生物多样性产生和维持机制一直都是生态学研究的核心问题。人们通过大量的实验观察和理论推导揭示了很多系统中生物多样性的共有模式，如物种多度分布、物种面积关系、群落相似度随地理距离的衰减等，并且提出了各种假说来解释这些模式。然而，相关的工作主要集中在大型生物中。微生物作为地球上数量最多、分布范围最广、生态功能不可替代的生物类群，被研究的并不充分；微生物群落是否遵循大型动植物里常见的生物多样性模式还未有定论。本文研究了内蒙古草原土壤细菌群落的生物多样性及其分布格局。在10厘米到200千米7个不同的空间尺度上嵌套取样，采集了90份土壤样品，通过高通量焦磷酸平行测序和实时定量PCR技术获取土壤细菌群落的物种组成和生物量数据。草原土壤细菌物种丰富度很高，达到上千个实用分类单元（OTU）；分属14个门，有5个门的细菌相对多度超过1%，8个门的细菌相对多度超过0.1%，其中放线菌、变形菌和酸杆菌相对多度之和达80%以上。细菌群落的生物量介于2.72×108 ~ 4.23×109（DNA拷贝数/克干土）之间。稀化处理后的细菌OTU数目、香侬-威纳指数和辛普森指数分别为105 ~ 306、3.70 ~ 5.48及15.80 ~ 171.05，土壤细菌群落的生物多样性与总生物量之间并不存在显著的相关关系。基于不同序列相似度（90%，95%或97%）构建了OTU数据库，对所有土壤细菌群落物种多度分布模式进行了对数正态和对数级数分布拟合。有一半的群落其常见OTU符合对数正态分布；而全部群落的稀有OTU都呈对数级数分布，暗示稀有细菌出现在群落中是随机的，不受环境、细菌间相互作用等因素的影响。分析了土壤细菌群落整体以及放线菌、变形菌和酸杆菌三个类别的群落相似性与地理距离的关系。在7个较狭窄的空间距离范围上，所有细菌及上述三类细菌没有表现出群落相似性随地理距离增加而衰减的模式。在跨7个空间尺度上（从10厘米到200千米），细菌群落整体和放线菌表现出群落相似性随地理距离的衰减模式，而变形菌和酸杆菌群落相似性没有呈现衰减模式。分析了细菌群落整体以及放线菌、变形菌和酸杆菌在100 m×100 m空间尺度上的物种面积关系。对于90%、95%和97%序列比对相似度构建的数据库，土壤所有细菌、放线菌、变形菌和酸杆菌得到的物种数-面积关系参数z值（对数尺度上物种数与生境面积关系的斜率）都相对较小，其范围分别为0.0059 ~ 0.0115、0.0040 ~ 0.0056、0.0134 ~ 0.0113及0.0122 ~ 0.0209，说明该尺度上细菌OTU周转率很低；并且除变形菌外，序列比对相似度越高——即OTU的分类阶元越低，z值越大。分析了所有土壤细菌、放线菌、变形菌和酸杆菌在7个各自空间尺度上的区域分布与局域多度的关系。无论是所有细菌还是上述三个类群，都表现出区域分布与局域多度的正相关关系，即局域多度高的细菌在区域内分布广，而局域多度低的细菌在区域内分布窄。基于本研究中观察到群落结构相似性衰减以及物种面积关系可以推测内蒙古土壤细菌群落受到很弱的扩散限制；同时，对群落物种多样性-生物量关系、物种多度分布以及物种区域分布-局域多度关系的分析表明，群落结构有着明显的随机化过程的信号。这暗示该地区土壤细菌群落中性或者近中性共存可能很普遍。

A central objective of ecology is to understand the mechanisms that generate and maintain biodiversity. Ecologists have conducted a lot of theoretical and empirical studies to unravel macroecological patterns of diversity and distribution of various organisms, such as species-abundance distribution, species-area relationship, distance-decay relationship, and to understand the underlying processes that determine these patterns. However, most of these studies focus on macroorganisms. While microorganisms are the most abundant and widespread organisms on earth and mediate important ecosystem processes, little is known about their diversity. Whether the microbial communities show the same patterns of diversity and distribution as plants and animals is still unknown. In this research, we implemented a field survey to describe the biodiversity patterns of soil bacteria in the Inner Mongolia grassland. A nested-sampling design was employed to collect 90 soil samples at seven different spatial scales, and the geographic distance between sampled sites was from 10 cm to 200 km. Taxa richness of soil bacteria and their relative abundance were measured by massively parallel high-throughput pyrosequencing and bacterial biomass was quantified by real-time PCR.Taxa richness of bacteria in the Inner Mongolia grassland was high, and thousands of operational taxonomic units (OTUs) were defined. Bacterial sequences belonged to 14 phyla, among which 5 were recorded with relative abundance over 1% and 8 over 0.1%. Actinobacteria, Proteobacteria, and Acidebacteria were dominant, comprising more than 80% of all the sequences. The biomass of soil bacteria was between 2.72×108-4.23×109 DNA sequences per gram dry soil. The number of OTUs, Shannon’s index and Simpson’s index of bacteria communities were 105-306, 3.70-5.48 and 15.80-171.05 respectively after rarefaction, and there were no significant relationships between them and bacteria biomass.Species-abundance distribution was studied based on different sequence similarities (90%, 95% and 97%), and log-normal distribution and log-series distribution were tested against the observed patterns. The common OTUs in half of the 90 communities were fitted well by log-normal distribution while the rare OTUs in these communities were explained by log-series distribution. The fact that rare OTUs in all the 90 communities were best characterized by log-series distribution suggests that rare bacteria are regulated randomly in communities rather than constrained by other deterministic factors such as environment.Distance-decay relationship was studied on all the bacteria, Actinobacteria, Proteobacteria, and Acidobacteria, respectively. At each spatial scale, there were no relationships between pairwise community similarity and geographic distance for all the bacteria, Actinobacteria, Proteobacteria, and Acidobacteria. Across all the 7 spatial scales (from 10cm to 200km), all the bacteria and Actinobacteria community similarity decreased significantly with increased geographic distance, but Proteobacteria and Acidobacteria yet did not show any pattern of distance-decay relationship.Species-area relationship was studied on all the bacteria, Actinobacteria, Proteobacteria and Acidobacteria at 100 m×100 m scale. The estimated z-values (the slope of regression between number of species and area) of the four groups at the 90%, 95% and 97% DNA resolutions were very small. Their ranges were 0.0059-0.0115, 0.0040-0.0056, 0.0134-0.0113 and 0.0122-0.0209 respectively, which suggested a low rate of taxa turnover across space at this scale. The z-value also varied by taxonomic resolution, increasing with increased taxonomic resolution for all the bacteria, Actinobacteria and Acidobacteria.Distribution-abundance relationship was studied on all the bacteria, Actinobacteria, Proteobacteria and Acidobacteria at different spatial scales. There was significantly positive interspecific relationship between regional distribution and local abundance of all the four groups, which suggested that locally abundant bacteria were widely distributed and locally rare bacteria were narrowly distributed.Taken all the above results together, we may infer that the dispersal of soil bacteria in the Inner Mongolia grassland is weakly limited. Biodiverstiy-biomass relationship, species abundance distribution and distribution-abundance relationship patterns show that random processes may drive the community structure of the soil bacteria, which in turn suggests widespread neutral or nearly neutral coexistence of soil bacteria in Inner Mongolia grasslands.