1 INTRODUCTION

Plant invasions are reshaping biotic interactions across all ecosystems and pose a major threat to biodiversity. Soil microbiota have the potential to either facilitate or hinder such invasions depending on the relative importance of the pathogens and mutualists involved (Reinhart & Callaway, 2006; Traveset & Richardson, 2014). In novel environments, introduced plants may no longer encounter the harmful soil-borne pathogens of their native range (Callaway et al., 2004; Reinhart et al., 2010), which may allow them to have a competitive advantage and/or higher fitness compared with native plants that are under attack by specialized pathogens (Inderjit & van der Putten, 2010). Alternatively, novel generalist pathogens in the introduced range may contribute to biotic resistance, potentially preventing invasion (Elton, 1958; Inderjit & van der Putten, 2010). Plant invasions can also be impeded by changes in interactions with mutualistic microbes, if specialized beneficial soil microbiota are absent in the novel environment (missed mutualists hypothesis; Mitchell et al., 2006; Alpert, 2006). When compatible mutualists are present, they can facilitate plant invasions (Hayward et al., 2015), for example, by enabling plants to tolerate abiotic stress (Afkhami et al., 2014). To more fully understand the mechanisms involved in plant invasions, it is necessary to explore the multiple ways that microbial communities can shape plant phenotypes.

Rhizobia are root nodule-inducing bacteria that convert atmospheric nitrogen (N₂) to a form that is usable by their legume hosts. In exchange for the fixed nitrogen, plants provide rhizobia with carbon, micronutrients and protection (Franche et al., 2009; Sprent, 2001). Although the association with rhizobia has been linked to the success of legume invasions in general (Rodríguez-Echeverría et al., 2009), rhizobia are mostly not transmitted via seeds and are likely dispersal-limited (Rout & Callaway, 2012). Therefore, legumes may lack suitable rhizobial partners in novel habitats (Parker, 2001; Parker et al., 2006). Availability of compatible rhizobia declined rapidly outside an established patch of the invasive legume Medicago polymorpha leading to decline in fitness at a spatial scale as short as 50 m away from the patch (Lopez et al., 2021). Indeed, symbiotic legume species are less likely to become invasive than non-symbiotic ones (Simonsen et al., 2017) and legumes that are more specialized in their rhizobial association have been found in fewer introduced ranges than their generalist counterparts (Harrison et al., 2018). Nonetheless, various legume species have become invasive across all environments and represent some of the most problematic species globally (Pyšek, 1998). The mechanisms that can alleviate the negative impacts of the uncoordinated dispersal of symbionts and hosts are symbiotic promiscuity, that is, the ability to form a symbiosis with a wider suite of symbiont species (Klock et al., 2015; but see Klock et al., 2016; Keet et al., 2017) and the ability to co-opt rhizobia from native legumes (Parker et al., 2006). Alternatively, a species can evolve to be less dependent on its mutualist soil microbiota through adaptive divergence in the introduced range (Seifert et al., 2009).

Because the association with rhizobia increases the amount of nitrogen available for the host plant (Sprent & Sprent, 1990), it can have notable effects on host traits related to leaf nitrogen content, such as palatability or nutritional quality to herbivores or plant phytochemical composition (Awmack & Leather, 2002; Bryant et al., 1983). For example, in the perennial herb Trifolium repens, a rhizobial association decreased resistance to the generalist moth Spodoptera littoralis (Kempel et al., 2009); similarly, soybeans (Glycine max) with rhizobia were more susceptible to spider mites than strains without rhizobia (Katayama et al., 2010). Nitrogen availability is also known to affect the emission of volatile organic compounds (VOCs) that serve multiple roles in plants, from within-plant signalling to plant–plant communication and indirect resistance against herbivores (Kessler & Heil, 2011; Li & Blande, 2017). As an example, in lima bean (Phaseolus lunatus), rhizobia altered the composition of induced VOCs by increasing the amount of N-containing indole while reducing the amount of C-containing compounds, which then deterred a specialist herbivorous beetle (Ballhorn et al., 2013). Terrestrial gastropods also use olfactory cues to assess the quality of their host plants (Kiss, 2017) and VOC profiles may be even more important determinants of host plant choice than leaf phytochemical content (Hanley et al., 2018). Such traits that mediate interactions with generalist herbivores can be especially important in invasive species that are likely to encounter generalist rather than specialist herbivores.

The North American perennial herb Lupinus polyphyllus (Lindl. Fabaceae) ranks among the top invasive plant species in Europe in terms of negative environmental and socioeconomic impacts (Rumlerová et al., 2016). The species negatively affects plant community diversity by increasing the proportion of competitively superior species relative to species with weaker competitive ability (Hansen et al., 2021) and diminishing species richness (Ramula & Pihlaja, 2012). In Finland, invasive L. polyphyllus has evolved resistance to a generalist snail, while simultaneously losing diversity in leaf alkaloid content (Kalske, Luntamo, et al., 2022). Populations with higher leaf mass per area (LMA) were more resistant to snails and therefore, LMA might be critical to leaf palatability (Kalske, Luntamo, et al., 2022). The species is able to nodulate profusely in sites outside its native range (Ryan-Salter et al., 2014), suggesting that it is capable of finding compatible symbiotic partners in novel areas. Generally speaking, Lupinus spp. mostly associate with Bradyrhizobium spp. (Andrews & Andrews, 2017), but the genera of rhizobia isolated from the nodules in different parts of the introduced range include Bradyrhizobium and Rhizobium (Stępkowski et al., 2018). It remains unknown whether the ability to associate with putatively novel rhizobia outside its native range is an innate characteristic (due to plasticity in the rhizobial association) or whether populations in the invaded area have adapted to co-opt local rhizobia.

We have begun to consider the importance of soil microbiota in influencing the outcome of plant invasions. To date, though, these efforts have largely focused on growth-related traits, and little is known about how the relationship between invasive plants and soil microbes may affect other traits such as resistance to herbivores. Herbivore resistance is one of the key traits that has been demonstrated to evolve in introduced plant populations, with the potential to determine the outcome of plant introductions (Rotter & Holeski, 2018). Because beneficial soil microbiota mediate herbivore resistance in plants in general (Ballhorn et al., 2013; Kempel et al., 2009), there is the intriguing possibility that the adaptive divergence of invasive plants may be mediated by their interactions with soil microbiota. In the current study, we used plants originating from native and invasive populations of L. polyphyllus to study how rhizosphere soil microbiota from an invasive population affects plant performance, herbivore resistance and VOC emissions. The use of soil microbiota from the introduced range enabled us to explore its benefits on plants from invasive populations and determine how native plants interact with putatively novel microbiota. We considered the following six performance traits: height, number of leaves, biomass, root: shoot ratio, nodule number and nodule viability. We evaluated herbivore resistance using a bioassay with a generalist snail and by measuring LMA. We predicted that the effect of symbiotic rhizobia and other beneficial microbes would be stronger than those of soil-borne pathogens, resulting in net positive effects of intact soil inoculum on overall performance and herbivore resistance. Furthermore, we predicted that plants from the invasive populations would benefit more from the soil microbiota than plants from the native populations in terms of both performance and herbivore resistance, suggesting adaptive divergence of the plants to the soil microbial community at the regional scale. We also predicted that this divergence would be due to the interactions between the plant and the symbiotic rhizobia, manifesting as more profuse nodulation in invasive compared with native L. polyphyllus. Finally, we hypothesized that differences in plant herbivore resistance between the native and invasive populations might be broadly reflected as differential VOC profiles.

中文翻译：

---

土壤微生物群解释了多年生草本植物的本地和入侵种群之间食草动物抗性的差异

1 简介

植物入侵正在重塑所有生态系统中的生物相互作用，并对生物多样性构成重大威胁。土壤微生物群有可能促进或阻碍这种入侵，这取决于所涉及的病原体和互惠互利者的相对重要性（Reinhart & Callaway,  2006 ; Traveset & Richardson,  2014）。在新环境中，引入的植物可能不再遇到其原生范围内的有害土壤传播病原体（Callaway 等人，  2004 年；Reinhart 等人，  2010 年），这可能使它们具有竞争优势和/或更高的适应性与受到特殊病原体攻击的本地植物相比（Inderjit & van der Putten,  2010）。或者，引入范围内的新型通用病原体可能有助于生物抗性，从而可能防止入侵（Elton，  1958；Inderjit & van der Putten，  2010）。如果新环境中不存在专门的有益土壤微生物群，植物入侵也可能受到与共生微生物相互作用的变化的阻碍（错过了共生者假设；Mitchell 等人，  2006 年；Alpert，  2006 年）。当存在相容的共生者时，它们可以促进植物入侵 (Hayward et al.,  2015 )，例如，通过使植物能够耐受非生物胁迫 (Afkhami et al.,  2014 )）。为了更全面地了解植物入侵的机制，有必要探索微生物群落塑造植物表型的多种方式。

根瘤菌是诱导根瘤的细菌，可将大气中的氮 (N ₂ ) 转化为可供豆科植物宿主使用的形式。作为固定氮的交换，植物为根瘤菌提供碳、微量营养素和保护（Franche 等人，  2009 年；Sprent，  2001 年）。尽管与根瘤菌的关联通常与豆科植物入侵的成功有关（Rodríguez-Echeverría 等人，  2009 年），但根瘤菌大多不通过种子传播，并且可能传播受限（Rout 和 Callaway，  2012 年）。因此，豆科植物在新栖息地可能缺乏合适的根瘤菌伙伴（Parker，  2001；Parker 等，  2006）。在已建立的侵入性豆科植物Medicago polymorpha斑块之外，相容性根瘤菌的可用性迅速下降，导致距离斑块仅 50 m 的空间尺度上的适应性下降（Lopez 等人，  2021 年）。事实上，与非共生豆科植物相比，共生豆科植物不太可能成为侵入性物种（Simonsen 等人，  2017 年），并且在根瘤菌关联方面更专业的豆科植物被发现的引入范围比其通才对应物更少（Harrison 等人，2017 年）。 ,  2018 年）。尽管如此，各种豆科植物已在所有环境中入侵，并代表了全球一些最有问题的物种（Pyšek，  1998）。可以减轻共生体和宿主不协调扩散的负面影响的机制是共生混杂，即与更广泛的共生物种形成共生关系的能力（Klock 等人，  2015 年；但参见 Klock 等人。 ,  2016 ; Keet et al.,  2017 ) 以及从本地豆科植物中选择根瘤菌的能力 (Parker et al.,  2006 )。或者，一个物种可以通过在引入范围内的适应性分歧进化为更少依赖于其互惠互利的土壤微生物群（Seifert 等人，  2009 年）。

因为与根瘤菌的结合增加了宿主植物可利用的氮量（Sprent & Sprent,  1990），它对与叶片氮含量相关的宿主性状具有显着影响，例如对食草动物的适口性或营养品质或植物植物化学成分。 Awmack 和皮革，  2002 年；Bryant 等人，  1983 年）。例如，在多年生草本植物Trifolium repens中，根瘤菌的结合降低了对泛光蛾Spodoptera littoralis的抗性（Kempel 等人，  2009 年）；同样，带有根瘤菌的大豆 ( Glycine max ) 比没有根瘤菌的菌株更容易感染红蜘蛛 (Katayama et al.,  2010）。众所周知，氮的可用性会影响挥发性有机化合物 (VOC) 的排放，这些挥发性有机化合物 (VOC) 在植物中发挥着多种作用，从植物内信号传导到植物-植物交流以及对食草动物的间接抗性 (Kessler & Heil,  2011 ; Li & Blande,  2017）。例如，在利马豆 ( Phaseolus lunatus ) 中，根瘤菌通过增加含氮吲哚的量同时减少含碳化合物的量来改变诱导 VOC 的组成，从而阻止了一种专门的食草甲虫 (Ballhorn 等人。 ,  2013 年）。陆生腹足动物也使用嗅觉线索来评估其寄主植物的质量（Kiss，  2017) 和 VOC 谱可能是比叶片植物化学成分更重要的寄主植物选择决定因素（Hanley 等人，  2018 年）。这种介导与通才食草动物相互作用的特征对于可能遇到通才而非专业食草动物的入侵物种尤其重要。

就负面环境和社会经济影响而言，北美多年生草本植物多叶羽扇豆(Lindl. Fabaceae) 在欧洲的入侵植物物种中名列前茅（Rumlerová 等人，  2016 年）。该物种通过增加竞争优势物种相对于竞争力较弱的物种的比例（Hansen 等人， 2021 年）和降低物种丰富度（Ramula 和 Pihlaja，  2012 年） ，对植物群落多样性产生负面影响 。在芬兰，入侵的多叶L. polyphyllus已经进化出对通才蜗牛的抗性，同时失去了叶片生物碱含量的多样性（Kalske、Luntamo 等人，  2022）。单位面积叶质量 (LMA) 较高的种群对蜗牛的抵抗力更强，因此，LMA 可能对叶片适口性至关重要 (Kalske, Luntamo, et al.,  2022 )。该物种能够在其原生范围以外的地点大量结瘤（Ryan-Salter 等人，  2014 年），这表明它能够在新区域找到兼容的共生伙伴。一般来说，羽扇豆属。主要与慢生根瘤菌有关。(Andrews & Andrews,  2017 )，但从引入范围不同部分的根瘤中分离出的根瘤菌属包括慢生根瘤菌属和根瘤菌属(Stępkowski et al.,  2018 )）。目前尚不清楚与假定的新根瘤菌在其原生范围之外的关联能力是否是一种先天特征（由于根瘤菌关联的可塑性），或者入侵地区的种群是否已经适应了吸收当地的根瘤菌。

我们已经开始考虑土壤微生物群在影响植物入侵结果方面的重要性。然而，迄今为止，这些努力主要集中在与生长相关的性状上，关于入侵植物和土壤微生物之间的关系如何影响其他性状，如对食草动物的抗性，我们知之甚少。草食动物抗性是已被证明在引入的植物种群中进化的关键性状之一，具有决定植物引入结果的潜力（Rotter & Holeski，  2018 年）。因为有益的土壤微生物群通常介导植物的食草动物抗性（Ballhorn 等人，  2013 年；Kempel 等人，  2009 年）），有一种有趣的可能性是，入侵植物的适应性差异可能是由它们与土壤微生物群的相互作用所介导的。在目前的研究中，我们使用了源自多叶乳杆菌本地和入侵种群的植物研究来自入侵种群的根际土壤微生物群如何影响植物性能、食草动物抗性和 VOC 排放。使用引入范围内的土壤微生物群使我们能够探索其对来自入侵种群的植物的益处，并确定原生植物如何与假定的新型微生物群相互作用。我们考虑了以下六个性能特征：高度、叶片数、生物量、根：茎比、根瘤数和根瘤活力。我们使用通才蜗牛的生物测定法和测量 LMA 来评估食草动物的抗性。我们预测，共生根瘤菌和其他有益微生物的影响将比土壤传播的病原体更强，从而导致完整土壤接种物对整体性能和食草动物抗性产生净积极影响。此外，我们预测，在性能和食草动物抗性方面，来自入侵种群的植物比来自本地种群的植物从土壤微生物群中受益更多，这表明植物在区域范围内对土壤微生物群落的适应性分化。我们还预测，这种差异是由于植物与共生根瘤菌之间的相互作用造成的，与原生根瘤菌相比，侵入性根瘤菌的结瘤更多。L. polyphyllus。最后，我们假设本地和入侵种群之间植物食草动物抗性的差异可能广泛反映为不同的 VOC 谱。