Insect seed-predator networks respond positively to restoration on a tropical island

When non-native species invade ecosystems, coevolved plant-animal interactions and associated ecological functions are altered, often to the detriment of local biodiversity. While mutualistic interactions can benefit from —and assist with— ecological restoration through the removal of non-native species, community-level changes in antagonistic interactions due to restoration are less well understood. Insect seed predators provide important ecosystem functions, but can also have antagonistic effects through pre-dispersal seed predation which can cause fruit abortion or premature fruit drop, potentially affecting plant community composition.

Here, we used a network approach and data on trophic interactions from a large-scale field experiment to investigate the differences in insect seed predator communities between native restored sites (where non-native plants were removed) and unrestored sites dominated by non-native plants on the island of Mahé, Seychelles.

Restored sites contained more individuals and species of seed predators (up by 81% and 50% respectively), and experienced a greater predation intensity (1.06 ± 0.53 vs. 0.27 ± 0.09), and predator specialisation compared to unrestored sites. The proportion of predated samples was similar between treatments (~14%). Beta diversity, measured as species and interaction turnover, increased at restored sites.

Synthesis and applications: Our findings imply that vegetation restoration generates marked biodiversity benefits on native seed predator communities and their interactions. However, removing non-native plants did not result in the elimination of non-native seed predators, which might require targeted control measures. Our results highlight the importance of considering the effects of restoration on antagonistic interactions, in addition to those reported for pollination and seed dispersal services. Management practices should aim to control plant invasions for the conservation of native plant communities that serve as resources and refugia both for mutualists (e.g. seed dispersers, pollinators) and native insect seed-predator antagonists, both of which benefit native ecosystem function.

Methods
The study was carried out on eight inselbergs (steep-sided granitic rock outcrops) of ca. 1 ha on Mahé, the largest granitic island of the Seychelles archipelago, Indian Ocean (- 4.6953°, 55.4999°, 154 km2, 900 m a.s.l.). To restore inselberg communities, all non-native woody plants were removed from four of the inselbergs between 2011 and 2012 by cutting individual plants and applying herbicide to cut stumps (Kaiser-Bunbury et al., 2015), with follow-up control thereafter. The other four inselbergs remained unchanged with both native and non-native species (Kaiser-Bunbury et al., 2017).

To estimate fruit availability at each site, we recorded monthly fruit abundance by counting all ripe fruits of woody fleshy-fruited plant species that are mainly dispersed by endozoochory along fixed linear transects.

To collect seed predators, we sampled fleshy fruits monthly at the eight study sites over one year (Sep. 2018 to Aug. 2019). Samples of ten fruits were collected opportunistically from all individuals of fleshy-fruited species with >10 fruits at each site.

For further information, please see the "Material and methods" section in the associated manuscript

当外来物种入侵生态系统时，协同演化的动植物相互作用及相关联的生态功能会发生改变，且往往会对本地生物多样性造成损害。尽管互利共生相互作用可通过移除外来物种助力生态修复并从中获益，但人们对修复过程中拮抗相互作用的群落级变化却知之甚少。昆虫种子捕食者（insect seed predator）虽能提供重要的生态系统服务功能，但也会通过扩散前种子捕食行为产生拮抗效应：这类捕食会引发果实败育或早落，进而可能改变植物群落组成。

本研究依托大型野外实验获取的营养相互作用数据，结合网络分析方法，针对塞舌尔马埃岛的两类生境——已移除外来植物的本地修复样地，以及以外来植物为主导的未修复样地——展开昆虫种子捕食者群落的差异分析。

与未修复样地相比，修复样地内的种子捕食者个体数与物种数均显著更高（分别提升81%与50%），捕食强度（1.06 ± 0.53 对比 0.27 ± 0.09）及捕食者特化程度也更高。两类样地的被捕食样本占比相近（约14%）。以物种与相互作用周转衡量的β多样性在修复样地中有所提升。

综合与应用：本研究结果表明，植被修复可显著提升本地种子捕食者群落及其相互作用的生物多样性效益。但移除外来植物并未完全清除外来种子捕食者，这或许需要针对性的防控措施。本研究结果同时强调，除授粉与种子传播服务相关效应外，还应重视修复过程对拮抗相互作用的影响。在管理实践中，应通过防控植物入侵来保护本地植物群落——这类群落既是互利共生生物（如种子传播者、传粉者）的资源与庇护所，也是本地昆虫种子捕食者拮抗物的栖息场所，而二者均有助于维持本地生态系统功能。

研究方法
本研究在印度洋塞舌尔群岛最大的花岗岩岛马埃岛开展，选取8处面积约1公顷的岛状花岗岩露头（inselberg，即陡峭边坡的花岗岩岩出露地）。2011至2012年间，研究人员通过砍伐植株并在伐桩施用除草剂的方式，移除了其中4处岛状露头的所有外来木本植物，并在后续进行了防控维护（Kaiser-Bunbury等，2015）。剩余4处岛状露头维持原状，同时保留本土与外来植物物种（Kaiser-Bunbury等，2017）。

为评估各样地的果实可获得性，研究人员沿固定线性样带，对主要通过内消化传播（endozoochory）的肉质果木本植物的所有成熟果实进行计数，以记录月度果实丰度。

为收集种子捕食者，研究团队在2018年9月至2019年8月的一年内，每月对8处研究样地的肉质果实进行采样。针对每个样地内果实数量超过10个的肉质果物种，随机采集10枚果实样本。

如需获取更多详细信息，请参阅相关论文的"材料与方法"部分。