Abstract: Based on data sources (e.g., SPOT-5 images, 1:1 million topographic maps, distribution maps of returning farmland to forest and Chongqing forest project, social and economic statistics information, etc.) this paper identified the characteristics and influencing factors of farmland marginalization. The results indicated that: (1) During the last 10 years (2002-2012), the rate of farmland marginalization was 16.18% in the study area. This phenomenon was mainly found in high mountainous areas of northern Qiyao mountain and middle mountain areas of southern Qiyao mountain. Moreover, farmland marginalization, to a great extent, will increase, associated with non-agriculturalization of rural labors and aging of remaining labors. (2) Elevation, relative distribution radius from village and road connection degree had a greater influence on farmland marginalization in the study area. Farmland marginalization rate showed an increasing trend with the increase of elevation, and 60.88% of the total farmland marginalization area is found at an altitude above 1000 m asl. The marginalization trend for slope and distribution radium was similar with altitude. (3) Farmland area per labor and average age of farm labor were major driving factors of farmland marginalization. Farmland transfer and small agricultural machinery sets affected the farmland marginalization in the aspects of management and productivity efficiency. (4) Farmland with "comparative disadvantage dominated marginalization" accounted for 55.32% of the total farmland marginalization area, followed by "location dominated marginalization"(33.80% ). (5) According to the specific real situation, different policies are suggested to mitigate the marginalization as follows: "continuous marginalization" policy will further exert the impact of returning farmland to forest on "terrain dominated marginalization"; "anti-marginalization" policy is suggested from arranging new rural residence and improving rural road system for "location dominated marginalization"; while "anti-marginalization" policy is planned from improving management and micro-mechanization point of view for "comparative disadvantage dominated marginalization". A new idea was promoted by integrating high resolution remote sensing and statistical data as well as survey information to identify land marginalization and driving forces in mountainous areas, and hence solved the scale problem on patches and villages.

Key words: farmland marginalization, characteristics, policy implications, influencing factors, mountainous areas