Taxon-and-Area Invariances, Maximum Entropy, and the Species-Area Relationship (A Reply to Harte et al.)
Abstract
Harte et al. (2009, 2013) argue that maximum entropy theory for ecology (METE) predicts a universal shape of species-area relationship (SAR), which is independent of the focal taxon. They argue that a “biological meaningful” splitting of the focal taxon just increases the amount of information available, which consequently leads to a more accurate prediction, which is understandably different from a less accurate prediction.
They state that such splitting does not violate the universality that was challenged by Sizling et al. (2011). Although we agree that the METE is mathematically consistent, assuming that the constraints used for calculating the most likely state are appropriate (namely, eq. [1c] in Harte et al. 2008, which provides the only biologically relevant constraints; the other constraints are mathematical necessities), we feel that Harte et al. (2013) miss our point.
While METE focuses on the most likely prediction of a single observation—which may be the same regardless of which group of species is used—our reasoning concerns possible patterns observed in data.