International Congress Series 1275 (2004) 209 – 217

www.ics-elsevier.com

Ontogenetic patterns in thermal adaptation of fish

vs. long-term temperature trends in large rivers
F. Schiemera,*, H. Keckeisa, H. Nemeschkalb, E. Schludermanna,
G. Winklera, I. Zweimqllera
a
Department of Limnology, Institute of Ecology and Conservation Biology, University of Vienna,
Althanstrasse 14, A-1090 Vienna, Austria
b
Department of Systematic Zoology and Developmental History, Institute of Zoology, University of Vienna,
Althanstrasse 14, A-1090 Vienna, Austria

Abstract. Rheophilic fish species characteristic of large river systems are finely tuned during early
ontogenetic development to the seasonal temperature regime of their main nursery habitats in the
inshore zones of rivers. River regulation and the construction of hydropower dams have disrupted the
balance between requirements and the field conditions. Frequently, temperatures fall below the
optimal range in regulated rivers due to faster runoff and reduced inshore retention capacity. Under
such conditions, growth through the bcritical periodQ during the life history is retarded resulting in
high mortality rates. This is a main reason for the decline of stocks and a critical state of recruitment of
many riverine species. The trends caused by long-term climatic changes have to be similarly
addressed from the point of view of match or mismatch between requirements and the predicted
changes in environmental conditions. Long-term trends in the temperature development have been
analyzed for the Austrian Danube. The potential effect of such trends is tested against the temperature
dependence of embryogenesis and a growth model developed for one of our target species, the
cyprinid Chondrostoma nasus. Field data on individual daily growth rates of larval fish, based on
otolith analysis, show that the present conditions are critical. The consequences of global change for
rheophilic fish species in large river systems are discussed. D 2004 Elsevier B.V. All rights reserved.

Keywords: Fish larvae; Critical stage; Ontogenetic niche; Differentiation; Energetics; Efficiencies; Ecophysiology;
Growth model; Global change; Match–mismatch

* Corresponding author. Tel.: +43 1 4277 54340; fax: +43 1 4277 9542.
E-mail address: friedrich.schiemer@univie.ac.at (F. Schiemer).

0531-5131/ D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.ics.2004.09.039
210 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217

1. Introduction
Along the river course, characteristic fish associations are adapted to prevailing
temperature regimes. The longitudinal thermal profile of rivers resembles to some extent
the latitudinal geographical pattern. Local success of species is dependent on the match or
mismatch between requirements and field conditions during the early life history, which is
the critical phase, the bottleneck, deciding on success or failure of a species. The fine
tuning between requirements and environmental conditions has been considerably
disrupted by river engineering (dams, reduction of inshore retention [1]). Global warming
is a further factor which has to be taken into account.
The study analyzes (a) the thermal ranges and ontogenetic niche shifts of
characteristic species during the early life history, (b) its optima and constraints with
regard to temperature, (c) identifies long-term trends in the temperature development of
the Danube river and (d) analyzes the potential effects of such trends on characteristic
fish species.

2. The thermal niche of riverine fish species

Three fish species are representative for the longitudinal distribution of fish in
European rivers. Brown trout (Salmo trutta) is characteristic for the upper rithral zone.
The nase (Chondrostoma nasus) is a cyprinid, typical for large and structured river
systems and a representative of the barbel-zone [2,3]. The roach (Rutilus rutilus), again
a cyprinid, is a eurytopic species occurring in the upper and lower potamal stretches and
different types of stagnant water bodies [4]. A comparison of the thermal niche
dimensions of the three species with respect to the position in the thermal gradient, the
optimum and tolerance range during embryogenesis (until hatching) and the early larval
(nase and roach) or juvenile phase (brown trout) shows that the viable range for
embryogenesis is similar in all three species (in the order of 9–10 8C) and widens with

Fig. 1. The viable (white) and the optimal temperature range (grey) of embryogenesis (lower bar) and larval
development (upper bar) of three species characteristic for different fish regions along the longitudinal range of
rivers. The upper insert presents a sketch on the hydromorphological and biocoentic zonation, rithral and potamal
of rivers as physiographically determined by a.o. current velocity pattern, substrate. Data on thermal ranges
according to: S. trutta, embryogenesis [5]; early juvenile period [6]. C. nasus, embryogenesis [7]; larval period
[8]. R. rutilus, embryogenesis [9], larval period [10]. The photo shows a characteristic nursery zone for rheophilic
fish in the Danube and a larval nase.
 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217 211

stage and size. There is a distinct shift in optimal temperatures from egg development to
the later stages which is more expressed in the two riverine species trout and nase than
in roach and which apparently reflects the rise of temperature after the spawning periods
in rivers (Fig. 1).

3. Microhabitat requirements and thermal niche shifts in the nase, C. nasus

C. nasus has become a flagship species for the ecology and conservation of large
European rivers. The thermal requirements of the species are closely linked to conditions
of larger and structurally rich river sections. A clear habitat shift in the course of the life
cycle is apparent: rithral gravel bars are required for spawning. The spawning period
occurs characteristically in April at river temperatures between 8 and 12 8C [11]. As larval
microhabitats richly structured inshore zones of rivers are required which offer a
combination of low current velocity, high production of zooplankton as larval food and
refuge possibilities in the case of floods. With increasing size, the preferred microhabitats
shift towards more lotic conditions.
Detailed experimental studies have been carried out on the early life history stages
recognizing that they are critical for recruitment. Fig. 2 provides information on the
temperature dependence of egg development [7]. Embryogenesis under constant
conditions takes place over a range between 10 and 19 8C. Mortality increases strikingly
below and above this range. It is of significance to note that temperature not only
determines duration but also changes the pattern of development: at higher values the
hatching of the embryos, the up-swimming of larvae and the onset of exogenous feeding
occur at a distinctly smaller body size (tissue weight), at a less advanced stage of lower
morphological differentiation but with a higher remaining yolk reserve (Fig. 2b). This
strategic shift indicating trade-offs between size, duration and the conservation of energy
reserves is schematically outlined in Fig. 2c.

Fig. 2. Temperature dependence of (a) duration of embryogenesis, full line: fertilization until hatching, broken
line: hatching until onset of feeding; (b) size of larvae at hatching (grey) and the remaining yolk reserve (white)
and (c) temperature induced shifts between size, duration and yolk reserve at hatching.
212 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217

Fig. 3. (a) Temperature dependence of duration of embryogenesis from fertilization until hatch and from hatch to
the end of the larval period. The rectangles delineate the temperature window (range of viable temperatures, range
of low mortality), the shaded zone indicates the optimal ranges (see text). (b) Monthly mean temperature (black
point) and range for April, May and June in a cold year (upper line), respectively a warm year (middle line) from
the series of 1951–1995. The lower line represents the prognosis for 2050 based on the trend analysis. Inserted is
the optimal temperature range for embryogenesis (April, May) and the larval period (June).

Fig. 3a illustrates the temperature dependence of stage duration of embryogenesis and

the larval period. It clearly shows the shift in the viable and the optimum range in the
consecutive stages, which occurs within a short period of time and developmental phase.
A distinct succession of thermal adaptations is evident. This shift is very pronounced—in
the order of 8–10 8C in the first phase of ontogeny—and occurs over a short period of time
(duration) in the order of days.

Fig. 4. Growth rates of three ontogenetic phases of C. nasus grown during larval phase over a wide range from 10
to 28 8C under constant temperatures and ad libitum food supply (for methodological details, see Refs. [7,8]).
Since embryogenesis is limited to a temperature range between 10 and 19 8C, larval growth at higher temperatures
were studied from larvae hatched at 16 8C.
 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217 213

Table 1
Size dependent growth rates in dependence of temperature, calculated as the allometric function G=aW b
A B
15 8C 0.0997 0.298
20 8C 0.1994 0.249
25 8C 0.3003 0.236

Temperature dependence of larval growth rates has been experimentally studied over a
wide experimental range from 10 to 28 8C. A Gompertz type model of the form:
   
FðT;DÞ ¼ Ac exp eaT Dc ¼ Ac exp eaT DclnD
could be fitted over the whole temperature range applied to the experimental data for the
larval and juvenile period in terms of body length and dry weight (dwt) by polynomial
regression with temperature and age (duration) as the independent variables.
Fig. 4 compares specific growth rates during embryogenesis, the larval and the early
juvenile period. Growth rates are particularly high during the egg development phase
(i.e. the conversion from yolk into tissue, data from Kamler et al. [7]) and decrease with
size and stage. The size dependence between the onset of exogenous feeding (1–5 mg
dwt) to the juvenile stage (over 100 mg dwt) at different temperatures are given in
Table 1. The figure also illustrates both the shift and an expansion of the thermal niche
from the larval to the early juvenile period by an increase of Q 10-values for growth with
size (Table 2).
It is well established that within the early life history the period from onset of feeding
until full yolk resorption of larvae (mixed feeding period) and the larval period of entirely
exogenous feeding is the most critical [12–14]. The morphological differentiation related
to the sensory, feeding, swimming and respiratory functions has to occur fast. Considering
the low metabolic scope and the high and competitive requirements for growth,
differentiation and maintenance, the temperature range providing for optimal fitness has
to be defined by a combination of characteristics:

! energetics, both power and efficiencies, determine the scope for growth and
development,
! the attained body size of characteristic stages are significant with regard to the
swimming, escape and feeding performances and determine the bpoint of no returnQ,
! duration, the time required to reach a certain stage relates to the cumulative risks larval
fish are exposed to in a stochastic nursery microhabitat.

There is indication that within the viable temperature range the various functions are
accelerated with temperature at different speeds [15] leading to a set of different solutions.

Table 2
Q 10-values of growth over the range of linear increase (13–25 8C)
Size class (mg dry weight) 1–5 5–30 30–50 50–80 80–200
Q 10 3.34 3.30 3.68 3.84 4.41
214 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217

Fig. 5. (a) Temperature dependence of size of larvae at the onset of exogenous feeding (white) and at the end of
the larval phase. Note that for hatching larvae and finite larval size different scales are used. (b) Assimilation
efficiency (A/C in percent): solid lines; production efficiency ( P/A): broken lines. Embryogenesis (data from
Refs. [1,7]): thin lines; exogenously feeding larvae [1,8]: thick lines.

At present, we cannot answer the question if they are the result of trade-offs and
constraints or represent an adaptive strategy in the sense that different solutions provide an
array of chances under stochastic environmental conditions with frequently occurring
suboptimal conditions (temperature, food, oxygen supply, water velocity), which will have
cumulative effects.
In order to define the boptimum rangeQ, Fig. 5 compares two major fitness qualities
during early ontogeny and its dependence from temperature. The upper panel compares
body size at hatching (white) and the finite larval size (grey; note the different scale). At
higher temperatures, both of these endpoints are attained at smaller body mass. The lower
panel compares bioenergetic efficiencies: it is interesting to note that temperature
dependence of assimilation efficiency (A/C 100) decreases in the upper viable range
during both phases while production efficiency ( P/C 100) increases. This comparison
illustrates that optimality cannot be narrowly defined.

4. Merging field ecology, global trends, bioenergetics and conservation

How do these requirements match with conditions larval fish encounter in the field?
In order to answer this question, we have to assess the relevant environmental
conditions and their spatiotemporal variability in the larval microhabitats in the inshore
zones of rivers [3]. Richly structured inshore zones are the main nurseries of riverine fish.
They are characterized by higher temperatures and a higher plankton (=food) production
compared to the main stem of the river, depending on the water retention (bInshore
Retention ConceptQ: [16]).
 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217 215

Fig. 6 provides an example of a detailed study on early larval growth in an inshore

location with high water retention of the free-flowing Danube downstream of Vienna in
Austria with comparatively good growth conditions. The individual growth pattern can be
traced-back from daily ring formations using the otolith microstructure. The growth
trajectories of two larvae show that the growth rates are not synchronized. The grey band
represents the growth variability encountered within a group of larvae analyzed from a
particular catch. The overall slope of the growth band of the cohort can be compared with
continuous temperature recordings in the nursery microhabitat. Inserted are the continuous
temperature recordings in the nursery microhabitat.
The overall slope of the growth band can be compared with the temperature dependent
growth rates in terms of larval length found in the experimental cultures. It becomes
apparent that suboptimal conditions are decisive for low growth rates in the field leading to
prolonged risks in a stochastic environment.
How critical are the present temperature conditions and what are the long-term
trends?
The trends illustrated in Fig. 7 are analyzed for the conditions in high retention inshore
zones. It has to be emphasized that the observed trends not only represent global climatic
changes but that they are additionally influenced by other impacts on the river system
(hydropower dams, connectivity of the river–floodplain system).

Fig. 6. (a) Early larval growth in an inshore location of the Danube (see text) back-calculated from daily
increments of the otoliths of larvae collected at a specific date. Two individual growth trajectories (solid line) and
the width of the bgrowth bandQ (shaded). Broken line: continuous temperature recordings in the nursery habitat.
(b) Experimental growth data compared to field growth.
216 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217

Fig. 7. (a) Long-term trends (1951–1996) of inshore temperature regime calculated from the daily readings and
the divergence pattern of inshore temperatures at highly structured inshore locations in the free-flowing Austrian
Danube downstream of Vienna. The trends (linear regressions) are given for the April, May and June when most
spawning activities and early fish development takes place. (b) Seasonal pattern of temperature changes in fish
nursery habitats calculated for a 10-year period.

In order to outline the match–mismatch between the physiological requirements and the
field situation in a high quality nursery situation, the temperature dependence of stage
duration are set in comparison to the conditions for April, May and June, when early life
history takes place in the field. Monthly means and the range of daily values have been
selected for a year with low and high temperatures from the series from 1951 to 1995. The
third line represents the prognosis for 2050 based on the trend analysis. The graph clearly
demonstrates that present temperature conditions are suboptimal and embryonic develop-
ment is retarded. The same holds good for the larval phase. River engineering, the
straightening of the river course and the loss of inshore retention zones have reduced the
overall spring temperatures and a disruption in the synchronization between the
physiological program of riverine species and the environmental temperature regime. At
a first glance, it would mean that over a time scale of approximately 100 years global
warming could improve the situation and counteract the deficiencies caused by
engineering. Even the predicted increase in temperature is outside the optimal range.
On the other hand, restoration programs are presently planned and implemented (e.g. for
the Danube) in order to return to more original conditions. This should additionally speed
up the temperature increase in large rivers over the next decades leading to the risks of
overshooting the narrow physiological windows.

5. Conclusions
Distinct ontogenetic niche shifts in the thermal adaptation of larval stages of riverine
fish are finely tuned towards the stage-specific microhabitat conditions.
Within a certain range of morphological differentiation, growth and metabolic
expenditures are accelerated with temperature at different speeds leading to a set of
different end-points in character combination.
River engineering and global warming are disrupting the tuning between physio-
logical requirements and field conditions, leading to an endangerment of characteristic
species.
Programs defined to analyze their ecophysiological requirements are essential for
guiding management, restoration and conservation.
 F. Schiemer et al. / International Congress Series 1275 (2004) 209–217 217

References

[1] F. Schiemer, H. Keckeis, E. Kamler, The early life history stages of riverine fish: ecophysiological and
environmental bottlenecks, CBP, A 133 (2003) 439 – 449.
[2] M. Penaz, Chondrostoma nasus—its reproduction strategy and possible reasons for a widely observed
population decline—a review, in: A. Kirchhofer, D. Hefti (Eds.), Conservation of Endangered Freshwater
Fish in Europe, Birkh7user Verlag, Bern, 1996, pp. 279 – 285.
[3] F. Schiemer, Fish as indicators for the assessment of the ecological integrity of large rivers. Proceedings on a
workshop on ecological integrity in Vienna, Hydrobiologia 422/423 (2000) 271 – 278.
[4] F. Schiemer, W. Wieser, Epilogue: food and feeding, ecomorphology, energy assimilation and conversion
ion cyprinids, Environ. Biol. Fishes 33 (1992) 223 – 227.
[5] A.F. Ojanguren, F. Brana, Thermal dependence of embryonic growth and development in brown trout,
J. Fish Biol. 62 (2003) 580 – 590.
[6] A.F. Ojanguren, F.G. Reyes-Gavilan, F. Brana, Thermal sensitivity of growth, food intake and activity of
juvenile brown trout, J. Therm. Biol. 26 (2001) 165 – 170.
[7] E. Kamler, H. Keckeis, E. Bauer-Nemeschkal, Temperature-induced changes of survival, development and
yolk-partitioning in Chondrostoma nasus, J. Fish Biol. 53 (1998) 658 – 682.
[8] H. Keckeis, et al., Survival, development and food energy partitioning of nase larvae and early juveniles at
different temperatures, J. Fish Biol. 59 (2001) 45 – 61.
[9] W. Wieser, et al., Growth rates and growth efficiencies in larvae and juveniles of Rutilus rutilus and other
cyprinid species: effects of temperature and food in the laboratory and in the field, Can. J. Fish. Aquat. Sci.
45 (1988) 943 – 950.
[10] W.M. Mooij, O.F.R. Van Tongeren, Growth of 0+ roach (Rutilus rutilus) in relation to temperature and size
in a shallow eutrophic lake: comparison of field and laboratory observations, Can. J. Fish. Aquat. Sci. 47
(1990) 960 – 967.
[11] H. Keckeis, Influence of river morphology and current velocity conditions on spawning site selection of
Chondrostoma nasus (L.), Arch. Hydrobiol., Suppl. 113 Large Rivers 10 (2001) 51 – 64.
[12] J.H.S. Blaxter, The effect of temperature on larval fishes, Neth. J. Zool. 42 (1992) 336 – 357.
[13] E. Kamler, Early Life History of Fish. An Energetics Approach, Chapman & Hall, 1992, 267 pp.
[14] W. Wieser, Energetics of fish larvae, the smallest vertebrates, Acta Physiol. Scand. 154 (1995) 279 – 290.
[15] C.C. Lindsay, A.N. Arnason, A model for responses of vertebral numbers in fish to environmental
influences during development, Can. J. Fish. Aquat. Sci. 42 (1981) 31 – 38.
[16] F. Schiemer, et al., The binshore retention conceptQ and its significance for large rivers, Arch. Hydrobiol.,
Suppl. 135 (2001) 509 – 516.