Data from: Fine-tuning the nested structure of pollination networks by adaptive interaction switching, biogeography and sampling effect in the Galápagos Islands

The structure of pollination networks, particularly its nestedness, contains important information on network assemblages. However, there is still limited understanding of the mechanisms underlying nested pollination network structures. Here, we investigate the role of Adaptive Interaction Switching (AIS), island area, isolation, age and sampling effort in explaining the nestedness of pollination networks across ten Galápagos Islands. The AIS algorithm is inspired by Wallace's elimination of the unfit, where a species constantly replaces its least profitable mutualistic partner with a new partner selected at random. To explain network structures, we first use a dynamic model that includes functional response of pollination and AIS, with only species richness and binary connectance as input (hereafter the AIS model). Thereafter, other explanatory variables (isolation, area, age and sampling effort) were added to the model. In four out of ten islands, the pollination network was significantly nested, and predictions from the AIS model correlated with observed structures, explaining 69% variation in nestedness. Overall, in terms of independent contribution from hierarchical partitioning of variation in observed nestedness, the AIS model predictions contributed the most (37%), followed by sampling effort (28%) and island area (22%), with only trivial contributions from island isolation and age. Therefore, adaptive switching of biotic interactions seems to be key to ensure network function, with island biogeographic factors only secondary. Although large islands could harbour more diverse assemblages and thus foster more nested structures, sufficient sampling proves to be essential for detecting non-random network structures.

传粉网络的结构，尤其是其嵌套性（nestedness），蕴含着关于网络群落组成的重要信息。然而，当前学界对传粉网络嵌套结构背后的形成机制仍知之甚少。本研究针对加拉帕戈斯群岛（Galápagos Islands）的10个岛屿，探究了适应性交互切换（Adaptive Interaction Switching, AIS）、岛屿面积、隔离度、岛龄以及采样强度对传粉网络嵌套性的解释效力。AIS算法的灵感源自华莱士的适者淘汰理论：物种会持续将自身互利性最低的共生伙伴替换为随机选取的新伙伴。为解释传粉网络结构，我们首先构建了包含传粉功能响应与AIS的动态模型，仅以物种丰富度和二元连接度作为输入参数（以下简称AIS模型）。随后，我们将其他解释变量（隔离度、面积、岛龄与采样强度）纳入模型之中。在10个岛屿中，有4个岛屿的传粉网络呈现出显著的嵌套结构，AIS模型的预测结果与观测到的网络结构具有良好相关性，可解释69%的嵌套性变异。总体而言，通过对观测到的嵌套性变异进行分层划分（hierarchical partitioning）得到的独立贡献度来看，AIS模型的预测贡献占比最高（37%），其次为采样强度（28%）与岛屿面积（22%），岛屿隔离度与岛龄的贡献则微乎其微。因此，生物交互的适应性切换似乎是保障网络功能的关键因素，而岛屿生物地理因素仅起到次要作用。尽管面积较大的岛屿可承载更多样的群落，进而促成更强的嵌套结构，但充足的采样量被证实是检测非随机网络结构的必要前提。