Understanding species interactions and the resulting natural selection are some of the biggest goals of ecology and evolutionary biology. Working with the narrowleaf cottonwood tree, a galling aphid and two gall predators (insects and birds), we examined attack rates on galls and natural selection on gall size and how they vary across host tree populations and with respect to the two predators. We conducted two studies. First, in an observational study, we visited three tree populations at low, mid and high elevations. We also used a common garden that included cottonwood populations from the same river system and elevation gradient as our observational study. At each population, we calculated the percentage of galls attacked and the pattern of natural selection caused by both insects and birds.
Results/Conclusions
In the wild, we found that both attack rates and patterns of natural selection on galls differed across sites and with respect to the two predators. For example, birds acted as an agent of natural selection, selectively foraging on small galls, but only at the mid elevation. Insects acted as an agent of natural selection, selecting small galls, at low elevation, but large galls at the mid elevation. In the common garden, attack rates on galls varied by tree population. For example, birds attacked roughly 40% of galls on trees from the mid and high elevation trees, but only 17% of galls on low elevation trees. Birds also acted as an agent of natural selection, selectively foraging on smaller galls. Some evidence suggested the strength of natural selection was weaker on trees from the low elevation site, but this trend was not statistically significant. In the common garden, we did not find evidence that insects acted as an agent of natural selection on gall size.
Our results show that both environmental and tree genetic variation contributes to variation in predator-prey interactions and patterns of natural selection on gall size. Further, because the substantial climate gradient among elevations, including tree populations that are known to have adaptively evolved along that gradient, our results suggest that understanding changes to complex species interactions and patterns of natural selection in the wake of climate change will require both ecological and genetic perspectives.