Description of a new species of Haemogamasus (Mesostigmata, Laelapidae, Haemogamasinae) from Chubut, Río Negro and Neuquén Provinces, Argentina

Abstract A new species of Haemogamasus mites is described from Chubut, Neuquén and Río Negro Provinces, Argentina. It was collected primarily from rodents of the genus Abrothrix (65 of 77 collections): Abrothrix longipilis (63), Abrothrix olivaceus olivaceus (1) and Abrothrix olivaceus xanthorhinus (1). Additional collections were made from Geoxus valdivianus (5) and Loxdomtomys micropus (5). Possibly incidental or contaminate collections were recorded from Oligoryzomys longicaudatus (1) and Dromiciops gliroides (1). Most collections came from Río Negro Province (36), with 22 from Neuquén and 19 from Chubut. An identification key is provided to distinguish ♀s of this new species from ♀s of species from the western hemisphere, based on morphological characters.

In his major taxonomic classification treatment of Haemogamasinae, Keegan (1951) separated the subfamily into two closely related genera, Haemogamasus and Euhaemogamasus. This was based on a major diagnostic character described by Keegan as follows: "The genus Haemogamasus is distinguished from Euhaemogamasus only in that its species possess accessory sternal setae." More recent authors have combined the two genera into the single genus Haemogamasus. Strandtmann and Wharton (1958) stated: "A fine review of this family [Haemogamasidae] was written by Keegan (1951), a paper from which we have borrowed freely. We have followed Asanuma (1952) and placed Euhaemogamasus in synonymy with Haemogamasus." The new species described herein, Haemogamasus alongipilis, fits within the designation of the genus Haemogamasus as defined by Keegan, as well as the classification followed by Asanuma and other subsequent authors.
In his studies of North American H. liponyssoides, Radovski (1960) found 4 species with a distinguishing character that differs from all other Haemogamasus species. He found that H. liponyssoides, H. harperi, H. occidentalis, and H. keegani all have slender, untoothed chelae, and concluded that "They appear to constitute a closely related group within the genus, and are referred to here as the H. liponyssoides complex." In a more recent treatment of the Haemogamasus species of North America, Wil liams et al. (1978) presented a new classification for the genus. They placed the North American species into two groups: The liponyssoides group and the reidi group based the same morphological character as Radovski, plus several additional characters. In their classification the liponyssoides group is separated from the reidi group by the fol lowing characters: Slender, untoothed chelae; anal shield elongate pyriform; and some setae of gnathosoma, sternal shield, anal shield and legs unbarbed. They placed the 4 Radovsky species (H. liponyssoides, H. harperi, H. occidentalis, and H. keegani), plus a new species, H. ghani, into the liponyssoides group and all other North American spe cies (H. ambulans, H. pontiger, H. longitarsus, H. thomomysi, H. onychomydis and H. reidi) in the reidi group. Species of the reidi group are differentiated by the following characters: Stout strongly toothed chelae; anal shield broadly pyriform; and some setae on gnathosoma, sternal shield, anal shield and legs barbed.
We have chosen to follow the classification proposed by Williams et al.. The new species described herein, H. alongipilis, clearly fits within the liponyssoides group based on the above described morphological characters. Haemogamasus alongipilis is easily distinguished from the previously described liponyssoides group by the following char acters: Significantly smaller body size (dorsal shield ca. 538 µ median long and ca.272 µ greatest width); anal shield with no accessory setae; genitoventral shield shorter, not expanded posterior to gv1 setae and with only 1 accessory seta.
The structural morphology and body setae nomenclature followed herein is basi cally from Keegan (1951), which was subsequently refined by Lindquist and Evans (1965) and Evans and Till (1965, 1966, 1979. The basic leg segmental setae no menclature and setae formula followed is from Evans (1963), which was subsequently revised by Evans andTill (1966, 1969) and Masan and Fender (2010).

Materials and methods
Mammals were trapped with Museum Special snap traps (Woodstream Corporation®) baited with oat meal. Traps were set out in the late afternoon and examined in the morning. Each trapped animals was placed in a plastic bag for later processing for ec toparasites. Ectoparasites were collected by exposing each animal to a stream of carbon dioxide while vigor ously brushing over a funnel supported by a ring stand. A 7.5 cm × 15 cm plastic zip-lock bag containing 2 ml of 80% ethanol was affixed with tape to the bottom of the funnel. Ectoparasites (fleas, lice, mites, and staphylinid beetles) fell into the slick-walled funnel and were captured and preserved in the zip-lock bag containing an RDS-field number for the respective mammal host. Sealed bags were subsequently stored for longer periods in a bottle containing 80% ethanol. The funnel was thoroughly cleaned after each animal to preclude cross contamination of ectopar asites from one animal to the next.
Mite specimens were cleared in a solution of KOH and then put through a series of: distilled water, 70%, 85%, 95% & 100% ethyl alcohol, then methyl salicylate and finally xylene. The specimens were then mounted on slides in Canada balsam, and the slides were cured in a warming oven for several weeks in preparation for examination and identification.
Digital images in TIFF format were prepared by photographing with an Olym pus BX61 Compound Microscope, Olympus CV12 digital camera using an Olympus Microsuite™ B3SV program. This photograph imaging system was also used to per form all measurements for the detailed descriptions of both ♀s and deutonymphs. All measurements are given in micrmeters (mµ) and all "µ" in the description. In many cases the measurement range is given as "ca. x". We found that there is about a 5-10% variation in measurements of the same structure from specimen to specimen.
The TIFF images were cropped, enhanced, sized appropriately and saved in TIFF as well as JPG format with Adobe Photoshop CS5 Extended version 11.1. Line draw ings illustrating diagnostic characters were prepared by printing the TIFF digital im ages, then tracing them on a light box, scanning the traced images and making neces sary adjustment, including setae labeling, with Adobe Photoshop CS5.
Differential diagnosis. ♀s of H. alongipilis are separable from all Western Hemi sphere Haemogamasus species by the following combination of characters: (1) Signifi cantly smaller body size (dorsal shield ca. 538 µ median long and ca.272 µ greatest width); (2) sternal shield (see Fig.1) approximately square in shape (length/width ratio 0.98); (3) No accessory setae on the anal shield (see Fig. 2); (4) paranal setae positioned at a level posterior to the anal orifice; and (5) genitoventral shield (see Fig. 3) shorter, not expanded posterior to gv1 or gv2 setae and with only 1 accessory seta centered at level posterior to gv2 seta. H. alongipilis and H. harperi share a single diagnostic char acter: Both have sternal shield accessory setae which distinguishes them from all other western hemisphere species.
Thus far, no ♂s of H. alongipilis are known from the 77 collections in Argentina, probably because males are rarely collected; however many deutonymphs were col lected and are described herein. It is interest to note that of the 77 collections, there were 99 deutonymphs collected whereas there were only 37 ♀s collected. This ratio of deutonumphs to females (3:1) is likely because the deutonymph stage is the migrat ing stage. This species is named after the predominant rodent host species, Abrothrix longipilis, from which 65 of the 77 collections were made.  The following two line drawings show the general shape of the body as well as the most distinctive morphological characters. The dorsal view (Fig. 5) shows the outline of the dorsal shield with an anteriorlateral and a posteriorlateral cutout showing the relative size and number of setae. The ventral view (Fig. 6) shows the outline of the ventral shields with the relative size and number of setae associated with each. Dorsum (Figs 7-8): Body elliptical with dorsal shield length/width ratio ca. 0.98; dorsal shield median length ca. 538 µ (515 -576 µ) and greatest width ca. 272 µ (250 -294 µ); anterior margin acutely rounded; lateral margins at level of anterior coxa I slightly concave; lateral margins at level between coxa I and coxa II moderately concave forming a rounded shoulder at level of coxa I; median lateral margins almost straight to slightly rounded; posterior lateral margins more strongly rounded; and posterior margin broadly rounded. Dorsal shield bearing numerous setae, far too many to count (see Figs 5-8). Marginal setae setaceous (ca. 24-26 µ long); medial setae setaceous (ca. 16-26 µ long); posterior terminal pair of setae (Z5) long and slender (ca. 54 µ); and J5 setae (just anterior to Z5) setaceous and longer (ca. 60 µ).
Sternal shield (Fig. 10) generally square in shape extending posteriorly to level of mid coxae III; length/width ratio 0.98; median length ca. 99 µ and width between coxae II ca. 101 µ. Anterior margin of sternal shield convex between st1 setae. Posterior margin slightly concave medially between st3 setae. Sternal shield bearing the usual 3 pairs of setae (st1, st2 and st3) plus 7 pairs of accessory setae and 1 single seta between and posterior to level of first pair of accessory setae. The usual 3 pairs of sternal setae are setaceous and subequal in length (ca. 36-42 µ). The accessory setae vary in length with anterior medial setae distinctly shorter than the posterior lateral setae.
Setae st1 on the anterior margin with concavities on both sides of each seta. Two pairs of distinct slit-like pores/ trichopores (lyriform in shape) associated with setae st1 and st2; the pair posterior to st1 generally parallel to anterior margin; the pair between st2 and st3 oriented at nearly 45 degree angle with medial end tipped up; third pair off shield on soft integument adjacent to st4. Distance between st1 setae ca. 65 µ; distance between st2 setae ca. 116 µ; and distance between st3 setae ca. 129-235 µ. Distance between st1 and st2 setae ca. 46 µ and distance between st2 and st3 setae ca. 38 µ. Metasternal setae (st4) ca. 38 µ long; situated on small, oval metasternal shields at level of mid coxae III and just posterior to posterior-lateral margins of sterna shield.
Genitoventral shield (Fig. 11) tongue-shaped with 5 setae (2 pairs plus 1 single seta posterior to level of 2 nd pair); gv1 setae positioned laterally adjacent to coxae IV; gv2 setae positioned a little more than half way between the gv2 setae and the posterior end of the genitoventral shield and not near margins; all 5 setae subequal in length (ca. 42-43 µ).
Legs The length and width of leg segments is given in Table 2 below. All meas urements were made ventrally. Length measurements were made midventrally and widths were made at the widest point. The measurements are approximate because there is about a 5-10% variation in measurements from specimen to specimen.
The leg segment chaetotaxy for females is shown in table 3 below. The chae totaxy formula for each leg segment is as follows: al -d/v -pl (anterior later als -dorsals/ventrals -posterior laterals). The total for each leg segment is in parenthesis bolded.

Figures 14-22
Diagnostic characters for deutonymphs: (1) Body elliptical and significantly larger than ♀s: length ca. 707 µ, and width at level of coxae III ca. 430 µ. (2) Dorsal shield divided at level between coxae III and IV, with both dorsal shields thickly beset with setae. (3) Ventral shield (no separate sternal and genitoventral shields) elongate, extend ing from level of coxae I to posterior 1/3 of coxae IV and tapering strongly between coxae IV to a blunt point posteriorly. adanal setae short not extending to base of postanal seta, which is very long (about 3½ times as long as paranal setae).
The following two line drawings show the general shape of the body as well as the most distinctive morphological characters. The dorsal view (Fig. 14) shows the outline of the dorsal shields showing the relative size and number of setae. The ventral view ( Fig. 15) shows the outline of the ventral shields with the relative size and number of setae associated with each.
Legs: The length and width of leg segments is given in table 4. All measurements were made ventrally. Length measurements were made midventrally and widths were made at the widest point. The measurements are approximate because there is about a 5-10% variation in measurements from specimen to specimen. Leg II: Coxae II greatest width ca. 135 µ and mid-ventral length ca. 45 µ. Anterior seta setaceous and ca. 39 µ long; posterior seta setaceous and longer, ca. 54 µ. Ventral setae of trochanter setaceous but with proximal seta much longer (ca. 50 µ) than distal (ca. 35 µ). Distal dorsal setae of femur stout, setaceous and short (ca. 43 long µ). Proxi mal anterior dorsal seta of genu short setaceous (ca. 30 µ long), whereas posterior seta longer setaceous (ca. 52 µ long). Tarsus with pre-apical seta all setaceous.
The leg chaetotaxy for deutonymphs is shown in table 5. The chaetotaxy formula for each leg segment is as follows: al -d/v -pl (anterior laterals -dorsals/ventrals -posterior laterals). The total for each leg segment is in parenthesis bolded.
There were a total of 128 H. alongipilis specimens mounted and identified. Of these, 33 (25.8%) were ♀s and 95 (74.2%) were deutonymphs. There were no male specimens found in the collections, likely because males are rare compared to females and deutonymphs. For the 3 Abrothrix host species there were 110 H. alongipilis speci mens mounted and identified. Of these 30 (27.3%) were ♀s and 80 (72.7%) were deutonymphs. Because Haemogamasus mites are primarily nest parasites and the deu tonymphs are the migrating stage, the ratio (about 1:3) of females to deutonymphs in the collection is not surprising. Our study shows that the range of H. alongipilis extends over more than three de grees of latitude (39°2' to 42°5'S) and a distance of 400 kilometers. All of the localities where the mice had this mite are located within the Valdivian rain forest phytogeo graphical region (Veblen et al. 1983). Presumably the mites occur on the same rodent species on the western side of the Andes Mountains in Chile in this same floristic re gion, thus is probably not of special conservation concern due to a small geographical distribution. But the total extent of the range of this mite remains to be determined. The main host species, Abrothrix longipilis, has a much larger total range than where we found this new mite. Further studies of the ectoparasites of this rodent will be re quired to determine whether the mite and this host have co-equal ranges. We have one limited bit of information suggesting that the mite may not co-occur with all popula tions of A. longipilis. Four individuals of this mouse that were collected at Parque Na cional Laguna Blanca (39°3'S, 70°21'W) did not have H. alongipilis. This national park is located in the Occidental District of the Patagonian semi-desert region (Soriano 1983) that is characterized by the growth of perennial bunch grasses and low-grow ing shrubs, rather than trees. In this arid environment the A. longipilis are limited to microhabitats of dense growths of the bunch grasses or along the margins of isolated streams and seepages. The absence of H. alongipilis from these four A. longipilis collec tions could be because the mite cannot survive the off-host phase of its life cycle in this general environment that is much less humid than that of the Valdivian rain forests.
A majority of the described species of Haemogamasus have been reported only from Europe and Asia. Of the twelve species reported from the western hemisphere, only 2 (H. ambulans & H. pontiger) are known to occur worldwide, and in North America from most of Canada and the United States. The other 10 species have been reported exclusively from North America.
The new species described herein, H. alongipilis, is the first Haemogamasus species described from South America. However, there is one literature report of H. pontiger (Berlese) having been collected in South America. This report is from Argentina by Sil  Keegan's (1951) description of the ♀ is compared with Mosquera's Argentina ♀ description, this appears to be a valid but incidental record. Keegan reported the distribution of H. pontiger to be cosmopolitan, and that it is apparently a facultative parasite, often freeliving and associated with various animal nests. He also stated that H. pontiger "were found in a variety of habitats: Wheat straw, flax tow, rice straw, rice hulls, in sod, and associated with the clothes moth." Mosquera reported that the Argentina specimens were found free living in association with Acaridae mites feed ing on grains and stored products. Thus, it is likely that this species does not occur as parasitic on small mammals in Argentina, but rather was introduced into the country with imported grains and stored products from elsewhere.