Our investigations cover three river catchment areas from three Central European countries (Fig. 1) representing three European ecoregions. Bükkösdi-víz (BUK) is located in the south-western part of Hungary (Pannonian ecoregion with Continental climate) with an area of 185 km² and forms a sub-catchment of Fekete-víz within the drainage area of the Drava River. The area is characterised by the hilly surroundings of Mecsek Mountains (sampling sites were located between 104–245 m a.m.s.l.). The Butižnica River (BUT) catchment is situated in the Dinaric Mountains (sampling sites were located between 235–591 m a.m.s.l.), North-Central Dalmatia, Croatia (Balkanic ecoregion, Mediterranean climate). Its drainage area is around 225 km² and belongs to the Krka River catchment. The Velička River (VEL) is a sub-catchment of the Morava River and it is located in the South Moravian part of Czechia in the White Carpathian Mountains (Continental ecoregion, humid continental climate). Velička catchment area of 172 km² was included in the sampling design of this project (sampling sites were located between 174–446 m a.m.s.l.) (Fig. 1).

Chironomidae larvae were collected using multihabitat sampling methods during two campaigns (Suppl. material 1): one in 2018–2019 at the Bükkösdi-víz catchment and another in 2021 at all catchments. Additionally, at some sites in Bükkösdi-víz catchment, floating chironomid pupal exuviae were collected from the water surface using a hand net (frame 25 × 25 cm, mesh size 250 µm) and a plastic tray in 2021.

In case of larvae and pupae, all individuals were identified from samples containing less than 300 individuals, while subsamples of ca. 300 individuals were taken from larger samples. We used multi-level identification methods that included: separating specimens based on macroscopic characteristics; conducting detailed investigations without preparation and using light microscope; and mounting specimens on microscope slides for examination at higher magnification. Exuviae were mounted on microscope slides in every instance. For morphological identification of larvae and pupae/exuviae, multiple keys were used. For further details, see Suppl. material 1.

In addition to morphological identification, a subset of larvae underwent molecular analysis using a traditional DNA barcoding protocol (Suppl. material 1). All sequences were uploaded to the Barcode of Life Data Systems (BOLD; Ratnasingham and Hebert 2007).

The first occurrence of 14 species, Macropelopia notata (Meigen, 1818), Eukiefferiella brevicalcar (Kieffer, 1911), Eukiefferiella claripennis (Lundbeck, 1898), Hydrobaenus distylus (Kieffer, 1915), Krenosmittia camptophleps (Edwards, 1929), Orthocladius calvus Pinder, 1985, Orthocladius luteipes Goetghebuer, 1938, Orthocladius saxosus (Tokunaga, 1939), Parakiefferiella triquetra (Pankratora, 1970), Rheocricotopus unidentatus Saether & Schnell, 1988, Rheosmittia spinicornis (Brundin, 1956), Microtendipes rydalensis (Edwards, 1929), Paratendipes nudisquama (Edwards, 1929) and Endochironomus donatoris Shilova, 1974 were recorded in Croatia. Additionally, the first occurrence of two species, Rheocricotopus unidentatus Saether & Schnell, 1988 and Microtendipes britteni (Edwards, 1929) were registered in Czechia.

DNA was extracted from a total of 31 specimens (14 from Hungary, 10 from Croatia and 7 from Czechia) that were uploaded to BOLD. Sequences had a median length of 658 bp, ranging from 569 to 659 bp, meaning all the sequences were high-quality barcodes (≥ 500 bp) being assigned to a Barcode Index Number (BIN; Ratnasingham and Hebert 2013). The 31 COI-sequences were assigned to 23 BINs, including eight unique BINs that are new to BOLD. Half of the specimens were identified to species level; however, in the case of 16 specimens, there were no matches passing the 1.6% threshold, preventing conclusive species-level identification.

In this section, we summarise detailed information on taxa with taxonomic issues and/or morphological peculiarities. Our aim was not to provide full taxonomic descriptions; therefore, we only highlight the relevant differential characteristics.

Clinotanypus – In most parts of Europe, C. nervosus can be found (Andersen et al. 2013), but from Russia and Romania, larvae identified as C. pinguis have also been recorded (Botnariuc and Cure 1999). The difference between the two species is that C. nervosus has a 5-tooth ligula, while C. pinguis has ligula with 6–7 teeth (Janecek 1998, Botnariuc and Cure 1999, Klink and Moller Pillot 2003, Andersen et al. 2013). However, C. nervosus is regarded as the only western Palaearctic representative of the genus, while C. pinguis is widely distributed in North America (Andersen et al. 2013) and, to our best knowledge, has never been found in the Palaearctic as adult. In our study, Clinotanypus larvae were found in Hungary and all of them had 6-toothed ligula. DNA barcoding revealed that this morphotype belongs to C. nervosus, proving that two morphotypes of this species exist and the new morphotype can be dominant in some populations. Furthermore, our results raise the possibility that the occurrence of C. pinguis in Europe might be based on misidentification.

Conchapelopia – In our study, larvae were identified to species level only when associated pupae were available. We distinguished two pupal morphotypes in our material: pupae of C. melanops are well recognisable, while there are small differences between C. pallidula and C. triannulata, species representing the other morphotype. In Hungary, the morphology of the collected specimens clearly fits the description of C. triannulata (Michiels and Spies 2002), which was also found in a previous study close to the Bükkösdi-víz catchment (Boóz and Móra 2020). However, the Croatian specimens were not unambiguously separable and were identified as a species pair.

Thienemannimyia – The species of the genus are not or hardly distinguishable as either larvae or pupae (Langton and Visser 2003, Vallenduuk and Moller Pillot 2007). In our study, larvae were morphologically identified to genus level, while one pupa from Czechia was identified as T. carnea, and one pupa from Croatia as T. lentiginosa/laeta. DNA barcoding was performed for the latter pupa and two larvae, but no match was found in BOLD systems in either case.

Zavrelimyia – According to the broadened generic conception of Zavrelimyia, the genus contains more former genera as subgenera (Silva and Ekrem 2016). In our study subgenera Zavrelimyia s. str. and Zavrelimyia (Paramerina) occurred, with larvae identifiable only to subgenera by morphological characters. Some pupae/exuviae were identified as Z. (Z.) barbatipes and as Z. (P.) cingulata. The pupal form Paramerina spec-Griechenland (Langton and Visser 2003) was found in Croatia and sequences from DNA barcoding matched (100% and 98.48% similarity) with one sequence under the name Zavrelimyia divisa in BOLD Systems. However, these sequences differed more than 9% from other sequences of Z. divisa, which questions the correctness of species identification. The clear morphological differences between Z. (P.) divisa and P. spec-Griechenland and the inconsistencies in molecular results suggest that the pupal form represents a different species.

Diamesa – There are many difficulties in separating Diamesa species as larvae (e.g. Janecek (1998), Rossaro and Lencioni (2015)) and as pupae (Langton and Visser 2003) by morphological characters. In our study, identification was possible only to the species group level in many cases (D. cinerella-Gr., D. zernyi-Gr.). Amongst D. insignipes/cinerella/tonsa, only pupae with very typical characters (see Langton and Visser (2003)) were identified as D. insignipes or as D. tonsa. In the case of larvae, D. tonsa was relatively easily recognisable, based on the colouration of the head (Rossaro and Lencioni 2015), but D. insignipes was only accepted when DNA barcoding confirmed the species identity (one specimen from Croatia and two specimens from Czechia).

Monodiamesa – Some larvae were found in Bükkösdi-víz catchment, Hungary. Only one species, M. nitida has been recorded in the country (Móra and Dévai 2004), but neither the larva nor the pupa of this species is known (Schmid 1993, Langton and Visser 2003). The other species that potentially might occur in Hungary is M. bathyphila, a widespread species in Europe. However, DNA barcoding revealed a high distance between M. bathyphila and our studied individual (only 89.74% similarity), suggesting that the collected specimens might belong to M. nitida.

Chaetocladius – The morphological identification of the larvae of this genus is only possible at species group level (Janecek 1998, Orendt and Bendt 2021), whereas species can be distinguished as pupal exuviae (Langton and Visser 2003). For this reason, in our study, only exuviae were identified to species (C. piger, C. laminatus). One pupa was identified as a pupal form (Chaetocladius pe1) because not all characters were clearly recognisable, but, most probably, it was an atypical individual of C. piger.

Corynoneura – From one site in Hungary, we identified the larval morphotype as C. cf. antennalis (Schmid 1993, Klink and Moller Pillot 2003), although this name is regarded as a synonym of C. celeripes by some authors (Evenhuis and Pape 2024), but questioned by other experts (Moller Pillot 2013). Accordingly, we used the name C. antennalis (as cf.) in our dataset. A larval morphotype (Corynoneura sp. BUT) was collected from many sites in Croatia that closely resembles C. lobata (three median teeth, strongly reduced first lateral teeth), but the central median tooth on the mentum is as large as the other median teeth (see Corynoneura species A in Cranston (1982)).

Cricotopus – Despite the available keys for Cricotopus species (e.g. Hirvenoja (1973), Langton and Visser (2003), Cuppen and Tempelman (2018)), specimens of this genus were identified to species level only occasionally due to the little morphological differences between species in many cases.

Epoicocladius – The Palaearctic E. ephemerae and Nearctic E. flavens are currently recognised as separate species (e.g. Ashe and O’Connor (2012), Andersen et al. (2013)); however, some authors suggest they are synonyms (e.g. Evenhuis and Pape (2024)). Accordingly, we used the name E. ephemerae for our records. Larvae collected in Butižnica River catchment were somewhat different from those in other populations, as the setae on body segments were remarkably less numerous and paler. DNA barcoding resulted in no match with any Epoicocladius sequences, raising the possibility that these larvae represent a different species within the genus.

Eukiefferiella – In the case of E. devonica, larvae were identified to species only as pupae and to species group as larvae. Eukiefferiella minor and E. fittkaui were separated neither as larvae nor as pupae due to very little morphological differences (Langton and Visser 2003, Orendt and Bendt 2021). A strange larval morphotype (Eukiefferiella sp. BUT) was found in the Butižnica River catchment: its morphology fits the generic description of Eukiefferiella (Andersen et al. 2013), but the thorax is covered with dense, relatively long (approx. half the width of the thoracic segment) setae (Fig. 2), which is unique in the genus. In the Velička River catchment, a peculiar pupa (Eukiefferiella sp. VEL) was collected. In the key of Langton and Visser (2003), the specimen could be identified as either E. claripennis or E. brevicalcar due to the swollen base of the thoracic horn, but the length of its apical filament separates this morphotype from both species (longer than in E. claripennis, but shorter than in E. brevicalcar). Although the pattern of the hooks on abdominal segments resembles E. claripennis, it is not clear whether it is a new pupal form or a non-typical representative of one or the other species.

Krenosmittia – The two common species of the genus (K. boreoalpina and K. camptophlebs) can be separated by the length of the 1^st antennal segment and the head (Schmid 1993), but this separation is considered as tentative (Janecek 1998). In the Croatian material, the dimensions of the head and the antenna fell in the range of K. camptophleps and pupae found here also confirmed the species identity. However, in the Czech material, only larvae occurred and their dimensions were ambiguous; therefore, we identified these specimens only to genus level.

Limnophyes – Larvae of the genus are morphologically indistinguishable (e.g. Janecek (1998), Orendt and Bendt (2021)). DNA barcoding performed on one of the many morphologically identical larvae from the Bükkösdi-víz catchment revealed that likely they belong to L. minimus.

Metriocnemus – Larvae of the genus are hardly distinguishable due to poorly-known morphological differences (Orendt and Bendt 2021) and many species with undescribed larvae (Moller Pillot 2013). DNA barcoding was performed on a larva from the Bükkösdi-víz catchment, but there was no match with valid species in BOLD systems, although the generic status was confirmed.

Orthocladius – Larvae and pupae belonging to the subgenus Orthocladius s. str. were identified to species level only occasionally due to unclear taxonomy and little morphological differences between the species (Cuppen and Tempelman 2022). Nearly all larvae and pupae belonging to the subgenus Euorthocladius were separated to species using multiple keys (Schmid 1993, Janecek 1998, Klink and Moller Pillot 2003, Cuppen and Tempelman 2022). One larva was morphologically different from the identifiable ones and either the DNA barcoding resulted in no match in BOLD Systems. Accordingly, this larva represents a new morphotype and a species that, at least as larvae, has not yet been described. In the subgenus Symposiocladius, larvae were identified as O. (S.) lignicola and as O. (S.) holsatus based on the L4 hair tuft and morphology of the mentum (Cuppen and Tempelman 2022). The occurrence of O. (S.) lignicola in Bükkösdi-víz catchment was also confirmed by pupae (Langton and Visser 2003). However, both larval types might represent other species as well, for example, the larva of O. (S.) ruffoi is not known (Cuppen and Tempelman 2022). Larvae from the Butižnica River catchment closely resembled O. (S.) holsatus, but the L4 hair tuft was much shorter than it is depicted in Cuppen and Tempelman (2022); accordingly, they only were identified to subgenus level.

Parametriocnemus – Larvae from all catchments were identified as P. stylatus, based on their morphology (e.g. Janecek (1998)) and co-occurrence of conspecific pupae (Langton and Visser 2003). However, DNA barcoding of a larva from the Bükkösdi-víz catchment resulted in no match with P. stylatus, suggesting that it is rather a species complex (see Orendt and Bendt (2021)). Some pupae collected in the Velička River catchment proved to be Parametriocnemus Pe1; DNA barcoding also resulted in no match in BOLD systems in this case.

Paraphaenocladius – From the Velička River catchment, larvae were identified as P. impensus (Klink and Moller Pillot 2003), but only pupae of P. irritus (Langton and Visser 2003) were found at the same site. The larvae of P. irritus is not known/not included in the keys available for European Paraphaenocladius species (e.g. Klink and Moller Pillot (2003), Orendt and Bendt (2021)). Accordingly, the possibility arises that the larvae of P. impensus and P. irritus are morphologically identical or hardly separable and only the latter species was collected in Czechia.

Rheocricotopus – At multiple sites in the Velička River catchment, a morphotype (Rheocricotopus sp. VEL) was found (Fig. 3), which has the general characteristics of the genus, but does not resemble any species with currently known larva or pupa (Schmid 1993, Janecek 1998, Langton and Visser 2003, Orendt and Bendt 2021).

Tvetenia – In the Butižnica River catchment, two larval morphotypes were found that differed from any species in the genus (Schmid 1993, Janecek 1998, Langton and Visser 2003, Orendt and Bendt 2021). Tvetenia sp. A is a known larval form (e.g. Cranston (1982), Schmid (1993), Klink and Moller Pillot (2003)) and was suggested to be linked with various species (Janecek 1998, Klink and Moller Pillot 2003, Orendt and Bendt 2021). Tvetenia sp. BUT clearly belongs to this genus, based on the morphology of the head and the presence of long setae on body segments, but does not resemble any other species with known larva.

Genus unknown – The morphology of some larvae from the Velička River catchment clearly matches with the taxon Orthocladiinae sp. “Berka vor dem Hainich, Thüringen” in Orendt and Bendt (2021). Our results proved that the distribution of this taxon is not limited to Germany. Orthocladiinae larvae and a pharate larvae with unique morphological characteristics were found in Bükkösdi-víz catchment (Orthocladiinae Gen. sp. III.). Neither larval nor pupal characteristics (Fig. 4) resemble any species or genera included in the keys used in our work. Some larval characters suggest that this morphotype belongs to Parakiefferiella, but the thoracic horn is completely different from that characteristic for the genus. DNA barcoding also resulted in no match in BOLD Systems.

Chironomus – Based on DNA barcoding, a larva from the Bükkösdi-víz catchment identified earlier as C. entis by morphological characters, proved to be C. plumosus. As sequences can also be found for C. entis in the BOLD Systems, we accept that this result is correct. In our larva, the upper surface of the antennal pedestal was completely light brown, which is a distinctive character for C. entis according to Vallenduuk (2017). Moreover, the dimensions of the parts of the head fell in the range of this species, but it must be noted that these characters can vary and overlap between the species. According to Orendt and Spies (2012), C. plumosus can be recognised by a distinct dark streak between the antennal pedestal and the upper eyespot. Our specimen proved that this pigmentation could spread further on the upper surface of the antennal pedestal. Accordingly, all characters used to separate the two species may overlap and should be regarded with caution. DNA barcoding performed on a larva identified as C. bernensis resulted in the closest match with a sequence under the name C. annularius. However, in this BIN, there are also specimens under the name C. bernensis MOTU7, which was analysed by Pfenninger et al. (2007), with the result that the nominal species C. bernensis might be a cryptic species complex. However, the taxonomy of C. annularius is also not clear with more biological species mentioned under this name by various authors (Spies and Sæther 2004). Nevertheless, the two larval morphotypes that represent these names can clearly be separated by the presence of lateral gills on abdominal segment VII in C. annularius, while these gills lack in C. bernensis (e.g. Orendt and Spies (2012), Vallenduuk (2017)). As the larvae we found had no lateral gills on segment VII, accordingly we listed them as C. bernensis.

Phaenopsectra – Larvae from the Bükkösdi-víz and Velička River catchments were identified as P. flavipes (Vallenduuk 2017), which was confirmed by DNA barcoding of an associated pupa. Another larval form was collected in the Butižnica River catchment, with the mentum and mandible being identical with those depicted as Phaenopsectra sp. in Andersen et al. (2013).

Polypedilum – Larvae identified as P. convictum from the Butižnica River catchment had elongate anal tubules which is unusual for this species (triangular tubules are characteristic, Vallenduuk (2017)). It is not clear if this difference is a population level variability or indicates a different species.

Micropsectra – We had difficulties with the identification of a larval Tanytarsini morphotype (cf. Micropsectra sp.) from Bükkösdi-víz catchment. The bifid premandible suggested Micropsectra, but the remarkably small body size, the lack of spur on antennal pedestal and the small number of claws on posterior parapods rather resemble Tanytarsus. The separation of the two genera is very difficult and probably the best character is the shape of the premandible (Andersen et al. 2013). Based on the bifid premandible, our larvae most probably belong to Micropsectra. In addition, co-occurring small Micropsectra pupae were also collected. These pupae resemble those of M. notescens, but are much smaller and less pigmented. The similar size of the larvae and pupae raises the possibility that they might represent the same species. The pupal morphotype Micropsectra sp. BUK1 was collected at one site in the Bükkösdi-víz catchment. This morphotype is most similar to M. auvergnensis due to the shagreen on the pleura of segments II-V, but, in the case of the latter species, pleura of segments II-VII have shagreen. Furthermore, male hypopygium was observable in one pupa and it clearly differed from that of M. auvergnensis (Reiss 1969). Further studies are in progress to clarify the taxonomic status of this species.