Data from: Experimental evidence for fundamental, and not realised, niche partitioning in a plant-herbivore community interaction network

Patterns of niche partitioning can result from local ecological interactions (e.g. interspecific competition) occurring within a contemporary time frame (realised niche partitioning). Alternatively they may represent the end-product of historical processes acting over long time frames (fundamental niche partitioning). Niche partitioning is often detected by analysing patterns of resource use within communities, but experiments are rarely conducted to test whether patterns of non-overlapping resource use reflect realised or fundamental niche partitioning. We studied a community of restio leafhoppers from the genus Cephalelus, and their host plants, the Restionaceae (restios). We used network and experimental approaches to determine whether network modularity (a measure of niche partitioning within local communities) reflects fundamental or realised niche partitioning. Using a weighted modularity index for two party networks (e.g. insect - plant) we determined whether the network of this community is modular (i.e. consists of groups of species interacting strongly, with weak interactions between groups). We also aimed to identify specific Cephalelus - restio modules (groups). Using knowledge of module membership to design experiments, we tested whether Cephalelus species from two different modules, C. uncinatus and C. pickeri, prefer and perform better on restios from their own modules versus restios from other modules. These experiments were performed under controlled conditions, eliminating the influences of competition and predation on host choices. The Cephalelus – restio community was modular, implying niche partitioning. Cephalelus also preferred and performed better on restios from their own modules in the absence of local contemporary factors. Most niche partitioning in the investigated Cephalelus community, is not caused by local interactions, and thus host use patterns represent fundamental niches. Our findings highlight the importance of understanding local community structure in the light of processes extrinsic to the local community context.

生态位分化（niche partitioning）格局的形成，可源于当代时间尺度下发生的本地生态相互作用（如种间竞争），此类分化被称为现实型生态位分化（realised niche partitioning）。此外，该格局也可能是长期历史过程作用的最终产物，即基础型生态位分化（fundamental niche partitioning）。生态位分化通常通过分析群落内的资源利用格局予以检测，但鲜有实验能够验证非重叠资源利用格局究竟对应现实型还是基础型生态位分化。我们针对隶属于Cephalelus属的帚灯草叶蝉及其宿主植物帚灯草科（Restionaceae，简称restios）群落展开研究。我们采用网络分析与实验结合的方法，探究网络模块化（network modularity，一种衡量本地群落内生态位分化的指标）是否能够反映现实型或基础型生态位分化。我们借助二部网络（two party networks，如昆虫-植物网络）的加权模块化指数（weighted modularity index），判断该群落的相互作用网络是否具备模块化结构——即由交互作用紧密的物种类群组成，类群间交互作用微弱。此外，我们还旨在识别特定的Cephalelus-帚灯草模块化类群。基于模块化类群的成员信息设计实验后，我们检验了来自两个不同模块化类群的Cephalelus物种——C. uncinatus与C. pickeri——相较于其他模块化类群的宿主植物，是否更偏好且更适配自身所属类群的帚灯草宿主。本实验在可控条件下开展，排除了竞争与捕食对宿主选择的影响。Cephalelus-帚灯草群落具备模块化结构，提示该群落存在生态位分化；且在排除本地当代相互作用的影响后，Cephalelus物种同样更偏好且更适配自身所属模块化类群的宿主植物。本次研究的Cephalelus群落中，多数生态位分化并非由本地相互作用所导致，因此宿主利用格局反映的是基础生态位。本研究结果凸显了结合本地群落外部过程来理解本地群落结构的重要性。