Premium
Identification of the most limiting factor for rice yield using soil data collected before planting and during the reproductive stage
Author(s) -
Wang MingMing,
Rengasamy Pichu,
Wang ZhiChun,
Yang Fu,
Ma HongYuan,
Huang LiHua,
Liu Miao,
Yang HaoYu,
Li JingPeng,
An FengHua,
Li YangYang,
Liu XiaoLong,
Liang ZhengWei
Publication year - 2018
Publication title -
land degradation and development
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.403
H-Index - 81
eISSN - 1099-145X
pISSN - 1085-3278
DOI - 10.1002/ldr.3026
Subject(s) - sowing , crop yield , yield (engineering) , soil water , environmental science , explanatory power , agronomy , crop , soil fertility , soil science , mathematics , biology , materials science , philosophy , epistemology , metallurgy
Soil parameters, measured before crop planting, are typically used to quantify the relationship between soil properties and crop yield and to identify factors limiting crop yield. However, soil properties during sensitive stages of crop growth may have a greater effect on crop yield than the initial values. We determined if inclusion of soil properties, measured during the reproductive stage, could improve the accuracy of crop yield prediction. Classification and regression trees were used to determine the explanatory power of different soil and rice variables for predicting rice yield in alkaline salt‐affected paddy fields in northeast China. The traditional method explained 77.5% of the rice yield variation and identified soil CO 3 2− in the 0‐ to 10‐cm soil layer, with single explanatory power of 53.4%, as the most important predictor. The whole explanatory power of the methods including soil variables during the reproductive stage and yield components, with/without soil variables before planting, increased to 81.3%. The residual sodium carbonate, measured in the 0‐ to 10‐cm soil layer during the reproductive stage, was identified as the most limiting factor due to its single maximum explanatory power of 60.5%. We conclude that inclusion of soil properties, measured during the reproductive stage, has potential for improving the rice yield prediction accuracy by enhancing the explanatory power in identification of the most limiting factor. These results encourage further investigation of the role of soil properties during sensitive stages of crop growth in crop yield prediction under different soil and climatic conditions.