Quantification of Faunal Remains
Vertebrate and invertebrate faunal remains are typically among the most abundant and informative categories of cultural remains recovered from local archaeological sites. An unresolved methodological issue concerns the most effective manner in which to describe and analyze these remains quantitatively.
Local faunal materials have usually been sorted initially by species, genus, higher taxon, or size class, and usually also have been segregated by provenience. Vertebrate remains have sometimes been sorted by anatomical element and portion of element, symmetry, age, sex, and condition. Quantification has been based on specimen counts or weights, or both.
There are several potential objectives in quantifying faunal recovery. These include expressing the relative richness in faunal remains from particular proveniences, assessing the relative abundance of particular taxa (or other categories, such as age or element), estimating the absolute nutritional or other value of the resources, indicating the statistical significance of any gaps in the range of materials that are represented in the assemblage, indicating the potential richness of the site in faunal remains for future investigations, assessing changes in the characteristics of a particular population through time, and maintaining the identification of curated collections.
To meet these objectives, several methods of quantification have been suggested (cf. Grayson 1984). Each has its advantages and drawbacks:
- Number of Identified Specimens (NISP). This has been the method that has been most widely used for reporting vertebrate remains. It has also been used fairly often for invertebrate remains, particularly in the case of small assemblages. One advantage is the low processing cost, except in the case of abundant and highly fragmented remains, such as some shell midden samples. However, the validity of comparisons between proveniences or between taxa are seriously compromised by the uncontrolled and often-important factor of specimen fracturing. Counts may also give a misleading impression of the relative importance of taxa because of size differences.
- Weight. This has been the primary method for reporting large invertebrate assemblages, although its validity has been questioned (Cerreto 1990:78). It has also been used for invertebrate remains. Weighing in addition to counting represents a small additional cost. Weights may be distorted by incomplete cleaning of the specimens or by postdepositional attrition of the bone or shell. Comparisons between taxa or between proveniences are generally more meaningful with weights than with NISP, but statistical tests of randomness versus patterning cannot be made directly from weights.
- Minimum Number of Individuals (MNI). This method, long popular in some regions, has occasionally been used locally for invertebrates (based on hinge counts, etc.) and vertebrates (based on the most abundant anatomical element). It overcomes some but not all of the deficiencies of the NISP statistic. Additional problems in the case of vertebrate MNI concern the choice of provenience units to be lumped together and the difficulty of projecting whole-site populations from limited excavation samples.
- Estimated Individuals by Weight (EIW). A rarely-used alternative to MNI is to divide the total weight of bone or shell (for a taxon and provenience) by an average weight for a whole individual belonging to the taxon (e.g. Laylander and Saunders 1993). Standard bone or shell weights for most taxa are not readily available, and there may be high variability in weight between individuals. However, this method converts weights into a form in which statistical patterns can be tested and projections of absolute resource value can be made. Vertebrate remains that are not identifiable beyond size class can be included in such analyses.
- Meat Equivalent by Element (MEE). A more precise method of estimating absolute resource value is based on the average amount of meat associated with particular anatomical elements. Increased costs of analysis are one drawback to the method. Another is the exclusion of the large proportion of specimens that are too fragmented to be identifiable to genus.
Multiple methods of quantification, used in combination, may also be able to shed light on issues that cannot be effectively addressed by any single method. For bivalve shellfish, the total weight of specimens assigned to a taxon divided by a hinge count provides an estimate of the average valve weight, even in severely fragmented assemblages in which whole-value measurements would be impracticable. Comparisons of average valve weight between different proveniences may provide information on such matters as harvesting strategies, resource overexploitation, or the natural robustness of the mollusk population.
PROSPECTS
Future archaeological investigations may be able to clarify the efficiency and analytical effectiveness of various methods of quantification, used singly or in combination.
| NISP | Weight | MNI | EIW | MEE | Weight / MNI |
|
| Comparisons between proveniences | 2 | 1 | 3 | 1 | 1 | 1 |
| Comparisons between taxa | 2 | 1 | 2 | 1 | 1 | 3 |
| Statistical tests for patterning | 2 | 3 | 2 | 1 | 1 | 1 |
| Estimation of whole assemblage from sample | 1 | 1 | 3 | 1 | 1 | 1 |
| Estimation of absolute resource importance | 3 | 1 | 1 | 1 | 1 | 1 |
| Added cataloging costs (for bone) | 1 | 2 | 3 | 2 | 3 | 3 |
| Added cataloging costs (for shell) | 2 | 1 | 1 | 1 | 1 | 2 |
| Frequency of local use (for bone) | 1 | 2 | 2 | 3 | 3 | 3 |
| Frequency of local use (for shell) | 2 | 1 | 2 | 2 | 3 | 2 |
Ratings: 1 = favorable method; 2 = somewhat unfavorable method; 3 = strongly unfavorable method
NISP = number of identified specimens; MNI – minimum number of individuals; EIW = estimated individuals by weight; MEE = meat equivalent by element