Updated check-list of the mayflies (Insecta: Ephemeroptera) of Iraq

Abstract Based on a recent field survey in Iraqi Kurdistan and a critical evaluation of previously published data, 37 mayfly species are listed as occurring in Iraq. We collected and identified nine species as new for the country and corrected some previously published records. For several species scarcely treated in the literature, we provide information allowing their identification in the larval stage to promote the acquisition of reliable faunistic data from Iraq in the future.

and Dia 1982, Sartori 1992, Malzacher 1992 and Iran , Hrivniak et al. 2020. Turkey has also been extensively studied (e.g. Kazancı and Türkmen 2012, Kazancı and Türkmen 2016, Salur et al. 2016, Sroka et al. 2019a, Hrivniak et al. 2019. In contrast to abovementioned regions, Iraq has been studied only poorly (Al-Zubaidi et al. 1987, Bojková and Soldán 2015, Abdul-Rassoul 2020. The stream network of Iraq includes the Tigris River basin which has several sub-basins (Khabur, Greater Zab, Lesser Zab, Adhaim and Diyala) and the Euphrates River basin (Al-Ansari et al. 2019). More than fifty watercourses flow from Turkey or Iran into Iraq (Yousuf et al. 2018). The discharge of both the Tigris and Euphrates is decreasing with time due to the construction of dams in the upstream part of the basins and climate change (Al-Gburi et al. 2017, Al-Ansari et al. 2018). In addition, the increasing stress of wastewater effluents has led to the degradation of many habitats, for example, in the southern wetlands and marshes (Al-Gburi et al. 2017). Mayfly larvae are well known for their sensitivity to oxygen depletion, therefore, they are commonly used as a suitable model in freshwater monitoring programmes Moog 2000, Menetrey et al. 2008). However, at present, there are not enough data for Iraq to analyse these processes and a good knowledge of the local fauna is a necessary prerequisite for such efforts.
Most of the published data concerning the mayflies of Iraq are restricted to the northern region of the country. Al-Zubaidi and Al-Kayatt (1986) presented information on the geographical distribution of five families, six genera and five newly-recorded species from northern Iraq. Then, Al-Zubaidi et al. (1987) published nine mayfly species as newly recorded for Iraq and described two new species, Oligoneuriella bicaudata Al-Zubaidi, Braasch & Al-Kayatt, 1987 and Isonychia arabica Al-Zubaidi, Braasch & Al-Kayatt, 1987. Four newly-recorded species from the middle region of Iraq were recorded by Carl (1989). Another new species, Prosopistoma helenae Bojková & Soldán, 2015 was discovered from the Tigris River in Mosul city. A remarkable mayfly species, Mortogenesia mesopotamica (Morton, 1921), has been thoroughly re-described by Soldán and Godunko (2013). Several other papers on the mayfly diversity of Iraq have been published quite recently and represent ecological studies focused on aquatic macroinvertebrates that are useful in water quality bioassessment (Shekha 2011, Hanna and Shekha 2015, Ali and Latef 2017and Hanna et al. 2019). The first checklist of the mayfly fauna of Iraq was summarised by Abdul-Rassoul (2020), based on literature data only. This checklist compiled 30 species in 18 genera and 10 families; however, some specimens were actually identified only on the genus level.
Most of the studies dealing with Iraqi fauna have been undertaken using the morphological approach only. The barcoding fragment of mtDNA (COI) was acquired for some locallycollected species by M. Al Saffar, but remains mostly unpublished (some of these sequences were incorporated in Sroka et al. 2019). Recently, Khudhur and Shekha (2020) have performed a molecular species identification (using 16S rDNA) of some Heptageniidae taxa.
Recent ecological studies or attempts to incorporate molecular data into mayfly research in Iraq are potentially very useful. They are, however, also hampered by a lack of suitable literature allowing accurate species identifications. Therefore, the samples are often identified using inappropriate literature sources, which might result in considerable confusion regarding faunistic data in the future (see ). Therefore, we aim to partly provide the necessary information to avoid this eventuality. Specifically, our study aims to: i) provide a list of the mayfly species occurring in Iraq, based on all published records to date and to critically evaluate these records; ii) based on our sampling, report the finding of a further nine species previously unknown from the country; iii) provide essential information about the diagnostic characters of some scarcely reported species to facilitate their identification and the future assembly of sound faunistic data from the country.

Material and methods
The samples were collected semi-quantitatively by a D-frame net using the kick-sampling method, sweeping through roots and submerged plants or picked manually from rocks and pebbles. All of the material was collected by F. Khudhur from January to October 2019. The samples were collected from 24 sites (one sample per site) which were selected to cover the mountainous region of the northern part of the country. For the list of localities, see Table 1. The larvae were stored in 96% EtOH at -20°C. The material totals 667 specimens and is housed at the Biology Centre CAS, Institute of Entomology (IECA). A map of all the localities sampled is provided in Fig. 1. To check minute morphological characters, parts of some specimens were mounted on microscopic slides using HydroMatrix (MicroTech Lab, Graz, Austria) mounting medium (soluble in water). Photographs were obtained using a Canon EOS 1200D camera mounted on a Leica M205 C stereomicroscope. All photographs were subsequently enhanced with Adobe Photoshop™ CS5.

Check-list of Mayflies in Iraq
Literature published up to now contains records of 37 species occurring in Iraq. Table 2 summarises all of the records (excluding eight additional species listed in Table 3, which were certainly reported, based on incorrect identification, see Discussion). Only taxa identified to the species-level are included. We report nine species as new for the fauna of Iraq. We collected fresh material only in the northern part of the country. The southern part of Iraq hosts a very different species composition, since the environment is totally different. Its further study would be highly beneficial, but at present, it is hampered by the difficult accessibility of these locations due to the unstable security situation.  Map of Iraq, localities sampled marked by red dots.  List of extra-limital mayfly species, reported from Iraq, based on misidentifications.

Diagnostic characters of selected species
Here, we list and illustrate the most pronounced diagnostic characters of several species (including species we report as new for the country) to facilitate accurate identification and to prevent the publication of confusing faunistic and ecological data, especially for nontaxonomists. Most of these species are not included in available identification keys, focusing on European fauna (e.g. Studemann et al. 1992, Engblom 1996,Bauernfeind and Humpesch 2001, Eiseler 2005.

Baetis (Rhodobaetis) braaschi Zimmermann, 1980
The species belongs to the subgenus Rhodobaetis Jacob, 2003. Its subgeneric placement can be readily confirmed by the presence of a row of articulated setae (spatulas) on the posterior margin of the abdominal terga. It mostly lacks spine-like setae on the margins of the gill plates, thus being easily recognisable from B. rhodani and B. ilex, other Rhodobaetis species frequently co-occurring in the region and always equipped with such setae. Furthermore, B. braaschi is distinct from all other Baetis species in the region by the presence of two rounded pale spots on the abdominal terga ( Fig. 2A). The species is not covered in European keys, its morphology being treated in Godunko et al. (2004) or . We suspect that B. braaschi is sometimes misidentified as Baetis vernus Curtis, 1834 (reported for Iraq by Shekha 2011), the presence of which in the Middle East we consider doubtful. Baetis vernus is a common European species with somewhat similar markings on the abdominal terga. This misidentification can be easily avoided by checking the presence of setae on the posterior margins of the abdominal terga, indicative of Rhodobaetis.

Baetis (Rhodobaetis) ilex Jacob & Zimmermann, 1978
The species represents another member of the subgenus Rhodobaetis, easily distinguishable by the presence of setae on the posterior margins of the abdominal terga ( Fig. 2B) and spine-like setae on the gill margins. It can be distinguished from all other Rhodobaetis species of the Caucasus and Middle East by: i) the presence of spine-like setae on both the inner and outer margins of the gill plates, although these on the inner margin tend to be scarce in specimens from Iraq. The most common Rhodobaetis species, B. rhodani, exhibits these setae on the outer margin only; ii) the elongated shape of the labrum (Fig. 2C) and relatively narrow labial palps (Fig. 2D); iii) the absence of triangular projections on the posterior margin of the abdominal terga (Fig. 2B). The illustrations in the original description of the species (Jacob and Zimmermann 1978) are of very good quality and can be reliably used for species identification.

Baetis (Baetis) samochai Koch, 1981
The species is distinct from all other Baetis species occurring in the region in several characters: i) a very broad outer tooth of the mandibular incisors, distinctly separated from the inner part of the incisors (Fig. 2E, F); ii) short and pointed setae on the posterior margin of the femora (Fig. 2G); iii) narrow pointed spines on the posterior margin of the terga (Fig.  2H). An overview of B. samochai morphology including some illustrations, based on material from Israel, is also available in Yanai et al. (2018).

Epeorus (Epeorus) zaitzevi Tshernova, 1981
Despite the species being relatively widespread and common in the region, existing descriptions are scattered and sometimes confusing. In the original description, Tshernova (1981) described only male imago. Braasch 1978 described a larva of E. zaitzevi, but erroneously under the name Epeorus znojkoi, as pointed out by Koch (1988) and Sartori (1992). Some information on E. zaitzevi larvae is also contained in Samocha (1972) and Demoulin (1973), where the species is treated as Epeorus sp. It is also included in the key of Kluge (1997).
E. zaitzevi larva is similar to European Epeorus assimilis Eaton, 1883-88, sharing the absence of paracercus and not having an enlarged gill I. Nevertheless, it can be distinguished by the following combination of characters: i) long and pointed posterolateral projections on the abdominal terga (Fig. 2I); ii) the presence of long, hair-like setae on the abdominal terga (Fig. 2J); iii) long, sparse teeth on the posterior margin of the abdominal terga. Sartori (1992) mentioned the presence of a narrow dark band on the posterior margin of each abdominal tergite, visible in both larvae and adults, clearly visible also on our material.

Ecdyonurus (Ecdyonurus) ornatipennis Tshernova, 1938
The morphology of this species is poorly known. In the original description, Tshernova (1938) studied only the male imago. The only description of larva was published by Braasch (1980), although he noted the association of his larval material with Tshernova's adults as questionable, based on the proximity of localities and not direct rearing. Nevertheless, the Braasch's concept of E. ornatipennis larvae was followed by subsequent authors Braasch (1988), Hrivniak et al. (2018)) and we follow it here as well, as a thorough revision of Caucasian Ecdyonurus is not yet available. From all other species of Ecdyonurus occurring in the area, E. ornatipennis can be distinguished by: i) posterolateral projections of the prothorax which are short and apically not pointed; ii) setae on the dorsal surface of hind femora which are apically not pointed (Fig. 2K, L); iii) abdominal sternites with a distinct dark medial band (Fig. 2M). These characters have already been described and/or depicted in and correspond with our observations.

Electrogena pseudaffinis (Braasch, 1980)
The species was described, based on larvae and the original description (Braasch 1980) is basically still the only source that describes its morphology, with some information also being scattered in . Electrogena pseudaffinis is distinct from other Electrogena species distributed around N. Iraq by: i) shape of pronotum, with posterior corners smoothly rounded, without abrupt step (Fig. 2N); ii) the colour pattern of head and abdominal terga (without distinct median dark band, Fig. 2O); iii) the shape of setae on the dorsal surface of the femora (Fig. 2P, Q); iv) shape of gill plates.

Caenis luctuosa (Burmeister, 1839)
The species is very similar to C. macrura, which is widely distributed in the region (being a dominant Caenidae species in neighbouring Iran, see . The main distinguishing character is the shape and arrangement of setae on the dorsal surface of the fore-femora. Caenis luctuosa possesses a line of short setae, bifurcated to almost half of their length (Fig. 2R), in contrast to the longer setae of C. macrura, bifurcated only apically. Both species are widely covered in European determination keys. However, a considerable morphological variability exists within these two species and the identification is not always without doubt. The most detailed description of intraspecific variability in C. macrura and C. luctuosa was published by Malzacher (1984). In Iraq, there is also the possibility of the occurrence of three more species, described from Israel, namely Caenis gilbonensis Malzacher, 1992, Caenis parabrevipes Malzacher, 1992 and Caenis antoniae Malzacher, 1992 (the latter two species were also reported from Jordan by ). Our specimens from Iraq differ from C. gilbonensis and C. parabrevipes by exhibiting a more deeply and acutely notched posterior margin of sternum X and the presence of more pronounced posterolateral spines on abdominal segments. As for C. antoniae, the shield shaped microtrichia on the surface of the larval wing pads are very small in this species ( fig. 5e in Malzacher 1992), whereas these structures are much larger in our material identified as C. luctuosa.

Discussion
We listed the records of 37 species for Iraq. From the nine species, we report as new for the country, all of these findings being expected, since these species occur in neighbouring regions and have not been previously mentioned from Iraq only because the area is understudied. The taxonomy of several species we collected is not clear (identified as "cf." in Table 2) . We refrain from providing more precise identifications, since these species require a separate taxonomic revision, based on a more extensive material than specimens merely from Iraq. For B. lutheri and B. vardarensis, local Caucasian subspecies were described (Zimmermann 1981). However, the delimitation from the nominal subspecies remains questionable, we thus refrain from identification on the subspecies level.
The occurrence of some earlier reported species we consider as highly unlikely and resulting from erroneous identification.  (2008)). In this case, all three species were identified by searching for the most similar 16S sequence in GenBank using BLAST search (Khudhur and Shekha 2020). However, the sequences of a vast majority of Middle Eastern species are not deposited in GenBank; thus, they cannot be associated with a query sequence using BLAST. Moreover, 16S is a relatively conservative gene, so a high degree of similarity does not necessarily imply conspecificity. Therefore, we suggest that all of the abovementioned species no longer be considered as a part of the Iraqi fauna.
The species composition of the area which we sampled in Iraqi Kurdistan is similar to that of the neighbouring mountainous areas of NW Iran and SE Turkey and exhibits some affinities to the Caucasian mayfly fauna. This is evidenced by the occurrence of B. ilex, B. braaschi, E. ornatipennis, E. pseudaffinis and E. nigripilosus. These represent Caucasian species, sometimes with a territorial extension to the Middle East and Central Asia. On the other hand, some typically Middle Eastern species have also been recorded, such as B. samochai (known from Turkey, Israel, Lebanon, Syria and Iran). We have also collected common Palaearctic or west Palaearctic species, some of which have unclear taxonomy and our specimens might actually represent separate Middle East lineages, different from the European ones (B. cf. pentaphlebodes, B. cf. gadeai, C. cf. dipterum).
The future prospects of mayfly research in Iraq include widening the sampling into the southern part of the country, with more affinities to the Levant and Arabian Peninsula. As for the taxa of the mountainous area in the north, it would be useful to include the material from Iraq in more general revisions of Caucasian taxa with complicated taxonomy (such as genera Baetis or Ecdyonurus).