Multilocus phylogeny of the parasitic wasps in the tribe Euphorini (Hymenoptera: Braconidae) with revised generic classifications

Background Parasitic wasps in the family Braconidae are important regulators of insect pests, particularly in forest and agroecosystems. Within Braconidae, wasps in the tribe Euphorini (Euphorinae) attack economically damaging plant bugs (Miridae) that are major pests of field and vegetable crops. However, the evolutionary relationships of this tribe have been historically problematic. Most generic concepts have been based on ambiguous morphological characters which often leads to misidentification, complicating their use in biological control. Methods Using a combination of three genes (COI, 28S, and CAD) and 80 taxa collected worldwide, we conducted Bayesian inference using MrBayes, and maximum likelihood analyses using RAxML and IQ-Tree on individual gene trees as well as the concatenated dataset. Results The monophyly of the tribe Euphorini and the two genera Peristenus and Leiophron were confirmed using maximum likelihood and Bayesian inference. The subgeneric classifications of Leiophron sensu lato were not supported, and the monotypic genus Mama was also not supported. Discussion Euphoriella, Euphoriana, Euphorus, and Mama syn. n, have been synonymized under Leiophron. Mama mariae syn. n was placed as a junior synonym of Leiophron reclinator. The generic concepts of Peristenus and Leiophron were refined to reflect the updated phylogeny. Further we discuss the need for revising Euphorini given the number of undescribed species within the tribe.

The tribe Euphorini Foerster contains koinobiont endoparasitoids of Hemiptera and Psocodea, which attack young nymphs (1st or 2nd instar) and feed internally on the hemolymph of their hosts (Loan, 1974b).Mature parasitoid larvae emerge from mature host nymphs or teneral adults, and overwinter as pupae in soil (Loan, 1974b).Several species of Euphorini have been extensively studied for their use in biological control programs because they attack many serious agricultural pests such as Lygus Hahn (Day, 1987;Haye et al., 2005;Haye et al., 2007).Despite the research interest using Euphorini wasps in applied entomology, the classification and identification of these parasitoids remains challenging.Variable morphological characters such as the degree of completion of the occipital carina, and the presence or absence of vein (RS + M)a and cu-a were used as defining characters (Chen & Van Achterberg, 1997;Stigenberg, Boring & Ronquist, 2015), leading to taxonomic uncertainty and misidentification.
Here, we extensively sample Euphorini wasps and reconstruct the evolutionary relationships among its members using a multi-locus dataset.We reassess the generic and subgeneric concepts of Euphorini and revise the classification to reflect the phylogeny.The results of this study provide a comprehensive framework for phylogenetic relationships among Euphorini wasps and we provide taxonomic clarity and identification resources to aid future applied research and biological control programs.
Outgroups included representatives of the most closely related tribes to Euphorini, based on Stigenberg, Boring & Ronquist (2015): Microctonus Wesmael (Perilitini), Townesilitus Haeselbarth & Loan (Townsilitini), and Chrysopophthorus Goidanich (Helorimorphini).A list of the specimens utilized in this study is provided in Table 1, and detailed locality information in Table S1.A map of the distribution of these specimens is depicted in Fig. 1, which was generated using ArcMap v10.5.1.For ease of interpretation of results, specimen information was added to taxon labels for the phylogenetic analyses, including country and lowest identification.The subgeneric names within Leiophron s.l. are used as specimen names to avoid confusion with Leiophron s.s.(eg.Leiophron (Leiophron) uniformis is listed as Leiophron uniformis, whereas Leiophron (Euphoriana) dispar is listed as Euphoriana dispar).

Terminology and image capture
Terminology used for most morphological characters follows Chen & Van Achterberg (1997) and Stigenberg, Boring & Ronquist (2015).However, wing venation terminology follows Sharkey & Wharton (1997).Specimens were photographed using a Canon 7D Mark II with a Mitutoyo M Plan Apo 10× objective mounted onto the Canon EF Telephoto 70-200 mm zoom lens, and the Canon MT-24EX Macro Twin Lite Flash (Tokyo, Japan) with custom-made diffusers to minimize hot spots.

DNA protocols
A total of 80 taxa were sampled out of which three species represented outgroups.The taxon sampling covers the entire range of Euphorini, which is found on all continents except in Antarctica and Australia outside of Papua New Guinea (Yu, Van Achterberg & Horstmann, 2012).This is the largest sampling of any of the euphorine tribes, and is comprised of all three Euphorini genera, as well as all four subgenera of Leiophron s.l.Specimens were extracted, amplified, and sequenced either at the Molecular Systematics Laboratory, Swedish Museum of Natural History following the protocols listed in (Stigenberg, Boring & Ronquist, 2015), or at the Insect Systematics Laboratory at the University of Central Florida following the DNeasy TM Tissue Kit protocol (Qiagen, Valencia, CA, USA).Petioles were separated from mesosomas to ensure buffer penetration during tissue lysis, and the two body parts were mounted onto the same point post-extraction for vouchering.Voucher specimens deposition are listed in Table 1.Three genes were amplified: partial 28S domain 2 and 3 (rDNA), partial CAD (Carbamoyl-Phosphate Synthetase 2, Aspartate Transcarbamylase, and Dihydroorotase) and the 5 region of mitochondrial COI.New Euphorini-specific primers were designed for CAD based on sequences from Sharanowski, Dowling & Sharkey (2011) and Stigenberg, Boring & Ronquist (2015).The faster rate of evolution of the mitochondrial genes is ideal for separating closely related species (Zhang, Ridenbaugh & Sharanowski, 2017), while the ribosomal and nuclear genes have slower rates of evolution and are more suitable for higher level phylogenetic relationships (Sharanowski, Dowling & Sharkey, 2011).All three genes are commonly used in Braconidae phylogenetics, including Euphorinae (Sharanowski, Dowling & Sharkey, 2011;Stigenberg, Boring & Ronquist, 2015;Zhang, Ridenbaugh & Sharanowski, 2017).
All PCRs were performed on a Bio-Rad MyCyclerTM thermal cycler, using approximately 1µg DNA extract, 1× Standard Taq Buffer (10 mm Tris-HCl, 50 mm   2. Amplicons of reaction products were cleaned with Agencourt CleanSEQ magnetic beads and sequenced in both directions using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, USA) and the Applied Biosystems 3730xl DNA Analyzer at the University of Kentucky, Advanced Genetic Technologies Center (UK-AGTC).Contigs were assembled and edited using Geneious version 8.18 (Kearse et al., 2012), and alignment was conducted using MAFFT server (Katoh et al., 2002) (https://mafft.cbrc.jp/alignment/server/).The protein coding genes were aligned using default MAFFT settings, and for 28S we used Q-INS-I strategy (Katoh & Toh, 2008) which takes secondary RNA structure into account.New sequences obtained from this study were deposited in GenBank (See Table 1).

Phylogenetic analyses
The three genes were analyzed separately and concatenated using Bayesian inference (BI) analysis with MrBayes v3.2.6 (Ronquist et al., 2012) on the CIPRES Science Gateway (Miller et al., 2009).Each analysis had two independent searches with four chains and  Tavaré (1986) were run for 10,000,000 generations, sampling every 1,000, with a 10% burnin discarded.

RESULTS
Here we present the most taxonomically comprehensive phylogeny of the euphorine braconid tribe Euphorini with all known genera and subgenera sampled.All genera and subgenera had multiple representatives except Euphoriana (only one exemplar included -E.dispar) and the monotypic Mama mariae.A total of 39 CAD, 71 28S, and 80 COI for a total of 190 sequences were used for the final analyses, 158 of which were newly generated for this study (Table 1).The summary statistics of all three genes can be found in Table 3.While we failed to amplify CAD sequences from some older specimens, the gene itself is informative (See Table 3) and should be used in other multilocus analyses of braconids.

Generic concepts of Peristenus and Leiophron s.l.
Our data supports the monophyly of Peristenus and Leiophron, corroborating the results of previous morphological studies (Loan, 1974a;Loan, 1974b;Shaw, 1985;Shaw, 1987;Chen & Van Achterberg, 1997).Additionally, with a much more focused taxon sampling we were able to delineate the finer relationships within the tribe Euphorini.Peristenus is largely uniform in morphology and exclusively attacks Miridae, while its sister taxon Leiophron is much more variable in both morphology and host breadth which likely has led to convergent morphology, and hence the subgeneric concepts and taxonomic confusion.Peristenus can be distinguished from Leiophron by the evenly setose 1st discal, basal, and subbasal cells in the forewing (Fig. 3A), and the 1st metasomal tergite, which is fused or touching basally (Fig. 4B).
Representatives of all four subgenera of Leiophron s.l.defined by Stigenberg, Boring & Ronquist (2015): Euphoriana, Euphoriella, Euphorus, and Leiophron s.s., were included in this analysis.These subgeneric relationships were not supported in any of our analyses, as they failed to form monophyletic clades (Fig. 2).This is not surprising given the lack of consistent morphological characters as specimens often exhibit characteristics of different subgenera, leading to similar suggestions for synonymizations in the past (Muesebeck, 1936;Loan, 1974b;Shaw, 1985;Stigenberg, Boring & Ronquist, 2015).Morphological characters such as the presence or absence of the forewing vein (RS + M)a, hindwing vein cu-a, and complete occipital carina are all too variable to be used as defining characteristics  (Chen & Van Achterberg, 1997;Stigenberg, Boring & Ronquist, 2015).These ambiguous distinguishing characters at the subgeneric level can easily lead to misidentification in ecological or applied studies.Euphoriella is morphologically unique based on the absence of radial cell, and the fusion of ventral side of petiole.However, this subgenus is rendered paraphyletic by Leiophron sp.YMZ364 and YMZ371.The type species of Euphoriella, Euphoriella incerta Ashmead, was originally collected from Florida (Muesebeck, 1936).However, the type specimen is too badly damaged to be compared with our equally damaged DNA voucher also collected from Florida to accurately confirm its identity.While there are no nomenclatural rules that requires subgenera to be monophyletic, we do not support paraphyly as these morphological similarities could be the result of convergent evolution and does not reflect evolutionary history.Therefore, based on our molecular evidence combined with the inconsistency of previously used morphological characters, we recommend treating Leiophron as a single genus without further subdivisions and synonymize the subgenera Euphoriana, Euphoriella, and Euphorus as junior synonyms of Leiophron.With this taxonomic update, Leiophron can be identified with the following combination of characters: 1st discal cell of the forewing is more setose than basal and subbasal cells (often glabrous) in Leiophron (Fig. 3B), but if not, then the ventral side of the 1st metasomal tergite (petiole) is not fused (Fig. 4C).
The exact age of the divergence between Peristenus and Leiophron is unknown, as the only known fossil record of Euphorini is a single specimen described as Euphorus indurescens Brues, found in Florissant, Colorado and dating back to Eocene at around 33.7-37 mya (Brues, 1910).In addition, both Leiophron and Peristenus have received little taxonomic attention outside of Europe, North America, and Asia (Loan, 1974a;Loan, 1974b;Chen & Van Achterberg, 1997;Belokobylskij, 2000b;Goulet & Mason, 2006;Stigenberg & Van Achterberg, 2016).We have included many undescribed species from Central and South America, Africa, and Papua New Guinea, which is unsurprising given the tremendous diversity of their major host Miridae.The first and second author are currently working on describing species from Papua New Guinea (J Stigenberg & YM Zhang, 2018, unpublished data), but a revision of the world Euphorini is needed.

Validity of the genus Mama
The validity of the enigmatic genus Mama, described based on a single species M. mariae from eastern Russia (Belokobylskij, 2000a), has been questioned before by Simbolotti, Villemant & Van Achterberg (2004).Both M. mariae and L. reclinator have long, compressed, and spiny scapes (Fig. 4A), but due to the poor condition of the L. reclinator lectotype no definitive conclusion was made (Simbolotti, Villemant & Van Achterberg, 2004).With the consistent placement of the two species as sister taxa with short branch lengths in all three genes (Figs.S1-S3) and concatenated dataset (Fig. 2), and identical in morphology (the second author has examined the holotypes of M. mariae and L. reclinator), we synonymize M. mariae syn.n. as a junior synonym of L. reclinator, thus effectively dissolving the monotypic genus Mama syn.n. as a junior synonym of Leiophron.The distribution of L. reclinator likely spans across Eurasia, as specimens are found from eastern Russia to Sweden and the United Kingdom (Stigenberg & Van Achterberg, 2016).

CONCLUSIONS
Using a multilocus phylogenetics approach and the most comprehensive taxon sampling of Euphorini to date, we were able to clarify the long standing taxonomic confusion within this tribe of economically important braconid wasps.The taxonomic uncertainty that has long impacted biological control studies of Euphorini is readily resolved with the revised generic concepts presented here, which reflects the strongly supported phylogenetic analyses, therefore providing clear distinguishing characters for the two genera Peristenus and Leiophron.With a phylogenetic framework to build upon, the next step should focus on the world revision of tribe Euphorini, with a strong alpha taxonomic component, as many of the species used in this study were undescribed or have an unknown biology.

Figure 1
Figure 1 Geographical distribution of specimens used in this study.Blue dots are published data from Stigenberg, Boring & Ronquist (2015), red dots are newly sampled taxa for this study.Full-size DOI: 10.7717/peerj.4783/fig-1

Figure 2 ZhangFigure 3
Figure 2 Concatenated gene tree for MrBayes, RAxML, and IQ-Tree.Peristenus is colored red, and Leiophron is colored in blue, with subgenera within Leiophron shown in different colors (Leiophron sensu stricto in blue, Euphorus in purple, Euphoriana in green, Euphoriella in orange, and Mama in brown).Asterisks indicate strong nodal support for all three analyses (≥0.98 posterior probability support for MrBayes; ≥90 for bootstrap support for RAxML; and ≥90 for ultrafast bootstrap support for IQ-Tree).Full-size DOI: 10.7717/peerj.4783/fig-2

Table 1 Specimen information. GenBank
accession numbers (new sequences generated for this study is in bold), collection localities, and voucher deposition institutes for all specimens used in this study.HIC (Hymenoptera Institute, University of Kentucky, Lexington), MNHN (French National Museum of Natural History, Paris), NHRS (Swedish Museum of Natural History, Stockholm), UCFC (University of Central Florida Collection of Arthropods, Orlando), ZIN (Zoological Institute of Russian Academy of Sciences, St. Petersburg).

Table 2
Primers.List of primers used in this study.