论文部分内容阅读
高光谱遥感能够快速无损地估测作物生长性状及产量,这为作物规模化育种的田间评价与选择提供了高效手段。选用生育时期相似、生长性状有差异的52份大豆品种(系)进行2年田间试验,在盛花期(R2)、盛荚期(R4)及鼓粒初期(R5)测定大豆冠层反射光谱,同步测定大豆叶面积指数(LAI)和地上部生物量(ABM),收获后测定产量。针对不同生育时期冠层光谱与生长性状及产量进行偏最小二乘回归(PLSR)分析。结果表明:不同生育时期LAI的PLSR模型可以解释LAI总变异的54.4%~61.0%;不同生育时期ABM的PLSR模型可以解释ABM总变异的65.5%~67.0%;R5期是利用冠层光谱估测产量的最佳生育时期,其PLSR模型可以解释产量总变异的66.1%。本研究结果可望为大豆规模化育种中大量试验材料的田间长势监测和产量估测提供快速无损预测的技术支持。
Hyperspectral remote sensing can quickly and nondestructively estimate crop growth traits and yields, which provides an efficient method for field evaluation and selection of crop large-scale breeding. Fifty-two soybean cultivars with similar growth periods and different growth traits were used for field experiments for two years. The canopy reflectance spectra of soybean were measured at the full flowering stage (R2), pod stage (R4) and initial drum stage (R5) Simultaneous determination of leaf area index (LAI) and aboveground biomass (ABM), yield was measured after harvest. Partial least squares regression (PLSR) analysis of canopy spectra, growth traits and yield at different growth stages was conducted. The results showed that the PLSR model of LAI explained 54.4% -61.0% of the total variation of LAI at different growth stages; PLSR model of ABM explained 65.5% -67.0% of total variation of ABM at different growth stages; R5 stage was estimated by canopy spectral analysis At the optimum fertility stage, the PLSR model explained 66.1% of the total variation in yield. The results of this study are expected to provide rapid and predictable technical support for field monitoring and yield estimation of a large number of test materials in large-scale soybean breeding.