Genetic variability of Chaerephon atsinanana (Chiroptera) within the context of the Afro-Malagasy Molossidae : a mitochondrial and nuclear perspective.

Abstract

This study has focused on genetic variability and structure in Chaerephon atsinanana, a newly-described molossid bat found in the mid to southern region of the eastern watershed of Madagascar. As these bats are strong fliers, and are able to traverse the riverine and mountain barriers within the landscape, it was hypothesized that they would show relatively low levels of intraspecific genetic structure, consistent with patterns shown for other Molossidae on Madagascar (Mormopterus jugularis, Mops midas, Mops leucostigma, and C. lecuogaster. Phylogenetic (neighbor-joining, parsimony and Bayesian inference) and population genetic analyses of maternally-inherited mitochondrial control region sequences revealed the presence of 6 distinct haplotype groups separated by genetic distances of up to 8.14% (mean 4.95%). There were high levels of genetic structure among the haplotype groups (overall FST= 0.994). Thus the hypothesis of low levels of genetic structure was rejected. Bayesian skyline analyses and significantly ragged mismatch distributions were consistent with ancient stable C. atsinanana populations which were of constant size during the last two major Pleistocene glacial periods. This made retreat into and expansion from glacial refugia an unlikely explanation for such high levels of structure. An alternative hypothesis is that C. atsinanana haplotype groups are spatially structured as a result of philopatry. As mitochondria are maternally-inherited, this data is consistent with the existence of female philopatry in C. atsinanana. The second aim of this study was to examine the genetic structure of C. atsinanana with nuclear sequence markers, which are biparentally-inherited, in order to provide information on the male contribution to gene flow and the possible presence of male philopatry in this molossid bat species. The initial objective was to amplify and sequence candidate nuclear markers in order to identify those which were variable among C. atsinanana samples. I attempted to amplify and sequence a set of 12 nuclear markers, identified from the literature, which had been reported to show high levels of variability, or which were untested and showed the potential for high levels of variability. Of these, the intron markers PNPO-3, SLC38A7-8, CARHSP1-1, GAD2-1, OSTA-5 had not previously been used in phylogenetic analyses while FES, GHR, RHO1 CHRNA1, STAT5, PRKC1 and THY had been. I was not able to amplify and/or sequence SLC38A7-8, CARHSP1-1, GAD2-1, OSTA-5, CHRNA1, STAT5 and THY across the range of the C. atsinanana samples. PNPO-3, FES, GHR, RHO1 and PRKC1were successfully amplified and sequenced, but showed no variability and very little polymorphism, and were therefore unsuitable for testing hypotheses related to genetic variability of C. atsinanana populations. These five nuclear sequence markers were further used to investigate phylogenetic relationships among 5 genera (Chaerephon, Mops, Mormopterus, Otomops and Sauromys) and 13 species of Afro-Malagasy molossid bats, and to provide a nuclear phylogenetic perspective on the newly-described C. atsinanana. PNPO-3 is a novel nuclear intron marker, previously unused in phylogenetic studies. This intron provides resolution primarily at the genus level, and is less informative at interspecific level. These five nuclear markers were combined with already existing mitochondrial cytochrome b (Cyt b) and nuclear Rag2 data retrieved from GenBank. This study provides strong support for the monophyly of the Chaerephon and Mops taxa included, with the exception of C. jobimena, which was weakly associated with this group. There was no support for the generic affiliation of C. jobimena or for the monophyly of either of the genera Chaerephon or Mops individually. This leads to the suggestion that Mops and Chaerephon be combined into a single genus, with crown age of 14.82 (6.44-25.54) MYA, or 21.97 (12.16-33.44) MYA if C. jobimena is included. Otomops forms a strongly supported clade consistent with its generic status, comprising two subclades corresponding to the recognised sister species O. martiensseni and O. madagascariensis, which last shared a common ancestor 8.35 (2.87-17.47) MYA. This study provides good nuclear support for the mitochondrially-defined subclades of O. martiensseni, which last shared a common ancestor 4.18 (1.08-9.96) MYA. It would appear appropriate to name the clade from north east Africa and Arabia as a new species of Otomops, as the clade from southern and western Africa includes the type locality. This study provides weak support based on individual gene regions for associations of Sauromys with Otomops and Mormopterus, although these do not stand up in the concatenated datasets which offer better resolving power, indicating that Sauromys is not phylogenetically associated with Chaerephon/Mops, Otomops and Mormopterus. These results provide some support for the membership of Mormopterus in the proposed Old World Molossid tribe, Tadarini, but also support Mormopterus as a basal genus within the Molossidae, consistent with its designation as a separate tribe, Mormopterini.