Genetic variance distribution of SSR markers and economically important quantitative traits in a progeny trial of Prosopis chilensis (Leguminosae): implications for the ‘Algarrobo’ management programme

Prosopis chilensis (Molina) Stuntz (Leguminosae) is a valuable native species in Argentina included in the Prosopis Management Programme. Natural provenances show important height and shape differentiation throughout their distribution in the Monte Desert. The availability of progeny trials provides an opportunity to quantify genetic differentiation among provenances and test the relative importance of demographic vs adaptive processes on morphological variation. We quantified both genetic and quantitative differentiation of neutral markers and five economically important traits, respectively, among four provenances in a provenance-progeny trial. We aimed to quantify the genetic basis of variations in height, basal diameter, tree shape, spine length and biomass. Two hypotheses were tested: (1) economically important traits have significant heritability, and (2) the phenotypic variation among provenances is the result of local adaptation to particular environmental conditions. Our results indicate that most morphological variation was found among individuals within families (~95 per cent). The h2 estimates were heterogeneous among traits and ranged from low (0 for number of stems) to moderate (0.22 and 0.28 for spine length and biomass, respectively). Variance among families (~5 per cent) was evenly distributed within and among provenances. Morphological differentiation among provenances was low, but significant, and could be attributed mainly to individuals from Villa Unión. Based on molecular markers, genetic differentiation among provenances was low and significant (FST = 0.03; P = 5 × 10−4) but was able to differentiate the groups from Villa Unión, Fiambalá and Mogna-Chilecito. Neutrality tests were conducted using the FST  –QST test and DJSOST and δGREGORIUS alternative coefficients of differentiation. Neutrality tests yielded no evidence of local adaptation and were rather consistent in showing a trend toward stabilizing selection, particularly for spine length. The selection strategy for breeding programmes should depend on the trait to be improved and should consider both provenance and familiar information. Considering an intra-familiar ranking is encouraged in order to maximize the genetic gain. Additionally, in order to recover germplasm provenance diversity, based on morphological and microsatellite results, our recommendation would be to include seeds from individuals from at least the Villa Unión, Fiambalá and Mogna-Chilecito areas.