Review of the Arabian Crematogaster Lund (Hymenoptera, Formicidae), synoptic list, distribution, and description of two new species from Oman and Saudi Arabia

Abstract The genus Crematogaster is one of the most species-rich and widespread groups of ants. Despite their often-high local abundance and important ecological interactions, the taxonomy of the genus is fragmentary and in great need of modern revisionary studies. As a first step towards a revision for the Arabian fauna of Crematogaster, a review of all known species with synoptic species accounts is provided. Seventeen species are recognized and illustrated from the Arabian Peninsula, of which two new species are described: C. jacindae Sharaf & Hita Garcia, sp. nov. from the Dhofar Governorate, Oman, and C. gryllsi Sharaf & Hita Garcia, sp. nov. from the Kingdom of Saudi Arabia (KSA) based on the worker caste. Crematogaster jacindaesp. nov. is easily separated from the remainder of the Arabian Crematogaster fauna due to its complete lack of propodeal spines, slit-shaped propodeal spiracles, and its distinct bicoloration, whereas C. gryllsisp. nov. is readily distinguished by its unlobed postpetiolar dorsum. Furthermore, new country records are presented: C. acaciae Forel for the KSA and Yemen, and C. delagoensis Forel and C. jehovae Forel for the KSAC. antaris for Qatar, whereas C. luctans Forel is excluded from the Arabian fauna. In addition, on the basis of morphological examination of original type material, C. affabilis Forel is proposed as junior synonym of C. chiarinii Mayr, and C. striaticeps is elevated to species rank stat. nov. Furthermore, a new identification key for the Arabian species is provided, as well as distribution maps for all species.


Introduction
The myrmicine ant genus Crematogaster Lund, 1831 is one of the most species-rich genera of the family Formicidae with 500 described species, 269 valid subspecies and two fossil species (Blaimer 2012a;Bolton 2019). The genus is widely distributed worldwide throughout most of the tropical and subtropical regions (Hölldobler and Wilson 1990;Longino 2003). Most species are arboreal and build nests in dead branches, under bark or in carton nest structures, or, less commonly, nest directly in soil (Hölldobler and Wilson 1990;Madden and Young 1992;Palmer and Brody 2013). When natural history information is available, these ants are generalized foragers or omnivores with numerous species also tending homopterans and Hemipterans (e.g., Brown 2000;Campbell 1994;Gullan et al. 2018;Longino 2003) for honey dew.
Despite the remarkable diversity, ecological importance, and often high local abundance of the genus, it is one of the most taxonomically neglected hyper-diverse ant genera. Presently, the taxonomic situation is only moderate to satisfactory for a few regions, such as North America (Buren 1959;1968), Costa Rica (Longino 2003), Madagascar (e.g., Blaimer 2010;2012b;, and parts of South East Asia (e.g., Hosoishi and Ogata 2008;2015;2017). However, apart from these, for most other regions, the taxonomy is in a very poor state, without modern revisionary treatments that cleared the taxonomic chaos from the 19 th and early 20 th century. As a consequence, the taxonomic validity of the majority of species and infraspecific taxa has not been comprehensively examined. We are certain that many nominally valid species and subspecies will have to be synonymized while at the same time there are numerous new species awaiting formal description. Therefore, species of Crematogaster are notoriously difficult to identify to species level in most parts of the world and the genus is in dire need of any modern, thorough revisionary study.
There are numerous, scattered, records of Crematogaster on the Arabian Peninsula. Prior to this study, the total number of species was 15 (plus one subspecies), recorded from the Kingdom of Saudi Arabia (KSA) (Collingwood 1985;Collingwood and Agosti 1996), the United Arab Emirates (UAE) (Collingwood and Agosti 1996;Collingwood et al. 2011), the Sultanate of Oman (Collingwood 1985;1988;Collingwood and Agosti 1996;Sharaf et al. 2018), Yemen (Collingwood and van Harten 1994;2001;2005), and Kuwait (Collingwood and Agosti 1996). However, there are no records known from Bahrain, nor the Socotra Archipelago.
The taxonomy of Arabian Crematogaster is even more challenging than in other parts of the world due to the geographic location of the Arabian Peninsula, which connects sub-Saharan Africa or the Afrotropical region with the Mediterranean/Middle East or Palaearctic region (Kreft and Jetz 2010). As a consequence, the Arabian Peninsula shares biogeographical affinities with both, the Afrotropical and Palaearctic regions, however, no modern revisions of Crematogaster exist for these regions around the Arabian Peninsula. While there are some treatments covering particular countries or smaller areas in Europe or the northern Mediterranean, these either just describe new taxa or provide identification resources without revising the genus (Collingwood 1978, Agosti and Collingwood 1987, Cagniant 2005, Seifert 2007, Karaman 2008). In addition, many Crematogaster species have large female reproductives and good dispersal abilities leading to often vast distribution ranges. Thus, to be certain about the identity of any Arabian Crematogaster material it would require examining and comparing hundreds of Afrotropical and Palaearctic type specimens from numerous natural history collections. This enormous task is unfortunately outside the scope of our study; however, we believe this study can present a first important step towards understanding the species distributions of Crematogaster on the Arabian Peninsula.
We present a synoptic list, species accounts for all species that include detailed taxonomic histories, as well as data and maps showing the currently known distribution ranges. In addition, we also present a new identification key to the Arabian species of Crematogaster and describe two new species: C. jacindae sp. nov. from Oman and C. gryllsi sp. nov. from the Asir Mountains, KSA. Furthermore, we examine and propose some status changes for Arabian taxa and discuss doubts about identifications and species records for some species.

Materials
Species names in this work follow the online catalogue of Bolton (2019) and in addition to published references (e.g., Collingwood 1985;Collingwood and Agosti 1996;Hita Garcia et al. 2013). The taxonomic histories of the treated taxa follow the online catalogue of ants of the world (Bolton 2019) and Antwiki (www.antwiki.org). Digital stacked color images were created using a Leica DFC450 digital camera with a Leica Z16 APO microscope and LAS (v3.8) software. These images are also available online on AntWeb (https://www.antweb.org) and are accessible through unique specimen identifiers attached (e.g., CASENT0872096). Distribution ranges were examined on Antmaps (https://antmaps.org; Guénard et al. 2017;Janicki et al. 2017) and general information of species on Antwiki. Distribution maps were made using DIVA-GIS (version 7.5.0.0).

Review of Arabian species
Crematogaster acaciae Forel Figure 5A-C Geographic range. Crematogaster acaciae was originally described from Ethiopia but is also known from Democratic Republic of Congo, Somalia, South Africa, and Zambia Janicki et al. 2017). For the Arabian Peninsula, it was previously only known from Oman (Collingwood 1985;Collingwood and Agosti 1996;Borowiec 2014;Sharaf et al. 2018;Monks et al. 2019) (Fig. 6). In this study, we provide the first records for the KSA and Yemen.
Remarks. This species represents a typical problematic taxon within the genus. The known distribution of C. acaciae and its three subspecies in the Afrotropics is patchy and there are some notable morphological differences between the infraspecific taxa. At present, it is likely that the taxonomic status of the involved taxa will change within the frame of a comprehensive revision of the Afrotropical fauna. Figure   Geographic range. This species was originally described from Egypt and can be found in most most countries of North Africa but also in Sudan and Kenya Janicki et al. 2017). For the Arabian Peninsula, it was recorded from the KSA, Oman, the UAE, and Yemen (Borowiec 2014;Collingwood 1985, Collingwood andAgosti 1996;Collingwood et al. 2011;Sharaf et al. 2018) (Fig. 6).

Crematogaster antaris Forel
Remarks. This is a very widespread and common species, which appears to be one of the most arid-adapted species within the genus. Our collections represent a new record for Qatar.   14; of C. inermis: Emery 1926: 3. Status as species : Forel 1894b: 24;Forel 1902: 152;Bondroit 1918: 115;Karavaiev 1927: 104;Finzi 1930: 15;Santschi 1937 Geographic range. This species was initially described from France and seems to have a broad distribution range from the south of France, the Iberian Peninsula, and the Canary Islands through all of North Africa to the Middle East, but is also found throughout the Balkans (Sharaf 2006;Borowiec and Salata 2012;Borowiec 2014;Guénard et al. 2017;Janicki et al. 2017). On the Arabian Peninsula it has been only recorded from the KSA (Collingwood 1985;Collingwood and Agosti 1996;Borowiec and Salata 2012;Borowiec 2014) (Fig. 11).

Crematogaster auberti Emery
Remarks. This is likely one of the most problematic species treated in this study. The taxonomic history provided above is complex with numerous status changes and infraspecific taxa. Borowiec (2014) states that all records from the eastern parts of the Mediterranean region are doubtful and likely misidentifications. If so, this would mean that the material from KSA is probably another species. However, at present and without a meaningful treatment of the European Crematogaster fauna, it is not possible for us to ascertain the genuine identity of this species. Thus, we maintain it as C. auberti for the moment and await further study of type material in European collections which might help elucidate its real identity.   3694N, 44.191E, 11-18.iii.1937(Dr. Carl Rathiens,B.M.1938  Geographic range. This species was originally described from Ethiopia and is widely distributed in the Afrotropical region. It seems to be a predominantly eastern African species but is also known from Central and South Africa Janicki et al. 2017). From the Arabian Peninsula, the species has been recorded from the southwestern mountains of the KSA and Oman (Wheeler 1922;Borowiec 2014;Collingwood 1985, Collingwood andAgosti 1996) (Fig. 13).

Crematogaster chiarinii Emery
Remarks. Crematogaster affabilis was originally described as a variety of C. chiarinii but subsequently elevated to species rank by Collingwood (1985) based on head width, length of the propodeal spines, and absence of the mesonotal tubercle. During the present study the type material of C. chiarinii and C. affabilis (MHNG) were examined and detailed morphological examinations of the type material shows that both are uniformly brown with the anterior half of the cephalic surface longitudinally striated; the frontal triangle is well-defined with a distinct posterior carina running back to the posterior level of the eyes; the propodeal spines long and acute, about 1.5 × longer than their base, making an angle of about 45 with the longitudinal axis of the body in profile view; and, the mesonotum in profile descending abruptly to a deep metanotal groove. The type of C. affabilis is somewhat larger but not more than major workers in both species are from minor workers. Consequently, even though both taxa are outside the focal region of Arabia, on the basis of any lack of significant phenotypical differences, we propose C. chiarinii as a senior synonym of C. affabilis.
Nevertheless, despite our synonymizing of both taxa, the taxonomic condition of C. chiarinii is still in need of a thorough revision. The taxonomic history above with all status changes, synonymic history, and numerous still valid infraspecific taxa shows clearly the complexity of this task. Based on superficial examination of material from the Afrotropical region, we are doubtful that the material from East Africa might remain conspecific with the one from Central and South Africa. Hopefully, a future revision of the Afrotropical fauna will resolve this situation.   Geographic range. This species was originally described from Mozambique and in the Afrotropical region seems to be restricted to Southern Africa Janicki et al. 2017). Our material represents the first record from the KSA since the species was previously only known from Yemen and Oman (Borowiec 2014;Collingwood and Agosti 1996;Sharaf et al. 2018) (Fig. 11).

Crematogaster delagoensis Forel
Remarks. Again, the disjunctive distribution raises some doubts about the identity of our material. Even though it is conspecific the material previously identified from Oman and Yemen, it needs to be proven that it is indeed C. delagoensis. The lack of records from East Africa and the southwestern parts of KSA could be due to insufficient sampling, but it could also be that the Arabian material is not similar to the South African "genuine" C. delagoensis. Geographic range. Crematogaster flaviventris was originally described from the Democratic Republic of Congo and is also found in Angola, Zambia, and southern Sudan Janicki et al. 2017). On the Arabian Peninsula, it seems to be restricted to Yemen (Collingwood and Agosti 1996;Blaimer 2012a;Borowiec 2014) (Fig. 16).

Crematogaster flaviventris Santschi
Remarks. This species record for Yemen is somewhat dubious. We have not examined the type material but preliminary examination of type images on AntWeb suggest that the material from Yemen is similar in color but there appear to be substantial differences in overall surface sculpture. This needs to be further investigated, ideally by comparing the Yemeni material with the type from Central Africa.   Head. Head as long as or little broader than long with convex sides and feebly concave posterior margin; antennae 12-segmented; in full-face view antennal scapes when laid back from their insertions reach or surpass posterior margin of head by the length of the first funicular segment; eyes of moderate size , located nearly at midlength of head in full-face view and with about 12 ommatidia in the longest row; anterior clypeal margin broadly convex. Mesosoma. Mesonotum in profile without a small tubercle on promesonotal suture; promesonotal suture feebly impressed; mesonotum convex in profile; median mesonotal carina absent; metanotal groove shallowly impressed but distinct; propodeal spines long, sharp and upward directed; propodeal spiracle circular, and located below the base of the propodeal spines; propodeal dorsum and propodeal spines forming a continuous concave curve in profile. Petiole. In profile petiole distinctly longer than high (PTHI 88-125; PTWI 145-267); approx. twice broader anteriorly than posteriorly in dorsal view; subpetiolar process well developed. Postpetiole. Postpetiolar node entire, not bilobed in dorsal view. Pilosity. Cephalic and clypeal surfaces with abundant scattered fine appressed pubescence; anterior clypeal margin and mandibles with abundant scattered long yellow hairs; antennae and legs with dense appressed pubescence; mesosoma without hairs; promesonotum and mesonotum dorsum with few appressed pubescence; petiole, postpetiole and the first three gastral tergites with appressed pubescence; few hairs on the terminal gastral segment. Sculpture. Mandibles longitudinally striated; clypeal surface smooth; area in front of eyes finely longitudinally striated; cephalic surface smooth; antennal fossae surrounded by fine and curved striolae; promesonotum side smooth; promesonotum dorsum faintly longitudinally rugulose; propodeal dorsum faintly longitudinally striated; mesopleura, petiole, and postpetiole distinctly densely punctate; metapleura faintly, irregularly striated; gastral tergites smooth. Color. Head, mesosoma, petiole, postpetiole, first gastral tergite, legs, and antennae uniform yellow, second half of gaster brown-yellow.
Etymology. The patronymic epithet has been selected in honor of Bear Grylls, the survival instructor in recognition of his remarkable efforts in spreading the culture of survival globally.
Ecological and biological notes. The microhabitat of Crematogaster gryllsi in the type locality (Fayfa Mountains) (Fig. 18) is the humid leaf litter under Acacia and Calotropis procera (Aiton) W. T. Aiton (Apocynaceae) trees. The specimens were collected by sifting leaf litter. The material from Raydah were collected by pitfall traps.
Geographic range. This new species is only known from the Asir Mountains, KSA (Fig. 19).
Remarks. This distinct new species immediately be separated from all Arabian Crematogaster species by the undivided postpetiolar node and the uniform yellow color. All other species have a divided postpetiolar node. Crematogaster gryllsi appears to be related directly to C. luctans Forel, 1907 from Kenya as it possesses an undivided postpetiolar dorsum as well. However, C. gryllsi can be separated by the following characters: metanotal groove shallowly impressed, propodeal dorsum and propodeal spines forming a continuous concave curve in profile; promesonotum side smooth; mesopleura distinctly densely punctate; whereas C. luctans has deeply impressed, U-shaped metanotal groove profile; convex propodeal dorsum in profile extending posteriorly to straight propodeal spines; mesosomal sides longitudinally striated.
Comments. The record of C. luctans given by Collingwood (1985) was based on material collected from Fayfa Mountains where the new species, C. gryllsi was collected. Considering that the shape of the postpetiole is uncommon in the Arabian fauna and the initial identification of C. luctans was probably based on this character, it is highly  likely that the record of C. luctans by Collingwood (1985) represents a misidentification. This hypothesis is further supported by the complete lack of any material of luctans in the WMLC. Therefore, the species was omitted from the list of Arabic species. Figure  Geographic range. Crematogaster inermis was originally described from Egypt and is widely distributed from the Iberian Peninsula through North Africa to the Middle East (Borowiec 2014;Guénard et al. 2017;Janicki et al. 2017). For the Arabian Peninsula, this species was only recorded from Yemen (Collingwood and van Harten 2001;Borowiec 20114).

Crematogaster inermis Mayr
Remarks. Knowing that the species is widespread in the Mediterranean and Middle East and the only record for the Arabian Peninsula is from Yemen, we think that it is likely that it is also present in the KSA. Futher sampling is necessary to verify this. Diagnosis. Crematogaster jacindae sp. nov. is distinguished from related congeners by the combination of the following characters: median mesonotal carina absent; propodeal spines absent; propodeal spiracle distinct in the form of a slit; area in front of eyes finely longitudinally striated; antennal fossae surrounded by fine and curved striolae; head black-brown or black, mesosoma, petiole and postpetiole dark brown, relatively lighter than head, gaster golden yellow.    Description. Worker. Head. Head as long as or little broader than long with convex sides and feebly concave posterior margin; antennae 12-segmented; in full-face view antennal scapes when laid back from their insertions fail to reach posterior margin of head; eyes of moderate size , located nearly at mid-length of head in full-face view and with ca. eleven ommatidia in the longest row; anterior clypeal margin broadly convex. Mesosoma. Promesonotum and mesonotum forming continuous curve in profile; median mesonotal carina absent; metanotal groove well developed; propodeal dorsum short forming curve with longer propodeal declivity; propodeal spines absent; propodeal spiracle distinct and slit-shaped. Petiole. In profile petiole distinctly longer than high (PTHI 44-68; PTWI 78-125); broader anteriorly than posteriorly in dorsal view. Postpetiole. Postpetiolar node distinctly bilobed in dorsal view; nearly as high as petiole in profile. Pilosity. Cephalic surface with abundant scattered fine pale hairs; anterior clypeal margin and mandibles with several long yellow hairs; posterior half of clypeus without hairs or pubescence; antennae and legs with abundant appressed pubescence; promesonotum and mesonotum each with single pair of hairs; promesonotum and mesonotum dorsum with appressed pale pubescence; no hairs or pubescence on propodeum; petiole and postpetiole each with single pair of posteriorly directed hairs; gastral pilosity restricted to few pairs on posterior margins of gastral tergites; gastral tergites with scattered appressed pubescence. Sculpture. Mandibles longitudinally striated; clypeal surface smooth; area in front of eyes finely longitudinally striated; cephalic surface feebly imbricate; antennal fossae surrounded by fine and curved striolae; promesonotum lateral side faintly imbricate; promesonotum dorsum faintly reticulate rugulose; mesopleura, metapleura, petiole, and postpetiole distinctly densely imbricate; gastral tergites faintly imbricate. Color. Head black-brown or black, mesosoma, petiole and postpetiole dark brown, relatively lighter than head, gaster golden yellow and strongly contrasting with remainder of body.

Crematogaster jacindae
Etymology. The patronymic epithet has been selected in honor of Ms. Jacinda Ardern, the Prime Minister of New Zealand in recognition of her humanitarian attitudes towards Muslim and minority communities in New Zealand.
Ecological and biological notes. The microhabitats where C. jacindae sp. nov. was encountered included leaf litter, soil, under stones, or on native vegetation, especially acacia trees (Fig. 22). The majority of specimens were collected foraging on plants using a beating sheet, but workers were also observed foraging on the ground and wild shrubs.
Geographic range. This new species is only known from Oman (Fig. 19).
Remarks. The distinctive golden yellow gaster and complete lack of propodeal armament of C. jacindae sp. nov. allows this Omani species to be immediately recognized from all other Arabian species. The closest relative of the new species is Crematogaster inermis Mayr, 1862 from Egypt. Both species are similar in body size and the lack of propodeal spines, but C. jacindae sp. nov. can be readily separated by the following characters: area in front of eyes finely longitudinally striated; cephalic surface feebly imbricate; eyes with about 11 ommatidia in the longest row; posterior half of clypeus without hairs or pubescence; mesopleura and metapleura distinctly densely imbricate; mesonotum with a single pair of hairs and without anterior tubercles; propodeal spiracles distinct in the form of a slit; body bicolored with head black-brown or black, mesosoma, petiole and postpetiole dark brown, relatively lighter than head, gaster golden yellow. By contrast, C. inermis has an unsculptured cephalic surface including the area in front of the eyes, eyes with ca. 14 ommatidia in the longest row, the posterior half of clypeus with fine appressed pubescence, mesosoma with a small anterior tubercle close to the promesonotal suture seen in profile; mesopleura and metapleura longitudinally striated, mesonotum without hairs, propodeal spiracle circular, unicolorous black-brown body. Figure 23A-C Taxonomic history. Crematogaster (Acrocoelia) auberti subsp. jehovae Forel, 1907c: 207 (w.) Israel: Menozzi 1933.    (Borowiec 2014;Collingwood 1985, Collingwood andAgosti 1996) (Fig. 25).

Crematogaster jehovae Forel
Remarks. This is a good example of a Crematogaster species with a complex and uncertain taxonomic situation. It is relatively widespread in the Mediterranean and Middle East and currently has nine subspecies. It is doubtful that they are conspecific and it is possible that this is a species complex. As outlined below, we consider C. laestrygon striaticeps as sufficiently different to raise it to the rank of species. Figure 26A-C Taxonomic history. Crematogaster arborea subsp. melanogaster Emery, 1895: 29 (w., q.)  Geographic range. This species was described from South Africa for the Afrotropical region and it seems to be restricted to the southern African countries of Botswana, Namibia, and South Africa Janicki et al. 2017). The only records for the Arabian Peninsula are from Oman (Collingwood and Agosti 1996;Borowiec 2014;Sharaf et al. 2018) (Fig. 16).

Crematogaster melanogaster Emery
Remarks. The species record from Oman appears doubtful based on the strange distribution pattern noted above. However, since we were unable to examine any material of this species, we consider it prudent to list it as an Arabian species for the moment. Figure 27A-C Taxonomic history. Crematogaster mimosae Santschi, 1914a: 87, fig. 11  Geographic range. Initially described from Kenya, in the Afrotropics this species is East African in its distribution found in Kenya, Uganda, Somalia, Sudan, and Tanzania Janicki et al. 2017). In the Arabian Peninsula, it was recorded from the KSA, Oman, the UAE and Yemen (Collingwood 1985, Collingwood andAgosti 1996;Borowiec 2014;Sharaf et al. 2018) (Fig. 25).

Crematogaster mimosae Santschi
Remarks. Crematogaster mimosae is one of four species of obligate acacia ants, which have been well studied in East Africa, mostly Kenya (e.g., Young et al. 1997;Martins 2010). From a taxonomic perspective, this is one of the "easy" cases within the genus in Arabia, thus very straightforwardly identifiable. Figure 28A-C Taxonomic history. Crematogaster (Acrocoelia) auberti st. oasium Santschi, 1911: 84 (w.)  Geographic range. Crematogaster oasium was described from Tunisia and can be found from Morocco east to Egypt (Borowiec 2014;Guénard et al. 2017;Janicki et al. 2017). In the Arabian Peninsula, it was recorded from Kuwait, KSA, Oman, and UAE (Collingwood 1985, Collingwood andAgosti 1996;Collingwood et al. 2011;Borowiec 2014;Sharaf et al. 2018) (Fig. 25).  Remarks. This species seems to be common and widespread in northern Africa. Since we have not collected or examined any material from Arabia we list this provisionally as an Arabian species for the time being since it was listed by several authors (see above). Figure 29A-C Taxonomic history. Crematogaster senegalensis Roger, 1863: 206 (w., q.)  Geographic range. While this species was originally described from Senegal, it seems to have a fairly disjunctive distribution since it is known from eastern and northwestern Africa without being recorded from countries in-between (Guénard et al. 2017;Janicki et al. 2017). In the Arabian Peninsula, it was recorded from the KSA, Oman, and Yemen (Collingwood 1985;Collingwood and Agosti 1996;Collingwood et al. 2011;Borowiec 2014;Sharaf et al. 2018) (Fig. 25).

Crematogaster senegalensis Roger
Remarks. The disjunct distribution of this species is a bit unusual and might require further attention in future studies of Afrotropical Crematogaster. It is likely that the odd distribution is just based on a sampling artifact or a preference of arid habitats, which are not as common in Central Africa. However, it could also be the case that this species, as currently defined, consists of several cryptic taxa. Without a thorough revision of the Afrotropical fauna it is impossible to be sure and we therefore list material examined as C. senegalensis.
Crematogaster striaticeps Forel, stat. nov.   Geographic range. This species was described from Algeria and is also found in the neighboring countries of Tunisia and Lybia (Borowiec 2014;Guénard et al. 2017;Janicki et al. 2017). In the Arabian Peninsula, it has been recorded from the KSA (Collingwood 1985).
Remarks. Crematogaster striaticeps was originally described as a subspecies of C. laestrygon but herein we treat this taxon as a good species. The main difference responsible for the decision to raise striaticeps to the specific rank is the presence of dense longitudinal striations on the entire cephalic surface whereas laestrygon has a smooth cephalic surface and longitudinal striations are feebly developed and restricted to the area in front of eyes.
We have been hesitant with this decision since both taxa are similar to other Afrotropical or Mediterranean Crematogaster taxa and it is not clear which taxonomic status they may assume after a thorough revision of the genus. However, since we established that C. laestrygon and C. striaticeps are not the same taxon, we propose to separate them by raising the latter to species rank to achieve clarity of their status for the Arabian region and future studies of its fauna.

Discussion
As can be seen from the species accounts presented above, the taxonomic histories of many species treated herein are complex and problematic. Many species have had numerous status changes and a changing number of infraspecific taxa. In some cases, it is likely that the species listed here will turn out to be senior or junior synonym of another taxon, and it is also very probable that some or many of the infraspecific taxa deserve to be treated as "good" species. As a consequence, except for the few species endemic to the Arabian Peninsula, for most others we suggest caution. Our review of species is based on literature records, material examined by the first author, and Arabian material examined in some European collections. We have pointed out which species we consider well identifiable and which ones are difficult. Overall, we consider previous identifications, as well as ours, as temporary. This study is meant as a first step stone towards a more comprehensive revision of the Arabian Crematogaster fauna. Since comprehensive taxonomic revisions of the genus are not to be expected from neither the Palaearctic nor the Afrotropical regions any time soon, the most sensible solution for the study of Arabian Crematogaster is to visit additional European collections and compare our material with as many types as possible in order to verify or improve the identifications of our material.
Notwithstanding the taxonomic problems lined out above, the treated fauna of Crematogaster exemplifies very well that the Arabian Peninsula shares substantial faunal elements with the Palaearctic (mostly Mediterranean species) and the Afrotropics, with a minority of species currently considered as Arabian endemics. At present, we recognize eight Afrotropical, seven Palaearctic, and two Arabian species, which we think fits fairly well with the biogeography of the Arabian Peninsula. However, this assessment might change with further studies and comparisons with types. We suspect that in some cases it might turn out that our material is not conspecific with any of the previ-ously identified species and might represent another undescribed endemic, but this requires further taxonomic work.
The ant genus Crematogaster is one of the most abundant myrmicine genera in the Arabian Peninsula, especially in the Asir mountains, Yemen, and Oman, particularly in areas with open forests and woodlands of Acacia (Martius, 1829) (Fabaceae) and Juniperus L. (Cupressaceae) trees. The close ecological association between ants and acacia trees has been documented by several authors (e.g., Hölldobler and Wilson 1990;Isbell et al. 2013;Madden and Young 1992;Palmer and Brody 2013;Young et al. 1997). Crematogaster mimosae is known to have mutualistic relationships with other ants in East Africa (Boyle et al. 2017;Palmer and Brody 2013) and we anticipate similar associations to be found in the vast areas of Acacia forests of the southwestern part of the Arabian Peninsula.
Although at present only 17 species of Crematogaster are known from the Arabian Peninsula, further targeted collecting may yield both additional species records and more new species. For example, no Crematogaster have been collected from Bahrain but it is very unlikely that the genus is absent from the country. We are sure that the current absence of Crematogaster records from Bahrain is due to the lack of any national myrmecological studies. The identification key to the Arabian Crematogaster presented herein serves as a foundation for further faunistic studies and taxonomic revisions of the genus.