中文摘要：

植被演替过程是有序进行还是随机发生，这不仅是植物生态学界一直关注的热点之一，也是植被恢复工作的一个关键问题．川西岷山地区频繁、强烈的地质构造活动常常引发滑坡，使得这些发生于不同年代的滑坡体坡面上生长着处于不同演替年限的植物群落，从而为验证植被演替模型理论提供了理想场所．选取海拔、坡度、坡向、土壤母质大致相同的17个坡面，对其植被进行群落调查，并将其划分为3个演替年限组——演替4年、演替30—60年、演替60年及以上．研究发现：（1）随着演替推进，群落的物种丰富度、均匀度指数、Shannon—Wiener指数、物种优势度均呈上升趋势，表明群落发展过程中物种显著增多；（2）从演替4年到30年这一演替早期阶段，群落明显存在着物种替代现象，但主要功能群（固氮、耐旱、耐贫瘠植物）的重要值在演替4年到演替60多年过程中并未有所下降，表明当地的环境资源条件（水分和养分）通过决定功能群构成强烈制约着坡面群落的物种集合；（3）在演替4年的群落组内，群落物种相似性系数为0．515，而在演替60多年的群落组内该值降为0．251，表明植被演替较晚阶段存在着随机过程．因此，机体论演替模型不足以完全解释当地植被演替过程．图5表3参30

英文摘要：

The deterministic or stochastic nature of vegetation succession has been a bone of contention in plant ecology, and at the same time a key to vegetation restoration. On the slopes of the Minshan Mountain in Western China, frequent and intense geological tectonic activities cause plant communities with various succession ages array closely together, providing an ideal platform to test vegetation succession models. The plant communities of 17 nearby slope sites with similar elevation, gradient, aspect and parent soil material were selected for plant community investigation. The communities were categorized into three succession age groups： early stage （4 years）, 30-60 years, and over 60 years. We found that （1） the species richness, species evenness, Shannon-Wiener index and species dominant index all increased with the succession, indicating an obvious species adding process in the community development; （2） the species substitutions were significant in early succession （early stage to 30 years）, but the importance values of major functional groups （nitrogen fixing species, dry-tolerant species and leanness- resistant species） showed no evidence of decrease from year 4 to year 60 and over, suggesting that physical conditions （water, nutrition） have strongly controlled the species assemblage through the functional group compositions on the sites; （3） the within-group community similarities decreased from over 0.515 in year 4 to 0,251 in year 60 and over, illustrating a tendency of stochastic direction of vegetation succession in the later stage. We concluded that the organismic model does not satisfactorily explain the succession process.