The term ‘beta diversity’ is surprisingly nebulous. According to its original definition by Whittaker, beta diversity refers to the variation in species composition along environmental or spatial gradients. But through the years there has been debate regarding the precise definition of the term, and even more debate about how to measure it. In our new paper, my friend Ben Gilbert and I show that the best way of measuring beta diversity among regions is through multivariate distance measures such as the Sørensen distance. Other methods, such as the classical forms proposed by Whittaker and a newer null model technique, provide biased results that are dependent on sampling effort. Even if two regions receive equal sampling effort, their beta diversity differences, as measured by classical or null-model techniques, are not directly comparable. Multivariate methods should therefore be the standard for comparison.
How to compare beta diversity among regions: