Measuring beta diversity with presence-absence data.

Six measures of beta diversity (5 from the literature, one proposed here) were compared and evaluated. Application was limited to measures suited for species presence-absence data along environmental gradients. Four ecological criteria of 'good' performance of beta diversity measures were developed: 1) conformity with the notion of community turnover ensures that the magnitude of a measure is meaningful; 2) additivity is the property that the sum of beta diversities betwen contiguous segments equals the beta diversity of the entire gradient; 3) independence from alpha diversity ensures useful application of a measure to systems with different alpha diversities; 4) independence from excessive sample size obviates any spurious effects of oversampling. Two measures of beta diversity (one proposed by Whittaker (1960) and one proposed here) came closest to fulfilling all 4 criteria.-from Authors