Tracy M. Rout

Optimal adaptive management for the translocation of a threatened species

Active adaptive management (AAM) is an approach to wildlife management that acknowledges our imperfect understanding of natural systems and allows for some resolution of our uncertainty. Such learning may be characterized by risky strategies in the short term. Experimentation is only considered acceptable if it is expected to be repaid by increased returns in the long term, generated by an improved understanding of the system. By setting AAM problems within a decision theory framework, we can find this optimal balance between achieving our objectives in the short term and learning for the long term. We apply this approach to managing the translocation of the bridled nailtail wallaby (Onychogalea fraenata), an endangered species from Queensland, Australia. Our task is to allocate captive-bred animals, between two sites or populations to maximize abundance at the end of the translocation project. One population, at the original site of occupancy, has a known growth rate. A population potentially could be established at a second site of suitable habitat, but we can only learn the growth rate of this new population by monitoring translocated animals. We use a mathematical programming technique called stochastic dynamic programming, which determines optimal management decisions for every possible management trajectory. We find optimal strategies under active and passive adaptive management, which enables us to examine the balance between learning and managing directly. Learning is more often optimal when we have less prior information about the uncertain population growth rate at the new site, when the growth rate at the original site is low, and when there is substantial time remaining in the translocation project. Few studies outside the area of optimal harvesting have framed AAM within a decision theory context. This is the first application to threatened species translocation.

Active adaptive conservation of threatened species in the face of uncertainty

Adaptive management has a long history in the natural resource management literature, but despite this, few practitioners have developed adaptive strategies to conserve threatened species. Active adaptive management provides a framework for valuing learning by measuring the degree to which it improves long-run management outcomes. The challenge of an active adaptive approach is to find the correct balance between gaining knowledge to improve management in the future and achieving the best short-term outcome based on current knowledge. We develop and analyze a framework for active adaptive management of a threatened species. Our case study concerns a novel facial tumor disease affecting the Australian threatened species Sarcophilus harrisii: the Tasmanian devil. We use stochastic dynamic programming with Bayesian updating to identify the management strategy that maximizes the Tasmanian devil population growth rate, taking into account improvements to management through learning to better understand disease latency and the relative effectiveness of three competing management options. Exactly which management action we choose each year is driven by the credibility of competing hypotheses about disease latency and by the population growth rate predicted by each hypothesis under the competing management actions. We discover that the optimal combination of management actions depends on the number of sites available and the time remaining to implement management. Our approach to active adaptive management provides a framework to identify the optimal amount of effort to invest in learning to achieve long-run conservation objectives.

How to Decide Whether to Move Species Threatened by Climate Change

Introducing species to areas outside their historical range to secure their future under climate change is a controversial strategy for preventing extinction. While the debate over the wisdom of this strategy continues, such introductions are already taking place. Previous frameworks for analysing the decision to introduce have lacked a quantifiable management objective and mathematically rigorous problem formulation. Here we develop the first rigorous quantitative framework for deciding whether or not a particular introduction should go ahead, which species to prioritize for introduction, and where and how to introduce them. It can also be used to compare introduction with alternative management actions, and to prioritise questions for future research. We apply the framework to a case study of tuatara (Sphenodon punctatus) in New Zealand. While simple and accessible, this framework can accommodate uncertainty in predictions and values. It provides essential support for the existing IUCN guidelines by presenting a quantitative process for better decision-making about conservation introductions.

Prevent, search or destroy? A partially observable model for invasive species management

Summary: The extensive impact of invasive species has motivated a growing field of research combining ecological and economic models to find cost-effective management strategies. Ecological systems are rarely perfectly observable, meaning decision-makers are usually uncertain about the current extent of an infestation and even whether an invasive species is present or absent. We show how to account for this uncertainty when providing decision support for invasive species management. We constructed the first partially observable model to analyse the trade-off between all three facets of invasive species management: quarantine, surveillance and control. We use a partially observable Markov decision process (POMDP) to determine how to allocate resources between these actions when the extent of an invasion is uncertain. We use a case study of potential black rat Rattus rattus invasion on Barrow Island, Western Australia. Our model shows it is often better to manage based on an uncertain belief in species presence than to spend money trying to confirm the presence or absence through surveillance. While it was never optimal to invest solely in surveillance to reduce uncertainty, it was often optimal to combine surveillance with quarantine or control. These mixed strategies, where multiple actions are implemented simultaneously, were more often optimal than for similar decision models where the extent of the infestation is known, suggesting an element of risk spreading. Optimal investments in each action were driven by their estimated efficacy, and the difference in the estimated impact of a localized and widespread invasion. For example, in our case study, it was often optimal to invest solely in control due to the low estimated efficacy of quarantine and the relatively small impact of a localized incursion. Synthesis and applications. Our analysis shows that the cost of reducing uncertainty through surveillance is not always accompanied by an improvement in management outcomes. By carefully analysing the benefits of surveillance prior to implementation of invasive species management strategies, managers can avoid wasting resources and improve management outcomes.

Uncertain Sightings and the Extinction of the Ivory-Billed Woodpecker

The extinction of a species can be inferred from a record of its sightings. Existing methods for doing so assume that all sightings in the record are valid. Often, however, there are sightings of uncertain validity. To date, uncertain sightings have been treated in an ad hoc way, either excluding them from the record or including them as if they were certain. We developed a Bayesian method that formally accounts for such uncertain sightings. The method assumes that valid and invalid sightings follow independent Poisson processes and use noninformative prior distributions for the rate of valid sightings and for a measure of the quality of uncertain sightings. We applied the method to a recently published record of sightings of the Ivory-billed Woodpecker (Campephilus principalis). This record covers the period 1897-2010 and contains 39 sightings classified as certain and 29 classified as uncertain. The Bayes factor in favor of extinction was 4.03, which constitutes substantial support for extinction. The posterior distribution of the time of extinction has 3 main modes in 1944, 1952, and 1988. The method can be applied to sighting records of other purportedly extinct species.

Optimal Allocation of Conservation Resources to Species That May be Extinct

Statements of extinction will always be uncertain because of imperfect detection of species in the wild. Two errors can be made when declaring a species extinct. Extinction can be declared prematurely, with a resulting loss of protection and management intervention. Alternatively, limited conservation resources can be wasted attempting to protect a species that no longer exists. Rather than setting an arbitrary level of certainty at which to declare extinction, we argue that the decision must trade off the expected costs of both errors. Optimal decisions depend on the cost of continued intervention, the probability the species is extant, and the estimated value of management (the benefit of management times the value of the species). We illustrated our approach with three examples: the Dodo (Raphus cucullatus), the Ivory-billed Woodpecker (U.S. subspecies Campephilus principalis principalis), and the mountain pygmy-possum (Burramys parvus). The dodo was extremely unlikely to be extant, so managing and monitoring for it today would not be cost-effective unless the value of management was extremely high. The probability the Ivory-billed woodpecker is extant depended on whether recent controversial sightings were accepted. Without the recent controversial sightings, it was optimal to declare extinction of the species in 1965 at the latest. Accepting the recent controversial sightings, it was optimal to continue monitoring and managing until 2032 at the latest. The mountain pygmy-possum is currently extant, with a rapidly declining sighting rate. It was optimal to conduct as many as 66 surveys without sighting before declaring the species extinct. The probability of persistence remained high even after many surveys without sighting because it was difficult to determine whether the species was extinct or undetected. If the value of management is high enough, continued intervention can be cost-effective even if the species is likely to be extinct.

Optimization methods to solve adaptive management problems

Determining the best management actions is challenging when critical information is missing. However, urgency and limited resources require that decisions must be made despite this uncertainty. The best practice method for managing uncertain systems is adaptive management, or learning by doing. Adaptive management problems can be solved optimally using decision-theoretic methods; the challenge for these methods is to represent current and future knowledge using easy-to-optimize representations. Significant methodological advances have been made since the seminal adaptive management work was published in the 1980s, but despite recent advances, guidance for implementing these approaches has been piecemeal and study-specific. There is a need to collate and summarize new work. Here, we classify methods and update the literature with the latest optimal or near-optimal approaches for solving adaptive management problems. We review three mathematical concepts required to solve adaptive management problems: Markov decision processes, sufficient statistics, and Bayes’ theorem. We provide a decision tree to determine whether adaptive management is appropriate and then group adaptive management approaches based on whether they learn only from the past (passive) or anticipate future learning (active). We discuss the assumptions made when using existing models and provide solution algorithms for each approach. Finally, we propose new areas of development that could inspire future research. For a long time, limited by the efficiency of the solution methods, recent techniques to efficiently solve partially observable decision problems now allow us to solve more realistic adaptive management problems such as imperfect detection and non-stationarity in systems.

Monitoring, imperfect detection, and risk optimization of a Tasmanian devil insurance population

Most species are imperfectly detected during biological surveys, which creates uncertainty around their abundance or presence at a given location. Decision makers managing threatened or pest species are regularly faced with this uncertainty. Wildlife diseases can drive species to extinction; thus, managing species with disease is an important part of conservation. Devil facial tumor disease (DFTD) is one such disease that led to the listing of the Tasmanian devil (Sarcophilus harrisii) as endangered. Managers aim to maintain devils in the wild by establishing disease-free insurance populations at isolated sites. Often a resident DFTD-affected population must first be removed. In a successful collaboration between decision scientists and wildlife managers, we used an accessible population model to inform monitoring decisions and facilitate the establishment of an insurance population of devils on Forestier Peninsula. We used a Bayesian catch-effort model to estimate population size of a diseased population from removal and camera trap data. We also analyzed the costs and benefits of declaring the area disease-free prior to reintroduction and establishment of a healthy insurance population. After the monitoring session in May-June 2015, the probability that all devils had been successfully removed was close to 1, even when we accounted for a possible introduction of a devil to the site. Given this high probability and the baseline cost of declaring population absence prematurely, we found it was not cost-effective to carry out any additional monitoring before introducing the insurance population. Considering these results within the broader context of Tasmanian devil management, managers ultimately decided to implement an additional monitoring session before the introduction. This was a conservative decision that accounted for uncertainty in model estimates and for the broader nonmonetary costs of mistakenly declaring the area disease-free.

Declaring Eradication of an Invasive Species

Imperfect detection methods mean that it is difficult to tell whether a species is absent from a site or remains undetected. For this reason, the decision to conclude an eradication program and declare a species successfully eradicated is fraught with uncertainty (Morrison et al., 2007). There are two errors that can be made (Regan et al., 2006). First, if the species is declared eradicated when it is still present, its population could grow undetected, causing large economic and environmental damages. There are costs associated with reinitiating the eradication campaign and reducing the species’ population to a low level. Second, monitoring cannot continue indefinitely, and continuing to survey when a species has already been eradicated uses resources that could be better deployed elsewhere. This chapter reviews statistical models that can be used to quantify the certainty that a species has been successfully eradicated from a site. It then describes how to analyse logically the decision to declare eradication, considering the risks and consequences of getting it wrong.

Should we move species threatened by climate change

One solution is to move species to places with a more suitable climate. But the idea of introducing species to areas where they have never occurred before is controversial, because species introduced to somewhere they’ve never lived could have devastating consequences for the species already there. Just think of foxes, lantana, cane toads and other invasive species in Australia. So how do we weigh up the costs and benefits? In a new study published recently in the journal PLOS ONE, we developed a way of finding the answer.