Introduction

The geographical distribution of plant species is determined by a number of biotic and abiotic factors, as well as the interactions between them that ultimately delineate species ranges worldwide. The same factors can act in concert to drive adaptive differentiation among populations belonging to the same plant species, a phenomenon known as local adaptation. A fitness advantage of local over nonlocal genotypes has been documented in many plant species (Leimu & Fischer, 2008; Hereford, 2009), including the model plant Arabidopsis thaliana. In A. thaliana, both reciprocal transplant experiments (Ågren & Schemske, 2012; Ågren et al., 2013; Postma & Ågren, 2016; Thiergart et al., 2020; Ellis et al., 2021) and common-garden experiments (e.g. Fournier-Level et al., 2011; Montesinos-Navarro et al., 2011; Exposito-Alonso et al., 2019) have found evidence of local adaptation across the native range in Europe. Correlations between genetically based variation in phenotype and environmental factors can suggest causes of divergent selection, but determining conclusively the respective contribution of abiotic and biotic factors such as climate, soil physicochemical properties and soil microbiome to differences in selection requires an experimental approach because these factors are typically correlated with one another. Specifically, it is difficult to uncouple the effect of soil physicochemical properties from the effect of the soil microbiome, and to identify the climatic variables that contribute the most to adaptive differentiation between populations. Recent evidence indicates that above-ground phenotypes and fitness in plants can be modulated by interactions with microbial root commensals (Friesen et al., 2011; Lau & Lennon, 2012; Wagner et al., 2014; Lu et al., 2018; Hou et al., 2021a; Van Nuland et al., 2021). However, there is limited knowledge on the extent to which variation in soil microbiome can drive adaptive differentiation among plant populations. A recent report shows that interactions with microbes can affect estimates of plant local adaptation, although effects may vary among environments (Petipas et al., 2020). To understand the importance of microbe-mediated local adaptation, it is essential to disentangle how biotic and abiotic factors act as selective agents and affect the relative fitness of local and nonlocal populations (Petipas et al., 2021).

Environmental conditions also drive geographical variation in below-ground soil microbial communities. Several studies identified a link between microbial community assembly and host distribution (Brundrett & Tedersoo, 2018; U'Ren et al., 2019), suggesting that the evolutionary history between root symbionts and their host plants has shaped plant populations worldwide (Tedersoo et al., 2020). Furthermore, relationships between microbial community assemblages and latitude (Vetrovsky et al., 2019; Thiergart et al., 2020) or soil physicochemical properties (Fierer & Jackson, 2006) have been reported. For example, climatic variables were shown to explain the global distribution of common soil fungi, as well as the composition and diversity of fungal communities, better than edaphic factors such as soil pH or bulk density (Tedersoo et al., 2014; Vetrovsky et al., 2019). By contrast, soil pH was repeatedly identified as the primary variable explaining bacterial community differentiation in soil at both small and large spatial scales (Fierer & Jackson, 2006; Rousk et al., 2010; Karimi et al., 2018). The large-scale sampling and substantial replication suggest that these associations are robust, but studies that subject predictions to experimental evaluation are lacking.

Here, we conducted an experiment under controlled conditions to examine the extent to which differences in climate and soil environment can explain divergent selection between two geographically distant and locally adapted populations of the model plant A. thaliana, from Italy and Sweden. In addition, we tested the effects of climate, soil matrix and plant genotype on the composition of the rhizosphere microbiome. A previous field experiment, in which plant genotypes and soil were reciprocally transplanted between the two source populations, indicated a strong effect of location but at most a weak effect of soil composition on the relative performance of the two genotypes (Thiergart et al., 2020). However, that experiment could not distinguish between the effects of soil matrix and microbial composition on plant fitness, which is challenging to perform under field conditions owing to the difficulty of keeping soil microbial communities independent of one another. Moreover, although the two genotypes differ in tolerance to freezing (Oakley et al., 2014), and there is a strong relationship between minimum temperature in winter and performance of the Italian genotype in Sweden (Ågren & Schemske, 2012), it is not clear whether differences in temperature, day length and light intensity are sufficient to explain the effect of location. In the present study, we experimentally examine: the independent and combined effects of soil matrix and soil microbiome on the relative fitness of the two genotypes in climate growth chambers that mimicked the seasonal changes in temperature, day length and light intensity at the two home sites; and the effects of soil matrix and plant genotype on the composition of the rhizosphere microbiome under the same chamber conditions (Fig. 1).

The results are consistent with the hypothesis that differences in climate between northern and southern Europe have driven much of the adaptive differentiation between the two A. thaliana populations, whereas differences in soil matrices and soil microbiome have been less important. By contrast, the results suggest that differences in below-ground soil physicochemical conditions, and to a lesser extent in above-ground climatic conditions, can explain differentiation in bacterial and fungal soil assemblages between the two sites.

中文翻译：

---

气候驱动根际微生物组变异和地理上遥远的拟南芥种群之间的不同选择

介绍

植物物种的地理分布由许多生物和非生物因素以及它们之间的相互作用决定，这些因素最终划定了世界范围内的物种范围。相同的因素可以协同作用，推动属于同一植物物种的种群之间的适​​应性分化，这种现象称为局部适应。许多植物物种（Leimu & Fischer, 2008 ; Hereford, 2009），包括模式植物拟南芥（ Arabidopsis thaliana ），都记录了本地基因型相对于非本地基因型的适应性优势。在拟南芥中，两个相互移植实验（Ågren 和 Schemske，2012 年；Ågren等人，2013 年；Postma 和 Ågren，2016 年；Thiergart等人，2020 年；Ellis等人，2021 年）和公共花园实验（例如 Fournier-Level等人，2011 年；Montesinos-Navarro等人，2011 年；Exposito-Alonso等人，2019 年) 在欧洲本土范围内发现了当地适应的证据。基于遗传的表型变异与环境因素之间的相关性可以表明不同选择的原因，但最终确定非生物和生物因素（如气候、土壤理化性质和土壤微生物组）对选择差异的贡献需要一种实验方法，因为这些因素是通常相互关联。具体而言，很难将土壤理化性质的影响与土壤微生物组的影响分开，并确定对种群间适应性分化贡献最大的气候变量。等人，2011 年；刘和列侬，2012 年；瓦格纳等人，2014 年；卢等人，2018；侯等人，2021a；Van Nuland等人，2021 年）。然而，关于土壤微生物组的变化可以在多大程度上推动植物种群之间的适​​应性分化的知识有限。最近的一份报告显示，与微生物的相互作用会影响植物局部适应的估计，尽管影响可能因环境而异（Petipas等人，2020）。To understand the importance of microbe-mediated local adaptation, it is essential to disentangle how biotic and abiotic factors act as selective agents and affect the relative fitness of local and nonlocal populations (Petipas et al ., 2021 ).

环境条件也推动了地下土壤微生物群落的地理变化。几项研究确定了微生物群落组装与宿主分布之间的联系（Brundrett & Tedersoo，2018 年；U'Ren等人，2019 年），这表明根共生体与其宿主植物之间的进化历史已经塑造了全球植物种群（Tedersoo等人） ., 2020 年）。此外，微生物群落组合与纬度之间的关系 (Vetrovsky et al ., 2019 ; Thiergart et al ., 2020 ) 或土壤理化性质 (Fierer & Jackson, 2006） 已经汇报过。例如，气候变量被证明可以解释常见土壤真菌的全球分布，以及真菌群落的组成和多样性，优于土壤 pH 值或容重等土壤因素（Tedersoo等人，2014 年；Vetrovsky等人） ., 2019 年）。相比之下，土壤 pH 值被反复确定为解释小空间尺度和大空间尺度土壤中细菌群落分化的主要变量（Fierer & Jackson，2006；Rousk等，2010；Karimi等，2018）。大规模抽样和大量复制表明这些关联是稳健的，但缺乏将预测纳入实验评估的研究。

在这里，我们在受控条件下进行了一项实验，以检查气候和土壤环境的差异在多大程度上可以解释来自意大利和瑞典的模式植物拟南芥的两个地理上遥远和当地适应的种群之间的差异选择。此外，我们测试了气候、土壤基质和植物基因型对根际微生物组组成的影响。先前的田间试验中，植物基因型和土壤在两个源种群之间相互移植，表明位置的影响很大，但土壤成分对两种基因型的相对表现的影响最大（Thiergart等人，2020）。然而，该实验无法区分土壤基质和微生物组成对植物适应性的影响，由于难以保持土壤微生物群落相互独立，因此在田间条件下进行这项实验具有挑战性。此外，尽管这两种基因型在耐冻性方面存在差异（Oakley等，2014 年），但冬季最低气温与瑞典意大利基因型的表现之间存在密切关系（Ågren & Schemske，2012 年）)，尚不清楚温度、日长和光照强度的差异是否足以解释位置的影响。在本研究中，我们通过实验研究：土壤基质和土壤微生物组对气候生长室中两种基因型的相对适应性的独立和综合影响，模拟了两个家庭地点的温度、日长和光照强度的季节性变化; 以及相同室条件下土壤基质和植物基因型对根际微生物组组成的影响（图1）。

结果与以下假设一致，即北欧和南欧之间的气候差异推动了两个拟南芥种群之间的大部分适应性分化，而土壤基质和土壤微生物组的差异则不太重要。相比之下，结果表明，地下土壤物理化学条件的差异，以及在较小程度上的地上气候条件，可以解释两个地点之间细菌和真菌土壤组合的差异。