Most abundance counts in paleontology and zooarchaeology use one of two major methodologies, Number of Identified Specimens (NISP) or Minimum Number of Individuals (MNI). However, in many situations, I have seen fossils from a given locality catalogued as several different specimens, even if it is obvious they belong to a single individual. For example, in several cases, a complete mammal skull will have the cranium be given one specimen number, the left dentary a second, and the right one a third, even if the jaws were found in occlusion. In these situations, especially if the counts are being done on a locality where elements are not found close to one another and therefore two given specimens are normally unlikely to belong to the same individual (e.g., a multi-square kilometer outcrop of a geological formation), does one count associated and articulated parts as a single "specimen"?
In addition to what Steve has mentioned concerning identification by association, there is also the issue of scale of measurement. Donald Grayson in his book, Quantitative Zooarchaeology (1984) mentions that our measurements of abundance and comparisons from these are ordinal scale at best. This means that when comparing abundances either within a single collection, or across collections, we can only make comparisons in terms of more/less abundant, rather than at ratio scale (such as stating an animal is 50% more abundant than another). Sometimes, we cannot make ordinal level comparisons and can only make nominal level comparisons such as presence or absence of a taxon. Grayson's book may still be out of print, but Lee Lyman has an entire chapter about measurement scale in his 2008 book Quantitative Paleozoology. Throughout his book Lyman also emphasizes that no mater which measurement is used, it is important to be explicit on how it was calculated, as some units have multiple ways to calculate them.
When comparing abundances, especially from multiple sites, it is important to account for the taphonomic histories of each collection. Some taxa may be less abundant or absent in a collection because of preservation or recovery practices, not because they weren't present in the past.
On a side note, see Lyman's comment in Driver's reprinted article in Ethnobiology Letters that Steve offers a link to in his answer. Lyman stresses the importance to explicitly report how identifications were made and the results of identification in a systematic paleontology.
@Jamie R. Wood
I am a little confused by this. I had always heard that in an MNI count, you ignore any individuals that cannot be tallied by the element you are using in question. I.e., if you have seven left adult humeri of species X at a site but you also have one right humerus that belongs to an individual that is clearly much younger than any of the others, you would still only report an MNI of 7.
@Laura Ellyson
As a side note, if I am reading your response right, the presence/absence of a taxon can be used as a form of comparison in the case of taxa that are rare?
The primary issue I am having is that I am working at a site where specimens are generally not found in close association with one another (i.e., an attritional assemblage, not a bonebed). Some specimens represent isolated bones, whereas others are associated or even articulated, but most are highly separated from one another (being primarily recovered by surface prospecting and not even necessarily from the same level of the stratigraphic column). Using a traditional NISP count would grossly overestimate the abundance of many groups (including one in particular that has a higher tendency to produce lots of tiny, well-preserved fragments) whereas using a traditionally MNI would also not accurately represent the data because it would not count associated specimens found kilometers apart as representing different individuals.
In the case of rare taxa, you can compare them based on presence or absence in an assemblage, but only if you can support that your sample is representative by accounting for the effects of taphonomic processes and sample size. One way to evaluate taxonomic composition and sample size is to produce a species area curve. An example of this can be seen in Figure 2 of Ronald Wolff's article in Paelobiology (1975 vol. 1 no. 2, "Sampling and sample size in ecological analyses of fossil mammals"). With increasing sample sizes, more taxa are usually added. When the number of taxa no longer increases with sample size (the curve becomes and remains flattened as larger samples are added), this is usually termed as having "sampled to redundancy", meaning that a sample is fairly representative of an animal community. With increasing sample sizes, rare taxa have a higher probability of being encountered, assuming that both natural and cultural taphonomic processes have not greatly impacted the sample in terms of presence/absence (if from an archaeological assemblage, it may not have been collected in the past, skeletal elements may have been modified beyond identification to a finer level such as genus, recovery bias during excavation, etc.).
In answer to the question asked, if there is a very good reason to believe that material belongs to one individual, standard curatorial practice would be to catalogue them with one specimen number. But if there is any doubt, it would likely be preferable to catalogue them with separate numbers, perhaps with a note in the catalogue indicating a possible association.
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology
- Ornithology