First described fossil representatives of the parasitoid wasp taxa Asaphesinae n. n. and Eunotinae (Hymenoptera: Chalcidoidea: Pteromalidae sensu lato) from Eocene Baltic amber

New Baltic amber species of Pteromalidae sensu lato are described, from two different subfamilies, Asaphesinae n. n. and Eunotinae. Asaphesinae is provided as a replacement name for Asaphinae Ashmead 1904, which is a junior homonym of the trilobite family Asaphidae Burmeister 1843. Coriotela lasallei gen. n., sp. n.. and Butiokeras costae gen. n., sp. n.. are described as the first known fossil species of Asaphesinae and Eunotinae, respectively. These species establish the minimum known age of both groups in the Eocene. Taxonomic changes are also proposed for some extant species. The genus Desantisiana Neder de Román syn. n.. is found to be a junior synonym of Notoglyptus, and its only described species is transferred as Notoglyptus jujuyensis (Neder de Román) comb. n.. The tribe Calyconotiscini, previously classified in Eunotinae, is abolished and Calyconotiscus Narendran & Saleem is transferred to Pireninae. http://www.zoobank.org/urn:lsid:zoobank.org:pub:7A107FF9-28E7-40AA-8A9B-71321E476C07 ARTICLE HISTORY Received 3 October 2019 Accepted 20 March 2020 Published online 23 September 2020 Published in print 23 September 2020


Introduction
Pteromalidae is well known as a polyphyletic 'dumping ground' of miscellaneous chalcidoid taxa (Noyes 1990;Gibson et al. 1999;Campbell et al. 2000;Krogmann and Vilhelmsen 2006;Munro et al. 2011;Heraty et al. 2013). No single molecular, morphological, or lifehistory trait can be used to reliably place species within Pteromalidae, and Pteromalidae has never been supported as monophyletic in any analysis that included more than one of its subfamilies (e.g. Munro et al. 2011;Heraty et al. 2013). However, we retain it in its current broad sense (sensu lato), in part because the phylogenetic relationships of most of its 31 subfamilies (Noyes 2019) are uncertain; they may be better combined to form a few distinct families, or may be better classified in other already existing families, but only after a thorough and robust revision of the family classification of Chalcidoidea.
Because higher-level classification of Pteromalidae imparts little information on how a group is distinct from and related to other chalcidoids, there is also little value in placing fossil taxa to Pteromalidae without further comment. Placing these taxa to subfamily or CONTACT Roger A. Burks burks.roger@gmail.com tribe communicates far more relevant information for Chalcidoidea (Burks et al. 2018). The discovery of fossil taxa also provides a better context for investigating the age of pteromalid subgroups.
Knowing the minimum age of these lineages provides a valuable calibration point for chalcidoid phylogenetics, and it may shed light on the evolution of the host associations of the lineages. Asaphesinae includes four genera, two that are associates of Hemiptera, and two Australasian genera that are egg parasitoids of Chrysomelidae. Species of Asaphes Walker are known primarily as hyperparasitoids attacking a variety of hosts in aphids, although some may be primary parasitoids of aphids or psyllids (see discussion in Gibson and Vikberg 1998). Hyperimerus Girault has been reared from psyllids (Bouček and Rasplus 1991). Ausasaphes Bouček and Enoggera Girault include endemic Australasian species that are egg parasitoids in Chrysomelidae, including the phoretic species Ausasaphes shiralee Naumann & Reid (Bouček 1988;Naumann and Reid 1990). Therefore, should Asaphesinae prove monophyletic, this would suggest that at least one major switch in life-history mode has occurred in its evolution. Eunotinae contains a variety of parasitoids and egg predators of scale insects (Hemiptera: Coccoidea) occurring in all continents except Antarctica, although some have been reported from other hosts (Graham 1969;Bouček 1988;Bouček and Rasplus 1991). Here again, at least one switch from egg predation to parasitism has apparently occurred. The life history of many species in both subfamilies is largely unknown, and most species have not been placed in a phylogenetic analysis, whether morphological or molecular. Therefore, additional research would prove invaluable in providing a context and foundation for the elucidation of their evolutionary histories.

Methods
Fossil holotypes are deposited in the American Museum of Natural History, New York, USA (AMNH). Stereoscope photographs were taken using Leica Imaging System Software with a Z16 APO A microscope. Terminology follows that of Heraty et al. (2013), with some additional terms or usage following  or Goulet and Huber (1993), and sculptural terms following Eady (1967 Coriotela Burks & Heraty n. gen. (Figure 1 Description. Head with coriaceous sculpture. Antenna with 12 flagellomeres, including a small anellus, 7 funiculars, and 4 clavomeres (including terminal button) (Figure 1(b)). Toruli above lower eye margin, slightly below centre of face. Maxilla with 4 palpomeres, labium with 3 palpomeres. Occipital carina weakly indicated as a uniform arch, and without the median peak found in some Asaphesinae. Gena posteriorly carinate near mandibular base.
Mesosoma predominantly with coriaceous sculpture. Mesoscutum with complete notauli. Axillula slightly recessed relative to surrounding areas; frenum indicated by complete frenal groove. Prepectus bare. Fore wing without speculum, uniformly setose; parastigma with constriction and hyaline area immediately beyond one pair of placoid sensilla (Figure 1(c)), marginal vein slightly longer than stigmal vein (when measured as indicated by , slightly over half as long as postmarginal vein. Fore leg with curved, stout, apically cleft fore tibial spur; basitarsal comb oblique, crossing basitarsal notch. Mid tibial spur not enlarged.
Metasoma length 1.6 × metasoma width, 1.4 × mesosoma length. Gastral terga of similar length, but Gt 3 slightly shorter than the others. Hypopygium short, reaching only to Gt 2 . Etymology. Named after John La Salle, who first introduced me (RAB) to the excitement of chalcidoid fossils.

Material examined
Discussion. Coriotela is classified in Asaphesinae because of shared features with Hyperimerus and to some undescribed Neotropical genera, including antenna position and flagellomere shape and count (Shender et al. Figure 14), pronotum shape (Shender et al. Figure 19), and fore wing venation (Shender et al. Figure 21). Because Asaphesinae is not yet defined phylogenetically, the key features used for this placement are likely a mix of plesiomorphies and synapomorphies. Some defining features of the subfamily are relatively rare across Chalcidoidea, and therefore should be investigated as potential synapomorphies of the subfamily, including: metasomal petiole short but with strong sculpture (Figure 1(b)), gaster strongly sclerotised and rigid, fore wing venation with a short marginal vein relative to the stigmal and postmarginal veins, parastigma with a constriction (Figure 1(c): constriction). Other defining features of the group are more likely plesiomorphic or only very locally informative, but serve to eliminate some other taxa from consideration: antenna with 12 flagellomeres, occipital carina present, pronotum with a relatively long but anteriorly rounded collar, notauli complete, frenal groove indicated.
A suite of similar taxa in Pteromalidae s. l. can be eliminated from consideration for placement of C. lasallei as follows: Pteromalinae either exhibit a demarcation between the pronotal collar and pronotal neck, or the pronotum consists only of an essentially vertical neck, and pteromalines do not have all the other key features of C. lasallei in combination. The lack of absolute morphological distinction between Pteromalinae and other subfamilies of Pteromalidae s. l. has been discussed elsewhere (Graham 1969;Bouček 1988), but molecular data support a monophyletic Pteromalinae, inclusive of some non-pollinator fig wasps, that is separate from most other subfamilies of Pteromalidae sensu lato (Heraty et al. 2013). The genera of Pteromalinae most similar to C. lasallei and other Asaphesinae include Coruna Walker, Oricoruna Bouček, Sphegipterosema Girault, and Yanchepia Bouček. However, each of these genera lacks key features mentioned in the diagnosis of C. lasallei and instead possess gestalt features that place them in particular subgroups of Pteromalinae (Graham 1969;Bouček 1988;Bouček and Rasplus 1991;Bouček and Heydon 1997).
A few subgroups of Pteromalidae s. l. are somewhat similar to C. lasallei, and are discussed here to help explain its subfamily placement. Austrosystasinae differs from all Asaphesinae in features of gestalt: a more arched and stout body, and a much larger metacoxa (Bouček 1988, figure 550). Diparinae usually possess an occipital carina, but other typically diparine features such as the cercal brush (Desjardins 2007, figure 17) and gestalt features of metacoxal sculpture and of the fore wing are absent from C. lasallei. Herbertiinae are part of a set of pteromalid taxa that differ by having fewer than 12 antennal flagellomeres. Keiraninae are similar to C. lasallei, but have a more distinct pronotal collar that is set off from the pronotal neck by a stronger change in curvature (Bouček 1988, figure 475). Distinction of Asaphesinae from Keiraninae, when limited to features typically visible on an amber fossil, relies on vague gestalt features. However, preliminary genetic data (Heraty et al. unpublished) support separation of extant species of these two taxa into widely separated clades in the superfamily phylogeny. Melanosomellini (in Ormocerinae) contains a few genera with an occipital carina, but these have a smaller axillula and a strong mesoscutellar rim, and they have a broader, less sculptured petiole (e.g. Bouček and Heydon 1997, figure 80). Parasaphodinae differ in having strongly advanced axillae (Bouček 1988 , figure 629).
Within Asaphesinae, C. lasallei is distinguished by a combination of features mentioned in the diagnosis of the genus. Asaphes differs chiefly in having two or more basal flagellomeres lacking multiporous plate sensilla (= anelli) and toruli situated very near the mouth margin (Bouček and Rasplus 1991;Bouček and Heydon 1997, figure 84;Gibson and Vikberg 1998). Ausasaphes Bouček possesses a dorsally indistinct frenal groove and a longer petiole (Bouček 1988 , figure 631). Enoggera Girault possesses a very differently shaped head, mesosoma, and metasoma, and also differs in having more anelliform basal flagellomeres and a much larger Gt 1 (Bouček 1988). Hyperimerus Girault differs in that its prepectus and Gt 1 basally are strongly setose (Gibson and Vikberg 1998;Schender et al. 2014 figures 11-12). The chiefly coriaceous sculpture of Coriotela is also distinctive within the subfamily, present mainly in some undescribed Neotropical taxa (Burks unpub.).

Asaphesinae: taxonomic changes
The type specimen of Desantisiana jujuyensis Neder de Román, described in Asaphinae (Neder de Román 1999), was examined and found to belong to Notoglyptus Masi, which is classified in the tribe Sphegigastrini of Pteromalidae sensu stricto. Because D. jujuyensis is the type and only species of Desantisiana n. syn., this name becomes a junior synonym of Notoglyptus. We suggest that comparison of Notoglyptus jujuyensis (Neder de Román) n. comb. with Neotropical species described by Heydon (1989) may result in synonymy of the species.

Diagnosis
Antennal flagellum in male with 1 anellus, 4 funiculars, and 3 clavomeres; all except the anellus with a row of many raised multiporous plate sensilla that span the length of their respective segment. Vertexal carina absent. Mesoscutum and mesoscutellum with fineraised reticulate sculpture. Fore wing densely setose, without speculum. Gaster not rigidly convex but apparently with more flexible terga; first gastral tergum short, not more than a third total gastral length, with lateral incision.
Description. Male. Head finely reticulate, with large subcircular eyes that are well separated and with ventral divergent medial margins, and gena short dorsally (thus head of shape typical for Eunotus Walker). Eye reaching posterior margin of head dorsally; lateral ocelli not reaching posterior margin of vertex. Malar sulcus present. Clypeus weakly convex, with subapical groove near ventral margin (as in Eunotus); labrum present and exposed, short and flap-like. Antennal flagellum with 1 very short anellus, 4 funiculars, 3 clavomeres, with each funicular and clavomere with raised multiporous plate sensilla extending the length of their flagellomere; flagellar setae short and appressed to their segment, therefore inconspicuous. Vertex somewhat abrupt posteriorly, but rounded and without vertexal carina.
Metasoma about as long as head plus mesosoma; terga not rigidly convex but apparently with more flexible terga. First gastral tergum short, less than a third total gastral length; with lateral incision of the type frequently found in other chalcidoids (e.g. Perilampidae) but not found in other Eunotini. Male genitalia with volsellar digiti and parameres.
Etymology. After the Greek words βυτίο = barrel, and κέρας = horn. Refers to the multiporous plate sensilla of the antenna. Gender neuter.
Head. Flagellomeres slightly broader than long, each with 1 row of multiporous plate sensilla [mentioned here because Eunotus species vary in the number or rows per segment].
Mesosoma very weakly arched anteriorly and posteriorly, but flat at scutoscutellar sulcus. Propodeum laterally with a strongly sculptured elevation. First tarsomere of each leg almost as long as the others combined. Etymology. After the Latin noun costa, meaning rib. Genitive case.

Material examined
Discussion. Females are unknown but presumably are similar to the known male except for the antenna and gaster. Butiokeras is classified in Eunotini, near Eunotus, because it shares features of the male flagellum and fore wing with extant species of Eunotus. The flagellomeres in males are distinctive, each with one or two rows of multiporous plate sensilla that give the appearance of ribs because they are conspicuously elevated and in total extend about the same length as their respective flagellomeres. These antennal features are best known from the Eunotus kocoureki Bouček species group, which is also defined by the presence of three mandibular teeth instead of two (Bouček 1972). Placement in the E. kocoureki species group was considered, but rejected because Butiokeras costae lacks a vertexal carina and possesses a short Gt 1 . Either of these character states would be unique within Eunotus, and the occurrence of both together and consideration of the fossil's age suggest that Butiokeras is best treated as a distinct genus.
Other genera of Eunotini are distinctly different from Butiokeras. Epicopterus Westwood and Mesopeltita Ghesquière differ in having a strong incision in the fore wing near the marginal vein base. Scutellista Motschulsky is known for its elongate mesoscutellum which typically extends dorsally over the metanotum, propodeum, and part of the gaster. Cephaleta Motschulsky and the apparently closely related Cavitas Xiao & Huang also have a relatively short Gt 1 , but they also have a relatively smooth and glossy mesoscutum and mesoscutellum instead of the fine raised sculpture present in Eunotus (Xiao & Huang 2001) and Butiokeras.

Eunotinae: taxonomic changes
Three other tribes may be correctly placed in Eunotinae: Idiopororini, Moranilini, and Tomocerodini, all distinctly different from Eunotini and potentially not forming a monophyletic group with it (Munro et al. 2011). Calyconotiscini was described by Narendran et al. (2012) in Eunotinae accommodate a species reared from galls of Cecidomyiidae. Examination of this species indicates that Calyconotiscus frontofasciatus Narendran & Saleem is a yellowish member of Pireninae that is otherwise not much different from Gastrancistrus Westwood, a conclusion upheld by its host relationships. We transfer Calyconotiscus Narendran & Saleem to Pireninae and abolish the tribe Calyconotiscini n. syn., because it likely does not represent a particularly informative group within Pireninae.

Discussion
An at least Eocene age for Asaphesinae and Eunotini is helpful for addressing some of the problems of pteromalid phylogenetics, indicating that some variation within Pteromalidae sensu lato is relatively old compared with that of already described families. Among extant families, only Mymaridae and Rotoitidae are definitively known from any fossils older than Eocene (Poinar and Huber 2011;Gumovsky et al. 2018). While other families such as Eulophidae and Trichogrammatidae were reported from putatively Cretaceous Ethiopian amber fossils by Schmidt et al. (2010), later data (Perrichot et al. 2016(Perrichot et al. , 2018 have shown Ethiopian amber to be Early Miocene in origin (16-23 Ma).
Therefore, it is possible that Asaphesinae and Eunotini, from at least the Eocene, are about as old as most other chalcidoid families. This possibility is upheld by molecular data, which place Asaphesinae and Eunotini in scattered positions among other Pteromalidae sensu lato and relatively well-sclerotised Chalcidoidea, such as Eucharitidae, Ormyridae, Perilampidae, and Torymidae (Heraty et al. 2013). Given that informative value of family groups would likely be damaged by lumping so many families into a broader Pteromalidae sensu lato, this suggests that some subfamilies of Pteromalidae may be better removed from Pteromalidae sensu stricto. This does not address the problem of family delimitation however, and not enough information exists in the fossil record to do so. Any change in rank or family placement for Asaphesinae or Eunotini would be premature without re-examination and equal-handed treatment of all other chalcidoid taxa using molecular and morphological data.
The John Heraty Laboratory and Lars Krogmann provided valuable comments on the manuscript. Daniel Aquino (Museo de la Plata, Argentina) provided valuable photographs of Desantisiana jujuyensis. Funding for this research was provided by grant NSF-DEB 1555808 and UCR Hatch project funding to JMH.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the National Science Foundation [NSF-DEB 1555808].