为探究自然因素和人类活动对水域景观格局演变产生的影响，以水域空间演变为主线，通过空间计算、景观 指数分析和统计分析等，对梁子湖流域 2000—2020 年水域转移状况、空间自相关特征、景观格局变化以及驱动 因子进行研究。结果表明：2000—2020 年梁子湖流域水域面积减少约 21%，减少面积中约 95% 转为耕地，建筑用 地面积增长超 4 倍，约 90% 来自耕地的转入；近 10 年水域空间聚集面积减少约 23%，水域分布关联性正在减弱； 水域景观特征在流域上下游差异显著，上游水域面积和形状基本稳定，聚集性和连通性均优于下游，下游水域面 积萎缩近一半，阻隔作用明显，岸线趋于规整；降水、气温、GDP、人口以及耕地和建筑用地面积比例解释了水域 景观变化的 54.7%，城镇化进程和农业生产活动是主要影响因子，其中人口增长和耕地开发对水域面积萎缩贡献 更大，GDP 增长和城镇化进程对水域边界复杂度和连通性的影响更明显。研究结果可为梁子湖流域水域景观保 护修复提供参考。

Landscape pattern and its changes are of great importance to ecological processes and security in basin or region. Liangzi Lake is a typical well protected grassy shallow lake in China. It plays important roles of flood storage, water supply, gene pool, freshwater products supply and so on. However, due to continuous reclamations for farmlands and urbanization, problems related to water area shrinkage, ecological functional degeneration and landscape fragmentation are raising continuously. Exploring the dynamic changes of water landscape and its driving factors from the perspective of landscape pattern in whole basin would obtain in-deep understanding of landscape characteristics and provide scientific references of Liangzi Lake protection from basin-wide perspective.The methods of spatial analysis, landscape metrics analysis and biostatistical analysis are applied in this paper. Taking water landscape as the research object, transition matrix analysis is used to describe transitions between water area and other types of land use during study phase. Aggregation feature and outlier distribution of water area are studied through spatial autocorrelation analysis. Landscape metrics analysis is used to express the features of area, shape, aggregation and connectivity quantitively, while redundancy analysis is employed to seek for the driving forces of dynamics changes.From 2000 to 2020, farmland area remained unchanged basically, while water area showed 21% decrease because of water coverage being reclaimed for use as farmlands. Urban area increased by more than 4 times by means of farmlands transferred into. Water area clustered spatially with a weakening trend. The distribution of High value aggregation and Low value aggregation in water area got more fragmented, and the area of Not Significant type increased by 40% in last decade. PLAND and LPI of water area had trend of increasing followed by trend of decreasing. PD dropped gradually with an obvious decline of LSI. There were significant differences between upper and lower basin - PLAND and LPI of water area in upper area were about 16% and 14% respectively, while PLAND and LPI in lower area decreased by more than 40% and 60% respectively. Shape of water area in lower area was more complicated than in upper area, and LSI of upper and lower area declined by 32% and 51% respectively. Aggregation and connectivity in upper area were better than those in lower area. Variation partitioning of redundancy analysis shows the 54.7% changes of water landscape are a result of natural and artificial factors, of these influences, 24.2% of the changes coming from precipitation and temperature, while 51.6% coming from contribution of human activities, the overlapping part is 21.0%.From 2000 to 2020, surface area of water in Liangzi Lake has decreased by 21% because of agricultural reclamation directly, which then being reclaimed for urbanized purposes, leading to 4-time area increases of artificial surfaces. Farmland area remains unchanged, water area decreased attributed to urbanized purposes indirectly. There is a steady water area aggregation in whole basin, while the relevance of water area in local region are becoming decreased or disappeared. Micro-water body are disappearing with the tendency of decreasing of complexity and irregularity. Water area distribution tends to be fragmented, however significant differences present between upper and lower basin areas. There are not much changes of spatial distribution of water landscape in upper basin area, but in lower basin area, water area reduced by nearly half. Shoreline tends to be normalized and unified with an obvious obstruction and cut of water body. 54.7% of the changes of water landscape are the result of precipitation, temperature, population, GDP and area ratio of farmland and artificial surface. Human activities are the main factors, whose explanatory power are twice of natural factors. Population and farmland area rate make more contributions to the water landscape area, meanwhile GDP and urbanization process have more effects on shape complexity and connectivity.