Construction and validation of core collections in Pisum sp. using different methodologies

Abstract

Core collections contribute to make conservation expenses more efficient and facilitate a better utilization of accessions in breeding programs. Eighty-five accessions from the working collection of Pisum germplasm belonging to cultivated species and subspecies were evaluated during 2015 and 2016 in the College of Agricultural Sciences, Rosario National University. Phenotypic values of 12 morphological traits were measured and best linear unbiased prediction (BLUP) of the adjusted means or genotypic values for these traits was calculated. Molecular characterization was performed assaying a total of 15 SSR primer combinations and 25 SRAP primer combinations on all accessions. With all the data collected, four Cluster Analysis were performed (phenotypic values, genotypic values, molecular markers and consensus) to divide the accessions into groups with similar characteristics. Four strategies to determine the number of accessions selected from each group (constant, logarithmic, proportional and maximization strategies) were applied to construct 16 core collections. Validation of the core collections were performed analyzing the parameters Mean difference percentage (MD), Variance difference percentage (VD), Coincidence rate of range (CR) Variable rate of coefficient of variation (VR), Shannon diversity index (SW) and the taxonomy coverage (TC). Considering all validation parameters together, the logarithmic strategy with genotypic values data was the best strategy, while the worst strategy was the proportional strategy with molecular marker data.