Management of a Walnut Germplasm Collection: Which of SSR or SNP Markers Are Most Suitable to Preserve Biodiversity?

The preservation of the maximum of diversity within the smallest number of accessions is one of the challenges of germplasm management. To construct core-collections, the assessment of the population structure and the relationships between the accessions represents a key step and the choice of suitable molecular markers is the starting point. Since the expansion of available SNP-based genomics tools, a debate has emerged regarding the usefulness of the widely used microsatellites (SSRs) markers. In this study, we analysed a part of the INRAE walnut germplasm collection of 150 accessions, unique in Europe for walnut biodiversity conservation, by comparing the power of both types of marker. We found that the first level of structure is equally detected using 13 SSRs or the Axiom J. regia 700K SNP array, and is in relation with the geographical origin of the accessions. For K=2, there was no exchange of accession between the two groups when both markers were compared. We also highlighted empirically that approximately 100 SNPs are needed to obtain similar clustering to SSRs in Principal Coordinate Analysis (PCoA). The neighbor-joining trees constructed were also consistent between both types of marker. The main differences lied in the upper levels of structure from K=3 to K=6, more powerful using the SNPs, and in the percentage of the explained variation in PCoA for K=2, higher using SSRs. We then constructed core-collections of 50 accessions, a crucial step in genetic resources management to reduce the costs and preserve the allelic diversity. Using two different construction methods, both SSR and SNP markers were suitable and able to keep at least 88.57% of the alleles. 32/50 accessions were in common between the two markers, for both methods. We concluded that the use of either marker is dependent on the researchers goal.