Fungus gnats (Diptera: Mycetophilidae) have gained considerable attention over the last decades and represent one of the most intensively studied groups of Diptera (Whitmore et al. 2021). Nevertheless, Meier et al. (2025) characterised the family as a “dark taxon”, with a highly uncertain global species richness. This is underscored by a recent integrative revision of Mycetophilidae from Singapore, which revealed 115 species new to science from a single, small geographic area (~730 km²), corresponding to an undescribed-to-described ratio of approximately 23:1 (Amorim et al. 2025). Indeed, the taxonomic deficit is not restricted to the tropics. Even in Europe, despite more than two centuries of taxonomic study, a good number of species remains undescribed (e.g. Kjærandsen and Søli 2020), emphasising that Mycetophilidae exemplify a globally under-characterised lineage across both well- and poorly studied regions.

It is noteworthy that several valid European Mycetophilidae species, described from material collected in the nineteenth century, have not been recorded since, including Boletina fuscula Holmgren, 1883, Leia maculosa Strobl, 1898, and Phthinia gracilis Winnertz, 1864 (Chandler 2005 and subsequent authors). Here, we present a new record of another very rare fungus gnat, Boletina kowarzi Stackelberg, 1943 (Fig. 1), originally described from German material collected in 1877 and not recorded from Central Europe since. We provide an updated diagnosis of the species, with emphasis on the male terminalia, supplemented by detailed photographs of the terminalia and habitus, as well as DNA barcode data and the diagnostic molecular combination.

The examined material was collected using Townes-type Malaise traps and preserved in 80% ethanol denatured with methyl ethyl ketone (MEK). The traps were installed in the Totengraben natural forest reserve within the Mangfall Forest protected area in the Bavarian Alps. The reserve covers 46.7 ha and includes a ca. 5 ha remnant of primeval forest embedded within a near-natural mixed mountain forest (Schnell 2004). It is situated on a steep northwestern slope at 970–1390 m a.s.l. and is characterised by a cool montane climate with a mean annual temperature of 4.5 °C and annual precipitation of 1800–2000 mm. Summers are cool, winters snowy, and the growing season lasts approximately 185 days. The vegetation corresponds to an Aposerido-Fagetum mixed mountain forest on Main Dolomite, dominated by fir, spruce, beech, and sycamore maple (Schnell 2004). The primeval forest stands contain trees approximately 200–450 years old. One trap (47.59395°N, 11.69079°E) was placed near an old, broken beech tree (Fagus sylvatica L.) heavily infested with fungi, whereas the other (47.59420°N, 11.69417°E) was placed about 260 meters from the first trap in a stand of old beech, fir, and spruce trees.

The collected insect material was sorted into selected groups, including Mycetophilidae, which were subjected to DNA barcoding and subsequently morphological examination following the reversed work flow of Meier et al. (2025). For species-level identification, characters of the male terminalia provide the most reliable diagnostic features. Accordingly, male terminalia were dissected, macerated in 10% KOH, neutralised with acetic acid, rinsed in distilled water, and examined in glycerine; they are preserved in glycerine in a microvial attached to the specimen. Photographs of the habitus and terminalia were taken using a Leica K5C camera mounted on a Leica 205C stereomicroscope, with image stacking performed in LAS X or Helicon Focus. Image plates were prepared in Adobe Photoshop CS5 (cf. Kjærandsen et al. 2022). Morphological terminology follows Søli (1997, 2017). The examined material is deposited in the Bavarian State Collection of Zoology (SNSB), Munich, Germany.

DNA extraction and barcode amplification followed the protocols described in Srivathsan et al. (2021). To amplify the 658 bp COI barcode, the forward primer LCO1490 (Folmer et al. 1994) and the reverse primer modified jgHCO2198 (Geller et al. 2013) were used, and 9 bp tags (Srivathsan et al. 2024) on the 5’ end of both primers were used for multiplexing the PCR products for sequencing. DNA sequencing was conducted on the Oxford Nanopore MinION platform following the protocol described in Srivathsan et al. (2024), using a MinION R10.4.1 flow cell in the GridION sequencing device and ‘super-accuracy’ basecalling mode. Demultiplexing of sequences was conducted with ONTbarcoder version 2.0 (Srivathsan et al. 2024) with the following settings: “Conventional barcoding”, “1” bp tag mismatch, and all other settings on default.

For generating a molecular diagnosis as recommended by Rheindt et al. (2023) and testing candidate diagnostic molecular combination (DMC), 9,097 COI sequences over 300 bp long were downloaded from BOLD (Ratnasingham et al. 2024; accessed 20 March 2026) for the genus Boletina and aligned with the COI sequences of the four specimens examined in this study using MAFFT (Katoh and Standley 2013). The sequences were divided into two datasets, with the first set containing 4,517 sequences used to generate the DMC and the second set containing 4,584 sequences used to test the DMC. The DMC was generated with the Find DMCs function in UITOTO (Torres et al. 2026) with the default parameters (iterations 5,000–20,000, minimum DMC length 8, minimum number of exclusive character states 4, refinement strength 0–0.25). Testing against aligned sequences was done with the default parameter maximum 1 mismatch allowed, and DMC visualization was done with the Visualizer function in UITOTO. DMC of the species is provided together with morphological diagnosis below.

To assess genetic differentiation among morphologically allied Boletina species, cytochrome c oxidase subunit I (COI) sequences were retrieved from the Barcode of Life Data Systems (BOLD; Ratnasingham et al. 2024). Sequences were aligned and pairwise genetic distances were calculated in MEGA X (Kumar et al. 2018) using both uncorrected p-distance and the Kimura 2-parameter (K2P) substitution model.

Although B. kowarzi was described by Aleksander Stackelberg (1897–1975) in 1943, the description was based on a single specimen collected much earlier, in 1877, by Ferdinand Kowarz (1838–1914) at Bad Liebenstein, Thuringia, Germany. Ferdinand Kowarz, employed by the Imperial Austrian Telegraph Department, was among the foremost European dipterologists of the second half of the nineteenth century (Osten Sacken 1903). Over the course of his career, Kowarz assembled several Diptera collections, which he sold at various times out of economic necessity (Pont 1998). These collections are now dispersed across European museums. Part of his Mycetophilidae material is housed in the Swedish Museum of Natural History, Stockholm (Kurina 2003), while other specimens are distributed among various European institutions (Kurina pers. obs.). The holotype of B. kowarzi is preserved in the collection of the Zoological Institute, St. Petersburg, Russia. Stackelberg (1943) provided illustrations of the male terminalia, which enable the unambiguous recognition of conspecific specimens. However, the general habitus of the species (Fig. 1) and the aedeagal complex (Fig. 2D–G) are illustrated here for the first time. Boletina kowarzi appears to be a genuinely rare species with a scattered distribution. Remarkably, despite intensive collecting activity in Central Europe, no additional records from the region have been reported until now. The species was recorded from two old-growth broadleaved forest localities in SE Norway in the late 1990s (Gammelmo and Søli 2006; GBIF 2026) and is red-listed as vulnerable (VU) in Norway (Gammelmo et al. 2021). It has not been rediscovered during the intensified study of Nordic fungus gnats in recent decades (Kjærandsen and Søli 2020) and has hitherto been absent from DNA barcode datasets including those generated by metabarcoding within the National Insect Monitoring in Norway (Åström and Davey 2025). The material examined here was collected in the Totengraben natural forest reserve in the Bavarian Alps, an area characterised by large quantities of dead wood (approximately 150 m³/ha) and a rich diversity of wood-inhabiting fungi (Schnell 2004). Such old-growth forests now persist only as remnants in Central Europe. Given the scarcity of fungus gnat records from Alpine forests, it remains unclear whether B. kowarzi is specifically associated with such old-growth habitats.

Boletina kowarzi occupies a distinct position among Palaearctic Boletina species. While the morphology of the male terminalia indicates the closest similarity of B. kowarzi to B. dubia and B. plana, both uncorrected p-distances and Kimura 2-parameter (K2P) distances reveal strong genetic differentiation of B. kowarzi from all other examined Boletina species. Interspecific divergence values range from approximately 10–14% based on p-distances and 11–15.5% using the K2P model, with B. dubia representing the nearest genetic neighbour under both metrics. These values substantially exceed the commonly applied minimal interspecific COI threshold for insects (≈ 2%; Zhang and Bu 2022) and the mean distance to the nearest neighbour (6.97%) in the Nordic region (Kjærandsen 2022). Although the recorded divergence levels are high, they are not unprecedented within Mycetophilidae. Jürgenstein et al. (2015) reported interspecific COI divergences of approximately 2.9–10.6% among closely related species in the Mycetophila ruficollis species group. Moreover, Kurina and Õunap (2023) documented unexpectedly high interspecific COI divergences (10.2–18.1%) among sympatric Neotropical Zygomyia Winnertz, 1864 species. These examples demonstrate that the high divergence values observed for B. kowarzi and its congeners fall within a plausible range for deeply diverged evolutionary lineages in Mycetophilidae. The concordance between distance measures and the magnitude of divergence supports the recognition of Boletina kowarzi as a well-defined and evolutionarily distinct lineage within Boletina.