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

Giardini Naxos, Italy - 18/21 September, 2002

ISBN 88-900622-3-1

Oral Communication Abstract - S4a

DEFINING THE ADAPTATION STRATEGY OF BREEDING PROGRAMMES

ANNICCHIARICO P.

Istituto Sperimentale per le Colture Foraggere

breeding strategy, genotype x environment interaction, plant adaptation, quantitative genetics

Breeding for large and/or diversified target regions may implies the definition of a breeding strategy to cope with, and possibly exploit, genotype x environment interactions. Genetically-based trade-offs between yield potential and tolerance to major abiotic stresses, the need to choose between incompatible levels of a key adaptive trait such as earliness of cycle, and the fact that many useful genetic markers are environment-specific, suggest that defining an effective adaptation strategy will remain of crucial importance for breeding programmes of major field crops.

Multi-environment yield trials performed for genotype selection or recommendation may also provide information for defining adaptation strategies, yield stability targets, indirect selection criteria (based on morphophysiological traits or genetic markers), parent germplasm and selection environments. Repeatable genotype x location (GL) interaction effects can be either exploited, by breeding material adapted to a specific sub-region, or minimized, by breeding material widely adapted to the region. Interfacing statistical modeling of genotype responses (e.g., by joint regression, AMMI or factorial regression techniques) with indirect selection theory allows for comparing different adaptation strategies based on yield gains predicted from selection of germplasm developed from the genetic base of which the tested genotypes are assumed to be a representative sample. Sub-regions can be characterized on the basis of climatic, soil, biotic and crop management factors associated with GL effects. A fine-tuned description of geographic boundaries may be obtained by a Geographic Information System (GIS). The chosen adaptation strategy can be complemented by decisions on selection environments (reproducing the yield responses across the region for wide adaptation, and in sub-regions for specific adaptation) and useful parent germplasm and adaptive traits. The possible identification of artificial environments capable of reproducing at one site the yield responses occurring across the region may open the way for a substantial reduction of costs due to multi-locational selection.

Examples are provided from two case studies. In the former, the generated information (supported by GIS data) contributes to decisions of the durum wheat national breeding programme of Algeria on key elements of its strategy. In the latter, the study of GL interaction for lucerne in northern Italy has allowed to define artificial environments that reproduce in Lodi the genotype responses occurring across the region.