1990—2020年淮河生态经济带耕地资源分布变化特征与驱动机制

探索淮河生态经济带耕地资源分布变化特征，可为保护耕地、保障国家粮食安全提供规划和决策依据。基于1990—2020年遥感影像数据，利用空间探索、统计分析等方法，揭示该区域耕地资源在过去30年的时空分布变化及其驱动机制，并进一步对2030年土地利用进行情景预测。结果表明：1）淮河生态经济带土地利用结构以耕地为主，平均占比为68.70%；近30年来研究区耕地面积剧烈缩减8.63×10⁵ hm²，年均减少2.88×10⁴ hm²，其中旱地减少占耕地总减少量的82%；耕地的流失主要在于建设用地的侵占。2）耕地资源具有明显的集聚特征，以淮河为分界线集聚分布，呈现“东南水田、西北旱地”的典型分布特征。3）耕地资源分布变化的主导驱动因子包括农业机械化水平、粮食产量、人口等社会驱动力；生态驱动力各要素对耕地资源变化的驱动力则相对较为稳定。4）自然发展情景下，2030年耕地面积持续减少1.10×10⁵ hm²，在积极的耕地保护情景下，耕地面积将显著增加5.29×10⁵ hm²；采取耕地保护情景，鲁中南低山丘陵区以及桐柏—伏牛山通过开垦部分疏林地和低覆盖草地将其整治为耕地，东部沿海可将沿海水域合理建设为建设用地，中部区域可通过减少建设用地的碎片化，较大程度上使耕地资源的连片化，从而在一定程度上提高耕地的利用效率。研究对于优化淮河生态经济带耕地资源结构以及空间优化配置具有重要意义。

Spatiotemporal variations and impacts on cultivated land resources in Huaihe Eco-Economic Belt during 1990-2020

Cultivated land is the foundational resource for grain production, particularly for the food security, ecological safety, as well as the economic and social stability. Therefore, it is crucial to explore the spatiotemporal distribution and driving mechanisms of cultivated land in the major grain-producing regions. Taking the Huaihe Eco-Economic Belt (HHEEB) as this research subject, this study aims to investigate the spatiotemporal distribution and driving mechanisms of cultivated land resources from 1990 to 2020. Remote sensing imagery data was also collected over the past 30 years. Firstly, four periods were divided for the remote sensing image data of land use in 1990, 2000, 2010, and 2020. Arcgis software was used to reclassify 25 second-level into the six first-level land categories. Spatial Analyst function was utilized to statistically analyze the quantity of cultivated land resources at varying scales, and then further analyze the temporal changes and transfer paths of cultivated land resources; Secondly, GeoDa spatial analysis software was also combined to clarify the spatial agglomeration of cultivated land resources; The land use data was overlaid for 1990 and 2020, according to the spatial variation in the cultivated land area at the urban scale. Further exploration was conducted on the spatial changes of farmland resources. 14 driving factors were selected from four aspects: natural driving force, environmental tolerance, agricultural productivity, and economic development. The driving mechanism of cultivated land resources was explored for the factor and interactive detection in the geographical detector; Finally, the FLUS model was used to optimize the allocation of cultivated land resources under both the natural development and protection scenario of cultivated land in 2030. The results show that: 1) The land use structure of the HHEEB was predominantly characterized by the cultivated land, with an average of 68.70% of the total area. Notably, the proportion of dryland was consistently exceeded that of paddy fields, with 71.81% of the annual average cultivated land area. There was the significant reduction in the cultivated land area over the past three decades, with a staggering decrease of 8.63×10⁵hm² and an average annual reduction of 2.88×10⁴hm². This decline was largely attributable to the loss of dryland, which was constituted 82% of the overall reduction in the cultivated land. The encroachment of construction land was identified as the primary source of cultivated land loss in the region. 2) The spatial exploration was also performed on the agglomeration of farmland resources. Distinct features were observed with the aggregation distribution divided by the Huaihe River, indicating the typical distribution pattern of southeast paddy fields and northwest dryland. 3) The factor detection showed that the leading driving factors of cultivated land resource distribution included the social driving forces, such as agricultural mechanization level, grain yield, and POP. There was the relatively stable driving force in each element of ecological driving force on the cultivated land resources. The interaction detection show that the pairwise interactions among factors in the agricultural productivity were generally greater than those in the ecological driving forces and the interactions within agricultural productivity. Agricultural productivity was consistently played a dominant role in the spatiotemporal changes of farmland resources. 4) The cultivated land area was projected to decrease by 1.10×10⁵ hm², if the natural development scenario was remained by 2030. However, the scenario cultivated land protection can be expected to offer the promising outlook with the significant increase of 5.29×10⁵ hm² in the cultivated land area. It is envisaged on the reclaimed cultivation in the hilly and mountainous areas, the parts of sparse forest and low-coverage grassland in the Tongbai-Funiu Mountains under the proactive protection scenario. Some suggestions can also propose to consider the potential development of coastal waters along the eastern coast for the construction purposes. In addition, it can be expected to reduce the fragmentation of construction land for the high utilization efficiency of agricultural land in the central region. The spatiotemporal variations in the arable land resources can provide a strong reference for the decision-making on the national food security.