Search

Many migratory shorebird species are undergoing severe population declines due to habitat loss. Selecting sites for protection along migratory shorebird flyways requires accounting for connectivity between sites and representing all migratory cycle stages within a protected area network. Site protection decisions often additionally account for the economic value of habitat-related ecosystem services, as shorebird habitats often provide services such as fisheries and recreational opportunities. In this study, we develop a new approach which combines graph theory, decision science, and stochastic dynamic optimization, to investigate whether habitat protection strategies aimed at safeguarding ecosystem services can also deliver efficient protection for migratory shorebirds. We use graph theory to empirically model migration routes for two shorebird species, the Curlew Sandpiper (Calidris ferruginea) and Eastern Curlew (Numenius madagascariensis), linking their migratory networks with data on habitat-related ecosystem services within the East Asian - Australasian Flyway (EAAF). Using dynamic optimization, we calculate the optimal habitat protection strategy for both ecosystem services and shorebird conservation objectives for both species. We further investigate if, and how, these two strategies can converge to achieve both objectives, and whether there is a trade-off between strategies. Our results show that the biodiversity outcome from strategies maximizing ecosystem services provided by shorebird habitat could align closely with a species-focused conservation plan. More specifically, the optimal ecosystem services oriented strategy maintains a similar level of shorebird populations as the optimal conservation strategy, while protecting an additional 45% (Curlew Sandpiper) and 42% (Eastern Curlew) of ecosystem services throughout the network. Conservation budget also matters, as small budgets would lead to inefficient protection for biodiversity. Our research shows that conservation planning can simultaneously achieve conservation and ecosystem services objectives for migratory species. By accounting for the value of ecosystem services within shorebird habitats, we provide quantitative evidence of a win-win scenario in which both shorebird conservation and ecosystem service provision can be achieved. Our model could be applied to other migratory species to improve spatially explicit planning for both biodiversity and ecosystem service targets.

Conservation of endangered species increasingly envisages complex strategies that integrate captive and wild management actions. Management decisions in this context must be made in the face of uncertainty, often with limited capacity to collect information. Adaptive management (AM) combines management and monitoring, with the aim of updating knowledge and improving decision-making over time. We provide a guide for managers who may realize the potential of AM, but are unsure where to start. The urgent need for iterative management decisions, the existence of uncertainty, and the opportunity for learning offered by often highly-controlled captive environments create favorable conditions for AM. However, experiments and monitoring may be complicated by small sample sizes, and the ability to control the system, including stochasticity and observability, may be limited toward the wild end of the spectrum. We illustrate the key steps to implementing AM in threatened species management using four case studies, including the management of captive programs for cheetah ( Acinonyx jubatus ) and whooping cranes ( Grus americana ), of a translocation protocol for Arizona cliffroses Purshia subintegra and of ongoing supplementary feeding of reintroduced hihi ( Notiomystis cincta ) populations. For each case study, we explain (1) how to clarify whether the decision can be improved by learning (i.e. it is iterative and complicated by uncertainty) and what the management objectives are; (2) how to articulate uncertainty via alternative, testable hypotheses such as competing models or parameter distributions; (3) how to formally define how additional information can be collected and incorporated in future management decisions.

International audience; The management of introduced species that are both invasive and commercially valuable is contentious. While such species provide substantial economic benefits to some, they pose considerable costs to others due to negative impacts on ecosystems. We propose a decision framework to help balance conflicting objectives and support the management of commercially valuable invasive species. We illustrate our framework using the buffel grass (Pennisetum ciliare syn. Cenchrus ciliaris L.) invasion in Australia. In rangelands worldwide, buffet grass is amongst a suite of commercially valuable invasive species, highly valued by some graziers as a pasture species but widely unpopular among those concerned with the threat it poses for native biodiversity. The framework comprises four components. First we develop a state and transition model to represent the invasion dynamics of buffel grass and the effect of management actions. Second we construct utility functions to represent the relative values of buffel grass cover in terms of production and biodiversity utilities (indicative of grazier's and conservationist's relative 'happiness' regarding buffel grass cover). We draw on expert and empirical data for the construction of the model and utility functions. Third, we present management strategies that minimize losses in production and biodiversity utilities in the absence of budget constraints. We use stochastic dynamic programming and multi-criteria decision analysis which explicitly account for trade-offs between production and biodiversity conservation concerns. Finally, we conduct a sensitivity analysis to assess the impact of assumptions of the model and utility functions. Management complexity arises from biodiversity-production trade-off situations in which the desirable state of buffel grass cover and effective actions to achieve this state depend on the utility functions. Management solutions were particularly sensitive to the shape of the utility functions for biodiversity. Solutions were less sensitive to uncertainty surrounding the effectiveness of management actions. We found in most biodiversity-production tradeoff situations a compromise solution led to management for an intermediate level (>0-50%) of buffel grass cover. However, when maximal values for both biodiversity and production are high, it may be more practical to manage for one value rather than find a compromise solution where concessions from both sides are high. Our decision framework does not attempt to optimize economic outcomes but provides a guide for formulating trade-offs between opposing views, something that is relevant to the management of any natural resource management problem where there are conflicting objectives.

For metapopulation management problems with small state spaces, it is typically possible to model the problem as a Markov decision process (MDP), and find an optimal control policy using stochastic dynamic programming (SDP). SDP is an iterative procedure that seeks to optimise a value function at each timestep by trying each of the actions defined in the MDP. Although SDP gives the optimal solution to conservation management questions in a stochastic world, its applicability has always been limited by the so-called curse of dimensionality. The curse of dimensionality is the problem that adding new state variables inevitably results in much larger (often exponential) increases in the size of the state space, which can make solving superficially small problems impossible. A large state space makes optimal SDP solutions computationally expensive to compute because optimal SDP techniques require the value function to be updated for the entire state space for every time step. The high computational requirements of large SDP problems means that only simple population management problems can be analysed. In this paper we present an application of the on-line sparse sampling algorithm proposed by Kearns, Mansour & Ng (2002), which can be used to approximate the optimal solution of a MDP for a given starting state. The algorithm is particularly attractive for problems with large state spaces as it has a running time that is independent of the size of the state space of the problem. We apply the algorithm of Kearns et al. (2002) to a hypothetical fish metapopulation where we define the management objective to be to maximise the number of occupied patches during the management horizon. Our model has multiple management options to combat the threats of water abstraction and waterhole sedimentation. We compare the performance of the optimal solution to the results of the on-line sparse sampling algorithm for a simple 3-waterhole case. We find that the approximation algorithm out-performs a random strategy for almost all combinations tested. In this case, the on-line algorithm gives the best approximation to the optimal SDP solution with three look-ahead steps and just two samples per approximation step. For this simple problem there does not appear to be much benefit in increasing the number of samples, although deeper simulation is required to confirm this. If only limited computing power is available, these results suggest that the best approximation will be obtained using three look-ahead steps and two samples per approximation step. General results for management are hard to obtain from this theoretical model, as the results are highly dependent on the arbitrary parameters used to generate the results. However the results obtained suggest that the continued presence of water in the system is critical to the survival of the metapopulation, and that for this parameter set, prevention of water abstraction is more important than catchment management for this system. Further work needs to be done to properly parameterize this model before general results can be drawn with more confidence. This paper introduces a new algorithm to conservation management that provides a way to avoid the effects of the curse of dimensionality. The work has the potential to allow us to approximate solutions to much more complex metapopulation management problems in the future.