International Journal of Horticulture, 2024, Vol.14, No.2, 78-88 http://hortherbpublisher.com/index.php/ijh 85 Figure 8 Genomic prediction of additive genetic variance Table 1 shows the genetic gains achieved for 545 hybrid strawberry varieties phenotyped during the 2017-18 and 2018-19 growing seasons in the Salinas region of California. The table lists the gain values (Δ G=EMM1–EMM2) for marketable fruit yield and quality traits, derived through linear contrasts of the estimated marginal means (EMMs). These EMMs were obtained from a linear mixed model analysis based on phenotypic observation data from 10 to 13 harvests per year and three clonal replicates per hybrid variety each year. The table also provides the REML estimates of broad-sense heritability ( h2), which indicate the genetic stability of traits based on clonal averages. Additionally, the ratio of genotype-by-year interaction variance to phenotypic variance is shown , reflecting the impact of environmental conditions in different years on trait expression. For each trait, the table lists two Δ Gestimates: one using a low-performing elite × elite hybrid as a reference, and the other using a low-performing elite × wild hybrid as a reference. The percentage change in ΔG (ΔG(%)) is calculated based on the ratio of the gain value to EMM2 multiplied by 100. The table indicates that, compared to elite × elite hybrids, elite × wild hybrids typically show larger genetic gains in traits such as yield, number, weight, firmness, total soluble solids (TSS), titratable acidity (TA), sugar-acid ratio (TSS/TA), and anthocyanin concentration. This may suggest that introducing genetic material from wild species can significantly enhance the performance of specific traits. The Pr(>t) values in the table indicate the statistical significance of non-zero genetic gains. A small value suggests sufficient evidence to reject the null hypothesis, confirming the presence of genetic gains. In most traits, this value
RkJQdWJsaXNoZXIy MjQ4ODYzMg==