As the initial step of the morphological analyses, we examined pinned dry specimens and ethanol-preserved specimens obtained from previous studies, from various regions including Germany (Baden-Württemberg, Bavaria, Rhineland-Palatinate, Saxony, Schleswig-Holstein), Austria, Hungary, and Southern Poland. The material was sourced from collections maintained by KD (Bayreuth, Germany), the Senckenberg Deutsches Entomologisches Institut (Müncheberg, Germany), and the Department of Hydrobiology at the University of Pécs (Pécs, Hungary).

Male specimens, regardless of their previous identifications as A. lotti or A. uliginosus [hereinafter A. uliginosus s. l. (sensu lato)], were examined for their primary external sexual organs. In cases of 29 male specimens, internal male sexual organs were dissected and analyzed in detail. Careful dissection was necessary to obtain intact phallobases, often involving the removal of tightly adhering basal parts of parameres, which were prone to damage during dissection. Age determination of freshly alcohol-preserved specimens from Hungary was conducted. The body weights of ethanol-preserved specimens from Hungary were measured one hour after deposition on filter paper using a Professional Digital Mini Scale, TL-series (1 mg–20 g; Shenzhen Union Technology, China). Additionally, a few data on ectadenia were recorded from freshly collected specimens from Bavaria.

In the next step, specimens were captured and kept alive from saline habitats of the Kardoskúti-puszta plain in southeastern Hungary (42 specimens) on May 1, 2023. The living Hungarian specimens were immediately sent to Germany by express mail using fishing bait boxes filled with heavily moistened filter papers. After arrival, the specimens were frozen, and fresh weights were measured 30 min after defrosting. Sixteen male specimens were selected for detailed morphological analyses. Soft translucent chitinous structures at phallobases were dyed darkened using a 1% aqueous or 1% ethanolic solution of pyrogallol (Seifert 1995) for microscopic investigation. Darkening was controlled after 1 h to 2 days to achieve best results.

To evaluate aedeagal morphology, measurements were taken on all available individuals, including the length of the penis from apex to distal corner of the basal part (a on Fig. 1A), the lengths of the penis base from the distal corner of the basal part to the basal part defined by the imaginary extension of the outer curvature of the penis on a line defined by a (b on Fig. 1A), the length of the sclerotized phallobase from the imaginary extension of the outer curvature of the penis to the basal sclerotized base of the phallobase in line with a + b (c on Fig. 1A), and the thickness of the non-sclerotized soft and translucent membrane of the phallobase (d on Fig. 1A). Basal pieces of both parameres and tegmina were also observed. Of the 45 male specimens of A. uliginosus s. l. examined in detail, fresh weight data were measured for 26 individuals (10 ethanol-preserved and 16 frozen-preserved, all from the Hungarian populations), while the most critical morphological parameters of the genitalia were recorded in 44 cases. All the beetles and genitalia were examined, dissected, and measured using a ZEISS binocular and a ZEISS microscope (Standard 16). Phase-contrast photographs were captured using an OLYMPUS BH2-RFCA microscope. All morphometric measurements and visual inspection results were compiled in Suppl. material 1: table S1.

Figure 1. 

Genital morphology and its changes from young and elder specimens of Agabus uliginosus sensu lato A, B morphology and aedeagus of an elder specimen in right side lateral (A) and dorsal (B) views, with measured parameters a-d. C–G Dorsal views of aedeagi from different age classes of A. uliginosus s. l. C corresponds to A. lotti, while D–G to A. uliginosus s. str. with increasing age C–E Hungarian population F, G old male specimen from Rudna, Mühlgast, Poland F, G show the same aedeagus isolated from a specimen preserved in ethanol, then dried for 5 min (F), and for 15 min or more (G). Black arrows on G symbolize movements of phallobases H–L side view of aedeagi and morphology of internal genital organs of an immature (H, K, #H04), an elder (I, #H27) and an old (J, L, #H30) specimen of A. uliginosus s. l. from the Hungarian population. H and K correspond to A. lotti, J and L to A. uliginosus s. str, while I shows a young specimen of A. uliginosus s. str. as a transition between them. Abbreviations: de: ductus ejaculatorius (checkered in H–J), ec: ectadenies (dotted in H–J), sd: spermatic duct (cross-striped in H–J), p: phallobase, t: testes (dotted in H–J), w.p: width of phallobase, w.de: width of ductus ejaculatorius. The triangles and the interrupted horizontal line in H–J illustrate the origin of the phallobase as indicated by b in A. In K (#H04) and L (#H30) anterior margin and origin of ductus ejaculatorius is marked by arrows. Scale bars: 1 mm (different scales refer to A–G and H–J). For K and L, 20-fold magnification was used.

To gain insight into the temporal distribution of records for both species, we conducted a comprehensive search for records with at least the month of collection. We utilized a variety of sources, including academic papers, specifically for A. lotti records: Turner et al. (2015), Boda et al. (2019, 2023), and Rocci et al. (2021). We accessed data from the GBIF database (for specific dataset citations see Suppl. material 1: table S2) and from the verified public database of “Izeltlabuak.hu” citizen science initiative (http://www.izeltlabuak.hu) for both species, and obtained collection records from the collection of KD (Bayreuth, Germany), the Senckenberg Deutsches Entomologisches Institut (Müncheberg, Germany), and the Collection of the Department of Hydrobiology of the University of Pécs (DHUP) for unpublished data on both species (Suppl. material 1: table S1).

Among the 45 male specimens of Agabus uliginosus s. l. subjected to detailed examination, a range of developmental stages were observed, including freshly hatched individuals with soft cuticles, very young, young, and more mature specimens (Suppl. material 1: table S1). Analysis of the genitalia revealed a correlation between age and the increasing lengths of phallobases (p), testes (t), and ectadenies (ec), which became progressively longer and thicker (Fig. 1H–J). The measurements (b+c) were found to be positively correlated with the lengths of ectadenies (Fig. 2B) and the fresh weights of the specimens (Fig. 2A), both in ethanol-preserved and frozen material, indicating that fresh weight roughly reflects the gradual enlargement of the phallobase. Additionally, the widths of phallobases in dorsal view (see Fig. 1C–G) significantly increased with age, with the widths of ejaculatory ducts (Fig. 2C). Moreover, total lengths of aedeagi also showed an increase (Fig. 1C–J). This significant swelling process (Fig. 1C–G), observed from young to elder specimens, is noteworthy because it is unusual among adephagous beetles, as diameters in dorsal views of phallobases and whole aedeagi typically represent species-specific values. Consequently, changes in diameters of aedeagi in dorsal views, as well as varying shapes of the entire aedeagi ranging from straight to curved and even asymmetrical organs, were observed (Fig. 1C–G). Furthermore, an old, freshly dissected specimen (Fig. 1F) exhibited a different shape compared to the same specimen after drying (Fig. 1G).

Figure 2. 

A Fresh weights and B length of accessory glands (= ectadenies, l.ectad. here and ec in Fig. 1H–J) as function of the size of the phallobase (b+c here and in Fig. 1A) C widths of the bases of the aedeagi in dorsal view (w.aed.) compared with diameters of ejaculatory ducts (d.ejac.) A–C measurements were made in ethanol-preserved (orange) and frozen (blue) male specimens of A. uliginosus sensu lato from Hungary D comparison of the body lengths and widths of the males showing the differences and the variation in body shapes in A. uliginosus sensu stricto (red) and A. lotti (black) types.

Significant correlations were seen in both types when comparing body length and width (Fig. 2D). However, the A. uliginosus s. str. specimens showed higher and more variable body lengths, with relatively constant widths. Conversely, in A. lotti type, both body length and width showed variability. We did not see significant differences in body ratio (length/width) between A. uliginosus s. str. (mean = 1.75, n = 11) and A. lotti types (mean = 1.77, n = 9) (Mann-Whitney U test, U = 36.5, z = 0.95, p = 0.34).

Another distinctive characteristic observed in male representatives of the A. uliginosus group is the soft and translucent external border of the phallobase (d in Fig. 1A), which is not sclerotized (Fig. 3A–C; arrows). Under higher magnification (Fig. 3C), radially arranged filaments were visible, indicating potential ongoing sclerotization processes along these structures. Elder males, as shown in Fig. 3D, showed nearly complete sclerotization of section d, which is greatly reduced. Additionally, the sclerotized phallobase displayed both radial and circular filaments (Fig. 3D). The widths of these non-sclerotized phallobase borders (d) in elder and old male specimens varied distinctly. This unique morphological detail, optimally visible after artificial darkening with reagents, may suggest an extension of the phallobase even in old males, potentially indicating multiple gonadal activities during several months or even years. Consistent with these findings, both basal parts of parameres and tegmina broadened (tegmina also lengthened) and became increasingly sclerotized with age from youngest to the oldest, as shown on Fig. 4A–D, respectively.

Figure 3. 

Translucent peripheral area of phallobase (d in Fig. 1A) of male A. uliginosus s. l. A–C #G06 specimen from Crailsheim (Germany) D nearly completely sclerotized basal area of aedeagus of an old male A. uliginosus sensu stricto #P01 specimen from Rudna, Mühlgast, Poland with both radial and circular filaments. Arrows indicate the outer margin of d in Fig. 1A, and radially arranged filaments between c and d are clearly visible, especially on C E spermatozoans from spermatic duct of elder male #H27 from Hungary (see also Figs 1I, 4C). Magnifications: A 32-fold B 100-fold C 200-fold D 100-fold E 400-fold.

Figure 4. 

Morphological characteristics (from left to right in all panels: foreclaws, aedeagi, parameres, tegmina) in Agabus uliginosus s. l. males of different age categories A #H19 B #H11 C #H27 specimens, all from Kardoskút, Hungary D #G02 from Salemer Moor, Ratzeburg, Germany. A represent A. lotti type B–D represent A. uliginosus s. str. type.

To sum up, during imaginal life, respective imaginal age classes of A. uliginosus s. l., the sclerotization of area c gradually increased, phallobases enlarged, parameres and tegmina lengthened, foreclaws became well-wore from immature specimens (see Figs 1H, 3A–C, 4A, B) to elder and highly aged adults (Figs 1I, J, 3D, 4C, D). Hereby, lengths below 0.6 mm (b + c) in immature specimens were tentatively associated with A. lotti (e.g., Fig. 1H, K). Conversely, males with b+c greater than 0.6 mm are interpreted as A. uliginosus s. str. (see Figs 1J, L, 3). Based on genital morphology and size proportions, approximately half of the examined males (20 specimens) showed characteristics typical of A. uliginosus s. str. type, 21 specimens were of A. lotti type, while three individuals showed completely transitional characteristics and values.

Altogether, we gathered more than 1600 dated records for A. uliginosus s. l. and 105 for A. lotti specifically. The majority of records for both species are concentrated in the months from March to June, with a notable peak in May. However, while data for A. uliginosus s. l. are spread across all seasons, with more than 300 records from August to February, records for A. lotti are limited to the period from March to July (Suppl. material 1: table S4). Notably, only two records for A. lotti are available in July, one from Austria and one from Slovakia, likely originating from higher elevation sites.

We successfully sequenced 16 specimens of A. uliginosus s. l. from Hungary. Comparison with all available A. uliginosus s. l. sequences in BOLD, revealed a relative homogeneity of the sequences regardless of whether our newly sequenced specimens morphologically classified to the A. uliginosus s. str. or A. lotti type. The Neighbor-Joining tree (Fig. 5) and Median-Joining network (MJ network) (Fig. 6) clearly demonstrate that the newly sequenced individuals classified as ‘A. lotti’ or A. uliginosus s. str are not distinct from each other or from the publicly available sequences in the BOLD database. While more ‘A. lotti’ sequences are located on the side branches of the MJ network, the two types remain intermixed throughout the entire network, with their haplotypes showing complete overlap (Fig. 6B).

Figure 5. 

Neighbor-Joining tree based on COI K2P pairwise distances for sequences of Agabus uliginosus s. l. and Agabus margareti (BOLD taxonomic backbone uses the incorrect name A. margaretae for this species, see introduction for more information) used as an outgroup. Names written in bold have genitalia examined for morphological discrimination between ‘lotti’ and ‘uliginosus s. str.’ types. Each sequence name consists of parts separated by a lower dash: (1) BOLD Process ID, and for the Hungarian specimens the specimen codes used in morphological analyses are also given, (2) species (types), and (3) country of origin. The results of the three species delimitation methods are indicated by vertical bars (BIN numbers given). Only bootstrap supports higher than 50% are shown.

Figure 6. 

Median-Joining network showing the relationships among the haplotypes of Agabus uliginosus s. l. in the context of A country of origin B morphotypes. Colors indicate individuals representing different countries of origin or morphological types. Each bar represents one substitution, whereas small black dots indicate undetected/extinct intermediate haplotype states. The sizes of the circles are proportional to the frequencies of haplotypes. Dashed lines encircle separate BINs.

In our sample from a single Hungarian population, we identified 12 haplotypes, with an additional seven haplotypes from public sequences in BOLD, resulting in a total of 19 known haplotypes and two intermediate haplotypes in the entire network (Fig. 6). From Germany and Finland, seven and two haplotypes, respectively, were identified. The central haplotype, comprising ten sequences, includes individuals from all three countries (Fig. 6A) and both types (Fig. 6B), i.e., the Hungarian individuals classified as both ‘A. lotti’ and A. uliginosus s. str., sequences that are 100% identical to each other and to sequences from Germany and Finland. Apart from the central haplotype, two haplotypes are represented by two copies, while the others are represented by only one.

Species delimitation methods yield consistent classifications for most sequences, with only a discrepancy observed in the case of three Hungarian sequences classified as ‘lotti’ type and one as uliginosus s. str type. The most conservative method (ASAP) assigns all sequences to a single Molecular Operational Taxonomic Unit (MOTU). For delimitation based on BINs, all but one new sequence aligned to the BIN BOLD:AAY8849, which already contained all 17 previously known A. uliginosus s. l. sequences. Within this relatively compact and homogeneous BIN, the average distance was 0.38%, and the maximum distance was 1.52%. However, the specimen #H27 was grouped separately and formed a singleton BIN (BOLD:AFI9740), which became the nearest neighbor of the A. uliginosus BIN, with a distance of 3.7%. In comparison, the only species of the species group whose sequence was available in the BOLD (A. margareti, BOLD:ACA6100 singleton BIN) showed a distance of 6.9% to its nearest neighbor, which, like that of the new #H27 BIN, was also the ‘original’ A. uliginosus BIN. Adding specimen #27, so including both A. uliginosus BINs, the average within-species distance increased to 0.72%, while the maximum within-species distance was 4.57% (Suppl. material 1: table S5.).

Among central European Agabus species, A. uliginosus is characterized by metacoxal lines fully reaching the hind margin of the metasternum and broad metasternal wings, according to Schaeflein (1971) and Nilsson and Holmen (1995). Additionally, the anterior pronotal puncture lines are continuous and not distinctly interrupted. The meshes of elytral sculpture are typically small, smaller than elytral punctures, as noted by Ganglbauer (1892). Characteristically, it exhibits a strongly convex dorsal surface and very broad pronotal lateral margins. Nilsson and Petrov (2006) separated A. uralensis from A. uliginosus, as a vicariant species in the western Palaearctic, based on its smaller body size and shorter, evenly tapering penis. As several authors before, recently Turner et al. (2015) noted high variance in several external morphological characteristics of A. uliginosus, including microsculpture and coloration. Stephens (1828) described stronger metallic specimens from England, while Queney (2002) observed a rufous form from France. Moreover, A. uliginosus, uniquely in the subfamily Agabinae, shows inter- and intrasexual dimorphism within part of its distribution area (Bilton et al. 2016).

The life cycle of A. uliginosus is characterized as an univoltine spring breeder with summer larvae and overwintering adults (Nilsson 1986). Galewski (1968, 1980) observed second-stage larvae from early May to late July in Poland, followed by third-stage larvae from late April to mid-September. He noted the presence of mostly third-stage larvae but also pupae and immature adults during May, suggesting a prolonged oviposition period. Nilsson (1986) suggested the possibility of an extended oviposition period as well. Similar bionomical data were reported by Balfour-Browne (1950). In contrast, Braasch (1989) proposed that A. uliginosus has two oviposition periods: the first from March to April and the second from July to August in Northeastern Germany. He demonstrated two peaks of adult beetles, one from January to March after adult hibernation and another from the end of May to October. Two peaks were also seen in both second- and third-stage larvae, with the first peak from April to the beginning of May and the second peak in July. Braasch (1989) also noted that A. uliginosus larvae are typically found in temporary waters, whereas adults mostly inhabit permanent water bodies and migrate between the two types of habitats at least for oviposition. Hendrich (2003) reported an interesting phenomenon: “After its habitats dry out in May and June, and shortly after hatching of the new generation, the species undertakes longer land migrations in large numbers, sometimes taking them more than 20 meters from the water. In the years 1985–1989, Ralph Platen was able to detect large numbers of A. uliginosus in various fen areas in Berlin using pitfall traps. Of the more than 1,500 diving beetles captured as part of a research project, more than 800 specimens were A. uliginosus”. Spitzenberg (2021) reported a phenological maximum for A. uliginosus during May in Saxony-Anhalt and observed avoidance of higher elevations. He also described the ecological preferences of the species for eutrophic ponds, backwaters, ditches, canals, salt lakes, and salt marshes in this area. Additionally, Spitzenberg (2021) reported flight activity at light, a behavior supported by Kehl and Dettner (2007), who associated this species with category 2a, indicating the presence or absence of flight muscles depending on individuals. Considering the above facts, that the larvae develop in seasonal waters, and migration is a significant investment for the adults, it is possible that the resulting time stress is the reason they evolved this special kind of genital development. Although it is only an assumption, it is likely to be beneficial to minimize the time of larval and pupal development, and therefore building up as little body biomass as possible in these periods, leaving some of the development to be enabled by the subsequent feeding of the adults taking place in more permanent waters.

Summarizing available records from Central Europe, we see a peak in records during March-April-May for both species, consistent with the phenological observations mentioned above. Additionally, the exclusive presence of specific records of A. lotti during these months supports the hypothesis that this species may actually represent young individuals of A. uliginosus returning to the water shortly after hatching, potentially contributing to the surge in density, and consequently in number of records. However, it is essential to consider that climatic variations and differences in altitude make objective phenological comparisons challenging. For instance, March marks the beginning of the breeding season in southern and lowland areas, whereas reproduction may commence later, typically in May, in higher elevations. Additionally, data on A. ‘lotti’ have been documented in mountainous regions of Italy, Austria, and Slovakia.

As demonstrated by Classen and Dettner (1983) and Dettner et al. (1986), age structures within dytiscid populations can exhibit significant fluctuations from month to month, during the course of one or several years, although the exact lifespan of many dytiscid species, including A. uliginosus, remains unclear. In species such as Agabus bipustulatus (Linnaeus, 1767), A. paludosus (Fabricius, 1801), or Platambus maculatus (Linnaeus, 1758), populations typically consist of varying numbers of individuals across different age classes at any given time, and these proportions are constantly changing. These age classes are delineated based on the developmental status of internal gonads, which can only be reliably assessed in freshly killed or recently ethanol-preserved specimens. Therefore, any additional records or notes accompanying regular distribution data regarding the first occurrences of larval stages or immature (soft, unsclerotized) adults or information on the status of the gonads provide crucial information for determining the life-cycle type of each species, in accordance with the findings of Nilsson (1986).

Within the uliginosus group, and likely also in the punctulatus group, Nilsson and Petrov (2006) suggested that during the imaginal stage, the degree of sclerotization and enlargement of the phallobase of the aedeagus might increase. After studying larger material of A. uliginosus and A. uralensis, they concluded that identification might be problematic if the phallobase did not enlarge since the development of the basal apodeme occurs later than that of the rest of the penis. This is unusual because sclerotized aedeagi of male insects typically represent morphologically stable, constant, and highly useful diagnostic structures, allowing species differentiation in most insect groups. The lock-and-key mechanism (Eberhard 1985) posits that the morphologically species-specific external genitalia of elder specimens correspond tightly between both sexes. However, the aedeagal morphology of freshly hatched male beetles has not been systematically analyzed before and has not been compared with appropriate structures of elder specimens, including the A. uliginosus group.

As reported by Classen and Dettner (1983) in A. bipustulatus and A. paludosus, it was expected, and now demonstrated for the first time in A. uliginosus s. l. (Fig. 1H–J), that the lengths and diameters of ectadenies gradually increase from one age class to another. Additionally, and in line with Dettner et al. (1986), the fresh weights also positively correlated with the lengths of “b+c” (Fig. 2A, see also Fig. 1A). At the same time, we clearly showed that aedeagal morphology gradually changes with age: the phallobase enlarges, its basal part becomes increasingly sclerotized, the parameres and tegmina lengthen from immature, young specimens (see Figs 1H, 3A–C, 4A, B) to elder and highly aged adults (Figs 1I–J, 3D, 4C, D). During this process, younger adults resemble A. lotti more, while elder and fully sclerotized adults resemble A. uliginosus s. str. more. Within Dytiscidae, beside the Agabus uliginosus group, Nilsson and Petrov (2006) mentioned a comparable case in Hygrotus tumidiventris (Fall, 1919), with a normal aedeagus figured by Larson (1975).

While they are not explicitly sexual organs, the foreclaws of males also play a role in the mating process, thus becoming worn out as specimens age, with their apex becoming blunted, deformed, and eventually breaking off. We hypothesize that the fixing and adhering of male foreclaws to the pronoti of females during one or more extended copulations, particularly with increasing age, may damage both the thin and sensitive arched apodemes situated at the base as well as the tips of the foreclaws.

Our results also underscore the urgent need for comparisons of aedeagal morphologies of young teneral males with those of adult specimens, not only in Dytiscidae but also in other taxa. This is particularly crucial for older descriptions where figures of aedeagi were provided in dorsal views. When large amounts of viscous ectadenial materials, especially proteins and mucosubstances, as seen in Carabidae (see Krüger et al. 2014; Schubert et al. 2017), pass through aedeagi, these structures, if not completely sclerotized, may swell enormously. Depending on the amounts and consistency of ectadenial secretions, a certain flexibility of aedeagal structures is essential.

DNA barcoding is already a well-established and powerful tool for molecular identification of the species (Grant et al. 2021; Antil et al. 2023). Beetles are among the flagship groups of invertebrates for which barcode reference libraries are built (Pentinsaari et al. 2014; Hendrich et al. 2015). Still, even in Europe, progress reached above 50% of known species for freshwater aquatic beetles (Csabai et al. 2023). Species delimitation methods revealed two (ASAP), three (BIN), and four (mPTP) MOTUs; these also considered morphological features, and the accepted approach can be considered as a rough proxy of species. We should consider the method limitations and interpretation of the obtained result. The conservative approach of the ASAP method seems congruent with our morphological investigation (Copilaş-Ciocianu et al. 2022), and the mPTP approach tended to over-split (Song et al. 2018; Goulpeau et al. 2022). While the phenomenon of bin-sharing may arise here (e.g., Raupach et al. 2018, 2020; Schattanek-Wiesmair et al. 2024), we can rule out this possibility since in some cases, the sequences are 100% identical for individuals classified morphologically as different types. This holds even when comparing Hungarian A. lotti individuals with German and Finnish A. uliginosus s. str. As a theoretical possibility, it could be argued that all previous COI sequences in BOLD may have originated from A. lotti. However, this can be clearly ruled out based on the attached photos of processed individuals in BOLD, especially with reference to the specimen with Process-ID COLFE1294-13, where the genitals (aedeagus and parameres) are visible, leaving no doubt that this old specimen belongs to A. uliginosus s. str.

The addition of the Hungarian sequences significantly increased the known haplotype diversity within the species. Large number of individuals are available from Germany, Finland, and Hungary in BOLD, and the Hungarian dataset stands out as the most heterogeneous. Only available molecular data from A. lotti are deposited in Venables’ thesis (2016). One specimen from Italy is closely related to A. uliginosus from Germany. However, it is noteworthy that Venables utilized COI-3P sequences, and unfortunately, the sequence corresponding to A. lotti is not publicly accessible in repositories. Despite this limitation, the observed genetic divergence appears relatively modest. Given the geographical context (Italy), it is plausible to consider the possibility of intraspecific variation induced by spatial isolation, although definitive conclusions warrant further investigation.

The comparison of sequences confirmed with high confidence that individuals classified as A. lotti and A. uliginosus belong to the same species based on COI markers. Although it nicely supports our single-species hypothesis in the present case, the COI analysis alone is insufficient for a comprehensive revision of the species complex. It would be highly beneficial to conduct a comprehensive multi-marker study encompassing most of the distribution range of the species group, covering various regions of the Holarctic region, including all known species (currently six without A. lotti), and employing integrated molecular and morphological approaches, including analysis of immature, young, and teneral individuals. Limited sampling size may also affect our findings.

As shown, specimen #H27 (morphologically identified as A. uliginosus s. str.) displayed molecular deviations from all other specimens, irrespective of their morphological identification as A. uliginosus s. str. or A. lotti. Therefore, detailed comparisons were made of the aedeagi, parameres, tegmina, male foreclaws, and internal gonads of the molecularly analyzed specimens. Young specimens were characterized by non-sclerotized, slightly colored, translucent genital sclerites, small and lengthened tegmina, as well as thin and small bases of parameres (Fig. 4A right; #H19). In contrast, elder specimens exhibited thickened and more sclerotized tegmina, with thickened and sclerotized bases of parameres (Fig. 4B right; #H11). Specimen #H27 (Fig. 4C right) was notably older compared to #H19 (Fig. 4A right), displaying typical features of A. uliginosus s. str., including an increased phallobase and parameres. However, #H27 had a lengthened tegmen that was only partially sclerotized in the center and not at the base. Exhibiting non-sclerotized internal organs, #H27 (c+d = 0.42 mm; d.ej.: 0.18 mm; ect.: 3.5 mm) appeared slightly younger than #H11 (c+d = 0.34 mm; d.ej.: 0.25 mm; ect.: 5.1 mm, see also Suppl. material 1: table S1). Nevertheless, based on its sclerotized and broader phallobase, #H27 was conclusively older than #H11. Age determination was further supported by the exclusive presence of sperms in the spermatic ducts (sd) of #H27 (Fig. 3E), contrasting with their absence in #H11 and #H19. Although specimens from the Hungarian population were not very old, external genitalia were compared to distinctly elder A. uliginosus specimens from other regions (despite the unavailability of internal genitalia). Two specimens (#G01, #G02) from Salemer Moor/Ratzeburg, Germany, both with c+d values of 0.4 mm, large phallobases, and thick and sclerotized basal pieces of parameres, showed characteristic lengthened tegmina similar to those seen in #H27. This comparison further confirmed that Hungarian specimen #H27 represented an elder specimen of A. uliginosus compared to #H11. Furthermore, the larger asymmetrical foreclaw was compared among these three Hungarian and one German male specimens (Fig. 4A–D left). The youngest specimen, #H19, displayed a highly arched basal apodeme with a very sharp tip (Fig. 4A left). In the somewhat elder specimen #H11, the basal apodeme of the sharp claw (only one leg was available) appeared slightly slender and frayed out (Fig. 4B left). Specimen #H27 was characterized by a worn-out basal apodeme and rounded or broken tips of the claws (Fig. 4C left), indicating that #H27 is the eldest specimen among the three Hungarian specimens depicted. The old specimen from Germany exhibited a male foreclaw with a rounded tip and a slightly frayed basal apodeme (Fig. 4D left). The case of specimen #27 unequivocally indicates that while current knowledge and examination of northern populations have revealed a relative genetic homogeneity, parts of the distribution area not yet genetically characterized, particularly the southern and eastern areas, may harbor unexpected discoveries, potentially exhibiting a broader diversity of haplotypes and even revealing previously unknown hidden species. As shown by Bergsten et al. (2012), in a comprehensive study 250 individuals were needed to demonstrate 95% of the molecular variation of the widespread species Agabus bipustulatus, and approximately 70 for other species with more restricted distributions. More extensive sampling in the Pannonian Basin may fill the gaps regarding A. uliginosus group, as this region is proven to be a biodiversity hotspot for several aquatic invertebrate groups (Csapó et al. 2020; Sworobowicz et al. 2020). With limited sampling, we cannot exclude or confirm the phylogeographic effect of molecular separation of specimen #27. Multiple BINs within species are not unique (Raupach et al. 2020; Geiger et al. 2021).

All our findings indicate that beetles identified as Agabus lotti represent young specimens of A. uliginosus. Therefore, Agabus lotti Turner, Toledo & Mazzoldi, 2015 must be regarded as a junior synonym (syn. nov.) of Agabus uliginosus (Linnaeus, 1761). We observed and demonstrated a gradual morphological transition of aedeagal features from young specimens (A. lotti) to elder individuals (A. uliginosus), corresponding with age-dependent variations of internal sexual organs. Obviously, Turner et al. (2015) was unable to analyze both sclerotized and soft structures across a series of male specimens of different ages. However, they noted in their description of A. lotti that the endophallus varies morphologically depending on the maturity of the beetles and referenced remarks by Nilsson and Petrov (2006), who observed enlarged aedeagi in mature A. uralensis specimens and underdeveloped ones in young teneral specimens. Additionally, the fact that nearly all localities of A. lotti are found within the geographic range of A. uliginosus (with exceptions possibly in Central Italy) further challenges the species priority of A. lotti. Our molecular analyses further reinforced our hypothesis that A. lotti and A. uliginosus are the same species, as we observed only minimal differences among most specimens, with all but one individual classified into a BIN demonstrating high homogeneity. However, the genetically distinct nature of a single individual also underscores the need for further investigations in the species group. Particularly, with regards to the possible existing but never analyzed southern populations, more significant differences may exist, or the possibility of discovering new, previously unknown species cannot be ruled out.

The authors have declared that no competing interests exist.

No ethical statement was reported.

ZK was supported by ERASMUS 22/1/KA131/HED000057355/SMT-094 and CEEPUS F-2324-179374 grants.

KD and ZC conceived the ideas and designed methodology; ZK and ZC conducted the fieldwork; KD, ZK, and TR conducted various morphological and molecular laboratory work to collect the data; ZC, KD and TR analyzed the data; KD and ZC led the writing of the manuscript. All authors contributed critically to the drafts and gave final approval for publication. Our study brings together authors from a number of different countries, including scientists based in the countries where the study was conducted. All authors were engaged with the research and study design to ensure that the diverse sets of perspectives they represent were considered. Where relevant, literature published by scientists from the region was cited.

Konrad Dettner https://orcid.org/0000-0002-4054-1878

Zsolt Kovács https://orcid.org/0009-0008-0101-3280

Tomasz Rewicz https://orcid.org/0000-0002-2085-4973

Zoltán Csabai https://orcid.org/0000-0003-1700-2574

All of the data that support the findings of this study are available in the main text or Supplementary Information.