Journal of the American Mosquito Control Association, 2O(I):91-93,200'4 Copyright @ 2OO4 by the American Mosquito Conffol Association, Inc.

SCIENTIFIC NOTE THE ENDANGERED ILLIDGE'S ANT BLUE BUTTERFLY (ACRODIPSASILLIDGEO FROM AN INTERTIDAL HABITAT MANAGED FOR MOSQUITO CONTROL

M. J. BREITFUSST ,qNo P E. R. DALE'

ABSTRACT. Acrodipsas illidgei is an endangered butterfly inhabiting mangrove forests in southeastern Queensland, Australia. Concern over the effects of mosquito control activities prompted a broad-scale survey for the species at Coomera Island, in southeastern Queensland. The butterfly was recorded on the edge of an old-growth mangrove forest in close proximity to mosquito control runnels. Other forms of mosquito control at Coomera Island are unlikely to impact on the species because of the mode of action of larvicides used and the fact larvae occur within ant colonies formed in hollow stems and branches of mangrove trees. Further studies are required to more fully understand the relationships between mosquito control activities and the population dynamics of endangered species such as A. illidgei.

KEY WORDS Butterfly, Queensland, mangrove, mosquito control

Environmental concern regarding the effects of and have rarely been observed in the wild because some mosquito control techniques on threatened of naturally low population numbers, the small size species has prompted debate at local, state, of adults, and flight activity (Braby 2000). national, and international forums. Of particular The aim of this paper is to discuss the habitat of concern in Australia has been the unsubstantiated A. illidgei at Coomera Island, an important nontarget threat to endangered species such as mosquito-breeding site that has experienced Illidge's ant blue butterfly (Acrodipsas illidgei mosquito control operations to reduce pest numbers (Waterhouse and Lyell)) from mosquito control for the past l8 years. activities in mangrove and adjacent salt-marsh Coomera Island (27.85'5. 153.38'E) is a 268-ha habitats (Beale 1995, Sands and New 2002). area located in the Southern Moreton Bay Marine Acrodipsas illidgei is listed as endangered under Park and was declared a Conservation Park in 1994. Queensland legislation (QPWS 1994) and is Coomera Island supports extensive areas of recognized internationally as an endangered species mangrove and salt marsh and is flooded by the by the International Union for the Conservation of highest spring high tides, exceeding 2.45 m. Nature (ruCN 2000). Additionally, activities that Dominant mangrove species include grey mangrove may affect nationally recognized endangered species and river mangrove (Aegiceras corniculatum (L.) invoke the application of the Commonwealth Blanco.) with salt-marsh species including marine Environmental Protection and Biodiversity couch (Sporobolus virginicus (L.) Kunth.) and Conservation Act (1999). Recently, Sands and New beaded glasswort (Sarcocornia quinqueflora (Bunge (2002) reviewed the IUCN listings and suggested the ex Ung.-Sternb.) A. J. Scott). Mangrove tends to formulation of a recovery plan to delist the species form a dominant community on the lowest regions to a level commensurate with the habitat quality of of the shore and is flooded regularly, whereas salt its known populations and levels of protection marsh occurs higher on the shore at sites less afforded to those habitats. one of which is Coomera frequently inundated. Island in southeastem Queensland, Australia. The island has experienced a range of human Larvae of A. illidgei live within ant colonies influences, including cattle grazing, dredging formed in hollow stems and branches of vegetation, outfall, and army training, and is currently managed including the grey mangrove (Avicennia marina for mosquito control to reduce breeding of the (Forssk.) Vierh.) (Braby 2000). Larvae are human viral-transmitting species Ochlerotatus myrmecophilous and feed on juvenile ants until vigilax (Skuse). One method of mosquito control eclosion, when adults emerge and seek partners. employed on the island to target mosquito larvae is Adults are thought to be active for only a few days runnelling. Runnelling involves the construction of shallow (<30-cm-deep) channels that allow regular I tidal access to isolated mosquito-breeding pools in University of Queensland, Australian Centre for the salt marsh. The mosquito larvae are flushed International Tropical Health and Nutrition, Queensland Institute of Medical Research, The Bancroft Centre, 300 from the salt marsh by tides and exposed to higher Herston Road, Herston, Queensland 4029, Australia. levels of predator pressure by fish and other 2 Australian School of Environmental Studies, Griffith organisms (Hulsman et al. 1989). Runnels are University, Brisbane, Queensland 4l I I, Australia. known to have some impacts on the surrounding

9l JounNnr-on rua Avpnrcax Moseurro Corvrnot AssocrerroN Vor-. 20, No. ecosystem, including aiding transport of mangrove that offer the highest standards of control and seeds to areas where they would not normally be environmental consideration. deposited (Breitfuss et al.2OO3) and influencing the The use of physical methods of control, such as distribution of some nontarget invertebrates runnelling, would be highly unlikely to result in (Chapman et al. 1998, Breitfuss 2001). In 1985, a degradation of identified butterfly breeding habitats. small area of approximately 0.5 ha was runnelled Runnelling can promote extension of mangroves into for mosquito control at a salt-marsh site identified areas of salt marsh where they previously have been as the major breeding area for Oc. vigilax on the restricted and this is the case at the runnelled area island (Hulsman et al. 1989). Although the area on Coomera Island (Breitfuss et al. 2003). However, modified was small, it was close to mangrove natural disasters and threatening processes may habitat relevant to the butterfly-a large ultimately compromise butterfly breeding habitats (approximately 5-ha) stand of mature A. marina. and recovery programs (Sands and New 2002). The remaining salt marsh not runnelled has Yanno (1996) referred to a number ofthreatening historically been treated with ground and aerial processes affecting the Schaus swallowtail butterfly applications of the larvicides Bacillus thuringiensis (Papilio aristodemus ponceanus) in southern var israelensis (Bti; Vectobac@), (S)-methoprene Florida tropical hardwood forests. In an effort to (Altosand@), and Temephos (Abate@). More preserve remaining breeding sites, the Monroe recently, use of Temephos has been discontinued. County Mosquito Control District discontinued use To address issues relating to potential impacts of aerial and ground application of mosquito from runnelling and larvicides on the endangered insecticides (not stated whether they are larvicides A. illidgei, a broad-scale survey of mangrove sites or adulticides) and a concerted effort to release was conducted. Initially, mangrove sites were captive-bred pupae and adults was initiated. categorized in terms of age and size, with the However, high predation on the I st installment of largest old-growth sites being separated from pupae by a flock of neotropical migratory birds smaller, younger habitats (age was determined by (Yanno 1996) highlights the fact that natural plant height and trunk diameter characteristics). phenomena also can constrain recovery efforts. Sampling, with a standard butterfly net, was In the case of A. illidgei, the current use of opportunistic and was conducted between 0900 and aerially applied larvicides (such as microbials and 1500 h on clear, sunny, and generally windless larval growth regulators [LGRs]) are unlikely to days. impact the breeding habitat of the butterfly. First, On September 9, 1999, a single female A. iUidgei because the butterfly larvae occur in ant colonies was collected from the exposed edge of an old-growth formed in hollow stems and branches, they would habitat directly adjoining and 400 m from the rarely be exposed to larvicides; and, 2nd, runnelled salt marsh (the specimen is deposited at the microbials such as Bti and LGRs such as (S)- Queensland Museum). Subsequent surveys revealed methoprene have a low toxicity to nontarget no further adults; however, 1 month after the initial organisms when applied at label rates (Brown et al. discovery of the adult female, 2 larvae of the butterfly 1999, Lawler et al. 1999, Boisvert and Boisvert were found in the nest of Crematogaster Lund 2000, Brown et al. 2000). (laeviceps group) ants (these ants tend the butterfly Although environmental concems associated with larvae) from the same mangrove tree from which the ungovemed use of broad-scale pesticides are valid, no adult female was collected. Ant surveys were information has been published to relate current conducted opportunistically because the species is mosquito control operations with deleterious impacts conspicuous on the tnrnk of inhabited mangroves. No on threatened butterflies in Australia. This preliminary A. illidgei were observed in any of the younger study provides evidence for broader investigations mangrove areas surveyed. into the relationships between operational and A number of other butterflies also were recorded ecological aspects of mosquito control and their from the mangrove edge, including the saltpan blue impacts on significant nontarget species. Specifically, (Theclinesthes sulpitius (Miskin)), Miskin's blue quantifled collections of adult and immature A. (Theclinesthes miskini (T. P Lucas)), the common illidgei (and its ant host) will be necessary to more grass-blue (Zizina labradus (Godart)), and the fully understand the threatening processes impacting copper jewel (Hypochrysops apelles apelles this species on Coomera Island, and if any mosquito (Fabricus)). The lst 3 species breed in salt marsh control activity is associated with changes in the but utilize mangroves for nutrition and mating sites, population dynamics of the species. whereas the latter species is associated with mangroves for breeding. REFERENCES CITED Modern mosquito control programs aim to deliver Beale JP. 1995. Report on the environmental impact of effective and relatively low-impact operational and biting midge control measures on lllidge's ant-blue technical options. In Australia, organizations such as bunerfly-part A Prepared for The Hervey Bay City the Mosquito and Arbovirus Research Committee Council. p 82. Available from QueenslandGovernment (Inc.) investigate current control options and provide Printers. unbiased recommendations for altemative approaches Boisvert M, Boisvert J. 2000. Effects of Bacillus ScreNrrrrcNores 93

thuringiensis var. israelensis on target and nontarget Chapman HE Dale PER, Kay BH. 1998. A method for organisms: a review of laboratory and field experiments. assessing the effects of runneling on salt marsh grapsid Biocontrol Sci Technol 10:5l7-561. crab populations. J Am Mosq Control Assoc l4:61-68. Braby MF 2UJJ. Buttetflies of Australia, their identiJication, Environmental Protection and Biodiversity Conservation biology and distribution Volume 2. Melboume, Australia: Act. 1999. Canberra, Australia: Australian Government. cstRo. Hulsman K, Dale PER, Kay BH. 1989. The runnelling Breitfuss MJ. 20O1. Predicting the effects of runnelling on method of habitat modification: an environment- non-target saltmarsh resources. Arbovirus Res Aust 8: focussed tool for salt marsh mosquito management. J 23-29. Am Mosq Control Assoc 5:226-234. for the Conservation of Breitfuss MJ, Connolly RM, Dale PER. 2003. Mangrove IUCN [International Union Naturel. 2OOO. Red list of threatened species Gland, distribution and mosquito control; transport of Switzerland: International Union for the Conservation Avicennia marina propagules by mosquito conffol of Nature. runnels in southeast Queensland saltmarshes. Estuarine Lawler SP, Jensen I Dritz DA, Wichtermhn G. 1999. Coastal Shelf Sci 56:573-579. Field efficacy and nontarget effects of the mosquito Brown MD, Thomas D, Mason P, Greenwood JG, Kay larvicides Temephos, methoprene, and Bacillus BH. 1999. Laboratory and field evaluation of four thuringiensis var. israelensis in Florida mangrove (Diptera insecticides for Aedes vigilax Culicidae) and swamps J Am Mosq Control Assoc 15:446-452. toxicity to the nontarget shrimp Leander tenuicomis QPWS [Queensland Parks and Wildlife Service]. 1994. (Decapoda: Palaemonidae). J Econ Entomol 92:7045- Queensland Nature Conservation (Wildlife ) Regulation 105l. (1994) Bisbane, Australia: Queensland Government Brown MD, Watson TM, Green S, Greenwood JG, Purdie Printers. D, Kay BH. 2000. Toxicity of insecticides for control Sands DPA, New TR 2OO2. The action planfor Australian of freshwater Culex annulirosrrls (Diptera: Culicidae) butterflies Canberra, Australia: Environment Australia. to the nontarget shrimp, Caradina indistincta Yanno M. 1996. New hope for the Shaus swallowtail. (Decapoda: Atyidae). J Econ Entomol 93:667-672. Endang Spec Bull 21(4):22-23.