Cross‐scale interactions mediate local drivers of nonnative species richness

Many well‐supported hypotheses seek to explain drivers of nonnative species richness across spatial scales, but evidence for common patterns among regions and taxa remains inconclusive. This study investigates why consistent patterns are elusive by estimating and assessing cross‐scale interactions, wherein large‐scale factors contextualize patterns measured at smaller scales. We investigated whether local relationships of disturbance and native species richness with nonnative species richness are moderated by regional native gamma diversity. Regions with higher gamma diversity, we hypothesized, would be unfavorable to nonnative species due to high levels of competition and reduced niche availability, thus mediating local effects of native richness and disturbance on nonnative species richness. Using a fine resolution stream fish community dataset covering 159 regional watersheds in the conterminous United States during 2000–2023, we quantified cross‐scale interactions using a two‐level Bayesian hierarchical model. In the first level, we estimated the effects of disturbance and native richness on nonnative richness in local stream segments indexed by region. In the second level, we used this regional index to estimate cross‐scale interactions of native gamma diversity (regional‐level richness) on the first‐level relationships. Local nonnative richness was generally positively associated with native richness and disturbance. However, these relationships were reduced in regions with more diverse native stream fish assemblages. Thus, native gamma diversity provided an important mechanistic context for local nonnative richness relationships across regional watersheds through a negative cross‐scale interaction. As large spatial datasets become increasingly available, accounting for cross‐scale interactions in inter‐regional observational studies will be critical for understanding ecological relationships and may provide a predictive framework for studies with conflicting support for differing conceptual models.