Biology of Pseudacteon Decapitating Flies (Diptera: Phoridae) That Parasitize Ants of the Solenopsis saevissima Complex (Hymenoptera: Formicidae) in South America

Pseudacteon flies (Diptera: Phoridae) parasitize individual ant workers, causing decapitation of the host during pupariation. Phorid flies that attack South American fire ants in the Solenopsis saevissima (Smith) complex are distributed across a wide range of habitats and climates associated with the geographical range of their hosts. Sympatric species sharing the same hosts often partition niche resources by season, active time of day, host size, and/or different host activities. They have the potential of being used for biological control of the imported fire ants in North America, Australia, and Asia.


Introduction
Pseudacteon decapitating flies are specific parasitoids of ants [1]. At least 22 species of Pseudacteon flies are known to attack South American fire ants in the Solenopsis saevissima (Smith) complex [2][3][4]. Seventeen species are parasitoids of the black imported fire ant, Solenopsis richteri, and the red imported fire ant, Solenopsis invicta [2]. These fly species are specific to S. saevissima complex fire ants and do not occur in populations of Solenopsis geminata complex fire ants native to North and Central America [2,[5][6][7]. Under forced laboratory conditions, some of these fly species (e.g., Pseudacteon curvatus, Pseudacteon obtusus, Pseudacteon tricuspis, and Pseudacteon wasmanni) can parasitize fire ants in the geminata complex; however, rates of successful parasitism are always very low [8][9][10][11]. A few flies of Pseudacteon pradei and P. wasmanni were observed attacking S. geminata workers in trays in the field, but these attacks did not result in successful parasitism [9,12]. First-generation flies reared from S. geminata complex ants in laboratory still maintained a strong preference for their normal S. saevissima complex hosts upon emergence [11]. Extensive field observations and forced laboratory tests have demonstrated that decapitating flies which parasitize Solenopsis fire ants are not capable of parasitizing ants in other genera [10]. Six highly specific phorid fly species in the genus Pseudacteon from native South America have been successfully released in the USA for biological control of the imported fire ant [13]. In this review, we focus on life history, distribution, phenology, and activity patterns of parasitic Pseudacteon flies, and their association with fire ants. Pseudacteon borgmeieri, Pseudacteon cultellatus, P. curvatus, Pseudacteon nudicornis, Pseudacteon nocens, P. obtusus, and P. tricuspis are able to successfully develop in both S. invicta and S. richteri [17,18,[20][21][22]. Other phorid species may attempt to oviposit on other saevissima complex fire ants such as Solenopsis macdonaghi and Solenopsis quinquecuspis, but the success of these attacks has not been confirmed [21]. Newly emerged flies are ready to mate and repeat the attack cycle within several hours of eclosion [7]. As completely starved flies rarely live beyond one day, provision of water increases longevity by two days in the laboratory [24]. Sugar feeding can further increase the longevity considerably depending on rearing temperature and fly activity [24][25][26]. The ability to digest monosaccharide sugars (fructose, glucose) and disaccharide sugars (sucrose, trehalose) may explain significant but modest increase in longevity of P. tricuspis feeding on cotton aphid honeydew [27,28]; therefore, phorid flies can be expected to feed on nectar and/or honeydew sources in field.
The number of eggs produced by a female Pseudacteon varies from about 200 to almost 300 [29]. Female flies attack at a rate of several workers per minute under laboratory conditions and can last for an hour or more with unlimited hosts [6]. Females may attack more frequently if potential hosts are abundant. However, some attacks are not successful. Only 11-35% of oviposition attempts result in successful parasitism and larval development in laboratory tests [6]. In native Argentina, the maximum parasitism rate per S. invicta colony by phorid flies is around 2.8% [30]. The low success rates may be attributed to the following possibilities: (1) oviposition strikes are too rapid to eject an egg, (2) flies are selective to quality of workers for egg laying through oviposition attempt, and (3) some of the oviposited eggs do not develop due to infertility or ant defenses [16].
Pseudacteon species with different sizes prefer different sizes of fire ant workers [31,32]. In some larger Pseudacteon species like Pseudacteon litoralis, P. nocens, P. obtusus, and P. tricuspis, male flies are smaller than female flies from the same species, but their size ranges may overlap to some degree [17]. Sex ratio of these flies is thought to be determined by the body size of the host with larger ants producing more females and smaller ants producing more males [33]. However, sexual size dimorphism in smaller fly species like P. cultellatus and P. curvatus does not appear to occur [17,20,34]. There are no gender differences in host sizes used in P. cultellatus [20] and P. curvatus [17]. Host size distributions in P. obtusus and P. nocens were predominately overlapping, but female flies did not develop in the smallest ants and males did not develop in the largest ants [17,18]. In one field test, P. curvatus and P. tricuspis preferred monogyne fire ant colonies, and more offspring are produced from monogyne than polygyne fire ant colonies, possibly due to poor adaption to polygyne fire ants, which are far less abundant than monogyne ones in South America [34,35]. Because worker size in polygyne form is smaller than in monogyne form [36], more male flies are expected to be produced from the polygyne form [37]. Sex ratios in Pseudacteon flies do not appear to be affected by environmental conditions [33].

Distribution
The geographic distribution of Pseudacteon flies in South America has been extensively sampled in the past two decades [3]. As host-specific parasitoids, the geographical range of these flies is largely determined by the range of their hosts, in addition to climatic factors and habitats [38]. The S. saevissima complex fire ants occur in vast areas from the Amazon Basin of Brazil, west to the Andes and south to ≈42 • S latitude in the Río Negro province, Argentina [39,40]. Early investigations showed that the range of S. richteri was southernmost Brazil (Rio Grande do Sui), Uruguay, and northern Argentina and that S. invicta occupied a 3000 km long, relatively narrow band centered on the headwaters of the Paraguay River, northward into the Amazon drainage along the Guapore River, and southward into Argentina and Paraguay [41][42][43]. Solenopsis richteri extended from approximately 30 • S to 38 • S in South America, whereas S. invicta extended from 10 • S to 33 • S between Porto Velho, Rondonia Territory, Brazil and Rosario, Argentina [42,44]. The two species are parapatric in their native South America and do not hybridize apparently even where their ranges overlap [44,45]. Their native ranges overlap only in a small area (southern Santa Fe Province, near Rosario) of central Argentina [44]. A large overlapping area seems likely because the southern boundary of the predicted possible range expansion for S. invicta matches the current southern range limit of S. richteri in South America [46]. While S. invicta occupies the tropical and subtropical areas, S. richteri occupies the more temperate ecological niche. Consequently, Pseudacteon flies follow the distribution of their hosts and are broadly distributed across a wide range of climates and habitats.
Fire ants with broader distributions often have Pseudacteon species with wider ranges [38]. Furthermore, a fly species using multiple hosts appears to have larger geographical ranges. In South America, P. cultellatus and P. obtusus have the broadest distributional ranges of the Pseudacteon species associated with S. saevissima complex [38]. These two flies attack many common South American fire ants including Solenopsis interrupta, S. invicta, S. macdonaghi, S. richteri, and S. saevissima [3,21]. Their ranges stretch from subtropical to temperate climates and habitats from Brazil to Argentina [38]. Pseudacteon obtusus occurs at the highest altitude (2280 m), the southernmost latitude (Corralito, Río Negro, Argentina, 40 • 43.786 S), and the westernmost longitude (Santa Cruz, Tucumán, Argentina, 65 • 46 W) [47]. Pseudacteon curvatus, P. litoralis, and P. tricuspis are the most abundant and common species in Argentina and Paraguay [47]. Pseudacteon curvatus and P. nocens are very abundant and widespread in northern Argentina in varying climatic conditions but rare in Brazil [3,38,47]. Pseudacteon species with narrower distributions could be more host-limited, utilizing single hosts with smaller ranges. Pseudacteon bulbosus, which paratisizes S. interrupta and a species close to Solenopsis electra, is found only in the Santiago del Estero province in Argentina [47][48][49]. Pseudacteon bulbosus and Pseudacteon comatus were often extremely rare in several extensive surveys [47,48,50]. The range of Pseudacteon conicornis is known only from S. saevissima along the Atlantic coastline of Brazil and appears to be restricted additionally by climate or habitat [3]. The known westernmost record for Pseudacteon fly species occurring in South America is a new species morphologically similar to P. obtusus Pseudacteon species are active at 16-37 • C, 20-90% RH, and 0-11.6 km/h wind speed [47]. Those flies with broader geographical distributions appear to have greater climatic (very dry and with extreme temperatures) tolerance [38]. Pseudacteon bulbosus, P. curvatus, P. litoralis, P. nocens, Pseudacteon obtusitus, and P. tricuspis were found to be active under greater climatic stress [47].
Local Pseudacteon communities typically consist of multiple fly species. Argentina has higher species diversity than the nearby countries [47]. Researchers have found 13 sympatric fly species attacking S. interrupta at sites in northern Argentina, 10 different species attacking S. invicta in northeastern Argentina, and 7 species attacking S. richteri in eastern Argentina [47,48]. Pseudacteon cultellatus, P. curvatus, P. litoralis, P. nocens, P. obtusus, and P. tricuspis are often codominant species [47,48]. Solenopsis saevissima (north climate) and S. invicta (south climate) have a parapatric distribution in the state of São Paulo, Brazil, apparently determined by the climate. Two different communities of decapitating flies were associated with S. saevissima in the north and with S. invicta in the south [51]. The most abundant species in the northern community were Pseudacteon affinis, P. cultellatus, Pseudacteon dentiger, Pseudacteon disneyi, and Pseudacteon fowleri, and in the southern community were P. litoralis, P. pradei, P. tricuspis, and P. wasmanni. The community structure of these flies largely depends on niche partitioning of common host resources. The rich phorid guild found in South America is apparently associated with species diversification of their host complex.

Phenology and Activity Pattern
Most phorid fly species are active throughout the year in their native range but with different seasonal and daily activity patterns. The abundance of flies in South America is generally high in spring, possibly due to significant host colony activity caused by the mating flights and/or because of the start of the rainy season [30,52]. The peak abundance varies with fly species throughout the year. Pseudacteon curvatus reaches peak abundance in summer, P. comatus, P. nudicornis, P. obtusus, and P. tricuspis in fall, and P. borgmeieri in winter [47,50]. In winter months, there are fewer species and lower abundance, and most fly species disappear during cold winter [50]. Furthermore, the abundance patterns of phorid flies usually vary among different areas. For instance, six species of S. invicta-decapitating flies were found active throughout the year in Corrientes [30], whereas only one species of S. richteri-decapitating fly was active throughout the year in Buenos Aires [50]. The warmer winter in Corrientes explains greater phorid activity throughout the year. Overall abundance for a given species can be high in spring and fall in northeastern Argentina and in summer in east central Argentina [47]. In Argentina, at least six common fly species, P. cultellatus, P. litoralis, P. nocens, P. obtusitus, P. obtusus, and P. tricuspis, are abundant and present in each month [30,48]. As they oviposit all year round in a temperature range of 18-40 • C, these flies have continual overlapping generations [5]. There can be multiple annual peaks of abundance. The fluctuations in abundance of most fly species can be interpreted, at least partially, by climatic variables related to temperature, moisture, and rainfall because some species are apparently favored by high humidity or warm temperatures [30,50,53]. In addition to environmental factors, the complex seasonal patterns of fly abundance may also depend on genetically based intrinsic physiology, host species, and habitat [30].
Fly species exhibit different activity patterns throughout the day. Host ants can be active all day and night, but Pseudacteon parasites are only active during the day. In some areas, fly abundance reaches high level at dusk, but species diversity reaches high level around midday. Pseudacteon affinis, P. disneyi, P. litoralis, and P. nocens are usually active in the early morning and late afternoon (more temperate species), whereas P. cultellatus, P. curvatus, P. dentiger, P. nudicornis, P. obtusitus, P. obtusus, and P. tricuspis are mainly active during the middle of the day [47,53,54]. Sometimes fly species (e.g., P. litoralis, P. nocens and P. obtusus) can be found attacking ants from morning until dusk in cool season [47,48,50]. Daily activity of phorid flies can be partially explained by climatic variables such as mean temperature, relative humidity, and light intensity. The influence of light intensity on fly occurrence is always associated with temperature changes. At low temperatures, greater numbers of flies are in the sun, whereas at higher temperatures, greater numbers of flies are in the shade. Approximately one-half of the species are found most commonly in full sun and the other half in full shade [55,56]. It seems reasonable that the relative abundances of phorid flies associated with shade are related to the overall effect of solar radiation on the microhabitat rather than to light preference.
Diversified sympatric species sharing the same host resources partition niche space by season and time of day. The species richness of phorid flies may increase gradually in summer time. The seasonal and daily activity patterns of phorid fly match its host activity patterns and are also constrained by the host's activities. The activity amplitudes of flies correlate to periodic ant foraging activity, i.e., host availability. The ant foraging activity generally increases over the course of the day, and species richness and abundance of phorid fly increase synchronically. Both Pseudacteon flies and their ant hosts tune their life cycles to seasonally changing abiotic conditions [50].

Phorid Fly-Fire Ant Association
A wide array of phorid species can be found attacking S. invicta and other saevissima complex fire ants [47,48,50,55]. Multiple Pseudacteon species in sympatric communities attack almost all sizes of fire ant workers in nature [31,32]. The coexistence of multiple fly species on a single host resource suggests that these flies may partition ecological niches with time of day or season, different host sizes, and attraction to different host locations [37,48,50,57]. In the presence of phorids, S. invicta workers usually adopt different foraging strategies like covering food resource with debris and dirt, foraging in tunnel systems, and/or increasing nocturnal activity. Some Brazilian colonies show a strong freezing response to the presence of phorid flies. The hiding behavior may allow fire ants to regain control of large food resources when the flies leave [58]. The presence of phorids, however, can also lead to turnover of resources from host species to competing ant species by reducing its competitive dominance in competitive interactions [59]. Phorid parasitism pressure is considered to be responsible for these behavioral adaptions.
In field, phorid flies were more likely to be found near disturbed ant mounds and/or trails. The flies of P. litoralis, P. tricuspis, and P. wasmanni, attacking ants near disturbed mounds, are expected to use alarm or defense compounds released by the ants as host location cues, and those of P. borgmeieri, Pseudacteon nuicornis, P. obtusus, and Pseudacteon solenopsidis, attacking near trails, are considered to use trail pheromone as a cue [60]. Therefore, workers are the primary source of long-range cues attracting phorid flies. Disturbance of fire ants causes the release of many defensive compounds, including alarm pheromone and venom alkaloids [61,62]. Electrical stimulation can be more effective in attracting phorid flies than mechanical mound disturbance presumably because it stimulates release of alarm pheromones and venom [63]. Defensive compounds other than alarm pheromones may also play an important role in this attraction since they are part of the alarm response and components of the complex alarm odor. Fire ant venom alkaloids prove to be key attractants for both male and female P. tricuspis [64]. Like the females, males of the phorid fly species are attracted to both fire ant alarm pheromone and venom alkaloids. Male flies of several species (e.g., P. obtusus and P. tricuspis) are also attracted to host ants apparently for finding mates [14]. The alarm pheromones shared by closely related ant species may serve as a general host location cue from long range, while the species-specific venom alkaloids are probably used as short-range cues for host location and host preference. The ratios of the cis and trans alkaloids and presence of some minor piperideine alkaloids [65,66], rather than abundance of the cis alkaloids, may be responsible for host preference by phorid fly for closely related ant species. Although alates contain a significantly higher concentration of the alarm pheromone than workers of S. invicta, P. tricuspis females rarely attack alates and are more attracted to worker ants compared to alates [7,[67][68][69]. Fire ant mating flights often attract large numbers of phorid flies due to release of alarm pheromone triggered by alate activity [69][70][71].

Conclusions
Phorid flies in the genus Pseudacteon are highly specific parasitoids of ants. A female fly lays an egg into the thorax of a live worker ant, and the larva eventually decapitates the host ant after consuming all head tissues. Some flies prefer large workers, whereas others tend to attack small to medium-sized workers [32,37]. Furthermore, many Pseudacteon flies are sympatric, exhibiting different annual and daily activity patterns. In addition to the direct effect of mortality on their host ants, Pseudacteon phorids affect fire ant foraging behavior and have population-level impacts on the survival of the host ants [2]. The presence of Pseudacteon phorids may weaken the competiting vigor of the host species relative to other ant species in the community. Because of these direct and indirect effects, Pseudacteon phorids are a promising group for biological control of invasive Solenopsis fire ants.