Data from: Inverse dispersal patterns in a group of ant parasitoids (Hymenoptera: Eucharitidae: Oraseminae) and their ant hosts

When postulating evolutionary hypotheses for diverse groups of taxa using molecular data, there is a tradeoff between sampling large numbers of taxa with a few Sanger sequenced genes or sampling fewer taxa with hundreds to thousands of next-generation sequenced genes. High taxon sampling enables the testing of evolutionary hypotheses that are sensitive to sampling bias (i.e. dating, biogeography, and diversification analyses), whereas high character sampling improves resolution of critical nodes. In a group of ant parasitoids (Hymenoptera: Eucharitidae: Oraseminae), we analyze both of these types of datasets independently (203 taxa with 5 Sanger loci; 92 taxa with 348 Anchored Hybrid Enrichment loci) and in combination (229 taxa, 353 loci) to explore divergence dating, biogeography, host relationships, and differential rates of diversification. Oraseminae specialize as parasitoids of the immature stages of ants in the subfamily Myrmicinae (Hymenoptera: Formicidae), with ants in the genus Pheidole being their most common and presumed ancestral host. A general assumption is that the distribution of the parasite must be limited by any range contraction or expansion of its host. Recent studies support a single New World to Old World dispersal pattern for Pheidole approximately 11–22 Ma. Using multiple phylogenetic inference methods (parsimony, maximum likelihood, dated Bayesian, and coalescent analyses), we provide a robust phylogeny showing that Oraseminae dispersed in the opposite direction, from Old World to New World, approximately 24–33 Ma, which implies that they existed in the Old World prior to their presumed ancestral hosts. Their dispersal into the New World appears to have promoted an increased diversification rate. Both the host and parasitoid show single unidirectional dispersals in accordance with the presence of the Beringian Land Bridge during the Oligocene, a time when the changing northern climate likely limited the dispersal ability of such tropically adapted groups.

在利用分子数据为多样类群（taxa）构建演化假说时，存在两类采样策略的权衡：要么对大量类群开展少量Sanger测序（Sanger sequencing）基因的采样，要么对少量类群开展数百至数千个下一代测序基因的采样。高类群采样量可用于检验易受采样偏差影响的演化假说（如分化时间估算、生物地理学与多样化速率分析），而高特征采样量则能提升关键节点的系统发育分辨率。本研究以蚁类寄生蜂类群——膜翅目（Hymenoptera）优特小蜂科（Eucharitidae）刻角蚜小蜂亚科（Oraseminae）为研究对象，分别独立分析两类数据集：一类包含203个类群、5个Sanger测序位点，另一类包含92个类群、348个锚定杂交富集（Anchored Hybrid Enrichment）位点；同时开展联合数据分析（包含229个类群、353个位点），以探究分化时间、生物地理学、寄主关系以及差异化多样化速率。刻角蚜小蜂亚科（Oraseminae）专性寄生切叶蚁亚科（Myrmicinae，膜翅目：蚁科（Formicidae））昆虫的幼态阶段，其中大头蚁属（Pheidole）昆虫是其最常见的寄主，同时被推测为该亚科的祖先寄主。学界普遍认为，寄生生物的分布范围会受其寄主的分布范围收缩或扩张所限制。近期研究表明，大头蚁属（Pheidole）的扩散模式为从新世界到旧世界，发生于约1100万至2200万年前（Ma）。本研究采用多种系统发育推断方法（简约法、最大似然法、时间校准贝叶斯法以及溯祖分析），构建了稳健的系统发育树，结果显示刻角蚜小蜂亚科（Oraseminae）的扩散方向恰好相反：从旧世界到新世界，发生于约2400万至3300万年前（Ma），这意味着该类群在其推定的祖先寄主出现之前，就已经存在于旧世界。该亚科向新世界的扩散似乎推动了其多样化速率的提升。寄主与寄生蜂类群均呈现单一的单向扩散模式，这与渐新世（Oligocene）白令陆桥的存在相契合——彼时北半球气候的变化可能限制了这类热带适应性类群的扩散能力。