首页关于本刊影响因子及获奖投稿须知订阅及广告专辑与专题学术会议绿色发表通道优秀论文 
说明:
此论文已被《生态学报》录用,现已优先出版,可查看文章PDF版本!
张力,闫文德,郑威,刘益君,梁小翠,高超,方晰.基于GA-BP人工神经网络的樟树林土壤呼吸对施氮响应的研究.生态学报,2017,(16).http://dx.doi.org/10.5846/stxb201605230990  
基于GA-BP人工神经网络的樟树林土壤呼吸对施氮响应的研究
Responses of Camphor forest soil respiration to nitrogen addition based on GA-BP network
投稿时间:2016-05-23  最后修改时间:2017-03-15
DOI: 10.5846/stxb201605230990
关键词土壤呼吸  施氮  GA-BP人工神经网络  响应曲面法  
Key Wordssoil respiration  nitrogen addition  GA-BP network  corresponding surface method
基金项目国家林业公益性行业科研专项(201404316);林业科技创新平台运行补助项目(2016⁃LYPT⁃DW⁃069);湖南省自然科学创新研究群体 基金(湘基金委字[2013]7 号);国家林业局软科学研究项目(2013⁃R09);湖南省教育厅一般项目(15C1431);湖南省研究生科研创新项目 (CX2015B296);中南林业科技大学研究生科技创新基金(CX2015B17);城市森林生态湖南省重点实验室资助
作者单位E-mail
张力 中南林业科技大学 woshizl1989@126.com 
闫文德 中南林业科技大学 生命科学与技术学院 csfuywd@hotmail.com 
郑威 广西壮族自治区林业科学研究院;广西壮族自治区林业科学研究院  
刘益君 中南林业科技大学 生命科学与技术学院  
梁小翠 中南林业科技大学  
高超 中南林业科技大学  
方晰 中南林业科技大学  
摘要点击次数 113
全文下载次数 27
摘要:
目前开展的施氮对土壤呼吸影响研究大多基于实验观测结果,受实验地自然条件的限制,不能研究在一定条件范围内土壤呼吸对施氮响应的连续变化。通过喷洒NH4NO3水溶液,设置对照(C ,no N added),低氮(L,5 g Nm-2 a-1),中氮(M,15 g N m-2 a-1),高氮(H,30 g N m-2 a1)4种处理水平,使用 GA-BP人工神经网络建立樟树林土壤呼吸对施氮响应的模型,并将模拟结果使用响应曲面法展示,研究土壤呼吸对施氮响应的变化。研究结果表明,施氮对樟树林土壤呼吸既有抑制作用又有促进作用,其程度是由土壤温湿度条件决定的,总体上使得施氮对土壤呼吸在低土壤湿度的条件下主要表现为促进作用,在高土壤湿度条件下主要表现为抑制作用,在一部分土壤温湿度组合下表现为无明显作用。GA-BP人工神经网络模型以其特性,可以模拟土壤呼吸对施氮响应的连续变化,并在一定程度上解释了施氮量、土壤呼吸、土壤温度和土壤湿度之间复杂的数学关系。
Abstract:
The current study on responses of soil respiration to nitrogen addition mostly based on the observations. Restricted by natural conditions between, it can''t study in certain conditions within the scope of soil respiration to the continuous variation of nitrogen response. A simulated nitrogen deposition experiment has been conducted in Camphor forest, which located in Hunan Forest Botanical Garden, subtropical China between May 2010 and June 2012. Soil respiration rate was measured twice a month under four levels of N treatments. By spraying NH4NO3 aqueous solution, this research set control (C, no N added), low nitrogen (L, 5 g Nm-2 a-1), middle nitrogen (M, 15 g N m-2 a-1), high nitrogen (H, 30 g N m-2 a-1) 4 kinds of processing level, using GA - BP artificial neural network to establish forest soil respiration of nitrogen addition model, and the simulation results showed by the corresponding surface method. The results showed that affected by the control factors such as solar radiation, precipitation, vegetation types and soil properties, the soil respiration showed a significant seasonal variation, the maximum value of soil respiration appeared in June and August, and the minimum value appeared in January and March, and nitrogen addition had an effect on soil respiration rate, but the seasonal dynamics of soil respiration was not changed. The application of nitrogen addition could not only inhibit the soil respiration, but also promote the soil respiration, and the degree determined by the conditions of soil temperature and humidity, soil respiration showed a significant increase, significantly decreased and no significant change of 3 in the total nitrogen application, under different conditions of soil temperature and humidity. The lower soil respiration rate mainly appeared in the lower soil temperature region, while the higher soil respiration rate mainly appeared in the region with higher soil temperature and humidity. At the maximum value of the soil respiration, the soil temperature was significantly affected by the effect of nitrogen addition, and the soil moisture change was not significant. But Soil temperature was not significantly(P>0.05)affected by nitrogen application when the soil respiration was minimum, however in this situation, the soil moisture behavior a significant change(P<0.05).The soil respiration in different humidity range was decreased with the amount of nitrogen addition increased. Nitrogen generally inhibited soil respiration, but this was the result of the interaction between nitrogen application on soil respiration and the promotion of two kinds of interaction, nitrogen addition to soil respiration under low soil moisture mainly to promote, under high soil moisture is mainly characterized by inhibition, under the combination of soil temperature and humidity of has no obvious effect. GA-BP artificial neural network model with its characteristics can be simulated soil respiration of nitrogen addition of the continuous changes and in a certain extent explains the complex mathematical relationships between the amount of nitrogen addition, soil respiration, soil temperature and soil moisture. At the same time, the input data of the model can be adjusted at any time according to the experimental conditions, so as to provide a great flexibility for the study of the relationship between soil respiration and multiple factors.
HTML 查看全文   查看/发表评论  下载PDF阅读器

您是本站第 57755479 位访问者

Copyright © 2005-2019   京ICP备06018880号
地址:北京海淀区双清路18号
  邮编:100085    电话:010-62941099
  E-mail : shengtaixuebao@rcees.ac.cn
本系统由北京勤云科技发展有限公司提供技术支持