[Ant Control Issue] Managing Ants in the National Zoological Park

Features - Ant Control Issue

Just how do you manage two non-indigenous ant species — the Asian needle ant and the big-headed ant — in an account as sensitive as a zoo?


The National Zoological Park in Washington, D.C., is home to many exotic animal species, all there by design. However, several years ago, two non-indigenous ant species showed up unsolicited and created significant management obstacles, one of which, to this day, has not resolved one of the problems. From a pest management standpoint, zoos present challenging ecological habitats and within them the most diverse range of microenvironments that seem to be designed for pests in mind, e.g., vast array of food, extensive harborages and in many cases year round humidity and temperature suited for mass reproduction. To further compound the problem, zoos located in northern climates are ripe for the introduction of many species that would only thrive in warmer climtes.

Big-Headed Ant. The big-headed ant, Pheidole megacephala, is a small, two-node ant most easily recognized by their two different sized (dimorphic) workers. Colonies have a major worker (1/8 in long) with a disproportionately large head, and a minor worker (1/10 in long) of normal proportions. Major workers comprise only about 2 percent of the population so they might not be observed upon an initial investigation of an ant colony.

Most species of big-headed ants are located in warmer climates in the United States where they are known to primarily live outside and are considered to be occasional invaders of buildings and homes. They normally nest in soil, under rocks and wood, but are also found within potted plants that can facilitate their transportation into other locations. They feed on a variety of food types, such as insects and honeydew produced by aphids, mealybugs and scale insects.

At the National Zoological Park (NZP), treatment of big-headed ants began in 2008 inside one of the bird breeding facilities at the Smithsonian Conservation Biology Institute (SCBI) in Front Royal, Va. SCBI is a 3,000-acre research facility located in the foothills of the Shenandoah Mountains, and serves as an education and breeding facility for rare and endangered species. Bird keepers at SCBI noticed the ants feeding on orange slices hung on large ficus trees for the amakihis (Loxops virens), a small, olive-green song bird native to Hawaii. With high humidity in the building, wooden cage frames and a desirable food source, the big-headed ants soon spread throughout the building. The birds could not be relocated, so to suppress the ant population, boric acid-based sugar solutions were placed around the perimeter and top of the exhibits.

Fig. 2. Ants forage for honeydew on palm tree in Bird House indoor flight cage.

In 2009, NZP’s Integrated Pest Management (IPM) staff began to notice small numbers of big-headed ants inside several buildings in Washington D.C., including the Great Ape, Small Mammal, and Bird Houses, and the Reptile Discovery Center. By 2011, the ants had become well established throughout both the Bird and Small Mammal Houses, with large populations occurring inside the large interior exhibits such as the Bird House’s indoor flight cage. Ants were found aggregating under rocks, plant pots, and underneath the root systems of woody plants in the exhibits, and nesting in the leaf sheaths of palm trees (see Figure 2 on the right). While some brood (eggs, larvae and pupae) were occasionally observed these areas did not appear to be nesting sites. Staff observed ants emerging from the bottoms of plant pots, and cracks in the cement walls and hollow exhibits. In the indoor flight cage, ants actively formed foraging trails up the trees to reach honeydew, (a common food source for big headed ants) from scale insects on the large palms.

To resolve the problem in both the Small Mammal and Bird Houses, IPM staff (see Figure 3 below) worked with Richard Kramer (Innovative Pest Management) to conduct a food preference trial using several boric acid sugar-based baits to determine the ant’s bait preferences (see table on the right). Initial trials considered boric acid baits due to their low toxicity to birds and mammals. Ants fed on all bait formulations, however they preferred the syrup and gel formulations. Three granular baits, containing abamectin, orthoboric acid, and hydramethylnon, were also tested mainly outside the exhibits because there was concern that the ants would cache the solid bait and potentially expose the animals in the exhibits.

Click on the picture above to see the full table.

For several years, boric acid carbohydrate baits were used in the bird breeding facility in SCBI with little success. While the population was suppressed, limited control only was achieved after the ants were excluded from their major alternate food source, oranges. Since alternate food sources were available in the Bird and Small Mammal Houses and the number of ants was significantly larger, products containing different active ingredients, indoxacarb and fipronil, were selected. In addition to different active ingredients the bait matrixes for both products were different, i.e., indoxacarb (recommended by Kramer as being highly attractive to ants) is carbohydrate based while the fipronil bait designed for carpenter ants is protein based.

Fig. 1. Holes were drilled into the tamper-resistant bait tubes so ants could access the tubes but non-target animals could not.

The zoo’s veterinary staff was familiar with fipronil, due to its animal registrations, but indoxacarb is a relatively new compound with limited toxicological data on exotic species. Therefore, the veterinary staff requested that animal contact with indoxacarb, either directly or indirectly via secondary exposure, be minimized. Ten tamper-resistant bait tubes (see Figure 1 on the right) were constructed using ½-inch diameter PVC pipe, tees and caps. Holes were drilled into the PVC tubes so that ants could access the tubes while birds, non-target animals, and other insects would not be able to reach the bait. The caps were not glued to facilitate the installation of bait.

For the test in the Bird House, 10 sites were selected based upon ant activity. Of those 10 test sites, five were randomly selected to receive treatment while the remaining five were used as controls (no treatment). Each treatment site was baited using 1 oz. of ant gel bait containing 0.05 percent indoxacarb that was injected into PVC bait tubes. Although an ideal control would have been the same ant gel bait matrix without the active ingredient, control bait tubes were not baited as the matrix was not available. All bait stations were placed under rocks, plant pots or other similar stationary items, to prevent contact by exhibit animals.

Fig. 3. Left to right: Anthony Hiza (NZP), Richard Kramer (IPM) and Greg Ose (NZP).

At six to seven day intervals, the 10 tubes were removed, weighed and cleaned to record amount of bait consumed at each of the five baited sites. Inspections of all 10 sites were made when bait tubes were removed to record ant activity levels, based upon visual estimates of the number of workers: observations were classified as being small (S) = 24 or fewer ants found; medium (M) = 25 to 1,000; or large (L) = 1,000+.

The same five treated tubes were again injected with 1 oz. of the same ant gel bait while the five controls remained untreated. All tubes were reinstalled at their original locations. Average daily and weekly gel bait consumption was calculated and recorded (see Fig. 4 below). During the fourth week, the 10 sites were inspected and the numbers of workers observed were recorded. Over the four-week baiting period with indoxacarb consumption initially decreased and then plateaued with almost all of the bait being consumed weekly. Ant activity at the baited sites increased dramatically whereas activity at control sites and on plants with honeydew decreased. Two possible explanations for this are that this is a megacolony and/or the bait was highly attractive and recruited foragers from the control sites as well as many of the plants originally serving as a food source.


There was little to no difference in the consumption rate of bait between each site treated with indoxacarb, i.e., most of the bait in each tube (1.0 ounce) was consumed each week. This led us to the belief that this was a megacolony or one that contained many satellite colonies as opposed to individual colonies. The other indicator is the number of foragers at each site varied little between locations as the result of recruitment.

After four weeks of baiting and our seeming inability to reduce the number of foragers at the baited sites the indoxacarb-based gel bait was replaced with ant gel bait containing 0.001% fipronil. The trial continued with the fipronil bait for an additional four weeks, with all methodology the same as discussed previously for the indoxacarb bait. The fipronil bait had good initial acceptance, but dropped off dramatically as baiting continued.

The number of foragers at the fipronil treatment sites decreased within the first week of changing the bait and remained markedly lower for the next three weeks. The control sites remained relatively unchanged during the entire period the fipronil bait was being used. It is difficult to assess the effectiveness of either bait, however it appears that the fipronil bait failed to recruit large numbers of foragers and at the end of four weeks was not being consumed in appreciable quantities and in amounts sufficient enough to suppress colony numbers.

After eight weeks of baiting the numbers of foraging ants are increasing in both the treated and control sites. Despite the fact that the humidity and temperature in the indoor flight cage remain relatively constant year round we suspect that bait matrix differences may play a role in foraging behavior, e.g., the carbohydrate- based indoxacarb bait would be more attractive in the fall and winter and the protein-based fipronil bait more attractive in the spring. The constant climatic conditions and bait preferences exhibited might imply a genetic based behavior.

If populations continue to rise with the current baiting regime, alternative methods will be addressed. Exhibit animals are housed inside the exhibits in both the Small Mammal and Bird Houses the entire year. For Small Mammals, the exhibits are routinely emptied approximately once a year for maintenance for two to three weeks. This could provide an opportunity to treat the ants via alternative methods, such as steam. The indoxacarb bait caused recruitment to the bait tubes, indicating that the use of a carbohydrate-based attractant this time of year could assist with direct application of a pyrethrin-based or other low toxicity insecticide to the underside and/or interior of rocks, logs and other features. Safety for the Small Mammals, however, would need to be ensured.

For the Bird House, zoo horticultural staff will be treating plants for control of the scale insects, which will eliminate a major food source. In contrast to the Small Mammal exhibit, birds in the exhibit are not removed for routine maintenance. Alternative controls will need to be devised which will include exclusionary measures, such as PVC bait tubes to prohibit birds from accessing treatment areas.

Asian Needle Ant. In the fall of 2008, animal keeper staff at the National Zoological Park (NZP), Washington, D.C., began complaining of stinging ants within the zoo’s Forest Carnivores section. The ants were living in rotten logs near the leopard cat’s exhibit and were foraging in the cat’s food. Keepers were being stung after handling the food bowls.

Michael Waldvogel, North Carolina State University, who found Asian needle ants Pachycondyla chinensis (family Formicidae, subfamily Ponerinae) in a zoo and state parks in North Carolina, determined the ant in the National Zoo was most likely the Asian needle ant. After confirming his identification, the zoo’s Integrated Pest Management (IPM) staff deemed the presence of a “tropical” ant in Washington, D.C., was unimportant as they were not expected to survive the winter.

The Asian needle ant is an invasive species believed to have been introduced into the United States from Asia during the 1930s and has since been found sporadically along the East Coast. Colonies have become well established in South Carolina and other southeastern states including Virginia, Georgia and North Carolina. Asian needle ants are relatively unknown in other areas of the U.S. although they have been reported in a few northern states, including New Jersey, New York and Connecticut. The presence of Asian needle ants had not been previously verified within the District of Columbia.

Asian needle ant colonies can contain multiple queens and are considered surface nesters, i.e., their nests usually are found underneath rocks, logs, mulch or leaf litter. While they are considered to be an ecologically dominant species which can displace native ants, needle ant behavior has not been thoroughly studied. It is known, however, they are scavengers that don’t form typical ant foraging trails.

Asian needle ants possess stingers and injected venom can cause allergic reactions, some severe. A Clemson University study (Nelder et al., 2006) reported the majority of people stung by Asian needle ants developed slight reactions (80 percent) while a small percentage (8 percent) displayed large, localized reactions. Three cases in North and South Carolina resulted in anaphylactic shock, and one case in Tennessee required hospitalization for emergency treatment. Clemson reviewed both the ant’s medical importance and biology in PCT in 2007 (Paysen et al).

At the National Zoo, in July 2010, animal keeper staff once again complained about painful stings from ants. This time reports were made from staff assigned to the Andean bear’s exhibit, which is located more than 400 feet from the original site. Most stings resulted in red welts the size of a mosquito bite. Sting-sites reportedly remained painful for a few days and then became ‘itchy’ for one to two weeks. While most reactions were mild, several keepers developed more severe reactions with swelling around the sting-site ranging in size from a golf ball to an orange.

Asian needle ants have been reported affecting staff within other zoos, specifically the Greenville Zoo (South Carolina), the Knoxville Zoo (Tennessee), and the North Carolina Zoological Park (Waldvogel, 2009). In North Carolina, it was suspected needle ants were accidentally transported to the zoo via potting soil, plants and landscaping mulch. Within the National Zoo in 2010, Asian needle ants were found foraging within the Andean bear exhibit yards, underneath logs and rocks, and along the concrete walls of the moat that enclosed the bears. IPM staff observed needle ants within the bear yards, particularly in shady areas and along keeper walkways. Thousands of individual ants were observed coming from multiple sources, with many ants originating from expansion joints and cracks located within the concrete moat.

Unfortunately, with so much of the exhibit made of concrete, only a few nests were located. The ants did not appear to have any discernible foraging trails but tended to move along cracks and the expansion joints to eventually aggregate at food items. Densities of ants ranged from just a few individuals per square foot up to several hundred around food.

IPM staff completed a zoo-wide survey, which included visual inspection of areas around the zoo considered to be favored habitat and travel ways and found that they were widely distributed throughout the zoo. Needle ant activity was most often noted along cracks in pavement and artificial rockwork, under leaf litter, rocks, logs, wood mulch, and inside rotten wood which might contain termites. Staff observed needle ants within termite tunnels throughout decayed logs and even found small colonies (and adults with pupae) within rodent burrows located approximately one foot below ground.

Throughout the National Zoo, a total of 66 Asian needle ant nests were located; nests were defined only as those containing brood since identifying or locating queens within any of the nests was not possible. The majority of nests were found underneath and in old logs.

Unlike their fire ant relatives (Solenopsis invicta), Asian needle ants did not appear to be aggressive, although they stung when picked up or pinched beneath clothing. Asian needle ants did not show aggression towards pavement ants (Tetramorium caespitum), but did attack termites (Reticulitermes spp.) and carpenter ants (Camponotus spp.).

The National Zoo determined treatment was necessary due to potential human safety concerns because no problems were previously reported with animals. Zoo staff consulted with Clemson University entomologists Patricia Zungoli and Eric Bensen, who had been researching control methods since the ant had been identified within their state.

Staff learned Clemson’s ongoing research showed baiting for Asian needle ants was largely ineffective with the baits tested. Instead of baiting, their recommendations for control included habitat modification (removal of nests) and methodical direct insecticidal spray applications. A number of pesticides were recommended, but the zoo chose products which contained active ingredients currently approved for use in the zoo. The zoo has an extensive chemical approval system where all chemicals, including pesticides, must be approved by a number of staff, including the pathologist, veterinarian, safety officer and animal curatorial staff. Staff reviews each pesticide for potential concerns, such as past history of use on or around animals, and known sensitivities to certain exotic species.

IPM staff evaluated two ant baits, a boric acid-based granular carbohydrate bait and a hydramethylnon-based granular protein bait. The ants appeared to accept the protein-based bait faster than the carbohydrate bait, but both baits were completely consumed after three days. However, there was no perceptible decrease in the population of ants for several weeks following treatment and ultimately, baiting was discontinued.

Inside animal exhibits, a 0.10 percent pyrethrin microemulsion water-based spray was used to treat exposed Asian needle ant nests and foraging workers. An aerosolized pyrethrin (0.50 percent) was applied inside hollow logs and other void spaces where nests could not be treated directly. A water-based pyrethroid (0.03 percent lambda-cyhalothrin) formulation was applied as a crack and crevice spray both inside and outside animal exhibits for residual control along suspected ant trails.

The Andean bear yards which had the largest population of ants and the greatest keeper concern for their personal safety were the only areas treated. No treatments were made inside other exhibits because the ant populations were much lower and only a small number of foraging ants were noted. Without an established method for control of foraging ants, the zoo chose not to pursue treatment.

Control throughout the National Zoo was most effective when Asian needle ant nests and travel routes could be treated. All ant activity ceased after one week when nests, or nests confined to specific areas (e.g., within logs) and foragers were sprayed and all cracks and crevices were treated with the residual insecticide. Retreatment was necessary the following year, but the size of the population was reduced by approximately 90 percent. For areas treated only with a residual crack and crevice spray, ant activity was greatly diminished the first year, but rebounded the following year with only a 40 percent population reduction.

The following are factors to consider when working in zoos and other accounts that house non-target animals or import animals and plants that may contain non-indigenous arthropods:

  • Inspect all incoming animals and materials for non-indigenous species.
  • Have a facility or area where infested materials can be quarantined until the infestation can be addressed.
  • In zoos, have protocols and products preapproved by veterinarians, keepers and entomological staff to address common arthropod pests that maybe introduced.
  • Be aware of sensitive areas, e.g., insect rearing facilities and exhibits that could be impacted by product applications either directly or by translocation.
  • Work with staff to minimize pest access to alternative and often preferred food sources. Keep in mind that bait acceptance may be dictated by the foods that attracted the pest to the area.
  • Often treated zones, e.g., the interiors of artificial trees and logs, the underside of stones and logs, can be utilized in exhibits. These areas often serve as harborages for pests but are inaccessible to animals.
  • Consider the potential for secondary poisoning particularly with omnivores and insectivores.

Ongoing research at universities, such as Clemson and North Carolina State, hopefully will identify additional tools (e.g., baits, biological controls) which can be adapted for use not only within the National Zoo, but other zoos and similar sensitive accounts.

Acknowledgements: The authors would like to thank Dr. Richard Kramer, Dr. Michael Waldvogel, Eleanor Rice, Dr. Eric Bensen, Dr. Patricia Zungoli and Dr. Karen Vail for their support and consultation.

Susan Alberts is an entomologist at the National Zoo. Gregory Ose is an assistant entomologist at the National Zoo. Anthony Hiza is an IPM technician at the National Zoo. Richard D. Kramer is a consulting entomologist with Innovative Pest Management.


For additional photos and charts regarding this research, visit www.pctonline.com and click “online extras.”