Objectives Grass pea (Lathyrus sativus) exhibits strong stress resistance and high nutrient content, making it one of the major leguminous green manure varieties promoted in the hilly regions of Sichuan. We investigated the effects of reduced nitrogen (N) fertilizer application combined with different incorporation rates of grass pea green manure on rice yield and paddy soil quality in the hilly areas of Sichuan, aiming to provide technical and theoretical support for the rational utilization of grass pea as green manure and reduced N fertilizer application in paddy fields within this region.

Methods A long-term field experiment was initiated in 2017 in Nanchong, Sichuan. It included two control treatments: winter fallow without N fertilizer application (CK) and winter fallow with 100% N fertilizer application (CF), as well as eight combined treatments involving a 40% (F₆₀) or 20% (F₈₀) reduction in N fertilizer application, coupled with the winter cultivation and subsequent incorporation of grass pea at rates of 15.0 (G₁₅), 22.5 (G_22.5), 30.0 (G₃₀), and 37.5 (G_37.5) t/hm². Rice yield was measured after harvest in 2022, along with the determination of nitrogen uptake by the above-ground parts of rice plants. Additionally, soil bulk density, porosity, chemical properties (pH, organic matter, total nitrogen, alkali-hydrolyzable nitrogen, available phosphorus, and readily available potassium) in the 0−20 cm soil layer, as well as microbial biomass carbon and nitrogen contents were analyzed.

Results Compared with CF, all treatments combining N fertilizer reduction with grass pea incorporation significantly increased rice yield by 7.5%−21.0%. Under 40% N reduction, treatments G₃₀F₆₀ and G_37.5F₆₀ were recorded significantly higher rice grain yields and above-ground part N uptake than G₁₅F₆₀, and under 20% N reduction, treatments G₃₀F₈₀ and G_37.5F₈₀ were recorded significantly higher rice grain yield and N uptake than G₁₅F₈₀ and G_22.5F₈₀, and there was no significant differences in rice yield and N uptake between treatment G_37.5F₆₀, and G_37.5F₈₀. Compared to CF, the treatments G_37.5F₆₀, G₃₀F₈₀, and G_37.5F₈₀ significantly reduced soil bulk density by 0.13−0.20 g/cm³ and increased soil organic matter, total nitrogen, available nutrient contents, as well as soil microbial biomass carbon and nitrogen contents (P<0.05). A comprehensive evaluation of soil quality using the minimum data set method indicated that appropriately increasing the grass pea incorporation rate was more conducive to enhancing soil quality. Under a 40% N reduction, G_37.5F₆₀ exhibited the highest soil quality index (SQI), significantly increasing by 14.5%−27.9% compared to G₁₅F₆₀, G_22.5F₆₀, and G₃₀F₆₀. Under a 20% N reduction, G_37.5F₈₀ had the highest SQI, significantly increasing by 24.5% and 6.9% compared to G₁₅F₈₀ and G_22.5F₈₀, respectively, but no significant difference was observed compared to G₃₀F₈₀. Partial least squares path modeling analysis revealed that incorporation rates of grass pea had positive (P<0.05) effects on soil organic matter, total nitrogen, available nutrients, and microbial biomass carbon and nitrogen contents, but had negative (P<0.05) effect on soil bulk density. These factors primarily influenced rice yield by regulating soil organic matter content.

Conclusions In the single-cropping rice region of the hilly areas of Sichuan, continuously reducing 20%−40% N fertilizer in rice season combined with incorporation of higher rates of grass pea green manure can enhance rice yield and soil properties. Overall, an incorporation rate of 30 t/hm² of grass pea is recommended under 20% N fertilizer reduction, while an incorporation rate of 37.5 t/hm² is recommended under 40% N fertilizer reduction.