Cost-effective management of invasive species: An application of info-gap decision theory

Liu, Yang (2022) Cost-effective management of invasive species: An application of info-gap decision theory. PhD thesis, Murdoch University.

Abstract

Increasing international trade and tourist activities raise the likelihood that invasive species will be introduced and result in damage to the economy, environment and society. There are primarily three biosecurity activities for managing invasive species; namely pre-border and border quarantine (quarantine, hereafter) to prevent incursion, post-border surveillance (surveillance, hereafter) to search for newly introduced species and, post-border eradication (eradication, hereafter) to remove the last individual of the invasive species. Preventing the introduction of invasive species is an effective and economical management approach. Intensive surveillance is necessary to prevent newly introduced populations from establishing and spreading, particularly in high-risk areas. Once the invasive species becomes established, eradication can be difficult and resource intensive. Appropriate budget allocation for future incursions, is important to ensure early detection and a rapid response in order to successfully eradicate the organism. When eradication is not feasible, containment and asset-based protection could be implemented. These management options are not considered in this research.

When confronted with biological invasions, decision-makers need to react quickly, however, they are often hindered by sparse and/or contradictory information. This is because ecological systems, and the species they contain, are variable and highly complex, resulting in severe uncertainty in ecological model parameters. The existence of severe uncertainty can lead to flawed and inefficient decisions. Within the limited time prior to an incursion response, it is usually impractical for decision makers to collect sufficient information and put forward more informed responses. Info-gap decision theory can be applied to model and manage such severe uncertainty.

Info-gap decision theory (IGDT) is developed for decision-making under severe uncertainty. Either the decision maker is risk seeking or risk averse, the uncertainties are thus favourable or detrimental respectively. While opportuneness searches for possibilities of extra gains for the optimistic decision makers, robustness can guarantee return for conservative decision makers, with minimum requirements always satisfied. The confidence in realizing desired goals under severe uncertainty is quantified and evaluated using robustness. Info-gap decision theory has been widely applied in ecosystem management. This research extends the application of IGDT by determining the estimated population threshold for surveillance detection and guiding robust budget allocation to biosecurity activities.

This research models the cost-effective management of a potential incursion of the Asian house gecko (Hemidactylus frenatus) (AHG) Duméril & Bibron, 1836 onto Barrow Island (BWI), Australia using info-gap decision theory. The AHG is an excellent hitchhiker and one of the most widespread reptiles world-wide. In its introduced range, this species has caused the extinction of several gecko species and is believed to carry novel parasites. Barrow Island is one of the most significant conservation reserves in Western Australia. The operation of industries on the island may increase the possibility of invasive species incursions. A strict quarantine management system (QMS) has been implemented on BWI to protect the valuable biodiversity of the island. AHG is one of the most frequently detected vertebrate invasive species at the border and was detected and eradicated on BWI in March 2015. BWI is currently free of the AHG (Chevron Corporation, 2021). Due to its widespread presence on the mainland of Australia and many overseas sites, its known invasiveness and its potential impact on the biodiversity of BWI, AHG has been identified as a priority invasive species for biosecurity management.

This research examines the robustness of expenditure that the decision makers would like to spend on surveillance and eradication. This research also enables decision makers to determine a robust portfolio of funds across the three biosecurity activities (quarantine, surveillance and eradication) against errors in the model parameter estimates. It is demonstrated that quarantine is more efficient than surveillance and eradication for managing the AHG. However, combining quarantine and surveillance works in a more robust manner against the underlying uncertainty in ecological modelling. Nevertheless, investment in quarantine should still outweigh that in surveillance. Increasing budget allocated to either quarantine or surveillance results in a larger annual budget, but decreases the total budget limit (i.e. the maximum total budget that decision makers may allocate to all three biosecurity activities) and increases robustness. More investment in eradication increases the robustness by increasing the probability of eradication success.

The model used to describe the spread of AHG after its incursion (spatial (i.e. spread among multiple locations) or local (i.e. spread at one location) dispersal model) and the estimated population threshold (i.e. estimated tolerable population size in which the surveillance programme is designed to detect at least one individual of AHG if it is present, ) both change the robustness achieved by the same budget limit. The spatial spread model has been demonstrated to work more efficiently than the local spread model. Transportation exploited by the AHG among locations could facilitate its spread and influence the invasion probability. Results vary when uncertainty is modelled in more fundamental parameters of model of detection probability. The situation in which is shown as robust-dominant and thus preferred in two of the data chapters and opportune-dominant in the first chapter of surveillance program design. Environmental damage and reputational, economic, ecological and societal impacts caused during the eradication campaigns and by the invasive species are not considered in this research. It is possible that the estimated population threshold could decrease when these are considered.

The model in this research can be extended to include additional activities (e.g. delimiting the infested area, monitoring to confirm eradiation success) to inform a more coordinated and comprehensive biosecurity system. This research model can also be extended to include temporal factors reflected by annual discount rate or by dynamic models related to varying population sizes (e.g. detection rate). Robust management of other invasive species (e.g. agricultural pests, marine invasive species and weeds), and conservation of endangered species in a data-poor environment can be achieved by adapting the three info-gap analysis models to the situations. A full decision analysis, weighing the costs and benefits of each biosecurity activity, is required to identify the risks and consequences of each biosecurity action on a case by case basis. This research provides a tool to assist in the completion of such decision analysis. IGDT as an alternative decision-making method has been applied in this research to present a quantitative bioeconomic modelling framework, thus providing a feasible and theoretical basis to assist in making informed decisions in a cost-effective and robust way. This research provides a reference for other islands and even mainlands for the purpose of managing other invasive species, but also provides scientific support for improving the rational biosecurity framework.

Keywords: invasive species, biosecurity, quarantine, surveillance, eradication, portfolio allocation, decision making, info-gap decision theory, robustness, opportuneness, spatial spread model, Asian house gecko, Barrow Island.

Item Type:	Thesis (PhD)
Murdoch Affiliation(s):	College of Science, Health, Engineering and Education Harry Butler Institute
Supervisor(s):	McKirdy, Simon, Thomas, Melissa, Coupland, Grey, Wang, Penghao and Zheng, D.
URI:	http://researchrepository.murdoch.edu.au/id/eprint/65047

Item Control Page

Downloads