Introduction

Acanthocephalans of the family Polymorphidae Meyer, 1931 are parasites of marine mammals, waterfowl and fish-eating birds (Schmidt and Hugghins 1973). At present, fifteen polymorphid genera are recognized worldwide (Amin 2013; Presswell et al. 2020), whose life cycles typically include crustaceans as intermediate hosts and, in some cases, fish as paratenic hosts (Schmidt 1985). Among them, Profilicollis Meyer, 1931 was considered as a subgenus of Polymorphus Lühe, 1911 for many decades, until Nickol et al. (1999) reinstated its generic status based on ecological evidences. These differences are mainly related to life cycle characteristics; indeed, whereas members of Polymorphus use amphipods as intermediate hosts, those of Profilicollis use decapods (Nickol et al. 1999, 2002). Amphipods were the ancestral intermediate hosts, while the association with decapods represents episodes of secondary colonization as shown by phylogenetic studies that support the monophyly and the validity of the genus Profilicollis (García-Varela and Pérez-Ponce de León 2008; García-Varela et al. 2013).

An additional difference between Polymorphus and Profilicollis seems to be the use of fish as paratenic hosts by the former, a feature shared by other polymorphid genera, i.e. Andracantha Schmidt, 1975, Bolbosoma Porta, 1908 and Corynosoma Lühe, 1911 (Aznar et al. 2006). Indeed, cystacanths of Polymorphus have been reported parasitizing the internal organs of several freshwater fish in America (Amin et al. 1995; Santos et al. 2008; García-Prieto et al. 2010; Alcántar-Escalera et al. 2013; Rauque et al. 2018). On the other hand, cystacanths of Profilicollis have occasionally been found in the stomach and intestinal contents of fishes (Alarcos and Etchegoin 2010).

Profilicollis chasmagnathi (Holcman-Spector, Mañé-Garzón & Dei-Cas, 1977) is the only representative of the genus so far reported at adult stage along the southwestern Atlantic coast (Lorenti et al. 2018). This species has been recorded at the estuaries of Buenos Aires Province and Patagonian coasts, infecting the gut of several bird species (Martorelli 1989; Vizcaíno 1989; Diaz et al. 2011; La Sala et al. 2013; Lorenti et al. 2018). Cystacanths of this species are common parasites of different crab species inhabiting estuarine and rocky intertidal marine habitats in Uruguay and Argentina (Holcman-Spector et al. 1977; Martorelli 1989; Alda et al. 2011; La Sala et al. 2012; Méndez Casariego et al. 2016; Rodríguez et al. 2017). These cystacanths occur at high prevalence in two varunid crab species, Cyrtograpsus angulatus (Dana, 1851) and Neohelice granulata (Dana, 1851) in soft bottom intertidal areas and salt marshes of the Mar Chiquita coastal lagoon, Buenos Aires Province, Argentina (Martorelli 1989; Méndez Casariego et al. 2016). These crabs have also been found in the stomach and intestine, along with cystacanths of P. chasmagnati, of several marine fish species that use the estuary as feeding grounds (Alarcos and Etchegoin 2010).

During parasitological studies on estuarine and freshwater fishes from the Mar Chiquita basin, cystacanths referable to Profilicollis were found in the intestine of the silverside Odontesthes argentinensis (Valenciennes, 1835) (Atherinopsidae) from estuarine areas. Similar larvae were also found in the mesenteries of the dientudo Oligosarcus jenynsii (Günther, 1864) (Characidae) from a freshwater tributary of the lagoon, along with larvae of the genus Polymorphus in the same microhabitat. This was an unexpected finding, because members of Profilicollis have never been recorded infecting the internal organs of fishes.

To the best of our knowledge, there is a single previous record of polymorphid acanthocephalans in the genus Oligosarcus, namely of cystacanths of Polymorphus sp. in O. hepsetus (Cuvier, 1829) from Brazil (Abdallah et al. 2004). The closely related genera Profilicollis and Polymorphus can be very difficult to distinguish based only on their morphology because they display some degree of overlapping in diagnostic features (Amin 1992). Furthermore, species within Profilicollis are very similar to each other (Rodríguez and D’Elía 2016; Rodríguez et al. 2016). Consequently, the aim of this research was to identify those larval acanthocephalans found parasitizing O. jenynsii to get insight on the potential departures from the typical life cycle (two-host and marine-estuarine) of Profilicollis, which could be driven by the variability in the ecological conditions imposed by an ecotonal freshwater-estuarine environment.

Materials and Methods

Study area

Fish samples were obtained at three different locations of the Mar Chiquita basin. This is a coastal lagoon in south-east Buenos Aires Province, Argentina. It is an elongated water body, parallel to the coastline, separated from the sea by a barrier of dunes and only connected to it by an inlet that seasonally varies both in width and in position (Isla 1997). Mar Chiquita is characterized by a marked salinity gradient, with values ranging from 29.15 (at the inlet area) to 2.8 (at its innermost section) (Marcovecchio et al. 2019). These authors identified three different areas, an external area functioning as a coastal marine system, an intermediate estuarine system and an inland water system, whose extents vary inter-annually depending on the rainfall. For example, the extent of the marine section varied between 12 and 71% of the total surface across rainy and dry years (Marcovecchio et al. 2019). The limit between these sections is also variable daily and seasonally depending on the amplitude of the tides, the meteorological conditions and the volume of fresh water present in the lagoon (Reta et al., 2001). The lagoon, with a total area of ~60 km2 receives waters from numerous streams (Marcovecchio et al. 2019) being also in connection with two Pampean shallow lakes, Nahuel Rucá and Hinojales. One of its main tributaries is Arroyo Grande stream, to the north of the lagoon. This system is therefore a vast ecotonal region of great biodiversity that combines freshwater environments, wetlands and grasslands with marshes, coastal dunes and beaches (Fig 1).

Fig. 1
figure 1

Map of the study area showing Mar Chiquita coastal lagoon and sampling sites (black dots), its main tributary streams, and Nahuel Rucá and Hinojales lakes, in Buenos Aires Province, Argentina. Locality References: AG: Arroyo Grande stream, HL: Hinojales lake NR: Nahuel Rucá lake, MC: Mar Chiquita coastal lagoon, CC: estuarine area near Cangrejo creek discharge

Full size image
Table 1 Prevalence and mean abundance, 95% confidence intervals in parenthesis, of Profilicollis chasmagnathi and Polymorphus sp. parasitizing Oligosarcus jenynsii, Odontesthes argentinensis and Neohelice granulata from Mar Chiquita basin, Buenos Aires Province, Argentina. Locality References: AG: Arroyo Grande, CC: estuarine area near Cangrejo creek discharge, MC: Mar Chiquita lagoon, NR: Nahuel Rucá lake
Full size table

Host samples

Fish were captured, by means of gillnets, at different locations during parasitological surveys in the lagoon. A total of 30 specimens of O. jenynsii were obtained at Arroyo Grande (AG: 37°31'14"S, 57°19'37" W; April 2018) and 40 O. argentinensis were caught inside the lagoon body (MC: 37°36'54"S, 57°23'26"W; April 2017-Agust 2018). Additionally, a sample of three decapod crabs, N. granulata, collected by hand in an estuarine area near the freshwater discharge of Cangrejo Creek (CC: 37°44'14"S, 57°26'18"W; March 2019), was used to obtain cysthacanths from the natural intermediate host. Also data from a sample of 214 O. jenynsii, caught at Nahuel Rucá lake (NR: 37°36'57"S, 57°25'29"W; July 2008-October 2012) was used for comparative purposes (Fig. 1).

Parasitological studies and morphology

All hosts were frozen at -20°C until examination. After thawing at room temperature, they were necropsied and the polymorphid acanthocephalans were collected. Prevalence and mean abundance (sensu Bush et al. 1997) were calculated for each parasite species in each host sample, with 95% bootstrap confidence intervals, following Rózsa et al. (2000), using Quantitative Parasitology software (QPweb) (Reiczigel et al. 2019). The microhabitat and state of preservation of each cystacanth was recorded, considering larvae as viable when entire and encysted, or as dead when partially reabsorbed (darkened and degraded).

Complete individuals were kept in distilled water for several hours in order to allow the eversion of the proboscis prior to fixation; they were later identified at generic level following Amin (1992) and Presswell et al. (2020). A single Profilicollis sp. from each host species and one Polymorphus sp. found parasitizing O. jenynsii were reserved for genetic identification (fixed and preserved in 96% ethanol). The remaining fully extended specimens were fixed in 4% formalin to later perform morphological studies under a stereoscopic microscope. Measurements included total length, proboscis length and width, neck length and width, somatic spines area length and width and hind trunk length and width. In addition, the number of hook rows and hooks per row was recorded. These variables were used in the taxonomical identification of the cystacanths, to the lowest level possible.

DNA extraction, PCR amplification, sequencing and sequence analysis

DNA extractions were carried out using the DNeasy Blood and Tissue® kit (QIAGEN, Hilden, Germany) on whole specimens. A region of the mitochondrial DNA was amplified by polymerase chain reaction (PCR). Mitochondrial cytochrome c oxidase subunit 1 (mtDNA cox1) was amplified using the universal DNA primers LCO1490 (forward) and HCO2198 (reverse) described by Folmer et al. (1994). All PCR reactions were set up in 25 μl reactions using 5 μl of DNA (≥10 ng) as the template, 0.5 μl (0.5 mM) of each primer and 12.5 μl (2X) HotStarTaq Master Mix (QIAGEN). The PCR was carried out using the following conditions: initial step for enzyme activation and denaturation at 95 °C for 15 min, followed by 35 cycles of amplification at 94 °C for 30 s, 50 °C for 30 s and 72 °C for 1:45 min, followed by 10 min of post amplification at 72 °C. Each PCR product was purified using QIAquick spin columns (QIAquick Gel Extraction Kit, QIAGEN). Sequencing was performed using ABI 3730XLs automated sequencer (Applied Biosystems, Macrogen, South Korea).

Sequences were edited and assembled in Proseq 3.5 (Filatov 2002). For identification, consensus sequences were compared against the NCBI database using the BLAST algorithm (Altschul et al. 1990). Curated contig sequences were deposited in GenBank (Accession numbers: MT580122, MT580123, MT580124, MT580125, Fig. 2)

Fig. 2
figure 2

Maximum likelihood tree inferred by mtDNA cox1 showing the taxonomic position of Profilicollis chasmagnathi and Polymorphus sp. from three host species (Neohelice granulata, Oligosarcus jenynsii and Odontesthes argentinensis). Nodal supports are indicated for ML, MP (1,000 replicates, only bootstrap values greater than 70% are shown) and BI (only posterior probabilities greater than 0.7 are shown). Names in bold correspond to sequences obtained in the present study. Asterisks indicate sequences deposited as P. antarticus that were posteriorly identified as P. chasmagnathi (KU928252, Rodríguez and D’Elia 2016; KX646756, Rodríguez et al. 2016)

Full size image

The obtained fragments from the mtDNA cox1 gene were aligned with sequences from other members of the genera Profilicollis and Polymorphus, retrieved from GenBank. Sequences from Bolbosoma turbinella (Diesing, 1851) and Corynosoma australe Johnston, 1937 were also aligned and used as outgroup. The alignment was performed by ClustalW (Thompson et al. 1994) as implemented in MEGA 7.0 software package (Kumar et al. 2016), using default parameters. The reading frame for the mtDNA cox1 sequences was determined by translating the sequences, specifying the appropriate gene code (invmtDNA) and by starting at different positions in the alignment and inspecting for stop codons.

The estimation of intra and interspecific genetic divergence among specimens was conducted in MEGA7.0 software package (Kumar et al. 2016), using the Tamura-Nei model (Tamura and Nei 1993).

Hypothesis on the specific identity of the cystacanths found in different host species were tested using evolutionary three inference methods, based on analysis of character-state data, as recommended by Nadler and Pérez-Ponce de León (2011), which has the advantage of revealing the particular changes in states supporting individual species. For that purpose, three different inference methods, namely maximum parsimony (MP), maximum likelihood (ML) and Bayesian inference (BI), were used to construct trees in order to visualize relationships between the sequences obtained in the present study and those from cystacanths and adults retrieved from Genbank. MP analyses were performed using PAUP∗ 4.0b10 (Swofford 2001), using a heuristic search with tree-bisection-reconnection (TBR), branch swapping and random addition of sequence. All characters were treated as unordered. ML analyses were performed using PhyML 3.1 (Guindon and Gascuel 2003).

Reliabilities of phylogenetic relationships were evaluated using nonparametric bootstrap analysis (Felsenstein 1985) with 1,000 replicates for MP and ML trees. Bootstrap values exceeding 70 were considered well supported (Hillis and Bull 1993). Bayesian inference was performed with MrBayes 3.1.1 (Ronquist and Huelsenbeck 2003). The Bayesian posterior probability analysis was performed with the MCMC algorithm where the number of chains was 4, the temperature of heated chains was 0.2 with 1,000,000 generations while the sub-sampling frequency was 100, with a burn-in fraction of 0.25. JModelTest (Posada 2008) was run to determine the best-fit model for the obtained data set, as implemented in the Akaike information criterion (AIC) (Posada and Buckley 2004). The best-fit model GTR + I + G was used for BI analysis and ML.

Results

Larval acanthocephalans, referable to Profilicollis, were found encysted in the intestinal mesenteries of O. jenynsii caught at AG. Similar cystacanths were found free in the intestinal contents of O. argentinensis from MC and in the body cavity of N. granulata (the natural host) from CC. A relatively high prevalence and low mean abundance of Profilicollis were recorded in both fish species (Table 1). In addition, a specimen of O. jenynsii from AG had two cystacanths referable to Polymorphus encysted in its intestinal mesenteries. In comparison, those O. jenynsii from NR were parasitized only by Polymorphus sp., at lower prevalence and mean abundance than in AG (Table 1).

The cystacanths found in O. argentinensis were in most cases associated with remnants of the crab C. angulatus, present in the intestinal contents. All of them had their proboscis invaginated and were therefore considered as viable. Some O. argentinensis presented encapsulated lesions in their intestinal wall, with dark, degraded contents that, in some cases, contained remains of proboscis that resembled those of Profilicollis sp. These lesions could indicate previous attachment sites; however, they were not quantified as Profilicollis sp., since their identity could not be asserted. On the other hand, all cystacanths found on O. jenynsii were encysted in the mesenteries and evidenced a strong reaction by the host, represented by brownish encapsulations. Despite this, 46% of them were identified as viable.

The Profilicollis sp. collected from the three host species (O. argentinensis, O.jenynsii and N. granulata) were morphologically similar to each other, with a constant number of hook rows in the proboscis (18) and of hooks in each row (8), although larvae from O. jenynsii showed slightly smaller body measurements than those from crabs and O. argentinensis (Table 2). Morphological and morphometric data allowed determining them as P. chasmagnathi according to Amin, 1992 and Vizcaíno, 1989.

Table 2 Measurements of Profilicollis chasmagnathi parasitizing Oligosarcus jenynsii, Odontesthes argentinensis and Neohelice granulata. Measurements are given in micrometres (μm)
Full size table

Further confirmation was accomplished through the genetic studies, which demonstrated that all specimens studied were in effect P. chasmagnathi. The mtDNA cox1 sequences (637 bp) from the cystacanths obtained from N. granulata crabs, O. jenynsii and O. argentinensis matched >99% with the mtDNA cox1 sequences of P. chasmagnathi deposited in GenBank. The genetic divergences between the P. chasmagnathi from the present study and those from GenBank were in average 1.0% (range: 0.9% - 1%). When compared with sequences from congeners available in GenBank (P. botulus (Van Cleave, 1916), P. altmani (Perry, 1942) and P. novaezelandensis Brockerhoff & Smales, 2002), the interspecific divergences ranged from 16 to 30%.

Regarding those larvae morphologically determined as Polymorphus sp., the closest match of a mtDNA cox1 sequence from the cystacanth found in O. jenynsii (631 bp) was 93.8% to a sequence of Polymorphus deposited en GenBank. The genetic divergence among species of Polymorphus retrieved from GenBank varied between 7.3, with P. brevis (Van Cleave, 1916), to 42%, with P. minutus (Zeder, 1800).

The results from all genetic analyses are summarized in a single tree with bootstrap and posterior probability values, as shown in Fig. 2. The ML analysis resulted in a single tree with -log likelihood: 4382.4844. The MP analysis revealed that 333 characters were constant; 293 were parsimony-informative, and 27 variable characters were parsimony-uninformative. For BI analysis, the average standard deviation of split frequencies was 3.29 × 10-3, after 1×106 generations. Trees based on ML, MP and BI analysis yielded similar topologies, as shown in the ML consensus tree (Fig. 2).

In summary, the mtDNA cox1 analysis allowed determining that the Profilicollis cystacanths sequenced in this study were in effect P. chasmagnathi, while the specific status of the Polymorphus sp. found in O. jenynsii could not be further asserted, although its sequence was included as a sister taxon to P. brevis in a well-supported clade.

Discussion

An integrative approach, combining morphological and genetic evidences allowed confirming that all cystacanths of the genus Profilicollis found in O. argentinensis from MC, O. jenynsii from AG and N. granulata from CC, are conspecific and belong to P. chasmagnathi. Those cystacanths of Polymorphus sp. found in O. jenynsii could not be morphologically determined beyond the genus level due to their larval stage, since the diagnostic characters at species level often rely on the structure and morphology of the reproductive system in both sexes (Presswell et al. 2020).

The genetic analysis showed that the relationships among Profilicollis and Polymorphus species, recovered in this study, are congruent with those from previous studies (García-Varela et al. 2013; Huston et al. 2020). Indeed, phylogenetic relationships among species of Polymorphus were poorly resolved, indicating that, as currently constituted, Polymorphus is not monophyletic (García-Varela et al. 2013), with the present material being placed genetically closer to P. brevis among the compared congeners. On the other hand, sequences of Profilicollis obtained in this study resulted in a well-supported clade with all sequences available of P. chasmagnathi (cystacanths and adults) from South American hosts. The low genetic divergence of P. chasmagnathi across host species was congruent with those values from previous studies in the region (Rodríguez et al. 2017; Lorenti et al. 2018). A diagnostic feature of the genus Profilicollis is the use of decapod crustaceans as intermediate hosts (Nickol et al. 1999, 2002) and up to the present, the lack of paratenic hosts, being directly transmitted to aquatic birds (its definitive hosts) through the consumption of crustaceans (Lorenti et al. 2018). Profilicollis chasmagnathi has been reported in several species of estuarine-dependent fishes (O. argentinensis, Paralichthys orbignyanus (Valenciennes, 1839), Micropogonias furnieri (Desmarest, 1823) and Pogonias cromis (Linnaeus, 1766)) from Mar Chiquita (Alarcos and Etchegoin 2010). The authors found these acanthocephalans in the gut of fishes, consequently, such infections should be considered as accidental and transient, being the result of predation on the decapod crabs inhabiting the lagoon. In fact, although crabs harbouring Profilicollis larvae are often predated by teleosts and elasmobranchs, it is considered that these fish do not play any role in the life cycle of the parasite (Oliva et al. 2008). In the present study, the presence of tissue lesions containing acanthocephalan hooks in the mucosa of some specimens of O. argentinensis, but not of attached worms, supports the idea of the transient nature of such infections in this hosts. This is supported by the morphometric similitude between worms from crabs and silversides, indicating that larvae were released from the crab’s body during digestion.

On the other hand, the presence of P. chasmagnathi in the mesenteries of O. jenynsii represents a long-term or permanent infection, with the fish acting as paratenic host, a symbiotic relationship that had not been reported yet for this genus. In the present study, crab remains were found neither in intestinal nor in stomach contents of any dientudo; however, the presence of cystacanths of P. chasmagnathi would indicate the previous consumption of infected crabs. Oligosarcus jenynsii is a generalist carnivorous and a freshwater species, tending to piscivory at larger sizes (Nunes and Hartz 2006), but freshwater decapods (shrimps and crabs) are frequently reported preys in other localities (Rodrigues et al. 2012). It is therefore possible that it feeds on estuarine crabs when this fish, visits myxo-oligohaline areas, at the north of the lagoon, mostly during winter (González-Castro et al. 2009). It is possible too that the infections occured in freshwater areas, thanks to the marked euryhalinity of the crabs N. granulata and C. angulatus, commonly parasitized by P. chasmagnathi, especially of the latter, which was abundant in Arroyo Grande (pers. obs.).

The absence of P. chasmagnathi in a large sample of O. jenynsii from Nahuel Rucá is therefore, explained by the rarity of both crab species in that lake. Indeed, only few specimens of C. angulatus were sporadically observed during fish sampling, all of them of a size too large to be prey of O. jenynsii (pers. obs.), supporting a crab-fish transmission in Arroyo Grande where small crabs can be eaten by the fish.

The generalized life-cycle pattern of acanthocephalans invariably includes an arthropod intermediate host (Kennedy 2006). In the case of polymorphids, these are crustaceans (García-Varela et al. 2013), which are infected by consuming the parasite’s eggs. After penetrating the gut wall of the crustacean host, the acanthor larvae develops into an acanthella and then into a cystacanth stage (Reish 1950; Rayski and Garden 1961). The acquisition of paratenic hosts, albeit facultative ones, is rare among acanthocephalans (Kennedy 2006); however, it enables them to ascend trophic levels and so move through food chains favouring transmission (Kennedy 2006) by bridging the trophic gap between intermediate and definitive hosts (Aznar et al. 2006). Beyond this evolutionary advantage, sometimes it is difficult to assess whether a host with acanthocephalans in the body cavity is actually a paratenic or an accidental host, in whose body cavity acanthocephalans may also occur (Kennedy 2006). A possible mechanism of the extra-intestinal infection by larval acanthocephalans in fish has been proposed for other genera; in fact, it is not uncommon to find partially excysted parasites in the body cavity and its organs of suitable definitive hosts (Kennedy 2006). According to this author, the age of cystacanths can play a role in the efficiency and site of infection. As examples, De Giusti (1949) and Nickol (1985) found that immature larvae could not effectively attach to their definitive fish hosts and would pass through the intestine wall to encyst (still as a cystacanth) in an extra-intestinal site, while older larvae are able to attach to the intestine and normally progress with their life cycle. This could be the case for those cystacanths found in dientudos, although they are not the definitive hosts, and could explain the smaller size of cystacanths infecting them relative to those found in crabs. Nevertheless, the fact that this fish could be a suboptimal host affecting the parasite’s development, as also indicated by the finding of dead worms, cannot be discarded.

Host capture, the colonization of new hosts by parasites (Holmes and Price 1980), seems to be an extended phenomenon in polymorphids, not only at the level of definitive (Kennedy 2006) and intermediate hosts (García-Varela et al. 2013), but also of paratenic hosts. Indeed, birds and mammals have captured Polymorphus species from fish, while marine mammals, may have captured Corynosoma and Bolbosoma species from birds (Kennedy 2006). Similarly, the acquisition of decapods as intermediate hosts for some genera (Profilicollis, Arhytmorhynchus Lühe, 1911, Ibirhynchus Garcia-Varela, Pérez-Ponce de León, Aznar & Nadler, 2011 and Hexaglandula Petrochenko, 1950) and of euphausiids for Bolbosoma, represent episodes of secondary colonization from amphipods, the ancestral intermediate hosts (García-Varela et al. 2013). Therefore, the incorporation of a paratenic host in the life cycle of Profilicollis, even in an apparently incipient stage, should not be surprising, since it has been frequent in related genera. The transmission of P. chasmagnathi cystacanths to a fish paratenic host implies their ability to penetrate the fish’s intestine and re-encyst in the mesenteries. This process is the same followed by the phylogenetically closely related genus Polymorphus (Alcántar-Escalera et al. 2013; Huston et al. 2020; Presswell et al. 2020), as well as by other polymorphids, such as Andracantha, Bolbosoma and Corynosoma (Aznar et al. 2006; García-Varela et al. 2013). On the other hand, the disability of P. chasmagnathi to parasitize the internal organs or tissues of other fish species, in whose guts is frequently found (Alarcos and Etchegoin 2010; present study), could represent a kind of host specificity and requires further research, although differential host defence mechanisms could be involved. The finding of dark lesions in the intestines of some silversides containing proboscis remains could be indicative of such defences. For example, the polymorphid Corynosoma strumosum (Rudolphi, 1802) may or may not respond secreting a protective thick layer of glycocalyx on its tegument (Skorobrekhova and Nikishin 2017) depending on the nature of the encapsulation defence mounted by the different species of paratenic fish hosts (varying from fibroblastic to leukocytal). In consequence, C. strumosum shows variable degrees of adaptation to disparate hosts (Skorobrekhova and Nikishin 2017; Nikishin and Skorobrekhova 2019).

Beyond the possible role of O. jenynsii as host of P. chasmagnathi could be considered as accidental or as an incipient paratenicity phenomenon, a high proportion of these larvae found were alive at the time of capture of these hosts. This implies that they could be viable for the infection of ichthyophagous birds, which would potentially enable P. chasmagnathi to widen its host range. Such facts are relevant, especially considering the low specificity of this parasite for its definitive hosts. Indeed, P. chasmagnathi has been reported in six families of birds, belonging to five orders (Martorelli 1989; Vizcaíno 1989; Torres et al. 1993; Diaz et al. 2011; La Sala et al. 2013; Lorenti et al. 2018), including podicipedids and phalacrocoracids, which are primarily piscivorous (Petracci et al. 2009; Josens et al. 2010) and abundant in Mar Chiquita basin (Favero et al. 2001; Ferrero and Iribarne 2001).

The capture of a fish paratenic host challenges the phylogenetic conservatism of the genus Profilicollis. However, host specificity cannot be considered a fixed trait, because in spite of being phylogenetically constrained to a large extent, it is strongly influenced by local environmental conditions (Mouillot et al. 2006), which cause considerable variation in realized host specificity (Wells and Clark 2019). Indeed, it has been proposed that estuaries and coastal-brackish lagoons are environments physically variable enough to select generalist genotypes of fish, in order they can adjust their morphology, physiology and behaviour to a wide range of conditions (Bamber and Henderson 1988). This selected plasticity would pre-adapt estuarine populations to invade, colonize and radiate into vacant niches in freshwater (Bamber and Henderson 1988), an eco-evolutionary mechanism that could explain a new host-parasite system, such as P. chasmagnathi-O. jenynsii.

The present findings represent the first record of the incorporation of a paratenic host in the life cycle of a member of the genus Profilicollis, and consequently an exception to the phylogenetic conservatism characteristic of this genus. Regarding the ecology of Profilicollis it would also imply a possible mechanism for the colonization of both a freshwater host and the freshwater environment, by an acanthocephalan genus ‘exclusive’ of marine and brackish habitats. This transition is probably promoted by the highly variable environmental conditions, typical of ecotonal environments between marine and freshwater realms, such as Mar Chiquita coastal lagoon.