Study of types of some species of “Filaria” (Nematoda) parasites of small mammals described by von Linstow and Molin

Parasitic nematodes from the Berlin (ZMB) and Vienna (NMW) Museum collections referred to the genus Filaria Mueller, 1787 by von Linstow or Molin were studied. Three samples were in good condition and the specimens redescribed. Litomosa hepatica (von Linstow, 1897) n. comb., sample ZMB Vermes Entozoa 3368, from the megachiropteran Pteropus neohibernicus, Bismarck Archipelago, resembles L. maki Tibayrenc, Bain & Ramanchandran, 1979, from Pteropus vampyrus, in Malaysia, but the buccal capsule differs. Both species display particular morphological characters which differ from species of Litomosa parasitic in microchiropterans. The remaining material originates from Brazil. The spicule morphology of Litomosoides circularis (von Linstow, 1899) Chandler, 1931, sample ZMB Vermes Entozoa 1059 from Hesperomys spec. (= Holochilus brasiliensis), Porto Alegre, confirms that it belongs to the sigmodontis group; the microfilaria presents characters of the genus Litomosoides, e.g. body attenuated at both extremities and salient cephalic hook. Taxonomic discussions by others confirm that species of Litomosoides belonging to the sigmodontis group and described subsequently are distinct from L. circularis. Litomosoides serpicula (Molin, 1858) Guerrero, Martin, Gardner & Bain, 2002, is redescribed, sample NMW 6323 from the bat Phyllostoma spiculatum (= Sturnira lilium), Ypanema. It is very close to L. brasiliensis Almeida, 1936, type host Moytis sp., but distinguished by a single ring in the buccal capsule, rather than two, supporting previous conclusions that the taxon L. brasiliensis, as generally regarded, may represent a complex of species. Samples NMW 6322 and NMW 6324, from other bats and also identified by Molin (1858) as Filaria serpicula, contain unidentifiable fragments of Litomosoides incertae sedis. Filaria hyalina von Linstow, 1890, sample ZMB Vermes Entozoa Q 3905 from Sorex vulgaris (= Sorex araneus), is incertae sedis because it contains two unidentifiable posterior parts of male, which might be an acuarid, Stammerinema sp. Filaria vesperuginis von Linstow, 1885, sample ZMB Vermes Entozoa Q 3929, from the bat Vesperugo serotinus (= Eptesicus serotinus), contains encysted nematode larvae and is a nomen dubium.


INTRODUCTION
F ilaria Mueller, 1787 was used generally for thread-like nematode parasites found in body cavities of hosts (Stiles & Hassal, 1922). The genus was later split into different genera distinguished by characters which often had not been considered in the original descriptions (Anderson & Bain, 2009). Analyzing ancient type material is useful to clarify Original contribution Parasite, 2011, 18, 151-161 their taxonomic positions. We studied here types of five species of "Filaria" (seven samples) parasitic in small mammals and described by von Linstow (1885,1890,1897,1899) and Molin (1858). None have been re-observed since, although two have previously been allocated to a more recent filarioid genus, Litomosoides Chandler, 1931(Chandler, 1931Guerrero et al., 2002).

MATERIAL AND METHODS
V on Linstow's types are from the Collection Vermes of the Museum für Naturkunde der Humboldt-Universität in Berlin (ZMB) and those of Molin are preserved in the Collection of Evertebrata Varia at the Naturhistorisches Museum Wien (NMW). Morphology and terminology of filarioids follow Bain (1966) and Guerrero et al. (2002). Specimens were cleared in lactophenol and drawn with a microscope equipped with a camera lucida. Particular attention was given to head papillae (the inner circle corresponds to the external labial papillae, and the external circle to the cephalic papillae), buccal capsule and area rugosa of the posterior region of the male. When possible, the male tail was placed in ventral view to assess the number and arrangement of caudal papillae: for a given pair, papillae are placed on a transverse line or they are more or less aligned on the mid-ventral line. If sufficient females were available, microfilariae were extracted from their uteri. ND indicates that the particular measurement was not determined for that specimen. Measurements were taken from drawings and are in micrometres, unless otherwise stated. When hololectotype and allolectotype were designated their measurements appear first, followed by the measurements of paralectotypes in parentheses. Regarding the hosts, the original name is given first followed by present name which is consistent with Wilson & Reeder (2005). The identification of the typehost of Filaria circularis von Linstow, 1899 presented a particular problem which was solved with the collaboration of the curator of the mammalian collection in Berlin. The type-host specimen, nro 1050, from Porto Alegre, had been identified by Hensel (1872) as Hesperomys? sp. Species of Hesperomys are at present contained in several genera (Musser et al., 1998) and, ealier, Hershkovitz (1955 had transferred specimens of Hesperomys from Porto Alegre to Holochilus. The type-host specimen itself is not in the Berlin collection but the eight "Hesperomys" specimens collected by Hensel in Porto Alegre and stored in the Berlin collection were identified to Holochilus brasiliensis (Desmarets, 1819) by Alfred L. Gardner when he revised the collection during his visit in Berlin (1978) Fig. 1): female body medium-sized, tapered at both ends. Head attenuated, rounded. Four externolabial papillae and four cephalic papillae in two rectangular cephalic shields expanded laterally, amphids tiny ( Fig. 1B, C). Buccal capsule slightly shorter than broad; anterior part with thin wall and posterior thicker ring embedded in oesophagus. Oesophagus with anterior muscular and posterior glandular parts of similar diameter. Vagina elongated, straight; ovejector directed anteriorly in its first half (Fig. 1A, H). Tail robust (Fig. 1D, E), extremity more or less truncated with three or four terminal lappets, one dorso-apical, one ventral and two lateral (Fig. 1E  maximum width 3.5-4.5, sheath not identified; anterior end slightly attenuated in dorso-ventral view, but not in lateral view; tiny cephalic hook; cephalic space longer than wide; last caudal nucleus far from tail tip, distal part of tail thin, undulated (Fig. 1F). • Taxonomic discussion These female specimens resemble those of Litomosa maki Tibayrenc, Bain & Ramanchandran, 1979described from Pteropus vampyrus (Linnaeus, 1758 in Malaysia (Ramachandran et al., 1966;Tibayrenc et al., 1979). Both have a complete set of head papillae (four externo-labial and four cephalic papillae), situated rather far from apex, a long oesophagus (≥ 1 mm) but a relatively small buccal capsule, and a truncated tail extremity with conspicuous conical lappets. No male was in the present sample but von Linstow (1897) gave the description translated here: "The 33.4 mm long and 0.31 mm wide male is distinctly narrowed posteriorly and the tail end is coiled in two close turns; the oesophagus is 1/22.8 (1,465 mm), the tail is 1/169 (0.198 mm) of the whole length; the spicules are very unequal, measuring 0.18 and 0.078 mm; at the tail end there are three teeth, one dorsal and two latero-ventral; papillae are not apparent". The characters of the male confirm the resemblance with L. maki. However, von Linstow's material differs in that the buccal capsule is composed of two segments (three in L. maki), the oesophagus is divided, the bodies of both sexes are longer (39.3-43.8 mm and 33.4 mm, in female and male, respectively, compared to 35 mm and 25 mm, in L. maki, as corrected by Tibayrenc et al., 1979). We thus propose the new combination Litomosa hepatica (von Linstow, 1897) n. comb. The microfilaria of L. maki has not been described. In L. hepatica the body of the microfilaria is not folded as is usual in species of Litomosa (Petit, 1980;Guerrero et al., 2002;Martin et al., 2006;Junker et al., 2009). In fact, the two species from megachiropterans display several morphological particularities not found in species of Litomosa parasitic in microchiropterans. Based on the long oesophagus, the arrangement of caudal papillae and the simple shape of the right spicule, Martin et al. (2006) suggested that L. maki from megachiropterans represented a primitive line in Litomosa. This is further supported by the number (8) and arrangement (far from mouth) of head papillae, in both species.
Locality: Porto Alegre, Brazil. Type specimens: hololectotype (male; ZMB Vermes Entozoa 1059), allolectotype (female), five male and two female paralectotypes (ZMB Vermes Entozoa 7429). Other paralectotypes (one male, a male posterior part, one female, and a female anterior part with microfilariae) in MNHN Paris collection, accession number 2 JW. General (Fig. 2): body very attenuated at both ends in both sexes. Males 1 ⁄ 3 length of females. Head rounded, head papillae grouped near apex, asymmetrically arranged, four small externo-labial papillae, two larger ventral cephalic papillae posterior to amphids. Oral opening tiny. Buccal capsule markedly longer than broad, posterior part embedded in oesophagus; irregular external aspect, faint asymmetrical or symmetrical ring in some specimens (Fig. 2D, K). Oesophagus divided into muscular and glandular parts, of equal lengths. Nerve ring located at mid-length of oesophagus. Male: four or five pairs of cloacal papillae, most anterior pair in adcloacal position; papillae of third and fifth pairs not symmetrically arranged (one papilla more posterior, placed on the median ventral line), papillae of fourth pair in transverse line but close to median ventral line; phasmids far from posterior end (Fig. 2I, M). Extremity of tail rounded. Lamina of left spicule with enlarged membranous part, terminal third rod-like with beveled extremity (Fig. 2H, J). Right spicule slightly sclerotized, with distal part elongated, supported by two fine cuticular rods and membranous extremity (Fig. 2L). Area rugosa composed of transverse bands of longitudinal crests. Measurements (hololectotype and extremes of six paralectotypes): total length 24.1 (21.5-24.6) mm, maximum width 149 (115-149), at nerve ring 58 (47-58) et al. (1989) distinguished two morphological groups in the genus, a sigmodontis and a carinii group. They placed von Linstow's species in the sigmodontis group, despite the original description of the spicules being unclear (see von Linstow's measurements and figure 73). The characters of the left and right spicules of von Linstow's type material as described above confirms that L. circularis belongs to the sigmodontis group. The morphological features that distinguish L. circularis from the 17 other species of the sigmodontis group subsequently described are listed in the following discussions (Mazza, 1928;Caballero, 1939Caballero, & 1947Esslinger, 1973;Muller, 1980;Bain et al., 1980Bain et al., & 1989Brant & Gardner, 1997;Notarnicola et al., 2000Notarnicola et al., & 2002Guerrero et al., 2002;Bain et al., 2003;Notarnicola, 2005;Notarnicola & Navone, 2009;. The single species of the sigmodontis group from marsupials, L. barretti Muller, 1980 is distinguished by a shorter buccal capsule (12 µm, measured in figure  1 of Muller), more cylindrical female tail, symmetrical arrangement of caudal papillae and higher spicular ratio, 3.0:1 (2.5:1 in L. circularis). Three species are parasites of bats. L. leonilavazquezae Caballero, 1939 is shorter than L. circularis in all measurements, except the buccal capsule; L. fosteri Caballero, 1947 has a buccal capsule with two thickened rings, a shorter right spicule and higher spicular ratio (4.8 compared to 2.44-2.86 in L. circularis); L. teshi Esslinger, 1973 has an asymmetrical buccal capsule and longer microfilariae 75-109 (compared to 62-63 in L. circularis). Thirteen species are parasites of rodents. L. hoplomyis Esslinger, 1973 in Eumysopinae (Echimyidae) is a very small species, 10-13 mm and 18-30 mm long respectively in males and females (21-25 mm and 60-80 mm long in L. circularis, respectively) and the female tail has a conical terminal part. In L. ctenomyos Brant & Gardner, 1997 from Ctenomyidae, the oesophagus is longer and undivided in the female and the caudal papillae are symmetrically arranged along the male tail to the tip.
The other species are parasites of Sigmodontinae (Muridae). L. patersoni (Mazza, 1928) redescribed by , has shorter, stout microfilariae (34-44), a pair of precloacal papillae, complete set of head papillae, and straight female tail with pointed extremity. Five species have prominent amphids; three of these have microfilariae longer than 75 ( Two species are known only from microfilariae (Marinkelle & Garcia-Castañeda, 1999): that of L. colombiensis Esslinger, 1973 is twice as long and tapers posteriorly to form a narrow tail tip while that of L. caliensis Esslinger, 1973 has a thicker tail with rounded extremity.
In conclusion, no species of Litomosoides falls into synonymy with L. circularis. Original contribution Molin (1858) described the species from three different hosts and two localities. However during examination it appeared that only one sample, NMW 6323, was adequate for study, whereas the other two samples were small fragments of females. Sample: NMW 6323. Host: Phyllostoma spiculatum (= Sturnira lilium) (Chiroptera: Phyllostomidae). Location: abdominal cavity. Locality: Ypanema, Brazil. Type specimens: hololectotype male, allolectotype female, a male and a female paralectotypes, three anterior and one posterior female paralectotypes. General (Fig. 3A-H): large filariae, females longer than males. At present, worms brownish, and head papillae not all observed (e.g. external-labial papillae) but, in lateral view, two large latero-ventral cephalic papillae identified. Buccal capsule markedly longer than broad, posterior half embedded in oesophagus; thickened ring at mid-length, at level of apex of oesophagus. Division of oesophagus into muscular and glandular parts distinct or not. Nerve ring located at mid-length of oesophagus. Tail hardly attenuated. Male: caudal papillae identified in lateral view: papillae of three anterior pairs on transverse lines, posteriorly three papillae longitudinally aligned. Right spicule with distal portion well sclerotized to its extremity, conspicuous dorsal keel posteriorly directed, subterminal rim and terminal cap (Fig. 3F). Measurements (hololectotype and a paralectotype): Total length 39.9 and 40.3 mm, maximal width 114 and 100 at mid-body, at oesophageal-intestinal junction 67. Buccal capsule 16 long, 9 wide; buccal ratio 1.88 and 1.78. Nerve ring to anterior end 196. Oesophagus 569 and 566 long, muscular anterior part 182 long. Left spicule partly visible, length of blade 240. Tail slightly bent ventrally, 207 long, 52 wide at cloaca. Area rugosa from 377 to 2,150 anterior to cloacal aperture, each band of ridges 24-26 long, distance between bands 11.1-12.6, ridges 3-4 high. Female: vulva well posterior to oesophago-intestinal junction. Vagina elongated, with distinct bulbous muscular part. Caudal end rounded, with a pair of phasmids. Measurements (allolectotype and extremes of a complete paralectotype, and three anterior and one posterior fragments of paralectotypes): total length 61.6 (78.1) mm, maximal width at mid-body 178 (169-185), at base of oesophagus 70, at vulva 105. [9][10] Guerrero et al. (2002) briefly observed the original specimens and found that these had the characteristic well sclerotized and segmented buccal capsule of Litomosoides, and resembled L. brasiliensis Almeida, 1936 with the pronounced distal cap and dorsal keel of the right spicule (Almeida, 1936). The taxon L. brasiliensis, originally described from Myotis sp. in Brasil, was subsequently assigned to specimens from diverse microchiropterans from South American geographic areas. However the redescriptions (Rego, 1961;Diaz-Ungria, 1963;Esslinger, 1973;Guerrero et al., 2002) were not all similar, particularly in the morphology of the buccal capsule and, when studied in ventral view, the arrangement of caudal papillae (Diaz-Ungria, 1963;Guerrero et al., 2002). Consequently, Guerrero et al. (2002) suggested that the taxon L. brasiliensis was very likely a complex of species and they prefered not to place L. brasiliensis in synonymy with the oldest taxon, L. serpicula.
The present study of Molin's specimens revealed a distinct buccal capsule: a single ring (Fig. 3B, H), instead of one small anterior and one larger posterior, as described for L. brasiliensis by Almeida (1936), Diaz-Ungria (1963), Esslinger (1973), Guerrero et al. (2002) for specimens from Carollia perspicillata, and . This difference supports our previous conclusions that Litomosoides serpicula and L. brasiliensis, both described from Brasil, are two distinct species, with Sturnira lilium and Myotis sp. as respective type hosts. Guerrero et al. (2002) studying specimens from Peruvian S. lilium had noted and illustrated a single ring in the buccal capsule, or a second very small and distant anterior one. However, to decipher the question of species diversity and host specificity, morphology of microfilariae will be useful as well as molecular analysis as has been done recently for some species of Litomosoides (Ferri et al., 2009). In addition we note here that the taxon Filaria spiculatum in the checklist of Ubelaker et al. (1977) is a lapsus calami for Filaria serpicula.
The sample NMW 6322, from the abdominal cavity of the phyllostomid bat Phyllostoma brevicaudum (= Carollia brevicauda), from Ypanema, Brazil, is composed of one anterior extremity and two fragments of female, in very poor condition. The sample NMW 6324, from the abdominal cavity of Phyllostoma sp. Incerta (sic) Rio Muria, Brazil is composed of pieces of females (one anterior extremity, another one without buccal capsule, two posterior extremities, and three Original contribution of Stammerinema Osche, 1955, a common acuarid in Soricidae, but no sound identification can be made in the absence of the characteristic dilated anterior part (Tiner, 1951;Soltys, 1951;Osche, 1955;Quentin, & Wertheim, 1986). Filaria hyalina is incertae sedis.