GENETIC AND GEOGRAPHIC VARIATION IN SEED TRAITS OF COMMON WALNUT (JUGLANS REGIA L.) AMONG TWENTY POPULATIONS FROM THE WEST OF SPAIN

Twenty Common Walnut (Juglans regia L.) populations were identified in the West of Spain. Seeds from ten walnut-trees were collected at each population. Twenty morphological traits of Juglans regia L. seeds were studied. The analysed seed traits were those related to size, weight, volume, shape (following UPOV descriptors), texture and colour (following IPGRI descriptors). Four of the traits (length, width, thickness and weight) were assessed in 30-45 seeds per walnut-tree, so there was 30-45 data per tree. The other traits were determined for a 25-seedlot, so there was just one data per tree. In February the seeds were planted in a greenhouse. The experimental design was a split-plot design. During the greenhouse experiment data of germination and some seedling traits were measured. The main objectives of this work are to: 1) study the variation among- and within- populations and the phenotypic correlation between traits; 2) determine if there are geographical variation in common walnut seed traits among populations. We have found highly significant differences among- and within- populations for width, thickness, weight and volume. For the other traits significant differences were found among populations (p<0.01), except for: shape in longitudinal section perpendicular to suture, shape in longitudinal section through suture, shape in cross section and texture. Geographic variations were found for some morphological traits, as a result of significant correlations (p<0.01) between several morphological traits and some geographic, climatic and edaphic traits. The most important pattern of geographic variation for some morphological traits was latitude. Hence, correlation analyses between morphological traits and geoclimatic factors indicated the existence of clinal variation in weight, width, thickness and volume. Seeds of the southern populations had a higher weight, width, thickness and volume than those of the northern populations. There were found also significant positive correlations (p<0.001) between height (H0) and following factors: annual precipitation (P), annual mean temperature (T) and Philippis index (PE_TC) and there were found negative correlations (p<0.001) between height (H0) and: number of frosty days (FD) and clay percentage (CL).