Christmas Island Yellow Crazy Ant Control Program

Large scale changes occur to the forest structure following the removal of red crabs. In healthy forest thatsupports a large number of red crabs the forest floor is cleaned of leaves which the crabs eat, and the crabsalso suppress seedling growth and till the soil. In a yellow crazy ant super‐colony there is a large amount ofleaf litter present which has implications for micro climates and nutrient cycling, and there are largenumbers of seedlings that eventually form an understory thickets. Only time will tell how these forestpatches will change over decades compared to forest that has never been invaded by ants.While the yellow crazy ants themselves donot affect tree health, the huge numbers ofyellow lac scale they farm can cause die‐back and even the death of large foresttrees. The stress of hundreds of thousandsof scale insects can stress the tree to thepoint where it defoliates. Over multipleseasons this ongoing and relentless stresscan kill trees. Excess honeydew that fallsfrom the trees onto the leaves of lowerbranches and trees promotes the growth ofsooty mould that suppressesphotosynthesis and can also eventually killtrees.Images: Peter Green & Dennis O’Dowd.The removal of red crabs from large areas of the forest has alsoenabled the persistence of a species that was controlledeffectively by red crabs prior to their removal from an area byyellow crazy ants. The introduced giant African land snail is nowcommon in areas of the island where red crabs have beenremoved.Giant African land snails are renowned as a pestspecies through the tropics. The implications forthe presence of the giant African land snail onnative snail assemblages in the forest are stillbeing examined.5

Island‐wide survey for yellow crazy ants and red crabsIn 2001, researchers from Monash and La Trobe universities designed an island‐wide survey for ParksAustralia to assess the distribution and density of yellow crazy ants and red crabs at 1000 sites acrossChristmas Island’s 135 km2. This survey has taken place every two years since and is the primary source ofdata on the spread of yellow crazy ants and the population of red crabs. The data enables a targeted controlprogram to be implemented and is also used to gauge the success or otherwise of control activities.Since its inception, the island‐wide survey hasevolved into a multi‐species distributionsurvey to assess the status of a variety ofintroduced and native flora and fauna thatwould not normally be monitored sothoroughly. More than 40 species now havetheir distributions recorded with minimaladditional effort. Surveyors are trained todetect the presence of many native andintroduced flora and fauna species, both atsurvey sites and while in transit.With its 1000 sites the IWS is one of, if not the most, intensive islandecological surveys on earth.The survey is based on a transect methodology where yellow crazy ant activity is measured every five metersalong a 50 metre transect and red crab burrows are counted and measured 1 metre each side of the sametransect tape (100m2). Past experience shows there is a critical abundance of ants where they begin to killred crabs. If the sum of ant counts from along the transect is greater than 37, the site is considered a supercolony, subject to further verification).Island‐wide surveys were completed in 2001, 2003, 2005, 2007, 2009, 2011 and 2013. Each survey has beenundertaken in slightly different ways and has developed in line with advances in technology, especially fornavigation.6

Options for yellow crazy ant controlChemical baitingSince 2001, the Christmas Island National Park incollaboration with La Trobe University and an advisorypanel of experts (the Crazy Ant Scientific Advisory Panel– CASAP) has implemented an annual yellow crazy antcontrol program to minimise the ecological impacts ofthe yellow crazy ant, especially on red crabs. Thisprogram uses insecticides to wipe out super‐coloniesonce they have formed.To date, Fipronil‐based ant bait remains the only provenoption for the control of yellow crazy ants on ChristmasIsland. Image: Director of National Parks.Ant control has been maintained annually since theprogram’s inception using a highly effective ant bait.The bait is called AntOff , and the attractant is a proteinbased granular product. Fipronil is the activeingredient.Research into additional control options has always been a large component of the control program, and avariety of other chemical ant baits have been trialled over the years. These have included other toxins suchas hydromethylnon and indoxacarb, but also insect growth regulators such as s‐methoprene andpyriproxyfen which are considered less harmful (i.e. have less potential for off‐target impacts). The baitmatrix used has also varied during the program with combinations of baits and matrixes tested periodically.Bait delivery methods and their effectivenessFipronil is a general invertebrate neurotoxin, meaning that it targets the nervous systems of any animals thatdo not have backbones. Not only does AntOff kill yellow crazy ants, but it will also kill a variety of otheranimals if they consume it. These effects are called non‐target impacts, and the first principal of chemicalcontrol in ecologically sensitive areas is to use the bait in such a way that non‐target impacts are kept to anabsolute minimum. Crazy ants are baited only when their densities are high enough that they willmonopolize most of the bait, denying it to other organisms and reducing non‐target impacts. Further, baitingis conducted during the dry times of the year when any red crabs living on the margins of supercolonies stayin their burrows to avoid the low humidity – not even the bait will entice them out of their burrows. Robbercrabs are more active than reds during these periods, and they are attracted to the bait. Before any baitingoccurs, Parks Australia rangers lure robber crabs away from the target or remove them by hand to anotherlocation, a process that makes an already labour‐intensive undertaking even more so. Thousands of crabshave been saved in this way.On Christmas Island the bait is applied in two ways, either through broadcast by hand or aerially using ahelicopter. For the most part, broadcast by helicopter is preferred as the machine can access parts of theisland that cannot be baited by hand, it is safer for staff and completed quickly. Although it’s a veryexpensive exercise, it’s more cost effective per hectare than ground baiting. A helicopter has been brought7

to the island three times since the program began. At other times national park staff hand‐bait supercolonies that are accessible on foot. These operations can also only be undertaken during the dry season.The supercolonies are mapped using GPS prior to bating operations, and regardless of delivery method, thebait is broadcast at a set rate per hectare.Whether by hand or by helicopter, the use of ant bait to control super‐colonies within the thick island forest is expensiveand a logistically difficult operation Images: Director of National Parks.The chemical baiting of yellow crazy ants has occurred over 5,500 ha of theChristmas Island land mass (13,500 ha) since 2000 with many areas treatedmultiple times as super‐colonies re‐invade.Image: Director of National Parks.Fipronil‐based ant bait is extremelyeffective at knocking out super‐colonies;ant densities are noticeably reducedwithin the first week, and the majority ofants are eliminated within a month oftreatment. In some cases 100 percent ofants are wiped out in what was previouslya super‐colony area. Unfortunatelythough, baiting will never completelyeliminate yellow crazy ants fromChristmas Island. The ants occur at verylow densities in many locations andcannot be baited safely, and thesepopulations, plus the occasional survivorsof baiting operations, serve as founderpopulations for new super‐colonies.8

Sustained baiting operations over more than a decade have slowed the decline of the red crab population,but not completely stopped it. The reason for this is straightforward – the ants have been forming andreforming super‐colonies and killing crabs at a greater rate than natural recruitment processes can replacethem. Although red crabs migrate to the sea every year, the return of bay crabs is much less frequent, theytoo can be wiped out by super‐colonies, and the survivors grow very slowly. At the heart of the problem isthat super‐colonies are difficult to find, and can only be baited safely once the ants have reached very highdensities – and already killed the resident red crab population. Baiting is not sustainable in the long term – itis reactive, requires constant vigilance, and significant financial resources to maintain.How can we control yellow crazy ants into the future? Is direct biological control an option?Research into alternative control options has always been an important part of the yellow crazy ant programon Christmas Island. Even before the difficulties of a program based on toxic bait were fully apparent,university researchers were considering if biological control could be used contain the yellow crazy ant.Classical biological control works on the principle that in their area of origin, native species are kept in checkby their natural enemies, be they predators, parasites or pathogens. There is a lot of scientific evidencedemonstrating that in many cases, species introduced outside of their native ranges become invasivebecause they have effectively left these enemies behind. The principle of classical biological control then isto re‐establish population control over invasive species by first identifying and then importing a naturalenemy – a biological control agent – from within the native range of the target organism. Unfortunately,ants are an especially difficult target group for biological control, despite their global ecological andeconomic impacts. A program for the biological control of the Red Imported Fire Ant (Solenopsis invicta)using a parasitic fly and a protozoan disease as agents is currently under development in the southeasternUnited States, but no species of ants have yet been controlled in the field using biological control agents.Classical biological control:In an unbalanced ecosystem, introduced plants and animalscan cause dramatic problems for native species. Biologicalcontrol aims to introduce organisms that have the potentialto bring the ecosystem into a balanced state where speciescoexist without significant detrimental impacts. Thediagram above demonstrates classical biological control.The fluctuating lines represent seasonality. Adapted iculture/WeedBiocontrol.aspx.9

Indirect biological control as an optionEarly on in the history of the yellow crazy ant control project, researchers noted that super‐colonies ofyellow crazy ants always occurred in association with large numbers of honeydew‐producing scale insects.This association suggested that a plentiful supply of the carbohydrates – sugars – in the form of honeydewwas critical to the formation of supercolonies, and that maybe a biological control agent could be used notagainst the yellow crazy ants themselves, but against scale insects, as a way of reducing their food supply.There has been long history of successfully using biological control agents against a variety of scale insects inagricultural situations, and in most cases the agents have been tiny wasps that parasitize the female scaleinsects and kill them.In 2009 a team of researchers from La Trobe University in Melbourne, funded by Parks Australia, began aresearch program to investigate the idea of indirect biological control on Christmas Island. The programquickly focussed on two key areas – investigating the importance of honeydew in yellow crazy ant biologyand supercolony formation, and identifying and investigating the biology and natural enemies of the mainscale insect species in supercolonies.Not a cane toadCane toads were released as a biological control agent in 1935 to control cane beetles. Infamously, the cane toadignored the cane beetle and ate or killed a huge number of native species. The damage caused by cane toads continuesto this day, especially in the Kimberley and Arnhem Land regions of northern Australia.However, the circumstances under which the cane toad was released in Australia are completely different to the highlyregulated environment in which we operate today. The cane toad was introduced without scientific consideration of itsbiology in relation to its cane‐beetle target, without peer‐reviewed research, and in the absence of a strict approvalprocess to ensure it would be an effective biological control agent that would not affect non‐target species. Today theprocess of nominating a target species and then importing a biological control agent is highly regulated, rigorous andsubject to multiple approval processes based on peer reviewed science and governmental policy.The cane toad is a spectacular example of a biological control program gone very badly wrong, but the regulatoryenvironment of modern times makes such a spectacular failure well and truly and thing of the past. In fact, there arenumerous examples of successful biological control agents being used in Australia and around the world. The mostnotable of these success stories was the introduction of the Cactoblastis moth in the 1920s to control prickly pear,which at the time was smothering more than 24 million hectares of north‐eastern Australia! The moth effectivelycontrolled the prickly pear in just a decade, removing it from large tracts of arable farmland.10

Research into biological control – Yellow crazy ant dependenceon honeydewSeveral years of study in the laboratory and the field supports the idea that sugars in honeydew from scaleinsects are critical to supercolony formation by yellow crazy ants, and that reducing the supply of thisresource could indirectly control the ants. The researchers found three lines of evidence.Examining honey dew use by yellow crazy ants using stable isotope analysisThe concept of the food chain has proven very useful in this part of the research. At the base of the foodchain are plants – they produce food for other organisms, without consuming other organisms themselves.Herbivores feed on plants, and predators feed on herbivores. Plants, herbivores and predators are known astrophic levels. Some animals, such as the yellow crazy ant, feed on more than one trophic level. Honeydewis a plant product, so ants act as herbivores when they feed on it. However yellow crazy ants also act aspredators when they kill and consume other animals, mainly other invertebrates. The point of interest is thebalance between those two dietary components – if honeydew is very important in supercolony formation,it’s reasonable to expect a high percentage of honeydew in ant diets, and therefore, that ants should behavemostly as herbivores when they occur at extremely high densities. Alternatively, when the ants occur atmuch lower densities, their diet will contain a lower of percentage of honeydew, and their foraging patternswill be less herbivore‐like and more predator‐like.Ever tried watching and measuring what an ant feeds on? Its next to impossible to observe exactly whatindividual ants eat from day to day, so researchers used an indirect method, known as isotope analysis, todetermine where yellow crazy ants sit on the trophic ladder, and how that position shifts depending on theirabundance. The researchers collected samples of ants in several supercolonies over a period of months, asthe supercolonies boomed and then busted naturally. The isotope analyses showed a subtle but very clearpattern – yellow crazy ants always have a high percentage of plant products in their diet regardless of theirdensity, but as predicted, the trophic position of the ants is more herbivore‐like when their supercolonieswhere booming, and significantly less herbivore‐like as the supercolonies declined. These patterns areconsistent with the idea that ants rely heavily on honeydew to sustain their high densities in supercolonies.Crazy ants from within a forming colonyconsumed more protein based foods thanants in a super‐colony whose proportion ofcarbohydrate based food, supplied from thehoney dew, was much larger Image: La TrobeUniversity, unpublished data.11

Proof of concept field experimentIn a large field experiment, the researchers attempted to mimic the effect of a successful introduction of abiological agent on the honeydew supply to ants. Most of the scale insects in super‐colonies live in thecanopies of the largest trees, and so removing the scales manually or with insecticides was impossible.Instead, the researchers prevented the ants from gaining access to the scale insects by banding all the treeson an experimental plot. From the ant’s point of view, this achieved the same effect as if a biological controlagent had killed the scale insects – a drastic reduction in their honeydew supply. Another plot was set up asan unbanded control next to the banded plot.The density of yellow crazy ants was at supercolony levels on the banded and unbanded plots for severalweeks before the experiment commenced. It remained high on the unbanded plot throughout theexperiment, but declined dramatically on the banded plot once the bands were put on the trees; density fellby around 80% within a matter of weeks. The rainforest canopy harbours other resources for ants besidesscale insects and their honeydew, and ants were blocked from these resources too. However, theresearchers were confident the decline in ant density on the banded plot was because the ants no longerhad access to all that honeydew, because only a tiny fraction of those ants that got caught in the log jamabove the plastic bands were carrying prey items in the mouthparts, but the abdomens of a very highpercentage of ants were obviously swollen with translucent honeydew.Plastic film sprayed with furniture polishwas wrapped around trees is an effectivebarrier to ant traffic on tree trunks. Alltree trunks and stems were banded andthis even extended to large multiplestemmed fig trees. Yellow crazy antheading up the trunk could not get pastthe bands, and coming down trunksformed logjams. These ants were gentlybrushed off the trunks each day until thecanopy was effectively empty of ants.Image: Sara Wittman.12

The effect of carbohydrate supply on yellow crazy ant growth and behaviourThe stable isotope study and field experiment both show that carbohydrates in honeydew are important tobuild and sustain high densities of ants in supercolonies. But what doe

Background The unique fauna and the ecological role of red crabs on Christmas Island The terrestrial landmass of Christmas Island is the top 361m of a 5km high seamount. It is a very remote island located 350km south of the island of Java, Indonesia and 2,600km north‐west of Perth.