State Key Laboratory of Soil and Sustainable Agriculture, Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008, China.
University of Chinese Academy of Sciences, Beijing 100049, China.
Sensors (Basel). 2017 Oct 8;17(10):2287. doi: 10.3390/s17102287.
Efficient and precise yield prediction is critical to optimize cabbage yields and guide fertilizer application. A two-year field experiment was conducted to establish a yield prediction model for cabbage by using the Greenseeker hand-held optical sensor. Two cabbage cultivars (Jianbao and Pingbao) were used and Jianbao cultivar was grown for 2 consecutive seasons but Pingbao was only grown in the second season. Four chemical nitrogen application rates were implemented: 0, 80, 140, and 200 kg·N·ha. Normalized difference vegetation index (NDVI) was collected 20, 50, 70, 80, 90, 100, 110, 120, 130, and 140 days after transplanting (DAT). Pearson correlation analysis and regression analysis were performed to identify the relationship between the NDVI measurements and harvested yields of cabbage. NDVI measurements obtained at 110 DAT were significantly correlated to yield and explained 87-89% and 75-82% of the cabbage yield variation of Jianbao cultivar over the two-year experiment and 77-81% of the yield variability of Pingbao cultivar. Adjusting the yield prediction models with CGDD (cumulative growing degree days) could make remarkable improvement to the accuracy of the prediction model and increase the determination coefficient to 0.82, while the modification with DFP (days from transplanting when GDD > 0) values did not. The integrated exponential yield prediction equation was better than linear or quadratic functions and could accurately make in-season estimation of cabbage yields with different cultivars between years.
精确高效的产量预测对于优化甘蓝产量和指导施肥至关重要。本试验采用 Greenseeker 手持光学传感器进行了为期两年的田间试验，旨在建立甘蓝产量预测模型。供试品种为健宝和坪宝，其中健宝连续 2 季种植，坪宝仅在第二季种植。共设置 4 个施氮量处理：0、80、140 和 200 kg·N·ha。移栽后 20、50、70、80、90、100、110、120、130 和 140 天分别采集归一化植被指数（NDVI）。采用 Pearson 相关性分析和回归分析方法，明确 NDVI 与甘蓝产量的关系。结果表明，110 天 NDVI 与产量显著相关，两年试验中健宝品种的产量变异可分别解释 87-89%和 75-82%，坪宝品种的产量变异可解释 77-81%。将累积生长度日（CGDD）纳入模型可显著提高预测模型的精度，决定系数提高到 0.82，而用移栽后有效积温（DFP）值进行校正则没有显著改善。综合指数产量预测方程优于线性或二次函数，能够准确估算不同品种甘蓝在不同年份的产量。
