Salt tolerance in Lycopersicon species. IV. Efficiency of marker-assisted selection for salt tolerance improvement.

The usefulness of marker-assisted selection (MAS) to develop salt-tolerant breeding lines from a F2 derived from L. esculentum x L. pimpinellifolium has been studied. Interval mapping methodology of quantitative trait locus (QTL) analysis was used to locate more precisely previously detected salt tolerance QTLs. A new QTL for total fruit weight under salinity (TW) near TG24 was detected. Most of the detected QTLs [3 for TW, 5 for fruit number, (FN) and 4 for fruit weight (FW)] had low R (2) values, except the FW QTL in the TG180-TG48 interval, which explains 36.6% of the total variance. Dominant and overdominant effects were detected at the QTLs for TW, whereas gene effects at the QTLs for FJV and FW ranged from additive to partial dominance. Phenotypic selection of F2 familes and marker-assisted selection of F3 families were carried out. Yield under salinity decreased in the F2 generation. F3 means were similar to those of the F1 as a consequence of phentoypic selection. The most important selection response for every trait was obtained from the F3 to F4 where MAS was applied. While F3 variation was mainly due to the within-family component, in the F4 the FN and FW between-family component was larger than the within-family one, indicating an efficient compartmentalization and fixation of QTLs into the F4 families. Comparison of the yield of these families under control versus saline conditions showed that fruit weight is a key trait to success in tomato salt-tolerance improvement using wild Lycopersicon germplasm. The QTLs we have detected under salinity seem to be also working under control conditions, although the interaction family x treatment was significant for TW, thereby explaining the fact that the selected families responded differently to salinity.