Narcine electric rays are small- to medium-sized, bottom-dwelling members of the order Torpediniformes, distinguished by paired electric organs derived from modified branchial musculature. The genus is especially challenging taxonomically because many characters used in conventional identification overlap or vary within species. A major revision proposed that Western Atlantic populations represented Narcine brasiliensis, Narcine bancroftii and a potentially undescribed form ranging from Suriname to northeastern Brazil. The first species was associated with diffuse blotches and broad horizontal, oval or circular markings, whereas the second was characterized by ocelli and incomplete rings. Research has since shown that tooth-row counts and dorsal coloration can vary substantially, undermining simple assignment of individuals to named species.

For the new analysis, the team sampled fresh muscle from seven rays captured as bycatch in artisanal fisheries at seven localities: two in the Gulf of Mexico, one on the Mexican Caribbean, one near Santa Marta in Colombia, and three in Brazil. Initial identifications used pigmentation, ocelli, snout markings and geographic distribution. The selection deliberately included unusual forms, such as a Gulf of Mexico individual with no dorsal spots or ocelli and a Brazilian specimen resembling the putative undescribed Narcine form. DNA was extracted, fragmented, indexed and sequenced as paired-end Illumina reads before complete mitogenomes were assembled and annotated. The seven new sequences, deposited in GenBank as PZ347752–PZ347758, were analysed with five public reference mitogenomes.

Phylogenetic analysis combined all 13 mitochondrial protein-coding genes in both nucleotide and amino acid datasets. Under partitioned maximum-likelihood inference, the two approaches produced broadly congruent trees. Narcine timlei from the Indo-West Pacific formed the deepest branch, while the eastern Pacific species Narcine entemedor was sister to the Western Atlantic group. Within the Atlantic radiation, all seven newly assembled mitogenomes joined a Florida reference sequence in a single lineage. Relationships among individual Atlantic samples received moderate to high support, and differences between nucleotide and amino acid trees were confined to shallow nodes. The agreement across sequence types, tree methods and distance calculations indicates that the central signal was not confined to one marker or analytical model.

Species-delimitation analyses produced different numbers of mitochondrial operational taxonomic units, illustrating the difficulty of applying generic thresholds to shallow genetic variation. Automatic Barcode Gap Discovery returned between three and eight units depending on its assumed level of within-species divergence, while Assemble Species by Automatic Partitioning favoured three primary clusters. A multi-rate Poisson tree process identified four. Despite these methodological differences, all three approaches placed the newly sampled Atlantic rays and the Florida reference together, separated them from the two Colombian reference mitogenomes and recognised the external species as distinct lineages. The authors interpret these outputs as hypotheses about mitochondrial diversity rather than formal species diagnoses because every analysis used mitochondrial rather than nuclear DNA.

The mitogenomes themselves were highly stable in structure. Each of the seven new sequences was about 17,071 base pairs long and contained the canonical vertebrate complement of 13 protein-coding genes, 22 transfer RNA genes, two ribosomal RNA genes and a noncoding control region. Gene order and strand orientation were conserved, with ND6 and eight tRNAs on the light strand and the remaining components arranged conventionally. Base composition was also nearly uniform: adenine averaged 36.23%, thymine 29.99%, cytosine 21.94% and guanine 11.84%, giving an overall A+T content of about 66.2%. Positive AT skew and negative GC skew were consistent across the new Atlantic genomes, matching the compositional asymmetry widely observed in vertebrate mitochondria.

Selection appears to have constrained the functional parts of these genomes. Comparisons of nonsynonymous and synonymous substitution rates produced Ka/Ks values well below one for every protein-coding gene, a pattern consistent with strong purifying selection. The cytochrome c oxidase genes COX1–COX3 were among the most constrained, while ATP8 and several NADH dehydrogenase subunits generally had higher values but remained below unity. Codon usage was non-random and closely followed the genomes’ A+T bias, with A- or T-ending codons generally favoured. The 22 tRNAs ranged from 58 to 72 base pairs and usually folded into the classic cloverleaf structure; tRNA-Ser (AGN) was the standard exception, lacking the D-arm. In the control region, newly assembled genomes contained mainly 46-base-pair tandem repeats, sometimes accompanied by 23-base-pair units.

A striking anomaly emerged from two Colombian reference mitogenomes. Although identified in GenBank as different nominal species—N. brasiliensis and N. bancroftii—they differed from each other by only about 1.74–1.76%. Conversely, their divergence from the newly sampled Atlantic lineage reached 12.93–13.94% in uncorrected protein-gene distances and 14.32–15.56% with Kimura two-parameter correction. Those values were greater than the 6.82–7.22% separation between the Western Atlantic lineage and N. entemedor. Their exclusion from the main Atlantic cluster was stable across every phylogeny, delimitation method and distance analysis. The pattern could reflect misidentification, an unrecognised lineage or another taxonomic problem, but it also makes the reference sequences a warning against the uncritical reuse of database identifications without voucher support.

For now, the results support mitochondrial continuity among the newly analysed Western Atlantic Narcine populations and provide no mitochondrial evidence for multiple evolutionary lineages corresponding to their contrasting dorsal patterns. That conclusion is important because coloration has long been central to numbfish identification, and the analysis suggests that at least some visually distinct forms may represent phenotypic variation within a common mitochondrial population. The sample nevertheless remains small, and mitochondrial ancestry records history rather than necessarily reflecting recent species boundaries. Low elasmobranch mutation rates, incomplete lineage sorting, past demographic events or historical introgression could also obscure recent divergence. Expanded geographic sampling, direct morphological reassessment, nuclear genomic evidence and carefully curated voucher-linked sequences will therefore be essential for an integrative taxonomy of the Narcine complex.

Subject of Research: Mitochondrial lineage structure and species delimitation in Western Atlantic Narcine electric rays

Article Title: Mitogenomic and phylogenomic analyses identify a cohesive Western Atlantic lineage within the Narcine complex (Torpediniformes: Narcinidae)

Article References: Palacios-Barreto, P., Boza, B. R., da Cruz, V. P., da Silva Ferrette, B. L., Oliveira, C., Rotundo, M. M., Sandoval-Laurrabaquio-Alvarado, N., Adams, D. H., & Díaz-Jaimes, P. (2026). Mitogenomic and phylogenomic analyses identify a cohesive Western Atlantic lineage within the Narcine complex (Torpediniformes: Narcinidae). Molecular Biology Reports, 53(1), Article 1530. https://doi.org/10.1007/s11033-026-12694-x

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12694-x

Keywords: Narcine electric rays, Western Atlantic, mitochondrial genomes, mitogenomics, phylogenomics, species delimitation, taxonomy, morphological variation, colombian references, elasmobranch evolution, Mitogenomic, phylogenomic