Phylogeography of the widespread African puff adder (Bitis arietans) reveals multiple Pleistocene refugia in southern Africa. Issue 4 (3rd January 2013)

• Record Type: Journal Article
• Authors: Barlow, Axel; Baker, Karis; Hendry, Catriona R.; Peppin, Lindsay; Phelps, Tony; Tolley, Krystal A.; Wüster, Catharine E.; Wüster, Wolfgang
• Is Part Of: Molecular ecology. Volume 22:Issue 4(2013)
• Page Start: 1134
• Page End: 1157
• Publication Date: 2013-01-03
• Subjects: Molecular ecology -- Periodicals; Molecular population biology -- Periodicals; 576
• DOI: 10.1111/mec.12157
• Languages: English
• ISSNs: 0962-1083

Abstract

<abstract abstract-type="main" id="mec12157-abs-0001"> <title>Abstract</title> <p>Evidence from numerous Pan‐African savannah mammals indicates that open‐habitat refugia existed in Africa during the Pleistocene, isolated by expanding tropical forests during warm and humid interglacial periods. However, comparative data from other taxonomic groups are currently lacking. We present a phylogeographic investigation of the African puff adder (<italic>Bitis arietans</italic>), a snake that occurs in open‐habitat formations throughout sub‐Saharan Africa. Multiple parapatric mitochondrial clades occur across the current distribution of <italic>B. arietans</italic>, including a widespread southern African clade that is subdivided into four separate clades. We investigated the historical processes responsible for generating these phylogeographic patterns in southern Africa using species distribution modelling and genetic approaches. Our results show that interior regions of South Africa became largely inhospitable for <italic>B. arietans</italic> during glacial maxima, whereas coastal and more northerly areas remained habitable. This corresponds well with the locations of refugia inferred from mitochondrial data using a continuous phylogeographic diffusion model. Analysis of data from five anonymous nuclear loci revealed broadly similar patterns to mtDNA. Secondary admixture was detected between previously isolated refugial populations. In some cases, this is limited to individuals occurring near mitochondrial clade contact zones, but in other cases, more extensive admixture is evident. Overall, our study reveals a complex history of refugial isolation and secondary expansion for puff adders and a mosaic of isolated refugia in southern Africa. We also identify key differences between the processes that drove isolation in <italic>B. arietans</italic> and those hypothesized for sympatric savannah mammals.</p> </abstract>