应用GIS技术和空间自相关分析方法,分析了1990年以来环渤海地区县域人均粮食占有量的空间演化特征。结果表明,研究区县域人均粮食占有量空间自相关特征明显,高-高聚集和低-低聚集趋势增强;人均粮食增长更多表现出空间分布的随机性,地理集中趋势不明显;县域人均粮食占有量以上升为主,增长类型以总产低、快速增长型为主,粮食生产重心逐步向平原区集中,辽中和鲁西地区的粮食生产功能日益凸显,生态脆弱区和快速城镇化区的人均粮食占有量明显下降。最后,通过空间计量经济模型SLM和SEM的回归分析,探讨了县域人均粮食占有量格局演化的主要驱动力。
The area along Bohai Rim, including Beijing, Tianjin, Liaoning, Hebei and Shandong, is one of the important grain production bases that guarantee food security in China. With the acceleration of area urbanization in the past 20 years, the grain production and consumption patterns along Bohai Rim have brought about significant changes. Agricultural statistics of the area along Bohai Rim at county - resolution level for the time period of 1990 -2008 is collected and the GIS technique as well as other tools such as the Moran's I and the Getis-Ord are introduced to describe the spatial changes of per capita grain possession at county level in this area. The conclusions are as the follows. First, per capita grain possession at county level in the area along Bohai Rim shows a significantly trend of positive spatial correlation and similar areas cluster in space. LISA cluster map demonstrates that counties with higher per capita grain possession gathered in Liaohe plain, the Yellow River floodplain in western Shandong Province and alluvial plain of Haihe river, while the lower per capita grain possession gathered in mountainous-hilly areas, tableland areas and densely populated city area. Second, the spatial framework of per capita grain possession growth is likely to be more stochastic and unstable in the aspect of spatial distribution. Hotspot areas are changing frequently without obvious appearance of geographical concentration. Third, Per capi- ta grain possession in most of the counties shows an upward trend, and lower production but faster growth is the main type. The grain production function has been improved significantly in plains and has been weakened as the implementation of Conversion of Cropland to Forest and Grassland Project in mountainous-hilly areas, and it has been shown a strong downward trend in cities and their surrounding counties as the sharp reduction of cultivated land and the swift growth of regional population. At last, the contributing factors for the variation of per capita grain possessi