广义可加模型在R中的快捷实现及蓝藻水华预测分析中的应用

生态学杂志

广义可加模型在R中的快捷实现及蓝藻水华预测分析中的应用

邓建明^1,2,秦伯强^1**,王博雯^3

(¹中国科学院南京地理与湖泊研究所，湖泊与环境国家重点实验室，南京 210008； ²中国科学院大学，北京 100049； ³南京师范大学生命科学学院，南京 210046)

出版日期:2015-03-10 发布日期:2015-03-10

Quick implementing of generalized additive models using R and its application in blue-green algal bloom forecasting.

DENG Jian-ming^1,2, QIN Bo-qiang^1**, WANG Bo-wen³

(¹State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing 210008, China; ² University of Chinese Academy of Sciences, Beijing 100049, China; ³College of Life Science, Nanjing Normal University, Nanjing 210046, China)

Online:2015-03-10 Published:2015-03-10

摘要/Abstract

摘要：

广义可加模型(generalized additive models，GAM)适用于响应变量与解释变量之间的关系是非线性或非单调的数据分析，近年来在生态学中受到越来越多的关注。本文利用太湖湖泊生态系统研究站的监测数据，在RStudio集成编程环境下通过R对数据进行预处理、确定连接函数、对模型进行筛选以及评估等步骤，运用GAM分析了环境因子对微囊藻生物量的影响。结果表明，水温、总磷、化学需氧量以及电导率是影响太湖微囊藻生物量的4个关键环境因子。分析过程中，通过自编函数能有效地减少编程过程中的代码输入工作；R+RStudio是高效、快捷的编程环境。

关键词: 食谱分析, 产量结构, 分蘖动态, 鱼活动, 叶片氮含量

Abstract: Generalized additive models (GAM) is applicable to analyze nonlinear and nonmonotonic relationships between the response variable and explanatory variables. Increased attention has been paid to GAM in ecology study recently. In this study, longterm monitoring data from Lake Taihu were used to analyze the effect of environmental factors on Microcystis biomass. Data screening, link function confirmation, model selection and assessment were carried out by R. The results indicated that water temperature, total phosphorus, chemical oxygen demand and electrical conductivity were the main environmental factors that affected Microcystis biomass in Lake Taihu. The use of selfcoding functions can be effective in reducing code numbers. In addition, R+RStudio might be the optimal programming environment.

Key words: tillering dynamics, yield components, leaf nitrogen concentration, fish activity, diet analysis

邓建明1,2,秦伯强1**,王博雯3. 广义可加模型在R中的快捷实现及蓝藻水华预测分析中的应用[J]. 生态学杂志.

DENG Jian-ming1,2, QIN Bo-qiang1**, WANG Bo-wen3. Quick implementing of generalized additive models using R and its application in blue-green algal bloom forecasting.[J]. cje.

[1]	刘晓萌,于凌飞,黄耀,孙文娟,刘超,陈健,胡正华. CO₂浓度升高下粳稻叶片光合作用对光强变化的响应 [J]. 生态学杂志, 2018, 37(4): 1051-1057.
[2]	宁金花1,3,陆魁东2,3**,宋忠华1,张艳贵1,周玉1,3. 超级双季晚稻产量构成与温光因子的关系 [J]. 生态学杂志, 2015, 34(3): 688-694.
[3]	宁金花1，2,陆魁东2**,霍治国3,解娜1,黄晚华2. 拔节期淹涝胁迫对水稻形态和产量构成因素的影响 [J]. 生态学杂志, 2014, 33(7): 1818-1825.