: core incubation of 1 M KCI and with 10 ml of 1 M MgCI2 (pH 7.6) for
Three cores collected from each station were transport- I hour on a multiple magnetic stirring platform. The
ed to the laboratory and after removal of the top bung extracted amounts were corrected for the contributions
placed in a tank filled with aerated water from the same from pore water. Sediment samples of about 1 ml were
site and equilibrated at the in situ temperature for two immediately fixed for iron determination in 5 ml of
hours prior to starting the incubation. The core tubes 0.5 M HCI when slicing the cores.
were then sealed with a rubber bung and left in the
dark for a period ranging from two to four hours with- Methods adopted for water analyses
out stirring the overlying water. This because of the Dissolved oxygen was determined following the Win-
occurrence of an easily resuspendable thick nepheloyd kler method (APHA, 1975). Ammonium was deter-
layer in the surficial sediment (Miller-Way et aI., 1994; mined using the indophenol-blue method (Koroleff,
Rizzo et aI., 1992). Oxygen, ammonium and reactive 1970) and reactive phosphorus was determined by the
dissolved phosphorus were determined at zero time ascorbic acid method (Valderrama, 1981). Total dis-
and at the end of the incubation period. solved nitrogen was determined reacting 10 ml of the
sample with 1.4 ml of a potassium persulfate solu-
Methods adopted for sediment analyses tion in an autoclave for I hour (Valderrama, 1981).
The organic matter content of the sediment was deter- The nitrate liberated was then reduced with a cadmium
mined on 2-3 g of finely powdered sediment, previ- column (APHA, 1975) and determined as nitrites (dia-
ously dried at 70°C followed by ignition in a muffle zotation; APHA, 1975). Total dissolved phosphorus
furnace at 550°C until constant weight was reached. was determined as reactive phosphorus after autoclave
Total nitrogen was estimated on fresh and dried oxidation (Valderrama, 1981). Ferric iron was reduced
sediment using the Kjeldhal method which only mea- to ferrous iron with hydroxylamine and then reacted
sures reduced forms of nitrogen. Approximately 15 ml with o-phenanthroline (APHA, 1975).
of 96% sulphuric acid and 10 ml of 40% hydrogen
peroxide were added to 1 g of sediment; the mixture Intersystem comparison
was brought to ebullition until the complete mineral- Intersystem comparisons were performed by Factor
ization of the sample. The liberated ammonium was Analysis using the Harris image procedure (Statview,
collected by distillation, fixed as ammonium sulphate Macintosh). Factor analysis was carried out on loga-
and determined by titration (APHA, 1975). In this type rithm [In( x + 1)1transformed values of seven represen-
of sediment NO;- and NO;- are usually below the level tative variables relating to the August-early September
of detection since oxygen only diffuses in the first few 1994 sampling period: SOD, SRP, NHt fluxes and the
millimeters (Sloth et aI., 1993). average content of organic matter, total nitrogen, total
Total phosphorus was extracted with 20 ml of 1 M phosphorus and total iron in the top 10 cm of sediment.
HCI from 1 g of dried sediment after ignition in a
muffle furnace at 550°C; total inorganic phosphorus
was extracted with 20 ml of 1 M HCI from 1 g of Results and discussion
dried sediment; total organic phosphorus was finally
calculated by difference (Aspila et aI., 1976). This The relationship between benthic fluxes of oxygen,
method underestimates the organic phosphorus pool dissolved ammonium, soluble reactive phosphorus and
mainly due to the quality of the organic matter and its physico-chemical characteristics of the surficial sedi-
capability of being hydrolysed by HCI (Barbanti et aI., ment was determined at five sites in the Sacca di Goro
1993). in 1994. Fluxes measured in late summer were then
Pore water was separated from the sediment by cen- compared with those determined at three sites located
trifugation of about 30 cm3 of wet sediment at 1800 x g in the Bassin d' Arcachon and at one site in the Etang
for 15 minutes; the supernatant was then filtered and du Prevost. The key differences at the sampling sta-
analysed for ammonium, soluble reactive phosphorus,
147
0.5
tions in these different lagoon systems were salinity, i
0.4
dominant macrophytes and trophic condition. - !
6' 0.3 ! ~
Sacca di Gora C
E 0.2 -"-
I fl
In 1994 seasonal trends in the hydrology of this lagoon E
0.1 I
were different from those observed in previous years. I
At the designated sampling sites, dissolved oxygen 0.0
concentrations were relatively constant and ranged
from 0.13 to 0.42 mM, reaching a minimum at station 4
(Figure 2). However, continuous monitoring conduct-
ed in the southernmost part of the lagoon from late June
onwards showed that there was almost constant hypox-
ia with several episodes of night anoxia (Viaroli et al.,
1996). This pattern was consistent with the expansion
of the Uiva rigida beds, which reached their maximal
development in the south-central part of the lagoon dur-
ing this period. Nevertheless, this area did not under-
go summer dystrophy, although oxygen concentrations
continued to be low until early August. This situation
was quite different from that observed from 1989 to
1992, where the sheltered part of the lagoon, east of
station 4, experienced severe summer anoxia accom-
80 T
60 -"- 0
panied by substantial recycling of phosphorus (Viaroli -
et al., 1992, 1993). However, since 1993 the summer 0
Z 40
I
-L x
dystrophic episodes have been reduced in intensity and 0
'0
duration, due to the increased seawater inflow through
a large channel cut through the southern sand bank.
E
:l.. 201 /j,
x
~
0
As a consequence, growth of Ulva was restricted to
the south-central area of the lagoon leading to several
oI
changes in benthic processes at the other sites. I
-<>-G -iJ-
Dissolved inorganic nitrogen was dominated by
nitrate mainly due to the discharge of the Po di Volano, T2.0 ~4 -x-SI8
the latter two sites nitrate was depleted in the warmer [O'S-I'~
0 4 ' "0' . "0.4
recent years of the P discharged into the lagoon from for station 8 (Figure 3). At the latter site, the amounts of
the Po di Volano, which has decreased from about 200 NHt and SRP released from sediment under dark con-
tons y-I in 1989 (Viaroli et aI., 1992) to 45 tons y-I ditions were almost undetectable. The highest values
in 1992 (Alvisi et aI., 1993). of SRP (4.3 mmol P m- 2 d- l ) and NHt (16.7 mmol
Variations in both dissolved inorganic P and N did N m- 2 d- l ) were determined in July at station 1 and
not follow a well defined seasonal trend, although oxy- at station G respectively. At the latter site, fluxes of
gen minima at stations 1, 4 and G, as well as nitrate ammonium and SRP did not follow a well defined
depletion at sites 5/8 and 8 occurred in the warmer peri- trend, being negative in April and June (i.e. from water
od. In the western area of the lagoon seasonal variations to sediment) or highly positive (i.e. from sediment to
might also have overlapped with short-term fluctua- overlying water). These results can be explained by
tions (e.g. daily changes) depending on increased salt considering tidal fluxes as well as mixing of fresh and
water intrusions which entered the lagoon when the marine water. At low tide ammonium and SRP con-
freshwater discharge was low. centrations were probably higher in the water overlying
Sediment oxygen demand (SOD) followed a clear the sediment than in the porewater. Thus, there would
seasonal trend with pronounced peaks in the warmer be a top down flux with NHt and SRP entering the
months (Figure 3). At station G, close to the Po di sediment, whereas at high tide, the opposite would
Volano discharge, SOD reached 80 mmol O 2 m- 2d- 1 occur since the overlying sea water is poorer in nutri-
in March and 365 mmol 02 m- 2 d- 1 at the beginning ents. However the peak observed in July can also be
of August. SOD followed a bi-modal trend consis- explained by enhanced microbial activities due to high
tent with that of phytoplankton chlorophyll-a (data not temperature as well as organic matter inputs.
shown). For example, at the beginning of March in Organic matter, nitrogen and phosphorus pools pro-
the surface waters of the Po di Volano chlorophyll-a files were determined in the surficial sediment (10 cm).
concentrations reached 109.2 J.Lg 1- I and particulate The values obtained were then averaged and the stan-
organic carbon 7.4 mg 1-1 . It is probable that organic dard deviations calculated. On average, organic matter
matter sustained microbial respiration in the sediment concentration did not show significant changes with
(Hansen & Blackburn, 1991; Enoksson, 1993). How- depth. At station G, the annual series revealed two
ever, the highest value found in August at station G peaks, in April and in August, which appear to be
can be explained by the richness of benthic infauna related to phytoplankton seasonality (Figure 4). More-
rather than by microbial activity. It was noted that the over, this trend was consistent with SOD fluctuations.
selected site was colonised by several thousands of The highest values were recorded at station 1, where in
benthic Harpacticoids per square meter (Ceccherelli, the late summer the organic matter content was 12.1 %
pers. comm.). In August, SOD values ranging from as dry weight, while the lowest average values were
200 to 350 mmol O 2 m- 2 d- 1 were measured at sta- found at station 8, a site dominated by sand (Dal Cin
tions 1, 4 and 5/8, while in the more sheltered site 8, & Pambianchi, 1991).
SOD never exceeded 150 mmol O 2 m- 2 d- l . Maxi- Total phosphorus content in the surficial sediment
mum values observed in the central part of the lagoon was clearly related to riverine discharge and, to a less-
(station 4) were significantly higher than those mea- er degree, to macroalgal growth. Maximum concentra-
sured both in 1991 (Viaroli & Naldi, 1992) and 1992 tions were measured at station G, while low concentra-
(Viaroli et aI., 1993). At station 5/8, which in 1994 was tions were recorded at station I (Figure 5). In May, at
affected to a slight degree by Viva growth, SOD val- all the sites sampled significant peaks occurred, with
ues were similar to those recorded in previous years. organic P accounting for approximately 50% of the
In contrast at station 8, the most sheltered which in the total. In the eastern part of the lagoon (stations 5/8 and
past was affected by severe summer anoxia and dystro- 8) organic P was quite important also in August, when
phy, oxygen demand was relatively low. It is evident it became the dominant form at station 8. Similarly, at
that the cutting of a new channel to the sea in 1993 station 1, organic P content continued to be high from
has led to a major change in the water quality of this May to September. The rapid increase in phosphorus
isolated area of the lagoon. concentrations at the beginning of May can be possibly
Ammonium and soluble reactive phosphorus (SRP) accounted for the increased riverine discharge follow-
regeneration was clearly affected by benthic oxygen ing heavy rainfall that occurred in mid April (see also
uptake and followed the same seasonal pattern as in Viaroli et aI., 1996) as well as by the deposition of
previous years, with pronounced summer peaks except phytoplankton blooms. However, the latter hypothesis
149
500 Station G 20 4400 T
r
400 15 I
300 2900
10
200
1400 +
100 I
o -100
i
500 T Station 1 20
t
t
400 15
300
lOOt
100
o
-5
10
2900 t
;~ ~F"~.L~~, ~
+-+f ,=--..,
5 1400
500
400
300
200
Station 8
20
15
10
:1
1400
II
+
5
FiKure 3. Benthic fluxes of dissolved oxygen (left), ammonium (central) and soluble reactive phosphorus at the five sites in the Sacca di Goro
(March-December 1994). Values are the average of three cores; bars indicate one standard deviation.
was not confirmed by a parallel increase in nitrogen might be explained by the burial of macroalgal detri-
concentrations (Figure 6). Increases in organic-P con- tus. It was observed that stations 1 and 5/8 were located
tent of the surficial sediment in the summer period at the edge of the zone where Viva had the maximum
150
IJlj tl
I
12
8
8
4
4
0
0
12 -
Stat ion 16
Station I
8 12
o -, 1 ,
12
Station 4 16 ~
Station 4
8
III
12
Station 8 16 T
o
-I ,
M A M J J A SON D M A M J J A s o
Figure 4. Organic matter content (% dry weight) in the top 10 cm Figure 5. Total inorganic and organic phosphorus in the top 10 cm
of sediment at the five sites in the Sacca di Goro (March-December of sediment at the five sites in the Sacca di Goro (March-October
1994). Values are the average of the first 10 cm of sediment; bars 1994). Values are the average of the first 10 cm of sediment; bars
indicate one standard deviation. indicate one standard deviation.
151
,~
On average, total nitrogen content in the top sed- Xl+
3
iment was higher in the sites influenced by riverine
0' 0
discharge, even if this gradient was not so clear as
it was for phosphorus. In a similar manner, seasonal
patterns were not well defined, even though the spring
maxima were followed by a sharp decrease in May and
~.~~
(Lapointe et a!., 1992) as well as in iron rich sediments !:O+
16
(Golterman, 1995). ro+
-3
In 1994, in the Sacca di Goro, concentrations of Xl+
iron in the sediment were quite high and reached val- 0', 0
ues greater than 200 Mmol cm- 3 (Figure 8). Seasonal J FMAMJ JASONO
L
o +--+--' 0+--+-
::~ II Station I
2
o .-B!JITIilIIIITialliiTiIIIlI1~. T T 10 +--+--
Station 4 2
o +--+--
450 Station 8 2
300
150
o +--+-- 0+--+-
1 FMAMll ASOND 1 F M A M 1 1 A SON D
Figure 7. Exchangeable reactive phosphorus and ammonium and ammonium at the five sites in the Sacca di Goro (March-December 1994).
Values are the average of the first 10 cm of sediment; bars indicate one standard deviation.
sediment as a possible mechanism for the observed even if they were affected by great variability. In the
P-remobilization. latter sites, slight seasonal trends were also observed,
The behaviour of nitrogen and phosphorus in the with peaks in spring and late summer. Considering
sediments has been tentatively explained by consider- these patterns, we can point out that stations G and 8
ing NIP ratios at different levels (Figure 9). At station were accumulating phosphorus, while stations 1,4 and
G, total N to total P ratio (TNffP) was always less 5/8 seemed to accumulate nitrogen mainly in spring
than 10, being usually around 4. Values in the same and autumn. Moreover the slight decrease of TNffP
range, although somewhat higher, were observed also ratio in the warmer months was consistent with nitro-
at station 8. At stations 1,4 and 5/8, TNffP ratios were gen depletion in the water column.
approaching the conventional Redfield ratio (NIP= 16),
153
.~
for TNrrp (Figure 10). At station G, N ExchlP Exch
~
never exceeded the balanced Redfield ratio, with most
~ 100
"- values less than 10. Ratios somewhat higher, but on
average less than 16, were also found at stations 8 and
0
5/8. At station 1 and 4 N Exch/P Exch followed clear
seasonal trends with pronounced peaks in spring and
late autumn, and very low values in July. These pat-
300 terns, show that the mobility of inorganic P relative to
Station I
ammonium was at its maximum in July, while the con-
.. -
:
200 verse occurred in spring as well as in late autumn. In
-
~~
-
B 'aE i:
tJ: FI the latter period the increased availability of exchange-
i
~ 100 able ammonium coincided with high concentration of
I
,
60
Stallon G
B
"-
- Wi
30
M A M A s o D M A M A s o D
60
Station I
B
""- 30
B
z
M A M A s o D
M A M A s o D
• W r n " '';0" 4
~ ] ;;DUn-, M A M
0 - ,-t-
n-h-,,---,..II.D
A s
f----l
o D
60
60
5 tat i on 5/8· .c
3 Stati on 5/8
.:l"
:: 30
§ ""
- 30
.c
z
.:l"
o z
0 ~--~==Y_--Y_--~--~--4_--4_--4_~
M A M A s o D
60
Station 8
E Station 8
""
- Wi
30
~ 30
M A M A s o D M A M A s o D
Figure 9. Total nitrogen (Ntot) to total phosphorus (Ptot) molar ratio Figure 10. Molar ratio of exchangeable ammonium (N Exch) to
in the top 10 em of sediment in the five sites of the Sacca di Goro exchangeable inorganic phosphorus (P Exch) at the five sites of the
(March-December 1994). Sacca di Goro (March-December 1994).
to be more healthy due to the large seagrass mead- tling of large amounts of epiphytes growing on Ruppia
ows (Bachelet et aI., 1994) while the impoundment stems. Station 11, free of rooted plants but general-
at Certes shows an intermediate behaviour (Labourg, ly invaded by floating beds of Viva, did not show a
1975; Escaravage, 1990; Castel et aI., 1996). significant OM profile probably because of sediment
The high content and the shape of the organic mat- disturbance from both infaunal and human activity.
ter profiles in the sediment at the sampling stations Total Kjeldhal nitrogen (TN) profiles were to a
in the Bassin d' Arcachon are probably determined by some extent overlapping those obtained for OM. Thus,
Zostera and Ruppia (Table 1). This was particularly it can be argued that most of the nitrogen was stored in
evident at station CI, where in the first 6 cm OM was the sediment throughout OM accumulation. As a mat-
over 20%; such high values can be explained by the set- ter of fact, in the surficial sediment at station CI, TN
155
Table 1. Organic matter (O.M.), total Kjeldhal nitrogen (TN), exchangeable ammonium (N Exch), total inorganic (TIP) and organic
phosphorus (TOP) and total ferrous and fenic iron content in the top 10 cm of sediment at the station II in the Etang du Prevost and
stations A, B and CI in the Bassin d' Arcachon. Average values and standard deviations are reported. (I, 4, 5/8 and G), Etang du
Prevost (II) and Bassin d' Arcachon (A, B, Cl).
11 10.1 ± 1.4 73.6± 9.5 2.4 ± 1.4 6.7 ± 1.4 0.9 ± 0.3 25.4 ± 11.0 16.6± 7.3
A 9.9 ± 5.0 124.2 ± 28.6 2.4 ± 0.4 6.2 ± 1.5 4.8 ± 1.3 27.0± 4.6 37.2 ± 4.9
B 16.2 ± 3.3 112.7 ± 14.7 2.1 ± 0.5 4.0 ± 0.5 3.2 ± 0.5 44.6 ± 22.4 46.8 ± 10.4
CI 19.9 ± 6.8 363.1 ± 157.0 3.0 ± 0.2 9.1±1.9 5.6 ± 1.9 99.3 ± 23.6 34.6 ± 15.3
:L__,
the organic P pool was generally low compared to the
inorganic one, while at stations A, Band Cl it account-
"=> Station 4
E ed for almost the 50% of the total. This may be due to
!;: different pathways in the degradation of organic P, in
~
"
turn following different quality of the organic matter:
z
M A M
,~ I'iii5'f'jJ
. easier degradable Ulva at station 11, more refractory
Zostera and Ruppia at stations A, Band Cl (Viaroli et
A S 0 0
aI., 1996).
60 In the sediment of the selected sites, inorganic
"=> Station 5/8 nitrogen was always represented by ammonium, since
E
Co
:; 30 nitrite and nitrate did not occur significantly due to
E
=>
both slow diffusion and high oxygen consumption. On
z
0
average, exchangeable NHt was higher at station Cl
M A M A S 0 D than at the other sites (Table 1). At station 11, pro-
files of both exchangeable and pore water ammoni-
60 urn may indicate a remarkable bottom regeneration
Station 8 of inorganic nitrogen, even if the values were much
d:"
"-
:; 30
lower than expected from preceding analyses. A pos-
E sible explanation is that the conditions at station 11 in
z August-September '94 were quite different from those
0
M A
of the previous field campaigns. As a matter of fact
M A S 0 0
in August 1994 the sediment surface, free of Ulva,
Figure 11. Molar ratio of ammonium (N Flux) to soluble reactive was light brown and well oxidized while in May and
phosphorus (P Flux) estimated from benthic fluxes at the five sites in August '93 and in May '94 the sediment surface,
ofthe Sacca di Goro (March-December 1994).
156
5/8 • A.•
tive (SRP) and dissolved organic (DOP) phosphorus
were almost undetectable in the pore water at stations oro
A and B, while at station Cl appreciable amounts of u. • 8•
DOP were detected in the top sediment layer. At sta- 11
tion 11, SRP was largely exceeding DOP and reached
concentrations up to 400 JLmoll- 1 in the deepest layer G •B
••
(8-10 cm) (Bartoli et a!., 1994). At the latter site pore
water profiles of SRP and NHt clearly overlapped
and are indicative of rapid mineralisation followed by 1 • 4
accumulation of the released inorganic Nand P. At the
sites in the Bassin d' Arcachon the mineralised P can Factor 1
Figure 12. August-early September 1994: Factor analysis of sedi-
be immediately immobilised by complexation with Ca ment variables and benthic fluxes at the different sampling stations
and Fe oxides, or be assimilated by plant roots. This in the Sacca di Goro (I, 4, ~ and G) Etang du Prevost (II) and
hypothesis dealing with the role played by iron as a Bassin d' Azcachon (A, B, Cl). First factor (FACTOR l) is inversely
sink for phosphorus is suggested by differences of Fe related to SRP fluxes, second factor (FACTOR 2) is inversely related
with total iron content of sediment and third factor (FACTOR 3) is
concentrations at the different sites. In particular, the directly related to total nitrogen and phosphorus content of sediment.
three stations of the Bassin d' Arcachon are, on aver-
age, richer in iron than station 11 in the Prevost lagoon
(Table 1). to SRP fluxes (Factor 1) and total iron content in the
top 10 cm of sediment (Factor 2). A weaker group-
Intersystem comparison ing results from the third factor which is related to the
In the Sacca di Goro as well as the French lagoons, sea- total nitrogen and phosphorus content of the surficial
grasses and macroalgae have a recognised and impor- sediment. The high SRP and NHt fluxes recorded at
tant role in driving fluxes on a vast scale (see for exam- station 11 in the Prevost lagoon can be explained in
ple Viaroli et a!., 1996). However, it may be expected terms of water renewal and macroalgal impact. In this
that the magnitude of variations (seasonal and inter- lagoon the water mass usually undergoes slow renewal
annual) of macrophyte-mediated processes are related and Viva biomass can exceed 5 kg m- 2 wet weight
to the size and availability of sedimentary nitrogen, (Bachelet et a!., 1994; Viaroli et a!., 1996). Thus the
phosphorus and iron pools. nutrient loads reaching this lagoon due to runoff or
This hypothesis has been tested by means of factor direct discharge of wastewater are slowly exported to
analysis (Figure 12). The first three factors explain the sea and cannot be totally removed either in the
84.4% of the total variance. The ordination of the algal biomass or in the sediment, especially in sum-
sites from the first two factors appears to be linked mer, when Nand P regeneration is rapid and the bottom
waters become anoxic. For example in 1993 the onset
157
Bachelet, G., X. de Montaudouin, I. Auby & P. J. Labourg, 1994. A Koroleff, F., 1970. Direct determination of ammonia in natural
comparative study of the seasonal changes in macrophytes and waters as indophenol blue. Information on techniques and meth-
macrozoobenthos assemblages in three coastal lagoons under ods for seawater analysis. I.C.E.S. Interlaboratory. Rep No.3:
varying degrees of eutrophication. In P. Caumette (Coord.) 19-32.
C.L.E.A.N. Progress Report 1994. EU Environment Programme Labourg, P. 1., 1975. Contribution it l'hydrologie des etangs
DG XII, Brussels: 353-367. saumatres de la region d' Arcachon: description des phenomenes
Barbanti, A, M. C. Bergamini, F. Frascari, S. Miserocchi & G. d'eaux blanches. Bull. soc. linn. Bordeaux 5: 3-8.
Rosso, 1993. Investigations on some critical aspects of sedi- Lapointe, B., M. M. Littler & D. S. Littler, 1992. Nutrient availability
mentary phosphorus chemical fractionation. J. envir. Qual. 23: to marine macroalgae in siliciclastic versus carbonate-rich coastal
1093-1102. waters. Estuaries 15: 75-82.
Barbanti, A., F. Frascari, D. Paltrinieri & G. Rosso, 1992. Transport Lavery, P. S. & J. A. Mc Comb, 1991. Macroalgal-Sediment nutrient
of nutlients in rivers: investigation on the Po river (Italy). Sci. interactions and their importance to macroalgal nutrition in a
Total. Envir. Suppl. 92: 337-344. eutrophic estuary. Estuar. coast. Shelf Sci. 32: 281-295.
Bartoli, M., M. Cattadori, G. Giordani & P. Viaroli, 1994. Oxy- Mann, K. H., 1988. Production and use of detritus in various freshwa-
gen, sulphide and nutrient fluxes in shallow eutrophic lagoons ter, estuarine, and coastal marine ecosystems. Limnol. Oceanogr.
with different primary producer communities. II. Superficial sed- 33: 894-910.
iment profiles of nutrients, iron and reduced sulfur pools. In P. Mesnage, V. & B. Picot, 1995. The distribution of phosphate in
Caumette (Coord.) C.L.E.AN. Progress Report 1994. EU Envi- sediment and its relation with eutrophication of a Mediterranean
ronment Programme DG XII, Brussels: 353-367. coastal lagoon. Hydrobiologia 297: 29-41.
Blackburn, T. H. & N. D. Blackburn, 1993. Rates of microbial Miller-Way, T., G. S. Boland, G. T. Rowe & R. R. Twilley, 1994.
processes in sediments. Phil. Trans. r. Soc., Lond. 344: 49-58. Sediment oxygen consumption and benthic nutrient fluxes on
Boynton, W. R. & W. M. Kemp, 1985. Nutrient regeneration and the Louisiana continental shelf: a methodological comparison.
oxygen consumption by sediments along an estuarine salinity Estuaries 17: 809-815.
gradient. Mar. Ecol. Prog. Ser. 23: 45-55. Nienhuis, P. H., 1992. Eutrophication, water management, and the
Buchsbaum, R., I. Valiela, T. Swain, M. Dzierzesky & S. Allen, functioning of Dutch estuaries and coastal lagoons. Estuaries 15:
1991. Available and refractory nitrogen in detritus of vascular 538-548.
plants and macroalgae. Mar. Ecol. Prog. Ser. 72: 131-143. O'Kane, J. P., M. Suppo, E. Todini & J. Turner, 1992. Physical
Castel, J., P. Caumette & R. Herbert, 1996. Eutrophication gradients intervention in the lagoon of Sacca di Goro. An examination
in coastal lagoons as exemplified by the Bassin d' Arcachon and using a 3-D numerical model. Sci. Total Envir. suppl. 92: 489-
the Etang du Prevost. Hydrobiologia 329 (Dev. Hydrobiol. 117): 510.
xi-xxx. Pugnetti, A., P. Viaroli & I. Ferrari, 1992. Processes leading to dys-
Caumette, P., 1986. Phototrophic sulphur bacteria and sulphate trophy in a Po River Delta lagoon (Sacca di Goro): phytoplancton-
reducing bacteria causing red waters in a shallow brackish coastal macroalgae interactions. Sci. Total Envir. suppl.: 445-456.
lagoon (Prevost Lagoon, France). FEMS Microbiol. Ecol. 38: Rizzo, W. M., G. J. Lackey & R. R. Christian, 1992. Significance of
113-124. euphotic, subtidal sediments to oxygen and nutrient cycling in a
Dal Cin, R. & P. Pambianchi, 1991. I sedimenti della Sacca di Goro temperate estuary. Mar. Ecol. Prog. Ser. 86: 51-61.
(Delta del Po). In S. Bencivelli & N. Castaldi (eds), Studio inte- Sand-Jensen, K. & J. Borum, 1991. Interactions among phytoplank-
grato dell'ecologia della Sacca di Goro, Francoangeli, Milano: ton, periphyton, and macrophytes in temperate freshwater and
253-263. estuaries. Aquat. Bot. 41: 137-175.
Enoksson, v., 1993. Nutrient recycling by coastal sediments: effects Sfriso, A., B. Pavoni, A. Marcomini & A. A. Orio, 1992. Macroal-
of added algal material. Mar. Ecol. Prog. Ser. 92: 245-254. gae, nutrient cycles, and pollutants in the lagoon of Venice. Estu-
Escaravage, v., 1990. Daily cycles of dissolvd oxygen and nutri- aries 15: 517-528.
ent content in a shallow fishpond: the impact of water renewal. Sloth, N. P., N. Risgaard-Petersen, S. Rysgaard, S. Pedret Pele-
Hydrobiologia 207: 131-136. gri, 1993. Nitrification, denitrification and nitrate ammonifica-
Giordani, G., M. Bartoli, M. Cattadori & P. Viaroli, 1996. Sul- tion in sediments of two coastal lagoons in southern France.
phide release from anoxic sediments in relation to iron availabil- In P. Caumette (Coord.), CLEAN Progress Report, European
ity and organic matter recalcitrance and its effects on inorganic Commission. EU Environment Programme DG XII, Brussels:
phosphorus recycling. Hydrobiologia 329 (Dev. Hydrobiol. 117): 159-185.
205-216. Valderrama, J. c., 1981. The simultaneous analysis of total nitrogen
Golterman, H. L. & A Booman, 1988. Sequential extraction of iron- and total phosphorus in natural waters. Mar. Chern. 10: 109-122.
phosphate and calcium-phosphate from sediments by chelating Val Klump, J. & c. Martens, 1981. Biogeochemical cycling in an
agents. Verh. int. Ver. Limnol. 23: 904-909. organic rich coa~tal marine ba~in-Il. Nutrient sediment-water
Golterman, H. L., 1995. The role of iron hydroxide-phosphate- exchange processes. Geochim. Cosmochim. Acta 45: 101-121.
sulphide system in the phosphate exchange between sediments Viaroli, P., A. Pugnetti & I. Ferrari, 1992. Viva rigida growth and
and overlying water. Hydrobiologia 297: 43-54. decomposition processes and related effects on nitrogen and phos-
Hansen, L. S. & T. H. Blackburn, 1991. Aerobic and anaerobic min- phorus cycles in a coastal lagoon (Sacca di Goro, Po River Delta).
eralization of organic material in marine sediment microcosms. In G. Colombo, I. Ferrari, V. U. Ceccherelli & R. Rossi (eds),
Mar. Ecol. Prog. Ser. 75: 283-291. Marine eutrophication and population dynamics. Olsen & Olsen,
Hopkinson, C. S. Jr., 1987. Nutrient regeneration in shallow-water Fredensborg: 77-84.
sediments of the estuarine plume region of the nearshore Georgia Viaroli, P. & M. Naldi, 1992. Ricerche sui cicli di azoto e fosforo
Bight, USA. Mar. BioI. 94: 127-142. in una laguna costiera eutrofizzata (Sacca di Goro, Delta del Po).
Johnston, C. A., 1991. Sediment and nutrient retention by freshwater S.IT.E ATT! 15: 95-115.
wetlands: effects on surface water quality. Critical Reviews in
Environmental Control 21: 491-565.
159
Viaroli, P., M. Naldi, R. R. Christian & I. Fumagalli, 1993. The & M. Naldi, 1996. Macrophyte communities and their impact
role of macroalgae and detritus in the nutrient cycles in a shallow on benthic fluxes of oxygen, sulpide and nutrients in shallow
water dystrophic lagoon. Verh. int. Ver. Limnol. 25: 1048-1051. eutrophic environments. Hydrobiologia 329 (Dev. Hydrobiol.
Viaroli, P., M. Bartoli, C. Bondavalli, R. R. Christian, G. Giordani (17): 105-119.
Hydrobi%gia 329: 161-174, 1996. 161
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
David T. Welsh 1,2*, Sophie Bourgues, Rutger de Wit2 & Rodney A. Herbert 1
1Department of Biological Sciences, University of Dundee, Miller's ll»nd, Dundee DD 1 4HN, Scotland
2Laboratoire d'Oceanographie Biologique, Centre d'Oceanographie et de Biologie Marine,
Universite Bordeaux 1, 2 rue Professeur Jolyet, 33120 Arcachon, France
* Present address. Fax: 33-56-835104
Key words: Acetylene reduction, nitrogen fixation, sulphate reduction, rhizosphere, Zostera noltii, root exudates
Abstract
Nitrogen fixation (acetylene reduction) rates were measured over an annual cycle in meadows of the seagrass
Z. noltii and uncolonised sediments of the Bassin d' Arcachon, south-west France, using both slurry and whole core
techniques. Measured rates using the slurry technique in Z. noltii colonised sediments were consistently higher than
those determined in isolated cores. This was probably due to the release of labile organic carbon sources during
preparation of the slurries. Thus, in colonised sediments the whole core technique may provide a more accurate
estimate of in situ activity. Acetylene reduction rates measured by the whole core technique in colonised sediments
were 1.8 to 4-fold greater, dependent upon the season, in the light compared with those measured in the dark,
indicating that organic carbon released by the plant roots during photosynthesis was an important factor regulating
nitrogen fixation. In contrast acetylene reduction rates in uncolonised sediments were independent of light.
Addition of sodium molybdate, a specific inhibitor of sulphate reduction inhibited acetylene reduction activity
in Z. noltii colonised sediments by > 80% as measured by both slurry and whole core techniques irrespective of the
light regime, throughout the year inferring that sulphate reducing bacteria (SRB) were the dominant component
of the nitrogen fixing microftora. A mutualistic relationship between Z. noltii and nitrogen fixing SRB in the
rhizosphere, based on the exchange of organic carbon and fixed nitrogen is proposed. In uncolonised sediments
sodium molybdate initially severely inhibited acetylene reduction rates, but the level of this inhibition declined
over the course of the year. These data indicate that the nitrogen fixing SRB associated with the Zostera roots and
rhizomes were progressively replaced by an aerobic population of nitrogen fixers associated with the decomposition
of this recalcitrant high C:N ratio organic matter.
Acetylene and sulphate reduction rates in the seagrass beds showed distinct summer maxima which correlated
with a reduced availability of NHt in the sediment and the growth cycle of Z. noltii in the Bassin. Overall, these
data indicate that acetylene reduction (nitrogen fixation) activity in the rhizosphere of Z. noltii was regulated both
by release of organic carbon from the plant roots and maintenance of low ammonium concentrations in the root
zone due to efficient ammonium assimilation.
Nitrogen fixation rates determined from acetylene reduction rates measured by the whole core technique ranged
from 0.1 to 7.3 mg N m- 2 d- 1 in the Z. noltii beds and between 0.02 and 3.7 mg N m- 2 d- 1 in uncolonised
sediments, dependent upon the season. Nitrogen fixation in the rhizosphere of Z. noltii was calculated to contribute
between 0.4 and 1.1 g N m -2 y-l or between 6.3 and 12% of the annual fixed nitrogen requirement of the plants.
Heterotrophic nitrogen fixation therefore represents a substantial local input of fixed nitrogen to the sediments of
this shallow coastal lagoon and contributes to the overall productivity of Z. noltii in this ecosystem.
high levels of primary production substantial inputs of ly been proposed as potentially the most important
fixed nitrogen are required (Patriquin, 1972). Whilst, heterotrophic nitrogen fixers in coastal marine sedi-
efficient recycling of organic nitrogen in the sediment ments (Herbert, 1975; Nedwell & Aziz, 1980). Where,
can supply a large proportion of this fixed nitrogen sulphate reduction is the dominant metabolic process,
(Iizumi et ai., 1982; Dennison et ai., 1987; Caffrey accounting for up to 50% of all organic matter miner-
& Kemp, 1992), several investigations have demon- alisation (J0rgensen, 1977; J0rgensen, 1982; Canfield,
strated that porewater concentrations of inorganic N 1989).
are insufficient to meet the growth requirements of
the plant communities (Patriquin, 1972; Short, 1983;
Moriarty et ai., 1985). Heterotrophic nitrogen fixation Materials and methods
in the phylosphere and rhizosphere of seagrasses may
therefore play an important role in regulating prima- Sampling site
ry production in these ecosystems. High rates of het-
erotrophic nitrogen fixation have been reported in sea- The sampling site used in this study is described by
grass colonised sediments and estimated to supply upto Welsh et ai., (1996) and corresponds to the CLEAN
50% of the nitrogen requirement of the plant communi- sampling Station A. Samples were collected between
ties (Patriquin & Knowles, 1972; McRoy et ai., 1973; March 1994 and February 1995, both from an area
Capone et al., 1979; Capone & Taylor, 1980; Wolfend- within the Z. noltii beds and an adjacent area where the
en & Jones, 1987; a'Donohue et ai., 1991a; Moriarty Zostera had recently died back but decaying roots and
& a'Donohue, 1993). However, there is a substantial rhizomes were still present.
energy cost associated with nitrogen fixation, estimat-
ed to be equivalent to 16 ATP per molecule of N2 Determination of acetylene reduction (nitrogen
fixed (Postgate, 1982) and thus rates of heterotrophic fixation) rates
nitrogen fixation in natural environments are generally
considered to be limited by the availability of suitable Nitrogen fixation rates were measured using the acety-
organic carbon substrates (Herbert, 1975; Zuberer & lene reduction technique of Stewart et ai., (1967), using
Silver, 1978; Nedwell & Aziz, 1980; Jones, 1982). The both slurry and whole core techniques as described
high rates of heterotrophic nitrogen fixation reported below.
in seagrass and salt marsh grass sediments have been
demonstrated to be associated with the excretion of Slurry experiments
organic compounds from the plant roots and close- Acetylene reduction rates in slurry experiments were
ly coupled to the photosynthetic activity of the plants determined as described by Welsh et ai. (1996).
(Capone et ai., 1979; Boyle & Patriquin, 1981; Whit-
ing et ai., 1986; a'Donohue et ai., 1991a). Whole core measurements
Previous studies on the rates of nitrogen fixation Large sediment cores were collected by inserting
in seagrass meadows and the relationship between the 5 x 25 cm (internal diameter) grey plastic core tubes
plants and the heterotrophic nitrogen fixing microflo- into the sediment until the sediment surface was lev-
ra in the rhizosphere have been undertaken primarily el with the rim of the core tube. The surrounding
in tropical or sub-tropical areas and thus these rela- sediment was removed, the sediment below the core
tionships are much less well characterised in tem- tube was sliced using a steel wire and the core care-
perate areas. In this study we have investigated fully transferred onto a perspex sheet for transporta-
the seasonal variation in nitrogen fixation (acetylene tion to the laboratory. In the laboratory cores were
reduction) rates in Zostera noltii Hornem. colonised stored under natural light conditions in a 50 cm (water
and uncolonised sediments in the Bassin d' Arcachon, depth) x 3 m (internal diameter) water bath (volume
South-West, France. The dependency of nitrogen fix- approx 600 litres), circulated with aerated seawater
ation on plant photosynthesis and the potential role from the Bassin for a maximum of 2 days before
played by nitrogen fixing sulphate reducing bacteria use. After equilibration the cores were transferred to
(SRB) in the rhizosphere was investigated. The abil- small water baths and the water level lowered to the
ity to fix nitrogen is widely distributed amongst SRB level of the sediment surface. Plexiglass core tubes
(Riederer-Henderson & Wilson, 1970; Postgate et ai., (20 x 5 cm, internal diameter) were inserted into the
1985; Postgate et ai., 1988), which have previous-
163
sediment, taking care not to damage the Zostera leaves, the first dilution tube and gently sonicated for 1 min to
when present. The core tubes were sealed with rubber release attached bacteria, decimal dilution series were
bungs and 10% of the heads pace volume was replaced prepared using anaerobic autocIaved filtered seawa-
with acetylene via a Suba-seal sampling port on the ter supplemented with Na2S. 9 H 20 (final concentra-
side of the core tube. The cores were incubated under tion 0.8 mM) in Hungate tubes previously flushed with
either natural light.or in the dark by covering the core N2/C02 90/10% v/v. Aliquot volumes (0.25 ml) of
tube with a double layer of aluminium foil. Molybdate each dilution were used to inoculate 8 replicate Veno-
treated cores were pre-incubated for 12-16 hours with ject tubes. The tubes were incubated in the dark at
an overlying water column of natural seawater supple- room temperature (20-25 0c) and regularly checked
mented with 25 mmol I-I sodium molybdate. During for the formation of black FeS precipitates which were
the incubation period, triplicate 1 ml samples of the scored as positive for the growth of SRB. MPN esti-
headspace gas were collected at 2-hour intervals over mates and their 95% confidence intervals were cal-
a 12-hour period and stored by inserting the needle into culated using the computer programme developed by
a butyl rubber bung. Clarke & Owens, (1983).
Gas samples were analysed for ethylene and acety-
lene within 1-2 hours of sampling by gas chromatogra- Determination of sulphate reduction rates and
phy, using a Perkin Elmer Autosystem Gas Chromato- sediment exchangeable NHt concentration
graph fitted with a 3 m x 2.2 mm (internal diameter)
Chromo sorb 101 (80/100 mesh) column with N2 as car- Sulphate reduction rates and sediment exchange-
rier gas and flame ionisation detection. Flow rates for able ammonium concentrations were determined as
N2, air and H2 were 30, 450 and 50 ml min- i respec- described by Welsh et al. (1996), except that sulphate
tively and the oven temperature was 30°C. Ethylene reduction was determined only for the 0-2 and 2-5 cm
concentrations were calculated by reference to known depth horizons.
standards and all the data were corrected for the small
quantities of ethylene present as a contaminant in the
acetylene used in this study. Results
Enumeration of sulphate reducing bacteria Data presented in Figure 1. shows depth profiles
of acetylene reduction activity (ARA) in slurries of
Populations of viable SRB were enumerated using the Z. noltii colonised sediments. ARA was detectable
Most Probable Number (MPN) technique. The growth throughout the 0-5 cm depth horizon during the sam-
medium consisted of filtered (0.22 /Lm pore size) sea- pling periods in March, July and October 1994 and Jan-
water (1000 ml), 5 mM NH4CI, 1 mM KH 2P0 4, uary 1995. The highest activities were recorded during
20 mM NaHC0 3 , 100 /LM Na2S204, 200 /LM FeS04, the summer and autumn sampling programmes, with
0.1 g yeast extract, 0.5 ml SL 12B trace elements solu- the peak of activity occurring in the 0-2 cm depth hori-
tion without EDTA (Pfennig & Triiper, 1992), l.0 ml zon (Figure 1). In winter and spring ARA was reduced
vitamin V7 solution (Pfennig et aI., 1981), 10 mM and the activity maximum occurred at greater depth
sodium lactate and 5 mM sodium acetate. The vitamin (Figure 1), presumably in response to changes in the
solution, sodium dithionite and iron solutions were fil- depth of O 2 penetration and carbon availability. The
ter sterilised (0.22 /Lm pore size) and aseptically added addition of 20 mmoll- I sodium molybdate a specif-
together with the autocIaved carbonate buffer to the ic inhibitor of bacterial sulphate reduction (Taylor &
bulk autocIaved medium when cool. The medium was Oremland, 1979; Smith & Klug, 1981; Oremland &
prepared according to the procedure described by Wid- Capone, 1988) to sediment slurries severely inhibit-
del & Bak, (1992) and the pH adjusted to pH 7.2 using ed ARA by between 75-95% throughout the sediment
sterile 1 M HCI and NaOH before aseptically dispens- profiles in all seasons (Figure 1).
ing 3 ml volumes into sterile 4 ml Venoject tubes. Acetylene reduction activity was also detectable
Populations of viable SRB were determined for 4 throughout the 0-5 cm depth horizon in slurries pre-
replicate sediment cores (3 cm internal diameter), the pared from uncolonised sediments (Figure 2). Rates
0-2 cm depth horizon of each core was sectioned, of acetylene reduction were always lower than those
transferred to a sterile petri dish and homogenised recorded in sediments colonised by Z. noltii and
using a sterile spatula. A 1 ml aliquot was transferred to showed a much lower seasonal variation, although a
164
Q
Q
N
Q
:..
Q
a.. Q
Oc
... ...N
0
a
~
0-1
•
=
0
-.: 2-3
N
1-2
•
•
J
0
.:::
.:::
a 3-4
•
•
~
0 4-5
-L.--'--
_.
,-. --I -- j
S 0-1 I
~
c 1-2
0
•
N
'': J
0 2-3
.::: I
....
.:::
Q.
3-4 I
~ J
0 4-5
~
summer maximum was apparent (Figure 2.). Sodium fixers_ Rates recorded by the whole core technique
molybdate additions inhibited ARA by >50% (Fig- for cores pre-incubated with sodium molybdate were
ure 2), indicating that SRB were also the dominant 15.5 and 28-fold lower during light and dark incuba-
component of the nitrogen fixing microflora present in tions respectively than those recorded in the absence
the uncolonised sediments. of sodium molybdate (Figure 3), indicating that SRB
A typical data set obtained using the whole core were responsible for the bulk of the recorded activi-
technique in July 1994 in Z. noltii colonised sediments ty. In molybdate treated cores the acetylene reduction
are shown in Figure 3. ARA was significantly influ- rate during light incubations was 6.7-fold greater than
enced by light with the rate in light incubated cores during dark incubations (Figure 3), inferring that the
being 4-fold greater than that recorded in the dark. molybdate resistant component of the nitrogen fixing
These data demonstrate that the photosynthetic activity microflora were also dependent upon the photosynthet-
of the Zostera was a major factor influencing nitrogen ic activity of the Zostera. In contrast, little difference
fixation (ARA) in the rhizosphere, since, nitrogen fix- was recorded between the acetylene reduction rates
ing cyanobacteria were not observed in the sediments during light and dark incubations in uncolonised sedi-
or as epiphytes on the Zostera leaves_ Additionally, ments indicating that the stimulation observed by light
data from the slurry experiments demonstrate that a in colonised sediments was a result of photosynthet-
large proportion of ARA occurred at depths where ic inputs by the Zostera rather than an effect of light
light was not available for photosynthetic nitrogen per se.
165
Acetylene reduction rate nnlOl. ml sedimenrl.h· 1
a 0-1
~
:: 1-2
0
N
.;::
0 2-3
.c
.c J-4
c..
~
c:l 4-5
'-------_. __ ........ _ - - - - '
E 0-1
~
1-2
2-3
3-4
4-5
Figure 2. Seasonal variation in depth profiles of acetylene reduction activity (nitrogen fixation) recorded in uncolonised sediments in March.
July and October 1994 and January 1995. using the slurry technique in the absence (open bars) and presence (solid bars) of20 mmoll- i sodium
molybdate. Data points represent the means of 5 replicate determinations. standard deviations were generally less than 5%.
Seasonal patterns of ARA measured in Z noltii pIing periods with rates measured in slurries being only
colonised sediments using both the slurry and whole 1.1 and 1.6-fold greater than those measured during
core techniques are presented in Figure 4. Acetylene light incubations of whole cores in July and October
reduction activity showed a mid-summer peak with respectively (Figure 4A & B).
highest rates recorded in July independently of which In Z noltii colonised sediments the acetylene reduc-
technique was used. However, the degree of seasonal tion rates recorded during light incubations of whole
variation observed differed considerably dependent on cores were always higher than those measured in
which technique was used. Acetylene reduction rates the dark (Figure 4B & C). The degree of stimula-
recorded by the slurry technique were always higher tion observed, however, varied between 1.8 and 4-
than those recorded by the whole core technique irre- fold dependent on the season, being greatest in July
spective of light regimes. These differences between and lowest in January, indicating that the influence of
the two techniques varied considerably with season Z noltii photosynthesis on ARA varied with the growth
with rates measured by the slurry technique being 9.4 phase of the plants. Addition of sodium molybdate
and 23-fold greater than those recorded using the whole severely inhibited ARA activity throughout the annual
core technique (light incubation) and 19.2 and 41.8- cycle by between 85 and 95%, irrespective of which
fold greater than whole core dark incubations in March technique was used (Figure 4), inferring that ARA was
and January respectively (Figure 4.). In contrast little primarily mediated by SRB.
difference was evident between values obtained by the Rates of acetylene reduction recorded in
two techniques during the summer and autumn sam- uncolonised sediments were substantially lower than
166
250 411 A
20
'E
C
:::
~
"'0
ISO
• • •
""::l
"'0
0
5. :: 40 B
il"
;:,
.:::
~
50
\l
8 12
40 c
rime OlOurs)
those recorded in Z. noltii colonised sediments (Fig- Fi/?ure 4. Seasonal variation in the rates of acetylene reduction mea-
sured in Zostera noltii colonised sediments between March 1994 and
ure 5). Rates of acetylene reduction recorded using January 1995. Rates were measured in the absence (open circles) and
the slurry technique showed only a small annual varia- presence (solid circles) of 20 mmoll- i sodium molybdate.
tion and were always inhibited by> 75% following the
A. Integrated rates for the 0-5 cm depth horizon deter-
addition of sodium molybdate (Figure 5A), inferring
mined using the slurry technique.
that SRB were responsible for the bulk of this activity. B. Rates determined under natural light conditions using
Little difference was recorded between light and dark the whole core technique.
incubations of the whole core technique (Figure 5B C. Rates determined during dark incubations using the
& C). Initially, sodium molybdate additions severe- whole core technique.
ly inhibited ARA by approximately 80-90% during
both light and dark incubations in March and July
respectively (Figure 5B & C). However, this level of concentrations were highest in winter with a concen-
inhibition decreased to 12 and 36% during light and tration of approximately 290 /Lmol per litre of sediment
dark incubations respectively in October and similar- recorded in January 1995. The exchangeable NHt
ly decreased in January (Figure 5B & C). These data concentration declined rapidly during spring and early
indicate a shift in the composition of the heterotroph- summer attaining a minimal value of approximately
ic nitrogen fixing microftora had occurred over this 190/Lmol 1 sedimenc I in May which was maintained
period. throughout the summer. The NHt pool slowly regen-
The mid-summer peak in nitrogen fixation (acety- erated during the autumn and winter period (Figure 6).
lene reduction) rates in Z. noltii colonised sediments This annual cycling of sediment exchangeable
correlated well with the availability of NHt in these NHt pool is consistent with the life cycle of Z. noltii
sediments (Figure 6). Sediment exchangeable NHt in the Bassin, which has a minimal standing crop dur-
167
...c:;
40 A
a 300
:.a
OJ
'"
'"
~
20
.::: "0
"!
E
--z
::t
8
200
'0 • •
8
:t ~
.... 40 B -;
..c
Q,)
(':I
'"
.:::....
~
u;<
u ~
.
:::l
"0 20 100
Q,)
1\1 A 1\1 J J A SON D J
Q,)
cQ,) Month
;:.
.... Figure 6. Seasonal variation in the sediment exchangeable ammoni-
Q,)
u um concentration between March 1994 and January 1995 in Zostera
< noltii colonised sediments (open circles) and uncolonised sediments
40 C (solid circles). Data points represent the means of triplicate determi-
nations, standard deviations have been omitted for the sake of clarity,
but were generally 15% (colonised sediments) and 5% (uncolonised
sediments) of the mean values.
20
2.0 25 7.0
T
I 1
"0
5 1.5 20
5 \-' ::; 6.0
~
B
..""
Co
=
:0
"'"
,....,
1
5
=
15 i
Il.
::;
5.0
~
0.5 10
M A ~1 J .1 A SON D J
Month
Figure 7. Seasonal variation of surficial sediment temperature (solid J\fAM,I,IASOND.J F
circles) and the integrated sUlphate reduction rate (open circles) for
the 0--5 cm depth horizon determined under natural light conditions Month
in Zostera noltii colonised sediments. Data points for sulphate reduc- Figure 8. Seasonal variation of the most probable number of sul-
tion rates represent the means of triplicate determinations ± standard phate reducing bacteria in the 0--2 cm depth horizon of Zostera noltii
deviation. The rate of sulphate reduction was not determined in Jan- colonised sediments determined between March 1994 and February
uary 1995. 1995. Data points represent the means of 4 replicate determina-
tions ± standard deviation.
Thus, acetylene reduction rates measured by the slurry all ARA increasing from 1.02 (light incubation) and
technique are more indicati ve of the potential ARA due 0.56 /lmol m- 2 h- I (dark incubation) in January to
to the elevated carbon status of the slurries, whereas in 32.41 (light incubation) and 8.70 /lmol m- 2 h- I (dark
situ rates will be lower due to the limited availability incubation) in July (Figure 4b & c). These differences
of suitable carbon sources. This hypothesis is support- in both the overall rates of ARA and the level of stimu-
ed by the fact that the addition of a range of different lation by light are consistent with the annual variation
carbon sources to slurries of Z. noltii colonised sedi- of the Z. noltii root and shoot biomass which is min-
ments at a final concentration of 15 mmoll- I resulted imal in winter and maximal in summer (Au by, 1991).
in only a minor 15-30% stimulation of ARA (Welsh Therefore, potential photosynthetic inputs to the rhi-
et aI., 1996). zosphere would be greatest in mid-summer, whereas
In contrast the intact core technique may provide in winter these would be substantially reduced due to
a more realistic estimate of the in situ rate of acety- the lower standing crop and shorter day length. Thus,
lene reduction, since, using this technique the distur- seasonal variations in plant biomass and rate of carbon
bance of physicochemical gradients of 02 sulphide and release by the plant roots, may be one of the major fac-
nutrients and the relationship between the plant roots tors regulating heterotrophic nitrogen fixation in the
and the rhizosphere microftora is minimal. Addition- rhizosphere. The hypothesis that the input of labile
ally, the use of the intact core technique in seagrass organic carbon to the rhizosphere via the Z. noltii roots
colonised sediments may preferentially measure ARA is a major determinant of ARA in the rhizosphere is
associated with the plant roots, since diffusion of acety- supported, if as discussed earlier, we consider that the
lene to the rhizosphere will be facilitated by transfer slurry technique is a measure of the potential ARA,
via lacunal air spaces within the plant leaves and roots whereas, intact core measurements approximate to the
(O'Donohue et aI., 1991b). This method may howev- in situ rate. Thus, in July. ARA measured by the whole
er, lead to a slight underestimate of the overall rate of core technique during light incubations was equivalent
acetylene reduction, since nitrogen fixing bacteria in to >90% of the potential (integrated slurry) activity
zones remote from the plant roots may not be subject (Figure 4a & b) whereas, in January rates measured
to a saturating acetylene concentration. by the whole core technique in the light represented
Acetylene reduction rates in Z. noltii colonised sed- <5% of the potential activity in the slurries. These data
iments measured by the whole core technique in the indicate that in January ARA was severely limited by
light were always higher than those measured in the availability of suitable organic carbon sources, where-
dark (Figures 3 & 4). In contrast, in uncolonised sed- as in July when the potential photosynthetic inputs of
iments the rates measured during light and dark incu- organic carbon from the Zostera roots was greatest the
bations were similar (Figure 5) and comparable with rate measured during light incubations of whole cores
rates measured during dark incubations of colonised approximates to the potential rate as measured by the
sediments (Figure 4). These data demonstrate that slurry technique.
heterotrophic nitrogen fixation (acetylene reduction) In contrast to colonised sediments acetylene reduc-
in the rhizosphere of Z. noltii was fuelled by inputs tion rates measured in uncolonised sediments, using
of organic carbon from the plant roots during pho- the whole core technique were not light dependent
tosynthesis, whereas, during dark incubations or in (Figure 5b & c) and were similar to those recorded
uncolonised sediments activity is linked to the degra- during dark incubations of Z. noltii colonised sedi-
dation of moribund organic matter in the sediment ments (Figure 4c). In these sediments ARA appears
and thus independent of light. These results are in to be linked to the degradation of high C:N ratio
agreement with previous studies of sea and saltmarsh ("" 38: 1, Welsh, unpublished data) moribund root and
grasses which have demonstrated that nitrogen fixa- rhizome biomass which was present. Previous, studies
tion/acetylene reduction rates are closely coupled to have similarly recorded heterotrophic nitrogen fixation
the photosynthetic activity of the plants and stimulated associated with decomposing seagrass roots (Kenwor-
by release of labile carbon from the plant roots (Boyle thy et aI., 1987; O'Donohue et aI., 1991a). Rates of
& Patriquin, 1981; Whiting et aI., 1986; O'Donohue acetylene reduction in these sediments also showed a
et aI., 1991a; Moriarty & O'Donohue, 1993). In this mid-summer peak (Figure 5) and are most probably
study the level of stimulation of ARA by light was vari- regulated by the quantity and quality of the residual
able with season, ranging from a 1.8-fold stimulation in organic matter.
January to a 4-fold stimulation in July, with the over-
170
The availability of fixed nitrogen sources in the sed- McClung, 1978; Capone & Taylor, 1980; Capone &
iment would also be expected to exert a major influ- Budin, 1982), conditions which would also stimulate
ence on nitrogen fixation rates as both the synthesis SRB.
and activity of nitrogenase are sensitive to feedback Sulphate reduction rates measured in Z. noltii
inhibition (Zumft & Castillo, 1978; Postgate, 1982; colonised sediments were high ranging from 15.6 to
Dixon, 1984; Postgate & Kent, 1984). Additions of 32.6 mmol SO~- m- 2 d- 1 and are comparable with
NHt have previously been demonstrated to inhib- those determined in other organically rich sediments,
it in situ ARA in the rhizosphere of the salt marsh including seagrass (Blackburn et aI., 1994), salt marsh
grass Spartina alternifiora Loisel (Yoch & Whiting, (Hines et ai., 1989; Hines et aI., 1994) and tropical
1986). Sediment exchangeable NHt concentrations in mangrove sediments (Hines & Lyons, 1982; Kris-
Z. noltii colonised sediments measured in this study tensen et ai., 1994; Nedwell et aI., 1994). Sulphate
showed a distinct annual cycle (Figure 6), with the reduction rates were highest in summer and approxi-
NHt pool exhibiting an inverse relationship to ARA mately double those measured in spring and autumn
(Figure 4). The exchangeable NHt pool decreased (Figure 7). The seasonal changes in sulphate reduction
during spring reaching a plateau concentration in May rate correlated well with changes in sediment tempera-
of approximately 190 tLmol 1 sediment- 1 which was ture (Figure 7) and acetylene reduction rates (Figure 4).
maintained until August, corresponding to the peri- This increase in sulphate reduction rate during summer
od of maximal ARA, thereafter the NHt pool slowly was primarily due to increased activity in the surficial
regenerated during the autumn and winter period and 0-2 cm depth horizon (data not shown) and thus cor-
this was accompanied by a corresponding reduction in relates with the shift in ARA towards the sediment
ARA. In contrast the sediment exchangeable NHt pool surface in summer observed in Z. noltii colonised sed-
in uncolonised sediments maintained a stable value of iments (Figure 1). However, the relationship between
between 280 and 300 tLmol 1 sedimenC 1 throughout the sulphate reduction rate and sediment temperature
the year (Figure 6), inferring that the changes observed may be coincidental, since these changes also correlate
in colonised sediments were a result of NHt- assimi- with the changes in the Z. noltii standing crop (Au by,
lation by the plants. Thus nitrogen fixation in the rhi- 1991) and thus potential carbon inputs from the plant
zosphere of Z. noltii appears to be regulated by the roots. It has previously been demonstrated in salt marsh
activity of the plant roots due both to the supply of sediments, colonised by Spartina alternifiora, that sul-
organic carbon as root exudates during photosynthesis phate reduction rates in the rhizosphere were closely
and the maintenance of low NHt concentration in the coupled to the growth phase of the plants and driven
rhizosphere during the growth season. by the release of dissolved organic carbon exudates
ARA in Z. noltii colonised sediments was severely from the plant roots during active growth (Hines et aI.,
inhibited (>75%) in the presence of sodium molyb- 1989). Similar seasonal variations in bacterial produc-
date, in both slurries and whole cores irrespective of tivity in the rhizosphere of seagrasses and plant growth
light regime throughout the year (Figures I, 3 & 4). ha ve also been observed and linked to the exudation of
Sodium molybdate is a potent inhibitor of sulphate organic carbon from the plant roots (Moriarty & Boon,
reduction (Taylor & Oremland, 1979; Smith & Klug, 1989). A similar relationship between the growth of
1981; Oremland & Capone, 1988) and these data pro- Z. noltii and the release of organic carbon by the plant
vide strong evidence that SRB were the dominant com- roots could account for the observed changes in sul-
ponent of the nitrogen fixing microflora in the rhizo- phate reduction rate in these sediments. Thus, a mutu-
sphere. In a previous study, O'Donohueet aI., (1991a), alistic relationship may exist between the Z. noltii and
estimated that approximately 90% of nitrogen fixation SRB in the rhizosphere.
in meadows of the seagrass Zostera capricornia in In uncolonised sediments sodium molybdate addi-
Moreton Bay, Australia was due to the activity of anaer- tions also inhibited ARA (Figure 5) indicating that
obic or microaerophilic bacteria in the rhizosphere, nitrogen fixing SRB were present. However, the degree
which would have included SRB, although no attempts of inhibition measured in whole core incubations (Fig-
were made to further identify the bacteria responsi- ure 5b & c) decreased over the year indicating a popula-
ble. Additionally, numerous studies have reported a tion change in the nitrogen fixing microflora, with SRB
stimulation of nitrogen fixation activity under anaer- being replaced by aerobic or possibly microaerophilic
obic or microaerophilic conditions in seagrass sed- bacteria, since sodium molybdate consistently inhibit-
iments or isolated roots and rhizomes (Patriquin & ed ARA by approximately 80 to 90% throughout the
171
Table 1. Comparison of nitrogen fixation rates recorded in the Bassin d' Arcachon with rates recorded in other seagrass
meadows. Data for the Bassin d' Arcachon are based on acetylene reduction rates from whole core experiments only and were
calculated using the theoretical 3: 1 ratio between acetylene reduction and nitrogen fixation.
year in anaerobic slurries (Figures 2 & Sa). These data fixing microorganisms possess uptake hydrogenases
indicate that in this area where the Zostera had recently (Adams & Mortensen, 1984; Houchins, 1984; Voor-
died back at the beginning of the experimental period, douw et a!., 1990) and thus in vivo e- lost due to
the initial population of nitrogen fixing SRB associated hydrogen evolution during nitrogen fixation may be
with the plant roots was progressively replaced by an efficiently recycled to nitrogenase. Additionally, there
aerobic or microaerophilic nitrogen fixing population is some controversy over the use of the theoretical
associated with the degradation of residual root and 3: I ratio since acetylene is not a natural substrate for
rhizome biomass. nitrogenase and inhibits other sedimentary processes
Rates of acetylene reduction can be used to calcu- (Payne, 1984; Oremland & Capone, 1987; Capone,
late rates of nitrogen fixation using the theoretical ratio 1988). Nevertheless, studies directly comparing acety-
of 3: 1 based on the provision of reducing equivalents lene reduction rates and 15N2 fixation rates in seagrass
as shown in equations 1 and 2. meadows or isolated seagrass roots have shown reason-
able agreement with the theoretical 3 : I ratio. (Patriquin
C 2H 2 + 2H+ + 2e- - t C 2H 4 (1) & Knowles, 1972; Capone & Budin, 1982; O'Donohue
N2 + 6H+ + 6e- - t 2N H3 (2) et a!., 1991 a) and we believe it is therefore appropriate
to use this ratio in this study.
However, in vitro studies of nitrogenase have Acetylene reduction rates measured in this study
demonstrated that e- are also shunted to H2 production using the whole core technique, yield rates equiv-
during nitrogen fixation as shown in equation 3, but this alent to nitrogen fixation rates of between 0.1 and
does not occur during acetylene reduction (Postgate, 7.3 mg N m- z d- I for Z. noltii colonised sediments and
1982; Capone, 1988). between 0.02 and 3.7 mg N m- 2 d- I in uncolonised
sediments, dependent on the season if the 3: 1 ratio is
N2 + 8H+ + 8e- -t 2NH3 + Hz (3) used. The rates recorded in the Z. noltii beds fall with-
This equation indicates that the ratio of acetylene in the lower region of those determined in previous
reduction to nitrogen fixation should be 4: 1, if e- studies and show a much higher degree of temporal
which are shunted to H2 production are accounted for. variation (Table 1). However, the majority of previ-
However, SRB in common with many other nitrogen ous studies reported in the literature have focused on
172
the role played by heterotrophic nitrogen fixation in for their kindness and assistance during the course of
tropical and sub-tropical seagrass ecosystems, gener- this study.
ally in oligotrophic coastal areas (Table 1). In contrast
the Bassin d' Arcachon is a semi-enclosed lagoon, sub-
ject to much greater seasonal variations in both tem- References
perature and light regimes, which receives substantial
anthropogenic inputs of fixed nitrogen, calculated to Adams, M. W W & L. E. Mortensen, 1984. The physical and catalyt-
ical properties of hydrogenase II of Clostridium pasteurianum. J.
be equivalent to 1000 tons Ny-I (Auby et aI., 1994). BioI. Chern. 259: 7045-7055.
It is therefore not surprising that the nitrogen fixation Auby, 1., 1991. Contribution a I' etude des herbiers de Zostera
rates measured in this ecosystem are somewhat low- noltii dans Ie Bassin d' Arcachon: dynamique, production et
er and seasonally more variable than those previously degradation, macrofaune associee. Ph.D. thesis, Universite de
Bordeaux I.
reported for tropical seagrass ecosystems. Auby, 1., F. Manaud, D. Maurer & G. Trut, 1994. Etude de la pro-
Auby, (1991) has previously calculated that the net liferation des algues vertes dans Ie Bassin d' Arcachon. Rapport
productivity of Z. noltii in the Bassin d'Arcachon is IFREMER-CEMAGREF-SSA-SABARC. 163 pp, IFREMER,
equivalent to between 74 and 111 g carbon m- 2 y-I Arcachon, France.
Blackburn, T. H., D. B. Nedwell & W J. Wiebe, 1994. Active mineral
for shoot biomass and between 53 and 70 g carbon cycling in a Jamaican seagrass sediment. Mar. Ecol. Prog. Ser.
m- 2 y-I for root and rhizome biomass. This biomass 110: 233-239.
has carbon to nitrogen ratios of 15.2:1 and 37.8:1 Boyle, C. D. & D. G. Patriquin, 1981. Carbon metabolism of Sparti-
na alterniflora Loisel in relation to that of a~sociated nitrogen-
for shoot and root and rhizome biomass respective-
fixing bacteria. New Phytol. 89: 275-288.
ly (Welsh, unpublished data, determined in October Caffrey, J. M. & W M. Kemp, 1992. Influence of the submersed
1994). Thus, by calculation Z. noltii in the Bassin plant, Potamogeton perf{)liatus, on nitrogen cycling in estuarine
d' Arcachon requires a fixed nitrogen input of between sediments. Limnol. Oceanogr. 37: 1483-1495.
Canfield, D. E., 1989. Sulfate reduction and oxic respiration in
6.3 and 9.2 g nitrogen m- 2 y-I. Based on the acety-
marine sediments: implications for organic carbon preservation
lene reduction rates measured in this study by the whole in euxinic environments. Deep Sea Res. 36: 121-138.
core technique, heterotrophic nitrogen fixation in the Capone, D. G., 1988. Benthic nitrogen fixation. In T. H. Blackburn
rhizosphere can provide between 0.4 and 1.1 g nitrogen & J. Slirensen (eds), Nitrogen cycling in coastal marine environ-
ments. John Wiley & Sons Ltd, Chichester. 85-123.
m- 2 y-I or between 6.3 and 12.0% of the fixed nitro- Capone, D. G. & J. M. Budin, 1982. Nitrogen fixation associated
gen requirement of the plants. Heterotrophic nitrogen with rinsed roots and rhizomes of the eelgrass Zostera marina.
fixation in the rhizosphere of Z. noltii, therefore rep- PI. Physiol. 70: 1601-1604.
resents a substantial local input of fixed nitrogen to Capone, D. G., P. A. Penhale, R. S. Oremland & B. F. Taylor,
1979. Relationship between productivity and N2 (C2H2) fixation
the surficial sediments of this shallow coastal lagoon in the rhizosphere of Thalassia testudinium. Limnol. Oceanogr.
and contributes to the overall productivity of Z. noltii 24: 117-125.
community in this ecosystem. Capone, D. G. & B. F. Taylor, 1980. N2 fixation in the rhizosphere
of Thalassia testudinium. Can. J. Microbiol. 26: 998-1005.
Clarke, K. R. & N. J. P. Owens, 1983. A simple and versatile micro-
computer program for the determination of 'most probable num-
Acknowledgments ber'. J. Microbiol. Meth. I: 133-137.
Dennison, W C, R. C. Aller & R. S. Alberte, 1987. Sediment ammo-
nium availability and eelgrass (Zostera marina) growth. Mar.
This research was carried out as part of the E.U. fund-
BioI. 94: 469-477.
ed ~oastal L.agoon Eutrophication and ANaerobic pro- Dixon, R. A., 1984. The genetic complexity of nitrogen fixation. J.
cesses (CLEAN) project, grant No EVSV-CT92-0080. gen. Microbiol. 130: 2745-2755.
Sophie Bourgues was supported by a grant from the Eppley, R. W, E. H. Renger & W G. Harrison, 1979. Nitrate and
phytoplankton production in southern California waters. Limnol.
French Ministry of Research (MESR). The authors
Oceanogr. 24: 483-494.
would like to thank all the participants in the CLEAN Herbert, R. A., 1975. Heterotrophic nitrogen fixation in shallow
project for their helpful discussions on the ecology of estuarine sediments. J. expo mar. BioI. Ecol. 18: 215-225.
this ecosystem and especially Isabelle Auby for invalu- Hines, M. E. & W B. Lyons, 1982. Biogeochemistry of nearshore
Bermudan sediments. 1. Sulfate reduction rales and nutrient
able local knowledge on the life cycle and ecology of regeneration. Mar. Ecol. Prog. Ser. 8: 87-94.
Zostera noltii in the Bassin d' Arcachon. We would also Hines, M. E., S. L. Knollmeyer & J. B. Tugel, 1989. Sulfate reduc-
like to thank Prof. Pierre Caumette for the use of labo- tion and other sedimentary biogeochemistry in a northern New
ratory facilities in his Institute and all the staff and stu- England salt marsh. Limnol. Oceanogr. 34: 578-590.
dents of the Laboratoire d'Oceanographie, Arcachon
173
Hines, M. E., G. T. Banta, A. E. Giblin & J. E. Hobbie, 1994. 0' Donohue, M. J., D. J. W. Moriarty & I. C. MacRae, 1991b. A com-
Acetate concentrations and oxidation in salt-marsh sediments. parison of methods for determining rates of acetylene reduction
Limnol. Oceanogr. 39: 140-148. (nitrogen fixation) by heterotrophic bacteria in seagrass sediment.
Houchins, J. P., 1984. The physiology and biochemistry of hydrogen J. Microbiol. Meths. 13: 171-183.
metabolism in cyanobacteria. Biochim. Biophys. Acta. 768: 227- Oremland, R. S. & D. G. Capone, 1988. Use of 'specific inhibitors'
255. in biogeochemistry and microbial ecology. Adv. Microbiol. Ecol.
Howes, B. L., J. W. H. Dacey & S. G. Wakeham, 1985. Effects of 10: 285-383.
sampling technique on measurements of pore water constituents Oremland, R. S., J. W. Gotto & B. F. Taylor, 1976. N2 (C2H2l fixa-
in salt marsh sediments. Limnol. Oceanogr. 30: 221-227. tion associated with the rhizosphere communities of the seagrass
lizumi, H., A. Hattori & c. P. McRoy, 1982. Ammonium regener- Thalassia testudinium. Abstracts of the annual meeting of the
ation and assimilation in eelgrass (Zostera marina) beds. Mar. American Society of Microbiology. abstr. 171.
BioI. 66: 59-65. Patriquin, D. G., 1972. The origin of nitrogen and phosphorus for
Jones, K., 1982. Nitrogen fixation in the temperate estuarine inter- the growth of the marine angiosperm Thalassia testudiniwn. Mar.
tidal salt marsh sediments of the river Lune. Limnol. Oceanogr. BioI. 15: 25--46.
22: 814-832. Patriquin, D. G. & R. Knowles, 1972. Nitrogen fixation in the rhi-
J0rgensen, B. B., 1977. The sulfur cycle of a coastal marine sediment zosphere of marine angiosperms. Mar. BioI. 16: 49-58.
(Limfjorden, Denmark). Limnol. Oceanogr. 22: 814-832. Patriquin, D. G. & c. McClung, 1978. Nitrogen accretion, and
J0rgensen, B. B., 1982. Mineralization of organic matter in the sea the nature and possible significance of N2 fixation (acetylene
bed: the role of sulphate reduction. Nature 296: 643-645. reduction) in a Nova Scotian Spartina alternifiora stand. Mar.
Kenworthy, W. J., C. Currin, G. Smith & G. Thayer, 1987. The abun- BioI. 47: 227-242.
dance biomass and acetylene reduction activity of bacteria asso- Payne, W. J., 1984. Influence of acetylene on microbial and enzy-
ciated with decomposing rhizomes of two seagrasses, Zostera matic assays. J. Microbiol. Methods. 2: 117-133.
marina and Thala.~sia testudinium. Aquat. Bot. 27: 97-119. Pereg, L. L., Y. Lipkin & N. Sar., 1994. Different niches of the
Kristensen, E., G. M. King, M. Holmer, G. T. Banta, M. H. Jensen, Halophila stipulacea seagrass bed harbor distinct populations of
K. Hansen & N. Bussarawit, 1994. Sulfate reduction, acetate nitrogen fixing bacteria. Mar. BioI. 119: 327-333.
turnover and carbon metabolism in sediments of the Ao Nam Bor Pfennig, N., F. Widdel & H. G. TrUper, 1981. The dissimilatory sul-
mangrove, Phuket, Thailand. Mar. Ecol. Prog. Ser. 109: 245-255. fate reducing bacteria. In M. P. Starr, H. Stolp, H. G. TrUper,
McRoy, C. P. & c. McMillan, 1977. Production, ecology and phys- A. Balows & H. G. Schlegel (eds), The Prokaryotes vol I.
iology of seagrasses. In C. P. McRoy & C. Helfferich (eds). Springer-Verlag, New York. 926-940.
Seagrass ecology. Marcel Dekker, New York. 53-87. Pfennig, N., H. G. Triiper, 1992. The family Chromatiaceae. In
McRoy, C. P., J. J. Goering & B. Chaney, 1973. Nitrogen fixation A. Balows, H. G. Triiper, M. Dworkin & K. H. Schleifer (eds),
associated with seagrasses. Limnol. Oceanogr. 18: 998-1002. The Prokaryotes, vol 4. Springer-Verlag, New York. 3200-3331.
Moriarty, D. J. W. & P. I. Boon, 1989. Interactions of seagrasses Postgate, J. R., 1982. The fundamentals of nitrogen fixation. Cam-
with sediment and water. In A. W. D. Larkum, A. J. McComb & bridge University Press, London.
S. A. Shepherd (eds), Biology of seagrasses. Elsevier, Amster- Postgate, J. R. & H. M. Kent, 1984. Derepression of nitrogen fixation
dam: 500-535. in DesultiJVibrio gigas and its stability to ammonia or oxygen
Moriarty, D. J. W. & M. J. O'Donohue, 1993. Nitrogen fixation in stress in vivo. J. gen. Microbiol. 130: 2824-1.
seagrass communities during summer in the Gulf of Carpentaria, Postgate, J. R., H. M. Kent, S. Hill & H. Blackburn, 1985. Nitrogen
Australia. Aust. J. mar. Freshwat. Res. 44: 117-125. fixation by DesultiJVibrio gigas and other species of Desulti)vib-
Moriarty, D. J. W., P. I. Boon, J. A. Hansen, W. G. Hunt, I. R. Poiner, rio. In P. W. Ludden & J. E. Burris (eds). Nitrogen fixation and
P. C. Pollard, G. W. Skyring & D. C. White, 1985. Microbial C02 metabolism. Elsevier, Amsterdam. 225-234.
biomass and productivity in seagrass beds. Geomicrobiol. J. 4: Postgate, J. R., H. M. Kent & R. L. Robson, 1988. Nitrogen fixation
21-51. by Desulti)vibrio. In J. A. Cole & S. J. Ferguson (eds). The
Moriarty, D. J. w., D. G. Roberts & P. C. Polard, 1990. Primary nitrogen and sulphur cycles. Society of General Microbiology
and bacterial productivity of tropical seagrass communities in Symposium 42. Cambridge University Press, Cambridge. 457-
the GulfofCarpentaria, Australia. Mar. ecol. Prog. Ser. 61: 145- 471.
157. Riederer-Henderson, M. A. & P. W. Wilson, 1970. Nitrogen fixation
Nazina, T. N., E. P. Rozanova & T. A. Kalininskaya, 1979. Fixation by sulphate reducing bacteria. J. gen. Microbiol. 61: 27-31.
of molecular nitrogen by sulfate-reducing bacteria from oil strata. Ryther, J. H. & W. M. Dunstan, 1971. Nitrogen, phosphorus and
Mikrobiologiya 48: 133-136. eutrophication in the coastal marine environment. Science 171:
Nedwell, D. & S. Aziz, 1980. Heterotrophic nitrogen fixation in an 1008-1013.
intertidal salt marsh sediment. Estuar. coast. mar. Sci. 10: 699- Short, F. T., 1983. The response of interstitial ammonium in eelgrass
702. (Zostera marina) beds to environmental perturbations. J. expo
Nedwell, D. B., T. H. Blackburn & W. J. Wiebe, 1994. Dynamic mar. BioI. Ecol. 68: 195-208.
nature of the turnover of organic carbon, nitrogen and sulphur in Smith, R. D., A. M. Pregnall & R. S. Alberte, 1988. Effects of
the sediments of a Jamacaican mangrove forest. Mar. Ecol. Prog. anaerobiosis on root metabolism of Zostera marina (eelgrass):
Ser. 110: 223-231. implications for survival in reducing sediments. Mar. BioI. 98:
Nixon, S. W. & M. E. Q. Pilson, 1983. Nitrogen in estuarine and 131-141.
coastal marine ecosystems. In J. E. Carpenter & D. G. Capone Smith, R. I. & M. J. Klug, 1981. Electron donors utilised by sulfate-
(eds), Nitrogen in the marine environment. Academic Press, New reducing bacteria in eutrophic lake sediments. ApI. envir. Micro-
York. 565-647. bioI. 42: 116---121.
0' Donohue, M. J., D. J. W. Moriarty & I. C. MacRae, 1991a. Stewart, W. D. P., G. P. Fitzgerald & R. H. Burris, 1967. In situ
Nitrogen fixation in sediments and the rhizosphere ofthe seagrass studies on N2 fixation using the acetylene reduction technique.
Zostera capricornia. Microb. Ecol. 22: 53-64. Proc. nat. Acad. Sci. U.S.A. 58: 2071-2078.
174
Taylor, B. F. & R. S. Oremland, 1979. Depletion of adenosine Widdel, F. & F. Bak, 1992. Gram-negative mesophilic sulfate-
triphosphate in DesultiJVibrio by oxyanions of Group IV ele- reducing bacteria. In A. Balows, H. G. Triiper, M. Dworkin,
ments. Curf. Microbiol. 3: 101-103. W. Harder & K. H. Schleifer (eds), The Prokaryotes vol 4.
Voordouw, G., V. Niviere, F. G. Ferris, P. M. Fedorak & Springer-Verlag, New York. 3352-3378.
D. W. S. Westlake, 1990. Distribution of hydrogenase genes in Wolfenden, J. & K. Jones, 1987. Seasonal variation of in situ nitrogen
Desuifovibrio spp. and their use in identification of species from fixation (CzHz reduction) in an expanding marsh of Spartina
the oilfield environment. ApI. envir. Microbiol. 56: 3748-3754. anglica. J. Ecol. 75: 1011-1021.
Welsh, D. T., R. Wellsbury, S. Bourgues, R. de Wit & R. Herbert, Yoch, D. C. & G. J. Whiting, 1986. Evidence for the NHt switch-
1996. Relationship between porewater organic carbon intent, sul- off regulation of nitrogenase activity by bacteria in salt marsh
phate reduction and nitrogen fixation (acetylene reduction) in the sediments and roots of the grass Spartina alternijiora. ApI. envir.
rhizosphere of Zostera noltii. Hydrobiologia 329 (Dev. Hydrobi- Microbiol. 51: 143-149.
01. 117): 169-177. Zuberer, D. A. & w. S. Silver, 1978. Biological dinitrogen fixation
Whiting, G. J., E. L. Gandy & D. C. Yoch, 1986. Tight coupling (acetylene reduction) associated with Florida mangroves. ApI.
of root -associated nitrogen fixation and plant photosynthesis in envir. Microbiol. 35: 567-575.
the salt marsh grass Spartina alternijiora and carbon dioxide Zumft, W. G. & F. Castillo, 1978. Regulatory properties of the
enhancement of nitrogenase activity. ApI. envir. Microbiol. 52: nitrogenase from Rhodopseudomonas palustris. Arch. Microbiol.
108-113. 117: 53-60.
Hydrobi%gia 329: 175-183, 1996. 175
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
David T. Welsh 1,3*, Peter Wellsbury2, Sophie Bourgues3, Rutger de Wit3 & Rodney A. Herbert 1
1 Department of Biological Sciences, University of Dundee, Miller's "'Ynd, Dundee DDI 4HN, Scotland
2 Department of Geology, Wills Memorial Building, University of Bristol, Queen's Road, Bristol BS8 1RJ England
3 Laboratoire d'Oceanographie Biologique et de Biologie Marine, Universite Bordeaux 1, 2 rue Professeur Jolyet,
33120 Arcachon, France
* Present address. Fax: 33-56-8351 04
Key words: Acetylene reduction, nitrogen fixation, sulphate reduction, acetate, seagrasses
Abstract
Depth profiles of nitrogen fixation (acetylene reduction), sulphate reduction, NHt concentration and porewater
volatile fatty acids concentrations were measured in Zostera noltii colonised sediments in the Bassin d' Arcachon,
France in March 1994. Acetylene reduction activity (ARA) was detectable throughout sediment profiles. Addition
of sodium molybdate (20 mmoll- 1) a specific inhibitor of sulphate reduction to slurries inhibited ARA by >75%
inferring that sulphate-reducing bacteria (SRB) were the dominant component of the nitrogen fixing microflora.
The peak of ARA was coincident with that of sulphate reduction and a relatively constant relationship of 40 mole
sulphate reduced per mole acetylene reduced was recorded throughout the profiles. From this ratio it was calculated
that at least 17% of the ATP yield from sulphate reduction would be required to support the measured rates of
nitrogen fixation (acetylene reduction).
Acetate was the dominant constituent of the porewater volatile fatty acids pool, accounting for >90% of the total
pool as measured by HPLC. Concentrations of porewater acetate recorded by HPLC were compared with those
measured using an enzymatic technique and these data indicate that approximately 10% of the total porewater
acetate pool was not available to microbial metabolism. Profiles of porewater acetate concentrations measured by
both techniques were similar to those recorded for both ARA and sulphate reduction and thus acetate oxidation
may fuel these activities.
-+-
noltii Hornem. These seagrass meadows cover an area
greater than 70 km 2 and are the largest in Western
Europe (Auby, 1991). The salinity of the Bassin nor-
mally ranges between 30-33%0, but can decrease to
20%0 following periods of heavy rainfall, whilst the 5 km
water temperature ranges from 4-6 °C in mid-winter
to 21-22 °C in August. A full description of the hydrol- Figure 1. A. Location of the Bassin d' Arcachon on the French
ogy of the Bassin is given by Robert et al. (1987). The Atlantic coast. B. Site diagram of the Bassin d' Arcachon at low
sampling station used during this study was situated tide showing the location of the study site (.) in the central region
of the Bassin and the major mudflats and channels.
in a large seagrass bed on the Ile aux oiseaux, a large
mudflat situated in the central region of the Bassin
(Figure 1) and corresponds to the CLEAN project Sta-
tion A. All the sediment samples were collected on the Nitrogenfixation (acetylene reduction) activity
8th March 1994.
Rates of nitrogen fixation were measured using the
acetylene reduction technique of Stewart et al. (1967).
Intact sediment cores were collected in 25 x 5 cm (i.d.)
177
Plexiglass core tubes and transported to the laborato- Determination of sulphate reduction rates
ry on ice. In the laboratory the cores were careful-
ly extruded, the Z. noltii leaves were removed and Sulphate reduction rates were determined using the
the upper 5 cm of sediment were sliced into 1 cm core injection method of J0rgensen & Fenchel (1974).
thick sections. The sediment slices were immediate- Sediment cores were collected in 20 x 4 cm (i.d.) Plex-
ly transferred into 360 ml volume glass bottles and iglass core tubes fitted with pre-drilled silicon rubber
20 ml of deoxygenated filter-sterilised (0.2 /-lm pore filled injection ports. Cores were injected at 2 cm depth
size) seawater was added. The bottles were sealed with intervals with 10 /-ll of carrier free 35S0~- [370 kbq]
screw tops fitted with rubber-sealed injection ports and (Isotopchim, France). The core tubes were covered in
immediately gassed out with oxygen-free nitrogen for a double layer of aluminium foil to the level of the
10 mins. Following degassing the sediment slurries sediment surface and incubated for 6 hours at in situ
were thoroughly mixed by vigorously shaking the bot- temperature under natural light conditions. Following,
tles. Experiments were initiated by replacing 10% v/v incubation the sediment was extruded and sliced into
of the headspace gas with acetylene and the slurries 1 cm thick sections which were immediately fixed by
were incubated for 24 hours at in situ temperature in the vigorous shaking with an equal volume of 2% w/v zinc
dark. Sodium molybdate was added when required pri- acetate to trap the sulphides and inhibit further activity.
or to degassing to gi ve a final concentration of 20 mmol The production of radiolabelled sulphides was
I-I and the slurries were pre-incubated at the in situ determined by the single step chromium reduction
temperature for 1 hour following degassing prior to the method ofFossing & J0fgensen (1989) and the released
addition of acetylene. Acetate when added was added sulphides trapped in 2% w/v zinc acetate. The specif-
prior to the addition of acetylene by injection of a filter ic activities of 35S2- and 35S0~- were determined in
sterilised stock solution to give a final concentration of aliquot volumes (1 ml) of the zinc acetate traps and the
5 mmoll- I . supernatant (following the initial centrifugation of the
In tests to determine the effect of carbon source sediment respectively) by liquid scintillation counting
additions on acetylene reduction activity a 1: 1 (sed- (Intertechnique SL-3ccm liquid scintillation counter).
iment: filter-sterilised seawater) slurry was prepared Sulphate concentrations were determined as previously
under nitrogen atmosphere from the pooled 0-3 cm described by Tabatabai (1974) and the sulphate reduc-
depth horizons of 8 cores. Aliquot volumes (50 m!) of tion rate calculated according to the method of Fossing
slurry were dispensed into 360 ml glass bottles and the & J0rgensen (1989).
slurries were thereafter treated as described previously.
Carbon sources were added by injection to give a final Determination of porewater and sediment
concentration of 15 mmoll- 1 prior to the addition of exchangeable ammonium concentrations
acetylene.
Following incubation the assays were terminated Sediment cores were collected in 20 x 5 cm (i.d.) Plex-
by the injection of 1 M trichloroacetic acid to give a iglass core tubes and transported to the laboratory on
final concentration of 5% w/v. Aliquot 1 ml volumes ice. Cores for porewater ammonium content were sec-
of the headspace gas were analysed for both acetylene tioned into I cm slices and immediately centrifuged
and ethylene by gas chromatography using a Perkin at 5000 x g for 10 min at 5 °C to extract the sedi-
Elmer Autosystem Gas Chromatograph fitted with a ment porewater. The collected porewater was passed
3 m x 2.2 mm (internal diameter) Chromosorb 101 through a 25 mm Whatman GFIF filter and frozen at
(8011 00 mesh) column with N2 as carrier gas and flame -40°C before analysis. Cores for sediment exchange-
ionisation detection. Flow rates for N2, air and H2 were able ammonium concentration were sectioned into
30, 450 and 50 ml min- i respectively and the oven 1 cm slices and the sediment extracted for 1 hour by
temperature was 30°C. Ethylene concentrations were shaking in an equal volume of I M KCI. The resul-
determined by reference to known standards and all the tant sediment slurry was centrifuged at 5000 x g for
data were corrected for the small quantities of ethylene 10 min at 5°C, the supernatant was decanted, passed
present as a contaminant in the acetylene used in this through a 25 mm Whatman GFIF filter and frozen at
study. -40°C. Thawed samples were suitably diluted with
distilled deionised water and analysed for ammonium
content using a Technicon continuous flow autoanal-
yser (Chemlab Ltd, England).
178
Table 1. Effect of a range of organic carbon sources on
Determination of pore water volatile fatty acid and
the rate of acetylene reduction recorded in slurries of
bioavailable acetate concentrations the 0-3 cm depth horizon of Zostera nolfii colonised
sediments. All carbon sources were present at a final
Sediment cores were collected using 25 x 5 cm (i.d.) concentration of 15 mmoll- I .
Plexiglass core tubes and transported to the laboratory Carbon Acetylene Control
on ice. The cores were extruded and sectioned into 1 cm source reduction rate
slices and the sediment centrifuged at 5000 x g for (mmol ml sed- I h- I )
10 min at 5 °C. The porewater was decanted and passed
No additions 0.786 100
through 25 mm Whatman GFIF filters and frozen at
Ethanol 0.988 126
-40°C.
Pyruvate 1.058 135
Porewater samples were analysed for total
Lactate 0.929 118
volatile fatty acid concentrations using the 2-
Glucose 1.031 131
nitrophenylhydrazine hydrochloride (2 NPH) derivi-
Galactose 0.991 126
tisation method of Parkes et al. (1989) as modified
by Wellsbury and Parkes (1995). Porewater bioavail-
able acetate concentrations were determined using the Acetylene reduction nmol ml sed- 1 h-1
enzymic method of King (1991), using HPLC separa-
tion and quantification of AMP and ATP as described o 0.2 0.4 0.6 0.8
by Wellsbury and Parkes (1995).
0-1
Results
--§.
.-.
Data presented in Figure 2 show profiles of acetylene S
1-2
reduction activity (ARA) in the surficial sediments of
the Z. noltii beds. ARA was detectable at all depth hori- .~
zons tested with peak activities of 0.55 and 0.43 nmol
:.. 2-3
~
ml- I h- I recorded in the 1-2 and 2-3 cm depth hori-
zons respectively. Ethylene accumulation was never -Q-.
<I.l 3-4
observed in control slurries to which no acetylene Q
was added (Data not shown). Additions of 20 mmol
I-I sodium molybdate (final concentration) a specific 4-5
inhibitor of bacterial sulphate reduction to sediment
slurries severely inhibited ARA by 75 to 85% (Fig-
ure 2) throughout the profile inferring that SRB were
Figure 2. Profiles of acetylene reduction activity in the surficial
responsible for the bulk of the recorded activity. Addi- sediments determined in slurry experiments with no additions (open
tion of sodium acetate to sediment slurries to give a bars), supplemented with 5 mmoll- I sodium acetate (hatched bars)
final concentration of 5 mmol I-I had little effect on and supplemented with 20 mmoll- I sodium molybdate (solid bars).
Data represent the means of 5 determinations, standard deviations
ARA with a maximal stimulation of 33% observed in
were less than 5%.
the 1-2 cm depth horizon (Figure 2). Similarly, addi-
tion of a range of carbon sources at a concentration
of 15 mmoll- ' only stimulated ARA between 18 and tion of approximately 40 (range 36 - 48) nmol sul-
35% when added to the slurries (Table 1). phate reduced per nmol acetylene reduced was record-
The depth profile of sulphate reduction activity ed throughout the depth profile with the exception of
was similar to that recorded for acetylene reduction the 0-1 cm depth horizon, where the ratio was 11.8: 1
with maximal activities of 20 and 17.5 nmol sulphate (Table 2).
reduced ml sedimenc I h -I recorded in the 1-2 and Acetate was the predominant volatile fatty acid
2-3 cm depth horizons respectively (Figure 3) and present in the sediment porewaters, representing
85% of total sulphate reduction occurred within the >90% of the total volatile fatty acid pool. Concen-
0-5 cm depth horizon (Data not shown). A relatively trations of total (measured by 2NPH method) and
constant ratio between sulphate and acetylene reduc- bioavailable acetate (enzyme method) were similar
179
Table 2. Relationship between sulphate and acetylene reduction activity with depth in
Zostera noltii colonised sediments.
4-5
NH~ concentration jlIDOI. rl not surprising that sulphate reduction rates in this zone
would be underestimated.
o 100 200 300 It has been estimated that the net ATP yield for
Desulfobacter postgatei growing on acetate and sul-
. I
__ i J - l - . phate is 0.8 mol ATP per mol sulphate reduced (See
Widdel & Hansen, 1991). Thus, using the recorded
~1-2~ ratio of 40: 1 between sulphate reduction and acety-
lene reduction and the theoretical ratio of 3: 1 (see
.S~2-3~
~_._ Welsh et aI., 1996) to convert acetylene reduction rates
to nitrogen fixation rates, the energy requirement for
~3-4~ this process in the Zostera noltii colonised sediments
"4-5~
can be determined. Postgate, (1982) calculated that
16 mol ATP are required to fix 1 mol of N2. On this
basis approximately 17% of the ATP generated by sul-
Figure 5. Depth profiles of pore water NHt (open bars) and sediment phate reduction would be required to fuel the record-
exchangeable NHt (hatched bars) concentrations. ed nitrogen fixation (acetylene reduction) rates in the
rhizosphere. This figure, however, represents a mini-
mum value since, the theoretical energy requirement
for nitrogen fixation does not take into account addi-
reduction (Taylor & Oremland, 1979; Smith & Klug, tional costs such as the synthesis of nitrogenase and
1981; Oremland & Capone, 1988). These data indicate the electron carriers. Reported ATP:N 2 ratios for pure
that SRB were responsible for the bulk of the recorded cultures vary considerably e.g. 30 in Klebsiella pneu-
ARA. The ability to fix molecular nitrogen has long moniae (Hill, 1976) and 42 for Rhizobium ORS 571
been recognised amongst SRB (Sisler & Zobell, 1951; (Starn et aI., 1984) but are generally much greater than
Le Gall et al.,1959; Postgate et aI., 1988) and these 16. Thus, it is probable that the true energy demand
bacteria have previously been proposed as potentially of nitrogen fixation in these sediments is substantially
the most important heterotrophic fixers of nitrogen in greater than 17% of the total ATP produced by sulphate
coastal marine sediments (Herbert, 1975; Nedwell & reduction.
Aziz, 1980). The addition of exogenous carbon sources to slur-
Depth profiles of sulphate reduction (Figure 3) ries had little effect on ARA of Z. noltii colonised
showed a similar overall pattern to those recorded for sediments (Figure 2, Table 1). These results are at vari-
ARA and with the exception of the 0-1 cm depth hori- ance with other studies of heterotrophic nitrogen fix-
zon, a relatively constant molar ratio of approximately ation in marine sediments, where carbon source addi-
40: 1 was recorded between ARA and sulphate reduc- tions were found to significantly stimulate nitrogen
tion rates at all depths (Table 2). These data provide fixation/acetylene reduction rates (e.g. Herbert, 1975;
further evidence to support the hypothesis that SRB Nedwell & Aziz, 1980). However, this may be due
are principally responsible for the observed acetylene to an inherent artifact of the slurry technique in veg-
reduction activity (nitrogen fixation) in the Zostera root etated sediments due to the release of organic carbon
zone. Deviation from the 40: 1 ratio in the 0-1 cm depth from the plant roots during the preparation of the slur-
horizon is probably due to an underestimation of the ries. Internal pools of labile carbon in the roots of
sulphate reduction rate in this zone due to the reoxi- Z. noltii represent a large pool of potentially avail-
dation of reduced 35 S by molecular oxygen during the able organic carbon within the sediment, since it has
6 hour incubation period. J0rgensen (1994) recently been demonstrated for Zostera marina L. that under
reported an 8-fold decrease in the calculated sulphate conditions of oxygen limitation the roots switch to a
reduction rates of oxic microbial mat sediments incu- fermentative metabolism producing CO 2, ethanol and
bated for 30 min and 3 hours, which he attributed to the lactate as metabolic end products, the latter accumulat-
reoxidation of radiolabelled sulphides. Oxygen pene- ing in the root tissue (Smith et aI., 1988). Such internal
tration in Z. noltii colonised sediments in the Bassin pools of labile carbon may then be released during the
d' Arcachon is spatially highly variable due to the pres- preparation of sediment slurries, since, previous stud-
ence of plant roots but is typically limited to a depth ies have shown that physical disturbance of salt marsh
of 6-13 mm (de Wit, unpublished data). It is therefore sediments can substantially alter the carbon composi-
181
tion of the sediment porewater (Hines et ai., 1994). the principal substrate for sulphate reduction in marine
Thus, organic carbon may be more readily available sediments (S0rensen et ai., 1981; Parkes et ai., 1989)
in slurries of Z. noltii colonised sediments due to root and to support nitrogen fixation in pure cultures of SRB
damage and perhaps it is therefore not surprising that (Widdel, 1987). This relationship is however difficult
carbon source additions result in only a minor stimu- to analyse, since depth profiles of acetate also correlate
lation of ARA. This explanation is supported by the with the distribution of the Z. noltii roots and thus may
observation that acetylene reduction rates recorded at have been influenced by the release of organic carbon
the same time by a whole core technique were substan- from the plant roots during porewater extractions as
tially lower than those recorded in slurries, indicating discussed previously. Additionally, it cannot be com-
activity was limited by carbon source availability in pletely excluded that the recorded porewater pools of
situ (Welsh et ai., 1996). acetate are the result of incomplete lactate oxidation.
Porewater acetate concentrations were determined However, this seems unlikely, since, in molybdate inhi-
by both HPLC following 2-NPH derivitisation and bition studies of sediments from the same sampling sta-
enzyme assay methods. The HPLC method mea- tion, where, carbon sources normally consumed during
sures the total acetate pool and involves the effi- sulphate reduction would accumulate, acetate was the
cient extraction and derivitisation of both bioavailable major VFA which accumulated and lactate was nev-
and non-bioavailable pools. In contrast the enzyme er detected (Finster et ai., 1994). These authors also
assay method potentially only measures the bioavail- calculated that VFA oxidation rates, based on accumu-
able acetate pool, which is converted to acetyl CoA lation rates during molybdate inhibition and assuming
by acetyl CoA synthetase according to the following steady state conditions could account for approximate-
equation: ly 90% of the recorded sulphate reduction rate at this
acetyl CoA synthetase station, indicating that other potential carbon and ener-
acetate + ATP + CoA --+ acetyl CoA + AMP + PP i
gy sources do not significantly contribute to sulphate
reduction and thus to nitrogen fixation by SRB.
This reaction represents the first step in the uptake Nitrogen fixation/acetylene reduction rates are like-
and metabolism of acetate by SRB (Widdel & Hansen, ly to be influenced by the availability of NHt, since,
1991) and methanogens (Whitman et ai., 1991) and nitrogenase is subject to feed-back inhibition by this
is thus thought to yield a more accurate estimate of source of fixed nitrogen (Postgate, 1982; Dixon, 1984;
the bioavailable acetate pool in sediment porewaters Capone, 1988). Porewater NHt concentrations in
(King, 1991; Kristensen et ai., 1994). However, the the Z. noltii beds ranged from 65 to 106 fLM (Fig-
recent study by Wellsbury & Parkes (1995), indi- ure 5), concentrations which would totally suppress
cates that this method may also overestimate the true nitrogenase activity of pure cultures growing in liquid
bioavailability of acetate in marine sediment porewa- media, due to both repression of nitrogenase synthe-
ters. sis and reversable 'switch-off' of preformed nitroge-
Depth profiles of porewater acetate concentrations nase (Postgate & Kent, 1984; Postgate et ai., 1985).
measured by both techniques were similar (Figure 4) However, a number of previous studies have reported
with maximal concentrations measured in the 1-2 and nitrogen fixation/acetylene reduction activity in marine
2-3 cm depth horizons. However, concentrations mea- sediments at several-fold higher porewater NHt con-
sured by the enzyme assay technique were general- centrations (see Capone, 1988 for review). Additional-
ly lower than those measured by HPLC, indicating ly, in these sediments, zones oflow NHt concentration
that approximately 10% of the porewater acetate pool may exist around the plant roots due to NHt assimila-
measured using the HPLC (2-NPH) technique was not tion by the plants and such microenvironments would
available to the indigenous microflora. Propionate was allow nitrogen fixation to continue due to the absence
the only other volatile fatty acid present in measurable of feedback inhibition.
concentrations but was present at much lower concen- In conclusion data presented in this paper indicate
tration than acetate which represented >90% of the that an association exists between nitrogen fixation
total volatile fatty acid pool (Figure 5). The depth pro- (acetylene reduction) rates and SRB in the rhizosphere
files of porewater acetate pools were similar to those of Z. noltii and that acetate may represent the princi-
recorded for ARA and sulphate reduction activity (Fig- pal carbon source fuelling nitrogen fixation by these
ures 3 & 4). Acetate could fuel these activities, since bacteria.
this volatile fatty acid has been demonstrated to be
182
rio. In W. Ludden & 1. E. Burris (eds), Nitrogen fixation and C02 Proc. natn. Acad. Sci. U.S.A. 58: 2071-2078.
metabolism. Elsevier Science Publishing, Amsterdam. 225-234. Tabatabai, M. A., 1974 Determination of SO~- in water samples.
Postgate, J. R., H. M. Kent & R. L. Robson, 1988. Nitrogen fixation Sulphur Inst. J. 10: 11-14.
by Desu!fiJVibrio. In J. A. Cole & S. J. Ferguson (eds). The Taylor, B. F & R. S. Oremland, 1979. Depletion of adenosine
nitrogen and sulphur cycles. Society for General Microbiology triphosphate in Desult(JVibrio by oxyanions of Group IV ele-
Symposium 42. Cambridge University Press, Cambridge: 457- ments. Curl'. Microbiol. 3: 101-103.
471. Wellsbury, P. & R. J. Parkes, 1995. Acetate turnover and bioavailabil-
Robelt, R., N. Guillocheau & Y. Collos, 1987. Hydrological param- ity in an estuarine sediment. FEMS Microbiol. Ecol. 17: 85-94.
eters during an annual cycle in the Arcachon Bassin. Mar. BioI. Welsh, D. T., S. Bourgues, R. de Wit & R. A. Herbert, 1996. Seasonal
95: 631-640. variation in rates of heterotrophic nitrogen fixation (acetylene
Short, F T., 1983. The response of interstitial ammonium in eelgrass reduction) in Zostera no/tii meadows and uncolonised sediments
(Zostera marina) beds to environmental pertubations. J. expo mar. of the Bassin d' Arcachon, South-west France. Hydrobiologia 329
BioI. Ecol. 68: 195-208. (Dev. Hydrobiol. 117): 161-174.
Sisler, F D. & c. E. Zobell, 1951. Nitrogen fixation by sulphate- Whiting, G. J., E. L. Gandy & D. C. Yoch, 1986. Tight coupling
reducing bacteria indicated by nitrogen/argon ratios. Science 262: of root -associated nitrogen fixation and plant photosynthesis in
209-210. the salt marsh grass Spartina alternijlora and carbon dioxide
Starn, H., H. W. van Versveld, W. de Vries & A. H. Stouthamer, enhancement of nitrogenase activity. Appl. en vir. Microbiol. 52:
1984. Hydrogen oxidation and efficiency of nitrogen fixation 108-113.
in succinate-limited chemostat cultures of Rhizobium ORS 571 Whitman, W. B., T. L. Bowen & D. R. Boone, 1991. The
Arch. Microbiol. 139: 53-60 methanogenic bacteria. In A. Balows, H. G. Triiper, M. Dworkin,
Smith, R. D., A. M. Pregnall & R. S. Alberte, 1988. Effects of W. Harder & K. H. Schleifer (eds), The Prokaryotes. Springer-
anaerobiosis on root metabolism of Zostera marina (eelgrass): Verlag, New York: 719-767.
implications for survival in reducing sediments. Mar. BioI. 98: Widdel, F, 1987. New types of acetate-oxidising, sulfate-reducing
131-141. Desult()bacter species, D. hydrogenophilus sp. nov., D. latus sp.
Smith, R. 1. & M. 1. Klug, 1981. Electron donors utilised by sulfate- nov., and D. curvatus sp. nov. Arch. Microbiol.
reducing bacteria in eutrophic lake sediments. Appl. envir. Micro- Widdel, F & T. A. Hansen, 1991. The dissimilatory sulfate-
bioI. 42: 116-121. and sulfur-reducing bacteria. In A. Balows, H. G. Triiper,
S~rensen, J., D. Christensen & B. B. J~rgensen, 1981. Volatile fatty M. Dworkin, W. Harder & K. H. Schleifer (eds), The Prokaryotes.
acids and hydrogen as substrates for sulfate-reducing bacteria in Springer-Verlag, New York: 583-625.
anaerobic marine sediment. Appl. envir. Microbiol. 42: 5-11. Wolfden, J. & K. Jones, 1987. Seasonal variation of in situ nitrogen
Stewart, W. D. P., G. P. Fitzgerald & R. H. Burris, 1967. In situ fixation (C 2H2 reduction) in an expanding marsh of Spartina
studies on N2 fixation using the acetylene reduction technique. anglica. J. Ecol. 75: 1011-1021.
Hydrobi%gia 329: 185-198,1996. 185
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Lucas J. Stal, Simone B. Behrens, Marlies Villbrandt, Stef van Bergeijk & Finn Kruyning
Laboratory for Microbiology, ARISE, University ofAmsterdam, Nieuwe Achtergracht 127,
NL-IOI8 WS Amsterdam, The Netherlands
Key words: Coastal lagoon, cyanobacteria, iron, phosphate, sulfide, nitrogen fixation
Abstract
This paper summarizes the results of a study on the biogeochemistry of two eutrophic marine lagoons. The
lagoons investigated were the Bassin d' Arcachon situated on the Atlantic coast, and Etang du Prevost on the
Mediterranean coast. The sites chosen for this study were characterized by the presence of dense communities of
microphytobenthos. Both lagoons receive a large input of nutrients but they differ in several aspects. The Bassin
d' Arcachon receives a large amount of iron. Iron is of great importance in reducing the effects of eutrophication.
Ferric iron is an efficient scavenger of phosphate and it has been proposed that this is one of the mechanisms that
controls primary productivity and algal growth in this lagoon. The mechanisms of phosphate mobilization were
studied by using sediment slurries. These experiments demonstrated that not only ferric iron but presumably also
calcium was responsible for phosphate binding. Another effect of the high iron content in the Bassin d' Arcachon
was the precipitation of sulfide as iron sulfide or pyrite. In the Etang du Prevost sulfate reduction resulted in the
accumulation of free sulfide. The relative low content of iron in Etang du Prevost not only allowed the formation
of free sulfide but may also have limited the binding capacity of phosphate in the sediment. On the other hand
sulfate reduction was not important for the release of phosphate from the sediment. In Etang du Prevost primary
productivity is nitrogen rather than phosphorus limited. In contrast in the Bassin d' Arcachon primary productivity
was presumably mostly phosphate limited. In Etang du Prevost the non-heterocystous cyanobacterium Oscillatoria
sp. was the dominant nitrogen-fixing species. Heterocystous species were excluded from this lagoon as a result
of the presence of free sulfide. It was demonstrated that heterocystous cyanobacteria are more sensitive towards
sulfide than non-heterocystous species. The absence of free sulfide explained the presence of the heterocystous
cyanobacterium Anabaena sp. in Bassin d' Arcachon.
planktonic cyanobacteria where the turbidity of the Dynamics of iron content of the sediments of
water is such that green algae can not proliferate. Bassin d' Arcachon and Etang du Prevost
In this investigation two coastal lagoons in France
were compared. The Etang du Prevost is a high- In Table 1 the iron content of the sediments of the sta-
ly eutrophic lagoon situated on the Mediterranean tions C (Bassin d' Arcachon) and 12 (Etang du Prevost)
coast, close to the city of Montpellier, whilst the are shown. Both stations were sampled in Septem-
Bassin d' Arcachon is situated on the Atlantic coast, ber 1993 and in May 1994. Two methods were used
60 km south-west of the city of Bordeaux. Whereas to assay iron. Total reactive iron was extracted with
the Mediterranean lagoon may experience dystrophic 10 M HCl and assayed with phenanthroline. Bioavail-
crisis, this phenomenon has never been recorded in the able iron was extracted from the sediment with 0.5 M
Bassin d' Arcachon. The processes that are involved in HCl and assayed using the ferrozine reagent (Lovley
causing dystrophy were investigated by studying one & Phillips, 1987a). The iron content was highest in
comparable site in each of these lagoons as a model sys- Station C of Bassin d' Arcachon. This is in accordance
tem. Shallow areas in both lagoons are characterized with the observation that the input of iron in the Bassin
by the occurrence of microphytobenthic communities. d' Arcachon via river discharge is much higher than
The role of iron in these sediments was investigat- in the Etang d Prevost (P. J. Labourg, pers. comm.).
ed. Iron reacts with sulfide to produce the virtually In September 1993 the iron content was particular-
insoluble precipitate FeS. It was postulated that iron ly high whereas in May 1994 the levels were much
might function as an efficient scavenger of sulfide what lower. This was the case in both stations but partic-
would prevent it from reaching toxic concentrations. ularly evident in Station C. The September:May iron
In contrast, oxygen production by the microphytoben- ratios in Station C were 3.2 and 3.1 for total reac-
thos may result in iron oxidation. Ferric iron is known tive iron and bioavailable iron, respectively. For Sta-
to form an insoluble complex with phosphate. Thus, tion 12 these ratios were 1.5 and 1.1. It is conceived that
iron may also playa role in regulating the availability iron accumulated during the spring and summer was
of phosphate for algal growth. mobilized during the autumn and winter. The water
Nitrogen fixation may be important in providing of Bassin d' Arcachon is subject to frequent exchange
an additional source of nitrogen. Among diazotroph- with the Atlantic Ocean because of the tides and the
ic organisms cyanobacteria are particularly important relatively large entrance of the lagoon to the open sea.
(Fay, 1992). As oxygenic photoautotrophic organ- In contrast the Etang du Prevost has only a narrow
isms they can provide the enzyme nitrogenase with connection with the Mediterranean Sea and the tidal
sufficient energy and reducing equivalents. In both range is small (~20 cm). Therefore, the water in Etang
lagoons benthic communities of cyanobacteria were du Prevost has a much longer residence time than in
present (microbial mats) in which nitrogen fixation Bassin d' Arcachon. This explains the smaller differ-
was important. In the Bassin d' Arcachon heterocys- ences in iron content between spring and summer in
tous cyanobacteria appeared to be the predominant the Etang du Prevost.
diazotrophs, whereas non-heterocystous species were The mechanisms of iron immobilization and mobi-
found in Etang du Prevost. Nitrogen fixation is particu- lization are not precisely known. The bulk of iron is
larly sensitive to oxygen and the oxygenic phototrophic present as ferrous sulfide which is virtually insoluble.
cyanobacteria have evolved special adaptations (Gal- Considerable iron is fixed as pyrite. Pyrite is a very
lon, 1992). Heterocystous cyanobacteria develop spe- stable compound and pyrite iron is not included in
cial cells, the heterocysts, that have lost the capacity the iron measurements presented in Table 1. Oxida-
for oxygenic photosynthesis and in which the nitrogen tion of ferrous iron results in the formation of insol-
fixation takes place (Haselkorn, 1978). Such cyanobac- uble ferric compounds (Phillips et aI., 1993). Benthic
teria can be considered to be the best adapted for dia- microorganisms such as cyanobacteria may also bind
zotrophic growth (Stal et aI., 1994). It has been shown and accumulate iron (Sta\' 1994). Thus, immobiliza-
that sulfide is important for the selection of cyanobac- tion of iron may be associated with microbial activities
teria (Howsley & Pearson, 1979). such as (i) sulfate- and sulfur reduction which pro-
duce the sulfide that will precipitate as FeS, (ii) oxy-
genic photosynthesis by microphytobenthos that pro-
duce the oxygen that will oxidize iron and form insol-
uble iron hydroxides either chemically or biologically,
187
Table 1. Summary of the results of iron determinations in the upper 5 em of the sediment of
Station C, the Bassin d' Areachon and Station 12, Etang du Prevost. Samples were collected in
September 1993 and in May 1994. Two methods of iron determination were used. Total reactive
iron was extracted with 10 M HCI and assayed with phenanthrolinechloride. Bioavailable iron
was extracted in 0.5 M HCI and assayed with ferrozine reagent. Iron is expressed either as
Jlmol cm- 3 or as molar ratios.
Table 2. Ferrous and ferric iron in the upper 5 em of the sediments of the sampling stations C
(Bassin d' Arcachon) and 12 (Etang du Prevost) in September 1993 and May 1994. Two methods
of iron determination were used. Total reactive iron was extracted with 10M HCI and assayed with
phenanthrolinechloride. Bioavailable iron was extracted in 0.5 M HCI and assayed with ferrozine
reagent.
(iii) binding to extracellular polysaccharide sheaths of May, respectively). In Station 12 only half of the total
benthic microorganisms (e.g. cyanobacteria) (Decho, reactive iron was determined as bioavailable (ratio 2.2
1990). Mobilization and liberation of the iron from and 1.5 in September and May, respectively). The rel-
the sediments is therefore possible when the activities atively low content of bioavailable iron in Station 12
described above cease. When acid volatile sulfide is in September coincided with a high content of ferric
subsequently oxidized (chemically and/or biological- iron, many forms of which are not readily available for
ly) iron will be liberated. organisms (Phillips et ai., 1993).
In September 1993 the sediment at Station C Data presented in Table 2 summarizes the amounts
contained almost three times (ratio Station C: Sta- of ferrous and ferric iron as well as their ratios at the two
tion 12=2.8) as much total reactive iron on a volume sampling stations in September 1993 and May 1994. It
basis than Station 12. In May 1994 this difference was is notable that the two methods applied give totally dif-
much smaller (ratio Station C: Station 12= 1.3). Anoth- ferent results with regard to the differentiation between
er remarkable difference between the two stations was ferrous and ferric iron. These data indicate that the
the ratio of total reactive iron and bioavailable iron. methods used cannot readily compared when ferrous
The ratio of bioavailable iron Station C: Station 12 and ferric iron are considered separately. Because it is
was 5.2 and 1.8 in September 1993 and May 1994, more likely that the bulk of the iron is in the reduced
respectively. In Station C virtually all iron was present form, only the ferrozine method will be considered
as bioavailable iron (ratio 1.2 and 1.1 in September and here since this clearly yielded the highest amounts
188
Table 3. Ratios of ferric:ferrous iron in sediment slurries from the Bassin
d' Arcachon (Station C) and i3tang du Prevost (Station 12) with and without
the addition of 25 mM Fe3+. The slurries were incubated in the dark for
up to 90 h. Experiments were carried out in May 1994.
0 24 48 90
ratio ferric:ferrous iron (Fe 3+ :Fe2+)
Table 4. The amounts of acid volatile sulfide (AVS) and pyrite in the upper 5 cm of the sediments
of the sampling stations C (Bassin d' Arcachon) and 12 (i3tang du Prevost). compared with the
iron content (amounts in jLmol cm- 3 ).
of ferrous iron. Station 12 shows the highest relative (Figure 1b). The iron content in May 1994 was lower
amounts of ferric iron (lowest ratios Fe2+ :Fe3+). The than in September 1993, particularly in the top lay-
sediment of Station 12 was covered by a dense layer er of the sediment. As stated previously, virtually all
(mat) of cyanobacteria. The oxygen produced by this iron was in the ferrous form. Ferrous iron (when not
community may have been responsible for the oxida- precipitated as metal sulfide) is very soluble and this
tion of the iron. Because the cyanobacterial mat was might explain the mobilization and dissolution of iron
most active in spring, the May ratio Fe2+ :Fe3+ was from the sediment, which would be most pronounced
lower than in September when cyanobacterial activity in the surface layer. The situation was different in Sta-
was low, presumably because evaporation resulted in tion 12 (Figure 1c, d). The iron profile in September
hypersaline conditions. In Station C the situation was 1993 showed a sharp increase with depth (Figure 1d).
totally different. The relative contribution of ferrous to In May highest concentrations were found in the top
total iron was much larger than in Station 12. In Sta- layers but the gradient was less clearly defined (Fig-
tion C the microphytobenthic community (cyanobacte- ure Ic). Although the highest iron concentrations were
ria and diatoms) developed during summer and photo- measured in September 1993, the depth profile of iron
synthetic oxygen production may therefore result in a at Station 12 indicated that it was in the process of
higher ferric iron content in September compared with being mobilized and dissolved. This was not only evi-
May. However, the highly reduced conditions in this dent from the gradient but also from the greater pro-
station ensured that ferrous iron was quantitatively the portion of ferrous iron. In contrast, in May iron was
most important species present both in September and apparently being immobilized. This is evident from the
in May. gradient and from the lower proportion of ferrous iron.
Depth profiles of iron revealed similar patterns Moreover, the immobilization of iron was also coinci-
regardless whether total reactive iron or bioavailable dent with the presence of an active cyanobacterial mat
iron was considered. The profile of iron (FeH + Fe3+) at this station during the spring. Oxygen produced by
in Station C showed an increase in content with depth the cyanobacteria may precipitate ferric iron and these
(Figure la, b). This gradient was more pronounced in organisms have also been shown to accumulate iron in
May 1994 (Figure la) compared with September 1993 their extracellular polysaccharide sheath (Stal, 1994).
189
4 A 4 B
E E
- -
.§. 10 §.
.c .c
Q.
Q) 25 Q.
Q) 25
0 0
50 50
0 30 60 90 120150180 0 30 60 90 120150180
Fe (Ilmol cm·3 ) Fe (Ilmol cm·3)
4 C 4 0
E E
-
§. 10
-
§. 10
.c .c
Q.
Q)
Q.
Q)
25
0 0
50 50
0 30 60 90 120150180 0 30 60 90 120150180
Fe (Ilmol cm·3) Fe (Ilmol cm"3)
Figure 1. Depth profiles of total (FeH + Fe3 +) reactive iron in the sediments of the Ba~sin d' Arcachon (Station C) (a,b) and Etang du Prevost
(Station 12) (c,d) in May 1994 (a,c) and September 1993 (b,d).
Experiments in which sediment slurries were used iron reduction, it was clearly less dramatic than at Sta-
demonstrated that ferric iron was rapidly reduced. The tion C. Sediment slurries were also amended with glu-
results of these experiments are summarized in Table 3. cose (20 mM), molybdate (20 mM) and a combination
The sediment slurries were incubated in the dark and of glucose and molybdate. By increasing the organic
the ratio of ferric:ferrous iron measured immediately matter content the reducing potential should increase
after the addition of 25 mM Fe3+ and after 1, 24, 48 (Lovley & Phillips, 1986). The addition of molybdate
and 90 hours. Sediment slurries without added iron inhibits sulfate reduction (Oremland & Capone, 1988)
were used as controls. In the control slurries, most of and it was expected that sulfide produced by sulfate
the iron was in the reduced state and very little change reducing bacteria would contribute significantly to iron
in the ferric:ferrous iron ratio was recorded in sediment reduction (see equations p. 11). However, the results
samples from the Etang du Prevost (Station 12) as well show that the iron-reducing potential of the sediment
as from Bassin d' Arcachon (Station C). The addition slurry was already maximal and could not be signif-
of 25 mM Fe3+ resulted in a substantial increase in icantly increased by the addition of glucose. Equally,
the ferric:ferrous iron ratio. At Station C this increase inhibition of sulfate reduction did not decrease this
was less dramatic since the iron content at this site potential (Figure 2). These data indicate that processes
was much higher than at Station 12 (Table 1). At Sta- other than sulfate reduction were responsible for iron
tion C ferric iron was already reduced after 1 h incuba- reduction. Ferric iron may serve as terminal electron
tion, after which the ratio ferric:ferrous iron returned to acceptor in the an<\erobic degradation of organic mat-
its original value. These results demonstrate the enor- ter (Lovley, 1987, 1991). Clearly, the organic matter
mous potential for iron reduction at this site. Sediment content of the sediments was sufficient to sustain these
slurries from Station 12 behaved slightly differently. processes since the addition of glucose had no effect.
Due to the lower iron content at this site the addition
of 25 mM ferric iron resulted in a much higher fer-
ric:ferrous iron ratio. This ratio increased somewhat Acid volatile sulfide in the sediments of Bassin
after 1 h incubation, probably due to oxidation by d' Arcachon and Etang du Prevost
oxygen introduced in the slurry during preparation.
Thereafter the ferric:ferrous iron ratio decreased but The concentrations of acid volatile sulfide (AVS) in
remained above the original value. Although it was the sediments of the Stations C (Bassin d' Arcachon)
evident that this site also possessed a high potential for and 12 (Etang du Prevost) were highest in September.
190
Table 7. Effects of sulfide on nitrogenase activity in a non-
3 heterocystous cyanobacterium (Oscillatoria sp.) and a heterocystous
... species (Anabaena sp.). The control (0 mM) was set to 100. Experi-
'"U.aI ments were performed at a pH of 8.
.'!; 2
/' Sulfide concentration (mM) Oscillatoria sp. Anabaena sp.
\
aI
U.
0 0 100 100
;::
ra
::1
0.5 154 46
'0
ra \ \,~ ......... .
1
5
104
94
27
13
E ~.. _ ._ _ • ...a..
.--=-.. :-- .... ...- 10 118 nd
0
0 1 24 48 90
Time (h)
Figure 2. Ratio ferric (Fe 3+) to ferrous (FeH ) iron after the addition
of 25 mM of Fe3+ at t =0 to a sediment slurry from the Etang du sulfur reducing microorganisms. These processes are
Prevost (Station 12). Slurries were incubated for up to 90 h in the only of importance after sufficient organic matter has
dark at ambient temperature. - . - : no further additions; -e- been produced, i.e. in summer or autumn, depending
: glucose added; -6-: molybdate added; -0-: glucose and
molybdate added.
on the geographical location. It is evident that during
winter and spring most of the sulfide in the sediment
Table 5. Comparison of total extractable phos- is oxidized due to either biotic or abiotic processes. In
phate and ferric iron (bio-available) in the upper September 1993 the sulfide concentrations were slight-
5 cm of the sediments of the sampling stations C 1y higher in Station C as compared to Station 12, but in
(Bassin d' Arcachon) and 12 (Etang du Prevost).
both stations acid volatile sulfide concentrations were
Amounts represent J1mol cm - 3 of sediment. Sam-
ples were collected in September 1993 and May very high. If these concentrations are compared with
1994. the total iron content of these sampling stations it was
evident that at Station C iron was present in consid-
Month Station Fe3+ PO!-
erable excess and therefore any sulfide produced at
September Arcachon C 11.0 0.7 this station would be immediately precipitated as FeS.
Prevost 12 4.3 4.1 Free sulfide (HS-) would not be expected to be present
May Arcachon C 1.2 1.2 at this site. The situation was different in Station 12
Prevost 12 5.7 1.2 where in September the concentration of total reac-
tive iron was only slightly higher than acid volatile
sulfide (Table 4). Moreover, when bioavailable iron
was considered acid volatile sulfide was in excess.
Acid volatile sulfide in the top 5 cm of Station was
Under these conditions free sulfide was expected to be
47 and 8 /Lmol cm- 3 in September and May, respec-
present. Pyrite was also present at both sampling sites
tively. In Station 12 these numbers were respectively
but was quantitatively unimportant. The dynamics of
35 and 1 (Table 4). Sulfide is produced by sulfate- and
pyrite followed that of acid volatile sulfide and iron,
i.e. it was also oxidized and leached during winter.
Table 6. Sediment slurries collected from the Bassin The profiles of acid volatile sulfide in the sediment
d' Arcachon (Station C) and Etang du Prevost (Sta- of the Stations C and 12 are shown in Figure 3. In Sta-
tion 12) amended with 100 J.LM phosphate. The tion 12 in May sulfide was apparently produced mainly
effect of iron was measured by adding 25 mM Fe3+ .
Incubation was for I h. Experiments were done in
at a depth of 4-8 mm (Figure Ic), while in September
May 1994. this had moved down to 12-25 mm (Figure ld). In Sta-
tion C a different pattern was recorded. In May AVS
Treatment Station C Station 12 concentrations increased with depth which suggests
J1M phosphate in water that sulfide oxidation was occurring at the sediment
surface (Figure la). In September virtually no gradient
no iron addition 10 24 was present and AVS was equally high at all depths
25 mM Fe3+ added 0.3 4.2 except the surface layer (Figure Ib). In the surface
layer presumably little sulfide was produced.
191
4~ A B
E E
g 10 g
....
.c
a.
....a.
.c
G)
25 G)
0 o
50
0 20 40 60 80 100 0 20 40 60 80 100
AVS (f.1mol cm"3) AVS (f.1mol cm"3)
E 4 c E 4 ~ D
g 10 g 10 ~
....a.
.c ....
.c
a. 25
G)
25 G)
0 0
50 50
0 20 40 60 80 100 0 20 40 60 80 100
AVS (f.1m ol cm"3) AVS (f.1mo l cm"3)
Figure 3. Depth profiles of acid volatile sulfide (AYS) in the sediments of the Bassin d' Arcachon (Station C) (a,b) and Etang du Prevost
(Station 12) (c,d) in May 1994 (a,c) and September 1993 (b,d)
Profiles of pyrite are shown in Figure 4. In Sta- 2FeH +S2- -+ 2Fe2+ +So
tion 12 in May highest concentrations of pyrite were 2Fe2+ +2S2- -+ 2FeS
found in the top layer indicating that pyrite formation FeS+So -+ FeS2
was occurring in this layer. In May Station 12 also had
a relatively high ferric iron content and it is probable
that this was important either directly or indirectly in
pyrite formation (see equation). Thus sulfide produc-
tion and pyrite formation appear to proceed in different Dynamics of phosphate in the sediments of Bassin
horizons in the sediment. In September sediment con- d' Arcachon and Etang du Prevost
centrations of pyrite was much higher but the depth
profile was reversed, indicating a possible oxidation In Table 5 the concentration of total extractable phos-
of pyrite at the surface (Figure 4d). At Station C the phate in the sediments of the sampling stations of
sediment content of pyrite was totally different. The Bassin d' Arcachon and Etang du Prevost are shown.
pyrite concentrations were higher at Station C com- At Station C highest phosphate concentrations were
pared with Station 12. Depth profiles of pyrite showed detected in May, while at Station 12 highest values
highest concentrations in the deeper lay~rs of the sedi- were recorded in September. These differences may
ment both in September and May. This suggested that be attributed to differences in growth season. In the
pyrite was formed in the deeper layers of the sedi- Etang du Prevost the cyanobacterial mats were ful-
ment, concurrent with sulfide production. The large ly developed in spring and early summer, while in
excess of ferrous iron probably prevented the forma- the Bassin d' Arcachon they did not appear until late
tion of elemental sulfur and thus pyrite formation, as summer. The active growth of the microphytobenthos
was probably the case at Station 12. Under these con- may also result in the lowering of phosphate concen-
ditions pyrite may have been produced here as a direct tration. In addition there are significant differences
consequence of sulfate reduction (Howarth & Merkel, in the phosphate load that the two lagoons receive.
1984). Highest phosphate concentrations were measured in
the sediments of Etang du Prevost. This lagoon clearly
received a much higher phosphate load. The phosphate
that was extracted from the sediments might not have
192
A 4 B
E 4
E
g 10 g
...
.c
Q. 25
...
.c 10 "
Q.
CD CD 25 "
c c 50 ~
50
0 2 4 6 8 10 12 0 2 4 6 8 10 12
E 4 c E 4 D
g 10 g 10
...g.
.c
25
...
.c
Q. 25
CD
c
C 50~~=F~~====~ 50
o 2 4 6 8 10 12 0 2 4 6 8 10 12
been available for growth of microphytobenthos. Phos- of the phosphate was immobilized at stations C and 12,
phate forms a complex with ferric iron (Ehrlich, 1990, respectively (Table 6).
Danenlouwerse et aI., 1993). In Table 5 the phosphate Because phosphate concentrations at both sites in
concentrations are compared with the concentrations May 1994 were comparable (1.2/Lmol cm -3, Table 5),
of ferric iron in the sediments. It is clearly evident that differences in phosphate immobilization between the
in all cases ferric iron equalled or exceeded the phos- stations C and 12 cannot be explained in terms of sed-
phate concentrations. At Station C in May phosphate iment phosphate content. Although total iron content
and iron concentrations were equal which suggests that in Station C was greater than at Station 12, more fer-
the microphytobenthos was supplied with sufficient ric iron was present at the latter station. Since phos-
phosphate. In September, however, ferric iron exceed- phate complexes with ferric rather than with ferrous
ed the sediment phosphate content by a factor of 15, iron it might be expected that more phosphate would
which is likely to have resulted in phosphate limita- be bound at Station 12. Furthermore, the addition of
tion. At Station 12 the situation again was reversed. 25 mM ferric iron would be expected to completely
In September ferric iron and phosphate showed equal bind phosphate. Notwithstanding the fact that, in con-
concentrations, indicating that phosphate supply prob- trast to Station 12, this ferric iron was almost instan-
ably was not limiting primary productivity. However, taneously reduced in Station C (Table 2), still more
in May ferric iron was in large excess (almost 5-fold). phosphate was bound at Station C. Thus, although
In order to study binding and mobilization of phos- iron was responsible for part of the phosphate bind-
phate at both sampling sites, sediment slurry experi- ing, the results cannot be solely explained in terms
ments were performed. Phosphate was added to sedi- of iron concentration. It is possible that the differ-
ment slurries was rapidly immobilized but the amount ences in physico-chemical conditions between the two
that was bound differed between the two sampling stations were responsible for the observed differences
sites. At Station C 90% of the phosphate added to the (Robertson & Alexander, 1992). Considerable differ-
slurry (100 p,M) was bound while at Station 12 this was ences have been observed in temperature, salinity and
only 76%. If the sediment slurries were enriched by the pH between these sampling sites. Moreover, phosphate
addition of 25 mM ferric iron (estimated final concen- can also be bound by calcium (Ehrlich, 1990, Hirschler
trations of ferric iron in the slurries of Stations C and et aI., 1990; Sukenik et aI., 1985).
12 are respectively 26 and 30 mM) 99.7% and 95.8% In order to understand the mechanisms of phos-
phate release from the sediments, experiments were
193
-
fact, highest phosphate release was observed in glucose :.
- 80- I
amended sediments in the presence of 20 mM molyb- G)
-
~
solution of calcium phosphate (Petterson et aI., 1988). .2: 80
G)
Fractionated extraction of phosphate from the sedi- «I 60
..c:
ments also confirmed that a large portion of phosphate Co
may be bound as calcium phosphate (Stal, unpublished
1/1
0 40
..c:
results). Thus the mobilization of phosphate from sed- a. 20 l - _ _ ....:..,.. r-
iments is a complex process in which the contribution 0 ~----~---------~------~--~
of iron and sulfate reduction in the decomposition of 0 1 24 48 90
organic matter plays a central role.
Time (h)
phosphate binding capacity. Therefore, primary pro- nitrogenase from oxygen inactivation are not precisely
ductivity in this system may be phosphate limited at understood (Gallon, 1992). It has been proposed that
particular times of the year. In contrast, Station 12 such cyanobacteria separate nitrogen fixation tempo-
in Etang du Prevost had a significantly lower phos- rally from oxygenic photosynthesis, i.e. nitrogen fix-
phate binding potential and particularly in summer a ation is confined to the dark (Gallon & Stal, 1992).
much higher phosphate content. This would almost However, nitrogen fixation in the non-heterocystous,
certainly result in nitrogen limitation. The sediments planktonic cyanobacterium Trichodesmium spp. is
of both stations were characterized by benthic commu- strictly light-dependent. Recent discoveries indicate
nities of phototrophic microorganisms. In the Etang du that confinement of nitrogen fixation to the dark in
Prevost the sediment at Station 12 was covered by many non-heterocystous cyanobacteria is determined
a microbial mat composed of Oscillatoria sp., a fil- by oxygen dynamics which cause supersaturation in
amentous, non-heterocystous cyanobacterium. In the the light (Sta!, 1995). Aerobic nitrogen-fixing non-
Bassin d' Arcachon the sediment of Station C was fre- heterocystous cyanobacteria are particularly known
quently colonized by benthic diatoms but sometimes from marine microbial mats which are characterized
mats of the heterocystous cyanobacterium Anabaena by oxygen supersaturation during the day and anoxic
sp. developed. Because environmental conditions at conditions during the night. It must be emphasized that
Station C undergo considerable fluctuations the ben- aerobic nitrogen-fixing, non-heterocystous cyanobac-
thic cyanobacterial communities were not particular- teria are not optimal adapted for diazo trophic growth
ly stabile. The occurrence of benthic communities of (Stal et aI., 1994). Nitrogenase in these organisms suf-
diatoms in Station C suggested phosphate rather than fer clearly from oxygen inactivation which is coun-
nitrogen limited primary productivity. However, mas- teracted by a high rate of de novo synthesis of the
sive developments of benthic diatoms may successive- enzyme. Therefore, notwithstanding their capability
ly lead to nitrogen limitation. It was considered that of diazotrophy, growth of these organisms may still be
under such conditions nitrogen-fixing cyanobacteria nitrogen limited. In contrast, heterocystous species can
may develop. provide all nitrogen necessary and are rarely nitrogen-
Cyanobacteria are oxygenic photoautotrophic limited.
prokaryotes and many species are capable of dia- Diel variations of nitrogenase activity at Station C
zotrophic growth (Fay, 1992). However since nitro- (Bassin d' Arcachon) showed a strict light dependence.
genase is extremely oxygen sensitive, nitrogen-fixing The cyanobacterial community in this station was com-
cyanobacteria have evolved different strategies in order posed of the heterocystous species Anabaena sp. in
to protect the enzyme from inactivation (Gallon, 1992). which nitrogen fixation was light dependent. In Sta-
Nitrogen-fixing cyanobacteria can be subdivided in 3 tion 12 (Etang du Prevost) nitrogenase activity was
groups. (i) Heterocystous cyanobacteria differentiate usually much higher and a clear diel variation was not
special cells, heterocysts, that have lost the capac- observed. The cyanobacterial community was com-
ity of oxygenic photosynthesis and in which nitro- posed of the filamentous non-heterocystous species
gen fixation takes place. Heterocysts possess special Oscillatoria sp. In Figure 7 the effects of light and
cell walls that provide a diffusion barrier that limit oxygenic photosynthesis on nitrogenase activity of
the entrance of oxygen. The interior of heterocysts is both communities are depicted. The Anabaena com-
therefore virtually anoxic. Among the cyanobacteria munity of Bassin d' Arcachon showed highest activ-
heterocystous species can be regarded as best adapted ity in the light (Figure 7a). In the dark activity was
for diazotrophic growth. (ii) Non-heterocystous, fila- much decreased while the .addition of 10- 5 M 3-
mentous or unicellular species fix nitrogen only under (3,4-dichlorophenyl)I,I- dimethylurea (DCMU), an
anoxic or extremely low-oxygen conditions. The strat- inhibitor of oxygenic photosynthesis (PS II), resulted
egy of these species is clearly the avoidance of oxy- in a slight decrease of activity. This result was expect-
gen since an adequate protection mechanism is absent. ed with heterocystous species since heterocysts do not
Although many cyanobacteria (may be up to 50% contain photosystem II and therefore DCMU has no
of all species) may belong to this group, their eco- effect. The slightly lower activity in the presence of
logical relevance has not been satisfactory elucidat- DCMU can be explained by an overall inhibitory effect
ed. (iii) Non-heterocystous, filamentous or unicellular on cyanobacterial metabolism or the limited supply
species that fix nitrogen under fully oxic conditions. with reducing equivalents from the vegetative cells.
The mechanisms by which these organisms protect The inhibition of photosystem II in the Oscillatoria
195
...
cyanobacteria dark activity is often higher than light :t: 0.2
incubations. This highlights the problems these organ- ::I:
isms have with photosynthetic oxygen production and ()
0.1
explains why nitrogen fixation in natural populations (5
E
is usually confined to the dark. The different behaviour ::1.
7 5
:c 6
u 4
en 5
E
:c 4 3
.z 3 2
u
(5
2
E 1
::t
o 10 20 50 o 0.5 1 5 10
mM Glucose mM Sulfide
0.25 0.5
";
:c
u 0.20
~
u 0.4
en 01
E 0.15 E
'k :t:
...
.z
u
0.10 :I:
U
0.2
mM Glucose mM Sulfide
Figure 8. The effect of the addition of glucose on nitrogenase Figure 9. The effect of the addition of sulfide on nitrogenase activity
activity (acetylene reduction) in sediment samples from the Bassin (acetylene reduction) in sediment samples collected from the Bassin
d' Arcachon (Station C) (A) and Etang du Prevost (Station 12) (B). d' Arcachon (Station C) (A) and Etang du Prevost (Station 12) (B).
Incubation was done under ambient light. Incubation was done under ambient light.
it is evident that iron plays a key role in the biogeo- The results also clearly demonstrate the mobilization
chemical processes occurring in these lagoons. The of iron. Mobilization of iron may proceed simulta-
Bassin d' Arcachon receives a large amount of iron neously with re-oxidation of sulfide and the libera-
which accumulates in the sediments and there is a tion of phosphate. In the Bassin d' Arcachon iron was
strong seasonal fluctuation in sediment iron content. in large excess of acid volatile sulfide. This means
The regulating mechanisms of iron (im)mobilization in that any sulfide that is produced will precipitate as
this lagoon are not precisely known and need further FeS and free sulfide will never reach toxic concentra-
investigation. Oxidation of ferrous iron may lead to tions. In addition, ferric iron was in excess of total
insoluble precipitates of ferric oxides. Also precipita- extractable phosphate in the sediment and addition of
tion ofFeS or pyrite (FeS2) is important for iron immo- phosphate resulted in instantaneous binding to the sed-
bilization. Benthic cyanobacteria may also be involved iment. Although ferric iron was certainly involved,
in binding of iron to their extracellular polysaccha- it was not the only mechanism of phosphate bind-
ride sheaths (Decho, 1990). Thus, immobilization of ing in these sediments. Calcium possibly may also
iron may be largely regulated by primary productiv- be important in the binding of phosphate. Iron reduc-
ity and sulfate reduction (Howarth & Merkel, 1984). tion did not necessarily result in phosphate liberation,
This explains why iron accumulation in these sedi- which would have been expected if ferric iron was the
ments seemed to occur predominantly during summer, only phosphate binding component in these sediments.
when primary productivity and sulfate reduction were Phosphate liberation was possible after enriching the
important. It is unlikely that the load of iron into the sediment with glucose however, the addition of molyb-
basin was higher in summer than in other seasons. date, which inhibits sulfate reduction, did not prevent
197
Danenlouwerse, H., L. Lijklema & M. Coenraats, 1993. Iron content Lovley, D. R., E. J. P. Phillips & F. Caccavo, 1992. Acetate oxidation
of sediment and phosphate adsorption properties. Hydrobiologia by dissimilatory Fe(lII) reducers. Appl. Envir. Microbial. 58:
253: 311-317. 3205-3206.
Ehrlich, H. L.. 1990. Geomicrobiology. Marcel Dekker, New York, Oremland, R S. & D. G. Capone, 1988. Use of 'specific' inhibitors
646 pp. in biogeochemistry and microbial ecology. Adv. Microb. Ecol.
Fay, P., 1992. Oxygen relations of nitrogen fixation in cyanobacteria. 10: 285-383.
Microbial. Rev. 56: 340-373. Oremland, R. S. & S. Polcin, 1982. Methanogenesis and sulfate
Gallon, J. R., 1992. Reconciling the incompatible: N2 fixation and reduction: competitive and noncompetitive substrates in estuarine
O 2. New Phytol. 122: 571-609. sediments. Appl. Envir. Microbiol. 44: 1270-1276.
Gallon, J. R. & L. 1. Stal, 1992. N2 fixation in non-heterocystous Pettersson, K., B. Bostrom & O.-S. Jacobsen, 1988. Phosphorus in
cyanobacteria: an overview. In E. 1. Carpenter, D. G. Capone & sediments - speciation and analysis. Hydrobiologia 170: 91-101.
1. G. Rueter (eds) Marine pelagic cyanobacteria: Trichodesmium Phillips, E. J. P., D. R. Lovley & E. E. Roden, 1993. Composition
and other diazotrophs. Kluwer Academic Publishers, Dordrecht, of non-microbially reducible Fe(III) in aquatic sediments. Appl.
NATO ASI Series C. Vol. 362: 115-139. Envir. Microbial. 59: 2727-2729.
Haselkorn, R., 1978. Heterocysts. Annu. Rev. Plant Physiol. 29: Pugnetti, A., P. Viaroli & I. Ferrari, 1992. Processes leading
319-344. to dystrophy in a Po River delta lagoon (Sacca Di Goro):
Hirschler, A., J. Lucas & J.-C. Hubert, 1990. Bacterial involvement phytoplankton-macroalgae interactions. Science Tot. Envir. Sup-
in apatite genesis. FEMS Microbial. Ecol. 73: 211-220. pI. 1992: 445-456.
Howarth, R W. & S. Merkel, 1984. Pyrite formation and the mea- Reynolds, A. & A. E. Walsby, 1975. Water-blooms. BioI. Rev. 50:
surement of sulfate reduction in salt marsh sediments. Limnol. 437-481.
Oceanogr. 29: 598-608. Riegman, R., A. A. M. Noordeloos & G. C. Cadee, 1992. Phaeocystis
Howsley, R. & H. W. Pearson, 1979. pH dependent sulphide toxicity blooms and eutrophication of the continental coastal zones of the
to oxygenic photosynthesis in cyanobacteria. FEMS Microbial. North Sea. Mar. BioI. 112: 479-484.
Lett. 6: 287-292. Robertson, B. K. & M. Alexander, 1992. Influence of calcium, iron,
Lovley, D. R, 1987. Organic matter mineralization with the reduc- and pH on phosphate availability for microbial mineralization of
tion of ferric iron: a review. Geomicrobiol. J. 5: 375-399. organic chemicals. Appl. Envir. Microbial. 58: 38-41.
Lovley, D. R., 1991. Dissimilatory Fe(III) and Mn(lV) reduction. Skyring, G. W., 1987. Sulfate reduction in coastal ecosystems.
Microbial. Rev. 55: 259-287. Geomicrobiol. 1.5: 295-374.
Lovley, D. R. & E. J. P. Phillips, 1986. Organic matter mineralization Stal, 1994. Microbial mats in coastal environments. In L. 1. Stal
with reduction of ferric iron in anaerobic sediments. Appl. Envir. & P. Caumette (eds), Microbial mats. Structure, development
Microbial. 51: 683-689. and environmental significance. NATO ASI Series 35 Springer
Lovley, D. R & E. J. P. Phillips, 1987a. Rapid assay for microbially Verlag, Heidelberg: 21-32.
reducible ferric iron in aquatic sediments. Appl. Envir. Microbial. Stal, 1995. Physiological ecology of cyanobacteria in microbial mats
53: 1536--1540. and other communities. New Phytol. 131: 1-32.
Lovley, D. R. & E. J. Phillips, 1987b. Competitive mechanisms for Stal, L. J., H. W. Paerl, B. Bebout & M. Villbrandt, 1994. Hete-
inhibition of sulfate reduction and methane production in the zone rocystous versus non- -heterocystous cyanobacteria in microbial
of ferric iron reduction in sediments. Appl. Envir. Microbial. 53: mats. In L. J. Stal & P. Caumette (eds), Microbial mats. Structure,
2636-2641. development and environmental significance. NATO ASI Series
Lovley, D. R. & E. J. P. Phillips, 1988. Novel mode of microbial 35 Springer Verlag, Heidelberg: 403-414.
energy metabolism: organic carbon oxidation coupled to dissim- Sukenik, A., W. Schroeder, J. Lauer, G. Shelef & c. J. Soeder,
ilatory reduction of iron or manganese. Appl. Envir. Microbial. 1985. Coprecipitation of microalgal biomass with calcium and
54: 1472-1480. phosphate ions. Water Res. 19: 127-129.
Hydrobio[ogia 329: 199-210,1996. 199
p. Caumette,l. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Key words: Sulfur bacteria, MPN counts, vertical distribution, marine sediments, oxygen, sulfide, eutrophication
Abstract
Enumeration of the functional groups of sulfur bacteria was performed in the sediments in the Bassin d' Arcachon,
a mesotidal lagoon with strong tidal currents and dominant populations of seagrass (Zostera noltii), and in the
Etang du Prevost, a shallow lagoon with moderate tidal fluctuations and dominant populations of floating seaweed
(Ulva sp.). In addition, data were collected on the distribution of oxygen and sulfide at the water-sediment
interface during diel cycles. Bacterial enumeration studies revealed highest numbers in the top two cm of the
sediments for three functional groups of sulfur bacteria, these being the sulfate-reducing bacteria (SRB), the
colorless sulfur bacteria (CSB), and the phototrophic sulfur bacteria (PSB). In both systems high numbers of SRB
were encountered, suggesting ample availability of organic matter. A comparison between different sites in each
ecosystem showed that sediments overlain by more stagnant water were dominated by PSB, whereas those overlain
by more oxygenated water were dominated by CSB. Important factors are the physical forces induced by tidal
currents and the degree of daily exchange of water between the lagoons and the sea. These factors may explain the
differences observed between the two systems with regard to the development of anoxic conditions, more so than
the level of eutrophication. It appears that rooted plants play an important role in the introduction of oxygen into
the sediments, thus enhancing the competitive position of CSB compared to PSB. Mini-electrodes studies revealed
high concentrations of free sulfide at the inner site of the Etang du Prevost but very low concentrations at the
inner station of the Bassin d' Arcachon, which may be explained by the high iron input of the latter, rather than by
differences in the rate of sulfide production.
when oxygen-consuming processes in the water col- oxygen and sulfide in both systems in the sediment and
umn exceed the supply of oxygen by production and at the water-sediment interface during a full diel cycle.
diffusion. This often coincides with the presence of Special attention was paid to the impact of Zostera beds
sulfide in the water column, resulting in suspended on the distribution of oxygen in the sediments and the
elemental sulfur (,white water'). During these periods, influence there of on the vertical distribution of CSB.
blooms of phototrophic sulfur bacteria in the water
column are frequently observed (Caumette, 1987). In
terms of eutrophication, the Etang du Prevost is consid- Materials and methods
ered to be more eutrophic than the Bassin d' Arcachon.
In coastal environments, when high mineralization Description of systems and sites
and respiration rates prevail, oxygen often becomes
limiting. Under these conditions a shift to alterna- The systems under study are both situated in France,
tive electron acceptors is observed of which sulfate, the Arcachon Basin (Bassin d' Arcachon) is situated
due to its abundance in the marine environment, is 80 km south-west of Bordeaux at the Atlantic coast,
the most important. In anoxic coastal sediments a the Prevost lagoon (Etang du Prevost) is situated 40 km
considerable part of the organic matter (>50%) is south-west of Montpellier on the Mediterranean.
mineralized via sulfate reduction by sulfate-reducing The Bassin d' Arcachon is a triangular-shaped
bacteria (SRB)(l¢rgensen, 1978; Thode Andersen & mesotidal lagoon with a total area of 156 km 2. At
l¢rgensen, 1989). The H2S produced is extremely tox- low tide only 40 km 2 of the basin is covered by water,
ic for most aerobic organisms; abiotically it can be from which an area of 10 km 2 is used for oyster farm-
trapped by iron and precipitated as ferrous sulfide or ing. In the system, the tidal amplitude ranges from
FeS2 (pyrite). Sulfide can also be oxidized by two 3.90 at spring tide to 2.10 m at neap tide. The Etang
functional groups of sulfur bacteria: the colorless sul- du Prevost is a relatively shallow lagoon covering an
fur bacteria (CSB) and the phototrophic sulfur bacte- area of 4.8 km 2 with a water depth ranging from 1.5
ria (PSB). The availability of electron acceptors, such to 2 m. In each system two locations were chosen, one
as oxygen or nitrate, is important for the growth of relatively close to the channel through which water
CSB, whereas phototrophic growth of PSB depends exchange takes place (the outer station), and one dis-
on the availability of light. Alternatively, PSB may tant from the channel (the inner station). In the Bassin
use oxygen during chemotrophic growth as has been d' Arcachon, the outer station (Station A) was situated
demonstrated for Thiocapsa roseopersicina (Bogorov, on an intertidal flat in the center of the basin where
1974; De Wit & Van Gemerden, 1987), and oth- coastal water was well mixed with basin water during
er Chromatiaceae (Kampf & Pfennig, 1980, 1986; each tidal cycle. The sediment at this site consisted
Overmann & Pfennig, 1992). Similarly the okenone- of muddy sand. Station B was located in the inter-
containing Thiocapsa strain 5811, isolated from the tidal area in the eastern part of the basin where little
Etang du Prevost has the potential for chemotrophic water exchange with coastal waters occurs. Sediment
growth (Caumette et aI., 1985). composition at this site was predominantly black mud
The abundance of these groups is largely deter- mixed with some sand. The sampling sites A and B
mined by the environmental factors described above. were both situated in eelgrass beds.
Sulfide oxidation by CSB and PSB is considered to In the Etang du Prevost samples were taken at Sta-
play an important role in these environments, in addi- tions X and 11. Station X was situated on a tidal flat
tion to sulfide removal by FeS formation. near the channel (Grau) to the Mediterranean. This sta-
As part of the research programme for the Coastal tion was devoid of a well developed flora. The sediment
Lagoon Eutrophication and ANaerobic processes consisted of grey-colored coarse sand. The second
(C.L.E.AN.), a project on the abundance of three func- sampling site (Station II) was situated in the western
tional groups of sulfur bacteria (SRB, CSB, and PSB) part of lagoon where the exchange of water is minimal.
in the sediments of the two lagoons was carried. The In these sediments (silt mixed with sand) a significant
hypothesis was put forward that in the Etang du Prevost infauna was present, revealed by the large quantities
these bacterial groups are more important than in the of shells. At this station a dense population of floating
Bassin d' Arcachon, on the basis of the occurrence of algae was encountered, consisting predominantly of
'white water' and blooms of phototrophic sulfur bacte- Ulva sp. Station 11 was continuously submersed and
ria. Data were collected on the vertical distribution of
201
situated at a depth of approximately 50 cm below the Vertical distribution of oxygen and sulfide
water surface.
More detailed information on the different sam- In 1994, data were collected during two field cam-
pling stations is given in the introductory chapter of paigns. During the June campaign attention was
this volume. For data on seasonal variations the reader focussed on the possibility of sulfide re-oxidation in
is referred to Bourgues et al. (1995) and De Wit et aI. Zostera beds at station B in the Bassin d' Arcachon,
(1994). while during the August campaign most attention was
paid to the measurement of oxygen and sulfide profiles
Enumeration offunctional groups, MPN methodology during full diel cycles.
The small-scale sensors for the measurement of
Data on the vertical distribution of CSB, PSB and SRB oxygen and sulfide were sheathed in stainless steel
were collected in the Bassin d' Arcachon in June and needles (outer diameter 0.9 mm) in order to prevent
September 1993, and June 1994. The Etang du Prevost damage to the sensors by shells and other small objects.
was sampled only during the two field campaigns in Particularly in the sediments of Station 11 in the Etang
1993. du Prevost are shells abundantly present. Details of the
Sediment samples were taken using 3.5 cm diame- construction and calibration of the mini-electrodes are
ter stainless steel cores. These cores were either taken given in Van Gemerden et aI. (1989) and Visscher et aI.
directly to the laboratory and processed immediately (1991).
upon arrival, or kept refrigerated and processed with- During the June 1994 campaign, a platform was
in one day. Based on introductory experiments, cores installed at station B in an area dominated by eelgrass.
were sliced in six consecutive layers, corresponding to At each side of the platform a 0.75 x 2 m area was
depth layers of 0-5 mm, 5-10 mm, 10-15 mm, 15- selected for sampling: in one part Zostera leaves were
20 mm, 20-30 mm and 30-50 mm. Each slice was carefully removed without removing the roots, while
suspended in sterilized, filtered and oxygen-free aged in the other part Zostera plants were left intact. During
seawater (1:4 vol/vol) and subsequently subjected to periods with no overlying water, cores were taken at
mild ultrasonic treatment (15 Hz for 30 s, Branson B3 regular intervals at each side of the platform. Oxygen-
bath) in order to detach the cells from sediment parti- and sulfide profiles were recorded at depth intervals
cles and to loosen clumps. The resulting slurry, which of 0.1-1 mm within 15 minutes after sampling, using
was used for the enumeration of all three function- a custom made nanoAmp- and m V-meter (Electronic
al groups of sulfur bacteria, was directly pipetted in Workshop, Biology Department, University of Gronin-
the first row of three microtiter plates (12 rows of eight gen).
wells, each maximally containing 250 ILl). The remain- During the August campaign electrode studies were
ing wells were previously filled with 200 ILl of growth performed using a computer controlled submersible
medium for the group to be enumerated. Ten-fold dilu- device developed for in situ measurements in shal-
tion steps were used. Microtiter plates for PSB and low marine environments (Mechanical and Electron-
SRB were incubated in an oxygen-free environment in ic Workshops, Biology Department, University of
gas-tight plastic bags (Anaerocult system, Merck) and Groningen). The software for the interactive control
incubated at room temperature in incandescent light was developed by TiePie Engineering, Jornwerd, The
and in darkness, respectively. Medium composition Netherlands. A brief description follows, details of the
for the enumeration of the different functional groups system will be published separately. On a I-m high
was as described previously (Visscher et aI., 1992a). stainless steel frame, an electrode block was mounted,
Growth of PSB was assessed by pigmentation of the which can be moved horizontally, along an X-and an
wells, growth of CSB was judged from acid forma- Y-axis for a distance of 50 cm. Vertically, the maxi-
tion and turbidity, whereas for the presence of SRB mum range was 30 cm with a minimal resolution of
blackening of the growth medium due to FeS precipi- 0.1 mm. Profiles were recorded after a depth sensor,
tation was taken as the criterion. Growth was checked mounted on the electrode block, has detected the sed-
frequently and final scores were obtained after six to iment surface. Needle electrodes for oxygen, sulfide,
ten weeks of incubation. The most probable numbers pH, redox and temperature (four each) were mounted
and 95% confidence limits were calculated according on the electrode block. Amplifiers and nanoAmp-mV
to the method of Klee (1993). converters were mounted directly on top of the elec-
202
trodes and millivolt signals sent to an on-land computer hand panels). Maximum cell densities for CSB were
for further analysis. found in the 5-10 mm depth layer (7 x 108 cells cm -3),
During a diel cycle, measurements were repeated but counts were still 106 cm -3 in the 30-50 mm layers.
each hour and the horizontal spacing between mea- As at Station A, low numbers of PSB were encoun-
surements was 2.5 cm. tered. The average popUlation density in the 0-2 cm
layer, containing 70% of the cells in the 0-5 cm layer,
was 3 x 103 cells cm- 3 . As at Station A, CSB domi-
Results nated over the PSB.
For the SRB, maximum counts were obtained in
Bassin d'Arcachon the 10-15 mm depth layer (6 x 107 cells cm- 3 ). As in
Station A, 90% of the popUlation of this group revealed
Station A to be present in the 0-2 cm depth layer, numbers close
Data on the vertical distributions of sulfur bacteria in to 106 cm- 3 were observed in the deeper layers.
the sediments of the Bassin d' Arcachon are presented In June 1994 a comparison was made between
in Figure 1. the numbers of CSB and PSB inside and outside the
At station A (Figure 1, left-hand panels) CSB were Zostera beds. In both conditions, the majority of CSB
present in relatively high numbers with cell densities and PSB were present in the upper 20 mm of the sedi-
up to 109 cells cm -3 sediment in the upper 5 mm of the ments, for CSB being 96% and 93%, respectively, and
sediment. The average cell count in the top two cm was for PSB being 84% and 98%, respectively. For CSB,
6 x 108 cm- 3 , which accounted for 99.5% of the total the average count for the top 20 mm inside the Zostera
population ofCSB in the 0-5 cm depth layer. In the 30- bed was 4.6 x 106 cm- 3 , and outside the seagrass bed
50 mm layer maximum counts were 2 x 106 cells cm - 3 30.4 x 106 cm -3. For the PSB the corresponding num-
sediment. PSB were detected in all depth layers from bers were 2.1 x 103 cm- 3 and 2.0 x 106 cm-3, respec-
0-5 cm, however, numbers were low when compared tively. The impact of the seagrass on the abundance
to CSB. Maximum population densities were found in of sulfide-oxidizing bacteria is elegantly illustrated by
the 10-15 mm depth layer (3 x 105 cells cm- 3 ), cell comparing the ratio CSB/PSB on a biomass basis. Out-
densities in the other depth layers were two orders of side the Zostera beds the contribution of CSB to the
magnitude lower. The average count in the top two cm total biovolume of the populations of CSB and PSB
of the sediment was 8 x 104 cells cm-3, accounting was 60.8%, inside the Zostera beds the corresponding
for 98% of the popUlation from 0-5 cm. At this site figure was 99.5%. It thus appears that the competitive
the CSB clearly outnumbered the PSB. The same was position of colorless sulfur bacteria was enhanced in
found during the campaigns in June 1993 and June seagrass beds.
1994. For a proper evaluation of functional groups, the During the June campaign 1994 data were collected
average individual cell size of each groups has to be on the profiles of oxygen and sulfide in sediment cov-
taken into account, for PSB this value is approximately ered with Zostera sp. (Figure 2, left-hand panels) and in
1 O-times higher than for CSB (Visscher & Van Gemer- an area from which the leaves of Zostera were removed
den, 1993; Van den Ende et aI., 1995). At Station A, the while the root system was left intact (Figure 2, right-
total biomass of CSB thus was approximately 100-fold hand panels). Cores were taken at 2-h intervals, except
higher than that of the PSB. when the sampling site was submerged. The differ-
Highest population densities of SRB also were ences observed between four different electrodes used
found in the 0-2 cm layer (2 x 107 cells cm- 3 ), repre- for each parameter were minimal, except for 18:50 h
senting 90% of the population present in the layer of when one electrode showed much deeper penetration
0-5 cm. In the deeper layers cell numbers decreased to of oxygen than the others. At the surface of the Zostera-
106 cm -3. covered area, oxygen concentrations were maximal in
the late afternoon (200-300 flmoll-I), but otherwise
Station B showed little variation with time. The depth of oxygen
The vertical distribution, as well as the abundance, of penetration was <5 mm during the night and up to
the three groups of sulfur bacteria was comparable to 10 mm during the day. The removal of the Zostera
Station A, i.e. highest population densities were found leaves resulted in lower maximum oxygen concen-
in the upper two cm of the sediments (Figure 1, right- trations in the superficial layers, and in much shal-
lower penetration. No clear differences were observed
203
~____~S~
ta~ti~
o~n _A______~11 ~I_______
ru_a_ti_on__B______~
I COLOUR LESS SULFUR BACTERIA
~ depth (mm) - cells (log N/cm 3) ~ depth (mm) - cells (log N/cm 3 )
3 458 7 8 9 10 3 4 5 8 7 8 9 10
0 ~•. ,7
······~
,: ~·'·~
'· ===7=2
=~=="--~---,,--,
10
20
"", I
30 =' ;===.===J:!.....-I
p:=--"'
.<
10§=
f--
10
Et-
20
30 30
40 40 t--
50
. iIffn 0_ 1im
SULFATE - REDUCING BACTERIA
O
~ depth (mm) - cells (log N/cm 3) ~ depth (mm) - cells (log N/cm 3)
3 4 5 8 7 8 9 10 3 4 5 8 7 8 9 10
10 10
20 20
30 30
Figure 1. Vertical distribution of sulfur bacteria counted by the MPN-method in the top 50 mm at the outer station (Station A) and the inner
station (Station B) in the Bassin d' Arcachon. Cores were collected on September 3Td 1993. Sulfate-reducing bacteria were enumerated with
lactate and acetate; colorless bacteria with thiosulfate and anoxigenic phototrophic sulfur bacteria with sulfide and thiosulfate. Error bars indicate
95% confidence intervals after Klee (1993).
204
at different times, indicating that oxygen production (:::::: 5 x 104 cells cm- 3 sediment) spread over the entire
was primarily due to Zostera together with the epi- depth layer analyzed.
phytes removed with the leaves of the latter. It thus Although maximum numbers of SRB were found
appears that sediments with intact Zostera plant are in the upper cm of the sediment (106 cells cm- 3 ), a
more oxygenated than those from which the leaves substantial number of viable cells was present in the
were removed. Free sulfide could not be detected, deeper layers, resulting in an average of 3 x 105 cells
neither underneath the Zostera beds, nor in the sed- cm- 3 over the entire depth horizon from 0-50 mm.
iment from which the Zostera leaves were removed. In August 1994 depth profile data were collected
The m V-signal of the sulfide electrode did not allow on the distribution of oxygen and sulfide. Oxygen con-
calculation of concentrations, nevertheless the trend centrations in the water phase just above the sediment
observed was that sediments without Zostera leaves ranged from 75 /lmoll- I to 250 /lmoll- I . Concentra-
were more reduced than the sediments without Zostera tions were low in the early morning and progressively
leaves (Figure 2). increased during the day to become maximal in the
In August 1994, data were collected at I-h inter- afternoon. The oxygen penetration depth into the sed-
vals on the distribution of oxygen and sulfide in the iments was fairly constant over time. In general, dur-
Zostera beds during a full diel cycle. Oxygen concen- ing the day anoxic conditions were observed at depths
trations, in general, were lower than in June (data not exceeding 5 mm, during the night oxygen penetration
shown). The oxygen concentration in the 10 mm water was less. The maximum oxygen concentrations at this
phase just above the sediments and in between the site never exceeded 100% saturation levels. During the
Zostera leaves ranged from 25 to 150 /lmoll- I , indi- 24-h period sulfide was detected incidentally at very
cating under-saturation. Maximum oxygen concentra- low concentrations.
tions were observed in the late afternoon, while min-
imum concentrations were found between midnight Station 11
and early morning. At all times, concentrations in the At Station 11 (Figure 3, right-hand panels), 8 x 107
sediments gradually decreased to zero. Again, oxygen cells cm -3 of CSB were found in the top 5 mm of the
penetration depth was highly variable and ranged from sediment. The average number in the top 20 mm was
only a few mm to more then 10 mm. On one occasion, 2 x 107 cells cm- 3 sediment, accounting for 90% of
concentrations up to 50 /lmoll- I O2 were observed the viable population of CSB from 0-5 cm.
in the deeper depth layers (8-30 mm) while oxygen in Relatively high PSB numbers (5 x 107 cm- 3 ) were
the 6-8 mm was absent. This phenomena, is proba- found in the top 10 mm of the sediment, and 96% of the
bly due to the presence of Zostera roots. In contrast to viable popUlation was enumerated in the top 20 mm.
measurements performed in September 1993 and June At this location, PSB were more numerous than CSB.
1994, sulfide could be detected in the sediments of Sta- When the numbers were transformed to biomass PSB
tion B, although at very low concentrations «6 /lmol dominated the CSB by a factor of 11. Highest cell
I-I S2-). densities for the SRB likewise were encountered in
the top layers, with a maximum of 6 x 108 cells cm- 3
Etang du Prevost sediment in the upper 5 mm. The numbers of SRB
showed a remarkable decrease of two to three orders of
Station X magnitude between the upper 5 mm and the 5-10 mm
The vertical distribution of sulfur bacteria at the outer layer. The high numbers of SRB in the surface layers
Station X is shown in Figure 3 (left-hand panels). of the sediment may explain the formation of 'white
CSB attained maximum population densities in the water' observed in the days before sampling. If so, it
upper 3 cm of the sediment with an average density may reflect a faster response of SRB to an increase in
of 107 cells cm- 3 , whereas in the 30-50 mm layer the substrate availability when compared to the response
population density was two orders of magnitude lower. of CSB and PSB to increased sulfide concentrations.
Of the entire population in the top 50 mm, 75% was Free sulfide was clearly present in Station 11, the
found in the top 20 mm. maximum concentration observed was 6 mmol I-I.
The popUlation of PSB exceeded that observed The data on oxygen concentrations in the water phase
in the Bassin d' Arcachon and was rather uniformly were comparable to those measured at Station X in the
Etang du Prevost, however the pattern in the diel cycle
205
• DEPTH (mm) - OXYGEN t/Jmolil), SULFIOE (mV) I DEPTH (mm) - OXYGEN (jm1oII1), SULFIDE (mV)
O~i=-_ _ _-='OO::::=~2OO:;;;,==".::300::::..._ ___;:-=:400::---~5OO 0 100 200 300 400 500
'0 i'~
I OXYGEN \
SULFIDE \ or'
'0 GEN
SULFIDE )
~~
t
t
t
tit
I
rot
30~
~
rf
30
~ !==========;;~~~=======;~~i~=!1=~=.OO==~'i ~ i~;;~~=======================!=l~~=25=='=:
,: ~ ~ ~ULFIDE ,:
XYGEN
SULFIDE J
t~ // i
::- I
20' r
~ i
i
,20
30 1/ i 30 ~ 18'.30 .'
~• i "8;~;;;;;;;;;:===================:::::;~==:::::I
!,
/ (I I
r
18:00
o ~ r
10 t
SULFIDE OXYGEN SULFlo:)
I lOr=- _ _S- I
20 f 2O!. ~ i
30 ~f'
~ i
\ i:1fl :
120:45 I
I
~ ~======- .;;:=;:::::==============~1:::::~1 ~ L. ___ .____ . _.. ==========~ ~
,: (~;:'GE>l .~ 1: rGE~:-~-=-'.=-·- SULFIC£ -n
~ ~ I
~rc::~~=======~ : I C
20 20 -------
10
:1
OXYGEN
~I :t
l.:r~'
[
= ~
=I
-.J'
I
o ~~;;:;;;;~;;;:;;:::::::=======::;:;~~ 0 ~;;;;;;;;:::::============~==~
OXYGEN
10FGE
20j
30
°
10
20
30
40
° OXYGEN
10 10
20 20
30 30
L -________________________________
40
~ ~
Figure 2. Concentration of oxygen and signal of sulfide needle sensor in cores taken in a Zostera bed at different times at Station B in the Bassin
d' Arcachon. Left column: intact Zostera plants; right column: leaves of Zostera cut off on June 2 1994 at 15-16 h, Data were collected on
2-3 June 1994. High tides were at 0:23 h and 13:05 at June 2, and at 1:33 and 14: 18 on June 3. Low tides were at 6:19 and 18:49 at June 2, and
at 7:23 and 19:54 on June 3.
206
Station X Station 11
COLOURLESS SULFUR BACTERIA
~ depth (mm) -+ cells (log N/cm 3 ) ~ depth (mm) -+ cells (log N/cm 3 )
345 6 7 8 9 10 3 4
8 95 10 6 7
==+1-
° CE:~~n
O F=~~================~'----'---'
10 ~ --+f-
10
..;:"-:.:.
2O =~~J
20
I',',' . :. ~
'>,.,
30 I====,.--rl 30 :,,::,............ '
40
10 10
20 20
I
30
I 30
I
,,'
40 40
50
SULFATE-REDUCING BACTERIA
~ depth (mm) -+ cells (log N/cm 3) ~ depth (mm) -+ cells (log N/cm 3 )
:E
3 4 5 6 7 8 9 10 4 5 6 7 8 9 10
0
I;'~"'
--
10
20
30 3O rij_ii
40 :..:..~,;: 40
50
Figure 3. Vertical distribution of sulfur bacteria counted by the MPN-method in the top 50 mm at the outer station (Station X) and the inner
station (Station II) in the Etang du Prevost. Cores were collected on September 6 th 1993. Sulfate-reducing bacteria were enumerated with
lactate and acetate; colorless bacteria with thiosulfate and anoxigenic phototrophic sulfur bacteria with sulfide and thiosulfate. Error bars indicate
95% confidence intervals after Klee (1993).
207
was more pronounced. During the night concentrations 5 X 103 cm- 3 (Visscher et aI., 1992a; Visscher & Van
decreased progressively to reach lowest values in the Gemerden, 1993).
very early morning (::::: 10 /Lmol I-I at 10 mm above The MPN counts obtained, particularly those of
the water-sediment interface), while highest oxygen SRB, are generally assumed to be gross underesti-
concentrations were observed between 18 and 20 h. mates due to the use of highly selective media (Her-
bert, 1985). Recovery percentages as low as 0.011 %
have been reported (Gibson et aI., 1987), however, a
Discussion careful choice of conditions, in particular with regard
to the composition of the growth medium, resulted in
Vertical distribution and abundance of sulfur bacteria much higher recoveries (Bak & Pfennig, 1991). Fur-
ther improvements were obtained by using water from
Two principal conclusions can be drawn from the data the sampling site in the preparation of media (Visscher
on the vertical distribution of sulfur bacteria. First- et aI., 1992a).
ly, the substantial numbers observed for the colorless High numbers of SRB were not only found at the
sulfur bacteria (CSB), the phototrophic sulfur bacte- inner station in the Etang du Prevost, but also at the
ria (PSB), and the sulfate-reducing bacteria (SRB), sampling sites in the Bassin d' Arcachon. This may
demonstrate that the sulfur cycle is of major impor- be regarded as indicative for the importance of the
tance both in the Bassin d' Arcachon and the Etang du input of organic matter. The prominent role of SRB in
Prevost. When cell counts were transposed to biomass, the mineralization of organic matter is demonstrated
it was generally observed that the sum of the biomass of by the high rates of sulfate reduction observed in the
CSB and PSB was roughly proportional to the biomass zones colonized by Zostera sp. and Ruppia sp. (Finster
of the SRB. Secondly, the fact that the bulk of the pop- et aI., 1994). No significant accumulation of reduced
ulation of all three groups were found in the surficial sulfur compounds was observed, suggesting that most
layers of the sediments, indicates that the processes of the sulfide produced by SRB was re-oxidized, either
primarily take place in the top layers of the sediments. biotically or abiotically. However, the rate of oxygen
The high popUlation density of SRB measured at the uptake, as measured by the diffusive flux of oxygen
continuously submerged Station 11 (Etang du Prevost), into the sediments, was too low to account for the re-
probably was the consequence of the massive decay of oxidation. In order to explain this discrepancy, Finster
macroalgae in the days before sampling. The observed et ai. (1994) proposed that rooted plants play an impor-
'white water' could be the result of sulfur particles, tant role in the influx of oxygen in these 'anoxic' sed-
either produced biotically by CSB and/or PSB, or abi- iments. In the sediments of stations devoid of rooted
otic ally. However, carbonate precipitation may also plants (e.g. Station 11, Etang du Prevost), bioturba-
have contributed to the formation of 'white water' tion by bivalves may act as the major way to introduce
(Caumette & Baleux, 1980). oxygen into the sediment.
The population densities of sulfur bacteria observed The hypothesis proposed by Finster et ai. (1994)
in general are high compared to other observations. may also explain the high numbers of CSB observed
Caumette (1987) reported maximum numbers of SRB at sites colonized by eelgrass (Figure 1), since oxygen
and PSB at an inner station in Etang du Prevost in the is a key environmental factor for the development of
order of 106 cm- 3 and 105 cm-3, whereas in the inner CSB.
station of Bassin d' Arcachon PSB counts maximally The simultaneous presence of light and reduced
were 3 x 104 ml- I , respectively. In the Ems-Dollard sulfur compounds is favourable for the development of
Estuary maximum numbers of SRB were present just phototrophic sulfur bacteria. The presence of PSB in
below the oxic/anoxic and accounted to 104 colony- cell densities upto 8 x 107 cm- 3 and 3 x 106 cm- 3 in
forming-unit (Laanbroek & Pfennig, 1981), whereas the sediments of the Etang du Prevost and the Bassin
in the Kattegat the populations of SRB were report- d' Arcachon, respectively, may be explained in this
ed to be no higher than 106 cm- 3 (Jorgensen & Bak, way. The re-oxidation of sulfide without the availabil-
1991). However, higher numbers were observed in ity of sufficient oxygen (Finster et aI., 1994), could be
a well developed Microcoleus-dominated microbial explained by the presence of phototrophically grow-
mat on the island of Texel (The Netherlands), in ing PSB, although growth and activities of PSB in
which the population densities ofCSB, PSB, and SRB these ecosystems likely are restricted by the limited
respectively were 2 x 109 cm- 3 , 5 x 107 cm- 3 , and availability of light. Tidal movements, resulting in
208
resuspension of sediment particles create a shadow- The oxygen concentrations observed in June 1994
ing effect. Also the small particle size of the sediment at the surface of the sediments in Station B, were
and the presence of high concentrations of FeS (data higher then those recorded in August. The commu-
not shown; Stal et al. 1994) are likely to reduce the nity at this site was less healthy in August com-
light intensity in deeper layers.Light penetration in sed- pared to June, when oxygen saturating conditions were
iments is restricted to the upper few mm (J0rgensen, measured at the sediment surface. At the end of the
1989; Lassen et aI., 1992; Lassen & De Wit, 1994;). At Zostera growing season (August-September), oxygen
station B, the shadowing effects of dense Zostera cov- concentrations in the sediment between the Zostera
erage, results in low photosynthetically available radi- leaves were lower, which can be explained by a high-
ation (PAR) in the surface layers of the sediments, how- er oxygen demand resulting from the accumulation of
ever, near infrared radiation was less affected (Lassen organic debris and decomposition the Zostera leaves.
et aI., 1994). The significance of the very low num- In August-September low oxygen concentrations or
bers of PSB (1 % of the surface abundance) detected even anoxia may be expected, and red colored patch-
in the deeper layers of the sediments, is marginal. Lit- es of phototrophic sulfur bacteria have been observed
tle is known about the survival characteristics of PSB between decaying Zostera leaves (P. 1. Labourg, pers.
under the prevailing environmental conditions. Possi- comm.).
bly, bioturbation could be responsible for the presence At Station 11, oxygen concentrations in the water
of phototrophic bacteria in deep and dark depth hori- phase were below 25 {1moll- 1 during the night. At this
zons. site the pelagic alga Viva, whose standing crop may
exceed 500 g dry weight m- 2 (Viaroli et aI., 1994)
Profiles of oxygen and sulfide constitutes an important source of organic matter, and
respiration and mineralization results in a high oxy-
Oxygen and sulfide are parameters of crucial impor- gen demand. Comparable low oxygen concentrations
tance, either stimulating or inhibiting the development were observed in 1993, and appear to be the result of
of CSB and PSB. In general, oxygen is required for the increased respiration rates, and not decreased rates of
proliferation of CSB. Nitrate, the alternative electron primary production (Viaroli et aI., 1994).
acceptor for these organisms, is only present at very At station X, variations in oxygen concentration
low concentrations and, consequently, nitrate plays a during the diel cycle were less pronounced. The organ-
minor role (Herbert & Welsh, 1994; Sloth et aI., 1994). ic matter content of the sediments was low in compar-
In addition to oxygen, CSB require sulfide, or another ison to site 11 (De Wit et aI., 1994). The absence of
form of reduced sulfur, as electron donor. a dense Viva community and the location adjacent to
CSB and PSB, although competitors for reduced the exchange channel may be important factors in this
sulfur compounds, might not be expected to thrive in respect.
the same habitat since CSB require oxygen, where- In general, oxygen penetration into the sediments of
as PSB primarily are anoxygenic phototrophs whose the Etang du Prevost was limited to the upper 10 mm
pigment synthesis is inhibited by oxygen (De Wit & of the sediments. However, oxygen penetration may
Van Gemerden 1990a, b). However, when oxygen con- extend to depths of 20 mm or more. Deeper oxygen
centrations show (die!) fluctuations and/or are limiting penetration may result from the presence of roots (e.g.
co-existence is feasible. The co-occurrence of CSB Zostera), animal burrows, or sea water seeping down-
and PSB in these sediments can be explained by the wards through the sediments due to tidal movements.
phenomenon that PSB can oxidize intermediate forms Although tidal differences are small in the Etang du
of sulfur - such as thiosulfate, elemental sulfur, and Prevost (about 20 cm), the coarse sediment will result
poly-sulfide - which are produced by CSB from sul- in exchange phenomena. In the Etang du Prevost ani-
fide upon oxygen limitation (Van Den Ende and Van mal burrows may be the most likely explanation. Fur-
Gemerden, 1993, 1994).It is evident from the data pre- thermore the surface structure of the sediments is high-
sented in Figures 1 and 2 that, the conditions required ly variable in these systems particularly at Station 11
by CSB for growth were met in sediments covered (Etang du Prevost).
with Zostera (Station A and B, Bassin d' Arcachon). The hypothesis put forward by Finster et al. (1994),
These data infer that oxygen was introduced into the that Zostera plants enrich the superficial layers of sedi-
sediments via the eelgrass roots. ments with oxygen, is substantiated by the vertical dis-
tribution of CSB inside and outside the Zostera beds.
209
Removal of Zostera leaves, which prevents transport the Bassin d' Arcachon cannot be explained by a lower
of oxygen from the leaves to the roots, was found level of eutrophication of the Bassin compared to that
to coincide with reduced concentrations of oxygen in of the Etang du Prevost.
the surface sediment layers, and, more importantly, A number of explanations can be invoked to explain
to result in reduced depth of penetration of oxygen the differences between the Bassin d' Arcachon and the
(Figure 2). Although numbers of CSB inside and out- Etang du Prevost, but it appears that two factors are
side Zostera beds were not substantially different, the of major significance. Firstly the strong currents in
ratio CSB/PSB was higher inside compared to outside the Bassin d' Arcachon, due to the large tidal differ-
Zostera beds. CSB were found to dominate ecosystems ence between low and high water, as compared to the
covered with Zostera (Stations A and B), whereas PSB Etang du Prevost, ensure a thorough mixing of the
dominated systems devoid of Zostera Beds (Station B water column, and as a consequence increased oxygen
and 11). availability. Secondly, the higher iron concentrations
The high oxygen-uptake rates of these systems in the Bassin d' Arcachon as compared to the Etang
(Sloth et al., 1994) are in agreement with the dense pop- du Prevost, which effectively and efficiently removes
ulations of CSB, although aerobic heterotrophic organ- free sulfide (Stal et al., 1994) and thus may limit the
isms can be expected to have contributed. Since the activities of PSB and CSB.
complete oxidation of the sulfide formed by the high
sulfate-reduction rates observed did not correspond to
the uptake of oxygen, Finster et al. (1994) postulated Acknowledgments
that sulfide was only partially oxidized, resulting in the
temporary accumulation of reduced intermediates. The authors would like to acknowledge Frank P. van
In this study the hypothesis that cutting of the den Ende, Henk M. Jonkers, Jutta Meyer and Mar-
Zostera leaves might limit aerobic respiration by elimi- cel van der Meer for help and stimulating discus-
nating an important way of introducing oxygen into the sions. We particularly like to thank Wim J. Beuke-
sediment was tested. Although the sediment became ma, Joost H. Roede (Electronic Workshop Biolo-
more reduced free sulfide did not accumulate under gy, Haren), Karel Visser, E. Joop Wittenaar and
these conditions. Bert L. A. Cazemier (Mechanical Workshop Biolo-
Free sulfide, in considerable concentrations, was gy, Haren) for the precision and craftsmanship during
only observed at Station 11 in the Etang du Prevost, the development and construction of the submersible
and not in the sediments of Station B in the Bassin multiple-electrode measuring device. This study was
d'Arcachon. While little organic matter was found at financed by a joint EU project on 'Coastal Lagoons,
Station X, Station B was characterized by high organic Eutrophication and ANaerobic processes (CLEAN)',
matter content and high rates of sulfate reduction (De contract number EV5V-CT92-0080.
Wit et al., 1994; Finster et al., 1994). At Station 11
(Etang du Prevost) the sulfide concentrations in the
sediments were undoubtedly related to the dense Ulva References
populations. Laboratory experiments showed that the
mineralization of 1 g dried Ulva resulted in the produc- Bak, F. & N. Pfennig, 1991. Sulfate-reducing bacteria in littoral
sediment of Lake Constance. FEMS Microbio!' Eco!. 85: 43-52.
tion of approximately 10 mmo1 sulfide (unpublished). Bogorov, L. v., 1974. About the properties of Thiocapsa roseop-
ersicina strain BBS, isolated from the estuary of the White sea.
Comparison between the two ecosystems Microbiology (Trans!.) 43: 275-2801.
Bourgues, S., O. Pringault & R. De Wit, 1995. Spatial and temporal
variations of oxygen uptake, sulfide production and population of
On the basis of the population densities of the function- sulfur cycle bacteria at the C.L.E.AN. stations. C.L.E.AN. Proc.
al groups of sulfur bacteria, one might expect the inner Prog. Rep. II: 109-132, E.U., Environment, Brussels.
stations of the Bassin d' Arcachon and the Etang du Caumette, P., 1987. Role des bacteries phototrophes et des bacteries
Prevost (Stations Band 11, respectively) to be compa- sulfato-reductrices dans les milieux lagunaires. Editions de
I'ORSTOM., Paris: 1-305.
rable with respect to eutrophication levels. Assuming Caumette, P., K. Schmidt, H. Biebl & N. Pfennig, 1985. Char-
that the number of SRB reflects the availability of sub- acterization of a Thiocapsa strain containing okenone as major
strates for these bacteria, which is substantiated by the carotenoid. System. app!. Microbio!. 6: 132-136.
numbers of CSB and/or PSB, it can be inferred that Caumette, P. & B. Baleux, 1980. Etude des eaux rouges dues 11
la proliferation des bacteries photosynthetiques sulfo-oxydantes
the apparent absence of free sulfide in the sediments of
210
dans I' etang du Prevost, lagune saumfitre mediterraneenne. Mar. Laanbroek, H. 1. & N. Pfennig, 1981. Oxidation of short -chain fatty
BioI. 56: 183-194. acids by sulfate-reducing bacteria in fresh water and in marine
De Wit, R. & H. van Gemerden, 1987. Chemolithotrophic growth sediments. Arch. Microbiol. 128: 330-335.
of the purple sulfur bacterium Thiocapsa roseopersicina. FEMS Lassen, c., H. Ploug & B. B. 1l'lrgensen, 1992. A fibre-optic scalar
Microbiol. Ecol. 45: 117-126. irradiance microsensor: application for spectral light measure-
De Wit, R. & H. van Gemerden, 1990a. Growth of the phototroph- ments in sediments. FEMS Microbiol. Eco112: 247-254.
ic sulfur bacterium Thiocapsa roseopersicina under oxic/anoxic Lassen, c., R. de Wit, L. Lorentzen & N. P. Revsbech, 1994. Light
regimens in the light. FEMS Microbiol. Ecol. 73: 69-76. penetration in the sediments of the Arcachon basin and Prevost
De Wit, R. & H. van Gemerden, 1990b. Growth and metabolism lagoon as measured by optic fibre microsensor. C.L.E.AN Proc.
of the purple sulfur bacterium Thiocapsa roseopersicina under Prog. Rep. II: 23--40, E.U., Environment, CG XII, Brussels.
combined light/dark and oxic/anoxic regimens. Arch. Microbiol. Overmann, 1. & N. Pfennig 1992. Continuous chemotrophic growth
154: 459--464. and respiration of Chromatiaceae species at low oxygen concen-
De Wit R., E. Buffan, S. Bourgues & H. Etcheber, 1994. Sea- trations. Arch. Microbiol. 158: 59-67.
sonal and diel variations of the physicochemical parameters at Sloth, N. P., N. Risgaard-Petersen, S. Rysgaard & S. P. Pelegri,
the C.L.E.AN. sampling station in Bassin d' Arcachon, Etang du 1994. Nitrification, denitrification, and nitrate ammonification in
Prevost and Domaine de Certes. C.L.E.AN. Proc. Prog. Rep. II: sediments of two coastal lagoons in southern France. C.L.E.AN.
15-21, E.U., Environment, DG XII, Brussels. Proc. Prog. Rep. II, E.U. Environment, DG XII, Brussels.
Finster, K., A. Gammelgard, N. Risgilrd-Petersen, S. Rysgilrd, S. Pel- Stal, L., S. B. Behrens & F. Kruyning, 1994. Interrelations of nitro-
gri & N. P. Revbech, 1994. Sulfate-reduction in eelgrass-covered gen, sulfur, and iron in benthic cyanobacterial communities in
sediments ofthe Arcachon Basin (France). C.L.E.AN Proc. Progr. two eutrophic marine lagoons. C.L.E.AN Proc. Prog. Rep. II:
Rep II: 364-379, E.U., environment, DG XII, Brussels. 231-253, E.U. Environment, DG XII, Brussels.
Gibson, G. R., R. 1. Parkes & R. A. Herbert, 1987. Evaluation Thode Andersen, S. & B. B. 1l'lrgensen, 1989. Sulfate reduction and
of viable counting procedures for the enumeration of sulfate- the formation of 35 S-Iabelled FeS, FeS2, and S 0 in coastal marine
reducing bacteria in estuarine sediments. 1. Microbiol. Meth. 7: sediments. Limnol. Oceanogr. 34: 793-806.
201-210. Van den Ende, F. P. & H. van Gemerden, 1993. Sulfide oxidation
Herbert, R. A., 1985. Development of mass blooms of photosynthetic under oxygen limitation by a Thiobacillus thioparus isolated from
bacteria on sheltered beaches in Scapa Flow, Orkney Islands. a marine microbial mat. FEMS Microbiol. Ecol. 13: 69-78.
Proc. R. Soc. of Edinburgh. 87B: 15-25. Van den Ende, F. P. & H. van Gemerden, 1994. Relationships
Herbert, R. A. & D. Welsh 1994. Nitrifying and denitrifying bacteria between functional groups of organisms in microbial mats. In
in the Arcachon basin and Prevost lagoon: abundance, activities L. 1. Stal & P. Caumette (eds), Microbial Mats: Structure, Devel-
and diversity. C.L.E.AN Proc. Prog. Rep. II: 199-229, E.U. envi- opment and Environmental Significance, NATO ASI Series vol
ronment, DG XII, Brussels. G35. Springer-Verlag, Heidelberg: 339-352.
Jl'lrgensen, B. B., 1978. A comparison of methods for the quantifi- Van Gemerden, H., C. S. Tughan, R. de Wit & R. A. Herbert, 1989.
cation of bacterial sulfate reduction in coastal marine sediments. Laminated microbial ecosystems on sheltered beaches in Scapa
III Estimation from chemical and bacteriological field data. Bio- Flow, Orkney Islands. FEMS Microbiol. Ecol. 62: 87-102.
geochem. 1. I: 49-64. Viaroli, P., M. Bartoli, C. Bondavalli, M. Cattadori, G. Giordani, M.
1l'lrgensen, B. B., 1989. Light penetration, absorption, and action Naldi & I. Ferrari, 1994. Macroalgae and macrophytes, oxygen
spectra in cyanobacterial mats. In Y. Cohen & E. Rosenberg and nutrient fluxes at the community level. C.L.E.AN. Proc. Prog.
(eds) Physiological ecology of benthic microbial communities, Rep. II: 255-279, E.U., Environment, DG XII, Brussels.
A.S.M., Washington: 123-137. Visscher, P. T., J. Beukema & H. van Gemerden, 1991. In situ
Jl'lrgensen, B. B. & F. Bak, 1991. Pathways and microbiology ofthio- characterization of sediments: measurements of oxygen and sul-
sulfate transformation and sulfate reduction in a marine sediment fide profiles with a novel combined needle electrode. Limnol.
(Kattegat, Denmark). Appl. Envir. Microbiol. 57: 847-856. Oceanogr. 36: 1476-1480.
Kampf, C. & N. Pfennig, 1980. Capacity of Chromatiaceae for Visscher, P. T., R. P. Prins & H. van Gemerden, 1992a. Rates of sul-
chemotrophic growth. Specific respiration rates of Thiocystis vio- fate reduction and thiosulfate consumption in a marine microbial
lacea and Chromatium vinosum. Arch. Microbiol. 127: 125-135. mat. FEMS Microbiol. Ecol. 86: 283-294.
Kampf, C. & N. Pfennig, 1986. Isolation and characterization of Visscher, P. T., F. van den Ende, B. E. M. Schaub & H. van Gemerden,
some chemoautotrophic Chromatiaceae. 1. Basic. Microbiol. 26: I 992b. Competition between anoxygenic phototrophic bacteria
507-515. and colorless sulfur bacteria in a microbial mat. FEMS Microbiol.
Klee, A. 1., 1993. A computer program for the determination of most Ecol. 101: 51-58.
probable number and its confidence limits. 1. Microbiol. Meth. Visscher, P. T. & H. van Gemerden, 1993. Sulfur cycling in laminat-
18: 91-98. ed marine microbial ecosystems'. In Biogeochemistry of global
Change: Radiatively Active Trace Gases. R. S. Oremland (ed.)
Chapman and Hall, New York: 672-690.
Hydrobiologia 329: 211-222,1996. 211
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and ANaerobic Processes (C.L.E.AN.).
©1996 Kluwer Academic Publishers.
Abstract
This research aims to analyse the sediment capacity to buffer free sulphide release in three coastal lagoons which
differ in terms of eutrophication level, tide influence and primary producer communities. A preliminary estimate
of soluble reactive phosphorus (SRP) regeneration coupled with sulphide fluxes is also made. Sediment profiles
of ferrous and ferric iron and reduced sulphur pools were determined in three stations in the Bassin d' Arcachon
(South West France), in one site in the Etang du Prevost lagoon (Southern France), and in three stations in the Sacca
di Goro lagoon (Northern Italy). Laboratory experiments were also conducted by incubating sediment slurries.
Slurries from the French lagoons were also enriched with about 2% d.w. of organic detritus obtained from the
dominant macrophytes of each site, namely Zostera noltii and Ruppia cirrhosa (Bassin d' Arcachon), and Ulva
rigida (Etang du Prevost). In the Sacca di Goro, slurry experiments were conducted at two sites with different
salinity range, sediment composition and hydrodynamics.
Field data showed that concentrations of available iron (Fe(II)+Fe(III» ranged from a minimum of 28.5 /lmol
cm- 3 (Etang du Prevost) to a maximum of 275.7 /lmol cm- 3 (Sacca di Goro). Moreover, in the French lagoons,
acid volatile sulphide (AVS) accumulation in the superficial sediment was related to ferrous iron concentrations.
Laboratory experiments showed that in spite of strong reducing conditions, sulphide and SRP release was weaker in
iron-rich sediments and in those enriched with the most refractory organic matter. The highest fluxes were detected
in sediment slurries from the Etang du Prevost, which had the lowest iron content, supplied by 2% of the labile
detritus from Ulva rigida. In this case, SRP release was directly related to sulphide production.
Two factors seem significant to evaluate the buffer capacity against free sulphide and SRP release from anoxic
sediment: organic matter biodegradability, which forces sediment toward reducing conditions, and iron availability,
which can affect sulphide mobility as well as the ironhydroxide-phosphate-sulphide system.
sition with sulphate reduction and associated fermen- in the French lagoons and at two sites in the Sacca di
tative reactions exceeding oxic respiration (J0rgensen, Goro.
1983; Giblin & Wieder, 1992). In the anoxic sediment,
sulphides produced by sulphate reduction are rapidly
precipitated, for the most part as iron sulphides (Can- Study area
field, 1989; Howarth & Stewart, 1992). The buffering
capacity exerted by iron against sulphide keeps free- The Bassin d' Arcachon is located on the French
sulphides at very low levels minimising their toxic Atlantic coast (Figure 1, see also Castel et aI., 1996).
effects, since free-sulphides are highly toxic to aquatic Sampling station A lies in the mesotidal centre of the
life (Koch et aI., 1990). However, in sheltered lagoons lagoon and station B in the inner tidal mud-flat. In
with excess macroalgal production (up to 103 g m- 2 both stations, the phanerogam Zostera noltii Hornem.
as dry weight), when water renewal decreases, the is forming large permanent meadows. The third site
sulphide produced from biomass decomposition can (station Cl) is located along the eastern shore of the
exceed the buffering capacity of the system. Thus, Bay, in the artificial impoundments of Certes covered
a considerable amount of free-sulphides can diffuse by large stands of Ruppia cirrhosa (Pet.) Grande.
upwards to the overlying water leading to the ultimate The Etang du Prevost is a sheltered lagoon locat-
stage of dystrophic crisis. ed on the Mediterranean coast of France. The cho-
During dystrophic events, noticeable fluxes of solu- sen station (station 11) lies in the western area of the
ble reactive phosphorus (SRP) have also been observed lagoon, where high production of the seaweed Ulva
as the result of detritus decomposition and strong anox- rigida C. Agardh occurs and since water renewal is
ia (Viaroli et aI., 1992). Even though most organ- limited results in the accumulation of large quanti-
ic phosphate compounds are readily metabolised by ties of macroalgal detritus (see also Caumette, 1986;
microbes, their mobility and availability appear to be Viaroli et aI., 1996).
controlled by the formation of complexes with met- The Sacca di Goro is a microtidal lagoon in the
al ions and hydroxides (De Groot, 1991; De Groot southern part of the Po Delta (Northern Italy). Fresh
& Golterman, 1993). Sediment can accumulate phos- water flowing into the lagoon from the Po di Volano
phate throughout adsorption on ferric hydroxides, pre- Canal and from the Po di Goro deltaic branch determine
cipitation with calcium ions and adsorption on clay par- the salinity gradient. The zonation of the lagoon also
ticles (Ponnamperuma, 1972; Golterman, 1984, 1995). depends on Viva rigida growth. We considered four
Among these complexing compounds, ferric oxides sites located in different zones of the lagoon. Station G
and hydroxides have been considered to playa key role is close to the Po di Volano mouth where mean depth
in marine environments, where iron reactivity is strict- is about 50 cm, salinity range is high and macrophyte
ly linked to sulphide production (Coleman et aI., 1993; cover is negligible. Station 4 (150 cm deep), located
Jensen & Thamdrup, 1993; Golterman, 1995). Thus in the central area of the lagoon, is affected mainly by
phosphorus cycling can be viewed as strictly related to Gracilaria spp. but also by Viva rigida growth. Station
reduced sulphur metabolism. 8 is located in the shallowest (about 30 cm deep) and
Data presented in this paper describes free-sulphide most sheltered area of the lagoon. Station 7, along the
fluxes in three coastal environments which differ in southern sand bank, is about 50 cm deep and is covered
terms of eutrophication level, tide influence and pri- in spring and early summer by Viva rig ida. The latter
mary producer communities. A preliminary estimate of site was considered only for slurry experiments.
soluble reactive phosphorus (SRP) regeneration cou-
pled with sulphide fluxes is also described.
Sediment profiles of ferrous and ferric iron and Materials and methods
reduced sulphur pools were determined in three sta-
tions in the Bassin d' Arcachon (South West France), Sediment profiles
in one site in Etang du Prevost lagoon, (Southern
France), and in three stations in the Sacca di Goro In the Sacca di Goro triplicate sediment cores (inner
lagoon (Northern Italy). diameter 50 mm, length 300 mm) were collected at
Laboratory experiments were also conducted by stations G, 4 and 8 with a hand corer. Samples were
incubating slurries of sediment collected from all sites collected on the 16 March, 15 July and 12 December
1994. In late summer 1994, in the French lagoons
Bassin d' Arcachon
.
N
Sacca di Gore
..
213
N
I.:. ..I,
~j'.,'.{i",~G
,'* .. .
4
.. ,..
. ..
r ;:; . '(
~ ,.- .~ \. ",
,,~
'\.. ..'\. ~~'~. .......... _.
.. ._":.:':,J.
Adriatic Sea
.N
1 km
Etang du Prevost
Figure 1. Map of the sampling stations.
station 11, Ruppia and epiphytes (50% each compo- blue method (Cline, 1969) after fixation with 2% w/v
nent) to the slurries from station C 1, and Zostera to the zinc acetate.
slurries from stations A and B. Reduced sulphur pools, namely acid volatile sul-
On 5 July 1994, similar experiments were conduct- phide (AVS) and chromium reducible sulphur (CRS),
ed with slurries obtained from stations G and 7 in the were determined by means of two sequential distilla-
Sacca di Goro. Sediment and slurries were treated iden- tions under anoxic conditions (Fossing & J0rgensen,
tically to those of French lagoons except for the organic 1989). AVS was detected as H2S released by the addi-
matter treatment which was omitted. Incubation times tion of 10 cm 3 of 5 M HCI to about 1 g of sediment
were also shorter than those adopted for experiments fixed in 2% w/v zinc acetate. CRS, which represents
carried out in the French sites. the remaining reduced inorganic sulphur, was extract-
Sediment manipulation and organic matter (O.M.) ed with lO cm 3 of 1 M CrCIz in 0.5 M HC!. The latter
addition resulted in small variations of O.M. concen- extraction was carried out at room temperature for 20
trations and redox of the slurries in comparison with minutes and at boiling temperature for 40 minutes. At
those normally observed in the field. each step, the sulphides produced were removed in a
. After the collection of the initial samples, the bot- N2 gas stream, fixed in 2% w/v Zn-acetate and then
tles were incubated in the dark under constant N2 flux measured by the methylene blue method.
and stirring for about 2 days in a water bath at con- Ferrous iron was measured after fixation of 1 cm3
trolled temperature. All the sulphide produced in the of sediment in 5 ml of 0.5 M HCI with the phenanthro-
slurry was carried in N2 gas flow and trapped in 100 ml line method (A.P.H.A., 1975). Reactive ferric iron was
of2% w/v Zn acetate. At selected time intervals during determined as ferrous iron after reduction with 10%
the experiment, 40 cm3 of slurry were withdrawn with a ascorbic acid.
syringe from a valve located in the N2 inlet tube. These Dry weight was determined at 70°C until a constant
samples were taken more frequently at the beginning weight was reached. Organic matter content (O.M.)
than at the end of the incubation period. Immediately was measured as weight loss after ignition of 2-3 g of
after collection, the Eh was determined and 3 cm3 of finely powered dry sample at 550°C for 24 hours.
slurry were fixed with 2% w/v zinc acetate for the anal- Total phosphorus was determined spectrophoto-
ysis of reduced sulphur pools. The remaining material metrically after extraction with 20 cm 3 of 1 M HCI
was centrifuged within a few minutes and the super- from 1 g of sediment ash (Aspila et aI., 1976).
natant was filtered through GF/C filters (Whatman) for
porewater analysis (pH, NHt and SRP). In parallel, the
zinc acetate trap was vigorously shaken and a subsam- Results
pie was taken for free sulphides analysis. Density, dry
weight, organic matter, total phosphorus and ferrous Sediment profiles of iron and reduced sulphur pools
iron as well as the above mentioned variables were
determined on the initial and final samples. Measurements conducted in the French lagoons are
restricted to short periods of observation. However,
Analytical determinations they cover the transitional phase from active growth
towards biomass decay of the dominant macrophytes.
Eh was measured with a pH/m V-meter (TIM 90, We thus assumed that this step would be representative
Radiometer) by using a platinum electrode (P1Ol, of the most critical conditions for the biogeochemistry
Radiometer) and saturated calomel reference (K701, of inorganic sulphur and iron.
Radiometer). pH was measured with a combined elec- In the Bassin d' Arcachon and in the Etang du
trode (GK240 1C, Radiometer) and a temperature probe Prevost, the first 10 cm of the sediment horizon were
(T801, Radiometer) for temperature compensation. characterized by a wide range of iron concentrations
Dissolved oxygen was determined by the Winkler (Fe(I!) and Fe(III)) (Figure 2). Maximum values were
method (A.P.H.A., 1975). Soluble reactive phospho- found at station Cl 006.5 to 152.7 /Lmol cm- 3 ) and
rus (SRP) was measured by means of the ascorbic acid to a lesser degree in the uppermost sediment layer
method (Valderrama, 1977), and ammonium by the (0-6 cm) of station B (93.9 to 130.6 /Lmol cm- 3 ).
indophenol blue method (Koroleff, 1970). Sulphide Noticeable differences were found between station B,
concentrations were determined with the methylene in the inner mud-flat, and station A, located in the
central sand bank. The lowest concentrations (28.5
215
Table 1. Experimental design of the slurry experiments. Initial values of organic matter (O.M.),
total phosphorus (TP) and Fe(lI) are reported.
-3 -3
!1mol em -3 !1mol cm !1molem I1ffioI em -3
0 50 100 150 0 50 100 150 0 50 100 150 0 50 100 150
Figure 2. Depth profiles of ferrous (light) and ferric (dark) iron concentrations in the sediments collected from stations A, Band CI in the
Bassin d' Arcachon, and station II in the Etang du Prevost.
to 48.6 /Lmol cm- 3 ) were detected in the dystrophic ed water renewal. In late August 1994, settling of
Etang du Prevost, which was already known to have large amounts epiphytes growing on Ruppia cirrhosa,
a very low iron content (Stal et aI., 1994). Neverthe- increased the O.M. content in the surface sediment
less, at this site the limited iron supply was coupled to 24.6 ± 1.4% d.w. (Bartoli et aI., 1996). Such an
to AVS concentration up to 33 Itmol cm- 3 (Figure 3). increase in O.M. availability would stimulate sulphate
In the Bassin d' Arcachon, AVS concentrations in the reducing bacteria activity which would account for the
first 10 cm of sediment followed a similar pattern to AVS accumulation (see also Viaroli et aI., 1996).
iron distribution. Maximum AVS concentrations (60.5 Low iron content in the surficial sediment at sites
to 79.9 /Lmol cm- 3 ) were detected in the uppermost A and 11 was also coupled to low CRS concentra-
sediment layer (0-6 cm) at station Cl, while minimum tions compared with those detected at stations Cl and
concentrations occurred at station B (less than 10 /Lmol B. At the latter site, CRS levels exceeding 500 /Lmol
cm- 3 ) and A (less than 20 /Lmol cm- 3 ). At these sites, cm- 3 could be achieved due to high iron availability
even though sulphate reduction rates reached the same and the reoxidation of reduced sulphur induced by the
order of magnitude as at station 11 (Bourgues et aI., sediment surface drying out at low tide and by oxygen
1994), AVS accumulation was low in the superficial diffusion from the plant roots (Canfield, 1989; Finster
sediment. This may be due to AVS reoxidation which et aI., 1994).
occurs at low tide when the sediment surface is exposed The relationship between iron and reduced sulphur
to the air. Station Cl is different because of the high pools was also analysed in the Sacca di Goro where
O.M. content in the surface sediment and the limit- three sites located in different zones of the lagoon were
216
-3 -3 -3 -3
/lmolem /lmol em /lmolem /lmo/em
0 10 20 0 10 20 0 4(; a~ 0 20 40
-3 -3 -3 -3
/lmolem flr.·,)! em ~1;'VI em /lmo/em
0 100 200 0 500 1000 0 500 1000 0 100 200
FiRure 3. Depth profiles of AVS (upper panels) and CRS (lower panels) concentrations in the sediments collected from stations A, Band CI in
the Bassin d' Arcachon, and station II in the Etang du Prevost.
-3 -3 -3
l'ITlolcm IUIlOI em l'ITlolem
0 200 400 0 200 400 0 200 400
"n:r
2- 4-6 4-6 4-6
n.d. 8-10
8-10 8-10
st. 8 st. 4 5tG
studied: station G near the freshwater outlet from the both freshwater inflow from the Po di Goro and marine
Po di Volano Canal, station 4 affected by tidal currents inputs determine the salinity regime. Total available
and station 8 in the sheltered part of the lagoon where iron (Fe(II) + Fe(III)) concentrations were on average
217
-3 -3 -3
j.lmolcm j.lmolcm j.lmolcm
o 25 50 75 o 25 50 75 o 25 50 75
51. G
0-2 laB_III 51. 8 51. 4
0-2
n.d.
8-1 0
h
~"[IlII_1I
"":'" ""
""'.
8-10 ~~~,... ~. . • •III
..• 8-10\ii.~~
........ ......... .
', ; ,
-3 -3
-3
j.lmolcm j.lmolcm j.lmolcm
0 200 400 0 200 400 o 200 400
8-10 • • •_ _
8-10 8-10 " ............... .
higher than those determined in the French lagoons, the course of the experiment, the Eh decreased to
with minimum values ranging from 100 to 110 /Lmol highly negative values, mostly in slurries with O.M.
cm- 3 and maximum values of about 310 /Lmol cm- 3 added (Table 2). In these experiments, wide differ-
(Figure 4). AVS concentrations showed wide seasonal ences between sites was also found, with the most
variations, with maximum values in July. CRS fol- negative value at station 11 (-315 mY). In the case
lowed similar patterns with maximum values in the of station C 1, O.M. addition resulted in only a slight
warmer summer months and minimum concentrations decrease in Eh in comparison with the untreated slur-
in December (Figure 5). This clear seasonal trend sug- ry. Moreover, slurries from station Cl showed the least
gests that reduced sulphur pools are controlled by 'iron negative Eh values under the experimental conditions
availability' as well as reoxidation processes. Howev- tested. The differences indicated by the Eh values, con-
er, the parallel decrease of AVS and CRS in December form to those shown by trends of free sulphide, AVS
appears to show that most of the reduced sulphur is and SRP (Figure 6). For all these variables, differences
reoxidised, minimising sulphur burial through pyrite between treated and untreated slurries were in the range
deposition. of approximately one order of magnitude, except for
the free sulphide and SRP at station Cl, which did not
Sediment slurry incubation appears to be affected by O.M. addition.
Sulphide production increased exponentially in the
The sediment capacity to buffer free sulphide release treated slurry for station 11, where the final concen-
was tested by incubating sediment slurries both with tration of 4.3 /'Lmol cm- 3 was reached. Considerable
and without organic matter (O.M.) added. During release of sulphide was also detected in the treated slur-
218
O~~~~~~~~
_-12"---a _O' .. _o--~_ . _. . . . .
~~
o---a
~
0
,'~o,--.,.---d
0
o 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
hours hours
hours
,,
11 ...... - ".d
2000
,
"
I
I , ,
,, " 10 100
.. ......
I
1000 I
'
.. ......................
0 0 0
0 10 20 30 40 50 0 ;0 20 30 40 50 0 10 20 30 40 50
hours hou~s hours
.G "OGtfDIJ ~ ~ ~ a-a---------a
f) 0 -·r---. 0
{) 5 10 15 20 0 5 10 15 20 oj 5 10 15 20
hr;urs hours hours
Figure 7. Changes of free sulphide (nmol cm- 3 ), AVS (J.imol cm- 3 ) and SRP (nmol cm- 3 ) concentrations during the incubation of sediment
slurries of stations 7 (open symbols) and G (black symbols). X-axis: incubation duration (hours), Y-axis: concentrations.
ries from stations A (2.5 /-Lmol cm- 3 ) and B (0.6 /-Lmol Free sulphide and AVS release was coincident with
cm- 3 ). significant SRP-ftuxes, except for station CI, which
AVS concentrations followed an almost linear did not show appreciable release of SRP.
increase under all the conditions investigated, show- Preliminary experiments were also conducted in the
ing a clear gradient with maximum values in the treated Sacca di Goro by incubating slurries of sediment col-
slurry from station 11. lected from two distinct sites: the sandy station 7 and
the muddy station G. Changes of free sulphide, AVS
and SRP concentrations conformed to those observed
219
Table 2. Main characteristics of slurries without (-) and with (+) organic matter added at the end of
incubation period. Data were adjusted to 40 hours of incubation for stations A, B, C I and II and to 15
hours for stations G and 7. Final values of Eh and CRS mean value of the whole incubation period were
reported.
in a number of dystrophic lagoons which are charac- Figure 8. Relationship between AVS concentration (Jimol cm- 3 )
terized either by limited water renewal or abnormal and the AVS/Fe(II) molar ratio. The horizontal line refers to the
primary production. theoretical ratio of iron monosulphide.
was less than 0.5. The latter value might indicate that ble free sulphide fluxes_ A significant sulphide flux
the buffering capacity of ferrous iron was greatly in without relevant AVS accumulation was found in the
excess of sulphide production_ However, at stations treated slurry from station A, similar to that observed in
A and B, sulphide mobility as well as AVS precipi- the untreated slurry_ In the case of station 11, sulphide
tation may not be simply related to iron availability production greatly exceeded the sediment capacity to
since the effects of tidal currents are also involved. precipitate AVS, leading to an AVSlFe(II) ratio of 1.2
In fact, drying out of the sediment at low tide allows and to 4_14 /Lmol cm -3 of free sulphide concentration_
oxidation of the superficial sediment, thus enabling This wide range of AVS and free sulphide concen-
sulphide to be reoxidised. The recorded sediment pro- trations can be explained to some extent by consider-
files do not conform completely to those obtained with ing ferrous iron availability as well as organic detritus
sediment slurry incubations (Table 2). In the sediment quality. Detritus from Viva can undergo more rapid
slurries without a.M. addition, we found that station decomposition than that from Zostera and Ruppia, due
Cl accumulated AVS, with an AVSlFe(II) ratio =0_15 chiefly to the higher content of soluble carbohydrates
resembling those of the sediment profiles_ Unexpect- (Lahaye & Jegou, 1993; Viaroli et aI., 1996).
ed significant production of free sulphide was found With regard to the slurry experiments we found a
at station A. This was probably due to the high sul- clear relationship between AVS and SRP concentration
phate reduction rates (SRR) (Bourgues et aI., 1994) (Figure 9). This aspect may be of considerable signifi-
as well as sediment porosity, which would make free cance in the analysis of lagoon eutrophication, since it
sulphide stripping by the N2 gas stream easier. More- can be considered as a form of positive feedback due to
over, free sulphide removal might stimulate SRR, since significant SRP recycling. Inorganic phosphate is usu-
this is controlled by sulphide concentration. Organic ally assumed to be present in the sediment as iron and
matter addition (approximately 1-2% by final concen- calcium bound phosphate (Golterman, 1995). Howev-
tration) enhanced microbial processes as evidenced by er, for a given sediment, the distribution of SRP within
the abrupt decrease in Eh_ However, station C I still the inorganic sedimentary pool is mostly controlled
maintained a good capacity to control sulphide release, by the iron hydroxide-sulphide system (Davison and
with an almost balanced AVSlFe(II) ratio and negligi- Heany, 1978; De Groot, 1991; Golterman, 1995). This
221
seems to be the case for all the sites considered in this (Coord.), C.L.E.A.N. Progress Report, European Commission:
study, except for station Cl. At the latter site, SRP \09-132. European Community Environment Programme, DG
XII, Brussels.
concentrations were always negligible despite high Canfield, D. E., 1989. Reactive iron in marine sediments. Geochim.
AVS concentrations. Accordingly, we postulate that Cosmochim. Acta 53: 619-632.
there, SRP fluxes are mainly controlled either by clay Castel, J .. P. Caumette & R. Herbert, 1996. Eutrophication gradients
(Ponnamperuma, 1972; Golterman, 1984) or by the in coastal lagoons as exemplified by the Bassin d' Arcachon and
the Etang du Prevost. Hydrobiologia 329 (Dev. Hydrobiol. 117):
calcium-carbonate system (Golterman, 1984, 1995). xi-xxx.
The significant differences among treated and untreat- Cline, J. D., 1969. Spectrophotometric determination of hydrogen
ed slurries may also be dependent on direct SRP recy- sulphide in natural waters. Limnol. Oceanogr. 14: 454-459.
cling from decomposition of the added detritus. Coleman, M. L.. D. B. Hedrick, D. R. Lovley, D. C. White & K. Pye,
1993. Reduction of Fe(lII) in sediments by sulphate-reducing
Although this investigation was restricted to a short bacteria. Nature 361: 436-438.
period of observation, we can conclude that iron avail- Caumette, P.. 1986. Phototrophic sulphur bacteria and sulphate
ability and O.M. quality can be considered as key indi- reducing bacteria causing red waters in a shallow brackish coastal
cators to predict the ability of a given lagoon to buffer lagoon (Prevost Lagoon. France). FEMS Microbiol. Ecol. 38:
113-124.
against dystrophic events. For example, the dystrophic Davison, W. & S. I. Heany. 1978. Ferrous iron-sulphide interactions
Etang du Prevost suffers from a severe iron limita- in anoxic hypolimnetic waters. Limnol. Oceanogr. 23: 1194-
tion and receives a high input of readily decompos- 1200.
De Casabianca. M. L., 1993. Macronutrients relative evolution in
able detritus from recurring 'green tides' (Ulvacean
Mediterranean macrophyte system subject to dystrophic crises. In
blooms). In contrast, dystrophic events are unusual in J. W. Rijstenbil & S. Haritonidis (eds). Macroalgae. eutrophica-
the sheltered impoundments of the Bassin d' Arcachon, tion and trace metal cycling in estuaries and lagoons. Proceedings
where iron concentrations are 2-3 times higher and of the COST-48 Symposium of Sub Group III, Thessaloniki 24-
26 September. EU-BRIDGE. 31-34.
Ulvacean blooms have a minor impact. However, to De Groot. C. J .• 1991. The influence of FeS on the inorganic phos-
what extent these measurements enable us to predict phate system in sediments. Verh. int. Ver. Limnol. 24: 3029-3035.
more accurately dystrophic events remains to be seen. De Groot. C. J. & H. L. Goiterman, 1993. On the presence of
organic phosphate in some Camargue sediments: evidence for
the impOitance on phytate. Hydrobiologia 252: 117-126.
Finster, K .. A. Gammerlaad & L. S. Hansen, 1994. Sulphate reduc-
Acknowledgements tion rates, inorganic sulphur pools and the accumulation of
volatile fatty acid in the Bassin d' Arcachon and the Prevost
We are indebted to Mr Tony Zappata (Battello Hydra, lagoon, southern France: a comparative study. In P. Caumette
(Coord.), C.L.E.A.N. Progress Report, European Commission:
Amministrazione Provinciale di Ferrara, Assessorato 223-252. European Community Environment Programme, DG
Ambiente) and Mr Vittorio Gaiani (Dipartimento di XII. Brussels.
Biologia, Dniversita di Ferrara) for the support during Fossing, H. & B. B. J~rgensen, 1989. Measurement of bacterial sul-
sampling in the Sacca di Goro. This study was financed phate reduction in sediment: evaluation of a single-step chromium
reduction method. Biogeochemistry 8: 205-222.
by the joint ED project on 'Coastal Lagoon Eutroph- Giblin, A. E. & R. K. Wieder. 1992. Sulphur cycling in marine and
ication and Anaerobic processes (CLEAN)', contract freshwater wetlands. In R. W. Howarth. J. W. B. Stewart & M.
number EV5V-CT92-0080. U. Ivanov (eds), Sulphur cycling on the continents: wetlands,
terrestrial ecosystems and associated water bodies. SCOPE 33, J.
Wiley & Sons. New York: 85-117.
Golterman, H. L.. 1984. Sediments, modifying and equilibrating
References factors in the chemistry of freshwater. Verh. int. Ver. Limnol. 22:
23-59.
A.P.H.A., 1975. Standard Methods For the Examination of Water Goiterman, H. L.. 1995. The role of the iron hydroxide-phosphate-
and Wastewater. 14th Edition. APHA, Washington, 1193 pp. sulphide system in the phosphate exchange between sediments
Aspila, K. I., H. Agemian & A. S. Y. Chau, 1976. A semiautomat- and water. Hydrobiologia 297: 43-54.
ed method for the determination of inorganic, organic and total Howarth, R. W. & J. W. B. Stewart, 1992. The interactions of sulphur
phosphate in sediments. Analyst 101: 187-197. with other element cycles in ecosystems. In R. W. Howarth. J.
Bartoli, M., M. Cattadori, G. Giordani & P. Viaroli, 1996. Benthic W. B. Stewart & M. U. Ivanov (eds), Sulphur cycling on the
oxygen respiration, ammonium and phosphorus regeneration in continents: wetlands, terrestrial ecosystems and associated water
surficial sediments of the Sacca di Goro (Northern Italy) and two bodies. SCOPE 33, J. Wiley & Sons. New York: 67-84.
French coastal lagoons: a comparative study. Hydrobiologia 329 Izzo. G. & V. Hull, 1991. The anoxic crises in dystrophic processes
(Dev. Hydrobiol. 117): 143-159. of coastal lagoons: an energetic explanation. In C. Rossi & E.
Bourgues, S., O. Pringault & R. de Wit, 1994. Spatial and temporal Tiezzi (eds), Ecological Physical Chemistry. Elsevier Sci. Pbls.
variation of oxygen uptake, sulphide production and population Amsterdam: 559-572.
of sulphur cycle bacteria at the CLEAN stations. In P. Caumette
222
Jensen, H. S. & B. Thamdrup, 1993. Iron-bound phosphorus in Sfriso, A., B. Pavoni, A. Marcomini & A. A. Orio, 1992. Macroal-
marine sediments as measured by bicarbonate-dithionite extrac- gae, nutrient cycles, and pollutants in the lagoon of Venice. Estu-
tion. Hydrobiologia 253: 47-59. aries IS: 517-528.
J(.1rgensen, B. B., 1983. Processes at the water-sediment interface. In Stal, L. J., F. Kruying, S. B. Behrens & M. Vilbrant, 1994. Spa-
B. Bolin & R. B. Cook (eds), The major biogeochemical cycles tial and seasonal variations of iron, phosphate and sulphide in
and their interactions. SCOPE 21, J. Wiley & Sons, New York: the sediments of two eutrophic lagoon. In P. Caumette (Coord.),
477-509. C.L.E.A.N. Progress Report, European Commission: 167-186.
Koch, M. S., I. A. Mendelssohn & K. L. McKee, 1990. Mechanism European Community Environment Programme, DG XII, Brus-
for the hydrogen sulphide growth limitation in wetland macro- sels.
phytes. Limno!. Oceanogr. 35: 399-408. Valderrama, J. c., 1977. Methods used by the Hydrographic Depart-
Koroleff, F., 1970. Direct determination of ammonia in natural ment of National Board of Fisheries, Sweden. In K. Grasshoff
waters as indophenol blue. Information on techniques and meth- (ed.), Report of the Baltic Intercalibration Workshop. Annex,
ods for seawater analysis. I.C.E.S. Interlaboratory Rep. No.3: Interim Commission for the Protection of the Environment of the
19-22. Baltic Sea: 14-34.
Labourg, P. J., 1975. Contribution Ii I' hydrologie des etang saumiitres Viaroli, P., A. Pugnetti & I. Ferrari, 1992. Viva rigida growth and
de la region d' Arcachon: description des phenomenes d'eaux decomposition processes and related effects on nitrogen and phos-
blanches. Bull. soc. linn. Bordeaux 5: 3-8. phorus cycles in a coastal lagoon (Sacca di Goro, Po River Delta).
Lahaye, M. & D. Jegou, 1993. Chemical and physical-chemical In G. Colombo, I. Ferrari, V. U. Ceccherelli and R. Rossi (eds),
characteristics of dietary fibres from Viva lactuca (L.) Thuret Marine eutrophication and population dynamics. Olsen & Olsen,
and Enteromorpha compressa (L.) Grev. 1. App!. Phyco!. 5: 195- Fredensborg: 77-84.
200. Viaroli, P., M. Bartoli, C. Bondavalli, R. R. Christian, G. Giordani
Ponnamperuma, F. N., 1972. The chemistry of submerged soils. Adv. & M. Naldi, 1996. Macrophytes communities and their impact
Agron. 24: 29-96. on benthic fluxes of oxygen, sulphide and nutrients in shallow
eutrophic environments. Hydrobiologia 329 (Dev. Hydrobio!.
117): 105-119.
Hydrobiologia 329: 223-225, 1996. 223
P. Caumette, J. Castel & R. Herbert (eds), Coastal Lagoon Eutrophication and Anaerobic Processes (C.L.E.AN.).
Subject index
Page numbers refer to the ftrst page of an article in which the entry is discussed. As regards biological
organisms, only genera and, when appropriate, higher taxonomic categories are indexed.
Thiobacillus, ix
Total kjeldahl nitro gene, 93
Trichodesmium, 185
Trophic relationship, ix, 79