Historical Cross-Site Association Based on Cultivar Performance in the Southern Cone

Abstract

Southern Cone wheat growing region is considered one large epidemiological zone separated by the Andes Mountain Range. Although local breeding programs have exchanged germplasm successfully and results have been published regularly, very little effort has been made to determine site similarities based on historical data. The present study uses 12 years of agronomic data on days to heading (DH), leaf rust reaction (LR) and septoria leaf blight infection (SB) from Southern Cone Wheat Advanced Lines Nursery (LACOS) to define and characterize cross-site associations. A modified pattern analysis using classification (cluster analysis) and ordination (Factor analysis) based on distances and similarities (average correlations between sites) were utilized for sites representing at least three years of data for each trait. Two factors accounted for very large proportion (73 percent) of the variation for DH. However, in the cases of LR and SB slightly lower variation, 45.6 and 53.5 percent respectively, was explained by two factors. Using three factors, 68.4 percent of the variation in SB and 53.5 percent of the variation in LR could be explained. Fourteen common sites were used classify the region in two distinct groups with regards to DH. The key factor seems to be the temperature regime during the vegetative phase of the crop in the two regions separated by the parallel 27 degrees South. The higher temperature regime of the north (Bolivia, Brazil and Paraguay) can be further subdivided to include very high temperature sites of Sao Paulo, Brazil, and drought prone site, San Benito, Bolivia, in one subgroup separating it from other sites in this group. Lower temperature locations grouped together represent sites in Argentina, Chile and Uruguay. LR represented a large variation in grouping the sites together. While all sites in Brazil grouped together as a unit, explaining the virulence similarity in this country, two Chilean sites, La Platina and Chillan grouped independently. Surprisingly, Chillan grouped with locations in Argentina. The fourth group includes sites in Paraguay, Uruguay and Mexico; all of them with heavy infection ratings over the years. Similar variation was observed in the groupings of seven SB sites. While Toluca, Mexico, and Marcos Juarez, Argentina, represented two independent groups, highland sites of Patzcuaro, Mexico, and Escalante, Bolivia, made up the third group. The southern locations of Chillan, Chile, and La Estanzuela, Uruguay, represented the high infection sites as an independent group. The site similarities demonstrated in this study should increase our understanding of the variation present in the Southern Cone Wheat Region and enhance cultivar adaptation across programs.