The soil water content (SWC) is a critical factor in agricultural production. To achieve real-time and nondestructive monitoring of the SWC, an experiment was conducted to measure the hyperspectral reflectance of soil samples with varying levels of water content. The soil samples were divided into two parts, SWC higher than field capacity (super-θ_f) and SWC lower than field capacity (sub-θ_f), and the outliers were detected by Monte Carlo cross-validation (MCCV). The raw spectra were processed using Savitzky–Golay (SG) smoothing and then the spectral feature variable of SWC was extracted by using a combination of competitive adaptive reweighted sampling (CARS) and random frog (Rfrog). Based on the extracted feature variables, an extreme learning machine (ELM), a back-propagation artificial neural network (BPANN), and a support vector machine (SVM) were used to establish the prediction model. The results showed that the accuracy of retrieving the SWC using the same model was poor, under two conditions, i.e., SWC above and below θ_f, mainly due to the influence of the lower accuracy of the super-θ_f part. The number of feature variables extracted by the sub-θ_f and super-θ_f datasets were 25 and 18, respectively, accounting for 1.85% and 1.33% of the raw spectra, and the variables were widely distributed in the NIR range. Among the models, the best results were achieved by the BPANN model for both the sub-θ_f and the super-θ_f datasets; the R²p, RMSEp, and RRMSE of the sub-θ_f samples were 0.941, 1.570%, and 6.685%, respectively. The R²p, RMSEp, and RRMSE of the super-θ_f samples were 0.764, 1.479%, and 4.205%, respectively. This study demonstrates that the CARS–Rfrog–BPANN method was reliable for the prediction of SWC.

Soil enzyme not only participates in soil nutrient cycling but affects soil structure. However, the response mechanism of soil enzyme to soil nutrient restriction and aeration at different altitudes is poorly understood. Therefore, In March, June, September, and December 2022, we determined and analyzed the forest soil at 1800, 2100, 2400, and 2700 m altitude in Liziping National Nature Reserve of Sichuan Province, China, and Mantel and Person correlation analyses were employed to analyze the relationship of soil physicochemical properties, CP : NCP and ecological stoichiometric ratios with soil enzyme activities on a seasonal scale. The data showed that the vertical distribution characteristic among soil enzyme activities along the soil layer at different altitudes was similar in each season, and the altitude, season, and their interaction had important effects on forest soil enzyme activities at different altitudes. With the altitude increasing, soil urease, sucrose, acid phosphatase, and catalase showed different variation trends in each season, season, and their interaction caused obvious differences in soil stoichiometric ratios. soil CP : NCP ratio varied with altitude was 1800 > 2100 > 2400 > 2700 m. In addition, the average values of soil C : N, C : P, and N : P ratio in a year were 29.26, 109.29, and 4.52, respectively. The soil CP : NCP ratio was the dominant factor of soil acid phosphatase in March and December (P < 0.01), soil CP : NCP ratio was the dominant factor of soil catalase in September, and soil C : P and CP : NCP ratios were the dominant factors of soil acid phosphatase in June. Furthermore, Soil bulk density and soil moisture content extremely significantly influenced soil urease, sucrase, and acid phosphatase in each season (P < 0.01), and soil noncapillary porosity had a significant effect on soil acid phosphatase in March, June, and December (P < 0.05). The entropy-weighted TOPSIS method obtained that soil enzyme activities, soil physicochemical properties, and CP : NCP-stoichiometric ratios were the worst at 2400 m and in March. This study provides theoretical support for enzyme activities and nutrient restriction characteristics and the driving factors of forest soil enzyme activities in forest soil at different altitudes in this nature reserve and also provides a basis for monitoring and management in this nature reserve in time and space.

To reveal the mechanism of heavy metal Zn migration in the irrigated paddy field reclaimed water, this study investigated irrigation quota of 75, 150, 225 and 300 m³/hm² for three consecutive years. The results showed that with the same irrigation quota, firstly the content of Zn, its variation and the rate of change in soil increased, and hereafter decreased with the increase of soil depth, and finally become stable of reclaimed water. Study results identified that when the irrigation quota was 75, 150, 225 and 300 m³/hm², the average content of Zn in the soil reached the maximum with the value of 9.60, 12.10, 16.75 and 18.50 mg/kg respectively at the depth of 30 cm. The average content of Zn in soil found maximum values of 13.51, 16.01, 19.02 and 20.98 mg/kg, respectively on the 120th day of cultivation. This study also identified that the content of Zn, its variation and the rate of change increased with the increase of irrigation quota at the same soil depth. Additionally, when the soil depth or plant growth time was the same, the content of Zn, its variation and the rate of change increased with the increase of irrigation quota. However, at the soil depth of 30 cm, the content of Zn in the irrigation quota of 75, 150 and 225 m³/hm² decreased by 48.11%, 34.59% and 9.46%. The fertility time of 120 days also decreased by 35.71%, 23.81% and 9.52% respectively compared to an irrigation quota of 300 m³/hm². All the findings are explored by a nonlinear regression under different situations and timing. The mean value of the standard error between the statistical and measured value is found insignificant. However, the correlation coefficient is found greater than 0.9400 and statistically significant. Thus, the findings by nonlinear regression reflected the migration law of soil Zn duly with soil depth and plant growth time in the rice field. This study provided theoretical support for the comprehensive treatment and ecological restoration of heavy metals to the farmland soil in China.