Hassalstrongylus dollfusi (Nematoda, Heligmonellidae): rediscovery in native South American rodents, six decades after its description

Hassalstrongylus dollfusi (Díaz-Ungría, 1963) Durette-Desset, 1971 was described in a wild house mouse, Mus musculus, from Venezuela and, since then, has never been reported again in the type host or in any other host. In this work, specimens assignable to H. dollfusi were found at 10 localities in Northeast Argentina, in five species of sigmodontine rodents. The nematodes were attributed to H. dollfusi based on diagnostic characters such as: synlophe with 22–31 subequal ridges; in males, hypertrophy of right ray 4 of the male bursa, thickening of the dorsal ray and bases of rays 8, distal tip of the spicules bent and spoon shaped; and, in females, presence of subventral postvulvar alae supported by hypertrophied struts. The new host recorded are: Oligoryzomys fornesi, O. flavescens, O. nigripes, Holochilus chacarius and Akodon azarae. The parasite showed a strong preference for host species of Oligoryzomys, which appear to act as primary hosts. The parasite could be present, parasitizing different species of Oligoryzomys, in a geographic area from the type locality in Venezuela southward to north Corrientes in Argentina. It has not been reported from populations of Oligoryzomys spp. of the Argentinean and Brazilian Atlantic Forest, nor south of 28° S, which may be explained by constraints in the environmental conditions required by the free-living stages of the parasite. This study provides the first identification and redescription of H. dollfusi in southern South America, from autochthonous hosts, six decades after its description.


Introduction
The Heligmonellidae are the most speciose family of the Trichostrongylina (Strongylida or bursate nematodes). Heligmonellids are distributed worldwide and are overwhelmingly parasites of rodents (ca. 350 species); very few species have been described from moles, lagomorphs, tragulids and dermopterans (< 20) [3,22]. Among the five subfamilies of heligmonellids, the Nippostrongylinae include the largest number of species (ca. 230); nippostrongylines are widespread all over the world and occur mainly in the superfamily Muroidea, frequently with more than one species per host [22].
In the Americas, nippostrongylines are parasitic in the Cricetidae (Muroidea), which are represented by four subfamilies: the Arvicolinae (with Holarctic distribution), the Neotominae and Tylomyinae (North or Middle American lineages); and the highly diverse Sigmodontinae (autochthonous to South America) [36]. Together, these subfamilies include ca. 650 species of rodents [43]. In contrast, less than 10% of these host species (ca. 60) have been investigated for nippostrongylines, which are known to be a highly diverse group [13].
Hassalstrongylus dollfusi (Díaz-Ungría, 1963) was originally described as Longistriata dollfusi Díaz-Ungría, 1963 in a wild Mus musculus from Venezuela [12]. Some years later, Durette-Desset (1969) [15] described the synlophe on reexamination of the type material, and finally Durette-Desset (1971) [16] transferred the species to the genus Hassalstrongylus. These three contributions are the only existing publications concerning H. dollfusi, there having been no new records since its description in 1963. It was, up to now, the only species of Hassalstrongylus reported exclusively from an exotic Murinae.
In this work, we report the finding of H. dollfusi in five species of sigmodontine hosts in 10 different localities of Northeast Argentina, providing a complete morphological description and enlarging the range of morphometrical data on this species. Additionally, we provide ecological parameters for the different surveyed populations and discuss the status of the different host-parasite associations.

Ethics
The research was performed in agreement with Argentine laws. The specimens, obtained using methods for live capture, were sacrificed following the procedures and protocols suggested by AVMA Guidelines on Euthanasia and approved by National Laws (Animal Protection National Law 14.346 and references in the provincial permits) and by the Ethics Committee for Research on Laboratory Animals, Farm and Obtained from Nature of the National Council of Scientific and Technical Research (CONICET). No endangered species were involved in this study.

Methods
Nematodes were recovered by observation of the gastrointestinal tracts under a stereomicroscope, subsequently fixed in 4% formalin and preserved in 70% ethanol. Nippostrongylines were studied in temporary mounts in Amman's lactophenol under a Leica DM 2500 microscope provided with a drawing attachment. Photographs were taken with a Leica DMC5400 camera. The synlophe was studied following Durette-Desset [18] and the nomenclature referring to the axis of orientation of the ridges follows Durette-Desset and Digiani [19]. Ridges are numbered from left to right and with respect to the oblique axis of orientation: 1 to n for right-dorsal ridges, and 1 0 to n 0 for left-ventral ridges. The nomenclature used for the study of the bursa (pattern of lateral lobes and symmetry) follows Durette-Desset et al. [21]. Measurements, unless otherwise stated, are provided in micrometres as the range followed by the mean in parentheses. SpL/BL and UtL/BL mean the proportion of the spicule length to the body length and of the uterus length to the body length, respectively. EP/OeL means the position of the excretory pore (in percentage) with respect to the oesophagus length. Ecological parameters (prevalence and mean intensity) were calculated according to Bush et al. [5]. With the aim of evaluating the status of the different hosts harbouring H. dollfusi, the relative dominance (RD) of the latter in relation to the other intestinal heligmonellid species was calculated for each host species and locality. Figure 1 was constructed with free and open code software QGIS 2.14 [37].  Localities (Fig. 1, Table 1): (0) Instituto de Investigación para la Pequeña Agricultura Familiar del Noreste Argentino (IPAF-NEA), Laguna Blanca, Departamento Pilcomayo, Formosa province (25°12 0 09.91 00 S, 58°07 0 13.71 00 W) (IPAF) (only H. chacarius examined from this locality. Not included in Table 1); (1) Estación de Animales Silvestres Guaycolec, Departamento Formosa, Formosa province (25°58 0 57.08 00 S, 58°10 0 04.00 00 W) (GUAY); (2) Selvas del Río de Oro, Departamento Libertador General San Martín, Chaco province, (26°47 0 23.35 00 S, 58°57 0 37.81 00 W) (SRDO); (3) Puerto Las Palmas, Departamento Bermejo, Chaco province, three sampling points (27°04 0 45.00 00 S, Prevalence and mean intensity: see Table 1. Material studied and deposited: see Table 2 Redescription (Figs. 2-6, Table 3) General: medium-sized nematodes, generally uncoiled, sometimes loosely coiled in 1-3 spirals. Cephalic vesicle present. Excretory pore always within posterior half of oesophagus (Table 3). Deirids at same level of excretory pore in most specimens (74% of 50 specimens), sometimes 3-25 lm anterior or posterior to it ( Fig. 2A).
Synlophe (studied in 6 males and 6 females): in both sexes, cuticle bearing longitudinal, uninterrupted ridges with welldeveloped struts. Ridges appearing gradually mainly on left side just posterior to cephalic vesicle up to oesophago-intestinal junction; disappearing just anterior to bursa in male and reaching distal cuticular inflation in females. Number of ridges variable in proportion to body diameter: at oesophago-intestinal junction, 22-26 in both sexes (Figs. 2C, 2D); at midbody 22-31 in males, 22-30 in females (Figs. 2E, 2F); within posterior third of body length 24-26 in males, 23-28 in females (Figs. 2G, 2H). All ridges medium sized, except those on right ventral quadrant (smallest) and those associated with lateral hypodermal chords (slightly larger). At oesophago-intestinal junction, especially in male, difference in ridge size more marked than at midbody. At midbody, ridge 2 0 largest. All ridges well oriented, determining a double axis of orientation of the ridges in both sexes: left axis inclined at 83°-84°with respect to sagittal axis, right axis at 70°-72°. Within distal third of body length: in both sexes, similar number and orientation than at midbody, differences in ridge size less marked (Figs. 2G, 2H). In females, between vulva and anus, presence of paired ventral postvulvar alae supported by hypertrophied struts (Fig. 3A).
Males (measurements reported by host in Table 3): bursa slightly dissymmetrical, with right lobe larger than left (Figs. 3B, 5A). Prebursal papillae not observed. Right lobe with pattern of type 2-2-1 tending to 1-3-1. Left lobe with pattern of type 2-3 tending to 2-2-1. Rays 2 shorter than rays 3 and curved toward median line. Rays 4 and 5 diverging at extremity, rays 4 longer, strongly curved anteriad, rays 5 straight. Right ray 4 markedly thicker than left one. Right ray 6 diverging from common trunk of rays 2-6 between rays 2 and 3; left ray 6 diverging distally in different degrees to ray 3: from just distal ( Fig. 3B 2 ) up to halfway between ray 3 and divergence of rays 4 and 5 (Fig. 3B 1 ). Rays 8 thickened at base, arising symmetrically from base of dorsal ray. Dorsal ray short, thickened at base, dividing within middle third into two branches, each one bifurcated distally into two papillae, external rays 9 and internal rays 10. Genital cone well developed. Telamon complex, composed of several sclerotized pieces: two dorsal simple branches forming a pincer or caliper, and two ventral bifid branches articulated distally (Fig. 4A). Gubernaculum inconspicuous, Spicules thin, subequal, alate. Each spicule tip spatulate or spoonshaped, bent at right angle with respect to main axis of spicule. Just before bending, spicular ala enlarged and ending widely, with the appearance of a heeled shoe (Figs. 4B and 5B, 5C).
Females (measurements reported by host in Table 3): reproductive tract monodelphic. Uterus less than 20% of BL, number of eggs variable (see Table 3). Infundibulum slightly longer than vestibule. Distal end, in most specimens, curved ventrally to different degrees, more rarely straight (on 50 females examined, 19% strongly curved, 30% curved at right angle, 28% slightly curved and 23% straight). Cuticle inflated to different degrees from level of sphincter or vestibule up to halfway between vulva and anus (Figs. 4C and 6). Two latero-ventral alae present between vulva and anus (Fig. 4C). Tail conical, not invaginated (Figs. 4C and 6).

Diagnosis
Diagnostic characters of H. dollfusi males are: the morphology of the caudal bursa (particularly hypertrophy of right ray 4 and thickening of dorsal ray and rays 8 at their base), and the peculiar shape of the distal tip of the spicules. Diagnostic characters of the female are: the dorsal cuticular inflation at ovejector level and the subventral postvulvar alae supported by hypertrophied struts. These characters were present in all specimens studied herein and, compared to those in the original description [12] and further redescription [15], did not show any differences. In addition, the synlophes at midbody of males and females examined were congruent with those described by Durette-Desset [15]. Related to the measurements, they were homogenous among specimens from the different host species, including the type material from M. musculus. The specimens parasitizing A. azarae and H. chacarius were slightly smaller (Table 3).
Slight variability with respect to the original description was found only in two characters: the level of divergence of left ray     Abbreviations: no = not observed, np = not provided. * Distance to apex. ** Distance to posterior extremity. 6 of the bursa (more distal in three of our males due to a larger right lobe, see Fig. 3B 1 ), and the general aspect of the female posterior end. Related to the latter, our specimens showed, independently of the host species, different degrees of curvature and different degrees of inflation of the cuticle (Fig. 6); whereas in the type material the posterior end is curved at a right angle and the cuticular inflation ends abruptly, almost perpendicular to the body wall (Fig. 4C 2 ). A degree of variability in these types of qualitative traits is expected when a large number of specimens are examined. Consequently, we identified our specimens as H. dollfusi.
Recently, Gómez-Muñoz et al. [24] reported, in two species of Oligoryzomys Bangs from Corrientes province, Argentina, specimens of nippostrongylines which were identified provisionally as Stilestrongylus sp., mainly based on characters of the synlophe. However, a more detailed study on those specimens indicated that they should be attributed to H. dollfusi and consequently they are included in this work. Values of prevalence, mean intensity and relative dominance of H. dollfusi (Table 1)  . In contrast, in H. chacarius, the prevalence was lower (P = 21.6%), and the intensity extremely low (MI = 2). In this host, the intensity of infection never exceeded 3 worms, and only one female worm was found in the whole sampling (Tables 1-3). In A. azarae, H. dollfusi was found in only one out of 28 examined hosts from the localities reported in this paper (P = 3.6%) ( Fig. 1; Table 1

Discussion
The finding of H. dollfusi in ten different localities in an area of about 19,000 km 2 , in five different host species and through several sampling events (see Table 2 for sampling details) allows us to consider H. dollfusi a usual component of the helminth fauna of sigmodontine rodents in this area. However, the values of prevalence and mean intensity, when considered globally by host species, greatly varied, being higher in the three species of Oligoryzomys and remarkably lower in H. chacarius and A. azarae (Table 1).
Data in Table 1 indicate that in the localities surveyed O. flavescens, O. fornesi, and O. nigripes act as primary hosts of H. dollfusi. The presence of the parasite in H. chacarius and A. azarae is, instead, less frequent, occurring when these latter share the habitat with the primary hosts; and even in such cases, H. dollfusi never appears as an important component of the helminth community. It is worth noting that in the localities reported herein, many other species of sigmodontine hosts were captured and examined for helminths. However, none of them was found to harbour H. dollfusi. These hosts were: Akodon montensis, Calomys callidus, C. callosus, Necromys lasiurus, N. obscurus, Oxymycyterus rufus and Sooretamys angouya.
As stated above, H. dollfusi was, up to now, only known from its type host Mus musculus, a non-native rodent. Díaz-Ungría [12] remarked that the host of H. dollfusi was captured in the wild (forest). In this sense, it is worth noting that the type locality (Tiara, Aragua Department, Venezuela) at the time of the study was a mining settlement (R. Guerrero, pers. comm.). Therefore, the presence of mice should not be surprising, and it is not unexpected that mice became infected with autochthonous nippostrongylines during incursions in the forest in the vicinities of the locality. Durette-Desset [15], when redescribing H. dollfusi, remarked that it was unlikely that the mouse was the primary host of the parasite, and that this latter should be found among the autochthonous cricetids. The fact that H. dollfusi was never reported again from mice supports this hypothesis. On the other hand, the lack of subsequent  records of the parasite even from native hosts should be attributed probably to both the paucity of helminthological surveys and the host preference showed by the parasite.
The primary host of H. dollfusi should undoubtedly be an autochthonous cricetid, but which one remains a question. In South America, the genus Oligoryzomys has 19 species distributed in a continuum from north Colombia and Venezuela to extreme southern Chile and Argentina (for a detailed distribution of the Oligoryzomys species see [42]). The specimens of H. dollfusi studied herein, compared to those from the type locality, show remarkable morphological homogeneity. Therefore, even when the present finding is far from the type locality of H. dollfusi, it would not be unreasonable to suppose that the parasite could be present in an area comprising at least from the type locality in Venezuela south to northern Corrientes province in Argentina, parasitizing different species of Oligoryzomys.
However, it is also worth noting the absence of H. dollfusi in several other localities where populations of Oligoryzomys spp. were present and abundant. This was observed in our study ( Fig. 1), but also in numerous studies carried out by Argentinian and Brazilian researchers regarding the parasitic fauna of different species of Oligoryzomys [6-8, 28, 31, 33, 38, 39].
Indeed, H. dollfusi was not found in localities south of 28°S (this work) where O. flavescens and O. nigripes were captured and examined for parasites (Fig. 1, localities F and H). Nor was the parasite found during an extensive study on the helminth assemblage of sigmodontine rodents (included O. flavescens and O. nigripes) from a broad wetland area of the Río de la Plata, Buenos Aires province, Argentina between 34°45 0 S, 58°06 0 W and 34°56 0 S, 57°42 0 W [31].
In a similar way, H. dollfusi has never been reported in any of the numerous studies on helminths of sigmodontine rodents in different localities from the Atlantic Forest from Brazil [6-8, 28, 38, 39] or Argentina [33], even when O. nigripes was one of the rodent species most commonly captured in such localities.
The Trichostrongylina are characterized by direct life cycles, in which the first three larval stages are free in the external environment and the stage infectious to the definitive host is the third-stage larva [2,22]. Therefore, the presence and persistence of the parasite in an area will be related both to the presence of populations of suitable definitive hosts and to the existence of adequate environmental conditions (temperature, humidity level, soil and vegetation type, etc.) for the free-living stages. Specific environmental requirements for the free-living stages of H. dollfusi are still unknown, but a probable constraint in these conditions may explain the absence of the parasite in well-surveyed populations of Oligoryzomys spp. south of 28°S in Argentina [31] and eastward in the Argentinean and Brazilian Atlantic Forest [6-8, 28, 33, 38, 39]. The extremely low value of RD of H. dollfusi in the unique positive locality corresponding to the Atlantic Forest biome (PPPI, Fig. 1, Table 1) is consistent with this hypothesis.
A significant sampling effort and thorough taxonomical work in the extensive intermediate area between the type locality and the area surveyed herein would greatly improve our knowledge of hosts and geographical distribution of H. dollfusi.
Meanwhile, this study provides the first identification and redescription of this species in southern South America, from autochthonous hosts, and more than 50 years after its original description.