Environmental factors are hypothesized to affect the functional diversity of assemblages hierarchically. First, ecological filters constrain the range of functional traits potentially displayed by an assemblage by determining its taxonomic composition. Second, some factors determine actual functional traits through the phenotypic plasticity of species. Little is known about the relative importance of each set of factors and, by using a priori functional information, most empirical studies report only the variability expected from species turnover and neglect that associated to phenotypic plasticity. Herein, we use structural equation models to assess the factors driving the functional richness, evenness, and divergence of a top-predator assemblage faced with high variability in resource availability and assemblage structure (i.e. species richness and abundance). We measured actual functional traits (i.e. diet composition and predation pressure) in the field and contrasted the effects of environmental filters and phenotypic plasticity (i.e. behavioral responses) by controlling for species turnover and sample size with a null model. We found that a priori estimations (i.e. null-model expectations) tended to significantly underestimate (richness and divergence) or overestimate (evenness) functional-diversity components, explaining just a fraction (13–63%) of the variability in observed values. Furthermore, while species richness strongly affected functional richness (positively) and evenness (negatively), and resource availability slightly affected functional divergence, via compositional effects, changes in functional-diversity components attributable to behavioral responses of predators showed little or no association with environmental variables. As a consequence, results indicated that in productive years, functionally-distinctive species exerting relatively low predation pressure entered the assemblage, increasing functional richness and decreasing functional evenness. However, the strong behavioral responses of functionally dominant species buffered such compositional changes, affecting to different extents the three functional-diversity components. Thus, we argue in favor of considering phenotypic plasticity in future studies of functional diversity.