Radfordia (Microtimyobia) mites infesting arvicoline voles in Japan comprise three new subspecies of Radfordia (M.) lemnina (Koch, 1841). These new subspecies are defined on the basis of the form of the male genitalia, the shape of the anterior gnathosomal setae, and the chaetotaxy of the female tritonymph. Radfordia (M.) lemnina hata ssp. n. is a parasite of Microtus montebelli (Milne-Edwards, 1872). Radfordia (M.) lemnina mikado ssp. n. infests only Clethrionomys rutilus mikado Thomas, 1905 in Hokkaido, and it differs from Radfordia (M.) lemnina rutila Fain and Lukoshus, 1977 parasitizing Clethrionomys rutilus (Pallas, 1979) in Europe. Radfordia (M.) lemnina japonica ssp. n. is found widely on the other Clethrionomys and Eothenomys voles in Japan. Among these host voles, Clethrionomys rufocanus bedfordiae Thomas, 1905 is designated as the type host of this new subspecies. Immature stages of the three new subspecies are described and illustrated. Radfordia mites parasitic on Eothenomys voles in Taiwan and Japan are much more different from each other than would be expected on the basis of the current systematics of the Arvicolinae proposed by Japanese mammalogists.
INTRODUCTION
Ono (1969) first carried out studies on Myobiidae parasitic on Japanese voles and recorded Radfordia lemnina (Koch, 1841) as the only myobiid of Arvicolinae (=Microtinae) (Uchikawa, 1977). Fain and Lukoschus (1977) summarized the 10 subgenera and 19 species and subspecies of the genus Radfordia Ewing, 1938 and tabulated the hosts and localities for the thus far known 69 taxa of 10 subgenera including the sugenus Microtimyobia Fain and Lukoschus, 1976 associated exclusively with Arvicolinae. The morphology of R. (M.) lemnina, which they designated as the type species of the subgenus, was redescribed; then, the nominate and three other subspecies within R. (M.) lemnina were characterized (Fain and Lukoschus, 1977). The idiosomal outline and chaetotaxy of the females are almost the same in all four subspecies; Fain and Lukoschus (Fain and Lukoschus, 1977) used these characters to define the species and other features for the subspecies.
Fain and Lukoschus (Fain and Lukoschus, 1977) also examined myobiids found on Microtus montebelli (Milne-Edwards, 1872) from Fukuoka, Japan, and Clethrionomys rufocanus bedfordiae Thomas, 1905 from Hokkaido, Japan, and identified the forms from both hosts with Radfordia lemnina lemnina (Koch, 1841). No detailed study on Radfordia associated with Japanese Arvicolinae has been made subsequently, although some differences dependent on host-species have been noticed. Meanwhile, material on which to base a discussion of the systematics of Japanese arvicoline-infesting myobiids has been accumulating. In the present study we compare Japanese specimens with European and Taiwanese species, using the type specimens or those taken from respective type hosts. As a result, all the putative subspecies of R. (M.) lemnina distributed in Japan are proven to be distinct as described below.
SUBGENUS Microtimyobia FAIN AND LUKOSCHUS, 1976
In the family Myobiidae, only the genera Myobia von Heyden, 1826 and Radfordia, both associated with Rodentia, bear a three-segmented leg I that lacks a claw. The claw formula of legs ll-IV is 2-1-1 in Radfordia, and 1-1-1 in Myobia.
In Radfordia, the subgenus Microtimyobia is defined as follows (Fain and Lukoschus, 1976, 1977): coxal setae 3-2-0-0; setae vi not very long; setae ve thick and striated; one long and unstriated dorsal seta each on trochanters III and IV; anterior gnathosomal setae flat or membranous, especially in females; associated mainly with Arvicolinae.
It is significant from a coevolutionary point of view that this combination of remarkable characters is found only in the mites infesting an allied host group, the Arvicolinae.
Radfordia (Microtimyobia) lemnina (Koch, 1841)
AND ITS SUBSPECIES
Fain and Lukoschus (1977) redefined R. (M.) lemnina as follows, using measurements taken from the specimens parasitic on Microtus agrestis (Linnaeus, 1761), the type host:
Female: Body (idiosoma + gnathosoma) 1.6–1.92 times as long as wide; setae vi 5–9 μm wide; sc i slightly thicker than sc e; l1 subequal in width to d1, d2, and l2; l3 about 1.5 times as long as d3; coxal setae ex l1 short and less than 3 μm in width; ic2 long; ic3 (75–100 μm long) as long as ic2.
Male: Genital orifice on a level slightly posterior to vi, but distinctly anterior to the level of leg III; penis 110 μm long; sc i thin and short (30 μm long); d4 and l4 not very short (33 and 24 μm long, respectively); cx l1 short; ic2 long; ic3 and ic4 short.
The nominate subspecies and three others, R. (M.) lemnina clethrionomys Fain and Lukoschus, 1977, R. (M.) lemnina rutila Fain and Lukoschus, 1977, and R. (M.) lemnina micromys Fain and Lukoschus, 1976, have been described.
Although M. agrestis from Holland was designated as the type host, R. (M.) lemnina lemnina reportedly occurs on M. agrestis and the other 8 species of Microtus voles from Europe, China, Japan, Russia, and Alaska, as well as on Clethrionomys rufocanus (Sundevall, 1846) from Sweden, Clethrionomys rufocanus bedfordiae from Hokkaido, and Clethrionomys gapperi (Vigors, 1830) from the USA (Fain and Lukoschus, 1977).
Radfordia (M.) lemnina clethrionomys differs from the nominate subspecies in the female having some dorsal setae (vi, sc i, sc e, l1, d2, l2, d3, and l3) shorter than those of the latter and elongate membranous gnathosomal setae, and in the male's genital cone abruptly widening anteriorly (Figs. 14–15, cf. Fig. 9) (Fain and Lukoschus, 1977). We add here the thick male genital setae to the above subspecific characters (Figs. 14 and 15). This subspecies occurs on Clethrionomys glareolus (Schreber, 1780) in Belgium (type locality), Italy, and Holland (Fain and Lukoschus, 1977).
Radfordia (M.) lemnina rutila is separable from the above two subspecies principally by the male's genital cone that bears prominent, triangular, lateral projections anteriorly (Fig. 12) and the female's elongate and bifurcate membranous gnathosomal seta (Fain and Lukoschus, 1977). The type host is Clethrionomys rutilus (Pallas, 1779) from N. Sweden (Fain and Lukoschus, 1977).
Fain and Lukoschus (Fain and Lukoschus, 1977) relegated R. (M.) micromys Fain and Lukoschus, 1976, taken from Micromys minutus (Pallas, 1771) from Holland, to a subspecies of R. (M.) lemnina. The original description was based only on the female, which bears setae vi different in form from those of R. (M.) lemnina. Moreover, Micromys is exceptional as the host in that it belongs not to the Arvicolinae but to the Murinae. Therefore, it is necessary to reevaluate the systematic position of this subspecies, using specimens of both sexes.
As noted above, R. (M.) lemnina has been divided into subspecies based on the shape of the idiosomal and gnathosomal setae of the female and the form of the male genitalia. Uchikawa (1991), on the other hand, proposed a subspecies of Myobia kobayashii Uchikawa and Mizushima, 1975 parasitic on Taiwanese Apodemus draco (Barrett-Hamilton, 1900) on the basis of the setation of nymphal stages. Thus, it would be good to compare immature stages as well, to understand subtle differences at the subspecies level, although immature stages have usually been ignored in descriptions of Radfordia species and subspecies.
DESCRIPTION
Radfordia (Microtimyobia) lemnina hata ssp. n.
Radfordia (Microtimyobia) lemnina lemnina (Koch, 1841)-Fain and Lukoschus, 1977: 63.
Types: Ex Microtus montebelli (Milne-Edwards, 1872). Holotype female, allotype male, 1 female (F) and 2 male (M) paratypes, Okada, Matsumoto City, Nagano Prefecture, 1983. IV. 14; 28 paratypes (9 female tritonymphs, TN(F), 7 male tritonymphs, TN(M), 5 deutonymphs, DN, 5 protonymphs, PN, and 2 larvae, L), Kirigamine, Suwa City, Nagano Pref., 1983. VI. 3.
The holotype, allotype, and 5 paratypes (TN(F, M), DN, PN, and L) are deposited in National Science Museum (Tokyo) (registration nos. NSMT-Ac 10701–10707); 7 paratypes (TN(F, M), 2DN, 2PN, and L) in A. Fain's collection, Brussels; and 17 pratypes (M, 12TN(7F, 5M), 2DN, and 2PN) in K. Uchikawa's collection, Matsumoto.
Female (Figs. 1–2, 4): Outline of body (idiosoma + gnathosoma) rather stout, with length/width ratios 1.55–1.93. Idiosomal chaetotaxy as in Figs. 1 (dorsum) and 2 (venter); dorsal setae anterior to l3 long and well striated basally; ventral setae ic2 and ic3 long; cx l1–2 thickened and longer than ic1. Measurements as in Table 1. Legs and leg chaetotaxy as in Figs. 1 and 2. Anterior membranous gnathosomal seta apically split into 4 points (Fig. 4).
Figs. 1–2
Female of Radfordia (Microtimyobia) lemnina hata ssp. n. (1) Dorsal view. (2) Ventral view. Bar: 100 μm.
Figs. 3–6
Anterior gnathosomal setae of female (left) and male (right). (3) Radfordia (Microtimyobia) lemnina lemnina. (4) R. (M.) lemnina hata ssp. n. (5) R. (M.) lemnina japonica ssp. n. (6) R. (M.) lemnina mikado ssp. n. Bar: 50 μm.
Table 1
Measurements in micrometers of R. (M.) lemnina hata ssp. n. and R. (M.) lemnina lemnina
Male (Figs. 7–8, 10): Dorsal setation of idiosoma as in Fig. 7, quite different on hysterosoma from that of female; d1, minute and on genital cone, d2 lacking; l2 and d3 not paired, and each on median line; l3 and d5 lacking. Ventral setation as in female, but only ic2 long and ic3 short (Fig. 8). Legs as in female. Genital opening at level slightly posterior to bases of vi; penis slides on genital cone extending anteriorly beyond basal level of vi; genital cone wide, with shoulders, and bearing pair of thin genital setae and d1 (Fig. 10). Penis straight. Measurements as in Table 1; setae ve, sc i, and l1 possibly slightly longer than those of the nominate subspecies. Anterior gnathosomal seta narrow and terminating in 2 points (Fig. 4).
Figs. 7–8
Male of Radfordia (Microtimyobia) lemnina hata ssp. n, (7) Idiosomal dorsum. (8) Idiosomal venter. Bar: 100 μm.
Figs. 9–16
Male genital regions. (9) Radfordia (Microtimyobia) lemnina lemnina. (10) R. (M.) lemnina hata ssp. n. (11) R. (M.) lemnina japonica ssp. n. (12) R. (M.) lemnina rutila. (13) R. (M.) lemnina mikado ssp. n. (14–15) R. (M.) lemnina clethrionomys. (16) R. (M.) eothenomys. Bar: 50 μm.
Immature stages (Figs. 17–20, 22): Consisting of larva, protonymph, deutonymph, and tritonymph. Body size variable according to developmental phase within each stage (Table 1), but ontogenetic changes discernible in idiosomal chaetotaxy and development of legs ll-IV.
Figs. 17–20
Idiosomal dorsum of immature stages of Radfordia (Microtimyobia) lemnina hata ssp. n. (17) Larva. (18) Protonymph. (19) Deutonymph. (20) Male tritonymph. Bar: 50 μm.
Figs. 21–24
Idiosomal dorsum (left) and venter (right) of the female tritonymphs at the same scale. (21) Radfordia (Microtimyobia) lemnina lemnina. (22) R. (M.) lemnina hata ssp. n. (specimen with the longest ic3). (23) R. (M.) lemnina japonica ssp. n. (24) R. (M.) lemnina mikado ssp. n. Bar: 100 μm.
Idiosomal chaetotaxy. Four pairs of propodosomal setae (vi, ve, sc i, and sc e) present in all immature stages, and pair of hysterosomal setae added at each stage; accordingly, hysterosomal setae in four stages from larva to tritonymph numbering 6 (d4, d5, and l4 lacking), 7 (d5 and l4 lacking), 8 (4 lacking), and 9 pairs (as in adult female), respectively (Figs. 17–20). Nature of all setae of larva different from that of more advanced stages (Fig. 17); each seta gaining in length in accordance with progress of stages (Table 1). Ventral setae also variable in number according to stage; intercoxal setae (ic series) from larva to tritonymph 1-0-0, 1-1-1-0, 1-1-1-1, and 1-1-1-1, respectively, with ic2 being long in nymphal stages; coxal setae (cx series) 0-0-0, 1-0-0-0, 1-0-0-0, and 2-1-0-0, respectively, with anterolateral cx I being shell-like in form (Figs. 21–24, right). Sexual dimorphism apparent only in tritonymph; intercoxal setae iC3 long, often with bilateral asymmetry, in female tritonymph (Fig. 14), but these setae short in male tritonymph (Table 1). Many dorsal setae of idiosoma distinctly longer in female tritonymph than in male tritonymph (Table 1).
Legs ll-IV: Fully developed leg of immature stages with four free segments (trochanter, femuro-genu, tibia, and tarsus) and single, strong terminal claw. Larva bearing complete legs II and primodia of legs III; protonymph with complete legs II and small, 3-segmented and sparsely setose legs III lacking terminal claws; deutonymph with complete legs II, 4-segmented legs III with no claws, and primodia of legs IV; and tritonymph having complete legs II and III, and 4-segmented legs IV without claws.
Notes: The immature stages of no Radfordia (Microtimyobia) species have been studied until now, so their specific characters can not yet be adduced. In the present study, a single female tritonymph of R. (M.) lemnina lemnina and all the immature stages of the other three subspecies were available. Among these specimens, the setae ve, sc i, sc e, d1, d2, and l2 seem to be longer, while setae l4 and ic3 are shorter, in the female tritonymph of R. (M) lemnina hata than in that of R. (M.) lemnina lemnina (Table 1). The relatively long setae d1, d2, and l2, the relatively short l4, and the very short ic3 are characteristic of this stage of R. (M.) lemnina hata. As for other stages, the setae sc e of the protonymph of R. (M.) lemnina hata are probably shorter than those of the other subspecies (Tables 1Table 2Table 3–4).
Table 2
Measurements in micrometers of R. (M.) lemnina japonica ssp. n. taken from C. rufocanus bedfordiae
Table 3
Measurements in micrometers of adults and female tritonymphas of R. (M.) lemnina japonica ssp. n. from diverse host voles from various localities
Table 4
Measurements in micrometers of R. (M.) lemnina mikado ssp. n. parasitic on C. rutilus mikado
Material examined: In addition to the types, the specimens listed below were taken from Microtus montebelli (type host): 2F and 1M, Kirigamine, Suwa City, Nagano Pref., 1983. VI. 6; 3F, 1TN(F), 1PN, and 2L, Tokusawa, Azumi Village, Nagano Pref., 1989. VII. 27; 1F, Kamikochi, Azumi Village, Nagano Pref., 1972. VIII. 4; 2F, Hakuba Village, Nagano Pref., 1971. VII; 1F, 2M, 4TN(F), 2TN(M), and 2DN, Gero Town, Gifu Pref., 1955. II. 26; 2F, 1M, and 1TN (M), Abiko, Chiba Pref., 1995; 1F, Fukuoka Pref., 1981. VII. 14; 1M, Fukuoka Pref., 1972. VII. 1, identified as R. (M.) lemnina lemnina by Fain (Fig. 10, left).
Comparative material:
Radfordia (Microtimyobia) lemnina lemnina ex Microtus agrestis (type host): 1F, 1M, and 1TN(F), Heer Agimont, Belgium, 1967. IX; 2F, Hatert, Holland, 1968. X. 22 and 1972. IX. 5, respectively; 1F, Hamert, Holland, 1972. VII. 9; 1M, Switzerland, ? date.
R. (M.) lemnina lemnina ex Clethrionomys glareolus: 1M, Carreglefn, Anglesey, N. Wales, 1980. VII. 22 (host record to be reconfirmed).
Differential diagnosis: It is very difficult to discriminate the females of R. (M.) lemnina sspp. from one another (Fain and Lukoschus, 1977). The shape of the anterior membranous gnathosomal setae sometimes differs among subspecies. The setae of R. (M.) lemnina hata ssp. n. are apically split into four or more points (Fig. 4), and they differ in shape from those of the female of the nominate subspecies, which bear three points (Fig. 3). However, the membranous setae are variable in form among individuals and, sometimes, even bilaterally on a single specimen. Moreover, the exact shape of these setae is usually difficult to observe. Accordingly, it is difficult to separate subspecies from each other based on subtle differences in these setae alone.
The structure of the male genital region is a more useful character than the gnathosomal setae for differentiating subspecies of R. (M.) lemnina. The male genital regions of European specimens parasitic on M. agrestis and the Japanese specimen from M. montebelli that was identified as R. (M.) lemnina lemnina by Fain are shown in Figs. 9 and 10 (left), respectively. It is difficult to find any distinct difference, although the genital setae are slightly longer on the Japanese specimen. Accordingly, Fain and Lukoschus (Fain and Lukoschus, 1977) had good reason to relegate the Radfordia mite parasitic on M. montebelli to R. (M.) lemnina lemnina. However, all the male specimens examined in the present study, exclusive of Fain's specimen, bear a genital cone as drawn in Fig. 10 (right), anteriorly broad and with shoulders and with genital setae longer than those of the nominate subspecies. No other distinctive characters and measurements that warrant subspecies rank have been found in the male.
In the female tritonymph the third intercoxal setae, ic3, are unusually short in all the specimens taken from M. montebelli (Fig. 22, Table 1), compared with those of the nominate subspecies (Fig. 21). With this character, R. (M.) lemnina hata is clearly defined, which confirms that the above subtle differences in the male genital region are significant for separating subspecies of R. (M.) lemnina. Further, some dorsal setae of the idiosoma (ve, sc i, sc e, d1, d2, and l2) are possibly longer in the female tritonymph of R. (M.) lemnina hata than in that of the nominate subspecies, and vice versa for the setae l4 (Figs. 21 and 22, Table 1).
Remarks: As shown above, some immature stages bear more obvious specific or subspecific characters than the adult stage does. In the subgenus Microtimyobia, all the systematic studies so far have been based on the morphology of the adult stage, ignoring immature stages. More attention should be paid to the latter, which are usually more abundant on any host animal than adults.
Etymology: The subspecies name, hata, is a noun in apposition. In Japanese it means cultivated field, one of the preferred habitats for M. montebelli (Japanese name, hatanezumi).
Radfordia (Microtimyobia) lemnina japonica ssp. n.
Radfordia (Microtimyobia) lemnina lemnina (Koch, 1841)-Fain and Lukoschus, 1977: 63.
Types: Ex Clethrionomys rufocanus bedfordiae. Holotype F, allotype M, 48 paratypes (6 each of F, M, TN(F), TN(M with long setae), TN(M with short setae), DN, PN, and L), Tonden, near Sapporo, Hokkaido, 1983. IX. 28–29.
The holotype, allotype, and 8 paratypes (TN(F), 1 of each form of TN(M), DN, 2PN, and 2L) are deposited in National Science Museum (Tokyo) (regestration nos. NSMT-Ac 10708–10713); 10 paratypes (F, M, TN(F), 1 of each form of TN(M), DN, 2PN, and 2L) in A. Fain's collection, Brussels; and 30 paratypes (5F, 5M, 4TN(F), 4 of each form of TN(M), 4DN, 2PN, and 2L) in K. Uchikawa's collection, Matsumoto.
Female (Fig. 5): General shape essentially as in preceding and nominate subspecies (cf. Figs. 1 and 2). Measurements as in Table 2. Idiosomal setae sc i slightly shorter than those of preceding subspecies, but ranges of setal length of both forms overlapping. Anterior gnathosomal setae relatively large, and split apically into numerous points (Fig. 5).
Male (Figs. 5 and 11): General shape as in preceding subspecies (cf. Figs. 7 and 8). Measurements as in Table 2. Setae l2 and ic4 longer than those of preceding subspecies. Genital region as in Fig. 11; genital cone narrow and situated posteriad of basal level of setae vi, penis slides not extending anteriorly beyond basal line of these setae. Anterior gnathosomal setae wide and split apically into at least 3 points (Fig. 5).
Immature stages: Formation and setation of idiosoma, gnathosoma, and legs as in R. (M.) lemnina hata. Characteristic properties discernible in each developmental stage other than larva when compared with corresponding stage of R. (M.) lemnina hata.
Female tritonymph (Fig. 23): Dorsal setae sc i, d1, d2, and l2 shorter than those of R. (M.) lemnina hata, and vice versa for l4 and ic3; ic3 distinctly thicker than ic2 and extraordinary long (100–150 μm) (Tables 1 and 2).
Male tritonymph: Lengths of idiosomal setae showing vast variation. When examined under the microscope, setae l1, d4, l3, and l4 apparently shorter on some specimens than others as if different kinds of male tritonymphs were mixed. Of 35 specimens examined, 12 possessing short setae, other 23 bearing long setae. Measurements of the two groups, TN (M, a) and TN (M, b), presented separately in Table 2. Among those with long setae, l1, d4, l3, l4, and ic2 distinctly longer than those of male tritonymph of preceding subspecies.
Deutonymph: Idiosomal setae d1, l2, and d4 slightly shorter, but ic2 longer, than those of R. (M.) lemnina hata (Tables 1 and 2).
Protonymph: Idiosomal setae d4 shorter, but ic2 longer, than those of R. (M.) lemnina hata (Tables 1 and 2).
Larva: As in R. (M.) lemnina hata (Tables 1 and 2).
Material examined: In addition to the types, many specimens were taken from arvicoline voles in Japan.
Ex Clethrionomys rufocanus bedfordiae (type host): 50F, 15M, 24TN(F), 23TN(M), 61DN, 18PN, and 3L, Tonden, near Sapporo, Hokkaido, 1983. IX. 28–29; 1M (identified as R. (M.) lemnina lemnina by Fain), Nopporo, Hokkaido, 1956. V. 30.
Ex Clethrionomys rex lmaizumi, 1971: 9F, 2M, 2TN(M), Kitakawaguchi, Teshio Town, Teshio-Gun, Hokkaido, 1989. X. 12–13.
Ex Eothenomys andersoni (Thomas, 1905): 1F, 1M, and 3TN(F), Takehara, Yamabe Town, Higashimurayama-Gun, Yamagata Pref., 1982. V-VI; 1F, Donko-Numa, Zaoh, Yamagata Pref., 1965. VII. 24; 1F, Mt. Iwaki, near Hirosaki City, Aomori Pref., 1978. VI. 27; 1F, and 1TN(F), Ohdate, Hakkoda Mts., Akita Pref., 1978. VII. 14; 13F, Tago, Nagano City, Nagano Pref., 1975. I.14; 7F, 1M, 1TN(F), 1YN(M), 1DN, and 1PN, Mt. lizuna, near Nagano City, Nagano Pref., 1975. VIII; 11F, and 7M, Asakawa, Nagano City, Nagano Pref., 1975. I. 11; 1M, Asahi-Yama, Nagano City, Nagano Pref., 1975. I. 13; 3F and 1M, Happo-One, Hakuba Village, Nagano Pref., 1972. X. 29; 4F, 10M, and 2TN(F), Kamikochi, Azumi Village, Nagano Pref., 1988. V. 21.
Ex Eothenomys smithii (Thomas, 1905): 2F, 2M, 9TN(F), 3TN(M), 7DN, and 1PN, Tashirohara, Kunimi-Cho, Shimabara City, Nagasaki Pref., 1995. IV. 12; 4M, 4TN(F), 1DN, Ohsaka Pass, Hikita Town, Kagawa Pref., 1958. XI. 25; 1M, Sai-ko (Lake), Yamanashi Pref., 1969. VII. 3; 2F, Ohtaga, Asuke Town, Aichi Pref., 1982. XII. 22; 1F, Iwakami, Asuke Town, Aichi Pref., 1983. II. 24; 2F, Goshomitsu, Inabu Town, Aichi Pref., 1983. III. 10.
Comparative material:
Radfordia (Microtimyobia) lemnina clethrionomys ex Clethrionomys glareolus: 1 M, Utrecht, Holland, 1975. VI. 25; 1M, Hamert, Holland, 1972. VII. 4; 1F, Pescasseroli, Holland, 1972. X. 16; 1F, Nijmegen, Holland, 1966. V. 2.
R. (M.) eothenomys Fain and Lukoschus, 1976 ex Eothenomys sp.: holotype F and allotype M, Chuei-feng, Taiwan (specimens of the National Museum of Natural History, Washington, deposited in the USDA, Beltsville).
Differential diagnosis: The female of R. (M.) lemnina japonica ssp. n. bears anterior gnathosomal setae that are larger and have more apical points than those of the nominate and preceding subspecies. However, this alone is not definitive, since the shape of these setae is variable, as noted above. The male gnathosomal setae are wide and split into 3 points, distinctly different from those of the subspecies R. (M.) lemnina lemnina and R. (M.) lemnina hata, both of which bear narrow, bifurcate setae (Fig. 5, cf. Fig. 4). The male genital cone is situated relatively more posteriad than those of these two subspecies, and the cone itself is narrow anteriorly. In some subspecies, as mentioned below, the anterior part of the genital cone is expanded laterally, which is a modification opposed to that of R. (M.) lemnina japonica. Accordingly, the position and form of the genital cone, together with the shape of the anterior gnathosomal setae, confirm that the male of R. (M.) lemnina japonica is distinct, although Fain and Lukoschus (Fain and Lukoschus, 1977) once identified it with the nominate subspecies.
In comparison to R. (M.) lemnina hata, the female tritonymph of R. (M.) lemnina japonica bears intercoxal setae ic3 that are long enough to warrant recognizing the form as distinct. Measurements of some other dorsal setae (sci, d2, l2, and l4) might also be useful to characterize the female tritonymph of the present new subspecies.
Remarks: Radfordia specimens found on Clethrionomys rex, Eothenomys andersoni, and Eothenomys smithii from various localities were all identified with R. (M.) lemnina japonica (Table 3). There are disturbing differences in the lengths of some setae on specimens of different origin (Table 3); however, it is reasonable to regard all such differences as intrasubspecific variation. With respect to this subspecies' wide host range, it is interesting to note that European Clethrionomys rufocanus is currently recorded as a host of R. (M.) lemnina lemnina, and that Taiwanese Eothenomys sp. is the host of R. (M.) eothenomys Fain and Lukoschus, 1976, which is different from R. (M.) lemnina at the species-level (Fig. 16). The problem of host relationships of R. (M.) lemnina sspp. and its allies is discussed further below.
Etymology: The subspecies name, japonica, is adopted to indicate that the mite is distributed throughout Japan in the range of arvicoline voles.
Radfordia (Microtimyobia) lemnina mikado ssp. n.
Types: Ex Clethrionomys rutilus mikado Thomas, 1905. Holotype F, allotype M, 41 paratypes (6 F, 6M, 6TN(F), 6TN(M), 6DN, 6PN, and 5L), Tonden, near Sapporo, Hokkaido, 1983. IX. 28–29.
The holotype, allotype, and 7 paratypes (TN(F), TN(M), DN, 2PN, and 2L) are deposited in National Science Museum (Tokyo) (regestration nos. 10714–10718); 9 paratypes (F, M, TN(F), TN(M), DN, 2PN, and 2L) in A. Fain's collection, Brussels; and 25 paratypes (5F, 5M, 4TN(F), 4TN(M), 4DN, 2PN, and 1L) in K. Uchikawa's collection, Matsumoto.
Female (Fig. 6): General shape essentially as in R. (M.) lemnina hata (cf. Figs. 1 and 2). Measurements as in Table 4. Anterior gnathosomal setae rather small and split apically into at least 3 points (Fig. 6).
Male (Figs. 6 and 13): General shape essentially as in R. (M.) lemnina hata (cf. Figs. 1 and 2). Measurements as in Table 4. Genital region as in Fig. 13; genital opening at level of bases of vi; genital cone with wide lateral projections anteriorly (Fig. 13); genital setae thick and conspicuous. Anterior gnathosomal setae narrow and bifurcate (Fig. 6).
Immature stages: General structure and setation similar to those of corresponding stages of R. (M.) lemnina hata. Measurements as in Table 4.
Female tritonymph: Ventral setae ic3 63–90 μm long, probably as long as those of nominate subspecies; l4 distinctly shorter than those of R. (M.) lemnina japonica (Tables 1, 2, and 4).
Male tritonymph: Lengths of setae on idiosomal dorsum not so variable as to separate individuals into groups, and ranging between those of short and long setae of male tritonymphs of R. (M.) lemnina japonica (Table 4).
Deuto- and protonymphs and larva: Similar to those of R. (M.) lemnina hata.
Material examined: In addition to the types, the following specimens were taken from Clethrionomys rutilus mikado (type host): 25F, 17M, 49TN(F), 39TN(M), 54DN, and 18PN, Tonden, near Sapporo, Hokkaido, 1983. IX. 28–29.
Conparative material:
Radfordia (Microtimyobia) lemnina rutila ex Clethrionomys rutilus: 2 paratypes (F and M) and 1M, Bjoerkliden, N Sweden, 1966. VIII. 11 (specimens in A. Fain's collection).
Differential diagnosis: The female of R. (M.) lemnina rutila bears anterior gnathosomal setae that are elongate and bifurcate (Fain and Lukoschus, 1977). In R. (M.) lemnina mikado ssp. n., the setae are split apically into at least 3 points (Fig. 6). However, as already noted, it is difficult to separate females at the subspecies level based on this character alone.
In the male of the present new subspecies, the anterolateral expansions of the genital cone are broad and triangular and the genital setae are thick (Fig. 13), compared with those of R. (M.) lemnina rutila (Fig. 12). In both features R. (M) lemnina mikado is intermediate between R. (M.) lemnina rutila and R. (M.) lemnina clethrionomys (Figs. 14 and 15). This level of difference in the genital region is thought to be enough to separate subspecies, as was noted for each of the preceding two new subspecies.
Some of the immature stages might be expected to have characters that corroborate the validity of the new subspecies. In the female tritonymph, R. (M.) lemnina mikado is easily separable from the other Japanese subspecies, R. (M.) lemnina hata and R. (M.) lemnina japonica, by the remarkable difference in the length of ic3, but not from R. (M.) lemnina lemnina, which probably bears setae ic3 of a similar length. However, setae ve and sc e are longer, and l4 is distinctly shorter, in R. (M.) lemnina mikado than in R. (M.) lemnina lemnina. The female tritonymph of R. (M.) lemnina mikado is thus separable from those of the nominate and two other Japanese subspecies. However, as suggested above by the shape of the genital cone, R. (M.) lemnina mikado might be more allied to R. (M.) lemnina rutila and R. (M.) lemnina clethrionomys than to the three subspecies just mentioned. The immature stages of the European subspecies have not been described yet, and are not available for comparison.
Etymology: The subspecies name, mikado, a noun in apposition, comes directly from the name of the host vole, C. rutilus mikado.
DISCUSSION
Immature stages:
Because Radfordia and Myobia mites share Mus musculus as their host, studies of immature stages of these two genera have been made repeatedly (summarized by Paran, 1995). However, the developmental stages of these two genera have often not been properly separated. Although some authors have shown only proto- and deutonymphs as the nymphal stages of Radfordia and Myobia, three nymphal stages are characteristic of the family Myobiidae.
Some immature stages have features that distingush subspecies within the genera Radfordia (Microtimyobia), as shown above, and Myobia (Uchikawa, 1991). Moreover, each larval or nymphal stage may also exhibit genus-level characters, although the definitions of genera have been made on the adult stage. The ontogenetic changes in legs ll-IV described above for three Japanese subspecies of Radfordia (Microtimyobia) lemnina are similar to those of Radfordia (Radfordia) affinis (Poppe, 1896) (Matuzaki, 1961; Paran, 1995), R. (R.) ensifera (Poppe, 1896) Matuzaki, 1964), and R. (R.) daltoni Scheperboer, Fain and Lukoschus (1987), but differ distinctly from those of Myobia musculi (Schrank, 1781) (Grant, 1942; Matsuzaki, 1961; Matuzaki, 1970), M. agraria Gorissen and Lukoschus (1982), M. apodemi Uchikawa (1973), M. nodae Matuzaki, 1965 (Matuzaki, 1965; Uchikawa et al., 1988), and M. kobayashii sspp. Uchikawa, 1991; Uchikawa et al., 1988). Legs IV are incomplete even in the tritonymph of Radfordia species, while legs of Myobia species become well developed in earlier immature stages. This is probably a genus-level ontogenetic character.
On the other hand, the dorsal chaetotaxy of the idiosoma of immature stages of Radfordia (Microtimyobia) represented by the above three Japanese subspecies is different from those of Radfordia (Radfordia) (Matuzaki, 1964; Paran, 1995; Scheperboer et al., 1987), and Myobia (Matuzaki, 1961, 1964; Paran, 1995; Uchikawa, 1991; Uchikawa et al., 1988). In Radfordia (Radfordia), the setae sc e are extraordinally long, and many nymphal hysterosomal setae have a bulbar expansion (cf. Figs. 18–20) (Matuzaki, 1964; Paran, 1995; Scheperboer et al., 1987). The number of hysterosomal setae seems to be different in Radfordia (Radfordia) and Radfordia (Microtimyobia) (Matuzaki, 1964; Paran, 1995; Scheperboer et al., 1987). The deuto- and tritonymphs of Myobia bear only 7 pairs of hysterosomal setae (Paran, 1995; Uchikawa, 1991; Uchikawa et al., 1988) instead of 8 and 9 pairs, respectively, in these nymphal stages of Radfordia (Microtimyobia). The one (d5) or respectively two (d5 and l4) missing pairs of setae of Myobia are only discernible as small circles or dots. Accordingly, the difference in the immature dorsal chaetotaxy may be subgenus-level as well as genus-level, but this requires confirmation based on more material.
The ventral setae of the cx series vary in number among the subgenera of Radfordia (Fain and Lukoschus, 1977). The ontogenetic development of the cx series of Radfordia (Microtimyobia), which advances more rapidly than in Myobia species (Paran, 1995; Uchikawa, 1991; Uchikawa et al., 1988), should be compared with those of the 9 other subgenera of Radfordia to understand the level at which this character is important.
As documented above, the characters of immature stages are sometimes more reliable for defining species or subspecies of Radfordia (and also Myobia as in Uchikawa, 1991) than those of adults that closely resemble each other. To adopt immature stages in the classification of any myobiid mite, it is necessary first to distinguish the intra- and intergeneric differences clearly.
Host relations:
Among the 18 known families of symbiotic Prostigmata, only the Myobiidae, Psorergatidae, and Demodicidae have specifically coevolved with the Mammalia (Nutting, 1985), but the known Myobiidae were formerly thought not to be host-species specific with their cosymbiotic mammals (Nutting, 1985). However, it has become evident that there are many monoxenic or host-species specific species as well as a few polyxenic species in the Myobiidae, and that both kinds of species serve as indicators suggesting mammalian phylogeny (Uchikawa, 1988). Since the monoxenic myobiid species are larger, elongate forms and polyxenic ones are smaller and rounded (Uchikawa, 1988), Radfordia and Myobia, being elongate and rather large, are probably host-species specific. Actually, the 69 species and subspecies of Radfordia available for Fain and Lukoschus (1977) comprise 49 monoxenic, 13 dixenic, 4 trixenic, 1 tetraxenic, and 1 hexaxenic species or subspecies and the exceptional R. (M.) lemnina lemnina, which has been recorded from 11 species of arvicoline voles (see Addendum) (Bochkov, 1994, 1995; Dusbábek, 1988; Fain and Lukoschus, 1977; Fain et al., 1979). Moreover, of the 17 congeneric taxa described thereafter, 16 are monoxenic and one is dixenic (Bochkov, 1994, 1995; Fain and Lukoschus, 1979a, b; Fain et al., 1979; Gill and Strandtmann, 1977; Lukoschus et al., 1981; Scheperboer et al., 1987; Vesmanis and Lukoschus, 1978). These data indicate that the majority of Radfordia species are host-species specific, and that the few oligoxenic species are shared by closely allied rodents. Accordingly, the host relations of the present oligo- and polyxenic species or subspecies should be thoroughly reevaluated.
When all the myobiids associated with Japanese arvicoline voles were identified as R. lemnina, it was reasonable to think that specialization in Radfordia has been retarded compared with myobiids of many other genera. Conversely, one could think that voles became fully established in Japan relatively late, thereby not allowing their parasites time to speciate enough to become species-specific. However, Fain and Lukoschus (Fain and Lukoschus, 1977) separated R. (M.) lemnina into several subspecies, and they recorded voles of the genera Microtus and Clethrionomys distributed over a wide geographic range (including M. montebelli and C. rufocanus bedfordiae from Japan) as the hosts of R. (M.) lemnina lemnina, while European C. glareolus and C. rutilus each yielded a different subspecies of R. (M.) lemnina.
Among all the subspecies, only R. (M.) lemnina lemnina is purportedly polyxenic as mentioned above, infesting the two host genera Microtus and Clethrionomys over a wide geographic range. There is thus little cause for surprise, based on the host relations of other myobiids, that R. (M.) lemnina sspp. parasitic on M, montebelli and C. rufocanus bedfordiae have proven to be separable from the nominate subspecies. Although we could not examine any specimens taken from European C. rufocanus, it is reasonable to suspect that this vole may be associated with an unrecognized subspecies of R. (M.) lemnina more similar to R. (M.) lemnina japonica than to the nominate subspecies. Accordingly, the present study suggests that specimens supposedly of the nominate subspecies taken from arvicoline voles other than European M. agrestis will be separable into several distinct subspecies by more detailed studies.
The presently understood host relations of R. (M.) lemnina japonica are arousing the same problem as those of the nominate subspecies. The host voles are Clethrionomys rufocanus bedfordiae, C. rex, Eothenomys andersoni, and E. smithii. Since the former two voles are thought to be allied to or, sometimes, synonymous with each other (Musser and Carleton, 1993), it is plausible that these hosts share the same subspecies of mite. On the other hand, it is inexplicable that Clethrionomys and Eothenomys should be infested with the same subspecies of R. (M.) lemnina. Since transfer of myobiid mites to new hosts can only have occurred with difficulty, and since the list of currently known hosts of R. (M.) lemnina lemnina including Microtus and Clethrionomys may be subject to revision as discussed above, there is no reason to expect that the two vole genera should share a subspecies of R. (M.) lemnina. Meanwhile, a Taiwanese Eothenomys vole is known to be the host of Radfordia (Microtimyobia) eothenomys Fain and Lukoschus, 1976 (Fain and Lukoschus, 1977). Accordingly, it is expected on the basis of all the other myobiid data that Japanese Eothenomys is probably different at the genus level from Taiwanese Eothenomys, although all are currently thought by Japanese mammalogists to be congeneric. Concerning this problem, Musser and Carleton (1993) reinstated Phaulomys for Japanese Eothenomys andersoni and E. smithii, in accordance with earlier and recent works of Japanese mammalogists, and suggested that Phaulomys and Clethrionomys are derived from a common ancestor. The present myobiid data give support to Musser and Carleton (1993). However, the genus Phaulomys has not gained acceptance in Japan. Renewed studies using all available methods on the myobiids themselves and their hosts are required to interpret the problem of Phaulomys.
The fact that European Clethrionomys rutilus and Japanese C. rutilus mikado harbor similar but distinct subspecies of R. (M.) lemnina is also unusual, since all subspecies of a host species are usually associated with the same myobiid species, which is not separable further. We think that Japanese C. rutilus on the island of Hokkaido has been isolated geographically apart from the European form for long enough to form a subspecies and for its myobiid to speciate. Based on these considerations, we accepted C. rutilus mikado as the host name. The same should be true for the relation between European C. rufocanus and Japanese C. rufocanus bedfordiae, which is one more reason to expect an unrecognized subspecies of R. (M.) lemnina parasitic on European C. rufocanus.
Acknowledgments
Dr. A. Fain (Institut Royal des Sciences Naturelles de Belgique, Brussels) sent us on loan invaluable specimens taken from European Microtus agrestis and Clethrionomys spp. The late Dr. F. S. Lukoschus, while still alive, donated three specimens of R. (M.) lemnina clethrionomys to KU. Dr. Robert L. Smiley (USDA, Beltsville) showed us the holotype and allotype of R. (M.) lemnina eothenomys. Dr. Hiroshi Suzuki (Nagasaki University, Nagasaki), Dr. Yukibumi Kaneko (Kagawa University, Takamatsu), and Dr. Hidenori Endo (National Science Museum, Tokyo) lent us vole specimens that harbored some of the mites recorded above. Cooperation of all these acarologists and mammalogists is deeply appreciated.
References
Appendices
Addendum:
After submitting the manuscript, we became aware of a recent paper by Bochkov (1995) dealing with the 9 taxa of Radfordia (Microtimyobia) associated with the Arvicolinae in Russia and surrounding countries. Bochkov (1995) raised R. (M.) lemnina clethrionomys Fain and Lukoshus to full species rank and reduced ft (M.) lemnina rutila Fain and Lukoschus to a subspecies of R. (M.) clethrionomys, because of the structure of the male genital cone (cf. Figs. 9,12–15). However, the females of ft. (M.) lemnina sspp. and ft (M.) clethrionomys are not separable from each other in Bochkov's key to species and subspecies. The female characters, which are generally thought to be more conservative in evolution than the male characters, are essential for separating species, as was shown by Fain and Lukoschus (1977). Radfordia (M.) rufocani Bochkov, 1995 was described as a parasite of Clethrionomys rufocanus (Sundevall) from Mt. Kykschik, Bashkhiria, Russia. Although this mite is dealt with by Bochkov (1995) as a full species, it corresponds to our unrecognized subspecies of ft. (M.) lemnina that has been expected above to be closer to ft. (M.) lemnina japonica than to ft. (M.) lemnina lemnina. The male of ft. (M.) rufocani has a narrow, subapicaly constricted genital cone, genital setae thinner than those of ft. (M.) lemnina lemnina, and membranous gnathosomal setae with only two apical points (Bochkov, 1995). By these characters, ft. (M.) rufocani is clearly separable from ft. (M.) lemnina japonica. The unique shape of the male genital cone of ft. (M.) rufocani suggests that this form is closer to ft. (M.) lemnina clethrionomys and R. (M.) lemnina rutila than to ft. (M.) lemnina japonica, contrary to our expectation. However, further comparison of ft. (M.) rufocani and R. (M.) lemnina japonica is necessary, using all the characters described in the present paper, for the clarification of the phylogenetic relationship between the two forms.
Bochkov (1995) tabulated the host species, localities and authors for ft (M.) lemnina (not ft. (M.) lemnina lemnina), adding 7 Microtus species and 3 Chionomys species to and excluding all the Clethrionomys species from the table of Fain and Lukoschus (1977). This indicates that detailed morphological and coevolutional studies of ft. (M.) lemnina or R. (M.) lemnina lemnina can not be done based on adult characters alone.
