Conservation organizations often rely on incentive payments to private landowners to "buy" conservation benefits. In evaluating the efficiency of such programs, conservation biologists have often assumed contracts can be acquired at landowners' willingness-to-accept. Were this possible it would represent something of a best possible outcome for conservation from a negotiation with private landowners. Drawing on game theory, optimization methods and agent-based simulations, I will examine how conservation outcomes would be affected if landowners instead were to hold out for payments over and above their willingness-to-accept to gain some surplus from the transaction.
Model uncertainty and limited data coverage are fundamental challenges to robust ecosystem management. These challenges are acutely highlighted by concerns that many ecological systems may contain tipping points. Before a collapse, we do not know where the tipping points lie, if the exist at all. Hence, we know neither a complete model of the system dynamics nor do we have access to data in some large region of state-space where such a tipping point might exist. These two sources of uncertainty frustrate state-of-the-art parametric approaches to decision theory and optimal control. I will illustrate how a non-parametric approach using a Gaussian Process prior provides a more flexible representation of this inherent uncertainty. Consequently, we can adapt the Gaussian Process prior to a stochastic dynamic programming framework in order to make robust management predictions under both model and uncertainty and limited data.
In this talk I will focus on the role of vector dynamics and its ecological and economic implications for disease control. Genetic methods of controlling insects and agricultural pests have advanced in recent years but their economic benefits remain largely unexplored. By integrating epidemiological and economic approaches we can build sensible mathematical models for exploring the implications of vector control on levels of disease burden. In this talk I will illustrate the ways that we have approached this for understanding Dengue dynamics and some of the subtle ways that understanding human movement and flows are important for implementing disease invention strategies.
Using stochastic dynamic programming techniques, we can optimize our decisions over time and space to maximize our chance of achieving our conservation goal at a minimum cost. In doing so, we assume that although uncertain, the future dynamics of the system can be represented using transition probabilities. When several future dynamics of the system are possible (model uncertainty), finding the best decision over time becomes an adaptive management problem. Unfortunately adaptive management problems suffer from a lack of solution methods. I will demonstrate how partially observable Markov decision processes (POMDP) can help solving large adaptive management problems over time and space.
Biological invasions are spatial-dynamic processes — they unfold over space and time, driven by a combination of reproduction and dispersal. Consequently, their management requires weighing not only how much and when to invest in control to reduce their damages, but where controls should be applied. Furthermore, invasions unfold in landscapes comprising numerous, independently managed properties such that their spread depends on the control choices of many landowners. Here I present three bioeconomic studies that address these complexities of bioinvasion management. They examine 1) optimal surveillance design for early detection of invasions, 2) optimal spatial control strategies, and 3) individual and cooperative invasion management.
Real landscapes are dynamic in space and time, and the scales over which such variation occurs can determine the success of different conservation strategies for resident species. Within such landscapes, real species rely on a variety of individual-level behaviors for movement and navigation. Movement behaviors such as long-distance searching and fine-scale foraging are often intermixed but operate on vastly different spatial and temporal scales. Individual experience, life-history traits, and resource dynamics combine to shape population-level patterns such as range residency, migration, and nomadism.
I will discuss how a combination of empirical movement data and powerful statistical approaches ("animal models" of pedigree effects; semi-variance functions) can be used to inform our understanding of animal movement and help guide conservation planning. Animal models can be used to control for genetic variation among individuals while exploring alternative hypotheses about other factors that influence animal movement. Semi-variance approaches can be used to identify multiple movement modes and solve the sampling rate problem for tracking data, allowing for the identification of critical scales for movement. Together these approaches can help reveal the relationships among individual movements, landscape dynamics, and population level patterns.
I will discuss a few major national and international initiatives that impinge on sustainability of resource management. Based on experiences with one of the larger projects on natural system restoration in the Everglades of South Florida, I will provide some insight and lessons learned in linking models to planning and policy. These will in part emphasize the difficulty in obtaining concensus on control objectives and the benefits of taking a rather less optimization- focused perspective but one considering robustness of relative rankings of alternatives. Finally I will point out some challenges in the topic areas of the workshop that may benefit from mathematical approaches.
I will give an overview of mathematical challenges that arise in developing management strategies for natural systems. The emphasis will be on issues that arise from the nature of the biological systems, including, but not limited to, limited data, nonlinearities, stochasticity, constraints that arise from biological issues, and time scales. I will illustrate the concepts by starting with some of the best studied examples, which come from fisheries, then discuss issues of invasive species, and finally move on to lesser studied and more poorly specified areas. The goal of the talk will be to set the stage for the workshop and initial discussions.
Management of stochastic renewable natural resources occurs in the presence of various forms of uncertainty (e.g., parametric, model, and state).While the implications of different types of uncertainty for management have been carefully analyzed individually, it is not clear when each different type of uncertainty is relatively more or less important for the resource manager.In this talk, I will attempt to define three different forms of uncertainty.I will then discuss some of the challenges of comparing the value of learning about these different uncertainties within the same resource management problem. I will then describe and simulate one candidate method for comparing two different types of uncertainty.
Risk analysis are increasingly recognized as important for prioritizing effort in biological invasions. I will review the state of the science on alien species risk assessment, identify where research has been focused, and identify gaps in the literature, which if filled may improve the science and application of risk analyses. I will walk through individual studies estimating components of risk, given data available. Finally, I will present the outcome of a pathway level bioeconomic risk analysis of an existing policy for invasive species.
My talk will be flexible and depend in the events of the first day. I can cover the application of Marxan and Marxan with zones to conservation planning with a focus on ecosystem services. The disadvantages of using Marxan and other spatial planning tools will be explained. I can also explain how we have calculated biodiversity benefits for offsetting.
Disturbances are ubiquitous in nature, and are believed to be strong drivers of both ecological diversity and species invasion. The need to address the impacts of environmental disturbance is increasingly urgent in the face of anthropogenic alterations to existing disturbance regimes. I will discuss how an ecological niche-based theory of disturbance, encompassing five interacting aspects (frequency, intensity, duration, extent and timing), can be used to study a wide range of issues related to disturbance regimes and their effects on biological systems. This conceptual framework allows an integrated study of disturbance across levels of biological organization: from the individual through to the population, the community and entire ecosystems. Ongoing theoretical and empirical research not only informs us about when disturbances are likely to pose a problem, but also lets us assess how we can manipulate disturbances to achieve desired management outcomes. As disturbances of many types are increasingly used to manage ecosystems, this approach therefore can be applied to a wide range of management issues. Managers and policy makers need to be able to make reliable predictions — only if we can anticipate them, can we avoid or ameliorate the impacts of such disturbances.
The national quarantine system in Japan has been very strict and extensive. All cargoes entering Japan are inspected, and any suspicious consignment is subject to admonitory cessation of import, restriction, exhaustive inspection, or clearance. Recently, partner countries and international organizations such as WTO have requested Japan to discontinue the superfluously strict regulations and to relax quarantine based on scientific risk-assessment in order to secure international trade currency. Thus, to rationalize the quarantine process, it is our urgent task to draw up a black list of invasive species based on scientific knowledge. However, how can we evaluate the potential risk of an invasive species that has NOT yet invaded Japan?
Nevertheless, invasive insect pests, particularly those of agricultural importance, have been inventoried by several international organizations, and this information can be retrieved via the internet.
We can now easily access information on non-native species from around the world; however, little attempt has been made to scientifically quantify the risk of potentially invasive insects and to utilize this information for local risk-assessment in actual quarantine administration.
In this study, we exhaustively collected information on invasive insect pests from various worldwide databases, and ca 5,000 species were indexed in our intra-network web application using a GIS database (a geographic information system-related database). Of the 5,000 species catalogued we initially focused on 500 species as candidates of primary importance. We obtained comprehensive details on the native and invasive distributions of these species and also on their basic biological characteristics. The invasion probabilities of these 500 species were calculated using Maxent, one of the statistical tools for predicting species distribution. The invasion probabilities were further weighted with the annual yield of agricultural products related to the focal insect. The total sum of the products of invasion probability and the value of all agricultural products related to the insect will be a used to predict the potential risk of a specific species in Japan. Using this procedure, we evaluated the risk for all 500 species.
Biodiversity in paddy fields and the quality of habitats for aquatic wildlife may directly or indirectly depend on wildlife-friendly agriculture conducted by farmers and the purchasing behavior of consumers. Wildlife-friendly agriculture could enhance biodiversity in agricultural landscapes, but the behavior of farmers and consumers is affected by the status of ecosystems as well as by other people's behavior. Thus, it is necessary to consider both the status of ecosystems and human behavior to properly manage agricultural landscapes and local economies and to increase biodiversity, the proportion of wildlife-friendly agriculture, and quantity of wildlife-friendly rice sales. Strategies for achieving these goals include conservation efforts for paddy fields, a subsidization scheme for certified wildlife-friendly rice, and improvement of rice quality. In this study, I developed a coupled ecological and social dynamics model to derive the optimal management strategies for agricultural landscapes in Sado Island, Japan. I found that increasing the attractiveness of wildlife-friendly rice is an effective management strategy. High conformist tendency in farmers could keep total biomass and the fraction of farmers practicing wildlife-friendly agriculture high even after the subsidy is reduced or terminated.