When Less is More: Element Selection as Sampling Strategy in Zooarchaeology

Abstract

Zooarchaeologists make a range of analytical decisions that affect taxonomic identifications, measures of abundance, and, by extension, socio-environmental interpretations. One decision that is often not documented and/or justified is the question of which skeletal elements, or element fragments, should be identified. This situation is especially concerning for, but not limited to, zooarchaeological analysis of fish remains, as ichthyoarchaeologists have historically adopted varied approaches to element selection for taxonomic identification, ranging from identification of a few distinctive jaw elements to attempted identifications for all elements. Here, we argue the process of element selection should be approached more intentionally as a sampling issue, which would provide a framework for evaluating whether identified samples are representative of the recovered assemblages from which they derive, and potentially improve the efficiency of analysis. We use nestedness analysis and sampling to redundancy to illustrate the effects of element selection on taxonomic richness and relative abundances in fishbone assemblages from three cultural and biogeographic regions: the northeast Pacific, the Hawaiian Islands, and Aotearoa New Zealand. We find that both methods are effective for evaluating the representativeness of identified assemblages based on varied element sets, and our results demonstrate that, in many assemblages, identifying all elements provides redundant information about taxonomic representation (richness and relative abundance). In such cases, one could select fewer elements for identification and spend more time examining remains from additional parts of a site or carrying out specialized analyses, which would provide new insights about past environments or human-animal relationships. The proposed analytical strategy aims to help zooarchaeologists justify and make informed decisions about how many and which specific elements to include in any given analysis. Our study adds to the growing discourse on best practices in zooarchaeology.