Abstract: In response to the degradation and grassland quality improvement requirements in Southern China,a large number of seeds of forage species have been introduced into the process of grassland restoration and reconstruction,forming a mixed plant communities together with native species. Little was known about the response of native communities to different seed sowing and tillage before sowing. Taking the degraded grassland in Zhaotong,Yunnan as the object,three treatments:sowing,tillage,tillage+sowing were designed and implemented,a quadratic saturated D-optimal design model was applied to design the weight and ratio combinations of the two grass and one bean species. Thus,14 treatments with different species sowing combination were arranged to detect the response of the local community to different combinations of sowing species. The results showed that tillage significantly reduced the diversity coefficient and changed the composition of native communities,in which annual species predominated. Sowing forage seeds with different quantity or ratio after tillage,the composition of native community did not change significantly,as well the abundance,aboveground biomass and four diversity indexes;but the internal community connectivity tended to be negative. Sowing seeds without tillage reduced the abundance and aboveground biomass of native plant communities,but the difference was not significant. The above results could provide a theoretical basis for the restoration and quality improvement of grassland in Southern China.

Key words: Seed sowing, Tillage, Native plant community

摘要： 为应对南方草地的退化与质量提升要求，大量的饲草物种被引入南方草地的退化改良与恢复中，与本地物种共同构成新的植物群落，但本地植物群落对饲草补播及其前期耕耙的响应尚不明确。本研究以云南昭通退化草地为对象，设置免耕补播、耕耙(不补播)、耕耙+补播三类处理，并利用二次饱和D-最优设计模型，在耕耙补播处理中布设14个处理的补播量梯度与配比试验。结果表明，耕耙显著降低群落多样性系数，改变了本土群落的组成，一年生物种占据优势;在耕耙后补播不同饲草组合，这种组成关系也未显著改变，且本地植物群落的多度、生物量和各群落多样性指数均未发生显著变化，但内部的物种关系发生了变化，群落联结性趋向负联结;免耕补播降低本土植物群落的多度和生物量，但差异不显著。上述结果可为昭通南方草地的退化恢复与质量提升提供理论依据。

关键词: 补播, 耕耙, 本地物种群落