Ecological processes have long been known to affect evolution, yet only recently has there been experimental evidence of evolution affecting ecological processes. We examine how trait variation (diet selectivity) in Trinidadian guppies affects stream ecosystems at local and landscape scales to address the question: Are top-down trait-mediated effects of locally adapted guppies masked by high environmental variation? We focus on guppies locally adapted to either low-elevation streams with high-predation pressure and high-resource availability, or high-elevation streams with low-predation pressure and low-resource availability. We ran an in situ experiment in two reaches of a single high-elevation stream with one reach dominated by introduced low-elevation guppies and a second reach dominated by native high-elevation guppies. Electric fish-exclosures (n=5) were nested in each 150 m experimental reach to isolate top-down effects on algal growth rates and invertebrate biomass. We then conducted a landscape-level study where we measured invertebrate and algal standing stocks in 16, 150 m stream reaches (n=112 samples), across 8 high-elevation watersheds characterized by either native high-elevation (n=4) or introduced low-elevation guppies (n=4) relative to paired guppy-free reaches (n=8).
Results/Conclusions
We found significantly different top-down effects of high-elevation versus low-elevation guppies on ecosystem properties (algal and invertebrate standing stocks) and processes (algae accrual rates). Moreover, these differential effects of guppy diet selectivity on ecosystem properties were evident at the landscape scale despite high environmental variability. In our in situ experiment, introduced low-elevation guppies did not have a significant top-down effect on algae but significantly reduced invertebrate biomass (3 x higher invertebrate biomass in fish-exclusion versus control treatments; p<0.01). In contrast, native high-elevation guppies had significant top-down effects on algae (50% higher algal standing stocks and accrual rates in fish exclusion versus control treatments; p<0.05) while invertebrate biomass was not significantly different between treatments. Data at the landscape-scale support our experimental results: high-elevation streams with introduced low-elevation guppies contained 25% lower invertebrate biomass relative to upstream controls (p<0.05), whereas streams containing native high-elevation guppies contained 20% lower algal biomass relative to upstream guppy-free controls (p<0.05). We suggest that local-adaptation can have significant effects on ecosystem-level properties despite high environmental variability across the landscape, highlighting the importance of the eco-evolutionary feedbacks in nature.