Geometric morphometrics and molecular systematics of Xanthocnemis sobrina (McLachlan, 1873) (Odonata: Coenagrionidae) and comparison to its congeners
Milen Marinov,Catalina Amaya-Perilla,Gregory I Holwell,Arvind Varsani,Katherine Van Bysterveldt,Simona Kraberger,Daisy Stainton,Anisha Dayaram,Nathan Curtis,Robert H Cruickshank,Adrian Paterson
DOI: https://doi.org/10.11646/zootaxa.4078.1.9
IF: 0.9
2016-02-09
Zootaxa
Abstract:The taxonomy of the damselfly genus Xanthocnemis is revised, with particular focus on populations inhabiting the North Island of New Zealand. Earlier studies revealed two species: X. sobrina, restricted to cool, shaded streams in kauri forests and other forested areas, and X. zealandica, a common species throughout New Zealand except the Chatham and subantarctic islands. A field study encompassing aquatic habitats throughout the whole North Island was carried out to establish the relationship between morphological variation (body size and various morphological traits over the entire body) observed by previous researchers with ecological conditions and/or geographical location. The main aim was to propose reliable diagnostic features that could be used in future studies. Morphological and molecular variation was assessed. Morphological examination included assigning landmarks for all body parts corresponding to the external morphological features that are usually used in Odonata taxonomy. Molecular analysis targeted fragments of the 28S and 16S rRNA genes. Congruence was sought between both types of data, statistical support for two morphological types previously described as different species and a maximum likelihood phylogenetic tree in conjunction with a pairwise genetic distance matrix constructed from the DNA sequences obtained from the sampled specimens. Geometric morphometrics revealed statistically significant differentiation between specimens identified as X. zealandica and X. sobrina for four traits: (1) dorsal view of the head for both sexes as well as male appendages from (2) dorsal, (3) ventral and (4) lateral views. Wings appeared different when analysed for males only. Molecular analysis, however, grouped all specimens into a single undifferentiated cluster with very low mean pairwise distance (<0.01) between them showing almost no variation at the molecular level among the sampled populations on the North Island. Therefore, an additional analysis of the mitochondrial cytochrome c-oxidase I gene was carried out comparing randomly selected North Island specimens to Xanthocnemis specimens targeted in other molecular studies (Nolan et al. 2007, Amaya-Perilla et al. 2014). The analysis of the COI gene confirmed that all North and South Island isolates of Xanthocnemis cluster together in a well-supported clade with pairwise identity >96% and ~93% pairwise identity with X. tuanuii sequences obtained from the Chatham Island specimens. A careful investigation of the thin plate spline deformations generated for the geometric morphometric landmarks showed that the significant variations in the appendages of the Xanthocnemis specimens appeared to be the result of size, rather than shape, differences. Therefore, X. sobrina is proposed as a synonym of X. zealandica. Recently Amaya-Perilla et al. (2014) synonymised X. sinclairi with X. zealandica and confirmed the status of the Chatham Island X. tuanuii as a distinct species. It is therefore proposed that the genus Xanthocnemis consists of two species only: zealandica occurring all over the North, South and Stewart Islands, and tuanuii, endemic to Chatham and Pitt islands. Considering several statistical tests involving body measurements and ecological variables recorded during the field study, as well as various discussion points from similar studies of other species of Odonata, two alternative hypotheses are proposed for future testing. The first hypothesis synonymises X. sobrina with X. zealandica and suggests a possible explanation for the evolution of the two morphological traits that have previously been considered diagnostic for these species. The second hypothesis suggests that as typical X. sobrina were not sampled during this study this could represent a species that is now extinct, unless future studies prove it otherwise.