The maintenance of genetic variability and geneflow are critical factors determining the long term survival of species. For wide-ranging, highly exploited and vulnerable species, such as Khaya senegalensis, strategies for their conservation and on-going utilisation can only be effective if detailed accounts of their geographic and genetic distributions are known. K. senegalensis is a eurytopic species of the northern savannah belt of tropical Africa (Sahelo-Sudano-Guinean bioclimatic region). Its life history traits have endowed it with the innate ability to tolerate a wide range of climatic and edaphic conditions. This versatile species produces one of the finest furniture timbers in the world (African mahogany) and is a provider of human and veterinary medicines, and fuelwood to tens of millions of people in this region. It is also a highly important fodder species, which may have contributed to this species decline. K. senegalensis is now classified as vulnerable on the IUCN Red List, and in some countries it is listed as endangered. The main objectives of this phylogeographic study were: (1) to ascertain the geographic distribution and conservation status of K. senegalensis (Meliaceae) across the northern savannah belt of tropical Africa using historic and current literature, herbarium specimens, and field data and observations; (2) to develop an extensive germplasm (with data) collection for molecular studies, ex situ conservation plantings in Australia and development of a practical conservation strategy; (3) to develop molecular markers (nSSR) for genotyping; (4) to identify patterns of genetic variability and geneflow in this species and its populations; and (5) to evaluate the role geography and climate may have played in shaping this species genetic variability and structure.
Field excursions were undertaken in 2005 to initially investigate the locations of seed collection sites. A second visit was made in 2007 to collect germplasm (seed) and to gather information about the species across its natural range. Due to the many access issues encountered in this region at the time, initial sampling focused on the West African countries. Additional germplasm collections were obtained with the support of government and non-government agencies in countries between Senegal and Cameroon. Central-east African (Chad to Uganda) samples were obtained from provenance trials in northern Australia (reported in Karan et al. 2011). Further detailed information on this species was obtained from botanical institutions (literature and herbarium) at Oxford, Kew and Montpellier.
The majority of DNA samples (n=505) were extracted from the leaf of progeny used as surrogates for the maternal trees in Africa. Further extraction trials revealed that DNA extracted from husk was maternal and was thereafter used in obtaining additional DNA samples. Microsatellite markers were developed specifically for this study to assess genetic variability and population structure in K. senegalensis. This exercise followed the testing of 23 microsatellite markers, which were known to amplify in closely related species. However, the majority of these markers provided little to no detail on the level of genetic variability in this species. Analysis of nSSR data revealed very weak genetic structure (Fst = 0.013) across the natural range of K. senegalensis, with most of the genetic variability contained within populations. This data supports the notion that K. senegalensis forms a single population across its entire range which may be associated with life-history characteristics that promote dispersal; including a flowering and fruiting period that coincides with strong trade winds (Harmattan) which sweep across this region, small seeds that are easily wind dispersed or through recent human mediated dispersal.
In an attempt to understand the dynamics of this species over space and time, fragments from three chloroplast genes (combined 1333bp sequence) were sequenced to infer genetic difference in a subset of individuals (n=92) from 14 populations across the natural range of this species. Multilocus phylogenetic analyses were conducted using maximum likelihood and Bayesian algorithms. Fourteen haplotypes were identified; two haplotypes were common throughout the natural range of this species, while twelve rare haplotypes occurred only in western Africa (Senegal to Cameroon). Only the two cosmopolitan haplotypes were found to exist in central-east Africa (Chad to Uganda). These findings suggest that climatic oscillations of the late Pleistocene not only compressed and fragmented the equatorial rainforests, but also caused significant fragmentation of the northern savannah belt in western Africa. The lower level of genetic diversity experienced in populations in central-east Africa may suggest that the overall integrity northern savannah belt in this region was largely maintained by the recolonisation of territory abandoned by the southward retreat of the Congolian rainforest during past glacial maximums and again with the northward retreat of the Sahara Desert during interglacial periods. Due to the low sample size used in this study, nuclear data was included in an attempt to increase the confidence level of this study, however, caution is still required with these interpretations and a greater sample size is desirable (if possible).
Finally, this study describes the botany and habit of this species, identifies the environments it inhabits, maps its natural distribution, identifies the threats to its survival, and provides a practical set of strategies for its long term conservation.