【作者】 吴则焰；

【导师】 刘金福；

【作者基本信息】 福建农林大学 ， 生态学， 2011， 博士

【摘要】 水松为稀有的松柏纲乔木,是我国特有的单种属植物,国家一级保护树种。作为古老的活化石植物,水松起源于中生代,经第四纪冰期后残存,现已成为罕见孑遗树种,仅零星分布于我国南方部分地区,在研究杉科植物的系统发育、古植物学和第四纪冰期气候等方面都有重要科学价值。由于历史气候变迁、人为干扰、生境被占、演替压力等因素,目前水松分布范围不断缩小且呈片断化,天然种群数量剧烈下降,大部分水松分布地仅剩几株孤立木,这一古老孑遗植物面临灭绝危险,如何保护水松稀有珍贵基因与珍稀种质资源,扩大水松种群规模已成为当务之急,开展必要的保护生物学和濒危机制及种群恢复研究刻不容缓。本研究综合运用种群生态学、保护遗传学、繁殖生态学等理论与方法,较全面、系统地开展水松自然地理分布、种群数量特征、遗传多样性、扦插繁殖技术等方面研究。通过种群生态学特性和数量特征研究,分析种群结构和动态规律及演化趋势;通过种群遗传多样性与遗传变异研究,分析水松物种水平和种群水平遗传多样性,比较现有水松人工种群与天然种群在遗传多样性和遗传结构方面的异同,探讨人工种群遗传结构是否得到恢复;从繁殖过程中播种苗生长过程探讨水松幼苗生长规律,通过试验不同因素对水松扦插繁殖生根率及生根性状的影响,建立良好的水松扦插繁殖体系;揭示水松濒危机制并提出科学保护对策。主要研究结果如下:（1）运用植被与气候关系指标探讨水松地理分布与气候因子相互关系,揭示水松主要种群的地理分布规律。结果表明,水松分布范围为中亚热带北纬13°21’～37°36’,东经102°40’～121°29’,主要分布在珠江三角洲、长江中下游及闽江流域,垂直分布范围为25～2000 m,不同分布区垂直分布差异很大。水松地理分布对气候因子要求是温暖湿润,光照充足,适宜分布区气候指标范围为:年平均温度12.10℃～22.60℃,年降水量1199.60 mm～2231.00 mm,年平均相对湿度62.00%～82.00%,年平均日照时数1589.20 h～2781.60 h。气候水热指标对水松地理分布影响很大,按影响力大小排序依次为低温因子>相对湿度>高温因子>年日照时数>年平均降水量。（2）运用种群生态学原理和方法,分析水松种群结构和动态规律及其演化趋势。天然种群方面:①运用直径分布模型拟合方法,探讨水松天然林种群直径分布规律。主要树种胸径和地径分布用正态分布函数拟合效果最好,β分布次之,Weibull分布、Γ分布和对数正态分布拟合效果不理想。水松中老龄径材较多,伴生树种以中小径阶林木为主,正态分布参数与林分直径呈显著线性关系,用林分特征因子预估正态分布参数值十分理想。②运用“空间代时间”方法,以胸径结构代替年龄结构分析水松种群结构特征,结果表明年龄结构为倒金字塔型,呈现大树过熟、幼龄木数量极少,种群天然更新差的现象,说明该种群属于衰退型,稳定性差严重影响种群发展,年龄结构与水松生物学特性及其生境条件有关。③以种群生命表及生存分析理论为基础,运用“横向代纵向”方法编制水松种群静态生命表,分析各龄级个体死亡数量、存活率及后代生长趋势,绘制种群存活曲线、死亡率曲线、死亡密度曲线、危险率函数曲线。水松种群存活曲线属于Deevey-Ⅲ型,反映了水松种群数量动态变化趋势。死亡密度函数与危险率动态变化规律基本一致,累计死亡率单调递增,生存率单调递减,累计死亡率和生存率上升或下降幅度均表现出前高后低,死亡密度函数表现出死亡个体数随着龄级曲线起伏,呈现两端高、中间低趋势。④应用谱分析方法揭示水松种群更新动态过程及其稳定性特征,结果表明水松种群天然更新过程通过不同龄级的株数分布波动表现,呈现出大周期内有小周期的多谐波迭加特征。基波A1 = 1.752最大,反映了水松种群最基本的周期波动,小周期波动性与高生长特性基本一致。⑤运用聚集度指标法、Iwao方程评价水松种群空间格局及其动态变化规律,结果表明水松种群空间格局呈聚集分布,形成机制与样地所处的地理环境关系密切。不同取样尺度下的分布格局变化研究表明,不同取样单元上,水松种群均表现为集群分布,且在25 m²取样单元上聚集度最高。⑥运用有限空间种群增长的逻辑斯谛模型探讨水松种群基面积增长规律,通过改进单纯型法进行最优拟合。研究结果表明,洪-Logistic新模型比其它种群增长模型更符合水松种群实际增长趋势。该模型残差平方和Q为1.7831,内禀增长率为0.7604,特征返回时间Tr为1.3116径级年,最大增长速率出现在第8径级,即胸径为48～54 cm时期,平衡位置在环境容量为58.1991 m²/hm²处。可见水松种群增长较慢,平衡性脆弱,受破坏后恢复时间较长。⑦采用Lotka-Volterra竞争模型研究种间竞争关系,水松及其伴生树种优势度分别为63396 cm²·hm^-2和15877 cm²·hm^-2,相对优势度分别为79.97 %和20.03 %,竞争系数达0.8231,种内竞争大于种间竞争。预测平衡时水松与伴生树种的相对优势度分别为69.21%和30.79%,呈现共优状态,说明群落相对稳定,但有逐渐被伴生树种取代趋势。⑧运用R型聚类分析方法对水松种群的25个数量性状进行聚类分析,探讨水松数量性状间的相互关系及其对水松生长、遗传及经济性状的影响。结果表明:水松材积与树高、冠长、胸径等性状呈显著正相关,生长性状与光合作用面积有关;树高是决定材积的第一要素;水松在干形遗传、自然整枝性方面的遗传性较为稳定。⑨将分形理论与地统计学原理相结合,计算水松不同种源树高和胸径生长的分形维数,揭示其空间分布变异规律和分形特征。结果表明,水松不同种源胸径、树高生长特性的分维值分别为1.635和1.824,胸径分维值小于树高分维值。为反映水松种源的空间差异性,在评价水松种源时应选取胸径生长指标。人工种群方面:①运用5种概率分布模型对斗门水松人工同龄纯林种群直径分布进行拟合,比较不同概率分布的拟合效果,探讨其直径结构特征。结果表明,水松人工林中小径阶的林木占多数,处于森林发育前期阶段,林分直径呈对称分布,变动范围不大。Weibull分布模型的或然误差值更小,拟合精度更高,为最优拟合模型。变动系数、偏度、峭度等特征值符合人工同龄林生长特点。与天然种群相比,人工同龄纯林表现出特有的直径结构规律,说明不同群落类型、不同密度、不同平均直径和生长阶段的水松种群,其直径分布规律存在差异。②运用回归分析方法,研究各竞争指标与林木胸径生长量的相关关系,在选择适合竞争指标基础上,建立水松人工林单木生长模型。结果表明,林木间距离对单木生长的影响不明显,相对树高、相对冠幅和冠长率与胸径生长量关系密切,是其主要影响因素。分别建立水松人工林相对树高、相对冠幅及冠长率与平均胸高断面积生长量的拟合方程、单木生长模型和林分蓄积量模型,拟合精度高。③运用森林自疏过程中密度变化规律模型,研究水松人工林自然稀疏机制,建立了林分总断面积与种群密度关系模型、单株材积与种群密度关系模型、种群自然稀疏规律模型,回归拟合精度高,可应用于指导水松人工林经营过程的密度管理与密度控制。④拟合水松树高胸径曲线方程,较好反映了树高与胸径的正相关性。运用Weibull分布能较好地拟合胸径和树高分布,建立的树高Weibull分布模型拟合效果良好,能进一步对树高生长进行预测。⑤探讨了水松人工林群落稳定性,阐述了斗门水松林保护现状,提出相应的保护措施。（3）采用ISSR分子标记技术对水松主要分布区的8个种群共136个体进行遗传多样性分析。用10条随机引物共检测到95个扩增位点,其中多态位点37个,多态位点百分率（P）为38.95%。与其它濒危裸子植物相比,水松在种群水平和物种水平都具有较低的遗传多样性。不同种群间遗传多样性差别较大,多态位点百分率（P）、Nei’s基因多样度（He）、Shannon信息指数（I）的平均值分别为33.56%、0.1078、0.1902,He和I大小变化趋势与P大致相同;物种水平Ae、P、He和I分别为1.2819、38.95%、0.1608、0.2360。各种群间基因分化系数G_ST为0.3982,基因流Nm仅0.3778,表明水松种群间存在一定程度的遗传分化,但种群内个体间变异在其遗传变异中占主导地位,UPGMA聚类分析反映出水松种群遗传距离与地理距离存在正相关性。水松具有较低的遗传多样性水平,与其地理环境与气候变迁以及人为活动干扰密切相关。人工种群遗传多样性明显低于天然种群,种群间遗传分化不如天然种群强烈。4个天然种群的P、He、I平均值分别为39.32%、0.1499和0.2202,远远高于人工种群平均值26.98%、0.1113和0.1558,天然种群遗传分化系数及平均遗传距离(G_ST = 0.4513,D = 0.0301)明显高于人工种群(G_ST = 0.2836,D = 0.0192)。遗传多样性时空变异研究表明,在一定范围内,水松种群遗传多样性随年龄增长而增加,随海拔升高而增加。基于水松种群遗传学和生态学的研究结果,提出应加大对遗传多样性高的水松种群保护力度,加强水松种群间基因交流,以最大限度地保存水松资源的遗传多样性。（4）运用扦插繁殖技术,探讨不同激素种类、激素浓度、处理时间、不同扦插基质及插穗的选择等因素对水松扦插生根率及生根性状的影响,建立较好的水松扦插繁殖体系,为水松无性系规模化生产提供技术保障。总结出水松扦插繁殖的最佳技术组合:于3月初选择生长健壮、母株年龄相对小的水松植株,取枝条基部1 a生以上枝条制穗,去除下部叶片,并使切口平滑下斜且靠近节基部,插条长度约8～12 cm,保留顶芽及少量侧枝,随采随插随处理,剪好的插条要下部齐整捆好,放入0.3%的多菌灵药液中进行消毒,用浓度为150 mg/L的ATB1浓液浸泡插穗2 h;扦插于泥炭土∶珍珠岩（2∶1）混合基质,株行距为3 cm×4 cm,扦插深度以顶端留3～4 cm为准,插后及时浇水、防晒、杀毒、灭菌。采用本技术,水松扦插生根率可达39.89%。从繁殖过程中播种苗生长过程探讨水松幼苗生长规律,利用Logistic方程拟合水松播种苗苗高、地径生长节律,相关指数均在0.95以上。水松苗高生长量遵循“S”型规律,地径生长量变化与苗高呈现正相关。（5）通过宏观和微观方面综合研究,阐述水松濒危特征及生态适应性,揭示其濒危机制。水松濒危是各方面因素综合作用的结果,是时间和空间长期的渐变过程。从现象上看可能是人类活动造成的生境片断化,滥砍滥伐导致水松数量减少,实质上是其演化过程中气候剧烈变化、遗传多样性低下、生境破坏、生态因子胁迫和树种竞争等多因素综合结果。水松濒危原因表明,该物种的保护是一个复杂系统工程,应综合考虑各种因素对其造成的影响,全面系统地开展水松珍稀资源的保护与恢复工作。更多还原

【Abstract】 Glyptostrobus pensilis（Staunt.）Koch,now unique in the Glyptostrobus,is at the risk of becoming extinct in the wild.It has been listed as a first-class threatened plant species,one of handful species of highest conservation concern in China.As a living fossil plant,G. pensilis was widely distributed in the Northern Hemisphere.After the Quaternary glaciation,because of natural geographical factors and human activities,majority populations have extincted except Southeast China.G. pensilis has important scientific value in many research fields such as the phylogenesis of family Taxodiaceae,paleobotany and the Quaternary glacial climate.Dues to the climatic change,human interference,habitat destruction,succession pressure and other factors,the distribution range of G. pensilis is still shrinking and become isolated and scattered.The population size are decreasing sharply,G. pensilis is in danger of extinction.How to protect rare gene resources,and expand population size of G. pensilis becomes an urgency matter.Therefor,research on conservation biology and recovery assessment of G. pensilis aims to provide necessary data for the establishment of scientific and effective conservation strategy for this endangered species.This study focus on nature geographic distribution,population quantitative characteristics,genetic diversity and cutting propagation technique of G. pensilis systematically and comprehensively from the views of population ecology,reproductive ecology and conservation genetics.Research on population ecology and quantitative characteristics could reveal population structure,dynamic law and trends of evolution of G. pensilis.ISSR markers was used to detect genetic diversity and genetic composition of wild and artificial G. pensilis populations to compare the differences of genetic diversity and genetic structure between wild and artificial G. pensilis populations,evaluate recovery genetic consequences of G. pensilis.The seedling growth rule was discussed through experiment of sowing,and the cuttage propagation techniques of G. pensilis was also explored by testing various factors.The endangered mechanism of G. pensilis was discussed and some suggestions on species recovery and conservation strategies were provided.The conclusions are as follows:（1）The relationship between geographical distribution and climate factors was discussed using relation index of vegetation and climate.In China,G. pensilis was disconnectedly distributed in Pearl River delta,the middle and lower reaches of Changjiang River and Minjiang River.It mainly distributed in north latitude 13°21’～37°36’ and longitude 102°40’～121°29’,in the altitude ranged from 25 m to 2000 m.It requires moist but well drained soil and a warm,sunny position,where annual mean temperature varied from 12.10℃～22.60℃,the annual mean precipitation varied from1199.60 mm～2231.00 mm,the annual mean relative humidity varied from 62.00%～82.00% and the annual average sunshine hours varied from 1589.20 h～2781.60 h.Low temperature,humidity,high temperature,sunlight and precipitation were the most important factors that affect the distribution of G. pensilis.（2）By using the principles and methods of population ecology,the population structure and dynamic law of G. pensilis were analyzed.Several aspects were discussed about the wild populations.①Five probability models were applied to analyze the structure features on diameter distribution of main tree populations in G. pensilis forest.The results showed that normal distribution model was the best fitness to express the structure of diameter distributions of G. pensilis population,which suggested that the range of diameter distributions were wider and the number of mid-size wood was in the majority.The fitting effect of Weibull distribution model,Γdistribution model and lognormal distribution model were not so perfect.There was obvious linear relationship between the parameters of normal distribution and tree diameter,and the model was effective by using forest characters to forecast the parameter values of normal distribution.②The population age structure characteristic of G. pensilis was studied using space for time method.The result showed that populations had opposite pyramid age structure,which had abundant old trees,relatively few seedlings and young trees.The age structure of G. pensilis in the community was declining,and ability to resist disturbance and regenerate was weak.Biological characteristics and habitat conditions were the critical elements for the age structure of G. pensilis.③In order to determine the population state of G. pensilis,analyze its past population structure and pre-disturbance situation and forecast its future population dynamics,the population life table were analyzed by space deducing time method.The static life table,survivorship curve,mortality rate curve and danger rate were compiled using the diameter at breast height class to represent the age class structure and smooth out technique.The survival cure of the population belonged to the type of Deevey-Ⅲ,and the dynamics trend showed that the mortality rate of population was high in early age,while the accumulative mortality and livability had emerged after the former high-low trend.The death density function had emerged high on edges and low in the center.④Spectral analysis of the population dynamics was made to revel the renewal process and stability of G. pensilis.The natural renewal process of G. pensilis population could be represented by the fluctuation of trees number distribution in different age grades . The natural renewal of G. pensilis indicated harmonics superposition characteristics,which large cycle contains small cycle.Minor cycle was coincident with the periodic fluctuation of annual growth of G. pensilis.⑤The spatial distribution pattern of G. pensilis population was determined using for aggregate indices and Iwao’s equation,and the causes of the evolved pattern was also determined.The results showed that the spatial distribution of G. pensilis population presented an aggregating distribution,and the degree of diffusion was related to the different habitats.The effects of different scales showed the highest aggregating intensity when quadrat was 25 m².⑥The growth law of basic area of G. pensilis population was studied by the methods of Logistic model.And a new model was optimized from the modified simplex method.The optimized results showed that the Hong-Logistic model was more suitable than other models to simulate the actual growth trend of G. pensilis population.The residual sum of squares was 1.7831,and the intrinsic growth rate was 0.7604,and the Tr value was 1.3116 diameter class year.The highest growth rate of basal area was the eighth age classes,namely the phase when breast diameter was 48～54 cm,and the basal area reached 58.1991 m²/hm².It was thus clear that the growth of G. pensilis population was retarded,and it needed long time to recover after suffering damages.⑦The competition of G. pensilis and its associated species were discussed with the Lotka-Volterrra equation.The result showed that dominant degree of G. pensilis and its associated species was 63396 cm²·hm^-2 and 15877 cm²·hm^-2,respectively.Interspecific competition modulus was 0.8231 by calculating importance value.The relative dominant degree of G. pensilis and associated species was 69.21% and 30.79% respectively when reaching their balance.The mixed trees could jointly dominate this community,and the whole community was controlled by G. pensilis population.But the number of G. pensilis would be constantly decline.⑧The method of R-cluster analysis was used to analyze the 25 quantitative traits cluster of G. pensilis population and the interaction of the quantitative traits.The quantitative traits’infection to growth,heredity and economic character of G. pensilis were also discussed.The results showed that the volume and tree height,crown length,diameter at breast height were significantly correlated.The growth traits were correlate with photosynthesis area,and tree height was the first element of volume,and the heredity was stable in tree shape and self-pruning.⑨The DBH and tree height mean growth of G. pensilis from different provenances were studied by combining geographical statistical methods with fractal theory.The fractal dimensions of DBH and tree height growth of G. pensilis were calculated in order to revealed the rule of spatial distribution variation.The results showed that the fractal dimensions of DBH and tree height were 1.635 and 1.824, respectively.The fractal dimensions of DBH was lower than fractal dimensions of tree height.The diameter breast height should be selected as evaluating indicator of different provenances.Several aspects were discussed about artificial population.①Five probability models were applied to analysis the structure feature on chest height diameter distribution of G. pensilis-aged and pure plantation.The results showed that the majority were middle and small diameter class.The population was currently in the primary stage of forest development,the symmetrical distribution of diameter change within a certain range.Weibull distribution model was the best fitting model for the smaller error probability and more accurate fitting.The eigenvalue such as coefficient of variation,skewness,kurtosis consistent with the characteristics of the artificial aged and pure plantation.The artificial population embodied special law on diameter distribution compared to natural populations.It showed that the populations different in communities,density,average diameter and growth phase had differences in diameter struture feature.②The interrelated relation between competitive index and growth of breast-height diameter was analyzed with method of regression analysis.The competitive index reflecting objectively the competition and growth pattern of G. pensilis plantation was selected,and stand growing models were established.The results showed that the relationship between diameter growth and distance was very weak,the main key elements composing competition index were relative tree height,relative crown and rate of crown length.The individual tree growing models can fitting the grow rule of G. pensilis accurately.The factors such as average height,average breast diameter,growing stock of sub-compartment and basal area would be estimated by using the growth model and estimation methods.③The self thinning model of G. pensilis population was studied by regression analysis.The relationship between total stand basal area and population density,and the relationship between singleplant volume and population density were also studied.The accuracy of models were significantly high and had a good prospect in plantation management density control.④In order to forecast accurately height of each diameter class of G. pensilis,the height-diameter model was established.The Weibull function was chosen to develop height-diameter models.By verified,this equation might be applied to estimate the tree height growth of G. pensilis.⑤The community stability of G. pensilis was discussed and conservation strategy was suggested according to the protection status of G. pensilis artificial population in Doumen.（3）Samples from the main distribution regions of G. pensilis were analyzed by ISSR（Inter simple sequence repeats）molecular marker.10 ISSR primers were used to amplify 136 individuals of 8 populations,a total of 95 discernible DNA fragments were detected.Of these,37（P = 38.95%）were polymorphic loci , which indicated that very low level of genetic variation existed in the populations.The difference of the genetic diversity among all the populations was evident.The percentage of polymorphic loci（P）,Nei’s gene index（He）and Shannon information index（I）were 33.56%、0.1078 and 0.1902,while those at the species level averaged 38.95%、0.1608 and 0.2360,respectively.Moreover,the genetic differential index(G_ST = 0. 3982)indicated that there was strong genetic differentiation among populations.The dendrograms of genetic relationships among populations was constructed based on Nei’s genetic distance.The mean genetic parameters of wild populations（P = 39.92%,He = 0.1499,I = 0.2202）were much higher than artificial populations（P = 26.98%,He = 0.1113,I = 0.1558）.Coefficient of gene differentiation and genetic distance of wild populations(G_ST = 0.4513,D = 0.0301)were much higher than artificial populations(G_ST = 0.2836,D = 0.0192).The change of genetic diversity had obvious positive correlation with population age and elevation in specific limit.The results indicated that although the quantity has been successfully restored,the genetic diversity and genetic structure of G. pensilis have not recovered appropriately,given the loss of genetic variation and low genetic diversity in artificial populations.The combined results of the analysis of population genetic structure and community investigation suggested that establishing a conservation spot in the optimal habitat of G. pensilis,and to cross-transplant adult plants or seedlings mutually among populations in order to enhance the gene flow.By these means,the genetic diversity resources of the species can be preserved to the greatest extent.（4）The cuttage propagation techniques of G. pensilis was explored by testing various factors including hormone types,soaking time,medium and concentrations and cuttings.The best way for cottage of G. pensilis was that cuttings came from the top part of 1-years-old branch,cuttings soaked in ATB1 150 mg/L solution for 2 h before inserting in the medium with combination of peat∶pearlite = 2∶1,spacing 3 cm×4 cm.The depth of the cutting insert for 3～4 centimeters distance to the top.Water the cuttings in time and keep sunshine resistant,antisepsis,sterilization and so on.The rooting ratio of cutting under these treatments above could reach 39.89%.The tree height and ground diameter growth rhythms of G. pensilis sowing seedling were fitted using Logistic equation to study seedling growth rule.The correlation coefficients was above 0.95.The seedling growth pattern displayed“S”curves for height while ground diameter growth had positive correlation with height. （5）The endangered characteristics,ecological suitability and endangered mechanism of G. pensilis was discussed after macroscopic and microscopic analysis.It was the process of temporal and spatial change and had been resulted from different factors.On the surface,the habitat fragmentation that human activities caused and deforestation was important factors threatened this species.Essentially,it resulted from the comprehensive effects of human activities and natural factors,such as drastic climate changes , low genetic diversity , habitat destruction , ecological stress factors and species competition.The endangered reasons of G. pensilis indicated that species conservation was a complex and systematic project.It should consider all the various factors to carry out the relevant protection and population recovery of G. pensilis systematically and comprehensively.更多还原