文章摘要
刘佳凤,田娜娜,赵玉杰,周其文,刘潇威,袁旭,郭新蕾.基于Cubist多元混合回归的稻米富集Cd模型构建研究[J].农业环境科学学报,2018,37(6):1059-1065.
基于Cubist多元混合回归的稻米富集Cd模型构建研究
Evaluation of cadmium accumulation in rice using a Cubist multivariate mixed regression model
投稿时间:2017-12-21  
DOI:10.11654/jaes.2017-1740
中文关键词: 稻米    Cubist
英文关键词: rice  cadmium  Cubist
基金项目:国家重点研发计划项目(2016YFD0800301-01);农产品质量安全环境因子风险评估重大专项(GJFP2017012);贵州省重大科技专项([2013]6024)
作者单位E-mail
刘佳凤 沈阳农业大学土地与环境学院, 沈阳 110866
农业部农产品质量安全环境因子控制重点实验室, 天津 300191 
 
田娜娜 沈阳农业大学土地与环境学院, 沈阳 110866
农业部农产品质量安全环境因子控制重点实验室, 天津 300191 
 
赵玉杰 农业部农产品质量安全环境因子控制重点实验室, 天津 300191 yujiezhao@126.com 
周其文 农业部农产品质量安全环境因子控制重点实验室, 天津 300191  
刘潇威 农业部农产品质量安全环境因子控制重点实验室, 天津 300191  
袁旭 贵州省草地试验推广站, 贵阳 550000  
郭新蕾 农业部农产品质量安全环境因子控制重点实验室, 天津 300191  
摘要点击次数: 2206
全文下载次数: 2820
中文摘要:
      为了给农产品质量安全预测提供技术支撑,基于土壤环境因子对稻米富集镉(Cd)的影响在不同条件下存在差异性这一原理,以湖南省典型区域为例,在土壤多参数检测获取基本大数据前提下,采用Cubist多元混合线性回归技术,构建了研究区土壤-稻米Cd传输模型。结果表明,在所筛选的23个土壤理化参数中,Ca、Fe、S、Cd、Cl和pH 6个参数对稻米富集Cd影响最为突出,并被纳入Cubist模型中。土壤Ca(以CaO计)与土壤pH在Ca含量小于1%时线性显著相关,S与土壤有机质(SOM)呈指数显著相关。土壤pH对稻米富集Cd的影响在不同值阈范围内差异明显。研究表明,混合线性回归技术克服了传统多元回归仅能构建单一性质模型的缺陷,可有效筛选及表征环境因子对稻米富集Cd的影响,是构建土壤-稻米Cd传输模型有效方法。
英文摘要:
      To provide technical support for ensuring the quality and safety of agricultural products, the influences of soil environmental factors on the accumulation of cadmium(Cd) in rice under different conditions were studied. Based on the base date and soil parameters determined for the Hunan Province, a Cubist multivariate mixed linear regression model of Cd for soil-rice systems was built. The results suggested that among the 23 soil parameters evaluated the soil environmental factors of Ca, Fe, S, Cd, Cl, and pH had a significant influence on the enrichment of Cd and were therefore incorporated into the Cubist model. The Ca concentration(calculated using CaO concentration) had significant correlation with pH when the Ca content of the soil was less than 1%. In addition, S was significantly correlated with the soil organic matter(SOM) value. The influence of soil pH on the Cd concentration in rice differed among different thresholds. The Cubist mixed linear regression model overcame the shortcomings of multivariate regression, which can only be used to construct a single model. The Cubist model can effectively screen and characterize the effect of environmental factors on Cd accumulation in rice. Hence, it is an effective method for constructing a soil-rice Cd transport model.
HTML    查看全文   查看/发表评论  下载PDF阅读器