﻿Three new species of Retusigaster Dangerfield, Austin & Whitfield, 1999 (Hymenoptera, Braconidae, Cardiochilinae) with an illustrated key to the New World species

﻿Abstract Retusigaster Dangerfield, Austin & Whitfield, 1999 is a genus of the subfamily Cardiochilinae Ashmead, 1900 and exhibits high species richness in the New World. Eight species of the genus were recorded before this work: five species from the Nearctic region, two species from the Neotropical region, and one species from the Palearctic region. In this article, three new species of New World Retusigaster are described based on morphological characters: R.pulawskiisp. nov.; R.purshisp. nov.; R.vanduzeeisp. nov. In addition, potential food sources of the members of R.arugosus (Mao, 1949) and R.purshisp. nov. are reported, and an illustrated key to the New World species of Retusigaster is provided. The number of species of Retusigaster in the New World is increased from seven to ten.


Introduction
Retusigaster Dangerfield, Austin & Whitfield, 1999 is a genus of Cardiochilinae Ashmead, 1900 (Ashmead 1900) with eight valid species (Yu et al. 2016). Seven species were recorded from the New World: R. arugosus (Mao, 1949), R. albopilosus Mercado, 2003, R. brevitarsus (Mao, 1949), R. dignus (Mao, 1949), R. noguerai Mercado, 2003, R. pullus (Mao, 1949) and R. rubidus (Mao, 1949). One species, R. eremita (Kokujev, 1904) was recorded from the Palearctic region. Genus-level phylogenetic analyses, based on morphological data, were conducted by Dangerfield et al. (1999) and Mercado and Wharton (2003) and validated the genus. In the phylogeny of Dangerfield et al. (1999), which included representatives of all the cardiochiline genera in the world, Retusigaster was resolved as a monophyletic group. In the phylogeny of Mercado and Wharton (2003), Retusigaster species groups, which were defined by the authors based on the degree of thickening of the apex of hind tibia were clustered with the members of Toxoneuron Say, 1836 (Say 1836) but made Toxoneuron paraphyletic. Regarding the result of their analysis, Mercado and Wharton (2003) described that "Nevertheless, Retusigaster, as defined by its type species rubidus, is readily identifiable and is accepted here as a monophyletic group regardless of the rank eventually accorded." In the current project, I follow the definition of Dangerfield et al. (1999) and describe three new species collected in the New World. Potential food sources of R. arugosus and R. purshi sp. nov. are reported, and an illustrated key to the New World species is included. In addition, the species groups defined by Mercado and Wharton (2003) with their diagnostic characters are re-evaluated and discussed, and the placement of R. eremita is discussed.

Specimen information
The specimens for this work were borrowed from the California Academy of Sciences (CAS; San Francisco, CA, USA), Hymenoptera Institute (HIC; Redlands, California, USA), Museum of Comparative Zoology (MCZ; Cambridge, Massachusetts, USA), and Texas A&M University Insect Collection (TAMU; College Station, Texas, USA). Types of the new species will be deposited in CAS and the National Museum of Natural History (NMNH; the Smithsonian Institution, Washington D.C., USA).

Morphological analyses
Specimens were examined using a Leica MZ75 stereomicroscope. The morphological terminology follows Dangerfield et al. (1999) and Sharkey and Wharton (1997). The terms used in this work can be found as synonyms on the website of Hymenoptera Anatomy and Consortium (2022). Terms for surface sculpture are based on Harris (1979). The following acronyms are used for morphological terms: POL: distance between posterior ocelli, T2: second metasomal tergum, and T3: third metasomal tergum. Using a Visionary Digital BK Plus imaging system (Dun, Inc.) with a Canon EOS 5DS DSLR, images were captured. Image stacking was performed via Zerene Stacker v.1.04 (Zerene Systems LLC.). Images were edited using Adobe Photoshop CS 6 and Photoshop CC 2022 v. 23.0 (Adobe Systems, Inc), and final image plates were produced using the same Adobe software. Body parts were measured using the same Adobe software mentioned. Numbers in parentheses in species descriptions indicate 0.01× the actual size of each body character. The unit of length is mm.

Retusigaster Dangerfield, Austin & Whitfield, 1999
Type species. Cardiochiles rubidus Mao, 1949 Diagnosis. Dangerfield et al. (1999) and Mercado and Wharton (2003) provided detailed diagnostic characters. Retusigaster can be easily distinguished from other cardiochiline genera by the combination of the following characters: eye seemingly bare ( Diagnosis. Members of Retusigaster albopilosus can be recognized by the combination of the following characters: body 3.5-5.5 mm; fore wing entirely infuscate with dark stigma; fore tibia pale; mid and hind femur entirely dark; T2 entirely dark. Description. See Mercado and Wharton (2003).

Diagnosis.
Retusigaster arugosus is nearly identical to R. pullus. The members of both species possess dark head and metasoma with pale metasoma. As Mao (1949) mentioned, R. arugosus is distinguished from R. pullus by having basally hyaline and apically infuscate wings (Key image 4A). Body ~ 5.5 mm.
Description. See Mao (1949). Male. Unknown. Biology (potential food source). Cotton (Gossypium sp.; Malvaceae; recorded on the label of one specimen collected in Brazos County, Texas).
Etymology. Named in honor of Dr Wojciech Jerzy Pulawski, Curator of Entomology, Emeritus, at CAS, the person who collected the specimen from Jamaica.
Etymology. Named in honor of Fredrick Traugott Pursh, a German American botanist. The genus of the potential food source was also named after him, Purshia.
Biology (potential food source). Mexican Cliffrose (Purshia mexicana (D. Don) S. L. Welsh; Rosaceae) Distribution. Retusigaster purshi sp. nov. is known from one female specimen collected in Telephone Canyon, Clark County, Nevada, USA. (Fig. 4)   Figure 3. Retusigaster purshi sp. nov. A lateral habitus B dorsal habitus C anterior head D dorsal mesosoma E forewing. Diagnosis. Members of Retusigaster rubidus are most similar to those of R. brevitarsis. R. rubidus can be distinguished from other members by the following characters: body ~ 7.5 mm. notauli crenulate (Key image 3B); mesoscutum mostly orange pale; fore wing with pale stigma (Key image 6A); metasoma mostly orange pale.

Discussion
Regarding the placement of R. eremita, Mercado and Wharton (2003) were unsure of the placement because the species is only a member of Retusigaster recorded from the Palearctic region. I was also inquisitive about its generic placement and reviewed the species descriptions by Telenga (1955) and Tobias (1995) to reconfirm the placement of the species. Telenga (1955) wrote that the members of R. eremita possess much thickened tips of hind tibiae and simple claws, which have not been recorded in other species of Retusigaster. Both of the characters suggest placement in Pseudcardiochilus Hedwig, 1957 among the Old World cardiochilines. According to the description by Tobias (1995), the members of R. eremita have simple claws but the apices of their hind tibia are not as expanded as those of Pseudcardiochilus acutus (Tobias & Alexeev, 1977). However, because I did not examine type specimens of R. eremita, I do not change the generic placement of the species. Erdoğan (2015) reported the first record of R. eremita from Turkey, but the species may not be R. eremita because of its extremely different body coloration.

Acknowledgments
I am grateful to Mr Chris Grinter at CAS, Ms Crystal Maier and Mr Charles Farnum at MCZ, and Dr Karen Wright at TAMU for specimen loan. I would like to thank Dr Michael Sharkey for reviewing the manuscript and Dr van Achterberg for sharing his documents about Pseudcardiochilus. Among the people at LSU, I would like to acknowledge Drs Chris Carlton, Michael Stout, and Rodrigo Diaz for help and support, and Ms Victoria Bayless for her encouragement. Finally, I am grateful to the LSU Agricultural Center and LSU Entomology Department for financial and logistical support.