城镇化带来的耕地快速非农化背景下耕地利用效率提高对确保粮食主产区粮食安全意义重大。运用Data Envelopment Analysis(DEA)模型、Malmquist指数和Geographically Weighted Regression(GWR)模型,从县域尺度分析了2000—2015年河北省耕地利用效率的时空特征、动态规律及影响因素。结果表明:从时间上看,2000年以来河北省耕地利用综合效率、纯技术效率和规模效率均处于0.8~1,保持较高水平,其中耕地利用综合效率先由0.9减少到0.8,2003年后波动上升,在2015年其值达到1。从空间上看,区域间耕地利用综合效率空间变化明显,区域同质性增强。从动态变化看,河北省耕地利用全要素生产率指数由1.03波动上升到1.18,其中耕地利用综合提高型主要分布在省域北部,技术进步提高型分布于冀中南部地区,技术变化提高型仅存在于衡水市辖区,综合下降型零散分布在中南部区域。从影响因素看,在一定概率水平下,河北省耕地利用综合效率与农业产值占比和复种指数呈正相关,与高程、坡度、农业人口占比以及非粮占比呈负相关。其中高程、坡度、复种指数和农业人口占比影响较小,农业产值占比和非粮占比影响较为明显,非粮占比影响最为显著,其相关系数为-0.520249。最后从耕地利用影响因素出发依据不同区域情况对如何提高耕地利用效率提出优化建议。

The improvement of arable land use efficiency is of great significance to ensure food security in the main grain production areas, given the rapid non-agriculturalization of arable land brought on by urbanization. This study used the DEA model, the Malmquist index, and the GWR model to analyze the temporal and spatial characteristics, dynamics, and influencing factors of the cultivated land use efficiency in Hebei Province, China from 2000 to 2015 at the county scale. The results showed the comprehensive efficiency, pure technical efficiency, and scale efficiency of arable land use was maintained at a relatively high level of 0.8-1. The comprehensive efficiency of arable land use first decreased from 0.9 to 0.8 by 2003, then had a volatility increase to reach 1 in 2015. From a spatial point of view, the overall efficiency of cultivated land utilization between regions changed significantly, and the regional homogeneity increased. In regards to dynamic changes, the total factor productivity index of arable land use fluctuated from 1.03 to 1.18. The comprehensive improvement of arable land use was mainly distributed in the northern part of the province, whereas the improvement of technological progress was distributed in central and southern Hebei, and the improvement of technological change only existed in the Hengshui City district, and the comprehensive descending type was scattered in the central and southern regions. From the perspective of influencing factors, the comprehensive efficiency of arable land under a certain probability level was positively correlated with the proportion of agricultural output value and the multiple crop index, and was negatively correlated with the elevation, slope, proportion of agricultural population, and proportion of non-grain crops. Among them, elevation, slope, the multiple cropping index, and the proportion of agricultural population had little influence. The proportion of agricultural output value and the proportion of non-grain were more obvious, with the proportion of non-grain having the most significant influence with a correlation coefficient of -0.520249. Based on these results, suggestions are given to improve the arable land use efficiency.