122 Plant Protection Quarterly VoI.11(3) 1996 The trees were scattered throughout the Activity of a hydramethylnon-based bait against the 10 ha block, separated by at least 70 m to preclude possible interaction of ant citrus ant pests, Iridomyrmex rufoniger gpo spp. and populations and randomly assigned as I. purpureus baited (5) or unbaited (5). Pre-treatment assessments of foraging ant activity were cond ucted for each tree on 26 October by David G. James, Mark M. Stevens, Karen O'Malley and Renay Heffer, counting ants moving past a fixed paint (in Yanco Agricultural Institute, NSW Agriculture, Yanco, New South Wales 2703, both directions) on the trunk during a two Australia. minute period. On 28 October 10 g of Arinosu-Korori™ (0.88% hydramethyl­ Summary non) was placed on the ant trail at the base A commercial granular ant bait, Arinosu­ insecticides, which were formerly used for of each treatment tree and covered with a Korori™, containing hydramethylnon ant control, are no longer available. Cur­ black, upturned flower pot which still al­ was readily collected by the citrus ant rent control of ants based on ground lowed ant access to the bait. This was nec­ pests, Iridomyrmex rufoniger gpo spp. and sprays of the organophosphate insecti­ essary due to the photosensitivity of I. purpureus, and provided good control cide, chlorpyrifos, is usually short-li ved hydramethylnon (Vander Meer el af. 1982) in small scale field experiments on single and may need to be repeated a number of and anticipated slow removal of bait by trees and nests. Ten grams of bait at the times during a season. An effective and 1. nlfaniger gpo spp. Assessments of forag­ base of ci trus trees containing foraging 1. environmentally acceptable method of ant ing ant activity on baited and unbaited rufoniger gpo spp. significantly reduced control in citrus groves is required in or­ trees were made at three-five day inter­ the number of foragers after three days, der to protect current biological control vals post-treatment by recording ant and for up to 75 days. The same quantity systems and to improve the biological travel on trunks as described above. of bait killed all individuals in a 1 m 2 nest control of honeydew-producing insects. In the second experiment, the unbaited of I . purpureus which remained ant-free One approach to ant control in citrus trees from the first experiment were re­ for at least 91 days. The bait also signifi­ groves is the use of slow-acting toxicants assigned as baited trees and vice versa cantly reduced numbers of ants in a 2 m 1 in baits, which are collected by foraging once ant activity had returned to pre­ nest for 35 days, however, no significant workers and spread to other colony mem­ treatment levels. Ten grams of Arinosu­ reduction in ant numbers occurred in a 4 bers by trophallaxis, thereby eliminating Korori™ were applied on 25 January in m 2 nest. Prospects for using this bait in all or most of the colony. The combination the manner described above. Assessments controlling ant populations in citrus of slow-acting toxicants and hi ghly a ttrac­ of ant foraging activity were made at groves are discussed. tive bait bases can be a very effective seven to nine day intervals fo r the first five method of ant control (Lofgren el af. 1975). weeks then at 51 and 75 days. Introduction The slow-acting stomach poison, Ants in the genus lridomyrmex Mayr, pri­ hydramethylnon, in the commercial bait, I. purpureus marily two species in the 1. rufoniger group Amdro™, provides good control of A single experiment using paired nests of and 1. pllrpllrells (F. Smith), cause serious 50lenopsis invicta Buren (red imported fire three sizes (1,2 and 4 mZ) was conducted in problems for citrus producers in eastern ant) (Collins el al. 1992) and Plteido1e native pasture near the Murrumbidgee Australia by d isrupting biological control lIIegacep'w'a (F. ) (bigheaded ant) (Reimer River at Yanco in southern New South of soft scales, aphids and whitefJies. The and Beardsley 1990) in the United States. Wales. Nests were separated by at least association of ants with honeydew­ Another commercial bait containing 50 m and were not interactive. A bait treat­ prod ucing insects has been well docu­ hydramethylnon, Ma xforce™, shows ment and control were assigned ran­ mented (e.g. Bartlett 1961, Way 1963, good activity against Lillepithema Ill/mile domly to each nest size pair. Pre­ Buckley 1987). Ants foraging in citrus (Mayr) (formerly Iridamyrmcx IlIImilis treatment assessments of ant activity were canopies reduce populations of the natural (Mayr» (Argentine ant) (Knight and Rust made for each nest on 23 November 1994 enemies of honeydew-producers, thereby 1991) but is not attractive to I. nlfonigeT gpo by counting the number of ants attending limiting their impact. This leads to a build spp. or I. pllrpureJ/s (James unpublished a bait station at the end of four consecutive up of sugar-rich honeydew, which coats observation). A third commercial granu­ 15 minute periods. Bait stations (small foliage and fruit and provides a substrate lar bait containing hydramethylnon, sheets of white p la stic with central inden­ for the growth of sooty mould fungi, Arinosu-Korori™, is used for controlling tations filled with 25% sugar /water solu­ downgrading fruit intended for the fresh household 'nuisance ants' including I. ti on) were placed near nest entrances. fruit market. Armoured scales such as red glaber (Mayr) in Japan and is attractive to 1. pllTpllreus is strongly attracted to and scale, Aallidiella aural/tii (Maskell), do not 1. Tilfoniger gpo spp. and I. plIrpllrclls. This feeds avidly on 25% sugar/water baits produce honeydew, but are also protected study was conducted to determine (James, Stevens and O'Malley unpublished from natural enemies by a nts foraging in whether Arinosu-Korori™ has the poten­ observation). Ten gram baits were applied ci trus trees (Samways and Tate 1984). Re­ tial to provide sustained control of 1. immediately after pre-treatment counts cent studies in the Murrumbidgee Irriga­ rufonigeT gpo spp. and 1. purpllreus in citrus were made, to a single location near the tion Area of southern New South Wales groves. major entrance of each nest. Other studies indicate that the removal of ants fr om cit­ indicated l. pllrpllrellS removed bait very rus trees can lead to low, non-economic Materials and methods rapidly so no attempt was made to protect populations of honeydew-prod ucing in­ it from sunlight. Ant activity assessments sects and red scale (James, Stevens and J. ru[oniger gpo spp. were made for each nest at approximately O'Malley in preparation). Two experiments were conducted in a Va­ weekly or fortnightly intervals using the The general reduction in insecticide in­ lencia orange grove at Farm 1777, Cudgel, procedure described for the pre-treatment puts for key pests (e .g. red scale) in Aus­ 16 km east of Leeton in southern New assessment. tralian citriculture over the past few dec­ South Wales during October-April Daily maximum and minimum tem­ ades has probably contributed to increas­ 1994-95. Ten trees with high levels of for­ peratures were recorded at the Yanco ing ant populations in citrus groves. In ad­ aging ant activity arising from single trunk Agricultural Institute «10 km from ex­ dition, th e highly effective organochlorine trails were selected for the experiments. periment sites). Ant activity data were Plant Protection Quarterly VoI.11(3) 1996 123 transformed to the log (X + 1) scale and there was no significant difference in ant on the 1 m2 nest compared to the unbaited subjected to analysis of variance and least numbers between baited and unbaited nest for 91 days (P<0.05) (Figure 2). Virtu­ significant difference procedures. Indi­ trees. Mean daily maximum and minimum ally no ants were recorded on this nest vid ual bait station counts were treated as temperatures were 27 and 14°C during the from day 16-91 . The bait caused a signifi­ replicates in the I. pllrpureus experiment. first half of the experiment and 33 and cant reduction in ant numbers on the 2 m 2 17°C during the second half. nest for 35 days (P<0.05) (Figure 2), but the Results In Experiment 2 the number of ants for­ population recovered thereafter and was aging on baited trees was significantly comparable to the population on the I. rufoniger gpo spp. lower than on unbaited trees for up to 75 unbaited nest for the remainder of the ex­ In Experiment 1 the bait significantly re­ days (P<0.05), although they were not ab­ periment. The bait had no significa nt im­ duced numbers of foraging ants for up to sent on any assessment date (Figure 1). pact on ant numbers on the 4 m2 nest 54 days, compared to unbaited trees Differences were significant on all dates (P>0.05) (Figure 2). The experiment was (P<0.05) (Figure 1). Total removal of bait except pre-treatment. All bait was re­ terminated after 9] days due to low levels did not occu r for about four weeks but a moved after 72 h. Mean daily maximum of ant activity caused by cool tempera­ significant reduction in ant activity was re­ and minimum temperatures were 32 and tures. corded on days 3 and 6, but not day 11 . 17°C during the first half of the experi­ Foraging activity was low at this time due ment and 27 and 12°C during the second Discussion to cool temperatures (8-21 °C). Ant activ­ half. Cool temperatures «21°C) after day This preliminary study indicates that the ity on baited trees was significantly lower 75 (mid April) substantially reduced ant hydramethylnon-based bait, Arinosu­ (P<0.05) than on unbaited trees from day activity on all trees and assessments were Korori™, has potential for controlling I. 14- 54. No ants were present on three discontinued. rllfoniger gpo spp. and I. purpureus in citrus baited trees on days 21-42, while less than groves. Limited quantities of bait pre­ 10 were recorded on each date on the I. purpureus vented us from conducting larger scale other two. Re-colonization occurred more The bait was totally removed by ants after experiments. Consequently, the results rapidly on the latter trees and by day 67 1 h and significantly reduced ant activity were necessarily compromised by the use of small test sites (single trees / nests) that permitted immigration of ants from 500 Experiment 1 healthy colonies, masking the true long term impact o f the bait on ant activity. --- baited -0- control Nevertheless, Significant control was 400 achieved for up to 11 weeks for I. rufaniger gpo spp. and up to 13 weeks for I. purpureus, suggesting that this bait should 300 provide control for more th an three months when applied to an entire grove. c: The bait was highly attractive to both E 200 ant species with I. purpureus removing it N within 1 h. I. rufoniger gpo spp. removed all ~ bait within three days in the summer (i t ~ 100 may have been removed earlier than this ~c: but no observations were made) but it Q) took about four weeks in spring. Arinosu­ E 0 Q) Korori™ is formulated largely from silk­ >a PT 3 6 11 14 17 21 26 28 31 35 38 42 46 54 67 7682 worm pupae (M. Jackson, Cyanamid Aus­ E 400 tralia Pty. Ltd., personal communication) ~ Experiment 2 but also contains other undisclosed ingre­ c: dients. Foraging I. rufoniger gpo spp. show "'c: --- baited a seasonal alternation of preference be­ : 300 -0- control tween carbohydrate and protein food ::;; sources, although during most of summer and autu mn, protein is preferred. I. p"rpllrells feeds on protein and carbohy­ 200 drate-based foods with a preference for the latter (James, Stevens and O'Malley unpublished observation). In this study bait was provided at 'sta­ 100 ti ons' protected from sunlight and rain. Apperson et al. (1984) found Amdro™ (0.88% hydramethylnon) bait for S. illvicta was more effective when broadcast than o when used as nest treatments. The broad­ PT 7 14 21 29 37 51 75 cast of bait granules in a citrus grove con­ taining many, evenly distributed ant nests, Days post-treatment may be a more attractive and practical proposition for growers than provision of Figure 1. Trailing activity of 1. rulonige. gpo spp (±SE) on trunks of citrus trees bait stations. However, the rapid degrada­ baited with 10 g of Arinosu-Korori™ bait (0.88% a.i. hydramethylnon). tion of hydramethylnon in sunlight (up to Experiment 1 conducted from 26 October to 18 January and Experiment 2 50% in 12 h according to Vander Meerel al. from 25 January to 10 April 1994-95. (1982)), would require ants to collect and 124 Pla nt Protecti on Quarterly Vo1.11 (3) 1996

30 receive a sub-lethal a mount because of di­ lution d uring trophallaxis. Stud ies on for­ 1 m' nest agi ng behaviour of l. rufoll iger gpo spp. and 25 I. purpureus indica te that d iurnal foraging is greatest d uring spring and a utumn 20 Oames, Stevens and O'Malley unpublished observati on) suggesting that bait removal and e fficacy may also be maximal at this 15 time. Collins et al. (1992) showed spring applicatio ns of Amdro™ produced the 10 ----- baited most rapid e ffects on populations of S. --0- control iuvicta. 5 Hydrame thylnon-based baits have proven to be an effecti ve and en viron­ mentall y acceptable stra tegy fo r control­ o ling at least three species of ants: S. invicta PT 7 16 21 35 49 6 2 69 77 84 91 (Collins el af. 1992), P. megacephala (Reimer and Beardsley 1990) and L. IlIImile (Knight E 50 and Rust 1991). The Arinosu-Korori™ bait It) ~ 2 m' nest fo rmula tion containing hydrame thymon ~ a ppea rs to show the same potential for 40 Co controlling 1. rufo lliger gp o spp. a nd 1. c: o pu rpureus in citrus groves. However, :;; 30 larger scale experiments w hich also con­ tV sider factors such as applica tio n technique -.. and ti ming a re necessary to determine the .'.:: tV 20 full potential of this product in ci trus ant .0 management. Cl c: 10 Acknowledgrnen ts c: " ----- baited We th ank Mike Jackson of Cyanamid Aus­ :::: --0- control trali a Pty. Ltd. for supplying the Arinosu­ tV o Korori™ bait used in these ex periments !l c: PT 7 16 21 35 49 62 69 77 84 9 1 and Paul Morris for allowing us to conduct tV the m in his orchard. Th is research was c: supported by fund ing from the Australian tV 50 Horticultural Research and Development ~ 4 m' nest Corporation.

40 References Apperson, CS., Leidy, R. B. and Powell , E.E. (1984). Effects of Amdro on the red 30 imported fi re ant (Hymenoptera: Formicid ae) and some non-ta rget ant species a nd persistence of Amdro on a pasture habita t in North Carolina. Jour­ 20 nal a/Economic Entomology 77,1012-8. Bartl ett, B.R. (1961). The influence of ants ----- baited --0- control on paraSi tes, predators and scale insects. 10 Anllals of the En tom%gical Society of America 54, 543-51. PT 7 16 2 1 35 49 62 69 77 84 91 Buckley, R.C (1987). Interacti ons involv­ ing pla nts, Homoptera and ants. Anttllal Days post-treatment Review of Ecology and Systema tics 18, 11 1-35. Figure 2. Number of I. purpureus (±SE) attending 25% sugar/water baits o n 1, 2 Collins, H.L., Callcott, A., Lockley. T.C and 4 m' nests baited with 10 g Arinosu-Korori™ bait (0.88% a.i. and Ladner, A. (1992). Seasonal trends in hydramethylnon) from 23 November to 1 March 1994-95. effectiveness of hydramethylnon (AMDRO) and fenoxycarb (LOGIC) for transfer bai t to nests within the first day of 7-10 days (Rei mer and Beardsley 1990) control of red imported fire ants (Hy­ exposure as is the case for Amdro™ when and appears to have a similar effect on I. menoptera: Formicidae). Journal of Eco­ used against P. megaceplzala or S. iflvicta rufoll iger gpo spp. and I. pllrp"rells. Al­ nomic Elltom%gy 85, 2131-7. (Reimer and Beardsley 1990, Colli ns el al. though 10 g of bait (containing less than Knig ht, R.L. and Rust, MK (1991). Effi ­ 1992). Clea rl y, the relative efficacy of 0.1 g of hyd ramethylnon) kill ed a Im' nest cacy of fo rmula ted baits for control of Arin osu-Korori ™ against ci trus ants when of }. p lI rpllre lls and had a significant but argentine ant (Hymenoptera: applied as a broadcast treatment or in bait short-term impac t on a 2 m2 nest, it had no Fo rmicidae). !oumal of Economic Elltomol­ stations, needs to be dete rmined . e ffect on a 4 m2 nest. Clearl y, the larger an ogy 84, 51 0-4. Hydramethylnon is a slow-acting stom ­ ant colony is, the greater the chance that Lofgren, CS., Banks, W.A. and Glancey, ach poison w hic h ki lls P. megacephala in some ants w ill not receive the toxicant or B.M. (1975). Biology and control of Plant Protection Quarterly VoI.1I (3) 1996 125 imported fire ants. Amwai Review of En ­ tomology 20, 1-30. Reimer, N.j. and Beardsley jr., j.W. (1990). Effectiveness of hydramethy ln on and fenoxyca rb for control of bigheaded ant (Hymenoptera: Formicidae), an ant as­ socia ted w ith mealybug w ilt of pinea p­ ple in Hawaii. Journal of Economic Ento­ mology 83,74-80. Samways, M. j. and rate, B.A. (1984). Eva luation of several trun k barriers used to prevent the movement of the pugnacious ant (Anoplo/epis custodiens (Smith)) into citrus trees. Citru s and Sub­ tropical jrllil/ollmaI608, 9-12, 20, 23, 25-6. Vander MeeT, R.K., Williams, D .F. and Lofgren, C.S. (1982). Degradation of the toxicant AC-217, 300 in Amdro im­ ported fire ant bait under field condi­ tions. Journal of Agricultural and Food Chemistry 30, 1045-8. Way, M.j. (1963). Mutualism between ants and honeydew-producing Homoptera. Annual Review of Entomology 8,307-44.