Understanding the patterns and processes underlying the origin and distribution of lineages, both at intra- and supra- specific levels, is fundamental for testing taxonomic hypotheses (Szymura and Barton 1986; Babik et al. 2005; Dufresnes and Martínez-Solano 2020; Sánchez-Vialas et al. 2020, 2021; Kamiński et al. 2021, Sainz-Escudero et al. 2022). In this context, phylogenetic and phylogeographic analyses have become key tools for delimiting evolutionary units, contribute to resolving long-standing taxonomic problems (Mas-Peinado et al. 2021; Sainz-Escudero et al. 2021; Cuesta-Segura et al. 2023).

Complex evolutionary dynamics, encompassing hybridization among morphologically similar species or lineages have usually driven to intense taxonomic debates (Dufresnes et al. 2020; Doniol-Vacroze et al. 2021). Gene flow across well-defined species can determine different outcomes along the speciation continuum, including reinforcement (strengthening preexisting reproductive isolating mechanisms between lineages) (Servedio and Noor 2003; Pfenning 2016), melting pots (Dinis et al. 2019), lineage fusion (Taylor et al. 2006), instant speciation (Ramsey and Schemske 1998; Gonçalves and Brito 2020), and introgression (Zieliński et al. 2013; Racimo et al. 2015). Within this context, the understanding of these processes is required to perform solid taxonomic revisions.

The extend of hybrids zones and the fate of the parental lineages depends on the strength of the barriers to gene flow (Dufresnes et al. 2020). In this sense, restricted and localized introgression is consistent with strong selection against hybrids, which also is of paramount importance to determine the taxonomic status of the evolving lineages (Dufresnes et al. 2020). The hallmark of gene flow between closely related species can be noticed as cyto-nuclear discordances (Wielstra and Arntzen 2020). This phenomenon is well-documented in regions of climatic instability, where range shifts promote secondary contact and lineages admixture (Hewitt 1999, 2004). This is the case of the western Palearctic, where cyclic climatic oscillations during the Pleistocene had a significant influence on the genetic diversity and distribution of the local biota (Hewitt 2004; Gómez and Lunt 2006). The integration of phylogeographic hypotheses and congruence analysis enables more accurate taxonomic investigations of morphologically difficult to interpret groups.

Scaurus Fabricius, 1775 (Coleoptera: Tenebrionidae) is a group of thermophilic darkling beetles comprised by 25 species widespread through the Mediterranean basin, with extensions into the Atlantic and Saharan regions (Escalera 1914; Labrique 2009, 2020; Iwan and Löbl 2020, Nabozhenko et al. 2020). Some species of this genus exhibit partially overlapping distributions, making them an excellent system for comparative phylogeographic studies. This applies to Scaurus uncinus (Forster, 1771) and S. gigas Waltl, 1835, two species with partially sympatric distributions across both sides of the Gibraltar Strait (Cartagena et al. 2002; Cartagena and Galante 2005; Español 1957, 1960, 1963; Labrique 1995a, 1995b). The former species extends over the western Mediterranean basin and the Atlantic coast of Morocco, whereas the latter, S. gigas, is restricted to the Atlantic coastal plains from the Oued Draa in Morocco to the southwestern corner of the Iberian Peninsula (Español 1960; Labrique 2004). Although no seasonal differences have been identified between these species, habitat segregation may be occurring; S. gigas inhabits sandy coastal areas and does not venture into inland regions, whereas S. uncinus occupies a broader range of open habitats and is rarer in coastal plains (Español 1960; López-Pérez 2010). Idiosyncratic patterns of distribution across the Gibraltar Strait have been reported in ecologically similar species of Tenebrionoidea, such as in Pimelia Fabricius, 1775, in which for some species the Strait of Gibraltar act as a geographical barrier but clearly not for others (Mas-Peinado et al. 2021). However, whether this barrier accounts for phylogeographical structuration in Scaurus is unknown.

The genus Scaurus is characterized by a conserved general morphology, which resulted in a complex taxonomic history for several species, including S. uncinus and S. gigas (see Labrique 2004). Whereas S. uncinus presents a relatively uniform morphology across most of its distribution, the southern Moroccan populations of S. gigas are diagnosable from other conspecifics (Español 1943, 1960). These populations were initially classified as subspecies, named Scaurus uncinus ferreri Español, 1943, based on specimens from Sidi Ifni, but were later transferred as a synonym of S. gigas (Labrique 2004). Currently, S. gigas encompasses two additional synonyms from northern Morocco: Scaurus hespericus (Solier, 1838) (Type locality: Tanger, Mahon and Gibraltar) and Scaurus maroccanus (Fairmaire, 1875) (Type locality: Mogador). Despite their partly overlapping geographic ranges on either sides of the Gibraltar Strait, no comparative phylogeographic studies have been conducted between the species pair S. uncinus and S. gigas, and consequently, their evolutionary history remains enigmatic. Given these shortfalls, a revision of the diversity and distribution of S. uncinus and S. gigas across most of their range is needed.

In this work, we perform comparative phylogeographic analyses by combining mitochondrial and nuclear gene fragments of S. uncinus and S. gigas sampled over an extensive area that covers most of their respective distribution ranges. We aim to explore (1) the intraspecific patterns of mitochondrial and nuclear genetic variation in both species, (2) the existence of shared phylogeographic patterns and barriers to gene flow, (3) the strength of the reproductive barriers between these pairs of species, and (4) the nature of the morphological differentiation among the populations of S. gigas.

The studied populations of Scaurus gigas from the Iberian Peninsula and North Africa are represented by 12 Cyt b haplotypes, three are only found in the Iberian Peninsula, eight exclusively in Morocco, and a single one is shared at both sides of the Strait of Gibraltar (Fig. 1; Table 1). The studied populations from Aglou and Deghaimis (Morocco) present two haplotypes separated between each other by 4-point mutations, but they are notably distant (26-point mutations) from all other haplotypes.

The studied populations of Scaurus uncinus presented eight Cyt b haplotypes. This species shows lower haplotype diversity than S. gigas. Four haplotypes are only found in the Iberian Peninsula, three only in Morocco, and one is shared across the Strait of Gibraltar. The studied population from Murcia (Spain) is separated by 10 mutations from the closest haplotype (Fig. 1).

Allele networks based on ITS2 for the studied populations of S. gigas (15 specimens) and S. uncinus (20 specimens) were represented by four and seven alleles, respectively (Fig. 2). The studied populations of S. gigas present one allele shared between North African and Iberian populations. All the other alleles are only present in Africa. The studied populations from Aglou and Deghaimis (Morocco) present a single allele separated by four-point mutations from the rest of the Iberian and African haplotypes (Fig. 2).

Scaurus uncinus, shows a higher allelic diversity than S. gigas, with seven ITS2 alleles observed. The most common allele is present both in the Iberian Peninsula, where it is widespread (at Faro, Cádiz, and Madrid), and in North Africa (at Ceuta). Also, three alleles are observed in North Africa and another three different alleles in the Iberian Peninsula. A single specimen morphologically identified as S. uncinus (CDAS0027, See Table 1) has mitochondrial DNA of S. gigas being represented by the Cyt b haplotype number 9 (Fig. 1; Table 1) and with nuclear DNA in concordance with its morphology being characterized by the ITS2 allele number 5 (Fig. 2; Table 1).

The Strait of Gibraltar, located between northwestern Africa and southwestern Europe, has acted as a geographical barrier for dispersion in different groups of small vertebrates and arthropods (Busack et al. 1985; Mas-Peinado et al. 2021; Sánchez-Vialas et al. 2020), although in some other cases it had no effect at all (Gutiérrez-Rodríguez et al. 2017; Mas-Peinado et al. 2022). The effect of the Strait and its evolutionary consequences are highly variable even for different species within the same genus (Mas-Peinado et al. 2021, 2022). The phylogeographic pattern shown by S. uncinus and S. gigas agree with those species that have been able to disperse across this marine strait, such as other tenebrionids like species of the subgenera Amblyptera Solier, 1836 and AmblypteracaMas-Peinado et al. 2018 of Pimelia (Mas-Peinado et al. 2022). The studied molecular markers of S. uncinus and S. gigas present shared haplotypes and alleles between the Iberian Peninsula and northern Morocco (Figs 1, 2).

The large degree of genetic structure found in the S. gigas lineage indicates a long history of geographic isolation and diversification in the African continent. This pattern is reversed in S. uncinus, which presents a pattern of subtle genetic structuring only among the European populations. The main nuclear allele of S. uncinus is widespread over the central and southwestern region of the Iberian Peninsula, but it is also present in northern Morocco. This allelic distribution could be related to a rapid range expansion into the African continent (Recuero and García-París 2011; Rodríguez-Flores et al. 2017; Cox et al. 2019). These data support the existence of recent intraspecific migration and gene flow through the Strait of Gibraltar for both species. However, the direction of the colonization seems to be different in each taxon. Hight levels of haplotypic diversity and internal recent speciation processes in the S. gigas lineage occur in the African continent suggesting that colonization of Europe took place from Africa; while in the S. uncinus lineage larger haplotypic diversity and geographic structuring occur within the European populations suggesting the opposite pattern of colonization, from Europe to Africa.

Scaurus gigas has its meridional limit at the surroundings of the Draa valley in Morocco (Español 1960). Populations of S. gigas from the Draa valley to Cabo Cantín (= Cabo Beddouza, 35 km North of Safi) were described by Español (1943) with subspecific status as S. uncinus ferreri (type locality: Sidi Ifni), morphologically characterized by presenting a poorly marked to absent dorsal carina in the anterior half of the elytra, which is well marked in the northern populations. Español (1943) ascription of these populations to S. uncinus was a result of an historical nomenclatural confusion stemming from early misidentifications. This confusion led to what is now recognized as S. gigas being incorrectly identified as S. uncinus, and S. punctatus being assigned to what is currently known as S. uncinus (Labrique 2004). During Labrique’s (2004) reassessment of the taxonomy of these species, S. uncinus ferreri was included in the synonymy of S. gigas (Labrique 2004; Iwan and Löbl 2020).

However, our results reveal the presence of a distinct haplotypic and nuclear lineage, closely related to S. gigas, found at the western foothills of the Anti-Atlas Mountains, south of the Souss Valley, specifically at Aglou Beach and Deghaimis. This population exhibits a morphology consistent with the type series of S. u. ferreri housed at the Museu de Ciències Naturals of Barcelona, confirming its alignment with specimens from Aglou and Deghaimis. The Aglou Beach population is located approximately 57 km north of the type locality of S. g. ferreri, well within the geographic range outlined by Español (1943) for the southern subspecies (Fig. 4F). The phylogeographic structure of the clade Scaurus gigas – S. ferreri in Morocco is likely derived from a latitudinal vicariant process of differentiation thought the Atlantic coast of Morocco, marked and more diverse in the South, and followed by a more recent northward expansion to the Iberian Peninsula, where populations are genetically more homogeneous (Figs 1, 2). Within this framework, it is noteworthy that in the southern distribution range of S. gigas – S. ferreri clade, there are no obvious current geographic barriers to dispersal that could prevent gene flow between populations located at both sides of the Souss Valley. Some studies suggest orography as the key factor influencing the maintenance of fine patterns of allopatric lineage structuration (Habel et al. 2012). However, S. gigas – S. ferreri are coastal species that do not occur far from the coastline, what makes difficult to infer if orography (Anti-Atlas) is accounting for their differentiation. The coastal area of the Souss Valley primarily comprises fluvial terraces with aeolian dunes (Bhiry and Occhietti 2004), flanked by the High Atlas Mountains to the North by and the Anti-Atlas to the South. Over time, this region has undergone significant environmental changes since the Miocene, including uplifted mountains and shifting environmental conditions (Aït Hssaine 1994; Bhiry and Occhietti 2004). These dynamic geological and ecological factors may have potentially promoted speciation among the reported lineages of Scaurus. A plausible similar pattern is evidenced by geckos of the genus Saurodactylus Fitzinger, 1843 and lizards of the Acanthodactylus pardalis species group. In the case of Saurodactylus, two lineages arised allopatrically (now species), which contact in the Souss valley without signs of genetic admixing (Rosado et al. 2017), while for the Acanthodactylus pardalis species group, the Anti-Atlas establishes a significant distribution gap between two sister species, one found in the Souss Valley, in the north of the Anti-Atlas, and the other to the South (Tamar et al. 2017). In our sampling we found strict concordance between nuclear and mtDNA markers depicting an isolated lineage south to the Souss Valley with no shared nuclear alleles between populations belonging to distinct mitochondrial populations in S. gigas – S. ferreri.

Although, our sampling across the Souss valley is very limited to test for the existence of complete reproductive isolation between the two lineages, the concordance between morphology (Figs 3, 4) and molecular markers made us considered them differentiated enough as to represent two independent evolutionary units. We, therefore, propose here the formal recognition of the southernmost populations of the S. gigas lineage, to be part of an independent species, and therefore revalidating the name Scaurus ferreri Español, 1943 stat. nov. for these populations (https://zoobank.org/C3150EA6-EE3D-447D-BB50-D22D80F4DF2C).

As expected by their morphological distinctiveness (Español 1943, 1960), genetic admixture between S. gigas and S. uncinus was almost absent across their contact zone. However, we observed a striking case of cyto-nuclear discordance in a single specimen morphologically identified as S. uncinus from Faro, southern Portugal. This specimen has a Cyt b haplotype of S. gigas, but its nuclear ITS2 allele as well as its morphology are within the variability of S. uncinus. This pattern of cyto-nuclear discordance observed in a single specimen, agrees with a possible hybrid origin for the specimen (Doniol-Vacroze et al. 2021). The low frequency of hybrids (n=1/30) across a wide potential area of sympatry (southwestern Atlantic coast of Europe and northwestern Atlantic coast of Africa), suggests that reproductive barriers between S. gigas and S. uncinus are strong, but not complete enough to prevent single episodes of hybridization (Dufresnes et al. 2020).

Transcontinental intraspecific patterns of colonization appear to be much common than previously expected between Africa and Europe for relatively large, flightless, xeric-adapted, species of terrestrial arthropods as darkling beetles (Condamine et al. 2013; Sánchez-Vialas et al. 2021; Mas-Peinado et al. 2018, 2022; Tinaut and Ruano 2021). Their low ability for dispersal is likely counteracted by large population densities and a strong survival ability in hostile environments. When darkling-beetle populations are located in coastal areas, as it is the case of the species of Scaurus studied here, they might be passively transported during exceptional climate disruptions (strong storms and tornados) or catastrophic volcanic episodes (ending in tsunamis) as it has been considered for other species as frogs (Hyla meridionalis Boettger, 1874) (Recuero et al. 2007) or tenebrionid beetles (Archaeoglenes Broun, 1893, Amblypteraca Mas-Peinado, Buckley, Ruiz & García-París, 2018) (Iwan et al. 2015; Mas-Peinado et al. 2022). While other representatives of Tenebrionoidea use complex strategies for dispersal, as blister beetles with larval phoresy (Sánchez-Vialas et al. 2021), darkling beetles seem to rely on passive dispersal favored by episodic events or even human accidental transport, as it seems likely the case for some populations of S. uncinus, resulting from heavy maritime transport from the Iberian coast to the Spanish city of Ceuta in northern Africa.

Conflict of interest. The authors declare no conflict of interest.

Data availability statement. The molecular data newly generated for this study is available in GenBank. Accession numbers P0438082–P0438110 for Cyt b and PQ435559-PQ435590 for ITS2 sequences.

Funding. The authors declare that his study was funded by the project-grant PID2019-110243GB-100/AEI/10.13039/501100011033 (Ministerio de Ciencia e Innovación) to Mario García-Paris.

We thank E. Recuero Gil, J. Gutiérrez Rodríguez, J. L. Ruiz García, P. Mas-Peinado, N. Percino Daniel, N. Rosas Ramos, A. Zaldívar, G. García Martín, F. Alda, I. Martínez-Solano, J. Miller, M. Carrillo, L. Puerta Rodríguez, J. Morales-García, C. Settanni, P. Jurado-Angulo, L. Saínz Escudero, K. López-Estrada, J.E. Uribe Arboleda, and M. Calvo Revuelta for their help with fieldwork. We especially thank J. L. Ruiz for his advice and assistance at the Instituto de Estudios Ceutíes. We are also grateful to Pablo Vicent Castelló for his support during the lab work, as well as E. Karen López Estrada and Lucía Sainz Escudero for their help with the molecular analyses. We express our deep gratitude to Mercedes París, curator of Entomology at the Museo Nacional de Ciencias Naturales (Madrid), for her invaluable work and availability. Special thanks to Dr. Berta Caballero López, curator of the arthropod collection at the Museu de Ciències Naturals (Barcelona), and Sergi Gago Carrión for their speed and efficiency in providing us with study materials when time was of the essence. We also thank Javier Diéguez-Uribeondo, director of the ‘Master en Biodiversidad en Áreas Tropicales y su Conservación, UIMP-CSIC,’ for his unwavering support, dedication to teaching, and for playing the role of scientific matchmaker by introducing the authors of this article.