Benefits and costs of using a lungworm parasite to control invasive toads in Australia



Reid N. Harris, Department of Biology, James Madison University, Harrisonburg, VA 22807, USA. Tel: +1 540 568 6665



Read the Feature Paper: Typhoid Mary in the frogpond: can we use native frogs to disseminate a lungworm biocontrol for invasive cane toads?

Other Commentaries on this paper:Is Typhoid Mary a good role model?

Response from the authors: New methods in the battle against cane toads: when should we move from research to implementation?

Pizzatto & Shine (2012) noted the serious impacts that the introduced cane toad Rhinella marina has caused to its predators in Australia. The toad's range is expanding at an increasing rate and additional negative consequences are predicted. Control measures have been ineffective or in some cases have not been tried. Female toads have a very high fecundity, and the species has a high potential growth rate. High removal rates of the toad population are needed to eliminate the species, suggesting that physical removal is not practical (Shine & Doody, 2010). One approach to controlling the toad species is biological control, which has been effective in some cases in Australia and elsewhere (Saunders et al., 2010).

Biological control usually involves introduction of a species that in its native range controls the population of the pest species. However, introducing an exotic species as a biocontrol agent can have negative effects on nontarget species and potentially on ecosystem function (Simberloff & Stiling, 1996). Use of a native species as a biological control agent eases these concerns, but may involve a protocol to increase the abundance and distribution of a native control agent. In the case of the cane toad, one goal of biocontrol can be slowing the rate of expansion. Individuals at the leading edge of expansion have greater locomotor endurance, and the authors suggest that this may be due to escape from a lungworm parasite that infects toads behind the leading edge. Therefore, the authors test aspects of a general hypothesis that posits that lungworm parasites can be used as a biocontrol agent to slow the toads' advance. This may not be an entirely satisfactory solution, but slowing the toads' advance might buy time to develop other ways to control the species' numbers. As the lungworm is already present in the toads and likely was present in the toads when they were introduced to Australia in 1935, it is not considered an exotic species.

The authors suggest that one way to increase the prevalence of lungworm infection is to infect a native tree frog species, Litoria caerulea, which does not seem to be negatively affected by the parasite. They also propose three primary assumptions necessary for biocontrol to work in this case: the parasite does not harm the tree frogs, the tree frogs can maintain an infection over the long term, and infective larvae of the parasite can be transmitted to cane toads and reduce viability of the toads. They test aspects of these assumptions in their paper.

Pizzatto and Shine showed that the tree frogs were infected at a 50% rate at the end of the experiment, and this was after placing the lungworm in the mouths of the frogs. All parasite-exposed frogs survived, and their growth rate was not different from that of uninfected controls. It would be important to compare the growth of infected frogs and uninfected frogs as the ‘parasite-exposed’ treatment apparently contained frogs that were not infected or had cleared the infection. Of those frogs that were infected, they had on average a low intensity (1.67 parasites per frog). It is unclear that this intensity would have the desired effect of transmission of larvae to toads.

To test for transmission to the toads, lungworm larvae were put in boxes housing five toads. About half of the toads were infected under these laboratory conditions. In terms of fitness effects on the toads, there was no survival difference between infected and uninfected, no difference in body mass and no difference in sprint speeds. The authors did find that toads exposed to lungworms had lower stamina than did controls.

Laboratory conditions are likely to maximize transmission of parasites as toads were under high density conditions, and their movements were constrained. However, under these conditions, parasite prevalence in frogs and toads averaged 50%. Importantly, only minor fitness effects were seen in the toads. Based on these results, it seems doubtful that control would be effective in nature.

There are other reasons to be skeptical that biocontrol would be effective in this case. Is the lungworm a factor in population control in other parts of the toads' range where prevalence is high? If not, why would it be expected to work at the leading edge? The relationship of stamina estimated in the laboratory may or may not be related to rate of advance of toads in nature. In addition, the laboratory experiments did not place the tree frogs and toads together under ecological relevant conditions that would provide a realistic test of parasite transmission. It seems possible that the tree frog species and the toad species do not co-occur in the same microhabitats as the tree frog species is more arboreal. In any case, transmission from the tree frogs' feces to the toads has not been demonstrated. To sum up, transmission under laboratory conditions achieved modest results in terms of long-term prevalence and intensity, infection did not affect survival, and transmission under natural conditions is less than certain. Given the tremendous fecundity of the toads and the role of compensatory density dependence, it is far from certain that biocontrol would have much of an effect in this case.

As always with a biocontrol protocol, effects on nontarget species and on ecosystem processes need to be considered. The authors note that another tree frog species, Litoria splendida, is affected by the lungworm. Thus, spill over of the parasite is a serious concern. It is also possible that L. caerulea may be negatively affected under relevant ecological conditions or that its health is negatively affected in ways that are not captured by survival and growth rate. All potential host species would need to be tested for their response to the lungworm parasite under ecological relevant conditions. Other effects are harder to predict. Perhaps, the predators of infected tree frogs are negatively affected by a bacterial symbiont of the lungworm.

The invasive cane toad causes damage to native predators in Australia so benefits and costs of biocontrol or any environmental manipulation need to be considered (Minteer & Collins, 2008). The lungworm as a biocontrol agent appears to provide little benefit, while the costs have not been fully evaluated. As such, additional research is needed before implementation is considered.