Lloyd W. Morrison and Lawrence E. Gilbert

Department of Zoology and Brackenridge Field Laboratory
University of Texas

Austin, TX 78712 USA

Abstract

We tested the host specificity of two South American Pseudacteon phorid flies, P. obtusus Borgmeier and P. borgmeieri Schmitz, on North American colonies of the red imported fire ant, Solenopsis invicta Buren, and the tropical fire ant, S. geminata (F.). Sequential host specificity tests conducted in the laboratory indicated that P. obtusus was highly specific to S. invicta. In individual trials, 20 females attacked S. invicta but none attacked S. geminata. Pseudacteon borgmeieri females, in contrast, attacked both Solenopsis species. Six of the eighteen known South American Pseudacteon species have now been tested for host specificity, and 4 of the 6 reveal a high degree of specificity to S. invicta, thus representing good biocontrol candidates.

Key Words: Everglades, ants, Brazilian pepper, Schinus, Strumigenys, exotic, Formicidae

Resumen

Se prob- la especificidad de dos moscas phorid Sudamericanas, P. obtusus Borgmeier y P. borgmeieri Schmitz, en colonias norteamericanas de la hormiga roja de fuego importada, Solenopsis invicta Buren, y la hormiga de fuego tropical S. geminata (F.). Pruebas secuenciales de especificidad de hospederos efectuadas en laboratorio indicaron que P. obtusus fue altamente específica a S. invicta. En pruebas individuales, 20 hembras atacaron a S. invicta pero ninguna atac- a S. geminata. En contraste, hembras de Pseudacteon borgmeieri atacaron a las dos especies de Solenopsis. Hasta la fecha se ha probado la especificadad de huespedes en seis de las 18 especies Pseudoacteon sudamericanas conocidas. Cuatro de las seis han mostrado un alto grado de especificidad a S. invicta, por lo cual se les considera buenos candidatos para el control biol-gico de la hormiga roja de fuego importada.

Phorid flies in the genus Pseudacteon are biocontrol candidates of pest Solenopsis ant species (Feener and Brown 1992, Orr et al. 1995, Porter et al. 1995b). An important component in any biocontrol program is the assessment of host specificity in the proposed biocontrol agents (Simberloff & Stiling 1996, Samways 1997). All known Pseudacteon species are ant parasitoids, and all Pseudacteon species that parasitize Solenopsis ants are apparently restricted to that genus (Disney 1994, Porter et al. 1995a).

Previously, we reported on the host specificity patterns of 4 South American Pseudacteon species that are known to parasitize workers in the S. saevissima complex (which includes the red imported fire ant, S. invicta Buren and the black imported fire ant, S. richteri Forel) (Gilbert and Morrison 1997). In laboratory tests, three of the four species (P. litoralis Borgmeier, P. tricuspis Borgmeier, and P. wasmanni [Schmitz]) rarely attempted to parasitize the tropical fire ant, S. geminata (F.), a species in the geminata complex and native to the southern U.S., but readily attacked populations of S. invicta. In contrast, P. curvatus Borgmeier attacked S. geminata relatively frequently.

At least 18 species of Pseudacteon phorid flies attack saevissima complex fire ants in South America (Porter 1998a), and are potential candidates for fire ant biocontrol in the U.S. Because variation does exist among species in their relative host specificity (Gilbert and Morrison 1997), it is critical that each species under consideration be screened before release in the U.S. This article describes the host specificity of two Pseudacteon species, P. obtusus Borgmeier and P. borgmeieri Schmitz.

Materials and Methods

Pseudacteon obtusus females were collected from a residential area near the University of Campinas (UNICAMP), São Paulo State, Brazil (n = 18) and near the Reserva Ecol-gica Costanera Sur in Buenos Aires, Argentina (n = 2). Pseudacteon borgmeieri females were obtained from roadsides in the foothills near Jundiai, São Paulo State, Brazil (n = 3), and near the Reserva Ecol-gica Costanera Sur in Buenos Aires, Argentina (n = 2). Solenopsis mounds were disturbed and aspirators were used to collect female phorids when they attempted to parasitize exposed worker ants. All female phorids were transferred to the containment facility at The University of Texas Brackenridge Field Laboratory where host specificity tests were conducted. Polygyne (multiple queen) colonies of S. invicta were obtained from the grounds at Brackenridge Field Laboratory and Circle C Ranch, Travis County, Texas. Polygyne colonies of S. geminata were obtained from Circle C Ranch and along Barton Hills Drive in Austin, and along the Colorado River in Mills County, Texas.

The methodology of host specificity testing was exactly the same as the sequential choice procedure described in Gilbert and Morrison (1997) for 4 other Pseudacteon species previously tested, and complete details of the procedure can be found there. Briefly, each female was evaluated individually by introducing it to a flight box of S. invicta, then to a flight box of S. geminata, and finally back to a flight box of S. invicta. If the female did not demonstrate a motivation to oviposit on S. invicta (i.e., did not attack within 20 minutes), the trial was aborted. If the female did begin to attack S. invicta within 20 minutes, the attack rate was recorded over a 5 minute period, beginning with the first attack, and then the female was immediately transferred into a flight box of S. geminata. After 20 minutes in the S. geminata flight box, the female was immediately transferred back into the original S. invicta flight box. By this method we could be certain that females that ignored S. geminata were motivated to oviposit before and after exposure to S geminata. In each case attack rates were based on the time between the first attack and the removal of the phorid. Once a phorid began to attack, it usually did not stop until it was removed from the tray.

The orientation time, or time elapsed between introduction of a phorid into the flight box and the first attack on an S. invicta worker, was compared for first and second exposures to S. invicta for P. obtusus females. A Wilcoxon matched-pairs signed-ranks test (Daniel 1990) was used.

Tests were conducted at intervals from June 1995 to June 1998. Voucher specimens of Pseudacteon spp. have been retained by the authors and deposited in the Natural History Museum of Los Angeles County, CA and the Museu de Hist-ria Natural, UNICAMP, São Paulo State, Brazil.

Results

Pseudacteon obtusus revealed a very high degree of host specificity for S. invicta. Of 20 females that attacked S. invicta, none attempted to oviposit on S. geminata when transferred to a flight box containing only S. geminata. Many P. obtusus females hovered low over S. geminata for a few minutes, but then turned away and flew near the top of the flight box or rested on the side. Eighteen of the 20 females that initially attacked S. invicta were transferred back to the S. invicta flight box after exposure to S. geminata, and 14 of these 18 resumed attacking S. invicta. The attack rate per individual female on initial exposure to S. invicta was 0.94 ± 0.66 attacks per minute (mean ± SD), compared with 0.98 ± 0.77 attacks per minute on subsequent exposure to S. invicta.

Total time spent with S. invicta (after the first attack) on the initial exposure for all 20 P. obtusus females considered collectively was 129.2 minutes, in which time 111 attacks occurred, for an overall attack rate of 0.86 attacks per minute. On the subsequent exposure, 18 P. obtusus females collectively spent a total time of 310.4 minutes with S. invicta and attacked 199 times, for an overall attack rate of 0.64 attacks per minute. Twenty P. obtusus females collectively spent a total of 386.1 minutes with S. geminata without attacking.

The orientation time for P. obtusus was 9.5 ± 6.0 (mean ± SD) minutes on the first exposure, compared with 5.1 ± 4.4 minutes on the second exposure. Although this difference was not significant (P = 0.20; Wilcoxon matched-pairs signed-ranks test), it follows the pattern observed in the 4 previously tested Pseudacteon species of having shorter orientation times on the second exposure to S. invicta (Gilbert & Morrison 1997).

In contrast, P. borgmeieri readily attacked S. geminata. Of the 5 P. borgmeieri females that attacked S. invicta, 4 (2 from Brazil and 2 from Argentina) also attacked S. geminata. Individual attack rates were higher on exposure to S. invicta than on subsequent exposure to S. geminata (0.63 ± 0.21 [mean ± SD] versus 0.22 ± 0.25 attacks per minute, respectively). Because P. borgmeieri frequently attacked S. geminata, this phorid species was not exposed to S. invicta a second time. Considering all females collectively, P. borgmeieri spent 39.5 minutes with S. invicta and attacked 24 times, for an overall attack rate of 0.61 per minute. Pseudacteon borgmeieri spent 93.4 minutes with S. geminata and attacked 16 times, for an overall attack rate of 0.17 per minute.

Discussion

The rationale for the design of the testing procedure and the conservative nature of the tests were described in detail in Gilbert & Morrison (1997). Previously, in similar tests with other Pseudacteon species, we found that P. litoralis, P. tricuspis, and P. wasmanni attacked S. invicta almost exclusively, with only a few oviposition attempts observed on S. geminata (Gilbert & Morrison 1997). Porter & Alonso (1999) obtained similar results in independent laboratory tests. A fourth species tested only by Gilbert & Morrison (1997), P. curvatus, was found to attack S. geminata relatively frequently. Complementary tests conducted in the field in Brazil confirmed the high degree of host specificity of P. litoralis and P. tricuspis, both of which parasitized saevissima complex species but ignored S. geminata (Porter 1998b).

The data reported here are directly comparable with the data presented for P. litoralis, P. tricuspis, P. wasmanni, and P. curvatus in Gilbert & Morrison (1997). Overall, P. obtusus had a lower rate of attack on S. invicta, both on an individual and collective basis, than the 4 species previously tested. It was also characterized by a longer orientation time, in both the initial and subsequent exposures to S. invicta. Attack rates of P. borgmeieri were the lowest of the six species tested, although the sample size for this species was small. Like P. curvatus, P. borgmeieri often attacked S. geminata, but attacked S. invicta more frequently (Gilbert & Morrison 1997). A summary of the relative host specificity of the 6 Pseudacteon species tested so far is presented in Table 1. The attack rate ratio illustrates, for each Pseudacteon species, the amount of parasitism pressure exerted on S. geminata relative to that exerted on S. invicta.

Although P. obtusus did not oviposit as frequently as the other S. invicta-specific species tested, it may still exert significant parasitism pressure on S. invicta, and therefore be useful in a biocontrol program, for two reasons: First, we have found a positive correlation to exist between size of Pseudacteon species and size of worker ants attacked (Morrison et al. 1997). Pseudacteon obtusus is a small phorid in most areas, with an average mesonotum width of 0.35 mm (Orr et al. 1997). The other three Pseudacteon species that are known to be relatively host specific to S. invicta (P. litoralis, P. tricuspis, and P. wasmanni) have average mesonotum widths ranging from 0.47 to 0.57 mm (Morrison et al. 1997, Orr et al. 1997) and are all significantly larger than P. obtusus (Orr et al. 1997). Although we did not quantify the size of worker ants attacked by P. obtusus, it was qualitatively obvious that P. obtusus preferred smaller workers than the larger Pseudacteon species. Second, P. obtusus was found to be more common along Solenopsis foraging trails than disturbed mounds in South America, while P. litoralis, P. tricuspis, and P. wasmanni were all found predominantly at mounds and only rarely along foraging trails (Orr et al. 1997).

Although relatively few P. borgmeieri were tested, preliminary results indicate that this species may attack S. geminata in nature. We did not attempt to rear offspring from S. geminata workers parasitized by P. borgmeieri. It is possible that some Pseudacteon species may attempt to oviposit on S. geminata but are unsuccessful (or less successful) in inserting an egg, or offspring may not complete development in this host. Porter & Alonso (1999), however, were able to get P. tricuspis females to attack S. geminata workers by surrounding them with S. invicta colony odors, and reared a single P. tricuspis to the adult stage in a S. geminata worker. Pseudacteon curvatus has also been reared to the adult stage in S. geminata (S. D. Porter, USDA-ARS, Gainesville, pers. comm.). Thus development in S. geminata is possible for some, if not all, South American Pseudacteon species.

Table 1. Comparison of host specificity among 6 South American Pseudacteon species tested in the laboratory.

Species1
Number of
females attacking
S. invicta
% of females
also attacking
S. geminata
Attack rate
ratio2
P. obtusus
20
0
0
P. tricuspis
25
4.0
0.001
P. litoralis
23
8.7
0.013
P. wasmanni
18
11.1
0.066
P. curvatus
20
65.0
0.301
P. borgmeieri
5
80.0

0.282

1Species are listed in order of decreasing specificity to S. invicta. Data on P. tricuspis, P. litoralis, P. wasmanni, and P. curvatus are from Gilbert & Morrison (1997).

2Attack rate ratio is the overall attack rate (= cumulative number of attacks/cumulative amount of time all females of the species spent in the flight box) on S. geminata / overall attack rate on S. invicta (on initial exposure). (For example: an attack rate ratio of 0.3 would indicate that, over the same period of time, 3 S. geminata workers were attacked for every 10 S. invicta workers.)

Even if Pseudacteon species that attempt to oviposit on S. geminata do not frequently complete development in S. geminata, the fact that these Pseudacteon species are attracted to S. geminata and will attempt oviposition indicates the need for caution in the selection of species for biocontrol agents. This is because in general Pseudacteon phorids appear to infect relatively few Solenopsis ants and thus have a very small direct effect on mortality of their hosts (Jouvenaz et al. 1981, Morrison et al. 1997). The mere presence of Pseudacteon phorids, however, has been documented to induce behavioral changes in Solenopsis workers, disrupting foraging efficiency and decreasing the competitive ability of Solenopsis species relative to other ants in the community (Feener & Brown 1992, Orr et al. 1995, Porter et al. 1995b, Morrison: 1999). Thus the major impact of Pseudacteon phorids on Solenopsis ants appears to be mediated through behavioral changes in workers when flies are present, and exotic Pseudacteon species that would potentially harass and attempt oviposition on S. geminata could have a negative impact on S. geminata populations even if the flies do not frequently complete development in the novel host.

Eighteen Pseudacteon species that attack S. saevissima complex workers in South America are now known (Porter 1998a). The four Pseudacteon species that show a strong preference for S. invicta (P. tricuspis, P. litoralis, P. wasmanni, and P. obtusus) all have trilobed ovipositors, whereas the two that are less specific (P. curvatus and P. borgmeieri) have an unlobed ovipositor. It remains to be seen whether this association will hold true for all species. In limited laboratory observations of P. solenopsidis, which has an unlobed ovipositor, we did not observe any oviposition attempts on S. geminata (unpublished data). This species, however, exhibits an attack behavior quite different than other South American Pseudacteon species that have been studied in this respect, and is characterized by a low attack rate (Orr et al. 1997). Further laboratory testing of host specificity preferences will be necessary for this species. The majority (13 of 18) of described South American Pseudacteon species have trilobed ovipositors (Porter 1998a).

If Pseudacteon phorid flies do prove to be effective biocontrol agents on imported fire ants, a potential strategy would be to introduce Pseudacteon species of different body sizes, to increase the range of worker sizes attacked (Morrison et al. 1997), and to introduce species that attack workers at both mounds and along foraging trails (Orr et al. 1997). Pseudacteon obtusus would be a welcome addition to the known ‘arsenal’ of larger, disturbed mound Pseudacteon species that are highly specific to S. invicta (P. litoralis, P. tricuspis, and P. wasmanni).

Acknowledgments

R. Guerra and E. A. Kawazoe (both University of Texas at Austin) assisted with this project in Texas. W. Benson, C. G. Dall’Aglio-Holvorcem, S. Seike (all Universidade Estadual de Campinas), and M. Orr (San Francisco State University), helped collect Pseudacteon spp. in Brazil. P. Folgarait collected Pseudacteon spp. in Argentina. We thank B. Brown (Natural History Museum of Los Angeles County) for verifying species identifications. CNPq and IBAMA (Brazil) and USDA APHIS supplied the necessary permits. This research was funded by grants from The Fondren Foundation, Ewing Halsell Foundation, R. J. Kleberg and H. G. Kleberg Foundation, the Houston Livestock Show and Rodeo Educational Fund, and the State of Texas Fire Ant Research and Management Committee (FARMAC).

References Cited

Daniel, W. W. 1990. Applied Nonparametric Statistics. 2nd Ed. PWS-Kent, Boston. 635 pp.

Disney, R. H. L. 1994. Scuttle flies: The Phoridae. Chapman & Hall, London. 467 pp.

Feener, D. H., Jr., and B. V. Brown. 1992. Reduced foraging of Solenopsis geminata (Hymenoptera: Formicidae) in the presence of parasitic Pseudacteon spp. (Diptera: Phoridae). Ann. Entomol. Soc. Am. 85: 80-84.

Gilbert, L. E., And L. W. Morrison. 1997. Patterns of host specificity in Pseudacteon parasitoid flies (Diptera: Phoridae) that attack Solenopsis fire ants (Hymenoptera: Formicidae). Environ. Entomol. 26: 1149-1154.

Jouvenaz, D. P., C. S. Lofgren, and W. A. Banks. 1981. Biological control of imported fire ants: A review of current knowledge. Bull. Entomol. Soc. Am. 27: 203-208.

Morrison, L. W. 1999. Indirect effects of phorid fly parasitoids on interspecific competition between fire ants (in press). Oecologia.

Morrison, L. W., C. G. Dall’Aglio-Holvorcem, and L. E. Gilbert. 1997. Oviposition behavior and development of Pseudacteon flies (Diptera: Phoridae), parasitoids of Solenopsis fire ants (Hymenoptera: Formicidae). Environ. Entomol. 26: 716-724.

Orr, M. R., S. H. Seike, W. W. Benson, and L. E. Gilbert. 1995. Flies suppress fire ants. Nature 373: 292.

Orr, M. R., S. H. Seike, and L. E. Gilbert. 1997. Foraging ecology and patterns of diversification in dipteran parasitoids of fire ants in south Brazil. Ecol. Entomol. 22: 305-314.

Porter, S. D. 1998a. Biology and behavior of Pseudacteon decapitating flies (Diptera: Phoridae) that parasitize Solenopsis fire ants (Hymenoptera: Formicidae). Florida Entomol. 81: 292-309.

Porter, S. D. 1998b. Host-specific attraction of Pseudacteon flies (Diptera: Phoridae) to fire ant colonies in Brazil. Florida Entomol. 81: 423-429.

Porter, S. D., and L. E. Alonso. 1999. Host specificity of fire ant decapitating flies (Diptera: Phoridae) in laboratory oviposition tests. J. Econ. Entomol. 92:110-114.

Porter, S. D., H. G. Fowler, S. Campiolo, and M. A. Pesquero. 1995a. Host specificity of several Pseudacteon parasites of fire ants in South America (Diptera: Phoridae; Hymenoptera: Formicidae). Florida Entomol. 78:70-75.

Porter, S. D., R. K. Vander Meer, M. A. Pesquero, S. Campiolo, and H. G. Fowler. 1995b. Solenopsis (Hymenoptera: Formicidae) fire ant reactions to attacks of Pseudacteon flies (Diptera: Phoridae) in Southeastern Brazil. Ann. Entomol. Soc. Am. 88: 570-575.

Samways, M. J. 1997. Classical biological control and biodiversity conservation: what risks are we prepared to accept? Biodiversity & Conserv. 6: 1309-1316.

Simberloff, D. S., and P. Stiling. 1996. Risks of species introduced for biological control. Biol. Conserv. 78: 185-192.