The interactions between species in ecological communities form highly connected webs. The removal of species from such webs can cause loss of vital interactions and cause a cascade of secondary extinctions. One approach currently taken to examine the consequences of extinction in these webs, emulating studies from other fields of network science, has been to look at the robustness of both real and model networks to random and ordered extinction sequences. Secondary extinctions are inferred from the topology of the network. In these studies a species was only deemed to suffer a secondary extinction when it was left with no remaining prey, i.e. species were allowed to loose up to 100% of their original energy requirement before they became extinct.

In this study we used allometric optimal foraging theory to look at how the robustness of model food webs changed if we varied the threshold of energy loss a species could sustain before going extinct. As a secondary aim we looked to see how the size-structure of the webs mediated the robustness of the webs and how it interacted with the different energy requirement thresholds.

Results/Conclusions

As might be expected, setting more stringent energy requirements for the species made the model food webs less robust to species loss, and extinction cascades were more easily triggered. These thresholds are as important for robustness as the extinction order and connectance of the webs. Interestingly, the degree of size-structure of a web affects its robustness to different extinction orders and possibly different thresholds. The implications of these results for the likelihood of a cascade of co-extinctions in real food webs will be discussed.