Revision of the genus Heteranassa Smith, 1899 (Lepidoptera, Erebidae, Omopterini)

Abstract Heteranassa Smith (Erebidae, Omopterini), native to the southwestern United States and Mexico, includes two recognized species, namely Heteranassa mima (Harvey) and Heteranassa fraterna Smith. These are separated mainly by subtle differences in wing color and pattern, leading to speculation about the validity of the described species. This study examines variation in external and internal morphology across the geographic range of the genus, aiming to clarify species limits, describe morphology, and provide a comprehensive assessment of variation within the genus. Results indicate that Heteranassa fraterna syn. n., is a junior synonym of Heteranassa mima.


Introduction
Heteranassa Smith, 1899, is a genus of moths native to warmer desert regions of southwestern United States southward to southern Mexico, currently containing two valid species. Heteranassa mima (Harvey, 1876), described from Texas, and H. fraterna Smith, 1899, described from Death Valley, California. Mustelin (2006 synonymized Specimens were also examined during visits to the United States National Museum (USNM, M. Pogue) and the McGuire Center for Lepidoptera and Biodiversity (MGCL, A. Warren). A complete list of specimens examined is included in Suppl. material 1. Dissected material was selected to represent the range of size and coloration found across the range of Heteranassa. A list of dissected specimens is included in Suppl. material 2.
Genitalic dissections follow techniques described by Hardwick (1950) and McCabe (1980). Terminology follows Forbes (1923Forbes ( , 1954 and Mikkola et al. (2009). Abdomens were removed from specimens by gently applying upward pressure near the end of the abdomen with a pair of angle-tipped forceps. Abdomens were cleared in 10% KOH overnight or in hot KOH for 15 minutes. The abdomens were then placed in a watch glass with distilled water, and scales were removed with a fine brush. Once clear of scales, the integument of the abdomen was cut along the left pleural membrane, and the genital capsule removed. On male specimens setae were carefully removed from the membranous costal region of valves with a fine camel's hair brush. The aedeagus was separated from the valves by grasping the distal end with fine-tipped forceps, and gently pulling to separate from the juxta. The ductus seminalis was then cut where it enters the side of the proximal part of the aedeagus. The vesica was then carefully teased out of the aedeagus with a #20 minuten with the tip bent to a right angle, held in a standard pin vice, and with water pressure from a syringe. A syringe with a modified 30 gauge needle ( Fig. 1) was used to force water into the opening of the ductus seminalis to help evert and inflate the vesica. The aedeagus was transferred to 95% ETOH to dehydrate. The vesica was inflated with ETOH, following the procedures described by McCabe (1980), for several seconds using the modified syringe. The valves were transferred to 95% ETOH, and the membranous lobes of the sacculus were inflated with 95% ETOH. Abdomens of female specimens were cleared in hot KOH. The abdomen was cut along the left pleural membrane, and then a circular incision was made around abdominal segment VIII to remove the female genitalia. Genitalia dissections were placed in Chlorazol Black® stain for ~10 seconds, then transferred to 95% ETOH to dehydrate overnight. The genitalia were placed overnight in orcein stain dissolved in 2-propanol. Genitalia and abdomen "pelts" were stored in pin-mounted glycerin vials. Material from the first author's personal collection was slide mounted following Winter (2000). The structure of the male genitalia did not allow the valves to be spread and flattened without damaging the sacculus, juxta, and transtilla. Whole specimens were cleared and stained following similar procedures as described for the genitalia.
Photographs of dissected specimens, genitalia, and adults were made using a Visionary Digital imaging system (http://www.visionarydigital.com, R. Larimer). Line drawings were made in Adobe Illustrator with a Wacom Intuos 4 drawing tablet from photographs or sketches made with a drawing tube attached to a Wild M5 stereomicroscope.
Eggs were obtained from gravid females collected in Box Canyon, Pima Co., AZ (18 July 2012). The females were placed in brown paper bags with Honey Mesquite (Prosopis glandulosa Torr.) foliage and bark. Larvae were reared on P. glandulosa foliage.
Specimens of Eubolina impartialis Harvey, Matigramma species, Acritogramma metaleuca (Hampson), Toxonprucha species and Coxina species are frequently misidentified as Heteranassa. Of these, A. metaleuca is the most similar to Heteranassa (Franclemont 1986). Acritogramma metaleuca can be most easily distinguished by the presence of spine-like setae on the mesothoracic tibia, and there are also subtle differences in wing pattern (Franclemont 1986). Acritogramma metaleuca has no brown lines or shading on the forewing, and the discal spot is distinctly lunulate. Eubolina impartialis is similar to both Heteranassa and A. metaleuca but has a brownish ground color on the hindwings, instead of grayish white, and spine-like setae on the mesothoracic tibia. From southern Texas into Mexico, Heteranassa may be confused with co-occurring Coxina species. This genus shows affinities to Heteranassa in forewing pattern and genitalia, but a lighter hindwing ground color serves to separate Heteranassa. Additionally, similarities in wing pattern and genital morphology suggest a relationship to the Caribbean and South American genus Elousa Walker. The ranges of Heteranassa and Elousa may overlap in southern Mexico. Elousa can be separated from Heteranassa by its smaller size, and the light gray to white mottling of the forewings. Toxonprucha species are generally smaller than Heteranassa, and they possess hindwings with a darker ground color and more distinct patterning than those of Heteranassa. A key to Heteranassa and similar species is provided below.
Taxonomic history. Harvey (1876) described Homoptera mima from a single female from Texas, listing Belfrage as the source of the specimen. He referred the species to the genus Homoptera Guenée, but did not mention any characters used to determine generic placement. Grote (1882), in a checklist, moved H. mima to the genus Eubolina Harvey, 1875, again without any mention of characters used. Smith (1899) described H. fraterna and H. minor and placed these species in the genus Campometra      Guenée, 1852. Campometra fraterna was described from a series of six lightly-marked specimens collected in Death Valley, California., and a single specimen from Catalina Springs, Arizona. Campometra minor was described from a series of five small female specimens collected in Arizona. Smith described these two species as new based on differences in size, coloration, and patterning. Smith (1899) proposed the genus Heteranassa to circumscribe H. mima, H. fraterna, and H. minor based on the absence of spine-like setae on the mesothoracic tibia in these three species. Smith (1899) wrote, "I prefer leaving them with Campometra temporarily, until all of the allied genera can be carefully studied, but suggest the term Heternassa in case generic separation seems desirable." These three species were formally referred to Heteranassa by Smith et al. (1903). Todd (1982) reviewed Smith's type series and designated lectotypes for H. minor and H. fraterna. The pupa of Heteranassa was first described by Comstock (1955), and Crumb (1956) gave a description of the larva of Heteranassa.
In his study of southern California Noctuoidea, Mustelin (2006) determined H. minor to be a synonym of H. fraterna. He found no differences in genital morphology between the types of H. fraterna and H. minor (Mustelin 2006). Mustelin (2006) did not examine the type specimen of H. mima, located in the Natural History Museum, London.
Eggs. Dark bluish gray, ~1/2 mm diameter; captured females laid eggs singly or in groups of less than five in crevices of host plant bark, or singly on sides of enclosing container.
Larvae. Variable in color; eggs developed into adults within five weeks; observations are consistent with Comstock (1955) and Crumb (1956). Larvae pupated before high-quality photographs could be taken.
Variation. Specimens tend to be larger in the eastern part of the range in Texas, and smaller specimens are more common in Arizona and California. Forewing coloration ranges from dark gray with some brown dusting to tan. Maculation ranges from  well-defined antemedial, postmedial, and subcostal lines to lightly marked specimens with only the subcostal line well defined. Lightly marked specimens are found most commonly in the Mojave Desert. Ground color of hind wings is lighter towards the western range of the species. Specimens from the eastern part of the range show distinctly marked discal spots and shading on the margins on ventral surface of the wings, and the undersides are more heavily dusted with darker scales. The size of the white patch in the reniform area varies from a narrow dash to a large spot, while forewing ground color ranged from dark gray to gray brown among moths reared from the same female collected in Southeast Arizona.
Barcode variation in Heteranassa is very conservative. Examination of more than 160 full-sequence (658 base-pair) barcodes from California, Arizona, New Mexico, Texas, and northern Mexico showed a maximum divergence of less than 0.8%. One haplotype * dominated the sample, representing more than half of the specimens; the other barcodes included 36 haplotypes that had no more than two base-pair differences from each other. One haplotype, restricted to central and southern Texas, departed from this pattern in being 0.8% different from those from farther west. This is most probably the haplotype that should be associated with the name Heteranassa mima, it being described from this part of Texas. However, this "eastern" haplotype is found with "western" haplotypes in central Texas and there is no indication in genital structural characters, or wing color or pattern, that Heteranassa includes more than a single species. The barcodes of Heteranassa are so divergent that they give no indication of a close relationship to any other erebid genus, other than belonging in the subfamily Erebinae, tribe Omopterini. Heteranassa specimens from Texas and Mexico are frequently confused with some species associated with the genus Coxina Guenée, which can have a similar superficial pattern, but the barcodes are more than 10% different and the two genera do not appear to be closely related. Distribution and habitat. Warm, arid habitats from California to Texas, northward to Oklahoma, and south as far as Oaxaca, Mexico (Fig. 16). A single specimen from Cartwright, Manitoba is in the LACM.
Discussion. The variation in Heteranassa wing pattern and coloration is continuous, with many specimens appearing intermediate to the phenotypes described by Smith (1899) and Harvey (1876). Genitalic morphology does not, however, correlate with wing pattern differences. These observations suggest that Heteranassa contains a single, highly variable species, H. mima. Studies of another erebine genus, Catocala, have shown that pressure from avian predators may drive high levels of polymorphism in forewing pattern and coloration (Ricklefs and O'Rourke 1975, Bond and Kamil 2002, Webster et al. 2009), and Heteranassa may be subject to similar evolutionary processes.
A series of Heteranassa from Death Valley, California collected in February, 2005, is the most variable in forewing pattern and coloration among the thousands of specimens observed to date. Heteranassa comprised roughly 90% of the moth specimens collected during this period, demonstrating that the genus is an abundant and likely ecologically important insect herbivore in North American desert biomes.
During the course of this research, we became aware of potential taxonomic affinities with the neotropical genera Elousa Walker (1857) and Coxina Guenée (1852). These genera have not been studied in a systematic framework since the turn of the 20th Century. A preliminary examination of male genitalia and wing pattern show significant overlap of characteristics between the genera. These three genera lack spines on the mesothoracic tibiae, and possess symmetrical male genitalia with membranous costal regions of the valves. These processes are larger in Heteranassa and Elousa albicans (Walker, 1857) than they are in Elousa schausi (Giacomelli, 1911) and the other Coxina species we have dissected. We have examined 10 species in these genera from the Caribbean and South America. Specimens belonging to this group that we collected in the Nicaraguan highlands appear more similar to E. schausi specimens from Argentina and Coxina specimens from Mexico, south Texas, and Florida, than they do to Caribbean Elousa or North American Heteranassa. Future research could test the monophyly of Coxina and Elousa with respect to Heteranassa, and how these genera speciated in North and South America and the Caribbean.