Robert M. Pringle, Tyler R. Kartzinel, Todd M. Palmer, Timothy J. Thurman, Kena Fox-Dobbs, Charles C. Y. Xu, Matthew C. Hutchinson, Tyler C. Coverdale, Joshua H. Daskin, Dominic A. Evangelista, Kiyoko M. Gotanda, Naomi A. Man in ’t Veld, Johanna E. Wegener, Jason J. Kolbe, Thomas W. Schoener, David A. Spiller, Jonathan B. Losos, Rowan D. H. Barrett
15 min
This study investigated how the introduction of top predators influences the coexistence of competing prey species. While classic ecological theory suggests that predators often promote biodiversity by preventing competitive exclusion (the keystone predation model), the authors tested whether predator-induced behavioral changes—specifically the fear-driven avoidance of predators—might instead collapse niche structure and destabilize coexistence.
Researchers conducted a six-year, whole-ecosystem manipulation on 16 small islands in the Bahamas. They used a 2x2 factorial design, introducing curly-tailed lizards (the top predator) and/or green anoles (a competitor) to islands already inhabited by brown anoles. The team tracked population trajectories, habitat use, and diet composition using DNA metabarcoding and stable-isotope analysis to determine how these species interacted and partitioned resources under different predation regimes.
Contrary to the keystone predation model, the presence of curly-tailed lizards did not facilitate the coexistence of the two anole species. Instead, the predator induced a behavioral shift: brown anoles, which typically occupy a more terrestrial niche, were forced into the arboreal habitat used by green anoles to avoid predation. This collapse of spatial niche partitioning forced the two species into direct competition for the same limited resources. Consequently, green anole populations were heavily suppressed and often went extinct on islands where the top predator was present. Furthermore, the study found that the top predator acted as a trophic omnivore, which shortened food chains by competing with the prey for arthropod resources.
This research challenges the long-standing assumption that top predators generally enhance biodiversity. It demonstrates that the non-consumptive effects of predation—such as the fear-driven restriction of prey to specific microhabitats—can be a powerful force that limits species coexistence. These findings highlight the importance of considering spatial structure and prey behavior when predicting the outcomes of biological invasions.
Sam: That's an interaction effect, not an additive one. If the predator were simply a stabilizer, I'd expect a longer chain wherever it was added.
Alex: And that non-additivity is what points to the spatial squeeze. The interaction was estimated with a generalized least-squares model that accounted for island area and the factorial treatments. That's a sensible control, since island size could otherwise confound the comparison. But it's a supporting layer. It shows the pattern survives accounting for area, not that the mechanism is proven. [[RP_SECTION:generalization-and-habitat-structure|Generalization and habitat structure]]
Sam: And how far does this generalize? Sixteen small islands is a tidy system.
Alex: That's the main constraint. The effect is demonstrated on small islands with particular vegetation. In larger or structurally richer habitats, prey could have more refuge options, or character displacement could ease the competition. The paper's implication is that the outcome depends on habitat architecture. When structure is simple, the predator's behavioral influence on prey becomes dominant.
Sam: Which matters for management. If you introduce or protect a predator to control one species, you can't assume the effects on the rest of the community are benign. They depend on whether prey can partition space.
Alex: Yes. The authors also suggest the same logic might be tested against other faunal collapses, the Lake Victoria cichlid extinctions among them, by looking for predator-induced spatial compression. That's a hypothesis for future work, not something this study shows.
Sam: If you want the figures and the method choices we skipped, you can generate a deep dive of this paper. The paper has the rest either way.
Alex: Thanks for listening.