中国东部暖温带刺槐花期空间格局的模拟与预测

模拟刺槐开花日期与气温之间的空间关系,对于揭示蜜源植物物候空间格局形成的生态机制和掌握养蜂生产的时宜,具有重要的科学意义.利用中国东部暖温带26个站点1986-2005年的刺槐开花始期、盛期和末期数据,建立了基于日均温的多年平均和逐年物候空间模型,模拟多年平均和逐年刺槐开花日期的空间格局,并对模型进行了空间外推检验.进而,将1986-2005年8 km×8 km分辨率的日均温格点数据代人多年平均和逐年物候空间模型,得到连续地理空间多年平均和逐年刺槐开花日期的空间格局,并尝试设计了研究区内转地放蜂的适宜路线.结果表明:中国东部暖温带1986-2005年多年平均及逐年最佳期间日均温的空间格局分别控制着多年平均和逐年刺槐开花日期的空间格局.各地多年平均刺槐开花日期的空间序列与最佳期间日均温的空间序列呈显著负相关(P＜0.001),多年平均气温—物候空间模型对刺槐开花始期、盛期和末期的方差解释量分别为87％、86％和77％,模拟的均方根误差(RMSE)分别为2.5、2.7d和4.1d.同样,各地逐年刺槐开花日期的空间序列与最佳期间日均温的空间序列均呈显著负相关(P＜0.05),逐年气温-物候空间模型对刺槐开花始期、盛期和末期的方差解释量分别介于44％-94％、57％-92％和39％-84％之间,模拟的平均RMSE分别为3.9、4.0d和5.4d.预测得到的连续地理空间多年平均刺槐开花日期呈现出自南向北、从平原向丘陵和山地逐渐推迟的空间演进特征.据此,中国东部暖温带地区转地放蜂可以沿西线、中线和东线进行,放蜂的大致持续时间可达40-50 d.此外,预测得到的连续地理空间1986-2005年期间刺槐开花始期、盛期和末期的线性趋势以提前为主,呈显著提前的面积分别占总面积的78％、26％和32％.

Simulation and prediction of spatial patterns of Robinia pseudoacacia flowering dates in eastern China's warm temperate zone

Modeling spatial relationships between Robinia pseudoacacia flowering date and air temperature is crucial for revealing the ecological mechanism of spatial patterns of honey plant phenology and knowing well the right time of bee keeping production. Using Robinia pseudoacacia phenology data of beginning of flowering (BF), full flowering (FF) and end of flowering (EF) at 26 stations of eastern China''s warm temperate zone during 1986 to 2005, we established daily mean temperature-based multi-year mean and yearly spatial phenology models to simulate multi-year mean and yearly spatial patterns of Robinia pseudoacacia flowering dates, and validated these models by extensive spatial extrapolation. Then, we substituted gridded daily mean temperature data with the spatial resolution of 8 km into the multi-year mean and yearly spatial phenology models, and obtained spatial patterns of multi-year and yearly Robinia pseudoacacia flowering dates over continuous geographic coverage during 1986 to 2005. Moreover, we attempted to project suitable routes of migratory beekeeping within the study area. The results show that spatial patterns of multi-year mean and yearly temperatures within the optimum length period control spatial patterns of multi-year mean and yearly Robinia pseudoacacia flowering dates, respectively in eastern China''s warm temperate zone during 1986 to 2005. Spatial series of multi-year mean Robinia pseudoacacia flowering dates correlate negatively with spatial series of multi-year mean temperature within the optimum length period (P < 0.001) at the 26 stations. The multi-year mean spatial phenology models explained 87% of variance in BF date with a root mean square error (RMSE) of 2.5 days, 86% of variance in FF date with a RMSE of 2.7 days, and 77% of variance in EF date with a RMSE of 4.1 days. Similarly, spatial series of Robinia pseudoacacia flowering dates correlate negatively with spatial series of daily mean temperature within the optimum length period in each year (P < 0.05) at the 26 stations. The explained variances of yearly spatial phenology models to BF, FF and EF dates are between 44%-94%, 57%-92% and 39%-84%, respectively. The mean RMSEs of yearly simulation of BF, FF and EF dates are 3.9 days, 4.0 days and 5.4 days, respectively. The predicted multi-year mean Robinia pseudoacacia flowering dates over continuous geographic coverage show a delayed spatial progression from south to north, and from plains to hills and mountains. Therefore, migratory beekeeping can be implemented along with west, middle and east routes in eastern China''s warm temperate zone with durations between 40 day and 50 days. In addition, the predicted linear trends in BF, FF and EF dates over continuous geographic coverage from 1986 to 2005 indicate a dominant advancement and the areas with significant advancement account for 78%, 26% and 32% of the total area, respectively.