Legume-grass mixtures are often used to increase forage yield and soil fertility in managed grasslands, but it remains unclear whether these benefits could be further improved by utilizing optimal legume-grass seeding ratios and phosphorus (P) fertilization. Here, we conducted a 5-year field experiment across 3 sites to investigate the effects of legume-grass seeding ratio and P fertilization on forage yield and soil quality. This experiment included mixtures comprised of two legume species and two grass species at five legume-grass ratios (3:7, 4:6, 5:5, 6:4, and 7:3), and monoculture of each species. P fertilizer was applied at the rate of 0, 9, 18, or 27 kg P ha^-1 year^-1. Plant diversity effects, e.g., complementarity and selection effects, were assessed by comparing yield of mixtures with monocultures. Our results show that a legume-grass mixture with a seeding ratio of 3:7 under moderate P fertilization resulted in the highest forage yield among all monocultures and mixtures, for each of the three sites, with persistent and consistent transgressive overyielding. Notably, greater soil organic matter, total nitrogen, enzymatic activity, and microbial biomass were observed with the legume-grass ratio of 3:7, compared to monocultures or other seeding ratios. Structural equation modeling indicated that the legume-grass ratio of 3:7 achieved highest yield through directly improving complementarity effect, and indirectly promoting selection effects because of increased fungal biomass. P fertilization directly enhanced soil nutrient and enzymatic activities, and further improved complementarity effect, resulting in high forage yield and soil quality. These results indicate that forage diversification practices with low legume-grass seeding ratios and moderate P fertilization can simultaneously benefit forage production and soil quality in managed grasslands. Overall, our study suggests that low legume seeding proportion in legume-grass mixtures combined with moderate nutrient management is a useful strategy for sustainable and highly productive managed grasslands.

豆科牧草混合物通常用于提高管理草地的饲料产量和土壤肥力，但目前尚不清楚是否可以通过利用最佳豆科牧草播种比例和磷（P）施肥来进一步提高这些效益。在这里，我们在3个地点进行了为期5年的田间试验，研究豆科牧草播种比例和磷肥对牧草产量和土壤质量的影响。该实验包括由两种豆科植物和两种禾本科植物按五种豆类-草类比例（3:7、4:6、5:5、6:4 和 7:3）组成的混合物，以及每个物种的单一栽培。磷肥施用量为0、9、18或27公斤P·公顷 ^-1 年 ^-1 。通过比较混合物与单一栽培的产量来评估植物多样性效应，例如互补性和选择效应。我们的结果表明，在适度施磷的情况下，播种比例为 3:7 的豆科草混合物在所有单一栽培和混合物中产生了最高的草料产量，对于三个地点中的每一个来说，都具有持续且一致的海侵超产。值得注意的是，与单一栽培或其他播种比例相比，豆科植物与草的比例为 3:7 时，土壤有机质、总氮、酶活性和微生物生物量更高。结构方程模型表明，豆草比为3:7时，通过直接提高互补效应，并因真菌生物量增加而间接促进选择效应，从而获得最高产量。施磷肥直接增强土壤养分和酶活性，进一步提高互补效应，实现牧草产量高、土壤质量好。 这些结果表明，低豆科牧草播种率和适度施磷的牧草多样化实践可以同时有利于管理草地的牧草产量和土壤质量。总体而言，我们的研究表明，豆类-草混合物中低豆类播种比例与适度的养分管理相结合是可持续和高产管理草原的有用策略。