In-season diagnosis of crop nitrogen(N) status is crucial for precision N management. Critical N(N_c) dilution curve and N nutrition index(NNI) have been proposed as effective methods to diagnose N status of different...In-season diagnosis of crop nitrogen(N) status is crucial for precision N management. Critical N(N_c) dilution curve and N nutrition index(NNI) have been proposed as effective methods to diagnose N status of different crops. The N_c dilution curves have been developed for indica rice in the tropical and temperate zones and japonica rice in the subtropical-temperate zone, but they have not been evaluated for short-season japonica rice in Northeast China. The objectives of this study were to evaluate the previously developed N_c dilution curves for rice in Northeast China and to develop a more suitable N_c dilution curve in this region. A total of17 N rate experiments were conducted in Sanjiang Plain, Heilongjiang Province in Northeast China from 2008 to 2013. The results indicated that none of the two previously developed N_c dilution curves was suitable to diagnose N status of the short-season japonica rice in Northeast China. A new N_c dilution curve was developed and can be described by the equation N_c = 27.7 W^(-0.34) if W ≥ 1 Mg dry matter(DM) ha^(-1) or N_c = 27.7 g kg^(-1) DM if W < 1 Mg DM ha^(-1), where W is the aboveground biomass. This new curve was lower than the previous curves. It was validated using a separate dataset, and it could discriminate non-N-limiting and N-limiting nutritional conditions. Additional studies are needed to further evaluate it for diagnosing N status of different rice cultivars in Northeast China and develop efficient non-destructive methods to estimate NNI for practical applications.展开更多
Soil diagnostic horizons, which each have a set of quantified properties, play a key role in soil classification. However, they are difficult to predict, and few attempts have been made to map their spatial occurrence...Soil diagnostic horizons, which each have a set of quantified properties, play a key role in soil classification. However, they are difficult to predict, and few attempts have been made to map their spatial occurrence. We evaluated and compared four machine learning algorithms, namely, the classification and regression tree(CART), random forest(RF), boosted regression trees(BRT), and support vector machine(SVM), to map the occurrence of the soil mattic horizon in the northeastern Qinghai-Tibetan Plateau using readily available ancillary data. The mechanisms of resampling and ensemble techniques significantly improved prediction accuracies(measured based on area under the receiver operator characteristic curve score(AUC)) and produced more stable results for the BRT(AUC of 0.921 ± 0.012, mean ± standard deviation) and RF(0.908 ± 0.013) algorithms compared to the CART algorithm(0.784 ± 0.012), which is the most commonly used machine learning method. Although the SVM algorithm yielded a comparable AUC value(0.906 ± 0.006) to the RF and BRT algorithms, it is sensitive to parameter settings, which are extremely time-consuming.Therefore, we consider it inadequate for occurrence-distribution modeling. Considering the obvious advantages of high prediction accuracy, robustness to parameter settings, the ability to estimate uncertainty in prediction, and easy interpretation of predictor variables, BRT seems to be the most desirable method. These results provide an insight into the use of machine learning algorithms to map the mattic horizon and potentially other soil diagnostic horizons.展开更多
基金supported by the Key National Research and Development Program (No. 2016YFD0200602)the National Basic Research Program (No. 2015CB150405)+1 种基金the National Natural Science Foundation (No. 31421092)the SINOGRAIN Project (No. CHN-2152, 14-0039) of China
文摘In-season diagnosis of crop nitrogen(N) status is crucial for precision N management. Critical N(N_c) dilution curve and N nutrition index(NNI) have been proposed as effective methods to diagnose N status of different crops. The N_c dilution curves have been developed for indica rice in the tropical and temperate zones and japonica rice in the subtropical-temperate zone, but they have not been evaluated for short-season japonica rice in Northeast China. The objectives of this study were to evaluate the previously developed N_c dilution curves for rice in Northeast China and to develop a more suitable N_c dilution curve in this region. A total of17 N rate experiments were conducted in Sanjiang Plain, Heilongjiang Province in Northeast China from 2008 to 2013. The results indicated that none of the two previously developed N_c dilution curves was suitable to diagnose N status of the short-season japonica rice in Northeast China. A new N_c dilution curve was developed and can be described by the equation N_c = 27.7 W^(-0.34) if W ≥ 1 Mg dry matter(DM) ha^(-1) or N_c = 27.7 g kg^(-1) DM if W < 1 Mg DM ha^(-1), where W is the aboveground biomass. This new curve was lower than the previous curves. It was validated using a separate dataset, and it could discriminate non-N-limiting and N-limiting nutritional conditions. Additional studies are needed to further evaluate it for diagnosing N status of different rice cultivars in Northeast China and develop efficient non-destructive methods to estimate NNI for practical applications.
基金supported by the National Natural Science Foundation of China (Nos. 41501229, 41371224, 41130530, and 91325301)the China Postdoctoral Science Foundation (No. 2015M581876)
文摘Soil diagnostic horizons, which each have a set of quantified properties, play a key role in soil classification. However, they are difficult to predict, and few attempts have been made to map their spatial occurrence. We evaluated and compared four machine learning algorithms, namely, the classification and regression tree(CART), random forest(RF), boosted regression trees(BRT), and support vector machine(SVM), to map the occurrence of the soil mattic horizon in the northeastern Qinghai-Tibetan Plateau using readily available ancillary data. The mechanisms of resampling and ensemble techniques significantly improved prediction accuracies(measured based on area under the receiver operator characteristic curve score(AUC)) and produced more stable results for the BRT(AUC of 0.921 ± 0.012, mean ± standard deviation) and RF(0.908 ± 0.013) algorithms compared to the CART algorithm(0.784 ± 0.012), which is the most commonly used machine learning method. Although the SVM algorithm yielded a comparable AUC value(0.906 ± 0.006) to the RF and BRT algorithms, it is sensitive to parameter settings, which are extremely time-consuming.Therefore, we consider it inadequate for occurrence-distribution modeling. Considering the obvious advantages of high prediction accuracy, robustness to parameter settings, the ability to estimate uncertainty in prediction, and easy interpretation of predictor variables, BRT seems to be the most desirable method. These results provide an insight into the use of machine learning algorithms to map the mattic horizon and potentially other soil diagnostic horizons.
