Ooderella Ashmead

Gibson, Gary A. P., 2017, Revision of world Ooderella Ashmead (Hymenoptera: Chalcidoidea: Eupelmidae), with description of the first males for the genus, Zootaxa 4289 (1), pp. 1-74 : 5-13

Ooderella Ashmead

Ooderella Ashmead, 1896: 6 (key), 10–11 (description). Type species: Ooderella smithii Ashmead , by monotypy and original designation.

Diagnosis. FEMALE. Brachypterous, with fore wing usually extending flat over mesosoma to propodeum or base of gaster, though sometimes bent upwards near posterior of mesosoma ( Figs 14, 15 ); membrane behind venation bare (e.g. Fig. 52 ) to variably densely setose (e.g. Fig. 7 ) but with hairlike setae and without hyaline cross-band or anterior and posterior spots with white setae. Head in lateral view variably highly convex (cf. Figs 11 , 46 , 91 ); sometimes with ocellocular mark between posterior ocellus and inner orbit (e.g. Figs 47 , 157 ), and sometimes with lighter-colored line or mediolongitudinal sulcus below anterior ocellus on one or both of frons and scrobal depression (e.g. Figs 156 , 165 ). Eye superficially bare but very sparsely microsetose. Antenna with fl1 short compared to fl2, but at least about as long as wide rather than strongly discoidal (e.g. Figs 5 , 13 ). Mandibles with acute tooth ventrally but apical margin above ventral tooth variably broadly truncate to deeply incurved or angularly incised so mandible variably distinctly bidentate (e.g. Fig. 23 : insert) to tridentate (e.g. Fig. 5 : insert). Pronotal collar transverse to somewhat longer than wide but more or less quadrangular with lateral margins parallel or only slightly anteriorly divergent, with anterolateral corners right-angled (e.g. Figs 3, 6 ) to more strongly, conelike produced (e.g. Figs 120, 124 ). Anterior margin of collar abruptly angled relative to inclined neck at least laterally, and dorsally collar divided mediolongitudinally by deep furrow and usually flat even if sometimes inclined on either side of furrow. Mesoscutum with smooth and shiny mesoscutal flange laterally, but otherwise meshlike reticulate at least anterior of posteriorly carinate portion of lateral lobes (except O. platyscapus where more transversely reticulate-strigose to reticulate-rugulose, Fig. 201 ); scutellar-axillar complex with scutellum extending to and acutely angled at transscutal articulation. Mesotibia with oblique groove apically between tibial spur and base of tarsus, and with row of dark apical pegs in region over base of tibial spur; mesotarsus with a single row of dark pegs along either side of basal four tarsomeres. Propodeum highly variable in structure. Gaster yellowish to dark brown, though often variably paler to white basally or subbasally; Gt1 distinctly incised medially but subsequent tergites more shallowly emarginate; syntergum narrowed posteriorly to often lighter-colored and/or more translucent and sometimes variably reflexed, posteriorly rounded or angulate bare region (syntergal flange) behind setae (e.g. Figs 9 , 18 , 204 ) (except O. ambigua , see Discussion).

MALE. New World. Head often with ocellocular mark evident as an angulation (e.g. Fig. 82 ) but only rarely as a lighter-colored line (e.g. Fig. 206 ) between anterior margin of posterior ocellus and inner orbit; scrobal depression often transversely Π-like above interantennal prominence (e.g. Figs 81 , 101 ), variably sculptured relative to smoother and shinier frons, and usually with mediolongitudinal light-colored sulcus or line of weakness (e.g. Figs 101 , 205 ), but not a circular depression or pit at apex of interantennal prominence (see Gibson 1995, figs 406–408); with uniformly short and sparse setae on gena lateral to malar sulcus. Antenna with torulus about midway between oral margin and anterior ocellus, the ventral margin obviously above lower orbit and distance between ventral margin of torulus and oral margin slightly greater than distance between dorsal margin of torulus and ventral margin of anterior ocellus (e.g. Fig. 101 ); scape compressed and broadest subbasally to medially, but variably distinctly narrowed apically; pedicel obviously longer than apical width, subtriangular, without row of differentiated setae along ventral length; flagellum conspicuously elongate-slender, filiform with fl1 (anellus) discoidal; funiculars much longer than wide with fl2 about 4× as long as wide and straight (ventral margin not curved) and subsequent funiculars of similar length or only slightly shorter apically, with sparse and comparatively inconspicuous, very short setae much shorter than width of funicle, and either with multiple overlapping rows of longitudinal, convex mps (multiporous plate sensilla) ( Figs 85 , 104, 105 ) (= Reikosiella group 3 sensu Gibson 1995, fig. 370) or with each mps having long, apically free, seta-like portion so mps superficially appear as comparatively dense setae subappressed to or projecting from surface at acute angle ( Figs 131, 132 ) (= Reikosiella group 4 sensu Gibson 1995, fig. 371); clava as slender as preceding flagellomeres and tapered to apex without distinct micropilose sensory region ventrally, at most only very obscurely differentiated into clavomeres by indistinct transverse sutures, and about as long as combined length of apical two funiculars. Mandibles variably distinctly tridentate. Fore wing hyaline or at most slightly infumate ( Fig. 109 ); not distinctively elongate-slender, at most length only about 2.7× maximum width; dorsally with variably large and distinct speculum behind parastigma beyond basal cell ( Fig. 86 ), but region often obscured by setae on ventral surface; marginal vein not distinctively long, less than 0.8× width of wing; postmarginal vein obviously longer than stigmal vein, though apex often difficult to distinguish accurately because vein gradually lightened apically. Mesoscutum with deep, distinct notauli; usually quite distinctly meshlike reticulate over at least anterior half (e.g. Figs 87, 88 ), though sometimes smoother, more coriaceous posteriorly, and small males sometimes almost entirely meshlike coriaceous-reticulate. Metapleuron bare. Propodeum with median carina.

Old World. Males of a single identified species differ from known New World males by the following: torulus obviously closer to oral margin than to anterior ocellus, with ventral margin at level of lower orbit ( Fig. 205 ); scrobal depression with light-colored line of weakness mediolongitudinally, though more elongate Π-like, extending almost to anterior ocellus ( Fig. 205 ); flagellum with fl2–fl8 obviously longer than wide and sparsely and inconspicuously setose, but at most only about 2× as long as wide and with elongate mps extending almost entire length of funicular ( Figs 208, 209 ); clava about equal to combined length of apical three funiculars and with much longer setae than preceding flagellomeres ( Fig. 208 : insert).

Hosts. Hosts are known only for O. americana , which was reared once from eggs of an unidentified species of Orchelimum Serville ( Orthoptera : Tettigoniidae ) (greater meadow katydids) on Daucus carota L. ( Apiaceae ) (wild carrot).

Diversity and distribution. Ooderella is restricted to the Neotropical and Afrotropical regions (Maps 1, 2) except for a single species, O. americana , in northeastern USA (Map 1A), though it is possible that the northern limits of at least one Neotropical species may extend into southeastern USA (see under O. setosa ). Within the Neotropical region , species are known from southernmost Mexico ( Fig. 2 A) through Central and South America , but not the islands of the West Indies other than Trinidad (Maps 1B–H, 2A–C). Of the 12 species recorded from the Neotropical region , only 3 are represented by more than 10 specimens . A total of 7 of the 12 species, 5 uniquely, are known from Costa Rica. This comparatively large number of species from Costa Rica undoubtedly reflects, at least in part, the greater collecting efforts in Costa Rica in modern times and the ready availability of specimens. Within the Afrotropical region, species are currently known only from south of about 10°S (Map 2D), though this likely is an artefact of collecting and specimen availability. Four of the five species recorded from the Afrotropical region are known from less than ten specimens and only one, O. botswanae , is recorded from more than one country. It therefore seems inevitable that more extensive collecting in both the Neotropical and Afrotropical regions will not only enlarge the known ranges of the currently described species, but result in other, if not many other, new species.

Generic limits and species relationships. FEMALES. Gibson (1995) provided a comparative description of Ooderella relative to other eupelmine genera based on females, but this did not include features of what is here described as O. ambigua . This newly described species is classified in Ooderella based primarily on the combination of female brachyptery and pronotal structure, though the pronotal collar is unusual in having the anterior margin most distinctly angulate or protuberant anteromedially on either side of the mediolongitudinal furrow ( Figs 12, 15 ) rather than anterolaterally (e.g. Figs 3 , 21 , 47 ). As such, the dorsal surface is shallowly concave ( Fig. 15 ), whereas in other species it is flat, even if sometimes it is inclined on either side of the furrow. Except for the absence of a transverse line of dark setae on the ridge that delineates the collar from the neck, the pronotal structure of O. ambigua is quite similar to that of typical Eupelmus ( Macroneura Dalman ) females, which was described as character state 15(4) by Gibson (1995). Other Ooderella females have a pronotal structure described as character state 15(2), and as either substate 2a or 2b depending on whether the anterolateral corners are right-angled (e.g. Figs 3 , 21 ) or more cone-like protuberant (e.g. Figs 47 , 120 ), respectively. These different pronotal structures were hypothesized to have been derived from a strongly transverse pronotum without a distinctly differentiated collar and neck (character 15, state 1), which is possessed by females of many eupelmine genera, including Brasema Cameron (see figures in Gibson 1995).

Syntergal structure is even more unusual for O. ambigua within Ooderella . Except for this species, females of Ooderella have the syntergum tapered into a posteriorly rounded or angulate region that often is lighter in color and/or translucent, but at least is bare beyond subapical setae whose apices extend only slightly beyond the posterior margin (e.g. Figs 100 , 164 , 193 ). The apical region is sometimes also variably abruptly recurved from a more convex or inclined basal region into a narrower and/or horizontal syntergal flange sensu Gibson (1995) (e.g. Figs 164 , 184 , 204 ). This type of syntergum (character 39, state 3) was stated as diagnostic for Ooderella and 16 other eupelmine genera that were hypothesized to form a monophyletic clade based on this shared structure ( Gibson 1995, fig. 410). The syntergum of O. ambigua differs by being uniformly dark and evenly tapered in one plane to a transverse posterior margin, and with a row of dark setae along the posterior margin that project conspicuously beyond the margin ( Figs 14, 18 ). This latter structure (character 39, state 1) and setal pattern were hypothesized as the groundplan states of the syntergum for Eupelminae by Gibson (2017), and the syntergal structures and setal patterns of other Ooderella as derived states that evolved through secondary development of the cuticle behind the apical-most line of setae characteristic of character state 39(1). The selective pressure for evolving and the functional advantage of having a flanged syntergum are unknown, as is the number of times similar structures may have evolved convergently. However, the putatively plesiomorphic non-flanged structure is possessed by females of some species of several eupelmine genera ( Gibson 1995, table 1), including those of Brasema (see figures in Gibson 1995), a genus that is by far most highly speciose and diverse in the New World. This latter genus is not defined by any autapomorphies and very possibly is paraphyletic relative to some other recognized genera such as Zaischnopsis Ashmead and possibly Anastatus Motschulsky ( Gibson 1995, 2016).

Females of O. ambigua also have the least reduced wings within the genus. The fore wing ( Figs 14, 15 ) is right-angle bent near the posterior of the mesosoma, with the basal portion consisting of a setose basal cell behind the costal cell, and the apical portion beyond the bend constituting the setose disc. The wing also retains the typical chalcid vein complex consisting of a submarginal, long marginal, stigmal, and postmarginal vein. The fore wings of other species of Ooderella extend at most only slightly beyond the base of the gaster and the venation either is reduced to just the submarginal vein or there are at most short remnants of the other veins (see further below). Similarly reduced fore wings to those of O. ambigua are possessed by females of some other eupelmine genera, including many Eupelmus (Macroneura) and at least one undescribed Neotropical species of Brasema . Interestingly, both types of fore wing rudiments occur also in different species of E. ( Macroneura ), some having the fore wings abruptly bent upwards near the posterior of the mesosoma as in O. ambigua and some having more severely reduced wings that extend flat over the mesosoma as for females of other species of Ooderella ( Gibson 1990; Fusu 2017).

Compared to some other species of Ooderella , females of O. ambigua also have relatively plesiomorphic head structures, including the head in lateral view being comparatively lowly convex with the anterior surface evenly curved ( Fig. 11 ), and in frontal view the scrobal depression being bell-shaped because of slightly incurved lateral margins that distinguish distinct parascrobal regions ( Fig. 10 ) (character 6, state 1). The frontovertex also lacks ocellocular marks (character 9, state 1), even though the frons is comparatively smooth and shiny with only fine coriaceous sculpture ( Figs 10, 12 ). Finally, the antenna has a tubular, uniformly thick, slightly curved scape (character 5, state 1) and a comparatively short-clavate flagellum ( Fig. 13 ) as compared to most other Ooderella . The acropleuron is also completely bare ( Fig. 17 ) and the mesoscutum lacks distinctly denser regions of white lanceolate setae posteriorly or anteriorly ( Fig. 15 ). All of these features are hypothesized as likely plesiomorphic compared to other states exhibited by females of some other Ooderella species. Of uncertain polarity is that in O. ambigua the mesoscutal lateral lobes are carinately angled along their entire length ( Fig. 15 ). Except for O.

platyscapus , other Ooderella have the mesoscutal lateral lobes carinately angled only within about the posterior third or less. Based on its atypical pronotal and syntergal structures as well as the combination of other putatively plesiomorphic features, O. ambigua could possibly represent an unusually modified brachypterous species of Brasema having a secondarily modified pronotal structure that is only superficially similar (convergent) to that characteristic of Ooderella , and thus be incorrectly classified to genus. However, propodeal structure in O. ambigua does not support such a hypothesis. Females of Brasema have a variably long but broad, flat to lowly convex plical region that is differentiated from the callar regions by comparatively obscure subparallel furrows closer to the spiracles than to the midline (character 27, state 2; Gibson 1995, figs 217–221), which was hypothesized as similar to the groundplan structure of the subfamily. Propodeal structure of O. ambigua is quite different, the propodeum having quite deep, anteriorly convergent depressions along the foramen that differentiate conspicuously convex callar regions from a medially inclined, short and narrow, somewhat bowtie-like plical region ( Fig. 16 ) (character 27, state 4). Although propodeal structures of other Ooderella are quite diverse, the structures are variably similar to, and could be derived from, an O. ambigua -like structure (see further below),

Even excluding O. ambigua , Ooderella is unusually diverse in some other features. Both sexes of most eupelmine genera usually have mandibles that can be assigned quite readily to one of two structures—three similar teeth (tridentate: character 1, state 1) or with a single ventroapical tooth and a broad, truncate to shallowly concave dorsoapical margin (bidentate: character 1, state 2a). Different species or even different females of the same species of Ooderella exhibit variably distinctly bidentate to tridentate mandibles, the difference resulting from how deeply incurved is the apical margin above the acute ventral tooth and whether the emargination is evenly incurved (e.g. Figs 23 , 35 ) or angularly incised to differentiate two more acutely angled teeth (e.g. Fig. 5 ). Mandibular structure is not clearly visible in most available females of O. ambigua , but there appears to be only one distinct tooth ventroapically plus a broad, truncate margin dorsoapically that at least sometimes is obscurely differentiated into two blunt teeth by a slight depression or transverse furrow ( Fig. 13 : insert). Tridentate mandibles were hypothesized as the groundplan structure for Eupelminae by Gibson (1995) and bidentate mandibles as apomorphic, and possibly correlated with egg parasitoidism ( Gibson 1995, 2016). Mandibular structure is clearly visible in only one female of O. americana , a putative egg parasitoid. Each mandible is quite distinctly tridentate because the emargination dorsal to the ventral tooth is angular ( Fig. 5 ). However, additional females with exposed mandibles are needed to determine whether there is intraspecific variation, similar to some other Ooderella species.

Ooderella is also unusual in that females of some species, including O. ambigua , have the gaster uniformly dark ( Figs 14, 18 ) (character 42, state 1), whereas in others it is variably paler to white dorsally and/or ventrally, most commonly subbasally (character 42, state 2). This latter, derived state is perhaps most characteristic of females of Anastatus , although it is also known for at least some species of several other genera ( Gibson 1995, table 1), particularly species with banded fore wings. Smith (1969) provided line drawings and commented on the similarity between females of Tachydromia lilaniensis Smith ( Diptera : Empididae ) and E. (Eupelmus) orthopterae (Risbec) , both of which have banded fore wings and are found in the same habitat. He suggested mimicry for the similarly banded wings, stating that the “egg-hunting Chalcid and the prey-hunting Empid would tend to exhibit a similar behavioural pattern and the banded wings, folded over the back would protect both in breaking up the insect outline which in their rather vulnerable position on lichen-covered boulders would give protection from would-be predators” ( Smith 1969: 75). Both the abdomen of the empid and the gaster of E. orthopterae are entirely dark. However, a white region basally to subbasally on the gaster of species with banded wings likely further visually emphasizes ‘breaking up’ of the habitus into two apparent body regions, regardless of how the wings are held at rest. Such a light-colored region appears to be prone to evolve also in brachypterous females. For example, the brachypterous females of E. ( Macroneura ) much more commonly have the gaster white ventrobasally ( Fusu 2017) than the macropterous females of E. ( Eupelmus ) ( Gibson & Fusu 2016), as do brachypterous compared to macropterous females of Brasema (personal observation). This disruptive color pattern in brachypterous females could also provide some protection against predators while exposed on plants or other substrates.

Ooderella species also have quite diverse female propodeal structures, although as noted above at least all New World species share structures that might be derived from an O. ambigua -like propodeum. Unstated in the above description of the propodeum of O. ambigua is that the posterior margin is carinate laterally, but near the midline each carina curves anteriorly, sinuately, to form a short median carina anteriorly. However, there is a triangular region of lighter-colored cuticle between the carina on either side behind the median carina so that the posterior margin is evenly, Π-like incurved ( Fig. 16 ). Females of the other New World species also have the posterior margin of the propodeum sinuately incurved on either side of the midline to form variably long contiguous margins or a median carina, but without any intervening cuticle posteriorly (e.g. Figs 7 , 60 , 97 ). Females of all Afrotropical species have the posterior propodeal margin evenly incurved, and in O. gymnosoma ( Fig. 183 ) and O. platyscapus ( Fig. 202 ) there is a variably distinct ^-like region of differentiated cuticle posteromedially similar to the posteromedial propodeal region of O. ambigua . Two of the five Afrotropical species ( O. botswanae and O. capensis ) have the foramen incurved almost to the medially emarginate anterior margin such that the plical region is transverse-lunate ( Figs 162 , 171 ), whereas two others ( O. gymnosoma and O. platyscapus ) have a somewhat longer but still relatively short plical region ( Figs 183 , 202 ), and O. kenyaensis has an unusually long propodeum with a median carina ( Fig. 190 ).

At least one feature could support all New World species except O. ambigua as a monophyletic lineage within Ooderella —the presence of dense white setae anteriorly on the acropleuron and usually slightly on the mesopectus anterior of the acropleural sulcus. Females of O. ambigua ( Fig. 17 ) and all Afrotropical species have the acropleuron bare posterior to the extensively setose mesopectus, whereas all New World species have setae anteriorly on the acropleuron. The setae are comparatively sparse and inconspicuous in O. flavida ( Fig. 26 ), but form a dense tuft posterior and/or dorsal to sparser white setae on the mesopectus in other New World species (e.g. Figs 8 , 41 , 51 ). A second feature that could support monophyly of all New World species except O. ambigua is the presence of a differentiated region of white lanceolate setae on the mesoscutum submedially to posteromedially. The mesoscutum of females of O. ambigua has mostly inconspicuous, posteriorly directed, hairlike setae except for some anteromedially directed, slender-lanceolate white setae over about the lower half of the inclined inner surfaces of the lateral lobes ( Fig. 15 ). Females of O. flavida have the mesoscutum inconspicuously setose with posteriorly directed hairlike setae except for a few posteriorly directed white lanceolate setae posteromedially ( Fig. 24 ). All other New World species have much more conspicuous, variably elongate-ovate regions of dense, lanceolate white setae posteromedially (e.g. Figs 6 , 33 , 40 ). Consequently, both the acropleural and mesoscutal setal patterns could support O. ambigua as the sister group of O. flavida + other New World Ooderella , with the setal patterns of O. ambigua representing the groundplan states and those of O. flavida intermediate states in development of denser, more conspicuous tufts of setae on the acropleuron and mesoscutum in other New World species. Further, all New World species have the mesoscutal setal tufts composed of setae that are in a ‘swirled’ pattern, i.e. consisting of anteriorly directed setae anteriorly, at least a few posteriorly directed setae posteriorly, and with a few medially directed setae medially (e.g. Figs 6 , 33 , 40 ) except for O. melanosceles ( Fig. 50 ), O. spinositegula ( Fig. 124 ) and O. thegalea ( Fig. 152 ) in which all the setae are directed posteromedially. This latter setal pattern may support the monophyly of these latter three species (see further below).

In addition to having white lanceolate setae posteromedially, several New World species have medially or posteromedially directed white lanceolate setae anteriorly on the mesoscutum on either side of the midline so as to form two, variably closely contiguous setal tufts (e.g. Figs 50 , 77 , 114 ). Although females of all Afrotropical species lack as dense regions of lanceolate setae on the mesoscutum as New World species, some do have somewhat denser regions of lanceolate setae anteriorly and posteriorly. The mesoscutum is extensively setose with posteriorly directed hairlike setae in O. gymnosoma ( Fig. 180 ) and O. platyscapus ( Fig. 201 ). Females of O. kenyaensis have a similar setal pattern except posteriorly some setae are directed anteriorly ( Figs 188, 189 ), whereas both O. botswanae ( Fig. 161 ) and O. capensis ( Fig. 170 ) have noticeably denser regions of white lanceolate setae both posteriorly and anteriorly, with the anterior setae directed posteriorly and uniformly distributed across the sclerite and the posterior setae directed anteriorly. The differences between the setal patterns in New World and Afrotropical species suggest that denser regions of lanceolate setae anteriorly and posteriorly on the mesoscutum evolved independently in the New World and Afrotropical region, though the selective pressure or adaptive function of such setal regions remains unknown. The groundplan state and polarity of the setal patterns in the Old World species are also uncertain, but the shared patterns of O. gymnosoma and O. platyscapus , and O. botswanae and O. capensis , could support these as species pairs. Further, among New World species, presence of posteromedially to medially directed white setae anteriorly on the mesoscutum could support monophyly of O. melanosceles + O. setosa + O. speculifrons + O. spinositegula + O. stenoptera + O. thegalea , and, possibly, O. americana and O. smithii . Females of O. americana differ from the first six species by having only one or two differentiated white setae that are widely separated on either side of the mesoscutal medial lobe ( Fig. 6 ). Females of O. smithii sometimes have several such posteromedially directed setae anteriorly on the mesoscutum, but even when most numerous they do not form as conspicuous regions of differentiated setae as for the first six species, and usually are not evident without correct angle of lighting ( Figs 95, 96 ). It is possible that the setal patterns of both O. americana and O. smithii represent initial stages in the evolution of more distinct setal tufts. However, the setal pattern of O. americana might more likely be indicated as independently derived from those of the other seven species based on females having a different head structure and sculpture pattern relative to the other species. Females of O. americana have a comparatively lowly convex, lenticular head ( Figs 3, 4 ) and a sculptured frons that curves smoothly into the scrobal depression ( Fig. 1 ) similar to that of O. ambigua ( Figs 10, 11 ) (see further below). Females of O. melanosceles ( Fig. 47 ), O. microptera ( Fig. 55 ), O. setosa ( Fig. 74 ), O. smithii ( Fig. 92 ), O. speculifrons ( Fig. 111 ), O. spinositegula ( Fig. 120 ), O. stenoptera ( Fig. 138 ), and O. thegalea ( Fig. 149 ) have a variably more highly convex head in combination with a smooth and shiny frons that is abruptly differentiated from a strongly sculptured scrobal depression by at least the sculptural difference (e.g. Figs 45 , 72 , 90 ). Consequently, head structure/sculpture pattern may support monophyly of these eight species. If so, apparent absence of differentiated setae anteriorly from the mesoscutum of the only known female of O. microptera and only obscurely differentiated setae in O. smithii could represent a transformation series indicating the following relationships: O. microptera + ( O. smithii + ( O. melanosceles + O. setosa + O. speculifrons + O. spinositegula + O. stenoptera + O. thegalea )).

As noted above, O. ambigua has the putative groundplan head states for Ooderella . Females of O. americana ( Fig. 1 ), O. flavida ( Fig. 19 ), O. hansoni ( Fig. 28 ), O. hyalipleura ( Fig. 35 ) and O. reticulifrons ( Fig. 63 ) all have the frons, at least in part, similarly sculptured as the scrobal depression, unlike the other eight New World species listed above with a smooth and shiny frons. The frons is entirely, similarly sculptured as the scrobal depression in all five of the above listed species except for O. hyalipleura , in which it is only partly sculptured in a longitudinal band below the median ocellus ( Fig. 35 ). Females of O. flavida also differ from the other species by having a more highly convex head in combination with a scrobal depression that is deep ventrally and shallowed dorsally toward the anterior ocellus. Further, the scrobal depression is abruptly divided by a transverse depression somewhat below mid-height, which differentiates below it a distinct interantennal prominence and separate scrobe above each torulus ( Fig. 19 ). This scrobal depression structure might represent an independent transformation from an O. ambigua -like structure or one that is intermediate in development of similar structures shared by O. americana ( Fig. 1 ), O. hansoni ( Fig. 28 ), O. hyalipleura ( Fig. 35 ) and O. reticulifrons ( Fig. 63 ). Females of the latter four species all have a strongly, transversely Π-shaped scrobal depression that has a variably distinct transverse line of demarcation separating the scrobes and interantennal prominence below it from a reticulate and strongly though poorly differentiated dorsal portion of the scrobal depression above it. The demarcation line is most noticeable in O. reticulifrons where it is evident as quite a distinct transverse furrow above each scrobe ( Fig. 63 ) similar to O. flavida ( Fig. 19 ), though this is less developed in the other three species ( Figs 1 , 28 , 35 ).

Excluding O. ambigua , females of O. americana , O. flavida , O. hansoni , O. hyalipleura and O. reticulifrons also have fore wings with the venation reduced to just the submarginal vein or at most the submarginal vein and a short remnant of the marginal/postmarginal veins apically (see Remarks section for O. flavida ), plus a costal cell and at most a very short membranous region beyond the costal cell. Available females of O. flavida appear to have the membrane behind the venation bare except for marginal setae in the apical half, whereas the membrane is sparsely setose in O. hansoni ( Fig. 32 ), O. hyalipleura ( Fig. 42 ) and O. reticulifrons ( Fig. 69 ) and densely setose in O. americana ( Fig. 7 ), but without any apparent discal region. Except for O. ambigua ( Figs 14, 18 ), females of the above species also have the gaster variably paler basally or subbasally dorsally and/or ventrally. Females of O. americana have the gaster paler basally but still brownish to yellowish-brown ( Figs 4, 9 ), whereas O. flavida uniquely has the gaster more or less unicolorous dorsally but almost entirely white ventrally ( Fig. 22 : insert). Females of O. hansoni ( Fig. 34 ), O. hyalipleura ( Figs 41, 43 ) and O. reticulifrons ( Figs 67, 71 ) have the gaster dorsally whitish subbasally as well as ventrobasally, which might support the monophyly of these three species, though O. speculifrons also has a gaster that is white ventrobasally ( Figs 112, 117 ). Females of O. hansoni ( Fig. 30 ), O. hyalipleura ( Fig. 41 ) and O. reticulifrons ( Fig. 67 ) also share long ovipositor sheaths, though the sheaths project quite conspicuously also in O. flavida ( Fig. 22 ). Females of the former three species also have a propodeum that is only comparatively sparsely setose with slender, more hairlike setae posterolaterally. Although a relative feature, other species, including O. ambigua ( Fig. 16 ), have more conspicuous, slender-lanceolate white setae that often form quite a dense region posterior to the spiracle.

Relationships among the group of eight New World species united by a smooth and shiny frons plus a comparatively highly convex head ( O. melanosceles , O. microptera , O. setosa , O. smithii , O. speculifrons , O.

spinositegula , O. stenoptera and O. thegalea ) remain largely unresolved. All share one putative symplesiomorphy—ovipositor sheaths that project only slightly beyond the syntergum, and except for O. speculifrons one other putative symplesiomorphy—an entirely dark gaster. All except for O. setosa ( Fig. 78 ) also share one putative apomorphy—a fore wing having the membrane behind the submarginal vein bare. As discussed above, differences in development of differentiated setae anteriorly on the mesoscutum could indicate O. microptera and O. smithii represent basal clades of the group. However, females of O. smithii , O. melanosceles , O. spinositegula and O. thegalea have slightly different head shapes than those of O. microptera , O. speculifrons and O. stenoptera . In lateral view, the first four species have a somewhat more highly convex head such that this is broadest dorsally ( Figs 46 , 91 , 119 , 148 ), whereas the last three species have the head broadest within the ventral half or close to the middle ( Figs 55 , 112 , 140 ). Regardless, O. melanosceles , O. spinositegula and O. thegalea almost certainly represent a monophyletic lineage. As noted above, this is supported by females of the three species having the posterior mesoscutal setal tuft composed entirely of posteromedially directed setae. It is also supported by females having more strongly, cone-like produced pronotal corners ( Figs 47 , 124 , 152 ) than those of the other species, an only coriaceous to coriaceous-imbricate rather than reticulate scrobal depression, and distinct tegular tufts of setae ( Figs 51 , 126 , 155 ), although females of O. americana also have quite distinct tegular setal tufts ( Figs 6, 8 ) and the setae are obviously denser along the inner margin posteriorly in some other species. Monophyly of the three species is likely further supported by their longer fore wings and ventrally bare costal cell. Unlike other New World Ooderella females with fore wings extending only to about the base of the gaster or less (i.e. excluding O. ambigua ), the submarginal vein extends at most about two-thirds the length of the wing and there is a comparatively long, narrower membranous region extending beyond the costal cell ( Figs 52 , 125 , 153 ). The absence of any vein remnant in this membranous portion beyond the submarginal vein, unlike for O. gymnosoma (see below), could indicate the apical membranous region is secondarily derived. If so, presence of the longest fore wings as well as quite distinct ocellocular marks in O. melanosceles ( Fig. 47 ) and at least larger O. spinositegula ( Fig. 120 ) females support a O. thegalea + ( O. melanosceles + O. spinositegula ) relationship, although females of O. smithii also have quite distinct ocellocular marks ( Fig. 92 ). Further, Afrotropical females have ocellocular marks whether or not the frons is smooth and shiny or distinctly sculptured ( Figs 157 , 167 , 176 , 186 , 196 ).

Monophyly of the five recognized Afrotropical species may be supported by females sharing a similar syntergal structure in which one or a few transverse rows of setae differentiate a bare, convexly inclined basal region from a more horizontal, bare syntergal flange ( Figs 164 , 184 , 193 , 204 ). All Afrotropical females also share a conspicuously highly convex head ( Figs 158 , 166 , 175 , 187 , 195 ) whether or not the frons is distinctly sculptured. This head structure must be convergent to that of some New World species if the New World species are monophyletic. Further, the gaster is variably paler subbasally, dorsally and/or laterally ( Figs 164 , 168 , 182 , 193 , 204 ) in all Afrotropical females except for some specimens of O. capensis . Again, similar subbasally pale gastral color patterns in some New World species must be convergent if the New World and Afrotropical species each represent monophyletic lineages. The ovipositor sheaths are short and the fore wings are setose in all Afrotropical species except O. gymnosoma ( Figs 177, 183 ), though the former two states likely are symplesiomorphic and the bare fore wing membrane and long ovipositor sheaths of O. gymnosoma likely convergent to similar states in some New World species. As noted above, the different mesoscutal setal patterns of females could support O. botswanae + O. capensis and O. gymnosoma + O. platyscapus as species pairs, though if the setal pattern of the last two species represent the groundplan state the different setal patterns would indicate the following relationship: O. gymnosoma and/or O. platyscapus + ( O. kenyaensis + ( O. botswanae + O. capensis )). The similarly reduced wing venation clearly not extending to the apical margin of the wing shared by O. botswanae and O. capensis females could further support their sister-group relationship. Females of the three other Afrotropical species have the venation extending to the apical wing margin with a variably distinctly differentiated apical portion possibly representing a remnant of the marginal/postmarginal veins or, in the case of O. gymnosoma , with the submarginal vein clearly not extending to the wing margin but with a detached remnant of the marginal/postmarginal veins still evident apically. Females of O. botswanae and O. capensis , and of O. gymnosoma and O. platyscapus , also share similar propodeal structures that could support the four species as two species pairs, but the groundplan propodeal structure and therefore likely structural transformation series remain uncertain.

Females of O. botswanae have a sulcus or light-colored line extending mediolongitudinally on the frons below the anterior ocellus into the scrobal depression ( Figs 156, 157 ) in addition to distinct ocellocular marks ( Fig. 157 ). Females of the Afrotropical genus Cervicosus Gibson have a similar head structure, including being highly convex with an even more distinct mediolongitudinal sulcus ( Gibson 1995, character 10, state 2; fig. 451) and ocellocular marks, though females are perhaps at least superficially more similar to females of O. platyscapus in having the scape conspicuously compressed, a more finely sculptured mesoscutum, and a comparatively long pronotum. In addition to having an elongate-pentagonal pronotum with subparallel sides, Cervicosus females are also brachypterous and have a strongly transverse propodeum that most closely resembles the propodea of O. ambigua and O. capensis . Unlike for Ooderella species, the dorsal surface of the pronotum is essentially in one plane without a distinctly differentiated collar and neck. Gibson (1995) distinguished two species in Cervicosus and although he hypothesized the two constituted a monophyletic lineage he also postulated that recognition of Cervicosus might render Ooderella paraphyletic ( Gibson 1995, fig. 519), simply representing two Afrotropical species in which the pronotum was secondarily ‘flattened’ in the common ancestor so that the abruptly differentiated collar and neck diagnostic of Ooderella species was lost. Another difference between species of the two genera is that fl1 is more strongly reduced, anelliform, in Cervicosus . The two genera could each be monophyletic sister-genera if the longer fl1 characteristic of Ooderella is secondarily derived, but unfortunately the present study provides no new information in resolving relationships between these two genera or other Eupelminae .

MALES. Even though males were unknown for the genus, Gibson (1995, fig. 519) included Ooderella within a clade of genera supported by four hypothesized synapomorphies, two of which based on males—ventral margin of torulus approximately in line with lower orbit (character 48, state 2), and flagellum filiform with both setae and mps short and dense, and the mps in multiple rows per segment (character 50, state 2b). However, New World males here interpreted as Ooderella have the toruli quite obviously above the lower orbit (character 48, state 1) (e.g. Figs 101 , 128 ), though this is affected by differences in the angle at which the head is viewed ( Fig. 81 ). Both New and Old World males also have a flagellum with very short and sparse setae, and either dense mps that are quite short and in multiple rows along each funicular ( Figs 85 , 105 , 209 ) (character 50, state 2c; fig. 370) or with a long, apically free portion such that each mps superficially appears as an apically directed seta ( Fig. 132 ) (character 50, state 2d; fig. 371). These two latter flagellar types are characteristic of what Gibson (1995) keyed, respectively, as males of “? Reikosiella group 3” (couplet 40) and “? Reikosiella group 4” (couplet 39). Gibson (2017) later synonymized Reikosiella under Merostenus Walker and considered true Merostenus males to have a flagellum with variably sparse but comparatively long setae at least about as long as the width of the funicular, and if with dense mps then these long and not in multiple rows per funicular. The above differentiation of Ooderella males from those of Merostenus is based on some males with type 3 and type 4 flagella being associated with females through identical label data, as well as the males sharing some morphological features with females. Although mesoscutal sculpture is variable, most New World males here identified as Ooderella have quite a coarsely meshlike reticulate mesoscutum, which is also characteristic of Ooderella females. Fusu (2013) described the males of three Palearctic species of Merostenus as imbricate-coriaceous and Gibson (2017) described males of some other Old Word Merostenus species as being variably more strongly reticulate in part. Some small individuals I identify as Ooderella males have quite a finely sculptured mesoscutum, but at least New World males I identify as Merostenus have the mesoscutum meshlike coriaceous to coriaceous-imbricate. Also, some of the males I identify as Ooderella have a transversely Π-like scrobal depression that resembles the scrobal depression of some females with identical collections records, and most have a vertical line of weakness or light-colored sulcus within the scrobal depression (character 47, state 2; Figs 81 , 101 , 128 , 205 ). Males identified as Merostenus often have a small depression or pit at the apex of the interantennal prominence ( Gibson 1995, figs 407, 408; Fusu 2013, fig. 43), but not a vertical sulcus or lighter-colored line within the scrobal depression. However, doubt remains concerning generic identity of males identified as Ooderella versus Merostenus because some New World males identified as one or the other genus have similar fore wing setal patterns. Some males assigned to both genera have a dorsally bare region (speculum) on the fore wing disc delineated basally by the basal setal line and posteriorly by the mediocubital setal line, though often the bare region is obscured by setae on the ventral surface. More reliable differentiation of males of Ooderella and Merostenus requires definitive association of the sexes through rearing or, more practicable and feasible, through COI analysis.

Those males of Ooderella having a flagellum with short mps in multiple overlapping rows and very short setae (e.g. Fig. 105 ) can be mistaken for males of Anastatus ( Gibson 1995, fig. 362) or those Zaischnopsis males with a similar flagellar structure that Gibson (1995, fig. 361) differentiated within the genus as “flagellar type 1”. Anastatus males always have bidentate mandibles whereas Zaischnopsis males have distinctly tridentate mandibles, though because of variation in Ooderella this is not necessarily helpful in correctly identifying Ooderella males. However, typical Anastatus and Zaischnopsis males differ from Ooderella males in having fl2 curved in lateral view or at least having the ventral surface slightly concave ( Gibson 1995, figs 361, 362). Further, such New World males have the toruli inserted obviously closer to the oral margin than the anterior ocellus at or somewhat below the lower orbits.

I have seen 43 males from the Neotropical region ( Bolivia, Brazil, Costa Rica, Ecuador, Peru, Trinidad and Venezuela) that I distinguish as Ooderella males. Of these, I assign 21 to three species— O. setosa , O. smithii and O. spinositegula , based primarily on the males being co-collected with females. Of the remaining 21 (all CNC) not assigned to species, 3 have flagellar structures similar to that described for O. spinositegula , whereas the other 19 have flagellar structures similar to that described for O. setosa and O. smithii . Several are most similar to males described as O. setosa and might be this species, but they are not included in the type series because they were not collected with females and mostly are from different localities than females. Although I have not seen any female Ooderella from southern USA, two males ( CNC) from Florida are very similar to O. setosa males. Because the latter species is known from Trinidad as well as Central America it is quite possible that O. setosa occurs also in subtropical Florida (see under this species). Males are known for only one of the species described from the Afrotropical region . They are similar to New World males assigned to Ooderella in having very short and sparse flagellar setae, but differ in having much shorter funiculars ( Fig. 208 ) and longer mps extending most of the length of the funicular ( Fig. 209 ). However, the two known males do have a yellowish line between the interantennal prominence and anterior ocellus ( Fig. 205 ), which supports an association with New World Ooderella . Relatively distinct ocellocular marks ( Fig. 206 ) also supports association of the sexes.

Figures

FIGURES 10 – 18. Ooderella ambigua n. sp. ♀. 10 & 11, head: 10, frontolateral (2016 - 77); 11, lateral (holotype). 12, head and pronotum, dorsal (holotype). 13, antenna [insert: mandibles] (2016 - 78). 14 & 15, holotype: 14, lateral habitus; 15, mesosoma, dorsolateral. 16 – 18, 2016 - 77: 16, propodeum, dorsal; 17, mesosoma, lateral; 18, gaster, dorsal [insert: syntergum, dorsal].

FIGURES 45 – 53. Ooderella melanosceles n. sp. ♀. 45 & 46, head (2016 - 83): 45, frontolateral; 46, lateral. 47, head and pronotum, dorsal (holotype). 48, antenna (2016 - 83). 49 – 53, holotype: 49, lateral habitus; 50, pronotum and mesoscutum; 51, mesosoma, lateral; 52, scutellar-axillar complex to propodeum plus wings; 53, gaster, dorsal.

FIGURES 1 – 9. Ooderella americana n. sp. ♀. 1 & 2, head: 1, frontolateral (2016 - 79); 2, lateral (holotype). 3, head and pronotum, dorsal (2016 - 79). 4 – 7, holotype: 4, lateral habitus; 5, antenna [insert: mandibles]; 6, pronotum and mesoscutum; 7, scutellar-axillar complex to propodeum plus wings. 8, mesosoma, lateral (2016 - 79). 9, gaster, dorsal (holotype).

FIGURES 90 – 98. Ooderella smithii Ashmead ♀. 90 – 93, 2016 - 86: 90, head, frontal; 91, head, lateral; 92, head and pronotum, dorsal; 93, lateral habitus. 94, lateral habitus (2015 - 4). 95, pronotum and mesoscutum (2016 - 86). 96 & 97, 2015 - 4: 96, pronotum and mesoscutum; 97, scutellar-axillar complex to propodeum plus wings. 98, mesosoma, lateral (2016 - 86).

FIGURES 156 – 164. Ooderella botswanae n. sp. ♀. 156, head, frontal (2016 - 92). 157 – 160, holotype: 157, head and pronotum, dorsal; 158, lateral habitus; 159, antenna; 160, mandibles. 161, pronotum and mesoscutum (2016 - 92). 162 – 164, holotype: 162, scutellar-axillar complex to propodeum plus wings; 163, mesosoma, lateral; 164, gaster [insert: syntergum], dorsal.

FIGURES 165 – 173. Ooderella capensis n. sp. ♀. 165, head, frontal (2016 - 94). 166 – 173, 2016 - 93: 166, head, lateral; 167, head and pronotum, dorsal; 168, lateral habitus; 169, antenna [insert: mandibles (2016 - 94)]; 170, pronotum and mesoscutum; 171, scutellar-axillar complex to propodeum plus wings; 172, mesosoma, lateral; 173, gaster, dorsal.

FIGURES 19 – 27. Ooderella flavida n. sp., holotype ♀: 19, head, frontolateral; 20; head, lateral; 21, head and pronotum, dorsal; 22, lateral habitus [insert: gaster, lateral (2016 - 81)]; 23, antenna [insert: mandibles]; 24, pronotum and mesoscutum; 25, scutellar-axillar complex to propodeum plus wings; 26, mesosoma, lateral; 27, gaster, dorsal

FIGURES 118 – 127. Ooderella spinositegula n. sp. ♀. 118 & 119, head, holotype: 118, frontal; 119, lateral. 120, head and pronotum, dorsal (2016 - 89). 121, lateral habitus (holotype). 122 & 123, 2026 - 91: 122, antenna; 123, mandibles. 124 & 125, 2016 - 89: 124, pronotum and mesoscutum; 125, scutellar-axillar complex to propodeum plus wings. 126, mesosoma, lateral (2016 - 90). 127, gaster, dorsal (holotype).

FIGURES 194 – 204. Ooderella platyscapus n. sp. ♀. 194, head, frontal (2016 - 96). 195, head, lateral (holotype). 196, head and pronotum, dorsal (2016 - 96). 197, lateral habitus (2015 - 5). 198, antenna (2016 - 97). 199, fore wing (2015 - 5). 200, mandibles (2016 - 98). 201 & 202, 2016 - 96: 201, pronotum and mesoscutum; 202, scutellar-axillar complex to propodeum plus wings. 203, mesosoma, lateral (2016 - 98). 204, gaster, dorsal (2016 - 96) [insert: syntergum, lateral (holotype)].

FIGURES 81 – 89. Ooderella setosa n. sp. ♂. 81 & 82, head: 81, frontal (allotype); 82, dorsal (2016 - 100). 83, lateral habitus (allotype). 84, mandibles (2016 - 101). 85, antenna [insert: pedicel to fl 3] (allotype). 86 – 88, 20165 - 100: 86, fore wing; 87; mesosoma, dorsal; 88, mesosoma, dorsolateral. 89, mesosoma, lateral (allotype).

FIGURES 205 – 213. Ooderella platyscapus n. sp. ♂ (2016 - 104): 205, head, frontal; 206, head, dorsal; 207, lateral habitus; 208, antenna; 209, pedicle to fl 3; 210, mesosoma, dorsal; 211, mesosoma, dorsolateral. 212, mesosoma, lateral; 213, fore wing.

FIGURES 101 – 109. Ooderella smithii Ashmead ♂. 101 & 102, head (2016 - 103): 101, frontal; 102, dorsal. 103 – 109, 2016 - 102: 103, lateral habitus; 104, antenna; 105, pedicel to fl 3; 106, mesosoma, dorsal; 107, mesosoma, dorsolateral; 108, mesosoma, lateral; 109, fore wing.

FIGURES 128 – 136. Ooderella spinositegula n. sp. ♂. 128 & 129, head: 128, frontal (allotype); 129, dorsal (2016 - 99). 130 – 134, allotype: 130, lateral habitus; 131, antenna; 132, pedicel to fl 3; 133, mesosoma, dorsal; 134, mesosoma, dorsolateral. 135, fore wing (2016 - 99). 136, mesosoma, lateral (allotype).

FIGURES 99 – 100. Ooderella smithii Ashmead ♀. 99, antenna [insert: mandibles (2016 - 88)]. 100, gaster [insert: syntergum], dorsal (2016 - 87).

FIGURES 185 – 193. Ooderella kenyaensis n. sp. ♀. 185 – 188, holotype: 185, head, frontal; 186, head and pronotum, dorsal; 187, lateral habitus; 188, mesosoma, dorsal. 189 – 192, 2016 - 95: 189, pronotum and mesoscutum; 190, scutellar-axillar complex to propodeum plus wings; 191, antenna; 192, mandibles. 193, gaster [insert: syntergum], dorsal (holotype).

FIGURES 174 – 184. Ooderella gymnosoma n. sp., holotype, ♀: 174, head, frontal; 175, head, lateral; 176, head and pronotum, dorsal; 177, lateral habitus; 178, antenna; 179, mandibles; 180, pronotum and mesoscutum; 181, mesosoma, lateral; 182, gaster, dorsal; 183, scutellar-axillar complex to propodeum plus wings; 184, syntergum, posterolateral.

FIGURES 35 – 44. Ooderella hyalipleura n. sp. ♀. 35 & 36, head: 35, frontal (holotype); 36, lateral (2016 - 82). 37, head and pronotum, dorsal (2016 - 82). 38, lateral habitus (holotype). 39 & 40, 2016 - 82: 39, antenna; 40, mesoscutum to propodeum. 41, mesosoma, lateral (holotype). 42 – 44, 2016 - 82: 42, fore wing; 43, gaster, dorsal; 44, syntergum, dorsal.

FIGURES 54 – 62. Ooderella microptera n. sp., holotype ♀: 54, head, frontal; 55, head, lateral; 56, head and pronotum, dorsal; 57, lateral habitus; 58, antenna [insert: mandibles]; 59, pronotum and mesoscutum; 60, scutellar-axillar complex to propodeum plus wings; 61, mesosoma, lateral; 62, gaster, dorsal.

FIGURES 28 – 34. Ooderella hansoni n. sp., holotype ♀: 28, head, frontolateral; 29, antenna and head, dorsolateral [insert: head, lateral]; 30, lateral habitus; 31, dorsal habitus; 32, scutellar-axillar complex to propodeum plus wings; 33, pronotum and mesoscutum; 34, gaster, dorsal.

FIGURES 147 – 155. Ooderella thegalea n. sp., holotype ♀: 147, head, frontal; 148, head and scape, lateral; 149, head and pronotum, dorsal; 150, antenna; 151 gaster, dorsal; 152, pronotum and mesoscutum; 153, scutellar-axillar complex to propodeum plus wings; 154, mesosoma, lateral; 155, mesosoma, lateral.

FIGURES 72 – 80. Ooderella setosa n. sp. ♀. 72, head, frontal (holotype); 73 – 75, 2016 - 84: 73, head, lateral; 74, head and pronotum, dorsal; 75, lateral habitus. 76, antenna (holotype) [insert: mandibles (2016 - 85)]. 77, pronotum and mesoscutum (2016 - 84). 78, scutellar-axillar complex to propodeum plus wings (holotype). 79, mesosoma, lateral (2016 - 84). 80, gaster, dorsal (holotype).

FIGURES 110 – 117. Ooderella speculifrons n. sp., holotype ♀: 110, head, frontolateral; 111, head and pronotum, dorsal; 112, lateral habitus; 113, antenna [insert: mandibles]; 114, pronotum and mesoscutum; 115, axillar complex to propodeum plus wings; 116, mesosoma, lateral; 117, gaster, dorsolateral.

FIGURES 137 – 146. Ooderella stenoptera n. sp., holotype ♀: 137, head, frontal; 138, head, dorsal; 139, scape, outer surface; 140, lateral habitus; 141, antenna; 142, mandibles; 143, pronotum and mesoscutum; 144, axillar complex to propodeum plus wings; 145, mesosoma, lateral; 146, gaster, dorsal.

FIGURES 63 – 71. Ooderella reticulifrons n. sp., holotype ♀: 63, head, frontal; 64, head, lateral; 65, head and pronotum, dorsal; 66, antenna; 67, lateral habitus; 68, pronotum and mesoscutum; 69, scutellar-axillar complex to propodeum plus wings; 70, mesosoma, lateral; 71, gaster [insert: syntergum], dorsal.