1 INTRODUCTION

There is compelling evidence that biodiversity positively influences ecosystem functioning across numerous experimental and real-world ecosystems (Cardinale et al., 2011; Flombaum & Sala, 2008; Gonzalez et al., 2020; Grace et al., 2016; Guerrero-Ramírez et al., 2017; Tilman et al., 2014). Specifically, positive effects of species coexistence through niche or resource partitioning and facilitation (Barry et al., 2019) promote higher ecosystem functioning through higher plant diversity and support the ability of more diverse plant communities to produce more biomass through complementarity (Cardinale et al., 2007; Hooper et al., 2005; Loreau & Hector, 2001). Alternatively, positive effects of plant diversity on biomass may be explained by more diverse plant communities having one or few species that are highly productive (Loreau & Hector, 2001). However, studies on the relationships between biodiversity and productivity in naturally assembled forests have shown contrasting results across environmental gradients (Paquette et al., 2018; Ratcliffe et al., 2017; but see Liang et al., 2016). This suggests that elucidating biodiversity-ecosystem functioning (BEF) relationships in real-world ecosystems requires considering environmental conditions and – possibly—the biogeographic context in which these relationships occur.

Biotic and abiotic conditions have been found to strongly influence BEF relationships in forests (e.g. Craven et al., 2020; Fei et al., 2018; Jing et al., 2022; Mina et al., 2018; Ratcliffe et al., 2017), with forest types, geographic regions and climatic conditions mediating the impacts of biodiversity on ecosystem functioning (Figure 1; Forrester, 2014; Grossiord et al., 2014; Jucker et al., 2016; Liang et al., 2016; Paquette & Messier, 2011; Pretzsch et al., 2013; Ratcliffe et al., 2016). For instance, water availability influences the strength of the BEF relationship in forests, with water-limited regions showing stronger positive BEF relationships than regions with higher water availability (Jing et al., 2022; Ratcliffe et al., 2017). Across latitudes, positive tree diversity-biomass relationships have been found consistently in temperate forests, while in tropical forests positive, negative and neutral relationships have been observed (van der Plas, 2019). In addition, biogeographic context, that is the contrasting geographical locations, their geological histories and their impact on shaping the processes generating biodiversity such as speciation and dispersal (Vellend, 2017), could be an important driver of the direction and magnitude of BEF relationships as they shape biodiversity patterns across spatiotemporal scales (e.g. Cai et al., 2023; Keil & Chase, 2019). Yet, the potential influence of biogeographic context on BEF relationships has rarely been assessed.

Theories that attempt to disentangle biodiversity spatial patterns may contribute directly or indirectly to understanding biodiversity effects on ecosystem functioning. For instance, the more-individuals hypothesis proposes that the total number of individuals in a community limits the number of species with viable populations, with a higher number of individuals resulting in higher species richness via passive sampling and therefore potentially higher productivity (Figure 1; Gaston, 2000; Srivastava & Lawton, 1998). Conversely, the species-energy hypothesis proposes that resource availability, for example temperature, water availability, limits population sizes in a given area, regulating the number of individuals, species richness and, consequently, the capacity of a given community to produce biomass via niche-based processes (Figure 1; Brown, 2014; Wright, 1983).

Early BEF studies focused on how species richness drives ecosystem functioning (van der Plas, 2019) and largely ignored the influence of species abundances, evolutionary history, and functional differences among species. Additionally, most studies in naturally assembled ecosystems investigate taxonomic diversity, despite the potential influence of other biodiversity facets on ecosystem functioning (Hagan et al., 2023; van der Plas, 2019). By embracing the multifaceted nature of biodiversity, we can examine ecological and evolutionary processes shaping species assemblages and diversity patterns beyond species counts, allowing us to better understand the impact of biodiversity change on ecosystem functioning (Cadotte et al., 2011; Díaz et al., 2007; Emerson & Gillespie, 2008). Specifically, functional diversity, via niche complementarity, not diversity per se influences ecosystem functioning (Dı́az & Cabido, 2001; Flynn et al., 2009; Loreau, 1998; Tilman et al., 1997). Further, phylogenetic diversity may explain variation in ecosystem functioning (Cadotte et al., 2008; Venail et al., 2015) by representing the diversity of phylogenetically conserved functional traits and integrating a greater number of traits than the soft ones usually used to estimate functional diversity (Nock et al., 2016). However, biodiversity-productivity relationships may be driven by functional traits that are not phylogenetically conserved, and would therefore be harder to capture by phylogenetic diversity indices alone (Craven et al., 2018), highlighting the importance of including additional facets of diversity (Figure 1).

Their limited area, varying levels of isolation, and contrasting geological histories make oceanic islands perfect study systems to understand how different ecological and evolutionary processes shape diversity patterns and species assemblages, and how biogeographic context influences them (Hagan et al., 2021; Warren et al., 2015; Weigelt et al., 2015; Whittaker & Fernández-Palacios, 2007). Processes such as rare dispersal events, environmental filtering, and in-situ speciation have generated unique, highly endemic plant assemblages on islands worldwide (Weigelt et al., 2015). The biased representation of higher taxa compared to the source pool on islands due to dispersal, environmental, and biotic filters, that is disharmony, (Carlquist, 1974; König et al., 2021; Kraft et al., 2015) may result in closely related species occupying different ecological niches, potentially increasing ecosystem functioning. Yet, unfilled niche space may persist on oceanic islands due to limited colonisation and low species diversity. Further, the high proportion of endemic species makes island ecosystems more susceptible to the naturalisation of non-native species (Moser et al., 2018; Sax et al., 2002), particularly on islands where phylogenetic relatedness among native species is higher (Bach et al., 2022). Moreover, there is a trend of disproportionate losses of island endemic species, as around 60% of all recorded extinctions took place on islands (Whittaker & Fernández-Palacios, 2007).

Their low species richness and high invasibility makes island ecosystems sensitive to changes in the abundance of keystone species, with potential repercussions on nutrient cycling, primary productivity and carbon storage (Worm & Duffy, 2003). In addition, anthropogenic impacts on island biodiversity and ecosystem functioning are a growing concern. Non-native species—particularly invasive species—and land-use change can modify ecosystem structure and function by negatively impacting native island ecosystems through changes in nutrient cycling, carbon storage, altering species composition, and increasing fire risk (Figure 1; Mascaro et al., 2012; Rothstein et al., 2004; Vitousek et al., 1996; Vitousek et al., 1997). However, non-native species may also increase certain ecosystem functions in highly degraded islands, enhancing soil structure and fertility and restoring forest cover (Lugo, 2004). Similarly, there is evidence that novel forests, that is forests resulting from a mixture of native and non-native species, can be more diverse and provide higher levels of ecosystem functioning than uninvaded native forests (Mascaro et al., 2012).

Here, we examine the relationship between taxonomic, phylogenetic, and functional diversity and above-ground productivity, an important ecosystem function, in forests on the Canary Islands and climatically similar areas in mainland Spain, and the extent to which multifaceted BEF relationships are mediated by biogeographic context. To this end, we used forest inventory data from Spain to first examine a simple version of the biodiversity-productivity relationship, that is not accounting for environmental conditions, for island and mainland forests. Later, we included the direct and indirect influence of environmental conditions and ecological processes that may affect this relationship (Figure 1). Specifically, we expected (1) positive multifaceted biodiversity-productivity relationships due to complementarity acting on both island and mainland forests, with phylogenetic diversity having a stronger influence on productivity than taxonomic or functional diversity as it can estimate the functional trait space of a community and also reflect species interactions (Srivastava et al., 2012), and phylogenetic diversity has been found to promote ecosystem functions and stability (Cadotte et al., 2012; van der Plas, 2019; Venail et al., 2015). We hypothesised that the magnitude of multifaceted biodiversity-productivity relationships differ between biogeographic contexts, with stronger relationships on islands because a large proportion of island biota evolved in these ecosystems and therefore developed specific traits to more efficiently use limited resources or to persist in harsh environments (Barajas Barbosa et al., 2023; Emerson & Gillespie, 2008); (2) environmental conditions, that is climate and soil properties, influence productivity on island and mainland forests, as bioclimatic variables and soil nutrients have been shown to strongly influence plant diversity (Kreft & Jetz, 2007; Lambers et al., 2011) and productivity (following the species-energy hypothesis); (3) the number of individuals positively influences productivity on island and mainland forests, as a higher number of individuals is expected to host a more biodiverse community (following the more-individuals hypothesis), which subsequently is expected to yield higher productivity (Gaston, 2000; Srivastava & Lawton, 1998); and (4) non-native species potentially influence productivity in both island and mainland forests, with the effect being stronger and positive on islands as non-native species likely perform different functions than native species (Rothstein et al., 2004; Vitousek et al., 1996). However, as the Canary Islands forests may not have been impacted by non-native species as extensively as other oceanic islands (Fernández-Palacios et al., 2023), the influence of non-native species on ecosystem functioning may be similar to that of mainland forests.

中文翻译：

---

生物地理环境调节岛屿和大陆森林的多方面多样性与生产力关系

1 简介

有令人信服的证据表明，生物多样性对众多实验和现实世界生态系统的生态系统功能产生积极影响（Cardinale 等，  2011；Flombaum 和 Sala，  2008；Gonzalez 等，  2020；Grace 等，  2016；Guerrero-Ramírez 等等人，  2017 年；蒂尔曼等人，  2014 年）。具体而言，通过生态位或资源分配和便利化实现物种共存的积极影响（Barry 等，  2019）通过更高的植物多样性促进更高的生态系统功能，并支持更多样化的植物群落通过互补性产生更多生物量的能力（Cardinale 等，2019）。 ，  2007 年；Hooper 等人，  2005 年；Loreau 和 Hector，  2001 年）。或者，植物多样性对生物量的积极影响可以通过更多样化的植物群落来解释，该群落具有一种或几种高产物种（Loreau & Hector，  2001）。然而，关于自然聚集森林生物多样性和生产力之间关系的研究显示了不同环境梯度之间的对比结果（Paquette et al.,  2018；Ratcliffe et al.,  2017；但参见Liang et al.,  2016）。这表明，阐明现实世界生态系统中的生物多样性与生态系统功能（BEF）关系需要考虑环境条件以及（可能）这些关系发生的生物地理背景。

已发现生物和非生物条件强烈影响森林中的 BEF 关系（例如 Craven 等人，  2020；Fei 等人，  2018；Jing 等人，  2022；Mina 等人，  2018；Ratcliffe 等人，  2017 ） ），森林类型、地理区域和气候条件调节了生物多样性对生态系统功能的影响（图 1；Forrester，  2014；Grossiord 等，  2014；Jucker 等，  2016；Liang 等，  2016；Paquette & Messier，  2011；Pretzsch 等，  2013；Ratcliffe 等，  2016）。例如，水资源可用性影响森林中 BEF 关系的强度，水资源有限的地区比水资源可用性较高的地区表现出更强的正 BEF 关系（Jing 等，  2022；Ratcliffe 等，  2017）。在不同纬度地区，温带森林中一致发现了树木多样性与生物量的积极关系，而在热带森林中则观察到了积极、消极和中性关系（van der Plas，  2019）。此外，生物地理背景，即不同的地理位置、地质历史及其对物种形成和扩散等生物多样性产生过程的影响（Vellend，  2017），可能是 BEF 关系方向和规模的重要驱动因素，因为它们塑造了跨时空尺度的生物多样性模式（例如 Cai 等人，  2023；Keil & Chase，  2019）。然而，生物地理环境对 BEF 关系的潜在影响却很少被评估。

试图理清生物多样性空间格局的理论可能直接或间接有助于理解生物多样性对生态系统功能的影响。例如，更多个体假说提出，群落中的个体总数限制了具有生存种群的物种数量，个体数量越多，通过被动采样就会导致更高的物种丰富度，从而潜在地提高生产力（图 1；加斯顿，  2000 年；斯里瓦斯塔瓦和劳顿，  1998 年）。相反，物种能量假说提出，资源可用性（例如温度、水可用性）限制了给定区域的人口规模，调节了个体数量、物种丰富度，从而调节了给定群落通过生态位生产生物量的能力基于流程（图 1；Brown，  2014；Wright，  1983）。

早期的 BEF 研究重点关注物种丰富度如何驱动生态系统功能（van der Plas，  2019），而很大程度上忽略了物种丰富度、进化历史和物种间功能差异的影响。此外，尽管其他生物多样性方面对生态系统功能存在潜在影响，但大多数自然组装生态系统的研究都调查了分类多样性（Hagan 等人，  2023 年；van der Plas，  2019 年）。通过接受生物多样性的多方面性质，我们可以研究塑造物种组合和物种计数之外的多样性模式的生态和进化过程，使我们能够更好地了解生物多样性变化对生态系统功能的影响（Cadotte等人，  2011年；Díaz等人，2011年）。 ，  2007 年；艾默生和吉莱斯皮，  2008 年）。具体而言，功能多样性通过生态位互补性而非多样性本身影响生态系统功能（Dı́az & Cabido,  2001；Flynn et al.,  2009；Loreau,  1998；Tilman et al.,  1997）。此外，系统发育多样性可以通过代表系统发育上保守的功能性状的多样性并整合比通常用于估计功能的软性状更多的性状来解释生态系统功能的变化（Cadotte et al.,  2008；Venail et al.,  2015）。多样性（Nock 等人，  2016）。然而，生物多样性-生产力关系可能是由系统发育上不保守的功能特征驱动的，因此更难以仅通过系统发育多样性指数来捕获（Craven et al.,  2018），强调了包括多样性的其他方面的重要性（图1）。

有限的面积、不同程度的隔离以及截然不同的地质历史使海洋岛屿成为完美的研究系统，以了解不同的生态和进化过程如何塑造多样性模式和物种组合，以及生物地理环境如何影响它们（Hagan 等人，  2021；Warren 等人）等人，  2015 年；Weigelt 等人，  2015 年；Whittaker 和 Fernández-Palacios，  2007 年）。罕见的扩散事件、环境过滤和原位物种形成等过程在世界各地的岛屿上产生了独特的、高度特有的植物组合（Weigelt 等，  2015）。由于分散、环境和生物过滤，与岛屿上的源池相比，较高类群的代表性存​​在偏差，即不和谐（Carlquist，  1974；König 等，  2021；Kraft 等，  2015）可能会导致密切相关相关物种占据不同的生态位，可能增强生态系统功能。然而，由于殖民活动有限和物种多样性较低，未填补的生态位空间可能会在海洋岛屿上持续存在。此外，高比例的特有物种使得岛屿生态系统更容易受到非本地物种归化的影响（Moser 等，  2018；Sax 等，  2002），特别是在本地物种之间的系统发育相关性较高的岛屿上（Bach）等人，  2022）。此外，岛屿特有物种有不成比例消失的趋势，因为所有有记录的灭绝事件中约 60% 发生在岛屿上（Whittaker & Fernández-Palacios，  2007 年）。

其低物种丰富度和高入侵性使得岛屿生态系统对关键物种丰度的变化敏感，对养分循环、初级生产力和碳储存产生潜在影响（Worm & Duffy，  2003）。此外，人为对岛屿生物多样性和生态系统功能的影响日益受到关注。非本地物种（尤其是入侵物种）和土地利用变化可以通过改变养分循环、碳储存、改变物种组成和增加火灾风险，对本地岛屿生态系统产生负面影响，从而改变生态系统结构和功能（图 1；Mascaro 等人） .,  2012 ; Rothstein 等人,  2004 ; Vitousek 等人,  1996 ; Vitousek 等人,  1997 )。然而，非本地物种也可能增加高度退化岛屿的某些生态系统功能，增强土壤结构和肥力并恢复森林覆盖（Lugo，  2004）。同样，有证据表明，新型森林，即由本地和非本地物种混合而成的森林，可以比未入侵的本地森林更加多样化，并提供更高水平的生态系统功能（Mascaro 等，  2012）。

在这里，我们研究了加那利群岛森林和西班牙大陆气候相似地区的分类学、系统发育和功能多样性与地上生产力（一种重要的生态系统功能）之间的关系，以及多方面 BEF 关系在多大程度上由生物地理背景。为此，我们使用西班牙的森林清查数据首先检查了岛屿和大陆森林的生物多样性与生产力关系的简单版本，该关系不考虑环境条件。后来，我们纳入了可能影响这种关系的环境条件和生态过程的直接和间接影响（图1）。具体来说，我们预计（1）由于岛屿和大陆森林的互补性，生物多样性与生产力之间存在积极的多方面关系，系统发育多样性比分类或功能多样性对生产力的影响更大，因为它可以估计群落的功能特征空间，并且也反映了物种之间的相互作用（Srivastava et al.,  2012），并且系统发育多样性被发现可以促进生态系统功能和稳定性（Cadotte et al.,  2012；van der Plas,  2019；Venail et al.,  2015）。我们假设，生物地理环境之间的多方面生物多样性与生产力关系的程度有所不同，在岛屿上的关系更强，因为很大一部分岛屿生物群在这些生态系统中进化，因此发展出特定的特征，以更有效地利用有限的资源或在恶劣的环境中持续存在。 Barajas Barbosa 等人，  2023；Emerson & Gillespie，  2008）； (2) 环境条件，即气候和土壤特性，影响岛屿和大陆森林的生产力，因为生物气候变量和土壤养分已被证明强烈影响植物多样性（Kreft & Jetz，  2007；Lambers 等，  2011）；生产力（遵循物种能量假说）； (3) 个体数量对岛屿和大陆森林的生产力产生积极影响，因为更多的个体预计会形成一个生物多样性更丰富的群落（遵循更多个体假设），从而预计会产生更高的生产力（Gaston，  2000 年；斯里瓦斯塔瓦和劳顿，  1998 年）； (4) 非本地物种可能影响岛屿和大陆森林的生产力，对岛屿的影响更为强烈和积极，因为非本地物种可能发挥与本地物种不同的功能（Rothstein 等人，  2004 年；Vitousek 等人） .,  1996）。然而，由于加那利群岛森林可能不像其他海洋岛屿那样受到非本地物种的广泛影响（Fernández-Palacios 等，  2023），因此非本地物种对生态系统功能的影响可能与大陆森林。