CAJ | 학술논문

为科学确定粮食主产区耕地产能大小，阐明各种因子对产能影响的作用强度及方向，该文以东北粮食主产区典型粮食大县巴彦县为研究区，采用潜力衰减法核算耕地产能大小，揭示其空间分异特征，运用遗传算法对神经网络模型进行修正和改进的因子降维法、通径分析和回归分析法，分析主要因子单一作用和因子间协同作用对耕地产能影响的作用强度及方向。结果表明：1）以玉米为基准作物，2009年研究区耕地理论单产、可实现单产和实际单产分别为2.33、1.58和1.08 kg/m2；理论产能、可实现产能和实际产能分别为547、372和254万t；理论利用潜力和可实现利用潜力分别为0.74和0.50 kg/m2；理论利用强度和可实现利用强度均为0.68；在空间分布上耕地理论单产、可实现单产及实际单产呈现出由西南向东北逐渐降低的分布特征，理论产能、可实现产能及实际产能呈现出由西北向东南逐渐降低的分布特征。2）差值植被指数、地貌类型、坡度、黑土层厚度、有机质、机械化程度、灌溉潜力7个主要因子的单一作用对耕地产能影响很大，其中机械化程度作用强度最大，黑土层厚度最小，除坡度对耕地产能有负向作用外，其他6个因子对研究区耕地产能有正向作用；主要因子间协同作用对研究区耕地产能的影响较大，其他主要因子与地貌类型的协同作用对耕地产能影响强度最大，方向为正；其他因子与主要因子形成的5组相关因子间协同作用对研究区耕地产能的影响同样也不容忽视，化肥施用量等因子与机械化程度协同作用对耕地产能的影响强度最大，方向为正。研究表明，主要因子的单一作用和因子间协同作用对研究区耕地产能具有重要影响。
위과학학정양식주산구경지산능대소，천명각충인자대산능영향적작용강도급방향，해문이동북양식주산구전형양식대현파언현위연구구，채용잠력쇠감법핵산경지산능대소，게시기공간분이특정，운용유전산법대신경망락모형진행수정화개진적인자강유법、통경분석화회귀분석법，분석주요인자단일작용화인자간협동작용대경지산능영향적작용강도급방향。결과표명：1）이옥미위기준작물，2009년연구구경지이론단산、가실현단산화실제단산분별위2.33、1.58화1.08 kg/m2；이론산능、가실현산능화실제산능분별위547、372화254만t；이론이용잠력화가실현이용잠력분별위0.74화0.50 kg/m2；이론이용강도화가실현이용강도균위0.68；재공간분포상경지이론단산、가실현단산급실제단산정현출유서남향동북축점강저적분포특정，이론산능、가실현산능급실제산능정현출유서북향동남축점강저적분포특정。2）차치식피지수、지모류형、파도、흑토층후도、유궤질、궤계화정도、관개잠력7개주요인자적단일작용대경지산능영향흔대，기중궤계화정도작용강도최대，흑토층후도최소，제파도대경지산능유부향작용외，기타6개인자대연구구경지산능유정향작용；주요인자간협동작용대연구구경지산능적영향교대，기타주요인자여지모류형적협동작용대경지산능영향강도최대，방향위정；기타인자여주요인자형성적5조상관인자간협동작용대연구구경지산능적영향동양야불용홀시，화비시용량등인자여궤계화정도협동작용대경지산능적영향강도최대，방향위정。연구표명，주요인자적단일작용화인자간협동작용대연구구경지산능구유중요영향。
The northeastern area is the production area of commercial grain, playing an important role in protecting national food security. To determine the size of cultivated land productivity and to clarify effect intensity and direction of various factors to the cultivated land productivity, the northeastern grain main production area Bayan county is used as the study area, digital topographic map and SPOT5 images from 2009 are used as the basic information source, and land detailed survey database and other statistics are used as the basic data source to analyze in the study. Based on the mechanism of formation of grain production capacity, synthesizing natural ecology, and socio-economic factors affecting cultivated land productivity, potential attenuation method is used to calculate cultivated land production capacity and analyze its spatial distribution pattern in the study area. While improved dimensionality reduction that backs propagation neural network is improved by genetic algorithm, path analysis and regression analysis methods are adopted to analyze single effects of each major factor and synergistic effects among factors to effect intensity and direction of cultivated land production capacity was also studied, which breaking the Agricultural Land Classification accounting methods, focused on natural factors in the past. The results show that:1) taking the corn crop as the basis in 2009, its theoretical, achievable, and actual cultivated land productivity per unit area in study area are respectively 2.33 kg/m2, 1.58 kg/m2, and 1.08 kg/m2; the theoretical, achievable, and actual total cultivated land productivity are respectively 5.47, 3.72, and 2.54 million tons. In addition, theoretical usage potentiality is 0.74 kg/m2, and achievable usage potentiality is 0.50 kg/m2, while theoretical usage intensity and achievable usage intensity both are 0.68. As for the spatial distribution of cultivated land productivity, the theoretical, achievable, and actual cultivated land productivity per unit area show the decreasing trend from the southwest to northeast, and the total theoretical, achievable, and actual cultivated land productivity show the decreasing distribution trend from northwest to southeast. 2) The single effects of 7 main factors including difference vegetation index (DVI), slope, geomorphic type, black soil thickness, organic matter, mechanization degree, and irrigation potential are great;the strongest action intensity to productivity is mechanization degree, and the weakest is black soil thickness. Except for the slope that has a negative effect, the other six factors have a positive effect to cultivated land productivity in the study area. The synergistic effects intensity among the 7 main factors to cultivated land productivity are strong, and the synergistic effect intensity between the other 6 main factors and geomorphological type is the strongest, while the synergistic effect intensity between the other 6 main factors and organic matter is the weakest. Besides, only the synergy action intensity between the main influential factors and slope to cultivated land productivity is negative and synergistic effects among the other 6 main factors all are positive. Meanwhile, according to the total 25 influential factors of cultivated land productivity, synergistic effects of 5-group-relative factors formed from the 7 main factors and the rest of the 18 influential factors (non-main factors) also play a significant part in the cultivated land productivity. Synergistic effects intensity between chemical fertilizer application and mechanization degree is the strongest, which is positive, while synergistic effects intensity between elevation and slope is the weakest and is negative. The effect mechanisms of influential factors show that single effects of main factors and synergistic effects of factors play an important role in the cultivated land productivity.