Winter survival of pea, faba bean and white lupin cultivars in contrasting Italian locations and sowing times, and implications for selection

Increasing grain legume yields via autumn sowing requires winter-hardy material. Forty-nine pea, 24 faba bean and 11 white lupin cultivars recently released by 29 breeding institutions worldwide were sown in autumn 2002 at Lodi (northern Italy) and Foggia (southern Italy) on two different dates at each location, with the objective of assessing the winter survival of species and cultivars, its consistency across locations and sowing times, its relationship with grain yield and various morphophysiological traits and the implications for selection. The winter seasons were representative for the sites and had 54 frost days and -7.3 degrees C absolute minimum temperature in Lodi, and 8 frost days and -3.4 degrees C absolute minimum temperature in Foggia. The species differed in optimal sowing time at each location. Mild winter temperatures, preventing a sufficient hardening against late frosts, led to higher proportions of plants killed over winter in Foggia than in Lodi (0.17 v. 0.12 for pea; 0.20 v. 0.16 for faba bean; 0.34 v. 0.11 for lupin). Variation in winter mortality was much larger within species than among species. Winter mortality and grain yield of cultivars were inversely correlated (r= -0.49, -0.43 and -0.74 for entry means over locations of pea, faba bean and lupin, respectively). The consistency across environments of genotype winter mortality, estimated by genetic correlations, was high across sowing times and low across locations for faba bean, moderate across sowing times and fairly low across locations for pea, and always fairly high for lupin. A visual cold tolerance score was always related to better winter survival. Winter survival tended to correlate with larger seeds in faba bean and lupin, and was associated with later flowering in lupin. A rosette-like winter growth habit (defined by lower height/number of leaves ratio of seedlings in January) was correlated with winter survival in all species (after partialling out the influence of seed size in faba bean and lupin). The optimal index of indirect selection for low winter mortality included the cold tolerance score, the seedling height/number of leaves ratio and the seed size for faba bean, the first two of these variables for pea, but only the first variable for lupin. The gain in predicted efficiency over direct selection was high for faba bean (29 %) and modest for pea and lupin (<= 10 %).