Changes in consumer assemblages caused by species invasions or losses have significant effects on ecosystem function. These perturbations can also alter organismal phenotypes, but the ecosystem consequences of phenotypic variability are largely unexplored. How important are the ecosystem effects of phenotypic diversification, especially when compared against variability in ecosystem processes that is generated by environmental heterogeneity? We address this question using a factorial mesocosm experiment that crosses guppy phenotype and light. Phenotypic evolution in Trinidadian guppies (Poecilia reticulata), which is caused by the removal of predators, has a significant effect on many ecosystem processes. In Trinidadian streams, light availability limits primary production, and influences many ecosystem processes. Light levels can vary dramatically on small spatial scales, but typically there are also significant differences in light levels between sites with and without predators. This suggests that phenotypic effects of guppies on ecosystem dynamics will be confounded with natural variability in light. In this experiment we crossed two guppy phenotypes with two light treatments that mimicked average light conditions in Trinidadian streams. We monitored a suite of ecological response variables including algal accrual, ecosystem metabolism, nutrient fluxes, and consumer-mediated nutrient recycling.
Results/Conclusions
Light had a significant positive effect on algal accrual (F1,15=100,P<0.001), gross primary production (F1,15=7.5, P=0.03), invertebrate biomass (F1,15=60.1,P<0.001), and guppy ammonium excretion (F1,15=58.5, P<0.001). High light also shifted the composition of algae towards larger diatoms, as indicated by non-metric multi-dimensional scaling analysis. Light did not have a significant effect on leaf decomposition (P>0.05). Guppy phenotype had a significant effect on algal accrual (F1,15=5.1, P=0.043), gross primary production (F1,15=6.1,P=0.02), ammonium excretion (F1,15=11.0, P<0.005), and leaf decomposition (P=0.04). There were no significant effects of guppies on algal diversity. Interactions between light and phenotype were weak, suggesting that the effects of light and predation were additive. We conclude that variability in light levels corresponding to natural variability in Trinidadian streams generated much stronger effects on ecosystem processes than phenotypic diversification, but that guppy phenotype can generate significant variability around the mean of the light effect.