Trait-dependency of trophic interactions in zooplankton food webs

Anthropogenic change in the abundance or identity of dominant top
predators may induce reorganizations in whole food webs. Predicting these
reorganizations requires identifying the biological rules that govern
trophic niches. However, we still lack a detailed understanding of the
respective contributions of body size, behaviour (e.g., match between
predator hunting mode and prey antipredator strategy), phylogeny and/or
ontogeny in determining both the presence and strength of trophic
interactions. Here, we address this question by measuring zooplankton
numerical response to fish predators in lake enclosures. We compared the
fit to zooplankton count data of models grouping zooplankters based either
on (i) body sizes, (ii) antipredator behaviour, (iii) body size combined
with antipredator behaviour, or on (iv) phylogeny combined with ontogeny
(i.e., different life stages of copepods). Body size was a better
predictor of zooplankton numerical response to fish than antipredator
behaviour, but combining body size and behaviour provided even better
predictions. Models based on phylogeny combined with ontogeny clearly
outperformed those based on other zooplankton grouping rules, except when
phylogeny was poorly resolved. Removing ontogenetic information plagued
the predictive power of the highly-resolved (genus-level) phylogenetic
grouping but not of medium-resolved or poorly-resolved phylogenetic
grouping. Our results support the recent use of phylogeny as a superior
surrogate for traits controlling trophic niches, and further highlight the
added value of combining phylogeny with ontogenetic traits. Further
improvements in our mechanistic understanding of how trophic networks are
shaped are bound to uncovering the trophic traits captured by phylogeny
and ontogeny, but that currently remain hidden to us.