Assessment of genetic variation in natural populations of marama bean (Tylosema esculentum) using molecular markers

Abstract

Abstract by author:

The level of genetic variation in four natural populations of the Marama bean taken from Namibia and Botswana was investigated using AFLP and RAPD molecular screening techniques. DNA was extracted from the leaf tissue of randomly chosen Marama bean individuals from the four sites, and the amount of genetic variation within population, between populations and regions was evaluated

A total of 1558 fragments were scored using both RAPD and AFLP marker systems of which 86 percent were polymorphic and 14 percent monomorphic, indicating that high variability exists in the Marama bean species. Proportion of genetic diversity within population is 85. 2 percent whereas the distribution of diversity between populations is 14. 8 percent. However no significant difference (p 0. 05) was detected in the mean genetic diversity between populations. Genotypes were not clustered into distinctive groups corresponding to their geographic origin. Moreover no significant difference could be detected in mean genetic diversity between regions (Namibia and Botswana), p 0. 05 indicating that regional differentiation did not occur. Also no significant differences (p 0. 05) were detected between red and white stem individuals. This information is substantial evidence that red stem and white stem individuals are not different varieties. The fact that little between-population variation occurs despite the vast area covered indicates that the Marama bean forms a continuous population rather than a population consisting of small restricted populations. This may also indicate that the Marama bean is a pre-dominantly outcrossing species

The rationale behind the development of molecular tools for diversity assessment studies is to find sufficient polymorphism considering the sample size. AFLP marker technique detected higher polymorphism (95 percent) than RAPD technique (77 percent). The two marker systems do not differ in the nature of data or their distribution in the genome, but the AFLP marker technique is an ideal system for detecting variation in the Marama bean because of its efficiency in detecting polymorphism. In addition, the number of polymorphic fragments scored is much higher in AFLP (1404) than in RAPD (63). AFLP system has a lower mean variance (0. 079) than RAPD (0. 832) showing higher precision between measurements. In addition between population significance level is higher in RAPD (0. 963) than in AFLP (0. 429), indicating that AFLP can discriminate better between populations than RAPD. However there is no significant difference between the mean genetic diversity estimates (p 0. 05) of the two techniques. Additionally the Spearman rank correlation coefficient between RAPDs and AFLPs was low in magnitude, but significant (r = 0. 638; p 0. 05). This demonstrates that even though AFLP technique can differentiate better between the closest Marama bean individuals, population and regions than RAPD, genetic distance estimates are positively correlated to that of RAPD. Hence the information gained is the same, but with slightly varying degrees

The present study revealed that there is high genetic variation within Marama bean populations rather than between populations. Moreover no genetically distinct groups reflecting their geographical segregation could be identified. The implications thereof are that Marama bean forms a homogenous population, rather than small restricted populations. RAPD and AFLP marker systems were successfully used to assess the amount of genetic variation between and within Marama bean populations. However, AFLP marker technique would be ideal for future genetic improvement of Marama bean