齿肋赤藓 (Syntrichia caninervis Mitt) 隶属于丛藓科赤藓属，是典型的耐旱耐寒藓类植物，在我国青藏高原、宁夏、新疆及毗邻的中亚部分地区均具有广泛分布。青藏高原的快速隆起及其引起的西北地区干旱化，显著影响了本地区植物的分布格局和群落结构，并逐渐成为了亲缘地理学研究的热点地区之一。然而，现有的大部分研究多集中在二倍体或多倍体的种子植物上，作为单倍体且以无性繁殖方式为主的藓类植物，在第四纪剧烈的地质和气候环境变迁中，是否具有和种子植物一样的遗传结构和种群动态历史，还缺乏系统性研究。因此，本研究以齿肋赤藓为研究对象，以青藏高原、帕米尔高原、西北及中亚干旱区为研究区，基于分子生态学、生态学、亲缘地理学和生态位模型等相关理论，重点探讨了齿肋赤藓在不同时空尺度下的遗传变异、谱系分化及种群动态历史，并初步分析了生境干扰对苔藓植物生长和遗传多样性的影响，为深入了解物种遗传多样性维持和物种形成机制提供了重要理论依据。主要研究结论如下： (1) 齿肋赤藓种群的遗传多样性及遗传结构 综合利用叶绿体片段 (psbD-trnT和trnL-rpL32) 和核微卫星 (19对) 数据对27个齿肋赤藓种群的遗传多样性和遗传结构进行了检测，研究结果显示，帕米尔高原 (hd = 0.635 ± 0.039; π= (1.110 ± 0.242)×10-3; uh = 0.570 ± 0.014) 及中亚 (hd = 0.696 ± 0.083; π= (0.980 ± 0.152)×10-3; uh = 0.521 ± 0.020) 分布区齿肋赤藓种群的遗传多样性显著高于青藏高原-腾格里区 (hd = 0.568 ± 0.089; π= (2.070 ± 0.538)×10-3; uh = 0.389 ± 0.033)。但青藏高原台面的唐古拉山南缘 (hd = 0.727; π= 5.200×10-3; uh = 0.520) 和柴达木盆地北缘 (hd = 0.564; π= (2.850 ± 0.538) ×10-3; uh = 0.540) 仍具有和帕米尔高原及中亚地理组相似的遗传多样性，这意味着这些区域可能是齿肋赤藓在青藏高原台面上的种源地。此外，基于叶绿体单倍型和微卫星数据的Structure聚类分析显示，齿肋赤藓遗传组分可被分为3个不同的地理组分，即中亚区 (CA)、帕米尔高原-天山区 (PT)、青藏高原-腾格里区 (QT)。Mantel test结果显示，不同齿肋赤藓种群间存在明显的谱系地理结构 (Nst = 0.369，Gst = 0.312，P < 0.05)。然而，对三个地理区的分层次AMOVA分析显示，齿肋赤藓的遗传变异主要发生在种群内部和不同地理组内的种群间，对各单倍型SAMOVA的分组和Bayes构建的一致树并没有将各单倍型很好的分组到各区域。同时，在整个分布区均有共享单倍型的分布，这意味着齿肋赤藓在整个分布区可能存在一定程度的基因流和繁殖体的远距离传播，近距离种群间的遗传分化和遗传结构可能由种群所处的微生境或气候因素造成。 (2) 齿肋赤藓的谱系结构与种群动态历史 通过筛选获得psbD-trnT和trnL-rpL32两个多态性较好的叶绿体片段，对27个齿肋赤藓种群的330个个体进行了序列变异分析，最终检测到15个单倍型。通过Network构建的单倍型谱系关系，可以明显地将中亚和青藏高原区域分开，帕米尔高原是二者的连接区域，这与分布区的地理分布结构相一致。基于Beast的分子钟分析，初步估算出各单倍型分化时间大约在0.2-6 Ma，主要集中在2 Ma年左右。最古老单倍型主要分布在青藏高原中东部的念青唐古拉山和横断山区，而次古老单倍型主要分布在帕米尔高原。结合不同齿肋赤藓种群之间的谱系结构、共享单倍型和特有单倍型分布，我们推测齿肋赤藓可能先由青藏高原中东部地区向帕米尔高原迁移，随后又由帕米尔高原在西风控制下向东北和东南两个方向迁移，腾格里沙漠的部分单倍型可能同时由青藏高原和帕米尔高原种群远距离传播而来。同时，基于齿肋赤藓物种分布信息和末次间冰期以来的气候数据，利用最大熵模型对齿肋赤藓的自末次间冰期以来的种群分布动态进行了模拟。结果显示，与大部分研究者对喜温种子植物在冰期出现大范围收缩、退居冰期避难所的研究结果不同，齿肋赤藓可能在冰期来临的时候出现了种群扩张，并在间冰期或冰期后喜温植物扩张时出现种群收缩。这为耐旱藓类植物在响应第四纪冰期、间冰期的剧烈环境变化下的种群动态历史提供了新证据。 (3) 微生境和干扰对齿肋赤藓生存的影响 上述结果显示微生境和干扰的变化可能会影响齿肋赤藓的生长和遗传多样性，为验证生境对齿肋赤藓的影响，我们在古尔班通古特沙漠进行了定点研究。结果显示，人为干扰和低降水可能会在一定程度上降低齿肋赤藓的遗传多样性。在荒漠地区水分是植物生长的主要限制性因子，降水可能会通过影响种群大小来影响种群遗传多样性。同时，短时间内灌丛的完全丧失可能会严重威胁积雪融化后其下苔藓植物的生长。灌丛的丧失会显著降低融雪后干燥期齿肋赤藓的叶绿素荧光活性和可溶性蛋白的含量，尤其对于灌丛完全丧失的种群。在资源短缺状态下，较渗透调节物质、POD和CAT而言，SOD可能是齿肋赤藓在抵御胁迫时更重要的抗氧化酶。但冬季的积雪会为苔藓植物的越冬提供保护。这侧面回应了苔藓植物在冰期缓慢降温过程中扩张的可能性。 (4) 齿肋赤藓叶绿体基因组结构 为更为准确地获取齿肋赤藓在叶绿体基因上的突变位点，基于高通量测序技术对齿肋赤藓叶绿体全基因组进行了测序、拼接和结构分析，结果显示：齿肋赤藓叶绿体基因组大小为123,131 bp，GC%含量为28.33%。其中反向重复区IR为9,965 bp，小单拷贝区SSC为18,539 bp，大单拷贝区LSC为84,663 bp。共包含130个基因，其中8个rRNA基因、37个tRNA基因和83个编码序列 (CDS)。值得注意的是，与其他湿生藓类和种子植物不同，齿肋赤藓叶绿体基因组缺少编码在植物光合作用电子传递中扮演重要作用的petN基因，这可能与其耐旱性进化有关。此外，齿肋赤藓叶绿体基因组含有SSR位点15个，通过6组齿肋赤藓叶绿体全基因组比对，获得潜在的碱基变异和缺失位点175处，其中碱基变异位点(SNP)139处，Indel位点36处，且主要发生在非编码区。为后续齿肋赤藓种群遗传多样性和遗传结构的分子标记选择提供了重要参考。为更好地了解苔藓植物的进化历史、苔藓植物的亲缘地理学研究奠定基础。

Syntrichia caninervis Mitt, Pottiaceae, which widely distributed in the arid northwestern region and the Qinghai-Tibetan Plateau, is a typical drought-tolerant plant. The rapid uplift of the Qinghai-Tibet Plateau and the aridification in the northwestern region caused by it have significantly affected the distribution pattern and community structure of the plants in the region, which has gradually become one of the hot spots in the study of phylogeography. However, most of the existing researches have focused on diploid or polyploid vascular plants, and the question has been unclear whether bryophytes that are haploid plants, has similar population historical dynamics with other vascular plants in the dramatic geological and climatic environment changes of the Quaternary. Therefore, this study uses S.caninervis as the research object, and the Qinghai-Tibetan Plateau, Pamirs Plateau and the arid region in Northwest China and central Asia as the research area, combined with theories of molecular ecology, ecology, phylogeography and ecological niche model, to investigate the genetic variation, lineage differentiation and population dynamic history of S.caninervis in different spatial and temporal scales. In the meantime, the effects of habitat disturbance on the growth and genetic diversity of bryophytes were analyzed. The study would provide an important theoretical basis for further understanding of species genetic diversity maintenance and species formation. The main research conclusions obtained are as follows: （1）Genetic diversity and structure of S.caninervis populations Based on two chloroplast fragments (psbD-trnT and trnL-rpL32) with high polymorphisms, and 19 pairs of microsatellite, the genetic diversity and genetic structure of 27 populations of S. caninervis in the distributed region were analyzed. The results show that the genetic diversity of S.caninervis in Pamirs-Tianshan(hd = 0.635 ± 0.039; π= (1.110 ± 0.242) ×10-3; uh = 0.570 ± 0.014) and Central Asia (hd = 0.696 ± 0.083; π= (0.980 ± 0.152) ×10-3; uh = 0.521 ± 0.020) regions are significantly higher than the Tibetan Plateau-Tengger Desert region (hd = 0.568 ± 0.089; π= (2.070 ± 0.538) ×10-3; uh = 0.389 ± 0.033). However, the southern margin of the Tanggula Mountains (hd = 0.727; π= 5.200×10-3; uh = 0.520) and the northern margin of Qaidam Basin (hd = 0.564; π= (2.850 ± 0.538) ×10-3; uh = 0.540) still have similar genetic diversity with that of other geographic clusters. This may mean that these regions are refuge areas or provenances of S.caninervis in Qinghai Tibet Plateau. Otherwise, Structure clustering analysis and haplotype distribution of chloroplast reveals three groups of S.caninervis populations in all distribution areas, Central Asia (CA), Pamirs_Tian mountain (PT) and Qinghai-Tibet_Tengger (QT). At the species level, there are significant phylogeographic structure between different S.caninervis populations (Nst = 0.369, Gst = 0.312, p < 0.05). AMOVA results of S.caninervis from three geographic regions demonstrate that geneic variations, occur within populations (75%) and between populations in different geographic clusters. However, the grouping results for each haplotype with SAMOVA and the consensus tree constructed by Bayesian Inference, does not group well the haplotypes into each region. At the same time, there is a wide distribution of shared haplotypes throughout the entire distribution area. It may mean there are extensive gene flow and long-distance dispersal of propagules throughout the whole distribution area. The genetic differentiation and genetic structure between short-distance populations may be caused by the micro-habitats in which the populations are located or climate factors. （2）Phylogenetic structure and population dynamics of S.caninervis After screening, two chloroplast DNA regions (psbD-trnT and trnL-rpL32), with high polymorphisms, are used for phylogeography research. 15 haplotypes are identified in 330 samples of 27 S.caninervis populations. The haplotype phylogenetic tree constructed by Network can clearly separate the CA and the QT region, and the PT is the connection area between them. This is in accordance with the geographical distribution structure of the distribution area, PT is the connecting area of CA and QT. Based on the analysis of molecular clocks used by Beast, it is initially estimated that the haplotype differentiation time is about 6-0.2 Ma, mainly in 2 Ma years, which shows that the genetic differentiation of S.caninervis populations may be mainly concentrated in the Pleistocene. The oldest haplotypes are mainly distributed in the middle east of Qinghai Tibet Plateau, while the sub-old haplotypes are mainly distributed on the Pamirs plateau. Combining the lineage structure, shared and unique haplotypes between the populations of different populations, we speculate that S.caninervis may migrate from the central and eastern parts of the Tibetan Plateau to the Pamirs first, and then from the Pamirs to the northeast and southeast direction under the control of the west winds. Some haplotypes in the Tengger Desert may have been transmitted from the Pamirs populations by long-distance diperal. The results of ecological niche modelling show that, contrary to most results of a large-scale shrinkage of the thermophilic vascular plants and retreation to refuge in the ice age, the population of S.caninervis may have expanded during the glacial period, and shrinkage occurs when thermophilic plants expand during the interglacial period or after the ice age. This provides new evidence for population dynamic history of drought-tolerant bryophytes in response to severe environmental changes during the Quaternary glacial and interglacial periods. （3）Influence of habitat on the survival of S.caninervis From the above shows that habitat changes may affect the growth and genetic diversity of S.caninervis, we have conducted a fixed study in the Gurbantunggut Desert in order to verify the effect of habitats. By studying the genetic diversity of S. caninervis and the relationship between S. caninervis and vascular plants, it is found that human disturbance and low precipitation may reduce the genetic diversity of S. caninervis to a certain extent. In desert areas, water is the main limiting factor for plant growth. Precipitation may have an effect on population genetic diversity by affecting population size. At the same time, the complete loss of shrubs in a short period of time may seriously threaten the growth of S. caninervis beneath the melting snow. However, the winter snow cover protects the wintering of bryophytes. The removal of shrubs significantly decreased chlorophyll ?uorescence activity and soluble protein content in S. caninervis, especially under the total shrub removal treatment. In resource-constrained conditions, SOD is an important antioxidant enzyme that of peroxidase (POD), catalase (CAT) and osmotic adjustment substances, for S.caninervis survival. This aspect responds to the possibility of expansion of S.caninervis during the slow cooling of the glacial period （4）Analysis of complete chloroplast genome of S.caninervis In order to accurately obtain the mutated locus of chloroplast gene, we based on high-throughput sequencing technology, sequences, splices, and analyzes the structure of the chloroplast genome of S.caninervis. The results shows that: we find that there are a total of 15 nuclear simple sequence repeat (SSR) loci, 175 potential base sites with mutations and deletions in the chloroplast genome of S.caninervis, containing 139 single nucleotide polymorphisms (SNPs) and 36 indel sites through comparison between 6 chloroplast genomes, which mainly occur in non-coding areas. It provides an important reference for the selection of molecular markers for the genetic diversity and genetic structure of the population of S.caninervis. In addition, the total genome size is 123,131 bp in length, containing a pair of inverted repeats (IRs) of 9,965 bp, which were separated by large single copy (LSC) and small single copy (SSC) of 84,663 and 18,539 bp, respectively. The overall GC contents of the plastid genome are 28.33%. A total of 130 genes are annotated, including 83 protein coding genes (CDS), 37 tRNA genes，8 rRNA genes. It is worth noting that, unlike other wet mosses and vascular plants, the chloroplast genome of S.caninervis lacks the gene petN that plays an important role in plant photosynthesis electron transport, which may be related to the evolution of drought tolerance.