Abstract
Understanding changes in biodiversity is a complex subject contingent on many interacting or poorly differentiated processes. As a result, it is desirable to organize processes in community ecology into a small number of high-level mechanisms that completely account for change in ecological communities. It has been suggested that all change in ecological communities can be partitioned into four mechanisms: 1) selection, 2) drift, 3) immigration and 4) speciation; however, the completeness of this framework requires testing. Here we use insights derived from one of the most fundamental equations in evolution, the Price equation, to quantify the strength of selection, drift, immigration and speciation in simulations. We show how the impacts of each of these mechanisms can be quantified using experimental and simulated data and find that these four mechanisms cannot account for large portions of the change in these simulated communities. This gap is a consequence of a fifth mechanism fundamental to evolutionary theory, transmission bias, which describes change in the measurements associated with organisms. Examples of transmission bias from evolution include phenotypic plasticity and selection within groups. Our results highlight the distinction between biodiversity change and the processes that change species' relative abundances. Selection, drift, immigration and speciation change species' relative abundances, but many diversity measures summarize an additional piece of information: a measurement of species' rarity. Species' rarity changes over time, and these changes are not included when the influence of selection, drift, immigration and speciation are quantified. Accounting for changes in species' rarity by adding transmission bias to our list of fundamental mechanisms leads to a complete accounting of biodiversity change.