Abstract
As climate regimes shift in many ecosystems worldwide, evolution may be a critical process allowing persistence in rapidly changing environments. Organisms regularly interact with other species, yet whether climate-mediated evolution can occur in the context of species interactions is not well understood. We tested whether a species interaction could modify evolutionary responses to temperature. We demonstrate that predation pressure by Dipteran larvae (Chaoborus americanus) modified the evolutionary response of a freshwater crustacean (Daphnia pulex) to its thermal environment over approximately seven generations in laboratory conditions. Daphnia kept at 21°C evolved higher population growth rates than those kept at 18°C, but only in those populations that were also reared with predators. Furthermore, predator-mediated selection resulted in the evolution of elevated Daphnia thermal plasticity. This laboratory natural selection experiment demonstrates that biotic interactions can modify evolutionary adaptation to temperature. Understanding the interplay between multiple selective forces can improve predictions of ecological and evolutionary responses of organisms to rapid environmental change.
1. Introduction
In light of projected rapid increases in mean global temperature [1], organisms that cannot migrate to more suitable habitats or evolve in place to adapt to the changing environment may face climate-induced population extinction [2]. Rapid adaptation to climate-related variables has been documented in a few single-species systems [3–7], but because virtually all organisms experience regular interspecific interactions [8], understanding how biotic interactions affect evolutionary responses is critical.
Community context can affect evolutionary responses by altering the availability of genetic variation or the strength of natural selection [8,9]. Additionally, biotic interactions can reduce evolutionary responses to climate-mediated selection if organisms that are well adapted to the abiotic environment are selected against by the interacting species [10]. Conversely, biotic interactions can increase the evolutionary response to climate-mediated selection if both the biotic pressure and the climate variable favour similar traits [11]. Given the ubiquity of biotic interactions in nature, our ability to predict whether populations can adapt to rapid environmental change hinges on a better understanding of whether, and how, biotic processes accelerate or hinder evolutionary adaptation to climate-related variables [2,8,10].
We tested the hypothesis that biotic interactions (predation) can alter evolutionary responses to abiotic factors (temperature) using a laboratory natural selection experiment. Chaoborus americanus larvae (hereafter Chaoborus) are predators of Daphnia pulex (hereafter Daphnia) and have been shown to exert strong selection on Daphnia [12]. We expected Daphnia reared with predators to evolve smaller body sizes and faster population growth rates [13,14]. Similarly, we expected Daphnia reared at warmer temperatures to evolve smaller body sizes and higher growth rates [15,16]. Because selection by predators appears to act on prey life history [13], whereas selection by temperature appears to act on cell and genome size [17], we assumed that these two mechanisms of selection were relatively independent, and thus in the presence of both selective forces, the overall evolutionary response would be greater than that to either temperature or predation alone.
2. Material and methods
We conducted a fully factorial laboratory natural selection experiment with three temperature (18°, 21° and 25°C) and two predator (with or without one Chaoborus larva) treatments. Details of the treatments and of the source and maintenance of organisms are available in the electronic supplementary material. We created 24 1 l microcosms (four per treatment), each stocked with 30 Daphnia and 900 ml filtered pond water. This density is similar to that used in [12]. Daphnia consumed algal populations introduced with the pond water. Although we did not quantify algal concentrations, we expected Daphnia population size to fluctuate with food availability, as is seen in nature. Nutrients were added to microcosms every two weeks via slow-release fertilizer pellets (Osmocote® Indoor-Outdoor Plant food, 19-6-12). Microcosms experienced full spectrum artificial lights on a 14 L : 10 D cycle.
At the end of week 14 (7.5 ± 1.5 Daphnia generations [12]), predators were removed and all microcosms were placed in a common environment (approx. 20°C) for three weeks (weeks 15–17, approx. two to three generations) to minimize maternal effects [12]. We expected negligible predator cues in the common garden environment [18]. Assessing Daphnia fitness traits after this common environment phase allows us to disentangle genetic change from phenotypic plasticity [19].
We assayed whether evolutionary change had occurred during weeks 0–14 by testing whether variation in two response variables—Daphnia population growth rate between weeks 17 and 19, and Daphnia body size at week 19—was significantly explained by selection treatment. To do this, at week 17, each microcosm was subdivided into subpopulations (n = 4–6 subpopulations per microcosm, depending on microcosm population size), and evolutionary change was assessed using these subpopulations. We used low starting subpopulation densities (0.001–0.025 individual ml−1) to minimize density-dependent effects. Subpopulations were kept at 18°C or 21°C (assay temperatures) for two weeks (weeks 18 and 19). We did not conduct the evolution assay with predators because of insufficient sample sizes.
We recorded Daphnia population size every two weeks from weeks 0 to 14 by averaging the number of Daphnia in three 100 ml samples taken from each microcosm. In the evolution assay, we recorded Daphnia subpopulation size and average body size (mm2) at the end of week 19. Daphnia lateral surface area was measured using a side-view photo of three individuals per subpopulation (Leica Application Suite v. 4.3.0). We calculated the per capita growth rate, r, per subpopulation, as population size at week 19 minus the subpopulation size at week 17, divided by the subpopulation size at week 17. We divided by this value by 14 days to calculate the number of new individuals produced per individual Daphnia per day.
We used ANOVA to examine whether temperature and predators explained variation in mean Daphnia population size over weeks 0–14. We used linear mixed-effects models to assess the importance of ‘selection temperature’, ‘selection predation’ and ‘assay temperature’ (all fixed effects) in predicting evolved Daphnia r and body size. The final model for both Daphnia traits included the three main effects, harmonic mean population size, microcosm (random effect) and the interactions of selection temperature × selection predation, and selection predation × assay temperature (see electronic supplementary material for model selection details). All statistical analyses were conducted in R [20] using the car and nlme packages [21,22].
3. Results and discussion
Overall, we found that (i) predators and temperature mediated Daphnia population dynamics, (ii) Daphnia evolved faster growth rates when reared with predators and at the warmer temperature, and (iii) selection with predators resulted in the evolution of smaller Daphnia body size.
Daphnia mean population size between weeks 0 and 14 was highest at 21°C (F2,18 = 7, p = 0.006; figure 1, inset). At all temperatures, predators reduced Daphnia population size by approximately 36% (F1,18 = 6.4, p = 0.02; figure 1, inset). All populations reared at 25°C went extinct by week 10 of the experiment. The predator-mediated reduction in prey population size was likely due to direct predation and predator avoidance behaviour [23].
Effect of selection regime on Daphnia population size. Light blue, dark blue and green symbols denote the 18°, 21° and 25°C temperature treatments, respectively. Circles/solid lines and triangles/dashed lines denote predator-absent and predator-present selection treatments. Error bars are ±1 s.e.m. Note that statistical analyses were conducted on mean values (inset) and that the weekly values are included here for completeness. (Online version in colour.)
Daphnia reared at 21°C evolved higher growth rates than those reared at 18°C, but only if they had also been reared with predators (figure 2a and table 1). This result may be due to elevated attack rates by Chaoborus at higher temperatures [24], resulting in stronger selection for increased growth rate at the higher selection temperature. It is unclear whether Daphnia reared in the absence of predators would have evolved to temperature change had we run the experiment for longer. We also observed that Daphnia selected with predators showed faster growth at the higher assay temperature (figure 2a and table 1). This evolutionary increase in plasticity supports the assertion that biotic interactions may be important drivers of evolutionary changes in reaction norms [25]. A potential mechanism for this result is that the predator treatment created a more heterogeneous environment with regard to spatial or temporal distribution of predation risk than did the temperature treatment, and variable environments have been linked to the evolution of increased plasticity [26].
Effect of selection temperature, selection predation and assay temperature on Daphnia (a) per capita growth rate, r, and (b) body size. Dark blue symbols (dashed lines) and light blue symbols (solid lines) denote selection with and without predators, respectively. Circles and triangles denote populations selected at 18°C and 21°C, respectively. Error bars are ±1 s.e.m. (Online version in colour.)
Effect of selection regime and assay temperature on Daphnia per capita growth rate (r) and body size. Wald χ2 and p-values from the mixed effects models are presented. Italicized values denote statistically significant parameters.
Daphnia reared with predators evolved smaller body sizes than those reared without predators (figure 2b and table 1). This result is inconsistent with some life-history theory of this predator–prey pair [27], but is consistent with theory for non-size-selective predation [14]. These results are also inconsistent with a similar experiment with Chaoborus and Daphnia [12], but there, Daphnia experienced constant food availability and were on average larger in size than those in this experiment. Future studies that examine predator-feeding behaviour at multiple temperatures would help to ascertain whether size-selective predation contributed to the body size pattern observed here. There was no effect of selection temperature on Daphnia body size, and Daphnia assayed at 21°C were slightly smaller than those assayed at 18°C (figure 2b and table 1). This body size plasticity is consistent with a previous study examining plastic responses of Daphnia in response to temperature change [28].
This experiment demonstrated that predation pressure mediated the evolutionary divergence of Daphnia in different temperature environments. Confirmation of these results in a field setting would help to assess their generality. Given that many predators are predicted to select for faster prey growth rate [14] and that elevated temperatures tend to increase ecotherm growth rate [15,16], synergistic interactions between temperature and predators may represent an important mechanism for organisms to adapt to a rapidly changing environment. Our results complement the growing ecological literature on population persistence in the context of species interactions in changing environments [29,30] by demonstrating the possible evolutionary benefits of trophic interactions for rapid adaptation in the face of warming temperature. Additionally, these results demonstrate that evolutionary change can occur in ecological time scales and thus strengthen the argument for incorporating evolutionary processes into explanations of short-term ecological patterns [31].
Ethics
Study organisms were collected from the UBC Experimental Ponds Facility. No permits were required.
Data accessibility
Data are available at the Dryad Digital Repository: http://dx.doi.org/10.5061/dryad.m64qc.
Authors' contributions
M.T. conceived the study. M.T. and M.I.O. designed the experiment. M.T. collected and analysed the data. M.T. wrote the manuscript. M.I.O. edited the manuscript. Both authors gave final approval for publication.
Competing interests
We have no competing interests.
Funding
Funding was provided by an NSERC (Canada) Discovery grant to M.I.O.
Acknowledgements
We thank N. Caulk for help with experimental set-up and maintenance, and L. Rieseberg and R. Stoks for helpful comments. We also thank J. Petkau, D. Dinsdale and the STAT551 Statistical Consulting graduate class at UBC for their review of this project.
- Received September 23, 2015.
- Accepted November 20, 2015.
- © 2015 The Author(s)