The genus Coreomyces (Laboulbeniaceae, Laboulbeniomycetes, Ascomycota) includes minute parasites on water boatmen (Corixidae, Hemiptera, Insecta). This taxonomic study is primarily based on freshly sampled corixids infected by Coreomyces from Sweden, although a few samples from Denmark and Turkey were also included. All records were verified using DNA sequence data from the internal transcribed spacer region and large subunit of the nuclear ribosomal DNA repeat region. We recognise four species, two of which are new to science: Coreomyces confusus H. Sundb. et al. sp. nov., C. dextrorsus H. Sundb. et al. sp. nov., C. macropus Thaxt., and C. corixae Thaxt. C. corixae is a new record for Denmark, Sweden, and Turkey, while C. macropus is a new record for Denmark and Sweden. All four species can inhabit two different yet distinct positions on the host. We observe that morphology is affected by the position on the host and that different species sharing the same position on the host tend to be difficult or impossible to separate on morphology only. We conclude that species circumscriptions in Coreomyces must be based on the integration of molecular and morphological data.

Corixids were collected from 15 localities in southern Sweden and one in Turkey. They were scanned for Coreomyces thalli under a dissecting microscope. In total, 88 mature thalli were sampled from 48 corixids of the genera Arctocorisa, Callicorixa, Cymatia, and Sigara. DNA sequences were generated following the methods of Sundberg et al. in Methods Ecol. Evol. 9: 148–158 (2018). We generated new sequences primarily from the internal transcribed spacer (ITS) region or, in some cases, from the large subunit of the nuclear ribosomal DNA (nrLSU) in case the ITS sequencing failed. New sequences were added to the alignments of Sundberg et al. in Fung. Syst. Evol. 1: 217–228 (2018) using the “--add” option of MAFFT version 7.475 with L-INS-i iterative refinement. Maximum likelihood trees were estimated using IQ-TREE version 2.1.2. A standard model selection (among models with 1, 2 or 6 rates) was performed and followed by a search for the best tree using default settings except that all nearest neighbour interchange (NNI) branch swaps were explored.

The data consist of two text files in NEXUS format, one containing the ITS data (Coreomyces_ITS.nex) and the other the nrLSU data (Coreomyces_LSU.nex). Each file contains the alignment in a characters block as well as the inferred maximum likelihood tree in a trees block.