Estimates of relatedness and heritability of quantitative traits using molecular markers in an experimental stand of Prosopis alba (Leguminosae)

BESSEGA C.; DARQUIER R.; SAIDMAN B. O.; EWENS M; VILARDI J. C.

Buenos Aires, Argentina

Congreso; XIII Congreso Forestal Mundial (CFM); 2009

FAO, Ministrerio de Agricultura, Ganadería y Pesca, Secretaría de Ambiente y Desarrollo Sustentable;

In the last decade efforts have been dedicated to the development of relatedness (r) estimators based on molecular  markers. They are potentially very useful in programs of conservation of natural variability and improvement of profitable quantitative traits of native species, where most material proceeds from natural stands without any genealogical information. Hyper variable codominant loci, such as microsatellites are the most informative molecular markers and the first approaches were based on them. However, in the last year dominant markers like RAPDs, AFLPs, and ISSRs were also included, through particular mathematical models, within the useful available loci for estimates of relatedness and heritability. In a stand of P. alba planted in 1991 at San Carlos, Santiago del Estero (Argentina) pairwise relatedness were estimated for all individuals, from the information proceeding from 6 microsatellite loci and 128 dominant markers (including 56 AFLPs and 72 ISSRs). Different relatedness estimation methods (Ritland, 1996, 2006; Hardy 2003) were applied using the programs MARK and SPAGEDI. Heritabilities of 16 quantitative traits (tree height, basal trunk diameter, biomass, basal trunk area, petiole length, number of pairs of leaflets per pinna, pinna length, spine length, number of pinnae, leaflet length, leaflet length/width, leaflet falcate, leaflet apex, leaflet  apex/total area, leaflet width, leaflet area) were then estimated from the regression of pairwise phenotypic distances on pairwise relatedness coefficients according to Ritland (1996). The r estimates obtained from different markers were highly correlated (P 10-4) to each other and with the relatedness expected from the family records. The traits with the highest heritability values were those associated with foliar morphology. In general terms, the h2 estimated using the relatedness estimated from microsatellite and dominant markers were lower than heritabilities estimated from the actual relatedness, however, in both cases marker-based h2 estimates were highly significantly (P 0.01) correlated with those based on family records. Indirect estimates of relatedness inherent to the use of molecular markers involve underestimation of h2. However, the information obtained from this approach allows predicting without familiar information which traits are most likely to respond to selection programs.