Abstract Chilean hazelnut (Corylus avellana L.) industry has been recently experiencing a solid development and nut production is consistently increasing. Corylus avellana is subjected to strong inter-annual yield fluctuations, which raise the issue of price predictability. In alternate bearing, a high fruit load hinders flower initiation thus leading to a smaller fruits production in the following year. This suggests the use of pollen airborne concentration as a proxy for flowers number to predict hazelnut irregular bearing. We addressed this question using high-resolution yield and pollen data collected on four hazelnut plantations located in Central Chile, and three hazelnut cultivars. Pollen parameters were calculated from modelled and non-modelled pollen seasons, after selecting the best modelling approach. An alternate bearing index was calculated to characterize hazelnut biennial bearing and to select plots with high yield variability. The calculated pollen parameters were then correlated with yield data. Significant correlations were found between pollen parameters and nut yields, with stronger relationship when alternate bearing was higher. The modelled peak of the pollen season presented the highest variability (average CV = 81 %) and its positive correlation with crop production (ρ = 0.45) was consistent across sites and cultivars. Our results support the integration of airborne pollen data in hazelnut yield prediction systems, as predictors of alternate bearing and to forecast annual yield.

中文翻译：

---

智利榛子园产量与花粉浓度的关系

摘要 智利榛子 (Corylus avellana L.) 产业近来发展稳健，坚果产量持续增加。Corylus avellana 受到强烈的年际产量波动的影响，这引发了价格可预测性的问题。在交替结果中，高果实负荷阻碍了开花，从而导致下一年的果实产量减少。这表明使用花粉空气传播浓度作为花数的代理来预测榛子不规则轴承。我们使用在位于智利中部的四个榛子种植园和三个榛子品种收集的高分辨率产量和花粉数据解决了这个问题。在选择最佳建模方法后，根据建模和非建模花粉季节计算花粉参数。计算了一个替代的承载指数来表征榛子两年一次的承载并选择具有高产量变异性的地块。然后将计算出的花粉参数与产量数据相关联。花粉参数与坚果产量之间存在显着相关性，当交替轴承较高时，相关性更强。模拟的花粉季节高峰表现出最高的变异性（平均 CV = 81 %），其与作物产量的正相关（ρ = 0.45）在不同地点和品种之间是一致的。我们的结果支持将空气传播的花粉数据整合到榛子产量预测系统中，作为替代轴承的预测因子和预测年产量。然后将计算出的花粉参数与产量数据相关联。花粉参数与坚果产量之间存在显着相关性，当交替轴承较高时，相关性更强。模拟的花粉季节高峰表现出最高的变异性（平均 CV = 81 %），其与作物产量的正相关（ρ = 0.45）在不同地点和品种之间是一致的。我们的结果支持将空气传播的花粉数据整合到榛子产量预测系统中，作为替代轴承的预测因子和预测年产量。然后将计算出的花粉参数与产量数据相关联。花粉参数与坚果产量之间存在显着相关性，当交替轴承较高时，相关性更强。模拟的花粉季节高峰表现出最高的变异性（平均 CV = 81 %），其与作物产量的正相关（ρ = 0.45）在不同地点和品种之间是一致的。我们的结果支持将空气传播的花粉数据整合到榛子产量预测系统中，作为替代轴承的预测因子和预测年产量。模拟的花粉季节高峰表现出最高的变异性（平均 CV = 81 %），其与作物产量的正相关（ρ = 0.45）在不同地点和品种之间是一致的。我们的结果支持将空气传播的花粉数据整合到榛子产量预测系统中，作为替代轴承的预测因子和预测年产量。模拟的花粉季节高峰表现出最高的变异性（平均 CV = 81 %），其与作物产量的正相关（ρ = 0.45）在不同地点和品种之间是一致的。我们的结果支持将空气传播的花粉数据整合到榛子产量预测系统中，作为替代轴承的预测因子和预测年产量。