Maize (Zea mays L.) for northern growing areas requires cold tolerance for extending the 2 vegetative period. Our objectives were to evaluate two large panels of maize inbred lines 3 adapted to Europe for cold tolerance and to estimate the effects of cold-related traits on 4 biomass production. Two inbred panels were evaluated for cold tolerance per se and in 5 testcrosses under cold and control conditions in a growth chamber and under field 6 conditions. Comparisons of inbreds and groups of inbreds were made taking into account 7 the SNP-based genetic structure of the panels, and the factors affecting biomass 8 production were studied. Eight flint and one dent inbreds with diverse origins were the 9 most cold tolerant. The most cold tolerant dent and flint groups were the Iodent Ph207 10 and the Northern Flint D171 groups, respectively. The relationships between inbred per se 11 and testcross performance and between controlled and field conditions were low.
12Regressions with dry matter yield in the field as dependent variable identified plant height combining ability, as well as between both groups and the Northern Flint D171 group.
20Key words: maize, cold tolerance, abiotic stress, germplasm. period that can potentially increase yield and stability, and the probability of escaping 8 summer drought stress (Kucharik, 2006). This is particularly true in some temperate areas, 9 where springs are cold and rainy and summers are hot and dry. But early sowing in 10 temperate areas requires cold tolerance and, consequently, the interest of breeders in cold 11 tolerance is increasing (Darkó et al., 2011; Frascaroli and Landi, 2013; Revilla et al., 2005; 12 Strigens et al., 2012 12 Strigens et al., , 2013. In the northwest of Spain, a breeding goal would be to advance 13 maize sowing two weeks, i.e. from May to mid April, because there are no late frosts. 14 However, in northern areas the gain could be just a few days.
15The main handicap for breeding programs intending to improve cold tolerance in 16 maize has been the narrow genetic base for this trait (Greaves, 1996; Revilla et al., 2005). were not more cold tolerant than the checks. Actually, they were similar to the improved 21 populations and checks already known, including commercial checks and the best cold explaining between 5.7 and 52.5% of the phenotypic variance for early growth and 10 chlorophyll fluorescence. They propose the use of whole genome prediction approaches 11 rather than classical marker assisted selection to improve the chilling tolerance of maize.
12Other major obstacles for breeders are that cold tolerance has large experimental 13 errors, a strong genotype by environment interaction, and a complex genetic regulation 14 (Revilla et al., 2000, Strigens et al., 2013. Moreover, evaluations for cold tolerance are not 15 accurate enough for a precise discrimination of cold tolerance. This is due to two facts. , 1981;Verheul et al., 1996). Given that most previous reports have faced 16 the problem by using limited resources, we belie...