Auchenipteridae belongs to the order Siluriformes and contains 25 genera and 126 species (Fricke et al., 2021) divided into two subfamilies: Centromochlinae (with nine genera) and Auchenipterinae, the latter including the species of Trachelyopterus Valenciennes, 1840 and Parauchenipterus Bleeker, 1862. Auchenipterids are small to medium-sized catfishes restricted to the Neotropical region. Several synapomorphies support the hypothesis of a monophyletic family, and its main characteristics are sexual dimorphism and internal insemination (Akama, 2004; Birindelli, 2014). Cytogenetic analyzes in this family are scarce and restricted to Ageneiosus Lacepède, 1803, Auchenipterus Valenciennes, 1840, Glanidium Lütken, 1874, Tatia Miranda Ribeiro, 1911 and Parauchenipterus (Tab. 1). Most species have 58 chromosomes (Ravedutti, Júlio, 2001; Fenocchio et al., 2008; Lui et al., 2010, 2013a), except for Ageneiosus and Tympanopleura Eigenmann, 1912 species that have 56 chromosomes (Fenocchio, Bertollo, 1992; Lui et al., 2013b), and Centromochlus heckelii (De Filippi, 1853) that has 46 chromosomes (Kowalski et al., 2020). Heterochromatin is distributed in the terminal regions of most chromosomes (Lui et al., 2010, 2015). Nucleolus Organizing Regions (NORs) are typically found on the short arm of one chromosome pair of most Auchenipteridae species studied (Lui et al., 2010, 2015). Exceptions to the most common in the group include Ageneiosus, Tympanopleura, Tatia neivai (Ihering, 1930) and Centromochlus Kner, 1858 (e.g., Fenocchio, Bertollo, 1992; Lui et al., 2013a,b; Kowalski et al., 2020). In Parauchenipterus, only P. galeatus (Linnaeus, 1766) has chromosomal data available, and all populations have 58 chromosomes (Ravedutti, Júlio, 2001; Lui et al., 2010; Araújo, Molina, 2013). However, small differences in the karyotype formula and in the AgNORs bearing pair were found. This diversity is supported by the putative chromosomal rearrangements, like pericentric inversions (Ravedutti, Júlio, 2001; Lui et al., 2009, 2010; Araújo, Molina, 2013). Moreover, according to Araújo, Molina (2013), this species is likely a species complex.

TABLE 1

Species	Locality	FN	2n	Karyotype formula	AgNORs/18S rDNA	5S rDNA	Ref.
Centromochlinae
Glanidium ribeiroi	Iguaçu River, Res. Salto Caxias, Paraná State	112	58	28m+16sm+10st+4a	pair 17, p, sm	-	1
	Iguaçu River, Res. Segredo, Paraná State	106	58	22m+16sm+10st+10a	pair 13, p sm	-	2
	Iguaçu River, Res. Salto Osório, Paraná State	106	58	22m+16sm+10st+10a	pair 13, p, sm	-	2
	Iguaçu River, Capanema, Paraná State	110	58	22m+20sm+10st+6a	pair 14, p, sm	pair 16, q, sm	3
Tatia neivai	Machado River, Denise, Mato Grosso State	116	58	26m+26sm+6st	pair 28, p, st	pair 4, p, sm / pair 21, p, sm pair 22, q, sm	4
Tatia jaracatia	Iguaçu River, Capanema, Paraná State	116	58	20m+26sm+12st	pair 28, p, st	pair 4, p, m / pair 18, p, sm pair 19, q, sm / pair 29, p, sm	4
Centromochlus heckelii	Solimões River, Manaus, Amazonas State	72	46	14m+6sm+6st+20a (♂) 15m+6sm+5st+20a (♀)	pair 20, p, a pair 13, p (W)	-	5
Auchenipterinae
Tympanopleura atronasus (cit. Ageneiosus atronasus)	Solimões River, Manaus, Amazonas State	100	56	16m+16sm+12st+12st	q, sm	-	6
Ageneiosus inermis (cit. Ageneiosus brevifilis)	Solimões River, Manaus, Amazonas State	102	56	20m+16sm+10st+10s	p, sm	-	6
	Araguaia River, Aragarças, Goiás State	108	56	32m+16sm+4st+4a	pair 20, p, sm	pair 4, p, m	7
Auchenipterus osteomystax (cit. Auchenipterus nuchalis)	Paraná River, Porto Rico, Paraná State	106	58	24m+14sm+10st+10a	pair 15, p, sm	-	1
Parauchenipterus galeatus	Paraná River, Porto Rico, Paraná State	98	58	22m+12sm+6st+18a	pair 23, p, a	-	1
	Paraná River, Três Lagoas, Mato Grosso do Sul State	108	58	24m+18sm+8st+8a	pair 25, p, st	pair 16, p, sm / pair 17, q, sm	8
	Piumhi River, Capitólio, Minas Gerais State	108	58	20m+16sm+14st+8a	pair 24, p, st	pair 15, p, sm / pair 16, q, sm	8
	São Francisco River, Lagoa da Prata, Minas Gerais State	108	58	22m+16sm+12st+8a	pair 23, p, st	pair 16, p, sm / pair 17, q, sm	8
	Pium River, Rio Grande do Norte State	108	58	24m+16sm+10st+8a	p, sm	-	9
	Araguaia River, Goiás State	108	58	20m+18sm+12st+8a	pair 23, p, st	pair 3, p, m	10
Parauchenipterus striatulus	Doce River, Minas Gerais State	106	58	18m+20sm+10st+10a	pair 23, p, st	pair 10, p, sm / pair 13, p, sm pair 15, q, sm	10
Trachelyopterus coriaceus	Araguaia River, Goiás State	108	58	20m+18sm+12st+8a	pair 24, p, st	pair 3, p, m / pair 16, q, sm	10

The lack of morphological diversity causes confusion in taxonomic classification in some fish groups and makes systematic organization difficult (for review, see Keat-Chuang Ng et al., 2017). The validation of Parauchenipterus has been the subject of extensive discussion for over two centuries, which has been considered valid by some researchers (Curran, 1989; Royero, 1999; Akama, 2004; Buckup et al., 2007; Graça, Pavanelli, 2007). Akama (2004) attempted to sort it out and pondered this genus valid, dividing it into two groups: “galeatus group”, composed of Parauchenipterus sp. n. and P. galeatus, which would have wide distribution occurring in the Amazon, Orinoco, Paraná-Paraguay, Guyana, São Francisco and Northeast Brazil drainages; and “striatulus group”, composed by Parauchenipterus ceratophysus (Kner, 1858), Parauchenipterus striatulus (Steindachner, 1877) and Parauchenipterus porosus (Eigenmann & Eigenmann, 1888), distributed in the Amazon basin, Paraná-Paraguay basin, drainages from eastern Brazil and possibly in the northeastern of Brazil. This classification was based on differences of osteological characters and male gonadal apparatus. Even though some researchers disagree with the validity of Parauchenipterus, following studies to the review by Akama (2004), suggest that Parauchenipterus is synonym of Trachelyopterus (Ferraris, 2007; Birindelli, 2014; Calegari et al., 2019).

The taxonomic conflict between Parauchenipterus and Trachelyopterus is due to the great morphological similarity between these two nominal genera. Therefore, three species of these genera were cytogenetically analyzed (P. striatulus from Doce River basin, and P. galeatus and Trachelyopterus coriaceus Valenciennes, 1840 from Araguaia River basin) to highlight the historical taxonomic problematic linked to this group of Neotropical fishes. In addition, the present work attempt to contribute to the taxonomic problem of Parauchenipterus, analyzing species from the “galeatus group” and “striatulus group”. These groups were defined in the last research, which focused in discussing Parauchenipterus validation (Akama, 2004), what did not happen until now.

Individuals of Parauchenipterus striatulus (7 males 5 females) were collected in Verde Lake, Doce River basin in Mariléia city, Minas Gerais State, Brazil, 19º49’44.5”S 42º37’52.5”W. Parauchenipterus galeatus (9 males 10 females) and Trachelyopterus coriaceus (6 males 9 females) were collected from the Marginal Lake to the Córrego do Medo, tributary of the Araguaia River in São Miguel do Araguaia city, Ana Maria farm, Goiás State, Brazil, 13º08’52.7”S 50º25’02.8”W, with fishing nets (Permanent License SISBIO 10538-1). The specimens were kept in aquaria and subsequently euthanized by clove oil overdose (Griffiths, 2000; Pereira-Da-Silva et al., 2009) (according to the Animal Experimentation Ethics Committee and Unioeste practical classes: 13/09 - CEEAAP/Unioeste), to remove tissues for cytogenetic preparations. The collected specimens were deposited in the Museu de Zoologia da Universidade de São Paulo, São Paulo, under the voucher numbers: MZUSP 109798 for P. striatulus, MZUSP 109793 and 110803 for P. galeatus, and MZUSP 106766 for T. coriaceus.

Metaphasic mitotic chromosomal cell suspensions were prepared with anterior kidney cells (Bertollo et al., 1978; Foresti et al., 1993). The best metaphases were selected for karyotyping. Chromosomal morphology was determined according to Levan et al. (1964). The heterochromatin distribution pattern was determined according to Sumner (1972) with modifications in the staining process (Lui et al., 2012). Nucleolus organizing regions (AgNORs) were identified using silver nitrate (AgNO₃) staining (Howell, Black, 1980). Both heterochromatin and NOR detection methods were carried out after Giemsa staining in order to follow a sequential analysis. Fluorescent in situ Hybridization (FISH) with 18S ribosomal (rDNA) and 5S rDNA probes was performed under stringency of 77% and followed the method by Pinkel et al. (1986), with modifications suggested by Margarido, Moreira-Filho (2008). 5S rDNA and 18S rDNA probes were obtained by DNA from Leporinus elongatus Valenciennes, 1850 (Martins, Galetti, 1999) and Prochilodus argenteus Spix & Agassiz, 1829 (Hatanaka, Galetti, 2004), respectively. Probes were labeled by nick translation with digoxigenin-11-dUTP (18S rDNA for P. striatulus; 5S rDNA for P. galeatus, and T. coriaceus) and biotin-16-dUTP (5S rDNA for P. striatulus; 18S rDNA for P. galeatus, and T. coriaceus), in accordance with the manufacturer’s instructions (Roche). Chromosomes were counterstained with DAPI (50 μg/mL). The slides were photographed through BX61 epifluorescence microscope (Olympus America Inc., Center Valley, PA, United States of America) with an attached Olympus DP71 digital camera and DP Controller 3.2.1.276 software.

Parauchenipterus striatulus: Doce River basin. Diploid number (2n) was 58 chromosomes and the karyotype was composed of 18 metacentric, 20 submetacentric, 10 subtelocentric, and 10 acrocentric chromosomes (Fig. 1A). AgNO₃ staining demonstrated simple NORs allocated in terminal region on the short arm of the submetacentric pair 23 (Fig. 1A, in box). C-banding technique evidenced pale heterochromatin in terminal regions of almost all chromosome pairs (Fig. 1D). 18S rDNA-FISH confirmed previous findings by AgNO₃ staining and revealed these sequences only on pair 23. 5S rDNA-FISH detected these sequences in three submetacentric pairs, on the short arm of the pair 10 and 13, and on the long arm of the pair 15 (Fig. 1G).

FIGURE 1

Parauchenipterus galeatus: Araguaia River basin. Diploid number was 58 chromosomes and karyotype was composed of 20 metacentric, 18 submetacentric, 12 subtelocentric, and 8 acrocentric (Fig. 1B). AgNO₃ staining revealed simple NORs allocated in the terminal region on the short arm of the submetacentric pair 24 (Fig. 1B, in box). C-banding evidenced pale heterochromatin in terminal regions of almost all chromosome pairs (Fig. 1E). 18S rDNA-FISH confirmed previous finding by AgNO₃ staining, simple NORs on pair 24. 5S rDNA-FISH detected these sites on short arm of the metacentric chromosome pair 3 (Fig. 1H).

Trachelyopterus coriaceus: Araguaia River basin. Diploid number was 58 chromosomes and karyotype was composed of 20 metacentric, 18 submetacentric, 12 subtelocentric, and 8 acrocentric (Fig. 1C). AgNO₃ staining revealed simple NORs allocated in terminal region on the short arm of the submetacentric chromosome pair 23 (Fig. 1C, in box). C-banding evidenced pale heterochromatin in terminal regions on almost all karyotype pairs (Fig. 1F). 18S rDNA-FISH also revealed only the pair 23 with these sites, corresponding to what was evidenced by the AgNO₃. 5S rDNA-FISH probe detected these sites on the short arm of the metacentric pair 3 and on the long arm of the submetacentric pair 16 (Fig. 1I).

The same diploid number of 58 chromosomes was found in all three species analyzed herein, which is considered basal for Auchenipteridae (Lui et al., 2013a, 2015), due to its prevalence within the species with available chromosome data (Tab. 1) and mainly because cytogenetic data in Doradidae (e.g., Eler et al., 2007; Milhomem et al., 2008; Baumgärtner et al., 2016; Takagui et al., 2017, 2019), sister-group of Auchenipteridae (Sullivan et al., 2006; Nelson et al., 2016). According Takagui et al. (2019), the most probable ancestor karyotype in Doradidae had 58 chromosomes. Ageneiosus inermis (Linnaeus, 1766) (Fenocchio, Bertollo, 1992; Lui et al., 2013b), Tympanopleura atronasus (Eigenmann & Eigenmann, 1888) (cited as A. atronasus) (Fenocchio, Bertollo, 1992) and C. heckelii (Kowalski et al., 2020) are exceptions, because they have lower diploid numbers. These reduced diploid numbers are probably result of fusion events (Lui et al., 2013b; Kowalski et al., 2020), which were confirmed by the presence of ITS (Interstitial Telomeric Sequence) in a metacentric chromosome pair detected in A. inermis (Lui et al., 2013b).

Small karyotype formula differences are found among the species analyzed in the present study. Parauchenipterus striatulus has higher amount of acrocentric (10 chromosomes) than P. galeatus and T. coriaceus (8 chromosomes). The presence of eight acrocentric chromosomes has already been reported in other three populations of P. galeatus (Lui et al., 2010), and their karyotype formulas are also very similar to the population from this study. Therefore, it shows irrelevant karyotype variation between these taxa, and these karyotype differences might be related to non-Robertsonian events, which do not imply diploid number changes. Nonetheless, the population of P. galeatus from Porto Rico in the Paraná River had 18 acrocentric chromosomes, and this difference is probably related to methodological problems, such as the chromosomal condensation pattern, which difficult the classification of chromosomes by different researchers, as already proposed for other fish groups (e.g., Carvalho, Dias 2005; Moraes-Neto et al., 2011). The two sympatric species (P. galeatus and T. coriaceus) had the same karyotype formula that indicates large chromosomal similarity between these two species. Moreover, researchers suggest that these species belong to different genera, supported by morphological data, as initially considered in this paper by Akama (2004); however, these chromosomal data reinforce the phylogenetic closeness between these two nominal genera.

Heterochromatin distribution was similar among the three analyzed species. Pale heterochromatin was found in terminal region of almost all chromosomes, as observed in other P. galeatus populations (Ravedutti, Júlio, 2001; Lui et al., 2009, 2010). This pattern is typically found in other Auchenipteridae species, such as Ageneiosus inermis (cited as A. brevifilis), Tympanopleura atronasus (cited as A. atronasus), Auchenipterus osteomystax (Miranda-Ribeiro, 1918) (cited as A. nuchalis), and Glanidium ribeiroi Haseman, 1911 (Fenocchio, Bertollo, 1992; Ravedutti, Júlio, 2001; Fenocchio et al., 2008; Lui et al., 2013b, 2015), suggesting to be a common family trait. On the other hand, small differences are found in some of these species: heterochromatin in terminal regions is strongly reported for A. inermis (Lui et al., 2013b); heterochromatin in centromeric region of chromosomes is observed for Tatia jaracatia Pavanelli & Bifi, 2009, and a conspicuous block in the interstitial region of a submetacentric pair is reported for T. neivai (Lui et al., 2013a).

AgNO₃ impregnation and FISH with 18S rDNA probe evidenced simple NORs in the terminal position on a subtelocentric pair 23 in P. galeatus and P. striatulus, and on a subtelocentric pair 24 in T. coriaceus. According to Ravedutti, Júlio (2001), simple NORs may be a common characteristic of Auchenipteridae, which has been confirmed by recent studies on this family (Tab. 1). Even though only one chromosomes pair bearing NORs is found in almost all auchenipterids, the location varies (terminal and interstitial), due to pericentric inversions, as aforementioned to explain the differences between karyotypic formulas. The three species in this study also had 18S rDNA present in subtelocentric chromosomes, as in most P. galeatus populations studied (Lui et al., 2009, 2010), however, there are populations with this marker located either in a pair of acrocentric (Ravedutti, Júlio, 2001) or submetacentric chromosomes (Araújo, Molina, 2013). As mentioned above, these differences might be associated with different chromatin condensation levels, which hinder their classification by researchers. Moreover, by analyzing the karyotypes of these last two papers, the pair bearing the NORs could have been classified as subtelocentric.

Despite the conservatism in the number of 18S rDNA sites for this family and small variation in the bearing pair, 5S rDNA has been found to be the chromosomal marker with the highest variability among Auchenipteridae species (Tab. 1). 5S rDNA has shown to be more dynamic, also evident in the data presented in this study. These cistrons were found in only one pair of chromosomes in G. ribeiroi and A. inermis (Lui et al., 2013b, 2015), but multiple 5S rDNA is more prevalent in the species of the genus Tatia: four pairs in T. jaracatia (metacentric pair 4, submetacentric pairs 18 and 19, and subtelocentric pair 29) and three pairs in T. neivai (metacentric pair 4 and submetacentric pairs 21 and 22) (Lui et al., 2013a). 5S rDNA submetacentric pairs reported for Tatia and Parauchenipterus species can be considered corresponding to the submetacentric pairs found in the species of this study, as they have similar measures, morphology, and position. Physical mapping of rDNA is considered a promising implement for evolutionary and taxonomic analyzes in fishes (Moraes-Neto et al., 2011), and the three species analyzed in this study indicate variation in the number of 5S rDNA bearing chromosome pairs. Therefore, this marker seems to be remarkable to understand chromosomal evolution of Auchenipteridae.

Regarding 5S rDNA data in Parauchenipterus and Trachelyopterus, FISH made it possible to identify differences among the three species of this paper. There are similar results found in other populations of P. galeatus, from different watersheds and all of these populations have two 5S rDNA submetacentric pairs (Lui et al., 2010). Trachelyopterus coriaceus also has two chromosome pairs carrying 5S rDNA sites, one metacentric and one submetacentric. Even though the morphology of these chromosomes is different from the two pairs of Parauchenipterus, these chromosomes can be considered corresponding. Similarly, P. striatulus also has two chromosomes (one with these clusters on short arm and the other one on the long arm) that might be corresponding to the same pairs. Moreover, the extra third 5S rDNA carrier submetacentric pair of P. striatulus, can be considered an autapomorphy of this species. It confirms that 5S rDNA is a good marker for the group, because these data differentiate P. striatulus from other phylogenetically close species. It is important to highlight that this species has distribution in coastal basins of the South American continent, different from P. galeatus and T. coriaceus, which are present in basins in the interior of the continent, thus there is no overlap in distribution and it suggests ancient diversification of P. striatulus from the other Trachelyopterus and explains the exclusive third 5S rDNA chromosome pair of this species.

Comparison of 5S rDNA distribution data among all populations of P. galeatus previously studied, T. coriaceus and P. striatulus shows greater similarity between P. galeatus from Araguaia River and T. coriaceus than with other P. galeatus populations. At a minimum, these data reinforce the phylogenetic proximity between these two nominal genera, and may even give rise to an interpretation contrary to Parauchenipterus validation. Thus, the differences in P. galeatus 5S rDNA data from Araguaia River basin, regarding the variation in number of sites from all other populations in this group, and the greater similarity with the sympatric population of T. coriaceus make it possible to suggest the existence of a new taxonomic unity. According to Birindelli et al. (2012), Trachelyopterus can be diagnosed by its gas bladder morphology among other characters. Trachelyopterus galeatus specimens analyzed showed differences in the gas bladder morphology (Birindelli et al., 2012), which could be interpreted as an indicative of cryptic diversity. Our data also suggest a population as a potentially new taxon in this fish group.

Considering the proposal of Akama (2004), the species belonging to Trachelyopterus analyzed in this study are the first report of cytogenetic data for this genus, and also the first record for a P. striatulus population. Comparing the data of Parauchenipterus and Trachelyopterus described here with those already found in the literature, it is possible to note the great similarity concerning chromosomal markers (diploid number, heterochromatin distribution, number and location of nucleolus organizing regions) and differences in 5S rDNA number and location. This way, our data reaffirm that Parauchenipterus should be considered synonymous of Trachelyopterus, contrary to what was suggested by last diagnosis reviewed of Parauchenipterus and closely related species (Akama, 2004), but in agreement with some other more recent studies in Auchenipteridae that were not focused on this issue (Ferraris, 2007; Birindelli, 2014; Calegari et al., 2019). It is noteworthy that maintaining Parauchenipterus as a junior synonym of Trachelyopterus is based on the greatest similarity between the sympatric species P. galeatus and T. coriaceus, compared to all others Auchenipteridae previously studied, mainly Trachelyopterus species, through the following chromosomal data: equal karyotype formula and metacentric pair 3 with marking on the short arm bearing rDNA 5S (for more detail, see Tab. 1). The 5S rDNA is an excellent marker to understand chromosomal evolution of Auchenipteridae and helped to suggest the existence of a new taxonomic unit distributed in the Araguaia River basin, which reaffirms the status of species complex in P. galeatus.