Is spatial variation in population size structures of a stream-dwelling caddisfly due to the altered effects of a predator by a third-party species?

Predators alter abundances and life history characteristics of prey, and effects of predator–prey interactions can be altered by third-party species. Here, we examine size structures of the caddisfly, Phylloicus hansoni , in Trinidadian streams with two distinct fish assemblages: upstream reaches where the predatory killifish, Anablepsoides hartii , is the only fish species ( Killifish - Only reaches), and downstream reaches where killifish and the omnivorous guppy, Poecilia reticulata , coexist ( Killifish – Guppy reaches). We asked: Do P. hansoni larvae exhibit differences in size structure between reaches with differing fish assemblages? We found that size distributions of larvae differed between reaches in the majority of replicate streams, with smaller median body lengths in Killifish - Only reaches. Killifish – Guppy reaches had higher proportions of the largest instar, but we did not find differences in body length within an instar. No evidence of size-selective predation was found through analysis of killifish stomach contents, and environmental variables were largely similar between upstream and downstream reaches of the five study streams, aside from higher killifish abundances in upstream reaches. Our results, coupled with previous evidence of guppies altering killifish populations, suggest that the mediating effects of a third-party species (guppies) on predator–prey (killifish–caddisfly) interactions can affect the population size structure of prey populations.