Patterns of morphological and genetic diversity within the fringe-toed lizards of the genus Acanthodactylus have puzzled systematists since the first assessments, and none more so than the Acanthodactylus erythrurus complex. A recent study combining multi-locus sequence data and morphological characters partially resolved the situation, identifying two new species in the southern part of the range in Morocco, but leaving an unresolved “Ibero-Moroccan” clade containing much of the genetic and morphological diversity. Here we sequenced a mitochondrial marker for new samples from across much of the distribution. Our data notably increase the known ranges of various species and lineages found in Morocco, and indicate a divergent genetic lineage within one of the newly described species. While far greater numbers of genetic markers will be needed to resolve taxonomic questions, greater geographic sampling is also still needed both to delimit the species, and to identify regions where potential genetic admixture may occur.

Acanthodactylus lacrymae, Acanthodactylus montanus, NADH dehydrogenase subunit 4, phylogeography

Acanthodactylus, or fringe-toed lizards, comprise the most specious genus within the family Lacertidae, with 44 recognized species (Uetz 2023). The genus is a member of the Eremiadini tribe within the subfamily Lacertini, which along with its phylogenetically closest members (including Mesalina; Garcia-Porta et al. 2019) primarily inhabits xeric habitats in North Africa and Asia (Sindaco and Jeremčenko 2008). Acanthodactylus ranges from the Iberian Peninsula, across North Africa and the Arabian Peninsula into western India, and northward to Turkey.

Acanthodactylus erythrurus (Schinz, 1838) is distributed across Morocco and Algeria and is the only species of the genus occurring in the Iberian Peninsula. It is usually found in habitats with a moderate supply of moisture in shrubland, mesic forests and rocky areas (Schleich et al. 1996). The Iberian Peninsula was colonized from North Africa (Harris et al. 2004), and the respective populations have shown high levels of admixture in their mitochondrial DNA (mtDNA), especially in the northern part of their distribution (Harris et al. 2019). Concerning the North African populations, Acanthodactylus erythrurus group is of great interest due to its unclear taxonomy and complex intraspecific relationships (Harris et al. 2004; Fonseca et al. 2009; Tamar et al. 2016; Beddek et al. 2018).

Intraspecific variation within Acanthodactylus erythrurus is extensive, and has led to notable differences in how studies have treated this. Salvador (1982) and Arnold (1983) accepted A. e. lineomaculatus and A. e. belli as the two subspecies occurring in North Africa, while Bons and Geniez (1995) added the subspecies A. e. atlanticus from the Atlas Mountains and suggested that A. e. lineomaculatus deserved specific status. On the other hand, Squalli-Houssani (1991) proposed that all Moroccan subspecies should be considered as one highly variable species with morphological differences reflecting adaptation to local habitats. Studies employing mtDNA sequences revealed that both the Tunisian A. blanci and A. (e.) lineomaculatus are nested within the A. erythrurus complex (Harris et al. 2004; Fonsesca et al. 2009; Tamar et al. 2016). A comparative phylogeographic study based on mtDNA data, including samples of A. erythrurus and A. blanci from Algeria, revealed considerable mtDNA diversity within the species, with one species-delimitation approach (Automatic Barcode Gap Discovery, ABGD; Puillandre et al. 2012) indicating potentially 15 taxonomic units, the most of any of the species examined (Beddek et al. 2018). At the same time, a strong East-West divergence pattern across the Maghreb was identified, with two main clades showing this motif with a third basal clade reported only from the southern Atlas Mountains.

Some taxonomic issues regarding the A. erythrurus complex were resolved by Miralles et al. (2020), employing multilocus sequence data and a morphological assessment. They recovered five major lineages, that were supported by both mtDNA and nuclear DNA (nDNA). They continued to find a complex pattern across most of an Ibero-Moroccan (IM) clade, with highly diverse mtDNA lineages that did not fully coincide either with the three subspecies accepted in Morocco, or with nuclear markers. However, in the Atlas Mountains of Morocco two divergent lineages were identified in both mtDNA and the nuclear markers, which also showed minimal differences with some morphological characters, and these were recognized as distinct species, Acanthodactylus montanus from the High Atlas Mountain region and Acanthodactylus lacrymae from the Middle Atlas region. These were more closely related to Algerian and Tunisian forms rather than the Ibero-Moroccan clade (Miralles et al. 2020). These authors also highlighted the need for further sampling to better assess the distribution of these forms. Rancilhac et al. (2023) attempted to resolve the situation within the IM clade of A. erythrurus using mtDNA and nine nuclear gene fragments. However, even with this enlarged dataset they were only able to distinguish four major groups of populations, separated by barriers to gene flow. A recent genome-wide RADseq approach to investigate a contact zone within the IM clade indicated that there was spatial restricted gene flow highlighting high levels of reproductive isolation, consistent with even more species-level diversity within the complex (Doniol-Valcroze et al. 2024).

To obtain new insights into the distribution of the different lineages and species of the A. erythrurus complex, here we sequenced a partial ND4 mitochondrial gene region for 42 individuals, primarily from Morocco. Since mitochondrial and nuclear DNA were completely congruent for A. montanus, A. lacrymae, and the major lineages within the A. erythrurus complex (Miralles et al. 2020), this should give additional information regarding the distribution of these species, and also help delimit the ranges of mtDNA lineages within the rest of the A. erythrurus complex.

We analyzed 42 samples of A. erythrurus (Table 1). Animals were caught in the field over a 20-year period, and a small piece of tail tissue was removed before releasing them at the collection site. Locations and codes of the samples are represented in Fig. 1. Since lineages and species cannot always be determined with certainty based on morphological characters of single individuals (see Miralles et al. 2020), all individuals were sequenced. Additional data from previous studies were retrieved from GenBank (Tamar et al. 2016; Beddek et al. 2018; Miralles et al. 2020) covering the distribution range of the species. Sequences of A. micropholis, A. blanfordii, A. grandi and A. boskianus (Heidari et al. 2014) were included as outgroups. All new sequences were submitted to GenBank (PQ000925–PQ000966).

Figure 1. 

Distribution map of lineages within the Acanthodactylus erythrurus complex within North Africa with complete range inset. Distribution outline follows the IUCN. Newly sequenced individuals in this study are numbered, others are from GenBank. Colour codes indicate the different forms – A. erythrurus complex IM clade (blue circles), A. montanus (red squares), A. lacrymae (pink diamonds), A. erythrurus complex Central Algeria Clade (purple triangles), A. erythrurus complex Algero-Tunisian clade (light blue triangles).

Total genomic DNA was extracted from alcohol-preserved tail tissue following standard high-salt protocols (Sambrook et al. 1989). We amplified a mitochondrial gene fragment, NADH dehydrogenase subunit 4 gene with the adjacent tRNAs (ND4+His, Ser, Leu) in order to allow comparison with previous published studies of A. erythrurus using this marker (Tamar et al. 2016; Beddek et al. 2018; Miralles et al. 2020). The fragment was amplified performing a Polymerase Chain Reaction (PCR) with primers ND4 and Leu from Arévalo et al. (1994), in a total volume of 25 μl, with 5 μl of 5x reaction Buffer, 3.2 μl of 25 mM MgCl₂, 1 μl of 5 mM dNTPs, 0.5 μl of 4.0 M of each primer, and 0.2 μl (1U) of Promega GoTaq DNA polymerase. PCR conditions were: pre-denaturation step of 94 °C (3 min), 33 cycles with 94 °C (30 s) denaturing, 47 °C (40 s) annealing, 72 °C (90 s) extension and with a final extension conducted at 72 °C for 5 min. Positive PCR products were sent to GENEWIZ (Germany) for sequencing.

Sequences were edited and aligned using ClustalW with default parameters in MEGAX (Kumar et al. 2016). Genetic uncorrected p-distances were also calculated in MEGAX (Kumar et al. 2016).

We employed two methods of phylogenetic inference based on the ND4 sequences, Maximum Likelihood (ML) and Bayesian Inference (BI). Best-fit partition schemes were selected using Partition Finder v1.1.1 (Lanfear et al. 2012). We used codon partitions and the tRNAs fragment was set as a fourth data block. ML analysis was performed with MEGAX with the built-in tool to choose the most appropriate model under the AIC (GTR+G), while nodal support was assessed by bootstrapping with 5,000 replicates. The BI analysis was carried out using MrBayes v3.2.7 (Ronquist et al. 2012) and separate models were set for the different partitions (in each case GTR+I+G). Two independent runs of 5×10⁶ generations were performed, with a sampling frequency of 1,000 and 25% of the trees were discarded as burnin. Trees were imported to FigTree v1.4.4 for visualization.

The dataset consisted of 42 newly sequenced members of the A. erythrurus complex, along with 105 previously published sequences from GenBank, with an aligned length of 769 bp. Both Bayesian Inference and Maximum Likelihood analysis for the mitochondrial fragment (ND4) produced almost identical topologies, differing slightly in the deeper nodes, with the BI tree revealing higher support on some nodes (Fig. 2). Our results are generally congruent with previous studies regarding the major clades (Beddek et al. 2018; Miralles et al. 2020). Following the taxonomy of Miralles et al. (2020), all five major clades – the species A. montanus (WHA) and A. lacrymae (EHA), the highly diverse Ibero-Moroccan (IM) clade, the Algero-Tunisian clade (AT) and the Central Algerian clade (CA) can all be identified (Fig. 2). However, when comparing the distribution of these clades (Fig. 1) with the newly sequenced specimens for this study there are some notable modifications. Regarding the IM clade, our new sequences from Debdou (Fig. 1: 40–45) fills a wide gap in sampling for this clade, which had previously been confirmed around the region of Taza and then a single sample over 200 km to the East in Algeria (Fig. 1). For A. lacrymae, our samples from just north of Aït Aïssa (Fig. 1: 35–39), extend the range over 100 km to the East of the previously confirmed populations. For A. montanus our sample 24 (Fig. 1) not only increases the apparent range of this lineage about 200 km to the northeast, it also means that the ranges of A. montanus and A. lacrymae can no longer be considered highly allopatric, since they are separated by at most around 10 km. This sample 24 is also interesting genetically, as although it is strongly supported as sister taxa to A. montanus (97% BPP), it is highly distinct from the samples from the southern part of the range (samples 33–34, 8±1% SE).

Figure 2. 

Estimate of relationships within the A. erythrurus complex in North Africa derived from a Bayesian analysis. Lineages are labelled following Miralles et al. (2020): A. erythrurus complex Ibero-Moroccan clade (IM), A. montanus (WHA), A. lacrymae (EHA), A. erythrurus complex Central Algeria Clade (CA), A. erythrurus complex Algero-Tunisian clade (AT). Stars indicate novel samples from this study, and numbers above and below nodes indicate Bayesian Posterior Probabilities and Bootstrap support from a Maximum Likelihood analysis respectively. Missing Bootstrap support values are due to different placement of the WHA clade with Maximum Likelihood, where it appears as sister-clade to EHA, CA, and AT.

Just as early assessments of morphological variation within the A. erythrurus complex identified high levels of complexity (Salvador 1982; Arnold 1983), so later assessments of genetic diversity have continued to perplex researchers. Miralles et al. (2020) managed to describe two species from the southeastern edge of the range, A. lacrymae and A. montanus, while leaving the bulk of the morphological and genetic diversity within an unresolved A. erythrurus IM clade. Our additional sampling further supports the distribution of this latter clade, across the Moulouya river valley region in Debdou, with these most closely related to a single published sequence from even further east (KY490390, Beddek et al. 2018), that indicates the IM clade reaches into Algeria (Fig. 1). The Moulouya region is often considered a biogeographical barrier separating herpetofauna into western and eastern forms (reviewed in Salvi et al. 2018), and the samples from the East of the Moulouya valley do form a subgroup within the IM clade, again highlighting the intricacy of the phylogeographic patterns within the A. erythrurus complex (Rancilhac et al. 2023).

Regarding the situation in the Atlas Mountains and the southeastern range of the distribution, Miralles et al. (2020) recognized A. montanus and A. lacrymae for the two highly genetically distinct lineages they recovered in this region. While these two forms can be morphologically separated from the IM clade of the A. erythrurus complex, Miralles et al. (2020) noted that A. montanus is “very similar to the allopatric Acanthodactylus lacrymae and single individuals are not always possible to separate”. However, the situation was simplified by the large distance between the ranges of the species – indeed these authors specifically presented networks of nuclear haplotype sharing between both A. montanus and A. lacrymae with the IM clade, but not between A. montanus and A. lacrymae (fig. 5 of Miralles et al. 2020). The finding of a distinct lineage apparently of A. montanus very close to the range of A. lacrymae (sample 24; Fig. 1) complicates this situation. Since these species cannot be easily separated in the field, extensive genetic sampling of individuals from the region of contact between A. montanus and A. lacrymae will be needed to confirm if there is genetic admixture, or if the two lineages are found in strict sympatry. Furthermore, the high degree of genetic differentiation between this sample (24) and A. montanus from the type locality (8% with this ND4 marker) is higher than between some accepted lacertid lizard species (e.g. Dinarolacerta mosorensis and Dinarolacerta montenegrina; 6.7%, Mendes et al. 2016). On the other hand, divergence levels were similar (7.8%) within another lacertid, Timon tangitanus, in Morocco which showed a lack of lineage sorting with nuclear markers (Abreu et al. 2020). While Rancilhac et al. (2023) were unable to fully delimit sublineages within the IM clade based on multiple nuclear markers, recent analyses of RADseq genome-wide data indicate that some of these at least correspond to cryptic species (Doniol-Valcroze et al. 2024). A similar approach will probably be needed to determine the taxonomic status of this new lineage.

To conclude, phylogeographic patterns within the A. erythrurus complex continue to slowly take shape. Our additional geographic sampling extends the ranges of some forms found in Morocco, the A. erythrurus IM clade, A. montanus and A. lacrymae. In particular, a genetically distinct apparent individual of A. montanus was found very close to the known range of A. lacrymae. Determination of genetic admixture between A. montanus and A. lacrymae will be necessary to confirm the genetic distinctiveness of these morphologically extremely similar forms. Our data highlights that, as well as the previously identified need for inclusion of greater numbers of genetic markers or even genomic level assessments (Doniol-Valcroze et al. 2024), increased geographic sampling is also needed, especially across the southeastern part of the range in Morocco where the distribution of the different species and lineages is still not fully determined.

Lizards were captured under permit of the Haut Commisariat aux Eaux and Forets of Morocco (HCEFLCD/ DLCDPN/DPRN/DFF No14/2010). We thank our colleagues from CIBIO for assistance during the fieldwork. Part of this work was carried out during an ERASMUS scholarship of DS, supervised by DJH. DJH is funded through the Fundação para a Ciência e Tecnologia (FCT), CEECINST/00104/2021/CP2819/CT0003. This work was also supported by the European Union’s Horizon 2020 Research and Innovation Programme under the Grant Agreement Number 857251.