Is there hidden genetic variability in the species of Steindachneridion Garavello, 2005 (Siluriformes: Pimelodidae)?

Sampling and D-loop PCR

Muscle samples were collected for the DNA assessment: ten Steindachneridion scriptum specimens from Rio Uruguay, six from Rio Tibagi, seventeen S. melanodermatum specimens from Rio Iguaçu, three S. parahybae specimens from Rio Paraíba do Sul, and two S. doceanum specimens from Rio Doce (Fig. 1), following the procedures outlined by Sambrook, Fritsch & Maniatis (1989).

The amplification of the D-loop region was conducted using five ng of DNA, 1.5 mM MgCl2, 1x buffer, 200 mM of each dNTP, 0.5 U of Taq DNA polymerase, and 0.5 mM of each primer (FTTF: 5′CAA AGC GCC GGT CTT 3′ and F12R 5′GTC AGG ACC ATG CCT TTG3′) (Sivasundar, Bermingham & Ortí, 2001). PCR was initiated with five cycles, comprising denaturation at 94 °C for one minute, hybridization at 53 °C for one minute, and elongation at 72 °C for one minute and 30 s, followed by 25 cycles at 50 °C for the hybridization step. PCR reactions were conducted using a PTC-100 thermocycler from MJ Research (Waltham, MA, USA). PCR products were subjected to electrophoresis in a 0.8% agarose gel utilizing phage l as the molecular mass marker. Samples were purified using the GFX PCR DNA and Gel Band Purification Kit (GE HealthCare, Chicago, IL, USA), followed by sequencing using the DYEnamic ET Terminator Cycle Sequencing kit (GE HealthCare) as per the manufacturer’s instructions. Bidirectional sequencing of all templates was carried out on a Sanger ABI 3130 sequencer (Applied Biosystems, Waltham, MA, USA), with images documented by the Kodak Electrophoresis Analysis and Documentation System (EDAS) 290. Sequences of Steindachneridion scriptum from Rio Uruguay and Rio Tibagi (accession numbers EU930030–EU930044), S. melanodermatum from Rio Iguaçu (accession numbers PQ871622–PQ871638), S. parahybae from Rio Paraíba do Sul (accession numbers PQ871639–PQ871641), and S. doceanum from Rio Doce (accession numbers PQ871620–PQ871621) were deposited in GenBank.

This research followed the international standards of animal experimentation, approved by the Ethics Committee for Animal Use of the Universidade Estadual de Ponta Grossa (CEUA process number 0769342/2021). The collection was authorized by the Ministério do Meio Ambiente (MMA/ICMbio number 15115-1).

Genetic diversity analysis

Genetic distances within and between groups were estimated based on the Kimura 2-parameter evolution model (K2P) (Kimura, 1980). A neighbor-joining (NJ) tree (Saitou & Nei, 1987) was constructed from this model to represent divergences between species to depict species divergences, graphically, and clade support was assessed by 1,000 bootstrap pseudo-replicates according to (Felsenstein, 1985), using MEGA 10.2.2 (Kumar et al., 2018). Genetic diversity parameters (S—polymorphic sites; Nh—number of haplotypes; h—haplotypic diversity; π—nucleotide diversity) were evaluated utilizing DnaSP 6.12.03 (Rozas et al., 2017) and employed neutrality tests including Tajima’s D (Tajima, 1989), Fu’s Fs (Fu, 1997) and R ₂ (Ramos-Onsins & Rozas, 2002).

Statistical significance was determined via 10,000 permutations under the coalescing process implemented in DnaSP 6.12.03 (Rozas et al., 2017). To elucidate relationships between species of the genus Steindachneridion and the population of Steindachneridion melanodermatum studied, haplotype networks were reconstructed using PopART 1.7 (Leigh & Bryant, 2015) through the TSC method (Clement, Posada & Crandall, 2000).

Phylogenetic analysis of mtDNA D-loop sequences

For the phylogenetic analyses, we retrieved a D-loop sequence from Brachyplatystoma rousseauxii with accession number DQ779046 (Batista & Alves-Gomes, 2006) and another sequence from Zungaro zungaro (accession number EU930046.1), as an outgroup from GenBank. These D-loop sequences underwent alignment and editing using the Clustal W program (Thompson, Higgins & Gibson, 1994). Subsequently, a Bayesian Inference (BI) analysis was conducted employing MrBayes v. 3.2.7 (Ronquist & Huelsenbeck, 2003) through Markov chain Monte Carlo (MCMC) searches on two simultaneous runs of four chains for 10⁶ generations, sampling trees every 10³ generations. A burn-in of 25% of the sampled trees was implemented, and consensus topology and nodal support were computed from the remaining trees, estimated as posterior probability values (Huelsenbeck & Ronquist, 2001). Visualization and finalization of phylogenetic trees were executed using FigTree v. 1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/). Additionally, the PAUP* program (Swofford, 2002) was utilized to determine the best evolutionary model via Modeltest 3.7 (Posada & Crandall, 1998) under the hierarchical likelihod ratio tests (hLRTs). The maximum likelihood (ML) analysis was performed using MEGA 11.0.13 (Tamura et al., 2007), employing the HKY model with a gamma distribution, and bootstrap resampling with 501 replications to assess node support.

Genetic diversity of Steindachneridion

Nucleotide sequencing retrieved a total of 38 partial D-loop sequences of Steindachneridion scriptum samples collected from the Rio Uruguay and Rio Tibagi. The calculated average nucleotide diversity across these sequences was 0.065, and S. scriptum specifically exhibited a nucleotide diversity of 0.017. Comparing of the nucleotide diversity and haplotype count within S. scriptum populations revealed that the Uruguay River harbored a greater number of haplotypes (N = 10). In contrast, the Tibagi River displayed higher nucleotide diversity (N = 6).

For S. doceanum, the determined nucleotide diversity was 0.030, while S. melanodermatun exhibited a nucleotide diversity of 0.011. Notably, a nucleotide diversity of 0.130 was observed in S. parahybae. The calculated R ₂ values were substantially higher and statistically significant, indicating evidence of a departure from neutrality. However, in contrast, the Fu’s (Fs) test metric showed relatively high values, which, despite their magnitude, did not achieve statistically significance.

It is important to note that the R ₂ test has been acknowledged as more effective in detecting demographic events in smaller samples sizes (Ramos-Onsins & Rozas, 2002), as observed in the present study. Conversely, Fu’s (Fs) test tends to be more suitable for samples with more of individuals, as emphasized in the comparison presented in Table 1.

In the context of haplotype composition, our analysis revealed the identification of novel haplotypes within different species. Specifically, two new haplotypes (Haplo 1 and 2) were identified in S. doceanum, while three distinct haplotypes (Haplo37, Haplo38, and Haplo39) were found in S. parahybae. Moreover, a total of fifteen haplotypes were detected in S. scriptum, and sixteen haplotypes were observed in S. melanodermatum, labeled as Haplo28 to Haplo44 (as depicted in Figs. 2–3).

In the haplotype network analysis, a closer relationship was observed between S. scriptum populations from Rio Tibagi and Rio Uruguay, with only nine mutational steps separating these two populations. Comparatively, when contrasting S. melanodermatum with S. scriptum populations from the Uruguay River and Tibagi River, 13 and 18 mutational steps were identified, respectively. In addition, a considerable distance was found between S. parahybae and the other species, with 72 mutational steps from S. parahybae to S. melanodermatum and 61 mutational steps to S. doceanum.

Regarding the outgroup species, the analysis revealed significant genetic distances. Specifically, between S. melanodermatum and the outgroup species, Brachyplatystoma rousseauxii and Zungaro zungaro, 104 and 113 mutational steps were observed, respectively. These findings provide insights into the genetic divergence and relationships among the studied species, emphasizing notable genetic distances and relationships within and between species.

Phylogenetic analysis

The topology resulting from the analyses using the BI and ML methods showed significant similarities (Fig. 4). The genus Steindachneridion was determined to be monophyletic for the four species examined in this study, grouped in a highly robust clade supported by a high posterior probability. Steindachneridion scriptum from the Tibagi and Uruguay rivers exhibited consistent clustering into two distinct clades, supported by bootstrap and posterior probability values. Steindachneridion doceanum grouped with S. parahybae, while S. melanodermatum formed a single clade. The species B. rousseauxii and Z. zungaro clustered together, comprising the sister group to all the analyzed species.

Annotation and characterization of haplotypes of S. melanodermatum

Partial sequencing of the mitochondrial D-loop gene resulted in an array of 849 base pairs, situated between positions 15,651 and 16,500 within the mitogenome of S. melanodermatum (ON408064) (Fig. 5). The 27 additional D-loop haplotypes (Haplo1 to Haplo27) retrieved from references MZ672137–MZ672163 encompassed positions 15,590 to 16,460 within the mitogenome (Haplo1), constituting a sequence length of 870 base pairs (depicted in Fig. 3).

Alignment of these obtained sequences with those available in GenBank revealed an overlap spanning 809 base pairs (from positions 15,651 to 16,460), flanked by two non-overlapping fragments situated upstream (positions 15,590 to 15,650) and downstream (positions 16,461 to 16,500) the overlapping region (as illustrated in Fig. 4). Notably, these non-overlapping fragments represent non-conserved regions within the sequences and did not influence the identification of S. melanodermatum haplotypes.

This comprehensive alignment and analysis clarify the specific regions used for sequence comparisons, underscoring the conserved nature of certain segments and their role in facilitating the accurate identification of S. melanodermatum haplotypes despite non-overlapping fragments.

Genetic diversity unveiled: a glimpse into Steindachneridion is enigmatic variability

The investigation of genetic diversity among the four Steindachneridion species has revealed a clustering mosaic of haplotypic variation, exhibiting substantial diversity levels (h = 0.97–1.00—Table 1). It is noteworthy that our methodologies, which are analogous to those employed in studies on S. scriptum from the Uruguay River (Paixão, Ribolli & Zaniboni-Filho, 2018) and S. melanodermatum from the Iguaçu River (Souza-Shibatta et al., 2021), are in agreement with the findings of these investigations. This revelation assumes particular intrigue given the status of these species as endemics (Garavello, 2005), which are believed to inhabit deep wells and are facing imminent extinction threats (ICMBio, 2018).

The integration of haplotype network data with information obtained from phylogenetic analysis, conducted through BI and ML, revealed the presence of two distinct evolutionary units for the species Steindachneridion scriptum. These molecular operational taxonomic units (MOTUs) were identified in the Tibagi and Uruguay rivers, respectively (Figs. 1 and 3). This finding is consistent with prior research conducted in the Upper Uruguay River (Paixão, Ribolli & Zaniboni-Filho, 2018). The presence of substantial genetic structure, coupled with evidence of departure from neutrality (as indicated by the R2 statistic and Fs statistic’s results, respectively), underscores the complexity within these populations.

Remarkably, discovering of numerous new haplotypes across the four Steindachneridion species signals an untouched reservoir of genetic variability within the genus. While previous studies, such as the one by Fonseca et al. (2017), employing microsatellite markers, revealed low heterozygosity rates in S. parahybae, the conundrum of how intrapopulation genetic variability is maintained in Steindachneridion remains unresolved.

The emerging kinship dynamics among Steindachneridion species paint a clearer picture. Our haplotype network asserts a robust genetic affinity between S. melanodermatum and S. scriptum from the Uruguay River. These findings echo the conclusions drawn by Souza-Shibatta et al. (2021), who, though employing different methodologies (six microsatellite markers in a larger sample size (N = 95)), reveled analogous connections between these species. Despite S. melanodermatum’s seemingly uncompromised genetic diversity, clues pointing to a recent bottleneck were evident in microsatellite data (Matoso et al., 2011). This study expands on previous knowledge, revealing 16 novel haplotypes for S. melanodermatum, illuminating the pressing need for deeper investigations into the species’ genetic diversity.

Our discovery of several new haplotypes within the studied Steindachneridion species underscores the potential undervaluation of their genetic diversity. The cohesive groupings observed in the haplotype network, coupled with BI and ML analyses, accentuate each species’ distinctiveness as a single operational taxonomic unit (OTU), warranting due consideration for conservation efforts. Additionally, our findings, aligning with those by Fonseca et al. (2017) in S. parahybae, highlight the conundrum posed by the combination of low nucleotide diversity and high haplotypic diversity, hinting at a demographic surge following a period of diminished effective population size.

Unraveling evolutionary enigmas: insights into Steindachneridion is the genetic tale

The investigation of morphological data and the genetic model data from nuclear and mitochondrial sources has revealed a remarkable evolutionary history within the Pimelodidae family (Lundberg & Littmann, 2003). This history is evidenced by the existence of a monophyletic clade (Sullivan, Muriel-Cunha & Lundberg, 2013). Similarly this monophyly is supported for the species of Steindachneridion (Garavello, 2005), which places these fish at the base of the pimelodid lineage (Lundberg, Sullivan & Hardman, 2011). Such evidence calls for a deeper comparative and kinship investigation, particularly within the Steindachneridion genus. Leveraging BI and ML analyses on 38 mitochondrial D-loop gene sequences from our studied species yielded robust clades, providing high reliability indices (Fig. 3). This analytical exploration was made possible by the comprehensive mitochondrial genome of S. melanodermatum (Silva et al., 2024), which was pivotal for concatenation of nucleotide sequences in our phylogenetic algorithms for BI and ML (Fig. 4).

Our investigations, rooted in the topology of the Pimelodidae (Lundberg & Akama, 2005), showcased Brachyplatystoma rousseauxii and Zungaro zungaro as outgroups, delineating the phylogenetic relationships among examined Steindachneridion species, distinctly positioned outside these clusters. Internally, these clusters identified within the four Steindachneridion species under scrutiny stand as conclusive evidence, reverberating uniform biogeographic patterns.

The geological dynamics of the Guiana and Brazilian continental shields and the frainage patterns along South America’s east coast have significantly shaped the river systems (Ribeiro, 2006; Lundberg, 1998). Notably, macrodome uplift, rifts, vertical movements, and erosive actions along the east bank have been pivotal in shaping the distribution of freshwater ichthyofauna in these regions. The observed biogeographic patterns (Ribeiro, 2006) underscore taxa with endemic coastal distributions maintaining phylogenetic ties with neighboring ichthyofauna, hinting at Steindachneridion is integration within this context. These fish, harboring endemic distributions across coastal basins (three of their six species) while spanning the Upper Paraná basins, including the Iguaçu River and Uruguay River, exhibit kinship connections with Zungaro, Sorubim, Pseudoplatystoma, and the Amazon basin-dwelling Brachyplatystoma.

The observed clustering between S. doceanum + S. parahybae and S. melanodermatum + S. scriptum could be attributed, in part, to biogeographic features along Brazil’s east coast, perhaps associated with the proximity of these species’ river headwaters. Comparable findings in the 16S gene sequences of Hoplias malabaricus (Dergam et al., 2002) unveiled intertwined relationships across basins like Doce, Paraíba do Sul, and Grande, remnants of a shared past during the Plio-Pleistocene. The ichthyofauna of these basins is currently a reflection of a relict fauna that still bears similarities to one another (Torres et al., 2008; Menezes et al., 2008). Erosive retreats and headwater capture events might have sculpted the ichthyofauna relationships across these basins (Ribeiro, 2006).

Steindachneridion is included in the National List of Aquatic Vertebrates and Fish Threatened with Extinction since 2013 (ICMBio, 2018) highlighting its conservation urgency. Species like S. punctatum and S. amblyurum, absent in our study due to sampling difficulties, emphasize their precarious status and the challenges in preserving them.

Since the diversification of Siluriformes in the Neotropical region came before the division of South America from other continental blocks (Pinna, 1998). Thus, comparative data with taxa beyond the Neotropical region are imperative to achieve a more comprehensive phylogenetic and biogeographic scenario. Future research on Sorubiminae’s systematics promises to elucidate the phylogenetic intricacies of basal lineages within this group, potentially fortifying conclusions derived herein.

Four of the six known Steindachneridion species were central to our study. S. punctatum, unfortunately, is no longer found in the wild. Beyond contributing to the evolutionary narrative of this endangered endemic genus, our primary aim was to unveil the latent genetic variability within these species. Our crucial takeaway underscores distinct MOTUs among Steindachneridion species, advocating for their conservation. Remarkably, recent events might have influenced genetic differentiation, hinting at intraspecific variability suggestive of speciation processes, demanding systematic consideration, as exemplified in the case of S. scriptum.