Diversity, stability and ecosystem functioning relationships across marine fish communities facing contrasting environmental and anthropogenic pressures

Understanding the mechanisms that underpin ecological stability has traditionally focused on the role of diversity, and particularly through biodiversity and ecosystem functioning (BEF) relationships. Yet, BEF relationships and the drivers of stability remain poorly explored across contrasting systems exposed to multiple pressures, such as exploited marine fish communities. Here, we used bottom trawl surveys from 54 marine demersal communities in the North Atlantic and the Mediterranean Sea to quantify the direction and strength of BEF relationships and to assess the relative importance of community structure, environmental, and anthropogenic pressures for different facets of stability across a broad geographic gradient. We combined linear mixed-effects models to analyze BEF relationships, and structural equation models to evaluate the direct and indirect drivers of stability measured through portfolio effect (PE) metrics that either ignore or account for the mean–variance relationship (Taylor’s power law) of the species within communities. Our results demonstrate how richness and evenness can have both positive and negative effects on demersal fish abundance along a macroecological gradient, with the strength of these relationships varying according to the community properties and the regional context. Stronger PE occurred when species fluctuated more asynchronously, presumably due to the increase in niche differentiation and species’ response diversity to the environment. Moreover, synchrony also channeled the indirect and positive effects of richness and evenness on PE, and all these community properties mediated the indirect effects of environmental conditions and anthropogenic pressures on stability. Accounting for the Taylor’s power law z-value when calculating stability metrics was fundamental because it determined the relationship with diversity and further showed how the PEs related differently with the z-value. Therefore, while biodiversity was linked to stability, the mean–variance scaling should not be ignored in order to better understand the mechanisms that operate the biodiversity–stability–functioning relationships in communities subjected to complex interplays among population dynamics, environmental forcing, and anthropogenic pressures.