Dissecting a vegetable landrace: Components of variation in Spanish ?Moruno? tomatoes as a case studio

Abstract

The variability of landraces is dissected and related to farmers' practices through surveys. Spanish 'Moruno' tomatoes, recognised for their excellent flavour, were selected as a case study. For this purpose, 30 populations were characterised in a four-year program. Higher intra-population than inter-population variability was found. Variability is generally reduced in the traits used by farmers as selection criteria. Farmers rarely used flavour as a selection criterion, but it is one of the main characteristics used by them to define the landrace. Seed exchange and growing different landraces simultaneously are commonplace, and outcrossing might occur, thus justifying in part the seed degeneration problems experienced by some farmers. At the same time, farmers select seeds for the next year on a per-fruit basis rather than a per-plant basis, justifying the maintenance of a high level of variability. In such cases, high pressure would be applied to key morphological traits but not to flavour. Accordingly, the sugars, acids and volatiles profiles related to flavour in the landraces' populations are highly variable, though the environment also exercises a high effect. It is necessary to make population selections to offer farmers materials combining the best organoleptic perception and a high stable yield. This would reduce the gap between the price premium received by farmers and the expected fair price. It would also be necessary to develop consumer information campaigns to exploit their willingness to pay for the extra value offered by landraces. Only then long-term on-farm conservation would be economically feasible. Although achieving a trade-off between yield and flavour is difficult, it is possible to identify populations that reach a compromise between them. In germplasm banks, it is impossible to evaluate all the materials in the same year. The use of hybrid controls, with no genetic variation, is helpful in considering the environmental effects. Still, genotype x environment interactions are evident, and using selected control populations of landraces is necessary to evaluate possible performances closer to the type of materials being evaluated.