Data from: Trapped in the web: network architectures spread coevolution and shape adaptation

Adaptation is critical for biodiversity to persist under global change.
Within ecological communities, species often face trade-offs between
adapting to shifting abiotic conditions and navigating the complex
selective pressures imposed by interaction networks. We hypothesize that
network architectures characterized by high interaction diversity and
overlap constrain coevolutionary dynamics, with asymmetric outcomes for
exploiters and victims. Specifically, we predict that exploiters, subject
to spread and conflicting selection imposed by their victims, will evolve
more slowly and show reduced capacity to track victims’ evolutionary
responses, with these constraints strongest for generalist exploiters. In
contrast, victims will show more variable dynamics depending on the
coherence of selection (i.e., whether pressures from different exploiters
push the victim’s trait in the same vs. different directions). To test
this, we simulated trait evolution in coevolving communities of exploiters
and victims across 91 empirical networks, and in artificial networks
designed to isolate specific structural effects. Our results show that
higher connectance, species richness, nestedness, and centrality
homogenize biotic effects and increase fluctuations in trait matching,
ultimately weakening coevolutionary coupling. Under these conditions,
exploiters face conflicting selection that slows evolution, whereas
victims either benefit from aligned selection that accelerates evolution
or are constrained by multiple pressures. Together, our findings suggest
that network architecture plays a fundamental role in shaping coevolution
and adaptation, and raises broader questions about its influence on
eco-evolutionary processes in more complex and environmentally variable
systems.