Abstract: The idea of conserving biodiversity is central to what biological conservation has come to mean – in the public imagination and for conservation organizations worldwide. This understanding of conservation as centrally about biodiversity conservation is largely built on an empirical thesis – that biodiversity is a significant causal determinant of ecosystem functioning – coupled with an economic thesis – that this biodiversity-caused functioning broadly underlies invaluable services, which ecosystems afford humankind. These coupled theses have fueled an enormous Biodiversity Ecosystem Function (BEF) research program dedicated to producing the requisite causal evidence for the empirical thesis.
I first review the data that are supposed to constitute direct evidence for biodiversity's causal influence. My method is to address a heretofore unasked, yet foundational, question, which should be asked about any causal hypothesis: Do the data meet basic requirements for credibility as causal evidence? I observe that, in this case, scientists have read causal significance into (i) massive numbers of causally irrelevant data points, (ii) an equation that merely presents a necessary equivalence between a stipulated definition and an algebraically equivalent expression, and (iii) correlations produced by computations that are formulated explicitly in order to produce those correlations. As a consequence, the data fall well short of meeting requirements for causal credibility.
Next I review the conservation argument for which biodiversity's hypothesized causal influence was supposed to be the crucial empirical premise. Informed assessment of this argument has heretofore been precluded for want of a consolidated account of the full complement of required premises and all required connecting logic. Here I spell out each step, which makes it possible to see that each one relies on a questionable assumption, invalid logic, or both.
I conclude with implications for conservation, biodiversity research, and scientific inquiry more generally.
