Revision of the New World Ceratoculicoides Wirth & Ratanaworabhan (Diptera, Ceratopogonidae, Ceratopogonini)

. The New World species of the genus Ceratoculicoides Wirth & Ratanaworabhan are described, illustrated and keyed in both sexes. Ceratoculicoides borkenti sp. nov., C. confusus sp. nov. , C. grogani sp. nov., C. pacificus sp. nov. and C. propinquus sp. nov. are described, with C. confusus being the first record of the genus from South America (Colombia). Ceratoculicoides blantoni is a junior synonym of C. virginianus and the identity of C. longipennis remains unclear as the male morphospecies associated in taxonomic literature with the female type series is not conspecific. A morphological phylogeny of all extant species in the genus is presented and characters discussed. The moravicus species group is recognized for a clade of species with the lateral margins of the aedeagus straight or concave.


Introduction
is a genus of small to mid-size (<1.8 mm wing length; Figs 1-2), poorly known predaceous midges that are collected only periodically, as adults and are entirely unknown as immatures. The genus is placed in the tribe Ceratopogonini, a paraphyletic grouping of genera (Borkent 2014) defined by the presence of palisade setae on the 1 st tarsomere of the hind leg and a petiolate or obsolete M 1+2 fork in the wing. Within the Ceratopogonini, Ceratoculicoides is phylogenetically related to several genera having 1-4 katepisternal setae including Brachypogon Kieffer, 1899 (Borkent 1995). Adults are most readily separated from all other Ceratopogonidae by a cluster of 2-3 setae on the posterior margin of the anterior anepisternum (Borkent 1992). Ceratoculicoides has been recorded from Europe (Havelka 1976(Havelka , 1980Knoz 1987;Delécolle & Schiegg 1998;Szadziewski & Knoz 2002), western Asia (Remm 1967), North America (Wirth 1951(Wirth , 1952Wirth & Ratanaworabhan 1971) and Central America (Huerta & Borkent 2005;, and here, for the first time, from northern South America. Life history and bionomic information is scant, with ecological studies collecting species in traps near streams (Havelka 1976(Havelka , 1980 and decaying wood (Delécolle & Schiegg 1998).
For the phylogenetic analysis morphological characters were directly scored from specimens, except Sinhalohelea gansi Grogan & Borkent, 1992, C. aliciae and C. tontoeguri, which were scored from the original descriptions and Delécolle & Schiegg (1998). Sinhalohelea gansi, Brachypogon (Brachypogon) canadensis Downes, 1976 and Brachypogon (Isohelea) sp. L (sensu W.L. Grogan, specimens in FSCA) were used as outgroups. Inapplicable character states were scored as "-" and missing data as "?;" none of the characters were weighted. The resulting dataset comprised 11 taxa (three outgroup, eight ingroup) and 12 characters (Table 1). Mesquite ver. 3.61 (Maddison & Maddison 2019) was used to build the character matrix and interpret character state distributions and polarity, with TNT 1.5 (Goloboff et al. 2008;Goloboff & Catalano 2016) used for a Maximum Parsimony analysis of the dataset via an Implicit Enumeration search (since the number of taxa is small enough to sample the entire treespace), and the Bremer Support function to find the branch support on the tree. Bootstrap values were also calculated but are not included in the results, as none exceeded the threshold value of 50. The tree figure was initially generated using iToL (Letunic & Bork 2019) and finalized with Inkscape. 11-12 with ring of laterally directed trichoid sensilla at base, 13 with apical trichoid sensillum, AR 0.65-1.02, FR 1.49-1.92. Eyes separated medially by diameter of 3-5 ommatidia, ommatrichia present. Palpus brown, with 5 segments, 3 rd with sensory pit.
Wing. Cells r 1 and r 2 reduced, M 2 base obsolete. Membrane with microtrichia, macrotrichia present only on C and R; membrane unpatterned, without macrotrichia.
PregeniTaL abdomen. Brown, without distinct coloration patterns, margins roughly parallel to genitalia.
geniTaLia. Epandrium with apicolateral processes present, cerci and proctiger near posterior margin of epandrium directed ventrally. Gonocoxite cylindrical; gonocoxal apodeme quadrate anteriorly, triangular posteriorly; gonostylus simple, subequal in length to gonocoxite, weakly curved at tip. Sternite 9 (hypandrium) slightly tapering anteriorly, length/width ratio 0.33, with medial emargination. Parameres FASBENDER A., Revision of New World Ceratoculicoides separate or weakly fused at base; articulating with anterior portion of gonocoxal apodeme, apical arms of paramere stylate, directed posteriorly at distinct angle from base, midpoint arched dorsally in lateral view, apex of paramere ventrally directed, extending to apex of aedeagus or beyond. Aedeagus heavily sclerotized, with or without apical hyaline incision medially, with distinct basal arms articulated on anterior portion of gonocoxal apodeme, posterolateral point present, of variable form.
Wing. Membrane with scattered macrotrichia along apical margin.
Legs. Fore-, midleg tarsal claws longer than those of hindleg, equal or slightly unequally sized, gently curving along length, apex with simple point.

Immature stages
Currently unknown.

Key to adult male New World Ceratoculicoides
The male lifestage of Ceratoculicoides longipennis from California is currently unknown. A poorly preserved male specimen from California that may be conspecific with C. longipennis is included as C. sp. M1 in the key. The distribution of each species is noted in brackets.

Description (male)
Aedeagus lateral margins straight or concave, posterior margin straight or concave. Hyaline posteromedial incision of aedeagus present. Posterolateral points of aedeagus directed laterally, adjacent or contiguous with apical points.

Remarks
The C. moravicus group is proposed for C. moravicus, C. aliciae, C. borkenti sp. nov., C. grogani sp. nov., C. propinquus sp. nov. and C. sp. M1. All of these species share similar aedeagal characters, most notably the synapomorphy of straight or concave lateral margins of the aedeagus. There are no female characters that can be used to diagnose this group. Huerta & Borkent, 2005 Figs 7a-b, 10d-e, 11a

Diagnosis
Male Ceratoculicoides aliciae can be separated from congeners by the following combination of characters: femora and tibiae yellow; apices of parameres acute, tapering; aedeagus lateral margins parallel basally, tapering apically, posterior margin with acute medial notch, apical and posterolateral point adjacent, posterolateral point rounded (Fig. 10e).

Remarks
As I did not examine the holotype, I cannot provide a full description of this species, but I did have the opportunity to examine the two female paratypes. Based on the description and illustrations of the male in Huerta & Borkent (2005), this species belongs in the C. moravicus group, similar to C. sp. M1, C. moravicus and C. propinquus sp. nov., based on the presence of accessory spines between the apical and posterolateral points of the aedeagus. Males of C. aliciae (Fig. 10e) can be recognized by the straight margins of the aedeagus narrowed conspicuously beyond their midpoint, the posterior margin with a distinct acute notch, and the apical and posterolateral points being adjacent. The females can be recognized by the largest spermatheca being 70-76 μm, a feature  (Wirth, 1951). Scale bars = 100 μm.
European Journal of Taxonomy 875: 159-202 (2023) found only in C. moravicus and an undescribed Colombian morphospecies (C. F2, Figs 7b, 12d; see taxonomic notes on C. confusus). Females of C. aliciae can be distinguished by the posterior branch of the 9 th sternite tapering to an acute point, unlike the broadly rounded posterior branch of C. sp. F2. Both species are distinguished from C. moravicus by the presence of two major spermathecae.

Etymology
This species is named in honor of Art Borkent for encouraging my study of Ceratopogonidae, including suggesting this project after I collected a specimen of this species.  THorax. Dorsocentral punctations inconspicuous, present only at posteriormost portion of dorsocentral setae near scutellum, may be absent in some specimens. Legs with femora and tibiae brown. geniTaLia ( Fig. 9d-f). Distal portion of parameres tapering gradually to acute apex. Aedeagus lateral margins concave, smoothly rounded, apical point posterolaterally directed, triangular with adjacent lateral obtuse spur-like posterolateral point without accessory spines, posterior margin emarginate, hyaline, hyaline medial incision extensive, margins parallel to lateral margins and basal arch. THorax. Legs with femora and tibiae brown. geniTaLia (Fig. 11b). 9 th sternite anterior branch acutely pointed, apices of each half touching medially or not; posterior branch evenly curving towards apex, spiniform, tip acute to rounded. 2 major spermathecae.

Remarks
Ceratoculicoides borkenti sp. nov. is one of the most distinctive species of Ceratoculicoides in the adult male life stage. As member of the C. moravicus group, the lateral margins of the aedeagus are heavily sclerotized, deeply concave, and taper in a smooth arc to a narrow apex (Fig. 9f). The median portion of the aedeagus is so hyaline as to be nearly transparent, and unlike any other Ceratoculicoides the hyaline incision expands above the basal arch subtrapezoidally, leaving only the lateral and anterior margins of the aedeagus sclerotized.
The females of C. borkenti sp. nov. (Fig. 11b) fall into a group of several species with the length of the largest spermathecae <70 μm. Of the described species, it can be differentiated from C. grogani sp. nov. by the medial margin of the 9 th sternite being deeply concave and rounded (vs nearly straight and sinuous, Fig. 11e) and C. virginianus (Fig. 12c) by the 9 th sternite's acute or triangular anterior branch (vs rounded or truncate) and evenly curved posteromedially directed spiniform posterior branch (base of branch directed posteriorly, curving medially in apical half in C. virginianus, usually tapering only in apical portion). However, there are a number of unassociated female specimens which display a similar morphology, including a species from Colombia (C. sp. F1, Fig. 12c) and another from Costa Rica (C. sp. F3, Fig. 12e). The Costa Rican C. sp F3 differs in having the base of the posterior branch directed posteriorly, vs evenly curving along the whole length in C. borkenti, but C. sp. F1 from Colombia cannot be reliably differentiated by the 9 th sternite. C. borkenti has a lower flagellum ratio (FR <1.5) than either of these provisional species. As there is only a limited sample of reliably associated females, and there are two male morphospecies without a female association sharing the range of C. borkenti (C. propinquus sp. nov. and C. sp. M1), the characters cited above may prove non-diagnostic upon examination of a broader range of material.

Diagnosis
Male Ceratoculicoides grogani sp. nov. can be separated from congeners by the following combination of characters: femora and tibiae yellow or lightly infuscate; apices of parameres acute, tapering distally; aedeagus lateral margins straight, posterolateral point an acute spine directed laterally (Fig. 9i).

Female
Only species of Ceratoculicoides with the following combination of characters: hind femur infuscate, other femora and tibiae brown, wing length ~ 1.4 mm; 2 spermathecae, largest 60; medial margin of 9 th sternite weakly concave, sinuous. THorax. Legs with femora and tibiae yellow or very lightly infuscate. geniTaLia ( Fig. 9g-i). Distal portion of parameres tapering gradually to acute apex. Aedeagus broad posteriorly, tapering anteriorly; lateral margins forming a smooth, shallow arc, divergent posteriorly, curving outward to acute laterally directed posterolateral point at apex, forming a smooth concave arc medially to triangular posteriorly directed apical point; posterior margin weakly concave, hyaline medial incision broad posteriorly, tapering anteriorly.

Remarks
This species belongs to the C. moravicus group. It can be differentiated from C. moravicus (Fig. 10h) and C. aliciae (Fig. 10e) by the aedeagus not conspicuously narrowing midway along its length, and from C. sp. M1 (Fig. 10c) and C. propinquus sp. nov. (Fig. 10k) by having the posterolateral point elongate and spiniform. It is readily distinguished from C. borkenti sp. nov. (Fig. 9f) by the weakly developed hyaline incision of the aedeagus. The female of this species is the most readily recognizable of any Ceratoculicoides, as the medial apex of the 9 th sternite is only weakly concave, instead of deeply emarginate (Fig. 11e).  Knoz, 1987 Figs 5b, 7e, 10f-h, 11f

Diagnosis
Male Ceratoculicoides moravicus can be separated from congeners by the following combination of characters: femora and tibiae brown; wing length 1.1 mm, apices of parameres acute, tapering distally; aedeagus lateral margins straight basally, strongly tapered at midpoint, apical and posterolateral points adjacent, 1-2 accessory spines between apical point and posterolateral point of aedeagus, posterolateral point subtriangular, apex directed anteriorly (Fig. 10h).

Female
Only species of Ceratoculicoides with the following combination of characters: FR 1.62-1.83; femora and tibiae brown; wing length ~ 1.2 mm; single major spermatheca 75-77; medial margin of 9 th sternite deeply concave.
THorax. Dorsocentral punctations possibly absent, at most a few spots of thinned cuticle among dorsocentral setae. Legs with femora and tibiae brown. geniTaLia ( Fig. 10f-h). Distal portion of parameres tapering gradually to acute apex. Aedeagus lateral margins tapering, constricted noticeably at midlength; posterior margin a smooth concave arc, hyaline medial incision broad posteriorly, tapering anteriorly; base of posterolateral point directed laterally, apex rounded, directed anterolaterally; apical point subacute, directed posterolaterally, adjacent to posterolateral point, with 1-2 accessory spines between apical and posterolateral points, similar in size and shape to apical point.

Distribution
Western Palaearctic.

Remarks
This extralimital species is diagnosed here to aid in differentiating it from other members of the C. moravicus group, particularly C. sp. M1 (see above). Delécolle & Schiegg synonymized C. havelkai and C. moravicus without examining either holotype, based solely on specimens from Switzerland and Germany (Delécolle & Schiegg 1998). Any future work on the Palaearctic fauna should involve examining both type specimens to confirm they are conspecific. Females of this species are the only known Ceratoculicoides with a single major spermatheca. The males are extremely similar to those of C. aliciae, C. propinquus sp. nov. and C. sp. M1. Ceratoculicoides moravicus (Fig. 10h) can be distinguished from C. aliciae (Fig. 10e) by the posterior margin of the aedeagus being evenly concave (vs with acute medial notch), from C. propinquus (Fig. 10k) by the accessory spines between the apical and posterolateral points (spines absent in C. propinquus), and C. sp. M1 (Fig. 10c) by having the aedeagus tapering distinctly at midlength (vs lateral margins straight). The differences between C. moravicus and C. sp. M1 may be the result of the overly compressed slide preparation of that specimen. Additional material is needed to ascertain whether the latter provisional morphospecies is conspecific.

Female
Unknown.

Material
THorax. Dorsocentral punctations possibly absent, at most a few spots of thinned cuticle among dorsocentral setae. Legs with femora and tibiae yellow or very lightly infuscate. geniTaLia (Fig. 10i-k). Distal portion of parameres tapering gradually to acute apex. Aedeagus lateral margins straight, becoming slightly divergent at apex; posterior margin a smooth concave arc, hyaline medial incision broadest posteriorly, tapering anteriorly; base of posterolateral point directed laterally, apex rounded, directed anterolaterally; apical point subacute, directed posterolaterally, adjacent to posterolateral point, without accessory spines between apical and posterolateral points, similar in size and shape to apical point.

Remarks
This species belongs to the C. moravicus group, and is very similar to C. aliciae, C. sp. M1 and C. moravicus. It can be recognized by the straight lateral margins of the aedeagus (vs distinctly tapering at midlength in C. aliciae and C. moravicus) and the apical and posterolateral points being adjacent but without accessory spines between them (spines present in C. sp. M1 and C. moravicus), and the posterolateral point apex rounded and directed anteriorly (Fig. 10k). The posterior margin of the aedeagus forms an evenly concave arc, while C. aliciae has a distinct acute medial notch along its posterior margin (Fig. 10e). Females have not been associated for this species.

Diagnosis
Male It can be separated from congeners by the following combination of characters: femora and tibiae brown; wing length 1.2 mm; apices of parameres acute, tapering distally; aedeagus lateral margins straight, apical and posterolateral points adjacent, 1-2 accessory spines between apical point and posterolateral point of aedeagus, posterolateral point subtriangular, apex directed anteriorly (Fig. 10c).

Female
Unknown.

Remarks
This is a provisional male morphospecies (not a formal binomen), as there is some evidence this may be the male of C. longipennis (see taxonomic note for that species). The genitalia of the only specimen were heavily compressed during mounting, splitting the aedeagus and distorting other structures, complicating its comparison with similar species. This morphospecies belongs to the C. moravicus group based on having the aedeagus lateral margins mostly straight and a concave posterior margin, deep hyaline medial incision, and the dorsolateral and apical points of the aedeagus adjacent at the posterolateral corner of the aedeagus. Ceratoculicoides sp. M1 (Fig. 10c) lacks the extensive subtrapezoidal medial incision of C. borkenti sp. nov. (Fig. 9f), while this species can be separated from C. grogani sp. nov. (Fig. 9i) by the posterolateral point length being subequal to its basal width (vs much longer than its basal width) and the apex of the posterolateral point is directed anterolaterally (vs slightly posterolaterally). It is distinguished from C. propinquus sp. nov. (Fig. 10k) by the presence of accessory spines between the apical and posterolateral points of the aedeagus, and larger adult size (1.25 mm wing length). It can be differentiated from the very similar C. aliciae (Fig. 10e) and C. moravicus (Fig. 10h) by the lateral margins of the aedeagus being relatively straight, weakly expanding apically (vs distinctly narrowing about midway along their length in those two species) and the emargination at the posterior apex of the aedeagus is much broader. However, it is possible that these differences result from the amount of compression the specimen experienced during slide mounting, as the aedeagus is partially split medially. Ceratoculicoides aliciae is smaller (male wing length 0.93 mm) than C. sp. M1 (suggesting its distinctness), but it is possible that C. moravicus is conspecific with C. sp. M1. This needs to be resolved by examining additional uncompressed specimens.  fig. 16.

Diagnosis
Male Ceratoculicoides confusus sp. nov. can be separated from congeners by the following combination of characters: femora and tibiae brown, apices of parameres acute, tapering distally; aedeagus lateral and posterior margins convex, medial apex of aedeagus without hyaline incision, apical points absent, FASBENDER A., Revision of New World Ceratoculicoides posterolateral points spiniform, emerging from dorsal surface of aedeagus, apex curving posterior, < 0.25 times width of apex of aedeagus (Fig. 8c, f).

Etymology
'Confusus' is the Latin participle for 'confusion', in reference to the past misapplication of the name C. longipennis to this species. THorax. Dorsocentral punctations inconspicuous, present among posterior third of dorsocentral setae, may be absent in some specimens. Legs with femora and tibiae brown. geniTaLia (Fig. 8a-f). Distal portion of parameres tapering gradually to acute apex. Aedeagus lateral margins convex, rounded, seamlessly transitioning into rounded posterior margin, medial apex without notch or hyaline incision; posterolateral point a posteriorly directed, hooked acute spine on dorsal surface of posterolateral margin.  (77); spermathecal neck 18-28(24); spermatheca/neck ratio 0.23.

Remarks
The male of this species was described as C. longipennis in Wirth & Ratanaworabhan (1971). It appears Wirth & Ratanaworabhan conflated three species in their description of C. longipennis, assigning female specimens with large spermathecae from the Pacific Northwest with C. longipennis based on wing length, then associating this eastern male morphospecies to C. longipennis based on the large spermathecae found in the females. The females associated with this eastern morphospecies are consistently smaller than the holotype of C. longipennis, with wing lengths less than 1.5 mm (vs 1.6 in said holotype); thus, the aforementioned male morphospecies represents a new taxon, C. confusus. Subsequent publication records of C. longipennis from eastern North America (Wilkening et al. 1985) represent this species. Female C. confusus can be recognized by being the only known species in eastern North America with a spermathecal length >80 μm. Only the western C. pacificus has similarly sized spermathecae, but it has wing lengths over 1.5 mm, while the female of C. confusus has a wing length between 1.2 and 1.5 mm. The males of this species can be distinguished by the convex lateral and posterior margins of the aedeagus, lack of a hyaline incision along the posterior margin, and absence of apical points (Fig. 8c).
I have assigned a male specimen from Colombia to this species based on its essentially identical genitalia features. The other morphological characters of this male fit within the range of North American material (the measurements of this specimen are included in brackets after the North American material in the description), although the dorsocentral punctations are more numerous and the antennal ratio is smaller than in the Nearctic specimens. I have also examined two females from the same locality (see Unassociated Ceratoculicoides female Ceratoculicoides specimens below), each apparently belonging to a separate species based on their wing lengths and the size of their spermathecae (C. sp. F1 with 1.25 mm wing length and 62 μm spermatheca length vs C. sp. F2 with 0.95 mm wing length and 76 μm spermatheca length). I do not feel confident associating either of these female morphospecies with the male based on the currently available material. If either of these female species is conspecific with this Colombian male, it would require reassessment of the species characters and boundaries of C. confusus. (Wirth, 1952) Helea longipennis Wirth, 1952: 201 (original description (not conspecific with holotype, see taxonomic notes).

Remarks
There has been considerable confusion with the identity of this species, as the holotype is a female described in Wirth's (1952) monograph of the California Ceratopogonidae. Wirth & Ratanaworabhan (1971) asserted that this species was conspecific with a morphospecies from eastern North America based primarily on the size of their spermathecae. Based on their material examined list, Wirth & Ratanaworabhan did not have access to any male specimens from California or nearby states. The broader range of material I have been able to examine demonstrates that the eastern morphospecies is not C. longipennis (described above as C. confusus). My comparison of the paratype and holotype of C. longipennis also found they are not conspecific, with the paratype having a wing length of 1 mm, while the holotype wing length is 1.6 mm. While Wirth & Ratanaworabhan described large (75-82 μm) spermathecae for this species, they never dissected or slide mounted the holotype. It appears from their material examined and my work in the USNM collection that they measured slide mounted females from Oregon and Washington belonging to C. pacificus as part of their description. I have not dissected the C. longipennis holotype to ascertain the size of its spermathecae, as I did not realize the depth of confusion surrounding the identity of this species when I had access to it. Even dissecting the holotype would offer few clues to the identification of this species based on the information currently available, as there are only a handful of characters which offer any diagnostic utility, and many female specimens are currently unplaceable to species (see Discussion).
Females of C. pacificus match the C. longipennis holotype in wing length, but they are known only from temperate rainforests in British Colombia, Oregon, and Washington. The nearest locality of this species is in the Willamette Valley of Oregon, over 900 km from the type locality in the Sierra Nevada Range. In the USNM collection, I found a specimen of a male morphospecies from Death Valley National Park, California, which fits the large size (~ 1.25 mm wing length) expected for the FASBENDER A., Revision of New World Ceratoculicoides male of C. longipennis. As size and loose geographic proximity are extremely weak evidence upon which to base an association, I have treated this male species as Ceratoculicoides sp. M1 above. Since the holotype remains entire and pinned, it may be possible to sequence its molecular barcode to associate it with future specimens. However, if additional material and further study deprecates the diagnosability of the holotype, it may be necessary to petition the International Commission of Zoological Nomenclature to set aside the holotype and designate a male neotype for C. longipennis (under article 75.5, ICZN 1999).

Diagnosis
Male Ceratoculicoides pacificus sp. nov. can be separated from congeners by the following combination of characters: femora and tibiae brown; apices of parameres acute, tapering distally; aedeagus lateral margins convex, medial apex of aedeagus with hyaline incision, posterolateral spines elongate, 0.5 width of apex of aedeagus (Fig. 8i).

Etymology
The specific epithet 'pacificus' is a Latin adjective referring to 'peacemaking' or 'peacefulness'. This epithet is in reference to this species inhabiting the coastal Pacific Northwest region.  Co., Okanagan-Wenatchee National Forest, DeRoux Forest Campground;11 Aug. 1971; Goeden and Gurney leg.; light trap; originally identified as C. longipennis; USNM • 1 ♂, 1 ♀; same collection data as for preceding; Olympic National Park; CNCI. THorax. Legs with femora and tibiae brown. geniTaLia ( Fig. 12a-b). 9 th sternite anterior branch truncate, apices widely separated; base of posterior branch directed posteriorly, distal half curving medially, tip usually rounded, sometimes with a minute hook at apex. 2 major spermathecae.

Remarks
Females of this species are similar in size to the female holotype of C. longipennis, but the geographic disjunction between the type locality and the range of C. pacificus argues against their conspecificity. I believe that the stability of nomenclature is better served by describing this taxon under a new name rather than assigning this morphospecies to C. longipennis based on spurious evidence, especially as another male morphospecies, sp. M1, has as much or more evidence suggesting it is the male of C. longipennis. Females of this species are the only Ceratoculicoides in western North America with spermathecae >80 μm long, but the other regional females are poorly known and this character may not be diagnostic. Males can be distinguished from all other Ceratoculicoides by their aedeagus having convex lateral and posterior margins, long (>0.5× width of main body of the aedeagus) posterolateral points, and acute hyaline medial incision flanked by short triangular apical points well separated from the posterolateral points (Fig. 8i).

Diagnosis
Male Ceratoculicoides virginianus can be separated from congeners by the following combination of characters: femora and tibiae yellow or brown; apices of paramere broadly rounded, not tapering, minute apical point displaced laterally; aedeagus subquadrate, slightly constricted apically (Fig. 9b-c).

Holotype
THorax. Dorsocentral punctations prominent and interspersed in posterior half of dorsocentral setae. Legs with femora and tibiae yellow or brown. geniTaLia (Fig. 9a-c). Distal portion of parameres not tapered, apex broadly rounded, with small subacute point displaced laterally. Aedeagus lateral margins convex, rounded, posterior margin straight, notched medially without hyaline incision, posterolateral point triangular, shifted medially nearly to lateral margin of medial notch, curving smoothly into dorsally directed, acute apical point. THorax. Legs with femora and tibiae yellow or brown. geniTaLia (Fig. 12c). 9 th sternite anterior branch truncate or obtusely rounded, apices widely separated; base of posterior branch directed posteriorly, distal half curving medially, tip rounded or pointed. 2 major spermathecae.

Remarks
Wirth & Ratanaworabhan (1971) described C. blantoni, distinguishing it from C. virginianus based on leg coloration (brown in C. blantoni, yellow in C. virginianus), orientation of the apex of the paramere (posterolateral in C. blantoni, ventroposterior in C. virginianus), and the aedeagal proportions (broader than long in C. blantoni, as long as broad in C. virginianus). My examination of additional material of both species found that these characters are not diagnostic. Leg color appears to be a spectrum; while some specimens have completely yellow and others completely brown legs, there are intermediates with solid brown femora and yellow tibiae or femora proximally brown and yellow distally. The legs of the holotype of C. blantoni are now yellow with brown bands on the femora, probably a result of the slide mounting process. The male genitalia characters used by Wirth & Ratanaworabhan appear to be artifacts of slide mounting. Specifically, the direction of the apex of the parameres is directly correlated to the compression of the genitalia by the coverslip during mounting: when there is little pressure the parameres point ventroposteriorly, but when compressed the apices are displaced laterally. The proportions of the aedeagus are similarly affected by the mounting of the specimen, the degree of tilt of the aedeagus in the dorsal plane changes its apparent proportions on the slide mount due to foreshortening. As no characters consistently separate C. blantoni from C. virginianus, C. blantoni is here considered a junior synonym of the latter species.
Currently, females of this species can be recognized by being the only known species in eastern North America with spermathecae < 70 μm long. There are several morphospecies from the western Nearctic European Journal of Taxonomy 875: 159-202 (2023) and Neotropics with similarly sized spermathecae. Ceratoculicoides virginianus (Fig. 12c) can be separated from C. grogani sp. nov. by the deeply concave and rounded medial margin of the 9 th sternite (vs weakly concave and sinuous, Fig. 11e), the base of the posterior branch being posteriorly from the base, curving medially in its apical half (vs directed posteromedially in an even curve in C. borkenti, Fig. 11b), the posterior branch tapering only in its apical half (vs C. borkenti sp. nov. and C. sp. F3 where it tapers to an acute spine in the basal half , Figs 11b, 12f). Accordingly, C. virginianus cannot be distinguished from C. sp. F1 based on morphology from our current knowledge. There is some variability in the shape of the apices of the anterior and posterior branches of the 9 th sternite. The anterior branch can be obtusely rounded or somewhat flattened and truncate, while the tip of the posterior branch may be either pointed or acutely rounded. Males can be separated from all other Ceratoculicoides by the broadly rounded apex of the paramere with subacute triangular point (Fig. 9b), the aedeagus with convex lateral margins, no hyaline medial incision, and the dorsolateral and apical points directed dorsally (Fig. 9c).

Female
Only species of Ceratoculicoides with the following combination of characters: femora and tibiae yellow, wing length ~ 1.2 mm; 2 major spermathecae, largest 56-66; medial margin of 9 th sternite deeply concave.

Remarks
This provisional species may be conspecific with C. virginianus, as there are no morphological characters which separate them in the female stage. I have refrained from assigning these specimens to that species FASBENDER A., Revision of New World Ceratoculicoides until the male life stage for the Colombian material is collected. This species can be differentiated from all other species besides C. virginianus by the spermathecal length being <70 μm and concave medial margin of the 9 th sternite having the posterior branch straight basally, tapering and curving only in its distal portion (Fig. 12d).

Remarks
Major spermathecae with a length between 70 and 80 μm are found only in this species and C. aliciae. Ceratoculicoides sp. F2 differs from that species in the posterior branch of the 9 th sternite being obtusely rounded (Fig. 12e), vs acutely pointed in C. aliciae (Fig. 11a).

Female characters and associations
The taxonomy of Ceratoculicoides has been complicated by the selection of Helea longipennis Wirth, 1952 as the type species, since the holotype of that species is a female and the paratype a second nonconspecific female. Neither of these types can be associated with a male morphospecies based on current evidence, especially as there are four Ceratoculicoides male morphospecies known from California, two of which have no female associations. The females of this genus are difficult to differentiate at the species level, with many characters used in other ceratopogonids being uniform or unreliable. The extent of spermathecal neck sclerotization is variable between otherwise identical specimens of a female series, suggesting it is not reliable to distinguish between species. I have discounted the use of a spermathecal ratio (length/width of spermathecae) as the width of the spermatheca varies depending on its orientation, and spermathecae are often collapsed and distorted during the clearing and slide mounting process. The chaetotaxy of the thorax also varies, especially the number of anepisternal and katepisternal setae. This is corroborated as intraspecific variability since conspecific males from the same site often have differing numbers of setae in the same positions. In contrast, the antennae, palpi and legs have proven largely identical throughout the genus.
I have found only four characters that potentially discriminate female morphospecies: flagellum ratio (FR), wing length, spermathecal length and the shape of the 9 th sternites's genital sclerotization. The flagellum ratio (the length of the antennal flagellum divided by the width of the head) can be quite variable within a species, possibly related to allometric changes related to environmental conditions like those reported in the Chironomidae (McKie & Cranston 2005). This seems to be corroborated by C. confusus and C. virginianus, whose ranges cover a wide climatic gradient from eastern Canada to Florida and have markedly variable antennal and flagellar ratios. A smaller FR separates C. borkenti sp. nov. females from those of C. virginianus, C. sp. F1 and C. sp. F3, though this is based on a very small sample of material. Wing length divides the females of this genus into two groups, those with wings >1.5 mm (C. longipennis and C. pacificus) and the remainder of the species (whose wing lengths fall between 0.9 and 1.5 mm). There is some intraspecific variation in wing length, with for example C. pacificus female wings falling between 1.5 and 1.8 mm or C. virginianus wings measuring between 0.9 and 1.3 mm. Spermathecal length differs in three general ranges, with those >80 μm (C. confusus, C. moravicus and C. pacificus), those 70-80 μm (C. aliciae and C. sp. F2) and the remaining species with spermathecae <70 μm. Finally, the shape of the 9 th sternite varies in the thickness and curvature of the dorsal and anterior branches . While the configuration found in C. grogani sp. nov. (Fig. 11e) is unique and recognizable due to its reduction of the posterior branch, examination of collection series for females confidently associated with the other species reveals there is significant variation in the thickness of both branches  and even the curvature of these branches may not be diagnostic. The apparent separation between the medial apices of the sternite also appears to be influenced by the method of preparation and amount of compression that occurs during slide mounting. For these reasons, the key I have presented for the females of the genus should be viewed circumspectly, as I had access to very limited numbers of individuals for several species and some of the characters used may not be diagnostic upon comparison with a broader sampling of material. I have also avoided naming several unassociated female specimens, though some undoubtebly represent new species, and instead give them provisional designations. Without a male association such names would certainly result in future taxonomic confusion and instability, such as seen with ongoing ambiguity regarding the identity of C. longipennis.
Associating male and female specimens in Ceratoculicoides has proven challenging. Of the informative characters only wing length (a correlate of general body size) is of any use in associations, and this is even quite tenuous. I have been forced to rely heavily on series with a single morphotype of males European Journal of Taxonomy 875: 159-202 (2023) and females collected at the same locality (the syntopic method sensu Hogue & Bedoya-Ortiz 1989), which is not completely reliable. Future studies with additional material are badly needed to clarify the characters for distinguishing females and may challenge the associations inferred herein. Molecular barcoding holds promise for solidifying the female taxonomy of Ceratoculicoides, as it has proven extremely useful in associating life stages in other groups (e.g., Stur & Ekrem 2011). (1), present (Fig. 10c, h)

Phylogeny
Though the phylogenetic structure of the Ceratopogonini is poorly resolved (Borkent 2014), Borkent (1995) placed Ceratoculicoides in a clade consisting of Brachypogon, Nannohelea Grogan & Wirth, 1980, Rhynchohelea Wirth & Blanton, 1970and Sinhalohelea Grogan & Borkent, 1992 based on the presence of synapomorphies consisting of setae on the katepisternum, with fusion of at least some of the male flagellomeres additionally uniting all these genera except Sinhalohelea. In describing Ceratoculicoides, Wirth & Ratanaworabhan (1971) noted the distinctiveness of the C-shaped sternite 9, suggested by A. Borkent as a synapomorphy of the genus (pers. comm.). Borkent (1992) proposed the presence of setae on the posterior margin of their anepisternum as another synapomorphy, noting it was unique within the Ceratopogonini. While some Palpomyiini and "Sphaeromyiini" s.l. (Borkent 2014) have setae on their anepisternum, they are more broadly distributed on the sclerite, indicating these are not homologous (Borkent 1992). Delécolle & Schiegg (1998) observed the presence of unique dorsocentral punctations on the posterior portion of the scutum in female Ceratoculicoides, which A. Borkent (pers. comm.) has hypothesized represents a third synapomorphy of the genus because it is unique in the Culicomorpha. The function of the punctations is unknown, though they may represent sensilla or pores (this merits histological study). The presence/absence and configuration of dorsocentral punctations in the males may have some phylogenetic signal, but this character has been difficult to interpret on slide mounted specimens (for some species, I had only one or two individuals to examine and typical lateral thoracic mounts obscure some of the dorsal surface of the scutum); thus, it was not included in this analysis. The morphological phylogeny produced in this study recovered a monophyletic Ceratoculicoides supported by four unambiguous synapomorphies: the aforementioned anepisternal setae (1;1), dorsocentral punctations on the scutum of the female (2;1), C-shaped sternite 9 of the female genitalia (3;1) and the presence of posterolateral points of the aedeagus (9;1,2), a feature noted by Wirth & Grogan (1988).
Interspecific relationships within Ceratoculicoides are well resolved. The best supported of these clades was a polytomy consisting of C. borkenti sp. nov. + C. grogani sp. nov. + (C. aliciae + C. propinquus sp. nov.+ (C. sp. M1 + C. moravicus)) , hereafter referred to as the "C. moravicus group". These FASBENDER A., Revision of New World Ceratoculicoides species are united primarily by possessing concave or straight lateral margins of the aedeagus (6;1), though this character state is shared with Brachypogon sp. L. I interpret this similarity to Brachpogon sp. L to be independently derived, as the other two outgroups have convex lateral margins and the C. moravicus group nests well within Ceratoculicoides, whereas all other species (including its sister taxa) have convex margins. Within the C. moravicus group, C. sp. M1 and C. moravicus are recovered as sister species based on the synapomorphic presence of accessory points between the apical and posterolateral points of the aedeagus (12;1); these species are included in a polytomy with C. aliciae and C. propinquus based on the synapomorphy of a rounded posterolateral point (10;1). The structure of all four species' male genitalia is extremely similar, differing only in the aforementioned accessory spines, the apical narrowing of the aedeagus in C. aliciae and C. moravicus and the presence of an acute apical notch along the posterior aedeagal margin in C. aliciae. The lack of narrowing towards the apex of the aedeagus in C. propinquus and C. sp. M1 may be a result of compression and distortion of the genitalia during slide mounting; see the taxonomic note for C. sp. M1 above regarding potential synonymy of that species with C. moravicus. Ceratoculicoides tontoeguri is supported as sister to the C. moravicus group based on the unambiguous synapomorphy of the posterolateral points of the aedeagus being directed laterally (9;2). Aditionally, the presence of a hyaline incision at the medial apex of the aedeagus (8;1) is shared between the C. moravicus group + C. tontoeguri, C. pacificus and Brachypogon sp. L, demonstrating the need for further outgroup comparisons. C. virginianus is weakly supported as sister to the C. moravicus group + C. tontoeguri by the straight or concave posterior margin of the aedeagus (7;1, also found in Brachypogon sp. L) and contiguous apical and posterolateral points (11;2). The final clade recovered was a sister group relationship between C. confusus and C. pacificus, based solely on the major spermathecae length >80μm (5;1). This is the same character that led Wirth & Ratanaworabhan (1971) to incorrectly associate the male of C. confusus with C. longipennis (see taxonomic notes for those species). A large spermatheca is also found in C. moravicus and C. tontoeguri, meaning this character state has independently arisen at least twice, therefore weakening its strength as a synapomorphy. In addition, a distinctively large spermatheca is present in numbers of other genera of Ceratopogonidae, suggesting this feature is susceptible to homoplasy.
The ground plan of several characters can be interpreted based on their distribution in the present phylogeny. Most Ceratoculicoides have two major spermathecae (4;0), including C. tontoeguri (Delécolle & Schiegg 1998, contra Havelka 1980), but C. moravicus has a single major spermatheca (4;1). Though Brachypogon canadensis also has a single major spermatheca, the other outgroup taxa combined with the distribution of this character within Ceratoculicoides indicates that two major spermathecae are the ground plan for the genus. Convex lateral (6;0) and posterior (7;0) margins of the aedeagus are another feature of the ground plan, being found in S. gansi and B. canadensis, though Brachypogon sp. L has the derived character state. Character 9, the posterolateral point of the aedeagus, can be separated into a type that emerges from the dorsal surface of the aedeagus (9;1) and another that emerges from the subapical lateral margin of the aedeagus (9;2). As the latter is a synapomorphy for the C. virginianus + C. moravicus group clade, the former state appears to be a symplesiomorphy within Ceratoculicoides. All Ceratoculicoides except C. sp. M1, C. aliciae, C. moravicus, C. propinquus sp. nov. and C. sp. M1 have a spiniform posterolateral point (10;0), another symplesiomorphy within the genus. The presence of apical points of the aedeagus appears to be part of the ground plan of Ceratoculicoides, as they are found in all species except C. confusus. In C. pacificus these points are well separated from the posterolateral points (11;1), but in the C. moravicus group and C. virginianus these points are adjacent and partially fused, typically with a concave arc forming their separation (11:2); it is unclear which of these character states is apomorphic. Brachypogon sp. L has similar apical points at the median apex of the aedeagus, and it is currently uncertain whether these are homologous with those in Ceratoculicoides.
European Journal of Taxonomy 875: 159-202 (2023) Since Brachypogon sp. L shares several character states with certain Ceratoculicoides, the morphological dataset was reanalyzed with that taxon removed. The same ingroup tree topology was recovered, though the Bremer support values for the branch supporting Ceratoculicoides improved to four. Brachypogon is extremely speciose and varied in morphology, and an understanding of the diversity and phylogeny of that genus will be critical to interpreting the evolution of related genera. Similarly, comparison of material of Nannohelea and Rhynchohelea should also improve our comprehension of the relationships within Ceratoculicoides.  (Wirth, 1952); open square = C. pacificus sp. nov.; solid triangle = C. propinquus sp. nov.; solid cross C. sp. M1. b. Eastern North America. Solid circles = C. confusus sp. nov.; open squares = C. virginianus (Wirth, 1951). c. Central and northern South America. Solid circle = C. confusus; open squares = C. sp. F1; open triangles = C. sp. F2; solid triangle = C. sp. F3.

FASBENDER A., Revision of New World Ceratoculicoides
There is a single fossil species referred to Ceratoculicoides from ~ 50 mya Baltic amber, Ceratoculicoides danicus Szadziewski, 1988. Known only from females, it has the combination of raptorial fore and midleg tarsal claws with small claws on the hindleg that characterize the genus but differs in a number of features from extant species of Ceratoculicoides. These include fourth and fifth segments of the maxillary palpus fused (vs separate), strongly unequal tarsal claws (vs equal or slightly unequal), and a large r 2 cell (vs reduced or absent) (Szadziewski 1988). The presence of anepisternal setae, dorsocentral punctations on the scutum or the structure of the 8 th sternite are not noted in the original description, making this species placement in Ceratoculicoides somewhat challenging to assess. Absent additional evidence, I suggest retaining C. danicus in Ceratoculicoides based on the tarsal claw character, though it does not conform to several other features of the extant fauna.

Bionomics and distribution
The natural history and ecology of Ceratoculicoides remains poorly understood. I have personally collected only C. borkenti and some unassociated females from a handful of sites in the Bitterroot Mountains near the Idaho-Montana border (USA). Each instance was near water: once along a series of beaver ponds, twice along streams approximately 6-10 m wide with shallow side channels. All specimens were collected by sweeping riparian vegetation, though specific adult microhabitats were not noted, as Ceratoculicoides cannot be reliably differentiated in the field without microscopy. Each incident resulted in only a single specimen collected, and most of the other material I have examined also consists of a handful of specimens from an individual collection event. An exception is several large series of C. pacificus that were collected by Neville Winchester in the temperate rainforest of Vancouver Island (Canada). The relatively coarse mandibular teeth and elongate, raptorial tarsal claws on the fore and midlegs of adult female Ceratoculicoides suggest they are aerial predators of small flying insects like many other members of Ceratopogonini (Borkent 1995). The immature stages and larval microhabitat remain unknown, as is the case for over half of the genera in Ceratopogonidae ( Borkent 2014).  Wirth & Ratanaworabhan, 1971 (unsupported branches collapsed). Bremer support values are noted above and character state changes (given as character # (state)) below branches. * Bremer support value increased to 4 in a reanalysis removing Brachypogon sp. L from dataset (see Discussion). 875: 159-202 (2023) Our current understanding of Nearctic species distributions must be considered preliminary, with several species (C. grogani sp. nov., C. longipennis, C. propinquus sp. nov. and C. sp. M1) known only from one or two localities, and two species (C. propinquus and C. confusus) with disjunctions of over 1200 km between populations. Ceratoculicoides borkenti, though known from more specimens and localities, exemplifies the patchy distribution of records within the genus. That species has been recorded from three distict ecoregions: the Desert Southwest of Arizona and California, the Northern Rocky Mountains of Idaho and Montana and the temperate rainforest of Vancouver Island. Separating these clusters are swathes of apparently suitable habitat spanning 700-1300 km, suggesting additional populations of C. borkenti sp. nov. await discovery. As the species fauna of the eastern and western Nearctic is disjunct for the genus, sampling in the center of the continent is of interest to establish the range limits of existing taxa and may uncover additional undescribed species. There are near total gaps in records for the genus across the Great Basin, Middle Rocky Mountains, Great Plains, Prairie Peninsula, Ozarks, and the Gulf Coast.

European Journal of Taxonomy
Our knowledge of Ceratoculicoides in the Neotropics remains extremely preliminary. For example, C. aliciae remains known only from the type series, while males have not been located for female material from Costa Rica (Huerta & Borkent 2005;) and whether it represents a distinct species remains obscure. The discovery of a C. confusus male in Colombia expands the range of the genus into the northern part of South America, where there is a paucity of knowledge on the ceratopogonid fauna (Borkent et al. 2018). Two females have been collected from a high elevation (> 4000 m) at the Páramo site within Sierra Nevada de Santa Marta National Park in northeast Colombia, though like the Costa Rican material, there are no associated males to illuminate their identity. Further sampling in Central and South America will undoubtably uncover additional populations and probably new species of Ceratoculicoides. As in the Nearctic, essentially nothing is known of the natural history of the Neotropical fauna for this genus. The report of female Ceratoculicoides specimens from North Korea by Szadziewski (1988) indicates that there is also an Asian fauna that remains totally unexamined.

Conclusion
The New World fauna of Ceratoculicoides has been expanded from four named species to eight, with five new species (C. borkenti sp. nov., C. confusus sp. nov., C. grogani sp. nov., C. pacificus sp. nov. and C. propinquus sp. nov. ). Ceratoculicoides confusus represents the first record of the genus from South America. Ceratoculicoides blantoni was synonymized with C. virginianus due to the original characters separating the species not proving to be diagnostic based on the examination of a broad range of material. Phylogenetic analysis supports the monophyly of the genus and the recognition of the C. moravicus species group. The natural history of the genus remains obscure, with the immature stages unknown and adult habits speculatively inferred.