A total of 14 literature sources containing substantial information about ants in Kamchatka, were reviewed. Species records were classified by reliability according to two categories: “reliable” (supported by multiple, including recent sources) and “questionable” (supported by single and/or dated sources, but contradicted by more recent work). Ants were collected manually in >95% ethanol in 8 localities in Kamchatka (Fig. 1A). Geographic coordinates and elevations were recorded for each sample using a Garmin etrex 10 GPS device. Habitats were photographed during field observations. Specimens were pinned for morphological study or kept in >95% ethanol at -20 °C. All specimens are stored in the private collection of the author (Dietikon, Switzerland). A key to the worker caste was developed based on comparative morphology of material in the author’s collection and inspired by literature (Radchenko and Elmes 2010; Seifert 2018). Pictures of specimens were taken using a Samsung Galaxy Xcover 5 cell phone in combination with a Leica EZ4 binocular at 16–70 × magnification. Focus stacking and image processing was performed using the programs align_image_stack v. 2019.0.0, Enfuse 4.2 and GIMP v. 2.10.30.

DNA-barcoding was performed performed de novo for 47 ant specimens by the company Sinsoma GmbH (Völs, Austria). For COI, the primer pair LCO1419 und HCO2198 (Folmer et al. 1994) were used, yielding 658 bp of the barcode region of the COI gene. The resulting chromatograms were edited in UGENE 46.0 (Okonechnikov et al. 2012). Sequences were aligned using MAFFT v7.490 (Katoh and Standley 2013). A total of 10 sequences published earlier (Schär et al. 2018) were added, yielding a data set of 57 sequenced specimens in total. A maximum likelihood tree was calculated in IQ-TREE v1.6.12 (Nguyen et al. 2015), selecting the TIM2+F+I model with ModelFinder (Kalyaanamoorthy et al. 2017). Monophyly of the subfamilies was enforced by constraint. Sequences were collapsed into haplotypes and a haplotype network was visualized using the R package “pegas” (Paradis 2010) in R v4.2.3 (R Core Team 2023). Within species raw genetic distances were calculated using the R package “ape” (Paradis and Schliep 2019). All retrieved unique COI sequences were compared to sequences on BOLD.

Worker specimens of suspected F. lemani from Kamchatka displayed some morphological traits intermediate between F. lemani and F. fusca. Therefore, a morphometric analysis was performed to clarify their identity. Measurements were taken using a Leica EZ4 binocular with graticule at 70× magnification. A total of 110 specimens from the author’s private collection were examined: 67 specimens of Formica lemani, 31 workers and 3 gynes from Kamchatka and 28 workers and 5 gynes from Europe. For comparison, 43 specimens (39 workers and 4 gynes) of F. fusca from Europe were measured as well. The following measurements were taken as suggested and defined by Seifert (2018): SL, EYE, OceD, nPn and nHFfl (workers) and SL, nPn and nMFfl (gynes). The discriminant given by Seifert (2018) to separate F. lemani (positive values) from F. fusca (negative values) was calculated for workers (D = 0.385 * nPn + 0.51 * nHFfl - 6.546 * SL - 27.22 * EYE + 38.46 * OceD + 10.578) and gynes (-D = 0.215 * nPn + 0.666 * nMFfl - 7.179 * SL + 8.657).

Literature review suggested a relatively clear distinction between reliable and questionable records among the 25 species of ants reported from Kamchatka (Table 1). For a total of 12 reported species, presence in Kamchatka seems reliable. These taxa were mentioned by at least 2 independent sources and including sources published within the last two decades. They belong to 4 genera (Camponotus Mayr, 1861, Formica Linnaeus, 1758, Leptothorax Mayr, 1855, and Myrmica Latreille, 1804). For the other 13 species and 1 genus (Lasius Fabricius, 1804), reports from Kamchatka appear questionable and these taxa are considered absent from Kamchatka for the remaining manuscript. These questionable records were reported by a single (n = 11) or multiple (n = 2) dated sources, but have all been contradicted by more recent work (Table 1).

A total of 9 species of ants were found across 8 localities in Kamchatka (Fig. 1A) during a search of the author in the year 2016 (Table 1). Of those, 3 species are restricted to the East Palearctic, 3 have a (trans-) Palearctic and 3 a Holarctic distribution range (Table 2). Ants were found from sea level up to an altitude of approximately 1000 m. At low and intermediate elevations, they inhabited meadows (Fig. 1D), stone birch- (Fig. 1E) and larch/birch forests (Fig. 1B). At higher altitudes they were found in the dwarf pine zone (Fig. 1C). Ant species compositions in these habitat types are shown in Table 2. Larch/birch forests at mid-elevations harbored the highest number of species, while the widespread stone birch communities at low elevations were species poor, with only 3 ant species observed (Table 2). No ants were observed in alpine tundra and plant-less habitats above 1200 m a. s. l.

DNA-barcoding succeeded in high quality for all 9 species and 57 specimens examined. For all but 2 specimens, the full length DNA-barcode (658 bp) could be retrieved. The exceptions were a specimen of M. kamtschatica and one of M. displicentia, for which sequences of 651 and 647 bp length were obtained. All 9 studied species were separated by a clear DNA-barcoding gap (Fig. 2). Interspecific genetic raw distance ranged from 2.3% (F. candida vs. F. gagatoides) to 22.5% (C. herculeanus vs. L. acervorum). Genetic diversity was low and for 6 of the 9 investigated species, only a single COI haplotype was found. Another 2 species (Formica lemani and Myrmica displicentia) were represented by 2 haplotypes each. The exception was Leptothorax acervorum, for which all 7 sequenced specimens displayed a different haplotype, and two haplogroups were detected (Fig. 3). Genetic raw distances within L. acervorum ranged between 0.15%–2.13% (Table 2). Genetic comparison to sequences on BOLD revealed that all except one encountered haplotypes of ants from Kamchatka were private, not found anywhere else so far. The exception is the only haplotype found for F. lugubris, which also occurs in Finland (Roslin et al. 2022). Closely related sequences were found for all observed haplotypes (Table 2), with the exception of the two Myrmica species endemic to East Asia. For those, the most closely related sequences were from unidentified Myrmica from the Primorje region of Far Eastern Russia (M. displicentia: 97.22%; M. kamtschatica: 97.41% similarity). Sequences were made accessible via the European Nucleotide Archive (see ESM for accession numbers).

Figure 2. 

Maximum-likelihood tree of 57 COI DNA-barcode sequences representing 9 species of ants from Kamchatka. Pictures: S. Schär.

Figure 3. 

Haplotype network for 18 COI DNA-barcode sequences of Leptothorax acervorum. Sequences not from Kamchatka are from Schär et al. (2018).

The discriminant D suggested by Seifert (2018) to distinguish F. lemani from F. fusca in Europe allowed almost complete separation between the two species for samples from Europe. Only a single F. fusca worker had a value for D slightly surpassing 0 (Fig. 4). For the investigated workers from Kamchatka, values of D were intermediate between European F. lemani and F. fusca (Fig. 4). About two-thirds (68%) of these workers had negative values of D, as European F. fusca, while the remaining third (32%) had positive values of D, as European F. lemani. None of the colonies from Kamchatka contained only workers with positive values of D, unlike European F. lemani. A different result was found for the 3 available gynes from Kamchatka, all of which yielded discriminant values suggestive of F. lemani. Additionally, they also showed a trait of F. lemani-gynes not included in the discriminant calculation, namely a shiny tergite I (covered by dense pubescence in F. fusca). Two of these gynes were found in the same nest with workers with partially negative values of D, as in European F. fusca (Fig. 4).

Figure 4. 

Scatterplot showing a morphometrical comparison of 34 Formica lemani ants from Kamchatka with 33 F. lemani specimens from from Europe and 43 F. fusca specimens from Europe. The discriminant D has been calculated according to Seifert (2018). For gynes, D was multiplied by -1 in order to match directionality of the discriminant used for workers. Symbols connected by lines are specimens from the same nests.

Based on literature review, field work and DNA-barcoding, I conclude that the ant species diversity of Kamchatka is lower than could be assumed based on the published record. Unreliable, earlier species records may be based on incorrect collection information, taxonomic changes or misidentifications of taxonomically difficult taxa. However, Kamchatka is a large peninsula and most of it is poorly explored. Thus, the 12 species recognized here may be a conservative estimate. The low observed number of ant species encountered may be explained by the cold climate of Kamchatka in combination with its geographic isolation. The only land bridge connecting it to the mainland is mountainous and located at around 60°N (Fig. 1A). Most ant species are sensitive to low temperatures and only exceptionally cold resistant species are likely to have colonized the peninsula via this route. The southern part of the landmass is separated from the mainland by several hundreds of kilometers of ocean. Therefore, in contrast to boreal species, species adapted to more temperate climate may not have been able to colonize Kamchatka. This may also explain why only moderate altitudinal variation in ant communities was observed. Of the 9 species encountered in the field, 4 (F. lemani, F. lugubris, L. acervorum and M. kamtschatica) were found across the entire altitudinal range observed to be inhabited by ants (0–1000 m). Two species (F. candida and M. displicentia) were observed only below and 3 (C. herculeanus, F. exsecta and F. gagatoides) only above roughly 500 m a. s. l. (Table 2). The highest number of species was found at an intermediate altitude of around 500 m a. s. l., where all of these species could potentially co-occur. The low species and genetic diversity of ants also support the previous notion, that the fauna of Kamchatka resembles that of an island rather than that of a mainland peninsula (Jones and Solomina 2015).

Figure 5. 

Morphological traits used for the identification of ants found in Kamchatka (see identification key).

Similar to the low observed species diversity, genetic diversity and within species divergence in mtDNA was low too (Table 2). This could be a result of an edge effect due to recent colonization of Kamchatka by the investigated species. The exception was L. acervorum, which displayed a surprising level of COI diversity, with every sequenced sample being of a different haplotype and genetic distance raging up to 2.13% (Table 2). The latter high genetic distance was found due to a divergent L. acervorum sample (Figs 2, 3), belonging to a haplogroup previously only known from North America (Smith et al. 2016; Schär et al. 2018) (Fig. 3). The high genetic diversity found within L. acervorum in Kamchatka is surprising. Possible explanations include that Kamchatka has been a long term refuge for this species. Alternatively, it could be suggestive of a strong dispersal ability of L. acervorum, i. e. wind dispersal of the small sexual castes over long distances. This latter theory is also supported by a previous observation of an L. acervorum-haplotype occurring in both Canada and Europe (Ojeda et al. 2021). For the 3 Holarctic species (C. herculeanus, F. gagatoides and L. acervorum), comparison of DNA-barcodes with sequences on BOLD revealed the existence of closely related haplotypes (<1% distance) in North America (C. herculeanus and F. gagatoides) or in Europe and North America (L. acervorum). This suggests more recent gene-flow across Beringia than between West and East Palearctic in these taxa. On the other hand, low or even lack of genetic distances were found for the three Formica species with trans-Palearctic distributions (F. exsecta, F. lemani and F. lugubris). This could mean that the latter have expanded their distribution ranges more recently than the former three Holarctic species. It is less likely to be caused by better dispersal ability of F. exsecta, F. lemani and F. lugubris, because they are absent from the Nearctic. The two Myrmica species endemic to East Asia (M. displicentia and M. kamtschatica) have not been represented on BOLD before. Comparison to other BOLD specimens supported the current placement of both taxa within the M. lobicornis species group (Radchenko and Elmes 2010) and their separate status from other sequenced members of this group.

Among the questionable ant species records from Kamchatka, F. fusca is the one mentioned by the the most (3) literature sources (Table 1). Two of these sources also mentioned F. lemani from Kamchatka. Among the worker specimens examined here, two-thirds showed morphological traits more suggestive of F. fusca, while one third was suggestive of F. lemani (Fig. 4). Thus, the presence of both species appears plausible at first. However, two observations lead to the conclusion that all of these specimens belong to the same species:

1. Specimens with both morphologies are found in the same nests (Fig. 4), and no nest contained exclusively specimens with the traits of F. lemani.
2. All but one sequenced colonies had the same COI haplotype (Fig. 2). The remaining sequence differed from the others by just a single nucleotide change (0.15%, Table 2).

Additionally, despite the majority of workers resembling F. fusca, it is argued here that the name of this species is F. lemani rather than F. fusca, unless F. lemani would be revealed as a junior synonym of F. fusca in the future. This conclusion is entirely based on the morphology of gynes. Gynes of the two taxa are more distinguishable than workers. Workers differ almost only by the number of macro-setae, a trait that can vary within species in Hymenoptera (Sandrock et al. 2011) and therefore should be viewed as insufficient for taxonomic decision making. In gynes, the pubescence density on the first tergite is an additional trait used for discrimination of the two taxa: the first tergite of F. fusca gynes is covered by dense pubescence dorsally, leading to a matt appearance, while this body part is smooth and shiny due to low pubescence density in F. lemani gynes (Seifert 2018). The three gynes found from Kamchatka all had a shiny first tergite, along with high numbers of setae typical for F. lemani (Fig. 4). Two of these gynes were found in the same nest with workers resembling F. fusca (Fig. 4). In previous investigations, F. fusca and F. lemani could not be separated based on nuclear microsatellite markers (Seppä et al. 2011) and mitochondrial COI sequences (Schär et al. 2018). Similarly, rather low support for heterospecificity of these two taxa was also found for the mitochondrial Cyt-b gene (Antonov and Bukin 2016). It is therefore possible that F. fusca and F. lemani are morphs of the same species and this should be investigated in the future. If so, the name F. fusca Linnaeus, 1758 would have priority over F. lemani Bondroit, 1917.