The measurement of taxonomic evenness in zooarchaeology

Zooarchaeologists frequently measure taxonomic evenness to document subsistence change and to understand the response of faunal communities to paleoenvironmental change. Although the measurement of evenness is commonplace, there are numerous challenges involved. Evenness indices are sensitive to changing richness, and by extension sample size, and various indices respond differently to changing taxonomic abundances (i.e., changing evenness). To refine protocol for comparing assemblages of varied sampling effort and identify indices that may be more useful for zooarchaeological applications, we examine the quantitative behavior of the widely used Shannon evenness and Simpson indices and two others more commonly used by ecologists. These indices are examined in relation to varied richness, sample size, and taxonomic abundances. We show that although zooarchaeologists are concerned with identifying and correcting for the effects of sample size on evenness, it may be appropriate to instead focus on how richness modulates evenness. Based on our analyses, we recommend the Simpson index for most zooarchaeological applications, except when comparing evenness across assemblages that are very even.