SEARCH

SEARCH BY CITATION

Keywords:

  • Colpidium striatum;
  • Didinium nasutum;
  • ecosystem functioning;
  • indirect and direct temperature effects;
  • trophic interactions;
  • viscosity

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

1. While much is known about the direct effect that temperature can have on aquatic communities, less is known about its indirect effect via the temperature dependence of viscosity and temperature-dependent trophic interactions.

2. We manipulated the temperature (5–20 °C) and the viscosity (equivalent to 5–20 °C) of water in laboratory-based bacteria–protist communities. Communities contained food chains with one, two or three trophic levels. Responses measured were population dynamics (consumer carrying capacity and growth rate, average species population density, and the coefficient of variation of population density through time) and ecosystem function (decomposition).

3. Temperature, viscosity and food chain length produced significant responses in population dynamics. Temperature-dependent viscosity had a significant effect on the carrying capacity and growth rates of consumers, as well as the average density of the top predator. Overall, indirect effects of temperature via changes in viscosity were subtle in comparison to the indirect effect of temperature via trophic interactions.

4. Our results highlight the importance of direct and indirect effects of temperature, mediated through trophic interactions and physical changes in the environment, both for population dynamics and ecosystem processes. Future mechanistic modelling of effects of environmental change on species will benefit from distinguishing the different mechanisms of the overall effect of temperature.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

For poikilotherms, internal body temperature largely depends on environmental temperature, which in aquatic systems is highly variable. Many key processes such as metabolic rate (Clarke 1991; Gillooly et al. 2001; Brown et al. 2004; Savage et al. 2004; Apple, del Giorgi & Kemp 2006), cilial and flagellar activity (Sleigh 1956) and digestion/food processing rates (Yee & Murray 2004), can be directly mediated via environmental temperature. Consequently, behaviours such as terrestrial isopod (Porcellio laevis) walking speed (Dailey et al. 2009) or mosquito fish (Gambusia holbrooki) swimming speed (Wilson 2005) as well as ecological processes [e.g. population growth rate (Rose & Caron 2007)] depend directly upon environmental temperature. Temperature-dependent feeding rates and therefore temperature-dependent trophic interactions (Ives 1995) can also lead to indirect effects of temperature on communities and ecosystem functioning (Beveridge, Humphries & Petchey 2010).

Temperature also alters the physical properties of an organism’s environment, e.g. fluid (i.e. air or water) density or viscosity. The physical properties of water are particularly temperature-dependent, the most notable being viscosity. A change in temperature from 20 to 5 °C is associated with an increase in viscosity (here we refer to kinematic viscosity, ν, which accounts for effects of fluid density) from 1 × 10−6 to 1 × 52−6 m2 s−1 (Vogel 1988; Denny 1993). The impact of a change in viscosity on an organism is mostly dependent upon its size and speed, with small slow-moving organisms being most affected. For example, Fuiman & Batty (1997) observed a 60% reduction in fish swimming speed when water temperature was reduced from 13 to 6 °C. For small fish, ≈43% of this reduction in swimming speed was attributed to a change in viscosity, with larger fish not significantly affected by a change in viscosity at 6 °C.

Viscosity-induced changes in swimming speed may contribute strongly to temperature-dependent trophic interactions. Encounter rates between organisms have been shown to be proportional to the square of an organism’s speed (Visser & Thygesen 2003; Visser & Kiorboe 2006; Kiorboe 2008). Viscosity-induced reductions in encounter rates between predators and prey, via changes in swimming speed, may reduce predator attack rates. Viscosity-induced reductions in feeding rate of small aquatic organisms have been previously observed (Podolsky, 1994; Bolton & Havenhand 2005; Abrusan 2004; Bolton & Havenhand 1998). Beveridge, Humphries & Petchey (2010) demonstrated that a reduction in temperature reduces predator-protist impact on consumer protists, and consumer-protist impact on resources. As viscosity negatively covaries with temperature and an increase in viscosity reduces feeding rates, one may predict that temperature-dependent viscosity accounts for a proportion of the reduction in predator/consumer impact associated with a decrease in temperature. Thus, temperature-dependent trophic interactions may in part be governed by temperature-dependent viscosity.

Recent modelling suggests that changes in species interactions, associated with an increase in temperature, decrease both population and community stability (Vasseur & McCann 2005). Furthermore, a decrease in viscosity has been observed (Luckinbill 1973) and modelled (Harrison 1995) to increase population stability of predator (Didinium nasutum) and prey (Paramecium aurelia) communities. However, the degree to which temperature-dependent viscosity contributes to a community’s overall response to temperature is poorly understood.

The aim of this study was to quantify whether viscosity significantly contributes to the temperature dependency of aquatic microbial population dynamics, via viscosity-induced alterations in trophic interactions. Bacteria–protist microcosms are ideal systems to quantify the importance of viscosity. The swimming speed response of two protists, Didinium nasutum and Colpidium striatum, has already been quantified and modelled, with temperature-dependent viscosity accounting for ≈20% of temperature-dependent swimming speed (Beveridge, Petchey & Humphries, unpublished). Additionally, the swimming speed of bacteria has also been observed as negatively or unimodaly viscosity-dependent (e.g. Schneider & Doestsch 1974; Greenberg & Canale-Parola 1977). Due to their small size and active swimming, ciliate protozoa are likely to be highly susceptible to changes in viscosity. Bacteria–protist microcosms can be readily manipulated, in terms of both environmental conditions and community composition. Generation times are short (≤1 day), thus allowing long-term population dynamic effects (over multiple generations) of temperature and viscosity to be quantified in a relatively short period of time. The ecosystem level consequences of organismal responses can be readily evaluated by measuring bacterial decomposition rate in these systems. Due to the previously discussed negative impact of viscosity on aquatic feeding rates, we predict that an increase in viscosity (while maintaining a constant temperature) will increase the average population density of consumer protists or bacterial resource when cultured in the presence of their respective predators or consumers.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Species culturing and experimental design

Experiments were conducted using bacteria–protist microcosms. Microcosms consisted of a 200-mL glass jar, containing 100 mL of Chalkley’s media (Tompkins et al. 1995), 0·55 g L−1 dissolved protist pellet (Carolina Protozoa pellets™, Burlington, NC, USA), with a foil lid (to allow gaseous diffusion but prevent contamination). Two wheat grains (of known dry mass) were added to provide a long-term energy source and allow an estimate of decomposition. All apparatus were sterilized prior to experimentation, and standard aseptic techniques were used throughout (Kemp et al. 1993).

Food chain length was manipulated from 0 (to assay abiotic mass loss of wheat grains) to 3. The first trophic level (resource) was a bacterial fauna, consisting of Bacillus cereus (Frankland and Frankland), Bacillus subtilis (Ehrenberg) and Serratia marcescens (Bizio). The second trophic level (consumer) was an obligate bacterivore protist, Colpidium striatum (Stokes). The third trophic level (predator) was Didinium nasutum (Muller), which in this system is an obligate predator of Colpidium. Inoculation protocol was: day 0, 200 μL bacterial fauna inocula (density assayed prior to inoculation); day 2, c. 200 Colpidium cells (stock cultures were assayed prior to inoculation, and volume of inocula estimated to obtain 200 cells); days 9 and 16, five Didinium were added via a direct count and micropipette transfer. This inoculation protocol ensured that consumers and predators had adequate sources of resource and prey prior to inoculation. The total duration of experimentation and microcosm culturing was 6 weeks.

The environment treatment had three levels: warm normal viscosity (temperature = 20 °C, viscosity = 1 × 10−6 m2 s−1), cold normal viscosity (temperature = 5 °C, viscosity = 1·52 × 10−6 m2 s−1) and warm adjusted viscosity (temperature = 20 °C, viscosity = 1·52 × 10−6 m2 s−1). Temperature was manipulated by placing microcosms in incubators (±0·5 °C; Qualicool-500, Oldham, UK). Viscosity was manipulated independently of temperature by dissolving 25 g mL−1 Ficoll (Type 400; Sigma-Aldrich) into the standard media. Ficoll is atoxic with regards to protists (Appendix S1, Supporting Information), readily soluble, maintains Newtonian fluid properties when dissolved in water and is not a food source for bacteria (Winet 1976; Berg & Turner 1979; Bolton & Havenhand 2005; Abrusan 2004; Loiterton et al. 2004). Viscosity was measured using an Ubbelohde-type viscometer (calibrated Cannon C457). All experimental treatments were conducted factorially (12 treatment combinations with 20 replicates per treatment combination).

Sampling

Bacterial densities were assayed via serial dilutions and agar plating (Mikrobiologie Nutrient Agar). Plates were cultured for 48 h at room temperature (≈22·5 °C) before colonies were counted. Bacteria identification was not possible by visual inspection. Bacterial densities were sampled once a week from day 1 to 42. Colpidium and Didinium were assayed by direct counting of aliquots under a stereo dissecting microscope (Nikon SMZ1000). In some communities, the density of Colpidium was high and required dilution prior to counting. Occasionally, aliquots of microcosms were observed to contain no Colpidium cells. However, these observations were succeeded by a positive count of Colpidium. Thus, such observations were treated as zero counts, rather than extinctions. Colpidium density, in communities where food chain length was 2, was sampled daily from day 1 to 9 (to allow an estimate of intrinsic growth rate). For the remaining communities and after day 9, Colpidium density was sampled once a week from day 16 to 37. Didinium density was sampled weekly from day 16 to 37. Decomposition rate was measured as the proportion dry mass loss of wheat grains from day 0 to 42.

Statistical analysis

All statistical analyses were conducted using the open source statistical software r (R Development Core Team, 2007). All data, unless otherwise stated, were found to be normally distributed and conform to the assumptions of the parametric analyses used.

The response variables considered were average population density (of each trophic level), intrinsic growth rate and carrying capacity (of Colpidium), the coefficient of variation (CV) of population density over time, and the proportional mass loss of wheat grains. An average population density was calculated for each microcosm after day 14 for each trophic level (data prior to day 14 were not used as species were either absent or populations were undergoing a growth phase). CV was used as a measure of community variability (McArdle, Gaston & Lawton 1990; Gaston & McArdle 1994). CV was estimated per microcosm, and provides an estimate of the variability of population density between observations within each microcosm. As with average population abundance CV was calculated per trophic level from data post day 14. Carrying capacity and growth rate were estimated via a nonlinear least squares (nls) fitting of the logistic growth model (eqn 1) to Colpidium population density (d) data from microcosms of chain length = 2. Estimates of growth rate (r) and carrying capacity (k) for Colpidium were made separately for each of the 60 microcosms containing only two trophic levels. Zero count data were removed before estimating r and k to allow the model to find a solution.

  • image(eqn 1)

A two-way analysis of variance (anova), with both environment (water temperature and viscosity) and food chain length (0, 1, 2 and 3) as explanatory categorical variables, was used to analyse decomposition (arcsine square-root transformed proportional dry mass loss), average density (log10 transformed) and CV of bacteria and Colpidium. Didinium average population density and CV, as well as Colpidium r and k were analysed with a one-way anova because of the lack of a food chain length treatment. Where a significant anova treatment effect was found, Tukey’s HSD tests were used to identify which pairs of treatment levels differed significant (Crawley 2007). If there were no significant pairwise differences between two levels of a treatment, these levels were merged into one level to provide a simplified explanatory variable. The explanatory power of this simplified explanatory variable was compared to the original explanatory variable by F-test (Crawley 2007). This provides a more sensitive test of the explanatory importance of the treatment levels than Tukey test.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Average population density

Environment and food chain length produced a significant interaction response in average bacterial density (Table 1, Fig. 1c). Post hoc tests indicate that bacterial density in the warm normal and warm adjusted levels of environment was not significantly different. A full list of multiple comparisons conducted on average density is provided in Fig. S1a–c (Supporting Information). Comparisons stated as significantly different have a P-value < 0·05. A new two-way anova was conducted on the data with the two treatment levels warm normal and warm adjusted combined. However, an F-test (d.f. = 1,171; F = 3·26; P = 0·023) indicated that this model simplification caused a significant increase in the unexplained variance. These apparently contradictory results of different statistical test result from the borderline significance of the patterns and the greater statistical power of the F-test relative to the Tukey’s HSD test.

Table 1.   Statistical summary of average species density log10(cells mL−1). Details of the model are discussed in the text
SpeciesSourced.f.FP-value
BacteriaEnvironment2250< 0·001
BacteriaChain length22·80·066
BacteriaEnvironment : chain length4670< 0·001
BacteriaError171  
ColpidiumEnvironment2260< 0·001
ColpidiumChain length135< 0·001
ColpidiumEnvironment : chain length223< 0·001
ColpidiumError114  
DidiniumEnvironment231< 0·001
DidiniumError57  
image

Figure 1.  Average density (left-hand column) and coefficient of variation (CV, right-hand column), ± one standard error, of protist (b and d, Didinium; c and e, Colpidium) and bacteria (c and f) populations. Lines connecting treatment points are present to aide visual interpretation of the data, and do not represent statistical models, statistical analyses are detailed in the main text.

Download figure to PowerPoint

Visual inspection of Fig. 1c shows clearly that the metabolic effects of temperature, rather than viscosity, are the main drivers of temperature effects on bacterial density. Bacterial density is less affected by temperature in the two and three trophic level treatments compared to the one trophic level treatment (Fig. 1c). In comparison, bacterial density in the one trophic level community was almost one million times greater in warm environments than in the cold environment (Fig. 1c). Overall, the addition of the consumer, Colpidium, to create a two trophic level system, reduced the density of bacteria in the warm environments, but increased density in the cold environments. The addition of the top predator, Didinium to create a three trophic level system had no significant effect on the average density of bacteria in warm normal and cold environments.

Environment, food chain length and their interaction also produced a significant response in average Colpidium density (Table 1, Fig. 1b). As with bacteria, post hoc multiple comparisons suggest no difference between warm normal and warm adjusted viscosity treatment levels. An F-test (d.f. = 1,114; F = 1·5; P = 0·23) indicated that the original and a simplified model (warm normal and warm adjusted combined) do not significantly differ. Thus, the minimum adequate model to explain Colpidium average density contains only two environmental levels: warm and cold. Colpidium density was greatest in the warm environments, and lowest in the cold environment in the absence of the top predator (Fig. 1b). The addition of Didinium reduced the average density of Colpidium in the warm environments, and moderately increased the density of Colpidium in the cold environment (Fig. 1b).

Average Didinium density was significantly different among all three environments (Table 1). At 5 °C, Didinium failed to form a viable population, thus its density was consistently zero. Viscosity had a significant negative impact on the density of Didinium, reducing its average density by 1·89 cells mL−1, or 68% (Fig. 1a).

Population temporal variability

Food chain length, environment and their interaction had a significant impact on the variability (CV) of species population density (Table 2, Fig. 1d–f). As with average population density, a full list of multiple comparisons conducted on CV is provided in Fig. S1d–f (Supporting Information). Comparisons stated as significantly different have a P-value < 0·05. Post hoc multiple comparisons suggested that bacterial CV did not significantly differ between warm normal and warm adjusted treatments, regardless of trophic chain length. Model simplification indicates that combining the warm adjusted and warm normal treatments produces no significant difference from the original model (d.f. = 1,170; F = 1·73; P = 0·16). The CV of the bacterial fauna in the absence of higher trophic levels was greatest in the cold environment (CV ≈ 1·5) compared to warm environments (CV ≈ 0·4; Fig. 1f). The addition of the second and third trophic levels significantly affected the CV of bacteria populations. At 20 °C, the CV of bacteria was higher in the two and three trophic level communities compared to bacteria fauna in the single trophic level system (Fig. 1f). This trend was reversed at 5 °C, with the CV of bacteria density being higher in the one trophic level system compared to the two or three trophic level system (Fig. 1f).

Table 2.   Statistical summary of the coefficient of variation of species density (cells mL−1) through time. Details of the model are discussed in text
SpeciesSourced.f.FP-value
BacteriaEnvironment23·80·02
BacteriaChain length29·3< 0·001
BacteriaEnvironment : chain length45·8< 0·001
BacteriaError170  
ColpidiumEnvironment250< 0·001
ColpidiumChain length160< 0·001
ColpidiumEnvironment : chain length254< 0·001
ColpidiumError113  
DidiniumEnvironment20·77< 0·001
DidiniumError27  

The CV of the consumer, Colpidium, was greatest at 5 °C (CV ≈ 0·9) compared to 20 °C (CV ≈ 0·4), when cultured without its predator (Fig. 1e). As with bacteria, the addition of a higher trophic level increased the CV of Colpidium at 20 °C (Fig. 1e). Viscosity had a significant impact on the CV of Colpidium, in the three trophic level community. The CV of Colpidium in the warm adjusted treatment was significantly higher than in the warm normal community, but still lower than the CV of Colpidium in the cold treatment (Fig. 1e). As Didinium did not successfully establish a community in the 5 °C community, Didinium CV could only be calculated for the warm treatments. An increase in viscosity had no significant impact on the CV of Didinium population densities (Fig. 1d).

Colpidium r and k

Colpidium populations expressed logistic growth dynamics in both the warm normal and warm adjusted viscosity treatments (Fig. 2). An increase in viscosity (independent of temperature) significantly increased both the carrying capacity (k) and growth rate (r) of Colpidium (Fig. 2). The results of a t-test for the parameter estimates of r and k from each microcosm were: mean radjusted viscosity = 0·75 cell divisions day−1; rnormal viscosity = 0·67 cell divisions day−1; d.f. = 34, t = −2·3, P = 0·030; kadjusted viscosity = 2000 cells mL−1, knormal viscosity = 1500 cells mL−1, d.f. = 35, T = −2·5, P = 0·018. The optimization algorithm used to fit the logistic growth model did not find a solution for Colpidium cultured at 5 °C.

image

Figure 2.  Growth dynamics of Colpidium in three different environmental treatments. (a) Colpidium density through time, points are individual sampling events of the density of Colpidium [log10(cells mL−1)] per day. Lines are fitted logistic growth models based on average parameter estimates of the logistic growth equation. (b) Average growth rate and carrying capacity of Colpidium ± one standard error. Model and statistics are discussed fully in the main text.

Download figure to PowerPoint

Decomposition

Food chain length, environment and their interaction had a significant impact on bacterial decomposition (Table 3, Fig. 3). Post hoc multiple comparisons suggest that viscosity did not have an effect on decomposition, as warm adjusted and warm normal environmental treatment levels did not significantly differ. A full list of multiple comparisons conducted on decomposition rate is provided in Fig. S2 (Supporting Information). A simplified (warm adjusted and warm normal treatments combined) model was significantly different (d.f. = 1,218; F = 3·12; P = 0·016) to the original model. Thus, the impact of viscosity is via the food chain length–environment interaction.

Table 3.   Statistical summary of the dry mass loss, arcsine (√(proportion mass loss)), of wheat grains
Sourced.f.FP-value
Environment2140< 0·001
Chain length3130< 0·001
Environment : Chain length633< 0·001
Error218  
image

Figure 3.  Decomposition measured as proportion dry mass loss of wheat grains. Data points are of the mean treatment combination dry mass loss ± one standard error. Lines connecting treatment points are present to aide visual interpretation of the data, and do not represent statistical models. Statistical (as detailed in the main text) analyses were conducted upon arcsine square-root transformed data, but untransformed data are presented here for ease of interpretation.

Download figure to PowerPoint

A temperature > 5 °C and food chain length > 1 was required before wheat grain mass loss was significantly increased, compared to abiotic mass loss of wheat grains (Fig. 3). The addition of Colpidium to a community did significantly increase the mass loss of wheat grains. The further addition of Didinium in the three trophic level system, did not significantly affect the decomposition of wheat grains. Temperature had a strong and positive effect on the decomposition of wheat grains across all communities (Fig. 3).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Food chain length, environment and their interaction significantly affected the size and variability of populations. Viscosity explained a significant but small proportion of consumer population variability and population dynamics. The impact of viscosity on the population density of bacteria was mediated via an interaction with food chain length. Viscosity also influenced the growth dynamics of the consumer, and the average density of the top predator. Our observations provide some of the first evidence that viscosity-based effects of temperature can alter populations. The effects of viscosity were not realized at all trophic levels. In isolation, bacterial density was not significantly affected by viscosity and viscosity did not affect the predator impact of Didinium on the consumer Colpidium.

Abrusan (2004) collected data on the feeding and growth rate of Daphnia when cultured with varying concentrations of filamentous bacteria. A 30% increase in viscosity caused a significant reduction in feeding and growth rates of Daphnia, but only at low food concentrations. Didinium expresses a Type II functional response to protist consumers (Hewett 1980, 1988), with saturation occurring at c. 60–70 cells mL−1 (for Didinium consuming Paramecium; Hewett 1980). At 20 °C, the density of Colpidium in our experiment remained in excess of 100 cells mL−1, even in the three trophic level system. At these high population levels, the consumption of Colpidium by Didinium would not be encounter limited, but directly related to the predator handling time. Therefore, an increase in viscosity resulting in a reduction in protist swimming speed and encounter rate would not have an impact on consumer density. This may explain why viscosity did not affect predator impact upon consumers in our study. Future studies may wish to repeat our experiment at differing consumer and resource densities, and observe if viscosity impacts become more profound as food availability is reduced.

Temperature, food chain length and their interaction had a significant impact on the temporal variability of population density. For both Colpidium and bacteria, the addition of their respective predators increased temporal variability at 20 °C. Unstable cyclic dynamics are regularly observed for microbial communities (e.g. Holyoak & Lawler 1996), and can be related to our observed increase in CV of prey and resources. Resource variability was high in the absence of consumers, and consumer variability was high in the absence of predators, at 5 °C. Surprisingly, at 5 °C the addition of a consumer decreased the CV of resource. This observation suggests that in cold environments the addition of a consumer increases the stability of resource species. Additionally, both the average density of consumers and resource species was increased in the presence of their respective consumer and predator, at 5 °C.

The combination of an increase in average density and a reduction in population variability supports a previous theorem resulting in the ‘Hydra Effect’. Abrams (2009) suggests several mechanisms where predation can increase a prey’s average population density. A fluctuating prey population may have an increased average density when a predator is added, as predation can reduce the amplitude of oscillating cycles and reduce the time a population spends recovering from a crash (Abrams 2009).

The temporal variability of a species’ population density can be an indicator of population stability (Petchey 2000; Ives & Hughes 2002). A reduction in interspecies interaction strength generally decreases population variability and increases population stability (Yodzis 1981; Emmerson & Yearsley 2004). Viscosity can decrease swimming speed (e.g. Winet 1976; Podolsky & Emlet 1993; Fuiman & Batty 1997), and therefore predator–prey encounter rates (Visser & Kiorboe 2006), and predator feeding rate (Podolsky, 1994; Bolton & Havenhand 2005; Abrusan 2004; Bolton & Havenhand 2005). Thus, one may predict that an increase in viscosity will reduce the strength of predator–prey interactions, and therefore increase stability. Our data suggest that the CV and therefore stability of Colpidium is affected by viscosity in the presence of its predator, but resource and predator CV are not affected by viscosity. An increase in viscosity has been demonstrated to increase the persistence of Didinium and when predating upon Paramecium aurelia (Luckinbill 1973; Harrison 1995). However, these studies manipulated viscosity to a level not expected in nature, and not linked to a temperature change. Our data suggest that the CV of Colpidium is increased by temperature-dependent viscosity.

Increased viscosity led to both a higher intrinsic growth rate and higher carrying capacity of Colpidium. Temperature-dependent intrinsic growth rate has previously been observed for Colpidium (Laybourn & Stewart 1975). Based on these findings, our viscosity-induced increase in Colpidium intrinsic growth rate is equivalent to ≈ 2 °C change in temperature. Intrinsic growth rate is an important component of population dynamics, variations in which may alter community dynamics, for example the coexistence of competitive ciliate consumers (Fox & Morin 2001; Jiang & Morin 2004). Therefore, it is possible that viscosity will alter equilibrium densities or the coexistence of competing consumers. It is important to note that viscosity can naturally increase independently of temperature. For example, the viscosity of seawater around spring blooms of Phaeocystis globosa can increases by up to 250% (Seuront, Vincent & Mitchell 2006). Such high viscosities are likely to alter population dynamics to a greater extent than the ≈ 50% change in viscosity discussed here, which requires the attention of future work.

A temperature > 5 °C and a food chain length > 1 was required to significantly increase the mass loss of wheat grains above that of abiotic mass loss. Thus, significant biological decomposition only occurred in communities containing a consumer with a temperature > 5 °C. Microcosms containing consumers at 20 °C had a decomposition rate approximately three times greater than microcosms containing only bacteria. An increase in bacterial decomposition via the addition of protozoa is well documented in the literature (Ribblett, Palmer & Coats 2005; Jiang & Krumins 2006; Krumins et al. 2006; Jiang 2007). Our findings suggest that increasing food chain length from 2 to 3 has negligible impact on decomposition. We propose that it is the presence of a protozoan bacterial consumer, rather than food chain length, which is important for determining decomposition. Further work is required to understand fully what is important in determining protist-facilitated bacterial decomposition.

Here, we demonstrate the importance of direct and indirect effects of temperature on microbial communities. Indirect effects accounted for a large proportion of the community’s overall response to temperature. Indirect temperature effects via temperature-dependent viscosity accounted for a significant, but small, proportion of the overall temperature dependency of aquatic communities. Indirect temperature effects via trophic interactions had a strong influence on both population dynamics and ecosystem functioning. Not accounting for indirect effects of temperature, particularly via trophic interactions, will lead to incorrect predictions of the response of populations and communities to varying temperature.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

O.S.B. was funded by a NERC studentship; O.L.P. is a Royal Society University Research Fellow; S.H. was supported by a NERC Advanced Fellowship (NE/B500690/1).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Abrams, P.A. (2009) When does greater mortality increase population size? The long history and diverse mechanisms underlying the hydra effect. Ecology Letters, 12, 462474.
  • Abrusan, G. (2004) Filamentous cyanobacteria, temperature and daphnia growth: the role of fluid mechanics. Oecologia, 141, 395401.
  • Apple, J.K., Del Giorgi, P.A. & Kemp, W.M. (2006) Temperature regulation of bacterial production, respiration, and growth efficiency in a temperate salt-marsh estuary. Aquatic Microbial Ecology, 43, 243254.
  • Berg, H.C. & Turner, L. (1979) Movement of microorganisms in viscous environments. Nature, 278, 349351.
  • Beveridge, O.S., Humphries, S. & Petchey, O.L. (2010) The interacting effects of temperature and food chain length on trophic abundance and ecosystem function. The Journal of Animal Ecology, 79, 693700.
  • Bolton, T.F. & Havenhand, J.N. (1998) Physiological versus viscosity-induced effects of an acute reduction in water temperature on microsphere ingestion by trochophore larvae of the serpulid polychaete galeolaria caespitose. Journal of Plankton Research, 20, 21532164.
  • Bolton, T.F. & Havenhand, J.N. (2005) Physiological acclimation to decreased water temperature an the relative importance of water viscosity in determining the feeding performance of larvae of a serpulid Polychaete. Journal of Plankton Research, 27, 875879.
  • Brown, J.H., Gillooly, J.F., Allen, A.P., Savage, V.M. & West, G.B. (2004) Toward a metabolic theory of ecology. Ecology, 85, 17711789.
  • Clarke, A. (1991) What is cold adaptation and how should we measure it? American Zoologist, 31, 8192.
  • Crawley, M.J. (2007) The R Book. John Wiley & Sons Ltd, Chichester, West Sussex.
  • Dailey, T.M., Claussen, D.L., Ladd, G.B. & Buckner, S.T. (2009) The effects of temperature, desiccation, and body mass on the locomotion of the terrestrial isopod, Porcellio laevis. Comparative Biochemistry and Physiology A-Molecular & Integrative Physiology, 153, 162166.
  • Denny, M.W. (1993) Air and Water: The Biology and Physics of Life’s Media. Princeton University Press, Princeton.
  • Emmerson, M. & Yearsley, M. (2004) Weak interactions, omnivory and emergent food-web properties. Proceedings of the Royal Society of London Series B-Biological Sciences, 271, 397405.
  • Fox, J.W. & Morin, P.J. (2001) Effects of intra- and interspecific interactions on species responses to environmental change. The Journal of Animal Ecology, 70, 8090.
  • Fuiman, L.A. & Batty, R.S. (1997) What a drag it is getting cold: partitioning the physical and physiological effects of temperature on fish swimming. The Journal of Experimental Biology, 200, 17451755.
  • Gaston, K.J. & McArdle, B.H. (1994) The temporal variability of animal abundances – measures, methods and patterns. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 345, 335358.
  • Gillooly, J.F., Brown, J.H., West, G.B., Savage, V.M. & Charnov, E.L. (2001) Effects of size and temperature on metabolic rate. Science, 293, 22482251.
  • Greenberg, E.P. & Canale-Parola, E. (1977) Motility of flagellated bacteri in viscous environments. Journal of Bacteriology, 132, 356358.
  • Harrison, G.W. (1995) Comparing predator prey models to Lucknbills’s experiment with Didinium and Paramecium. Ecology, 76, 357374.
  • Hewett, S.W. (1980) The effect of prey size on the functional and numerical responses of a protozoan predator to its prey. Ecology, 61, 10751081.
  • Hewett, S.W. (1988) Predation by Didinium-nasutum– efects of predator and pey size. Ecology, 69, 135145.
  • Holyoak, M. & Lawler, S.P. (1996) Persistence of an extinction-prone predator-prey interaction through metapopulation dynamics. Ecology, 77, 18671879.
  • Ives, A.R. (1995) Predicting the response of populations to environmental-change. Ecology, 76, 926941.
  • Ives, A.R. & Hughes, J.B. (2002) General relationships between species diversity and stability in competitive systems. The American Naturalist, 159, 388395.
  • Jiang, L. (2007) Negative selection effects suppress relationships between bacterial diversity and ecosystem functioning. Ecology, 88, 10751085.
  • Jiang, L. & Krumins, J.A. (2006) Consumer vs environmental productivity control of bacterial diversity and bacteria-mediated organic matter decomposition. Oikos, 114, 441450.
  • Jiang, L. & Morin, P.J. (2004) Temperature-dependent interactions explain unexpected responses to environmental warming in communities of competitors. The Journal of Animal Ecology, 73, 569576.
  • Kemp, P.F., Sherr, B.F., Sherr, E.B. & Cole, J.J. (1993) Handbook of Methods in Aquatic Microbial Ecology. Lewis Publishers, Boca Raton, FL.
  • Kiorboe, T. (2008) A Mechanistic Approach to Plankton Ecology. Princeton University Press, Woodstock.
  • Krumins, J.A., Long, Z.T., Steiner, C.F. & Morin, P.J. (2006) Indirect effects of food web diversity and productivity on bacterial community function and composition. Functional Ecology, 20, 514521.
  • Laybourn, J. & Stewart, J. (1975) Studies on consumption and growth in ciliate Colpidium-campylum Stokes. Ecology, 44, 165174.
  • Loiterton, B., Sundbom, M. & Vrede, T. (2004) Seperating physiological effects of temperature on zooplankton feeding rate. Aquatic Science, 66, 123129.
  • Luckinbill, L.S. (1973) Coestience in laboratory populations of Paramecium aurelia and its predator Didinium nasutum. Ecology, 54, 13201327.
  • McArdle, B.H., Gaston, K.J. & Lawton, J.H. (1990) Variation in the size of animal populations – patterns, problems and artifacts. The Journal of Animal Ecology, 59, 439454.
  • Petchey, O.L. (2000) Prey diversity, prey composition and predator dynamics in experimental microcosms. The Journal of Animal Ecology, 69, 874882.
  • Podolsky, R.D. & Emlet, R.B. (1993) Seperating the effects of temperature and viscosity on swimming and water movement by sand dollar larvae (Dendraster excentricus). The Journal of Experimental Biology, 176, 207221.
  • Podolsky, R.D. (1994) Temperature and water viscosity: physiological versus mechanical effects on suspension feeding. Science, 265, 100103.
  • R Development Core Team (2007) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN: 3-900051-07-0. Available at: http://www.R-project.org (accessed on 12/09/2009).
  • Ribblett, S.G., Palmer, M.A. & Coats, D.W. (2005) The importance of bacterivorous protists in the decomposition of stream leaf litter. Freshwater Biology, 50, 516526.
  • Rose, J.M. & Caron, D.A. (2007) Does low temperature constrain the growth rates of heterotrophic protists? Evidence and implications for algal blooms in cold waters. Limnology and Oceanography, 52, 886895.
  • Savage, V.M., Gilloolly, J.F., West, G.B. & Charnov, E.L. (2004) Effects of body size and temperature on population growth. The American Naturalist, 163, 439441.
  • Schneider, W.R. & Doestsch, R.N. (1974) Effect of viscosity on bacterial motility. Journal of Bacteriology, 117, 696701.
  • Seuront, L., Vincent, D. & Mitchell, J.G. (2006) Biologically induced modification of seawater viscosity in the eastern English Channel during a Phaeocystis globosa spring bloom. Journal of Marine Systems, 61, 118133.
  • Sleigh, M.A. (1956) Metachronism and frequency of beat in the peristomial cilia of Stentor. Journal of Experimental Biology, 33, 1528.
  • Tompkins, J., Deville, M.M., Day, J.G. & Turner, M.F. (1995) Culture Collection of Algae and Protozoa: Catalogue of Strains 1995. Titus Wilson & Son Ltd, Kendal.
  • Vasseur, D.A. & McCann, K.S. (2005) A mechanistic approach for modeling temperature-dependent consumer-resource dynamics. The American Naturalist, 166, 184198.
  • Visser, A.W. & Kiorboe, T. (2006) Plankton motility patterns and encounter rates. Oecologia, 148, 538546.
  • Visser, A.W. & Thygesen, U.H. (2003) Random motility of plankton: diffusive and aggregative contributions. Journal of Plankton Research, 25, 11571168.
  • Vogel, S. (1988) Life’s Devices: The Physical World of Animals and Plants. Princeton University Press, Princeton.
  • Wilson, R.S. (2005) Temperature influences the coercive mating and swimming performance of male eastern mosquitofish. Animal Behaviour, 70, 13871394.
  • Winet, H. (1976) Ciliary propulsion of objects in tubes – wall drag on swimming Tetrahymena (Ciliata) inpresence of mucin and other long-chain polymers. Journal of Experimental Biology, 64, 283302.
  • Yee, E.H. & Murray, S.N. (2004) Effects of temperature on activity, food consumption rates, and gut passage times of seaweed-eating Tegula species (Trochidae) from California. Marine Biology, 145, 895903.
  • Yodzis, P. (1981) The stability of real ecosystems. Nature, 289, 674676.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Appendix S1. Summary of preliminary experiments to test the toxicology of Ficoll with respect to protists (Colpidium striatum).

Fig. S1. Visual summary of post hoc multiple comparisons (Tukey’s HSD), average density (left-hand column) and CV (right-hand column).

Fig. S2. Visual summary of post hoc multiple comparisons (Tukey’s HSD), of proportion mass loss of wheat grains.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JANE_1741_sm_AppendixS1.pdf115KSupporting info item
JANE_1741_sm_FigureS1.pdf46KSupporting info item
JANE_1741_sm_FigureS2.pdf29KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.