【作者】 奥海玮；

【导师】 谢应忠； 李永宏；

【作者基本信息】 宁夏大学 ， 草业科学， 2015， 博士

【摘要】 本研究以宁夏固原市农科所头营科研基地、宁夏同心县涝塘乡河草沟村、宁夏贺兰山农牧场、宁夏盐池县马儿庄乡、宁夏盐池科技局农科所农牧试验基地、宁夏海原县贾塘乡西北农林科技大学宁南半干旱试验区试验数据为支撑,应用澳大利亚农业生产系统模型(APSIM模型),从APSIM苜蓿模型在宁夏苜蓿种植区的适应性、APSIM模型对不同灌溉情形的适应性和APSIM模型对宁夏海原地区草粮轮作系统的适应性3个方面,全面系统地对APSIM模型在宁夏苜蓿种植区的适应性进行了验证分析,并利用经过验证的模型研究分析了不同气候变化与水分条件下苜蓿生产力与土壤水分环境变化及时空分布。主要研究结果如下：(1)完成了APSIM模型在宁夏苜蓿种植区的主要参数的本地化。首次确定了固原紫花苜蓿、美国大叶紫花苜蓿和E3006紫花苜蓿品种参数。实现了影响模型模拟苜蓿生长、灌溉及轮作精度的主要参数的本地化。(2)利用宁夏固原市农科所头营科研基地、宁夏同心县涝塘乡河草沟村、宁夏贺兰山农牧场田间试验数据。首次通过了APSIM在固原地区的参数校准以及在银川和同心两个地区对苜蓿生育期、苜蓿干物质产量和土壤含水量的验证。验证结果表明校准后的APSIM苜蓿模型模拟研究区苜蓿生长生育期、干物质产量和土壤含水量的模拟值与实测值决定系数R2范围在0.71-1之间,D值范围在0.85-0.99之间,体现出了较好的相关性和一致性,表明校准后的APSIM苜蓿模型在宁夏苜蓿种植区具有较好的适应性。干物质产量模拟值比实测值平均高估7.1%与土壤整体含水量模拟值比实测值平均高估10.1%相关。(3)利用宁夏盐池县马儿庄乡、宁夏盐池科技局农科所农牧试验基地试验数据,对APSIM模拟不同灌溉情形下的苜蓿生长能力进行了验证。各验证项模拟值和实测值表现出较好的相关性和一致性,决定系数R2范围在0.56至0.89之间,D值范围在0.96至0.99之间,NRMSE范围在17%至48%之间。同时,灌溉量与灌溉次数对干物质产量的影响趋势相同,校准后的APSIM模型可以较为准确地模拟不同灌溉情形下苜蓿的生产情况。(4)利用宁夏海原县贾塘乡西北农林科技大学宁南半干旱试验区田间试验数据,采用苜蓿-小麦-谷子8种不同轮作方式中的小麦-小麦-小麦(WWW)和谷子-谷子-谷子(MMM)2种进行APSIM草粮轮作系统的校正,确定相应作物品种参数。利用另外6种方式进行验证分析。验证结果表明6种轮作方式下产量实测值和模拟值的决定系数R2值范围在0.83至0.98之间,D值范围在0.94至0.99之间,6种轮作方式中5种方式的NRMSE值范围在21.1%至29.8%之间。土壤含水量实测值和模拟值的决定系数R2值范围在0.52至1之间,D值范围在0.94至0.97之间,6种轮作方式中5种方式的NRMSE值范围在27.4%至38.3%之间,产量和土壤含水量表现出了较好的相关性和一致性。首次实现了APSIM对宁夏海原地区草粮轮作系统的适应性验证。(5)基于APSIM苜蓿模型模拟研究了宁夏4个研究区1981～2010年间无灌溉情形下气候变化对苜蓿的雨养产量以及土壤含水量影响情况。30年间,4个研究区苜蓿雨养产量模拟值差异较大。生长季降雨量与苜蓿雨养产量之间的相关系数都很小,表明苜蓿的产量与生长季的降水无明显的线性关系。土壤含水量与苜蓿生长季降水量之间表现出一定的线性相关关系,尤其是同心和银川决定系数分别达到0.7和0.6,表明在宁夏干旱、半干旱地区土壤含水量主要来源于年季降雨。(6)利用校验好的APSIM灌溉模型,进行了不同灌溉情形的应用尝试,得出一些基本结论。灌溉是宁夏干旱、半干旱地区苜蓿增产的重要因素,可大幅度增加苜蓿干物质产量。灌溉量较低时,蒸发蒸腾量(ET)也较低,灌溉量较大时,ET也较大,两者变化趋势基本一致,ET值与灌水量呈线性显著相关(P<0.01),相关系数达到0.826。苜蓿水分利用效率与灌溉量不成正比,呈先增长后下降的趋势,对苜蓿栽培具有一定指导意义。(7)利用校验好的APSIM草粮轮作模型,分别对2000-2011年12年间和2039-2050未来12年间不同气候条件下草粮轮作系统土壤含水量和作物产量的变化情况进行了模拟分析和对比,预测未来气候条件下草粮轮作系统相对于历史气候条件下草粮轮作系统土壤含水量有较为明显的下降,但作物产量却有所增加。土壤含水量下降主要与未来降雨有所下降有关,作物产量增加可能与未来气候太阳辐射量增大有关。(8)通过将GIS与APSIM相结合,利用21个气象观测站点的气象数据,运用反距离插值法,有效地分析了宁夏干旱、半干旱地区苜蓿全生育期内以及不同生育阶段平均降雨量、日照时数、最高、最低气温等气候要素以及其中8个地区的苜蓿产量的时空分布特征,得出降雨量是影响苜蓿产量的主要气候要素,但不是唯一决定性因素。分枝至开花阶段适当补充苜蓿土壤水分,对苜蓿增产具有重要的影响。固原、西吉、海原大部以及盐池、中宁、同心部分地区苜蓿产量较高,比较适宜于苜蓿种植,可以作为苜蓿产业发展的主要地区。更多还原

【Abstract】 The study aims at a systematic evaluation and the application of Agricultural Production System Simulator (APSIM) to lucerne planting zones in Ningxia Hui Autonomous Region, China. With data collected from Touying Base of Guyuan Agricultural Institute; Hecaogou Village, Laotang Township, Tongxin County; Helanshan Pasture Farm; Ma’erzhuang Township, Yanchi County; Pilot Base of Yanchi Agricultural Institute; and the Northern Ningxia Semi-Arid Pilot Zone at Jiatang Township, Haiyuan County, developed by Northwest Agriculture and Forestry University, the model was verified in terms of how well the model was applied to lucerne production in the above zones, under different irrigation conditions, and involved in grass-crop rotation systems at Haiyuan County especially. The verified model was further analyzed in terms of how lucerne production and soil moisture varied with climate. Through a careful analysis, major findings include the following:(1) APSIM’s main parameters were determined specifically for the lucerne planting zones in Ningxia, including those for Guyuan variety, Meiguodaye and E3006 variety for the first time, thereby leading to the model with localized parameters to simulate the plant’s growth, irrigation and rotating accuracy;(2) the parameters were adjusted and verified in Guyuan for the first time and the plant’s period of growth, production of dry plant and soil moisture were also verified through an analysis of data collected from Touying Base, Hecaogou Village and Helanshan Pasture Farm; and results show that the R2 was between 0.71 and 1 and D between 0.85 and 0.99, which indicates that the simulated and observed data are significantly correlated and consistent; correlation is also found between production of dry plant and overall soil moisture when the former had 7.1%higher simulated data and the latter 10.1% on average, which indicates that the verified model can be well applied to the Ningxia planting zones;(3) irrigation conditions were verified in Yanchi through an analysis of data collected from Ma’erzhuang Township and the Pilot Base of Yanchi Agricultural Institute; and results show that the R2 was between 0.56 and 0.89, D between 0.96 and 0.99 and NRMSE between 17% and 48%, which indicates that the simulated and observed data are significantly correlated and consistent; and that irration volume and frequency tend to take the same effects on the production of dry plant and the verified model can work well for simulating the plant’s production in different irrigation conditions;(4) the parameters for the crop varieties were determined for the first time in Haiyuan through an analysis of data collected from the Northern Ningxia Semi-Arid Pilot Zone at Jiatang Township; of the model’s eight Lucerne-wheat-rice rotation systems, WWW and MMM were adjusted; other six systems were verified and analyzed and results show that the production had the R2 between 0.83 and 0.98, D between 0.94 and 0.99, but NRMSE between 21.1% and 29.8% simply for five of them; and that the soil moisture had the R2 between 0.52 and 1, D between 0.94 and 0.97, but NRMSE between 27.4% and 38.3% simply for five of them; it is indicated that the simulated and observed data are significantly correlated and consistent in both the production and soil moisture;(5) through the model simulation, effects of climate changes was analyzed on the plant production and soil moisture with no irrigation between 1981 and 2010 in the four zones of Ningxia and significant differences are found; little correlation is found beween the rainfall during the period of growth and the production relying simply on the rainfall, which indicates no significant linear correlation between them; but such significance is found between the rainfall during the period of growth and the soil moisture due to the R2 of 0.7 and 0.6 in Tongxin and Yinchuan repectively, which indicates that the annual rainfall is the major contributor to soil moisture in such arid and semi-arid zones;(6) the verified model was put into use in different irrigation conditions and results show that irrigation is a major factor in an increased production of Lucerne in Ningxia arid and semi-arid zones and a sharp increase is possible; changes in both volume and ET showed similar tendency and significant correlation is found (p<0.01) with a coefficient of 0.826; but efficiency of Lucerne’s water consumption is found not proportional to irrigation volume and it is a tendency that a fall goes after a rise at the start, which is significant to the planting of Lucerne.(7) with the verified model simulation, correlation between soil moisture and crop production in grass-crop rotation systems under different climates were analyzed between 2000 and 2011 and a comparative analysis was then for a prediction from 2039 till 2050; results show that soil moisture will significantely decrease while crop production will increase to some extent, which can attribute to a possible reduction of rainfall as well as a possible increase of sun radiation in the future;(8) through a combination of GIS and APSIM, climate factors, in terms of average rainfall, sunlight hours, and minimal and maximal temperatures, were well determined in the whole period of growth as well as in different phrase of the period with data collected from the 21 meteorological observations by using the inverse distance weighted method; and the space-time distribution of Lucerne production in the eight zones were analyzed and results show that rainfall is a major but not the only decisive climate factor in the plant’s production; a proper increase in soil moisture will contribute significantly to the plant’s increased production during its branching and blooming phrases; the majority of Guyuan, Xijie and Haiyuan and some parts of Yanchi, Zhongning and Tongxin are considered to be dominant Lucerne planting zones for a higher production in better conditions.更多还原