Biting midges of the genus Culicoides in South Carolina zoos.

Biting midges of the genus Culicoides (Diptera: Ceratopogonidae) were collected during the summer of 2007 at the Greenville and Riverbanks Zoos in South Carolina with Centers for Disease Control and Prevention (CDC) traps equipped with ultraviolet or incandescent lights and baited with carbon dioxide. Sixteen species of Culicoides were collected, four of which represented more than 80%. They were Culicoides guttipennis (Coquillett), Culicoides mulrenanni Beck, Culicoides obsoletus (Meigen), and Culicoides sanguisuga (Coquillett). C. guttipennis was found on a dead colobus monkey and a dead golden-headed lion tamarin; Culicoides husseyi Wirth & Blanton was collected from an unidentified, abandoned bird's nest. Ultraviolet light-equipped traps captured significantly more Culicoides specimens than traps with incandescent light. Half of the collected species previously have been associated with vertebrate pathogens, indicating a potential risk to captive animals.

Introduction

Zoos can provide a valuable surveillance system for a variety of zoonoses. The initial isolation and identification of West Nile virus in the United States, for example, was performed by employees of the Bronx Zoo, New York, NY, in 1999 (Centers for Disease Control and Prevention 1999). This initial isolation and recognition of West Nile virus became the impetus for a mosquito-surveillance program in the Bronx Zoo. However, most other American zoos have not initiated similar surveys to detect potential epizootics (Neider, unpublished). Eastern equine encephalitis virus was detected in whooping cranes, Grus americana, from the Patuxent Wildlife Research Center, Laurel, Maryland (Dein et al. 1986). The etiological agent of tularemia, Francisella tularensis, was detected in a squirrel monkey, Saimiri sp. (ca. 1989) and another unidentified monkey species (ca. 1988) in the Sacramento Zoo, California (Farlow et al. 2001). Epizootics in zoos, therefore, offer excellent opportunities to study the epidemiology of diseases with the ultimate goal of preventing further epizootics and epidemics.

While the presence of arthropods such as mosquitoes and cockroaches has been documented in zoological settings (Greenberg and Sanati 1970), biting midges have been overlooked, though there have been reports of the pathogens transmitted by these flies. For example, viral pathogens of birds (Avipoxvirus spp.), transmitted by biting midges and other biting flies, have been reported from two species of captive owls in two facilities in Florida (Deem et al. 1997). Species of Haemoproteus, avian blood parasites transmitted by biting midges and other biting flies, have been recorded from infected birds in zoos of California (Schrenzel et al. 2003; Ziman et al. 2004), Michigan (DeJong and Muzzall 2000), Oklahoma (Halpern and Bennett 1983), and Texas (Ferrell et al. 2007). These cases demonstrate that zoos can be foci of ceratopogonid-borne zoonoses, indicating the need for collaboration between veterinarians and entomologists.

To protect endangered animals and those in conservation and breeding programs, potential threats posed by biting midges should be evaluated. The objectives of this research were to survey biting midges and assess CDC traps equipped with ultraviolet or incandescent lights for their capture efficiency in two zoos in South Carolina: the Greenville Zoo in Greenville and the Riverbanks Zoo in Columbia.

Materials and Methods

Study sites

The Greenville Zoo (N 34° 50.58′ W 82° 23.24′; ∼294 MASL) is situated in the Western Division of the Piedmont on 4.5 hectares bordered by the Reedy River (5–10 m wide), a 1 m-wide unnamed stream, businesses, and residential areas. The Riverbanks Zoo and Botanical Gardens (N 34° 00.58′ W 81° 04.56′; ∼51 MASL) occupies 69 hectares in Columbia near the confluence of the Saluda and Broad Rivers. It is in the Sandhills Ecoregion on the Fall Line of the extreme western edge of the Atlantic Coastal Plain and is bordered by Interstate 126, business properties, and recreational areas.

Collection of biting midges

Biting midges (Culicoides spp. (Diptera: Ceratopogonidae)) were collected with two types of traps: CDC traps with incandescent light (4-Watt, ABC trap, Clarke Mosquito Control, www.clarkemosquito.com) and CDC traps with ultraviolet light (Trap Model 1212, John W. Hock Company, www.johnwhock.com). All traps were suspended from trees at 1.5–2.0 m above ground in forests with approximately 50% canopy cover dominated by oaks, pines, and bamboo Phyllostachys aurea Carr. ex A. & C. Rivière (Cyperales: Poaceae). Ground cover consisted of woody plants such as poison ivy Toxicodendron radicans L. (Sapindales: Anacardiaceae) and Virginia creeper Parthenocissus quinquefolia L. (Vitales: Fitaceae). To provide a source of CO2, all traps were baited with 2.0–2.5 kg of dry ice in insulated dry-ice containers placed immediately above the traps (Igloo, Trap Model 812, John W. Hock Company). All traps were powered by 6-V, 12-amps per hour, rechargeable batteries (Model PS-6100F1, Power Sonic®, www.power-sonic.com).

Biting midges were collected from May to August 2007, using incandescent and UV traps. Two traps of each type were used in the Greenville Zoo, and three of each type were used in the Riverbanks Zoo. Incandescent and UV traps were run simultaneously for four consecutive nights per month, beginning at 16:00 h, and trap-collection containers (containing live biting midges) were retrieved at 08:00 h each of the four consecutive mornings. A trap night was defined as the period from deployment of a single trap until collection the next morning. Dead animals and nests, when available, also were examined for flies.

All specimens were fixed in 95% ethanol for subsequent clearing in phenol-alcohol and slide mounted in phenol-balsam, following the methods of Wirth and Marston (1968). Culicoides species were identified with keys and illustrations by Battle and Turner (1971), Blanton and Wirth (1979), and Wirth et al. (1985) and by comparison with specimens in the collection of WL Grogan. Representative specimens were deposited in the Clemson University Arthropod Collection and the synoptic collection of ceratopogonids maintained by WLG at Salisbury University.

Statistical analyses

The mean numbers of biting midges per trap night were analyzed using a two-way ANOVA, comparing means between trap types, using month as a block, and following the general statistical methodology of Zar (1996). Tests were considered significant at p <0.05.

Results

Sixteen species of Culicoides (Diptera: Ceratopogonidae) were collected from the zoos, with 10 species from the Greenville Zoo and 12 species from the Riverbanks Zoo (Table 1). Six species were present at both zoos: Culicoides guttipennis (Coquillett), C. obsoletus (Meigen), C. paraensis (Goeldi), C. sanguisuga (Coquillett), C. scanloni Wirth and Hubert, and C. stellifer (Coquillett). A total of 101 biting midges were collected at the Greenville Zoo, 90% of which were represented by two species, C. guttipennis (n = 59) and C. obsoletus (n = 32). At the Riverbanks Zoo, 88 specimens of Culicoides were captured, 81% of which belonged to four species: C. sanguisuga (n = 31), C. mulrenanni Beck (n = 25), C. stellifer (n = 8), and C. guttipennis (n = 7).

Table 1.

Biting midges collected from the Greenville Zoo (GVZ) and the Riverbanks Zoo (RBZ), South Carolina, with notes on the traps used to collect each species.

The incandescent traps, for both zoos combined, collected 38 individuals, representing 20% of the total number of Culicoides trapped. In the Greenville Zoo, the mean number of C. guttipennis per trap night was significantly higher in UV traps than in incandescent traps (F2, 45 = 12.8, p = 0.001) (Table 2). In the Riverbanks Zoo, the mean number of C. mulrenanni per trap night was not significantly higher in UV traps than in incandescent traps (F2, 45 = 3.7, p = 0.06). Mean numbers of C. obsoletus and C. sanguisuga caught per trap night were not analyzed statistically because they were not collected in incandescent traps.

Table 2.

Mean number of Culicoides per trap night for the two most commonly collected species at the Greenville and Riverbanks Zoos, South Carolina, for two CDC-trap types, May-August 2007.

Culicoides species were not observed feeding in the zoo, but females of C. guttipennis were collected from a dead colobus monkey, Colobus guereza Rüppell (Haplorrhini: Cercopithecidae), (one of two animals examined) and a dead golden-headed lion tamarin, Leontopithecus chrysomelas (Kuhl) (Haplorrhini: Cebidae), (one of three animals examined) in the Greenville Zoo. C. husseyi Wirth & Blanton was collected from one of four unidentified, abandoned bird nests in the Greenville Zoo.

Discussion

The 16 species of Culicoides collected in both zoos were typical of those in more rural areas of South Carolina. Suitable habitats for the immature stages of Culicoides are often abundant in zoos. The relatively higher numbers of C. obsoletus and C. sanguisuga in this study are likely due to the abundance of larval habitats of these two species in the zoos, including moist leaves and dung-laden straw (Jamnback 1965). C. guttipennis inhabits water-filled treeholes, a habitat found throughout the zoos and surrounding properties.

The collection of the coastal, salt marsh species, Culicoides furens (Poey), with a UV trap in the Riverbanks Zoo was unexpected. Adults of this species in the US Virgin Islands might be carried by prevailing winds for at least 6.4 km over low mountains (366 m) (Williams 1962), and adults in Panama could be carried up to 3.2 km by winds (Breeland and Smith 1962).

To our knowledge, C. furens has never been collected as far inland as central South Carolina; it possibly reached this locale by following the Congaree River inland from coastal habitats. The salt marsh mosquito Ochlerotatus taeniorhynchus Wiedemann (Diptera: Culicidae) is routinely collected in the vicinity of the Riverbanks Zoo (Neider, unpublished data).

Several species of Culicoides in this study have been associated with vertebrate pathogens (Table 3). For example, C. stellifer harbors Bluetongue virus in Alabama (Mullen and Anderson 1998) and West Nile virus in Louisiana (Sabio et al. 2006). Several species of Culicoides (e.g., C. crepuscularis, C. haematopotus, and C. hinmani) are vectors of Haemoproteus spp. and filarial parasites. Some species also might transmit nematodes, protozoans, and viruses to captive animals in zoos. C. obsoletus is a major vector of Bluetongue virus-8 in Europe and is the cause of epidemics throughout northern Europe (Mehlhorn et al. 2007). However, in the United States and Canada, C. sonorensis Wirth & Jones is the primary vector of Bluetongue virus (Borkent 2005), and C. obsoletus does not appear to be an effective vector in the Nearctic Region. Several other pathogens, if introduced into the United States, could possibly be transmitted by native species of Culicoides. For example, C. paraensis was collected from both zoos and is the only known vector of Oropouche virus, which causes nonfatal, febrile, flu-like infections in humans in South America and has been isolated from other primates and sloths (Mullen 2002).

Table 3.

Larval habitats, host-feeding records, and pathogens associated with biting midges (Culicoides) collected at the Greenville and Riverbanks Zoos, South Carolina.

In zoos of South Carolina, surveillance for biting midges would best be served using UV traps because of the relatively greater abundance of biting midges caught by these traps and because no species was collected exclusively in incandescent traps. Competition from other incandescent light sources in and around a zoo likely reduces the collection efficiency of incandescent traps. The efficiency of traps for capturing biting midges differed geographically, and surveillance measures might need to be tailored for specific zoos. Surveillance programs should take into account the height at which biting midges seek hosts in the canopy. Some species in our study, such as C. crepuscularis Malloch, C. haematopotus Malloch, C. hinmani Khalaf, C. scanloni, and C. villosipennis Root & Hoffman, more frequently are collected higher in the canopy (Swanson 2007). Surveillance programs at zoos should consider the target species and vary the height of traps to sample the regional biting midge community.