Abstract: Plants play important roles as habitat and food for a tremendous diversity of specialist animals and fungi. The disappearance of any plant species can lead to extinction cascades of its associated biota. In consequence, documenting the diversity and specificity of plant-associated organisms is of high practical relevance in biodiversity conservation. Here, we present the first large-scale molecular investigation into the diversity, host specificity, and cophylogenetic congruence of an especially rich plant–fungal association, the rust fungi (Pucciniaceae) of Cyperaceae and Juncaceae. Using the largest rust fungi DNA barcoding dataset published to date (252 sequences, 82 taxa), we reject the presence of a global ITS2-28S barcode gap, but find a local gap in Cyperaceae–Juncaceae rusts, and suggest the existence of many cryptic species in North America, with some broadly circumscribed species possibly corresponding to >10 cryptic species. We test previous hypotheses of correlations between the phylogenies of rust fungi and their Cyperaceae–Juncaceae hosts using a combination of global-fit and event-based cophylogenetic methods. A significant cophylogenetic signal is detected between rusts and their hosts, but the small number of cospeciations argues for preferential host jumps as the driving process behind these correlations. In addition, temporal congruence between the origin of major Carex clades and their rusts suggests that host diversification may have promoted parasite diversification. Finally, we discuss the relevance of rust infection patterns to the systematics of Cyperaceae, highlight some taxonomic problems uncovered by the analyses, and call attention to the promise of DNA barcoding for bridging knowledge gaps in poorly studied plant-associated microorganisms.

Key words: coevolution, host–parasite, molecular phylogenetics, phytopathology, Pucciniomycotina, rushes, sedges, Pucciniales