Mutualistic network architecture and eco-evolutionary feedbacks modulate the occurrence of transitions and stability in response to rising temperature

1. Ecological networks comprising of mutualistic interactions can suddenly transition to undesirable states, such as collapse, due to small changes in environmental conditions such as a rise in local environmental temperature. 2. However, little is known about the capacity of such interaction networks to adapt to changing temperatures and thereby impact the occurrence of critical transitions. 3. Here, combining quantitative genetics and mutualistic dynamics in an eco-evolutionary framework, we evaluate the resilience of mutualistic networks to critical transitions as environmental temperature increases. Specifically, we model the dynamics of a phenological optimum trait that determines the tolerance to local environmental temperature as well as temperature-dependent species interaction and evaluate the impact of trait variation and evolutionary dynamics in the occurrence of tipping points and community collapses. 4. We found that mutualistic network architecture, i.e., community size and the arrangement of species interactions, interacted with evolutionary dynamics to impact the onset of network collapses. In addition, some networks had more capacity to track the rise in temperatures than others and thereby delay the occurrence of threshold temperatures at which the networks collapsed. 5. However, such a result was modulated by the amount of heritable trait variation species exhibited, with high trait variation in the mean optimum trait value delaying the environmental temperature at which the network collapses. 6. Our study argues that mutualistic network architecture modulates the capacity of networks to adapt to changes in temperature and thereby impact the occurrence of community collapses.