金沙江是长江上游，流域内地形复杂，动植物资源丰富，生态安全作用突出，其中干热河谷区降水少而蒸发强烈，生境恶劣，植被稀疏，是典型的生态脆弱区。为增加金沙江流域植被盖度，保障生态安全，国家先后实施了多个植被恢复工程，整体上改善了当地的生态环境，但局部地区生态修复效果不理想。特别是近年来，随着金沙江流域社会经济发展和梯级水电开发，人类活动强度逐渐加大，深刻理解多重人类活动对植被演替的影响及作用机制成为迫切需求。 本研究选择龙开口库区为典型研究区，通过野外调查采样、遥感数据解译、模拟模型构建等多种方法与手段，以坡柳和云南松为研究对象，探讨了龙开口库区典型植被与环境因子的关系，以识别的关键环境因子土壤水分为驱动，构建了基于系统动力学的植被个体生长模型和包含植物关键生态过程的植被空间演替模型，模拟了库区典型植被生长演替过程，揭示了多重人类活动影响下龙开口库区植被的演替规律及空间格局。研究成果为在西南干热河谷区进行植被生态修复提供了科学支持，研究得出以下主要结论： 2003-2018年，龙开口库区土地利用类型存在由天然牧草地向灌木林地（以坡柳为主）和乔木林地（以云南松为主）转变的趋势，随着水电站建成并蓄水，区域内水域面积明显增大，降雨量有所增加。 随着研究区海拔上升，坡柳生物量在阴坡先增加后降低，在阳坡持续降低；云南松生物量在低海拔（< 1940 m）区域的阴坡较高，在高海拔（≥1940 m）区域的阳坡较高。研究区雨季降水量和土壤水分明显高于旱季，海拔引起的旱季末期土壤水分的变化对干热河谷植被分布起决定作用，保证土壤水分充足是植被恢复的关键。 以土壤水分为驱动力，利用植物的相对生长曲线及种间竞争模型，构建了基于系统动力学的植被个体生长模型，模拟了不同种间竞争模式下坡柳和云南松生物量在土壤水分梯度上的变化趋势。种间竞争作用降低了坡柳和云南松的生物量；种间竞争不对称时，坡柳生物量最大值对应的土壤水分由0.062降低至0.054，云南松的由0.102增加至0.105。土壤水分增加，坡柳生物量减小，云南松生物量先增加后减小。 考虑植物生长、结实、种子传播、发芽和成苗的关键生态过程，构建了典型植被空间演替模型。模型稳定时，研究区植被呈带状分布，坡柳和云南松占优势的面积分别为54.34km2和70.75km2，共占总面积的89%。研究区植被更新扩散能力较强，8个步长后，坡柳和云南松分别占据了适宜空地的83%和94%。 模拟了典型人类活动影响下龙开口库区植被的演替过程和空间格局变化。结果显示，采伐和放牧等人为干扰下库区植被总面积减少了0.04 km2，坡柳占优势的面积增加了6.18 km2，但研究区植被的空间格局变化不大。播种使研究区植被扩散能力增强，坡柳和云南松植物群落稳定时间分别缩短了45%和44%。多重人类活动下，坡柳和云南松的总面积略有增加，特定条件下，人类活动对植物生物量的影响会相互抵消。 本研究以坡柳和云南松为对象，构建了植被个体生长模型和空间演替模型，模拟了龙开口库区典型植被的生长演替过程，探讨了多重人类活动对植物群落演替和空间格局的影响，有助于深入理解干热河谷典型植物群落的演替规律，及其对多重人类活动的具体响应，对金沙江及西南地区有针对性的开展植被恢复措施，加强生态安全建设有重要意义。

The Jinsha River is the upper reaches of the Yangtze river, with complex terrain and abundant plants and animals, which plays an important role in the ecological security of the southwest of china. The Arid-hot Valley of the Jinsha River has less precipitation and strong evaporation and difficult in vegetation growing and is a typical ecologically fragile area. In the Jinsha River basin, aim to increase vegetation coverage and ensure ecological security, the state has implemented a variety of vegetation restoration projects, which has improved the ecological environment as a whole, but the effect of ecological restoration is still unsatisfactory in some areas. Especially in recent years, with the development of cascade hydropower and social economy in Jinsha River basin, it is increasing that the strength of human activities on the local environment, making it more necessary about understand the impact of multiple human activities on vegetation urgent, especially of the degree and mechanism. In this study, we selected the Longkaikou reservoir as a typical study area. Through the methods of surveying and sampling, remote sensing data interpretation and simulation model construction, we analysed the relationship between environmental factors and the typical vegetation, Dodonaea viscosa and Pinus yunnanensis, and identified soil moisture as the key environmental factors. Soil moisture was used to build an individual plant growth model based on system dynamics and a spatial model of vegetation succession including key ecological processes of plant. We simulated the processes of plant community succession to reveal the succession law and spatial pattern of vegetation affected by multiple human activities in the Longkaikou reservoir. The study will provide a scientific basis for the ecological restoration in the Arid-hot Valley of southwestern China, which draws the following main conclusions: During 2003-2018 years, due to the completion of hydropower construction and reservoir impoundment, rainfall and the area of the water body were increased and the land use showed a trend of transformation from natural grass land to shrub land (mainly by D. viscosa) and forest land (mainly by P. yunnanensis) in the study area. In the study area, with the elevation increasing, the D. viscosa biomass increased first and then decreased on shade slope and decreased continuously on sunny slope. For P. yunnanensis, the biomass was higher on the shady slope at the low elevation (<1940 m). but it was higher on sunny slope at high elevation (≥1940 m). The precipitation and soil moisture in the rainy season were significantly higher than that in the dry season. The change of soil moisture caused by elevation at the end of the dry season had a decisive effect on the vegetation distribution in the arid-hot valley, and rich soil moisture is the key to vegetation restoration. We used the relative growth curve and interspecific competition of different plants, an individual plant growth model was established which was driven by soil moisture and based on system dynamics, and the vegetation biomass changes were simulated under different interspecific competition. Comparing with the no interspecific competition, the interspecific competition reduced the biomass of D. viscosa and P. yunnanensis. When the competition was asymmetric, the soil moisture for the maximum biomass of D. viscosa decreased from 0.062 to 0.054, but it increased from 0.102 to 0.105 for P. yunnanensis. With the increasing of soil moisture, the biomass of D. viscosa decreased and it increased first and decreased then for P. yunnanensis. Considering the key ecological processes, like plant growth, fruiting, seed dispersal and germination, a typical spatial model of vegetation succession was built. When the model was stable, dominant D. viscosa and P. yunnanensis had zonal distribution and their area respectively were 54.34km2 and 70.75km2, which accounted for 89% of the total area. Vegetation in the study area has a strong ability to renew and diffuse. After 8 steps, D. viscosa and P. yunnanensis accounted for 83% and 94% of the suitable open space, respectively. we simulated the succession process and spatial pattern of vegetation under the influence of typical human activities in the Longkaikou reservoir. The results shown that vegetation total area decreased by 0.04 km2 due to logging and grazing, the area of dominate D. viscosa increased by 6.18 km2, however, the spatial distribution of vegetation has not changed much. Sowing increased the ability of vegetation dispersal in the study area, and the time to stabilize the plant communities of D. viscosa and P. yunnanensis was respectively shortened by 45% and 44%. Under multiple human activities, the area of D. viscosa and P. yunnanensis increased by 0.11 km2 and 0.04 km2, respectively, and in particular situation, the effects of human activities on vegetation were mutual offset. In this study, taking D. viscosa and P. yunnanensis as study objects, we planted an individual vegetation growth model and a spatial succession model to simulate the succession process of vegetation and explore the effects of multiple human activities on plant community succession and spatial pattern in the Longkaikou reservoir. The study is helpful to understand the succession rule of typical plant communities and their responses to multiple human activities, which is great significance to carry out vegetation restoration measures and protect ecological security in the Jinsha River basin and the southwest of China.