Disparity is a measure of the range or significance of morphology in a given sample of organisms, as opposed to diversity, which is expressed in terms of the number (and sometimes ranking) of taxa. At present there is no agreed definition of disparity, much less any consensus on how to measure it. Two possible categories of metric are considered here, one independent of any hypothesis of relationship (phenetics), the other constrained within an evolutionary framework (cladistics).
The Early Cambrian radiation was clearly a period of significant morphologic and taxonomic diversification. However, we question the interpretation of its first generation products as numerous body plans at the highest level. Four phenetic and two cladistic measures have been used to compare disparity among Cambrian arthropods with that in the living fauna. Phenetic methods assessing character-state variability and the amount of morphological attribute space occupied yield similar results for Cambrian and Recent arthropods. Assessments of disparity within a taxonomic framework rely on the identification of particular characters that delineate higher level body plans. This requires a phylogenetic interpretation, a cladistic investigation of hierarchical structure in the data. Both sets of arthropods fall within the same major clades, and within this cladistic framework the amount of character-state evolution in the two groups is comparable. None of these methods identifies markedly greater disparity among the Cambrian compared with the Recent taxa.
Although measures of disparity are applied here to a consideration of the Cambrian radiation, the metrics clearly have a much wider potential for estimating macroevolutionary trends independently from existing taxonomic frameworks. Geometric morphometry is ideal for measuring morphological variety at lower taxonomic levels, but it requires the recognition of homologous landmarks in all the forms under comparison, or the identification of entire homologous structures. Conventional phenetics has much wider application as it can operate on data coded as discrete homologous character states (this facility is also a requirement of cladistics), which are a more appropriate basis for comparing disparity in markedly dissimilar forms.