四川凉山州孙水河流域耕地撂荒时空格局分析

Spatiotemporal Pattern of Abandonment of Arable Land in Sunshuihe River Basin of Liangshan Prefecture in Sichuan Province

[目的]提取长期、精确的撂荒地空间分布信息,摸清流域耕地撂荒分布规律,进而为大凉山区域土地资源管理、乡村振兴战略规划等提供数据支撑。[方法]以凉山州孙水河流域为研究区域,应用CART决策树分类方法提取了研究区2009—2016年的撂荒地分布数据,分析了近10年来孙水河流域撂荒地时空格局、演变动态与影响因素。[结果](1)2009—2016年流域耕地撂荒情况呈现两个阶段:2009—2011年缓慢降低,2012—2016年快速增加。撂荒地多分布在中高山区域,并向河谷平原地区逐渐聚拢。多年平均撂荒面积为5357.88 hm^(2),2016年达到最大撂荒率(24.50%),年均撂荒面积:中游(2260.50 hm^(2))>上游(1716 hm^(2))>下游(1381.37 hm^(2)),年均撂荒率:上游(19.17%)>中游(15.46%)>下游(8.50%)。(2)撂荒地标准差椭圆分布呈现两个发展阶段,2009—2012年表现为“东南—西北”的方向分布,2013年后重心回移,撂荒更为聚集。(3)撂荒地多分布在海拔>2800 m,坡度15°~25°,距水系500 m内的地形单元。[结论]地形条件、生产水平、政治因素等均会导致区域撂荒时空格局变化,未来应加强辨析人类活动及自然条件对撂荒地影响机制的深入研究。

[Objective]The aims of this study are to extract long-term and precise monitoring of abandoned farmlands,to understand the spatiotemporal distribution patterns of abandoned farmlands in river basins,and to provide data support for the management of land resources and the strategic planning of rural revitalization in Liangshan Prefecture.[Methods]The Sunshuihe River Basin in Liangshan Prefecture was taken as the research area.The Classification and Regression Tree(CART)model was used to extract abandoned farmlands from 2009 to 2016,and the spatiotemporal pattern and evolution dynamics of abandoned farmland in the past eight years were analyzed using the standard deviation ellipse method.[Results](1)Farmland abandonment in the basin showed two changing trends during the study period:slow decrease from 2009 to 2011 and rapid increase from 2012 to 2016.Abandoned farmlands mostly distributed in the middle and high mountain areas and gradually decreased in the valley plain.The average abandoned area for the entire river basin was 5357.88 hm^(2),and the maximum abandonment rate(24.50%)was observed in 2016.The average annual abandoned area followed the order:middle reaches(2260.50 hm^(2))>upstream(1716.01 hm^(2))>downstream(1381.88 hm^(2)).The average annual abandonment rate increased with stream reaches:upstream(19.17%)>middle reaches(15.46%)>downstream(8.50%).(2)The standard deviation ellipse distribution of abandoned farmland showed two stages of development.From 2009 to 2012,it showed the distribution of‘southeast-northwest’.After 2013,the center of gravity moved back,and the abandoned land was more concentrated.(3)Along topographic and proximity factors,abandoned farmland mostly distributed with an altitude of>2800 m,and a slope gradient of 15°~25°,and 500 m away from the river system.[Conclusion]The spatiotemporal pattern of abandoned farmlands can be influenced by various factors such as topographical conditions,production levels,and political factors.In the future,it is important to enhance the in-depth study of the impact mechanisms of human activities and natural conditions on the land abandonment.