Some three-species motifs (unique patterns of interactions between three species) are both more stable when modelled in isolation and over-represented in empirical food webs. This suggests that these motifs may reduce extinction risk for species participating in them, ultimately stabilizing the food web as a whole. We test whether a species' time to extinction following a perturbation is related to its participation in stable and unstable motifs and assess how motif roles covary with a species' degree or trophic level. We found that species' motif roles are related to their times to extinction following a disturbance. Specifically, having a larger proportion of the motif role made up by the omnivory motif was associated with longer times to extinction, even though the omnivory motif is less stable than the others in when modelled in isolation. While motif roles were associated with extinction risk, they also varied strongly with degree and trophic level. This means that these simpler measures of a species' role may be sufficient to roughly predict which species are most vulnerable to disturbance (though motif roles can be used to refine these predictions), but that studies of species' motif participation can also reasonably comment on vulnerability to extinction. Previous research showed that certain motifs appear more frequently in stable webs; we show that this is at least in part because participating in these motifs provides protection from extinction.

Food webs were simulated using the script simulate_initial_foodwebs.jl. Species removals were simulated using the script simulate_removals.jl.

Both scripts require Julia and were written to be compatible with Julia version 1.6.3. Code used to simulate food webs and disturbance dynamics are available from the Zenodo Digital Repository: https://doi.org/10.5281/zenodo.6377974 (Cirtwill and Wootton 2022).