Supplementary material 3 from: Onuferko TM (2018) A revision of the cleptoparasitic bee genus Epeolus Latreille for Nearctic species, north of Mexico (Hymenoptera, Apidae). ZooKeys 755: 1-185. https://doi.org/10.3897/zookeys.755.23939

Abstract Herein, the cleptoparasitic (cuckoo) bee genus Epeolus (Hymenoptera: Apidae) is revised for species occurring in North America, north of Mexico, and an updated checklist of all species known to occur in Canada and the United States of America is provided with comprehensive descriptions, diagnoses, and a single dichotomous key (using the same couplets for both sexes) to aid in their identification. To increase their recognition among North American naturalists, English common names are also proposed for all North American Epeolus. A total of 43 species is confirmed as present in the region, 15 of which are newly recognized. The following new species are proposed based on unique morphological (and in most cases also molecular) attributes: E. andriyi sp. n., E. attenboroughi sp. n., E. axillaris sp. n., E. basili sp. n., E. brumleyi sp. n., E. chamaesarachae sp. n., E. deyrupi sp. n., E. diadematus sp. n., E. ferrarii sp. n., E. gibbsi sp. n., E. inornatus sp. n., E. nebulosus sp. n., E. packeri sp. n., E. splendidus sp. n., and E. tessieris sp. n. Of the 15, six (E. axillaris, E. brumleyi, E. chamaesarachae, E. diadematus, E. splendidus, and E. tessieris) were identified as new species under different names (nomina nuda) in an M.Sc. thesis by Richard L. Brumley in 1965, but until now they have not been formally described. Detailed morphological comparisons with some evidence from DNA barcoding support the following synonymies, one of which C was first proposed by Brumley (1965): a) E. melectimimus Cockerell and Sandhouse, syn. n., under E. asperatus Cockerell; b) E. crucis Cockerell, syn. n., under E. compactus Cresson; c) E. mesillae palmarum Linsley, syn. n., under E. mesillae (Cockerell); and d) E. weemsi Mitchell, syn. n., and e) E. vernalis Mitchell, syn. n., under E. ilicis Mitchell. Only one member of the almost entirely Neotropical “Trophocleptria group” (Epeolus bifasciatus Cresson) is confirmed as occurring north of Mexico, and is widespread East of the Rocky Mountains. Known floral associations are indicated for each species, as are suspected or known host species of Colletes Latreille. Evidence is presented that suggests further investigation into the possible synonymy of Colletes wickhami Timberlake under C. scopiventer Swenk is warranted.


Introduction
Epeolus Latreille (Hymenoptera: Apidae, subfamily Nomadinae) is one of the most widespread genera of cleptoparasitic bees (commonly referred to as cuckoo bees), occurring on all continents except Antarctica and Australia. The genus is also absent from Madagascar, Oceania, and parts of Southeast Asia, regions in which their host genus Colletes Latreille (Hymenoptera: Colletidae: Colletinae) is not present (Michener 2007). Other genera in the tribe Epeolini are largely restricted to the Americas, mostly to the Neotropical region. The similarly diverse bee genus Triepeolus Robertson has only two representatives in the Palearctic region, whereas Epeolus is represented across Africa, Asia, and Europe by about 48 species (Ascher and Pickering 2017). However, the genus is most diverse in North America, with 32 valid species confirmed as occurring north of Mexico before the date of this publication.
For North American species, the taxonomy of Epeolus has been in need of revision for some time. While  treatment of the Eastern United States fauna was fairly comprehensive, the Western species have been in much need of attention. In his M.Sc. thesis, Richard L. Brumley (1965) recognized several new species from the Western United States, but his names were never published and are therefore not considered valid. Recently, Onuferko (2017) identified 14 redundant names (most are of Western "species"), which were synonymized under the names of four valid species, but this treatment was limited to the Canadian fauna. The purpose of the present study is to resolve the taxonomy of Epeolus occurring in Canada and the USA by naming and describing new species and identifying which accepted names are valid and which are not, thereby standardizing name use, as well as to provide a user-friendly dichotomous identification key. To help amateur and professional entomologists become more familiar with these bees, English common names are proposed for all North American species of Epeolus. An additional objective is to present ecological information in terms of floral and Colletes hosts and phenology wherever possible, as well as comprehensive occurrence records to aid those interested in locating and identifying representatives of the species treated herein for further research.

Materials and methods
To revise Epeolus an integrative biosystematics approach was followed, using morphological and molecular evidence to distinguish intraspecific from interspecific variation (as in Gibbs 2009, 2010, 2011, Pauly et al. 2014, Rocha-Filho and Packer 2015, Ferrari 2017, Onuferko 2017. Morphological evidence was prioritized over molecular evidence when the two were not in agreement, as in Gibbs (2009). Sequence data from a 658 bp segment of the mitochondrial cytochrome c oxidase subunit I (COI) gene (DNA barcode, Hebert et al. 2003a, b) were obtained from specimens of nearly all (42 out of 43) species, and 37 have sequences that are barcode compliant (i.e., have met the criteria to be assigned automated barcode index numbers (BINs) given to unique barcode clusters, Hebert 2007, 2013). One or two legs were removed from each specimen to be "barcoded", and sent to the Canadian Centre for DNA Barcoding in Guelph, Ontario, Canada for DNA extraction and gene amplification and sequencing. A neighbor-joining (NJ) tree, based on Kimura's twoparameter distance model (Kimura 1980), was used to compare short, non-compliant and barcode-compliant sequences for the purpose of validating species designations of sequenced specimens and checking for contamination errors. Partial and BIN-compliant sequences are published in the "Epeolus of North America project" on the Barcode of Life Data Systems website (http://www.barcodinglife.org/) and have been deposited in the GenBank database (see Suppl. material 1 for accession numbers).
Terminology used herein is consistent with that used in the recent treatment of Canadian Epeolus (Onuferko 2017), which generally followed Michener (2007), except the terms frontal area and vertexal area are used instead of frons and vertex, respectively. Acronyms used herein (in bold) are as follows. Puncture density is described in terms of interspaces (i) relative to the diameters (d) of punctures. Median ocellar diameter (MOD) is a comparative unit of measurement for smaller structures. F followed by a number represents one of 10 (for female) or 11 (for male) flagellomeres of the antenna. T followed by a number represents one of six (for female) or seven (for male) exposed metasomal terga. S followed by a number represents one of six (for female) or eight (for male) metasomal sterna. Several terms used in Onuferko (2017), some of which were taken from Rightmyer (2008), are defined here again for clarity, and are indicated in bold. Length refers to measurements made along the longitudinal axis of the bee, except in reference to the longitudinal extent of the transverse metasomal fasciae, for which the term breadth is used, and width refers to measurements made along the lateral axis. The length and width of an anatomical feature refer to its longest and widest margins, respectively, and were recorded at the highest magnification that allowed measurement in ocular micrometer units. The scape was measured without the radicle. In Epeolus, the frontal line extends into the supraclypeal area as a pronounced carina on a convex surface, referred to herein as the frontal keel. Paramedian bands are the paired lines of off-white or yellow tomentum located anteriorly on the mesoscutum of most Epeolus species (Fig. 1). The term bigibbous is an adjective used in reference to the biconvexities present on the mesoscutellum of Epeolus species. The basal and apical fasciae of T1 are often connected by a longitudinal band of pale tomentum of varying width. Discal patch refers to the discal region of T1 that is typically covered in dark tomentum and is bordered by bands of pale tomentum. This area is not always clearly delineated because the surrounding bands of pale tomentum may be reduced or missing entirely.
The species of Epeolus are, with the usual exceptions (differences in the number of antennal flagellomeres, number of exposed metasomal terga, length of the S4 and S5 subapical hairs [usually longer in males], and terminalia) and a few atypical ones, sexually monomorphic. For this reason, separate keys for females and males are not presented, and the few sex-specific features used to distinguish species are indicated as such in the couplets. The key to Nearctic Epeolus is heavily based on the structure of the axilla and the bands of pale tomentum forming the basal and apical fasciae on the metasomal terga. To limit the number of steps required to identify all species, efforts were made to make the key as close to fan shaped (evenly bifurcated) as possible, following the recommendations of Walter and Winterton (2007; see also Packer et al. 2016). When possible, couplets were based on more than a single feature (ideally one per tagma) should one be obscured or lost in the specimen being identified. However, avoiding monothetic couplets was not always possible. In such cases couplets were usually based on mesosomal features that should be visible even in damaged pinned specimens. In couplets that list multiple features, the most important (i.e., reliable) one for achieving a diagnosis is given first whereas features that do not always result in a positive identification (e.g., integument black vs integument black or ferruginous will resolve species with ferruginous but not black integument) are included but given at the end and always preceded by at least one feature that is fully contrasted between both halves of the couplet. The features referenced in the key were imaged. Quite often a single image or image plate was used to illustrate more than one feature, so a number of figures were cited two or more times within the key and elsewhere in the present monograph. As a result, it was not possible to put most illustrations near the couplets without duplicating them, and for practical reasons multiple versions of the same figures are not included herein. Many couplets rely on precise comparative measurements, and the key is meant to be used with the aid of an eyepiece graticule. None of the couplets require specimens to be dissected. Although the male S7, S8, and genital capsules of nearly all species were examined (except those represented by very few male specimens), the variation among them is minimal (illustrated in part in plates 2 and 3 in Onuferko 2017), and the terminalia have not proven useful in separating similar-looking species. Consequently, they have not been illustrated or imaged. The illustrations presented to aid in the identification of Epeolus species are my own. Images were taken with a digital camera (Canon EOS 40D SLR) using the Visionary Digital macro-imaging BK PLUS Lab System, focus stacked in Helicon Focus, and edited in Adobe Photoshop and PaintShop Pro.
Species descriptions follow the format of Onuferko (2017). A full description of the primary type specimen of each species is provided, except for the species occurring in Canada that were recently redescribed in Onuferko (2017). The physical name-bearing type specimens of all described North American Epeolus were seen and thoroughly examined, including those whose names are no longer considered valid, except in the case of E. mercatus Fabricius, for which the original type material cannot be traced and description is so insufficiently detailed that it is unclear if the species is an Epeolus or Triepeolus (Rightmyer 2008). Since most Epeolus species to date were described from female specimens, new species described herein are generally represented by a female holotype, male allotype, and paratypes. Given that Epeolus is a genus of largely sexually monomorphic species, descriptions of the sex opposite that of the name-bearing type list only key differences to avoid unnecessary duplication of text. In many, but not all, cases it is the female that is fully described. I have opted to propose new names for the species Brumley (1965) discovered rather than validate the ones he used. This will ensure that it is clear who made designations of type specimens (i.e., specimens used as types by Brumley (1965) and me have both our type labels, those unavailable to me but designated as types by Brumley (1965) have only his labels, and those seen exclusively by me and given type status have only my labels). This will also eliminate any possible confusion that could arise if Brumley's (1965) names are published and registered in ZooBank long after their first appearance in his thesis.
The proposed common name for each species reflects its scientific name, which in most cases was easy to translate into English. Since there are many genera of cuckoo bees, epeolus is used herein as the common name for the genus instead of cuckoo bee or more specific but cumbersome names like Colletes cuckoo bee or polyester bee cuckoo bee.
Among the material examined were representatives of Epeolus from all Canadian provinces and territories except Newfoundland and Labrador and Nunavut, and all but six (Connecticut, Delaware, Kentucky, Rhode Island, Tennessee, and West Virginia) of the 49 states in the continental U.S. where the genus is expected to occur. Also examined were Epeolus records from 17 states in Mexico, and their data are included for species confirmed as occurring north of the Mexico-United States border. All examined records are presented in Suppl. material 1. Specimens were made available for study by curators and collections managers (in parentheses) from the following institutions: likely to be confused with vespid wasps (hence the root 'odynerus') rather than Epeolus. Comprehensive overviews of the distinguishing features of Epeolus in reference to all other Epeolini are provided in Rightmyer (2004) and Michener (2007).

Diagnosis for Epeolus in North America (Canada and the United States).
Diagnostic for female Epeolus is a very distinct S6, which is usually retracted except sometimes for a pair of convergent spatulate lateral apical processes bearing setae modified into minute, pointed denticles (Onuferko 2017, Fig. 2A, B). Basally, the processes are separated by a large lobe-like disc, which in Triepeolus is reduced to a narrow transverse bar. In both Triepeolus (Onuferko 2017, Fig. 2C, D) and Odyneropsis, the lateral apical processes are subparallel and bear coarse, spine-like setae. Additionally, females may be separated on the basis of the pseudopygidial area (the apicomedial region of T5 that changes slope from the rest of the tergum), which in Epeolus is covered in a silvery band of short apically rounded setae. In Triepeolus, the pseudopygidial area is usually longer than in Epeolus and in most species the setae reflect a golden color. The T5 in female Odyneropsis is unique in that it is broadly notched posteriorly and has a distinct middorsal depressed area in the shape of a pointed oval outlined by ridges (Rightmyer 2004, fig. 180A).
Male Epeolus are more difficult to diagnose. As in females, the body lacks integumental white or yellow areas but the mesosoma and usually other tagmata have short appressed plumose white and/or yellow setae; the maxillary palpus is two or three segmented; the inner margins of the compound eyes are distinctly convergent below; the axilla is produced to a rounded lobe or angle or spine (i.e., not continuing the contour of the mesoscutellum); the distitarsi of all legs have arolia; the fore wing usually has three submarginal cells (if with two, then the second is at least nearly as long as the first), and the marginal cell is apically removed from the wing margin and much longer than the stigma; and a pygidial plate is present. In male Epeolus, the pygidial plate in most species is broadly rounded posteriorly (Fig. 2B); in Odyneropsis and Triepeolus it is usually more elongate and with a median constriction (Fig. 2F). It should be noted that males of some species of Epeolus in North America (notably E. australis Mitchell, E. flavofasciatus Smith, and some males in the "americanus group") have a very narrow and distinctly Triepeoluslike pygidial plate ( Fig. 2A, C, D), as opposed to the more broadly rounded/subtruncate pygidial plate typically associated with male Epeolus (Fig. 2B). The presence of a preapical tooth of the mandible (Fig. 3B, C, D, F) (often hidden from view because the mandibles are usually closed) confirms these and other species as Epeolus; all Triepeolus and only some Epeolus (in North America E. ainsliei, E. erigeronis, E. ilicis, E. inornatus, and E. zonatus) lack one (Fig. 3A, E) (Rightmyer 2004).

List of species with their proposed common names
Epeolus ainsliei Crawford, 1932 -Ainslie's epeolus Epeolus americanus (Cresson, 1878) -American epeolus Epeolus andriyi Onuferko, sp. n. -Andrew's epeolus Epeolus asperatus Cockerell, 1909 -rough epeolus Figure 2. Pygidial plate (in dorsal view) of male A E. australis (longer than wide and apically narrowed) B E. brumleyi paratype (nearly as long as wide and apically rounded) C E. flavofasciatus (longer than wide, with the lateral margins parallel) D E. asperatus (longer than wide and apically narrowed) E E. barberiellus (somewhat longer than wide and apically narrowed), and F T. concavus (longer than wide, with the lateral margins somewhat concave). Scale bars 1 mm. attenboroughi holotype with an inconspicuous, obtuse preapical tooth C E. carolinus with an inconspicuous, obtuse preapical tooth D E. gibbsi paratype with an obtuse angle appearing like a tooth E E. vernalis holotype (herein synonymized under E. ilicis) without a preapical angulation or tooth, and F E. compactus with a distinct preapical tooth with sides forming a right triangle. Scale bars 0.5 mm.
Distribution. Great Plains to southwestern Ontario (Fig. 5). Ecology. HOST RECORDS: Epeolus ainsliei has been collected with possible host species Colletes susannae Swenk in Birds Hill Provincial Park (Gibbs et al. 2017) and Spruce Woods Provincial Park (J. Gibbs, personal communication, 2017), Manitoba, Canada and Spring Green Preserve in Sauk County, Wisconsin, USA (Wolf and Ascher 2009). In all cases at least one other species of Colletes was observed at the same locality and time as C. susannae and E. ainsliei, but observations of other Colletes were limited to one or two localities.
Discussion. Detailed morphological and taxonomic remarks about this species are given in Onuferko (2017).
Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. americanus apart from all other North American Epeolus except E. asperatus and E. barberiellus: in females, F2 is not more than 1.1 × as long as wide; the mesoscutum has distinct paramedian bands; the axilla is small to intermediate in size, not extending beyond the midlength of the mesoscutellum and the free portion is less than 1/4 as long as the entire medial length of the axilla, and like the mesoscutellum black; the mesopleuron is closely (i≤1d) and evenly punctate; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least as wide as the breadth of the apical fascia; and the T1 and T2 apical fasciae are interrupted or at least greatly narrowed medially. Whereas in E. barberiellus the pronotal lobe and legs, at least from the tibiae to tarsi (sometimes the trochanters and femora as well), are reddish orange, in E. americanus the pronotal lobe and legs are brown or black. Epeolus americanus is also very similar to E. asperatus, but in E. asperatus the mesopleuron has much denser punctures ventrolaterally (most i<1d) than that of E. americanus and the T3 and T4 fasciae are never complete but broken or at least greatly narrowed laterally, as well as medially into separated or narrowly connected oval patches.
Redescription. This species was recently redescribed (Onuferko 2017).  Distribution. Widely distributed across Canada and the United States, including Alaska; not known to occur in parts of northeastern North America, the southeastern United States, or the high arctic (Fig. 7).

Epeolus andriyi sp. n.
http://zoobank.org/97D5B971-2314-4E0A-BCF1-0E938C0EDA25 Figs 8, 9 Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. andriyi apart from all other North American Epeolus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin, dilated laterally but relatively straight along the medial margin, and like the mesoscutellum ferruginous; the axilla's free portion is clearly less than 2/5 as long as its entire medial length; the mesopleuron is closely (i≤1d) and evenly punctate; the metasomal terga are black; T1 has a distinct basal fascia, which may be narrowly interrupted medially; the mesoscutum and metasomal terga have bands of bright or pale yellow short appressed setae; at least the T1-T3 apical fasciae are distinctly interrupted medially; and the pseudopygidial area of the female is lunate with the apex <2 × the medial length. Epeolus andriyi is most similar to E. howardi, but in E. howardi the axillae extend further posteriorly, as far back as or beyond the posterior margin of the mesoscutellum, and both the axillae and mesoscutellum are entirely red whereas in E. andriyi the mesoscutellum is dark brown or black along the anterior margin. Epeolus andriyi is also similar to E. scutellaris, but in E. scutellaris the T1-T3 apical fasciae are complete or only very narrowly interrupted medially, and the pseudopygidial area of the female is lunate with the apex >2 × the medial length.
Description. FEMALE: Length 8.2 mm; head length 1.9 mm; head width 2.6 mm; fore wing length 5.5 mm (margins of both worn in holotype).
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, axilla, mesoscutum, mesoscutellum, mesopleuron, and legs. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex. Antenna brown except scape, pedicel, and F1 extensively orange. F2 with orange spot basally. Pronotal lobe and tegula pale ferruginous to amber. Mesoscutum with reddish-brown spot anterolaterally between pronotal lobe and tegula. Wing membrane dusky subhyaline, slightly darker at apex. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Clypeus, upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma  and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half hairy, except beneath base of fore wing (hypoepimeral area); ventrolateral half nearly bare. Metanotum with tomentum sparser medially, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally by few sparsely scattered pale hairs. T1-T3 with apical fasciae interrupted medially and narrowed before becoming somewhat broader laterally; T2 with fascia without anterolateral extensions of tomentum, although few sparsely scattered pale hairs present. T4 with fascia narrowed medially. T5 with two patches of pale tomentum (both quite faint in holotype because much of pubescence discolored or rubbed off) lateral to and contacting pseudopygidial area. T5 with pseudopygidial area lunate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Preapical tooth inconspicuous, blunt and obtuse. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.5). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) half as long as mesoscutellar width (W) (L/W ratio = 0.5) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, not noticeably longer than wide (L/W ratio = 1.1); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures more or less evenly spaced throughout, with the interspaces shining.
Etymology. This species is named in honor of my father, Rev. Andriy Onuferko, in gratitude for encouraging my interests in the natural world and for his assistance in collecting Epeolus in the field.
Distribution. Presently known from a single location along the Patuxent River in Maryland, USA (Fig. 9).
Ecology. HOST RECORDS: The host species of E. andriyi is/are presently unknown. FLORAL RECORDS: Unknown.

Discussion.
Epeolus andriyi and E. howardi are very similar to one another, and both species have been collected in Maryland, USA in late August. Although E. andriyi is known from only two specimens, in both the axillae are shorter than in any examined specimen of E. howardi. The status of E. andriyi as a separate species is further supported by a separate BIN, but unusually its nearest neighbor is E. lectoides, from which E. andriyi exhibits a large barcode sequence divergence (7.1%).

Epeolus asperatus
Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. asperatus apart from all other North American Epeolus except E. americanus and E. barberiellus: in females, F2 is not more than 1.1 × as long as wide; the mesoscutum has distinct paramedian bands; the axilla is small to intermediate in size, not extending beyond the midlength of the mesoscutellum and the free portion is less than 1/4 as long as the entire medial length of the axilla, and like the mesoscutellum black; the mesopleuron is closely (most i<1d) and evenly punctate; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least as wide as the breadth of the apical fascia; and the T1 and T2 apical fasciae are interrupted or at least greatly narrowed medially. Whereas in E. barberiellus the legs, at least from the tibiae to tarsi (sometimes the trochanters and femora as well), are reddish orange and the metasomal terga are fasciate, in E. asperatus the legs are brown or black and the T3 and T4 fasciae are broken or at least greatly narrowed laterally, as well as medially into separated or narrowly connected oval patches. Epeolus asperatus is most similar to E. americanus, but in E. americanus the mesopleuron has sparser punctures ventrolaterally (i≤1d) than that of E. asperatus, with the interspaces shining, and the T3 and T4 fasciae are complete or broken medially and/or laterally, but rarely into separated oval patches. Redescription. FEMALE: Length 7.8 mm; head length 2.0 mm; head width 2.8 mm; fore wing length 5.4 mm.
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, and legs. Mandible with apex darker than rest of mandible; preapical tooth lighter than mandibular apex (difficult to see in the E. asperatus holotype; described from non-type specimens). Antenna brown except F1 and F2 orange in part. Flagellum slightly lighter than conspicuously dark brown scape and pedicel, primarily due to extensive pilosity on flagellum. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs with brown or black more extensive than reddish orange.
Pubescence. Face with tomentum densest around antennal socket. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half hairy, ventrolateral half nearly bare. Metanotum with tomentum rubbed off medially in the E. asperatus holotype, but somewhat sparser medially and uniformly off white in non-type specimens. T1 with median quadrangular black discal patch enclosed by pale tomentum, except for medial separation at apex, and narrow, such that longitudinal band nearly half as wide as width of discal patch in dorsal view. T2-T4 with fasciae interrupted medially and with anterolateral extensions of sparser tomentum. T3 and T4 with fasciae also interrupted laterally, appearing as pair of oval patches between medial and lateral interruptions. T5 with two patches of pale tomentum lateral to and separate from pseudopygidial area (difficult to see in the E. asperatus holotype because T5 mostly retracted; described from non-type specimens). T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Preapical tooth with blunt tip. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5-2 MOD at its terminal (difficult to see in the E. asperatus holotype; described from non-type specimens). Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and tip not extending beyond midlength of mesoscutellum; axilla with tip visible, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with second submarginal crossvein incomplete in the E. asperatus holotype; with submarginal cells two or three and closed or second submarginal crossvein incomplete in non-type specimens. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, nearly as long as wide (L/W ratio = 0.8); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate V-shaped but apically rounded, with large deep, well-separated punctures, with the interspaces shining.
Distribution. Central and southern California (Fig. 11). Ecology. HOST RECORDS: Nine representatives of this species were collected at the Robert J. Bernard Biological Field Station in Claremont, California, USA in the spring of 2016 (see Material studied), and the only Colletes collected or observed was a single female of a predominantly black species with pale pubescence limited to the mesosoma. The collected female of the possible host species was barcoded, and using Stephen's (1954) key identified as C. californicus Provancher. However, its sequence clusters with sequences of specimens collected in New Mexico (also in the spring of 2016) and identified as C. sphaeralceae Timberlake (with entirely/predominantly pale pubescence) through the use of Stephen's (1954) key, dissection of the male terminalia, and collection from red Sphaeralcea A. St.-Hil. (Malvaceae) flowers, and all were assigned the same BIN (BOLD:ABZ4529). Another predominantly black female specimen from the San Diego National Wildlife Refuge Otay-Sweetwater Unit in California Figure 11. Approximate geographic range of E. asperatus (orange) based on occurrence records known to the author (yellow circles).
was barcoded (its image and 601 bp sequence are available on the Barcode of Life Data Systems website [http://www.barcodinglife.org/]), and was assigned the same BIN as the female from Claremont and specimens from New Mexico.
FLORAL RECORDS: Labels of examined voucher specimens indicate floral associations with Lasthenia Cass. (Compositae) and Plagiobothrys.
Discussion. Brumley (1965) synonymized E. asperatus and E. melectimimus under E. americanus, but current evidence suggests that the holotypes of E. asperatus and E. melectimimus belong to a cryptic species within the "americanus group", distinct from E. americanus and E. barberiellus. In addition to the subtle diagnostic morphological features that separate E. asperatus from E. americanus and E. barberiellus, the status of E. asperatus as a separate species is supported by a separate BIN and large barcode sequence divergence (4.4%) from its nearest neighbor, E. barberiellus.
Epeolus melectimimus, with three submarginal cells, was described by Cockerell and Sandhouse (1924), who claimed that the species resembles a small Pseudomelecta Radoszkowski (a subgenus of Melecta Latreille in Michener 2007), from which it can be readily distinguished based on differences in the marginal cell. In the E. asperatus holotype, the second submarginal crossvein on each side is incomplete and inconspicuous. A series of E. asperatus was collected from the Robert J. Bernard Biological Field Station in Claremont, California, USA, which is in the same county as the type locality (Los Angeles). In some specimens, the fore wing has three submarginal cells whereas in others, the second submarginal crossvein is incomplete or lacking entirely. In some specimens, one fore wing has three submarginal cells and the other has an incomplete second submarginal crossvein. The male holotype of E. melectimimus was examined, and excluding sex-specific features the specimen with few exceptions agrees with the present redescription based on the female holotype of E. asperatus. Along with the abovementioned differences in wing venation, the pronotal lobe and tegula are darker in the holotype of E. melectimimus than in that of E. asperatus, but these differences fall within the range of observed intraspecific morphological variation among sequenced specimens. Although both E. americanus 13.iv.2016, T.M. Onuferko (2♀, 1♂, PCYU), 14.iv.2016, T.M. Onuferko (1♀, PCYU), 26.iv.2016; W L Jepson Prairie Preserve (TNC) (13 mi S Dixon, Solano County), 20.v.1983, J.D. Barbour (1♂, UCBME).

Epeolus attenboroughi sp. n.
http://zoobank.org/FD2EAACB-3D7A-477C-9B7D-A7EABE7DE10B Figs 3B, 12, 13, 94B, 95B, 96A Diagnosis. The following morphological features in combination can be used to tell E. attenboroughi apart from all other North American Epeolus except E. rufulus: the mandible has a blunt, obtuse preapical tooth; the preoccipital ridge does not join the hypostomal carina; the mesoscutum is largely obscured by pale tomentum; the axilla is elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin, and the free portion is distinctly hooked; the mesopleuron is closely (most i<1d) and evenly punctate; and T1-T4 have complete apical fasciae. Whereas in E. rufulus the discal patch is so wide that the longitudinal band is barely visible in dorsal view and in females F2 is noticeably longer than wide, in E. attenboroughi T1 has a comparatively narrow discal patch (the longitudinal band is more than half as wide as the breadth of the apical fascia in dorsal view) and in females F2 is less than 1.2 × as long as wide. Epeolus attenboroughi is also similar to E. ainsliei in that in both species the axilla is dilated laterally and the free portion is distinctly hooked, and the T1-T4 apical fasciae are complete; however, in E. ainsliei the mandible is simple, the preoccipital ridge joins the hypostomal carina, and the mesoscutum has distinct paramedian bands.
Pubescence. Face with tomentum densest around antennal socket, slightly sparser on clypeus, upper paraocular and frontal areas, and vertexal area. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum, mesoscutellum, and axilla largely obscured by pale tomentum. Mesopleuron densely hairy, except for sparsely hairy circular patch occupying much of ventrolateral half of mesopleuron. Metanotum with tomentum uninterrupted, uniformly off white. T1 with median quadrangular reddish-brown discal patch entirely enclosed by pale tomentum and narrow, such that longitudinal band more than half as wide as breadth of apical fascia in dorsal view. T2-T4 with fasciae complete, T2 with fascia with anterolateral extensions of sparser tomentum. T5 with two patches of pale tomentum lateral to and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD.
Structure. Preapical tooth blunt and obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 not noticeably longer than wide (L/W ratio = 1.1). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.6) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion approximately half its medial length; axilla with lateral margin arcuate and carinate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, as long as wide (L/W ratio = 1.0); mesopleuron almost entirely obscured by white tomentum; S4 and S5 with much longer coppery to silvery subapical hairs, which individually are often darker apically; pygidial plate apically rounded, with large deep, well-separated punctures, with the interspaces shining.
Etymology. This species is named in honor of English broadcaster and naturalist Sir David Attenborough in recognition of his inspiring books and television programs on natural history.
Distribution. New Mexico and southern Colorado (Fig. 13). Ecology. HOST RECORDS: The host species of E. attenboroughi is/are presently unknown.
FLORAL RECORDS: Unknown. Figure 13. Occurrence records of E. attenboroughi known to the author (yellow circles).

Discussion.
Epeolus attenboroughi is similar in overall appearance to E. ainsliei and E. rufulus, and the ranges of the three species overlap to some extent. Although BINcompliant sequences are presently not available for E. attenboroughi, partial sequences 421 bp and 289 bp in length are available for two specimens (male and female respectively) collected at the same locality and within one day of each other, and there is virtually no divergence (<1%) between the two. Moreover, the 421 bp sequence does not cluster closely with any sequences from other Epeolus species in a NJ tree of sequences >300 bp in length (Suppl. material 2). The longer of the two partial sequences is most similar (95.2%) to sequences from E. glabratus and E. lectoides (very different species).
In general, there is little morphological variation among examined specimens except in integument coloration; the axillae and mesoscutellum range from entirely black to partially ferruginous. Based on known records, adults of E. attenboroughi are active in summer.
Pubescence. Face with tomentum densest around antennal socket, slightly sparser on clypeus, upper paraocular and frontal areas, and vertexal area. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half densely hairy, except beneath base of fore wing (hypoepimeral area); ventrolateral half sparsely hairy. Metanotum with tomentum uninterrupted, uniformly off white. T1 with discal patch elliptical and very wide, the basal and apical fasciae only narrowly joined laterally. T1 with basal and apical fasciae and T2-T4 with apical fasciae complete, T2 with fascia with basomedially convergent anterolateral extensions of tomentum. T5 with two large patches of pale tomentum lateral to and separate from pseudopygidial area, enclosing pseudopygidial area in triangle, except for medial separation at base. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD.
Structure. Preapical tooth inconspicuous, blunt and obtuse. Labrum with pair of small subapical denticles (approximately at 1/4 length of labrum from apical margin) not preceded by carinae. Frontal keel strongly raised. Scape (missing in holotype) with greatest length 1.6 × greatest width in paratype. F2 (missing in holotype) not noticeably longer than wide (L/W ratio = 1.1) in paratype. Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla intermediate in size, its lateral margin (L) nearly half as long as mesoscutellar width (W) (L/W ratio = 0.4-0.5) and tip not extending beyond midlength of mesoscutellum; axilla with tip visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, as long as wide (L/W ratio = 1.0); S4 and S5 with much longer coppery to silvery subapical hairs, which individually are often darker apically; pygidial plate unusually narrow (Triepeolus-like) and apically rounded, with large deep punctures closely clustered.
Distribution. Mid-Atlantic states to Texas and presumably Mexico, given the close proximity of some collection localities (e.g., Eagle Pass, Texas) to the Mexico-United States border (Fig. 15).
Ecology. HOST RECORDS: The host species of E. australis is/are presently unknown. FLORAL RECORDS:   Discussion. This southeastern species displays minor sexual dimorphism in the coloration of the mesoscutellum, which is bright ferruginous in females and dark ferruginous to black in males. Otherwise, there is very little morphological variation among examined specimens. Although BIN-compliant sequences are presently not available for E. australis, 422 bp sequences were obtained from two male specimens (one from New Jersey, USA and one from South Carolina, USA), and there is virtually no divergence (<1%) between the two. Moreover, these sequences do not cluster with any sequences from other Epeolus species in a NJ tree (Suppl. material 2). Based on known records, adults of E. australis are active in spring.  Robertson, 1902 Figs 16, 17 Epeolus autumnalis Robertson, 1902. Entomol. News 13: 81 (♀, ♂). Webb, 1980. Ill. Nat. Hist. Surv. Bull. 32: 108 (♀) [lectotype designation (by W.E. LaBerge)].

Diagnosis.
The following morphological features in combination can be used to tell E. autumnalis apart from all other North American Epeolus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin, dilated laterally, and like the mesoscutellum black; the mesopleuron is closely (i≤1d) and evenly punctate; the T1 discal patch is so wide that the longitudinal band is barely visible in dorsal view; and the T2 fascia lacks lobe-like anterolateral extensions of tomentum, although a few sparsely scattered pale hairs are sometimes present. Epeolus autumnalis is similar to E. scutellaris in terms of surface sculpture and the patterns of pubescence on the mesosoma and metasoma, but in E. scutellaris at least the axilla is partially to entirely ferruginous (as is often the mesoscutellum), and the axilla is more elongate, extending to or beyond the band of pale tomentum along the posterior margin of the mesoscutellum.
Ecology. See Onuferko (2017) for host and floral records. Floral associations are also indicated in Suppl. material 1.   Diagnosis. Epeolus axillaris can be differentiated from all other Epeolus species in North America by the distinct posteromedial depression of the metanotum; in all other species the metanotum is flat, strongly convex, or weakly convex. Epeolus axillaris closely resembles E. banksi, E. minimus, and E. olympiellus in that the axilla (except sometimes the tip) and mesoscutellum are black; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least half as wide as the breadth of the apical fascia; and the T2 fascia has lobe-like anterolateral extensions of tomentum. However, in all three species the metanotum is flat and the axilla does not extent much beyond the midlength of the mesoscutellum, whereas in E. axillaris the axilla is more elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin. Description. FEMALE: Length 10.0 mm; head length 2.1 mm; head width 2.9 mm; fore wing length 6.9 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, axilla, legs, T5, and pygidial plate. Mandible with apex darker than all but extreme base; preapical tooth slightly lighter than mandibular apex (difficult to see in holotype because mandible closed; described from paratypes). Flagellum brown and (except F1) slightly lighter than partially dark brown (otherwise orange) scape, pedicel, and F1, primarily due to extensive pilosity on flagellum. Axilla only with tip orange. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs, except reddish-orange mesotibia, metatibia, and tarsi, with brown or black more extensive than reddish orange.
Pubescence. Face with tomentum densest around antennal socket. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band wider and joined posteriorly. Mesopleuron densely hairy, except for two sparsely hairy circular patches (one behind pronotal lobe, a larger one occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum uninterrupted except for median bare patch in posterior half, uniformly off white. T1 with median quadrangular black discal patch enclosed by pale tomentum, except for medial separation at apex. T2-T4 with fasciae interrupted medially and narrowed before becoming somewhat broader laterally, T2 with fascia with anterolateral extensions of equally dense tomentum. T5 with two patches of pale tomentum bordering and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD.
Structure. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5-2 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla large, its lateral margin (L) half as long as mesoscutellar width (W) (L/W ratio = 0.5) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin relatively straight and without carina. Metanotum with posteromedial depression beneath overhanging anterior portion. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, not noticeably longer than wide (L/W ratio = 1.1); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures more or less evenly spaced throughout, with the interspaces shining.
Etymology. The name is in reference to the axillae of this species, which are distinctly longer than those of the similar E. minimus and E. olympiellus.
Distribution. California and western Nevada. According to Brumley (1965), this species also ranges into Oregon, but its presence in that state could not be verified in the present study (Fig. 19).
Ecology. HOST RECORDS: The host species of E. axillaris is/are presently unknown. Discussion. This species is most similar to E. minimus and E. olympiellus, and there is overlap in the ranges of all three species. Brumley (1965) recognized E. axillaris as a separate species in which the axilla is more elongate and the metanotum is uniquely depressed posteromedially. The morphological distinction is supported by molecular Figure 19. Approximate geographic range of E. axillaris (orange) based on occurrence records known to the author (yellow circles). data, as sequenced specimens exhibiting these attributes were assigned a separate BIN from either of the other two species.
Material studied. Type material.  (Cockerell, 1907) Figs 20,21,96F Triepeolus banksi Cockerell, 1907a. Entomologist 40: 135 (♂). Epeolus banksi Mitchell, 1962. N. C. Agric. Exp. Stn. Tech. Bull. 152: 442. Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. banksi apart from all other North American Epeolus except E. minimus and E. olympiellus: in females, F2 is at least 1.2 × as long as wide; the mesoscutum has distinct paramedian bands; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is more than 1/4 as long as the entire medial length of the axilla, and the axilla and mesoscutellum are black; the mesopleuron is closely (most i<1d) and evenly punctate; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least half as wide as the breadth of the apical fascia; and the T2 fascia has anterolateral extensions of tomentum. Whereas in E. minimus and E. olympiellus the mesoscutum and metasomal terga have bands of off-white to pale yellow short appressed setae, in E. banksi the mesoscutum and metasomal terga have bands of gray short appressed setae. In E. banksi, the integument is entirely dark brown or black. In E. olympiellus, at least the pronotal lobe is ferruginous. In E. minimus from California, the integument is often entirely dark brown or black, but throughout most of its range E. minimus exhibits reddish-orange coloration on the labrum, antenna, pronotal lobe, and/or legs, except foreleg, from trochanters to tarsi. Both sexes of E. banksi are larger (~10 mm in length) on average than E. minimus or E. olympiellus (7-8 mm in length).
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, antenna, tegula, and legs. Mandible black except apex reddish brown; preapical tooth same color as mandibular apex (difficult to see in holotype; described from non-type specimens). Flagellum, except right F1 and F2, missing in holotype, but brown and (except F1) slightly lighter than conspicuously dark brown scape and pedicel, primarily due to extensive pilosity on flagellum, in non-type specimens. Wing membrane subhyaline, apically dusky. Legs, except reddish-orange tarsi, with brown or black more extensive than reddish orange.
Pubescence. Face with tomentum densest on clypeus and around antennal socket, sparser on upper paraocular area and vertexal area. Dorsum of mesosoma and metasoma with bands of off-white to pale gray short appressed setae. Mesoscutum with paramedian band. Mesopleuron densely hairy, except for two sparsely hairy circular patches (one behind pronotal lobe, a larger one occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum uninterrupted, uniformly off white. T1 with median quadrangular black discal patch enclosed by pale tomentum, except for medial separation at apex. T2-T6 with fasciae interrupted medially, those of T2-T4 narrowed before becoming somewhat broader laterally, T2 with fascia with anterolateral extensions of sparser tomentum. S4 and S5 with long coppery to silvery subapical hairs, which individually are often darker apically.
Structure. Labral apex with pair of small denticles, each preceded by longitudinal carina. Frontal keel not strongly raised. Scape with greatest length 1.6 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.2). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5-2 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla intermediate in size, its lateral margin (L) nearly half as long as mesoscutellar width (W) (L/W ratio = 0.4-0.5) and tip not extending much beyond midlength of mesoscutellum (extending to <2/3 its length); axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin relatively straight and without carina.
Fore wing with three submarginal cells. Pygidial plate apically rounded, with large deep punctures closely clustered.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 even longer than wide (L/W ratio = 1.4); T5 with two patches of pale tomentum bordering and separate from pseudopygidial area present only in female; T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on flat disc of apicomedial region elevated from rest of tergum; S4 and S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD); pygidial plate apically truncate, with small, denser punctures.
Distribution. Maryland to North Carolina (Fig. 21). Ecology. HOST RECORDS: The host species of E. banksi is/are presently unknown. Discussion. Most of the specimens of this species that were examined were collected in the Washington metropolitan area. While  indicated Epeolus banksi as being quite prevalent across the Eastern United States, reportedly ranging from Minnesota to New Jersey and North Carolina, it seems that the name has been commonly misapplied to specimens of E. minimus (as in MacKay and Knerer (1979) for example, and probably by  as well). Epeolus banksi is much larger than E. minimus, and has completely black integument, but unlike similarly dark specimens of E. minimus from California, E. banksi has gray as opposed to pale yellow bands of tomentum on the mesosoma and metasoma. Unfortunately, no recently collected material was available for barcode sequencing, and the specimens seen are all from the early 1900s. The absence of this species from recent collections has not gone unnoticed (e.g in Colla et al. 2012 it is listed among the bee species not collected since 1990). Increased urbanization in and around Washington D.C. may have resulted in the extirpation of this species there, and perhaps it has even disappeared entirely throughout its earlier range. Hence, extensive efforts should be made to rediscover this species, by sampling its apparent historical range between North Carolina and Maryland, to assess its conservation status. The flight season of E. banksi appears to be late summer/early autumn.  Cockerell, 1907Figs 2E, 22, 23, 96E Epeolus barberiellus Cockerell, 1907b.

Diagnosis.
The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. barberiellus apart from all other North American Epeolus except E. americanus and E. asperatus: in females, F2 is not more than 1.1 × as long as wide; the mesoscutum has distinct paramedian bands; the axilla is small to intermediate in size, not extending beyond the midlength of the mesoscutellum and the free portion is less than 1/4 as long as the entire medial length of the axilla, and like the mesoscutellum black; the mesopleuron is closely (i≤1d) and evenly punctate; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least as wide as the breadth of the apical fascia; and the T1 and T2 apical fasciae are interrupted or at least greatly narrowed medially. In E. asperatus the mesopleuron has much denser punctures ventrolaterally (most i<1d) than that of E. barberiellus and the T3 and T4 fasciae are never complete but broken or at least greatly narrowed laterally, as well as medially into separated or narrowly connected oval patches. Epeolus barberiellus is most similar to E. americanus, but in E. americanus the pronotal lobe and legs are brown or black, not reddish orange. Redescription. FEMALE: Length 5.7 mm; head length 1.8 mm; head width 2.3 mm; fore wing length 5.0 mm.
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, mesopleuron, metapleuron, propodeum, legs, metasomal terga (including pygidial plate), and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth as dark as mandibular apex (difficult to see in holotype because mandible closed; described from non-type specimens). Pedicel and flagellum brown and orange in part, slightly lighter than dark brown scape. Pronotal lobe reddish brown. Tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black. T5 and pygidial plate reddish orange.
Pubescence. Face with tomentum densest around antennal socket. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band and moderately dense pale tomentum along margins. Mesopleuron densely hairy, except for almost entirely bare circular patch occupying much of ventrolateral half of mesopleuron. Metanotum with tomentum uninterrupted, uniformly off white. T1 with median quadrangular reddish-brown discal patch enclosed by pale tomentum, except for medial separation at apex, and narrow, such that longitudinal band more than half as wide as width of discal patch in dorsal view. T2 with fascia interrupted medially and without anterolateral extensions of tomentum, although fascia broader laterally with hairs sparser basally. T3 and T4 with fasciae complete and narrowed laterally. T5 with two patches of pale tomentum lateral to and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5-2 MOD at its terminal (difficult to see in holotype; described from non-type specimens). Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.3) and tip not extending beyond midlength of mesoscutellum; axilla with tip visible, but unattached to mesoscutellum for less than 1/4 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, nearly as long as wide (L/W ratio = 0.8); S4 and S5 with much longer coppery to silvery subapical hairs, which individually are often darker apically; pygidial plate orange and V-shaped but apically rounded, with large deep punctures closely clustered.
Distribution. Arizona to west Texas (Fig. 23). Ecology. HOST RECORDS: The host species of E. barberiellus is/are presently unknown.
FLORAL RECORDS: Labels of examined voucher specimens indicate floral associations with Aster (possibly in reference to a plant that is in a different genus now) (Compositae) and Sphaeralcea.

Epeolus basili sp. n.
http://zoobank.org/764C92DA-591F-4302-9337-C9C32D2AD4D8 Figs 24,25,97D,98B Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. basili apart from all other North American Epeolus except E. nebulosus, E. novomexicanus, and E. pusillus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum but at most to the band of pale tomentum along its posterior margin, dilated laterally, and usually ferruginous to some degree (rarely all black) whereas the mesoscutellum ranges from entirely black to partially ferruginous; the axilla's free portion is clearly less than 2/5 as long as its entire medial length; the mesopleuron is closely (most i<1d) and evenly punctate, that of the female is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ) whereas that of the male (excluding the hypoepimeral area) is entirely obscured by white tomentum; the T1-T3 apical fasciae are complete or only very narrowly interrupted medially; the T2 fascia has lobe-like anterolateral extensions of tomentum; and the pseudopygidial area of the female is lunate with the apex at least 2 × and clearly <2.5 × the medial length. Epeolus basili, E. nebulosus, E. novomexicanus, and E. pusillus are all extremely similar to one another. Whereas in E. pusillus the flagellum, except sometimes F1, and metasomal sterna are consistently brown or black and clearly not the same reddish-orange color as the legs (tibiae to tarsi), in E. basili the flagellum, at least ventrally, is the same reddish-orange color as the legs (tibiae to tarsi) as are usually the metasomal sterna. In E. nebulosus and E. novomexicanus the T2-T4 fasciae are on or very little removed from the apical margin, and in both species as well as in E. pusillus the pseudopygidial area of the female is commonly less and no more than 2 × the medial length. By contrast, in E. basili the T2 and T3 (for female) or T2-T4 (for male) fasciae are narrowed medially and removed from the apical margin, and the pseudopygidial area of the female is ≥2 × the medial length. Epeolus basili is also similar to E. scutellaris in that the axilla is large, with the lateral margin arcuate, and that the apical fasciae are complete or only very narrowly interrupted medially. However, in E. scutellaris the pseudopygidial area of the female is even wider (the apex ~2.5-3 × the medial length) than in E. basili, and the mesopleuron of both the female and male is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ).
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, axilla, legs, and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex (difficult to see in holotype; described from paratypes). Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black. S1-S5 reddish orange. Pubescence. Face with tomentum densest around antennal socket, slightly sparser on clypeus, upper paraocular and frontal areas, and vertexal area. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron densely hairy, except for sparsely hairy circular patch occupying much of ventrolateral half of mesopleuron. Metanotum with tomentum uninterrupted, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally. T1-T3 with apical fasciae complete (basal fascia of T1 also), narrowed medially, and removed from apical margin, most noticeably at midline; T2 with fascia with anterolateral extensions of tomentum. T4 with fascia complete. T5 with large, continuous patch of pale tomentum bordering and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex twice as wide as medial length, indicated by silvery setae on flat disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD.
Structure. Preapical tooth obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) half as long as mesoscutellar width (W) (L/W ratio = 0.5) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, but still longer than wide (L/W ratio = 1.2); mesopleuron (excluding hypoepimeral area) entirely obscured by white tomentum; S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep, well-separated punctures, with the interspaces shining.
Etymology. This species is named in honor of my brother, Basil V. Onuferko (1986Onuferko ( -2013. Distribution. Northwestern Mexico and southwestern United States (Fig. 25). Discussion. Structurally, this species is indistinguishable from the other three members of the "pusillus group", and although consistent, the features (differences in integument coloration and patterns of pubescence) that in combination may be used to distinguish E. basili from E. nebulosus, E. novomexicanus, and E. pusillus are subtle. Its status as a separate species is supported by a separate BIN and large barcode sequence divergence (>7.3%) from its nearest neighbor, E. pusillus. In the United States, Epeolus basili appears to be restricted to parts of the American Southwest, east of California. DNA barcoded material with BIN-compliant sequences. Available. BOLD:ACR5356. See Type material for specimens examined and sequenced (indicated by unique CCDB-plate and well number). Diagnosis. Unique to E. bifasciatus among North American species of Epeolus are each of the following morphological features: the frontal area bears a pair of granulose protrusions, each located near the upper mesal margin of the compound eye; the pronotal collar is elongate, dilated laterally to about 2 × the medial length in dorsal view; and the dorsum of the metasoma has at most two bright orange-yellow fasciae (usually a basal fascia on T1 and always an apical fascia on T2). Similar species occur in Mexico and Central America, but their occurrence in Canada and the United States has not been confirmed.
Ecology. See Onuferko (2017) for host and floral records. Floral associations are also indicated in Suppl. material 1, which includes newly discovered associations with Coreopsis tinctoria Nutt. (Compositae) and Verbena hastata L. (Verbenaceae) based on labels of examined voucher specimens.
Discussion. Epeolus bifasciatus is the only species within the "Trophocleptria group" verified as occurring north of Mexico. Originally a genus, Trophocleptria Holmberg was later considered a subgenus of Epeolus (Michener 2000). Although its constituent species seem to form a natural group, a phylogenetic study by Rightmyer (2004) found that maintaining the subgeneric designation rendered Epeolus (Epeolus) paraphyletic, so Michener (2007) treated Trophocleptria as a distinct species group within Epeolus.
Epeolus fumipennis Say has been listed as occurring in Kansas (Snow 1879, in which E.T. Cresson was acknowledged for aiding in identification), but was probably confused with E. bifasciatus, a species that is common in that state (Ascher and Pickering 2017). Brumley (1965) examined specimens at the ANSP and KUNHM from the Midwestern and Southeastern United States labelled as E. fumipennis that according to him were clearly E. bifasciatus. The primary type of E. fumipennis was probably destroyed along with much of Thomas Say's insect collection (LeConte 1859:v-vi, xix [footnote]), but the medially-narrowed ferruginous pronotal collar and yellow fasciae on T1 and T2 (contrasting with the whitish fasciae on the remaining terga), as well as  its occurrence in Mexico, strongly suggest that this species is in the "Trophocleptria group". However, in E. fumipennis the mesoscutum has distinct paramedian bands, which are absent in E. bifasciatus, and no specimens from Canada or the United States fitting such a description were seen.
Material Diagnosis. The following morphological features in combination can be used to tell E. brumleyi apart from all other North American Epeolus: the frontal carina is weakly convex, such that the supraclypeal area is barely protuberant in lateral view; the mesoscutum has distinct paramedian bands; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is at least 1/4 as long as (and less than 2/5) the entire medial length of the axilla, relatively straight along the medial margin, and ferruginous to some degree whereas the mesoscutellum is typically all black; the fore wing has three submarginal cells; the T1 basal and apical fasciae are subparallel; T2-T4 have complete fasciae; and the T2 fascia has a pair of anterolateral extensions of tomentum that are weakly convergent basally. Epeolus brumleyi most closely resembles E. australis, but in E. australis the frontal carina is strongly convex and the pygidial plate of the male is narrower (the medial length is ~1.5 × the basal width) than in E. brumleyi (the medial length ≈ the basal width). Description. FEMALE: Length 7.6 mm; head length 1.9 mm; head width 2.7 mm; fore wing length 5.8 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, axilla, legs, metasomal terga (including pygidial plate), and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex (difficult to see in holotype because mandible closed; described from paratypes). Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Clypeus, upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron densely hairy, except for two almost entirely bare patches (one beneath base of fore wing (hypoepimeral area), a larger circular patch occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum rubbed off medially in holotype, but uninterrupted and uniformly off white in paratypes. T1 with discal patch elliptical and very wide, the basal and apical fasciae only narrowly joined laterally. T1 with basal fascia complete and apical fascia interrupted medially, T2-T4 with fasciae complete, T2 with fascia with anterolateral extensions of sparser tomentum. T5 with two large patches of pale tomentum lateral to and contacting pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by much more than 1/4 MOD.
Structure. Preapical tooth blunt and obtuse. Labrum with submedial pair of small denticles, apex edentate. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal (difficult to see in holotype; described from paratypes). Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and tip not extending beyond midlength of mesoscutellum; axilla with tip visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, nearly as long as wide (L/W ratio = 0.9); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered. Etymology. This species is named after its discoverer, Richard L. Brumley, who recognized it and five other Epeolus formally described here (E. axillaris, E. chamaesarachae, E. diadematus, E. splendidus, and E. tessieris) as new species.
Distribution. Arizona to Texas and presumably Mexico, given the close proximity of some collection localities (e.g., Douglas, Arizona) to the Mexico-United States border (Fig. 29).
Ecology. HOST RECORDS: Four representatives of this species were collected at a single site in southeast Arizona in the spring of 2016 (see Material studied), from or flying near patches of Chamaesaracha (A. Gray) Benth. (Solanaceae), which were visited by large numbers of Colletes (presumably the host species). Using Stephen's (1954) key, collected females were identified as C. scopiventer Swenk (a species known only from females) whereas males were identified (based in part on examination of the terminalia, which were excised) as C. wickhami Timberlake (a species known only from males), and sequenced specimens of both sexes were assigned the same BIN (BOLD:AAJ7578).
Discussion. Epeolus brumleyi is a southwestern species that exhibits very little intraspecific morphological variation. Adults have been collected in every month from March to September, and barcoded specimens collected in early May, June, and late August were assigned the same BIN. Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. canadensis apart from all other North American Epeolus except E. compactus and E. ferrarii: in females, F2 is at least 1.2 × as long as wide; the mesoscutum has a small anteromedial patch of pale tomentum; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is more than 1/4 as long as the entire medial length of the axilla, and the axilla (except sometimes the tip) and mesoscutellum are black; the mesopleuron is closely (most i<1d) and evenly punctate; and the T2 fascia lacks lobelike anterolateral extensions of tomentum, although it may be broader laterally. Epeolus canadensis differs from E. compactus and E. ferrarii in the shape of the T1 discal patch, which in E. canadensis is distinctly triangular or semicircular (the basal fascia is conspicuously arched and fully continuous with the longitudinal band) and its medial longitudinal extent is more than 1/3 the lateral extent. In E. compactus and E. ferrarii the shape of the T1 discal patch is variable but typically quadrangular with the basal and apical fasciae subparallel and separated by a distinct longitudinal band. In E. compactus, the medially-interrupted T1 basal and apical fasciae may be so broad laterally that they are joined, resulting in a diamond shape with concave sides. In E. ferrarii the discal patch may be trapezoidal or almost semicircular, but if at all semicircular its medial longitudinal extent is at most 1/3 the lateral extent and the basal fascia and longitudinal band are at least joined at somewhat of an angle.

Epeolus carolinus
Diagnosis. The following morphological features in combination can be used to tell E. carolinus apart from all other North American Epeolus: the mandible has a blunt, obtuse preapical tooth; the axilla is elongate, extending well beyond the midlength of the mesoscutellum but not beyond its posterior margin, and the free portion is distinctly hooked; the mesopleuron is closely (most i<1d) and evenly punctate; and the metasomal fasciae are yellow to orange and interrupted medially. Epeolus carolinus resembles E. deyrupi in general appearance, but in E. deyrupi the axilla is larger, extending as far back as or beyond the posterior margin of the mesoscutellum, and dilated laterally but relatively straight along the medial margin, and the mesopleuron commonly has sparser punctures ventrolaterally (i≤2d) than that of E. carolinus, with the interspaces shining or somewhat dull due to tessellate surface microsculpture.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, axilla, mesoscutum, mesoscutellum, legs, and pygidial plate. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex (difficult to see in holotype; described from paratype). Antenna brown except scape, pedicel, and F1 extensively orange. Pronotal lobe and tegula pale ferruginous to amber. Mesoscutum with orange spot anterolaterally between pronotal lobe and tegula. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white and yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron densely hairy, except for two sparsely hairy circular patches (one behind pronotal lobe, a larger one occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum sparser medially, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally by few sparsely scattered pale hairs (not joined in paratype and multiple non-type specimens). T1-T5 with apical fasciae interrupted medially, those of T2-T4 somewhat broader laterally, T2 with fascia without anterolateral extensions of tomentum. T6 with fascia complete. S4 and S5 with long coppery to silvery subapical hairs.
Structure. Preapical tooth inconspicuous, blunt and obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by less than 1 MOD at its terminal (difficult to see in holotype; described from non-type specimens). Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.6) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin arcuate and carinate. Fore wing with three submarginal cells. Pygidial plate apically rounded, with large deep punctures more or less evenly spaced throughout, with the interspaces shining.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 even longer than wide (L/W ratio = 1.7); T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on flat disc of apicomedial region elevated from rest of tergum; S4 and S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD); pygidial plate apically truncate, with small, denser punctures.
Ecology. HOST RECORDS: The host species of E. carolinus is/are presently unknown.  Discussion. This southeastern species is quite variable in terms of integument coloration and pubescence on the metasomal terga. The mesoscutellum and disc of T1 range from entirely black to entirely ferruginous. The axillae appear to be at least partially ferruginous. Whereas T1 and T2 have prominent yellow fasciae, the fasciae on the remaining terga range from prominent to reduced or even absent. Adults of Epeolus carolinus are active in September and October.
Material studied. Type material. Primary: USA: North Carolina: Kill Devil Hills, 12.ix.1956 Diagnosis. Epeolus chamaesarachae does not closely resemble any other species of Epeolus except E. diadematus. Unique in the genus to both species are each of the following morphological features: the vertexal area has two pairs of shiny (usually impunctate) protrusions, the mesoscutum is distinctly ornamented with mostly separate patches of (but some intermixed) pale and ferruginous tomentum, and the T2 fascia has two pairs of anterolateral extensions of tomentum. The difference is that in E. chamaesarachae the mesopleuron has sparser punctures ventrolaterally (most i>1d) whereas in E. diadematus the mesopleuron has denser (most i≤1d) and more numerous punctures ventrolaterally. Description. FEMALE: Length 7.0 mm; head length 2.0 mm; head width 2.6 mm; fore wing length 5.7 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal collar, pronotal lobe, tegula, axilla, mesoscutellum, and legs. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex (difficult to see in holotype; described from paratypes). Antenna dark brown except scape, pedicel, and F1 brownish orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Vertexal area with tomentum mostly ferruginous. Dorsum of mesosoma with bands of off-white and ferruginous short appressed setae. Dorsum of metasoma with bands of off-white to pale yellow short appressed setae. Pronotal lobe entirely obscured by pale tomentum. Pronotal collar with tomentum black medially, pale and ferruginous laterally. Mesoscutum with paramedian band of pale tomentum; ferruginous and pale tomentum encircling black spots medially and laterally, respectively, on anterior margin; and fer-ruginous tomentum along medial mesoscutal line and parapsidal line. Mesopleuron with upper half densely hairy, although scrobe visible; ventrolateral half nearly bare. Metanotum with tomentum uninterrupted, off white laterally and black medially. T1 with median diamond-shaped black discal patch enclosed by pale tomentum, except for medial separation at apex. T1 with apical fascia with black spot posterolaterally. T2-T4 with fasciae interrupted medially, T2 with fascia with paired anterolateral extensions of tomentum. T3 and T4 with fasciae interrupted laterally, with medial portion on apical margin and lateral portion encircling black tomentum on apical margin. T5 with two large patches of pale tomentum lateral to and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD.
Structure. Labral apex with pair of small denticles (preceded by submedial pair of small denticles) separated by shallow concavity and between second pair of apical lobes. Frontal keel strongly raised. Vertexal area with two pairs of impunctate shiny protrusions. Scape with greatest length 1.6 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5 MOD at its terminal. Mesoscutellum strongly bigibbous. Axilla intermediate in size, its lateral margin (L) nearly half as long as mesoscutellar width (W) (L/W ratio = 0.4-0.5) and tip not extending beyond midlength of mesoscutellum; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, nearly as long as wide (L/W ratio = 0.8); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered apically and sparser basally, with the interspaces shining.
Etymology. The name is in reference to the genus of flowers (Chamaesaracha) on which the holotype was collected.
Ecology. HOST RECORDS: The female PCYU paratype collected by H.T. Ngo (see Material studied) is labelled with the same collection information as three Colletes specimens (2♀, 1♂) of the presumed host species, which were barcoded and all share the same BIN (BOLD:AAJ7578). Using Stephen's (1954) key, the two females were identified as C. scopiventer (a species known only from females) whereas the male was identified (based in part on examination of the terminalia, which were excised) as C. wickhami (a species known only from males). Discussion. This species and the very similar E. diadematus are unusual among Epeolus in that the vertexal area has two pairs of shiny (usually impunctate) protrusions, and dorsally the mesosoma and metasoma have unique patterns of ferruginous (mesosoma only) and off-white to pale yellow short appressed setae. Epeolus chamaesarachae occurs in the Southwestern United States, and its flight season, based on material examined, is late summer.
Material studied. Type material. Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. compactus apart from all other North American Epeolus except E. canadensis and E. ferrarii: in females, F2 is at least 1.2 × as long as wide; the mesoscutum has a small anteromedial patch of pale tomentum; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is more than 1/4 as long as the entire medial length of the axilla, and the axilla (except sometimes the tip) and mesoscutellum are black; the mesopleuron is closely (most i<1d) and evenly punctate; and the T2 fascia lacks lobe-like anterolateral extensions of tomentum, although it may be broader laterally. Epeolus compactus is most similar to E. ferrarii, and in both species the T1 discal patch is typically quadrangular with the basal and apical fasciae subparallel and separated by a distinct longitudinal band, but in E. ferrarii the T2-T4 fasciae are not broadened medially into rounded lobes (as in E. compactus) but evenly broad or tapering until separated medially. Epeolus canadensis differs from both species in that the T1 discal patch is distinctly triangular or semicircular (the basal fascia is conspicuously arched and fully continuous with the longitudinal band) and its medial longitudinal extent is more than 1/3 the lateral extent. In E. compactus, the medially-interrupted T1 basal and apical fasciae may be so broad laterally that they are joined, resulting in a diamond shape but with concave sides; in E. canadensis the lateral sides are straight or convex.
Ecology. See Onuferko (2017) for host and floral records. Floral associations are also indicated in Suppl. material 1.
Discussion. Epeolus compactus is a commonly collected species, widespread in Western North America. It is most similar to E. canadensis and E. ferrarii. In the original description of E. crucis Cockerell, the holotype was said to have been initially identified as E. compactus by W.J. Fox, but Cockerell (1904) considered it to be distinct, mainly because of differences in coloration and pubescence. The specimen (unusually) has abundant pale tomentum on the discs of the metasomal terga (Fig. 38A), but representatives of several species (e.g., E. ainsliei, E. minimus, and E. novomexicanus) exhibiting atypical abundance of pale tomentum on the mesosoma and metasoma were also observed. Despite the presence of pale tomentum, the discal patch is quadrangular/diamond-shaped ( Fig. 38A) as is typical for E. compactus (Fig. 38B), and the fascia of T2 is separated medially into rounded lobes. In the E. crucis holotype, the axillae and mesoscutellum are (unusually) ferruginous, but it is not unprecedented for species of the genus to have representatives displaying atypical integument coloration. Interestingly, Brumley (1965) treated E. crucis as distinct, but the features listed as unique for that species are evident only in the holotype of E. rufulus. In fact, his key does not work for the holotypes of E. crucis and E. novomexicanus, which Brumley believed to be the same species. Unlike in E. rufulus, in the E. crucis holotype the axillae do not extend beyond the midlength of the mesoscutellum, and the axilla is not conspicuously diverging from the side of the mesoscutellum -the free portion is less than 1/3 as long as the entire medial length of the axilla.  Diagnosis. The following morphological features in combination can be used to tell E. deyrupi apart from all other North American Epeolus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum, dilated laterally, and like the mesoscutellum ferruginous; the mesopleuron commonly has sparser punctures ventrolaterally (i≤2d) than in upper half, with the interspaces shining or somewhat dull due to tessellate surface microsculpture; and the T1-T3 apical fasciae are interrupted and (to varying degrees) brownish orange medially and off white laterally. Epeolus deyrupi resembles E. andriyi, E. floridensis, E. howardi, and E. packeri in that the axilla is large, with the lateral margin arcuate, and like the mesoscutellum ferruginous, and that the T1-T3 apical fasciae are interrupted medially. However, in E. deyrupi the pseudopygidial area of the female is wider (the apex >2 × the medial length) than in E. andriyi, E. floridensis, or E. howardi (the apex <2 × the medial length), and the T1 basal fascia is absent or reduced to a pair of small patches of pale tomentum whereas in E. andriyi, E. floridensis, and E. howardi T1 has a distinct, although often medially-interrupted, basal fascia. Epeolus deyrupi closely resembles E. packeri, but in E. packeri the mesopleuron has denser punctures ventrolaterally (most i<1d) than that of E. deyrupi and the metasomal terga have pale but not brownish orange pubescence.
Pubescence. Face with tomentum densest around antennal socket. Dorsum of mesosoma and metasoma with bands of off-white and brownish orange short appressed setae. Mesoscutum with paramedian band. Mesopleuron mostly bare, but tomentum moderately dense ventrally as well as between two almost entirely bare patches (one beneath base of fore wing (hypoepimeral area), a larger circular patch occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum uninterrupted except for median bare patch in posterior half (also bare along posterior margin), uniformly off white. T1 with basal fascia reduced to pair of small patches of off-white tomentum; T1-T4 with apical fasciae brownish orange medially and off white laterally, and medially interrupted and removed from apical margin; T2 with fascia without anterolateral ex-tensions of tomentum. T4 with fascia interrupted laterally. T5 with two patches of pale tomentum bordering and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Preapical tooth obtuse. Labral apex with pair of small denticles, each preceded by longitudinal carina. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.2). Preoccipital ridge not joining hypostomal carina, from which it is separated by less than 1 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.6) and tip nearly extending as far back as apex of horizontal dorsal portion of mesoscutellum; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, not noticeably longer than wide (L/W ratio = 1.1); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate with apex slightly concave and large deep punctures closely clustered basally and sparser apically, with the interspaces shining.
Etymology. This species is named after its discoverer, Dr. Mark A. Deyrup, who recognized it as a new species and brought his discovery to my attention.
Distribution. Florida and coastal Georgia (Fig. 40 protrusions, the mesoscutum is distinctly ornamented with mostly separate patches of (but some intermixed) pale and ferruginous tomentum, and the T2 fascia has two pairs of anterolateral extensions of tomentum. The difference is that in E. diadematus the mesopleuron has denser punctures ventrolaterally (most i≤1d) whereas in E. chamaesarachae the mesopleuron has sparser (most i>1d) and fewer punctures ventrolaterally. Description. FEMALE: Length 6.9 mm; head length 2.0 mm; head width 2.6 mm; fore wing length 6.0 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal collar, pronotal lobe, tegula, axilla, mesoscutellum, and legs. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex. Antenna dark brown except scape, pedicel, and F1 brownish orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Vertexal area with tomentum mostly ferruginous. Dorsum of mesosoma with bands of off-white and ferruginous short appressed setae. Dorsum of metasoma with bands of off-white to pale yellow short appressed setae. Pronotal collar with tomentum black medially, pale and ferruginous laterally. Mesoscutum with paramedian band of pale tomentum; ferruginous and pale tomentum encircling black spots medially and laterally, respectively, on anterior margin; and ferruginous tomentum along medial mesoscutal line and parapsidal line. Mesopleuron with upper half densely hairy, although scrobe visible; ventrolateral half nearly bare. Metanotum with tomentum uninterrupted, off white laterally and black medially. T1 with median diamond-shaped black discal patch enclosed by pale tomentum, except for medial separation at apex. T1 with apical fascia with black spot posterolaterally. T2-T4 with fasciae interrupted medially, T2 with fascia with paired anterolateral extensions of tomentum. T3 and T4 with fasciae interrupted laterally, with medial portion on apical margin and lateral portion encircling black tomentum on apical margin. T5 with two large patches of pale tomentum lateral to and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD. Surface sculpture. Punctures dense, but those of head and mesosoma sparser in some areas, larger, deep, and distinct. Labrum mostly with larger and sparser punctures (i=1-2d) than clypeus (i≤1d). Upper paraocular area and vertexal area sparsely punctate (i=1-2d), the interspaces shining. Mesoscutum, mesoscutellum, and axilla coarsely and densely rugose-punctate; the interspaces shining. Tegula densely punctate mesally (i=1-2d), much less so laterally (i>2d). Mesopleuron with denser (i<1d) punctures in upper half than ventrolateral half, although ventrolateral half with most interspaces small (i≤1d); the interspaces shining. Metasomal terga with punctures very fine, dense (i≈1d), evenly distributed on disc.
Structure. Labral apex with two pairs of small denticles (the middlemost pair preceded by submedial pair of small denticles and separated by shallow concavity). Frontal keel strongly raised. Vertexal area with two pairs of nearly impunctate shiny protrusions. Scape with greatest length 1.6 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum strongly bigibbous. Axilla intermediate in size, its lateral margin (L) nearly half as long as mesoscutellar width (W) (L/W ratio = 0.4-0.5) and tip not extending much beyond midlength of mesoscutellum (extending to <2/3 its length); axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin somewhat arcuate. Fore wing with three submarginal cells. Pygidial plate mostly hidden in holotype, but apically truncate in paratypes.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, nearly as long as wide (L/W ratio = 0.8); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered apically and sparser basally, with the interspaces shining.
Etymology. The name is in reference to the four shiny, usually impunctate, tubercles on the vertexal area of the head of this species. From the Latin, "diadema" (royal headband).
Distribution. Texas and presumably Mexico, given the close proximity of some collection localities (e.g., Southmost, Texas) to the Mexico-United States border (Fig. 42).
Ecology. HOST RECORDS: The host species of E. diadematus is/are presently unknown.
FLORAL RECORDS: The label of one examined voucher specimen indicates a floral association with Engelmannia pinnatifida A.Gray ex Nutt. (Compositae). This species has also been collected from Aphanostephus riddellii Torr. & A. Gray (Compositae) (J. Neff, personal communication, 2016).
Discussion. This species and E. chamaesarachae are very similar in terms of integument coloration, pubescence, and structure, and are presumably sister species. Specimens of E. diadematus are distinct from those designated as E. chamaesarachae in that the mesopleuron has much denser punctation. The status of E. diadematus as a separate species is further supported by a separate BIN and large barcode sequence divergence (3.2%) from its nearest neighbor, E. chamaesarachae (Suppl. material 2). The ranges and flight seasons of these species also differ. With one exception, examined specimens of E. diadematus were collected in spring, and all are from Coastal or South Texas. By contrast, E. chamaesarachae occurs further west in the United States, and adults are active in late summer.
Material studied. Type material. Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. erigeronis apart from all other North American Epeolus except E. ilicis and E. inornatus: the mandible is simple; the axilla does not attain the midlength of the mesoscutellum but the free portion is distinctly hooked, with the tip unattached to the mesoscutellum for more than 1/3 of the entire medial length of the axilla; the pronotal collar and metasomal terga are black; the metasomal terga have rather fine punctures; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length and not in contact with two large patches of pale tomentum (one on each side) throughout its length (in contact only at apex, diverging basally). Although in all three species the mesopleuron is closely and evenly punctate, in E. erigeronis the punctures are more variable in size, with many smaller punctures among large ones, and most interspaces are narrower such that the surface appears to be very coarsely and densely rugose-punctate. By contrast, in E. ilicis and E. inornatus the mesopleuron has punctures that are similar in size and shiny interspaces that are commonly equal to the puncture diameters.  Redescription. FEMALE: Length 8.6 mm; head length 2.2 mm; head width 3.0 mm; fore wing length 6.3 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, and legs. Mandible with apex darker than all but extreme base. Antenna brown except scape, pedicel, and F1 orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half hairy, except beneath base of fore wing (hypoepimeral area); ventrolateral half nearly bare. Metanotum with tomentum uninterrupted except for median bare patch in posterior half, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally. T1 and T2 with apical fasciae interrupted medially, those of T2 and T3 somewhat broader laterally, T2 with fascia with faint anterolateral extensions of sparser tomentum. T3 and T4 with fasciae complete. T5 with two large patches of pale tomentum lateral to and separate from pseudopygidial area. T5 with pseudopygidial area campanulate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by 1/3 MOD.
Structure. Mandible without preapical tooth. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.6). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla intermediate in size, its lateral margin (L) nearly half as long as mesoscutellar width (W) (L/W ratio = 0.4-0.5) and tip attaining midlength of mesoscutellum; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin arcuate and carinate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, but still longer than wide (L/W ratio = 1.3); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered basomedially and sparser apically and laterally, with the interspaces shining.
Ecology. HOST RECORDS: The host species of E. erigeronis is/are presently unknown. Discussion. Epeolus erigeronis exhibits very little intraspecific morphological variation. However, in some specimens the axillae are partially ferruginous whereas in others they and the mesoscutellum are entirely black. Based on examined records, adults of E. erigeronis are active throughout spring.
Although BIN-compliant sequences are presently not available for E. erigeronis, four partial sequences (three 422 bp and one 394 bp in length) are available for specimens from North and South Florida, and these sequences form a distinct cluster that does not include any sequences from other Epeolus species in a NJ tree (Suppl. material 2).
Material studied. Type material. Primary: USA: Florida: Levy County, 13.iv.1955 Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. ferrarii apart from all other North American Epeolus except E. canadensis and E. compactus: in females, F2 is at least 1.2 × as long as wide; the mesoscutum has a small anteromedial patch of pale tomentum; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is more than 1/4 as long as the entire medial length of the axilla, and the axilla (except sometimes the tip) and mesoscutellum are black; the mesopleuron is closely (most i<1d) and evenly punctate; and the T2 fascia lacks lobe-like anterolateral extensions of tomentum, although it is broader laterally. Epeolus ferrarii is most similar to E. compactus, and in both species the T1 discal patch is typically quadrangular with the basal and apical fasciae subparallel and separated by a distinct longitudinal band, but in E. compactus the T2-T4 fasciae are not evenly broad or tapering until separated medially (as in E. ferrarii) but broadened medially into rounded lobes, which may be joined or separated. Epeolus canadensis differs from both species in that the T1 discal patch is distinctly triangular or semicircular (the basal fascia is conspicuously arched and fully continuous with the longitudinal band) and its medial longitudinal extent is more than 1/3 the lateral extent. In E. ferrarii the discal patch may be trapezoidal or almost semicircular, but if at all semicircular its medial longitudinal extent is at most 1/3 the lateral extent and the basal fascia and longitudinal band are at least joined at somewhat of an angle.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, legs, and pygidial plate. Mandible with apex and preapical tooth darker than all but basal quarter. Antenna brown except F1 extensively orange. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs from tibia to tarsus extensively reddish orange. Pygidial plate orange along apical margin, otherwise dark brown.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with anteromedial horseshoe-shaped patch of pale tomentum. Mesopleuron densely hairy, except for two sparsely hairy circular patches (one behind pronotal lobe, a larger one occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum uninterrupted, pale yellow laterally and black medially. T1 with median elliptical verging on semicircular discal patch. T1-T3 with apical fasciae medially interrupted, narrowed (broader laterally), and removed from apical margin; T2 with fascia without anterolateral extensions of tomentum. T4-T6 with fasciae complete, those of T4 and T5 somewhat narrowed medially. S4 and S5 with long coppery to silvery subapical hairs, which individually are often darker apically.
Structure. Labral apex with pair of small denticles, each preceded by longitudinal carina. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width.
F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5 MOD at its terminal (difficult to see in holotype; described from paratypes). Mesoscutellum weakly bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and tip not extending much beyond midlength of mesoscutellum (extending to <2/3 its length); axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically rounded, with large deep punctures closely clustered medially and sparser laterally, with the interspaces shining.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 slightly but not noticeably longer than wide (L/W ratio = 1.1); T5 with large, nearly continuous patch of pale tomentum bordering and separate from pseudopygidial area present only in female; T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum; S4 and S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD); pygidial plate apically truncate, with small, denser punctures.
Etymology. This species is named in honor of my colleague, Rafael Ferrari, with whom I collected this species in Southwestern New Mexico, USA.
Distribution. Arizona and New Mexico to southeastern Mexico (Fig. 46). Ecology. HOST RECORDS: The host species of E. ferrarii is/are presently unknown. FLORAL RECORDS: Labels of examined voucher specimens indicate a floral association with Melilotus albus.
Discussion. Epeolus ferrarii is a cryptic species that most closely resembles E. canadensis and E. compactus, and can only be differentiated from these two species on the basis of very subtle differences in the patterns of pubescence on the metasomal terga. Its status as a separate species is supported by a separate BIN, but unusually its nearest neighbor is E. splendidus (a very different species, although presumably in the same species group), from which E. ferrarii exhibits a large barcode sequence divergence (3.9%). Although most species of Epeolus were described from a female namebearing type, a male specimen is designated as the holotype of E. ferrarii because a barcode-compliant sequence is associated with it and because the collection locality is more precise than for the available female specimens, one of which is herein designated as the allotype.    Cockerell, 1907c. Ann. Mag. Nat. Hist. 20: 60 (♂).

Epeolus flavofasciatus
Diagnosis. The following morphological features in combination can be used to tell E. flavofasciatus apart from all other North American Epeolus: the dorsum of the mesosoma and metasoma have bright or pale yellow pubescence, the mesoscutum has distinct paramedian bands, the axilla does not attain the midlength of the mesoscutellum, and T1 has a median triangular or semicircular discal patch. Epeolus canadensis resembles E. flavofasciatus in that the integument is mostly black, the axilla does not attain the midlength of the mesoscutellum, and T1 has a median triangular or semicircular discal patch, but in E. canadensis the mesoscutum has a distinct anteromedial patch of pale tomentum instead of paramedian bands. Epeolus flavofasciatus is quite large for Epeolus (≥9 mm in length), and the pygidial plate of the male is narrower than that in most species, so males may be confused with Triepeolus. However, in E. flavofasciatus the mandible has a blunt, obtuse preapical tooth, whereas in all Triepeolus the mandible is simple.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, axilla, legs, and pygidial plate. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex (difficult to see in the E. flavofasciatus lectotype because mandible closed; described from non-type specimens). Antenna brown except scape, pedicel, and F1 extensively orange. F2 with orange spot basally. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane dusky subhyaline, slightly darker at apex. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white and bright to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron sparsely hairy except mesally with densely hairy sigmoid patch and ventrally. Metanotum with tomentum uninterrupted, uniformly black (uniformly pale yellow in the E. agaricifer holotype and multiple non-type specimens, uniformly black or to varying degrees bright or pale yellow laterally and black medially in other non-type specimens). T1 with median semicircular black discal patch enclosed by pale tomentum (basal fascia widely separated medially and with much tomentum rubbed off in the E. flavofasciatus lectotype, but conspicuously arched and narrowly interrupted medially in nontype specimens). T2-T4 with fasciae complete, T2 with fascia without anterolateral extensions of tomentum. T5 with two large patches of pale tomentum lateral to and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by much more than 1/4 MOD.
Structure. Labral apex with pair of small denticles preceded by submedial pair of small denticles and separated by shallow concavity. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5-2 MOD at its terminal (difficult to see in the E. flavofasciatus lectotype; described from non-type specimens). Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and tip not extending beyond midlength of mesoscutellum; axilla with tip clearly visible, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, but still longer than wide (L/W ratio = 1.2); S3-S5 with much longer coppery to silvery subapical hairs, which individually are often darker apically; pygidial plate unusually narrow (Triepeolus-like) and apically rounded, with large deep punctures closely clustered.
Distribution. Mexico, excluding the Baja California Peninsula, and southwestern United States to central America (Fig. 48).
Ecology. HOST RECORDS: The host species of E. flavofasciatus is/are presently unknown.
FLORAL RECORDS: Labels of examined voucher specimens indicate floral associations with Heterotheca subaxillaris and Vicia L. (Leguminosae).
Discussion. Smith (1879) described E. flavofasciatus from both sexes, represented by two syntypes (one female and one male) deposited at the NHMUK. Both specimens were examined, and the female is herein designated as the lectotype because it is in better condition, because most Epeolus spp. are represented by female name-bearing types, and because Smith (1879) provided a more complete description of the female. The male syntype at the NHMUK is herein designated as the lectoallotype. Cockerell (1907)   specimen of T. agaricifer, and agree with Rightmyer's treatment. Two specimens (both males) were barcoded, one of which is from Southeast Arizona, USA (nearer the type locality of T. agaricifer: Beulah, New Mexico, USA) and the other is from Jalisco, Mexico (nearer the type locality of E. flavofasciatus: Oaxaca, Mexico), and both were assigned the same BIN. Brumley also described this species under the manuscript name Epeolus artus [nomen nudum] in 1965.
There is some intraspecific variation in the pubescence on the metanotum, which ranges from entirely yellow to medially or mostly black, and T1, in which the apical fascia is either complete or interrupted medially, with differences not conforming to any discernable geographic pattern. Based on examined records, the range of this species appears to be quite continuous from the American Southwest to Central America.
Among the examined specimens of this species is one that appears to be the first known example of bilateral gynandromorphism in Epeolus (see Material studied). Descriptions and images of the aberrant features exhibited by the specimen are published separately (Onuferko 2018

Diagnosis.
The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. floridensis apart from all other North American Epeolus: the axilla is large, with the tip extending as far back as or beyond the posterior margin of the mesoscutellum, dilated laterally, and like the mesoscutellum ferruginous; the mesopleuron is closely (i≤1d) and evenly punctate; T1 is (with few exceptions) ferruginous and with a distinct, although sometimes medially-interrupted, basal fascia; the mesoscutum and metasomal terga have bands of pale gray to white short appressed setae; at least the T1-T3 apical fasciae are distinctly interrupted medially; and the pseudopygidial area of the female is lunate with the apex <2 × the medial length. Epeolus floridensis is similar to E. howardi, but in E. howardi the mesoscutum and metasomal terga have bands of bright or pale yellow short appressed setae and the metasomal terga (including T1) are black. Epeolus floridensis is also similar to E. packeri, but in E. packeri the T1 basal fascia is absent or reduced to a pair of small patches of pale tomentum, the metasomal terga (including T1) are black, and the pseudopygidial area of the female is lunate with the apex >2 × the medial length.
Integument coloration. Black in part, at least partially ferruginous on mandible, labrum, clypeus, antenna, pronotal collar, pronotal lobe, tegula, axilla, mesoscutum, mesoscutellum, metanotum, mesopleuron, metapleuron, propodeum, legs, T1, T5, pygidial plate, and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex. Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Mesoscutum almost entirely reddish brown. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black. Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale gray short appressed setae. Mesoscutum with paramedian band. Mesopleuron sparsely hairy, but tomentum moderately dense along margins. Metanotum with tomentum uninterrupted, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally by few sparsely scattered pale hairs. T1-T4 with apical fasciae interrupted medially and somewhat broader laterally, T2 with fascia without anterolateral extensions of tomentum. T5 with two patches of pale tomentum lateral to and contacting pseudopygidial area. T5 with pseudopygidial area lunate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Preapical tooth inconspicuous, blunt and obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.6). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5 MOD at its terminal (difficult to see in holotype; described from non-type specimens). Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.6) and tip extending as far back as apex of horizontal dorsal portion of mesoscutellum; axilla with tip clearly visible, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: upper paraocular area very finely and sparsely punctate in part, the interspaces shining; F2 shorter, but still longer than wide (L/W ratio = 1.3); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered basomedially and sparser apically and laterally, with the interspaces shining.
Distribution. Florida peninsula (Fig. 50). Discussion. Epeolus floridensis exhibits very little intraspecific morphological variation. However, one specimen was observed in which T1 is as dark as the remaining terga rather than bright ferruginous, the usual state. Also, in some males the upper paraocular area has comparatively fewer punctures than in females while in other specimens punctures are similarly dense between the sexes. Based on examined records, adults of E. floridensis appear to be most active in spring, although  lists some paratypes that were collected in mid-July.
Material studied. Type material.    apart from all other North American Epeolus: the mandible has a blunt, obtuse preapical tooth; in females, F2 is less than 1.2 × as long as wide; the axilla does not attain the midlength of the mesoscutellum but the free portion is distinctly hooked, with the tip unattached to the mesoscutellum for more than 1/3 of the entire medial length of the axilla; the mesopleuron is closely and evenly punctate (i≤1d), with the interspaces shining and punctures similar in size; the legs are usually darker, at least from the metacoxa to metatibia; the metasomal terga have rather fine punctures; S4 and S5 of the male have long curved coppery to silvery subapical hairs; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length and in contact with two large patches of pale tomentum (one on each side [the two are parallel to each other]) throughout its length. Epeolus gibbsi most closely resembles E. ilicis and E. inornatus, but in males of the latter S4 and S5 have short straight subapical hairs and in both E. ilicis and E. inornatus the mandible is simple, and in females of both species F2 is more than 1.2 × as long as wide and the pseudopygidial area is not in contact with two large patches of pale tomentum (one on each side) throughout its length (in contact only at apex, diverging basally).
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, and legs. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex (difficult to see in holotype; described from paratype). Antenna dark brown except scape and F1 reddish brown in part. Pronotal lobe dark brown to black. Tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron densely hairy, except for two sparsely hairy circular patches (one behind pronotal lobe, a larger one occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum uninterrupted except for median bare patch in posterior half, uniformly off white. T1 with median elliptical verging on semicircular discal patch. T1 and T2 with apical fasciae interrupted medially, those of T2 and T3 somewhat broader laterally, T2 with fascia with anterolateral extensions of sparser tomentum. T3 and T4 with fasciae complete. T5 with two large patches of pale tomentum parallel to and contacting pseudopygidial area throughout its length. T5 with pseudopygidial area campanulate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD.
Structure. Preapical tooth blunt and obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 not noticeably longer than wide (L/W ratio = 1.1). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1 MOD at its terminal (difficult to see in holotype; described from paratype). Mesoscutellum strongly bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and tip attaining midlength of mesoscutellum; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, as long as wide (L/W ratio = 1.0); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered.
Etymology. This species is named after its discoverer, Prof. Jason Gibbs, who collected the specimen herein designated as the holotype, recognized it as an unusual find, and brought his discovery to my attention.
Distribution. Upper midwest and adjacent Canada (Fig. 52). Ecology. HOST RECORDS: The holotype of E. gibbsi was collected in an area where Colletes brevicornis and C. kincaidii were in abundance, the latter of which is likely associated with E. minimus, which was also present at the site, as was E. ainsliei and its tentative host C. susannae (J. Gibbs, personal communication, 2017).
FLORAL RECORDS: Unknown. Discussion. What Romankova (2004) identified as E. ilicis, which constituted a new record of that species in Canada, might actually be E. gibbsi and/or E. inornatus. Unfortunately, the vouchered material from that study (three specimens from Ontario) cannot be traced, so the presence of E. ilicis in Canada has not been confirmed in the present study. Epeolus ilicis has been reported from the New England states, though the only examined specimen from that region (a male from Massachusetts) that has been identified as E. ilicis (by Richard L. Brumley) appears to actually be E. inornatus based on the very short straight subapical hairs on S4 and S5. In Canada, E. gibbsi is only confirmed from southern Manitoba, so the specimens from southern Ontario studied by Romankova could represent any of the three species. The key presented in Onuferko (2017)  Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. glabratus apart from all other North American Epeolus except E. lectoides: the axilla is elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin, and the free portion is distinctly hooked; the mesopleuron has sparser punctures ventrolaterally (most i>1d) than in upper half, with the interspaces shining; the metasomal terga have minute, shallow punctures; T2-T4 are medially bare; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length. Whereas in E. lectoides the pronotal collar is black, as are sometimes the axilla and mesoscutellum, and the metasomal terga are black and fasciate, in E. glabratus the pronotal collar, axilla, mesoscutellum, and discs of T1 and T2 are ferruginous and the pale pubescence on the metasomal terga are commonly reduced to discrete lateral patches.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron sparsely hairy, but tomentum dense ventrally as well as between two sparsely hairy patches (one beneath base of fore wing (hypoepimeral area), a larger circular patch occupying much of ventrolateral half of mesopleuron). Metanotum with tomentum uninterrupted except for median bare patch in posterior half, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally. T1 with basal and apical fasciae and T2-T4 with apical fasciae widely separated medially, the apical fasciae reduced to pairs of small patches somewhat broader laterally; T2 with fascia without anterolateral extensions of tomentum, although few sparsely scattered pale hairs present. Remaining metasomal terga mostly hidden in holotype, but T5 and T6 with complete or narrowly interrupted fasciae in non-type specimens. S4 and S5 with long coppery to silvery subapical hairs.
Structure. Preapical tooth blunt and obtuse. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.3). Preoccipital ridge not joining hypostomal carina, from which it is separated by less than 1 MOD at its terminal (difficult to see in holotype; described from non-type specimens). Mesoscutellum moderately bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.6) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin relatively straight and with tip carinate. Fore wing with three submarginal cells. Pygidial plate mostly hidden in holotype, but apically rounded, with large deep punctures more or less evenly spaced throughout with the interspaces shining in nontype specimens.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 even longer than wide (L/W ratio = 1.5); T5 with two large patches of pale tomentum lateral to and separate from pseudopygidial area present only in female; T5 with pseudopygidial area campanulate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum; S4 and S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD); pygidial plate apically truncate, with small, denser punctures.
Distribution. Florida and coastal Georgia (Fig. 54). Ecology. HOST RECORDS: The host species of E. glabratus is/are presently unknown. Discussion. Sequenced specimens of E. glabratus share the same BIN as those of E. lectoides. There is virtually no divergence (<1%) between the barcode sequences of the two species, but the morphological differences are pronounced. Structurally, E.
glabratus and E. lectoides are identical, but in E. glabratus the pronotal collar, axilla, mesoscutellum, and discs of T1 and T2 are ferruginous, whereas in E. lectoides at least the pronotal collar and metasomal terga are entirely black. Epeolus glabratus appears to be restricted to Florida and parts of Georgia, and the prevalence of red integument coloration among Florida Hymenoptera is a well-known unexplained phenomenon (Deyrup and Eisner 2003). Except in some examined specimens from Georgia, in E. glabratus the metasomal fasciae are lacking; the pale pubescence is instead reduced to discrete lateral patches. By contrast, in E. lectoides the metasomal terga are always fasciate. Although both species inhabit Florida, E. glabratus (with red coloration and reduced pubescence on the metasomal terga) appears to be present only on the peninsula whereas E. lectoides (with fasciae and black metasomal terga) appears to be restricted to the Florida panhandle. Since the marked abundance of red coloration is coupled with a general loss of pubescence in E. glabratus, and since these are features restricted to specimens from a particular geographical region, I have opted to treat E. glabratus and E. lectoides as heterospecific, despite the lack of evidence of genetic divergence.

Diagnosis.
The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. howardi apart from all other North American Epeolus: the axilla is large, with the tip extending as far back as or beyond the posterior margin of the mesoscutellum, dilated laterally, and like the mesoscutellum ferruginous; the mesopleuron is closely (i≤1d) and evenly punctate; the metasomal terga are black; T1 has a distinct, although sometimes medially-interrupted, basal fascia; the mesoscutum and metasomal terga have bands of bright or pale yellow short appressed setae; at least the T1-T3 apical fasciae are distinctly interrupted medially; and the pseudopygidial area of the female is lunate with the apex <2 × the medial length. Epeolus howardi most closely resembles E. andriyi and E. floridensis, but in E. andriyi the axillae are shorter, not extending as far back as the posterior margin of the mesoscutellum, and in E. floridensis the mesoscutum and metasomal terga have bands of pale gray to white short appressed setae and T1 is (with few exceptions) ferruginous. Epeolus howardi is also similar to E. scutellaris, but in E. scutellaris the T1-T3 apical fasciae are complete or only very narrowly interrupted medially, and the pseudopygidial area of the female is lunate with the apex >2 × the medial length.
Integument coloration. Black in part, at least partially ferruginous on mandible, labrum, clypeus, antenna, pronotal collar, pronotal lobe, tegula, axilla, mesoscutum, mesoscutellum, metanotum, mesopleuron, legs, T1, pygidial plate, and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex. Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Mesoscutum reddish brown along lateral margin and with pair of reddish-brown markings near posterior margin between midline and parapsidal line. Wing membrane dusky subhyaline, slightly darker at apex. Legs more extensively reddish orange than brown or black. T1 dark in general, not contrasting strongly with remaining metasomal terga, but reddish brown laterally.
Pubescence. Face with tomentum densest around antennal socket. Clypeus, upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron sparsely hairy, but tomentum moderately dense along margins. Metanotum with tomentum uninterrupted, uniformly off white. T1 with dis-cal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally by few sparsely scattered pale hairs. T1-T4 with apical fasciae interrupted medially and narrowed before becoming somewhat broader laterally, T2 with fascia without anterolateral extensions of tomentum. T5 with two patches of pale tomentum lateral to and contacting pseudopygidial area. T5 with pseudopygidial area lunate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Preapical tooth inconspicuous, blunt and obtuse. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.7). Preoccipital ridge not joining hypostomal carina, from which it is separated by less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.7) and tip extending beyond apex of horizontal dorsal portion of mesoscutellum; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, but still longer than wide (L/W ratio = 1.3); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures more or less evenly spaced throughout, with the interspaces shining.

Epeolus ilicis
Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. ilicis apart from all other North American Epeolus except E. erigeronis and E. inornatus: the mandible is simple; the axilla does not attain the midlength of the mesoscutellum but the free portion is distinctly hooked, with the tip unattached to the mesoscutellum for more than 1/3 of the entire medial length of the axilla; the pronotal collar and metasomal terga are black; the metasomal terga have rather fine punctures; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length and not in contact with two large patches of pale tomentum (one on each side) throughout its length (in contact only at apex, diverging basally). Epeolus ilicis is most similar to E. inornatus, and in both species the mesopleuron has punctures that are similar in size and shiny interspaces that are commonly equal to the puncture diameters. By contrast, in E. erigeronis the punctures are more variable in size, with many smaller punctures among large ones, and most interspaces are narrower such that the surface appears to be very coarsely and densely rugose-punctate. Whereas in E. inornatus the legs (and sometimes the pronotal lobe and tegula) are usually darker, at least from the metacoxa to metatibia, the dorsum of the mesosoma and metasoma have gray short appressed setae, and S4 and S5 of the male have short straight subapical hairs, in E. ilicis the pronotal lobe and legs are more extensively reddish orange than brown or black (at least the anterior surface of the metatibia and metatarsus are the same reddish orange color), the dorsum of the mesosoma and metasoma have gray but also usually some pale yellow short appressed setae, and S4 and S5 of the male have long curved coppery to silvery subapical hairs. Epeolus ilicis is also similar to E. gibbsi, but in E. gibbsi the mandible has a blunt, obtuse preapical tooth; in females F2 is less than 1.2 × as long as wide (it is more than 1.2 × as long as wide in female E. ilicis); and the pseudopygidial area of the female is in contact with two large patches of pale tomentum (one on each side [the two are parallel to each other]) throughout its length.
Redescription. This species was recently redescribed (Onuferko 2017). Distribution. Southeastern United States (Fig. 58). Ecology. HOST RECORDS: Rozen (1989) described first instar E. ilicis based on two larvae recovered from the nest of Colletes brimleyi Mitchell on St. Catherines Island in Georgia, USA, from where conspecifics of the former have been recorded (see Material studied).
FLORAL RECORDS: Onuferko (2017) lists associations with five plant genera based on  and a record on Discover Life Pickering 2017 [then 2016]). Since the discovery of E. inornatus, a cryptic species very similar to E. ilicis whose name applies to at least one of Mitchell's paratypes of E. ilicis (see Material studied under E. inornatus), my taxon concept of E. ilicis has changed. As a result, I have only been able to determine that records of Ilex glabra and Prunus angustifolia Marshall (Rosaceae), taken from the collection labels of the holotypes of E. ilicis and E. weemsi respectively, are associated with what is here understood to be the true E. ilicis. Discussion. Both the holotype of E. ilicis and the holotype of E. vernalis were examined, and the two appear to be the same species. In  key, the two species were differentiated on the basis of whether or not (and if so to what degree) the metasomal fasciae are interrupted medially, but the T1-T3 apical fasciae are interrupted medially (those of T1 and T2 are somewhat more widely separated medially) in both holotype specimens and the T4 fascia is complete in the E. ilicis holotype and only very narrowly interrupted in the E. vernalis holotype. Moreover, the type locality is the same for both (Holly Shelter [Pender County], North Carolina, USA), and the two specimens were collected only 12 days apart.
Presently, only a single 422 bp sequence is available for E. ilicis (a male specimen from Florida, USA), which clusters with sequences of E. zonatus (Suppl. material 2), and all were assigned the same BIN. The Florida specimen is most similar to the holotype of E. weemsi, which  described before noting that it might be the male of E. vernalis. In both the sequenced specimen and E. weemsi holotype, S4 and S5 have long curved coppery to silvery subapical hairs, which are absent in the very similar E. inornatus but present in all other North American male Epeolus. Whereas I have opted to treat E. ilicis and E. zonatus as heterospecific based on remarkably consistent differences in integument coloration coupled with a general loss of pubescence in E. zonatus, despite the apparent lack of evidence of genetic divergence, the extremely subtle differences in integument coloration and pubescence among the holotypes of E. ilicis, E. vernalis, and E. weemsi seem to fall within the range of intraspecific variation, and therefore E. vernalis and E. weemsi are herein synonymized under E. ilicis. Although the three names were published simultaneously, priority of the name should be given to E. ilicis because the holotype is in the best condition (those of E. vernalis and E. weemsi have broken antennae and in the latter much of the pubescence is discolored or rubbed off), it is female and most Epeolus spp. were described from female name-bearing types (the holotype of E. weemsi is male), and because an allotype and paratypes were designated for E. ilicis but not E. vernalis or E. weemsi. This species appears to be quite common in the Southeastern United States, where it may be confused with E. erigeronis or E. inornatus.

Epeolus inornatus sp. n.
http://zoobank.org/AFC50A58-43E8-4BC2-A71B-0F85C431B390 Figs 59, 60, 92G, 93C, 96D, 100B Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. inornatus apart from all other North American Epeolus except E. erigeronis and E. ilicis: the mandible is simple; the axilla does not attain the midlength of the mesoscutellum but the free portion is distinctly hooked, with the tip unattached to the mesoscutellum for more than 1/3 of the entire medial length of the axilla; the pronotal collar and metasomal terga are black; the metasomal terga have rather fine punctures; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length and not in contact with two large patches of pale tomentum (one on each side) throughout its length (in contact only at apex, diverging basally). Epeolus inornatus is most similar to E. ilicis, and in both species the mesopleuron has punctures that are similar in size and shiny interspaces that are commonly equal to the puncture diameters. By contrast, in E. erigeronis the punctures are more variable in size, with many smaller punctures among large ones, and most interspaces are narrower such that the surface appears to be very coarsely and densely rugose-punctate. Whereas in E. ilicis the pronotal lobe and legs are more extensively reddish orange than brown or black (at least the anterior surface of the metatibia and metatarsus are the same reddish orange color), the dorsum of the mesosoma and metasoma have gray but also usually some pale yellow short appressed setae, and S4 and S5 of the male have long curved coppery to silvery subapical hairs, in E. inornatus the legs (and sometimes the pronotal lobe and tegula) are usually darker, at least from the metacoxa to metatibia, the dorsum of the mesosoma and metasoma have gray short appressed setae, and S4 and S5 of the male have short straight subapical hairs. Epeolus inornatus is also similar to E. gibbsi, but in E. gibbsi the mandible has a blunt, obtuse preapical tooth; in males S4 and S5 have long curved coppery to silvery subapical hairs, as in E. ilicis and all other Nearctic Epeolus; in females F2 is less than 1.2 × as long as wide (it is more than 1.2 × as long as wide in female E. inornatus); and the pseudopygidial area of the female is in contact with two large patches of pale tomentum (one on each side [the two are parallel to each other]) throughout its length.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, and legs. Mandible with apex darker than all but extreme base. Antenna dark brown except F1 reddish brown in part. Pronotal lobe dark brown to black. Tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs with brown or black more extensive than reddish orange.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale gray short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half hairy, ventrolateral half nearly bare. Metanotum with tomentum uninterrupted except for median bare patch in posterior half, uniformly off white. T1 with median quadrangular black discal patch enclosed by pale tomentum, except for medial separa-tion at apex. T2 with fascia interrupted medially and with faint anterolateral extensions of sparser tomentum. T3 and T4 with fasciae complete. T5 with two large patches of pale tomentum lateral to and contacting pseudopygidial area at apex, diverging from pseudopygidial area basally. T5 with pseudopygidial area campanulate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by more than 1/4 MOD.
Structure. Mandible without preapical tooth. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and tip not extending beyond midlength of mesoscutellum; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin relatively straight and carinate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, not noticeably longer than wide (L/W ratio = 1.1); pygidial plate apically rounded, with large deep punctures closely clustered.
Etymology. The name is in reference to the grayish pubescence and largely monochromatic dark brown or black integument of this species. From the Latin, "inornatus" (unadorned).
Distribution. Mid-Atlantic states to Texas (Fig. 60). Ecology. HOST RECORDS: The host species of E. inornatus is/are presently unknown. FLORAL RECORDS: Labels of examined voucher specimens indicate floral associations with Quercus laevis Walter (Fagaceae) and Vaccinium arboreum Marshall.
Discussion. The specimens from Texas, USA that Brumley (1965) identified as E. ilicis are probably E. inornatus. Although BIN-compliant sequences are presently not available for E. inornatus, a single 421 bp sequence is available for a female specimen (the holotype) from East Texas, which does not cluster with the single sequence (422 bp in length) available for what is herein considered to be the true E. ilicis (a male specimen from Florida, USA) based on its greater resemblance to the holotype of that species (Suppl. material 2). Instead, the sequence from the Florida specimen clusters with sequences of E. zonatus, which is a visibly different bee, and all were assigned the same BIN. Whereas male E. inornatus are unique among Epeolus in having very short straight subapical hairs on S4 and S5 instead of the usual long curved coppery to silvery subapical hairs, females are practically indistinguishable from E. ilicis in terms of surface sculpture and structure. Although consistent, the features (differences in integument coloration and pubescence) that in combination may be used to distinguish female E. inornatus from E. ilicis are subtle. Based on known records, adults of E. inornatus appear to be most active in spring, the same time of year when adults of E. ilicis and E. zonatus are active.
Material studied. Type material.

Diagnosis.
Unique to E. interruptus among North American species of Epeolus are each of the following morphological features: the metanotum has a blunt median process and T1 has a wide triangular discal patch with concave lateral sides. Epeolus interruptus most closely resembles E. tessieris in that the mesoscutum has short paramedian bands; the axilla does not attain the midlength of the mesoscutellum and like the mesoscutellum is ferruginous (although both are occasionally black in E. interruptus); the mesopleuron commonly has sparser punctures ventrolaterally than in upper half, with the interspaces shining; and T1-T4 have medially-interrupted metasomal fasciae. However, in E. tessieris the metanotum is flat and T1 has a trapezoidal to nearly semicircular discal patch.
Redescription. This species was recently redescribed (Onuferko 2017). Distribution. Central and western Canada, east of the Rocky Mountains, to northern Mexico (Fig. 62).
Ecology. See Onuferko (2017) for host and floral records. Floral associations are also indicated in Suppl. material 1, which includes a newly discovered association with Heterotheca villosa (Pursh) Shinners based on the label of one examined voucher specimen.
Discussion. Detailed morphological and taxonomic remarks about this species are given in Onuferko (2017).

Diagnosis.
The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. lectoides apart from all other North American Epeolus except E. glabratus: the axilla is elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin, and the free portion is distinctly hooked; the mesopleuron has sparser punctures ventrolaterally (most i>1d) than in upper half, with the interspaces shining; the metasomal terga have minute, shallow punctures; the T2-T4 fasciae are conspicuously narrowed or interrupted medially; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length. Whereas in E. glabratus the pronotal collar, axilla, mesoscutellum, and discs of T1 and T2 are ferruginous and the pale pubescence on the metasomal terga are commonly reduced to discrete lateral patches, in E. lectoides the pronotal collar is black, as are sometimes the axilla and mesoscutellum, and the metasomal terga are black and fasciate. Epeolus lectoides is also similar to E. lectus, but in E. lectus the metasomal terga have coarse, deep punctures and the T2-T4 fasciae are complete and evenly broad. Redescription. This species was recently redescribed (Onuferko 2017). Distribution. Eastern North America (Fig. 64).  Diagnosis. The following morphological features in combination can be used to tell E. lectus apart from all other North American Epeolus: the mesopleuron has sparser punctures ventrolaterally (most i>1d) than in upper half, with the interspaces shining; the metasomal terga have coarse, deep punctures; and T2-T4 have complete and evenly broad fasciae. Epeolus lectus is most similar to E. lectoides, and in both species the free portion of the axilla is distinctly hooked and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length, but in E. lectoides the metasomal terga have minute, shallow punctures and the T2-T4 fasciae are conspicuously narrowed or interrupted medially.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, tegula, axilla, mesoscutellum, legs, and metasomal sterna. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex (difficult to see in the E. lectus holotype; described from non-type specimens). Flagellum brown and (except F1) slightly lighter than partially dark brown (otherwise orange) scape and F1 and entirely dark brown pedicel, primarily due to extensive pilosity on flagellum. F2 with orange spot basally. Wing membrane dusky subhyaline, slightly darker at apex. Legs from trochanter to tarsus extensively reddish orange, coxae brown.
Pubescence. Face with tomentum densest around antennal socket. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half sparsely hairy, ventrolateral half nearly bare. Metanotum with tomentum sparser medially, uniformly off white. T1 with discal patch elliptical and very wide, the basal and apical fasciae only narrowly joined laterally. T1 with basal and apical fasciae and T2-T3 with apical fasciae complete (T4 entirely retracted in the E. lectus holotype, but with complete fascia in non-type specimens), T2 with fascia with faint anterolateral extensions of sparser tomentum. T5 with two large patches of pale tomentum lateral to and contacting pseudopygidial area at apex. T5 with pseudopygidial area campanulate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by 1/3 MOD.
Structure. Preapical tooth blunt and obtuse. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.5). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal (difficult to see in the E. lectus holotype; described from non-type specimens). Mesoscutellum moderately bigibbous. Axilla large, its lateral margin (L) half as long as mesoscutellar width (W) (L/W ratio = 0.5) and tip not extending much beyond midlength of mesoscutellum (extending to <2/3 its length in the E. lectus holotype and all examined non-type specimens; extending to ~2/3 its length in the E. agnathus holotype); axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion approximately half its medial length; axilla with lateral margin relatively straight and carinate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, but still longer than wide (L/W ratio = 1.2); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered medially and sparser laterally, with the interspaces shining.
Distribution. Great Plains and Mountain states east of the Continental Divide (Fig. 66). Ecology. HOST RECORDS: In late July 2015, I collected several specimens of this species near the Poudre River in the Roosevelt National Forest, Colorado, USA, where large numbers of Colletes females were collected and observed foraging on purple Dalea flowers. Using Stephen's (1954) key, collected specimens were identified as being either C. robertsonii Dalla Torre or C. timberlakei Stephen, the females of which cannot be reliably distinguished morphologically, although the short triangular mesosomal spines and fine punctation on the tegulae of examined specimens coupled with their collection locality suggest they are C. timberlakei.
FLORAL RECORDS: The label of one examined voucher specimen indicates a floral association with Cryptantha cinerea var. jamesii (Torr.) Cronquist (Boraginaceae).
Discussion. The names Epeolus agnathus and E. lectus were published simultaneously, although  remarked that E. agnathus may be the male of E. lectus as the two specimens are structurally similar.  synonymized E. agnathus under E. lectus, and separated both specimens from E. lectoides based on differences in metasomal pubescence and punctation (see diagnosis). I have examined the holotype specimens of E. lectus and E. agnathus, and agree with Robertson's treatment. Although  did not provide any justification for selecting the name E. lectus over E. agnathus, the holotype of the former is in better condition (that of E. agnathus is missing an antenna) and is female, the sex upon which most Epeolus species descriptions have been based. While Cresson's Epeolus types include remarkably little collection data, the type locality of E. agnathus (Dakota Territory) is even more vague than that of E. lectus (Kansas).
In contrast to the similar and presumably closely related E. lectoides, E. lectus has a much more restricted range and is rare in collections. Both species are known from the Great Plains, although the range of E. lectus extends further west. In E. lectus, the metasoma has much coarser punctures than that of any other North American species in the genus, including E. lectoides, in which the metasoma has much finer and sparser punctures. In addition to this and other clear morphological differences (see diagnosis), the distinction between E. lectus and E. lectoides is supported by separate BINs for the two species.

Diagnosis.
The following morphological features in combination can be used to tell E. mesillae apart from all other North American Epeolus: the axilla does not attain the midlength of the mesoscutellum and like the mesoscutellum is black, the fore wing has two submarginal cells, and T1-T4 have complete fasciae. Only in E. americanus and E. asperatus is the fore wing commonly with two submarginal cells, but in both species at least the T1 and T2 apical fasciae are interrupted or at least greatly narrowed medially. Epeolus brumleyi is similar to E. mesillae in axillar structure; in that in females F2 is shorter, as long as wide; and in that T1-T4 have complete fasciae. However, in E. brumleyi the axilla is commonly ferruginous in part and the fore wing has three submarginal cells.
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, antenna, pronotal lobe, tegula, and legs. Mandible orange between dark brown base and reddish-brown apex; preapical tooth slightly lighter than mandibular apex (difficult to see in the P. mesillae neotype because mandible closed; described from non-type specimens). Flagellum brown, except F1 extensively orange, and slightly lighter than dark brown scape and pedicel. Pronotal lobe reddish brown. Tegula pale ferruginous to amber. Wing membrane hyaline throughout. Legs, except tarsi, with brown or black more extensive than reddish orange.
Pubescence. Face with tomentum densest on clypeus and around antennal socket, sparser on upper paraocular area and vertexal area. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band partly obscured by surrounding pale tomentum. Mesopleuron almost entirely obscured by white tomentum, except where rubbed off in the P. mesillae neotype. Metanotum with tomentum uninterrupted, uniformly off white. T1 with discal patch elliptical, narrow, and short. T2-T6 each with complete fascia, those of T2 and T3 somewhat broader laterally, T2 with fascia with anterolateral extensions of sparser tomentum. S3-S5 with long coppery to silvery subapical hairs.
Structure. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.7 × greatest width. F2 nearly as long as wide (L/W ratio = 0.9). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5-2 MOD at its terminal (difficult to see in the P. mesillae neotype; described from non-type specimens). Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.3) and tip not extending beyond midlength of mesoscutellum; axilla with tip visible, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with two submarginal cells. Pygidial plate apically rounded, with large deep punctures closely clustered.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 slightly longer, as long as wide (L/W ratio = 1.0); wing membrane subhyaline, apically dusky; T5 with large, continuous patch of pale tomentum bordering and separate from pseudopygidial area present only in female; T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum; S3-S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD); pygidial plate apically truncate, with small, denser punctures.
Distribution. Known to occur in all major North American deserts (Fig. 68). Ecology. HOST RECORDS: Colletes clypeonitens Swenk is the presumed host of E. mesillae (Hurd and Linsley 1975). Personal observations support such an association. In Whitewater Discussion. Epeolus mesillae was originally described under the now defunct genus Phileremus because the fore wing in this species has two rather than three submarginal cells, the typical state for most Epeolus species. Among North American Epeolus, E. mesillae exhibits unusual sexual dimorphism in that in females the fore wing and (to a lesser extent) hind wing are apically dusky whereas in males the wings are hyaline throughout. There is some variability in the pubescence on the metasomal terga among specimens, with some exhibiting more grayish-white than yellowish fasciae. Linsley (1939) recognized specimens from southern California as a distinct subspecies (E. mesillae palmarum) based on a larger body size and the presence of pale tomentum interspersed with darker tomentum on the discs of the metasomal terga, especially laterally. Specimens from across the range of this species exhibiting these features have been examined, as well as specimens from southern California in which the metasomal fasciae are clearly distinct from the all-dark discs. Specimens from near the type locality of E. mesillae palmarum were barcoded, and their sequences cluster closely with those from specimens from Southeast Arizona and adjacent Sonora, nearer the type locality (Las Cruces, New Mexico) of E. mesillae mesillae. Hence, I do not consider these to be distinct subspecies, and herein synonymize E. mesillae palmarum under E. mesillae, a change in taxonomic status first proposed by Brumley (1965).
I have not seen the male holotype of P. mesillae and do not know where it is housed, despite personally searching through the entomological collections where T.D. Cockerell deposited the types of other Epeolus species he described. In Brumley (1965), no reference was made to Cockerell's holotype of P. mesillae, suggesting Brumley too was unable to find it. Moreover, no references in the literature to Cockerell's type since the species' original description could be found. In the same publication, another species was described under Phileremus -P. verbesinae (now Neolarra verbesinae (Cockerell)) -, which was redescribed by Michener (1939) who indicated that the type was in the T.D.A. Cockerell Collection. It is unclear if either specimen has since ended up in an institution that maintains a research collection, but that the holotype of E. mesillae has not been referenced since its original description strongly suggests it is unlikely to turn up in the future and to all intents and purposes has been lost. In my search for the holotype at the CUM, a male specimen of E. mesillae (labelled as Phileremus mesillae Ckll.) from Mesilla Park (the original type locality) collected by Cockerell from Dimorphocarpa wislizeni on May 7 th was discovered. The specimen, which is the property of the CUM, agrees with the original description, and was used to write the present redescription and diagnosis. Given that a synonymy under E. mesillae is proposed herein, it is sensible to have a neotype to serve as a point of reference for any future comparisons. Aside from the collection date, the specimen selected as the neotype of Phileremus mesillae fits the description of the original, which can no longer be traced. Hence, in this particular case the qualifying conditions for designating a neotype as listed under

Diagnosis.
The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. minimus apart from all other North American Epeolus except E. banksi and E. olympiellus: in females, F2 is at least 1.2 × as long as wide; the mesoscutum has distinct, evenly broad paramedian bands that may be joined posteriorly; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is more than 1/4 as long as the entire medial length of the axilla, and the axilla (except sometimes the tip) and mesoscutellum are black; the mesopleuron is closely (most i<1d) and evenly punctate; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least half as wide as the breadth of the apical fascia; and the T2 fascia has lobe-like anterolateral extensions of tomentum. Whereas in E. banksi the mesoscutum and metasomal terga have bands of gray short appressed setae, in E. minimus the mesoscutum and metasomal terga have bands of off-white to pale yellow short appressed setae. In this respect, E. minimus more closely resembles E. olympiellus, but in E. olympiellus the T3 and T4 fasciae are broken or at least narrowed laterally, as well as medially, whereas in E. minimus the T3 and T4 fasciae are not broken laterally, and are complete or narrowly interrupted medially. Epeolus minimus is also similar to E. axillaris, but in E. axillaris the metanotum has a distinct posteromedial depression (as opposed to being flat) and the axilla is more elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin. Redescription. This species was recently redescribed (Onuferko 2017). Distribution. Widely distributed across Canada and the United States, although apparently more common in the west; not known to occur in parts of northeastern North America or the high arctic (Fig. 70). Also, the single (perhaps mislabelled) examined specimen from Florida is an extreme outlier, and given the lack of other examined material from the Southern United States the record should be treated with some skepticism.
Ecology. HOST RECORDS: Graenicher (1906) associated E. minimus (as A. minima) with C. eulophi Robertson based on detailed observations of a female of the former inspecting and entering the nest of a female of the latter in Lake Woods, Wisconsin, USA. However, according to Stephen (1954) Graenicher's record of C. eulophi in Wisconsin is based on observations of C. kincaidii. Epeolus minimus has been collected with C. kincaidii in Birds Hill Provincial Park and Spruce Woods Provincial Park, Manitoba, Canada where no C. eulophi were collected or observed (J. Gibbs, personal communication, 2017), so the association between E. minimus and C. kincaidii seems likely.
FLORAL RECORDS: See Onuferko (2017). Floral associations are also indicated in Suppl. material 1, which includes newly discovered associations with Ericameria nauseosa var. nauseosa and Medicago L. (Leguminosae) based on labels of examined voucher specimens.
Discussion. In Onuferko (2017), E. minimus is said to be similar to a Californian species yet to be formally recognized, which herein is formally described under the name Epeolus axillaris. Detailed morphological and taxonomic remarks about this species are given in Onuferko (2017). Epeolus minimus is among the most widespread and commonly collected Epeolus species in North America.  Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. nebulosus apart from all other North American Epeolus except E. basili, E. novomexicanus, and E. pusillus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum but at most to the band of pale tomentum along its posterior margin, dilated laterally, and ferruginous to some degree whereas the mesoscutellum is typically all black; the axilla's free portion is clearly less than 2/5 as long as its entire medial length; the mesopleuron is closely (most i<1d) and evenly punctate, that of the female is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ) whereas that of the male (excluding the hypoepimeral area) is entirely obscured by white tomentum; T2-T4 have complete and evenly broad fasciae; the T2 fascia has lobe-like anterolateral extensions of tomentum; and the pseudopygidial area of the female is lunate and wider than long (the apex ≤2 × the medial length). Epeolus basili, E. nebulosus, E. novomexicanus, and E. pusillus are all extremely similar to one another. Epeolus nebulosus is most similar to E. novomexicanus, but in E. novomexicanus the mesoscutum usually has distinct paramedian bands and at least the integument beneath the T1 apical fascia is ferruginous, as are sometimes the rest of the tergum and other terga, whereas in E. nebulosus the mesoscutum is entirely obscured by pale tomentum and the metasomal terga (excluding the brown translucent apical margins) are entirely black. In E. basili the metasomal terga are also ferruginous to some degree, and the T2 and T3 (for female) or T2-T4 (for male) fasciae are narrowed medially and removed from the apical margin (in E. nebulosus the T2-T4 fasciae are on or very little removed from the apical margin), and the pseudopygidial area of the female is ≥2 × the medial length. Whereas in E. pusillus the flagellum, except sometimes F1, and metasomal sterna are consistently brown or black and clearly not the same reddish-orange color as the legs (tibiae to tarsi), in E. nebulosus the flagellum, at least ventrally, is the same reddish-orange color as the legs (tibiae to tarsi) as are usually the metasomal sterna. Epeolus nebulosus is also similar to E. scutellaris in that the axilla is large, with the lateral margin arcuate, and that the apical fasciae are complete. However, in E. scutellaris the pseudopygidial area of the female is much wider (the apex ~2.5-3 × the medial length) than in E. nebulosus, and the mesopleuron of both the female and male is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ).
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, antenna, pronotal lobe, tegula, axilla, legs, pygidial plate, and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex. Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black. S1-S6 reddish orange.
Pubescence. Face with tomentum densest on clypeus and around antennal socket, slightly sparser on upper paraocular area and vertexal area. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum largely obscured by pale tomentum. Mesopleuron (excluding hypoepimeral area) entirely obscured by white tomentum. Metanotum with tomentum uninterrupted, uniformly off white. T1 with narrow and short discal patch largely obscured by pale tomentum. T2-T6 each with complete fascia, T2 with fascia with wide basomedially convergent anterolateral extensions of tomentum. S4 and S5 with long coppery to silvery subapical hairs, which individually are often darker apically.
Structure. Preapical tooth obtuse. Labrum with pair of small subapical denticles not preceded by carinae (difficult to see in holotype; described from paratypes). Frontal keel not strongly raised. Scape with greatest length 2.0 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.2). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) half as long as mesoscutellar width (W) (L/W ratio = 0.5) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically rounded, with large deep, well-separated punctures, with the interspaces shining.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 even longer than wide (L/W ratio = 1.5); mesopleuron densely hairy, except for two almost entirely bare patches (one beneath base of fore wing (hypoepimeral area), a larger circular patch occupying much of ventrolateral half of mesopleuron); T5 with large, continuous patch of pale tomentum bordering and contacting pseudopygidial area present only in female; T5 with pseudopygidial area lunate, its apex twice as wide as medial length, indicated by silvery setae on disc of apicomedial region elevated from rest of tergum; S4 and S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by much more than 1/4 MOD); pygidial plate apically truncate, with small, denser punctures.
Etymology. The name is in reference to the pale tomentum obscuring much of the integument of this species. From the Latin, "nebulosus" (hazy).
Ecology. HOST RECORDS: The host species of E. nebulosus is/are presently unknown. FLORAL RECORDS: Labels of examined voucher specimens indicate a floral association with Ericameria nauseosa.
Discussion. Epeolus nebulosus is a cryptic species within the "pusillus group" that closely resembles some specimens of E. novomexicanus, and the ranges of the two spe- cies overlap to some extent. The morphological differences (in integument coloration and patterns of pubescence) among the four members of the "pusillus group" are subtle. The status of E. nebulosus as a separate species is further supported by a separate BIN and large barcode sequence divergence (>3.2%) from its nearest neighbor, E. novomexicanus. Although most species of Epeolus were described from a female namebearing type, a male specimen is designated as the holotype of E. nebulosus because a barcode-compliant sequence is associated with it and because much of the pubescence is discolored or rubbed off in the available female specimen, which is herein designated as the allotype. Since this species is described from very few specimens, efforts should be made to collect additional representatives of E. nebulosus for DNA barcoding to determine if the morphological differences between it and E. novomexicanus reported here are consistent.

Epeolus novomexicanus
Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. novomexicanus apart from all other North American Epeolus except E. basili, E. nebulosus, and E. pusillus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum but at most to the band of pale tomentum along its posterior margin, dilated laterally, and ferruginous to some degree whereas the mesoscutellum is typically all black; the axilla's free portion is clearly less than 2/5 as long as its entire medial length; the mesopleuron is closely (most i<1d) and evenly punctate, that of the female is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ) whereas that of the male (excluding the hypoepimeral area) is entirely obscured by white tomentum; T2-T4 have complete and evenly broad fasciae; the T2 fascia has lobe-like anterolateral extensions of tomentum; and the pseudopygidial area of the female is lunate and wider than long (the apex ≤2 × the medial length). Epeolus basili, E. nebulosus, E. novomexicanus, and E. pusillus are all extremely similar to one another. Epeolus novomexicanus is most similar to E. nebulosus, but in E. nebulosus the mesoscutum is entirely obscured by pale tomentum and the metasomal terga (excluding the brown translucent apical margins) are entirely black whereas in E. novomexicanus the mesoscutum usually has distinct paramedian bands and at least the integument beneath the T1 apical fascia is ferruginous, as are sometimes the rest of the tergum and other terga. In E. basili the metasomal terga are also ferruginous to some degree, but the T2 and T3 (for female) or T2-T4 (for male) fasciae are narrowed medially and removed from the apical margin (in E. novomexicanus the T2-T4 fasciae are on or very little removed from the apical margin), and the pseudopygidial area of the female is ≥2 × the medial length. Whereas in E. pusillus the flagellum, except sometimes F1, and metasomal sterna are consistently brown or black and clearly not the same reddish-orange color as the legs (tibiae to tarsi), in E. novomexicanus the flagellum, at least ventrally, is the same reddish-orange color as the legs (tibiae to tarsi) as are usually the metasomal sterna. Epeolus novomexicanus is also similar to E. scutellaris in that the axilla is large, with the lateral margin arcuate, and that the apical fasciae are complete. However, in E. scutellaris the pseudopygidial area of the female is much wider (the apex ~2.5-3 × the medial length) than in E. novomexicanus, and the mesopleuron of both the female and male is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ).
Integument coloration. Mostly black; notable exceptions as follows: at least partially ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, axilla, legs, metasomal terga (including pygidial plate), and metasomal sterna. Mandible with apex darker than rest of mandible; preapical tooth slightly lighter than mandibular apex (difficult to see in holotype because mandible closed; described from non-type specimens). Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black. S1-S6 reddish orange.
Pubescence. Face with tomentum partly rubbed off in holotype, but white and densest around antennal socket in non-type specimens. Tomentum slightly sparser on clypeus; upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band partly obscured by surrounding pale tomentum. Mesopleuron (excluding hypoepimeral area) entirely obscured by white tomentum (except where rubbed off in holotype). Metanotum with tomentum uninterrupted, uniformly off white. T1 with narrow and short discal patch partly obscured by pale tomentum. T2-T5 each with complete fascia (T6 mostly retracted in holotype, but with complete fascia in non-type specimens), T2 with fascia with wide basomedially convergent anterolateral extensions of tomentum. S4 and S5 with long coppery to silvery subapical hairs, which individually are often darker apically.
Structure. Preapical tooth obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) half as long as mesoscutellar width (W) (L/W ratio = 0.5) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically rounded, with large deep punctures closely clustered.
FEMALE: Description as for male except for usual secondary sexual characters and as follows: F2 noticeably longer than wide (L/W ratio = 1.5); mesopleuron densely hairy, except for two sparsely hairy circular patches (one behind pronotal lobe, a larger one occupying much of ventrolateral half of mesopleuron); T5 with large, continuous patch of pale tomentum bordering and contacting pseudopygidial area present only in female; T5 with pseudopygidial area lunate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum; S4 and S5 with much shorter hairs (S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~1/3 MOD); pygidial plate apically truncate, with small, denser punctures.

Epeolus olympiellus
Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. olympiellus apart from all other North American Epeolus except E. banksi and E. minimus: in females, F2 is at least 1.2 × as long as wide; the mesoscutum has distinct, evenly broad paramedian bands that may be joined posteriorly; the axilla is small to intermediate in size, not extending much beyond the midlength of the mesoscutellum (extending to <2/3 its length) but the free portion is more than 1/4 as long as the entire medial length of the axilla, and the axilla (except sometimes the tip) and mesoscutellum are black; the mesopleuron is closely (most i<1d) and evenly punctate; T1 has a quadrangular discal patch, in dorsal view the longitudinal band is at least half as wide as the breadth of the apical fascia; and the T2 fascia has lobe-like anterolateral extensions of tomentum. Whereas in E. banksi the mesoscutum and metasomal terga have bands of gray short appressed setae, in E. olympiellus the mesoscutum and metasomal terga have bands of off-white to pale yellow short appressed setae. In this respect, E. olympiellus more closely resembles E. minimus, but in E. minimus the T3 and T4 fasciae are not broken laterally, and are complete or narrowly interrupted medially, whereas in E. olympiellus the T3 and T4 fasciae are broken or at least narrowed laterally, as well as medially. Whereas throughout most of its range E. minimus exhibits reddish-orange coloration on the labrum, antenna, pronotal lobe, and/or legs, except foreleg, from trochanters to tarsi, in E. olympiellus the labrum, antenna, and legs from coxae to femora are brown or black. Epeolus olympiellus is also similar to E. axillaris, but in E. axillaris the metanotum has a distinct posteromedial depression (as opposed to being flat) and the axilla is more elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin.
Description. This species was recently redescribed (Onuferko 2017). Distribution. United States west of the Rocky Mountains to southern British Columbia (Fig. 76).
Ecology. See Onuferko (2017) for host and floral records. Floral associations are also indicated in Suppl. material 1.
Discussion. Detailed morphological and taxonomic remarks about this species are given in Onuferko (2017).
Material studied. Type material. Primary: Canada: British Columbia: Nanaimo (Nanaimo Biological Station), 24.vi.1920    predominantly ferruginous; the axilla is large, with the tip extending as far back as or beyond the posterior margin of the mesoscutellum, dilated laterally, and like the mesoscutellum ferruginous; the mesopleuron is closely (most i<1d) and evenly punctate; the metasomal terga have pale but not brownish orange pubescence; and the T1-T3 apical fasciae are interrupted medially and commonly reduced to discrete lateral patches. Epeolus packeri resembles E. andriyi, E. deyrupi, E. floridensis, and E. howardi in that the axilla is large, with the lateral margin arcuate, and like the mesoscutellum ferruginous, and that the T1-T3 apical fasciae are interrupted medially. However, in E. packeri the pseudopygidial area of the female is wider (the apex >2 × the medial length) than in E. andriyi, E. floridensis, or E. howardi (the apex <2 × the medial length), and the T1 basal fascia is absent or reduced to a pair of small patches of pale tomentum whereas in E. andriyi, E. floridensis, and E. howardi T1 has a distinct, although often medially-interrupted, basal fascia. Epeolus packeri closely resembles E. deyrupi, but in E. deyrupi the mesopleuron commonly has sparser punctures ventrolaterally (i≤2d) than that of E. packeri, with the interspaces shining or somewhat dull due to tessellate surface microsculpture, and the T1-T3 apical fasciae are (to varying degrees) brownish orange medially and off white laterally. Epeolus packeri is also similar to E. scutellaris, but in E. scutellaris the pronotal collar is pre-dominantly black and the T1-T3 apical fasciae are complete or only very narrowly interrupted medially.
Pubescence. Face with tomentum densest on paraocular area around antennal socket, otherwise almost entirely bare. Mesoscutum without pale tomentum. Dorsum of metasoma with bands of off-white short appressed setae. Mesopleuron nearly bare, except along margins. Metanotum with tomentum uninterrupted except for median bare patch in posterior half, uniformly off white. T1 and T2 with apical fasciae medially interrupted, narrowed (broader laterally), and removed from apical margin; T2 with fascia without anterolateral extensions of tomentum. Metasoma otherwise without fasciae, although T3 and T4 with few sparsely scattered pale hairs present on apical impressed areas. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on flat disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by 1/3 MOD.
Structure. Preapical tooth blunt and obtuse. Labral apex with pair of small denticles, each preceded by longitudinal carina. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by less than 1 MOD at its terminal (difficult to see in holotype; described from paratype). Mesoscutellum moderately bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.7) and tip extending slightly beyond apex of horizontal dorsal portion of mesoscutellum; axilla with tip clearly visible, but unattached to mesoscutellum for less than 2/5 the medial length of axilla; axilla with lateral margin arcuate. Fore wing with three submarginal cells. Pygidial plate apically truncate. MALE: Description as for female except for usual secondary sexual characters and as follows: face with more abundant pale tomentum, densest from midlength of clypeus to upper paraocular and frontal areas; F2 shorter, but still longer than wide (L/W ratio = 1.2); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered basally and sparser apically, with the interspaces shining.
Etymology. This species is named in honor of my dissertation adviser, Prof. Laurence Packer, who collected the first specimen of this species I have seen.
Ecology. HOST RECORDS: The host species of E. packeri is/are presently unknown. FLORAL RECORDS: Labels of examined voucher specimens indicate a floral association with Solidago.
Discussion.  keys to female and male Epeolus, this species comes out as E. floridensis in which T1 is not bright ferruginous but black. However, in E. floridensis the dorsum of the mesosoma and metasoma has more abundant pale pubescence, and the pseudopygidial area is conspicuously narrower. Moreover, all examined specimens of E. floridensis (adults) were collected in spring whereas all those identified as E. packeri were collected in October.
In terms of surface sculpture, structure, and the width of the pseudopygidial area, E. packeri is most similar to E. scutellaris, and sequenced representatives of both forms share the same BIN. The two are considered to be heterospecific based on the marked abundance of red coloration coupled with a loss of pubescence (the same rationale for treating E. glabratus as distinct from E. lectoides) in E. packeri, features that are common in Florida Hymenoptera and constitute an unexplained regional phenomenon (Deyrup and Eisner 2003 Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. pusillus apart from all other North American Epeolus except E. basili, E. nebulosus, and E. novomexicanus: the axilla is large, with the tip extending well beyond the midlength of the mesoscutellum but at most to the band of pale tomentum along its posterior margin, dilated laterally, and usually ferruginous to some degree (rarely all black) whereas the mesoscutellum is entirely black; the axilla's free portion is clearly less than 2/5 as long as its entire medial length; the mesopleuron is closely (most i<1d) and evenly punctate, that of the female is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ) whereas that of the male (excluding the hypoepimeral area) is entirely obscured by white tomentum; the T1-T3 apical fasciae are complete or only very narrowly interrupted medially; the T2 fascia has lobe-like anterolateral extensions of tomentum; and the pseudopygidial area of the female is lunate and wider than long (the apex ≤2 × the medial length). Epeolus basili, E. nebulosus, E. novomexicanus, and E. pusillus are all extremely similar to one another. Whereas in E. basili the flagellum, at least ventrally, is the same reddish-orange color as the legs (tibiae to tarsi) as are usually the metasomal sterna, in E. pusillus the flagellum, except sometimes F1, and metasomal sterna are consistently brown or black and clearly not the same reddish-orange color as the legs (tibiae to tarsi). Whereas in E. nebulosus and E. novomexicanus the longitudinal extent of the T1 discal patch is less than or equal to the breadth of the apical fascia and the T2-T4 fasciae are on or very little removed from the apical margin and more or less evenly broad, in E. pusillus the longitudinal extent of the T1 discal patch is no less (and usually greater) than the breadth of the apical fascia and the T1-T3 apical fasciae are removed from the apical margin and commonly narrowed or narrowly interrupted medially. Epeolus pusillus is also similar to E. scutellaris in that the axilla is large, with the lateral margin arcuate, and that the apical fasciae are complete or only very narrowly interrupted medially. However, in E. scutellaris the pseudopygidial area of the female is much wider (the apex ~2.5-3 × the medial length) than in E. pusillus, and the mesopleuron of both the female and male is obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ). Despite the species name 'pusillus', meaning very small in Latin, the size range overlaps too much with other species to be diagnostic.
Description. This species was recently redescribed (Onuferko 2017). Distribution. Eastern North America to Mexico (Fig. 80). Ecology. HOST RECORDS: Rozen and Favreau (1968) associated E. pusillus with C. compactus compactus Cresson based on observations of a female of the former entering and emerging from a nest of a female of the latter and subsequent discovery of an Epeolus egg upon excavation of the nest. Ascher et al. (2014) noted that the small size and flight season of E. pusillus suggest and additional or alternative association with C. americanus Cresson. Cockerell, 1941Figs 81, 82, 96B Epeolus rufulus Cockerell, 1941.

Diagnosis.
The following morphological features in combination can be used to tell E. rufulus apart from all other North American Epeolus except E. attenboroughi: the mandible has a blunt, obtuse preapical tooth; the preoccipital ridge does not join the hypostomal carina; the mesoscutum is covered in pale tomentum, which is densest anteromedially; the axilla is elongate, extending well beyond the midlength of the mesoscutellum but not as far back as its posterior margin, and the free portion is distinctly hooked; the mesopleuron is closely (most i<1d) and evenly punctate; and T1-T4 have complete apical fasciae. Whereas in E. attenboroughi T1 has a comparatively narrow discal patch (the longitudinal band is more than half as wide as the breadth of the apical fascia in dorsal view) and in females F2 is not noticeably longer than wide, in E. rufulus the discal patch is so wide that the longitudinal band is barely visible in dorsal view and in females F2 is more than 1.2 × as long as wide. Epeolus rufulus is also similar to E. ainsliei in that in both species the axilla is dilated laterally and the free portion is distinctly hooked, and the T1-T4 apical fasciae are complete; however, in E. ainsliei the mandible is simple, the preoccipital ridge joins the hypostomal carina, and the mesoscutum has distinct paramedian bands.
Redescription. FEMALE: Length 7.6 mm (difficult to gauge in holotype because head detached and glued to collection label, and much of pronotum missing; given instead for non-type specimen most similar in size); head length 1.9 mm; head width 2.6 mm; fore wing length >5.1 mm (margins of both very worn in holotype).
Integument coloration. Black in part, at least partially ferruginous on mandible, labrum, clypeus, antenna, pronotal lobe, tegula, axilla, mesoscutum, mesoscutellum, metanotum, mesopleuron, metapleuron, propodeum, legs, metasomal terga (including pygidial plate), and metasomal sterna. Mandible with apex darker than all but extreme base; preapical tooth lighter than mandibular apex (difficult to see in holotype; described from non-type specimen). Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Mesoscutum orange along lateral margin and with pair of orange markings near posterior margin between midline and parapsidal line. Wing membrane subhyaline, apically dusky. Legs entirely reddish orange (both forelegs missing in holotype, but entirely reddish orange in non-type specimens).
Pubescence. Face with tomentum densest around antennal socket. Clypeus, upper paraocular and frontal areas, and vertexal area mostly exposed. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum sparsely covered in pale tomentum. Mesopleuron with upper half sparsely hairy; ventrolateral half nearly bare, except along margins. Metanotum with tomentum rubbed off medially in holotype, but uninterrupted and uniformly off white in nontype specimens. T1 with discal patch quadrangular and very wide, the basal and apical fasciae only narrowly joined laterally. T1 with basal and apical fasciae and T2-T4 with apical fasciae complete, those of T2 and T3 somewhat broader laterally, T2 with fascia without anterolateral extensions of tomentum. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD.
Structure. Preapical tooth blunt and obtuse. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.6). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal (not visible in holotype because head detached and glued to collection label; described from non-type specimens). Mesoscutellum weakly bigibbous. Axilla large, its lateral margin (L) more than half as long as mesoscutellar width (W) (L/W ratio = 0.6) and tip extending well beyond midlength of mesoscutellum but not as far back as its posterior margin; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin arcuate and carinate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, as long as wide (L/W ratio = 1.1); mesopleuron almost entirely obscured by white tomentum; S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep, well-separated punctures, with the interspaces shining.
Distribution. Great Plains to American southwest and presumably Mexico, given the close proximity of one collection locality (near Cloverdale, New Mexico) to the Mexico-United States border (Fig. 82).
Ecology. HOST RECORDS: The host species of E. rufulus is/are presently unknown. FLORAL RECORDS: The label of one examined voucher specimen indicates a floral association with Heterotheca subaxillaris ssp. latifolia.

Discussion.
In his unpublished thesis, Brumley (1965) synonymized Epeolus rufulus under E. crucis, treating the latter as a valid species. Herein, E. crucis is synonymized under E. compactus for reasons described in the Discussion of E. compactus. Also synonymized under E. crucis was E. novomexicanus, but morphological comparisons suggest that the type of E. novomexicanus belongs to the "pusillus group". Epeolus rufulus is similar in overall appearance to E. ainsliei and E. attenboroughi, and the ranges of the three species overlap to some extent.
Epeolus rufulus appears to be uncommon, or at least uncommonly collected. The male of E. rufulus is described here for the first time. There is very little morphological variation among the few examined specimens, and in all the mesoscutum lacks distinct paramedian bands and is instead sparsely covered in pale tomentum.
Material studied. Type material.   Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. scutellaris apart from all other North American Epeolus: the pronotal collar is predominantly black; the axilla is large, with the tip extending to or beyond the band of pale tomentum along the posterior margin of the mesoscutellum, dilated laterally, and ferruginous to some degree whereas the mesoscutellum ranges from entirely black to entirely ferruginous; the mesopleuron is closely (most i<1d) and evenly punctate and obscured by white tomentum only in the upper half (with a large, sparsely hairy circle occupying much of the ventrolateral half ); the T1-T3 apical fasciae are complete or only very narrowly interrupted medially; and the pseudopygidial area of the female is lunate with the apex clearly >2 × the medial length. Epeolus scutellaris resembles E. basili, E. nebulosus, E. novomexicanus, and E. pusillus in that the axilla is large, with the lateral margin arcuate, and that the apical fasciae are complete or only very narrowly interrupted medially. However, in E. scutellaris the pseudopygidial area of the female is wider (the apex ~2.5-3 × the medial length) than in the four members of the "pusillus group" (the apex clearly <2.5 × the medial length). In all four members of the "pusillus group", the mesopleuron of the male (excluding the hypoepimeral area) is entirely obscured by white tomentum and lacks the sparsely hairy circular area present in both sexes of E. scutellaris. Epeolus scutellaris is most similar to E. packeri in terms of surface sculpture and structure, but in E. packeri the pronotal collar is predominantly ferruginous, the T1 basal fascia is absent or reduced to a pair of small patches of pale tomentum, and the T1-T3 apical fasciae are interrupted medially and commonly reduced to discrete lateral patches. Epeolus scutellaris is also similar to E. andriyi and E. howardi, but in E. andriyi and E. howardi the T1-T3 apical fasciae are distinctly interrupted medially, and the pseudopygidial area of the female is lunate with the apex <2 × the medial length.

Epeolus scutellaris
Description. This species was recently redescribed (Onuferko 2017). Distribution. Widely distributed across the contiguous United States, excluding peninsular Florida and the west coast, and southern Canada (Maritime to Prairie provinces) (Fig. 84).
Ecology. See Onuferko (2017) for host and floral records. Floral associations are also indicated in Suppl. material 1, which includes newly discovered associations with Chrysothamnus (possibly in reference to plants that now are in the genus Ericameria), Erigeron, and Heterotheca subaxillaris based on labels of examined voucher specimens.
Discussion. In Onuferko (2017), E. scutellaris is said to be similar to two species from Florida yet to be formally recognized, which herein are formally described under   Description. FEMALE: Length 8.4 mm; head length 2.1 mm; head width 3.0 mm; fore wing length 6.6 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, antenna, pronotal lobe, tegula, legs, metasomal terga (including pygidial plate), and metasomal sterna. Mandible with apex and preapical tooth darker than rest of mandible. Antenna brown except scape and pedicel orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs with brown or black more extensive than reddish orange.
Pubescence. Face with tomentum densest around antennal socket. Dorsum of mesosoma and metasoma with bands of off-white and bright yellow short appressed setae. Pronotal collar with tomentum sparser medially, uniformly bright yellow. Mesoscutum with anteromedial chevron-shaped patch of bright yellow tomentum. Mesopleuron with upper half densely hairy, except beneath base of fore wing (hypoepimeral area); ventrolateral half sparsely hairy. Metanotum with tomentum uninterrupted, uniformly off white. T1 with broad, off-white basal fascia, complete bright yellow apical fascia, and narrow and extremely short discal patch of dark brown tomentum. T2-T4 each with complete bright yellow fascia, T2 and T3 with fasciae with anterolateral spots of sparser off-white tomentum. T5 covered in off-white tomentum except for line of separation from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on flat disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD.
Structure. Labrum with pair of small subapical denticles, each preceded by small discrete longitudinal ridge. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 as long as wide (L/W ratio = 1.0). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum weakly bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.3) and tip not extending beyond midlength of mesoscutellum; axilla with tip clearly visible, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, nearly as long as wide (L/W ratio = 0.95); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered basomedially and sparser apically and laterally, with the interspaces shining.
Etymology. The name is in reference to the uniquely smooth, shiny propodeum of this species. From the Latin, "splendidus" (bright).
Distribution. Known to occur in all major hot North American deserts (Fig. 86). Ecology. HOST RECORDS: The female PCYU paratype (see Material studied) was collected in the spring of 2015 along the Catalina Highway in Pima County, Arizona, USA where possible host Colletes visiting Eriogonum Michx. were collected and observed. Using Stephen's (1954) key, collected females were identified as C. wootoni Cockerell (one of which was sequenced and assigned the same BIN [BOLD:AAI9255] as a male from New Mexico whose terminalia were excised for identification) whereas collected males (one of which was sequenced and assigned the following BIN: BOLD:ABZ4837) were identified (based in part on examination of the terminalia, which were excised) as C. eulophi. Discussion. This southwestern species was identified as unique by Brumley (1965), and the colors and patterns of pubescence on the mesosoma and metasoma clearly set it apart from other Epeolus in North America. There is very little morphological variation among examined specimens, and sequenced material was assigned the same BIN. Based on known records, adults of E. splendidus are active in spring.
Material studied. Type material. Diagnosis. The following morphological features in combination can be used to tell E. tessieris apart from all other North American Epeolus except E. interruptus: the axilla does not attain the midlength of the mesoscutellum, its tip is unattached to the mesoscutellum for less than 1/3 of the entire medial length of the axilla, and like the mesoscutellum is ferruginous; the mesopleuron has sparser punctures ventrolaterally (most i≥1d) than in upper half, with the interspaces shining; and T1-T4 have medially-interrupted metasomal fasciae. Whereas in E. interruptus the metanotum has a blunt median process and T1 has a wide triangular discal patch with concave lateral sides, in E. tessieris the metanotum is flat and T1 has a trapezoidal to nearly semicircular discal patch. Description. FEMALE: Length 5.8 mm; head length 1.7 mm; head width 2.3 mm; fore wing length 4.8 mm.
Integument coloration. Mostly black; notable exceptions as follows: partially to entirely ferruginous on mandible, labrum, antenna, pronotal lobe, tegula, axilla, mesoscutellum, and legs. Mandible with apex darker than rest of mandible; preapical tooth lighter than mandibular apex (difficult to see in holotype because mandible closed; described from paratypes). Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Wing membrane subhyaline, apically dusky. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Dorsum of mesosoma and metasoma with bands of off-white to pale yellow short appressed setae. Mesoscutum with paramedian band. Mesopleuron with upper half hairy, except beneath base of fore wing (hypoepimeral area); ventrolateral half nearly bare. Metanotum with tomentum sparser medially, uniformly off white. T1 with median trapezoidal verging on semicircular black discal patch enclosed by pale tomentum, except for medial separations at base and apex. T2-T4 with fasciae interrupted medially and narrowed before becoming somewhat broader laterally, T2 with fascia with anterolateral extensions of sparser tomentum. T5 with two large patches of pale tomentum anterolateral to and separate from pseudopygidial area. T5 with pseudopygidial area lunate, its apex more than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs not extending beyond apex of sternum by much more than 1/4 MOD.
Structure. Preapical tooth blunt and obtuse. Labrum with submedial pair of very small denticles, apex with pair of small points separated by shallow concavity (difficult to see in holotype; described from paratypes). Frontal keel not strongly raised. Scape with greatest length 1.8 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by about 1.5 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla small to intermediate in size, its lateral margin (L) less than half as long as mesoscutellar width (W) (L/W ratio = 0.4) and not extending beyond midlength of mesoscutellum; axilla with tip visible, but unattached to mesoscutellum for less than 1/3 the medial length of axilla; axilla with lateral margin relatively straight and without carina. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, not noticeably longer than wide (L/W ratio = 1.1); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures closely clustered basomedially and sparser apically and laterally, with the interspaces shining.
Etymology. This species is named in honor of my wife, biologist Stéphanie Tessier. The name is in the genitive case and declined as mulier, a Latin noun with a consonant stem.
Ecology. HOST RECORDS: The host species of E. tessieris is/are presently unknown.   Brumley (1965) identified as new, this appears to be the least commonly collected species. Among examined specimens, there is notable variability in punctation density of the mesopleuron, but the smooth, shiny interspaces are usually greater than puncture diameters. Although BIN-compliant sequences are presently not available for E. tessieris, 421 bp sequences are available for two specimens (a female from Arizona, USA and a male from Coahuila, Mexico), and there is virtually no divergence (<1%) between the two. Moreover, these sequences do not cluster closely with any sequences from other Epeolus species in a NJ tree (Suppl. material 2).

Discussion. Of the Epeolus
Material studied. Type material. Primary: USA: Arizona: 3 mi W Marana (Pima County), 13.ix.1962 Diagnosis. The following morphological features in combination (excluding any that are specific to the opposite sex of the one being diagnosed) can be used to tell E. zonatus apart from all other North American Epeolus except E. erigeronis, E. ilicis, and E. inornatus: the mandible is simple; the axilla does not attain the midlength of the mesoscutellum but the free portion is distinctly hooked, with the tip unattached to the mesoscutellum for more than 1/3 of the entire medial length of the axilla; and the pseudopygidial area of the female is distinctly campanulate with the apex <2 × the medial length. Whereas in E. erigeronis, E. ilicis, and E. inornatus the pronotal collar and metasomal terga are black, as are sometimes the axilla and mesoscutellum, in E. zonatus the pronotal collar, axilla, mesoscutellum, T1, and T2 are ferruginous. Also, in E. zonatus the dorsum of the mesosoma and metasoma is commonly with much less pale pubescence. Redescription. FEMALE: Length 9.7 mm; head length 2.3 mm; head width 3.1 mm; fore wing length 6.2 mm. Integument coloration. Black in part, at least partially ferruginous on mandible, labrum, clypeus, antenna, pronotal collar, pronotal lobe, tegula, axilla, mesoscutum, mesoscutellum, metanotum, mesopleuron, legs, T1, T2, and metasomal sterna. Mandible with apex darker than all but extreme base. Antenna brown and orange in part. Pronotal lobe and tegula pale ferruginous to amber. Mesoscutum reddish-brown along lateral margin and with pair of reddish-brown markings near posterior margin between midline and parapsidal line. Wing membrane dusky subhyaline, slightly darker at apex. Legs more extensively reddish orange than brown or black.
Pubescence. Face with tomentum densest around antennal socket. Clypeus, upper paraocular and frontal areas, and vertexal area mostly exposed. Mesoscutum without pale tomentum. Dorsum of metasoma with bands of off-white short appressed setae. Mesopleuron nearly bare, except along margins. Metanotum with tomentum sparser medially, uniformly off white. T1 with discal patch quadrangular and very wide, the basal and apical fasciae at most only narrowly joined laterally (not joined in lectotype and multiple non-type specimens). T1 with basal and apical fasciae and T2-T3 with apical fasciae widely separated medially, the apical fasciae reduced to pairs of small patches somewhat broader laterally, T2 with fascia without anterolateral extensions of tomentum. T4 with fascia much more narrowly interrupted medially than on preceding terga. T5 with two faint patches of pale tomentum lateral to and contact-ing pseudopygidial area at apex, diverging from pseudopygidial area basally. T5 with pseudopygidial area campanulate, its apex less than twice as wide as medial length, indicated by silvery setae on impressed disc of apicomedial region elevated from rest of tergum. S5 with apical fimbria of coppery to silvery hairs extending beyond apex of sternum by ~2/5 MOD.
Structure. Mandible without preapical tooth. Labrum with pair of small subapical denticles not preceded by carinae. Frontal keel not strongly raised. Scape with greatest length 1.9 × greatest width. F2 noticeably longer than wide (L/W ratio = 1.4). Preoccipital ridge not joining hypostomal carina, from which it is separated by no less than 1 MOD at its terminal. Mesoscutellum moderately bigibbous. Axilla intermediate in size, its lateral margin (L) nearly half as long as mesoscutellar width (W) (L/W ratio = 0.4-0.5) and tip not extending beyond midlength of mesoscutellum; axilla with tip conspicuously diverging from side of mesoscutellum, distinctly hooked, and axilla with free portion 2/5 its medial length; axilla with lateral margin relatively straight and carinate. Fore wing with three submarginal cells. Pygidial plate apically truncate.
MALE: Description as for female except for usual secondary sexual characters and as follows: F2 shorter, not noticeably longer than wide (L/W ratio = 1.1); S4 and S5 with much longer coppery to silvery subapical hairs; pygidial plate apically rounded, with large deep punctures more or less evenly spaced throughout, with the interspaces shining.
Distribution. Florida and coastal Georgia (Fig. 90). Ecology. HOST RECORDS: The host species of E. zonatus is/are presently unknown.       A E. attenboroughi paratype, with F2 not noticeably longer than wide B E. rufulus holotype, with F2 noticeably longer than wide C E. gibbsi holotype, with F2 not noticeably longer than wide D E. inornatus paratype, with F2 noticeably longer than wide E E. barberiellus, with F2 as wide as long, or nearly so, and F E. banksi, with F2 noticeably longer than wide. Scale bars 0.5 mm.
Discussion. Smith (1854) described E. zonatus from both sexes, represented by three syntypes (all females) deposited at the NHMUK. The male description is actually based on a female specimen (see E. zonatus paralectotype [catalog number: 010812211] under Type material) of another species (E. bifasciatus). All three specimens were examined, and one of the two females of the true E. zonatus is herein designated as the lectotype, the one that is in better condition that fits Smith's (1854) original description of the female.
Structurally, E. zonatus and E. ilicis are identical, but in E. zonatus the pronotal collar, axilla, mesoscutellum, and discs of T1 and T2 are ferruginous, whereas in E. ilicis at least the pronotal collar and metasomal terga are entirely black. These Figure 97. Pseudopygidial area (in dorsal view) of female A E. packeri paratype (lunate and wider than long) B E. floridensis (lunate and nearly as long as wide) C E. scutellaris (lunate and wider than long) D E. basili paratype (lunate and wider than long) E E. novomexicanus (lunate and somewhat wider than long) F E. gibbsi paratype (campanulate and nearly as long as wide) G E. ilicis (campanulate and nearly as long as wide) H E. zonatus (campanulate and nearly as long as wide), and I E. australis (lunate and wider than long). Scale bars 1 mm. The pseudopygidial area is the apical portion of T5 that changes slope from the rest of the tergum and is covered in short, silvery hairs uniform in length (posteromesad the light blue lines).  . Metasoma (in dorsal view) dampened with water to show differences in integument coloration between T1 of male A E. nebulosus paratype, which is entirely black, and B E. novomexicanus, which is red beneath the apical fascia. Scale bars 2 mm. Note that lightly wetting the terga with ethanol allows for this feature to be seen without having to remove the tomentum.   Head (in lateral view) of female A E. australis, in which the frontal keel is strongly raised, and B E. brumleyi paratype, in which the frontal keel is only weakly protuberant. Scale bars 1 mm. Note that the supraclypeal area is usually covered in dense white tomentum, which was partially removed in these specimens to show the maximum extent of the keel.  Mesoscutum with paramedian band. Metasomal fasciae bright yellow to brownish orange and interrupted medially. Fig. 32B  F2 of female less than 1.2 × as long as wide (Fig. 96A). T1 in dorsal view with longitudinal band more than half as wide as breadth of apical fascia (Fig.  12B)  F2 of female more than 1.2 × as long as wide (Fig. 96B). T1 in dorsal view with discal patch so wide that longitudinal band barely visible (its width less than half the breadth of apical fascia) (Fig. 81B)  Axilla with tip not extending as far back as posterior margin of mesoscutellum, mesoscutellum dark brown or black basally (Fig. 8D) .E. andriyi sp. n. -Axilla with tip extending as far back as or beyond posterior margin of mesoscutellum, axilla and mesoscutellum entirely red (Fig. 55D)  Axilla with tip well short of band of pale tomentum along posterior margin of mesoscutellum (Fig. 16D), axilla and mesoscutellum entirely black. T2 fascia without anterolateral extensions of tomentum ( Fig. 16A-C) Mesopleuron of male obscured by white tomentum only in upper half (although hypoepimeral area usually with sparser tomentum), with a large, sparsely hairy circle occupying much of ventrolateral half (Fig. 83C). T5 with pseudopygidial area of female with apex clearly more than twice as wide as medial length (~2.5-3 × the medial length) (Fig. 97C). Axilla with tip extending to or beyond band of pale tomentum along posterior margin of mesoscutellum, mesoscutellum entirely black to entirely ferruginous (Fig. 83D)  Flagellum, except sometimes F1, and metasomal sterna (excluding apical margins) brown or black, clearly not the same reddish-orange color as legs from tibiae to tarsi (Fig. 98A). T1 with longitudinal extent of discal patch no less (and usually greater) than breadth of apical fascia (Fig. 79B). T1-T3 with apical fasciae removed from apical margin, commonly narrowed or narrowly interrupted medially ( T2 and T3 (for female) or T2-T4 (for male) with fasciae removed from apical margin, commonly narrowed or narrowly interrupted medially ( Fig.  24A-C). T5 with pseudopygidial area of female with apex at least twice as wide as medial length (Fig. 97D). T1 with longitudinal extent of discal patch greater than breadth of apical fascia, at least medially (Fig. 24B)  T2-T4 with fasciae on or very little removed from apical margin, more or less evenly broad (Figs 71A-C, 73A-C). T5 with pseudopygidial area of female with apex commonly less and no more than twice as wide as medial length (Fig. 97E). T1 with longitudinal extent of discal patch variable, but commonly less than breadth of apical fascia (Figs 71B, 73B)  Axilla with free portion about 2/5 its medial length or longer and distinctly hooked (i.e., concave, not relatively straight along medial margin) (minimum free extent shown in Fig. 65D) (see also Figs 43D, 51D, 57D, 59D, 89D). T5 with pseudopygidial area of female distinctly campanulate, with apex less than twice as wide as medial length (Fig. 97F-H)  Axilla with free portion less than 2/5 its entire medial length (usually ≤ 1/3) and relatively straight along medial margin (maximum free extent shown in Mesoscutum with anteromedial patch of bright or pale yellow tomentum, usually chevron-, horseshoe-, or V-shaped and narrowed anterolaterally (Figs 30B, 36B, 38, 45B) but sometimes semicircular (Fig. 85B) (Fig. 101B); IF joined posteriorly (i.e., U-or V-shaped), THEN not distinctly narrowed anterolaterally (Fig. 101A)  Propodeum with posterior surface highly polished and (except along lateral margin) hairless (Fig. 102A). T1 with broad, transverse off-white basal fascia, discal patch greatly reduced or absent; T1-T4 with complete bright yellow apical fasciae, terga otherwise covered in brown (and laterally sometimes offwhite) tomentum ( Fig. 85A-C)  Propodeum with posterior surface dull due to surface microsculpture and with long erect hairs submedially (Fig. 102B). T1 with median black or nearly black discal patch surrounded by pale tomentum; T1-T4 with complete or medially-interrupted pale yellow apical fasciae, terga otherwise covered in black or nearly black tomentum (Figs 30A-C, 36A-C, 38B, 45A-C) ....... 31 31 T1 discal patch triangular or semicircular (with lateral sides straight or convex), basal fascia fully continuous with longitudinal band AND discal patch more elongate, its medial longitudinal extent (measured as if apical fascia were complete) more than 1/3 the lateral extent. Fig. 30B  T1 with median triangular or semicircular discal patch (basal fascia conspicuously arched, apical fascia straight) AND longitudinal band at least half as wide as breadth of apical fascia in dorsal view (Fig. 47B) ...... E. flavofasciatus Smith -T1 not as above; IF discal patch triangular, THEN so wide that longitudinal band barely visible in dorsal view (its width less than half the breadth of apical fascia) (Fig. 61B)  This work would not have been possible without loans of specimens and help from the curatorial staff of the institutions listed in the Methods -I thank you all very much. I am especially grateful to the curators and collections managers of the abovementioned institutions who willingly sent primary type specimens to me for study, without which this revision would have been impossible. I thank Prof. John Ascher and Prof. Jason Gibbs for sharing their insights and observations with me regarding potential Colletes host associations. Some field work was conducted with permission on protected lands. I thank the Claremont Colleges and in particular Dr. Wallace Meyer (Pomona College) for giving me access to collect Epeolus and other bees at the Robert J. Bernard Biological Field Station in Claremont, California. The area hosts several uncommon/ rare cuckoo bees, and its importance as a refugium for native species in an otherwise completely human-transformed environment cannot be overstated. Several specimens used in this study were collected at Rondeau Provincial Park, and I am grateful to the Ministry of Natural Resources and Forestry for granting me a permit to collect there. Most figures presented herein were made possible with the use of the imaging system at the PCYU, which was purchased through Canadensys with funds from the Canadian Foundation for Innovation and the Ontario Research Fund. While working on this revision, I was a recipient of the Susan Mann Dissertation Scholarship (issued by the Faculty of Graduate Studies at York University), through which I received generous financial support, for which I am very grateful. This study was funded by Laurence Packer's discovery grant from the Natural Sciences and Engineering Research Council of Canada (NSERC). Funds for DNA barcoding were generously donated by Robert and Cecily Bradshaw. I thank Liam Graham for his help in preparing barcode plates and databasing and imaging specimens for BOLD, and the Biodiversity Institute of Ontario, University of Guelph for processing the plates. This article was reviewed for ZooKeys by Dr. Molly G. Rightmyer, whose thorough assessment and constructive remarks are very much appreciated. Lastly, the studied material and associated data were made available through the contributions of countless collectors, whom I thank for helping further the knowledge of this most wonderful and fascinating group of bees.