Based on historical soil data, this paper studied the spatial variability of soil texture and its relationship with environmental factors in southern Ningxia by using machine learning.

Classification and regression tree (CART), random forest (RF) and traditional statistical methods were used to explore the main environmental factors that affected the soil texture types and predict the spatial distribution of soil texture types in southern Ningxia, based on 428 soil profiles from the second soil survey in the 1980s, combined with topographic factors, soil types, and normalized vegetation index. The accuracy of the models were verified by the validating set of soil profiles and the soil samples in Haiyuan County, Ningxia.

(1) The accuracy rates of RF and CART on the soil texture type of the verification set of soil profiles were 62.36% and 55.29%, respectively; the area under the receiver operating characteristic (ROC) curve of them (area under roc curve, AUC) were 0.7515 and 0.6933, respectively; the accuracy rates of them on soil samples in Haiyuan County were 54.10% and 48.36%, respectively; the AUC of them were 0.6599 and 0.5981 respectively. (2) Soil type (ST) was the most important predictor variable, followed by elevation (Ele). The higher elevation was, the heavier the soil texture was. The effects of wind exposition index (WEI) and slope (Slo) on soil texture were lower. (3) The results predicted by two methods showed a spatial distribution trend that the soil texture was heavy in the southern area but light in the northern area of southern Ningxia.

The prediction accuracy of RF for soil texture type in southern Ningxia was higher than CART. Making full use of historical data, combined with field sampling, could meet the accuracy requirements of digital mapping. In the loess region, soil types and elevation were the environmental factors which had strong correlation with spatial variation of soil texture.

This study aimed to explore the spatial-temporal variation characteristics and influencing factors of relative yield gap of wheat and maize in China during the past 15-20 years and the response of relative yield gap to nitrogen fertilizer under different soil productivity levels, so as to provide a theoretical basis for rational application of nitrogen fertilizer and the realization of high and stable yield of wheat and maize.

Based on the long-term monitoring database, the difference of wheat and maize yield between fertilized area and non-fertilized area was used to represent the relative yield (RY). The highest relative yield (HRY), the average relative yield (ARY) and the relative yield gap (G_RY) were obtained by using the statistical of high-yielding households, the effects of fertilization and soil factors on the relative yield gap were determined used the random forest model, and soil productivity level was divided according to the yield of non-fertilized area. The relationship between the relative yield gap of wheat and maize and nitrogen application rate under different soil productivity levels was quantified.

HRY of wheat in China was 3.83-6.75 t·hm^-2, ARY was 2.10-3.42 t·hm^-2, and G_RY was 1.73-3.33 t·hm^-2, G_RY accounting for 44.64%-49.06% of HRY. HRY, ARY and G_RY of wheat were north China>middle-lower Yangtze Plain>northwest China>southwest China. HRY of maize in China was 6.53-8.20 t·hm^-2, ARY was 3.37-4.12 t·hm^-2, and G_RY was 3.16-4.08 t·hm^-2, G_RY accounting for 44.78%-50.52% of HRY. HRY, ARY and G_RY of maize were northeast China>north China>southwest China>northwest China. Except for north China, HRY and G_RY of wheat and maize increased with time. Except in northwest China, the G_RY decreased with the increase of nitrogen application rate in low and medium soil productivity, and the decrease amplitude was more significant in low soil productivity level, while the decrease of G_RY with nitrogen application rate in high soil productivity was not significant. Regionally, the balance points of nitrogen fertilizer application were found in wheat and maize in North China, wheat in middle-lower Yangtze Plain, and maize in northeast China at low and medium soil productivity. Overall, the nitrogen application rate and soil organic matter were relatively important influencing factors of G_RY for wheat and maize at low and medium soil productivity. Potassium application had a significant impact on the G_RY in middle-lower Yangtze Plain and north China, while organic matter had a significant impact on the G_RY in the northwest and southwest China under high soil productivity.

N application and soil organic matter were important factors affecting the relative yield gap. The higher soil productivity level, the lower the effect of nitrogen fertilizer on reducing the relative yield gap. N fertilizer should be reduced appropriately in high productivity soil. In order to increase yield and avoid the waste of resource and environmental risks, it was suggested that the application rate of nitrogen fertilizer should not exceed its balance point. The recommended application rates of nitrogen fertilizer were 260.6 and 159.2 kg·hm^-2 for wheat and 262.5 and 246.0 kg·hm^-2 for maize at low and medium productivity levels in north China, respectively. In the middle-lower Yangtze Plain, 199.5 and 187.5 kg·hm^-2 were recommended for nitrogen application at low and medium productivity levels, respectively. In northeast China, the recommended amount of N fertilizer application was 259.5 and 228.0 kg·hm^-2, respectively. Under low and medium productivity levels in southwest and northwest China, N fertilizer should be appropriately increased. The potassium fertilizer reasonable application should be paid more attention at high soil productivity in north China and middle-lower Yangtze Plain. The improvement of soil organic matter should be as the main measures to achieve high and stable yields in southeast and southwest China.