Recognizing morphospecies in the heliolitid coral Plasmoporella

Abstract

Multivariate analysis, which has often been used effectively in paleontological work, was performed on 17 specimens of Plasmoporella collected from a bed of the Xiazhen Formation in the Zhuzhai section of Yushan area, China. Frequency histograms, simple correlation analysis, and principal component analysis (PCA) showed that five morphological characters, i.e., diameter of tabularium (DC), distance between neighboring septa (DNS), distance between opposite septa (DOS), distance between corallites (DBC), and number of corallites per cm2 (N) were suitable for cluster analysis. Moreover, PCA suggested that the coralla could be divided into two groups. Cluster analysis with squared Euclidean distances calculated based on the selected five morphological characters was conducted on the raw matrix to discriminate “morphospecies”. As a result, two groups were extracted in the dendrogram. The validity or distinctiveness of the groups was tested by discriminant analysis (DA). The analysis confirmed that the assemblage of coralla can be divided successfully into two groups, which are interpreted as two morphospecies. A qualitative comparison between two reported species and the studied morphospecies showed that morphospecies A is Plasmoporella shiyanshanensisLin and Chow, 1977 and B is Plasmoporella kasachstanicaBondarenko, 1958. In addition, DA indicated that DOS and DNS were effective to distinguish the two morphospecies. Bivariate plots based on the two morphospecies were used to illustrate the range of variation, which showed that intraspecific variation in these characters was obvious but interspecific variation was not.