Agroforestry is an ecological agricultural model that promotes the coordinated development of agriculture and animal husbandry. Exploring appropriate water and nitrogen management strategies for forage grasses in agroforestry systems is of great significance for improving productivity. This study aims to investigate the effects of different water and nitrogen management practices on the growth, nitrogen uptake, and utilization efficiency of intercropped alfalfa in a goji berry-alfalfa system. It is assumed that moderate water deficiency combined with appropriate nitrogen fertilizer can optimize the growth of alfalfa in the intercropping of wolfberry and alfalfa. This study was based on a 2-year (2021 and 2022) field trial, focusing on alfalfa in a goji berry||alfalfa system. Four irrigation levels [full irrigation (W0, 75–85% θ_fc), mild water deficit (W1, 65–75% θ_fc), moderate water deficit (W2, 55–65% θ_fc), and severe water deficit (W3, 45–55% θ_fc)] and four nitrogen application levels [no nitrogen (N0, 0 kg·hm⁻²), low nitrogen (N1, 150 kg·hm⁻²), medium nitrogen (N2, 300 kg·hm⁻²), and high nitrogen (N3, 450 kg·hm⁻²)] were set up to systematically analyze the effects of water and nitrogen regulation on biomass allocation, nitrogen translocation, hay yield, and nitrogen use efficiency of alfalfa. The results showed that (1) irrigation and nitrogen application levels significantly affected the stem-to-leaf and root-to-shoot ratios of alfalfa (p < 0.01). The smallest stem-to-leaf ratio (0.758) was observed under W1N2, while the smallest root-to-shoot ratio (0.595) was observed under W0N2. (2) Irrigation and nitrogen application levels significantly affected nitrogen accumulation and nitrogen translocation in alfalfa (p < 0.05). The maximum nitrogen accumulation was observed under W0N2, which was 43.39% higher than that under W0N0. The maximum nitrogen translocation was observed under W1N2, which was 15.1% and 33.4% higher on average than that under W0N0 and W3N0, respectively. (3) Irrigation and nitrogen application had highly significant effects on alfalfa hay yield (p < 0.01). The highest hay yield (8325 kg·hm⁻² and 12,872 kg·hm⁻²) was achieved under W0N2. The nitrogen productivity of alfalfa increased with increasing water deficit and initially increased, then decreased with increasing nitrogen application. The nitrogen use efficiency of alfalfa followed the order N2 > N1 > N3 and W1 > W0 > W2 > W3, with the highest value of 9.26 under W1N2. Based on the comprehensive evaluation of alfalfa in agroforestry systems under water and nitrogen regulation using the entropy weight-TOPSIS method, mild water deficit combined with medium nitrogen application (W1N2) can optimize the stem-to-leaf ratio, root-to-shoot ratio, and nitrogen use efficiency of alfalfa without significantly reducing yield and nitrogen production efficiency. This water-nitrogen combination is suitable for use in goji berry||alfalfa systems in the Yellow River irrigation area of Gansu Province and similar ecological zones.

Water resources are fundamental to economic and social development. Improving agricultural water-use efficiency is essential for alleviating water scarcity, ensuring food security, and fostering sustainable growth. This study examines the effects of irrigation levels (severe water deficit, W0: 45–55% θ_FC; moderate water deficit, W1: 55–65% θ_FC; mild water deficit, W2: 65–75% θ_FC; full irrigation, W3: 75–85% θ_FC) and nitrogen application rates (N0: 0 kg·hm⁻², N1: 150 kg·hm⁻², N2: 300 kg·hm⁻², N3: 450 kg·hm⁻²) on soil environment, crop yield, and water–nitrogen use efficiencies in Lycium barbarum under integrated water–fertilizer drip irrigation. The coordinated application of water and nitrogen significantly influenced yield and efficiencies (p < 0.05) by modifying rhizosphere conditions such as soil moisture, temperature, salinity, and enzyme activities. Soil temperature increased with nitrogen application (N1 > N2 > N0 > N3), with N1 raising soil temperature by 4.98–8.02% compared to N0, N2, and N3. Electrical conductivity was lowest under N0, showing a 7.53–18.74% reduction compared to N1, N2, and N3. Urease activity peaked under N3 (31.84–96.78% higher than other treatments), while alkaline phosphatase and catalase activities varied across treatments. The yield was highest under N2, at 6.79–41.31% higher than other nitrogen treatments. Water use efficiency (WUE), growth use efficiency (GUE), and nitrogen agronomic efficiency (NAE) peaked under N2, while nitrogen use efficiency (NUE) decreased with higher nitrogen rates. Among irrigation levels, W0 showed the highest soil temperature, while W3 exhibited the lowest conductivity in the 0–40 cm layer. W2 had the highest soil enzyme activities, yielding 4.41–42.86% more than other levels, with maximum efficiencies for WUE, GUE, NUE, and NAE. The combination of mild water deficit (65–75% θ_FC) and 300 kg·hm⁻² nitrogen application (W2N2) resulted in the highest yield (2701.78 kg·hm⁻²). This study provides key insights for implementing integrated drip irrigation in northwest China’s arid regions.

Timely access to soil moisture conditions in farmland crops is the foundation and key to achieving precise irrigation. Due to their high spatiotemporal resolution, unmanned aerial vehicle (UAV) remote sensing has become an important method for monitoring soil moisture. This study addresses soil moisture retrieval in alfalfa fields across different growth stages. Based on UAV multispectral images, a multi-source feature set was constructed by integrating spectral and texture features. The performance of three machine learning models—random forest regression (RFR), K-nearest neighbors regression (KNN), and XG-Boost—as well as two ensemble learning models, Voting and Stacking, was systematically compared. The results indicate the following: (1) The integrated learning models generally outperform individual machine learning models, with the Voting model performing best across all growth stages, achieving a maximum R² of 0.874 and an RMSE of 0.005; among the machine learning models, the optimal model varies with growth stage, with XG-Boost being the best during the branching and early flowering stages (maximum R² of 0.836), while RFR performs better during the budding stage (R² of 0.790). (2) The fusion of multi-source features significantly improved inversion accuracy. Taking the Voting model as an example, the accuracy of the fused features (R² = 0.874) increased by 0.065 compared to using single-texture features (R² = 0.809), and the RMSE decreased from 0.012 to 0.005. (3) In terms of inversion depth, the optimal inversion depth for the branching stage and budding stage is 40–60 cm, while the optimal depth for the early flowering stage is 20–40 cm. In summary, the method that integrates multi-source feature fusion and ensemble learning significantly improves the accuracy and stability of alfalfa soil moisture inversion, providing an effective technical approach for precise water management of artificial grasslands in arid regions.