Fig 1.
Environmental context of the sampled populations.
a- Principal component analysis (PCA) of 22 environmental variables at Ugandan C. canephora native sites. For the BIOx (see S3 Table), temperature-related variables are shown in grey-brown; precipitation-related in blue. Elevation, aridity index (AI) and potential evapotranspiration (PET) are shown in green. The first two axes, i.e. PC1 and PC2, account for 64.7% and 21.3% of the total variation, respectively. b- The seven natural forests are reported on the Ugandan map of average annual rainfall from the Ugandan Meteorological Service (NEMA, 2009) [34]. The five climatic zones based on precipitation [Matete et al. 2010] [33] are delimitated in grey. They include: 1. Karamoja, 2. Acholi-Kyoga, 3. Lake. Victoria, 4. Ankole-Southern Uganda and 5. Western Uganda areas. Climatic zones:3, 4 and 5 are important with regard to the presence of wild and cultivated of C. canephora populations. National Environment Management Authority (NEMA). 2009. Uganda: Atlas of our changing environment. NEMA, Kampala, Uganda.
Table 1.
Ugandan C. canephora sample origin.
Mean and standard deviation of selected environmental variables across locations. All variables were significantly different across locations based on a one-way ANOVA (p-value < 2.2e-16). Pairwise t-test results between locations are also presented.
Table 2.
Allelic pattern across populations.
Values are provided for each native forest, the feral population and the two collections: Number of alleles, number of effective alleles, number of private alleles, allelic richness, observed and expected heterozygosities, and fixation index. Since Mabira, Malabigambo and Kalangala belong to the same south-central genetic group [SC], data are also provided for this group.
Fig 2.
Genetic structure of Ugandan C. canephora native populations.
a- Percentage membership of the 191 wild accessions to each of K = 3 or 4 population clusters as inferred by STRUCTURE based on 19 SSR loci. The substructure is presented below for Itwara and Kibale individuals with K = 2 sub-clusters. b- Geographical distribution on the Ugandan map with the annual precipitation variable (BIO12) of SSR genetic groups considered according to the STRUCTURE analysis at K = 4. Each location is depicted as a pie chart with the proportional membership of its alleles to each one of the four groups with colours according to graph a. c- Unrooted dendrogram produced using the unweighted neighbour-joining method based on genetic dissimilarity among the 191 accessions. The branch colours indicate accessions corresponding to the five clusters from the population structure analysis, as in a-. Individuals from Mabira, Kalangala and Malabigambo forests are intermixed within a same large southern-central population (SC-pop).
Fig 3.
Ugandan C. canephora diversity relative to the species’ global diversity structure.
a- Neighbour-joining of Ugandan wild C. canephora material together with feral accessions from Kalangala islands (noted “Kl-F” in black) and cultivated material maintained in Kituza and Kawanda collections (in red). The six elite clones, KW13, KW14, KW15, KW16, KW18 and KW19 are indicated. Wild material was collected from Zoka, Budongo, Kibale, Itwara, Malabigambo, Mabira and Kalangala islands (Kl). The colours of branches of wild material correspond to the five clusters from the population structure analysis presented in Fig 2. Individuals representative of other genetic groups (A, B, C, D, E, R) from the whole species diversity are also presented for reference. b- Eight divergent genetic groups of C. canephora and their geographical distribution (adapted from [23]).
Fig 4.
Constrained redundancy analysis (RDA) of the effects of environmental parameters on the allelic diversity of Ugandan C. canephora wild populations.
Twelve out of the 22 environmental variables used in this study significantly and collectively explained 16.3% of the total variation in allelic diversity. Colour codes are the same as in Fig 1. The first and second constraining axes explained 4.63% and 3.33% of the total genetic diversity, respectively.