学位论文题目：《兴都库什－喜马拉雅地区红豆杉属植物物种界定与保护遗传学研究》

论文摘要：

红豆杉属（Taxus L.）的物种是第三纪孑遗植物，也是兴都库什－喜马拉雅地区[Hindu Kush-Himalaya(HKH)]经济价值最高和受威胁最大的植物之一。喜马拉雅和青藏高原的隆升导致局域气候变化及其对物种演化历史的影响已受到广泛关注。然而，对气候变化和人类活动对该地区植物演化动力和脆弱性的影响却知之甚少。深入理解第四纪冰期对现存物种分布格局的影响，将有助于我们预测濒危物种红豆杉的可能响应策略和命运。为了对该地区红豆杉属植物进行有效保护和管理，迫切需要对该属植物开展分类学、谱系地理学和居群遗传学的研究。本研究在居群水平上利用形态学、叶绿体DNA（cpDNA）、核糖体DNA（nrDNA）、微卫星标记和气候数据对这一地区红豆杉属植物开展了相关研究，主要研究结果和结论如下：

1. 兴都库什-喜马拉雅地区红豆杉属植物的分类和物种分布区的准确界定

本研究澄清了兴都库什－喜马拉雅地区（HKH）分类地位不明确，形态鉴定困难的红豆杉属植物的物种数目及其在尼泊尔的准确分布区界线。在大规模居群采样的基础上，利用形态学、分子生物学和气候数据对红豆杉属植物开展了综合的分类学研究。通过对兴都库什-喜马拉雅地区及邻近地区790份样品（包括47份馆藏标本）的27个形态性状的主成分分析（PCA），发现该地区分布有密叶红豆杉T. contorta，喜马拉雅红豆杉T. wallichiana和南方红豆杉T. mairei三个物种。基于33条ITS序列和36条trnL-F序列的系统发育分析，上述三种红豆杉分别构成单系分支并得到强烈支持。对三种红豆杉生态位的一致性和背景数量化检测表明，三个物种具有不同的气候偏好：如密叶红豆杉更喜欢干燥寒冷的环境，喜马拉雅红豆杉则为湿润寒冷，而南方红豆杉的生境为湿润温暖。此外，基于11个生物气候因子变量的分布区模拟分析，其结果与三个物种的地理分布区相一致，即密叶红豆杉分布于西喜马拉雅地区（阿富汗东部到西藏西南部的吉隆），喜马拉雅红豆杉则分布于东喜马拉雅地区（从尼泊尔中部至云南西北部），而南方红豆杉分布于中喜马拉雅及印度-缅甸地区。本研究的结果表明红豆杉属植物在尼泊尔中部平行分布，但其分布区并不重叠，喜马拉雅红豆杉的分布范围比通常认为的要狭窄。南方红豆杉分布于这一地区为首次发现，该种曾被认为只分布于中国南部地区。同时描述了鉴别分布于该地区的三种红豆杉的分类检索表。

2. 基于cpDNA和nrDNA数据的谱系地理学研究

利用采自兴都库什-喜马拉雅地区43个红豆杉居群的叶绿体trnL-F 和核基因ITS片段的序列开展该地区红豆杉属植物的谱系地理学研究。从776个个体的trnL-F片段中得到22个cpDNA单倍型，而在235个个体的ITS序列中发现了8个nrDNA单倍型。除了2个杂交个体外，物种间无叶绿体单倍型/核单倍型共享。三个物种的存在得到了cpDNA和nrDNA数据的支持。基于trnL-F 片段进化速率的估算，密叶红豆杉大约在5.9 - 9.3百万年前 (mya)与喜马拉雅红豆杉和南方红豆杉分开，而后两者的分化时间大约在3.9 - 5.9 mya。中新世-上新世喜马拉雅和青藏高原隆升及随后的气候和生态环境变化对这一地区红豆杉属植物的分化产生了重要影响。在物种水平上，密叶红豆杉（GST = 0.085, NST = 0.056, p = 0.056）和喜马拉雅红豆杉（GST = 0.260, NST = 0.283, p = 0.283）单倍型的分布没有显著的谱系地理学结构。和同属的其他物种相比，密叶红豆杉和喜马拉雅红豆杉的遗传多样性较低。尽管在这两个物种具有显著的居群结构，但其遗传分化系数低于同属其他物种。遗传多样性高的地区，如巴基斯坦北部的兴都库什地区，尼泊尔中部以北地区以及中国西藏南部的吉隆一带可能是密叶红豆杉的第四纪冰期避难所。东喜马拉雅的东部（西藏东南部）地区和尼泊尔东部地区也呈现出较高的遗传多样性，他们可能是喜马拉雅红豆杉的冰期避难所。在兴都库什-喜马拉雅地区，密叶红豆杉和喜马拉雅红豆杉经历了不同的扩张历史。密叶红豆杉可能在更新世晚期（130 kya）发生过居群数量的扩张，而在全新世（6 kya）发生了一次空间扩张。基于失配分布分析和中性检测，密叶红豆杉在末次冰盛期（LGM）可能发生了严重的瓶颈效应，少数居群在避难所中存活下来；在末次冰盛期后，从避难所经历快速的居群扩张（奠基者效应），迅速回迁到现代的分布区。然而在东喜马拉雅地区，喜马拉雅红豆杉可能在95.8 kya 前先发生居群数量的扩张，接着在85.4 kya 经历空间扩张。根据单倍型结构和第四纪冰期规模的研究结果，表明喜马拉雅红豆杉可能受末次冰盛期的影响较小，很多居群在不同的避难所存活，并在冰期和间冰期的反复作用下经历了多次扩张-收缩过程，可能频繁发生不同避难所居群间的基因交流。遗传多样性沿西藏东南部至尼泊尔中部逐渐递减，暗示了喜马拉雅红豆杉很可能从东喜马拉雅向西沿喜马拉雅山向尼泊尔中部地区迁移。

3. 基于微卫星标记的居群遗传学分析

利用10对微卫星标记对兴都库什-喜马拉雅地区的42个居群共761个个体进行了居群遗传学研究。遗传多样性分析表明，三个物种分别形成明显不同的遗传实体，在物种水平上具有显著遗的传分化（FST = 0.514 P < 0.01）。在尼泊尔中部发现三个种间具有强烈的地理和生态障碍。该地区的红豆杉属植物在物种水平具有中等的遗传多样性水平，密叶红豆杉具有较高的遗传多样性(NE = 2.510, HO = 0.221, HS = 0.480, HT = 0.588)，喜马拉雅红豆杉的遗传多样性次之(NE = 1.747, HO = 0.290, HS = 0.402, HT = 0.442) ，而南方红豆杉的最低(NE = 1.516, HO = 0.271, HS = 0.260, HT = 292)。红豆杉属的三个物种在居群间均具有明显的居群结构。密叶红豆杉的遗传分化最高(FST = 0.196, GST = 0.183, G'ST = 0.405)，其次为南方红豆杉(FST = 0.152, GST = 0.107)，而喜马拉雅红豆杉的遗传分化最低(FST = 0.113, GST = 0.089, G'ST = 0.171)。与同属其它物种相比，这种较低的遗传分化可能是由于景观效应，生境片段化和遗传漂变等所导致。遗传距离和地理距离在密叶红豆杉中无相关性，而在喜马拉雅红豆杉中却显著相关。居群遗传多样性沿东喜马拉雅到尼泊尔中部逐渐递减，表明喜马拉雅红豆杉很可能从东向西迁移到达尼泊尔中部地区。微卫星分析的结果和cpDNA的结果高度一致，密叶红豆杉, 喜马拉雅红豆杉和南方红豆杉之间遗传多样性的差异可能由于它们在兴都库什-喜马拉雅地区分别经历了不同的第四纪冰期作用和不同程度的人类活动影响所致。

4. 兴都库什-喜马拉雅地区红豆杉属植物的保护

本研究界定的三种红豆杉的分布区不连续，形态区别明显并具有不同进化历史，在尼泊尔中部异域分布。谱系地理学和居群遗传学的研究结果表明，与同属其他物种相比，分布于该地区的三种红豆杉具有较低的遗传多样性。由于红豆杉属植物沿兴都库什-喜马拉雅山系分布在的不同国家，因此每个物种都可看做是一个独立的显著进化单元。基于本研究的结果，我们建议沿兴都库什-喜马拉雅地区分布的三种红豆杉属植物应首先采取就地保护的策略，居群具有高遗传多样性的地区可作为就地保护的优先保护区域。而且很多红豆杉居群已在兴都库什-喜马拉雅地区的保护地区。然而，对保护策略和管理还需进一步完善，特别要控制过度放牧以促进资源再生。我们特别推荐利用森林群落学的方法促进不同隔离居群之间的基因流。此外，迁地保护可作为就地保护的有力补充，包括活体植物、种子和遗传材料的种质资源收集和保存应该优先开展，对那些具有高遗传多样性和私有单倍型的居群和个体都应给予特别关注。红豆杉属植物各种规模的人工种植是实现减少因日趋增长的对紫杉醇需求对野生居群采集压力的有效途径。本文也对不同国家不同情况下几种可能的保护措施进行了详细讨论。
关键词：保护遗传学，兴都库什-喜马拉雅地区、谱系地理学，物种界定，红豆杉属 

ABSTRACT：Yews the Tertiary relic species are one of the economically most important and at the same time most threatened plants along the Hindu Kush-Himalaya (HKH) region. The uplift of the Qinghai-Tibet Plateau (QTP) and its effects on local climate and the evolutionary history of species have been widely appreciated. However, our understanding on the evolutionary dynamics of plants, their vulnerability from climate change and human activities along the southern belt of the HKH region is very limited. Detailed understanding on the effects of Quaternary glaciation on the pattern of current distribution enables to predict the possible response or fate of highly endangered species such as Taxus in future. Therefore, to maximise the impact of conservation and management efforts, more insights into the taxonomy, phylogenetics and population genetics of yews are needed. Thus, this study has investigated yews populations along the HKH region using morphology, cpDNA, nrDNA, microsatellite markers and climatic data. The major findings are given below.

1. Taxonomy of yews and their precise boundary in the HKH region.

Taxonomic uncertainties regarding the number of morphologically identifiable taxonomic units occurring in the HKH region and their precise boundaries in Central Nepal was clarified in this study. An integrative taxonomic approach was adopted, using sets of morphological, molecular and climatic data generated predominantly from fine-scale collections. A principal component analysis performed using 27 morphological characters on 790 individuals, including 47 herbarium specimens, representing the entire distribution range of yews throughout the HKH and adjacent regions revealed three distinct species of Taxus to occur there, namely T. contorta, T. mairei and T. wallichiana. Phylogenetic analyses based on 33 ITS and 36 trnL-F sequences showed three distinct highly supported monophyletic clades, representing the three species. Identity and background tests quantitatively defined the ecological niche of the three yew species and suggested different climatic preferences for each species: cold and dry for T. contorta, cold and wet for T. wallichiana and warm and humid for T. mairei. Moreover, distribution modeling of the climatic preferences based on eleven bioclimatic variables corresponded well with the geographic distribution of the three yews, i.e. T. contorta in the western Himalayas [E. Afghanistan to SW Xizang (Tibet)], T. wallichiana in the eastern Himalaya (C. Nepal to NW Yunnan) and T. mairei in the lesser Himalaya and Indo-Burma region. This study unraveled a distinct non-overlapping parallel distribution of the Taxus species in Central Nepal, indicated a much less widespread distribution of T. wallichiana than generally assumed and reported for the first time the presence of T. mairei in the HKH region, the latter a species previously considered to occur exclusively in South China. A taxonomic key based on morphological characters found to be suitable for the reliable separation of the three species was given.

2. Phylogeography of yews in the HKH region based on cpDNA and nrDNA

The phylogeography of the yews occurring in the HKH region was investigated using chloroplast DNA (cpDNA) trnL-F and nuclear ribosomal DNA (nrDNA) ITS sequences generated from 43 populations distributed throughout the study area. A total of 22 cpDNA haplotypes and 8 nucleotypes were identified from 776 trnL-F and 235 ITS sequences respectively. Both markers revealed strong phylogenetic structure among the haplotypes of T. contorta, T. mairei and T. wallichiana. None of the haplotypes / nucleotypes were shared between the species, except for two hybrid plants. Three distinct species were defined by both the cpDNA and nrDNA data. Based on the mutation rate estimation of trnL-F, T. contorta diverged from T. wallichiana / T. mairei about 5.9 - 9.3 million years ago (mya), and the later two species diverged around 3.9 - 5.9 mya. The uplift of the QTP during the Miocene-Pliocene and subsequent climatic and ecological changes can be attributed to the divergence of the yews in the HKH region. At the species level, no significant phylogenetic structure was detected in the populations of T. contorta and T. wallichiana (T. contorta: GST = 0.085, NST = 0.056, p = 0.056; T. wallichiana: GST = 0.260, NST = 0.283, p = 0.283). A low genetic diversity was observed in T. contorta and T. wallichiana compared to other congeneric species. Similarly, despite a significant population structure detected among the populations in these two species, their genetic differentiation was lower than that of congeneric species. Areas with a high genetic diverstiy in the Hindu-Kush region of Pakistan and North of Central Nepal and South Xizang (Jilong) of China imply possible Quaternary refugia for T. contorta. The East of the eastern Himalayas (Southeast Xizang) and East Nepal also showing a high genetic diversity, might represent Quaternary glacial refugia for T. wallichiana. The two species experienced different expansion histories in the HKH region. For T. contorta, a demographic expansion likely happened in the late Pleistocene (130 kya), and a spatial expansion took place in the Holocene (6 kya). Based on mismatch distribution analysis and neutrality tests, T. contorta possibly underwent severe bottleneck events during the LGM and survided in refugia. After the LGM, it experienced a sudden expansion and quickly colonized its current distribution range. While for T. wallichiana, demographic expansion 95.8 kya was followed by a spatial expansion 85.4 kya in the eastern Himalaya. The observed haploytype structure and studies on the extent of Quaternary glaciation in eastern Himalaya suggests that T. wallichiana did not experience serious impacts during the LGM. Populations of this species likely survived in different refugia and experienced several expansion / contraction cycles during the glacial and inter-glacial cycles with gene exchanges between populations. The gradual decrease of haplotype diversity from Southeast Xizang to Central Nepal suggested that populations of T. wallichiana possibly migrated westward along the East Himalaya to Central Nepal.

3. Population genetics of yews in the HKH region based on microsatellite markers

Population genetic analyses were conducted using 10 nuclear microsatellite loci on 761 individuals representing 42 Taxus populations of yews across the HKH region. The genetic diversity was found to be structured at the species level (FST=0.514 P < 0.01), and three distinct genetic entities were observed representing T. contorta, T. mairei and T. wallichiana. Strong geographic and ecological barriers played crucial roles in Central Nepal for the lack of a distinct admixture zone previously suggested. A moderate species-level genetic diversity was observed in T. contorta (NE = 2.510, HO = 0.221, HS = 0.480, HT = 0.588), followed by T. wallichiana (NE = 1.747, HO = 0.290, HS = 0.402, HT = 0.442) and T. mairei (NE = 1.516, HO = 0.271, HS = 0.260, HT = 292). A significant population structure among populations of the three Taxus species was detected. Taxus contorta showed the highest genetic differentiation (FST = 0.196, GST = 0.183, G'ST = 0.405), followed by T. mairei (FST = 0.152, GST = 0.107), with the lowest in T. wallichiana (FST = 0.113, GST = 0.089, G'ST = 0.171). These values were low compared to those of congeneric species. This may be due to landscape effects, habitat fragmentation, and genetic drift. No significant correlation was detected between genetic and geographic distance for T. contorta, while a significant positive correlation was observed in T. wallichiana. Decreasing of population genetic diversity from eastern flank of Eastern Himalaya to central Nepal suggests the migration of T. wallichiana westwards to central Nepal from the eastern Himalaya (i.e. Southeast Xizang). The results obtained from microsatellite analyses were highly consistent with those for cpDNA, therefore the differences of genetic diversity among T. contorta, T. mairei and T. wallichiana could be a result of highly vulnerable and asynchronous Quaternary glaciation in addition to varying degree of impacts from anthropogenic activities experienced by each yew species distributed in the western and eastern Himalayas along the HKH region.

4. Conservation of Taxus species

This study has identified three spatially discrete and morphologically distinct and identifiable species of Taxus with different evolutionary histories, but now occurring in allopatry in Central Nepal. Phylogeography and population genetic investigations of the three species revealed a low level of genetic diversity compared with congeneric species. Given the distibution of yews in different nations along the HKH region, each species can be recognized as an independent evolutionary significant unit (ESUs). Based on the results of this study, we suggest that in-situ conservation should take priority for all three Taxus species along the HKH region. Populations that harbour high genetic diversity need to be conserved in-situ. Many populations of yews are already in conservation areas along the HKH region. However, an improved management particularly to control overgrazing and enhance regenerations are needed. Community forest approach is highly recomended to bridge gene flow between several fragmented populations. In addition, ex-situ conservation is an effective complemant to in-situ conservation. Germplasm preservation, including living plants, seeds and genetic material should be carried out with priority. Populations and individuals from areas with a high genetic diversity and private haplotypes should be taken into account. Small to large scale farming of Taxus is the only alternative to fulfil the growing demand for Taxol, the anticancer property isolated from the bark of yews and can reduce the pressure on wild populations. Several possible measures of conservation for each nation in different scenarios are discussed in detail. 

Keywords: Conservation genetics, Hindu Kush-Himalaya region, Phylogeography, Species delimitation, Taxus