Cluster Analyses of Ostracoda Based on Dimensions of Body Structures: Implications for Taxonomic Classification

Dayou Zhai,Robin J. Smith,Ping Peng,Na Yu,Shunxin Ma,Xiangzhong Li
DOI: https://doi.org/10.1163/15685403-00003667
2017-01-01
Crustaceana
Abstract:We measured selected podomeres, setae and claws in different ostracods and calculated the between-specimen morphological difference, which is expressed as a Canberra dissimilarity index. Our data indicate that morphological differences between ostracods increase with their taxonomic distance. Cluster analyses of ostracod specimens based on Canberra dissimilarity are able to discriminate different species and concur with existing classifications. We suggest that the dimensions of body structures are taxonomically valuable, and that ostracod species identification can be assisted based on the dimensional data of body structures. Species discrimination with such a method does not rely on explicit morphological hiatuses, such as the presence/absence of particular setae, but instead utilizes measurable morphological differences. Our numerical methods also show good potential for studying phenotypic diversity. Analyses on ostracod populations from isolated temporary pools and those from permanent but geographically distant habitats indicate that dispersal improbability is responsible for the observed morphological differentiation.
What problem does this paper attempt to address?