4.1 Phylogenetic relationships
Phylogenetic trees obtained through single-locus mitochondrial data were almost identical since all mitochondrial markers are inherited as a single locus. In addition, and as expected, the combination of the mtDNA markers in a multi-locus dataset improved the statistical support of the phylogenetic trees obtained when compared with the single-locus ones (Janko et al. 2011).
The phylogenetic tree obtained with mtDNA showed longer branches and better statistically supported clusters than those obtained with nDNA. These results could be explained in a scenario of recent divergence among some of the species and by the higher substitution rate observed in mtDNA compared to nDNA (Brown et al. 1979). Nevertheless, both mtDNA and nDNA strongly acknowledged the basal position of G. mediterraneus and G. guttatus in the Gaidropsarus lineage in agreement with previous studies, with the remaining species in a second group (Francisco et al. 2014).
4.1 Gaidropsarus biscayensis – Gaidropsarus macrophthalmus
Previously, it has been proposed that G. biscayensis could be distinguished morphologically from G. macrophthalmus by the number of second dorsal fin rays and by the anal fin ranges (Iwamoto and Cohen, 2016). However, in further bibliographical revisions these magnitudes were found to overlap, invalidating them as distinctive (Barros-García et al. 2018). This scenario was previously observed in the synonymy between the Atlantic Lepidion eques (Günther, 1887) and the Mediterranean Lepidion lepidion (Risso, 1810), resulting in the latter as the only valid species with an Atlantic-Mediterranean distribution (Bañón et al. 2013; Barros-García et al. 2016). In both cases, G. biscayensis – G. macrophtalmus and L. eques – L. lepidion, the slight morphological differences recorded could be explained by the different environmental conditions, particularly temperature, between the Atlantic Ocean and the Mediterranean Sea (Bañón et al. 2013).
The analyses carried out in the present study showed a general agreement in considering G. biscayensis and G. macrophthalmus as a single species, despite the different approaches and dataset combinations tested. It is true that, since mtDNA has been included in the analyses, the possibility of hybridization/introgression events must be considered (Harrison and Larson, 2014). However, no evidence of mixed alleles has been found in the nuclear markers considered, so these possibilities can be discarded. Present results are in agreement with previous results obtained by combining morphological data and DNA Barcoding (Barros-García et al. 2018). Therefore, G. biscayensis should be considered a junior synonym of G. macrophthalmus, resulting in a single species with an Atlantic-Mediterranean distribution.
4.2 Gaidropsarus guttatus – Gaidropsarus mediterraneus
During their original descriptions, the high similarity between G. guttatus and G. mediterraneus was reported (Collet, 1905). Recently, a bibliographical revision of the morphological data available showed overlaps in the main characters measurements used to distinguish between these two species (Barros-García et al. 2018). Only colour patterns and distribution areas remained as distinguishing features (Cohen and Russo, 1979). Therefore, it is not surprising that a DNA barcoding analysis was not able to differentiate them, producing a single cluster where individuals of both species were mixed (Barros-García et al. 2018).
Far from helping to clarify the taxonomic problem of these two species, the nDNA analysis showed incongruent results when compared with mtDNA, misplacing some of these individuals with other species. It is well known that nDNA possess specific characteristics such as a lower mutation rate and higher effective population sizes than compared with mtDNA which slower the time required for divergence (Torres-Hernández et al. 2022). The different responses of the two sets of DNA data could be due to a variety of biological phenomena such as introgression of the mtDNA or incomplete lineage sorting at the nDNA level. Similar results with mixed individuals in nDNA based trees have been found in other fishes like Nemacheilidae (Cypriniformes) (Dvořák et al. 2022). Taking into account the mtDNA results, the former is to be considered unlikely, but further analyses will be required to clarify the evolutionary phenomenon behind these results (Kornilios et al. 2020).
In any case, the undeniable fact is that, apart from colouring, neither the examination of morphological characters nor the mtDNA sequences comparisons can distinguish between specimens of one species or another. Without prejudice to whether this phenomenon is due to adaptive processes of population type, or even to an imminent speciation event, the data provided seem to indicate that G. guttatus and G. mediterraneus are, in fact, a single species.
4.3 Gaidropsarus argentatus – Gaidropsarus ensis
The boreal Gaidropsarus argentatus and G. ensis are deep-sea species with an overlap in their North Atlantic Ocean distribution areas. Although their colouration and morphology are similar, they can be easily differentiated by the length of the first ray of the first dorsal fin, which is longer than its head in G. ensis and shorter in G. argentatus (Svetovidov, 1986b). Previous DNA barcoding and phylogenetic studies indicated a close relationship between these two species (Francisco et al. 2014; Barros-García et al. 2018). Interestingly, species delimitation analyses based on mtDNA clearly distinguished them, whereas nDNA markers did not, since phylogenetic analyses based on nDNA grouped them similarly to G. guttatus and G. mediterraneus In a recent speciation scenario, nDNA trees may fail to differentiate between species since not enough time has passed to accumulate substitutions in nuclear markers compared to the faster-evolving mtDNA genes (Brown et al. 1979). Therefore, this discrepancy observed in the species delimitation analyses performed with mitochondrial and nuclear DNA could be interpreted as evidence of a recent speciation phenomenon between G. argentatus and G. ensis.
4.4 Gaidropsarus granti – Gaidropsarus vulgaris
While morphological examination places G. granti as a species close to G. mediterraneus and G. guttatus (Svetovidov 1986a), molecular data place it close to G. vulgaris (Francisco et al. 2014; Barros-García et al. 2018). The latter two species can be differentiated from each other by their colouration pattern, habitat and distribution. Gaidropsarus vulgaris is found in the European continental shelf up to 120 meters of depth, while G. granti is a far more uncommon species, mainly inhabiting islands and seamounts at greater depths (Bañón et al. 2020). Analyses relying on mtDNA data clearly distinguished between the two species in a similar way to those performed for the boreal G. argentatus and G. ensis. Despite clustering almost all sequences together, nDNA indicated the independence of G. granti, which could be indicative of a speciation event older than in the previous case of the boreal species, in a way in which nDNA markers would have had some time to begin to acquire species-level substitutions (Brown et al. 1979). Since it is a rare species, the sample size of G. granti in this investigation is low and its related results should be interpreted cautiously. Nevertheless, the analysis of the data set used suggests that its status as an independent species must be maintained although in close relationship with G. vulgaris.
4.5 Gaidropsarus sp.
Six deep-water specimens of Gaidropsarus captured in two independent locations in the North Atlantic Ocean did not fit with any current description of Gaidropsarus species, but they showed common morphological features and, therefore, were appointed as Gaidropsarus sp. Both mtDNA and nDNA data, when employed in the phylogenetic and species delimitation analyses, indicated that this set of individuals belongs to the same species, different from the other eight nominal species considered in this investigation. Further analyses will be required, including morphological comparisons with the relatively unknown species from the South Hemisphere, to clarify if these specimens belong to a completely unknown Gaidropsarus taxon or if their capture constitutes evidence of the existence of a new distribution range in the North Hemisphere for a known southern species.
4.6 Final Remarks
The aim of this study was to shed new light on several taxonomic incongruities regarding the known Gaidropsarus species of the North Atlantic Ocean and the Mediterranean Sea. Thus, two synonymies among the already recognised species, and the presence of an unidentified species for the northern hemisphere have been detected. Gaidropsarus biscayensis and G. guttatus were described under the premises of the existence of phylogeographic barriers between the Mediterranean Sea and the Atlantic Ocean for the former, and between European continental waters and the Macaronesian region for the latter. These assumptions have led to an overestimation of Gaidropsarus species in the shallow waters of the North Atlantic Ocean and the Mediterranean Sea. Conversely, the recently discovered Gaidropsarus mauli (Biscoito and Saldanha, 2018), and the previously unknown specimens included in this investigation indicate the presence of a certain number of yet unknown Gaidropsarus species in the deep-sea ecosystems of the North Atlantic Ocean, despite the relatively good knowledge of the fish fauna of this geographic area (Barros-García et al. 2018; Bañón et al. 2021).
The existence of shared nuclear alleles among different Gaidropsarus species could be evidence of a far more complex evolutionary history than expected for this lineage of teleost fishes. Increasing the number of individuals from more areas and examining more regions of their genomes using next-generation sequencing techniques may be necessary to try to clarify the evolution, and thus the taxonomy, of the North Atlantic and Mediterranean species of the genus Gaidropsarus.