Abstract

Three new species of the genus Pseudacteon are described, all from Chiapas, Mexico, and all of which are parasitoids of the ant Azteca instabilis. Sternite 6 of Pseudacteon dorymyrmecis Borgmeier is illustrated for the first time, and P. confusus Disney is synonymized with this species. The natural history of the Azteca-Pseudacteon interaction is described.

1. Introduction

The species of the phorid fly genus Pseudacteon Coquillett have been under intense scrutiny lately because of their potential to control invasive species of fire ants (Solenopsis invicta and S. saevissima in North America; S. geminata elsewhere) [13]. Other lesser-studied species of Pseudacteon, many of them undescribed, attack different hosts, including species of Crematogaster, Lasius, Liometopum, Nylanderia, Pseudolasius, and other small ants.

In a series of papers [47], a new host record, with Azteca instabilis Fr. Smith, has been documented from southern Mexico. Below, the so-far-known species of Pseudacteon associated with these ants are described, their identification clarified, and natural history summarized.

2. Materials and Methods

Specimens were collected into 70% alcohol and dried using hexamethyldisilazane [8]. They were deposited in the following collections:CEET: El Colegio de la Frontera Sur, Collección de Insectos Asociados a Plantas Cultivadas en la Frontera Sur, Tapachula, Chiapas, Mexico,LACM: Natural History Museum of Los Angeles County, California, USA,MCZC: Museum of Comparative Zoology, Harvard University, Massachusetts, USA,MUCR: Universidade de Costa Rica, San Jose, Costa Rica,USNM: Smithsonian Institution, Washington, DC, USA.

3. Systematics

Pseudacteon Coquillett [9]; full synonymy in Borgmeier, 1968 [10]; type species. Pseudacteon crawfordi Coquillett, original designation.

Note on gender: the word Actaeon is a Greek name for a (male) hunter; thus, the name Pseudacteon means “false Actaeon” and is masculine in gender.

Pseudacteon laciniosus new species (see Figures 1(a) and 2(a)).

Diagnosis
The last general key to adults of Pseudacteon is that of Borgmeier [11]. This new species keys to couplet 10, where a user is given the alternatives of “ovipositor lanceolate” versus “ovipositor subcylindrical, tapering at apex.” Since the protruding stylet could fool users into accepting the first alternative, these flies could key out to either P. dorymyrmecis Borgmeier in the first lead or P. onyx Steyskal in the second. Both of these species are markedly smaller than P. laciniosus, and both have narrower oviscapes. Further, based on examination of the holotype specimen, sternite 6 of P. dorymyrmecis has a long pair of medial setae originating basally on the segment (see Figure 2(b)). The oviscape of P. onyx was illustrated by Borgmeier [12] and is much more ventrally curved than that of P. laciniosus. Finally, unlike both of the other species, P. laciniosus has a dark brown body that strongly contrasts with its yellowish legs.

Disney (in [13]) described a new species, Pseudacteon confusus, that also keys to P. dorymyrmex. We examined a paratype female of P. confusus, comparing it to the holotype female of P. dorymyrmex, and conclude the two are the same species. Therefore, P. confusus is a junior subjective synonym of P. dorymyrmex (new synonymy).

Description (Female)
Body length 1.2–1.5 mm (mean = 1.3). Frons dark brown, with 2-4-4-4 setae and one pair of proclinate supraantennal setae. Flagellomere 1 dark brown, rounded, flat; length of arista about two times that of flagellomere 1. Palpus light brown, setulae thick. Thorax dark brown. Scutellum with two pairs of large setae, anterior pair 0.8 length of posterior pair. Legs yellowish brown, except forecoxa lighter, mid- and hind coxae darker brown. Wing with mean costal length 0.36 wing length, range 0.35–0.37. Halter yellow. Abdominal tergites dark brown, ventral abdominal membrane dark gray. Tergite 6 with large posterior emargination and lateral pair of setae. Sternite 6 with one large (0.18 mm) pair of slightly divergent setae; more lateral pair much smaller (0.06 mm), but larger to subequal in size to median pair on one or both sides in some specimens; with scattered setulae (see Figure 2(a)). Oviscape (Figure 1(a)) with narrow anterior process, convex dorsally, with large lateral flanges anteriorly, terminating bluntly, although preserved specimens often with stylet protruding, making oviscapes appear pointed. Minute setae on dorsal surface of oviscape arranged in laterally concave lines, with larger, divergent, apical pair.Holotype. ♀, Mexico: Chiapas: Finca Irlanda, 15°11′ N, 92°20′ W, July 2010, S. Philpott, over Azteca instabilis (barcode LACM ENT 139561) (CEET).

Paratypes
16♀, same data as holotype (CEET, LACM, MCZC, USNM).Etymology. The specific epithet is from a Latin word for “fringed,” referring to the enlarged lateral margin of the oviscape.

Pseudacteon planidorsalis new species (see Figures 1(b) and 2(c))

Diagnosis
The species is similar to Pseudacteon dorymyrmecis, but differs in the presence of longer setae flanking the relatively shorter medial pair on sternite six.

Description (Female)
Body length 1.1–1.4 mm (mean = 1.3). Frons dark brown, with 2-4-4-4 setae and one pair of proclinate supra-antennal setae. Flagellomere 1 dark brown, rounded, flat; length of arista about two times that of flagellomere 1. Palpus yellow, setulae thick. Thorax brown. Scutellum with two pairs of large setae, anterior pair 0.47–0.60 length of posterior pair. Legs yellowish brown, except forecoxa lighter. Wing with mean costal length 0.38 wing length, range 0.37–0.39. Halter yellow. Abdominal tergites dark brown, ventral abdominal membrane dark gray. Tergite 6 with large posterior emargination and lateral pair of setae. Sternite 6 with one large (0.09 mm) pair of large setae and smaller lateral setae in basal transverse row; smaller lateral setae scattered more posteriorly (see Figure 2(c)). Oviscape (Figure 1(b)) lacking narrow anterior process, flat dorsally, with large lateral flanges anteriorly, terminating in rounded point (which is difficult to see in Figure 1(b) because the oviscape is downturned apically). Minute setae on dorsal surface of oviscape arranged in laterally convex lines, with slightly larger apical pair.

Holotype
♀, Mexico: Chiapas: Finca Belen, 15°15′ N, 92°23′ W, 8.ii.2003, S. Philpott, over Azteca instabilis (barcode LACM ENT 294148) (CEET).
Paratypes
3♀, same data as holotype (LACM).

Etymology
The specific epithet is from a Latin word for “flat backed,” referring to the surface of the oviscape.

Pseudacteon pseudocercus new species (see Figures 1(c), 1(d) and 2(d))

Diagnosis
This species can be recognized by the strongly downturned, lightly sclerotized oviscape with the pair of cercuslike apical processes. Other species of Pseudacteon with a bilobed oviscape have the structure strongly sclerotized and dark brown in color.

Description (Female)
Body length 1.2 mm. Frons dark brown, with 2-4-4-4 setae and one pair of proclinate supra-antennal setae. Flagellomere 1 dark brown, rounded, flat; length of arista about two times that of flagellomere 1. Palpus light brown, setulae small, thin. Thorax brown. Scutellum with two pairs of large setae, anterior pair 0.47–0.60 length of posterior pair. Legs yellowish brown, except forecoxa lighter. Wing with costa 0.40 wing length. Halter yellow. Abdominal tergites brown, ventral abdominal membrane gray. Tergite 6 with large posterior emargination, pair of lateral setae near midline, and longer, thicker seta more laterally. Sternite 6 anteriorly emarginate, with scattered small setae (see Figure 2(d)). Oviscape (see Figures 1(c) and 1(d)) without narrow anterior process, convex dorsally, apically downturned with pair of cercuslike lobes.

Holotype
♀, MEXICO: Chiapas: Finca Belen, 15°15′ N, 92°23′ W, 8.ii.2003, S. Philpott, over Azteca instabilis (barcode LACM ENT 294147) (CEET). No other specimens preserved.

Etymology
The specific epithet is from Latin words for “false circus,” referring to the apex of the oviscape.

4. Natural History

Little is known about the life cycle of P. laciniosus, P. planidorsalis, and P. pseudocercus, as the three species are only known from adults. Further, most natural history information available treats the flies as a genus, rather than as individual species, so more work will be necessary to distinguish between them. Adults have been observed in a range of shaded coffee plantations varying in canopy cover from ~25 to 100% in the Soconusco region of Chiapas, Mexico. Specifically, P. laciniosus, P. planidorsalis, and P. pseudocercus have been observed in Finca Irlanda, Tapachula municipality (15°11′ N, 92°20′ W), between 800–1100 m elevation, Finca Hamburgo, Tapachula municipality, between 800 to 1100 m elevation (15°10′ N, 92°19′ W), and in Finca Belen, Huixtla municipality, between 800 to 1200 m elevation (15°15′ N, 92°23′ W). The shade coffee habitats from which the phorids have been seen range from shaded monocultures with relatively low levels of canopy cover, tree diversity, and density to rustic coffee plantations with a high diversity and density of shade trees and nearly 100% canopy cover [14]. Preliminary work indicates that the relative abundance of the three species in a range of coffee agroecosystems is similar, even as canopy conditions change [15].

Within the shade coffee habitats, females of P. laciniosus, P. planidorsalis, and P. pseudocercus have only been observed when hovering over or ovipositing in the host ant species, A. instabilis, or flying out of leaf litter collected from the ground or on tree trunks and branches near to A. instabilis nests. Males of the three species have not been collected or identified. As a group, the flies are attracted to the host ant by an alarm pheromone (1-acetyl-2-methylcyclopentane) released from the dorsal section of the ant gaster, but they do not attempt to oviposit without ant movement [16]. Work is underway to determine whether visual and similar chemical cues are used by each species in host location and host selection processes. Once a female fly locates host individuals, it will remain in the area for up to several minutes, closely hovering over and following moving ant individuals. Individual flies have been observed to attempt to oviposit at least a dozen times before disappearing from view; actual oviposition has not yet been quantified. Several individuals of the three species (up to 8–10) have been observed simultaneously around the same A. instabilis nest.

Phorids strongly modify the behavior of the A. instabilis ants and thereby indirectly affect other insects in coffee agroecosystems. Azteca instabilis is an aggressive, canopy-dominant ant that has important impacts on many members of the coffee insect food web [17]. In the presence of Pseudacteon flies, A. instabilis ant foraging is reduced (by about 50%) for up to 90 min. after the first appearance of the phorid [4]. Once the Pseudacteon arrives near an A. instabilis nest, the ants will either (1) run back to their nest, or to hiding places under tree bark or (2) remain motionless with their heads tilted back [17]. This reduction in ant activity allows other species of ants to gain access to food resources [3, 5] and reduces the predatory effects of ants on lepidopteran larvae [4] and the coffee berry borer (Hypothenemus hampei Ferrari) [18]. Furthermore, A. instabilis normally prevent adults of the coccinellid beetle Azya orbigera Mulsant from feeding on scale insects (Coccus viridis Green), a keystone mutualist of the ant. When the A. instabilis are under attack by the phorids, A. orbigera greatly increase their feeding rates [19] and oviposition rates (Hsieh and Perfecto, unpublished data). The host ant, A. instabilis, is patchily distributed within coffee agroecosystems, and one force maintaining this distribution and relative abundance of colonies within sample areas may be attacks from the phorid flies [7]. Thus the Pseudacteon flies, through their influence on the activity and distribution of this keystone species, likely have widespread impacts on the coffee insect food web. This result contrasts with conclusions of studies with Pseudacteon tricuspis Borgmeier and P. curvatus Borgmeier and fire ants (S. invicta). At least some studies have concluded that phorids attacking S. invicta do not have long-term impacts on the ants or associated arthropods (e.g., [20]). Phorids that attack A. instabilis do not lower population sizes, similar to findings with P. tricuspis and P. curvatus, but likely impact coffee food webs due to their role in maintaining a similar number of colonies and reducing ant behavior.

Field evidence and observations suggest that P. laciniosus, P. planidorsalis, and P. pseudocercus adults probably live near their host, in leaf litter on the ground. Data suggest that the fly population is likely distributed in a density dependent manner, as number of attacks on A. instabilis individuals are more frequent and more numerous where A. instabilis densities are greater [6, 7]. Likewise, field evidence indicates that Pseudacteon adults are usually located within the leaf litter on the ground because time to first oviposition attempt on A. instabilis adults placed on the ground is much less than for ants placed at 1.5 m above ground [6]. However, more rapid arrival on the ground could mean that the phorids primarily search for hosts at ground level.

5. Key to Females of Pseudacteon Attacking Azteca in Southern Mexico

This key is intended for ant ecologists who need to identify phorid parasitoids from known ant hosts. A new general key to New World Pseudacteon is needed, as researchers currently must use a combination of Borgmeier [11], Porter and Pesquero [21], Plowes et al. [1], the key below, and reference to species not covered in the previously listed keys [2226].

(1) Apex of oviscape with lobelike processes (Figures 1(d) and 2(a)); venter of segment 6 with short
setae only (Figure 2(d)) ……………………………………………………………………………P. pseudocercus new species
- Apex of oviscape without lobe-like processes (Figures 1(a) and 1(c); venter of segment 6 with some
long setae…………………………………………………………………………………………………………………………………………………2
(2) Oviscape anteriorly with narrow process, dorsally domelike, with small setulae scattered posteriorly,
but longer pair near apex; apex of oviscape truncate (Figure 1(a)); enlarged ventral setae of segment 6
placed posterior to midpoint of sternite, much longer than other ventral setae (see Figure 2(b))
……………………………………………………………………………………………………………P. laciniosus new species
- Oviscape anteriorly with broad apex; dorsally flattened on apical third (except for downturned tip);
small setulae in laterally convex rows; apex of oviscape pointed; enlarged ventral setae of segment 6
placed anterior to midpoint of sternite, only slightly longer than those directly lateral (see Figure 2(d))
………………………………………………………………………………………………………………P. planidorsalis new species

Acknowledgments

The authors thank Brian Koehler and Inna Strazhnik for expertly rendering the illustrations and Vladimir Berezovskiy and Giar-Ann Kung for technical assistance. Dr. R.H.L. Disney kindly loaned them a paratype specimen of P. confusus for comparison, and Dr. Carlos Lamas loaned them the holotype of P. dorymyrmecis. This research was supported by NSF Grant DEB-1025922 to B. Brown and P. Smith, DEB-1020096 to S. Philpott, and DEB-0349388 to I. Perfecto and J. Vandermeer.