Environmental Variance and Broad-sense Heritability of Nut Traits in Japanese Chestnut Breeding

Abstract

We evaluated the nut harvesting date (NHD), nut weight (NW), pericarp splitting (PS), and infestation by insects (II) in eight cultivars/selections of Japanese chestnut, including a Japanese–Chinese hybrid, over 6 years. Data were analyzed by analysis of variance (without transformation for NHD, after log-transformation for NW and PS, and after square root transformation for II). The among-tree variance accounted for only 1.1% to 8.5% of the total variance. The variance component resulting from residual factors for the tree × year interaction and sampling errors was the largest component for NW, PS, and II, accounting for 46% to 54% of the total environmental variance. Because tree replication is costly and time-consuming in chestnut breeding, increasing the number of yearly repetitions is more efficient than increasing the number of tree replicates. Broad-sense heritability was 0.84 for NHD, 0.27 for NW, 0.48 for PS, and 0.17 for II in evaluations with one tree without yearly repetition. It increased to 0.91 for NHD, 0.40 for NW, 0.62 for PS, and 0.29 for II in evaluations with one tree in 2 years. For NHD, the heritabilities are sufficiently high to distinguish genetic differences among cultivars/selection. In contrast, the low heritability of II suggests that this trait should not be evaluated with an emphasis on the initial selection stage but rather with an emphasis on the secondary selection stage based on testing at several locations with a large number of yearly and tree replications.