Species delimitation, phylogeography and population genetics of the endemic Malagasy dwarf lemurs (genus Cheirogaleus)

Abstract

This dissertation addressed aspects of species delimitation, phylogeography and population genetics of the endemic Malagasy genus Cheirogaleus. Morphometric, molecular and distributional data at multiple scales from all available data sources, including museum collections, sequence data from GenBank and newly collected specimens from the field, were used to answer the following questions: Is the current taxonomy, which recognizes seven species, supported by new data from the field? Morphological and genetic data are concordant in the support of the existence of three Cheirogaleus species and thus the current taxonomy is not fully supported by our analyses. To what conclusion regarding the number of Cheirogaleus species does the analysis of morphological data from field and museum specimens lead ? Analyses of morphological data, from six external and 32 cranio-dental characters of 120 museum specimens and 36 individuals from the field, revealed three distinct morphs, which correspond to C. medius, C. major and C. crossleyi. Cheirogaleus adipicaudatus and C. ravus are proposed to be synonymous with C. medius and C. major, respectively. The sample size for C. minusculus is too small for decisive inferences and results for the status of C. sibreei are inconclusive. What can be concluded from the analyses of mitochondrial DNA and nuclear DNA markers in terms of species delimitation? Analyses based on mtDNA (cytb+cox2) and three independent nuclear markers (adora3, fiba and vWF) from 48 individuals caught in the field, 17 museum specimens and 24 haplotypes from GenBank, support three evolutionary lineages, which correspond to C. medius, C. major and C. crossleyi. Population genetic cluster analyses revealed a further layer of resolution within the three lineages and identified two distinct sets of populations/individuals per lineage. How are the extant Cheirogaleus species distributed geographically and how did those patterns arise? The three recognized species, as assessed by the morphological and genetic analyses, are not local endemics, but have distributions spanning the island. Cheirogaleus medius is found along the west coast from the southeastern to the northeastern tip of the island. Cheirogaleus major is present along the east coast from the southeastern tip up to the Masoala peninsula and C. crossleyi is found in the eastern rainforests from the southeastern to the northernmost tip and also on the northwestern coast as far south as Ampijoroa. Can existing biogeographic hypotheses for Madagascar be supported by new distributional data and a reassessed Cheirogaleus taxonomy? Neither the model of biogeographic zones based on phytogeography, nor the `centers-of-endemism' hypothesis is concordant with, or explains the contemporary distribution of the currently recognized species of dwarf lemurs. When did the different Cheirogaleus species begin to differentiate and how do their time divergence estimates compare to closely related taxa? The age estimates for dwarf lemur species range from 1.5 to 8.1 mya (95% credibility intervals) and are comparable to estimates for mouse lemur species at 1.5 to 12.8 mya (95% credibility intervals) and thus species level divergences in both genera largely predate the Quaternary. This indicates that models, such as the `centers-of-endemism' hypothesis, which assume that climatic shifts in the Quaternary were the driving force of speciation, do not apply. What can be deduced from the population genetic structure of a C. medius population from western Madagascar? Evidence for equal dispersal by both sexes in a pair-living primate species, C. medius, was found. This dispersal seems to occur to a limited degree, creating local populations that are potentially quite differentiated behaviorally from nearby populations. On the one hand the confounding factor of genetic signatures due to potentially asymmetrical dispersal patterns of the sexes can thus be disregarded in future mtDNA based studies. On the other hand population differences were found on a small spatial scale and it is therefore questionable how representative the conclusions of this study are. What does the present population genetic structure of a C. medius population in western Madagascar reveal about the demographic history of that population? Analyses of mtDNA variability at the population level, i.e. of 140 individuals from two subpopulations, revealed very low levels of genetic variability combined with high haplotype diversity, which is indicative of a recent population bottleneck. Which sex disperses in C. medius and what is the consequence in terms of population genetic structure? No evidence for spatial clustering of same-sexed individuals with identical haplotypes within each of two C. medius subpopulations was found, indicating equal dispersal by both sexes. Comparison with a closely related sympatric species demonstrated that differences in social organization and dispersal patterns result in different genetic structures: In Microcebus murinus females had a larger aggregation index and FST than males, whereas in C. medius the sexes showed equal variances in the number of individuals representing each haplotype, as well as equal levels of aggregation of identical haplotypes and almost equal FST estimates. Are these data representative for the whole species? We discovered population differences in density and sex ratio at the smallest spatial scale, indicating behavioral and social flexibility in this species, which suggests that any randomly chosen study site may not be representative for the social system of a local population, or even species.