Robustness to secondary extinctions: Comparing trait-based sequential deletions in static and dynamic food webs

The loss of species from ecological communities can unleash a cascade of secondary extinctions, the risk and extent of which are likely to depend on the traits of the species that are lost from the community. To identify species traits that have the greatest impact on food web robustness to species loss we here subject allometrically scaled, dynamical food web models to several deletion sequences based on species’ connectivity, generality, vulnerability or body mass. Further, to evaluate the relative importance of dynamical to topological effects we compare robustness between dynamical and purely topological models. This comparison reveals that the topological approach overestimates robustness in general and for certain sequences in particular. Top-down directed sequences have no or very low impact on robustness in topological analyses, while the dynamical analysis reveals that they may be as important as high-impact bottom-up directed sequences. Moreover, there are no deletion sequences that result, on average, in no or very few secondary extinctions in the dynamical approach. Instead, the least detrimental sequence in the dynamical approach yields an average robustness similar to the most detrimental (non-basal) deletion sequence in the topological approach. Hence, a topological analysis may lead to erroneous conclusions concerning both the relative and the absolute importance of different species traits for robustness. The dynamical sequential deletion analysis shows that food webs are least robust to the loss of species that have many trophic links or that occupy low trophic levels. In contrast to previous studies we can infer, albeit indirectly, that secondary extinctions were triggered by both bottom-up and top-down cascades.