基于积温的蓝花楹开花物候期预测模型构建

Construction of a Prediction Model of Jacaranda mimosifolia Flowering Phenology Based on Cumulative Temperature

建立不同地区蓝花楹开花物候期预测模型,为蓝花楹的选择育种、园林景观营造及其产业开发应用提供理论依据。以早期引种栽培在中国的蓝花楹为研究对象,选取17个花期物候较为稳定的地区,采用蓝花楹初花期、盛花期和末花期三个物候期的观测数据,结合气象数据对这三个物候期进行时间和空间尺度分析。结果表明:17个观测点的积温年际变化差异显著,蓝花楹花期物候差异均呈显著或极显著;不同地区之间同一积温指数各物候期所需积温相差较大,不同积温指数所需积温呈显著性差异;不同积温指数和不同物候时期都对物候所需积温影响差异极显著,两个因素交互作用影响差异显著;≥0℃积温指数与物候期日序具有高度相关性,可用于蓝花楹花期预测;17个观测点蓝花楹初花期、盛花期和末花期≥0℃积温的日序与经纬度、海拔呈极显著或显著的多元线性回归关系,各观测地点日序的回归模拟值与观测值单因素方差分析证实该回归模型可用于花期预测;用普通克里金插值法,采用上述预测模型,绘制了中国分布区蓝花楹开花物候期的时空分布图。综合分析得出,基于≥0℃积温指数建立的积温模型可用于蓝花楹花期预测。

The objectives of this study were to establish a floral phenology model of Jacaranda mimosifolia in different regions and to provide a theoretical basis for the breeding and industrial development of J.mimosifolia and landscape creation.Taking early introduced J.mimosifolia in China as research subjects,17 regions in China with stable floral phe-nology were selected,and the three phenological stages of early,full and late floral stages were analyzed in time and space by combining with meteorological data.The results showed that the inter-annual variation of cumulative tempera-ture at the 17 observation sites was significant,and differences of J.mimosifolia flowering associated with weather varia-tion was significant or highly significant.Also,the cumulative temperature required for each weathering period varied significantly among regions with the same cumulative temperature index,and the cumulative temperature required for different cumulative temperature indices was significantly different;the difference in the cumulative temperature required for different cumulative temperature indices and different weathering periods was highly significant,and the interaction of the two factors was significant;the difference in the cumulative temperature required for the weathering period can be used for flowering prediction because of its high correlation with the day sequence of the phenological period.The spatial and temporal distribution of the flowering phenology in the distribution area of J.mimosifolia was mapped using the above prediction model and it was concluded that the cumulative temperature model based on the 0℃cumulative tem-perature index can effectively be used to predict the flowering period of J.mimosifolia.