Grammia, Rambur, 1866

Schmidt, B. Christian, 2009, Taxonomic revision of the genus Grammia Rambur (Lepidoptera: Noctuidae: Arctiinae), Zoological Journal of the Linnean Society 156 (3), pp. 507-597 : 523-525

GRAMMIA VIRGO (LINNAEUS) (FIGS 8, 45, 77, 101)

Bombyx virgo Linnaeus, 1758: 501 .

Euprepia virgo var. citrinaria Neumögen & Dyar, 1893: 142 .

Grammia virgo var. simplex Stretch, 1906 : pl. 5 f. 14, pl. 8 f. 4.

Apantesis virgo (L.); Franclemont, 1983: 117.

Grammia virgo (L.); Ferguson & Opler, 2006: 10.

Type material: Bombyx virgo : female lectotype in the Museum Ludovicae Ulricae Reginae, Uppsala, Sweden, designated and illustrated by Mikkola & Honey (1993) [photo examined]; type locality restricted to Philadelphia, PA, USA by Mikkola & Honey (1993).

Euprepia virgo var. citrinaria : female holotype in USNM, type no. 33683 [not examined]; type locality not stated; citrinaria is an unavailable infrasubspecific name, based on the rare yellow-hindwing form.

Grammia virgo var. simplex : types probably lost; type locality not stated; simplex is an unavailable infrasubspecific name, referring to variants lacking the forewing medial transverse band. Although Stretch (1906) did not provide a description of simplex, the illustrated specimens are identical to illustrations of virgo except that simplex lacks the medial band. Stretch (1906) did not indicate simplex as being a new name, despite indicating other new names as ‘n. var.’. This may have been an accidental omission on his part, as there are no other sources that indicate the name was published earlier.

Diagnosis: Grammia virgo is the largest member of the genus, although there is some size overlap with G. parthenice where the two occur together. Grammia virgo is distinctive, but occasionally confused with G. parthenice ; the hindwing medial discal spot of virgo is however absent in G. parthenice (very rarely lacking in G. virgo ). The forewing medial vein line is wider than the anal vein line in virgo , both equally thin in G. parthenice , and the angle of the ‘break’ in the postmedial band is more pronounced in G. virgo than G. parthenice ; although there is size overlap between the two species as a whole, G. virgo is always larger when specimens from the same geographical region are compared. Grammia virgo usually flies 3–4 weeks earlier in the season than G. parthenice .

Description: Head – Palps black, frons and vertex yellowish to creamy buff; male antennae moderately bipectinate, rami averaging 6.40 ¥ 10 - 1 mm, (N = 10); female antennae slightly bipectinate; dorsal scales dark brown to black with scattered buff scales; eyes well developed. Thorax – Vestiture black with yellowish to creamy buff borders on vertex, patagia, and tegulae; entirely dark brown to black ventrally, rarely brownish yellow tuft at base of forecoxa; legs entirely or predominantly black, tibia sometimes buff dorsally. Abdomen – Dorsal ground colour orange to orangepink, very rarely yellow in yellow-hindwing forms; apex pale buff; medial and lateral markings black; medial band usually narrow, sometimes reduced to series of dots; ventrally usually entirely black, lateral black spots sometimes incompletely confluent such that a pale medial line is evident, rarely with distal margins of segments pale. Forewing – Male forewing length averaging 24.2 mm (N = 6), averaging smallest for ssp. borea , largest for ssp. gigas ; black dorsally, vein lines complete, lines and bands yellowish buff to creamy buff; cubital band expanded slightly relative to other vein lines, radial vein line sometimes so; antemedial band usually absent entirely, or visible as slight pale bar along costa; postmedial band variable, most often present as rectangular bar across discal cell onto costa, rarely extending slightly beyond cubital vein; postmedial and subterminal bands well developed; postmedial angled strongly at M 3, not extending through postcubital stripe; fringe, costa, and anal margin concolourous with pale markings; ventral markings similar, but dark markings slightly paler, and with a distinct yellowish cast overall; sexes similar. Hindwing – Ground colour pink to orangepink, rarely yellow; black markings moderately variable in extent; antemedial spots consisting of costal, discal, and anal spots, discal spot nearly always present and well developed; postmedial and subterminal spots well developed, not confluent except near tornus; ventral markings similar, but colours more washed-out and with overall yellowish cast; sexes similar. Male genitalia – Distal portion of valve gradually tapering to rounded or slightly pointed apex; clasper poorly to moderately developed, median ridge moderately developed; uncus broad-based, process tapered to point, two ¥ as long as width of base; juxta 2–2.5¥ wider than long; aedeagus with dorsad curve at 2/3 of distance beyond base; vesica with basal and medial chamber approximately equal in length and width, not scobinate; second medial diverticulum elongate, prominent; distal chamber medium-sized, twice as long as width of medial chamber, kidney-shaped, finely scobinate; vesica longer than aedeagus. Female genitalia – Ductus bursae unsclerotized, narrowed slightly distad of appendix bursae; corpus bursae globose, four ¥ width of ostium bursae; signa round to rounded triangular, relatively small, moderately scobinate; coil of appendix bursae evenly rounded; posterior apophysis 1.5–2¥ longer than papillae anales.

Biology: Grammia virgo is essentially a species of mesic deciduous and mixed forests, associated with natural forest openings such as wetlands and meadows, and disturbed habitats such as pastures and roadsides. Adults are often common to abundant, and fly in a single yearly generation. In the northern parts of its range, virgo flies in July, and September to October in the south-eastern USA. The flight season peaks about two to three weeks earlier than G. parthenice . Both sexes come to light, although females very rarely so.

Distribution: Occurs across the boreal region from north-eastern BC, east to NS, and south to FL. Absent south and west of the northern Great Plains, and west of the Rocky Mountains, occurring as far south as central AB and at least to ND ( Fig. 101 ), with a single record from east-central NE ( Ferguson et al., 2000). There are historical records of a specimen each from Victoria and Vancouver, BC [CNC] but these appear to represent accidental introductions because there are no known extant populations west of the Rocky Mountains. There are isolated literature records for Provo, UT, and Durango, CO ( Smith, 1938a), but I have not seen these or other specimens from this region.

Molecular variation: Twenty-five samples of G. virgo representative of 12 localities from across the species’ range exhibited 14 haplotypes ( Table 2). All but two of the G. virgo haplotypes formed a relatively divergent lineage with a minimum between-species divergence of 5.6% ( Virgo lineage, Fig. 133 ); divergence within the Virgo lineage was moderate. One G. virgo haplotype fell within each of the Kodara and Western lineages, ( Figs 133 , 134 ), both highly divergent from the Virgo lineage, and resulting in unusually high within-species divergence of 7.1% ( Table 2). These haplotypes may represent introgressed mtDNA as a result of hybridization ( Schmidt & Sperling, 2008), although additional molecular data should be sought to substantiate the validity of these divergent haplotypes.

Remarks: Grammia virgo exhibits considerable geographical variation, here segregated into two subspecies: the nominal subspecies throughout Canada and most of the eastern USA, and the extremely large, vividly coloured populations of the south-eastern USA, described as a new subspecies below.

The level of mtDNA divergence within G. virgo is consistent with or exceeds that observed between some Grammia sister species, in that samples of the south-eastern USA populations ( ssp. gigas ) showed two haplotypes at least 0.9% divergent from all other G. virgo haplotypes. There was no evidence of haplotype sharing between G. virgo and other species (although two of 25 haplotypes fell well outside of the virgo lineage, see above), and I therefore consider the genotypic variation of G. virgo to reflect the genealogy of virgo rather than novel introgressed genotypes. Other than the larger size and somewhat brighter colours of specimens from these southern populations, I could find no structural differences among northern and southern specimens. I therefore treat the southeast populations as a subspecies until additional specimens can be examined for both phenotypic and genotypic variation from populations geographically intermediate between the two putative subspecies. The phylogeography of G. virgo is suggestive of more than one refugium during the last glacial maximum, with the VI12–13 lineage ( Fig. 133 ) haplotypes restricted to NC and TN.

Figures

Figure 133. Neighbour-joining tree of cox1 ‘barcode’ haplotypes of Grammia species and Holarctia obliterata. Two-letter codes correspond to the species name in a given lineage. The Kodara Lineage includes Grammia quenseli (QE1), Grammia williamsii (WI1), and Grammia virgo (VI14) haplotypes. Haplotypes of the Western and Eastern lineages are shown separately in Figs 134 and 135, respectively. Numbers in parentheses indicate number of sampled specimens.

Figure 134. Neighbour-joining tree of Western lineage haplotypes of Grammia species. Multiple symbols per row indicate multiple species and/or samples for a given haplotype. Squares indicate species with haplotype complements primarily or exclusively in the Western lineage, circles indicate predominantly Eastern lineage species. Grammia virgo haplotype VI15 (triangle) is the sole G. virgo haplotype outside of the Virgo lineage.

Figures 96–101. Distribution of examined specimens of Holarctia and Grammia.