Dactylogyridae (Monogenoidea, Polyonchoinea) from the gills of Auchenipterus nuchalis (Siluriformes, Auchenipteridae) from the Tocantins River, Brazil

Two species of Cosmetocleithrum Kritsky, Thatcher & Boeger, 1986 (both new) and two species of Demidospermus Suriano, 1983 (one new) are reported from the gills of the catfish Auchenipterus nuchalis, popularly known as “mapará”, from the Tocantins River and tributaries, North Region of Brazil. Cosmetocleithrum berecae n. sp. differs from all other species presently known in the genus by the morphology of the anchors presenting an elongate shaft and short recurved point, a coiled male copulatory organ (MCO) with three rings, and an elongate slender accessory piece with a bifurcated distal end. Cosmetocleithrum nunani n. sp. differs from its congeners by the combination of the following features: (1) Ventral and dorsal anchors with moderately long curved shaft and short point; (2) Hooks with poorly developed thumb; (3) Hook pairs 5 and 6 similar to each other, but morphologically distinct from remaining hook pairs; and (4) MCO coiled, with approximately 1.5 rings. Demidospermus tocantinensis n. sp. is easily distinguished from other species of the genus by presenting an inverted-G-shaped MCO with a median knee-like expansion. Demidospermus osteomystax Tavernari, Takemoto, Lacerda & Pavanelli, 2010 is redescribed based on paratypes and specimens from the gills of A. nuchalis from the Tocantins River, a new host and locality records for this species. The monotypic Paracosmetocleithrum Acosta, Scholz, Blasco-Costa, Alves & Silva, 2017, the only other Neotropical genus reported in siluriforms besides Cosmetocleithrum with species presenting two ribbon-like projections on the posterior margin of the dorsal bar, is considered a junior subjective synonym of Cosmetocleithrum.


Introduction
Siluriformes is a large and diverse order of fishes, collectively known as catfishes. Most of them are omnivores, unlike most freshwater fishes, nocturnal, and depend mainly on senses other than sight, such as tactile and chemo-sensitive barbels to explore their surroundings [20]. Siluriforms are hosts to an extraordinarily rich and diverse fauna of gill monogenoids, and this host-parasite system represents an attractive model for phylogenetic studies in the Neotropical Region [27]. The "mapará", Auchenipterus nuchalis (Spix & Agassiz), inhabits lower courses of the larger rivers of South America [12]. The global fauna of the Dactylogyridae infecting catfishes is very diverse and includes around 379 species belonging to 31 genera [25]. Almost half of the genera (14) and about 75 species are native to the Neotropical Region [25]. Demidospermus Suriano, 1983 and Cosmetocleithrum Kritsky, Thatcher & Boeger, 1986 are the most diverse genera among dactylogyrids of Neotropical catfishes.
Cosmetocleithrum was proposed to accommodate species of dactylogyrids from Oxydoras niger (Valenciennes) and Pterodoras granulosus (Valenciennes), all from the Amazon River basin [19]. The genus includes species characterized in part by the presence of two submedial ribbon-like projections on the dorsal bar, a copulatory complex comprising a variably coiled MCO with counterclockwise rings and an elaborate nonarticulated accessory piece [19]. Subsequently, several other species were proposed [1,38]. Presently, 15 species of Cosmetocleithrum are known from Brazil, Argentina, and Peru [22,27,34,36,41] from siluriform hosts belonging mainly to the Doradidae but also the Auchenipteridae and Pimelodidae. A single report of a species of Cosmetocleithrum in Hoplias malabaricus (Bloch) (Erythrinidae, Characiformes) [13] requires confirmation due to the unexpected association.
During studies on the helminth fauna of freshwater fish from Brazil, specimens of A. nuchalis have been examined from the Tocantins River and some tributaries in Tocantins and Maranhão, states of the North and Northeast Region of Brazil, respectively. During this study, two new species of Cosmetocleithrum and one new species of Demidospermus were described and Demidospermus osteomystax was redescribed from this host.

Materials and methods
During August 2010, 67 specimens of A. nuchalis from the Tocantins River and two of its tributaries (Itaueiras and Arraias Rivers), and one from the Rio dos Mangues, State of Tocantins, Brazil were examined for helminths. Fishes were captured with gill nets and hook and line, the gills were removed and placed in vials containing hot water (65°C) that were shaken; formalin was added to reach a concentration of 5%. Monogenoids were picked from the sediment and gill arches in the laboratory with the aid of a stereoscopic microscope. Some specimens were mounted in Hoyer's medium to study the sclerotized parts and others were stained with Gomori's trichrome to study the internal organs of the parasite [17]. Measurements are presented in micrometers; range values are followed by mean and number of structures measured in parentheses. Dimensions of organs and other structures represent the greatest distance; lengths of curved or bent structures (anchors, bars and accessory piece) represent the straight-line distances between extreme ends [19], except for Demidospermus tocantinensis n. sp. (see Fig. 5C). The numbering of hook pairs follows Mizelle [28] (see also [31]). Values of prevalence, mean intensity (range of intensity) and mean abundance (range of abundance) of infestation follow Bush et al. [8]. Specimens were illustrated with the aid of a camera lucida or a microprojector attached to an Olympus BX 50 microscope (both phase contrast and differential interference contrast  Cosmetocleithrum berecae n. sp. closely resembles members of the first morphological group, differing, however, from all known species of this group by presenting anchors with an elongate shaft and a copulatory complex comprising a coiled MCO with three rings, and a slender accessory piece with well-defined and blunt elements of the bifid distal end.   Copulatory complex comprising MCO, accessory piece. MCO sclerotized, a sinuous tube, margin of base sclerotized with a conspicuous flap directed posteriorly. Accessory piece varying in shape (rod-shaped to walking-stick-shaped), non-articulated, serving as guide for MCO. Germarium pre-testicular. Oviduct, ootype, uterus not observed. Seminal receptacle anterior to germarium; globose. Vagina sclerotized; vaginal aperture dextral. Vitellaria scattered throughout the body, except for the region of reproductive organs. Eggs not observed. Haptor with dorsal, ventral anchor/bar complexes, seven pairs of hooks with ancyrocephaline distribution. Anchors similar, each with tapering superficial root, reduced deep root, short shaft, elongate point. Ventral bar articulated, strongly U/V-shaped; dorsal bar variably V-shaped, articulated medially, distal end blunt or slightly bifid. Hooks pairs similar, with short point, protruding thumb, delicate shank comprised by a single subunit.

Remarks
Demidospermus osteomystax was proposed by Tavernari et al. [40] parasitizing Auchenipterus osteomystax from the upper Paraná River floodplain, Brazil. The discovery of D. osteomystax in A. nuchalis represents a new host record for this species and in a new river system, the Araguaia-Tocantins basin.
The specimens collected in the Tocantins River were considered a member of this species, despite small differences (Table 1, Figs. 3 and 4) that were interpreted as intraspecific variations or artifacts of the preparation method. For instance, bars and anchors of the type specimens studied were overly flattened and may have resulted in the broad appearance of the articulation of the bars. Small differences in the morphology of the copulatory complex may represent real intraspecific features (e.g. comparatively reduced fringe around the MCO aperture in specimens of the Paraná River) (cf. Figs. 3 and 4). However, the study of specimens available from the Tocantins River and re-examination type specimens from the Paraná River clearly show that the vaginal aperture of the species is dextral, as opposed to the sinistral position described by Tavernari et al. [40].
Due to the uncertain/questionable phylogenetic status of Demidospermus (as presently composed) and the absence of an emended morphological diagnosis allowing a more precise assignment of species, we tentatively retain D. osteomystax within this genus, while recognizing that it may represent a member of a genus-group not yet formally proposed.

Remarks
Like D. osteomystax, D. tocantinensis n. sp. also presents a dextral vagina. Similar to D. osteomystax, the new species is provisionally allocated to Demidospermus simply for the lack of more compatible generic taxon. However, similar morphology between these species (especially in the position of the vagina and the morphology of the haptoral sclerites) appear to indicate their phylogenetic proximity, which may result in subsequent proposal and assignment to another generic group. Demidospermus centromochli, originally assigned to Demidospermus and that also parasitize auchenipterids, depicts similar morphological features and may represent an additional member of this morphological (perhaps monophyletic) group. Among these species and all other species presently allocated to Demidospermus, D. tocantinensis n. sp. is easily distinguished by the unique morphology of the MCO (inverted-G shaped with a knee-like protuberance near the first third of its length).

Discussion
It is evident from phylogenetic hypotheses (e.g. [25]) that the classification of the Dactylogyridae from freshwater Neotropical siluriforms is far from stable. This reflects the incipient knowledge about the richness and composition of species in the group. Generic groupings are being recognized continuously, representing approximately 60 years of efforts of different scientists and research groups [3,19,29,30,35,36]. At early stages of faunal characterization, it is common to generate "catch-all" genera to accommodate morphologically similar species. This is well exemplified by the scenario revealed by Kritsky et al. [19] for the Neotropical Urocleidoides Mizelle and Price, 1964. Since the recognition of the morphological identity of the type species of the genus in the abovementioned study, new generic groups have been proposed for species [15,18,19,23] that could be (and were) allocated in the past to Urocleidoides due to the exceedingly general original diagnostic features [32].
In the present study, four species have been described or redescribed from A. nuchalis and allocated to two previously known genera: Cosmetocleithrum and Demidospermus. The study of the specimens from museum and freshly collected specimens of previously known species revealed further problems associated with the generic classification of the parasites of siluriforms from the Neotropics.
The fauna of monogenoids from the gills of Neotropical siluriform fishes includes species of Demidospermus, comprising 31 known species. According to Braga et al. [7], species of this genus display the broadest host range among siluriforms, occurring in species of Loricarioidei and Silurioidei. Other monogenoid genera with species parasitizing Neotropical  freshwater siluriforms are Cosmetocleithrum, with eight species described from members of the Doradidae, two from Auchenipteridae and one from a pimelodid host from Brazil, Peru and Argentina; Ameloblastella, with 11 species from Pimelodidae, Hypophthalmidae and Heptapteridae; and Vancleaveus, with five species, described from species of Pimelodidae, Doradidae, and Loricariidae [4,9,22] [2]. Among the genera mentioned above, only species of two taxa are known to present two ribbon-like projections on the posterior margin of the dorsal bar: Cosmetocleithrum and Paracosmetocleithrum. The monotypic Paracosmetocleithrum was diagnosed by Acosta et al. [2] to include species presenting "a well-developed ornamentation in the middle portion of the ventral bar, and a sclerotized patch on the surface of the dorsal bar with an inconspicuous medial process that possesses two submedial projections arising from the tapered ends of this patch" and a dextro-marginal vagina. Acosta et al. [2] further justified the Paracosmetocleithrum on its position in a phylogenetic analysis based on 28S rDNA.
The morphological variability of known species of Cosmetocleithrum may suggest that the genus contains several subordinate clades that can be seen in the future as distinct genera following phylogenetic reconstruction of the species. Among other variable features, there are species presenting somewhat straight bars and others with V/U-shaped bars, similar to some species of Demidospermus (compare species proposed by Kritsky et al. [19]. However, P. trachydorasi is similar to the species of Cosmetocleithrum proposed herein and many others previously described, except for the position of the vagina. The vagina in almost all species of Cosmetocleithrum is reported as sinistralexcept for C. tortum (dextral)while the single species of Paracosmetocleithrum was also reported as having a dextral vagina.
A trivial comparison of the type species of Cosmetocleithrum, C. gussevi, and P. trachydorasi reveals their great similarity, including many of the putative features considered diagnostic for Paracosmetocleithrum suggested by Acosta et al. [2]. For instance, both species present similar sclerotizations on the ventral and dorsal bars, the similar general organization of the copulatory complexwith accessory piece non-articulated and distally bifurcatedegg sub-globose with a short appendix, lacking eyes, and by the presence of two ribbon-like posterior projections on the dorsal bar. Furthermore, restudy of available type specimens of P. trachydorasi (CHIOC 38.881 a-d) demonstrated that the original description failed in determining the ventral and dorsal orientation of the specimens studied, resulting in an error of the position of the vaginal aperture. The vaginal pore in the species is sinistral (and not dextral as originally described). For the same reason, the illustration mistakenly depicts the vas deferens looping the right caecum. While both species appear unique, they are clearly congeneric. Since Cosmetocleithrum has priority over Paracosmetocleithrum, P. trachydorasi is transferred to Cosmetocleithrum as C. trachydorasi comb. nov. Paracosmetocleithrum is a junior subjective synonym of Cosmetocleithrum.
Additionally, the phylogenetic independence of Paracosmetocleithrum suggested by the hypothesis of Acosta et al. [2] is questionable. The reduced support for its sister-group relationships and the impossibility of testing its monophyly are deceptive. By having a single species, Paracosmetocleithrum is monophyletic by assumption. Also, the low values of posterior probability and/or bootstrap of several putative ancestral branches of the phylogeny used to justify the identity of Paracosmetocleithrum (see Figure 3 of Acosta et al. [2]) suggest a soft polytomy (see [21]) which includes P. trachydorasi, two species of Cosmetocleithrum, and two clades including species of Demidospermus and unidentified dactylogyrids.
As suggested by Franceschini et al. [11], Demidospermus likely represents an unnatural taxon due to the recurrent polyphyletic nature of its current members in several phylogenetic hypotheses based on molecular data [11,25,26]. Unfortunately, previous phylogenetic studies did not explore the evolution of morphological features, which could provide support for a temporary or a definitive resolution to the problem. The integration of molecular and morphological data into a more comprehensive phylogenetic hypothesis can potentially provide diagnostic features for generic groups, as demonstrated by Mendoza-Palmero et al. [26]. However, even in the absence of molecular data, a comprehensive analysis of morphological features can provide important evidence for yet undisclosed supraspecific groupings within the apparently polyphyletic Demidospermus. The critical analysis of museum and freshly collected specimens, and descriptions of the known species reveals distinctive features (synapomorphies?) of dactylogyrids of auchenipterid hosts presently allocated to Demidospermus.
For instance, all five species of Demidospermus from auchenipterid hosts have a morphology similar to other species of Demidospermus, except for two species, D. osteomystax and D. tocantinensis, both of which present a dextral vagina. Examination of the whole mount illustrated in the original description of D. centromochli (Fig. 1 of [24]) suggests that the vagina may also be dextral, based on the pathway of the vas deferens that invariably loops the left caecum in Dactylogyridae [5,6]. Since the report of D. uncusvalidus on an Auchenipteridae is likely erroneous [16], the only species presently allocated to Demidospermus from this family of hosts that depicts sinistral vagina is D. bidiverticulatum. Sharing of a dextral vagina and similarities on the morphology of the MCO appears to suggest common ancestry of these species and may result in the proposal of a new genus.
However, to avoid additional instability to the already confused classification of dactylogyrids of freshwater Neotropical siluriforms, the proposal of a new genus is not considered desirable at this moment. For this reason, D. osteomystax and D. tocantinensis are provisionally retained in Demidospermus until both morphological data and molecular markers are available for them and several representatives of present and future known species from these hosts, which may result in the recognition of monophyletic groups and their respective morphological synapomorphies that will allow the establishment of a more robust classification of the group. While the use of molecular markers may generate information important to the reorganization of the classification of these dactylogyrids (and any other group), this should not signify that morphological analyses and interpretations can be ignored or belittled. Unfortunately, this seems to be the general pattern of many recent publications on Neotropical Monogenoidea. The published errors in determination of the dorsoventral axis of mounted specimens resulting in the wrong allocation of the vagina and other sclerotized structures in recent taxonomic accounts of the group and other serious mistakes in morphological interpretations, are consistent with this.