The knowledge of trees diversity is very important for the planning of conservation programs. Genetic diversity is fundamental for the evolution of species and it made the plants to be adapted to environment changes since thousand years. Fire, forest destruction and factors like exploitation, affect continuously agro forestry genetic resources (Achard et al., 2002). Most of them are still under-utilised, meanwhile sustainability use of genetic diversity of agroforest contributes to the achievement of the new challenges and the promotion of economic, social and cultural values as well as services and environmental incomes (FAO, 2010). Unfortunately, advances in the fields of genomics and molecular genetics of agroforestry’s species are not remarkable although biochemical and DNA markers are developed for an increasing number of tropical species (Baird et al., 1996; Mhameed et al., 1997; Ude et al., 2006). These tools increases knowledge on forests tree and allows for better plan of the domestication process.

The characterisation of many groups of agroforestry plants is now well established. Among them, is Dacryodes edulis (Don.), a multipurpose tree known for its potential dietary uses and economic properties with a great potential of industrial applications. The plant is a source of edible oils and nutritious fruits (Vivien and Faure, 1996; Ajibesin, 2011). It has been classified as priority species for domestication by the International Centre for Research in Agroforestry (ICRAF) (Tchoundjeu et al., 2002). In central Africa, cultivated forms are integrated into the cropping systems in a region still dominated by subsistence agriculture and is the centre of diversity for the species (Mbeuyo et al., 2013). The ecological and cultural conditions under which traditional agriculture is carried out have helped to preserve large amounts of their diversity (Kengue, 2002). The cultivated forms consist of heterogeneous genetic material, resulting from empirical selections in different ecological regions (Youmbi et al., 2010). Actually, the selection pressure especially among Cameroonian populations, for the choice of plant material by farmers is high and may leads to the erosion of the genetic base of the species. Some populations grown are wild but the species has evolved very little over time (Okafor, 1983; Kengue, 2002). The species is essentially allogamous; pollination is entomophilous and achieved by insects (Fohouo et al. 2002). The pollen can be spread over a long distance (Kengue, 1990). The wild genetic materials could have many important influences on their cultivated relatives, and store great amounts of genetic variation which may be of interest for future crop improvement programs (Kengue, 2002). In addition, they can provide information on the domestication process, and could also represent a potential source of gene flow with their domesticated materials. Thus there is need to enhance genetic richness of this important crop. This will help in identification, germplasm collection, improvement and conservation. The aim of this study was to assess genetic diversity, population structure and gene flow of D. edulis.

Six microsatellites loci (CB09, CC01, CE09, CG11, LB12 and LD06) previously developed by Benoit et al. (2011) were employed. The allele size ranged from 125 to 375 and the repeats motifs concerned were guanine-adenine (GA) and cytonine-thymine (CT). PCR was carried out in two thermocyclers (Ependort and Corbet) in 0.2 mL individual tube and 96 well plates. It started with an initial step of 95°C for 5 min, followed by 35 cycles of 30 s at 95°C, 1 min at 57°C and 30 s at 72°C. Final elongation was carried out at 72°C for 20 min. The products were separated by electrophoresis in 1.8% agarose gels with a running time of 40 min at 100 V in 0.5X TBE buffer. A 50 bp DNA ladder (Invitrogen) was used as molecular size marker. Gels were stained with Gelred and the DNA bands were visualized under ultraviolet light; digital images were recorded using the software Image Aide, version 3.06.04.r®.

Genotyping by capillary gels

For each genotyping sample, 1 µL of PCR product was diluted with 9 µL mixture of 1 mL deionized Formamide and 15 µL ROX-labeled y internal sizing standard with a total reaction volume of 10 µL. The fragments of the DNA were denatured (3 min at 95°C with thermocycler and immediately cooled in ice bath) and size fractioned using capillary electrophoresis on an ABI 3730 automatic DNA sequencer. The GeneMapper software Version 4.1 was applied to size peak patterns, using the internal GS500LIZ 3730 size standard and for allele calling.

Statistical analysis

For each loci site, the number of alleles, expected heterozygot (He), polymorphism index content (PIC) and imbreeding coefficient were calculated using Popgen32 software (Yeh et al., 2000). The genotype frequency within samples was tested for agreement with Hardy-Wemberg expectation by Chi square test to compare observed versus expected outcomes.

Wright’s F-statistics (1965) were calculated by analysis of molecular variance (AMOVA) to evaluate the level of population structure. Values and  were estimated through variance among populations; variance among and within individuals.

The rate of inter-population gene flow among all populations was estimate using the procedure based on Wright’s , where  (Wright,1965). This method estimated  which is the average number of migrants into a population per generation. The calculation of genotype frequency, Wright’s F-statistics and the rate of inter-population gene flow were done by GenAlEx_6.4 (Peakall and Smouse, 2006).

Genetic population differentiation and gene flow

Wright’s F-statistics expressed by F_IS F_ST and F_IT were used to assess population structure by analysis of molecular variance. The probability (F_IS) that two alleles in an individual are identical by descent (relative to the subpopulation from which they are drawn) was equal to 0.058 and showed that there was a high average heterozygotes in each D. edulis population. The probability (F_IT) that two alleles in an individual are identical by descent (relative to the combined population) was equal to 0.069 and showed that there was a high average heterozygotes in a 3 groups of populations. The last probability (F_ST) for which two alleles drawn at random are identical by descent (relative to the combined populations) was equal to 0.012 and showed that the degree of gene differentiation among populations in terms of allele frequencies was low. Otherwise, there were no clear differences with only 1% variation among the three populations, 6% variation among individuals within populations and 93% within individual of the populations. The level of gene flow was estimated by the number of migrant. Both the number of migrant/generation  and the  values suggest that there was substantial gene flow among the populations.

Private alleles were assessed within individuals of the 3 populations. Private alleles in the context of this study are alleles unique to a single population; all the other alleles totaled were shared between them. These alleles were found in 17 individuals of the three populations (8 for Mbakwa supe, 4 for Santchou and 5 for Yaounde). The total number of loci (9) and loci sites (5) in wild forms with private allele was highest as compared to those of the two cultivated forms which were (10) and (5), respectively.

The main areas where microsatellite markers are being applied in forest trees include studies of genetic diversity in natural and breeding populations, particularly in species with low levels of isozyme variation, gene flow, pollen and/or seed dispersal and mating systems. As these parameters are relevant to the conservation of forest genetic resources, microsatellites are being used to monitor genetic impacts of forest management practices and of fragmentation. It is the case of Eucalyptus sieberie (Glaubitz et al., 1999) and Pinus radiate (Echt et al., 1999). In the present study, the number of alleles was high and the loci sites were polymorphic. The calculation of polymorphism index contain suggest that we can get a reliable results of the population genetic estimate using the six SSRs.

There was not structure in the populations. This cannot be understood without considering that there is no genetic drift occurring in some of its subpopulations. In addition, migration might be uniform throughout the population, or mating is random throughout the population. A population’s structure might affect the extent of genetic variation and its patterns of distribution

A high degree of polymorphism was observed within all the groups and particularly in wild forms. In terms of genetic resource, this suggests that wild forms present a significant genetic diversity. In addition, most of the species like D. edulis have a greater genetic diversity among wild population in terms of number of alleles (Saghai-Maroof et al., 1984). Both the number of migrant/generation (Nm=20) and the F_ST=0.012 values suggest that there was substantial gene flow among the populations. Our very low  estimates are lower than the average reported for outcrossing seeds plants (Hamrick, 1989; Vigouroux et al., 2008). These results can be explained by great potential of pollen dispersal and dominant cross-pollination within populations. In addition, in a separate study regarding growing of D. edulis under different cropping conditions in Makenene region of Cameroon, it was found that the majority of D. edulis plants (88%) are planted from seeds derived from adult individuals, while 12% are from conventional breeding techniques (aerial layering). The seeds used are selected from trees bearing large, nonacid fruits that are rich in oils (Mbeuyo et al., 2013). Meanwhile, gene flow between cultivated plants and their wild relatives may have increased likelihood of rapid decrease of wild relatives. In fact, during this surveys, populations in the field reported their decreasing and coexistence with cultivated form in certain regions. However, studies that examine the extent of introgression between wild and cultivated populations are necessary.

There was no significant difference (p < 0.01) between expected and observed genotype frequency in wild’s population. For these reasons, Hardy-Wemberg equilibrium was respected, assuming no unbalanced external factor is affecting wild's population. In contrast, 05 cases of deficiencies in heterozygotes were mentioned in cultivated forms. In the domestication process of D. edulis, farmers grow plants requiring only agronomic and food traits (Kengue, 2002). These cultural practices increased the selection pressure for the choice of planting material by farmers and might progressively lead to the erosion of the genetic base of the species from where deficiencies in heterozygotes is observed in some cultivated individuals.

It is noted from the concluding remarks that there was no difference in the three population’s structure. In contrast, variability and heterozygosis were very high within the populations. There could be substantial gene flow among the populations and the greatest number of private alleles was found in wild’s individuals. For breeding programs, self-pollinations of selected individuals followed by crossing should be useful to produce hybrids of economic and agronomic interest such as size, taste and oil content of fruits. Some wild individuals could be a source of pollen for crossing their cultivated relatives.

The authors have not declared any conflict of interests.

The authors gratefully acknowledge the financial support provided to the Biosciences Eastern and Central Africa Hub at the International Livestock Research Institute (BecA-ILRI Hub) by the Australian Agency for International Development (AusAID) through a partnership between Australia's Commonwealth Scientific and Industrial Research Organization (CSIRO) and the BecA-ILRI Hub; and by the Syngenta Foundation for Sustainable Agriculture (SFSA); the Bill and Melinda Gates Foundation (BMGF); and the Swedish Ministry of Foreign Affairs through the Swedish International Development Agency (SIDA), which made this work possible.

EDTA, Ethylenediaminetetraacetic acid; PVP, polyvinylpyrrolidone; DNA, deoxyribonucleic acid; PCR, polymerase chain reaction; He, expected heterozygoty; PIC, polymorphism index content.