Conditioned Taste Aversion and Northern Quoll Survival

Source Details

  • Authors: Stephanie O’Donnell, Jonathan K. Webb, Richard Shine

  • Article: “Conditioned taste aversion enhances the survival of an endangered predator imperilled by a toxic invader”

  • Journal: Journal of Applied Ecology, Vol. 47, No. 3 (June 2010), pp. 558-565

  • DOI: 10.1111\/j.1365\text{-}2664.2010.01802.x

  • Publisher: British Ecological Society

  • Geographic focus: Northern Australia (including Kakadu National Park and Kimberley region)

Global & Australian Context

  • Human-mediated movement of species is a leading driver of biodiversity loss (Williamson 1996).

  • Australia: suffers \tfrac{1}{2} of all global mammal extinctions in the last 2 centuries.

  • Invasive predators (cats Felis catus, foxes Vulpes vulpes) traditionally targeted via eradication/exclusion, but large-scale, long-term control is costly and often infeasible.

  • Cane toad (Bufo marinus = Chaunus marinus) introduced in 1935; now occupies 60\% of northern quoll range and projected to fill the entire range within \sim 20 years (Urban et al. 2007).

Focal Species & Threat

Northern quoll (Dasyurus hallucatus)

  • Size: cat-sized, nocturnal, omnivorous/carnivorous marsupial.

  • Life history: mature at 11 months, single intense breeding season per year; females can live multiple years but high post-reproduction mortality; males often semelparous.

  • Shelter: tree hollows, logs, rock crevices, termite mounds, dense grass, burrows.

  • Conservation status: listed “Endangered” (EPBC 1999) after rapid declines post-toad invasion (Watson & Woinarski 2004).

Cane toad impact mechanism

  • Carry bufotoxins unlike native frogs; predators die rapidly after oral contact.

  • Documented crashes in varanid lizards, freshwater crocodiles, and northern quolls.

Management Innovation: Conditioned Taste Aversion (CTA)

  • Definition: association of a novel food taste with post-ingestion malaise → long-lasting avoidance (Garcia et al. 1974).

  • Strengths: single exposure suffices; most powerful with novel prey.

  • Prior conservation uses: reducing coyote/wolf attacks on lambs, crow egg-raiding, deterring nuisance bears.

Study Objectives & Hypotheses

  1. Can CTA be induced in captive-reared juvenile quolls by feeding a toxin-laced dead toad?

  2. Does CTA increase short-term post-release survival in toad-infested habitats?

  3. Are survival effects sex-specific given male boldness?

Predictions tested in Program MARK models:

  • P_1 (toad-smart > toad-naive overall),

  • P_2 (treatment effect within each sex),

  • P_3 (effect stronger in males),

  • P_4 (females survive better regardless of treatment).

Methods

Captive CTA Protocol

  • Subjects: N=62 juvenile quolls (31 toad-smart, 31 toad-naive).

    • Sex ratios: toad-smart 15 ♂ + 16 ♀; toad-naive 17 ♂ + 14 ♀.

  • Housing: individual outdoor enclosures 6 \times 4.5 \times 2\,\text{m} at Territory Wildlife Park (TWP), NT.

  • Aversion induction: one non-lethal minced dead toad piece <2\,\text{g} mixed with thiabendazole 400\,\text{mg kg}^{-1} body mass.

  • Behavioural test (1–2 days later): exposure to live <2\,\text{g} toad in container; responses scored “attack”, “reject”, “ignore” on hidden video.

Field Release & Monitoring

  • Five sites (GPS UTM WGS-84): TWP, Lake Bennett, Beatrice Hill Farm, Fogg Dam, Mary River Park.

  • Release window: Nov 2008 – Feb 2009 at dusk 17{:}00$–$19{:}00\,h.

  • Radiocollars <5\% body mass, mortality sensor; topical anti-parasite (Frontline).

  • Tracking: continuous night observation \approx 1\,h, GPS fixes thereafter day/night; helicopter sweeps at 30–42 days.

  • Minimum monitoring: 10 days per individual (mean 7 days, range 0–24).

  • Trapping (Elliott & day-old chick baits) near dens allowed collar recovery and live resight.

Statistical Analyses

  • Feeding response: multinomial logistic regression \chi^2{2}=11.300,\ P<0.004 for treatment; \chi^2{2}=11.334,\ P<0.004 for sex.

  • Toad encounter rate post-release: two-factor ANOVA; treatment F{1,17}=0.79,\ P=0.38; sex F{1,17}=1.43,\ P=0.25; interaction NS.

  • Survival estimation: Burnham’s live-encounter\/dead-recovery within Program MARK.

    • Parameters: S (survival), F (site fidelity), p (recapture\/resight), r (dead recovery).

    • Goodness of fit: \hat{c}=1.07 from 1000 bootstraps \Rightarrow minimal over-dispersion; QAICc used.

Key Results

CTA Alters Behaviour

  • Only males attacked toads in pre-release tests: 4 of 10 toad-naive ♂ bit mesh-caged toads; 1 toad-smart ♂ bit then spat out.

  • All females either ignored or approached then rejected.

Post-Release Toad Mortality

  • Confirmed deaths n=8; 7 due to toad poisoning, 1 by feral cat.

  • 5\,/\,7 poisoned were toad-naive males.

Survival Statistics

  • Mean time alive (days):

    • Toad-naive ♂ =2.17 (range 0–14) vs toad-smart ♂ =12.3 (1–31).

    • Toad-naive ♀ =5.42 (0–20) vs toad-smart ♀ =10.17 (0–20).

  • Model hierarchy (∆QAICc): best model S(group)\,p(group)\,r(const)\,F(const).

  • Model-averaged daily apparent survival:
    S{♀,\,smart}=0.94\,(SE=0.02) S{♀,\,naive}=0.84\,(SE=0.04)
    S{♂,\,smart}=0.88\,(SE=0.04) S{♂,\,naive}=0.58\,(SE=0.09)

  • Toad-smart survival higher in both sexes; effect size larger in males.

Discussion & Significance

  1. Proof-of-concept: single thiabendazole-laced toad carcass induces robust aversion with tangible survival benefit.

  2. Remaining caveats:

    • CTA efficacy <100\% (2 trained ♂ still attacked large toads) — dosage or stimulus generalisation needs refinement.

    • Thiabendazole’s anthelmintic side-effects may confound long-term survival comparisons.

  3. Sex differences: male boldness ⇒ higher toad-induced mortality; yet population-level risk mitigated as few males needed for fertilisation.

  4. Management translation:

    • Aerial or trap-based deployment of “toad sausages” ahead of invasion front could pre-arm wild quolls.

    • CTA complements but does not replace fire management, predator control, and habitat quality measures.

  5. Wider applications: CTA could buffer endangered prey from non-eradicable invasive predators by exploiting predator naïveté and prey novelty.

Practical & Ethical Considerations

  • Approval: University of Sydney ACEC; NT Parks & Wildlife permits.

  • Animal welfare: collars designed to drop within \approx1 month; non-lethal toad pieces; anti-parasite treatment provided.

  • Caution: need to ensure non-target species not harmed by CTA baits; evaluate landscape-scale logistics and cost-benefit.

Key Numerical Summary

  • Total quolls: N=62 (31 smart + 31 naive).

  • Mortality by toad poisoning: n=7 confirmed (( \geq ) all male except 2).

  • Apparent daily survival benefit of CTA:
    \Delta S{♀}=0.10; \Delta S{♂}=0.30.

  • Logistic regression treatment effect: \chi^2=11.300,\ df=2,\ P<0.004.

  • Best MARK model weight w_{i}=0.5824.

Connections & Theoretical Foundations

  • Reinforces Garcia effect (taste-malaise) and predator foraging theory (risk vs reward).

  • Aligns with “behavioural rescue” concept: rapid learning can buffer populations before genetic adaptation.

  • Echoes previous successful CTA in coyotes, crows, mongooses, bears; extends to marsupial predators.

Future Directions

  • Optimise bait size, toxin dose, and pre-exposure interval.

  • Longitudinal monitoring >6 months to estimate true survival St and reproductive success R0.

  • Test landscape-scale aerial bait deployment; model cost-effectiveness.

  • Examine non-target uptake and community-level consequences.

  • Explore CTA for other predator–prey invasive interactions (e.g., cats vs island birds).

Source Details

  • Authors: Stephanie O’Donnell, Jonathan K. Webb, Richard Shine

  • Article: “Conditioned taste aversion enhances the survival of an endangered predator imperilled by a toxic invader”

  • Journal: Journal of Applied Ecology, Vol. 47, No. 3 (June 2010), pp. 558-565

  • DOI: 10.1111\/j.1365\text{-}2664.2010.01802.x

  • Publisher: British Ecological Society

  • Geographic focus: Northern Australia (including Kakadu National Park and Kimberley region)

Global & Australian Context

  • Human-mediated movement of species, especially through introduced or invasive alien species, is recognized as a profound and leading driver of biodiversity loss globally (Williamson 1996), impacting ecosystem structure and function.

  • Australia: Tragically, Australia accounts for approximately frac{1}{2} of all global mammal extinctions that have occurred in the last 2 centuries, a disproportionately high figure largely attributable to the impacts of invasive species.

  • Invasive predators (European red foxes Vulpes vulpes and feral cats Felis catus) have traditionally been targeted via broad-scale eradication or exclusion fencing strategies, but large-scale, long-term control of these highly mobile and adaptable species is often prohibitively costly and frequently infeasible for effective population management.

  • Cane toad (Bufo marinus = Chaunus marinus): These highly toxic amphibians were deliberately introduced to Queensland in 1935 from Hawaii, with the misguided aim of biological control for cane beetles. They have since spread extensively, now occupying an estimated 60\% of the northern quoll’s historical range and are projected to colonize the entire habitat within approximately \sim 20 years (Urban et al. 2007) if their current expansion rates continue.

Focal Species & Threat
Northern quoll (Dasyurus hallucatus)

  • Size: A medium-sized, cat-sized (typically 0.3-1.2\,\text{kg} for males and 0.2-0.8\,\text{kg} for females), nocturnal marsupial, which is highly agile and primarily arboreal.

  • Life history: Northern quolls mature rapidly at around 11 months of age, engaging in a single, highly intense breeding season per year during the dry season (June-July). While females are polyestrous and can potentially live for multiple years, there is high post-reproduction mortality, particularly in males, which often exhibit semelparous reproduction (breeding once and then dying due to physiological stress and exhaustion).

  • Shelter: They utilize a variety of crucial daytime refuges including tree hollows, fallen logs, rock crevices and caves, disused termite mounds, dense grass tussocks, and burrows (often abandoned ones of other animals).

  • Conservation status: The species is currently listed as “Endangered” under the Australian Environment Protection and Biodiversity Conservation Act (EPBC 1999) following documented rapid and severe population declines (up to 80\% in some areas) immediately post-cane toad invasion (Watson & Woinarski 2004).

Cane toad impact mechanism

  • Bufotoxins: Unlike native Australian frogs and toads, cane toads possess potent cardiotoxic and neurotoxic bufotoxins within their parotoid glands and skin secretions. These toxins are rapidly absorbed through mucous membranes (mouth, eyes) upon oral contact or ingestion.

  • Lethality: Predators that attempt to consume or mouth cane toads die rapidly, typically within minutes to hours, due to cardiac arrest or neurological impairment caused by the bufotoxins.

  • Documented crashes: The swift and devastating impact of cane toads has been well-documented, leading to severe population crashes in numerous native Australian predators, including varanid lizards (goannas), freshwater crocodiles (Crocodylus johnstoni), and critically, northern quolls.

Management Innovation: Conditioned Taste Aversion (CTA)

  • Definition: Conditioned Taste Aversion (CTA) is a powerful form of classical conditioning involving the association of a novel food taste or prey item with a subsequent experience of gastrointestinal malaise (nausea or illness), which then leads to a rapid, profound, and long-lasting avoidance of that specific food or prey (Garcia et al. 1974). This learning is often attributed to specialized neural pathways evolved for detecting and avoiding toxins.

  • Strengths: A key strength of CTA is its efficacy after only a single exposure to the noxious stimulus; repeated exposures are generally not necessary for the aversion to form and persist. The effect is most powerful and readily induced when the target prey is novel to the predator, as was the case with cane toads for native Australian predators.

  • Prior conservation uses: CTA has been successfully applied in a variety of conservation and management contexts globally, including reducing coyote and wolf attacks on livestock (e.g., lambs), deterring crow egg-raiding behavior, and mitigating nuisance behaviors in large carnivores like bears.

Study Objectives & Hypotheses

This study aimed to evaluate the potential of CTA as a conservation tool for northern quolls:

  1. Induction of CTA: Can a robust and lasting conditioned taste aversion be successfully induced in captive-reared juvenile northern quolls by feeding them a non-lethal, toxin-laced piece of dead cane toad?

  2. Short-term Survival: Does the induction of CTA significantly increase the short-term post-release survival rates of these quolls when reintroduced into their natural habitats that are already infested with cane toads?

  3. Sex-specific Effects: Are the survival effects of CTA sex-specific, given the documented differences in boldness and foraging behavior between male and female quolls?

Predictions tested through the analysis of capture-recapture and recovery data using Program MARK models:

  • P_1: Quolls assigned to the 'toad-smart' treatment group are hypothesized to exhibit generally higher overall survival rates compared to 'toad-naive' individuals across the study population.

  • P_2: A positive treatment effect (higher survival for toad-smart individuals) is predicted to be observed within each sex, indicating the general efficacy of CTA.

  • P_3: The survival benefit derived from the CTA treatment is hypothesized to be stronger and more pronounced in male quolls, who typically face a higher risk of toad-induced mortality due to their greater exploratory behavior and tendency to attack toads.

  • P_4: Regardless of the treatment group, female quolls are expected to survive better than males due to their generally less risky foraging strategies and lower inherent post-breeding mortality rates compared to semelparous males.

Methods
Captive CTA Protocol

  • Subjects: A total of N=62 juvenile northern quolls were used in the study, carefully divided into two experimental groups: 31 individuals designated as 'toad-smart' (receiving the CTA treatment) and 31 individuals as 'toad-naive' (control group).

  • Sex ratios: The 'toad-smart' group comprised 15 males (♂) and 16 females (♀), while the 'toad-naive' group consisted of 17 males (♂) and 14 females (♀), ensuring a balanced sex distribution across treatments.

  • Housing: All quolls were housed individually in spacious outdoor enclosures, each measuring 6 \times 4.5 \times 2\,\text{m} (\text{length} \times \text{width} \times \text{height}) at the Territory Wildlife Park (TWP) in the Northern Territory. These enclosures provided a naturalistic environment while allowing for controlled experimental conditions.

  • Aversion induction: For the 'toad-smart' group, CTA was induced by providing a single, small non-lethal minced piece of dead cane toad (weighing less than 2\,\text{g}) mixed with thiabendazole at a dose of 400\,\text{mg kg}^{-1} body mass. Thiabendazole is an anthelmintic agent that induces temporary gastrointestinal malaise without causing lasting harm, simulating the toxic effects of ingesting a live toad.

  • Behavioural test (1–2 days later): To assess the efficacy of the CTA, each 'toad-smart' quoll, along with all 'toad-naive' controls, was individually exposed to a live, small (typically less than 2\,\text{g}) cane toad placed within a secure, mesh-caged container. Their behavioral responses—categorized as