Proceedings of the XLVII Italian Society of Agricultural Genetics - SIGA Annual Congress

Verona, Italy - 24/27 September, 2003

ISBN 88-900622-4-X

Poster Abstract - 2.42

MOLECULAR ANALYSIS OF ANCESTRY IN ELITE DURUM WHEAT GERMPLASM BY SSR MARKERS

M. MACCAFERRI*, C. XIE**, M.C. Sanguineti*, R. TUBEROSA*, S. SMITH**

*) Department of Agroenvironmental Science and Technology, Via Fanin 44, 40127 Bologna, Italy

**) Pioneer-Dupont, Johnston, IA, USA

SSRs, genetic similarity, ancestry, haplotypes, durum wheat

Information on the genetic diversity structure of elite materials is useful for many issues, in particular for the planning and management of breeding programs, for monitoring the whole level of genetic diversity present in a crop gene pool and for germplasm conservation. Also, the molecular (fingerprinting) profiles are useful to elucidate the genetic relationships occurring among cultivars, which ultimately have to be considered for the protection of breeder’s rights (e.g. for essential derivation issues). Recently, molecular markers have also been considered by UPOV as particularly useful to investigate the relatedness between varieties. Among molecular markers, SSRs (Simple Sequence Repeats), in particular dinucleotide genomic SSRs, represent useful molecular markers for such purposes. Durum wheat is an important crop world-wide and is currently the object of a renewed interest: in Italy, more than 100 varieties are registered and commercialised.

A collection of 134 durum wheat accessions was assembled, mainly comprising modern varieties, representative of the modern durum wheat gene pools currently cultivated in some of the most important durum-producer regions world-wide (Italy, France, Canada and USA); additionally, a number of landmark genotypes from the CIMMYT/ICARDA breeding program and cvs. from Austria and Australia were included, as well as cvs or accessions recognised as important foundation genotypes from different breeding pools. Accessions were profiled with 70 dinucleotide markers, which produced a molecular data set with an average of 6.9 different allelic variants per SSR locus.

Genetic relationships (co-ancestries) among cvs. were investigated by calculating a matrix consisting of all the pair-wise genetic similarities and compared and integrated with the probabilities of ancestry for all 134 cvs. (the latter calculated on the basis of a Bayesian algorithm). Cluster analysis pointed out that the genetic diversity of the elite germplasm herein analysed is highly structured and a number of groups and subgroups of different breeding lineages, mostly consisting of highly related cvs. (in accordance with pedigree information) were identified. Ranking of genetic similarities and probabilities of ancestry were used to predict the most probable parents and the closest relatives of each cv., among those profiled. Case-studies consisting of 15 triplets (i.e., one cv. and both its putative parents) were present in the collection and profiled with molecular markers, as long as 35 cases where a cv. (obtained by single or multiple cross) and one of its putative parents. In most cases, results agreed to what expected on the basis of the corresponding registered pedigree. However, incorrect pedigrees were also identified; for such cases, robust molecular evidences indicated the involvement of genotypes different from those listed in the pedigree. The analysis of graphical genotypes (map-based haplotypes) allowed for the identification and the evaluation of the extent of the single chromosome regions possibly directly passed through breeding cycles to modern cvs.. Rare alleles at highly variable SSR loci were also considered as signatures of identity by descent.