【作者】 廖慧敏；

【导师】 吴超；

【作者基本信息】 中南大学 ， 安全管理工程， 2010， 博士

【摘要】 转基因植物大面积商业释放会给生态环境带来潜在而巨大的风险。目前对转基因植物商业释放的风险分析与安全评估研究还处于初级阶段,缺乏行之有效的安全评价方法用于生产实践。为建立一套转基因植物生态环境风险评估体系,本文对其风险分析与安全评价方法进行了初步探索,其研究内容主要包括以下几方面：(1)运用安全系统工程中的因果分析法与事故树分析法,结合转基因植物生态风险问题进行了系统风险源辨识。根据因果分析图与层次分析法对风险主干因素的分析,结果显示基因漂流是引起转基因植物生态环境风险的首要主干原因；而后运用事故树分析对处于底层的基本原因事件进行分析,其中自播转基因植株在结构重要度分析中占有重要地位。据此风险源辨识结果将基因流与自播植株的生态风险作为转基因植物风险分析与安全评价的重点。(2)利用复杂系统的网络拓扑结构分析与社会网中的小团体理论,构建了转基因植物基因流风险的网络分析法。该方法主要包括转基因植物基因流网络构建、网络结构特性分析、小团体分析与结构同型性分析四个基本步骤。以1989-2008年二十年间的常规油菜与十字花科植物杂交文献实验数据为基础对该方法进行了实例演示,研究表明油菜基因流网络的节点度服从幂律分布,具有无标度特性,网络特性分析显示该网络在受到随机攻击时具较好的鲁棒性,而受到恶性攻击时具极弱的抗攻击性,小团体和结构同型性分析可将网络中298个节点22类十字花科植物划分为2个小团体和5类结构角色,其中甘蓝型油菜在油菜基因流网络小团体中具有关键性作用。这些演示结果表明基因流风险的网络分析法可有效进行转基因植物基因流风险规律划分。(3)在基因流风险的网络分析基础上,为进一步细化转基因植物与同源物种之间风险规律关系,运用TOPSIS排序法与粗糙集综合权重确定法相结合,构建了转基因植物基因流风险排序集成模型。基于甘蓝型油菜与其近缘种间的杂交组合研究文献与《中国植物志》中甘蓝型油菜的近缘植物的生物学特性描述,运用该模型对甘蓝型油菜与其近缘植物的基因流风险排序进行了演示。结果显示该模型能将甘蓝型油菜与其近缘植物之间的基因流风险按从大到小或从小到大的规律进行一一排序,进一步深化了基因流风险的网络分析法结论。(4)利用模糊集重心理论与熵值权重确定法相结合,构建了转基因植物营养生长速率等级安全评价模型。模型包含方差分析、建立营养生长速率等级分类标准值、评价指标正态隶属函数确立、模糊集重心理论计算指标重心向量集、熵值法计算评价指标权重向量和营养生长速率等级划分等六个基本步骤。并利用湖南省具代表性的芸苔属作物杂交组合和营养生长指标实测值对该模型进行了演示。论证了该模型能较好区分芸苔属作物各物候期的营养生长速率的安全等级。(5)基于模糊数学中模糊聚类分析与模糊模型识别,并结合生物统计学相关知识,构建了转基因植物基因流安全等级评价模型。该方法包含转基因植物种间组合亲和性的模糊聚类分析、模糊聚类结果F检验、种间组合风险等级划分和风险等级模式识别四个基本步骤。并在等级模式识别步骤中针对样本数据收集情况的不同,构建了不同的等级识别方法,与传统聚类的硬性划分相比更具灵活性。以常规油菜杂交研究文献的实验数据对该方法进行了演示,将这些杂交组合聚类形成六个风险等级模式,并用研究文献资料中的一杂交组合数据为待识别样本进行风险等级识别,识别结论与文献结论相符,论证了该方法具可操作性。最后,基于风险规律与安全评价方法构建的基础上,提出了转基因植物风险控制的网格化管理模式和生物安全立法建议。更多还原

【Abstract】 With the commercial release of transgenic plants in large range, there are many potential and large risks of ecological environment. Currently, the risk analysis and safety assessment methods research of transgenic plants are still at the primary stage, and there are almost not effectual methods applied to agriculture practice. To build a system of risk analysis and safety assessment methods of transgenic plants, some preliminary study is developed in this paper, which the main contents were as following:Firstly, based on the combination of the causality of analysis with the fault tree analysis, the risk sources of transgenic plants were system idetified. According to the analysis results of causality analysis, gene flow was the most important reason among transgenic plants ecological environment risks. The basic events were analysised by fault tree analysis, in which, volunteer crops played an important role in structure importance. According to the result of system identification, the gene flow and volunteer crops were two main research contents in this paper.Secondly, to understand the gene flow rules of transgenic plants, the gene flow risk complex network analysis method was built based on the theory of complex network and small group. The method included four steps:transgenic plants gene flow network construction; network structure characteristics analysis; small group analysis and structure isomorphism analysis.The method was demonstraded based on the documents of normal rapeseed hybridizing with cruciferae from 1989 to 2008. The transgenic rapeseed gene flow topological graph was constructed and the structural characteristics were analyzed. The result showed that node degree of network followed power-law distribution and had the scale-free properties. The network robustness was analyzed from two aspects of random attacks and intentional attacks. Analysis showed the gene flow network was robust against random failures of nodes but fragile to intentional attacks when the removal of node no more than 10% nodes. At last, the subgroup and structural position were analyzed, which showed 22 kinds cruciferae plant were divided into two subgroups and five structural positions, and Brassica napus played a key role in subgroups and structural position. These research results can provide new approach for transgenic plants gene flow research.To further study on the risk rule of the transgenic plants based on the gene flow risk complex network analysis method, the gene flow risk ranking integration model was built based on TOPSIS ranking and rough set theory. Then, the model was demonstrated by using the cross combination documents and the biological characteristics data in China flora, the result showed the model can preferably identify the rank between Brassica napus and cruciferae and further deepened the results of the complex network analysis method.Two aspects research were mainly to be developed on safety assessment method. Firstly, to predict the law of vegetative growth of transgenic plants and study corresponding agricultural cultivation method when they compete with volunteer crops, an evaluation model of transgenic plants was proposed base on theory of barycenter of fuzzy set and entropy method. The model mainly included four steps:The normal membership function establishment of evaluation indexes, calculating index barycenter vector set by theory of barycenter of fuzzy set, calculating index weight by entropy method and identifying the vegetative growth rate rank. Subsequently, the evaluation model was demonstrated by using the cross data of normal rapeseed. These results proved that the model proposed could identify plants vegetative growth rate rank in various phonological phase and was operational. Furthermore, the agriculture activities of transgenic plants could be inducted by it. Secondly, to predict the environmental risk of the gene flow of transgenic plants after commercialization, a protocol for assessing its compatibility risks was proposed by the fuzzy clustering analysis and fuzzy model identification along with the biometric theory. The protocol included four steps:the fuzzy clustering analysis of interspecific crossing compatibility of transgenic plants, the F test of the outcome of the fuzzy clustering analysis, ranking of the risk of the interspecific crossing compatibility, and identifying the risk ranks. At the final stage of the protocol, various methods for identifying risk ranks were proposed to accommodate different data collection situations, therefore this approach was more flexible than the rigid ranking by the traditional clustering analysis. Subsequently, the protocol was demonstrated by using the the documents of normal rapeseed, and six ranks of risks were established, and the risk rank of the interspecific compatibility was identified successfully in a crossing data of Brassica napus. These results proved that the protocol proposed was operational.At last, based on these methods built above, the grid management pattern and some legislative suggestions of transgenic plants risk control were put forward.更多还原