Species distribution models are commonly used for predicting the effects of climate change including potential extinctions, however, such models treat species independently and neglect to consider interactions between species.
Based on a rich trait dataset of seven European pollinator networks, this study explores potential climate-induced extinctions while considering mutual dependencies and thus co-extinctions. Adding GBIF-mediated occurrences, authors progessively projected models of 244 plant species into future climates, assigning a co-extinction probability each time an interaction would be lost.
The results show substantial variability across networks with dramatic increases in co-extinction rates in the Mediterranean networks—compared to the other networks. In fact, the 2080 projection shows a near-doubling of climate-induced extinctions when potential co-extinctions are considered.
Not surprisingly, some trait profiles are less likely to be driven to climatic extinction than others, however, the same profile may indeed have a significantly larger risk of dissappearing through subsequent co-extinctions.