Flowering date in perennial fruit trees is an important trait for fruit production. Depending on the winter and spring temperatures, flowering of olive may be advanced, delayed, or even suppressed. Deciphering the genetic control of flowering date is thus key to help selecting cultivars better adapted to the current climate context. Here, we investigated the genetic determinism of full flowering date stage in cultivated olive based on capture sequencing data of 318 genotypes from the worldwide olive germplasm bank of Marrakech, Morocco. The genetic structure of this collection was organized in three clusters that were broadly attributed to eastern, central, and western Mediterranean regions, based on the presumed origin of genotypes. Flowering dates, collected over seven years, were used to estimate the genotypic best linear unbiased predictors, which were then analyzed in a genome-wide association study. Loci with small effects were significantly associated with the studied trait, by either a single- or a multi-locus approach. The three most robust loci were located on chromosomes 01 and 04, and on a scaffold, and explained 7.1%, 6.2%, and 6.5% of the trait variance, respectively. A significantly higher accuracy in the best linear unbiased predictors of flowering date prediction was reported with Ridge- compared to LASSO-based genomic prediction model. Along with genomic association results, this suggests a complex polygenic determinism of flowering date, as seen in many other fruit perennials. These results and the screening of associated regions for candidate genes open perspectives for further studies and breeding programs targeting flowering date.

中文翻译：

---

一系列栽培橄榄树开花日期的全基因组关联分析


多年生果树的开花期是水果生产的重要性状。根据冬季和春季的气温，橄榄的开花可能会提前、延迟，甚至受到抑制。因此，破译开花日期的遗传控制是帮助选择更好地适应当前气候环境的品种的关键。在这里，我们根据来自摩洛哥马拉喀什全球橄榄种质库的 318 个基因型的捕获测序数据，研究了栽培橄榄全花期阶段的遗传决定论。根据推测的基因型起源，该集合的遗传结构分为三个簇，大致归因于地中海东部、中部和西部地区。七年来收集的开花日期被用来估计基因型最佳线性无偏预测因子，然后在全基因组关联研究中进行分析。通过单基因座或多基因座方法，影响较小的基因座与所研究的性状显着相关。三个最稳健的基因座位于 01 号和 04 号染色体以及支架上，分别解释了 7.1%、6.2% 和 6.5% 的性状变异。据报道，与基于 LASSO 的基因组预测模型相比，Ridge 的开花日期预测的最佳线性无偏预测器的准确度显着更高。与基因组关联结果一起，这表明开花日期存在复杂的多基因决定论，正如许多其他多年生水果植物中所见。这些结果以及对相关区域候选基因的筛选为进一步研究和针对开花日期的育种计划开辟了前景。