【作者】 鲁明星；

【导师】 吴礼树； 贺立源；

【作者基本信息】 华中农业大学 ， 植物营养学， 2007， 博士

【摘要】 耕地是土地的精华，是农业生产最重要的资源。耕地地力的好坏直接影响到农业的可持续发展和粮食安全。开展耕地地力评价，摸清耕地地力及其影响因子的变化和条件，是加强耕地质量建设和合理利用土地的重要基础。本研究采用传统土壤调查方法与现代“3S”（GIS、GPS、RS）技术相结合，选择鄂东沿江丘陵区（鄂州市）、鄂中丘陵区（钟祥市）、江汉平原区（江陵县、荆州区）等3个主要农作区的4个县（市、区）作为评价区域，对3个试区两个时段（1980-1981年和2003-2005年）的耕地地力及其时空变化特征进行了研究与探索。同时，利用耕地地力评价成果，开展了土壤养分空间插值与取样密度比较、土壤速效磷丰缺指标修订、农田精细农作氮素营养快速诊断等应用研究，获得了以下结果：1．在GIS支持下，利用土壤图、土地利用现状图叠置划分法确定区域耕地地力评价单元，分别建立了鄂州、江陵（荆州）、钟祥等3个地貌类型区耕地地力评价指标体系及其模型，运用层次分析法和模糊数学方法对耕地地力进行了综合评价。结果表明，鄂州市现有耕地共划分为1-6等，其中高产地（1、2等）占耕地的22.57％，而低产地（4等以下）占38.22％；江陵县（含荆州区）和钟祥市现有耕地划分为1-4等，其中高产地（1、2等）分别占77.75％和72.12％，中低产地（3、4等）占22.25％和27.88％。不同区域间表现出江汉平原及鄂中丘陵区，高产田面积大（占70％以上），而鄂东沿江丘陵区中低产田所占比例大（占75％以上）。20年间不同区域耕地地力变化表现出，丘陵区耕地面积大幅减少，耕地质量明显下降；平原区耕地面积虽有所减少，但减幅不大，耕地数量和质量保持基本稳定。2．对不同地貌类型区、水田、旱地和菜地耕层土壤主要理化属性及其时空变化特征进行了分析，比较、归纳了试区不同土壤属性的变化规律，发现20年间水田土壤养分变化特征为，土壤OM、全N和速效K含量呈下降趋势，而土壤有效P则明显上升；旱地则表现为土壤OM、全N、有效P和速效K含量呈升高趋势；不论水田还是旱地均表现出明显的土壤酸化趋势。研究按照土类进行了变异特征分析，获得了不同土壤主要理化属性的变化规律。对菜地的研究考虑了养分在耕层（0-20cm）和亚耕层（20-40cm）的分布特征，发现土壤OM、全N、有效P、速效K和有效B、Zn、Fe、Mn、Mo等营养元素的含量呈现较明显的垂直分布规律，耕层大于亚耕层；而pH则呈现相反特征，耕层pH平均值为6.65，亚耕层则为7.14。3．通过对kriging（克吕格）插值法、样条函数法、距离权重倒数法在不同空间尺度下土壤养分含量的插值效果及按不同土壤特性对合理采样密度的分析，发现kriging插值法与距离权重倒数法的插值精度要比样条函数法高，插值结果的离散程度比实际测定值小，样条函数法插值结果的离散程度较大；合理的采样密度与土壤利用类型和养分元素含量的变异大小有关；土壤质地影响养分元素的插值误差，旱地的插值误差较水田要大；土壤有效P的插值误差最大，pH的插值误差最小，铵态N和速效K的插值误差居中。本研究样区中，随着采样密度增加，有效P和pH的插值结果变化不明显，铵态N和速效K的插值误差则随着采样密度的增加而减小。4．针对耕地土壤有效P含量大幅上升的现状，于2003-2005年在湖北省油菜主产区进行了27个油菜施磷试验，对当前土壤速效P的分级标准进行了研究，发现湖北省油菜施磷增产效果显著，有92.6％的试验施磷增产5％以上，每千克P₂O₅平均增收油菜籽6.2kg，产投比平均达3.10；以不施磷对照处理产量占施磷处理产量的50％、75％和90％作为判断土壤缺磷标准，则土壤速效P极度缺乏、严重缺乏和缺乏的指标分别为9.9 mg／kg、14.3 mg／kg和20.2 mg／kg，约高出20年前推荐指标的1倍。试验结果表明，研究修订的新指标可用于指导当前的油菜生产。5．通过田间氮肥小区试验，结合离子树脂膜和数码相机快速获取土壤和水稻氮素营养状况，配合图象处理技术进行叶色和封行指数提取及相关分析研究，结果表明水稻叶色指数与施氮量、封行指数与施氮量的相关均达到显著水平，离子交换树脂法获得的土壤有效养分与最高分蘖数的相关达到显著水平；初步建立了水稻生长过程氮素营养田间诊断模式，为后续利用GIS软件对这些信息进行动态管理，进而建立鄂东沿江丘陵区精细农作模式进行了有益探索。更多还原

【Abstract】 Farmland is the most important resource of agricultural production and the prime of land, fertility of which has a direct influence on food security and sustainable development of agriculture. Therefore, it is fundamental work of improving quality of farmland and properly making use of it to evaluate farmland fertility grade and analysis the factors and their change conditions affecting on soil fertility.Adopting traditional method of soil survey and geological information system （GIS）, geological position system （GPS）, remote system （RS） techniques, choosing 4 counties （or cities） in 3 agricultural areas, including Ezhou City in Edong hilly area along Yangtze River located at eastern Hubei province, Zhongxiang City in central area of Hubei province, both Jiangling County and Jingzhou District in Jianghan Plain located at south Hubei province, investigate and study farmland fertility and its variation characters with time and space within period of 1980-1981 and 2003-2005. Using farmland fertility evaluation achievements, it is also to carry out special study to such research hotspots as soil nitrogen testing and fertilizer recommendation, indexes of abundance and shortage of soil available P, models of precision farming in south China, improvement and utility of middle-and-low-yield cropland, food security, compare soil nutrient interpolation with different sampling density. Conclusions were as follows:1. The evaluating units of farmland fertility established by overlapping soil maps and soil utilizing maps with GIS, index system and evaluation model were set up respectively according to the physiognomy of three areas above, and the farmland fertility grade was comprehensively evaluated. The result showed that farmland in Ezhou City could be divided into one to six grades, in which grade one to grade two （high-yield farmhmd） accounted for 22.57% and grade four or below （low-yield farmland）38.22% in area. The farmland in Jiangling County （including Jingzhou District） and Zhongxiang City could be divided into one to four grades, in which mid-and high-yielding farmland took up 77.75% and 72.12%, and mid- and low-yielding farmland 22.25% and 27.88%. However, in near twenty years, farmland in different areas had an obvious change that the amount and quality of farmland in hill district decreased obviously, while the area in plain district had a little bit decrease, and the quantity and quality were almost stable.2. Soil physico-chemical properties and variability in space and time in tillaged-layer of paddy, dry field and vegetable field in different physiognomic areas were analyzed. The variation characteristics for various soil nutrient elements in the three different types of farmland before and after twenty years were explored and compared. It were revealed that content of soil OM, total N, available K showed a decline in paddy fields, soil available P increase obviously, that contents of soil OM, total N, fast K exhibited an increase in dry fields. And the same trend of soil acidification was showed in both paddy and dry fields. For vegetable field, the distribution properties of soil nutrients in both tillaged-layer（0-20cm） and sub-tillaged-layer （20-40cm） were studied. The research discovered that contents of OM, alkali-hydrolyzable N, available P and K, and available B、Zn、Fe、Mn、Mo took on vertical distribution （tillaged-layer＞sub-tillaged-layer）, while pH value showed reverse development （tillaged-layer＜sub-tillaged-layer）-pH6.65 in tillaged-layerand and pH7.14 in sub-tillaged-layer. So, the above results indicated that soil nutrient elements accumulated in tillaged-layer and acidification developed into under-layer. According to the various types of soil classification, the main variation rules of chemic.-physical properties of them were gained at the same time.3. By using interpolation methods of Kriging, spline function, distance-and-weight-countdown, comparing interpolation effects for contents of soil nutrient in different space, and analyzing proper sampling density in line with soil properties, the research found out that interpolation methods of Kriging and distance-and-weight-countdown was more precise than method of spline function, the dispersion degree of interpolation results lower than actually, that of method of spline function higher, proper sampling density had relation to soil utility types and variability of contents of soil nutrient elements, and that soil texture had influence on interpolation error—the interpolation error of dry farmland was larger than of paddy field, and that of soil available P was the largest, that of pH the smallest, and that of NH₄⁺—N and available K in middle. The interpolation error of available P and pH had little change with the increase of soil sampling density and that of NH₄⁺—N and available K decreased as sampling density increasing in the study area.4. In light of fact that content of available P in farmland soil rose to a great extent, twenty-seven rape applying P experiments were carried out in year 2003-2005 in the main rape plantation area of Hubei province to examine grading standard for current soil available P. The results showed that application of P to rape could markedly increase its production, 92.6% experiments for fertilizing P got 5% increase of production, each kilogram of P₂O₅ could produce 6.2 kg of rapeseed and output/input ratio reach 3.10 in average. If taking 50%, 75% and 90% of CK production （no application of P） as standard for P deficiency, the indexes of extreme lack, serious lack and lack were 9.9 mg/kg, 14.3 mg/kg和20.2 mg/kg, respectively, which showed the new indexes could be taken as a guidance for present rape production.5. Nutrition status of N in soil and rice plant was quickly obtained from plot trial for N-fertilizer treatment. By membrane of ion-exchange resin and digital camera, the data of leaf color index and vegetation Index were collected and studied with the help of image manipulation technique, which indicated that indexes for both leaf color and vegetation had significant correlation with amount of fertilization and so did soil available nutrients obtained by method of ion-exchange resin with the maximum number of tillering. Furthermore, diagnosing model for N-nutrient was preliminarily established, which laid foundation for establishing precision farming model in Edong hilly area along Yangtze River by making use of GIS software to manage these information dynamically.更多还原