Objectives  We compared the effects of different green manure production and their turnover on soil fertility and maize yield to propose efficient and suitable green manure management for sustainable and high maize yield in the Yunnan Plateau.

  Methods  A site-specific field experiment was conducted on maize using the green manure produced in red soil in the Yunan Plateau from 2011 to 2019. The experiment had five treatments, namely winter fallow crop land + maize (CK), planting radish and ploughing in its fresh grass + maize (R), planting vetch (Vicia villosa Roth. var.) and ploughing in its fresh grass + maize (V), ploughing in external sourced vetch fresh grass + maize (EV), planting vetch and reaping its stems and leaves away + maize (RV). Soil samples were collected at 0–20 cm depth to determine physiochemical properties after harvesting maize. Also, maize yield was measured every year.

  Results  Compared with CK, the soil bulk density reduced by 7.20%–10.46%, and soil organic matter content increased by 3.92%–14.03% under green manure production and utilization. We observed increased variation in soil organic matter across the treatments until the fourth year. The highest soil organic matter was recorded in treatment V. The alkali-hydrolyzed nitrogen in red soil increased by 24.21%, 14.66%, and 13.76% in treatment V, EV, and RV; there was no significant change in R. Available phosphorus was higher in EV, reduced by 8.23% in R, and no change was observed in the other treatments. Available potassium increased by 55.31% and 40.18% in EV and V but did not change in other treatments. Compared with the initial nutrient contents in red soil, the alkali-hydrolyzed nitrogen content decreased in the first five years but increased in the sixth and seventh years of the green manure-maize system. Available phosphorus and potassium decreased as a result of maize harvest.

  Conclusions  Continuously returning green manure crops to the soil could increase organic matter content and decrease bulk density of red soil. This practice affects maize yield differently due to changes in the soil available nutrient concentrations. Returning vetch fresh grass (EV) from an external source could ensure high maize yield as it increased soil available N, P, and K. Planting and returning vetch could increase soil available N and K content but consumes available P; radish could increase soil available K but consumes available P. Planting vetch and harvesting the aboveground biomass could result in loss of available K.