Beyond variance: simple random distributions are not a good proxy for intraspecific variability in systems with environmental structure

Abstract

AbstractThe role of intraspecific variability (IV) in shaping community dynamics has been intensively discussed over the past decade and modeling studies have played an important role in that respect. However, a major, but often implicit, assumption typically made by these studies, that IV can be represented by independent random draws around species-specific mean parameters, has largely remained undiscussed. A great part of observed IV is however structured in space or time, in particular when resulting from environmental dimensions that influence individual performance but are imperfectly characterized or unobserved in the field. To test this assumption, we designed a simulation experiment where we varied the level of knowledge of the environment in virtual communities, resulting in different relative importance of explained vs. unexplained individual variation in performance. We used a community dynamics simulator to generate communities where the unexplained individual variation is, or is not, added as an unstructured random noise. Communities simulated with unstructured IV never reached the community diversity and composition of those where all the variation is explained and structured (full knowledge model). This highlights that incorporating unstructured IV (i.e. a random noise) to account for unexplained (but structured) variation can be misleading. Also, comparing communities simulated with the same level of knowledge of the environment, but adding unstructured IV or not, we found that the effects of incorporating unstructured IV depended on the relative importance of structured versus unstructured IV. In particular, increasing the proportion of unstructured IV into the model moved from a positive to a negative effect on community diversity and similarity in composition with the full knowledge model. This suggests that it is crucial to account for the sources and structure of observed IV in real communities to better understand its effect on community assembly and properly include it in community dynamics models.