Genetic structure and mating system of wild cowpea populations in West Africa

Abstract

Background: Cowpea is a highly inbred crop. It is part of a crop-weed complex, whose origin and dynamics is unknown, which is distributed across the African continent. This study examined outcrossing rates and genetic structures in 35 wild cowpea (Vigna unguiculata ssp. unguiculata var. spontanea) populations from West Africa, using 21 isozyme loci, 9 of them showing polymorphism.

Results: Outcrossing rates ranged from 1% to 9.5% (mean 3.4%), which classifies the wild cowpea breeding system as primarily selfing, though rare outcrossing events were detected in each population studied. Furthermore, the analyses of both the genetic structure of populations and the relationships between the wild and domesticated groups suggest possibilities of gene flow that are corroborated by field observations.

Conclusions: As expected in a predominantly inbred breeding system, wild cowpea shows high levels of genetic differentiation and low levels of genetic diversity within populations. Gene flow from domesticated to wild cowpea does occur, although the lack of strong genetic swamping and modified seed morphology in the wild populations suggest that these introgressions should be rare.

Figure 1

Geographical locations of the 35 sampling sites reported in Table1. All locations correspond to wild cowpea populations, excepted BEN05, BUR03, GHA26, NIG04, NIG10, NIG12, and NIG13 where wild and domesticated populations were sampled.

Figure 2

Pattern of isolation by distance. Pattern of isolation by distance in the wild populations. Correlogram plot of the genetic autocorrelation coefficient as a function of the geographical distance classes; dotted lines define the 95%-confidence interval based on 999 permutations.

Figure 3

Principal coordinate analysis. PCO (Principal coordinate) map (Axes 1 and 2) of the sampled populations based on their genetic distances. White dots correspond to domesticated populations; black dots, wild populations. The two first axes accounted for 73.6% of the total variation (43.2% and 30.4% for axis 1 and axis 2, respectively).

Figure 4

Autocorrelation analysis. Autocorrelation analysis of wild-domesticated pairs of populations (r_W-D) performed on asymmetrical distance matrices. Plot of the genetic autocorrelation coefficient as a function of the geographical distance classes; dotted lines define the 95%-confidence interval based on 999 permutations.