This subgenus is widely distributed in the Old World (Trunz et al. 2016). Of the criteria given here, not all will apply to Anodonteutricharaeaoutside the Palearctic; characters that appear constant in this subgenus throughout its range are the elongate basal seta on the claws in the female (Fig. 7), which is the most salient diagnostic trait to identify the females of Anodonteutricharaea; and the absence of an inferior projection on the mandible and the characteristic shape of the gonostylus in the male (Fig. 50). Females: Palearctic females of Anodonteutricharaea agree with Eutricharaea in their small to medium size, mostly white or yellowish white scopa (although it is orange in Megachile albohirta). The sterna do not have continuous, apical fasciae beneath the scopa. In contrast to Eutricharaea, the mandible is similar to that of the subgenus Megachile s. str. (Figs 30–31), with a partial cutting edge in the second interspace and no visible cutting edge in the third interspace in the Palearctic species; often in Anodonteutricharaea, the cutting edges are entirely recessed and invisible in front view (Fig. 31). Also as in Megachile s. str., the mandible is 5-toothed, or if 4-toothed, the upper tooth is truncate and weakly divided into two teeth, thus approaching the 5-toothed condition (Fig. 30). In some species, there is a straight, carinate ridge in the middle of the interacetabular interspace; this ridge is placed above the outer ridge (see Gonzalez and Engel 2012: Fig. 5) and extends form the mandibular base to the middle of interspace 1 (Figs 30–31). In most other group 1 subgenera, this ridge is less visible, less carinate or regularly curved. Males: Males of Anodonteutricharaea can be diagnosed by the absence of projection along the inferior margin of the mandible, unlike most other group 1 subgenera; the mandible is either 3- or 4-toothed. The gonostylus is distinct and diagnostic (Fig. 50): it is broadened apically, with numerous hairs on the external surface. In the Palearctic, the few known species are superficially similar to Eutricharaea in the light body vestiture, the presense of a front coxal spine (spine short in M. villipes), the T6 with disc covered by dense, white tomentum and multidentate preapical carina, and the small, unmodified T7 (except in M. albohirta, where T7 is produced to a short median tooth). In most Palearctic species, the front tarsi are slightly enlarged, yellowish-white or yellowish-brown and have a black macula ventrally on tarsal segments 2 (most species) or 1 and 2 (M. albohirta). Palearctic Anodonteutricharaea are readily distinguished from Eutricharaea in the hyaline apical margin of S4 (Fig. 12).

Figures 16–23. 

Female clypeus and mandibles, front view. 16Megachile (Creightonella) amabilis17M. (Maximegachile) maxillosa; the white line indicates the hypostomal tooth 18M. (Callomegachile) sculpturalis19M. (Chalicodoma) hungarica20M. (Pseudomegachile) ericetorum21M. (Pseudomegachile) incana22M. (Pseudomegachile) sp. aff. incana23M. (Pseudomegachile) foersteri.

The Palearctic species of Anodonteutricharaea form a rather homogenous group; how this group relates to the eight species groups recognized by Pasteels (1965) for the African fauna remains to be investigated. The following western Palearctic species belong to this subgenus: Megachile albohirta, M. inornata, M. thevestensis, and M. troodica. M. mandibularis and M. villipes are probably restricted to Central Asia and records of the latter from the western Palearctic (Schulthess 1924, Zanden 1989, Özbek and Zanden 1994) likely refer to other species. I have seen at least three additional, undescribed species in the western Palearctic.

Very little is known on the biology of this subgenus. Ferton (1920) described nests of Megachile thevestensis. The cells were placed individually or in groups of two in existing cavities between rocks or in burrows in the soil; it was unclear whether the female had dug the burrow or used a preexisting burrow. All cells investigated consisted of an external, rigid layer of leaf fragments (not circularly cut as in most other group 1 subgenera) and a layer of petals inside. I have captured M. villipes on Alhagi(Fabaceae) in Uzbekistan, while M. inornata and a closely related, undescribed species appear oligolectic on Lamiaceae. M. troodica also appears to have a distinct or exclusive preference for Lamiaceae (Mavromoustakis 1953), and in M. thevestensis and another, undescribed species from Morocco, the hairs on the median parts of the clypeus are short and simple, suggesting pollen collection from nototribic flowers such as Lamiaceae (see comments under the subgenus Pseudomegachile).

In the few species present in the Palearctic, the preoccipital carina is always well developed and constitutes a good diagnostic trait. Females: Females can be recognized by the mandible without cutting edges and by the dull mandibular surface with comparatively few punctures or ridges [the mandible is also dull in some Chalicodoma of the montenegrensis group, and to some extend in Megachile (Pseudomegachile) foersteri Gerstäcker, 1869 (Fig. 23)]. The clypeus is always modified in the species of Callomegachile present in the Palearctic, either broadly truncate, with truncation wide, shiny, vertical or overhanging the base of the labrum (Fig. 18). Males: Males of the species present in the Palearctic always have a front coxal tooth and a quadrate projection along the inferior margin of the mandible. In many Callomegachile (but not in M. sculpturalis Smith, 1853), the first segment of the front tarsi forms a ventral concavity. The preapical carina of T6 is entire, without denticulation. The clypeus is mostly glabrous basally and with a dense fringe of hairs apically; in most other Palearctic Megachile, the clypeus is entirely covered by hairs.

Megachile sculpturalis, an Oriental species, has recently been introduced to western Europe (Vereecken and Barbier 2009, Amiet 2012, Quaranta et al. 2014, Westrich et al. 2015); otherwise, only a few species of Callomegachile occur on the Arabian Peninsula, including M. simonyi Friese, 1903 and M. cephalotes Smith, 1853 among others. At least two further species occur on the Arabian Peninsula; they probably belong to previously described African species. As mentioned above, M. breviceps likely does not belong to the Palearctic fauna.

Most species of the subgenus Callomegachile investigated so far predominantly use resin to build their nests (Michener 2007), although other materials (wood chips, mud or even leaf fragments) are sometimes used (Piel 1930). Megachile sculpturalis nests in existing cavities in wood and uses resin to separate the cells; the nest plug is also built with resin but sometimes covered with a thin layer of mud (Quaranta et al. 2014, Westrich et al. 2015). Whether mud is also used for the cell partitions is unclear. In contrast, M. cephalotes appears to build entire cells with mud; the cells are placed in hollow stems (Gupta et al. 2004). Apparently, a “brown-colored detachable membranous layer” (Gupta et al. 2004: 58) lines the inside of the mud walls, but it is unclear what this layer refers to as it exhibits a nipple-shaped projection apically, which seems to refer to the cocoon. The floral choices of the widely polylectic species M. sculpturalis have been studied using pollen analyses (Quaranta et al. 2014, Westrich et al. 2015).

Females: Females can be recognized by the combination of the convex, denticulate apical margin of the clypeus and the elongate mandible with a straight margin above the two apical teeth (Fig. 19) (rarely with minute tooth 3, e.g., in the montenegrensis group). Only some species of Pseudomegachile have similarly elongate mandibles, for example Megachile ericetorum (Fig. 20) and M. lanata. The former can easily be separated from Chalicodoma by the different apical margin of the clypeus (Fig. 20) and the long ocelloccipital distance, which is markedly longer than the interocellar distance; M. lanata has a short clypeus with apical margin entire (as in Fig. 24), not denticulate. In some species of the cyanipennis group of Pseudomegachile, notably M. saussurei Radoszkowski, 1874, the apical margin of the mandible is nearly straight, with reduced teeth (Fig. 25), thus approaching the condition found in Chalicodoma. In such species however, the mandible is less elongate, with the outer margin strongly convex. In M. (Pseudomegachile) incana Friese, 1898, the mandible is 5 to 6 toothed (Fig. 21), and in old specimens the teeth may be little visible and the condition thus similar to that seen in Chalicodoma; as in Chalicodoma, the ocelloccipital distance is shorter than the interocellar distance in M. incana. All species of the incana group can easily be diagnosed by the large body size, the light-grey metasomal vestiture without dense tergal fasciae and the comparatively broad hind basitarsus (Fig. 27). Males: Males Chalicodoma fall into three distinct species groups and there are few diagnostic characters common to all. In all species the mandible is comparatively elongate, 3-toothed and without inferior projection, and the preapical carina of T6 is denticulate.

Figures 24–29. 

Female clypeus and mandibles, front view. 24.Megachile (Pseudomegachile) seraxensis25M. (Pseudomegachile) saussurei26–29 Female hind basitarsus, lateral view 26M. (Pseudomegachile) ericetorum27M. (Pseudomegachile) incana28M. (Eutricharaea) giraudi29M. (Eurymella) patellimana.

Tkalců (1969) and Rebmann (1970) have independently divided the subgenus Chalicodoma into the same four groups, to which they gave subgeneric rank. I recognize three groups, not four, because Megachile hirsuta, unknown to both Tkalců and Rebmann at that time, renders the distinction between two of their groups difficult in the female sex. Recognizing these groups as subgenera appears little useful for species identification in the Palearctic and I recognize them as species groups.

1. montenegrensis group (Euchalicodoma Tkalců, 1969; Xenochalicodoma Tkalců, 1971; Allomegachile Rebmann, 1970; Katamegachile Rebmann, 1970). Males: Front coxa with large, spatulate tooth. T6 with lateral tooth (Fig. 48) [small in Megachile rufitarsis (Lepeletier, 1841)]. T7 mostly produced to large, rounded, median tooth (Figs 15, 48) (tooth small in M. rufitarsis), or trifid. Gonostylus simple, slightly broaden apically (Fig. 48). In this species group, the front tarsi are variously modified. Unusual characters of males of some or all species of this group are the partly exposed S5 (character not visible in Fig. 15), the apically strongly convex margin of S6 (Fig. 15), as in Creightonella, and, in some species, in the lack of hairs laterally on S8, unlike other group 2 subgenera. Females: Surface of mandible mostly dull, with few shiny ridges or punctures, except in M. rufitarsis, in which the mandible is as in the parietina group. S6 with depressed apical zone, with strong preapical carina separating the elevated, basal part from the depressed apical zone, except in M. montenegrensis Dours, 1873 and M. hirsuta, both of which have dull mandibles. In M. montenegrensis, the vertex is slightly concave laterally, a unique feature in Palearctic Chalicodoma (see Tkalců 1969).

2. lefebvrei group (Allochalicodoma Tkalců, 1969; Heteromegachile Rebmann, 1970). Males: Front coxa without tooth. T6 with a small lateral tooth (sometimes reduced to a mere angle, as in Fig. 45). T7 rounded apically, unmodified. Gonostylus tapering apically, thickened preapically and without preapical, projecting lobe (Fig. 45). Females: Surface of the mandible covered with numerous shiny ridges and punctures (Fig. 19), as in the parietina group. S6 with depressed apical zone; carina separating the elevated part from the depressed marginal area interrupted medially and only visible laterally. In addition, all females of the lefebvrei group have conspicuously modified vestiture on the clypeus (Fig. 19) and the frons: the hairs are short, simple and bent apically (Müller 1996). Such modified hairs are not found in the montenegrensis group and only rarely found in the parietina group.

3. parietina group. Males: Front coxa without tooth. T6 without lateral tooth. T7 rounded or truncate. Gonostylus with preapical lobe (Fig. 44). Females: S6 mostly not divided in two zones (weakly so in some species, such as Megachile nasidens Friese, 1898), without preapical carina. Hairs on clypeus mostly branched, except in some rare species [e.g. M. marina Friese, 1911 and M. palaestina (Tkalců, 1988)].

Females of this subgenus are sculpturally uniform and frequently exhibit mimetic color evolution; hidden sternites of males are mostly diagnostic but these structures have only been described for few species (e.g. Tkalců 1969, 1974). Consequently, the taxonomic status of numerous “geographic” forms within Chalicodoma remains unclear and a complete species list is not given here. In the West Palearctic, there are at least five species in the montenegrensis group [Megachile hirsuta, M. manicata Giraud, 1861, M. mauritaniae (Tkalců, 1992), M. montenegrensis and M. rufitarsis), two to five species in the lefebvrei group [M. heinii Kohl, 1906, known only in the female sex; and depending on the species concept adopted one to four additional, parapatric species: M. albocristata Smith, 1853, M. hungarica Mocsáry, 1877, M. lefebvrei (Lepeletier, 1841) and M. roeweri (Alfken, 1927)]; and approximately 20 species in the parietina group, of which three are undescribed.

The nesting biology of Megachile parietina has been described in detail (reviewed in Westrich 1989 and Müller et al. 1997). This species builds exposed nests made of hard mud in rock crevices, more rarely on twigs (Rebmann 1969, Vereecken et al. 2010). These exposed nests are particularly hard and resistant; Kronenberg and Hefetz (1984) have demonstrated that females of M. sicula (Rossi, 1794) add labial gland secretions to the mud; these secretions rapidly harden and render the nest hydrophobic. Accounts of the nesting biology of the few species of the parietina group investigated so far indicate that the cells are build in a similar way: in M. pyrenaica Lepeletier, 1841, the cells are often hidden in holes in walls or under stones (Le Goff 2007), or placed in existing holes in steep, hard soil slopes, under overhanging rocks (Müller et al. 1997), or as described by Fabre (1879, 1882) under the roof tiles of old barns; sometimes the nests are exposed on stones as in M. parietina (Le Goff, 2007); in M. rufescens (Pérez, 1879) the nests appear to be mostly placed on twigs (Fabre 1882). M. sicula builds nests both on twigs and on rock surfaces (Kronenberg and Hefetz 1984, Vereecken et al. 2010). Few studies have documented the nesting biology of species of the other species groups: M. manicata appears to nest exclusively in existing holes in rocks (Le Goff 2012, Gogala 2014, C. Praz, unpublished data); Le Goff (2012) described one nest containing two cells made of hard mud mixed with pebbles; the nests observed were closed with hard mud. Nests of M. lefebvrei have been described in detail by Ferton (1908, 1920); the biology of this species slightly deviates from the typical nesting biology seen in the subgenus. Females build 2–4 cells in holes in rocks; these cells have the general appearance of those built by M. parietina, thus they consist of mud mixed with “salivary secretions” (Ferton 1908: 545), without pebbles. Once several cells are built, they are covered with a thin (1 mm), concave layer of hardened mud; this layer is located inside the hole of the rock and its outer surface is a few millimeters beneath the external surface of the rock. Subsequently, the female fills the space above the thin mud layer with a mix of pebbles and masticated plant material. According to Ferton, the masticated plant material contains salivary secretions (but no resin), and it hardens quickly. Ferton (1920) reports a nest of M. lefebvrei from southern France; the nest structure and the material used were similar but the nest had been built in an empty snail shell.

Many species of Chalicodoma, including Megachile hirsuta, M. montenegrensis, M. manicata, M. parietina, and M. pyrenaica have a distinct or exclusive preference for Fabaceae (Westrich 1989, Müller et al. 1997, Gogala 2014, C. Praz, unpublished data). All species of the lefebvrei group are likely polylectic with a preference for Lamiaceae (Müller 1996; C. Praz, unpublished data). The pollen spectrum of the other species remains poorly investigated.

The subgenus Creightonella is in many ways intermediate between the leafcutter and the dauber bees. Females: Females can easily be diagnosed by the shape and structure of the mandible (Fig. 16): the mandibular surface is distinct, with numerous, elongate punctures but comparatively few long ridges, and is covered by numerous hairs. The apical margin has 5 or 6 teeth, with tooth 1 broad and larger than the other teeth, and the teeth 2–5 (-6) becoming progressively smaller. There is a conspicuous, partial cutting edge in the second interspace and a small, little visible cutting edge in the third and sometimes fourth. In spite of being a leafcutter, Creightonella does not have the typical tapering metasoma of most other group 1 members and the hind basitarsus is slender (as in Fig. 26). The hind claw has only one elongate seta, a unique condition in Palearctic Megachile. Males: Males of Creightonella have a strong front coxal tooth and a rounded projection along the inferior margin of the mandible (Fig. 9). The front tarsi are unmodified although in Megachile albisecta (Klug, 1817) they are yellowish-brown (Fig. 9) and the second tarsal segment has a dark spot on the ventral side. The preapical carina of T6 is mostly denticulate (weakly so in M. arabica Friese, 1901) and laterally the carina extends at right angle towards the base of the tergum. T7 is mostly triangular in dorsal view, with a strong longitudinal carina (Fig. 46), except in M. doriae, where T7 is truncate, but with a spine on the basal part of the disc (Fig. 47). The apical margin of S5 is exposed in repose and the apical margin of S6 is convex and rests on the apical margin of T7 in repose (Fig. 11).

Figures 30–37. 

Female clypeus and mandibles, front view. 30Megachile (Anodonteutricharaea) albohirta31M. (Anodonteutricharaea) thevestensis32M. (Megachile) versicolor33M. (Megachile) lapponica34M. (Megachile) bombycina35M. (Xanthosarus) lagopoda36M. (Xanthosarus) nigriventris37M. (Xanthosarus) willughbiella.

Most Palearctic species fall into a rather homogenous group of species referred to here as the albisecta group (Metamegachile Tkalců, 1967), even if the structure of T7 is rather different in Megachile doriae; there are at least four valid species in this group: M. albisecta, M. amabilis Cockerell, 1933, M. doriae and M. morawitzi Radoszkowski, 1876; a fifth is undescribed (A. Monfared and C. Praz, in prep). The status of M. ghigii Gribodo, 1924, described from Libya, and M. aurantiaca Rebmann, 1972 from Iran (a junior homonym of M. aurantiaca Friese, 1905) as well as of numerous Central Asian taxa related to M. albisecta remains unclear. On the Arabian Peninsula, additional species with Afrotropical affinities occur: Pasteels (1979) placed M. arabica in the aurivilli group (see Pasteels 1965) and M. felix (Pasteels, 1979), known only in the female sex, in the angulata group. Mentions of M. sudanica Magretti, 1898 from Jordan (Zanden 1989) possibly refer to M. arabica, and both species may be conspecific (Pasteels 1965: 26). In addition, I have seen a single specimen putatively from Yemen (BMNH) of an African species with a modified clypeus, possibly either M. bicornuta Friese, 1903 or M. cornigera Friese, 1904, both in the African cornigera group.

The nesting biology of Megachile albisecta has been described in detail (Ferton 1901). This species nests in burrows in the ground; whether the females dig their own burrows or rent existing cavities as suggested by Ferton is not clear, although a brief description of a nest by Grandi (1961) mentions a nest “excavated in very hard soil” [the Oriental species M. frontalis (Fabricius, 1804) consistently digs its own burrows in hard, heavy soils (Michener and Szent-Ivany 1960, Willmer and Stone 1989]. Reports of nests in dead wood (Benoist 1940) and in stems (Banaszak and Romasenko 2001) are likely identification errors. Unlike the circular leaf discs cut by M. frontalis (see Michener and Szent-Ivany 1960: Plate 1), the leaf fragments used by M. albisecta are irregular; in M. frontalis, the leaf fragments used for the outer layer of the cells are also irregular (Michener and Szent-Ivany 1960: 31 and Plate 3; Willmer and Stone 1989). In M. albisecta the fragments are stuck together by masticated leaf material, and the entire cell rests on a solid plug of masticated leaf pulp mixed with pebbles. The nest plug consists of circular leaf fragments and pebbles cemented together by masticated leaf pulp. Nests of M. albisecta do not appear to include resin according to the description by Ferton (1901), although Ferton’s subsequent articles (Ferton 1908: 547) suggest that resin was included in the nest. Possibly the leaf pulp is hardened with secretions but not resin. Resin was not found in the nests of M. frontalis (Michener and Szent-Ivany 1960; Willmer and Stone 1989) but was reported from the nest of the African M. cornigera (Michener 1968). M. arabica has been observed cutting leaf discs (H. Priesner, cited in Alfken 1934). M. albisecta (and possibly all species of the albisecta group) is oligolectic on Asteraceae, with a preference for Carduoideae (C. Praz, unpublished data).

Females: The females of most Eurymella have a distinct, robust mandible (Fig. 38), with tooth one markedly larger than the other teeth; the mandible is 4-toothed in Palearctic and Arabian species and divided into two plane surfaces by the outer ridge; the brushes of orange hairs in the grooves near the base of tooth one are lacking or reduced (Fig. 38). A complete, well-visible cutting edge is present in the third interspace (Fig. 38). Some Eutricharaea (Megachile deceptoria Pérez, 1890 and a few related species; see under the subgenus Eutricharaea) also have a robust mandible, with tooth one comparatively large, and reduced brushes of hairs near the base of tooth one (Fig. 40). In Eurymella, the hind basitarsus is comparatively broad, its length approximately 2.3 times its maximal width (Fig. 29). In Eutricharaea, the hind basitarsus is usually less broad (Fig. 28) (length approximately 2.7 times its maximal width), except in a few species such as M. marginata. In Eurymella the claw of the hind leg bears two thin setae (Fig. 6); the condition is not as clear as in Anodonteutricharaea because the basal seta is short in Eurymella. As in Eutricharaea but unlike most other group 1 subgenera, the sterna have conspicuous apical fasciae beneath the scopa in the Palearctic species. Males: Males of M. patellimana Spinola, 1838, the only Palearctic species outside the Arabian Peninsula, are easy to diagnose using the criteria mentioned in the key, especially the uniquely shaped T6 (Fig. 49) and the absence of inferior mandibular tooth (see below for the males of other species from the Arabian Peninsula).

Figures 38–41. 

Female clypeus and mandibles, front view. 38Megachile (Eurymella) patellimana39M. (Eutricharaea) rotundata40M. (Eutricharaea) deceptoria41M. (Eutricharaea) giraudi.

This subgenus is diverse in Africa, where it forms numerous, morphologically distinct species groups (Pasteels 1965); very few morphological features characterize all males of Eurymella, while the females are more homogeneous. In the Palearctic, Eurymella is probably represented by only one distinctive species, Megachile patellimana, included in the patellimana group (Pasteels 1965). Pasteels (1979) mentions two species of the eurimera group (see Pasteels 1965) from the southern parts of the Arabian Peninsula: M. gibboclypearis Pasteels, 1979 and “M. aff. eurimera Smith, 1854”.

Both Arabian species of the eurimera group are known only in the female sex, although the male of the widely distributed Afrotropical species Megachile eurimera is well known (Pasteels 1965, Eardley 2013). The males of the eurimera group differ from those of the patellimana group in the pointed, inferior projection of the mandible and the simple T6 with preapical carina not produced posteriorly. In the present key, males of M. eurimera would run to couplet 18 (Eutricharaea and Xanthosarus); they differ from Eutricharaea in the strikingly different genitalia (see Pasteels 1965: p. 83), the presence of a wide, hyaline margin apically on S4 (as in Fig. 14), and the apical margin of T6 (beneath the preapical carina) with two teeth (as in Fig. 14). Xanthosarus is presumably absent from the Arabian Peninsula.

Little is known on the biology of Eurymella; only a brief account is given by Pasteels (1965: 127) for Megachile semifulva Friese, 1922: the nests of this species are placed in burrows in the ground and consist of leaf discs. Gess and Roosenschoon (2017) described nests of M. patellimana in the United Arab Emirates. The nests were placed in excavated burrows in compacted sand; it was not clear whether the burrows had been dug by M. patellimana or were pre-existing. A female was captured carrying a cut leaf piece, while another was captured carrying a piece of tough green plastic.

Females: In females the scopa is mostly white at least basally (often dark apically) and the sterna have distinct apical fasciae beneath the scopa (lacking in Megachile giraudi and M. hohmanni Tkalců, 1993). In a few rare cases, the scopa is orange or entirely black (some populations of M. melanogaster, M. hohmanni). The upper mandibular tooth is usually truncate (e.g. M. leachella Curtis, 1828), although it is sometimes acute (e.g. M. deceptoria; Fig. 40) or conversely clearly divided into two teeth [e.g. M. rotundata (Fabricius, 1787); Fig. 39]. In some species, such as M. orientalis, the mandible is clearly 5-toothed (Fig. 2). There is no visible cutting edge in the second interspace (Figs 39–41), in contrast to the subgenera Xanthosarus, Megachile or Anodonteutricharaea (in the latter all cutting edges are often strongly reduced); even in the third interspace, the cutting edge is often small and partly hidden behind the margin in many Eutricharaea. Males: In the Palearctic, males of this subgenus can be recognized by the following criteria: the preapical carina of T6 is mostly denticulate (weakly so in some species such as M. rotundata) and the disc of T6 covered with dense, light vestiture entirely hiding cuticula (except in some populations of M. giraudi). T7 is small and little visible. The mandible is 3-toothed, always with triangular inferior projection. In contrast to the subgenera Anodonteutricharaea and Megachile and to most Xanthosarus and Eurymella, S4 does not have a hyaline rim apically (Fig. 10). The front tarsi may be modified or not, and the front coxal tooth is present (although minute in the apostolica group).

Figures 42–49. 

Apex of male metasoma; 42–46, 48T5-T7 and apex of genital capsule, dorsal view 47T5-T7 and apex of genital capsule, lateral view 49T4-T6, dorsal view 42Megachile (Megachile) melanopyga43M. (Pseudomegachile) sp. aff. incana44M. (Chalicodoma) parietina45M. (Chalicodoma) hungarica46M. (Creightonella) albisecta47M. (Creightonella) doriae48M. (Chalicodoma) manicata49M. (Eurymella) patellimana.

1. rotundata group (Neoeutricharaea Rebmann 1967). Males: Gonostylus simple, not bifid apically, s-shaped in lateral view (Fig. 54). Front coxa with tooth. T2 and sometimes T3 with fovea laterally. Apical margin of S4 unmodified, medially without tubercule or spot of dense, golden hairs (Fig. 10). Pretarsal claws of all legs with two similar, thin setae (as on Fig. 4). Front legs modified or not. Females: T2 and sometimes T3 with fovea laterally. Ocelloccipital distance often larger than diameter of lateral ocellus. Clypeus apically with comparatively wide, impunctate margin. Scopa usually white, black on S6, sometimes also S5, rarely entirely black.

Note. This group is particularly diverse and additional species groups may be recognized for isolated, divergent species: Megachile giraudi, with a particularly long interspace 3 in the female sex (Fig. 41) and front basitarsus modified, with a long apical projection in the male sex; M. orientalis, with the female mandible 5-toothed with cutting edges hidden behind the mandibular margin (Figs 2–3) (the male is a regular member of the rotundata group); and M. hohmanni, included in a distinct, monotypic subgenus by Tkalců (1993). Another group includes species with robust, 4-toothed mandible (Fig. 40), with tooth one larger than the other teeth and with reduced tufts of hairs apically, approaching the condition seen in Eurymella; males of this group are mostly characterized by the strongly convex and medially produced apical margin of T5. In the Palearctic, this group includes M. communis Morawitz, 1875, M. deceptoria, M. dohrandti Morawitz, 1880, M. sedilloti, M. jakesi Tkalců, 1988, probably M. arachosiana Gonzalez, Engel and Hinojosa-Díaz, 2010 as well as at least one undescribed species. Because all these species do not have particularly distinct gonostylus, I do not recognize separate species groups for them. Of note, the rotundata group also occurs in Africa; Pasteels (1965) referred to it as the malangensis group. I have examined the type material of M. malangensis Friese, 1904 (ZMHB); this species is a member of Eutricharaea and not of Paracella (=Anodonteutricharaea) as suggested by Eardley (2013).

2. naevia group. Males: Similar to males of the rotundata group, but gonostylus apically shortly bifid (Fig. 55). Front tarsi modified, yellowish-white. T2 with a weak fovea laterally. Females: As in the rotundata group, but mandible always red. Apical margin of clypeus straight with comparatively wide impunctate area. Ocelloccipital distance shorter than diameter of lateral ocellus. T2 but not T3 with fovea laterally.

3. leachella group. Males: Gonostylus either bifid (Fig. 56) with long (e.g. Megachile leachella) or short (e.g. M. walkeri Dalla Torre, 1896) preapical process, or simple (e.g. M. concinna Smith, 1879), in all cases apically with rounded emargination (Fig. 56). Front coxa with tooth. T2 and T3 without fovea laterally (although the corresponding area on T2 is often finely punctured and covered with brown hairs which contrast with white hairs on rest of the tergum). Apical margin of S4 either with a patch of yellow hairs medially, or a minute tubercle covered with numerous short, white hairs. Pretarsal claw with basal hair short, thickened, as in the female (as in Fig. 5), and apical hair long. Front legs not modified although front tarsi sometimes yellowish brown. Females: T2 and T3 without fovea laterally (although the corresponding area on T2 is often finely punctured and covered with brown hairs which contrast with white hairs on rest of the tergum). Ocelloccipital distance commonly equal to the diameter of the lateral ocellus. Clypeus apically with narrow impunctate zone. Scopa white on S1-S5, black on S6, sometimes partly or entirely orange-red, never black on S5.

Note. Megachile ventrisi Engel, 2008, from Saudi Arabia, Yemen (Engel and Schwarz 2011) and at least Kenya (C. Praz and L. Packer, unpublished) belongs to another African and Oriental group in which the male front tarsi are modified (the ventral surface of the first tarsal segment is concave) and the gonostylus simple but differently shaped, without apical emargination. This group was included in the leachella group (as the “argentata” group) by Pasteels (1965); it includes the African species known as M. frontalis Smith, 1853, which is a junior homonym of M. frontalis (Fabricius, 1804); “M. frontalis Smith” is a member of Eutricharaea and not of Paracella (=Anodonteutricharaea) as suggested by Eardley (2013).

4. apostolica group. Males: As leachella group, with the following exceptions: gonostylus bifid apically (as in Megachile pilidens Alfken, 1924), but apex conspicuously slender and preapical process long. Front coxa with minute tooth (see comments above). Front tarsi yellowish-white. Females: As leachella group, with the following exceptions: all legs predominantly orange; terga brown; hairs on ventral side of mid and hind femora modified, short, apically thickened. M. walkeri (included in the leachella group) also has similarly modified hairs (Engel 2008), but in M. walkeri the cuticula of T1 and T2 is orange.

5. leucomalla group. Males: Male gonostylus as in rotundata group, but bent apex longer (Fig. 57). Front tarsi greatly enlarged, yellowish-white. S2 and S3 (but not S4) medio-apically with dense patch of yellow hairs. T2 and T3 without clearly delimitated fovea. Pretarsal claw of all legs with two similar hairs. Females: Body length above 13 mm; punctation of mesonotum coarse and sparse, interspaces shiny and nearly as large as puncture diameters. Apical clypeal margin truncate, premarginal, impunctate zone comparatively narrow. T2 and T3 without clearly delimitated fovea.

This is a large and taxonomically complex subgenus in need of revision. The number of unpublished synonymies is large and a list of species is not given here. There is only one species in the leucomalla group, M. leucomalla Gerstäcker, 1869. I know one Palearctic species in the apostolica group, Megachile soikai; this species is possibly conspecific with one of the African species related to M. apostolica Cockerell, 1937 and listed by Gonzalez et al. (2010: 65). In the naevia group, I know two species: M. naevia Kohl, 1906 from the Socotra Archipelago and an undescribed species from the Arabian Peninsula, North to Israel. There are more species in the leachella group [possibly seven; see Soltani et al. (in press) for a treatment of the taxa allied to M. concinna], and a large number of species, a few of which undescribed, in the rotundata group, for a possible total of 29 species in the western Palearctic.

Megachile rotundata, a European species introduced into North America for the pollination of alfalfa, has been studied in detail (reviewed in Pitts-Singer and Cane 2011). All species of Eutricharaea use leaf discs to build their brood cells, although M. rotundata (and M. giraudi, see below) sometimes use petal fragments and not leaves (Westrich 1989, and references therein). In M. minutissima Radoszkowski, 1876, the cell walls are sometimes omitted when the nest is located in cavities of small diameters, so that leaf discs are only used to build the cell partitions (Krombein 1969). The outer nest plug of M. marginata consists of a series of leaf discs covered by a layer of mud, possibly mixed with “saliva” (Ferton 1914). Nests of Eutricharaea are mostly built in existing cavities such as stems or beetle burrows (e.g. M. rotundata), under stones or in existing cavities in the ground (e.g. M. pilidens: Müller et al. 1997; M. marginata: Ferton, 1914; M. minutissima: Alqarni et al. 2014), or more rarely in burrows dug by the bee in sandy soil (M. leachella: Westrich 1989; note that this species may accept trap nests under artificial conditions: Holm and Skou 1972) or in hard soil (M. deceptoria: Benoist 1940, Grandi 1961, Zettel et al. 2005, Mazzucco and Mazzucco 2007). M. giraudi appears to nest specifically in holes in rocks and to use petals for cell construction (Maneval 1939, Banaszak and Romasenko 2001; C. Praz and A. Müller, unpublished). The description of nesting aggregations in the soil for M. leucomalla in Central Asia (Marikovsakya 1968; see also Banaszak and Romasenko 2001) likely refers to another species, probably M. communis, based on Fig. 1 in Marikovsakya (1968).

The pollen preferences of species included in the subgenus Eutricharaea are varied; most species are probably polylectic with a preference for Fabaceae (e.g. Megachile rotundata and M. pilidens: Westrich 1989; see also O’Neill et al. 2004; M. minutissima: Alqarni et al. 2014); Müller and Bansac (2004) have shown that some species have a distinct (e.g. M. apicalis) or exclusive preference for Carduoideae (M. marginata, M. melanogaster and M. flabellipes Pérez, 1895), and in these species a modified pollen collecting device is found under the hind trochanter and femur. Soltani et al. (in press) analyzed the pollen preferences of the taxa allied to M. concinna; all taxa were polylectic, although M. leucostoma Pérez, 1907 and M. anatolica Rebmann, 1969 showed a preference for Plantaginaceae (Linaria-type) and Lamiaceae, respectively. Two species have long and conspicuous hairs on the galea, M. dolosa Alfken, 1936 and M. posti; all known specimens of M. posti from Cyprus were collected on Noaea mucronata (Forssk.) Asch. & Schweinf. (Chenopodiaceae) (Mavromoustakis 1952). Strikingly similar hairs are found on the galea and first labial palpi of some Hoplitis [e.g. H. karakalensis (Popov 1936); A. Müller, pers. comm., January 2016]; this pollen collecting apparatus likely constitutes an adaptation for collecting pollen from small flowers of Chenopodiaceae (Müller 2016).

This species-poor, group 2 subgenus is probably represented in the Palearctic by one conspicuous species that is well characterized in the key. In both sexes, Maximegachile consists of large, elongate bees with a typical vestiture pattern: hairs are predominantly black except snow white on the propodeum, on T1 and the basal part of T2. No other Palearctic species has such a vestiture pattern. Females: In addition, females have an elongate, 3-toothed mandible, a modified clypeus and a strong hypostomal tooth (Fig. 17). Males: The males are characterized by the absence of both front coxal spine and inferior mandibular projection, by the bilobed carina of T6 and the strong preoccipital carina (see additional comments to the male morphology below).

Figures 50–53. 

Apex of male metasoma, dorsal view. 50Megachile (Anodonteutricharaea ) sp. aff. inornata51M. (Pseudomegachile) ericetorum52M. (Pseudomegachile) saussurei53M. (Pseudomegachile) foersteri.

Two species have been mentioned for the Palearctic region and for the Arabian Peninsula: Pasteels (1979) mentions Megachile maxillosa Guérin-Méneville, 1845 from the Arabian Peninsula and describes M. esseniensis (Pasteels, 1979) from a single male collected in Southern Israel. Based on the examination of numerous specimens of M. maxillosa from Africa, it seems that M. maxillosa occurs as two distinct morphs in the male sex: a large morph (body length approximately 20 mm) with clypeus entirely glabrous and covered by coarse punctures on its disc; and a smaller morph (body length approximately 15 mm) with clypeus covered by dense vestiture at least apically and with finer punctures. A similar conclusion seems to have been reached by Eardley (2012: 26), who mentions that the lower clypeus is “densely pubescent (naked in very large specimens)”. I have seen both morphs from the Arabian Peninsula but the limited material that I have seen from Israel were of the small morph, which corresponds to the description of M. esseniensis. For now, I thus place M. esseniensis as a junior synonym of M. maxillosa (syn. n.), in line with the treatment of M. maxillosa in Africa, and consequently I recognize only one species in the Palearctic region.

Gess and Roosenschoon (2017) provide a description of the nesting biology of Megachile maxillosa in the United Arab Emirates. Nests of this species were found in trap nests. Entire cells were built using a mixture of sand and resin and the completed nests were closed with a plug of sand and resin. According to Gess and Gess (2003), the same species visits plants from various families but shows a preference for the Fabaceae, notably Crotalaria, in South Africa and Namibia.

Females: In females of this subgenus, the mandible always has a conspicuous, partial cutting edge in the second interspace (reduced in Megachile genalis, which is easy to diagnose due to the thickened buldge at the base of the mandible: Amiet et al. 2004, Scheuchl 2006), and no visible cutting edge in the third interspace (Figs 32–34). The mandible is mostly 5-toothed with the distance between both upper teeth subequal to the distance between the other teeth (not clearly so in M. genalis) (Figs 1, 32); in some species teeth 4 and 5 are poorly separated (Fig. 33). In M. bombycina, the mandible is elongate and 4-toothed but the partial cutting edge in the second interspace is well visible (Fig. 34). In most species, the clypeus is flat or slightly depressed preapically, and there the punctures are coarse and the interspaces wide, shiny (Fig. 32); this condition is not found in M. lapponica Thomson, 1872, in which the clypeus is densely punctured (Fig. 33), and in M. bombycina, in which the clypeus is modified, short, laterally truncate and medially produced to a blunt triangular process, and with its apical margin forming a wide, smooth and slightly concave area (Fig. 34). In the female sex, many species of Megachile s. str. are sculpturally similar to Anodonteutricharaea; in the former, the scopa is mostly orange basally and rarely white, while in Anodonteutricharaea the scopa is mostly white to yellowish white; and Anodonteutricharaea can be distinguished by the presence of two unmodified setae on the pretarsal claws; in Megachile s. str., the basal seta is modified to a thicken process, as in Eutricharaea or Xanthosarus (as in Fig. 5). Males: Males of the subgenus Megachile s. str. are rather homogenous both in their sculpture and in genitalic structures, and often difficult to identify. All males of the subgenus Megachile s. str. lack a front coxal spine, although in some species (M. ligniseca, M. melanopyga Costa, 1863), the front coxa forms a weak angle. The mandible is always 3-toothed, mostly with inferior projection (lacking in M. pilicrus and M. armenia Tkalců, 1992). In most species the disc of T6 is devoid of white or light pubescence (except in M. pyrenaea Pérez, 1890, M. armenia, M. pilicrus and M. melanopyga), the preapical carina of T6 is never denticulate, although there is a lateral tooth at the lateral margin of T6 (Fig. 42). T7 is either small or exposed, but mostly unmodified (without tooth apically). The gonostylus is always simple, never bifid apically (Fig. 42).

Figures 54–57. 

Genital capsule, lateral view. 54Megachile (Eutricharaea) apicalis55M. (Eutricharaea) sp. aff. naevia56M. (Eutricharaea) leachella57M. (Eutricharaea) leucomalla.

There are at least 14 valid species in the Western Palearctic: Megachile alpicola Alfken, 1924, M. armenia, M. bombycina, M. calloleuca, M. centuncularis (Linnaeus, 1758), M. genalis, M. lapponica, M. ligniseca, M. melanopyga, M. melanota Pérez, 1895, M. octosignata Nylander, 1852, M. pilicrus, M. pyrenaea and M. versicolor. I have not been able to locate the type of M. dacica Mocsáry, 1879; Schwarz et al. (1996) and Westrich (2011) recognized this species as valid; in contrast, according to identified material in his collection (OLML), B. Tkalců considered dacica as a valid subspecies of M. lapponica. The status of M. melanota is unclear, as it may represent a dark color form of M. octosignata.

All species are leafcutters and build brood cells made of leaf fragments (e.g. Malysheva 1958, Westrich 1989, Ruhnke 2000); no Palearctic species appears to use petals as documented in the Nearctic Megachile montivaga (Michener, 2007, and references therein; Orr et al. 2015). The nests are mostly located in existing, above-ground cavities; some species such as M. centuncularis are flexible and nest both in above-ground cavities and in the soil (Westrich 1989, and references therein). Nests of M. pyrenaea and M. octosignata are placed under stones or in the soil (Ferton 1909, Grandi 1961, Westrich 1989), and a nest of M. melanopyga was found “loose in the grass” (Friese 1898). M. genalis exclusively nests in standing stems, favoring fresh stems, thus from plants of the same year; the female digs an opening and places the brood cells vertically above or below the entrance (Ruhnke 2000, and references therein). According to a brief account by Friese (1898), M. pilicrus nests in dry stems of thistles.

Most species of the subgenus Megachile are polylectic, often with a preference for Fabaceae and Asteraceae (Westrich 1989). Megachile lapponica is an oligolege on Epilobium (Westrich 1989, Kühn et al. 2006), while M. genalis is possibly oligolectic on Asteraceae (Westrich 1989); M. pilicrus is oligolectic on Carduoideae and its hind trochanter and femur are covered by modified, stiff hairs, as observed in some Eutricharaea species specialized on these plants (Müller and Bansac 2004).

This subgenus is morphologically diverse and in both sexes few diagnostic traits are common to the entire subgenus. The description is given for the each species group. The rather distinct incana group has previously been recognized as a distinct subgenus, Parachalicodoma, but in Trunz et al.’s phylogeny (Trunz et al. 2016), this group was nested within Pseudomegachile, with Megachile foersteri being sister to a clade formed by the cyanipennis and the incana groups. Based on its morphology, the rhodoleucura group (not included in Trunz et al. 2016) appears to build a transition between the incana and the cyanipennis groups. Similarly, the Oriental group of species previously known as Largella (considered here to form the semivestita group) is rather distinctive, although M. lanata is intermediate between the semivestita group and other, regular looking Pseudomegachile. Thus although Pseudomegachile could be split into several subgenera, numerous species are intermediate and the recognition of several subgenera does not seem practical or necessary: the number of species is not particularly large in the Palearctic or in the Oriental region, and having numerous subgenera would not be of much utility.

1. foersteri group: Males: Large species (body size above 15 mm) with metasoma densely covered by grey to yellow-brown vestiture. Front coxal spine present. Front tarsi modified, first tarsal segment concave inferiorly, segment 2 and 3 brownish-yellow and with black maculae on ventral surface. Mandible 4-toothed with quadrate inferior projection. Preapical carina of T6 multidentate and T7 produced into a long median tooth (Fig. 53). Females: Large (body size above 17 mm) and robust species with metasoma entirely covered by grey to yellow-brown vestiture. Mandibular surface comparatively dull (Fig. 23); mandible slightly elongate (condition intermediate between that seen in the cyannipennis and ericetorum groups), with apical margin weakly 4-toothed. Apical margin of clypeus truncate, laterally with a blunt tooth (Fig. 23). Ocelloccipital particularly long, nearly three times longer than the interocellar distance.

2. cyanipennis group (Xenomegachile Rebmann, 1970): Males: Front coxal tooth present. Mandible 4-toothed (teeth sometimes blunt or small), with quadrate inferior projection. Front tarsi slightly enlarged, brownish-yellow to yellowish-white, first segment with one dark spot on ventral surface (except in Megachile cyanipennis). Preapical carina of T6 multidentate, laterally mostly without tooth (a small tooth is present in M. saussurei). T7 large, well visible from above, produced to a short spine (Fig. 52), or trifid, or broadly truncate. Females: Mandible comparatively short, external margin strongly rounded, apical margin 4-toothed (Fig. 24), although sometimes teeth nearly absent and margin almost straight (Fig. 25); impunctate, premarginal area along apical margin comparatively wide (Fig. 24). Scutellum elevated medially. Ocelloccipital distance shorter than interocellar distance.

3. rhodoleucura group: Males: As in cyanipennis group, with the following differences: front coxal tooth small; mandible without inferior projection; front tarsi unmodified although first segment with dark spot on ventral surface; T7 with apical margin emarginate medially (Alqarni et al. 2012: Fig. 12). Females: As in cyanipennis group, but apical clypeal margin apically swollen, projected over base of labrum (Alqarni et al. 2012: Figs 3–4); modified vestiture consisting of simple, short hairs on clypeus, supra-clypeal area and frons; scutellum not elevated medially.

4. incana group (Parachalicodoma Pasteels, 1966): Males: Medium to large species (body length above 12 mm) with pale metasomal vestiture not forming distinct tergal fasciae. Front coxa without tooth. Front tarsi unmodified. Mandible 4-toothed, without inferior projection. Preapical carina of T6 multidentate, laterally with strong tooth (Fig. 43). T7 produced to rounded or truncate spine (Fig. 43). S5 partly exposed in repose (in all other groups of Pseudomegachile, S5 is hidden under S4). Females: Large, robust species (body length above 13 mm) with pale metasomal vestiture not forming distinct tergal fasciae. Mandible 5 to 6 toothed, teeth either blunt (Fig. 21) or acute (Fig. 22); width of impunctate, premarginal area along apical margin intermediate between cyanipennis and ericetorum groups. Ocelloccipital distance shorter than interocellar distance. Hind basitarsus broad, 2.7 times as long as wide (Fig. 27). Scutellum not elevated medially. In one species the facial vestiture is modified as in the rhodoleucura group.

4. ericetorum group: Males: Front coxa with tooth. Mandible 3-toothed, without inferior projection, although inferior margin swollen medially. Front tarsi not particularly modified although partly yellow-brown, second segment ventrally with weak, black spot. Preapical carina of T6 multidentate (Fig. 51). T7 produced to a median tooth that is compressed laterally (Fig. 51). Females: Mandible long and slender, almost as in Chalicodoma, weakly 4-toothed, impunctate area along apical margin narrow (Fig. 20). Scutellum not elevated medially. Clypeus flat in profile, medially produced to a small tooth (Fig. 20), laterally weakly denticulate. Ocelloccipital distance longer than interocellar distance.

5. flavipes group (Archimegachile Alfken, 1933): Males: As in ericetorum group, but front tarsal segments 1–3 or 1–4 maculated ventrally, maculations sometimes reduced to thin lines. Females: As in ericetorum group, with following differences: clypeus apically without median tooth, or tooth minute; ocelloccipital distance either subequal to or shorter than interocellar distance.

6. lanata group: Males: Front coxal spine short; front tarsi unmodified. Mandible weakly 4-toothed, without inferior projection although the inferior margin is swollen medially. Preapical carina of T6 bilobed, not denticulate. T7 small, little visible. Females: Mandible as in ericetorum group, 4-toothed and comparatively elongate, impunctate area along apical margin narrow. Clypeus short, apical margin straight, without medial tooth. Ocelloccipital distance slightly shorter than interocellar distance.

There is only one Western Palearctic species each in the foersteri, rhodoleucura, ericetorum and lanata groups (Megachile foersteri, M. riyadhensis, M. ericetorum and M. lanata; see above for comments on the taxonomic status of M. riyadhensis); three species in the incana group, of which two are undescribed (Dorchin and Praz, in prep.). At least M. cinnamomea Alken, 1926, M. farinosa Smith, 1853, M. flavipes Spinola, 1838, M. sanguinipes Morawitz, 1875 and M. tecta Radoszkowski, 1888 are valid species of the flavipes group; the status of M. rubripes is unclear: both M. flavipes and M. rubripes are parapatric and sculpturally very similar and have been considered as two subspecies of the same species by some authors (e. g., Özbek and Zanden 1994). At least M. cyannipennis, M. nigripes Spinola, 1838, M. saussurei, M. schnabli Radoszkowski, 1893 and M. seraxensis Radoszkowski, 1893 are valid species within the saussurei group, which includes at least three undescribed species (Dorchin and Praz, in prep).

The nesting biology of Megachile ericetorum has been described in detail (references in Westrich 1989). Ferton (1895) provides a description of the nests, which are located in existing cavities such as existing holes in soil or clay banks, or in reed stems. The cells are arranged linearly and consist of mud; Ferton observed females gathering moist soil close to streams, a behavior not observed in bees of the subgenus Chalicodoma, which add secretions to dry mud to build their nests (see under that subgenus). The outside of the cell is coarse and irregular, while the inner wall is lined with a thin (up to 0.5 mm) layer of resin. The cell is closed with a plug of soil; the cell cap is not lined with resin on the inside of the cell, although resin is found on the outer surface, which constitutes the base of the nest cell. The nest closure consists of a layer of mud covered by a thin layer of resin, with sometimes an additional layer (2 mm) of mud. Rozen and Kamel (2007) described the biology of M. nigripes, a member of the saussurei group. The species nests in tunnels dug by the females in “adobe”, i. e., hard, dry, mud walls. The cells, oriented almost vertically in the oblique tunnels, are elongate and not arranged linearly. The cell walls probably consist of the excavated mud wall and the inside of the cell is uneven but appears to have been lined with dark coating that is “water retardant but not truly waterproof”. The authors conclude that it may consist of “very fine soil particles bound together by partly dried nectar”. The nests of species of the flavipes group have been briefly described several times (Gutbier 1914, Alfken 1934, Mavromoustakis 1939, Krombein 1969). The most precise account describes the nests of M. rubripes in Turkmenistan (Ponomareva 1958). Nests of this species were found in reed stems. Two cells were arranged linearly and made of mud “moistened with water, apparently without using secretions (...); the dry cells instantly disintegrate on submergence in water”. As in M. ericetorum, the outside of the cell is irregular, but the inside is smooth; in M. rubripes resin or secretions were apparently not used to line the inner wall.

Megachile ericetorum is likely oligolectic on Fabaceae (Westrich 1989, Müller et al. 1997), although according to Westrich (1989)Lamiaceae pollen is very occasionally also used (see also Gogala 2014). Limited evidence based on floral records suggests that M. foersteri has a preference for Carduoideae (Asteraceae) (C. Praz, unpublished data). According to Rozen and Kamel (2007), the main host of M. nigripes in Egypt is Trifolium alexandrinum L. I observed M. saussurei on Medicago in Uzbekistan and in Iran (see also Popov 1946), M. flavipes and the closely related M. rubripes only on Fabaceae. As indicated by Alqarni et al. (2012), many unrelated species of Pseudomegachile have conspicuously modified facial hairs; these modified hairs are always found on the clypeus but may also be found on the supraclypeal area and the frons. In the species associated with Fabaceae listed above, the hairs on the face are plumose at least in fresh specimens, while in species with modified hairs, the hairs are short, simple and often bent downwards apically, or wavy (see Müller 1996). Alqarni et al. (2012) listed several species of Pseudomegachile with such modification; among the Palearctic species, they cite the following species. M. riyadhensis, which they collected exclusively on Blepharis (Acanthaceae); I examined the pollen present in the scopa of the holotype of M. rhodoleucura (with identical facial vestiture as M. riyadhensis; see comments above) and the pollen also belongs to Acanthaceae. M. farinosa and M. cinnamomea (the condition is not clear in the latter species), both in the flavipes group also have modified facial hairs; according to my field observations, M. farinosa has a preference for the flowers of Vitex (Lamiaceae), while M. cinnamomea is likely polylectic. In the saussurei group, Alqarni et al. (2012) cite M. transgrediens Rebmann, 1970 as having modified facial hairs; this mention likely refers to one of two undescribed species from Turkey and Iran with modified hairs (M. transgrediens is probably a junior synonym of M. saussurei; C. Praz, unpublished); these two undescribed species are sculpturally close to M. saussurei, in fact differing from the latter in the female sex mostly by the modified facial vestiture. Host plants of these two species are unknown. M. seraxensis also has modified facial hairs; I collected one female of this species visiting Blepharis. Lastly, I collected in Israel several specimens of an undescribed species of the incana group collecting the pollen from Blepharis; this species has modified hairs on the clypeus, supraclypeal area and frons, exactly as in M. riyadhensis. Another closely related species from the incana group from Oman and the UAE has branched hairs on the clypeus; it was observed foraging on Crotalaria (Fabaceae; Sarah Gess, pers. comm., September 2016).

Females: Most females of Xanthosarus have a typical, broad mandible (Fig. 35) with tooth 1 larger than other teeth, without apical brush of hairs in the grooves around the base of tooth 1. In such females there is a well-visible, continuous cutting edge in the third interspace and a partial cutting edge in the second interspace (Fig. 35), the third interspace is deeper than the other interspaces, and the two upper teeth close together or poorly separated, so that the mandible is not clearly 5-toothed. Such females will be easy to identify using the present key. The mandible of Megachile nigriventris Schenck, 1868 (Fig. 36) [and to some extend of M. willughbiella (Kirby, 1802): Fig. 37] is different, less robust, with the cutting edges little visible in front view; the upper tooth is broadly truncate and the condition is intermediate between the 4-toothed and the 5-toothed conditions. In all Palearctic species of Xanthosarus, the clypeus is finely and densely punctured, regularly convex (Figs 35–37) unlike the condition found in most females of Megachile s. str. The scopa is mostly orange-red basally, sometimes nearly entirely black (M. nigriventris) or yellow-white on sternites 2 and 3 [e.g. M. maritima (Kirby, 1802)], rarely entirely yellowish-white. The sterna lack the apical fringe of hairs beneath the scopa, unlike in most Eutricharaea or in M. (Eurymella) patellimana. Males: In males of Xanthosarus the front coxal spine is always well developed; often there is a field of modified, short, orange bristles on the surface anteriorly to the spine, but such field may be lacking (e.g. in M. analis Nylander, 1852). The front tarsi are always yellow or white, from relatively narrow [e.g. M. analis, M. circumcincta (Kirby, 1802)] to conspicuously enlarged [e.g. M. lagopoda (Linnaeus, 1761), M. maritima]; the first tarsal segment is mostly strongly concave interiorly, except in M. analis. The mandible is either 3-toothed (M. lapogoda, M. maritima) or more commonly 4-toothed; the inferior margin always has a pointed process directed posteriorly; in some large species, this process is particularly large and glabrous, except for some orange hairs apically. The disc of T6 is mostly not covered by white vestiture; the preapical carina of T6 is bilobed or weakly denticulate. The apical margin of T6 often lacks a lateral tooth (a small tooth is sometimes present, e.g. in M. maritima). The apex of T7 is usually produced to a small tooth medially, although the tooth is small in some species; in other species T7 is weakly trifid. The gonostylus is variable, either simple or strongly bifid, but this character does not clearly segregate groups.

There are at least seven species in the western Palearctic: Megachile analis, M. circumcincta, M. diabolica Friese, 1898, M. lagopoda, M. maritima, M. nigriventris and M. willughbiella. M. fulvimana Eversmann, 1852 has been mentioned from Southeastern Europe (Banaszak and Romasenko 2001); this species is known to me only from Central Asia. M. mguildensis Benoist, 1940, from Algeria and Morocco, apparently closely related to M. nigriventris, may either represent a distinct species or merely a color morph of M. nigriventris. Özbek and Zanden (1994) further cite M. metatarsalis Morawitz, 1894 from Turkey; this species is unknown to me. The status of M. fulvescens Smith, 1853 from Sicily is unclear; it is unlikely to represent a species distinct from those listed above; its description (“the pollen brush is of golden hue towards the base, becoming bright fulvous at the apex”) suggests M. maritima. Similarly, the status of M. maacki Radoszkowski, 1874 remains in doubt; Scheuchl (2006) provides a description of the male and the female and differentiates this species from M. nigriventris. However, no lectotype has been designated. Of the possible syntypes that I could examine (ISZP), one male agrees with the original description and does not appear specifically distinct from M. nigriventris. Possibly, M. maacki represents an Eastern Palearctic, geographic form of M. nigriventris. Xanthosarus appears particularly diverse in the mountains of Central Asia and in Mongolia, where several additional species occur.

Most species of the sugbenus Xanthosarus place their brood cells made of cut leaves in underground burrows or more rarely under stones; the females appear to dig the burrows themselves (Westrich 1989, and references therein; Hartmann and Arens 1998). Megachile nigriventris digs burrows only in decaying wood (Westrich 1989, Dubitzky 2000, Reichholf 2002), while M. willughbiella either digs burrows in decaying wood or uses existing cavities such as vacant Anthophora cells (Westrich 1989, and references therein; Müller et al. 1997). Megachile analis is unusual in that it uses thin bark fragments and not leaves for the construction of the cells (Westrich 1989, and references therein); leaf discs are also used by this species.

Megachile nigriventris is likely oligolectic on Fabaceae (Westrich 1989, Müller et al. 1997) and M. diabolica possibly on Campanulaceae (Hartmann and Arens 1998); M. analis and M. willughbiella are polylectic but show a preference for Fabaceae and Campanulaceae, as well as for the genus Epilobium by M. willughbiella (Westrich 1989, Müller et al. 1997). Megachilelagopoda and M. maritima are likely polylectic with a preference for Fabaceae and Asteraceae.