1 INTRODUCTION

Mutualistic interactions, such as those between plants and their seed dispersers, are vital for maintaining the structure, stability and functioning of ecological communities (Bascompte & Jordano, 2014). Invasive non-native species can alter the composition of recipient communities (Hejda et al., 2009; Levine et al., 2003) with effects on diversity and on the intricate network of biotic interactions that sustain ecological communities (Heleno, 2020; Vilà et al., 2011). For instance, the introduction of non-native frugivores can change the recruitment patterns of native plant species both directly, e.g. by dispersing the seeds of non-native plants (Mandon-Dalger et al., 2004), or indirectly, e.g. by reducing the services provided by native seed dispersers (Traveset & Richardson, 2014).

The magnitude of such disturbances is highly variable and is contingent on multiple factors. There is increasing evidence that the relative timing of non-native and native fruiting phenologies plays a critical role in the rate at which non-native species invade native communities as it may determine the potential for diverting seed dispersal services (Heleno, 2020). For example, when non-native species fruit synchronously with natives they can directly compete for seed dispersal services. This competition can divert frugivore visits away from native plants (Heleno, 2020; Kueffer et al., 2009), thereby reducing native seed rain and recruitment (Rowles & O'Dowd, 2009; Traveset et al., 2012). Conversely, non-native plant species with phenologies that exploit unoccupied temporal niches in native communities (i.e. those that fruit outside of the main native fruiting season), will likely attract a high number of seed dispersers which might facilitate the invasion (Heleno, Olesen, et al., 2013; Williams & Karl, 1996). The time of fruiting can also affect the assembly of available seed dispersers and even the main direction of dispersal (González-Varo et al., 2021). However, to date, few studies evaluated the importance of (a-) synchronous fruiting phenology on seed dispersal services and recruitment patterns. Furthermore, most of our knowledge on the effects of non-native species on seed dispersal interactions is derived from static (i.e. temporally aggregated) seed dispersal interaction networks (e.g. Heleno, Olesen, et al., 2013; Heleno, Ramos, et al., 2013; Vizentin-Bugoni et al., 2019), and it remains uncertain how plant–seed disperser interactions are affected by the variable pressure of biological invasions under natural conditions.

Species interactions vary considerably across multiple time-scales (CaraDonna et al., 2021; Costa et al., 2020; Trøjelsgaard & Olesen, 2016). Within a season, the addition or loss of interactions will be determined by which species temporarily co-occur, along with species relative abundance and the availability of suitable resources (phenological matching; CaraDonna et al., 2021; Olesen et al., 2010; Vázquez et al., 2009). For instance, given that frugivores show high plasticity in their foraging behaviour and fruit preferences (Carnicer et al., 2009), seed removal rates and the number and identity of plant–frugivore interactions, strongly depend on the fruiting neighbourhood (Albrecht et al., 2015; Carlo & Morales, 2008; Rumeu et al., 2019). This temporal variation in the identity and strength of interactions arises even in tropical ecosystems (e.g. Kaiser-Bunbury et al., 2014; Ramos-Robles et al., 2016), where seasonality is less marked than in temperate regions. In this context, the analysis of aggregated networks that span several seasons cannot fully capture the dynamics of interactions and is likely to provide an incomplete overview of the seasonal effects of biological invasions on species interactions and community dynamics (Arroyo-Correa et al., 2020). Therefore, explicitly incorporating the temporal dimension, is likely very important to understanding how non-native species integrate and affect native seed dispersal networks.

Here we investigated the impact of plant invasions on the seasonal dynamics of seed dispersal function and networks across a large-scale experimental setup and a plant invasion gradient on a tropical island. We primarily focus on the direct effects of plant invasion on native plant and frugivore communities and their interactions, but also explore indirect effects mediated through native and non-native frugivores. We collected data on fruit and frugivore abundances and feeding interactions over time to address the following questions:

1. How does the proportion of non-native fleshy fruits alter seed dispersal frequency of native plants, and do these effects differ between seasons?
2. How do seasonality and the level of plant invasion affect the structure of seed dispersal networks and determine the roles of native and non-native frugivores?

Given the tropical climate and relatively small size of the island and the lack of migratory bird species in the frugivore community, we anticipated little changes in the frugivore assemblages between seasons. We, therefore, predicted that both fruit traits of non-native plants and relative densities of fruits of native and non-native plants are primary drivers of changes in frugivore movement and foraging behaviour and that these drivers differ between seasons. Specifically, we expected that non-native plants fruiting outside the main fruiting period have a higher probability of attracting seed dispersers (Muñoz & Ackerman, 2013) and that those non-natives fruiting alongside with native plants, both inside and outside the main season, will compete with, and divert seed dispersal services away from native plants (Heleno, Ramos, et al., 2013; Kueffer et al., 2009). These indirect effects between native and non-native plants could equally be mediated through native and non-native frugivores. Finally, we predicted strong temporal variation in plant–frugivore interactions as a result of changes in the fruit availability and plant community composition (Arroyo-Correa et al., 2020; CaraDonna et al., 2021), with more links, and greater diversity and overlap of plant–frugivore interactions during the fruiting peak (Albrecht et al., 2015).

中文翻译：

---

非本地物种对热带种子传播网络影响的季节性变化

1 简介

诸如植物及其种子传播者之间的相互作用，对于维持生态群落的结构、稳定性和功能至关重要（Bascompte & Jordano，  2014 年）。外来入侵物种可以改变受体群落的组成（Hejda et al.,  2009 ; Levine et al.,  2003），对多样性和维持生态群落的复杂的生物相互作用网络产生影响（Heleno,  2020 ; Vilà et al., 2003）等人，  2011 年）。例如，引入非本地食果动物可以直接改变本地植物物种的补充模式，例如通过分散非本地植物的种子（Mandon-Dalger et al.,  2004)，或间接地，例如通过减少本地种子传播者提供的服务（Traveset & Richardson，  2014 年）。

这种干扰的幅度是高度可变的，并且取决于多种因素。越来越多的证据表明，非本地和本地结果物候的相对时间在非本地物种入侵本地群落的速度中起着关键作用，因为它可能决定转移种子传播服务的潜力（Heleno，  2020 年）。例如，当非本地物种与本地物种同步结果时，它们可以直接竞争种子传播服务。这种竞争可以将食果动物的访问从本地植物转移（Heleno，  2020；Kueffer 等人，  2009），从而减少本地种子雨和补充（Rowles & O'Dowd，  2009；Traveset 等人，  2012）。相反，具有利用本地群落中未被占用的时间生态位的物候的非本地植物物种（即在主要本地结果季节之外结果的那些）可能会吸引大量种子传播者，这可能会促进入侵（Heleno，Olesen，等人，  2013 年；威廉姆斯和卡尔，  1996 年）。结果的时间也会影响可用种子分散器的组装甚至传播的主要方向（González-Varo et al.,  2021）。然而，迄今为止，很少有研究评估（a-）同步结果物候对种子传播服务和补充模式的重要性。此外，我们关于非本地物种对种子传播相互作用影响的大部分知识来自静态（即时间聚合）种子传播相互作用网络（例如 Heleno，Olesen 等人，  2013 年；Heleno，Ramos 等人。 ,  2013 ; Vizentin-Bugoni 等人,  2019 )，目前尚不确定植物-种子分散器相互作用如何受到自然条件下生物入侵的可变压力的影响。

物种相互作用在多个时间尺度上差异很大（CaraDonna 等人，  2021 年；Costa 等人，  2020 年；Trøjelsgaard 和 Olesen，  2016 年）。在一个季节内，相互作用的增加或丧失将取决于哪些物种暂时共存，以及物种的相对丰度和合适资源的可用性（物候匹配；CaraDonna 等，  2021；Olesen 等，  2010； Vázquez 等人，  2009 年）。例如，鉴于食果动物在觅食行为和水果偏好方面表现出高度可塑性（Carnicer 等，  2009)、种子去除率以及植物-食果动物相互作用的数量和特性在很大程度上取决于结果邻域（Albrecht 等人，  2015 年；Carlo & Morales，  2008 年；Rumeu 等人，  2019 年）。这种相互作用的特性和强度的时间变化甚至出现在热带生态系统中（例如 Kaiser-Bunbury 等人，  2014 年；Ramos-Robles 等人，  2016 年）)，那里的季节性不如温带地区那么明显。在这种情况下，对跨越多个季节的聚合网络的分析无法完全捕捉相互作用的动态，并且可能无法完整概述生物入侵对物种相互作用和群落动态的季节性影响（Arroyo-Correa 等，  2020）。因此，明确地结合时间维度，对于理解非本地物种如何整合和影响本地种子传播网络可能非常重要。

在这里，我们通过大规模实验装置和热带岛屿上的植物入侵梯度，研究了植物入侵对种子传播功能和网络的季节性动态的影响。我们主要关注植物入侵对原生植物和食果动物群落及其相互作用的直接影响，但也探索通过原生和非原生食果动物介导的间接影响。我们随着时间的推移收集了有关水果和食果动物丰度和摄食相互作用的数据，以解决以下问题：

1. 非本地肉质果实的比例如何改变本地植物的种子传播频率，这些影响在不同季节是否有所不同？
2. 季节性和植物入侵水平如何影响种子传播网络的结构并确定本地和非本地食果动物的作用？

鉴于热带气候和岛屿相对较小的面积以及食果动物群落中缺乏候鸟物种，我们预计季节之间的食果动物组合变化不大。因此，我们预测，非本地植物的果实性状以及本地和非本地植物果实的相对密度都是食果动物运动和觅食行为变化的主要驱动因素，并且这些驱动因素在季节之间有所不同。具体来说，我们预计在主要结果期之外结果的非本地植物有更高的概率吸引种子传播者（Muñoz & Ackerman,  2013) 并且那些在主季内外与本地植物一起结果的非本地植物将与本地植物竞争并转移种子传播服务（Heleno、Ramos 等人，  2013 年；Kueffer 等人，  2009 年）。本地和非本地植物之间的这些间接影响同样可以通过本地和非本地食果动物来介导。最后，我们预测，由于水果可用性和植物群落组成的变化，植物与食果动物之间的相互作用会出现强烈的时间变化（Arroyo-Correa 等人，  2020 年；CaraDonna 等人，  2021 年），具有更多的联系和更大的多样性以及结果高峰期植物-食果动物相互作用的重叠（Albrecht 等人，  2015 年）。