Semi-arid Chernozem farmland in Northeast China faces the dual challenges of water scarcity and excessive nitrogen (N) fertilisation, yet quantitative understanding of coupled water–N use efficiency and environmental trade-offs under climatic variability remains limited. Using the WHCNS (Water, Heat, Carbon and Nitrogen Simulator) model calibrated and validated against field data (2022–2023) from Qiqihar, we quantified crop yield, water use efficiency (WUE), System-Level Nitrogen Conversion Efficiency (SLNCE) and N losses under five management regimes encompassing conventional fertilisation, straw mulching and organic manure substitution. Long-term climate adaptability was evaluated using precipitation scenarios from 2001 to 2023. Straw mulching reduced evapotranspiration by 12.2% and increased WUE to 2.09 kg·m⁻³. Combined chemical and organic fertiliser application (T223) achieved the highest yield (9745.9 kg·ha⁻¹) and WUE (2.14 kg·m⁻³), whereas a 30% reduction in chemical fertiliser with straw return (T323) reduced yield and SLNCE, indicating insufficient N supply. Organic substitution achieved the highest SLNCE (35.9 kg·kg⁻¹), exceeding the straw-only treatment by 29.1%. Nitrate leaching below 1 m was negligible, whereas gaseous N losses were the dominant loss pathway and increased with rainfall. SLNCE responded nonlinearly to precipitation, peaking at 420–451 mm; extreme drought years constrained yield through nitrogen limitation, whereas wet years increased gaseous N losses while maintaining relatively high yields. These results show that organic substitution can simultaneously enhance WUE and SLNCE, providing a climate-adaptive strategy for sustainable water–N management in semi-arid Chernozem soils.

To investigate how steam-explosion pretreatment affects humification during sawdust composting, an aerobic composting experiment was conducted using sawdust, chicken manure, and spent mushroom substrate as feedstocks. Two treatments were established—a steam-explosion-pretreated sawdust group (SEW) and an untreated sawdust control (CK)—each with three replicate reactors. Samples were collected dynamically at five key composting stages (initial, heating, thermophilic, cooling, and maturation) for physicochemical, enzymatic, and microbial community analyses. Linear mixed-effects model analysis revealed that enzyme activities were significantly affected by treatment, composting time, and their interaction. SEW significantly enhanced cellulase and polyphenol oxidase activities, and increased laccase and peroxidase activities at specific stages. Compared with CK (humic substances, 75.30 g/kg), SEW promoted higher humic substance accumulation (120.80 g/kg) and altered the dynamics of dissolved organic carbon. Microbial co-occurrence networks in SEW (50 nodes, 602 edges) were more complex than CK (49 nodes, 464 edges), indicating tighter microbial interactions. Path analysis revealed that HS in CK was mainly influenced by DOC and temperature, while HS in SEW was associated with enzyme activities, microbial diversity, and Pseudogracilibacillus. These results suggest that steam-explosion pretreatment enhances substrate transformation and humic substance formation during composting.