The experiment was conducted at the Shihezi University Experimental Station (44°18′N, 86°03′E) from 2023 to 2024. The site is located at an elevation of 435 m, with a mean annual precipitation of 208 mm and mean annual potential evaporation of 1,660 mm. The region has a temperate continental arid climate, characterized by limited rainfall, abundant sunshine, and sufficient thermal resources. Temporal variations in temperature and precipitation during the soybean growing season are shown in Figure 1 . The physicochemical properties of the soil (0–60 cm depth) in the experimental field are presented in Table 1 .

A split-plot design with two factors was used. The main plot factor was tillage method: conventional tillage (CT) and no-till (NT). The subplot factor consisted of four nitrogen application rates: 0 kg·ha^-¹ (N0), 105 kg·ha^-¹ (N1), 150 kg·ha^-¹ (N2), and 195 kg·ha^-¹ (N3). The eight treatments were replicated three times. Each plot measured 20 m² (4 m × 5 m). Soybeans (cv. Haojiang 35, growth period 92 days) were sown on 2 July 2023 and 4 July 2024 at a depth of 2–3 cm, with row spacing of 30 cm, plant spacing of 6 cm, and a density of 5.55×10⁵ plants·ha^-¹. Chemical fertilizers included urea (46% N), potassium dihydrogen phosphate (52% P₂O₅, 34% K₂O), and potassium sulfate (51% K₂O). Irrigation was applied nine times during the growing season, synchronized with fertilizer application. Detailed fertilization rates are provided in Table 2 . Other field management practices followed local standards.

Data were organized using Microsoft Excel 2021. Statistical analyses were performed using IBM SPSS Statistics 20.0. A two-way analysis of variance (ANOVA) appropriate for the split-plot design was applied to test the significance (P < 0.05). Prior to ANOVA, assumptions of normality (Shapiro-Wilk test) and homogeneity of variances (Levene’s test) were verified. Mean separation was conducted using Tukey’s Honestly Significant Difference (HSD) test. Data visualization was performed using Origin 2021.

Analysis of variance indicated that the main effects of year (Y) (p < 0.001), tillage (T) (p < 0.001), nitrogen (N) (p < 0.001), and the interactions of Y×T (p = 0.032), T×N (p < 0.001), and Y×T×N (p < 0.001) significantly affected soybean yield. The interaction of Y×N was not significant (p = 0.262) ( Table 3 ). Specifically, under N0 conditions, yield under CT was 7.7–36.5% higher than under NT (P < 0.05) ( Figure 2 ). Nitrogen application markedly increased yield in both systems up to the N2 rate (150 kg ha^-¹). Compared to CTN0, CTN2 yield increased by 14.5–42.6%, while the response was even greater under NT, with NTN2 yield increasing by 75.7–83.4% relative to NTN0. However, increasing the rate to N3 (195 kg ha^-¹) did not improve yield further; both CTN3 and NTN3 yields were significantly lower than those of their respective N2 treatments. The two-year average yield for NTN3 was numerically 8.9% higher than CTN3. These results suggest that while higher nitrogen rates can partially compensate for yield reduction under NT, the optimal N rate is system-dependent, and maintaining yield under NT requires precise N input management combined with long-term soil improvement.

The hundred-seed weight and number of seeds per plant were significantly influenced by the main effects of year (Y) (p < 0.001), tillage (T) (p = 0.001 for hundred-seed weight; p < 0.001 for seeds per plant), nitrogen (N) (p < 0.001 for both), and the interactions of Y×N (p < 0.001 for both), T×N (p < 0.001 for both), and Y×T×N (p < 0.001 for hundred-seed weight; p = 0.005 for seeds per plant). The Y×T interaction was not significant for hundred-seed weight (p = 0.777) but was significant for the number of seeds per plant (p < 0.001) ( Table 3 ). Overall, CT resulted in significantly higher hundred-seed weight and more seeds per plant than NT when averaged across nitrogen levels ( Table 4 ). Nitrogen application up to the N2 rate consistently improved yield components in both tillage systems. The NTN2 treatment demonstrated substantial improvements, increasing the number of seeds per plant by 35.8–51.1% compared to NTN0 across both years (P < 0.05). Under CT conditions, the N2 rate also significantly increased yield components compared to CTN0. Conversely, increasing the nitrogen application to the N3 rate led to a consistent reduction in both hundred-seed weight and seeds per plant relative to the N2 rate in both tillage systems, indicating a negative effect of excessive nitrogen on yield formation. Notable interannual variability was observed in the response of hundred-seed weight to tillage practices. Nevertheless, the NTN2 treatment effectively enhanced yield components, with the number of seeds per plant reaching 37.09 and 64.50 in 2023 and 2024, respectively, representing significant improvements over the NTN0 treatment.

Soil water storage (SWS) was not significantly influenced by any main effects or interactions. The main effects of year (Y) (p = 0.119), tillage (T) (p = 0.854), nitrogen (N) (p = 0.996) and the interactions of Y×T (p = 0.932), Y×N (p = 0.703), T×N (p = 0.593), and Y×T×N (p = 0.223) were not significant ( Table 5 ). Despite the lack of statistical significance, observed values showed considerable variation between years under different treatments ( Figure 3 ). In 2023, under the N0 condition, SWS was numerically higher under CT than under NT across the 0–60 cm soil profile. By contrast, in 2024, NTN0 showed numerically higher SWS than CTN0 at the same depths. Nitrogen application at the N2 rate increased SWS relative to N0 in both tillage systems. Notably, the NTN2 treatment in 2024 maintained numerically higher SWS in the 20–60 cm layers compared to CTN2, with increases ranging from 12.5% to 21.7%. Additionally, the highest SWS values under nitrogen application were consistently observed in the N2 and N3 treatments. These results demonstrate that nitrogen application significantly influenced SWS, with the NTN2 regime exhibiting improved water retention in subsurface layers during the second year.

Soil total nitrogen (STN) content was significantly influenced by the main effect of year (Y) (p = 0.001) and the interaction of Y×N (p < 0.001). The main effects of tillage (T) (p = 0.232), nitrogen (N) (p = 0.407) and the interactions of Y×T (p = 0.898), T×N (p = 0.561), and Y×T×N (p = 0.304) were not significant ( Table 5 ). Considerable interannual variation was observed under the N0 condition. In 2023, STN under CTN0 was significantly higher than under NTN0 across all sampled soil layers (P < 0.05) ( Figure 4 ). Conversely, in 2024, this trend was reversed in the upper soil layers (0–40 cm), where NTN0 surpassed CTN0 by 36.0–49.4%. The application of nitrogen at the N2 rate significantly increased STN under both tillage systems (P < 0.05). The NTN2 treatment consistently demonstrated substantial improvements in STN compared to NTN0, with increases of 51.1–135.1% observed across the two years. Notably, in 2024, NTN2 also resulted in numerically higher STN values than CTN2 in the subsurface layers (20–60 cm). These results indicate that while tillage practice alone did not significantly affect STN, its interaction with nitrogen application played a crucial role in determining soil nitrogen content, with the NTN2 combination showing particularly beneficial effects on soil nitrogen status.

Leaf area index (LAI) was significantly influenced by the main effects of tillage (T) (p < 0.001) and nitrogen (N) (p = 0.007). The main effect of year (Y) (p = 0.637) and the interactions of Y×T (p = 0.395), Y×N (p = 0.523), T×N (p = 0.427), and Y×T×N (p = 0.464) were not significant ( Table 6 ). CT resulted in significantly higher LAI than NT across growth stages and nitrogen levels ( Figure 5 ). Nitrogen application markedly improved LAI in both tillage systems, with the N2 rate consistently producing the highest LAI values across all growth stages. The most substantial improvements were observed under the NT system with N2 application. NTN2 increased LAI by 32.3–76.8% at the R4 stage and by 36.4–36.8% at the R6 stage relative to NTN0. Similarly, under CT, CTN2 increased LAI by 18.3–101.8% at R2, 22.2–42.9% at R4, and 25.0–79.7% at R6 compared to CTN0. These results demonstrate that while CT generally maintained higher LAI than NT under equivalent nitrogen conditions, the application of nitrogen at the N2 rate effectively mitigated this difference, particularly during the critical R4 and R6 growth stages.

SPAD values were not significantly influenced by any of the main effects or their interactions ( Table 6 ). This suggests that leaf chlorophyll content, as estimated by SPAD, was stable across the different management practices and experimental years. Despite the lack of significant effects, descriptive comparisons revealed trends. In 2023 under N0, CT exhibited numerically higher SPAD values than NT at all growth stages. Conversely, in 2024, NT showed numerically higher values than CT at the R2 and R6 stages ( Figure 6 ). Nitrogen application markedly increased SPAD values in both tillage systems. The N2 treatment consistently produced the highest SPAD values across different growth stages, significantly outperforming the N0 treatment. Under NT, the NTN2 treatment increased SPAD values by 17.7–24.3% at the R4 stage and by 4.4–18.6% at the R6 stage compared to NTN0.

Significant positive correlations (P < 0.05) were observed between grain yield (GY) and multiple physiological traits, including chlorophyll content (SPAD), net photosynthetic rate (Pn), stomatal conductance (Gs), transpiration rate (Tr), intercellular CO₂ concentration (Ci), leaf area index (LAI), along with the yield component hundred-seed weight (HSW) ( Figure 7 ). Conversely, neither soil water storage (SWS) nor total soil nitrogen (STN) showed significant correlations with GY (P > 0.05).

Integrated tillage and nitrogen (N) management is widely recognized as an effective strategy to overcome the limitations of single-factor agricultural practices, ultimately enhancing crop yield (Jiang et al., 2022; Zhang et al., 2023). This integrated approach is particularly critical during the transition from CT to NT, which often involves a period of yield decline (Li Z, et al., 2024). Under the N0 condition, the NT treatment significantly reduced seeds per plant and overall yield compared to CT, this overall yield decline was primarily driven by a consistent and significant reduction in the number of seeds per plant under NT (Imani et al., 2022). These reductions may be attributed to the initial immobilization of nitrogen under NT systems due to surface residue retention, which can limit early nutrient availability and impede root development during critical growth stages. Nitrogen optimization plays a pivotal role in maintaining soil fertility and maximizing yield potential (Fernández-Ortega et al., 2023). However, excessive N application—exemplified by the N3 rate in this study—can disrupt source-sink balance, leading to inefficient translocation of photoassimilates and impaired grain filling processes (Liu et al., 2018). Our results demonstrate that both CT and NT systems attained peak productivity at the N2 rate, beyond which yield losses occurred primarily due to reduced hundred-seed weight. This suggests that optimal N fertilization is essential for balancing vegetative and reproductive growth, regardless of tillage regime. Tillage practices profoundly influence the root zone environment by altering soil structure, residue distribution, and microbial community dynamics (Du et al., 2023). While CT facilitates nitrogen mineralization and root proliferation in the short term, promoting grain filling and seed weight (Ding et al., 2021), NT combined with optimal N (N2) achieved statistically comparable yields to CT in this study. This indicates that NT can effectively compensate for its early-growth limitations through improved water and nutrient retention over time. Moreover, NT under the N2 regime resulted in higher levels of soil moisture and nitrogen availability in the 2024 growing season, suggesting that long-term NT systems may foster a more resilient rhizosphere environment conducive to root-microbe interactions and sustained yield formation. The synergistic benefits of integrated tillage and N management highlight the importance of adapting agronomic practices to enhance system sustainability and productivity under evolving climatic conditions.

Tillage practices fundamentally alter soil physical structure and biogeochemical cycling, with contrasting mechanisms driving system performance. CT improves short-term aeration and water infiltration by mechanically loosening the soil profile; however, long-term use disrupts aggregate stability, accelerates organic matter mineralization, and can lead to subsurface compaction (Hu et al., 2018). In contrast, NT minimizes mechanical disturbance, preserves biopore-derived porosity, and enhances surface residue retention. These features collectively mitigate erosion and reduce nitrogen loss by slowing organic matter decomposition and mineralization rates (Brevilieri et al., 2024; Song et al., 2024; Zhang et al., 2025). Our results revealed notable interannual variability in SWS and STN under NT without nitrogen application (NTN0). The higher SWS under CTN0 in 2023 can be attributed to improved initial infiltration in the freshly tilled and loosened soil. The reversal in 2024, however, suggests a progressive improvement in NT’s soil structural properties and hydrological functioning. Over time, continued residue accumulation, fungal hyphae development, and earthworm activity enhance pore connectivity and water-stable aggregation, thereby increasing water retention capacity—a process consistent with models of ecosystem development in reduced-disturbance systems (Wen et al., 2024). Furthermore, the positive correlation between STN and SWS in upper soil layers (0–40 cm) under NT underscores the role of coupled hydro-nutrient retention in surface horizons, likely mediated by organic material accumulation and reduced evaporative loss. However, the significant reduction in STN in deeper soil layers (40–60 cm) under NTN0 compared to CT highlights a potential limitation of NT in regulating nutrient leaching. This may be attributed to reduced root exploration in the subsoil under NT during early transition periods, along with the development of preferential flow paths through biopores and cracks, which can facilitate the downward movement of soluble nitrogen before plants establish extensive root systems. Application of nitrogen at the N2 rate ameliorated these constraints through multiple mechanisms: NTN2 not only retained more water in the 20–60 cm depth than CTN2 but also maintained higher STN, likely due to improved plant growth, enhanced root-derived carbon inputs, and stimulated microbial activity that promotes nutrient immobilization and cycling (Fashi et al., 2019; Wang et al., 2021). The deeper rooting system encouraged under stabilized NT environments also facilitates nutrient uptake from subsurface layers, reducing leaching losses. Conversely, the N3-induced reductions in STN across all depths reflect exacerbated leaching risks, particularly below 40 cm. This is consistent with reports of increased nitrate mobility under excessive N inputs, which overwhelm plant uptake and microbial immobilization capacity, leading to displacement of nitrogen beyond the root zone (Carvalho et al., 2024; Fang et al., 2006). The deeper leaching under N3 also implies that NT systems, despite their surface advantages, are not immune to nitrogen loss through leaching under imbalanced fertilization, highlighting the necessity of coupling tillage practices with precise nitrogen management to achieve both agronomic and environmental objectives.

Photosynthetic performance is intimately linked to soil management practices through their effects on water availability, nitrogen supply, and root system functioning. The observed superiority in LAI and SPAD values under CTN0 compared to NTN0 can be attributed to improved early-stage root exploration in the mechanically loosened soil of CT systems, facilitating greater access to soil moisture and nutrients—particularly nitrogen—which in turn promotes canopy development and chlorophyll accumulation (Farhangi-Abriz et al., 2021; Hafeez et al., 2019). By contrast, the compacted and less-aerated soil typical of initial NT systems can physically constrain root growth and limit early nutrient uptake, resulting in delayed canopy establishment. However, NT systems exhibited a pronounced and compensatory response to nitrogen addition. The significant increases in LAI, SPAD, Pn, and Gs under NT with N2 indicate that nitrogen application effectively mitigated these early physiological constraints. The mechanism likely involves improved leaf nitrogen status, facilitating enhanced chlorophyll biosynthesis and greater photosynthetic enzyme activity, which collectively elevate light capture and carbon assimilation capacity (Nkebiwe et al., 2016; Qiang et al., 2025; Savala et al., 2021). The depression in Pn coinciding with elevated Ci values across treatments in 2024, particularly under NT, suggests the presence of non-stomatal limitations to photosynthesis. This may reflect metabolic impediments such as reduced Rubisco activity, diminished electron transport capacity, or feedback inhibition due to carbohydrate accumulation—all of which can occur under suboptimal source-sink balance or abiotic stress (Buczek et al., 2022). Notably, the superior photosynthetic performance in NT during 2024 indicates that the system underwent functional adaptation over time. We propose that this enhancement arises from improved microbial-mediated nitrogen mineralization and stabilization (Zhang et al., 2020), combined with the development of a deeper and more branched root architecture that better accesses subsurface water and nutrients (Jeelani, 2017). Furthermore, the accumulation of soil organic matter and stabilization of soil aggregates under NT may enhance moisture retention and microclimate conditions, supporting more sustained photosynthetic activity during critical growth stages. These findings illustrate that while CT can create initially favorable physical conditions for crop growth, NT—when paired with optimized nitrogen input—fosters a more resilient and efficient photosynthetic apparatus. The results underscore the importance of biological adaptation and nutrient–water synergism in no-till systems, highlighting their capacity to not only compensate for early limitations but also to achieve sustained photosynthetic productivity under variable environmental conditions.