The rapid and accurate acquisition of soil moisture (SM) information is essential. Although Unmanned Aerial Vehicle (UAV) remote sensing technology has made significant advancements in SM monitoring, existing studies predominantly focus on developing models tailored to specific regions. The transferability of these models across different regions remains a considerable challenge. Therefore, this study proposes a transfer learning-based framework, using two representative small agricultural watersheds (Hongxing region and Woniutu region) in Northeast China as case studies. This framework involves pre-training a model on a source domain and fine-tuning it with a limited set of target domain samples to achieve high-precision SM inversion. This study evaluates the performance of three algorithms: Random Forest (RF), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM) network. Results show that the fine-tuned model significantly mitigates the decline in prediction accuracy caused by regional differences. The fine-tuned LSTM model achieved the highest retrieval accuracy, with the following results: 10% samples (R = 0.615, RRMSE = 15.583%), 30% samples (R = 0.682, RRMSE = 13.97%), and 50% samples (R = 0.767, RRMSE = 16.321%). Among these models, the LSTM model exhibited the most significant performance improvement and the best transferability. This study underscores the potential of transfer learning for enhancing cross-regional SM monitoring and providing valuable insights for future UAV-based SM monitoring.

Crop roots selectively recruit certain microbial taxa that are essential for supporting their growth. Within the recruited microbes, some taxa are consistently enriched in the rhizosphere across various locations and crop genotypes, while others are unique to specific planting sites or genotypes. Whether these differentially enriched taxa are different in community composition and how they interact with nutrient cycling need further investigation. Here, we sampled bulk soil and the rhizosphere soil of five soybean varieties grown in Shijiazhuang and Xuzhou, categorized the rhizosphere-enriched microbes into shared, site-specific, and variety-specific taxa, and analyzed their correlation with the diazotrophic communities and microbial genes involved in nitrogen (N) cycling. The shared taxa were dominated by Actinobacteria and Thaumarchaeota, the site-specific taxa were dominated by Actinobacteria in Shijiazhuang and by Nitrospirae in Xuzhou, while the variety-specific taxa were more evenly distributed in several phyla and contained many rare operational taxonomic units (OTUs). The rhizosphere-enriched taxa correlated with most diazotroph orders negatively but with eight orders including Rhizobiales positively. Each group within the shared, site-specific, and variety-specific taxa negatively correlated with bacterial amoA and narG in Shijiazhuang and positively correlated with archaeal amoA in Xuzhou. These results revealed that the shared, site-specific, and variety-specific taxa are distinct in community compositions but similar in associations with rhizosphere N-cycling functions. They exhibited potential in regulating the soybean roots’ selection for high-efficiency diazotrophs and the ammonia-oxidizing and denitrification processes. This study provides new insights into soybean rhizosphere-enriched microbes and their association with N cycling. Key points: • Soybean rhizosphere affected diazotroph community and enriched nifH , amoA, and nosZ. • Shared and site- and variety-specific taxa were dominated by different phyla. • Rhizosphere-enriched taxa were similarly associated with N-cycle functions.