CARACTERÍSTICAS PARTICULARES DEL ÁREA DE ESTUDIO

All of the examined sanctuary populations represent a mixture of former historic populations with varying levels of admixture between historic populations. The mixture during the 1970s of animals originally from the Chyulu Hills region and the central highlands within the Solio Game Reserve had a profound effect on the levels of admixture in the current sanctuary populations. Animals from SGR have been used as founders for many of the existing sanctuaries in Kenya including OPC, MRS, LWC and LNP. Examination of the founders from SGR for MRS and OPC indicates that extensive admixture occurred amongst the small number of founding SGR rhinoceros. The current national park populations however show little evidence of admixture even for the four cluster model of genetic structure.

Interestingly the former extensive populations from the west and the north of the country are not represented in the sampled sanctuary populations (Brett 1993, Walpole et al. 2001). A single male at OPC is confirmed as having originated from the north of country, however due to misidentification in the field, this animal was not sampled as part of the study. An absence of haplotypes from these areas in the current study is possibly a reflection of localised management practices and these historic populations may well be represented in other protected areas, such as the Maasai Mara and Lake Nakuru National Park. Although ten animals from Lake Nakuru National Park which were moved to MRS in 2006, were genotyped as part of this study and had haplotypes corresponding to the geographic regions already identified.

It is apparent that distances measures between populations support the broad differentiation of current populations based on relative compositions of central and southern historic populations. FST values for microsatellite data show there is no significant differentiation

between the national park populations and only little differentiation between LWC and OPC, possibly due to the relatively high content of central highland genes in the OPC population. MRS shows moderate differentiation from LWC and NOPC probably due to

the relative lack of Nairobi area genes in both of these populations compared to MRS. However differentiation between the complete OPC population and MRS is low, presumably due to the inclusion of the Nairobi genes contained within the original SWG population. Population differentiation for mtDNA data supports the microsatellite FST

values, although moderate population differentiation based on microsatellite FS T values are

shown to be strong or very strong for ΦST, which is presumably a function of the admixture

of nucleic genes between historic populations.

Bottleneck events of the magnitude experienced by the Kenyan black rhinoceros population are usually accompanied by reductions in genetic diversity resulting in elevated risks of extinction (Gilpin & Soule 1986). The results of this study demonstrate that high levels of genetic diversity are preserved within the remnant eastern black rhinoceros population in Kenya, despite over a 99% reduction in numbers in 17 years. Previous genetic studies on the black rhinoceros have produced conflicting results, with some studies indicating low levels of genetic diversity within populations based on allozymes (Merenlender et al. 1989), mtDNA (Ashley et al. 1990) and microsatellites (Nielsen et al. 2008), whilst other studies have found high levels of diversity in allozymes (Swart et al. 1994), mtDNA (Brown & Houlden 2000) and microsatellites (Garnier et al. 2001, Harley

et al. 2005). Discrepancies in the results from previous studies might be due to differing

demographic histories of sample populations, unrepresentative sample sizes or due to differences in marker variability in the case of the mtDNA studies. No previous study has however examined the effect of the mixing of historic populations on diversity measure. Two previous studies with the largest sample sizes found levels of microsatellite diversity comparable to those shown in this study (Garnier et al. 2001, Harley et al. 2005). Harley et

al (2005) using the same microsatellite markers as used in this study, found expected

heterozygosity of 0.675 and allelic diversity of 5.56 in a population of 19 D. b. michaeli in South Africa founded from just 6 individuals from Kenya. Similarly Garnier‟s et al (2001) study of a single population of 35 D. b. minor founded with 12 individuals from a variety of locations in the Zambezi Valley found expected heterozygosity measures of 0.614 and an allelic diversity of 4.0 for ten microsatellite loci. Our results (HE = 0.689, A = 5.0)

although based on much larger sample sizes are comparable with both of these studies. Brown and Houlden (2000) analysed a 450 bp segment of the same mtDNA control region examined in this study and found higher levels of haplotype diversity (h = 0.86) amongst a captive population of D. b. minor (N= 9) also originally from the Zambezi Valley. Previous studies have suggested that the Zambezi Valley rhino population is particularly

important due to its high levels of genetic variation and historic large population size (Swart et al. 1994, Garnier et al. 2001). The apparent recolonisation of Kenya by black rhinoceros during the middle Pleistocene indicates a possible bottleneck of the eastern subspecies which could potentially account for the differences in mtDNA diversity measures between the D. b. minor populations in the Zambezi Valley and the D. b.

michaeli examined in this study.

The relationship between mtDNA haplotype diversity and microsatellite diversity indicates that the levels of genetic diversity within populations are related to the amount of admixture of historic populations. Examination of the studbook and mtDNA analysis for LWC show that the founders of the population included rhinoceros from the north of the country, Chyulu Hills, Amboseli and the central highlands. Demographic stochasticity, variance in reproductive output and lineage sorting (Allendorf & Luikart 2007) has resulted in a low haplotype diversity, with 77% of the population having a haplotype from the central highlands. The levels of LWC nucleic diversity are consequently low with an allelic richness of 3.87 and an expected heterozygosity measure of 0.637. Whilst these diversity indices are comparable with the studies that have reported high levels of genetic diversity in other black rhinoceros populations, they are the lowest amongst the sampled Kenyan populations (Garnier et al. 2001, Harley et al. 2005). Although LWC is the oldest sanctuary in the study (est. 1984) the time between its foundation and 2006 when the study was undertaken represents only about 3 generations of black rhinoceros (Goetting-Minesky & Makova 2006). Therefore whilst demographic stochasity and lineage sorting appears to have had an impact on diversity indices, there is no evidence of inbreeding (FIS = 0.024)

within this population given the relatively small number of generations since its foundation. In contrast to LWC, the SWG population which was established in 1989 is comprised of mtDNA genotypes representing historic populations from the central highlands (50 %), the south east of the country (24%) and the areas around Nairobi (26%). The newly introduced population (NOPC) is comprised of 44% (N= 11) of animals with a south eastern haplotype (DB03) and 56% (N=14) with a central highlands haplotype (DB06). Consequently the total OPC population represents a considerable period of rhinoceros evolution within Kenya, which is reflected in the diversity indices for both nucleic and organelle genomes. The link between diversity of geographic origin and genetic diversity within the admixed Kenyan populations is further reinforced by the relatively new MRS population which is comprised of mtDNA haplotypes from all sampled geographic areas with the exception of Bura. The MRS population has the highest

levels of genetic diversity of any of the sampled black rhinoceros populations with a haplotype diversity of 0.822, an allelic richness of 4.57 and an expected heterozygosity of 0.757.