Understanding the origins and diversification of Indian blindsnakes

Abstract

Much of systematics done in the past was based on morphological data. One of the drawbacks of using morphological data is that in groups with conserved morphology, the diversity is often underestimated and relationships difficult to infer. One such group are the fossorial blindsnakes (superfamily Typhlopoidea, infraorder Scolecophidia). The diversity, life histories and ecology of such groups remain unexplored due to their sub-terranean and secretive habit. India has around 17 species, spanning four genera and two families. However, many of these species have not been discovered since their first descriptions, and many are suspected to be synonyms. For my PhD, I studied the diversity of the four typhlopoid genera in India, from a phylogenetic and biogeographical perspective, and attempted to understand diversification patterns within one species group, which gave rise to the Indotyphlops braminus, a blindsnake with a surprisingly cosmopolitan distribution. In the first chapter, I explore the phylogenetic positions of the four genera- Gerrhopilus, Indotyphlops, Grypotyphlops and Argyrophis, in the global phylogeny. I use a concatenated dataset of five nuclear markers to reconstruct the phylogenies using maximum likelihood and Bayesian methods. The results show that Indian typhlopoids have very diverse affiliations. Gerrhopilus is retrieved as sister to all other typhlopoids. Grypotyphlops is nested with the African genera, which corroborates its classification based on morphology. Indotyphlops was shown to be polyphyletic, with Indotyphlops porrectus being sister to all southeast Asian typhlopoids, and thus requiring a taxonomic revision. We also find an Indian radiation of Indotyphlops. Argyrophis from India is sister to the Argyrophis from southeast Asia. The second chapter looks at the biogeographic origins of the four genera. The divergence times were estimated using the concatenated five gene dataset used in the first chapter, with nine fossils used to calibrate the time-tree. The time-tree was then used for ancestral area estimation in BioGeoBEARS, implemented in R. I carried out a time stratified analysis to accommodate the dramatic changes in the position of the Indian landmass over geological time and model fitting to compare multiple dispersal and vicariance hypotheses. The best fit model invokes both dispersal and vicariance as the explanations for the current distribution of typhlopoids in India. India harbours an ancient Gondwanan group of typhlopoids as well as comparatively younger dispersals from Africa and Asia. The third chapter looks at more fine scale diversification of blindsnakes in peninsular India, particularly Indotyphlops braminus. Indotyphlops braminus is an enigmatic blindsnake that has a pan-tropical distribution, most probably due to human mediated dispersal. It is also proposed to be the only obligate parthenogenetic snake as of now. Karyotyping studies have shown Indotyphlops braminus to be triploid, which has been proposed to be a result of hybridization, and a possible cause of the parthenogenetic reproduction. I investigated the discordance between mitochondrial and nuclear phylogenies to understand whether Indotyphlops braminus is a result of hybridization between two Indian species. I also explored additional lines of evidence by looking at the discordance between gene trees and species trees and a statistical test for hybridization. The results strongly suggest that this cosmopolitan, triploid, parthenogenetic taxon is indeed a hybrid of two Indotyphlops species found in India, but it is not the true Indotyphlops braminus. This new, hybrid species, therefore, merits a taxonomic revision. The underlying theme of my thesis is understanding the origins and diversification patterns in the Indian typhlopoids at broad and fine taxonomic scales, from phylogenetic and biogeographical perspectives.