早稻氮磷钾施肥类别归属的贝叶斯判别方法研究
|
摘要
【目的】在测土配方施肥中,"测土"是实现合理施肥微观指导的关键技术手段。本研究针对高度分散经营的耕地代表性土样采集难、测试费用高、花费时间长等"测土"难题,探讨早稻氮磷钾施肥类别归属的统计模式识别技术。【方法】利用闽东南80个早稻氮磷钾田间肥效试验资料作为样本,采用欧氏距离–离差平方和法进行系统聚类分析,建立区域早稻施肥类别。在此基础上,根据田块氮磷钾施肥量及其产量结合地力信息等数量指标,应用贝叶斯判别准则探讨具体田块的施肥类别归属。【结果】在确保空白区产量和平衡施肥产量均值在两两施肥类别间具有统计显著差异的前提下,闽东南早稻最多可划分为6个氮磷钾施肥类别。多元统计检验表明,6个施肥类别的类内协方差矩阵差异不全相等,据此建立了各个施肥类别的贝叶斯判别函数。结果表明,原始试验数据标准化处理可大幅度提高判别准确性,监测样本的回代误判率和交叉误判率分别只有1.2%和1.3%,预留6个试验点的84个处理的氮、磷、钾施肥量和产量试验资料的施肥类别归属的判别正确率达到81.0%。【结论】统计模式识别原理及其贝叶斯判别准则可用于早稻氮磷钾施肥类别归属决策,判别准确性能够满足推荐施肥的精度要求。
【Objectives】In soil testing and formulated fertilization, soil testing is the key to realize rational fertilization through practical guidance. However, collection of representative soil samples is often difficult because of highly decentralized farmland management, and soil sample analysis is cost and time-consuming.Therefore, statistical pattern recognition techniques were studied in this paper to explore category attribution of nitrogen(N), phosphorus(P) and potassium(K) fertilization without soil testing for early rice.【Methods】Data from eighty field experiments in southeast of Fujian Province, China, were used in this study. Based on the response of early rice to N, P and K fertilizers, the paddy fields were divided into regional fertilization categories,using clustering analysis method of Euclideana distance-sum of squares of deviations. Then the NPK fertilization category for a field was calculated based on the statistical pattern recognition principle, the application rates of N,P and K fertilizers and the outputs.【Results】On condition of ensuring that the average yield of the blank area and that of balanced fertilization had statistical significant differences between any two fertilization categories, the80 paddy fields were divided clearly into six fertilization categories. Multivariate statistics showed that the differences of covariance matrix of six categories were not all equal, and the Bayesian discrimination function of each category was established based on that. The standardization of the original data greatly improved the discrimination accuracy, with the back substitution misjudgment rate and cross misjudgment rate of training samples of only 1.2% and 1.3% respectively, and the category discrimination accuracy of the 84 treatments in the6 reserved experimental sites reached 81.0%.【Conclusions】Statistical pattern recognition principle and its Bayesian discriminate analysis method may provide an effective technical approach for the attribution decisions of N, P and K fertilization category of early rice without soil testing, and its results could meet the accuracy requirements of fertilizer recommendation.
【Objectives】In soil testing and formulated fertilization, soil testing is the key to realize rational fertilization through practical guidance. However, collection of representative soil samples is often difficult because of highly decentralized farmland management, and soil sample analysis is cost and time-consuming.Therefore, statistical pattern recognition techniques were studied in this paper to explore category attribution of nitrogen(N), phosphorus(P) and potassium(K) fertilization without soil testing for early rice.【Methods】Data from eighty field experiments in southeast of Fujian Province, China, were used in this study. Based on the response of early rice to N, P and K fertilizers, the paddy fields were divided into regional fertilization categories,using clustering analysis method of Euclideana distance-sum of squares of deviations. Then the NPK fertilization category for a field was calculated based on the statistical pattern recognition principle, the application rates of N,P and K fertilizers and the outputs.【Results】On condition of ensuring that the average yield of the blank area and that of balanced fertilization had statistical significant differences between any two fertilization categories, the80 paddy fields were divided clearly into six fertilization categories. Multivariate statistics showed that the differences of covariance matrix of six categories were not all equal, and the Bayesian discrimination function of each category was established based on that. The standardization of the original data greatly improved the discrimination accuracy, with the back substitution misjudgment rate and cross misjudgment rate of training samples of only 1.2% and 1.3% respectively, and the category discrimination accuracy of the 84 treatments in the6 reserved experimental sites reached 81.0%.【Conclusions】Statistical pattern recognition principle and its Bayesian discriminate analysis method may provide an effective technical approach for the attribution decisions of N, P and K fertilization category of early rice without soil testing, and its results could meet the accuracy requirements of fertilizer recommendation.
引文
[1]杨卓亚,毛达如,黄金龙,卢志光.基于年景变化的施肥决策[J].北京农业大学学报,1995,21(Suppl.):19-22.Yang Z Y,Mao D R,Huang J L,Lu Z G.Fertilization strategy for different crop yield year type[J].Acta Agriculturae Universitatis Pekinensis,1995,21(Suppl.):19-22.
[2]耿兴元,毛达如,曹一平.土壤肥力模糊量化评价(判)系统的建立[J].北京农业大学学报,1995,21(Suppl.):23-28.Geng X Y,Mao D R,Cao Y P.A fuzzy system for quantitative evaluation of soil fertility[J].Acta Agriculturae Universitatis Pekinensis,1995,21(Suppl.):23-28.
[3]金耀青,张中原.配方施肥方法及其应用[M].沈阳:辽宁科学技术出版社,1993:35-158.Jin Y Q,Zhang Z Y.The method of formula fertilization and its application[M].Shenyang:Liaoning Science and Technology Press,1993:35-158.
[4]李娟,章明清,孔庆波,等.福建早稻测土配方施肥指标体系研究[J].植物营养与肥料学报,2010,16(4):938-946.Li J,Zhang M Q,Kong Q B,et al.Soil testing and formula fertilization index for early rice in Fujian Province[J].Plant Nutrition and Fertilizer Science,2010,16(4):938-946.
[5]杨俐苹,白由路,王贺,等.测土配方施肥指标体系建立中“3414”试验方案应用探讨-以内蒙古海拉尔地区油菜“3414”试验为例[J].植物营养与肥料学报,2011,17(4):1018-1023.Yang L P,Bai Y L,Wan H,et al.Application of“3414”field trial design for establishing soil testing and fertilizer recommendation index[J].Plant Nutrition and Fertilizer Science,2011,17(4):1018-1023.
[6]侯彦林,郭喆,任军.不测土条件下半定量施肥原理和模型评述[J].生态学杂志,2002,21(4):31-35.Hou Y L,Guo Z,Ren J.Summarization of principles and models for semi-quantitative fertilization without soil testing[J].Chinese Journal of Ecology,2002,21(4):31-35.
[7]Kim J W,Ahn Y J,Lee K C,et al.A classification approach for genotyping viral sequences based on multidimensional scaling and linear discriminant analysis[J].BMC Bioinformatics,2010,11:434.
[8]Zhao X J,Pei Z Y,Liu J,et al.Prediction of nucleosome DNAformation potential and nucleosome positioning using increment of diversity combined with quadratic discriminant analysis[J].Chromosome Research,2010,18:777-785.
[9]Li X L,Yi S L,He S L.Identification of pummelo cultivars by using Vis/NIR spectra and pattern recognition methods[J].Precision Agriculture,2016,17:365-374.
[10]罗汝英.江西省低山区杉林土壤障碍性条件的判别分析[J].土壤学报,1980,17(2):145-155.Luo R Y.Discriminant analysis of inhibiting soil conditions of Chinese fir stands in low mountain district of Jiangxi Province[J].Acta Pedologica Sinica,1980,17(2):145-155.
[11]熊国炎.两组线性判别分析在土壤分类中的应用[J].土壤,1980,(5):177-181.Xiong G L.Two groups of linear discriminant analysis applied for the soil classification[J].Soils,1980(5):177-181.
[12]Brus D J,Bogaert P,Heuvelink G B.Bayesian maximum entropy prediction of soil categories using a traditional soil map as soft information[J].European Journal of Soil Science,2008,59(2):166-177.
[13]邱琳,李安波,赵玉国.基于Fisher判别分析的数字土壤制图研究[J].土壤通报,2012,43(6):1281-1286.Qiu L,Li A B,Zhao Y G.Digital soil mapping based on Fisher discriminant analysis[J].Chinese Journal of Soil Science,2012,43(6):1281-1286.
[14]吴海波,赵晓慎,王治国,张超.基于Bayes判别分析模型的水土保持区划[J].水土保持科学,2012,10(2):88-91.Wu H B,Zhao X Z,Wan Z G,Zhang C.Regional division for soil and water conservation based on Bayes discriminant analysis model[J].Science of Soil and Water Conservation,2012,10(2):88-91.
[15]鲁如坤.土壤农业化学分析方法[M].北京:中国农业科技出版社,2000:146-196.Lu R K.Soil agricultural chemical analysis method[M].Beijing:China Agricultural Science and Technology Press,2000:146-196.
[16]Andrew R W,Keith D C(王萍译).统计模式识别(第三版)[M].北京:电子工业出版社,2015:163-201.Andrew R W,Keith D C(Translated by Wang P).Statistical pattern recognition(3rd ed.)[M].Beijing:Electronic Industry Press,2015:163-201.
[17]李娟,章明清,孔庆波,等.构建县域早稻氮磷钾施肥的系统聚类方法研究[J].植物营养与肥料学报,2017,23(2):531-538.Li J,Zhang M Q,Kong Q B,et al.Building fertilization categories of N,P and K fertilization for early rice using systematic clustering method in county territory[J].Journal of Plant Nutrition and Fertilizer,2017,23(2):531-538.
[18]袁志发,宋世德.多元统计分析(第二版)[M].北京:科学出版社,2009:278-293.Yuan Z F,Song S D,Multivariate statistical analysis(2nd ed.)[M].Beijing:Science Press,2009:278-293.
[19]李柏年,吴礼斌.MATLAB数据分析方法[M].北京:机械工业出版社,2015:81-104.Li P N,Wu L B.MATLAB data analysis method[M].Beijing:Mechanical Industry Press,2015:81-104.
[20]毛达如,张承东.多点肥料效应函数的动态聚类方法[J].北京农业大学学报,1991,17(2):49-54.Mao D R,Zhang C D.Dynamic clustering method of multipoint fertilizer response function[J].Acta Agriculturae Universitatis Pekinensis,1991,17(2):49-54.
[21]章明清,林仁埙,林代炎,姜永.极值判别分析在三元肥效模型推荐施肥中的作用[J].福建农业学报,1995,10(2):54-59.Zhang M Q,Lin R X,Lin D Y,Jiang Y.Function of distinguish analysis on extreme value in recommendatory fertilization for threefertilizer efficiency model[J].Fujian Journal of Agricultural Sciences,1995,10(2):54-59.
[22]周鸣铮.测土施肥的科学基础[J].土壤通报,1984,15(4):156-160.Zhou M Z.Scientific foundation of soil testing and recommendation fertilization[J].Chinese Journal of Soil Science,1984,15(4):156-160.
[23]周鸣铮.中国的测土施肥[J].土壤通报,1987,18(1):7-13.Zhou M Z.Soil testing and recommendation fertilization in China[J].Chinese Journal of Soil Science,1987,18(1):7-13.
[24]金耀青.配方施肥的方法及其功能-对我国配方施肥工作的述评[J].土壤通报,1989,20(1):46-49.Jin Y Q.Formula fertilization methods and its functions-review of formula fertilization work in China[J].Chinese Journal of Soil Science,1989,20(1):46-49.
[25]陆允甫,吕晓男.中国测土施肥工作的进展与展望[J].土壤学报,1995,32(3):241-251.Lu Y P,LüX N.Progress and prospect in fertilizer recommendation based on soil testing in China[J].Acta Pedologica Sinica,1995,32(3):241-251.
[2]耿兴元,毛达如,曹一平.土壤肥力模糊量化评价(判)系统的建立[J].北京农业大学学报,1995,21(Suppl.):23-28.Geng X Y,Mao D R,Cao Y P.A fuzzy system for quantitative evaluation of soil fertility[J].Acta Agriculturae Universitatis Pekinensis,1995,21(Suppl.):23-28.
[3]金耀青,张中原.配方施肥方法及其应用[M].沈阳:辽宁科学技术出版社,1993:35-158.Jin Y Q,Zhang Z Y.The method of formula fertilization and its application[M].Shenyang:Liaoning Science and Technology Press,1993:35-158.
[4]李娟,章明清,孔庆波,等.福建早稻测土配方施肥指标体系研究[J].植物营养与肥料学报,2010,16(4):938-946.Li J,Zhang M Q,Kong Q B,et al.Soil testing and formula fertilization index for early rice in Fujian Province[J].Plant Nutrition and Fertilizer Science,2010,16(4):938-946.
[5]杨俐苹,白由路,王贺,等.测土配方施肥指标体系建立中“3414”试验方案应用探讨-以内蒙古海拉尔地区油菜“3414”试验为例[J].植物营养与肥料学报,2011,17(4):1018-1023.Yang L P,Bai Y L,Wan H,et al.Application of“3414”field trial design for establishing soil testing and fertilizer recommendation index[J].Plant Nutrition and Fertilizer Science,2011,17(4):1018-1023.
[6]侯彦林,郭喆,任军.不测土条件下半定量施肥原理和模型评述[J].生态学杂志,2002,21(4):31-35.Hou Y L,Guo Z,Ren J.Summarization of principles and models for semi-quantitative fertilization without soil testing[J].Chinese Journal of Ecology,2002,21(4):31-35.
[7]Kim J W,Ahn Y J,Lee K C,et al.A classification approach for genotyping viral sequences based on multidimensional scaling and linear discriminant analysis[J].BMC Bioinformatics,2010,11:434.
[8]Zhao X J,Pei Z Y,Liu J,et al.Prediction of nucleosome DNAformation potential and nucleosome positioning using increment of diversity combined with quadratic discriminant analysis[J].Chromosome Research,2010,18:777-785.
[9]Li X L,Yi S L,He S L.Identification of pummelo cultivars by using Vis/NIR spectra and pattern recognition methods[J].Precision Agriculture,2016,17:365-374.
[10]罗汝英.江西省低山区杉林土壤障碍性条件的判别分析[J].土壤学报,1980,17(2):145-155.Luo R Y.Discriminant analysis of inhibiting soil conditions of Chinese fir stands in low mountain district of Jiangxi Province[J].Acta Pedologica Sinica,1980,17(2):145-155.
[11]熊国炎.两组线性判别分析在土壤分类中的应用[J].土壤,1980,(5):177-181.Xiong G L.Two groups of linear discriminant analysis applied for the soil classification[J].Soils,1980(5):177-181.
[12]Brus D J,Bogaert P,Heuvelink G B.Bayesian maximum entropy prediction of soil categories using a traditional soil map as soft information[J].European Journal of Soil Science,2008,59(2):166-177.
[13]邱琳,李安波,赵玉国.基于Fisher判别分析的数字土壤制图研究[J].土壤通报,2012,43(6):1281-1286.Qiu L,Li A B,Zhao Y G.Digital soil mapping based on Fisher discriminant analysis[J].Chinese Journal of Soil Science,2012,43(6):1281-1286.
[14]吴海波,赵晓慎,王治国,张超.基于Bayes判别分析模型的水土保持区划[J].水土保持科学,2012,10(2):88-91.Wu H B,Zhao X Z,Wan Z G,Zhang C.Regional division for soil and water conservation based on Bayes discriminant analysis model[J].Science of Soil and Water Conservation,2012,10(2):88-91.
[15]鲁如坤.土壤农业化学分析方法[M].北京:中国农业科技出版社,2000:146-196.Lu R K.Soil agricultural chemical analysis method[M].Beijing:China Agricultural Science and Technology Press,2000:146-196.
[16]Andrew R W,Keith D C(王萍译).统计模式识别(第三版)[M].北京:电子工业出版社,2015:163-201.Andrew R W,Keith D C(Translated by Wang P).Statistical pattern recognition(3rd ed.)[M].Beijing:Electronic Industry Press,2015:163-201.
[17]李娟,章明清,孔庆波,等.构建县域早稻氮磷钾施肥的系统聚类方法研究[J].植物营养与肥料学报,2017,23(2):531-538.Li J,Zhang M Q,Kong Q B,et al.Building fertilization categories of N,P and K fertilization for early rice using systematic clustering method in county territory[J].Journal of Plant Nutrition and Fertilizer,2017,23(2):531-538.
[18]袁志发,宋世德.多元统计分析(第二版)[M].北京:科学出版社,2009:278-293.Yuan Z F,Song S D,Multivariate statistical analysis(2nd ed.)[M].Beijing:Science Press,2009:278-293.
[19]李柏年,吴礼斌.MATLAB数据分析方法[M].北京:机械工业出版社,2015:81-104.Li P N,Wu L B.MATLAB data analysis method[M].Beijing:Mechanical Industry Press,2015:81-104.
[20]毛达如,张承东.多点肥料效应函数的动态聚类方法[J].北京农业大学学报,1991,17(2):49-54.Mao D R,Zhang C D.Dynamic clustering method of multipoint fertilizer response function[J].Acta Agriculturae Universitatis Pekinensis,1991,17(2):49-54.
[21]章明清,林仁埙,林代炎,姜永.极值判别分析在三元肥效模型推荐施肥中的作用[J].福建农业学报,1995,10(2):54-59.Zhang M Q,Lin R X,Lin D Y,Jiang Y.Function of distinguish analysis on extreme value in recommendatory fertilization for threefertilizer efficiency model[J].Fujian Journal of Agricultural Sciences,1995,10(2):54-59.
[22]周鸣铮.测土施肥的科学基础[J].土壤通报,1984,15(4):156-160.Zhou M Z.Scientific foundation of soil testing and recommendation fertilization[J].Chinese Journal of Soil Science,1984,15(4):156-160.
[23]周鸣铮.中国的测土施肥[J].土壤通报,1987,18(1):7-13.Zhou M Z.Soil testing and recommendation fertilization in China[J].Chinese Journal of Soil Science,1987,18(1):7-13.
[24]金耀青.配方施肥的方法及其功能-对我国配方施肥工作的述评[J].土壤通报,1989,20(1):46-49.Jin Y Q.Formula fertilization methods and its functions-review of formula fertilization work in China[J].Chinese Journal of Soil Science,1989,20(1):46-49.
[25]陆允甫,吕晓男.中国测土施肥工作的进展与展望[J].土壤学报,1995,32(3):241-251.Lu Y P,LüX N.Progress and prospect in fertilizer recommendation based on soil testing in China[J].Acta Pedologica Sinica,1995,32(3):241-251.