Top predator removals have consistent effects on large species despite high environmental variability

Top predator losses affect a wide array of ecological processes, and there is growing evidence that top predators are disproportionately vulnerable to environmental changes. Despite increasing recognition of the fundamental role that top predators play in structuring communities and ecosystems, it remains challenging to predict the consequences of predator extinctions in highly variable environments. Both biotic and abiotic drivers determine community structure, and manipulative experiments are necessary to disentangle the effects of predator loss from other co-occurring environmental changes. To explore the consistency of top predator effects in ecological communities that experience high local environmental variability, we experimentally removed top predators from arid-land stream pool mesocosms in southeastern Arizona, USA, and measured natural background environmental conditions. We inoculated mesocosms with aquatic invertebrates from local streams, removed the top predator Abedus herberti (Hemiptera: Belostomatidae) from half of the mesocosms as a treatment, and measured community divergence at the end of the summer dry season. We repeated the experiment in two consecutive years, which represented two very different biotic and abiotic environments. We found that some of the effects of top predator removal were consistent despite significant differences in environmental conditions, community composition, and colonist sources between years. As in other studies, top predator removal did not affect overall species richness or abundance in either year, and we observed inconsistent effects on community and trophic structure. However, top predator removal consistently affected large-bodied species (those in the top 1% of the community body size distribution) in both years, increasing the abundance of mesopredators and decreasing the abundance of detritivores, even though the identity of these species varied between years. Our findings highlight the vulnerability of large taxa to top predator extirpations and suggest that the consistency of observed ecological patterns may be as important as their magnitude.