基于叶片及冠层叶绿素参数的冬小麦籽粒蛋白质含量预测研究

小麦籽粒蛋白质含量是衡量小麦营养品质的重要指标，实现小麦品质快速的预测预报对于粮食收购部门和加工企业具有重要意义。研究基于作物叶绿素/氮素速测仪SPAD及Multiplex 3，获取冬小麦不同生育期叶片及冠层叶绿素参数，从小麦个体及群体参量两方面进行冬小麦收获期籽粒蛋白质含量及蛋白产量的预测研究。试验于2012年4—6月在国家精准农业研究示范基地开展，研究结果表明，冬小麦返青至灌浆初期，小麦冠层氮素密度与籽粒蛋白质含量的相关性优于叶片氮素含量与蛋白质含量的相关性，灌浆中期两者与籽粒蛋白质含量相关性差别不大;小麦叶片SPAD值与叶片氮素含量相关性总体优于其与冠层氮素密度的相关性，而叶绿素荧光参数SFR_G, SFR_R与冠层氮素密度的相关性优于其与叶片氮素含量的相关性;叶片SPAD与籽粒蛋白质含量的相关性在拔节期最弱，在灌浆中期最强，小麦冠层叶绿素荧光参数SFR_G, SFR_R与籽粒蛋白质含量相关性在返青至拔节期不显著，但孕穗期开始显著相关，在灌浆中期相关性最强且明显优于同期叶片SPAD与籽粒蛋白质含量的相关性;冬小麦籽粒蛋白产量与叶片SPAD值在小麦孕穗期至灌浆期显著相关，与SFR_G和SFR_R在小麦灌浆期显著相关;研究基于灌浆中期SPAD值及SFR_R值，构建了冬小麦籽粒蛋白质含量及籽粒蛋白产量的预测模型，其中，籽粒蛋白质含量预测模型复相关指数分别为0.426和0.497，模型标准误差分别为0.060%和0.055%，籽粒蛋白产量预测模型复相关指数分别为0.366和0.386，模型标准误差分别为125.367和123.454 kg·ha-1。研究表明，利用叶片SPAD值及冠层叶绿素荧光信息，在小麦收获前进行品质的快速预测是可行的。

Winter Wheat GPC Estimation Based on Leaf and Canopy Chlorophyll Parameters

The present study focused on the wheat harvest grain protein content (GPC) estimation based on wheat leaf and canopy chlorophyll parameters, SPAD and SFR, which were acquired by two hand-held instruments, SPAD and Multiplex 3. The wheat GPC estimate experiment was applied on a wheat field of the Scientific Observation and Experiment Field Station for Precision Agriculture at suburb of Beijing in 2012. The wheat leaf SPAD and canopy SFR value were measured in field for all 110 wheat sample points at five different wheat growth stages from April to June. The wheat plant sample for each point was then collected after the SPAD and SFR measurement and sent to lab for leaf nitrogen content (LNC) and canopy nitrogen density (CND) analysis. Analysis results showed that the correlation coefficients of wheat GPC with wheat CND were much higher than that from wheat tillering stage to early milking stage. They were similar at the wheat middle milking stage. While the wheat leaf SPAD value was highly correlated with wheat LNC at wheat tillering, heading and early milking stage. Wheat canopy chlorophyll parameters SFR were highly correlated with wheat CND at wheat tillering, jointing, heading and milking stage. It can be seen from the study that SFR is more sensitive to the wheat CND compared with wheat LNC. The analysis also indicated that leaf SPAD value at wheat tillering, heading and milking stage was highly correlated with wheat GPC and yield of grain protein (YGP). The wheat canopy parameters, SFR_G and SFR_R were significantly correlated with wheat GPC and YGP at wheat milking stage. Then the optimal GPC and YGP estimation model was established. The R2 of GPC estimation models established by SPAD and SFR_R are 0.426 and 0.497, and the standard errors of the estimate are 0.060% and 0.055%, respectively. The R2 of YGP estimation models established by SPAD and SFR_R are 0.366 and 0.386 and the standard errors of the estimate are 125.367 and 123.454 kg·ha-1, respectively. The study reveals that SPAD value is a good indicator of single plant activity while SFR_G and SFR_R are better indicators for the wheat group activity. Wheat leaf SPAD value and canopy chlorophyll fluorescence information SFR are all feasible and valuable for GPC estimation before wheat harvesting.