节点文献

小清河流域农业面源氮素污染模拟研究

Model Simulation on Agricultural Non-point Source Nitrogen Pollution in Xiaoqinghe Watershed

【作者】 高懋芳

【导师】 邱建军;

【作者基本信息】 中国农业科学院 , 农业遥感, 2011, 博士

【摘要】 氮肥的广泛使用是近现代农业发展的重要推动力,对提高农产品产量起到了至关重要的作用;然而随着氮肥投入量增加,土壤中过量的氮素以各种形式进入周围环境。畜禽养殖业的快速发展也产生了大量氮素流失,农业生态系统氮流失是农业面源氮素污染的主要原因。本文选择包含径流和畜禽养殖过程的最新版Manure-DNDC模型,在山东小清河流域典型农田和畜禽养殖场进行验证;借助遥感与地理信息系统(GIS)技术,利用多种来源数据建立以乡镇为单位的小清河流域数据库;模拟评价2008年小清河流域农业生态系统氮平衡状况,分析农业面源氮素污染时空分布特征,并提出有效地优化管理措施。主要研究结果如下:(1)实地验证了改进版农业生态系统生物地球化学模型——Manure-DNDC。利用山东小清河流域典型轮作系统冬小麦/夏玉米、冬小麦/大葱、设施蔬菜地田间试验数据对Manure-DNDC进行了验证,结果表明Manure-DNDC模型能够较好地模拟农田土壤气候、作物生长、氮淋溶等动态过程。利用广饶县肉牛养殖场观测数据验证表明,模型能够较好地动态模拟畜禽养殖过程中排出的粪便和尿的氮素含量,以及圈舍氨挥发量。区域验证也取得了较好地效果,模型能够合理预测作物产量和径流量,其中对冬小麦和夏玉米总产量模拟的R~2值均在0.8以上。(2)建立了以乡镇为单元的小清河流域农业面源氮素污染评价基础数据库,构建了小清河流域农业生态系统氮平衡定量评价系统。针对小清河流域地势低平的特点,研究提出了遥感图像识别与GIS空间分析相结合的数据库构建方法,首先利用高分辨率遥感图像识别中下游主干河道,修改DEM后提取流域边界以及主要河流网络,再计算坡度、坡长、最大水流长度等地形因子。同时通过查表获取主要土壤侵蚀因子,利用模型估算土壤水分常数,结合实测土壤有机质含量、作物播种面积和管理措施、气象观测数据、以及畜禽养殖数据构建小清河流域以乡镇为单元的农业面源氮素污染数据库。将数据库与经验证后的Manure-DNDC模型结合,综合考虑地表产汇流与土壤侵蚀规律,构建了流域农业生态系统氮平衡综合评价系统,为定量评价奠定了基础。(3)小清河流域农田种植与畜禽养殖过程中产生大量氮素盈余,农业面源氮素污染负荷很高。2008年农田氮素投入总量为25.99万吨,其中化肥氮18.42万吨,有机肥氮6.66万吨,相当于每公顷农田施用化肥282.06 kg N、有机肥101.95 kg N。农田淋溶和径流损失氮素分别为2.38万吨、0.71万吨,平均淋溶和径流损失分别为36.42 kg N/hm~2、10.82kg N/hm~2。畜禽养殖过程中通过径流损失氮素4.66万吨,堆肥场所有3.50万吨氮素残留。农业生态系统氮素径流损失总量是淋溶损失总量的2.26倍,其中86.83%发生在畜禽养殖以及粪便处理过程中。(4)农田氮素淋溶具有明显的区域差异,水肥管理是影响氮素淋溶强度的主要因素。济南市南部、章丘市中北部、以及寿光市部分地区农田氮素淋溶比较严重,其中有41个乡镇超过30 kg/hm~2。氮素淋溶与施肥量的关系非常密切,当施肥量低于300 kg/hm~2时,70%以上的乡镇淋溶强度小于10 kg/hm~2,当施肥量大于400 kg/hm~2时,约20%的氮肥受到淋溶。降水与灌溉是农田氮素淋溶的主要驱动力,流域内约70%的氮素淋溶发生在六、七月份。从不同的轮作模式看,蔬菜三季轮作的氮素淋溶损失量最高,达到1.07万吨,单位面积氮素淋溶量为285.32 kg/hm2。小麦/玉米轮作模式的平均淋溶强度为18.72 kg/hm~2,因为播种面积较大,淋溶总量也达到6192.25吨氮。集约化粮食作物和设施蔬菜地是农田氮素淋溶的主要来源。(5)氮素径流损失强度主要受坡度、降水、土壤性质以及畜禽养殖量的影响,径流损失氮素以有机氮为主。农田氮素径流损失强度比较大的乡镇主要位于流域上游的济南市历城区、章丘市、邹平县南部的,流域中部的广大地区农田径流损失一般都低于5 kg/hm~2。农田氮素径流损失总量最大的是小麦/玉米轮作,其次是棉花以及小麦、玉米单作。畜禽养殖以及粪便处理过程中的氮素径流损失在整个流域内都很严重,有31个乡镇损失强度超过100 kg/hm~2,其中济南市南部和章丘市最突出,主要原因是畜禽养殖规模大。牛排出的氮素中以气体成分损失的比例最高,家禽排出的氮素施入农田的比例最高。(6)农田氮素投入过量、畜禽养殖规模与农田面积不匹配、畜禽粪便处理方式不合理是造成本地区农业面源氮素污染的主要原因。减轻小清河流域农业生态系统氮流失与农业面源氮素污染负荷的有效措施主要有控制集约化粮食作物和设施蔬菜地的氮肥投入量,适当减少蔬菜地灌溉量,限制坡耕地种植,控制畜禽养殖密度,使畜禽养殖规模与农田面积相匹配,完善畜禽粪便处理技术等。优化管理措施分析表明,当化肥氮总量降低15%时,氮素淋溶量降低41%,小麦、玉米和蔬菜产量没有明显下降。减少单位面积农田上的畜禽养殖量,可以有效降低土壤氮盈余量,控制农田氮流失,同时减少畜禽养殖过程中的氮素损失。

【Abstract】 Application of chemical fertilizer substantially increased crop production in the 20th Century. However, overuse of nitrogenous fertilizer has introduced excessive nitrogen (N) into the environment in various forms. During the past decades, livestock industries rapidly grew to meet the increasing demand for dairy and meat products in China that has led to more N discharged into the air, soil and rivers. The excessive N loads formed the agricultural non-point source N pollution in the country. This study was to specify the sources of N pollution by means of a modeling tool. To handle the complexity of N cycling in the agroecosystems, a process-based model, Manure-DNDC, was adopted and modified to quantify N releases from the various agricultural sources in the Xiaoqinghe watershed. The main results were as follows:(1) Validation tests were conducted at site and regional scales to verify the applicability of Manure-DNDC for the domain watershed. The modified Manure-DNDC model is capable of simulating soil climate, crop biomass, N leaching parameters in typical winter wheat/summer maize, winter wheat/scallion and vegetable field. The model can also get good results for excretion N from livestock and ammonia from farm. The modified Manure-DNDC model is also capable of simulating crop product, surface runoff and soil erosion that drive the soil or manure N to move at horizontal dimension. R square for the production of winter wheat and summer maize simulation are higher than 0.8.(2) Agricultural non-point source N pollution evaluation database for every town and agricultural eco-system N balance evaluation system was established in Xiaoqinghe watershed. The database was constructed combined with remote sensing image identifying and Geographical Information System (GIS) spatial analysis as the watershed was flat in middle and downstream. High resolution remote sensing data was used for indentifying main channels and then modify the DEM based on the channels. Then watershed boundary, river network, slope, slope length and longest flow path will be picked up using GIS spatial analysis. Soil erosion parameter was get from lookup table. Soil moisture constant was get from model simulation. Agricultural non-point source N pollution database for Xiaoqinghe watershed in every town will be available combined with measured soil organic matter, crop sown area and management, meteorological data, and livestock data. Agricultural eco-system N balance evaluation system was constructed integrate the database, modified Manure-DNDC, surface runoff and soil erosion.(3) Large amount of N surplus appeared in agricultural planting and livestock breeding in Xiaoqinghe watershed, which caused serious agricultural non-point source N pollution. Based on the baseline simulations for 2008, 259.9 million kg N was added to the agricultural soils in the Xiaoqinghe watershed, including 184.2 million kg N from synthetic fertilizer application and 66.6 million kg N from livestock manure. It means that 282.06 kg N from synthetic fertilizer and 101.95 kg N from manure was added to every hectare cropland. Driven by the rainfalls and irrigation application, 23.8 million kg N was leached from the cropping systems in the year. Driven by surface runoff, 46.6 and 7.1 million kg N were lost from livestock operation and cropping systems, respectively. Average leaching and runoff N lost from cropland were 36.42 kg N/hm~2 and 10.82kg N/hm2, respectively. At the end of year, there was 35.0 million kg N manure left in the livestock operation systems, mainly in form of compost. N lost by mean of runoff was 2.26 times of that lost by means of leaching, with 86.83% runoff lost happened in livestock breeding system. In general, the major agricultural non-point source of N in this watershed were livestock manure runoff losses and fertilizer leaching from the cropland.(4) Main factor affecting N leaching was water and fertilizer management. Regions with high leaching mainly located in south Jinan, central and north Zhangqiu, and Shouguang, with 41 towns higher than 30 kg/hm2. N lost from the cropping system was mainly related to fertilizer application rates. If fertilizer rate was lower than 300 kg/hm2, N leaching for more than 70% towns was lower than 10 kg/hm2. However, 20% of N will be leached if fertilizer rate was higher than 400 kg/hm~2. Precipitation and irrigation are main drivers for N leaching in cropland. About 70% N leaching happened in June and July. Rotations with three vegetable in a year suffered from the most serious N leaching with 10.7 million kg N and 285.32 kg/hm~2. Average leaching intensity for winter wheat and summer maize was 18.72 kg/hm~2, with 6.19 million kg N in total. Intensive grain crop and vegetable planting was main sources for cropland N leaching.(5) Land slope, precipitation and soil property played key roles in determining the intensity of the surface runoff and soil erosion. The N lost from surface runoff and soil erosion was mainly in organic forms. Towns with high intensity of soil erosion mainly locate in upstream of Xiaoqinghe watershed, such as Licheng in Jinan, Zhangqiu and Zouping. N leaching from surface runoff in middle of the watershed was mainly lower than 5 kg/hm2. Winter wheat and summer maize rotation, cotton, wheat, and maize were major planting mode for N runoff lost. N lost from livestock operation systems was heavy all over the watershed with 31 towns higher than 100 kg/hm~2. South Jinan and Zhangqiu was the most serious area because of the large livestock scale.(6) The main reason for agricultural non-point source N pollution in Xiaoqinghe watershed was too much synthetic fertilizer application, too much livestock compared to the cropland, and inappropriate manure operation system. Optional management practices were tested with the modeling approach. The results indicated that the effective measures for mitigating N loading for in the target watershed include optimizing fertilizer application rates for cropping systems, adjusting irrigation intensity for vegetable fields, abounding cultivation on the slop lands, and improving manure management practices. An assessment was conducted to identify the best management practices for mitigating N loads. The results indicated that N leaching loss could be decreased by 41% if synthetic N fertilizer application rate is reduced by 15%, which will maintain the crop yields. Reducing the livestock herd size in the counties with intensified animal farms would significantly reducing N loading rates.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络