Objectives To provide a theoretical basis for the promotion of efficient green manure and rice straw returning technology, the effects of different returning methods of green manure and rice straw on rice yield and soil fertility were studied.

Methods Field experiments were conducted consecutively for 6 years in typical red soil area of southern China from 2013 to 2018. There were six treatments: neither green manure planting nor rice straw returning (CK), green manure planting without rice straw returning (MV), all early rice straw returning without green manure planting (RS1), all early and late rice straw returning without green manure planting (RS2), green manure planting with all early rice straw returning (RS3), green manure planting with all early rice straw and high stubble of late rice straw returning (RS4). All treatments wereapplied with the same amount of chemical fertilizer.The rice yields were recorded, and the coefficient of variation and sustainable index were calculated. After late rice harvest in 2017, soil samples were collected in 0–20 cm layer, the nutrient and water-stable aggregate contents were determined.

Results Compared with CK, the average yields of early, late rice and their sum in all treatments, except for RS1,were increased significantly by 6.6%–24.6%, 11.9%–28.8% and 9.0%–26.5%, with the highest in RS4, in which the average yield of late rice and the sum were significantly higher than those in other treatments. The coefficients of variation of early rice yield in RS1 and late rice yield in MV were increased, while those of both early and late rice decreased in the others, compared to the CK, with the lowest in both early and late rice in RS3. Compared with CK, the sustainable yield index of late rice was decreased in MV treatment while those of both early and late rice remained unchanged or increased in all the others, with the highest in RS3 treatment. After 5 years, the soil organic matter contents in all the treatments were increased than that at initial, and except in RS1, the organic matter contents in all the green manure and straw treatments were significantly higher than that in CK, and that in RS4 was significantly higher than in RS1. In spite of that the soil total N contents of all the treatments were reduced, compared to the initial, those in green manure and straw treatments decreased less than in CK, and those in MV and RS3 were significantly higher than in CK. The soil readily available K contents were all higher than in initial, except that in MV and RS3. The water-stable aggregate of size > 0.25 mm was the highest in RS2, up to 68.9%.

Conclusions Planting green manure with with all early rice straw and high stubble of late rice straw returning is a relatively good tillage system for high and stable yield of rice in red soil area of southern Hunan, where the long-term effect of increasing yield, soil organic matter content and stabilizing soil nitrogen supply is remarkable.