Acomys cilicicus was described from Silifke, Mersin by Spitzenberger (1978). Until now, we had fairly limited knowledge about its phylogeny, and even its taxonomic status was unclear (Spitzenberger, 1978; Macholán et al. 1995; Kıvanç et al. 1997; Barone et al. 2001; Kryštufek and Vohralík, 2009; and Kıvanç et al. 2013 Aghova et al. 2019, Renaud et al. 2020). A recently published study determined its distribution in the same region (Çetintaş et al. 2018). The distribution of this species was very narrow and under considerable anthropogenic pressure, and thus the genetic distance calculations resulted in low genetic diversity both within the species and among subpopulations. We reveal two haplotypes. These two haplotypes (Karahmetli and Narlıkuyu) were separated from the others in all analyses. While the genetic distance calculations within Acomys cilicicus showed relatively small variation, the distance between the two studied populations of Acomys cilicicus was found to be notably higher than the distances between Acomys cilicicus and the other two species, Acomys minous. This suggests that the genetic divergence between these two localities of Acomys cilicicus is comparatively significant. Based on the genetic distance analyses, it may not be supported as a distinct species; however, the indication of their location as a separate branch in both phylogenetic tree topology and network analysis, combined with the presence of geographical isolation and reproductive isolation, supports considering them as separate species. Additionally, the comparison of genetic distances between subspinosus, spinosissimus, and cahirinus, which were used as an outgroup, revealed a much lower divergence compared to the distances observed between species in other taxa, with a value close to 2%, which is commonly considered as the threshold for speciation. This suggests that the rate of genetic differentiation within this group of Acomys species is relatively slow. Furthermore, according to the chromosome number analysis, while Acomys cilicicus originated from Acomys minous (Crete), the chromosome number of Acomys cilicicus was found to be the same as that of Acomys cahirinus (2n = 36) and no population with 2n = 36 chromosomes was found in Crete. This also suggests the presence of geographical isolation or reproductive isolation. Based on observations made in the field, the species is relatively sedentary with a highly restricted distribution range in Anatolia, which may contribute to the slow rate of genetic differentiation. All these indications support calling them different species.
Our phylogenetic tree showed that the closest ancestral species of A. cilicicus was A. minous, and the bootstrap values supported these results. The tree also showed that both species originated from cahirinus. Aghová et al. (2019) grouped the spiny mouse in five major clades: subspinosus, spinosissimus, russatus, wilsoni, and cahirinus. They placed the Mediterranean species within the cahirinus clade (cah9). They also suggested that cilicicus was a commensal population of cahirinus. The migration analysis results provided three possible scenarios for the migration route of Mediterranean spiny mice with radiation from cahirinus. The analysis resulted in a route in which cahirinus first invaded Crete, where it evolved to minous. The second wave occurred from Crete to Anatolia through an invasion by minous. After this second wave of invasion to Anatolia, the evolution and spread of cilicicus occurred. Acomys originated in southern Africa and dispersed northward by speciation (Aghová et al. 2019). In the northern part of Africa, cahirinus was derived around 9 mya. Our molecular clock estimated that the divergence time between cahirinus (cah10) Ethiopia and cahirinus (cah9) Egypt was 1. cahirinus/minous (Table 2, Add. Figure 4) separation as 1.6 mya, which was earlier than the estimation made by Barome et al. (2001), which was based on a 650-bp long data set. Giagia-Athanasopoulou et al. (2011) stated that colonization might have occurred earlier than 0.4 mya by evaluating chromosome variation results combined with cyt-b data. Later, the most recent study on Acomys (Aghová et al. 2019) dated the split time of cahirinus (cah9)/cahirinus (cah10) to approximately 1.7 mya and minous/cahirinus 1.59 mya, which aligned with our results and pointed out that cahirinus was the ancestral species for Mediterranean spiny mice. nesiotes could not be grouped as a separate species, probably due to the insufficient number of specimens and the very recent separation of the species. For this reason, detailed analysis is required to understand the status of nesoites.
The migration analysis results showed that both cilicicus and nesiotes evolved from minous, so that all Mediterranean species were grouped within the cahirinus clade in our tree as revealed by other studies (Barome et al. 2001, Aghová et al. 2019). Also, the tree given in Barome et al. (2001) placed minous and nesiotes within a group that was displayed as a sister group of minous and cilicicus. A similar chromosome topology variation for Crete spiny mouse was high when compared to those of theother island species. This might be the consequence of rapid chromosomal rearrangement and fast fixation of new rearrangements since the effect of a small population on selection appears faster. In other scenarios, the colonization occurred in multiple waves because nesiotes-minous and cilicicus-minous were grouped into different branches in every study (Barome et al. 2001; Aghová et al. 2019; Giagia-Athanasopoulou et al. 2011; and our results). As estimated by the migration analysis regarding the possible route for colonization of nesiotes and cilicicus, they originated and evolved from minous right after the occupation of Crete by humans, and because of introduction followed by speciation. The species limitation analysis also proved that minous become sources for the two different species in the Mediterranean region. Multiple species delimitation analysis showed that there were two divergence times from minous at two dates. One of them aroused cilicius while the other was nesiotes. The source of that was probably the invasion of Acomys with human movements from different places from the Mediterranean coasts of Africa, where cahirinus was distributed, to Crete. Castaneda et al. (2010) reported that the invasion of Santa Catalina Island by Peromyscus fraterculus resulted in the better adaptation of the invasive species compared to the native species P. slevi. Although the three species of Acomys found in Mediterranean islands form separate groups in phylogenetic trees, some individuals of the ancestral species, minous, enter other branches according to the Migrate analysis. This suggests that populations introduced by humans to Cyprus and Mersin have different levels of adaptation to their respective origins. Kaya Ozdemirel et al. (2022) modeled the potential distribution areas of cilicicus; however, during our fieldwork since 2002, no individuals belonging to this species were captured (unpublished personal data). This may indicate a low adaptive capacity associated with the introduction of cilicicus by humans. This situation raises another question: is the low adaptive capacity of cilicicus due to its low genetic diversity resulting from human transportation, or is it because the species is naturally docile, and therefore intentionally brought aboard ships by humans? Uncertainty remains regarding whether cilicicus was consciously introduced by humans due to its docile nature. Renaud et al. (2022) conducted an analysis showing that cilicicus has the smallest skull size among the species found in Mediterranean islands, and when captured in the field, this species exhibits a calm demeanor (unpublished personal observation between 2015–2018). Compared to the sympatric species Apodemus mystacinus, cilicicus is significantly less mobile in trap-closure areas. Consequently, alongside its low genetic diversity, the presence of a docile species raises suspicions that it may have been intentionally transported by humans on ships, potentially serving as a valuable food source during long sea voyages.
Gafney (2021) reviewed and stated that seafering in the Mediterrenean may have started during the Pleistocene from North Africa (figure). In any case, as Barome et al. (2001), Giagia-Athanasopoulou et al. (2011), and Aghová et al. (2019) emphasized minous was introduced by humans to Crete, as a descendant of cahirinus from Egypt. Humans unintentionally introduced this species (Barome et al. 2001), and the colonization might have happened in either a single wave or multiple waves (Giagia-Athanasopoulou et al. 2011). Multiple waves seem more likely, considering the chromosomal variation of The Crete spiny mouse. If multiple waves were indeed the case, another question spawned: “Did humans really carry along the spiny mouse unintentionally?”. The absence of Mus and other Rattus species in the study area has led to the emergence of this issue. Because Mus and Rattus species are small mammals that are frequently transported during invasion events, often leading to the eradication of other species in the introduced area (Harris 2008). In fact, genetic analyses of the Mus genus have shed light on human movements in the Mediterranean (Rodrigues et al. 2018). However, our molecular clock calculations have shown that cilicicus was introduced to Anatolia earlier. During fieldwork conducted between 2015 and 2018, no other species were found in the area (Çetintaş et al. 2019). The introduction of Acomys to Crete indicates a much older history than the arrival of Mus, as calculated by Cucchi et al. (2005) (Table 2). In light of this evidence, it appears plausible that Acomys was transported by humans during the time periods indicated by molecular clock calculations. However most archaeologists and paleoanthropologists believed that long distance hominin migration must have occured across land and land bridges (Anton and Swisher, 2001; Bar-Yosef and Belfer-Cohen, 2001). Gafney (2021) reviewed the potential harbors during the Pleistocene. Early homonid in North Africa may have invaded the island in the Mediterrenean as our migrate analysis suggested (Fig. 4).
The colonization pattern of the spiny mouse was similar to that of the caprine and house mouse, including a stopping station in the Mediterranean. However, from Crete, the transition to Cyprus and Turkey are showing a similar story in Mus. In Crete Barome et al (2001) noted Acomys minous b haplotype A turkey, a haplotype is the kind of resources that has emerged in Cyprus and our results also support this.
As a result, the occupation of Crete in the Mediterranean occured at different times and only certain routes were traveled from the cities established here. For this reason, cilicicus probably originated from minous's haplotype b, while nesiotes probably originated from haplotype a. We have shown this with migrate analysis. We show cilicicus as a separate species in both the phylogenetic tree by settling in separate branches and supporting these branches. Although it emerges as a separate species, partial microsatellite results and population genetics analysis performed with cytb sequences revealed genetic similarities within cilicicus; therefore, we suggest more detailed genetic analysis (SNP, ddRAD) should be performed to determine the genetic structure of cilicicus and other Mediterrenean species.