Water mites of the genus Hygrobates Koch, 1837 (Acari: Parasitengona: Hygrobatidae) from Hokkaido, northern Japan

Nine species of Hygrobates (Acari: Parasitengona: Hygrobatidae) from Hokkaido, Japan are described or redescribed from newly collected material and historical specimens. Treated herein are eight species in the subgenus Hygrobates, including one new species, H. bibi sp. nov., as well as H. calliger Piersig, 1896; H. foreli (Lebert, 1874); H. japonicus Uchida, 1931; H. longipalpis (Hermann, 1804); H. longiporus Thor, 1898; H. nigromaculatus Lebert, 1879; and H. sokolowi Thor, 1927. Also treated is H. ezoensis Uchida, 1934 in the subgenus Rivobates. A lectotype and paralectotype are designated for H. japonicus Uchida, 1931. Hygrobates (s. str.) heteropalpis Imamura, 1954 is synonymized with H. calliger Piersig, 1896. The name H. ezoensis Uchida, 1934 is resurrected from synonymy with H. diversiporus Sokolow, 1927. Six species previously known from Hokkaido were collected in the study: H. foreli, H. japonicus, H. longipalpis, H. longiporus, H. diversiporus, and H. ezoensis. New records for both Hokkaido and Japan include H. nigromaculatus and H. sokolowi. A species previously recorded from Hokkaido, H. taniguchii Imamura, 1954, was not found in this study. Three new characters are proposed as useful for the taxonomy of the genus Hygrobates: the ratio of the distance between the P‐4 ventral setae to P‐4 length, the ratio of the length of the longest terminal seta on IV‐L‐5 to the length of IV‐L‐5, and the nature of the outer border of the genital plates.


Introduction
So-called water mites, members of the Phalanx Hydrachnidia Krantz, 1978 (Acari: Parasitengona), are the most diverse group of mites living in aquatic habitats. More than 3000 species in seven superfamilies are assigned to the Hydrachnidia worldwide (Viets 1987). Over 250 species in 58 genera distributed among all the superfamilies have been recorded from Japan. The genus Hygrobates Koch, 1837 is one of the largest genera of the family Hygrobatidae Koch, 1842. Most species of Hygrobates occur in still or slowly flowing waters, such as lakes, ponds, and the mid-and downstream parts of rivers, where they live among vegetation and between and underneath rocks. Hygrobates has a diversity of more than 150 species in 11 subgenera worldwide (Cook 1974;Tuzovskij and Gerecke 2003). In Japan, 14 nominal species from two subgenera have been previously recorded: 12 species of the subgenus Hygrobates Koch, 1837, including H. biwaensis Tuzovskij, 2003;H. calliger Piersig, 1896; H. foreli (Lebert, 1874); H. heteropalpis Imamura, 1954;H. japonicus Uchida, 1931; H. longipalpis (Hermann, 1804); H. longiporus Thor, 1898; H. minutus Imamura, 1953;H. papillosus Imamura, 1953;H. rarus Tuzovskij, 2003; H. sinensis Uchida and Imamura, 1951;and H. variabilis Tuzovskij, 2003; and two species of the subgenus Rivobates Thor, 1897, including H. diversiporus Sokolow, 1927 andH. taniguchii Imamura, 1954. Despite much previous work on water mites in Japan, the taxonomy of many groups is poorly resolved due to insufficient original descriptions or inadequate evaluation of taxonomic characters; as a consequence, it is often difficult to identify specimens and recognize new species. To begin to address these problems, we conducted a taxonomic study of the genus Hygrobates on Hokkaido, Japan's northernmost island. Among the 14 species of Hygrobates known from Japan, all except H. biwaensis, H. minutus, H. papillosus, H. rarus, H. sinensis, and H. variabilis were known to occur on Hokkaido (Uchida 1931(Uchida , 1934(Uchida , 1936bEnami 1940;Imamura 1950Imamura , 1953aImamura , 1953bImamura , 1954Imamura , 1955Imamura , 1960Imamura , 1980Tuzovskij 2003).
In this paper, we provide original descriptions or redescriptions of nine species of Hygrobates from Hokkaido detected through examination of extensive newly collected material and specimens in the collections of two previous workers, Uchida and Imamura. The species treated here comprise eight species of the subgenus Hygrobates, including one new species and two new records for Japan, and one species of the subgenus Rivobates. Among the species previously known from Hokkaido, this study identifies all but one, H. taniguchii. We recommend two nomenclatural changes: removal of the name H. heteropalpis through synonymy with H. calliger, and resurrection of the name H. ezoensis for Hokkaido specimens previously identified as H. diversiporus. We also nominate a lectotype and paralectotype for H. japonicus Uchida, 1931.

Material and methods
Most of the specimens examined in this study were collected by the first author during 2002-2003 at 35 localities across Hokkaido ( Figure 1; Table I). Qualitative sampling was conducted in freshwater habitats (rivers, ponds, springs, and lakes) with a net of 250 mm mesh size. Specimens were either fixed and preserved in Koenike's fluid, or collected into 80% ethanol and later transferred to Koenike's fluid. There was no visible difference in effect on integument pattern between these two treatments. Specimens were dissected in lactic acid before mounting in polyvinyl alcohol (Danielsson 1985) or Euparal.
Specimens in Uchida's collection in the Hokkaido University Museum and Imamura's collection at Ibaraki University were also examined. Information on specimens examined in this study is noted in the ''Material examined'' section for each species.
Observations and measurements were made with a differential interference microscope (Olympus BH-2), and illustrations were prepared with the aid of a camera lucida. Measurements are given in micrometres; ratios are without units. Measurements of paired  Table I.     Figure 2D). Antenniform setae thickened, each located on a small, rounded base ( Figure 2E), 45 in length. Without dorsalia.
Capitulum ( Figure 2A). Broadly fused with the first coxae, 139 in width; anterior margin of each half of capitulum with a rounded notch.

Remarks
Hygrobates bibi sp. nov. is most similar to H. calliger, but differs from the latter in four of 18 characters (Table II). Hygrobates bibi has a rounded projection on P-2, the ventral margin of P-3 convex, the pair of ventral setae on P-4 very close together, and an integument pattern of fine striation. In contrast, H. calliger has a sharper, triangular P-2 projection, the ventral side of P-3 flat, the pair of ventral setae on P-4 farther apart, and an integument pattern of strong lineation. Examination of the old collections revealed that the new species had already been collected from Japan prior to our study. Three of about 20 specimens labelled H. japonicus in Uchida's collection and both specimens labelled H. japonicus in Imamura's collection proved to be H. bibi. Imamura (1954) noted some differences in the morphology of P-3 between his specimens and the original description of H. japonicus (Uchida, 1931), but concluded that these differences were due to the immaturity of his specimens. Examination of more than 20 specimens newly collected from Hokkaido has revealed that P-3 morphology is diagnostic between H. japonicus and the new species. The character state Uchida (1931, p 266) described for H. japonicus as ''Der gezackte Hö cker auf den Beugeseite des dritten Gliedes ist niedrig, aber ziemlich breit'' (serrated knob on the ventral side of P-3 is low, but relatively broad; our translation) is separable into ''ventral side of P-3 with a low knob in the middle'' ( Figure 11B) in H. japonicus and ''ventral side of P-3 convex without a distinct knob'' ( Figure 2B) in H. bibi, without any transitional states.
Other characters previously used to distinguish species of Hygrobates include the morphologies of P-2 and the genital plates, and the distribution of the ventral setae on P4. Uchida (1931, p 265) mentioned for H. japonicus, ''auf den Beugeseite des zweiten Palpengliedes befindet sich ein mit Spitzen besetzter stumpfer Zapfen'' (on the ventral side of P-2 is a blunt projection occupied by points; our translation), but provided a poor illustration. Once again, the character state described by Uchida can be separated into two states: ''projection narrow, with a truncated end'' ( Figure 11B) for H. japonicus and ''projection broad, with a round end'' ( Figure 2B) for H. bibi, also without any transitional states.
A comparison of character states between H. bibi sp. nov. and H. japonicus is summarized in Table II. The two species differ in 10 of the 18 characters compared, including the morphology of the P-2 projection and the ventral side of P-3, the P-4 setae ratio, the ratio of the longest terminal seta on IV-L-5 to length of IV-L-5, the shape of the male genital plate, and secondary sclerotization of the coxoglandularia. Uchida (1931) noted that H. japonicus has the two setae on ventral side of P-4 widely separate ( Figure 11B). However, his specimens labelled H. japonicus include specimens of H. bibi that have these setae close together ( Figure 2B). The ratio of the distance between the two setae to P-4 length does not overlap between H. japonicus and H. bibi, and the means are significantly different ( Figure 4). Finally, as a novel taxonomic character for distinguishing species of water mites, we propose here the ratio of the longest terminal seta on IV-L-5 to the length of IV-L-5. This ratio does not overlap between H. japonicus and H. bibi, and the means are significantly different ( Figure 5, also compare Figures 3D and 12D).

Remarks
Imamura (1954) described Hygrobates heteropalpis from Hokkaido as very similar to H. calliger, noting that the former has a longer ventral projection on P-2 compared to the latter. Two P-2 projections in the original illustration of H. heteropalpis (Imamura 1954, p 78, Figure 45b, c) certainly seem to be longer than those of H. calliger calliger illustrated by Imamura (1953a). However, the P-2 projections of three subspecies of H. calliger (H. calliger calliger, H. calliger obtusipalpis, and H. calliger latilaminata) illustrated by Viets (1930, p 378, Figures 13-15) do not seem to be shorter than those of H. heteropalpis. We addressed this problem by examining the shape variation of the P-2 projection among 11 specimens of the two species, including the holotype and paratype of H. heteropalpis, three specimens of H. calliger from Sicily and North Tuscany, Italy, and six specimens from a population at Nanashi-gawa river, Niseko, southern Hokkaido. As is clearly shown in Figure 8, there are no diagnostic differences in the P-2 projection between the two nominal species. Both the length and shape of the projection in the types of H. heteropalpis fall within the range of variation of the Hokkaido population of H. calliger. The P-2 projection of the European specimens of H. calliger tends to be broader than that of either H. calliger or H. heteropalpis from Japan, but again the character overlaps (e.g. compare Figure 8B with 8E). Except for uncertainty regarding the male characters of H. heteropalpis (the male has not been described), H. heteropalpis and H. calliger overlap completely in P-2 morphology and other characters examined (Table II). With little doubt, H. heteropalpis Imamura, 1954 is a junior synonym of H. calliger Piersig, 1896.
Hygrobates calliger is most similar to H. bibi sp. nov. but differs from the latter in four of 18 characters (Table II). The former has a narrow, spike-like P-2 projection, the flat ventral side of P-3 flat, and a strongly lineated integument, whereas the latter has a broad, rounded P-2 projection, the ventral side of P-3 convex (compare Figures 6B and 2B), and a finely striated integument. The P-4 setae ratio is somewhat greater in H. calliger than in H. bibi.

Remarks
Hygrobates foreli has previously been reported twice from Hokkaido (Uchida 1934;Imamura 1954). This species and H. longipalpis are both larger in overall body size than the other species treated. However, H. foreli is most similar to H. nigromaculatus; the two share 14 of 18 characters (Table II)
Capitulum ( Figure 11A). Broadly fused with the first coxae. Anterior border of each half of capitulum is irregularly truncate, without a notch.

Description of male
Paralectotype; ZIHU-2331. Characters as given for female, except for genital field. Chelicerae and palps missing. Genital field ( Figure 11G). Genital plate length/width 194/240; outer border irregularly crenate with three distinct lobes. Middle third of anterior border raised as a rounded projection; posterior border irregular, with a wide, shallow median notch; apodemes lacking. Genital plate with 22 setae on right side and 23 on left. Three genital acetabula on each side, arranged in an obtuse triangle. Distance between Ac1 and Ac2, and Ac2 and Ac3, less than width of Ac1. Length/width Ac1 65/45, Ac2 75/39, Ac3 71/36.

Distribution
Middle and northern Japan; eastern Russia (Primorsk, Sokolow 1940). Uchida (1936a), in his redescription of H. japonicus erroneously based on specimens of H. longipalpis (see below), noted that ''This species is common also in Sapporo''. In fact, his specimens labelled H. japonicus from Sapporo include both H. longipalpis and H. bibi, but not H. japonicus.
Remarks Uchida (1931) described this species based on two specimens from Tokyo, Japan. We located these specimens in Uchida's collection by comparing the collection data cited in the original description with information on the specimen labels. Since Uchida did not designate types for the species, we have designated one of the specimens as the lectotype and the other as a paralectotype for H. japonicus. The male specimen designated as the paralectotype lacks palps; therefore, we used another specimen from Uchida's collection, confirmed by us to be H. japonicus, for observation of the male palps. These three specimens had been labelled by Uchida only as ''Hygrobates sp.'' The collection contained another 25 specimens labelled as H. japonicus. Among these, only one specimen proved to be H. japonicus; the rest were either H. bibi or H. longipalpis. Uchida (1936a) redescribed H. japonicus on the basis of over 20 specimens collected from Kunashiri Island, off the north-west tip of Hokkaido. By comparing the collection data cited in his text with information on specimen labels, we located in Uchida's collection the specimens he used for the redescription, all labelled H. japonicus. Interestingly, all of these proved to be H. longipalpis! Uchida's redescription is far from current standards and of no use in discriminating H. japonicus from H. longipalpis, because nine characters we have found useful to distinguish between the two were poorly described, with no illustrations. For example, Uchida did not mention whether the posterior end of the anterior coxal group is round versus triangular, nor whether the suture line between Cx3 and Cx4 is complete versus incomplete, both of which are diagnostic between H. japonicus and H. longipalpis. Likewise, he indicated the shape of the P-2 projection only as ''blunt'' and failed to note the distribution of denticles, and he failed to describe adequately the shape of the male genital plate and the position of the pregenital sclerite of the female.
Hygrobates japonicus is, in fact, most similar to H. longipalpis; the two share nine of 18 characters (Table II). In particular, only these two species have P-3 with a prominent ventral convexity bearing a compact, limited denticulate patch. The most important character diagnostic for H. japonicus is the flat end of the P-2 projection, which separates it from the other six Japanese species in the subgenus Hygrobates.

Remarks
There are several previous records of H. longipalpis from Hokkaido (Uchida 1936b;Imamura 1950Imamura , 1954. Among the eight species from Hokkaido, this species is most similar to H. foreli, H. japonicus (see Remarks section for that species) and H. longiporus which shares nine of 18 characters (Table II). H. longipalpis and H. longiporus exclusively share one coxal and two genital features: a complete suture line between the Cx3 and Cx4, a notched anterior border of the male genital plate, and the pregenital sclerite located posterior to the anterior ends of the female genital plates. This species was common in ponds, and in slowly flowing rivers close downstream from springs. It co-occurred with H. bibi sp. nov. in the Bibi-gawa River, Bibi, Chitose.

Material examined
One adult female from Hamamasu, 31 July 2001; one adult female and one adult male from Gosen-gawa River, Horokanai, 7 July 2002; three adult females and three adult males from Horonobe, 20 August 2002; one adult female from Otoineppu, 7 July 2002.
Capitulum ( Figure 15A). Broadly fused with the first coxae and 140 (130-149, n55) in width. Anterior portion of each half of capitulum with a rounded notch.

Remarks
Hygrobates longiporus has been previously recorded once from Hokkaido (Imamura 1954). It is most similar to H. longipalpis (see Remarks section for the latter) sharing nine of 18 characters with each (Table II). Among the eight Japanese species of this subgenus, only H. longiporus and H. sokolowi have the genital plate with the distance between Ac1 and Ac2 greater than the width of Ac1 in the female, and greater than or equal to the width of Ac1 in the male. Hygrobates longiporus is easily distinguished from the other seven Hygrobates species by two unique characters: heavy secondary sclerotization of coxoglandularium 2 and the long patch of denticles on P-2 occupying nearly the entire ventral segment length. This species was common among stones on the bottoms of rivers.

Material examined
Two adult females and two adult males collected from Horonobe, 20 August 2002.

Description of male
Based on two specimens, ZIHU-3093 and 3094, collected by the first author.
Capitulum ( Figure 21A). Fused with the first coxae, 104 in width. Anterior portion of each half of capitulum with a rounded notch.
Coxae ( Figure 21A). Length/width anterior coxal group 318/402. Length/width posterior coxal group 337/279. Posterior end of anterior coxal group broadly rounded; posterolateral apodemes extend slightly beyond sclerotization, 188 from tip to tip. Suture line between first coxae and capitulum irregular, not curved. Coxoglandularium I on Cx2. Suture line between Cx3 and Cx4 incomplete, extending to near glandularium of Cx4. Medial margin of Cx4 variable (evenly rounded or with a rounded angle projecting in middle), without angle apodemes. Genital field ( Figure 21F). Entire genital field 318 in width. Genital plates with somewhat irregular outer border. Right genital plate with 16 setae, left with 16. No setae on membranous integument near genital plate. Length/width genital plates 149/91. Genital acetabula closely packed in each genital plate, seven on left side, six on right ( Figure 21F is from a female specimen other than allolectotype). Pre-and postgential sclerites each with a median apodeme. No setae on pregenital sclerite. Pregenital sclerite located anterior to the anterior ends of genital plates. Postgenital sclerite located anterior to the posterior ends of genital plates. Genital opening between pre-and postgenital sclerites.

Description of male
Lectotype, ZIHU-2362. Characters same as for female, except for genital field. Antenniform setae missing from specimen.
Coxae. Length/width anterior coxal group 324/389. Length/width posterior coxal groups 324/311. Posterolateral apodemes of anterior coxal group 188 from tip to tip. diagnostic differences in characters among these specimens from Uchida's collection and about 30 specimens we collected from various localities on Hokkaido for our study. Thus we confirmed that the species of the subgenus Rivobates that we collected on Hokkaido is the same that Uchida identified from both Hokkaido and Sakhalin. The question remained as to the species' identity.
The taxonomy of H. ezoensis/diversiporus has been convoluted. In his original description, Uchida (1934, p 95) noted for Rivobates ezoensis, ''The species is easily distinguished … from R. diversiporus (Sokolow) by the form of the third epimera and genital plates''. However, in later synonymizing Rivobates ezoensis with R. diversiporus, Uchida (1936b, p 314) concluded that ''these differences are due to individual variations'' and remarked that ''Sokolow's specimens are probably somewhat aberrant ones of the species''. Imamura (1954), on the other hand, did not agree with this synonymy and stated, ''it [H. ezoensis] is distinguished from the latter [H. diversiporus] in the shapes of genital plates of both sexes''. Nonetheless, Viets (1956) accepted the synonymy, and Imamura (1980) listed H. diversiporus in an illustrated encyclopaedia of Japanese mites and ticks, to the exclusion of H. ezoensis.
We examined the male genital plate of 18 specimens from Sakhalin and Hokkaido, of which 11 are illustrated in Figure 23A-K, including three of Uchida's specimens of H. ezoensis from Chitose, Hokkaido (Figure 23C, D, G); three of Uchida's specimens from Sakhalin labelled as H. diversiporus ( Figure 23H, J, K); and five specimens we collected from northern ( Figure 23A, B, E) and central ( Figure 23F, I) Hokkaido. In all 18 specimens examined, the male genital plate is irregularly ovoid, with the anterior border a rounded obtuse angle and the posterior border straight, irregularly convex, or broadly curved (that in Figure 23H is atypical, likely due to a developmental anomaly or healed wound). This form of the male plate is just as originally figured by Uchida (1934, p 93, Figure 26) for Rivobates ezoensis. It differs from the male plate ( Figure 23L) of H. diversiporus originally described by Sokolow (1927) as ''verkehrt-herzfö rmiger'' (reverseheart-shaped; our translation), with an acute anterior angle and a depressed posterior border having a broad, rounded median notch. We thus conclude that H. ezoensis (Uchida, 1934) is a valid species, distinct from H. diversiporus Sokolow, 1927, and that the name H. ezoensis should be restored. Unfortunately, the male genital plate seems to be the salient difference between the two species, as we detected no clear diagnostic differences in the female genital plates or other characters.
Since Uchida (1934) did not designate types for the species, one of the male specimens from his collection is here designated as lectotype (ZIHU-2362) and one of the female specimens (ZIHU-2368) is designated as allolectotype of H. ezoensis (Uchida, 1934).
Another Japanese species in the subgenus Rivobates, H. taniguchii, was described by Imamura (1954) on the basis of the two specimens collected from Kamishokotsu, Hokkaido. Despite extensive collecting across Hokkaido, we did not find this species.

Discussion
The confused state of Hygrobates taxonomy in northern Japan prior to our study is reflected in a number of misidentifications we detected in historical collections. For example, 25 specimens labelled H. japonicus in Uchida's collection included three species: nine specimens were H. bibi, 15 were H. longipalpis, and one was H. japonicus. Similarly, two specimens labelled Hygrobates japonicus in Imamura's collection turned out to be H. bibi. There are likely a number of reasons why these previous workers failed adequately to diagnose species among their material. One was that the standard of description was somewhat lower then than now. Uchida (1931, p 266), for example, originally described the P-2 projection of H. japonicus only as a blunt projection occupied by points, and his original figure (1931, Figure 6) shows a P-2 projection more like that of H. longipalpis than H. japonicus. In his redescription of H. japonicus, Uchida (1936, p 178) again described the P-2 projection as a blunt process covered with denticles and provided no new illustration. Our study shows that the P-2 projection is a diagnostic character among H. bibi, H. longipalpis, and H. japonicus, with a morphology that is broad and rounded, narrower and sharper, and squarely truncated, respectively.
Failure to separate species correctly, however, was due to more than what we would now consider as inadequate descriptions. In fact, Uchida's original figure of H. japonicus accurately represents the forms of P-3 and the male genital plate, which taken together unambiguously diagnose this species. A further shortcoming was likely an overly broad estimation of intraspecific variation, that is, a failure to distinguish correctly inter-from intraspecific variation. Three of the most important characters for distinguishing among H. japonicus, H. bibi, and H. longipalpis are the forms of the P-2 projection, P-3, and the male genital plate. For each of these characters, it is possible to view variation as continuous, if one assumes a priori a broad range of intraspecific variation. For example, the ventral side of P-3 might seem to range from flat (H. longipalpis) to uniformly convex (H. bibi) to having a convex protrusion in the middle (H. japonicus), or the outline of the male genital plate from smooth (H. bibi) to moderately crenate (H. japonicus) to highly crenate (H. longipalpis). In a similar vein, Imamura (1953b) mistakenly attributed the discrepancies in palp morphology between his specimens and the original description of H. japonicus to ontogenetic variation within H. japonicus; thus, he failed to recognize the new species here described as H. bibi.
A final drawback of previous studies may have been examination of only limited material from a few localities. In all, our study examined approximately 300 newly collected specimens, 127 of which were prepared on slides, from 35 localities across Hokkaido.
Difficulties in correctly partitioning inter-from intraspecific variation will be magnified if undue weight is given to a single character. Here we have taken a polythetic approach, describing in detail numerous characters from each species encountered. Suites of characters, as shown in Table II, allow reliable identification of all species we have treated. However, many of the traditionally used characters are qualitative and in some cases require experience in interpretation. To improve species definition, we propose two new quantitative characters for the taxonomy of the genus Hygrobates that have not been considered significant in the previous literature. These are the ratio of the distance between the P-4 ventral setae to P-4 segment length, and the ratio of the length of the longest terminal seta on IV-L-5 to the length of IV-L-5. Although not diagnostic among all the species treated in our study, these characters are distinct, for example, between H. japonicus and H. bibi (Figures 4, 5). In addition, we have for the first time noted the sculpturing of the outer border of the genital plates, expecting an important role of this character in future taxonomy.
Representatives of the genus Hygrobates occur worldwide, on many islands and all continents, apparently except Antarctica. Some of the species on Hokkaido are broadly distributed in the Palaearctic (H. calliger, H. foreli, H. longiporus, H. nigromaculatus) or Holarctic (H. longipalpis) regions (Viets 1978). Prior to our study, all of these but H. nigromaculatus were known in Japan from Kyushu, Honshu and Hokkaido Islands (Enami 1940;Imamura 1950Imamura , 1953aImamura , 1953bImamura , 1954Imamura , 1955Imamura , 1960Uchida 1931Uchida , 1934Uchida , 1936aUchida , 1936bN. Matsumoto, unpublished data). We recorded H. nigromaculatus for the first time from Hokkaido and also from Japan. This suggests that with further research, other widely distributed water mite species can be expected in Japan.
Hygrobates ezoensis and H. bibi are examples of species with narrow known distributions. Although H. ezoensis is common in springs in Hokkaido and occurs on adjacent Sakhalin Island, it has not been recorded elsewhere. Hygrobates bibi is so far known only from Hokkaido and central Honshu Islands, Japan.
Hygrobates sokolowi was previously thought to be a rare species with a limited distribution within Europe. Our records from a single locality on Hokkaido considerably extend the range to eastern Eurasia. Thus H. sokolowi may have a considerably broader distribution than was previously known, extending across northern Eurasia. If this broad range has previously gone undetected, it suggests much remains to be learned about the biogeography even of widely distributed species.