A graphical causal model for resolving species identity effects and biodiversity-ecosystem function correlations: comment.

Abstract

In a recent paper, Schoolmaster, Zirbel, and Cronin (2020) (SZC) claim "Formal causal analysis show[s] that biodiversity-ecosystem function (BEF) correlations are non-causal associations." If this conclusion is accepted as true, it suggests a reconsideration of much of our current understanding of how biodiversity relates to the functioning of ecosystems. On the surface, it is easy to spot clear signs of something problematic with SZC's presentation. They claim, for example, that (a) species richness is incapable of having a causal effect on ecosystem functioning on theoretical grounds and (b) that trait diversity cannot be causally influenced by species diversity. These remarkable claims are counter to existing thought and evidence. We point to logical errors that lead them to a misapply causal analysis and produce erroneous conclusions.