Phytoptidae s.str. is a lineage of eriophyoid mites associated with angiosperms. Based on representative taxon sampling and four gene markers (COI, HSP70, 18S, and 28S), we inferred the molecular phylogeny of this group and performed comparative analyses of cuticle-lined female internal genitalia. Although basal relationships were unclear, several well supported clades were recovered. These clades were supported by geography, host associations, and female genital anatomy, but contradicted the current morphology-based systematics. The monophyly of each of five conventional supraspecific groupings (Fragariocoptes, Phytoptus, Phytoptinae, Sierraphytoptinae, and Sierraphytoptini) is rejected based on a series of statistical tests. Additionally, four morphological characters (the absence of tibial solenidion φ and opisthosomal seta c1, presence of telosomal pseudotagma, and ‘morphotype’) were found to be homoplasies that cannot be used to confidently delimit supraspecific lineages of phytoptids. However, our molecular topology was highly congruent with female genital characters. Eight molecular clades were unambiguously supported by the shapes and topography of the spermathecal apparatus and genital apodemes. This suggests that the female genital anatomy could be an important factor affecting cladogenesis in Phytoptidae, a conclusion contrasting with the general expectation that host characteristics should be a major macroevolutionary force influencing the evolution of host-specific symbionts. Indeed, despite the high host-specificity, there were no apparent cophylogenetic patterns. Furthermore, we show that gall-inducing ability evolved multiple times in phytoptids. Because gall formation creates nearly instantaneous niche partitioning and the potential loss or reduction of gene flow, we hypothesize that it could be an important evolutionary factor affecting speciation within different host-associated clades of phytoptid mites.

疫霉科 是与被子植物相关的类类螨虫谱系。基于代表性的分类单元采样和四个基因标记（COI，HSP70、18S和28S），我们推断了该组的分子系统发育，并对表皮内衬女性内生殖器进行了比较分析。尽管基础关系尚不清楚，但恢复了几个支撑​​良好的进化枝。这些进化枝受到地理，宿主关联和女性生殖器解剖学的支持，但与当前基于形态学的系统学相矛盾。五个常规超特异性分组（Fragariocoptes，Phytoptus）中的每一个的单系。，Phytoptinae，Sierraphytoptinae和Sierraphytoptini）基于一系列统计检验而被拒绝。此外，还有四个形态学特征（胫骨节理素φ和阿片梭菌c1不存在，端粒假性假单胞菌的存在和'形态型'）被发现是同质异型，不能用来自信地界定植物优化的超特异谱系。但是，我们的分子拓扑与女性生殖器特征高度一致。八个分子进化枝明确地由精算器的形状和外形以及生殖器的顶点支撑。这表明女性生殖器解剖可能是影响疫霉科成虫的重要因素，这一结论与普遍预期相反，宿主期望宿主特征应是影响宿主特异性共生体进化的主要宏观进化力。确实，尽管宿主特异性很高，但没有明显的系统发生模式。此外，我们表明诱导胆汁的能力在植物黄连中多次进化。