Heilongjiang Bayi Agricultural University

The application of leguminous green manure (GM) can enhance the soil inorganic phosphorus (Pi) pool, offering considerable benefits for crop cultivation in slightly and moderately saline-alkali soils. To optimize its agronomic potential, systematic and science-based fertilization strategies are required. In this study, we researched the changes in the content, movement distance, and accumulation of Pi fractions at the GM microsites in coastal saline-alkali soils of differing salinity levels (slightly vs. moderately) following the application of Sesbania GM at two rates (30 and 60 t ha−1) over 14- and 28-day incubation periods. The results indicated that GM application significantly (p < 0.05) increased the accumulation of all Pi fractions—including aluminum-bound phosphorus (Al-P), iron-bound phosphorus (Fe-P), occluded phosphorus (O-P), and forms of calcium-bound Pi (Ca-P: Ca2-P, Ca8-P, and Ca10-P)—at the manure microsite, with the magnitude of increase declining with distance from the manure site. Further analysis revealed positive correlations between GM rate, two incubation periods and Pi-fraction movement distance, indicating that the observed effects were significantly influenced by incubation period, GM rate, and soil salinity-alkalinity. While temporal dynamics governed the rates of Pi movement and transformation, elevated salinity-alkalinity partially inhibited these processes. This study provides practical insights for improving GM utilization efficiency on saline-alkali soils. These results support optimized GM application to enhance P efficiency and reduce fertilizer reliance in saline systems.

Natural antisense transcripts (NATs) correspond to nearly 60% of annotated rice loci, however their functions are largely unknown. In this study, we characterise a rice cis-NAT (NAT1850) that completely overlaps with a rice-specific primary miRNA, pri-miR1850. Pri-miR1850, but not its mature miR1850 products, promotes the accumulation of NAT1850, while NAT1850 overexpression in turn reduces the accumulation of pri-miR1850 transcripts. A 21-nt siRNA (siR1850) derived from the pri-miR1850 transcripts is generated by cleavage of pri-miR1850-NAT1850 dsRNA and overlaps in sequence with miR1850.1 and miR1850.2. Both NAT1850 and siR1850 negatively regulate cold tolerance at both the young-seedling and booting stages. Interestingly, siR1850 targets and represses NPR3, which is also a target of miR1850.1. NPR3 interacts with the WRKY76 transcription factor and acts as a co-transcriptional activator of WRKY76 to trigger DREB1B under cold stress. Genetic evidence shows the NAT1850-siR1850 module functions in cold stress response via an NPR3-dependent manner. Furthermore, NAT1850 and siR1850 control nitrogen assimilation and rice yield in a miR1850.1-NPR3-independent pathway. Our findings reveal a regulatory mode for a pri-miRNA and its cis-NAT, and uncover their roles in balancing the cold-stress response and rice yields.

To address issues such as difficulty in capturing subtle movements and complex long-term temporal dependencies in the recognition of food-deprived chewing behavior in pigs, this study proposes an identification method based on an improved SlowFast network. The study integrates spatial attention, temporal attention, and depthwise separable convolution modules into the SlowFast network. Compared with the original SlowFast network (used as the baseline): the spatial attention module increases the mean Average Precision (mAP) by 5.2 percentage points to 73.5%; the depthwise separable convolution module achieves lightweight design with "near-lossless precision", reducing the number of parameters by 60% and increasing the inference speed by 157%; the combination of dual attentions (spatial + temporal) reaches the performance peak with an mAP of 76.2%. The fully improved network (integrating all three modules) achieves an accuracy of 83.6%, an mAP of 75.1%, contains 12.8 million parameters, and has an inference speed of 21.3 FPS. This performance not only meets the storage and real-time requirements of edge devices but also ensures the accuracy of multicategory recognition. This study clarifies the core value and collaborative logic of each improved module, providing empirical support and an optimization scheme for the edge deployment of the SlowFast network in the field of agricultural action recognition. It also shows that the proposed method can accurately identify subtle and continuous food-deprived chewing behaviors, offering a reliable technical means for the automated monitoring of stereotypic behaviors in livestock and poultry.

The concentrations of nitrogen (N), phosphorus (P), potassium (K), organic matter (OM), and pH in soil are critical markers of fertility that influence crop growth and yield. Traditional wet-chemical analyses are labor-intensive, time-consuming, and costly, thereby constraining timely soil information acquisition for precision agriculture. This study evaluates whether multi-output Gaussian process regression (MOGPR) can enhance the prediction accuracy of multiple soil nutrients by exploiting their intrinsic correlations, in comparison with single-output Gaussian process regression (SOGPR). Near-infrared (NIR) spectroscopy was applied to 622 typical black soil samples collected from the Farm 855 (45°43′ N, 131°35′ E), Heilongjiang Province, China. Corresponding MOGPR and SOGPR models were developed for systematic performance comparison. Results indicated that MOGPR significantly outperformed SOGPR for nutrients exhibiting moderate-to-strong intercorrelations (N, P, K, and OM), yielding R2 improvements of 0.070.28 and RPD increases of 16–40%, whereas only limited gains were observed for pH due to its weak correlations with other nutrients. These findings indicate that combining NIR spectroscopy with MOGPR offers significant potential for rapid, nondestructive assessment of multiple soil nutrients. This study further establishes a correlation-aware multi-output modeling framework that links shared spectral responses with an inter-nutrient dependency structure, providing methodological guidance for multi-nutrient soil prediction.