New Weapons in the Toad Toolkit: a Review of Methods to Control and Mitigate the Biodiversity Impacts of Invasive Cane

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/316915049

New Weapons in the Toad Toolkit: A Review of Methods to Control and Mitigate the Biodiversity Impacts of Invasive Cane...

Article in The Quarterly Review of Biology · June 2017 DOI: 10.1086/692167

CITATIONS READS 2 207

13 authors, including:

Reid Tingley Robert J Capon University of Melbourne University of Queensland

47 PUBLICATIONS 1,042 CITATIONS 359 PUBLICATIONS 5,007 CITATIONS

SEE PROFILE SEE PROFILE

Richard Shine University of Sydney

1,087 PUBLICATIONS 40,853 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Accounting for imperfect detection in environmental DNA surveys View project

Novel P-glycoprotein substrates and inhibitors in application to Multidrug Resistance View project

All content following this page was uploaded by Simon Clulow on 14 May 2017.

The user has requested enhancement of the downloaded file. Volume 92, No. 2 June 2017 THE QUARTERLY REVIEW of Biology

NEW WEAPONS IN THE TOAD TOOLKIT: A REVIEW OF METHODS TO CONTROL AND MITIGATE THE BIODIVERSITY IMPACTS OF INVASIVE CANE TOADS (RHINELLA MARINA)

Reid Tingley School of BioSciences, University of Melbourne Melbourne, Victoria 3010 Australia e-mail: [email protected]

Georgia Ward-Fear School of Life and Environmental Sciences, University of Sydney Sydney, New South Wales 2006 Australia e-mail: [email protected]

Lin Schwarzkopf College of Science and Engineering, James Cook University Townsville, Queensland 4811 Australia e-mail: [email protected]

Matthew J. Greenlees School of Life and Environmental Sciences, University of Sydney Sydney, New South Wales 2006 Australia e-mail: [email protected]

123 124 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

Benjamin L. Phillips School of BioSciences, University of Melbourne Melbourne, Victoria 3010 Australia e-mail: [email protected]

Gregory Brown School of Life and Environmental Sciences, University of Sydney Sydney, New South Wales 2006 Australia e-mail: [email protected]

Simon Clulow School of Environmental and Life Sciences, University of Newcastle Callaghan, New South Wales 2308 Australia e-mail: [email protected]

Jonathan Webb School of Life Sciences, University of Technology Sydney Sydney, New South Wales 2007 Australia e-mail: [email protected]

Robert Capon Institute for Molecular Bioscience, University of Queensland St. Lucia, Queensland 4072 Australia e-mail: [email protected]

Andy Sheppard CSIRO Canberra, Australian Capital Territory 2601 Australia e-mail: [email protected]

Tanja Strive CSIRO Canberra, Australian Capital Territory 2601 Australia e-mail: [email protected]

Mark Tizard Australian Animal Health Laboratory, CSIRO Geelong, Victoria 3220 Australia e-mail: [email protected] June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 125

Richard Shine School of Life and Environmental Sciences, University of Sydney Sydney, New South Wales 2006 Australia e-mail: [email protected]

keywords Bufo marinus, conditioned taste aversion, containment, genome engineering, mitigation, suppression

abstract Our best hope of developing innovative methods to combat invasive species is likely to come from the study of high-profile invaders that have attracted intensive research not only into control, but also basic biology. Here we illustrate that point by reviewing current thinking about novel ways to control one of the world’s most well-studied invasions: that of the cane toad in Australia. Recently developed methods for population suppression include more effective traps based on the toad’s acoustic and pheromonal biology. New tools for containing spread include surveillance technologies (e.g., eDNA sampling and automated call detectors), as well as landscape-level barriers that exploit the toad’s vulnerability to desiccation—a strategy that could be significantly enhanced through the introduction of sedentary, range- core genotypes ahead of the invasion front. New methods to reduce the ecological impacts of toads include conditioned taste aversion in free-ranging predators, gene banking, and targeted gene flow. Lastly, recent advances in gene editing and gene drive technology hold the promise of modifying toad pheno- types in ways that may facilitate control or buffer impact. Synergies between these approaches hold great promise for novel and more effective means to combat the toad invasion and its consequent impacts on biodiversity.

Introduction fundamental biology of invaders by review- PREVENTING nonnative species from ing current ideas (most of them developed reaching new areas via quarantine is very recently) on ways to control and re- typically the most effective strategy to com- duce the ecological impact of the world’s bat biological invasions, although eradicating most intensively studied amphibian invasion: newly established populations can be feasi- the deliberate release and subsequent spread ble with intensive effort (Simberloff 2003). of the cane toad (Rhinella marina, formerly Once an invader begins to spread, however, Bufo marinus) in Australia (Figure 1). eradication success declines markedly, and Native to Latin America, these exception- substantial population reductions may be dif- ally large (up to more than 1 kg) and highly ficult to achieve. Effective control methods toxic anurans were released along the north- may take decades to develop and, with a few eastern coast of Australia in 1935 in a futile exceptions (Shanmuganathan et al. 2010), attempt to control insect pests in commer- are unlikely to succeed (Saunders et al. 2010). cial sugarcane plantations (Turvey 2013). Nonetheless, the potentially devastating eco- The toads have since spread at an increas- logical impacts of invasive species (Kraus 2015; ing rate across more than 1.2 million km2 Bellard et al. 2016) mean that we urgently of tropical and subtropical Australia (Urban need to develop and apply new control and et al. 2008), and have not yet filled their fun- mitigation methods. Our best hope of suc- damental niche, either in Australia (Kearney cess may come with high-profile pest spe- et al. 2008) or globally (Tingley et al. 2014). cies that have attracted intensive research Over the course of their Australian invasion, not only into control, but also basic biology cane toads have had devastating impacts on (Simberloff 2003). In this paper, we illus- native biodiversity (Shine 2010). In particu- trate the importance of understanding the lar, large anuran-eating predators (such as 126 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

Figure 1. A Male Cane Toad From the Northern Territory, Australia Photo by Ben Phillips. See the online edition for a color version of this figure. marsupial quolls, freshwater crocodiles, var- ter et al. 2008, 2011), other viral-based, gene anid and scincid lizards, and elapid snakes) deletion and RNAi-based control strategies are fatally poisoned when they ingest the (Shanmuganathan et al. 2010), and sex-bias- toxic invaders. In many areas, toad invasion ing mechanisms (Koopman 2006; Mahony has caused precipitous declines in predator and Clulow 2006) have been considered and abundance (Letnic et al. 2008; Doody et al. occasionally (in the case of virally vectored 2009, 2014; Shine 2010; Jolly et al. 2015; strategies) extensively investigated (Hyatt et al. Fukuda et al. 2016). These declines have, in 2008). Their likelihood and context of suc- turn, had flow-on effects to other species cess for managing toads have also been ana- via trophic cascades (Brown et al. 2011b; lyzed (McCallum 2006; Thresher and Bax Doody et al. 2013, 2015). 2006), but technical obstacles (and concerns Despite considerable financial investment about unintended consequences) prevented (e.g., more than $20 million from 1986– their successful development and implemen- 2009; Commonwealth of Australia 2011), tation (Hyatt et al. 2008; Shannon and Bayliss the spread of cane toads across Australia has 2008). continued unabated. Initial control efforts in Consequently, when the state of efforts to the wet-dry tropics of northern Australia fo- control cane toads was last reviewed by the cused on manually collecting adult toads and Australian Federal Government in 2011, it have been largely unsuccessful. The feasi- seemed likely that the species would fully bility of more sophisticated genetic-based colonize its potential range before any suit- biocontrol methods, such as virally vectored able technology for broad-scale control be- autoimmunity (Robinson et al. 2006; Pallis- came available (Commonwealth of Australia June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 127

2011). In its threat abatement plan, the Aus- trapping of adults to chemical suppression tralian Commonwealth therefore decided of larval development. to move away from broad-scale control and eradication, and focus instead on the pro- manual removal tection of key biodiversity assets, such as offshore islands and priority native species and The earliest terrestrial stages of the cane ecological communities (Commonwealth toad are largely diurnal, and restricted to the of Australia 2011). Suitable approaches for margins of natal ponds during dry weather achieving that aim were largely unavailable conditions, particularly in the wet-dry tropics at the time, but a recent ( June 2016) work- (Child et al. 2008a,b; Pizzatto et al. 2008). shop to review the threat abatement plan Metamorph toads can be collected or killed revealed substantial advances in our under- in large numbers if located before dispersing standing of chemical ecology, genetic tech- into the surrounding habitat matrix. Com- nologies, mechanisms of toad impact, and munity groups have sprayed chemicals to kill fundamental cane toad biology. The partic- metamorph toads, but the adverse ecologi- ipants of that workshop identiﬁed exciting cal impacts of this practice (e.g., on native opportunities to capitalize on these recent ants) have led to it being banned by environ- advances to develop innovative management mental-management authorities in some parts approaches for a variety of applications, in- of the toad’s Australian range (Kelehear et al. cluding at and behind the invasion front, 2012). Chemicals that were commonly used and on offshore islands (Table 1). for this purpose (e.g., DettolTM) also kill lung- In this paper, we provide an overview of worm parasites that might otherwise impair these recently proposed approaches and de- toad viability (Kelehear et al. 2012). velopments. We consider three main types Juvenile and adult toads are primarily noc- of management goals: suppressing toad pop- turnal, and often prefer relatively open hab- ulation sizes; containing toad spread; and itat at night when foraging (González-Bernal mitigating toad impacts on biodiversity. For et al. 2015). Adult males typically gather each goal, we summarize available manage- close to waterbodies, whereas adult females ment tools and provide, where available, are often dispersed through the surrounding evidence for their effectiveness. We also spec- habitat (González-Bernal et al. 2015). Thus, ulate on the effectiveness of each manage- hand collecting can result in male-biased ment tool in different climatic regions of the captures if removal is focused exclusively toad’s Australian range, and highlight poten- on waterbodies. Radiotelemetric monitor- tial barriers to implementation. We conclude ing suggests that on nights with unsuitable with potential integrated strategies that draw weather conditions, a high proportion of on multiple management approaches. Our adult toads remain inactive and, hence, are examples are focused on the Australian cane unavailable to collectors (Schwarzkopf and toad invasion, but the approaches we pres- Alford 1996, 2002). Nonetheless, hand col- ent here will be applicable to other countries lecting can remove a signiﬁcant proportion throughout the Caribbean Sea and Indian of adult toads within an area (Somaweera Ocean where cane toads have established non- and Shine 2012). The spatial and temporal native populations, as well as to amphibian in- impact of such activities differs among sites vasions more generally (Kraus 2015; Measey and seasons. Manual removal of adult toads et al. 2016). Additionally, the ideas generated has a minor and short-term impact on toad from intensive research on cane toads may abundances in areas close to the invasion suggest novel approaches to control a wide front in the tropics, where fast-dispersing variety of other invasive organisms. toads from nearby areas soon replenish local densities (Somaweera and Shine 2012). However, manual collection has longer-term Population Suppression impacts during the tropical dry season (when Most efforts at controlling cane toads have toads are sedentary; Brown et al. 2011a) and in focused on reducing toad abundance; these southern (cool climate) areas (M. J. Greenlees range from direct manual collection and and R. Shine, unpublished data). Manual 128

TABLE 1 Management goals and tools for controlling cane toads in Australia Management goal Action Life stage Time frame Barrier(s) to implementation

Population suppression Manual removal Postmetamorphic/tadpoles/eggs Short-term High dispersal and reproductive rates mean long-term reductions are difficult Fencing waterbodies Postmetamorphic Short-term Fence maintenance; collateral impacts Traps for adults Postmetamorphic Short-term High dispersal and reproductive rates mean long-term H UREL EIWO BIOLOGY OF REVIEW QUARTERLY THE reductions are difficult; checking traps remotely Traps for tadpoles Tadpoles Short-term High dispersal and reproductive rates mean long-term reductions are difficult Suppression pheromones Tadpoles Medium-term High dispersal and reproductive rates mean long-term reductions are difficult Using native species to reduce Tadpoles Short-term Effects on nontarget species toad numbers Using pathogens for toad control Postmetamorphic/tadpoles Medium-term Effects on nontarget species; current lack of suitable agent Containment Environmental DNA sampling Postmetamorphic/tadpoles/eggs Short-term Does not remove toads Automated call detection devices Postmetamorphic Short-term Does not remove toads Pilbara Line (waterless barrier) Postmetamorphic Medium-term Stakeholder willingness; design of leak-proof tank and trough systems Genetic backburning Postmetamorphic Medium-term Public acceptance Impact mitigation Conditioned taste aversion N/A Short-term Public acceptance Targeted gene flow N/A Short-term Public acceptance Gene banking and assisted N/A Short-term to Public perception that conservationists have reproduction medium-term “given up” Gene editing and genome Postmetamorphic Medium-term Proof of concept; public acceptance engineering (reducing toad toxicity)*

Listed for each action is the targeted life stage (postmetamorphic, tadpole, or egg), the amount of time needed for implementation (short-term: now–2 years; medium-term: 2–5 years), and

important barriers to implementation and/or success. Volume * Detoxified toad, sex change, or sex-specific mortality/infertility, inducible mortality or disease susceptibility. 92 June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 129 collection has been used to locally eradicate for cattle, pastoralists have created numer- an island population of cane toads off Ber- ous artificial watering points (AWPs; James muda (Wingate 2011), and appears to have et al. 1999). These typically consist of earthen contributed to the decline and disappear- bore-fed dams, which gravity feed water to ance of two separate, extralimital populations raised cattle troughs fitted with float valves. south of the main distributional range on In areas devoid of natural water, earthen the eastern Australian seaboard (in Port Mac- dams serve as dry season invasion hubs from quarie: White 2007; and more recently in which cane toads disperse during wet sea- Sydney: M. J. Greenlees and R. Shine, un- son rains (Letnic et al. 2015). published data). Cane toads cannot survive without fre- In summary, manual collection and re- quent access to water, or at least damp sub- moval of juvenile and adult cane toads can strates (Florance et al. 2011; Tingley and be useful in some places at some times, most Shine 2011; Jessop et al. 2013; Letnic et al. notably to eradicate satellite populations. It 2014; Webb et al. 2014) and, thus, fencing also has potential as a conservation tool in dams during the dry season is a simple way restricted areas (such as islands) where the to eradicate toads in arid and semiarid land- conservation benefit may be high (e.g., to scapes (Florance et al. 2011). Fences can be prevent sympatry with particularly sensitive constructed with star pickets, fencing wire, species). However, in order to be most effec- and shade cloth (Figure 2A); a team of work- tive, manual collection needs to be integrated ers can erect a fence in approximately three with other approaches (e.g., to capture indi- hours. An effective toad fence consists of two viduals over longer periods or to eliminate wires, one 600 mm high, and one flush with breeding). Manual collection of adult toads the ground, to which a roll of shade cloth has garnered sustained enthusiasm by the is attached with metal fencing clips. Fences public in various parts of Australia and re- should be constructed on the flat crest of the quires little expertise or equipment and, thus, dam wall, and should include a soil-covered is well-suited to community engagement (as 400 mm flange of shade cloth extending out- long as humane euthanasia protocols are in ward to prevent toads from burrowing under place; Shine et al. 2015). Most large-scale re- the fence (Letnic et al. 2015). Fences not moval efforts have occurred at the northern only exclude toads from water, but also fa- and southern invasion fronts, but popula- cilitate the hand collection of toads trapped tion reduction in long-established areas can inside the fenced area. also be important for rallying public support Fences are particularly effective for erad- for toad control, as human population den- icating toads in arid to semiarid environ- sities can be much higher in these regions. ments if they are constructed during the Disadvantages of manual removal include: late dry season when toads experience sig- the risk that untrained community mem- nificant thermal and hydric stress (Schwarz- bers will inadvertently collect and kill native kopf and Alford 1996; Jessop et al. 2013). frogs (Somaweera et al. 2010); and the risk In a replicated field experiment in the Vic- that if employed in isolation, or without stra- toria River district, local eradication of toads tegic consideration of spatial-temporal con- was achieved at fenced dams after seven days text, ineffectiveness of manual collecting will (Florance et al. 2011). Fences had no ad- render the public cynical about the possibil- verse effects on native wildlife; frog mortality ity of controlling toads and, thus, reluctant was negligible, and birds, pythons, kanga- to adopt other methods (although support roos, and dingoes easily traversed the fences has been sustained in some regions where (Florance et al. 2011). Despite no mainte- toads are long established, such as Towns- nance for a year, the fences remained func- ville, Queensland). tional. One year later, the fenced dams acted as ecological traps, attracting toads to a water source they could no longer access. Thus, fencing waterbodies fences not only eradicated toads, but pre- The dominant land use in semiarid Aus- vented them from reestablishing in the area tralia is cattle grazing, and to provide water the following dry season (Letnic et al. 2015). 130 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

Figure 2. Examples of Approaches to Suppress Cane Toad Populations (A) A cane toad exclusion fence erected around a farm dam in the Northern Territory, Australia. Photo by Jonathan Webb. (B) Tanks used to store water for cattle on a pastoral station in the Northern Territory. Such tanks reduce toad hydration and breeding opportunities. Photo by Jonathan Webb. (C) A solar-powered cane toad trap, set up as part of a surveillance strategy to prevent toads from establishing on Groote Eylandt, Northern Territory. Photo by Deborah Bower. (D) Illustration of the effects of chemical suppression on tadpole growth. Pictured are two cane toad tadpoles from the same clutch: a control tadpole (left), and a tadpole that has been exposed to the suppressant chemical (right). Photo by Greg Clarke. See the online edition for a color version of this ﬁgure.

Fencing waterbodies is most useful in arid leaks, toads will not be able to access the wa- landscapes; in well-watered regions, fencing ter, and so will be unable to persist in the area. isunlikelytohavemeasurableeffectsontoad Replacing dams with tanks will be more ex- numbers. Fencing is well suited to small-scale pensive than fencing, and will require the eradication attempts, but could also be used participation and cooperation of landholders to exclude toads from large areas, provided (Southwell et al. 2017). However, enclosed that fences are maintained over multiple water storage tanks reduce evaporation, and years (Letnic et al. 2015). Community groups thus present a potential benefit to pastoral- comtemplating using fences for toad control ists in arid regions. might consider the use of more rigid fencing materials than the shade cloth fences traps for adults used in previous research, although the impacts of such fences on native wildlife have For vertebrate pests, especially those with not been studied. An alternative solution is high reproductive rates such as cane toads, to replace earthen dams with metal or plas- it is often strategic to remove reproductive tic tanks (Figure 2B; Florance et al. 2011; females (Reidinger and Miller 2013), thereby Tingley et al. 2013). Providing that such in- reducing future population size. Male cane frastructure is well constructed and free of toads, like males of most anuran species, June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 131 make advertisement calls to attract females, motely. Methods to remove large numbers and this aspect of their biology can be ex- of toads humanely, and to monitor traps re- ploited to attract females to specific loca- motely, are in development. Traps should be tions, including traps (Figure 2C). Recent re- available for commercial purchase in 2018. search suggests that the call deployed on the trap can be engineered, using principles from traps for tadpoles studies of sexual selection in toads, to attract a higher proportion of females (calls attract The high fecundity of cane toads (up to both males and females), and to attract mostly 30,000 eggs per clutch; Lever 2001), com- gravid females (B. Muller et al., unpublished bined with their preference for anthropogen- data). For example, in call choice experiments, ically disturbed habitats (González-Bernal female toads chose the largest male available et al. 2016), generates high levels of intra- that was smaller than themselves, and pre- specific competition. That competition may ferred calls with a low pulse rate and domi- be especially intense during the larval phase, nant frequency, whereas males did not show when large numbers of tadpoles compete preferences for any particular call character- for access to limited food resources in small istics (Yasumiba et al. 2015). waterbodies (Cabrera-Guzmán et al. 2013b). Amphibian traps can also use lights to at- Recent research has revealed that cane toad tract invertebrates as a food source (Schwarz- larvae exhibit complex chemical communi- kopf and Alford 2007). UV “black” lights cation systems, which appear to have evolved work best to attract invertebrates to catch in the context of intraspecific competition. toads, as toads are repelled by white incan- Toad tadpoles reduce the recruitment of descent and fluorescent lights, but not by subsequently laid clutches by actively search- UV “black” lights (Davis et al. 2015). Traps ing out and consuming freshly laid toad with UV lights captured 10 times as many eggs (but not the eggs of native frogs; Cross- toads as those equipped with white fluores- land et al. 2012). The tadpoles locate those cent lights (Davis et al. 2015). Traps deploy- eggs by following chemical trails, specifically ing both a toad call and a light were more toxins (bufagenins) that are released as effective than those using a call alone (Yeager eggs develop (Crossland et al. 2012). Fun- et al. 2014). nel traps baited with toad toxins (obtained As suggested for manual removal, trap- by squeezing toxins from the parotoid glands ping in the dry season can be an effective of adult toads) can attract and remove vast time to remove adults. Preliminary mark-re- numbers of toad tadpoles from spawning capture studies conducted with trapping at waterbodies, with minimal collateral impact this time of year suggested that traps can re- on native frog larvae (although potential move up to 40% of the adult population (B. impacts on other taxa, such as invertebrates Muller, unpublished data). Future studies and fish, warrants additional research; Cross- could determine whether traps can be used land et al. 2012). To advance the development to eradicate extralimital incursions, or toad of a commercial tadpole trapping solution, populations on very small islands. the chemical attractant in cane toad eggs Humane disposal of large numbers of has been isolated and characterized, and trapped toads can pose barriers to the use a surrogate formation, prepared from adult of traps, as could inadvertent capture of non- toad parotoid secretion (BufoTab), has been target (especially, native) species of fauna; evaluated and patented (R. J. Capon, A. however, thus far, adult toad traps have not Salim, R. Shine, and M. R. Crossland. “Chem- captured any other species in Townsville, ical Attractant and Use Thereof, Patent.” Queensland, or surrounds, after thousands PAT-02059-AU-02. University of Queensland of hours of trapping (L. Schwarzkopf, per- and University of Sydney, Australia, 2012). sonal observation). In addition, any trapping Pilot studies, including those implemented program requires regular visits to traps to by the public, contractors, and local gov- ensure that individuals are removed, and ernment environment officers, have proved this can be challenging if traps are placed re- highly successful, with capture rates exceed- 132 THE QUARTERLY REVIEW OF BIOLOGY Volume 92 ing 10,000 tadpoles per trap (R. Shine and field experiment to date reported that eggs R. J. Capon, unpublished data). The patent were significantly suppressed after immer- for tadpole trapping technology was recently sion in a pond containing toad tadpoles, but licensed to a U.S.-based company, Springstar not after immersion in ponds without toad Inc., which will establish a Brisbane-based tadpoles (Clarke et al. 2016). sister company to further develop, manu- facture, and market a commercial product. using native species to In concert with this commercial approach, reduce toad numbers the University of Queensland has initiated a nationwide citizen science and community Laboratory studies have shown intense engagement program, the Cane Toad Chal- competition between cane toad tadpoles lenge (http://canetoadchallenge.net.au), to and the larvae of native frogs; in many cases build community awareness of cane toad con- the small, slow-swimming toad tadpoles are trol solutions, to marshal support, and to outcompeted by larger and more active tad- deliver tadpole trapping technology direct poles of native species (Alford et al. 1995; to the public. Alford 1999; Cabrera-Guzmán et al. 2011). Tadpole trapping is well suited to water- Thus, we can suppress toad larval survival ways abutting urban areas. The technology by encouraging native frogs to breed in the is cheap, semiautomated (set and leave), and ponds used by toads, or by introducing eggs with the provision of baits, easily and safely or tadpoles of native frogs directly (Cabrera- implemented by the public. Traps need to Guzmán et al. 2011, 2013b,c; Shine 2011). be checked daily, however, to avoid nontar- Similarly, many native invertebrates are vo- get impacts on aquatic fauna, particularly raciouspredatorsoftoadtadpoles(e.g.,water native tadpoles. The long-term ecological beetles and dragonfly larvae) or metamorph benefit of tadpole traps has yet to be estab- toads(ants), and wemight beable toincrease lished, but remains a work in progress. densities of such predators by manipulat- ing habitat attributes (Ward-Fear et al. 2009, 2010a,b; Cabrera-Guzmán et al. 2012, 2013a, suppression pheromones 2015b). Additionally, dense growth of vege- Intense competition between larval cane tation around waterbody edges discourages toads (discussed above) has resulted in the toad breeding, such that replanting these evolution of a species-specific chemical sup- areas might curtail toad recruitment (Hag- pression system. As soon as they have trans- man and Shine 2006; Semeniuk et al. formed into free-swimming tadpoles, toad 2007). Local government authorities in larvae begin to produce a substance that some parts of Australia insist on revegeta- interferes with embryonic development of tion around potential toad-spawning sites younger conspecifics (Crossland et al. 2012; following major building works, but there Clarke et al. 2015, 2016). Eggs or hatchlings arenofield studies of the effectivenessof this that encounter this substance exhibit se- method for toad control. Many Australian verely reduced rates of growth and survival predators also consume adult toads or parts (Figure 2D). The substance has no detect- thereof (typically leaving the parotoid glands able impact on the native frog species that uneaten), suggesting that high densities of have been tested to date, and thus may offer native rodents (Cabrera-Guzmán et al. 2015a) a species-specific suppressant of cane toad and raptorial birds (Beckmann and Shine reproduction (Crossland et al. 2012; Clarke 2011) might also contribute to controlling et al. 2015, 2016). Current studies are at- toad numbers. tempting to identify the nature of the suppressant substance, with a view to deploying using pathogens for toad control it in natural waterbodies to prevent toads from breeding. There are few documented cases of path- At present, evidence for the effectiveness ogens successfully controlling vertebrate of chemically mediated suppression is based pest populations; in all instances, they are primarily on laboratory work. The single introduced viruses controlling introduced June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 133 mammals (Shanmuganathan et al. 2010; Di Toad populations often undergo dramatic Giallonardo and Holmes 2015). Important declines that could plausibly be caused by characteristics of a potential biocontrol agent an unidentified pathogen (Freeland 1986; include host specificity to the target organ- Freeland et al. 1986; Speare 1990). Because ism and negligible impacts on native species toads congregate at high densities around (Simberloff 2012). In the case of cane toads, shrinking waterbodies during the dry season, additional aspects of an effective control there is ample opportunity for pathogen agent include self-dissemination and target- transfer among toads (Freeland 1986). Thus, ing of adults (Shanmuganathan et al. 2010). the search for new, toad-specificpathogens Introduced species often lose many of the remains an important pursuit (Shannon and pathogens they carry in their native range Bayliss 2008; Shanmuganathan et al. 2010). (Blossey and Notzold 1995; Torchin et al. Investigation into a recent (September 2003) and this appears to be the case in cane 2014) mortality event in wild toads observed toads (Barton 1997). Systematic surveys of near Darwin in the Northern Territory has cane toad pathogens have found that toads implicated an undescribed species of Ent- carry an array of potentially pathogenic or- amoeba as the cause of death (C. Shilton, J. ganisms, but that few are associated with Slapeta, and G. Brown, unpublished data). actual disease (Freeland et al. 1986; Delvin- Although this is the first pathogen identi- quier and Freeland 1988; Speare 1990; Free- fied to cause mortality in wild toads, further land 1994; Speare et al. 1997; Hyatt et al. study is needed to ascertain where the toads 1998; Linzey et al. 2003; Peacock 2006). acquired the organism, rates of infection and Several known or potential pathogens (e.g., disease, the circumstances that make the or- Mucor, Ranavirus, Myxidium)poseriskstona- ganism pathogenic, and the risk it poses to tive frog species. No promising candidate native species. pathogens have yet been found that might widely suppress toad populations the way Containment myxomatosis or rabbit calicivirus successfully suppressed feral rabbit populations (Shan- Where eradication is not an option (as muganathan et al. 2010). has so far proven to be the case for cane The finding that the lungworm infecting toads in Australia), the most powerful tool Australian cane toads was not Australian but for minimizing the impact of an invasive rather a South American parasite (Rhabdias species is to contain its spread (Epanchin- pseudosphaerocephala) brought with the foun- Niell and Hastings 2010). In most cases, this der toads, suggested a possible biocontrol amounts to creating or maintaining barriers mechanism (Dubey and Shine 2008). Ini- to natural spread. Such barriers may play tially, it seemed that strategically releasing out at several scales, from local asset protec- the lungworm in advance of the toad inva- tion (e.g., a wildlife park wanting to keep sion front might serve to infect and slow van- toads out of enclosures) to maintaining toad- guard toads, which are typically parasite-free free islands and keeping toads out of large (Phillips et al. 2010b; Kelehear et al. 2011; areas of the Australian mainland. Pizzatto and Shine 2012). Unfortunately, al- The price of containment is eternal vigi- though the lungworm does not establish in- lance. A barrier that has been effective for fections in most of the native frog species years can be undone by a single coloniza- tested, it is deadly to one species, the mag- tion event comprising a male and female nificent tree frog Litoria splendida (Pizzatto toad. It is currently unclear how commonly et al. 2010; Pizzatto and Shine 2011). Thus, such colonization events occur, nor the pri- the use of Rhabdias as a biocontrol agent mary modes by which they occur (although could have collateral negative impacts on see Massam and Gray 2011). Toads some- native fauna. Evidence that toads infected times colonize offshore islands if such is- with Rhabdias tend to disperse at greater, lands are close to shore, if they are in the not lesser, rates than uninfected individuals plume of a major river, and particularly in also argues against its utility in slowing toad the monsoon tropics where large volumes of spread (Brown et al. 2016). rain can leave freshwater lenses on seawater 134 THE QUARTERLY REVIEW OF BIOLOGY Volume 92 that persist for weeks. Establishment events densities (Thomsen et al. 2012; Smart et al. driven by accidental human introduction 2015). This sampling technique could, there- seem substantially rarer. Successful establish- fore, provide an early warning signal for new ment in extralimital areas on the mainland, toad incursions on offshore islands or help for example, is surprisingly rare (Lever 2001). identify extralimital populations on the main- Indeed, there are possibly billions of toads land. Environmental DNA sampling may on the continent, and huge flows of traffic prove particularly effective when toads are along the eastern seaboard where toads have congregated around waterbodies (e.g., in been present for more than 85 years, yet arid landscapes or during the dry season there have been only three nonintentional in more mesic environments). successful colonization events recorded ahead To the best of our knowledge, no previ- of the main invasion front. All of these have ous study has attempted to detect cane toad been at the southern invasion front in New eDNA, but numerous studies have applied South Wales (one to Port Macquarie; one this approach to detect amphibians in wet- to Yamba; and one to Sydney; although lands and streams (Goldberg et al. 2011; De- smaller-scale introductions have occurred jean et al. 2012; Thomsen et al. 2012; Pilliod near Yamba; M. J. Greenlees, unpublished et al. 2013; Rees et al. 2014). Furthermore, data). This low frequency of successful colo- the wealth of genetic sequence data from nization events occurs despite a large num- across the cane toads’ Australian range (Es- ber of individual toads being accidentally toup et al. 2004; Rollins et al. 2015; Trumbo transported (White and Shine 2009). Such et al. 2016) coupled with the fact that there transportation events typically involve only are no native toads in Australia, means that a single animal (and so cannot establish a the development of suitably specific genetic population); where they do involve multiple primers should be straightforward. The key animals, the natural tendency of toads to seg- to implementing this sampling technique regate by sex (Zug and Zug 1979; González- would be to identify likely cane toad breed- Bernal et al. 2015) means that these multiple ing sites near locations with a high likeli- animals are quite likely to be all the same hood of human-assisted colonization (e.g., sex. The practice of conducting “toad races” maritime shipping ports and airports on off- as a tourist pastime in Queensland has, how- shore islands; caravan parks, camping sites, ever, introduced toads to offshore islands and plant nurseries ahead of the mainland with resorts (L. Schwarzkopf, personal ob- invasion front). servation). Any containment strategy requires vigilant automated advertisement monitoring and a plan for rapid eradication call detection if a small population is detected. Upon detection, the tools for local population reduc- Male cane toad advertisement calls could tion and eradication (above) can be brought be used to detect toad presence ahead of into play, but successful eradication is more the mainland invasion front, or on offshore likely if an incursion is detected early. Be- islands. Automated detectors that use com- low we review the new tools available for puter algorithms to scan acoustic data for early detection of incursions, and sketch a cane toad calls were designed and estab- promising containment strategy for main- lished prior to 1996 (Taylor et al. 1996), land Australia. but both hardware and software for such tasks are improving rapidly, and new methods are constantly under development (Hu environmental dna sampling et al. 2009). The most useful versions of Environmental DNA (eDNA) sampling— these methods automatically detect calls in the detection of genetic material released the field, either in real time or with a short by organisms into the environment—can delay (less than one minute), and then up- be a remarkably sensitive method for de- load that information to a server or tele- tecting freshwater species at low population phone to allow an appropriate response. June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 135

Figure 3. Approximate Mainland Distribution of Cane Toads in Australia as of 2016 The box in Western Australia represents the approximate extent of the Kimberley-Pilbara corridor.

This approach successfully detected the junction with other detection methods, in- cane toad front as it advanced across the cluding eDNA detection and, possibly, adult Northern Territory (Grigg et al. 2006), but traps. was plagued by equipment failure. More generally, detecting specific sounds in a natural the pilbara line setting using algorithms is a difficult com- putational problem because there are other In arid landscapes, cane toads rely critically sounds, such as wind, rain, and leaves flut- on artificial watering points (AWPs), such tering, that can interfere with or mask animal as farm dams, for hydration and breeding calls, reducing the probability of detection. (see the section Fencing Waterbodies; Flo- Detection algorithms must be tuned for their rance et al. 2011; Tingley and Shine 2011). location of application and ground truthed Thus, excluding toads from a large number to ensure effectiveness (Towsey et al. 2012). of AWPs in an area ahead of the invasion In addition, application of such systems in front could potentially contain their spread. remote locations is hampered by high cost Florance et al. (2011) highlighted a promis- and inconsistent network coverage, which in- ing area for such a strategy, between the terferes with data upload. Acoustic detec- Kimberley and the Pilbara regions in West- tion would be best applied in locations at ern Australia (see inset of Figure 3). Here high risk of cane toad arrival, and in con- the Great Sandy Desert meets the coast, cre- 136 THE QUARTERLY REVIEW OF BIOLOGY Volume 92 ating a chokepoint with very few natural pe- were first introduced to Australia, they spread rennial waterbodies. Simulation modeling of around 10 km per year; on the invasion front, cane toad spread along this arid corridor sug- they now spread closer to 50 km per year gests that the invasion front could be halted (Urban et al. 2008). there by a waterless barrier approximately The vast difference in dispersal ability be- 80 km wide, for a total cost of about $4.5 mil- tween invasion front and long-established lion (Southwell et al. 2017). The benefitof toad populations suggests a radical idea for such an action would be that we keep cane improving the effectiveness of the waterless toads out of the Pilbara—an area with ex- barrier. What if, as the invasion front ap- tremely high endemism—and 268,000 km2 proached the Pilbara Line, we introduced of the toads’ potential range in Western Aus- toads from long-established populations on tralia (Tingley et al. 2013). the nearside of the waterless barrier? This The current proposal is to fit all AWPs action would mean that we would be asking with leak-free tank and trough systems (Fig- the barrier to stop the much less dispersive ure 2B), with troughs that are inaccessible to toads of the long-established populations toads (Southwell et al. 2017). This infrastruc- (Alford et al. 2009) and, if the introduction ture could then be monitored by pastoralists was done correctly, the highly dispersive (as part of their normal farm activities), and toads of the invasion front would never make any leaks repaired as a conservation cost. it to the barrier. That is, for the very modest Thus, we can engineer mutual benefit for cost of introducing toads from Queensland both pastoralists and conservationists. This ahead of the invasion front, we could make “Pilbara Line” represents an astonishingly the waterless barrier substantially more ef- cost-effective strategy to reduce the impact fective. of toads and secure many of the last popu- This is a radical idea, and would likely face lations of toad-impacted predators, such as considerable community and political op- yellow-spotted monitors Varanus panoptes and position. Nonetheless, early modeling work northern quolls Dasyurus hallucatus. There suggests that genetic backburning can sub- is some uncertainty around the future of wa- stantially increase the effectiveness of land- ter use in the area of the Pilbara Line, with scape barriers (Phillips et al. 2016). Although a government-subsidized push to increase more detailed modeling work is warranted, the use of groundwater in the area for crop- all of the evidence at hand suggests that— ping; a move that would make the Pilbara in the absence of unexpected surprises such Line substantially more expensive to imple- as strong assortative mating between inva- ment. Other than this ongoing uncertainty sion front and range core animals—the idea around land use, the idea is very well devel- should work. oped: it needs an endowed trust fund to fi- nance it and a small amount of research to Impact Mitigation design an effective tank and trough system. Even if attempts to reduce toad abundances (above) are successful, we are un- genetic backburning likely to eradicate toads completely over a The waterless barrier required to stop large part of their range. Even low densities toads at the Pilbara Line has to be approxi- of toads may be ecologically catastrophic mately 80 km wide because invasion-front for native predator populations: a predator toads move astonishing distances: up to 60 km (such as a quoll or varanid) may be killed if in a wet season. The reason they move such it encounters a single adult toad (O’Donnell huge distances (an order of magnitude fur- et al. 2010). As a result, substantial research ther than most other anurans move in their has been devoted to an alternative tactic: to lifetime; Smith and Green 2005) is because buffer the impact of toad invasion by changing toads on the invasion front have evolved to the outcomes of encounters between toads become hyperdispersive (Phillips et al. 2006, and anuran-eating predators (O’Donnell 2008, 2010a; Alford et al. 2009). When toads et al. 2010; Ward-Fear et al. 2016). Other June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 137 researchers have focused on retaining the of CTA learning in wild predators will likely genetic diversity likely to be lost when pred- differ depending on their physiology and ecol- ator populations collapse after toad inva- ogy. This is a necessary consideration in the sion, and developing ways to restore that design of management strategies. For go- lost genetic diversity to wild populations annas, live toads were significantly more ef- postdecline. fective in buffering toad impact than were nausea-inducing sausages of toad flesh (which conferred no survival benefit; Ward-Fear conditioned taste aversion training et al. 2016). Smaller predators generally are unaffected The coordinated deployment of toad-aver- by toad invasion at the population level, de- sion baits and “teacher toads” ahead of the spite occasional mortality when toads are in- toad-invasion front could help prevent local gested (Shine 2010). Toxin content increases extinctions of numerous vulnerable species, rapidly with toad body size, such that a small such as northern quolls and goannas. Toad toad may not contain enough toxin to kill aversion baits could be aerially deployed in a predator that ingests it (Phillips et al. remote rocky habitats favored by quolls 2003). Instead, some predators that ingest (O’Donnell et al. 2010), whereas teacher small toads become ill, and rapidly learn toads could be deployed on floodplains or to avoid toads thereafter (Webb et al. 2008, near waterbodies frequented by yellow-spot- 2011; Somaweera et al. 2011; Ward-Fear et al. ted monitors (Ward-Fear et al. 2016). Teacher 2016)—a phenomenon known as “condi- toads presumably have the potential to train tioned taste aversion” or CTA (Garcia et al. a much wider array of fauna; their perceived 1974). The reason that large predators (e.g., benefit is, therefore, likely currently under- quolls, varanids, and freshwater crocodiles) estimated. However, the window of time to are at high risk is that they attack large toads execute this strategy in the Kimberley re- (which dominate the invasion front; Phil- gion of Western Australia is rapidly closing. lips et al. 2006) and thus are killed by their Field trials to determine which nontarget first encounter with a toad. If we could ex- species consume toad-aversion baits are cur- pose toad-naïve predators to small toads rently underway in this region (D. Pearson, first, they might learn to avoid the lethally personal communication). large toads when they arrive (Ward-Fear It is important to note, however, that the et al. 2016). Additionally, we could deploy CTA approach depends upon predators gen- “toad aversion” baits—sausages of toad flesh eralizing from small toads or sausages to infused with an odorless, nausea-inducing large toads; the degree to which that condi- chemical—ahead of the toad invasion front tion is satisfied will depend upon the cues (Webb et al. 2008; O’Donnell et al. 2010; used for prey recognition. A sausage-trained Price-Rees et al. 2013). predator that cues in on the visual stimulus Laboratory and field trials with small offered by a large moving object may still at- “teacher toads” or toad-aversion baits have tack a large toad, whereas one that cues in produced encouraging results. In three of on chemical substances may avoid a large the predator species most imperiled by toad toad when it encounters one (provided that invasion, CTA training has significantly in- large live toads smell similar to a small toad creased survival rates. Initial studies on quolls or toad sausage). Northern quolls that were (D. hallucatus) used captive-raised individ- trained with small dead toads infused with a uals that were later released into the wild nausea-inducing chemical subsequently in- (O’Donnell et al. 2010); studies on blue- vestigated and sniffed large toads when re- tongued skinks (Tiliqua scincoides intermedia) leased back into the wild, but did not attack used wild-caught individuals briefly retained them ( J. Webb, personal observation). Vara- in captivity for CTA learning trials (Price-Rees nid lizards also rely heavily on chemical cues et al. 2011, 2013); and trials on yellow-spot- priortoattackingprey,soprovidedthatthe ted monitors (V. panoptes) exposed free-rang- lizards become ill after ingesting a small toad ing lizards to small live toads. The pathway or sausage, they may be less likely to attack a 138 THE QUARTERLY REVIEW OF BIOLOGY Volume 92 large toad compared to a toad-naïve animal. genome banking and assisted Future work could usefully explore the nature reproductive technologies of cues for prey recognition by potentially vulnerable predators, and hence the feasi- Permanent loss of genetic diversity is one bility of prey-stimulus-generalization in CTA of the greatest conservation challenges as- trials. sociated with declining populations (Frank- The largest implementation barrier to the ham 2010). Although the ideal solution for deployment of aversion baits is their viability managing declining populations is to arrest in the field. Nontarget species, such as ants the cause(s) of decline, this is not always and some bird species, may reduce bait avail- possible in the time frame required to pre- ability; whereas extreme temperatures in vent the erosion of genetic diversity. In such northern Australia may lead to rapid bait dis- cases, genome banking and assisted repro- integration and loss of cues. The largest hur- ductive technologies could play a pivotal role dle to the release of teacher toads is likely to in preventing genetic diversity loss, and in be public opinion. It is important, however, reversing the effects of inbreeding and ge- to remember that any mortality caused by netic bottlenecks through the production teacher toads would have been inevitable of live animals from genetic material stored once the toad-invasion front arrived; the ben- prior to decline (Clulow and Clulow 2016). efits of this strategy, therefore, substantially The genetic rescue of wild populations outweigh the risks. has been demonstrated (Madsen et al. 1999), and reintroducing lost genes to wild populations from cryopreserved spermatozoa has targeted gene flow been achieved for an endangered species Although toads have had major impacts (Howard et al. 2016). Madsen and colleagues on many Australian predators, populations demonstrated genetic rescue in a small, de- of these predators can still be found in areas clining snake population through transloca- where toads have been present for more tion of males from a separate population, than 50 years. The persistence of predator resulting in a dramatic reversal of popula- populations in these long-colonized areas tion decline (Madsen et al. 1999). Further, is often due to rapid adaptation to the pres- Howard et al. demonstrated that restoring ence of toads. These adaptations can be lost genes to a population of endangered seen in predator morphology, but most pow- black-footed ferrets could be achieved by erfully take the form of behavioral avoid- artificial insemination (AI), using cryopre- ance of toads as prey (Phillips and Shine served sperm that had been stored for de- 2004, 2006; Llewelyn et al. 2014). If there cades from the founders of a captive breeding is a genetic basis to toad-smart behavior, population (Howard et al. 2016). there is no reason why we should not be able Encouragingly, work has already begun to use targeted gene flow (Kelly and Phillips to develop protocols for collection and cryo- 2016) to introduce those genes into toad na- preservation of varanid spermatozoa (one ïve populations. This can simply be achieved of the taxa most heavily impacted by toads, by breeding predators from long-colonized and a logical starting point for reptile ge- areas with populations soon to be impacted, nome storage; see Clulow and Clulow 2016) or by using assisted reproductive techniques and on the AI procedures required to use to move genes into populations via gametes cryopreserved sperm to restore genes. Tar- (see section below), eliminating the need to geted collection and storage of spermatozoa move live animals long distances or across bor- is a realistic goal in the short to medium time ders, thereby minimizing the transmission frame available before toads cause further of unwanted pathogens. Potential outbreed- loss of genetic diversity in many wild popu- ing depression in native predators could be lations of predatory reptiles and marsupials mitigated by sourcing individuals from pop- (Clulow and Clulow 2016). Preliminary stud- ulations that are climatically and ecologically ies have successfully collected spermatozoa similar to the recipient location. from live animals via electroejaculation, and June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 139 from animals postmortem. Optimization of peat (CRISPR) and the associated nuclease, these techniques in the near future will en- CRISPR-associated protein 9 (Cas9). The able collection and permanent storage of CRISPR/Cas9 system is derived from a form spermatozoa from wild populations in ad- of “adaptive antiviral immune mechanism” vance of the toad-invasion front, in facilities found in bacteria. It has been recoded for that are now available in Australia and else- use in animals, and uses the nucleotide se- where. This material will then be available quence complementarity of a short guide to restore genetic diversity postdecline via RNA to enable the Cas9 endonuclease to AI into wild females, potentially providing find, match, and cut the DNA of a target genetic rescue (Clulow and Clulow 2016). gene in an animal’s genome. Cellular repair One potential barrier to implementation mechanisms, including homology-directed of this approach is public and scientific resis- repair and nonhomologous end joining, can tance to genome storage and assisted repro- then be utilized and manipulated to either duction for wildlife conservation. However, disrupt, remove, edit, or insert genetic traits this resistance is greatly at odds with the gen- with exquisite precision (Doudna and Char- erally positive attitudes displayed toward the pentier 2014). same approaches in plants (i.e., seed banks, The CRISPR/Cas9 system has already been both for conservation and agriculture), in successfully applied to the western clawed frog industry (horse, cattle, and livestock breed- Xenopus tropicalis to generate gene deletions ing and movement), and in human repro- that have provided insights into pigment bi- duction (e.g., sperm storage and in vitro ology (Shigeta et al. 2016). When adapting fertilization). It is also at odds with the simi- these methods to the cane toad, an obvious lar, albeit more costly and limited approach, first approach would be to target a signifi- of captive assurance colonies and reintro- cant pathway of impact: the toad’s toxin. ductions of live animals, often championed As the structure, composition, and gene ex- by conservation scientists and the public. This pression pathways of the toxins are revealed, is likely a perception issue around the use of key enzymes in the pathway to their pro- more sophisticated technologies (e.g., using duction may be identified in the genome se- artificial insemination with animals versus quence, and these could be targeted using planting a seed in the ground), and should CRISPR/Cas9 to reduce or eliminate their ease as these technologies become more production. Recent research (R. J. Capon widely used. et al., unpublished data) has identified one key enzyme that, if removed using a CRISPR/Cas9-mediated knockout approach, gene editing and genome engineering holds great promise for generating a toad Advances in DNA sequencing have enabled that produces a nonlethal product from the generation of full genome sequences in the toxin pathway. reasonable time frames and at reasonable The CRISPR/Cas9 system can also be cost (PacBio sequencing; Rhoads and Au used to assemble a so-called “gene drive” 2015). There is a current concerted effort (see Esvelt et al. 2014 for details). At every to determine a full genome sequence for reproduction event, the gene drive dupli- the cane toad (P. White, L. Rollins, and cates itself, and all offspring of that repro- E. Holmes,unpublished data),which will pro- ductive event will carry the gene drive vide the information needed to probe crit- (Figure 4; Esvelt et al. 2014). To this simple ical aspects of cane toad biology with gene gene drive cassette other genes can be added editing. (often referred to as “payload”). These an- In recent years, exciting new molecular cillary genes would lead to a changed biolog- technologies have emerged that enable the ical trait, e.g., reduced fitness (Esvelt et al. specific and directed editing of the genome 2014), as they are pushed into the gene of an organism (Gaj et al. 2013). The most pool alongside the CRISPR/Cas9 cassette. widely used technique involves the Clustered Over a number of generations, the theoret- Regularly Interspaced Short Palindromic Re- ical expectation is that all members of the 140 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

Figure 4. Spread of a Gene Drive in a Population by Sexual Reproduction When a gene drive carrying animal (dark gray) mates with a wild type animal (light gray), all offspring will inherit the gene drive. This way the gene drive and its associated trait (e.g., reduced or no toxin production) can penetrate quickly into an entire population. Modified from Esvelt et al. (2014). population will carry the modified trait. This sequences of releasing gene drives are yet to outcome is expected even if that trait has a be understood. One risk is that an animal fitness penalty, providing it does not block carrying a gene drive could accidentally or sexual reproduction, on which it relies for maliciously reach its native range (in this case propagation. The use of CRISPR/Cas9 gene South America). There is currently a very drives is already well advanced in insects, active international debate regarding how where its scope is mainly in rendering dis- risks can be assessed and mitigated (Oye ease vectors incapable of spreading human et al. 2014; Webber et al. 2015). However, and veterinary pathogens (Gantz et al. 2015). while this debate continues, proof of con- More recently, gene drive approaches have cept work can be carried out using precision been actively discussed for use in reducing genome editing via CRISPR/Cas9, without pest animal populations and impacts (Esvelt the added gene drive function (e.g., to assess et al. 2014). the viability of detoxified cane toads). With The production and potential release of a few nucleotides removed and no added ge- genetically modified (GM) animals would netic material, such an organism might not involve community engagement—to ensure be considered (or regulated) as a GM organ- that there is social licence to undertake de- ism. If this is effective, a modification of this ployment—and regulation and policy ap- approach involving an integrated transgene provals from relevant state and federal (not a gene drive) to interfere with toxin governments. Engagement with indigenous production could be generated that would communities on whose land the cane toad segregate and breed into the population as is impacting would be critical. All of this a dominant Mendelian genetic trait (such would need to be preceded by appropriate a toad would be regulated as GM). Each of modeling of postrelease gene flow in tar- these approaches could be steps to provide get populations, and a thorough risk assess- data for the evaluation and consideration ment. As transmission of gene drives is via of gene drive to push the “detoxified” trait sexual reproduction, there is no risk to non- into the wider toad population, such that the target species; however, the longer-term con- impact on key predators is all but removed. June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 141

Figure 5. Synergies Between Proposed Threat Abatement Strategies and Management Actions for Cane Toads in Australia Each box represents an action that can be carried out independently under one of three broad strategies: population suppression, containment, and impact mitigation. Arrows depict directions of “synergies” (i.e., actions that can be achieved concurrently or an action that enhances the effectiveness of another). Brackets at the right group the management actions that can be applied synergistically within each of the three regions of cane toad occupation (nondispersive, established regions, and the dispersive invasion fronts of the North and South).

Genome engineering could be used not ratios to eventually achieve reproductive pop- only to develop nontoxic toads, but could ulation crashes or rendering a species sus- also be used to suppress cane toad popula- ceptible to a lethal treatment (e.g., a prodrug) tions. Genome engineering could, for ex- that is otherwise harmless to nonmodified ample, enable the modification of genes animals or other species. controlling growth, physiology, life-stage tran- sitions, or sexual development. This could Synergies empower functional studies of the biology of the cane toad in the search for its Achil- Excitingly, many of the tools proposed in les heel, and the pinch point through which this paper create opportunities to design in- a strong functional biological control might tegrated management strategies that exploit be exerted, some of which might not be ge- synergies, thereby maximizing both effec- netic. Theoretically, there is a wide variety tiveness and resource efficiency (Figure 5). of approaches that can be used with a gene Below, we provide examples of integrated drive system, including biasing offspring sex management strategies that capitalize on 142 THE QUARTERLY REVIEW OF BIOLOGY Volume 92 some of these synergies, and discuss their removal (M. J. Greenlees and R. Shine, un- applicability to invasion front versus long- published data). established toad populations. Finally, the promise of new genetic technologies (CRISPR) is one of widespread control or mitigation. If such a promise is met, population suppression then it provides an endpoint to containment efforts. Containment need only work for as Although we currently lack the tools needed long as it takes us to develop powerful genetic to eradicate toads from mainland Australia, techniques for toad control. we can substantially increase the effectiveness of local population control at waterbodies by implementing multiple tools concurrently. impact mitigation For example, fences could be erected around Conditioned taste aversion (CTA) with waterbodies to increase the effectiveness of wild predators and targeted gene flow cur- manual removal of juveniles and adults. Sur- rently present the best options to mitigate vival and recruitment could simultaneously toad impacts on native predators ahead of be reduced by trapping both adults and tad- invasion fronts. These techniques could also poles, while applying suppressor pheromones be used to create “toad-smart” predators for and encouraging native predators of toads reintroduction behind the invasion front, in and around waterbodies. This suite of tools where predators were historically abun- could be applied across the toad’s entire dant. These predators could be produced range, although fencing is more effective from the same mainland populations using in arid landscapes than in landscapes with banked genetic stock and AI, enabling the greater water availability. restoration of predecline genetic diversity. Alternatively, predators for CTA training and reintroduction could be collected from containment offshore islands that are currently being kept As outlined above, the effectiveness of the toad-free via population suppression and/or Pilbara Line at the northern invasion front containment methods. could be enhanced substantially via genetic Recent advances in genome engineering backburning. The effectiveness of this strategy present additional synergies—if we can suc- also could be bolstered by reducing propa- cessfully reduce toad toxicity at all life stages, gule pressure behind, or within, the waterless we could potentially engineer “teacher toads” barrier, using population suppression tools. with drastically reduced toxicity for release Similarly, population suppression tools could into the wild. Engineered toads would need be applied to reduce propagule pressure to remain sufficiently unpalatable but non- near areas that are likely to serve as sources lethal, such that toads of any size, including of island incursions (e.g., shipping ports). full-grown adults, could be used to “teach” Some of these tools, such as adult traps and predators ahead of the invasion front to avoid tadpole traps, could be coupled with contain- consuming wild-type (lethal) toads as they ment tools, such as eDNA sampling and auto- arrive in a new area. The benefits of this mated call detectors, to preemptively survey would be threefold: opportunities for learn- areas that are currently toad-free but at high ing in native predators would persist in the risk of invasion. landscape for longer; risk of collateral dam- Population reduction tools could also im- age to nontarget species would be diminished; prove the likelihood of containment at the and the “teacher toad” methodology would southern invasion front. In New South Wales, be more appealing to the public. for example, satellite populations have been Gene banking and assisted reproductive successfully eradicated using a strategy of sus- technologies could further assist in the spread tained and systematic population reduction, of engineered toad genotypes; sperm from i.e., a combination of adult and tadpole less toxic toads could be banked to store and trapping, fencing waterbodies, and manual maintain lines for subsequent release into June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 143 wild populations. Samples of banked sper- as release of “teacher toads”) may well be matozoa can be easily transported to any lo- amenable to rapid resolution. cation and subsequently used to produce tens When evaluating the strategies outlined of thousands of offspring via IVF, a technol- here, the question should not only be, “what ogy that exists for amphibians and has been isthecostoftrying?” but also, “what is the applied globally (Clulow et al. 2012; Lawson cost of doing nothing?” The cane toad sys- et al. 2013; Clulow and Clulow 2016). tem has attracted very detailed research, providing a diverse array of novel methods for invader control. For some of those methods Conclusion (such as trapping and CTA induction), we The cane toad’s conquest of Australia is have extensive field data to demonstrate ef- one of the best-documented examples of the fectiveness and lack of collateral impact. wide-ranging ecological impacts of an inva- Now is the time for implementation: toads sive species. However, recent advances in will have fully colonized the Kimberley within our understanding of cane toad biology have a decade. led to the development and (in some cases) More broadly, the Australian cane toad application of exciting new methodologies invasion illustrates the importance of under- to suppress toad populations, contain their standing the fundamental biology of invad- spread, and mitigate their ecological im- ers, and how we can exploit that knowledge pacts. Although many of these methods are to develop effective new strategies for in- practical, some (such as CTA, the Pilbara vader control and mitigation of invader im- Line, and genome engineering) may be con- pact. The cane toad example lends credence troversial for certain groups among the gen- to the exhortation to know one’senemy;itis eral public. The obstacles to implementation from such knowledge that creative strategies thus involve the need for public engagement can spring. Investment in basic research on strategies to disseminate information to the invader biology will, therefore, be a power- public (in order to facilitate informed de- ful strategy for success. bate), as well as scientific issues. We need to clearly identify potential collateral risks of any control measures, and encourage pub- acknowledgments lic discussion of the way forward. Issues asso- We thank the Australian Department of the Environ- ciated with genome engineering likely will ment and Energy for hosting the workshop that brought require sustained discussion, whereas poten- the authors together, and all workshop participants for tial objections to the simpler methods (such their input.

REFERENCES

Alford R. A. 1999. Ecology: resource use, competition, Barton D. P. 1997. Introduced animals and their par- and predation. Pages 240–278 in Tadpoles: The Biol- asites: the cane toad, Bufo marinus, in Australia. ogy of Anuran Larvae, edited by R. W. McDiarmid Australian Journal of Ecology 22:316–324. and R. Altig. Chicago (Illinois): University of Chi- Beckmann C., Shine R. 2011. Toad’s tongue for break- cago Press. fast: exploitation of a novel prey type, the invasive Alford R. A., Cohen M. P., Crossland M. R., Hearnden cane toad, by scavenging raptors in tropical Austra- M. N., Schwarzkopf L. 1995. Population biology of lia. Biological Invasions 13:1447–1455. Bufo marinus in northern Australia. Pages 172–181 Bellard C., Cassey P., Blackburn T. M. 2016. Alien spe- in Wetland Research in the Wet-Dry Tropics of Australia, cies as a driver of recent extinctions. Biology Letters edited by C. M. Finlayson. Canberra (Australia): 12:20150623. Supervising Scientist Reports. Blossey B., Notzold R. 1995. Evolution of increased Alford R. A., Brown G. P., Schwarzkopf L., Phillips B. L., competitive ability in invasive nonindigenous plants: Shine R. 2009. Comparisons through time and a hypothesis. Journal of Ecology 83:887–889. space suggest rapid evolution of dispersal behaviour Brown G. P., Kelehear C., Shine R. 2011a. Effects of in an invasive species. Wildlife Research 36:23–28. seasonal aridity on the ecology and behaviour of 144 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

invasive cane toads in the Australian wet-dry trop- tion of threatened amphibians and reptiles: bring- ics. Functional Ecology 25:1339–1347. ing the ARTs up to speed. Reproduction, Fertility and Brown G. P., Phillips B. L., Shine R. 2011b. The eco- Development 28:1116–1132. logical impact of invasive cane toads on tropical Clulow J., Clulow S., Guo J., French A. J., Mahony M. J., snakes: field data do not support laboratory-based Archer M. 2012. Optimisation of an oviposition pro- predictions. Ecology 92:422–431. tocol employing human chorionic and pregnant Brown G. P., Kelehear C., Pizzatto L., Shine R. 2016. mare serum gonadotropins in the barred frog Mix- The impact of lungworm parasites on rates of dis- ophyes fasciolatus (Myobatrachidae). Reproductive Bi- persal of their anuran host, the invasive cane toad. ology and Endocrinology 10:60. Biological Invasions 18:103–114. Commonwealth of Australia. 2011. Threat abatement Cabrera-Guzmán E., Crossland M. R., Shine R. 2011. plan for the biological effects, including lethal toxic Can we use the tadpoles of Australian frogs to re- ingestion, caused by cane toads. Canberra (Austra- duce recruitment of invasive cane toads? Journal lia): Department of Sustainability, Environment, of Applied Ecology 48:462–470. Water, Population and Communities. Cabrera-Guzmán E., Crossland M. R., Shine R. 2012. Pre- Crossland M. R., Haramura T., Salim A. A., Capon dation on the eggs and larvae of invasive cane toads R. J., Shine R. 2012. Exploiting intraspecific com- (Rhinella marina) by native aquatic invertebrates in petitive mechanisms to control invasive cane toads tropical Australia. Biological Conservation 153:1–9. (Rhinella marina). Proceedings of the Royal Society B: Cabrera-Guzmán E., Crossland M. R., González-Bernal Biological Sciences 279:3436–3442. E., Shine R. 2013a. The interacting effects of ungulate Davis J. L., Alford R. A., Schwarzkopf L. 2015. Some hoofprints and predatory native ants on metamorph lights repel amphibians: implications for improving cane toads in tropical Australia. PLOS ONE 8:e79496. trap lures for invasive species. International Journal of Cabrera-Guzmán E., Crossland M. R., Shine R. 2013b. Pest Management 61:305–311. Mechanisms of competition between tadpoles of Aus- Dejean T., Valentini A., Miquel C., Taberlet P., Belle- tralian frogs (Litoria spp.) and invasive cane toads main E., Miaud C. 2012. Improved detection of an (Rhinella marina). Freshwater Biology 58:2584–2600. alien invasive species through environmental DNA Cabrera-Guzmán E., Crossland M. R., Shine R. 2013c. barcoding: the example of the American bullfrog Competing tadpoles: Australian native frogs affect Lithobates catesbeianus. Journal of Applied Ecology 49: invasive cane toads (Rhinella marina) in natural 953–959. waterbodies. Austral Ecology 38:896–904. Delvinquier B. L. J., Freeland W. J. 1988. Protozoan Cabrera-Guzmán E., Crossland M. R., Pearson D., parasites of the cane toad, Bufo marinus, in Austra- Webb J. K., Shine R. 2015a. Predation on invasive lia. Australian Journal of Zoology 36:301–316. cane toads (Rhinella marina) by native Australian Di Giallonardo F., Holmes E. C. 2015. Exploring host– rodents. Journal of Pest Science 88:143–153. pathogen interactions through biological control. Cabrera-Guzmán E., Crossland M. R., Shine R. 2015b. PLOS Pathogens 11:e1004865. Invasive cane toads as prey for native arthropod Doody J. S., Green B., Rhind D., Castellano C. M., Sims predators in tropical Australia. Herpetological Mono- R., Robinson T. 2009. Population-level declines in graphs 29:28–39. Australian predators caused by an invasive species. Child T., Phillips B. L., Brown G. P., Shine R. 2008a. Animal Conservation 12:46–53. The spatial ecology of cane toads (Bufo marinus) Doody J. S., Castellano C. M., Rhind D., Green B. in tropical Australia: why do metamorph toads stay 2013. Indirect facilitation of a native mesopredator near the water? Austral Ecology 33:630–640. by an invasive species: are cane toads re-shaping Child T., Phillips B. L., Shine R. 2008b. Abiotic and biotic tropical riparian communities? Biological Invasions influences on the dispersal behavior of metamorph 15:559–568. cane toads (Bufo marinus) in tropical Australia. Jour- Doody J. S., Mayes P., Clulow S., Rhind D., Green B., nal of Experimental Zoology Part A: Ecological Genetics Castellano C. M., D'Amore D., Mchenry C. 2014. and Physiology 309A:215–224. Impacts of the invasive cane toad on aquatic rep- Clarke G. S., Crossland M. R., Shilton C., Shine R. tiles in a highly modified ecosystem: the importance 2015. Chemical suppression of embryonic cane toads of replicating impact studies. Biological Invasions 16: Rhinella marina by larval conspecifics. Journal of Ap- 2303–2309. plied Ecology 52:1547–1557. Doody J. S., Soanes R., Castellano C. M., Rhind D., Green Clarke G. S., Crossland M. R., Shine R. 2016. Can we B., McHenry C. R., Clulow S. 2015. Invasive toads shift control the invasive cane toad using chemicals that predator–prey densities in animal communities by have evolved under intraspecific competition? Eco- removing top predators. Ecology 96:2544–2554. logical Applications 26:463–474. Doudna J. A., Charpentier E. 2014. The new frontier Clulow J., Clulow S. 2016. Cryopreservation and other of genome engineering with CRISPR-Cas9. Science assisted reproductive technologies for the conserva- 346:1258096. June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 145

Dubey S., Shine R. 2008. Origin of the parasites of an González-Bernal E., Brown G. P., Crowther M. S., invading species, the Australian cane toad (Bufo ma- Shine R. 2015. Sex and age differences in habitat rinus): are the lungworms Australian or American? use by invasive cane toads (Rhinella marina)andana- Molecular Ecology 17:4418–4424. tive anuran (Cyclorana australis) in the Australian Epanchin-Niell R. S., Hastings A. 2010. Controlling es- wet–dry tropics. Austral Ecology 40:953–961. tablished invaders: integrating economics and spread González-Bernal E., Greenlees M. J., Brown G. P., dynamics to determine optimal management. Ecol- Shine R. 2016. Toads in the backyard: why do inva- ogy Letters 13:528–541. sive cane toads (Rhinella marina) prefer buildings to Estoup A., Beaumont M., Sennedot F., Moritz C., bushland? Population Ecology 58:293–302. Cornuet J.-M. 2004. Genetic analysis of complex Grigg G., Taylor A., MacCallum H., Fletcher L. 2006. demographic scenarios: spatially expanding popu- Monitoring the impact of cane toads (Bufo marinus) lations of the cane toad, Bufo marinus. Evolution 58: on Northern Territory frogs—a progress report. 2021–2036. Pages 47–54 in Science of Cane Toad Invasion and Esvelt K. M., Smidler A. L., Catteruccia F., Church G. M. Control, edited by K. Molloy and W. R. Henderson. 2014. Concerning RNA-guided gene drives for the Canberra (Australia): Invasive Animals Cooperative alteration of wild populations. eLife 3:e03401. Research Centre. Florance D., Webb J. K., Dempster T., Kearney M. R., Hagman M., Shine R. 2006. Spawning site selection by Worthing A., Letnic M. 2011. Excluding access to feral cane toads (Bufo marinus) at an invasion front invasion hubs can contain the spread of an invasive in tropical Australia. Austral Ecology 31:551–558. vertebrate. Proceedings of the Royal Society B: Biological Howard J. G., Lynch C., Santymire R. M., Marinari Sciences 278:2900–2908. P. E., Wildt D. E. 2016. Recovery of gene diversity Frankham R. 2010. Challenges and opportunities of using long-term cryopreserved spermatozoa and ar- genetic approaches to biological conservation. Bio- tificial insemination in the endangered black-footed logical Conservation 143:1919–1927. ferret. Animal Conservation 19:102–111. Freeland W. J. 1986. Populations of cane toad, Bufo Hu W., Bulusu N., Chou C. T., Jha S., Taylor A., Tran marinus, in relation to time since colonization. Aus- V. N. 2009. Design and evaluation of a hybrid sen- tralian Wildlife Research 13:321–329. sor network for cane toad monitoring. ACM Trans- Freeland W. J. 1994. Parasites, pathogens and the im- actions on Sensor Networks 5:4. pact of introduced organisms on the balance of na- Hyatt A., Pallister J., Hardy C., Halliday D., Shanmuga- ture in Australia. Pages 171–180 in Conservation Biology nathan T., Venables D. 2008. The development of in Australia and Oceania,editedbyC.Moritzand a cane toad biological control (2007–2008). Geelong J. Kikkawa. Sydney (Australia): Surrey Beatty and Sons. (Australia): Australian Animal Health Laboratory, Freeland W. J., Delvinqueir B. L. J., Bonnin B. 1986. CSIRO. Food and parasitism of the cane toad, Bufo marinus, Hyatt A. D., Zupanovic Z., Parkes H. 1998. Identification, in relation to time since colonization. Australian Characterisation and Assessment of Venezuelan Viruses Wildlife Research 13:489–499. for Potential Use as Biological Control Agents Against Fukuda Y., Tingley R., Crase B., Webb G., Saalfeld K. the Cane Toad (Bufo marinus) in Australia. Geelong 2016. Long-term monitoring reveals declines in an (Australia): Australian Animal Health Laboratory, endemic predator following invasion by an exotic CSIRO. prey species. Animal Conservation 19:75–87. James C. D., Landsberg J., Morton S. R. 1999. Provi- Gaj, T., Gersbach C. A., Barbas C. F. III. 2013. ZFN, sion of watering points in the Australian arid zone: TALEN, and CRISPR/Cas-based methods for ge- a review of effects on biota. Journal of Arid Environ- nome engineering. Trends in Biotechnology 31:397– ments 41:87–121. 405. Jessop T. S., Letnic M., Webb J. K., Dempster T. 2013. Gantz V. M., Jasinskiene N., Tatarenkova O., Fazekas A., Adrenocortical stress responses influence an inva- Macias V. M., Bier E., James A. A. 2015. Highly ef- sive vertebrate’s fitness in an extreme environment. ficient Cas9-mediated gene drive for population mod- Proceedings of the Royal Society B: Biological Sciences 280: ification of the malaria vector mosquito Anopheles 20131444. stephensi. Proceedings of the National Academy of Sciences Jolly C. J., Shine R., Greenlees M. J. 2015. The im- of the United States of America 112:E6736–E6743. pact of invasive cane toads on native wildlife in Garcia J., Hankins W. G., Rusiniak K. W. 1974. Behav- southern Australia. Ecology and Evolution 5:3879– ioral regulation of the milieu interne in man and 3894. rat. Science 185:824–831. Kearney M., Phillips B. L., Tracy C. R., Christian K. A., Goldberg C. S., Pilliod D. S., Arkle R. S., Waits L. P. 2011. Betts G., Porter W. P. 2008. Modelling species dis- Molecular detection of vertebrates in stream water: a tributions without using species distributions: the demonstration using Rocky Mountain tailed frogs cane toad in Australia under current and future and Idaho giant salamanders. PLOS ONE 6:e22746. climates. Ecography 31:423–434. 146 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

Kelehear C., Brown G. P., Shine R. 2011. Influence of NRM&W Cane Toad Workshop, edited by K. Molloy lung parasites on the growth rates of free-ranging and W. Henderson. Canberra (Australia): Invasive and captive adult cane toads. Oecologia 165:585–592. Animals Cooperative Research Centre. Available at Kelehear C., Cabrera-Guzmán E., Shine R. 2012. Inadver- http://www.pestsmart.org.au/wp-content/uploads tent consequences of community-based efforts to con- /2010/03/CaneToadProc.pdf. trol invasive species. Conservation Letters 5:360–365. Massam M., Gray G. 2011. Cane toad (Bufo marinus) Kelly E., Phillips B. L. 2016. Targeted gene flow for pathway analysis for Western Australia. Perth (Aus- conservation. Conservation Biology 30:259–267. tralia): Department of Agriculture and Food, West- Koopman P. 2006. Daughterless cane toads. Pages ern Australia. 111–116 in Science of Cane Toad Invasion and Control: McCallum H. 2006. Modelling potential control strat- Proceedings of the Invasive Animals CRC/CSIRO/Qld egies for cane toads. Pages 123–133 in Science of Cane NRM&W Cane Toad Workshop, edited by K. Molloy Toad Invasion and Control: Proceedings of the Invasive and W. Henderson. Canberra (Australia): Invasive Animals CRC/CSIRO/Qld NRM&W Cane Toad Work- Animals Cooperative Research Centre. Available at shop, edited by K. Molloy and W. Henderson. Can- http://www.pestsmart.org.au/wp-content/uploads berra (Australia): Invasive Animals Cooperative Re- /2010/03/CaneToadProc.pdf. search Centre. Available at http://www.pestsmart Kraus F. 2015. Impacts from invasive reptiles and am- .org.au/wp-content/uploads/2010/03/CaneToad phibians. Annual Review of Ecology, Evolution, and Sys- Proc.pdf. tematics 46:75–97. Measey G. J., Vimercati G., de Villiers F. A., Mokhatla Lawson B., Clulow S., Mahony M. J., Clulow J. 2013. M., Davies S. J., Thorp C. J., Rebelo A. D., Kum- Towards gene banking amphibian maternal germ schick S. 2016. A global assessment of alien am- lines: short-term incubation, cryoprotectant toler- phibian impacts in a formal framework. Diversity ance and cryopreservation of embryonic cells of and Distributions 22:970–981. the frog, Limnodynastes peronii. PLOS ONE 8:e60760. O’Donnell S., Webb J. K., Shine R. 2010. Conditioned Letnic M., Webb J. K., Shine R. 2008. Invasive cane taste aversion enhances the survival of an endan- toads (Bufo marinus) cause mass mortality of fresh- gered predator imperilled by a toxic invader. Jour- water crocodiles (Crocodylus johnstoni) in tropical nal of Applied Ecology 47:558–565. Australia. Biological Conservation 141:1773–1782. Oye K. A., Esvelt K., Appleton E., Catteruccia F., Letnic M., Webb J. K., Jessop T. S., Florance D., Church G., Kuiken T., Lightfoot S. B.-Y., McNamara Dempster T. 2014. Artificial water points facilitate J., Smidler A., Collins J. P. 2014. Regulating gene the spread of an invasive vertebrate in arid Austra- drives. Science 345:626–628. lia. Journal of Applied Ecology 51:795–803. Pallister J., Voysey R., Boyle D., Halliday D., Hyatt A., Letnic M., Webb J. K., Jessop T. S., Dempster T. 2015. Shanmuganathan T., Venables D., Robinson T. Restricting access to invasion hubs enables sus- 2008. The immune approach to cane toad control. tained control of an invasive vertebrate. Journal of Page 131 in 14th Australasian Vertebrate Pest Confer- Applied Ecology 52:341–347. ence. Bruce (Australia): Invasive Animals Coopera- Lever C. 2001. The Cane Toad: The History and Ecology of tive Research Centre. Available at http://pestsmart a Successful Colonist. Otley (United Kingdom): West- .org.au/wp-content/uploads/2010/03/14thVPCPro bury Academic and Scientific Publishing. ceedings2008.pdf. Linzey D., Burroughs J., Hudson L., Marini M., Robert- Pallister J. A., Halliday D. C. T., Robinson A. J., son J., Bacon J., Nagarkatti M., Nagarkatti P. 2003. Venables D., Voysey R. D., Boyle D. G., Shanmuga- Role of environmental pollutants on immune func- nathan T., Hardy C. M., Siddon N. A., Hyatt A. D. tions, parasitic infections and limb malformations 2011. Assessment of virally vectored autoimmunity in marine toads and whistling frogs from Bermuda. as biocontrol strategy for cane toads. PLOS ONE 6: International Journal of Environmental Health Research e14576. 13:125–148. Peacock D. 2006. Investigation of a pathogen for Bufo Llewelyn J., Schwarzkopf L., Phillips B. L., Shine R. marinus in northern Argentina: could it be a try- 2014. After the crash: how do predators adjust fol- panosome? Pages 98–105 in Science of Cane Toad In- lowing the invasion of a novel toxic prey type? Aus- vasion and Control: Proceedings of the Invasive Animals tral Ecology 39:190–197. CRC/CSIRO/Qld NRM&W Cane Toad Workshop, ed- Madsen T., Shine R., Olsson M., Wittzell H. 1999. Con- ited by K. Molloy and W. Henderson. Canberra servation biology: restoration of an inbred adder (Australia): Invasive Animals Cooperative Research population. Nature 402:34–35. Centre. Available at http://www.pestsmart.org.au Mahony M., Clulow J. 2006. Control of cane toads by /wp-content/uploads/2010/03/CaneToadProc.pdf. sterile male release and inherited sterility. Pages Phillips B. L., Shine R. 2004. Adapting to an invasive 134–150 in Science of Cane Toad Invasion and Control: species: toxic cane toads induce morphological Proceedings of the Invasive Animals CRC/CSIRO/Qld change in Australian snakes. Proceedings of the Na- June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 147

tional Academy of Sciences of the United States of Amer- aversion in an imperilled native reptile predator. ica 101:17150–17155. Animal Conservation 16:386–394. Phillips B. L., Shine R. 2006. An invasive species induces ReesH.C.,BishopK.,MiddleditchD.J.,PatmoreJ.R.M., rapid adaptive change in a native predator: cane Maddison B. C., Gough K. C. 2014. The application toads and black snakes in Australia. Proceedings of of eDNA for monitoring of the Great Crested Newt the Royal Society B: Biological Sciences 273:1545–1550. in the UK. Ecology and Evolution 4:4023–4032. Phillips B. L., Brown G. P., Shine R. 2003. Assessing Reidinger R. F. Jr., Miller J. E. 2013. Wildlife Damage the potential impact of cane toads on Australian Management: Prevention, Problem Solving, and Conflict snakes. Conservation Biology 17:1738–1747. Resolution. Baltimore (Maryland): Johns Hopkins Phillips B. L., Brown G. P., Webb J. K., Shine R. 2006. University Press. Invasion and the evolution of speed in toads. Na- Rhoads A., Au K. F. 2015. PacBio sequencing and its ture 439:803. applications. Genomics, Proteomics and Bioinformatics Phillips B. L., Brown G. P., Travis J. M. J., Shine R. 13:278–289. 2008. Reid’s paradox revisited: the evolution of dis- Robinson T., Siddon N., Tarmo S., Halliday D., persal kernels during range expansion. American Shanmuganathan T., Venables D. 2006. CSIRO bio- Naturalist 172:S34–48. control project: concept and progress. Pages 86–88 PhillipsB.L.,BrownG.P.,ShineR.2010a.Evolutionarily in Science of Cane Toad Invasion and Control: Proceedings accelerated invasions: the rate of dispersal evolves of the Invasive Animals CRC/CSIRO/Qld NRM&W upwards during the range advance of cane toads. Cane Toad Workshop, edited by K. Molloy and Journal of Evolutionary Biology 23:2595–2601. W. Henderson. Canberra (Australia): Invasive Ani- Phillips B. L., Kelehear C., Pizzatto L., Brown G. P., mals Cooperative Research Centre. Available at Barton D., Shine R. 2010b. Parasites and patho- http://www.pestsmart.org.au/wp-content/uploads gens lag behind their host during periods of host /2010/03/CaneToadProc.pdf. range advance. Ecology 91:872–881. Rollins L. A., Richardson M. F., Shine R. 2015. A Phillips B. L., Shine R., Tingley R. 2016. The genetic genetic perspective on rapid evolution in cane backburn: using rapid evolution to halt invasions. toads (Rhinella marina). Molecular Ecology 24:2264– Proceedings of the Royal Society B: Biological Sciences 2276. 283:20153037. Saunders G., Cooke B., McColl K., Shine R., Peacock Pilliod D. S., Goldberg C. S., Arkle R. S., Waits L. P. T. 2010. Modern approaches for the biological 2013. Estimating occupancy and abundance of control of vertebrate pests: an Australian perspec- stream amphibians using environmental DNA from tive. Biological Control 52:288–295. filtered water samples. Canadian Journal of Fisheries Schwarzkopf L., Alford R. A. 1996. Desiccation and and Aquatic Sciences 70:1123–1130. shelter-site use in a tropical amphibian: comparing Pizzatto L., Shine R. 2011. The effects of experimentally toads with physical models. Functional Ecology 10: infecting Australian tree frogs with lungworms (Rhab- 193–200. dias pseudosphaerocephala) from invasive cane toads. Schwarzkopf L., Alford R. A. 2002. Nomadic movement International Journal for Parasitology 41:943–949. in tropical toads. Oikos 96:492–506. Pizzatto L., Shine R. 2012. Typhoid Mary in the frog- Schwarzkopf L., Alford R. A. 2007. Acoustic attractants pond: can we use native frogs to disseminate a lung- enhance trapping success for cane toads. Wildlife Re- worm biocontrol for invasive cane toads? Animal search 34:366–370. Conservation 15:545–552. Semeniuk M., Lemckert F., Shine R. 2007. Breeding- Pizzatto L., Child T., Shine R. 2008. Why be diurnal? site selection by cane toads (Bufo marinus) and na- Shifts in activity time enable young cane toads to tive frogs in northern New South Wales, Australia. evade cannibalistic conspecifics. Behavioral Ecology Wildlife Research 34:59–66. 19:990–997. Shanmuganathan T., Pallister J., Doody S., McCallum Pizzatto L., Shilton C. M., Shine R. 2010. Infection dy- H., Robinson T., Sheppard A., Hardy C., Halliday namics of the lungworm Rhabdias pseudosphaeroce- D., Venables D., Voysey R., Strive T., Hinds L., Hyatt phala in its natural host, the cane toad (Bufo mari- A. 2010. Biological control of the cane toad in Aus- nus), and in novel hosts (native Australian frogs). tralia: a review. Animal Conservation 13:16–23. Journal of Wildlife Diseases 46:1152–1164. Shannon M. F., Bayliss P. 2008. Review of the CSIRO bi- Price-Rees S. J., Webb J. K., Shine R. 2011. School for ological control of cane toad program to April 2008. skinks: can conditioned taste aversion enable blue- Canberra (Australia): Department of the Environ- tongue lizards (Tiliqua scincoides)toavoidtoxiccane ment, Water, Heritage and the Arts, Commonwealth toads (Rhinella marina) as prey? Ethology 117:749– of Australia. 757. Shigeta M., Sakane Y., Iida M., Suzuki M., Kashiwagi Price-Rees S. J., Webb J. K., Shine R. 2013. Reducing K., Kashiwagi A., Fujii S., Yamamoto T., Suzuki K. T. the impact of a toxic invader by inducing taste 2016. Rapid and efficient analysis of gene function 148 THE QUARTERLY REVIEW OF BIOLOGY Volume 92

using CRISPR-Cas9 in Xenopus tropicalis founders. E. 2012. Monitoring endangered freshwater biodi- Genes to Cells 21:755–771. versity using environmental DNA. Molecular Ecology Shine R. 2010. The ecological impact of invasive cane 21:2565–2573. toads (Bufo marinus) in Australia. Quarterly Review of Thresher R. E., Bax N. 2006. Comparative analysis of Biology 85:253–291. genetic options for controlling invasive populations Shine R. 2011. How can we ensure that conservation of the cane toad, Bufo marinus. Pages 117–122 in Sci- policies are based on science, not emotion? Pacific ence of Cane Toad Invasion and Control: Proceedings of Conservation Biology 17:6–10. the Invasive Animals CRC/CSIRO/Qld NRM&W Cane Shine R., Amiel J., Munn A. J., Stewart M., Vyssotski A. L., Toad Workshop, edited by K. Molloy and W. Hender- Lesku J. A. 2015. Is “cooling then freezing” a hu- son. Canberra (Australia): Invasive Animals Coop- mane way to kill amphibians and reptiles? Biology erative Research Centre. Available at http://www Open doi:10.1242/bio.012179. .pestsmart.org.au/wp-content/uploads/2010/03 Simberloff D. 2003. How much information on popu- /CaneToadProc.pdf. lation biology is needed to manage introduced spe- Tingley R., Shine R. 2011. Desiccation risk drives the cies? Conservation Biology 17:83–92. spatial ecology of an invasive anuran (Rhinella marina) Simberloff D. 2012. Risks of biological control for con- in the Australian semi-desert. PLOS ONE 6:e25979. servation purposes. BioControl 57:263–276. Tingley R., Phillips B. L., Letnic M., Brown G. P., Shine Smart A. S., Tingley R., Weeks A. R., van Rooyen A. R., R., Baird S. J. E. 2013. Identifying optimal barriers McCarthy M. A. 2015. Environmental DNA sampling to halt the invasion of cane toads Rhinella marina in is more sensitive than a traditional survey technique arid Australia. Journal of Applied Ecology 50:129–137. for detecting an aquatic invader. Ecological Applica- Tingley R., Vallinoto M., Sequeira F., Kearney M. R. tions 25:1944–1952. 2014. Realized niche shift during a global biolog- Smith M. A., Green D. M. 2005. Dispersal and the ical invasion. Proceedings of the National Academy of metapopulation in amphibian ecology and conser- Sciences of the United States of America 111:10233– vation: are all amphibian metapopulations? Ecogra- 10238. phy 28:110–128. Torchin M. E., Lafferty K. D., Dobson A. P., McKenzie Somaweera R., Shine R. 2012. The (non) impact of in- V. J., Kuris A. M. 2003. Introduced species and vasive cane toads on freshwater crocodiles at Lake their missing parasites. Nature 421:628–630. Argyle in tropical Australia. Animal Conservation 15: Towsey M., Planitz B., Nantes A., Wimmer J., Roe P. 152–163. 2012. A toolbox for animal call recognition. Bio- Somaweera R., Somaweera N., Shine R. 2010. Frogs acoustics 21:107–125. under friendly fire: how accurately can the general Trumbo D. R., Epstein B., Hohenlohe P. A., Alford public recognize invasive species? Biological Conser- R. A., Schwarzkopf L., Storfer A. 2016. Mixed pop- vation 143:1477–1484. ulation genomics support for the central marginal Somaweera R., Webb J. K., Brown G. P., Shine R. 2011. hypothesis across the invasive range of the cane Hatchling Australian freshwater crocodiles rapidly toad (Rhinella marina) in Australia. Molecular Ecology learn to avoid toxic invasive cane toads. Behaviour 25:4161–4176. 148:501–517. Turvey N. 2013. Cane Toads: A Tale of Sugar, Politics and Southwell D., Tingley R., Bode M., Nicholson E., Phillips Flawed Science. Sydney (Australia): Sydney University B. L. 2017. Cost and feasibility of a barrier to halt the Press. spread of invasive cane toads in arid Australia: incor- Urban M. C., Phillips B. L., Skelly D. K., Shine R. 2008. porating expert knowledge into model-based deci- A toad more traveled: the heterogeneous invasion sion-making. Journal of Applied Ecology 54:216–224. dynamics of cane toads in Australia. American Nat- Speare R. 1990. A review of the diseases of the cane uralist 171:E134–E148. toad, Bufo marinus, with comments on biological Ward-Fear G., Brown G. P., Greenlees M. J., Shine R. control. Wildlife Research 17:387–410. 2009. Maladaptive traits in invasive species: in Austra- Speare R., Berger L., O’Shea P., Ladds P. W., Thomas lia, cane toads are more vulnerable to predatory ants A. D. 1997. Pathology of mucormycosis of cane toads than are native frogs. Functional Ecology 23:559–568. in Australia. Journal of Wildlife Diseases 33:105–111. Ward-Fear G., Brown G. P., Shine R. 2010a. Using a Taylor A., Watson G., Grigg G., McCallum H. 1996. native predator (the meat ant, Iridomyrmex reburrus) Monitoring frog communities: an application of to reduce the abundance of an invasive species (the machine learning. Pages 1564–1569 in Proceedings cane toad, Bufo marinus) in tropical Australia. Jour- of the Eighth Annual Conference on Innovative Applica- nal of Applied Ecology 47:273–280. tions of Artificial Intelligence, edited by H. Shrobe and Ward-Fear G., Brown G. P., Shine R. 2010b. Factors af- T. Senator. Palo Alto (California): AAAI Press. fecting the vulnerability of cane toads (Bufo ma- Thomsen P. F., Kielgast J., Iversen L. L., Wiuf C., Ras- rinus) to predation by ants. Biological Journal of the mussen M., Gilbert M. T. P., Orlando L., Willerslev Linnean Society 99:738–751. June 2017 NEW WEAPONS IN THE TOAD TOOLKIT 149

Ward-Fear G., Pearson D. J., Brown G. P., Rangers B., Mosman (Australia): Royal Zoological Society of Shine R. 2016. Ecological immunization: in situ New South Wales. training of free-ranging predatory lizards reduces White A. W., Shine R. 2009. The extra-limital spread of their vulnerability to invasive toxic prey. Biology Let- an invasive species via “stowaway” dispersal: toad to ters 12:20150863. nowhere? Animal Conservation 12:38–45. Webb J. K., Brown G. P., Child T., Greenlees M. J., Wingate D. B. 2011. The successful elimination of cane Phillips B. L., Shine R. 2008. A native dasyurid toads, Bufo marinus, from an island with breeding predator (common planigale, Planigale maculata) habitat off Bermuda. Biological Invasions 13:1487– rapidly learns to avoid a toxic invader. Austral Ecol- 1492. ogy 33:821–829. Yasumiba K., Alford R. A., Schwarzkopf L. 2015. Why Webb J. K., Pearson D., Shine R. 2011. A small dasyurid do male and female cane toads, Rhinella marina, re- predator (Sminthopsis virginiae) rapidly learns to spond differently to advertisement calls? Animal Be- avoid a toxic invader. Wildlife Research 38:726–731. haviour 109:141–147. WebbJ.K.,LetnicM.,JessopT.S.,DempsterT.2014. Yeager A., Commito J., Wilson A., Bower D., Behavioural ﬂexibility allows an invasive vertebrate to Schwarzkopf L. 2014. Sex, light, and sound: loca- survive in a semi-arid environment. Biology Letters 10: tion and combination of multiple attractants affect 20131014. probability of cane toad (Rhinella marina) capture. Webber B. L., Raghu S., Edwards O. R. 2015. Opinion: Journal of Pest Science 87:323–329. Is CRISPR-based gene drive a biocontrol silver bul- Zug G. R., Zug P. B. 1979. The marine toad, Bufo let or global conservation threat? Proceedings of the marinus: a natural history resumé of native popula- National Academy of Sciences of the United States of tions. Smithsonian Contributions to Zoology 284:1–58. America 112:10565–10567. White A. W. 2007. Living with Bufo. Pages 16–29 in Pest or Guest: The Zoology of Overabundance, edited by Associate Editor: Andrew Sih D. Lunney, P. Eby, P. Hutchings, and S. Burgin. Handling Editor: John J. Wiens

View publication stats