中国不同类型耕地撂荒时空分异及其影响因素研究——基于全国346个山区县的调查数据  

Spatial and temporal differentiation and influencing factors of fallowing in different types of farmland in China:Based on survey data from 346 mountainous counties

作　　者：王亚辉 杨遨郗 辛良杰[4] 董世杰 李秀彬[4,5] WANG Yahui;YANG Aoxi;XIN Liangjie;DONG Shijie;LI Xiubin(School of Geographical Sciences,Southwest University,Chongqing 400715,China;New Liberal Arts Laboratory for Sustainable Development of Rural Western China,Chongqing 400715,China;Chongqing Jinfo Mountain Karst Ecosystem National Observation and Research Station,Chongqing 400715,China;Institute of Geographic Sciences and Natural Resources Research,CAS,Beijing 100101,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]西南大学地理科学学院,重庆400715 [2]西部乡村可持续发展新文科实验室,重庆400715 [3]重庆金佛山喀斯特生态系统野外科学观测研究站,重庆400715 [4]中国科学院地理科学与资源研究所,北京100101 [5]中国科学院大学,北京100049

出　　处：《地理研究》2024年第8期1977-1993,共17页Geographical Research

基　　金：国家自然科学基金项目(41930757、42271263、41901232);中央高校基本科研业务费项目(SWU-KT22008);重庆金佛山喀斯特生态系统国家野外科学观测研究站开放基金项目(JFS2023A01)。

摘　　要：系统性探寻耕地撂荒现象,有助于因地制宜地稳步拓展农业生产空间。基于全国346个代表性山区县的抽样调查,本文利用空间分析和分层线性模型等方法,系统揭示不同类型耕地撂荒的时序、规模、空间分异特征及其决定因素,以期为分类推进撂荒地治理提供科学依据。研究发现:不同类型耕地的撂荒时间节点存在显著差异,坡耕地、旱作梯田、水作梯田、平地水田和平地旱地依次撂荒,耕作条件较差的耕地更早遭遇撂荒,特别是坡耕地自20世纪80年代起就开始撂荒。目前,26.64%的农户经历过耕地撂荒,撂荒耕地的面积占比达到9.88%,呈现出“南高北低”的空间分布特征。水田撂荒主要集中在东南沿海,而坡耕地和旱地则多发于黄土高原、东南沿海和西南山区。撂荒情况受村落和农户特征影响,这两个层面分别对撂荒变异解释了44.83%和55.17%。从事非农工作、年长及身体状况较差的农户更可能撂荒耕地;同时,遭受野生动物侵害、土地细碎化严重且灌溉条件不佳的区域更易撂荒。劳动力成本的持续上升压缩了农业收益,使得农户种粮收益偏低,旱作梯田甚至面临亏损。因此,撂荒地的治理应聚焦坡耕地和梯田,政府应积极推动土地流转与整治,普及适合丘陵山区的小型机械,以降低耕地经营成本,提高农耕效益。Identifying and repurposing abandoned farmland is a crucial strategy for the sustainable expansion of agricultural production spaces.This study utilizes a representative sample of 346 mountainous counties in China to examine the patterns of abandonment,spatial differentiation,and underlying determinants of five distinct types of abandoned farmland:flat paddy field,flat dryland,sloping farmland,rice terrace,and dryland terrace.Through comprehensive spatial analysis,hierarchical linear modeling(HLM),and cost-benefit analysis,we systematically elucidate the sequence of abandonment and provide scientifically grounded recommendations for targeted management strategies.The findings reveal significant variability in abandonment initiation across different farmland types,with the sequence being sloping farmland,dryland terrace,rice terrace,flat paddy field,and flat dryland.Notably,the earliest abandonment occurred in 1980 for sloping farmland,with less productive land typically abandoned first.Currently,over 25%of farming households have ceased agricultural operations on their land,with the abandoned farmland area totaling 9.88%.This abandonment exhibits a geographical pattern characterized by higher rates in southern regions and lower in the north,predominantly affecting paddy fields in the southeastern coastal areas and sloping farmlands and drylands in the Loess Plateau and southwestern mountainous regions.The decision to abandon farmland is significantly influenced by both village and household characteristics,which together account for approximately 44.83%and 55.17%of the variance in abandonment rates,respectively.Factors such as long-term non-agricultural employment,elder household heads with poorer health,and land that suffer from wildlife damage,severe fragmentation,or poor irrigation conditions are more likely to be abandoned.Additionally,escalating labor costs have diminished agricultural profitability,particularly affecting the economic viability of sloping farmlands and terraces,with dryland terraces experien

关 键 词：耕地撂荒 时空分异 影响因素 分层线性模型 山区 

分 类 号：F323.211[经济管理—产业经济]