The analysis of interaction networks across spatial environmental gradients is a powerful approach to investigate the responses of communities to global change. Using a combination of DNA metabarcoding and traditional molecular methods we built bipartite Drosophila-parasitoid food webs from six Australian rainforest sites across gradients spanning 850 m in elevation and 5° Celsius in mean temperature. Our cost-effective hierarchical approach to network reconstruction separated the determination of host frequencies from the detection and quantification of interactions. The food webs comprised 5-9 host and 5-11 parasitoid species at each site, and showed a lower incidence of parasitism at high elevation. Despite considerable turnover in the relative abundance of host Drosophila species, and contrary to some previous results, we did not detect significant changes to fundamental metrics of network structure including nestedness and specialisation with elevation. Advances in community ecology depend on data from a combination of methodological approaches. It is therefore especially valuable to develop model study systems for sets of closely-interacting species that are diverse enough to be representative, yet still amenable to field and laboratory experiments.

中文翻译：

---

分子分析揭示了雨林果蝇 - 寄生蜂群落与海拔高度一致的食物网结构

跨空间环境梯度的交互网络分析是研究社区对全球变化的反应的有力方法。使用 DNA 元条形码和传统分子方法的组合，我们在跨越海拔 850 m 和平均温度 5°C 的梯度的六个澳大利亚热带雨林地点构建了果蝇-寄生蜂食物网。我们具有成本效益的网络重建分层方法将主机频率的确定与交互的检测和量化分开。食物网在每个地点由 5-9 个寄主和 5-11 个寄生物种组成，并且在高海拔地区的寄生发生率较低。尽管宿主果蝇物种的相对丰度发生了相当大的变化，并且与之前的一些结果相反，我们没有检测到网络结构的基本指标发生重大变化，包括嵌套性和随高度的专业化。社区生态学的进步取决于来自方法论方法组合的数据。因此，为一组密切相互作用的物种开发模型研究系统特别有价值，这些物种的多样性足以具有代表性，但仍然适合现场和实验室实验。