Introduction
The tachinid genus Trigonospila Pokorny 1886, is a small cosmopolitan genus in the tribe Blondeliini of the Exoristinae (Diptera, Tachinidae) (Wood 1985, O'Hara and Wood 2004). The blondeliines are a polyphyletic group (Wood 1985) widely distributed throughout the New World, making up approximately 10% of the tachinid fauna in North America. The concept of the tribe Blondeliini was originally proposed by Mesnil (1939) and has remained largely unchanged to date. Wood (1985) provided a concise diagnosis of the tribe based on Mesnil's work: 1) prosternum setose (though this character has been noted to be variable within Trigonospila), 2) first postsutural supraalar bristle, smaller than first postsutural dorsocentral, 3) bend of wing vein M, rounded and obtuse angled, 4) subapical scutellar bristles, long stout and divergent (Wood suggested this character may be variable with respect to tribal placement), and 5) wing vein R4+5 ending at or near the wing margin (Wood 1985).
The genus Trigonospila is widely distributed, occurring throughout the Old World, in Eurasia, Australia, Oriental and in the Afrotropical regions (Crosskey 1973, Crosskey 1980, O'Hara 2014), and seven species included in the New World, three species occurring north of Mexico, and four in South America (Wood and Zumbado 2010). The genus was originally described from a male specimen, under the type species Trigonospila picta, collected in southeastern Styria in what was then referred to as 'bundesland'. Some authors have interpreted Pokorny's original text as having the specimen originating from "monte Wechsel", but this was an erroneous translation as this is referring to a mountain pass, and not a proper noun. It was originally characterized as a small gray and black fly with an elongate cone-shaped abdomen, adorned with silver or gold tomentose bands of varying thickness. Males and females are dimorphic, with only females possessing two proclinate fronto-orbital bristles, apically pointed and downwardly incurved abdomen, as well as significant differences in coloration and pattern. Host preference in Trigonospila has been poorly understood, to date there have been no rearing records for any of the New World species with only few specimens being represented in collections. Crosskey (1973) recorded two lepidopterous hosts, belonging to Oecophoridae and Gelechiidae, for an Australian species, T. brevifacies (Hardy). Shima (2006) reported the emergence of T. vittigera from a coleopteran host in the family Tenebrionidae. Soto and Ocampo (2011), also suggested the emergence of an alleged species of Trigonospila from a coleopteran host in the family Curculionidae, however examination of these specimens by both DMW and AJF was not consistent with a diagnosis of Trigonospila (Unpublished data).
Herein we describe three new species of the genus Trigonospila, and describe the male of T. panamensis (Townsend), all reared from wild-caught caterpillars collected from Area de Conservación Guanacaste, in northwestern Costa Rica. The decision that these three are previously undescribed is based on examination of New World Trigonospila, and observation of differences in external morphology, and CO1 gene sequences. By coupling CO1 data with morphological descriptions we are able to show that the coloration patters of the abdomens are not only differences between males and females but they are consistent within species making them useful in visual species identification. This paper adds to the existing knowledge of Trigonospila by providing new records relating to distribution and host preference.
Materials and methods
As this paper forms part of a larger series dedicated to naming the tachinid fauna of ACG, the methods described herein are in referenced and adapted from earlier works by the authors (Fleming et al. 2014a, Fleming et al. 2014b, Fleming et al. 2015).
Acronyms for depositories.
BMNH - The Natural History Museum, London, United Kingdom
CNC - Canadian National Collection of Insects, Arachnids and Nematodes, Ottawa, Canada
USNM - U.S. National Museum of Natural History, Washington, D.C., USA
INBIO - Instituto Nacional de Biodiversidad, Santo Domingo de Heredia, Costa Rica
Geographic area of the study and rearing intensity
All flies and rearing information described here were found by the 35+ year–old ongoing inventory of the caterpillars, their food plants and their parasitoids of the dry forest, rain forest, cloud forest, and intergrades, in the 125,000+ ha terrestrial portion of Area de Conservación Guanacaste (ACG) in northwestern Costa Rica (Fernandez-Triana et al. 2014, Janzen and Hallwachs 2011, Janzen et al. 2009, Janzen et al. 2011, Rodriguez et al. 2013, Smith et al. 2006, Smith et al. 2007, Smith et al. 2008, Smith et al. 2009, Smith et al. 2012). The parasitoid rearing methods are described at http://janzen.bio.upenn.edu/caterpillars/methodology/how/parasitoid_husbandry.htm. In brief, caterpillars (and sometimes pupae) are found in the wild at all instars by a wide variety of search methods, and reared in captivity on the food plant species on which they were found, until they produce an adult, a parasitoid, or die of other causes. Each caterpillar is documented as an individual, as are the adult parasitic flies. The caterpillars used in this study are a result of the tireless efforts of ACG staff and collaborators. Through these efforts, science has garnered an unprecedented amount of data, providing invaluable information on parasitoid biology and that of the associated hosts.
This inventory has reared about 600,000 wild-caught caterpillars since 1978. All frequencies of parasitism reported here need to be considered against this background inventory. Equally, it is patently obvious that the inventory searches some kinds of vegetation and height off the ground much more thoroughly than others, and it also searches throughout the year. Comparison of reared species of parasitoids with those collected by net or Malaise traps demonstrates that to date, the caterpillar inventory has so far encountered well less than half the species of caterpillar parasitoids present in ACG. The largest unsampled void is the upper foliage of the canopy above about 3–4 m above the ground.
The treatment reported here is focused on placing names on the species reared, thereby preparing them for later detailed ecological and behavioral accounts and studies that will normally extend across ACG ecological groups, whole ecosystems, and taxonomic assemblages much larger than a genus.
Imaging
Our descriptions of new species are deliberately brief and only include some differentiating descriptions of body parts and colors that are commonly used in tachinid identification. These brief descriptions are complemented with an extensive series of color photos of every species to illustrate the readily-observed differences among them.
Habitus photographs were taken using a Canon T3i digital SLR, equipped with a 65mm Macro Photo Lens 1:2.8 (MP–E 65mm), mounted on a microscope track stand (AmScope, Model: TS200) modified to accept a Manfrotto QR 200PL–14 quick release plate. Images were shot in aperture priority, allowing the camera to control shutter speed at f/4.5 and take 40 images at equal distance increments. Illumination was provided with a homemade reflective dome (instruction for dome creation can be found at: http://www.cdfa.ca.gov/plant/ppd/entomology/Dome/kd-200.html) placed over a 144 LED ringlight (AmScope, Model: LED–144–YK).
The photographic series was processed from RAW format using Photoshop CS6, and digitally stacked. Each final composite image was created using Zerene Stacker Software v1.04 maximizing image quality and depth of field.
All specimens listed as examined are considered paratypes, except for the holotype which is noted separately.
Wherever a specimen label has been examined, the information is presented using the following symbols: /, indicates the end of a line; //, indicates the end of a label. Labels are presented from top (closest to the specimen) to bottom, with any comments about the label being given in square brackets.
Voucher specimen management
All caterpillars reared from the ACG efforts receive a unique voucher code in the format of yy–SRNP–xxxxx. Any parasitoid emerging from this caterpillar receives the same voucher code, and then if/when later the parasitoid is dealt with individually, it receives a second voucher code unique to it, in the format of DHJPARxxxxxxx. The voucher codes and collateral data assigned to both host and emergent parasitoids are available at http://janzen.bio.upenn.edu/caterpillars/database.lasso. To date, all DHJPARxxxxxxx coded tachinids have had one leg removed for attempted DNA barcoding at the Biodiversity Institute of Ontario (BIO) in the University of Guelph, with all collateral data and all successful barcodes permanently and publically deposited in the Barcode of Life Data System (BOLD, http://www.boldsystems.org) (Ratnasingham and Hebert 2007), and later migrated to GenBank as well. A neighbor–joining (NJ) tree (Saitou and Nei 1987) for all Trigonospila reared and DNA barcoded by this inventory through 2013 is included as a Suppl. material 1. The inventory grows continually and new specimens can be found by searching the genus Trigonospila in BOLD. Each barcoded specimen also has an accession code in the Barcode of Life Data System (BOLD) and GenBank.
Inventoried Tachinidae were collected under Costa Rican government research permits issued to DHJ since 1978, and likewise exported under permit by DHJ from Costa Rica to Philadelphia, and then to the final depository in the Canadian National Insect collection in Ottawa, Canada. Tachinid identifications for the inventory were done by DHJ in coordination with a) visual inspection by AJF and DMW, b) DNA barcoding by BIO, MAS, and BOLD, and c) correlation with host caterpillar identifications by DHJ and WH through the inventory itself. Dates of capture of each reared fly in the inventory are the dates of eclosion of the fly, and not the date of capture of the caterpillar. This is because the fly eclosion date is much more representative of the time when that fly species is on the wing than is the time of capture of the caterpillar or (rarely) finding a parasitized pupa. However, the collector listed is the parataxonomist who found the caterpillar, rather than the person who retrieved the newly eclosed fly from its rearing bag or bottle, and processed it by freezing, pinning, labeling and oven–drying. Fly biology and degrees of parasitization by these flies will be the detailed subject of later papers.
DNA barcoding
DNA barcodes (the standard 5’ region of the mitochondrial cytochrome c oxidase I (COI) gene) for all ACG inventory specimens were obtained using DNA extractions made from single legs using a glass fiber protocol (Ivanova et al. 2006). Total genomic DNA was re-suspended in 30 μl of dH2O, and a 658-bp region near the 5’ terminus of the COI gene was amplified using standard primers (LepF1–LepR1) following established protocols (Smith et al. 2007, Smith et al. 2006, Smith et al. 2008). All information for the sequences associated with each individual specimen (including GenBank and BOLD accession) can be retrieved from the Barcode of Life Data System (BOLD) (Ratnasingham and Hebert 2007) via the publicly available dataset: http://dx.doi.org/10.5883/DS-ASTRIOGO.
Generic synonyms of <italic><tp:taxon-name><tp:taxon-name-part taxon-name-part-type="genus">Trigonospila</tp:taxon-name-part></tp:taxon-name></italic> Pokorny
Trigonospila Pokorny 1886: 191. Type species: Trigonospila picta Pokorny 1886 (=Tachina ludio Zetterstedt 1849), by monotypy.
Zosteromyia Brauer and Bergenstamm 1891: 376 (72). Type-species: Myobia cingulata Macquart sensu Brauer and Bergenstamm (misidentification) (=Zosteromyia braueri Townsend 1933), by original designation. Townsend's belief that this species was misidentified needs confirmation (Crosskey 1976). [Type material not examined].
Succingulum Pandellé 1894: 52 [no included species]; Pandellé 1896: 148. Type species: Succingulum transvittatum Pandellé 1896, by subsequent monotypy. [Type material not examined].
Panacemyia Townsend 1919: 164. Type species: Panacemyia panamensis Townsend 1919, by original designation. Synonymy by Wood 1985: 85.
Gymnamedoria Townsend 1927: 283. Type species: Gymnamedoria medinoides Townsend 1927 [=Succingulum transvittatum Pandellé 1896, by original designation]. [Type material not examined].
Zosteromyiopsis Townsend 1933: 456. Type species: Myobia cingulata Macquart 1851, by original designation. [Type material not examined].
Nimiocauda Reinhard 1943: 78. Type species: Nimiocauda erilis Reinhard 1943, by original designation. Synonymy by Thompson 1963. Wood (1985) was in error in citing Nimiocauda as a new synonym, as noted by O'Hara and Wood 2004: 108.
New World species previously included in <italic><tp:taxon-name><tp:taxon-name-part taxon-name-part-type="genus">Trigonospila</tp:taxon-name-part></tp:taxon-name></italic> Pokorny
In the process of species determination, specimens provided from ACG were examined in comparison to the entire known seven-member fauna of the New World Trigonospila by both AJF and DMW. These comparisons were made based on geographical proximity of the species, as well as any similarities in life history and morphology. It was found only one species reared in ACG matches any of the known species, Trigonospila panamensis (Townsend). Differentiating comparisons are discussed in the descriptions, when necessary. Wherever possible, holotypes were compared to ACG specimens. However, it should be noted if holotype material was unavailable, direct comparisons were made with specimens present at the CNC.
erilis Reinhard 1943: 79 (Nimiocauda). Holotype female (CNC) [examined by DMW & AJF]. Type locality: United States, New York, Long Island. Type label: Holotype #F: “Wading River/L.I., N.Y./VII-1924//HOLOTYPE/Nimiocauda/erilis/Reinhard//Nimiocauda/erilis/R Rnh.//Panacemyia/erilis/R. Rnh”
melaleuca Wulp 1890: 145 (Hypostena). Holotype male (BMNH) [examined by DMW]. Type locality Mexico, Tabasco, Teapa. Type label: Holotype #M: “HOLOTYPE//small label with male symbol//Teapa,/Tabasco./March. H.H.S.//B.C.A. Dipt. II./Hypostena/melaleuca/v.d.W.//Central America./Pres. by/F.D. Godman/O. Salvin/1903-172.//HOLOTYPE/of Hypostena/melaleuca Wulp/examined 1979/D.M. Wood”
pallipes Reinhard 1953: 246 (Panacemyia). Holotype female (CNC) [examined by DMW & AJF]. Type locality: United States, Texas, College Station. Type label: Holotype #F: “College Station/May 16, 1946 Tex//H J Reinhard/Collector//HOLOTYPE/Panacemyia/pallipes/Reinhard//Panacemyia pallipes/ R. Rnh”
panamensis Townsend 1919: 164 (Panacemyia). Holotype male (USNM) [examined by DMW]. Type locality: Panama, Taboga Island. Type label: Holotype #M: “Taboga I Panama/ 26 Febr. 1912/ ABusck coll// Type No./22066 U.S.N.M.//Panacemyia panamensis”
solitaria Curran 1926: 107 (Hypostena). Holotype male (BMNH) [examined by DMW]. Type locality: Jamaica, [Panama], Blue Castle. Type label: Holotype #M: “HOLOTYPE//TYPE/Hypostena/(Tachinophyto)/solitaria/Curran/No.//C.C. GOWDEY//Blue Castle,/ JAMAICA/6.vi.1923/NO.820//Pres. by/Imp. Bur. Ent./Brit. Mus./1927-158.”
trinitatis Thompson 1963: 459 (Panacemyia). Lectotype female by designation of Wood 1985: 86 (CNC) [examined by DMW & AJF]. Lectotype locality: Trinidad, St. Augustine, ICTA. Type label: Lectotype #F: “ICTA/St Augustine/Trinidad/Aug 1959/Ex Coll. Off Samanea Flowers//F.D. Bennett Collector//LECTOTYPE #f/of Panacemyia/trinitatis Thompson/ designated 1984/D.M. Wood”
verticalis Reinhard 1953: 247 (Panacemyia). Holotype male (CNC) [examined by DMW & AJF]. Type locality: United States, Ohio, Amherst. Type label: Holotype #M: “Amherst, Ohio/VI-15 1924//H J Reinhard/Collector//HOLOTYPE/Panacemyia/verticalis/Reinhard//Panacemyia/verticalis/R. Rnh”
Diagnosis of the genus <italic><tp:taxon-name><tp:taxon-name-part taxon-name-part-type="genus">Trigonospila</tp:taxon-name-part></tp:taxon-name></italic> Pokorny
Wood (1985) provided a diagnosis and review of the genus which is adapted here. The specimens reared from ACG presented in this study conform to the keys found in Wood (1985) and Wood and Zumbado (2010).
Head: male lacking proclinate fronto-orbital bristles; reclinate orbital bristles of male indistinguishable from frontals; ocellar seta hair-like, nearly parallel to each other in male, parallel or divergent in female; eye bare, or with minute inconspicuous hairs; parafacial bare, extremely narrow; lower margin of face at level of vibrissa not visible in profile; facial ridge with a few small recumbent setae on lower third or less; subvibrissal ridge short, usually with 3 or fewer bristles; anterior margin of postgena concave anteriorly, sloping anteroventrally toward vibrissal angle, without genal dilation; first flagellomere of male about as long as that of female; arista minutely to short pubescent, thickened on basal fourth to fifth.
Thorax: prosternum bare (this character state, a rarity in Blondeliini, is contrary to Mesnil 1939, diagnosis of the tribe, however as with most biological systems, the tribe is not defined by a single character state, but rather the combination of the other diagnostic characters has maintained its present placement; proepisternum bare; postpronotum usually with 2 bristles, or if 3, the inner basal bristle is usually small forming a slightly curved row; katepisternum with 3 bristles, the anteroventral sometimes small; lateral scutellar bristles shorter than subapical bristles, curved medially; apical scutellar bristles usually lacking; fore tibia with 1 posterior bristle; mid tibia with 1 anterodorsal bristle; vein R2+3, usually with a single bristle at base of R4+5.
Abdomen (Figs 1, 2): abdominal mid-dorsal depression not extending to hind margin of syntergite (ST) 1+2; ST1+2, T3 and T4, each bearing median marginal bristles; median discal bristles almost as long as marginal bristles of same segment; in some species male an extra pair of discal bristles present in front of main pair; abdomen of female strongly curved ventrally.
According to Wood (1985) members of Trigonospila share many similarities with Dolichotarsus Mesnil, 1977, Embiomyia Aldrich, 1934, Steleoneura Stein, 1984, and Pararondania Villeneuve, 1916. The genus is distinguished from Steleoneura by the antennal structure, in which the pedicel of Steleoneura is longer than the first flagellomere, and the scapes are separate at their bases; Steleoneura also possessing a single relatively long bristle at base of R4+5, a single long straight bristle on prosternum, and a small bristle on katepimeron. In Dolichotarsus the female abdomen is more laterally compressed, and the membranous ovipositor is fully withdrawn into the abdomen. In the New World species of Trigonospila, the arista is less noticeably pubescent than in most of the Old World species, but the female abdomen appears to be similar in all of them.