Loricariidae is one of the most species-rich fish families, with 1,056 valid species grouped in 115 genera (Fricke et al., 2024). The broad species diversification in Loricariidae makes it difficult for accurate determination of species relationships, leading to doubts in subfamilies, tribes, and genera (Isbrücker, 1980; Armbruster, 2004; Reis et al., 2006; Lujan et al., 2015; Covain et al., 2016; Pereira, Reis, 2017; Roxo et al., 2019). Another crucial problem in Loricariidae is the taxonomic uncertainties that occur in paired species belonging to isolated hydrographic basins (Anjos et al., 2021; de Sousa et al., 2021; Lustosa-Costa et al., 2022) or species resident in tributaries and with low vagility (Albert et al., 2020; de Sousa et al., 2021).

The genus Isbrueckerichthys (Loricariidae: Siluriformes) was proposed by Derijst (1996), justified on the discovery of an earlier type-species designation by Regan (1920), where Pareiorhaphis duseni (Miranda Ribeiro, 1907) and Pareiorhaphis alipionis (Gosline, 1947) were moved to the new genus. Their representatives are small-sized species of armored catfishes distinct from the sister taxon Neoplecostomus Eigenmann & Eigenmann, 1888 by several diagnostic morphological characteristics, such as a small naked area behind the pterotic-supracleithrum, abdomen with small platelets embedded in the skin between pectoral girdle and pelvic-fin insertions, one spine and seven branched rays in the dorsal fin, caudal peduncle ovoid in cross-section, and lacking a series of papillae on the lower lip of the dentaries (Derijst, 1996; Pereira, Oyakawa, 2003; Jerep et al., 2006). Isbrueckerichthys species inhabit headwater streams, usually rapids (<1 m deep), with well-oxygenated water and substrates consisting of rocks (Pereira, Reis, 2002). The first species recognized as Isbrueckerichthys (I. duseni, I. alipionis, and I. epakmos Pereira & Oyakawa, 2003)occur in the plain coastal lands of the Atlantic and are restricted to the Ribeira de Iguape River basin (Derijst, 1996; Pereira, Oyakawa, 2003; Dagosta et al., 2024). Afterward, Jerep et al. (2006) described two species (I. saxicola Jerep, Shibatta, Pereira & Oyakawa, 2006and I. calvus Jerep, Shibatta, Pereira & Oyakawa, 2006) from upland rivers of the Tibagi basin, belonging to the upper Paraná River.

For some authors, a taxonomic system based strictly on morphology underestimates the cryptic species diversity, which is common in many groups (Knowlton, 1993; Jarman, Elliott, 2000; Hebert et al., 2003). The essence of this limitation is the phenotypic plasticity and the genetic variability in the characters that can lead to erroneous identifications (Hebert et al., 2003). In Neotropical fishes, molecular analyses involving species delimitation or population biology have contributed to the understanding of distribution, taxonomy, and systematics in several groups (do Nascimento et al., 2018; Santos et al., 2019; Argolo et al., 2020; Herrera-Collazos et al., 2020; Traldi et al., 2020; Ariza et al., 2022; Azambuja et al., 2022; Silva-Santos et al., 2023; Souza et al., 2023). Previous studies have stipulated parameters for standardizing values for genetic divergence for molecular markers utilized for species identification (Ward, 2009). For several authors, the interspecific divergence must be ten times higher than the intraspecific divergence for a taxa validation (Hebert et al., 2004). Ward (2009) proposed that fish specimens that are strictly related and present genetic divergences above 2% belong to different species. Although the genetic divergence values associated with molecular markers used for molecular identification are standardized for some organisms, it is known that the trigger values for intra/interspecific divergence are variable and are influenced by the variation in the substitution rate through the lineages (Barraclough et al., 2009).

In this context, integrative taxonomy/phylogenetics is a contemporary strategy addressed to questions in which single data-based approaches reveal less consistent conclusions rather than those by combined multisource evidence in terms of detecting the species richness within supposedly homogeneous taxa (Padial et al., 2010; Schlick-Steiner et al., 2010; Grković et al., 2017). Distinct species delimitation methods, such as ABGD, ASAP, and bPTP, have been proposed to analyze a single locus (Pons et al., 2006; Zhang et al., 2013). In general, these analyses have demonstrated that the combination of different methods, as well as integrative taxonomy, allows more representative results for the identification of lineages in species (Tang et al., 2014; Kapli et al., 2017; Luo et al., 2018; Serrano et al., 2018; Cardoso et al., 2023; Souza et al., 2023). Isbrueckerichthys is a Loricariidae genus with reformulations and species descriptions in the last three decades (Derijst, 1996; Pereira, Oyakawa, 2003; Jerep et al., 2006). However, molecular species delimitation studies are lacking in the genus. Therefore, this study aimed to conduct a molecular delimitation analysis in all Isbrueckerichthys species inhabiting the lowlands and uplands in south and southeast Brazilian drainages.

Biological samples. Specimens of Isbrueckerichthys were collected in different tributaries of the lowlands from the Ribeira de Iguape basin and in tributaries of the uplands from the Tibagi basin (Fig. 1). Specimens were deposited as vouchers in the ichthyological collection at Núcleo de Pesquisas em Limnologia, Ictiologia e Aquicultura of Universidade Estadual de Maringá, Maringá (NUP) and in the Laboratório de Biologia e Genética de Peixes of the Universidade Estadual Paulista “Júlio de Mesquita Filho”, Botucatu (LBP) (Tab. 1).

Molecular analyses. Genomic DNA of thirty-one individuals of Isbrueckerichthys were extracted from liver samples using the CTAB (cetyltrimethylammonium bromide) method (Murray, Thompson, 1980). The barcode region of the mitochondrial gene cytochrome c oxidase subunit I (COI) was amplified by PCR using the primers Fish F1 and Fish R1 (Ward et al., 2005). The reaction was composed of 1x Taq Reaction buffer (200 mM Tris pH 8.4, 500 mM KCl), 0.2 mM dNTPs, 1 mM MgCl2, 1 U of Taq DNA polymerase (Invitrogen), 0.4 µM of each primer, and 40 ng of DNA. The following reaction program was used: initial denaturation for 10 min at 94 °C, 35 cycles of 94 °C for 1 min, 51.8 °C for 45 sec and 72 °C for 1 min, and a final extension at 72 °C for 10 min. PCR products were purified with the Illustra GFX PCR DNA and Gel Band Purification (GE Healthcare) and sequenced in an ABI-Prism 3500 Genetic Analyzer (Applied Biosystems).

Sequences were analyzed using Geneious v. 7.1.9 (Kearse et al., 2012) software and submitted to GenBank. Five sequences of I. saxicola (GU701901 – GU701903; GU701905 – GU701906) from the Jacutinga River were recovered from the GenBank. The sequences were aligned with the ClustalW algorithm integrated into the Geneious program. Nucleotide (π) and haplotype (Hd) diversity were computed by DnaSP v. 5 (Librado, Rozas, 2009). Sequences were separated into groups, and genetic distances were calculated in MEGA X (Kumar et al., 2018) under the Kimura-2-Parameters evolution model and 1,000 bootstrap replications. A haplotype network was generated through the median-joining criterion (Bandelt et al., 1999) in the PopArt v. 1.7 software (Leigh, Bryant, 2015). Structural analysis was performed by assigning each individual to the respective populations using Bayesian Analyses of Population Structure – BAPS 6, allowing for a range of 1–6 (Corander et al., 2004, 2008). Population structuring and analysis of molecular variance (AMOVA) (Excoffier et al., 1992) were performed by Arlequin v. 3.5.2.2 software (Excoffier, Lischer, 2010).

One Pareiorhaphis splendens (Bizerril, 1995) (Neoplecostomini) COI sequenceEU359449.1 was aligned to Isbrueckerichthys with ClustalW and used to root trees, according to previous molecular phylogenetic studies (Roxo et al., 2012). The alignment was submitted to jModelTest 2 (Darriba et al., 2012) using the corrected Akaike information criterion (AICc) to select the best-fit nucleotide evolution HKY model for downstream analyses. A Bayesian inference tree was generated in the MrBayes 3.2 program (Huelsenbeck, Ronquist, 2001; Ronquist, Huelsenbeck, 2003), applying 100,000,000 interactions of MCMC, sampling trees every 10,000 generations and burn-in of 10,000,000.

Three methods for species delimitation were applied: Automatic Barcode Gap Discovery (ABGD) (Puillandre et al., 2012), Assemble Species by Automatic Partitioning (ASAP) (Puillandre et al., 2020), and Bayesian Poisson Tree Processes (bPTP) (Zhang et al., 2013). The alignment generated for the sequences was used as an input file for the delimitation of species in the ABGD web server (https://bioinfo.mnhn.fr/abi/public/abgd/abgdweb.html) and in the ASASP web server (https://bioinfo.mnhn.fr/abi/public/asap/asapweb.html), the Kimura (K80) TS/TV was selected for distance mode and the other parameters were kept at default. For bPTP analysis, the Bayesian tree generated in MrBayes was used as an input file in a bPTP web server (https://species.h-its.org/ptp/). There were 500,000 MCMC generations run (thinning = 500), and the other parameters were kept at default.

Sequences of the COI gene were obtained from the Isbrueckerichthys specimens and deposited in GenBank (PP391830–PP391860). All sequences were of high quality and presented no evidence of insertions, deletions, or stop codons. The obtained sequences were aligned with the sequences of I. saxicola from the GenBank database, and a matrix of 36 sequences with 612-bp was generated. The nucleotide diversity (π) was 0.01369, and haplotype diversity (Hd) was 0.806. A total of ten haplotypes were observed: Isbrueckerichthys alipionis had four haplotypes; I. calvus and I. saxicola shared a unique haplotype; individuals of I. duseni and I. cf. duseni had one exclusive haplotype to each species; and in I. epakmos three haplotypes were observed (Fig. 2A).

The intraspecific distance for each group ranged from 0 to 0.23%, and the interspecific distance ranged from 0 to 3.63% (Tab. 2). The BAPs results indicated an optimal partition of four clusters (k = 4; marginal log-likelihood = -240.4536). Individuals of I. calvus and I. saxicola were assigned as a single group, I. duseni,and I. cf. duseni were separated into a second cluster, and I. epakmos and I. alipionis represent independent groups (Fig. 2B). The AMOVA analysis, considering the clusters delimited by BAPs, shows a variance value of 83.98% among groups (Tab. 3), with an FST value of 0.96692 (p < 0.05). The pairwise FST values between each population ranged from 0.00 to 1.00 (Tab. 4).

The Bayesian phylogenetic tree revealed five supported clades: I. epakmos, I. alipionis, Isbrueckerichthys cf. duseni, I. duseni, and I. calvus + I. saxicola (Fig. 2C). Species delimitation methods returned different results (Fig. 2C). The ABGD method resulted in seven partitions ranging from two to nine species, with partition 4 (prior maximal distance P = 0.004642) indicating six species, including the outgroup. ABGD recognized I. epakmos, I. alipionis, I. duseni,and Isbrueckerichthys cf. duseni as single groups and grouped I. calvus and I. saxicola. The best partition identified by the ASAP method (ASAP-score: 3.50; p-value: 1.74e-02; W: 3.13e-05; Threshold dist.: 0.004094) recognized the same six species as the ABGD method. The ML solution of bPTP delimited eight species, including the outgroup. bPTP separated I. alipionis into three species and grouped I. calvus and I. saxicola.

Evolutionary biologists agree that species are lineages or metapopulations that evolve separately, with disagreements remaining only about where, along the continuum of divergence, separate lineages should be considered distinct species (Wiley, 1978; Hey, 2006; Mallet, 2008). Dias Filho et al. (2020) argued that molecular techniques should complement classical taxonomic identification for a detailed description of organisms. Data analysis showed a restricted relationship among Isbrueckerichthys species, with some divergences concerning morphological identification for the samples obtained in the delimitations between the First (lowlands in Ribeira de Iguape basin) and the Second (uplands in Tibagi basin) Paraná State plateaus, Brazil.

The data obtained from COI sequences analysis demonstrated that I. epakmos, I. alipionis, and I. duseni, all inhabitants of the Ribeira de Iguape River basin, present K2P divergence values greater than or close to 2%, well-supported branches in the phylogenetic tree, and pairwise Fst close to 1, corroborating to the absence of gene flow and the occurrence of the nominal taxa. The Ribeira de Iguape River basin is in an intricate mountainous region with high isolating potential for aquatic organisms (Ribeiro, 2006), and Roxo et al. (2012) proposed that the ancestor of Isbrueckerichthys inhabited tributaries from this basin at least 20 million years ago. Langeani (1990) stated that most Neoplecostomini (Neoplecostominae in that work) species have low vagility and difficulties adapting to streams that occur at low altitudes due to the low oxygen concentration in the water compared to rivers in higher regions. Thus, streams in valleys between mountains (low altitude) could hinder the dispersal of Neoplecostomini species into different tributaries of the Ribeira de Iguape River basin (Fig. 1). These geological and ecological factors and evolutionary times allow for congruence between the molecular and morphological divergence data for I. epakmos, I. alipionis, and I. duseni.

The methods for species delimitation included I. cf. duseni (D’Areia stream), sampled in the border of the Ribeira de Iguape basin, the limit with Tibagi basin in Paraná State, as a new taxon for Ribeira de Iguape basin. Comparing specifically I. dusenivs.I. cf. duseni, 0.49% of K2P genetic distance was observed. However, the pairwise Fst showed gene flow restriction, and branches well-supported in the phylogenetic tree topology for I. dusenivs.I. cf. duseni were shown. Some Neotropical fish species demonstrate distinct taxonomic entities with genetic distances lower than 2%, although they are arranged in monophyletic groups (Bellafronte et al., 2013; Traldi et al., 2020). Based on molecular data, several studies have highlighted cryptic diversity for neotropical freshwater fish (Melo et al., 2016; Guimarães et al., 2019; Mateussi et al., 2019; Souza et al., 2021, 2023; Cardoso et al., 2023). In addition, as mentioned above, the specimens of Isbrueckerichthys could have gene flow restriction due to their biological features associated with the geological formation of the Ribeira de Iguape tributaries. There is a valley between the sample localities of I. duseni and I. cf. duseni (Fig. 1), which, according to Langeani (1990), makes dispersal difficult and, in this case, could contribute to genetic isolation. The data suggests that a molecular operational taxonomic unit for I. cf. duseni in the Ribeira de Iguape basin or a case of incipient speciation should be considered.

The genetic comparison among I. duseni, I. cf. duseni, I. saxicola,and I. calvus also demonstrated low K2P genetic distance (0.49–0.66%), high Fst, and branches well-supported in the phylogenetic tree. Despite low K2P values, I. duseni and Isbrueckerichthys cf. duseni are geographically isolated from I. saxicola and I. calvus by the Ponta Grossa Arch, an uplift of the relief that isolated the Ribeira de Iguape basin from upland rivers (Fig. 1), such as Tibagi basin (Ribeiro, 2006). Isbrueckerichthys duseni and I. cf. duseni were considered valid taxa in species delimitation methods, but a morphological and molecular data incongruence was observed between I. saxicola and I. calvus. These two species were described from the Tibagi basin by Jerep et al. (2006) having morphological differences with congeners including: bicuspid teeth, hypertrophied odontodes, and a more extended spine in the pectoral fin. The morphological differentiation between I. calvus and I. saxicola is based on the number of odontodes on the abdominal platelet (six or more in I. saxicolavs. up to six in I. calvus), by the absence/presence of a flattened area in the inferior portion of the two first plates of the lateral line, by the degree of exposition of the cleithrum and by the degree of convexity and exposition of the supraoccipital plane (Jerep et al., 2006). Although the morphological description of Jerep et al. (2006) is well-detailed, the morphological divergence is not accompanied by a genetic divergence. All methods based on the molecular species delimitation used in this study, using a single locus analysis, agree with I. saxicola and I. calvus as a single taxon. In addition, a pairwise Fst 0 was estimated between the localities of the I. saxicola and I. calvus samples, indicating no population structure.

In turn, it has been proposed that recent speciation in newly inhabited areas may present an intermediary stage of speciation due to the short space of evolutionary time for diversification, especially cases of incomplete lineage sorting and lack of divergence (Roxo et al., 2012, 2014, 2017). Neoplecostomini species occurring in the uplands of the Tibagi basin, like the case of Isbrueckerichthys, were proposed to inhabit this new area from dispersion of fishes through headwaters capture events between the coastal regions and those more elevated in the borders of the crystalline shield (Ribeiro, 2006; Roxo et al., 2012, 2014; Dagosta et al., 2024). Roxo et al. (2012) also suggested that the dispersal occurred around 4.8 (2.3–7.8), 5.9 (2.9–9.5), and 10.6 (5.5–16.5) million years ago. Based on this information, it is possible to verify that the morphological diversification between I. calvus and I. saxicola is not accompanied by a rate of nucleotide substitution, probably due to the short time of dispersal and distribution of Isbrueckerichthys in the Tibagi River basin.

Even with the emergence of combined multiple data sets strategies applied to diversity analysis, according to some authors, it is a challenge to evaluate species delimitation in considering demes that comprise geographically widespread series of similar individuals (de Queiroz, 2007; Willis, 2017), or those that emerged from a short period of dispersal in evolutionary diversification. However, also is expected a degree of phenotypic (morphological, ecological, and behavioral) divergence among sub-populations due to genetic drift and local adaptation (López-Fernández et al., 2005; de Queiroz, 2007; Zapata, Jiménez, 2011; Willis et al., 2017). At this point, the morphological differences between I. calvus and I. saxicola could be explained by their dispersal in distinct tributaries (environments) in the Tibagi basin. To solve these cases of incomplete lineage sorting and lack of divergence, as here demonstrated in Isbrueckerichthys,some authors argue that a multilocus analysis, phylogenomics methods, and multiple datasets are important for both phylogenetic resolution and in-depth investigation of speciation (López-Fernández et al., 2005, 2010; Ilves et al., 2018).

Integrative studies of molecular and morphological investigations may be considered for species delimitation because they represent complementary approaches practical for the group’s phylogeny, systematics, evolution, and phylogeography. However, groups with recent diversification (as revealed by similar or identical genetics coupled with morphological differentiation) still challenge integrative taxonomy. Here, there is low genetic divergence among Isbrueckerichthys that inhabit the border of Ribeira de Iguape basin in Paraná State and those from uplands in the Tibagi basin, which can be considered recent speciation cases. In I. calvus and I. saxicola in the Tibagi basin, molecular analysis indicates that morphological diversification between the species is not accompanied by genetic diversification.