Genetic resources of teak (Tectona grandis Linn. f.)—strong genetic structure among natural populations

Abstract

Twenty-nine provenances of teak (Tectona grandis Linn. f.) representing the full natural distribution range of the species were genotyped with microsatellite DNA markers to analyse genetic diversity and population genetic structure. Provenances originating from the semi-moist east coast of India had the highest genetic diversity while provenances from Laos showed the lowest. In the eastern part of the natural distribution area, comprising Myanmar, Thailand and Laos, there was a strong clinal decrease in genetic diversity the further east the provenance was located. Overall, the pattern of genetic diversity supports the hypothesis that teak has its centre of origin in India, from where it spread eastwards. The analysis of molecular variance (AMOVA) gave an overall highly significant F st value of 0.227-population pairwise F st values were in the range 0.01-0.48. Applying the G″ st differentiation parameter, the estimated overall differentiation was 0.632, implying a strong genetic structure among populations. A neighbour-joining (NJ) tree, using the pairwise population matrix of G″ st values as input, contained three distinct groups: (1) the eight provenances from Thailand and Laos, (2) the Indian provenances from the dry interior and the moist west coast and (3) the provenances from northern Myanmar. The provenances from southern Myanmar were placed close to the root of the tree together with the three provenances from the semi-moist east coast of India. A Bayesian cluster analysis using the STRUCTURE software gave very similar results, with three main clusters, each containing two sub-clusters, while Bayesian cluster analysis in the Geneland software, exploiting the spatial coordinates of the provenances, resulted in five clusters in accordance with the former results. The implications of the findings for conservation and use of genetic resources of the species are discussed.

Worldwide translocation of teak—origin of landraces and present genetic base

“…At the same time, large differences in the level of genetic diversity were observed in the different regions. In the eastern part of the distribution area (Myanmar, Thailand and Laos), a strong clinal decrease in diversity was observed towards the east, where the average number of alleles of the southern Myanmar populations was around three times higher compared to the eastern Laos and Thai populations (Hansen et al 2015). The findings draw attention to the importance of the genetic origin of native as well as planted teak, but the comprehensive study also opens opportunities for the use of so-called assignment methods, which are designed to determine population membership for individuals or groups of individuals using genetic information (Manel et al 2005).…”

“…This latter point is important, since almost half of the remaining natural forests are found in Myanmar (Kollert and Cherubini 2012) and because historical records suggest that Myanmar has been source for several of the first translocations of teak to Africa and America as discussed above. Recently, Hansen et al (2015) completed the first study of teak's genetic structure comprising provenances from the total distribution area of the species, including Myanmar. Using simple sequence repeat (SSR) markers on 721 individuals from 29 provenances, a very strong genetic structure was revealed, showing an overall F st value of 0.23 with population pairwise F st values in the range 0.01-0.48.…”

“…The 29 teak populations originating from the natural distribution area and analysed by Hansen et al (2015) using six SSR markers were used as a reference population to constitute a 'genetic map' to which we could try to assign the landraces described above. Further details about the origin and sampling of the 29 populations are given in Hansen et al (2015) and references therein.…”

“…Further details about the origin and sampling of the 29 populations are given in Hansen et al (2015) and references therein. Bayesian cluster analysis of the 29 populations analysed by Hansen et al (2015) resulted in three main clusters and within those, six sub-clusters-the latter are depicted together with the geographic origin of the 29 populations in Fig. 1.…”

“…Ancestry of individuals with PopFlag = 0 is updated according to the admixture or noadmixture model, as specified by the user (Pritchard et al 2010). As in the analyses of natural populations by Hansen et al (2015), two rounds of STRUCTURE analyses were run. The first one comprised data from all 46 populations (1166 trees) and used the 'no admixture ancestry' model in combination with the assumption of independent allele frequencies among populations, due to the strong genetic structure observed in the overall natural distribution area.…”