【作者】 李冰；

【导师】 叶志鸿；

【作者基本信息】 中山大学 ， 环境科学， 2012， 博士

【摘要】 水稻（Oryza sativa L.）对甲基汞（Methylmercury, MeHg）（一种神经毒素）具有很强的积累能力，人为活动造成越来越多的稻田受到汞（Mercury, Hg）污染，从而使得稻米遭受Hg/MeHg污染，影响人类健康。如何管理和利用受Hg污染的稻田，减少水稻地上部（尤其是米粒）总汞（Total Mercury, THg）和MeHg的积累是亟需解决的粮食安全和生产问题。目前尚未有一种廉价高效的修复方法适用于大面积受Hg污染的水稻田。已有室内实验显示，不同基因型水稻对Hg的吸收有很大差异，这暗示可通过Hg低积累基因型的选用来达成减少水稻对THg/MeHg积累的目的。但目前相关研究，尤其是机理的研究鲜有报道。因此，本论文在野外调查的基础上，通过野外大田和室内盆栽实验相结合，比较不同基因型水稻对THg/MeHg的吸收差异及其稳定性，并从外部因素（土壤因子）和内部因素（水稻渗氧—铁膜）去揭示这些差异形成的原因。研究结果将为筛选和培育对THg/MeHg低积累、高耐性的水稻基因型去解决Hg污染稻田的利用和保证稻米安全提供理论依据和方法。主要研究结果如下：1、调查了贵州、湖南和广东三省16个地区的稻田（13个为矿区，3个为非矿区），其中，13个矿区的稻田全部遭受不同程度Hg污染，最严重的贵州铜仁Hg矿区稻田THg浓度达到136.5mg kg^-1。从三个省的污染区来看，贵州省调查区稻米THg平均浓度为28.6ng g^-1，湖南省为20.5ng g^-1，广东省为5.98ng g^-1，其中，稻米THg浓度最高达到54.4ng g^-1（贵州丹寨Hg矿区），是食品安全标准(20ng g^-1)的2倍以上。2、在贵州万山、湖南新晃和广东凡口Hg污染稻田种植26个水稻基因型，研究其对THg和MeHg的吸收特征（首次在野外开展这方面研究）。结果发现，在三块实验田，大米对THg和MeHg的积累都存在显著（p <0.01）的基因型差异。在万山、新晃和凡口，大米THg平均浓度分别为25.8ng g^-1、28.6ng g^-1和14.5ng g^-1；大米MeHg平均浓度分别为9.4ng g^-1、11.4ng g^-1和1.2ng g^-1。在污染比较严重的万山和新晃实验田，中花11、日本晴、越光、老黄稻、南丰糯和苏玉糯6个基因型THg浓度是低于国家食品安全标准限量的。在凡口实验田，有19个基因型大米THg浓度是在安全标准之下的。大米THg浓度在万山和新晃（r=0.81, p <0.0001），万山和凡口（r=0.64, p <0.001），凡口和新晃（r=0.76,p <0.0001）都呈现一个极显著的正相关；同样，谷粒产量在这三个实验田也呈现两两显著（p <0.05⁰.01）正相关；大米MeHg浓度在万山和新晃（r=0.67, p<0.001）呈现极显著正相关。日本晴和南丰糯这两个基因型同时低积累THg和MeHg。3、在广东凡口实验田，存在严重的复合重金属污染。因此，本次工作也研究了种植在凡口实验田的32个水稻基因型对毒性元素镉（Cd）、铅（Pb）和矿质营养元素铁（Fe）、锌（Zn）、镍（Ni）和锰（Mn）的吸收情况及其相互影响。结果发现：水稻对这6种元素的吸收都存在显著（p <0.001）的基因型差异。大米中Fe和Mn浓度随着Cd浓度的增加而显著降低；Fe、 Mn和Ni浓度随着Pb浓度的升高而增加，Ni和Cd也存在相同的趋势。五丰优2168、天优196和桂农占这三个基因型能同时低积累Cd和Pb，且有较高的矿质元素含量和谷粒产量。4、利用来自11个地方（贵州5个：万山四坑、万山五坑、铜仁、丹寨和三都；湖南5个：新晃、保靖、机建、茶树喇和牛斗坪；广东1个：凡口）的Hg污染稻田土，研究4个水稻基因型对THg和MeHg的吸收情况及土壤因子对其吸收THg和MeHg的影响。结果发现，种植在不同土壤的同一基因型水稻组织对THg和MeHg吸收存在显著（p <0.01）差异。种植在相同土壤的不同基因型水稻组织对THg和MeHg的吸收也存在显著（p <0.01）差异，即水稻对THg和MeHg的吸收同时受到基因型和不同来源土壤（具有不同性质）的影响，此外，还受到二者交互作用（p <0.001）的影响。土壤因子中，有机碳（TOC）、土壤Fe和Ni浓度与水稻THg浓度呈显著（p <0.05⁰.01）负相关；而土壤总钾（TK）和THg浓度与水稻THg浓度呈显著（p <0.05⁰.001）正相关。土壤pH和水稻MeHg浓度呈显著（p <0.05⁰.01）负相关。5、采用贵州万山Hg污染稻田土结合河砂进行根际袋土培实验，研究4个水稻基因型在不同生长期（分蘖期、拔节期、孕穗期、扬花期、灌浆期和成熟期）水稻根系渗氧和根表铁膜的动态变化，水稻组织对THg和MeHg的吸收变化及其相互作用。结果发现，4个基因型在不同生长期根系渗氧率和根表铁膜（Fe浓度）都存在显著性差异（p <0.001）。天优196、华优广抗占、南粳35和紫香糯生长期平均渗氧率分别为5.92mmol O2kg^-1root d. w. h^-1、5.27mmol O2kg^-1root d. w. h^-1、8.48mmol O2kg^-1root d. w. h^-1和3.4mmol O2kg^-1root d. w. h^-1；根表铁膜平均Fe浓度分别为20474mg kg^-1、16979mg kg^-1、24224mg kg^-1和14952mg kg^-1。4个基因型都呈现根系渗氧率随着水稻的生长逐渐降低的趋势。在水稻生长前三个阶段（分蘖期、拔节期和孕穗期），根表铁膜随着时间显著增加；到后三个阶段（扬花期、灌浆期和成熟期），根表铁膜变化较小。4个基因型茎叶THg和MeHg浓度都是随着生长逐渐降低，根THg浓度在前四个时期是逐渐增加，后两个时期变化幅度不大；根MeHg浓度从分蘖期到拔节期有微弱的上升后就逐渐降低；相关分析发现，4个水稻基因型根系渗氧率和根表铁膜呈现显著（p<0.01⁰.0001）正相关。水稻根系渗氧率、根表铁膜与茎叶、根THg浓度都呈显著（p <0.05⁰.0001）负相关。扬花期的花序，灌浆期的谷粒、成熟期的谷壳和糙米THg浓度与相应时期和分蘖期的根系渗氧率和根表铁膜都是呈现显著（p<0.01⁰.0001）负相关。虽然不同基因型水稻在不同生长期其组织MeHg浓度都存在显著性差异，但是MeHg浓度与根系渗氧和根表铁膜都没有显著相关性。以上野外大田实验和室内盆栽实验结果都显示水稻对THg和MeHg的吸收都存在显著的基因型差异，在严重的Hg污染稻田中，依然存在稻米THg和MeHg浓度是在食品安全限量以下的基因型。且水稻对THg的吸收具有基因型稳定性，对MeHg的吸收虽然受到环境的影响程度高于THg，但在土壤污染程度相似的贵州万山和湖南新晃依然具有同步的吸收特征。一些基因型（日本晴和南丰糯）能同时低积累THg和MeHg；同样也存在对复合重金属（如Cd、Pb）同时低积累的水稻基因型。本研究还表明，土壤TOC、Fe、Ni和K浓度能显著影响水稻对THg的吸收；土壤pH能显著影响水稻对MeHg的吸收；土壤Hg污染程度是影响水稻吸收THg和MeHg的重要因素。说明水稻对THg和MeHg的吸收是受到土壤因子（外部因素）和基因型（内部因素）共同影响的。对水稻内部因素—渗氧的研究表明，在不同生长期，不同基因型水稻根系渗氧能力显著不同，且与水稻组织THg浓度呈显著负相关，说明水稻渗氧能力不同是导致水稻对THg吸收存在显著基因型差异的内部因素之一。此外，孕穗期是水稻对Hg（尤其是MeHg）吸收的一个关键时期，一些潜在的可降低水稻对THg/MeHg积累的方法（如水分管理）如果在该时期进行，可能效果会更好。总的来说，本研究表明通过筛选基因型和改变土壤因素能降低水稻对THg和MeHg的积累，这是一种潜在的解决Hg污染稻田的利用和保证稻米安全的有效且可行的方法。更多还原

【Abstract】 Mercury （Hg） is a toxic element, and is now considered a key global pollutant. Incertain environmental conditions （e.g. paddy field）, Hg was methylated tomethylmercury （MeHg） which is more toxic （neurotoxin） than the inorganic form.Nowadays, more and more paddy fields have been contaminated by Hg due toanthropogenic Hg emission, resulting elevated Hg levels in rice and subsequently inhumans foods posing human health threat. Recent studies have demonstrated that riceseed has the highest ability to accumulate MeHg compared to other tissues. There istherefore urgency in management and utilization these paddy fields contaminated byHg, to develop mitigation measures to reduce total mercury （THg） and MeHg in ricegrain and to ensure food safety. However, cost-effective methods are not available toremediate large areas of Hg-polluted paddy fields. Prevoius studies reported that therewere significant differences in THg and MeHg uptake and tolerance among ricegenotypes grown in soils added with Hg under greenhouse conditions. The resultssuggest that it is possible to screen rice genotypes to ensure rice safety inHg-contaminated fields. However, currently the genetic controls of THg and MeHg accumulation in rice remain largely unknown. In this study, field and greenhouseexperiments have been conducted to investigate the genotype variation and stablity ofrice in THg and MeHg accumulation, and to explore the mechanisms of this variationfrom the aspects of radial oxygen loss （ROL） and iron plaque. The effects of edaphicfactors in THg and MeHg accumulation in rice genotyes were also investested inpresented study. The major results and conclusions are as follows:Firstly, THg concentration in soils of paddy fields and rice grains collected from13mines sites and3control sites across Guizhou, Hunan and Guangdong Provinces wereinvestigated. The results showed that （1） the paddy fields in the13mines sites havebeen seriously contaminated by Hg, up to136.5mg kg^-1in paddy soils at Tongren（TR） Hg mine site in Guizhou Province;（2） the rice was also polluted by Hg, meanTHg concentrations of rice grain in Guizhou, Hunan and Guangdong were28.6ng g^-1,20.5ng g^-1and5.98ng g^-1, respectively. The highest concentration of THg in ricegrain was54.4ng g^-1（Danzhan Hg mine of Guizhou） which is exceed the permissablelimit (20ng g^-1) of the Maximum Levels of Contaminants in Foods of China （MLCFGB2762-2005）.Secondly, twenty-six rice genotypes were cultivated in three paddy fieldscontaminated Hg across Guizhou （Wanshan Hg mine）, Hunan （Xinhuang Hg mine）and Guangdong （Fankou Pb/Zn mine） Provinces. Results showed that there wassignificant genotype variation （p <0.01） for grain THg and MeHg at the three fieldsites. The mean THg and MeHg concentrations of rice grain in Wanshan, Xinhuangand Fankou were25.8ng g^-1,28.6ng g^-1and14.5ng g^-1;9.4ng g^-1,11.4ng g^-1and1.2ng g^-1, respectively. At Wanshan and Xinhuang field sites (paddy soil THgexceeded40mg kg^-1in the two sites), there were6genotypes （Zhonghua11,Ribenqing, Yueguang, Laohuangdao, Nanfengnuo and Suyunuo） with the lower grainTHg which were under the permissable limit of MLCF. At Fankou site, concentrationsof THg in grains of19genotypes were lower than the permissable limit of MLCF. Forthe26genotypes at the three sites there were highly significant correlation betweenthe concentrations of THg between Wanshan and Xinhuang （r=0.81, p <0.0001）, Wanshan and Fankou （r=0.64, p <0.001）, Fankou and Xinhuang （r=0.76, p <0.0001）. Moreover, there were also significant positive correlations （p <0.05⁰.01）between grain yields at the three field sites. A significant positive correlation betweenthe concentrations of MeHg between Wanshan and Xinhuang （r=0.67, p <0.001）.Two rice genotypes （Ribenqing and Nanfengnuo） were identified with a combinationof low grain THg and MeHg.Thirdly, the paddy fields in Fankou Pb/Zn mining area were co-contaminated by acocktail of mixed toxic heavy metals （Hg, Cd and Pb）. This study indentified ricegenotypes with both low Cd and Pb accumulation under Cd-and Pb-contaminatedfield conditions, and the interactions of toxic elements Cd and Pb with micronutrientelements iron （Fe）, zinc （Zn）, manganese （Mn） and nickel （Ni） were also studied.Among32rice genotypes tested, there were significant differences （p <0.001） inconcentrations of6elements of brown rice. Significant decreases in concentrations ofFe and Mn were detected with increasing Cd concentrations and a significantelevation in Fe, Mn and Ni with increasing Pb concentrations. A similar result wasalso shown by Cd and Ni. Three genotypes were identified with a combination of lowbrown rice Cd and Pb, high micronutrient and grain yield （Wufengyou2168, Tianyou196and Guinongzhan）.Fourthly, four rice genotypes were cultivated in11soils contaminated by Hgcollected from different paddy sites （10Hg mine and1Pb/Zn mine across threeprovinces）. The pot experiment was conducted under greenhouse condition toinvestigate effect of genotypes and edaphic factors on THg and MeHg accumulationin rice. Results indicated that genotypes as well as locations （soil sources） had asignificant effect on grain THg and MeHg. Moreover, there was also a significantgenotype by location interaction for rice THg and MeHg. There were significantnegative correlations between rice THg and concentrations of total organic carbon（TOC）, Fe and Ni in soil; a significant positive correlation was observed between riceTHg and soil total potassium （TK）. There was a negative correlation between riceMeHg and soil pH. Finally, a pot trial was conducted using soil-sand combination rhizosbag systemwith four rice genotypes, to study dynamic variation of radial oxygen loss （ROL） andiron plaque on root surface, and THg/MeHg uptake in six rice growing stages（tillering, elongation stage, booting, flowering, filling stage and mature stage）. Resultsshowed that there were significant differences （p <0.01） in rates of ROL amongdifferent rice genotypes, and among different growing stages. There were alsosignificant variations for degree of iron plaque formation on root surface amongdifferent genotypes and different growing stages. The mean rates of ROL in the fourgenotypes （Tianyou196, Huayouguangkangzhan, Nanjing35and Zixiangnuo） were5.92mmol O2kg^-1root d. w. h^-1、5.27mmol O2kg^-1root d. w. h^-1、8.48mmol O2kg^-1root d. w. h^-1and3.4mmol O2kg^-1root d. w. h^-1in six rice growing stages,respectively. The mean concentrations of Fe on root surface （DCB-Fe） of the fourgenotypes were20474mg kg^-1、16979mg kg^-1、24224mg kg^-1and14952mg kg^-1,respectively. Straw THg/MeHg and root MeHg were gradually decline in rice growingperiod. However, root THg was gradually increased from tillering to flowering.There was a significant negative correlation （p <0.05⁰.0001） between rate of ROLand rice THg, similarly, a significant negative correlation （p <0.05⁰.0001） betweeniron concentration on root surface and rice THg was also found. However, there wereno significant correlations between rice MeHg and rate of ROL, iron concentration onroot surface.In summary, present results showed that there were significant genotype variationsfor rice THg and MeHg concentrations under both field and greenhouse conditions.Some rice genotypes accumulated lower THg and MeHg in grains （under thepermissable limit of food safety） even if they were cultivated the paddy fieldsseriously contaminated by Hg. Moreover, rice genotypes tended to stabilize in THgaccumulation under different field conditions, and in MeHg uptake in similarenvironment conditions. Ribenqin and Nanfengnuo were potential useful genotypeswith simultaneously low grain THg and MeHg. Similarly, Wufengyou2168, Tianyou196and Guinongzhan were identified with a combination of low grain mixed toxic heavy metals （Cd and Pb）. The degree of soil Hg-contamination is an important factorfor THg and MeHg accumulation in rice. Some edaphic factors （e.g., TOC, Fe, Ni andK） had significant effects on rice THg, and soil pH on rice MeHg accumulation. Theresults from rhizobag trial suggest that ROL-iron palque combination play animportant effect on THg accumulation in rice. Booting is a key stage for Hg（especially MeHg） accumulation in rice growing period, imply that some measures（e.g. water management） could be carry out to mitigate THg and MeHg accumulationin rice at this stage. As a whole, present results indicate that selection of rice genotypeand control of edaphic factor were effective ways in reducing THg and MeHgaccumulation in rice, this may have potential to be applied in Hg-contaminatedregions.更多还原