Godoy, O.; Calleja-Solanas, V.; Soler-Toscano, F.; Portillo, J. R.; Garcia-Fuentes, J.; Stouffer, D.; Langa, J. A.

The assembly of ecological communities depends on both species' demography and the interactions among them. Ecological theory assumes that both ecological processes vary independently. However, empirical evidence demonstrates otherwise. How the structural stability of ecological communities changes when species' intrinsic growth rates and interactions are coupled remains unknown. To address this knowledge gap, we present several seemingly disconnected examples, such as the stress-gradient hypothesis, the covariance between environment and competition, higher-order interactions, and niche packing, that actually fall along a continuum of changes ranging from positive to negative interactions. Although the specific details of each scenario differ, they all share a common feature. When demography and interactions are coupled in multispecies models of population dynamics, the limits of structural stability in ecological communities shift non-linearly, and the domain of viability increases or decreases compared to a non-coupling scenario. We mathematically demonstrate that whether these changes are positive or negative depends on how the strength of intraspecific interactions shifts relative to interspecific interactions as species' growth rates increase. These new non-linear relationships imply two important shifts of thinking. Theoretically, the definition of structural stability must be generalized and its estimation requires the use of probabilistic measures based on ecological limits. Empirically, if abiotic and biotic barriers operate simultaneously, we need to account for such non-linear processes that can significantly modify population dynamics. Otherwise, we will not make accurate predictions of how ecological communities are assembled and biodiversity is maintained.