Some Sampling Properties of the Fager Index for Recurrent Species Groups

Abstract

A Monte Carlo technique was used to produce sets of samples from a series of artificial 2—species communities. The Fager r ecurrent group index was calculated for these samples. The effects of sample numbers were assessed by comparing the sample indices to size—corrected population index values. A small negative bias was noted for indices calculated from the smaller sample sets, particularly in those communities where the species were rare. The index appears to give conservative estimates of affinity when only a small number of samples are taken.