Are type 3 functional responses just statistical apparitions?

Abstract

Type 3 (sigmoidal) functional responses can protect rare prey from predation, but such functional responses are not often described. It is also not clear that factors that are hypothesized to promote type 3 functional responses, such as predator–prey size ratios and higher temperatures, are actually associated with type 3 curves, especially given experimental design concerns. Here, we assess support for type 2 or type 3 functional responses for a compilation of functional response experiments and evaluate the role of experimental design and biology in generating type 3 functional responses. We find that experimental design considerations generate support for type 3 models more than biological factors, and after accounting for experimental effects, both predator–prey size ratio and temperature increase the support for type 3 models, in line with predictions. However, no factor explains much variation in support for functional response type (full model R² = 0.08). We also show that the standard model selection cutoff for support for one model over the other is inconsistent across bootstrapped samples, indicating that support for type 3 models may emerge stochastically due to noise in foraging trial data. Finally, we suggest a more conservative approach is needed for assessing support for type 3 models.