Reeves' muntjac: a molecular genetic study of an invading species

The possibility of screening large numbers of individuals directly for genetic variability at the DNA level has only become feasible within the last decade. Prior to this, gel electrophoresis was employed to resolve small differences in the amino acid sequences of proteins. This thesis describes the application of three molecular genetic techniques to elucidate the relationships within sub-populations as well as the distribution and relationships between groups of sub-populations of an invading species of deer; Reeves muntjac (Muntiacus reevesi). Samples were taken from eighteen U.K. sub-populations of muntjac, representing increasing geographic distances centred on Woburn Abbey (the putative centre of origin) and one group from Taiwan were investigated.

The polymerase chain reaction was used to amplify the control region of muntjac mitochondrial DNA prior to digestion with restriction enzymes. The restriction fragments produced were found to be sufficiently informative to identify eight maternal lineages and demonstrate that genetically closely related groups are in fact geographically widely separated, a finding counter to the hypothesis of a regular mode of dispersal. Genetic partitioning between sub-populations, a result which may be expected given the short time since the introduction of muntjac to the U.K.

An appropriate DNA fingerprinting protocol was established, based on published protocols and two commercially available minisatellite probes. Detailed analysis of the banding patterns produced by these probes allowed an estimate of the level of inbreeding within sub-populations to be established. Tentative relationships were also able to be established between the sub-populations sampled. Although there was some overlap between probes, they generally detected independent loci. The mean band sharing coefficients were found to range from 0.19 to 0.33 (probe 33.15) and 0.15 to 0.34 (probe 33.6).

The polymerase chain reaction was used to amplify the control region of muntjac mitochondrial DNA prior to digestion with restriction enzymes. The restriction fragments produced were found to be sufficiently informative to identify eight maternal lineages and demonstrate that genetically closely related groups are in fact geographically widely separated, a finding counter to the hypothesis of a regular mode of dispersal. Genetic partitioning between sub-populations was assessed by three agglomerative clustering methods which demonstrated that there is very little geographic partitioning between sub-populations, a result which may be expected given the short time since the introduction of muntjac to the U.K.

Genetic variability between sub-populations was also assessed by an investigation of the geographic distribution of seven microsatellite loci. Most microsatellite loci were found to be highly polymorphic and there was substantial variation in the number of alleles detected per sub-population. Heterozygosity values were found to be high, ranging from 0.48 to 0.74. Fourteen 'rare' alleles were uncovered distributed between ten of the eighteen sub-populations. Wright's F\(_{IS}\) was calculated as a metric of the level of inbreeding within sub-populations. The mean values of this estimator ranged from -0.181, indicating heterozygosity, to 0.222, indicating homozygosity. Nei's G\(_{ST}\), an analogue of Wright's F\(_{IS}\) was calculated to give an indication of the level of genetic sub-structuring between sub-populations. These calculations indicated that differentiation by distance is not significant over the geographic range of the sample area, again indicating that the muntjac have no dispersed in a 'natural' way.

The use of three different molecular techniques has allowed comparisons to be made between their relative merits in terms of the level of genetic information they provide and in their ability to define population structure.