干旱区生态环境知识资源中心(Arid): 利用PPR新方法建立草地产量增长模型及生态成因机理研究

Arid

	利用PPR新方法建立草地产量增长模型及生态成因机理研究
其他题名	The Study on Establishing Growth Patterns of Grassland Yields and Analysis of its Ecological Formations Using New PPR
	李建龙; 蒋平; 朱明; 宋子健; 赵德华
来源期刊	生态学报
ISSN	1000-0933
出版年	2002
卷号	22 期号:7 页码:973-981
中文摘要	利用新型的投影寻踪回归（PPR）统计方法，对新疆天山北坡蒿属荒漠草地上观测的3a草地产量和其它6个相关生态因子数据，进行了多元回归统计分析和PPR分析，建立了PPR新型牧草增长模型，并进而做了产量生态成因机理分析和理论解释。研究结果表明，在一般年份下，干草产量在不同年代不同季节均随降水量的分配状况而发生有规律的变化，产量分配呈现双峰模式；在降水量、积温和光照时数3个影响产量形成和高低的生态因子中，尤以降水量影响最明显，是主导生态因子，但牧草生长发育状况和增长类型则是水热因子、质地、牧草生态习性和种类组成共同作用的结果；在建立草地产量增长模型和成因机理分析过程中，充分运用PPR统计方法成功地克服了非正态线性高维数据点稀疏所造成的维数祸根和自动排除了无关因子及人为因素的干扰，提高了建模的精度和效能，并给出自变量因子对建模贡献率大小的效果，科学地揭示了用其它多元回归统计方法，所以难以发现的数据内在的结构特征和规律性。进而，运用建立的PPR模型，较好地估测和拟合实测产量值，使模型估产精度达90%以上，并通过了效果检验和给出了贡献统计分析。在6个自变量因子中，对牧草产量增长贡献值最大的是生长时期（T），其次为土壤含水量（SW）、降水量（PR）、气温（AT）和土壤温度（ST），最小为草地类型（GT）值。多年研究结果为合理利用天山草地和深入了解草地生态系统形成机理，提供了坚实的科学依据和建模的新方法等。
英文摘要	Using the Projection Pursuit Regression (PPR) method, the multivariate regression statistical and PPR analysis were made between the grassland yields and six other relative ecological factors observed on the desert steppe in the northern slope of Tianshan, Xinjiang in 1992 ～ 1994. The PPR regression models of herbage growth were set up, and the formations of grassland yields were analyzed and theoretically explained about them. The results showed that hay yields regularly varied according to precipitations and possessed a bimodal distribution in common years. Among the three factors (precipitation (PR), accumulated temperature (AT) and light duration affected the herbage yield formations), precipitation was the chief ecological one. However, the growth conditions and patterns of herbages were resulted jointly by hydrothermal factors, soil textures, ecological habits and species compositions. While setting up the growth models of grassland yields and analyzing the yield formations with PPR method, the dimensional curses brought about by rarefaction of nonnormal and nonlinear higher dimensional data points and the interference of independent factors and anthropogenic factors were automatically eliminated. Thus, the precision and effectiveness of setting up growth models were enhanced by the PPR method. The study also offered the results and contribution of each variance in the PPR model, and scientifically brought out the internal structural characters and regularity of the data indiscoverable than other multivaiate regression statistical methods. The real grassland yields were estimated and that fitted well with the PPR regressive estimating model in 5 types. The estimating precisions were above 90%. The models passed the effective verification and the statistical analysis of contributive rates was done. Among the 6 variances, growth periods (T) had the greatest contributive rate to herbage yields, and soil water contents (SW), precipitations (PR), accumulated temperatures (AT), soil temperatures (ST), grassland types (GT) followed decreasingly in succession. Our results provided foundations for reasonable utilization of grassland and understanding of the yield formative mechanism in grassland ecological systems.
中文关键词	PPR方法 ; 天山草地 ; 建立PPR模型 ; 产量成因机理分析
英文关键词	PPR new method grassland in Tianshan setting up of PPR model yield formative effect analysis
语种	中文
国家	中国
收录类别	CSCD
WOS类目	AGRICULTURE MULTIDISCIPLINARY
WOS研究方向	Agriculture
CSCD记录号	CSCD:1067468
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/203387
作者单位	南京大学生命科学学院应用生态研究所, 南京, 江苏 210093, 中国
推荐引用方式 GB/T 7714	李建龙,蒋平,朱明,等. 利用PPR新方法建立草地产量增长模型及生态成因机理研究[J],2002,22(7):973-981.
APA	李建龙,蒋平,朱明,宋子健,&赵德华.(2002).利用PPR新方法建立草地产量增长模型及生态成因机理研究.生态学报,22(7),973-981.
MLA	李建龙,et al."利用PPR新方法建立草地产量增长模型及生态成因机理研究".生态学报 22.7(2002):973-981.