1 INTRODUCTION

Biological invasions are a major characteristic of human-induced global environmental change. Invaded ecosystems and communities in many parts of the world have been affected severely (Fei et al., 2014; Vilà & Hulme, 2017). For this reason, much effort in ecological research has aimed at understanding what determines invasion success. Furthermore, biological invasions represent natural experiments that offer real-time opportunities to study the assembly of communities (Shea & Chesson, 2002). Specifically, as many alien organisms have integrated into native resident communities, biological invasions are key to understanding the major ecological processes that determine the formation of novel interactions.

As closely related species are likely to be more similar, they should show strong niche overlap and compete for shared resources. Based on this premise, Darwin (1859) posed that relatedness between alien and native species could impede the success of alien species (Darwin's naturalization hypothesis). At the same time, if there are related native species, this indicates that the environment will most likely be suitable for the alien species too (preadaptation hypothesis). These two hypotheses, with opposing predictions on how relatedness between alien and native species should affect naturalization success, are now known as Darwin's naturalization conundrum (Diez et al., 2008). Indeed, the results of previous studies testing these hypotheses are largely inconsistent, suggesting that the two mechanisms act at different spatial scales and stages of invasion (Cadotte et al., 2018; Omer et al., in press; Thuiller et al., 2010). Moreover, most previous studies were correlative, based on floristic lists and field observations (Cadotte et al., 2018; Gallien & Carboni, 2017; Sheppard et al., 2018), whereas manipulative experiments, in which species are introduced into communities to provide causal insights (e.g. Malecore et al., 2019), are scarce. Furthermore, most previous studies considered direct interactions between alien and native plant species (e.g. competition for space or nutrients), but very few studies have tested indirect interactions such as those mediated by pollinators (but see Bezeng et al., 2015; Burns et al., 2011). However, as about half of flowering plant species relies on pollinators for at least 80% of their seed production (Rodger et al., 2021), the ability to attract resident pollinators can play a major role in the integration of alien plants in novel communities.

To reproduce in the non-native range, alien plants, which are often decoupled from their historical pollinators can use resident pollinators (Razanajatovo et al., 2015; Razanajatovo & van Kleunen, 2016; Traveset & Richardson, 2014), and thereby form novel plant-pollinator interactions. When the principles of community assembly are extended to pollinator-mediated interactions, communities governed by filtering and facilitation are expected to show more similar floral traits, and those governed by competition are expected to show more divergent floral traits (Sargent & Ackerly, 2008). In plant invasions, alien plants can negatively impact floral visitation of native plants, but native plants may also facilitate the early establishment of closely related alien plants by providing a preadapted pollinator community. Therefore, there are two apparently contradicting major concepts on the roles of plant-pollinator interactions in the assembly of invaded communities. First, a co-occurring native plant can have a positive influence on the floral visitation of an alien plant with similar floral traits (pollinator facilitation) (Brown & Kodric-Brown, 1979; Rathcke, 1983; Thomson, 1978; Waser & Real, 1979). Second, a co-occurring native plants can have a negative influence on the floral visitation of an alien plant with similar floral traits (pollinator-mediated competition; Schemske et al., 1978; Waser, 1982; Zimmerman, 1980), providing an advantage to alien plants with different traits to native plants. Furthermore, to attract flower visitors that are potential pollinators, floral traits should match the pollinator traits (Darwin, 1862). Therefore, the formation of novel plant-pollinator interactions between alien plants and resident pollinators should depend on the degree of floral trait similarity between the alien and native plants.

Floral traits, such as flower symmetry, colour and size, can act as signals for flower visitors to locate floral rewards, and have therefore been considered as important traits that mediate plant-pollinator interactions (Fornoff et al., 2017; Junker et al., 2015; Ortiz et al., 2021; Reverté et al., 2016). Furthermore, nectar production has been shown to influence the indirect interactions between co-flowering plants with shared pollinators (Carvalheiro et al., 2014). Nevertheless, the floral traits involved in pollinator attraction are still generally missing from studies in plant-community ecology (E-Vojtkó et al., 2020; Sargent & Ackerly, 2008). A meta-analysis seeking to understand the impacts of alien species on the pollination and reproductive success of native species considered floral traits, and showed that similarities in flower symmetry and colour in alien and native plants increased competition for pollinators (Morales & Traveset, 2009). Therefore, similarity in these floral traits may play critical roles in pollinator-mediated alien-native plant interactions.

The patterns of novel plant-pollinator interactions may differ with regard to phylogenetic and floral trait distances. While phylogenetic relatedness is frequently assumed to be a proxy for trait similarity, some floral traits may not be evolutionarily conserved (Sargent & Ackerly, 2008). For example, closely related species frequently have different flower colours (Eaton et al., 2012; Shrestha et al., 2014), and flower colour can vary within species (Coetzee et al., 2021). It is therefore important to consider both phylogenetic distance and trait dissimilarity (Cadotte et al., 2013; Lemoine et al., 2015) and to disentangle the effects of both (Marx et al., 2016). Thus, the patterns of novel plant-pollinator interactions might be more complex than those predicted by Darwin's Naturalization Conundrum (Diez et al., 2008). Furthermore, different mechanisms may act simultaneously to drive the success of alien species, as has been shown for direct plant–plant interactions (Malecore et al., 2019; Sheppard et al., 2018). When both pollinator facilitation and competition for pollinators play a role, floral visitation to alien species should be highest or lowest at intermediate phylogenetic and floral trait distances to native species. It is therefore important to consider nonlinear relationships between floral visitation to alien species and phylogenetic and floral trait distances to native species.

In a field experiment in which we simulated recently invaded communities by adding potted alien plants into co-flowering native communities, we tested whether floral visitation to alien plants depended on phylogenetic relatedness and functional similarity to native plants. If floral visitation to alien plants was facilitated by co-flowering native plants with similar floral traits, we predicted a negative relationship between floral visitation to alien plants and phylogenetic or floral trait distance to natives. If floral visitation to alien plants was hampered by co-flowering native plants with similar floral traits (pollinator-mediated competition), we predicted a positive relationship. If both facilitation and competition were operating, we predicted non-linear relationships with either high or low floral visitation at intermediate phylogenetic or floral trait distances (i.e. hump- or U-shaped relationships). More specifically, we asked (a) whether floral visitation to alien plants was related to phylogenetic and floral trait distances between alien and native species, and (b) whether the similarity in flower visitor composition was related to phylogenetic and floral trait distances between alien and native species.

中文翻译：

---

外来植物的花卉访问与其系统发育和花卉与本地植物的相似性呈非线性关系

1 简介

生物入侵是人类引起的全球环境变化的主要特征。世界许多地区被入侵的生态系统和社区受到了严重影响（Fei 等人，  2014 年；Vilà & Hulme，  2017 年）。出于这个原因，生态研究中的许多努力旨在了解决定入侵成功的因素。此外，生物入侵代表了自然实验，为研究社区的组装提供了实时机会（Shea 和 Chesson，  2002 年）。具体而言，由于许多外来生物已融入当地居民社区，因此生物入侵是了解决定新相互作用形成的主要生态过程的关键。

由于密切相关的物种可能更相似，它们应该表现出强烈的生态位重叠并争夺共享资源。基于这个前提，达尔文（1859）提出外来物种与本土物种之间的相关性可能会阻碍外来物种的成功（达尔文的归化假说）。同时，如果有相关的本地物种，这表明环境也很可能适合外来物种（预适应假设）。这两个假设，以及关于外来物种和本地物种之间的相关性如何影响入籍成功的相反预测，现在被称为达尔文的入籍难题（Diez 等，  2008）。事实上，先前测试这些假设的研究结果在很大程度上是不一致的，这表明这两种机制在不同的空间尺度和入侵阶段起作用（Cadotte 等人，  2018 年；Omer 等人， 出版中；Thuiller 等人，  2010 年） ）。此外，大多数先前的研究都是相关的，基于植物区系列表和实地观察（Cadotte 等人，  2018 年；Gallien 和 Carboni，  2017 年；Sheppard 等人，  2018 年），而操纵性实验，其中将物种引入群落以提供因果洞察（例如 Malecore 等人，  2019)，稀缺。此外，以前的大多数研究都考虑了外来植物物种和本地植物物种之间的直接相互作用（例如空间或营养竞争），但很少有研究测试间接相互作用，例如由传粉媒介介导的相互作用（但参见 Bezeng 等人，  2015 年；Burns 等人） .,  2011 年）。然而，由于大约一半的开花植物物种至少 80% 的种子生产依赖传粉媒介（Rodger 等人，  2021 年），因此吸引常驻传粉媒介的能力可以在将外来植物整合到新群落中发挥重要作用.

为了在非本地范围内繁殖，通常与其历史传粉媒介脱钩的外来植物可以使用常驻传粉媒介（Razanajatovo et al.,  2015 ; Razanajatovo & van Kleunen,  2016 ; Traveset & Richardson,  2014），从而形成新颖植物与传粉者的相互作用。当群落组装的原则扩展到传粉媒介介导的相互作用时，受过滤和促进控制的群落预计会表现出更多相似的花性状，而受竞争控制的群落预计会表现出更多不同的花性状（Sargent & Ackerly,  2008）。在植物入侵中，外来植物会对本地植物的花卉访问产生负面影响，但本地植物也可以通过提供预先适应的传粉媒介群落来促进密切相关的外来植物的早期建立。因此，关于植物-传粉媒介相互作用在入侵群落组装中的作用，有两个明显矛盾的主要概念。首先，一种同时出现的本地植物可以对具有相似花性特征的外来植物的花访问产生积极影响（传粉者促进）（Brown & Kodric-Brown,  1979 ; Rathcke,  1983 ; Thomson,  1978 ; Waser & Real,  1979）。其次，同时出现的本地植物会对具有相似花性状的外来植物的花访问产生负面影响（传粉媒介介导的竞争；Schemske 等人，  1978 年；Waser，  1982 年；Zimmerman，  1980 年），提供了优势外来植物与本地植物具有不同的性状。此外，为了吸引作为潜在传粉者的花卉访客，花卉性状应与传粉者性状相匹配（达尔文，  1862 年）。因此，外来植物与常驻传粉者之间新的植物-传粉者相互作用的形成应取决于外来植物与本地植物之间的花性状相似程度。

花的特征，例如花的对称性、颜色和大小，可以作为花卉参观者定位花奖励的信号，因此被认为是介导植物-传粉媒介相互作用的重要特征（Fornoff 等，  2017；Junker 等。 ，  2015 年；Ortiz 等人，  2021 年；Reverté 等人，  2016 年）。此外，花蜜生产已被证明会影响共花植物与共享授粉媒介之间的间接相互作用（Carvalheiro 等人，  2014 年）。尽管如此，植物群落生态学研究中仍然普遍缺少与传粉媒介吸引有关的花性特征（E-Vojtkó 等人，  2020 年；Sargent 和 Ackerly，  2008 年））。一项旨在了解外来物种对本土物种授粉和繁殖成功影响的荟萃分析考虑了花卉特征，并表明外来植物和本土植物在花朵对称性和颜色方面的相似性增加了对传粉媒介的竞争（Morales & Traveset，  2009 年） . 因此，这些花性状的相似性可能在传粉媒介介导的外来植物相互作用中起关键作用。

新的植物-传粉媒介相互作用的模式可能在系统发育和花性状距离方面有所不同。虽然系统发育相关性经常被认为是性状相似性的代表，但一些花性状可能在进化上不是保守的（Sargent & Ackerly,  2008）。例如，密切相关的物种通常具有不同的花色（Eaton 等人，  2012 年；Shrestha 等人，  2014 年），并且花色在物种内可能会有所不同（Coetzee 等人，  2021 年）。因此，重要的是要同时考虑系统发育距离和性状差异（Cadotte 等人，  2013；Lemoine 等人，  2015）并解开两者的影响（Marx 等人，  2016）。因此，新型植物-传粉媒介相互作用的模式可能比达尔文的归化难题（Diez 等人，  2008 年）所预测的更为复杂。此外，不同的机制可能同时发挥作用来推动外来物种的成功，正如植物与植物之间的直接相互作用所证明的那样（Malecore 等人，  2019 年；Sheppard 等人，  2018 年）。当传粉媒介促进和对传粉媒介的竞争都发挥作用时，在与本地物种的中间系统发育和花卉性状距离处，对外来物种的花卉访问量应该最高或最低。因此，重要的是要考虑到外来物种的花卉访问与本地物种的系统发育和花卉性状距离之间的非线性关系。

在一项田间实验中，我们通过将盆栽外来植物添加到共同开花的本地社区中来模拟最近入侵的社区，我们测试了对外来植物的花卉访问是否取决于与本地植物的系统发育相关性和功能相似性。如果通过共同开花具有相似花卉特征的本地植物促进对外来植物的花访问，我们预测对外来植物的花访问与与本地人的系统发育或花特征距离之间存在负相关关系。如果对外来植物的花卉访问受到具有相似花卉特征（传粉媒介介导的竞争）的共同开花本地植物的阻碍，我们预测了一种正相关关系。如果促进和竞争都在起作用，我们预测了在中间系统发育或花卉性状距离（即驼峰或 U 形关系）处与高或低花卉访问的非线性关系。更具体地说，我们询问（a）对外来植物的花卉访问是否与外来物种和本地物种之间的系统发育和花卉性状距离有关，以及（b）花卉访问者组成的相似性是否与外来和本地物种之间的系统发育和花卉性状距离有关。本土物种。