Colonization, covariance and colour: Environmental and ecological drivers of diversity–stability relationships

Understanding the mechanisms that underlie the relationship between community diversity and biomass stability is a fundamental topic in ecology. Theory has emphasized differences in species-specific responses to environmental fluctuations as an important stabiliser of total biomass fluctuations. However, previous analyses have often been based on simplifying assumptions, such as uniform species abundance distributions, uniform environmental variance across species, and uniform environmental responses across species pairs. We compare diversity–stability relationships in model communities, based on multi-species Ricker dynamics, that follow different colonization rules during community assembly (fixed or flexible resource use) forced by temporally uncorrelated (white) or correlated (red) environmental fluctuations. The colonization rules generate characteristic niche-dependent (hierarchical, HR) environmental covariance structures, which we compare with uncorrelated (independent, IR) species’ environmental responses. Environmental reddening increases biomass stability and qualitatively alters diversity–stability patterns in HR communities, under both colonization rules. Diversity–stability patterns in IR communities are qualitatively altered by colonization rules but not by environmental colour. Our results demonstrate that diversity–stability patterns are contingent upon species’ colonization strategies (resource use), emergent or independent responses to environmental fluctuations, and the colour of environmental fluctuations. We describe why our results arise through differences in species traits associated with niche position. These issues are often overlooked when considering the statistical components commonly used to describe diversity–stability patterns (e.g., Overyielding, Portfolio and Covariance effects). Mechanistic understanding of different diversity–stability relationships requires consideration of the biological processes that drive different population and community level behaviours.

► We model competitive communities to investigate diversity–stability relationships. ► Most previous theory points to positive diversity–stability relationships. ► Direction of these relationships depends on community assembly. ► Environmental colour and correlation interact to alter diversity–stability patterns. ► Our results contradict earlier work based on various simplifying assumptions.