3 présentations

• 
  14h00 - 14h30

  Optimal harvesting of a regime-switching stochastic growing renewable resource: a utility theory approach

  • Jules Sadefo Kamdem, prés., Université de Guyane (DFR SJE)

  In this paper, following Smith (1978) and Ludwig (1979), we examine the optimal rate of extraction from a stochastically regime-switching growing resource stock (typically to maximize the expected flow of utility from net revenue). Assuming a continuous-time stochastic regime switching for the biomass Gilpin-Ayala growth function, and CRRA utility function for firms preferences, we find the optimal extraction rate that maximizes the expected integral of the utility function of net revenue of firms.

• 
  14h30 - 15h00

  Probability of misidentifying population declines and recoveries using linear models

  • Matthew Holden, University of Queensland
  • Jerzy Filar, prés., University of Queensland

  Accurately estimating trends in population abundance is critical for developing ecological theory, performing environmental assessments, and advising natural resource management. While the error and power of statistical methods for detecting population declines and recoveries are well studied, they rarely consider the issue of density dependence. If population size time series data occurs in an area where the species is abundant, density dependence may cause the over-prediction of a population decline. In this paper, we provide simple analytic formulae for the probability of misestimating population growth rates above or below a specified threshold. We then use the formulae in two applied contexts (1) the probability of falsely predicting a threatened species is declining or recovering and (2) the use of linear population models for predicting species occurrence. In the latter case, we derive simple rules of thumb for the critical population abundance, in relation to carrying capacity, after which density dependence interferes with accurate predictions of persistence. The critical abundance can be used as a guideline for when it may be appropriate to use linear population process models to predict species occurrence in a density-dependent world.

• 
  15h00 - 15h30

  Risk measures for natural resources management strategy assessment

  • Michel De Lara, prés., ENPC, CERMICS
  • Maia Douillet, Ecole des Ponts ParisTech
  • Martin Quaas, Leipzig University
  • Rudi Voss, Christian-Albrechts-Universität, Kiel

  Strategy assessment is part of natural resources management.
  Uncertainty is ever present, as ecosystem functioning is often poorly
  understood and modelized, often depends upon variable environmental conditions,
  and as a time spans can extend over decades.
  We propose to introduce risk measures, originally designed for mathematical
  finance, in strategy assessment for natural resources management.

  In a first part, we set the mathematical framework for natural resources
  management, and outline the traditional methods of strategy assessment in this
  framework. Then, we introduce a novel approach based on risk measures. The
  second part is a case study of selective fishery of Baltic cod (Gadus morhua).
  We compare the strategies evaluated as "sustainable" according to the novel
  approach and to the traditional approaches, with numerical simulations.
  In the third part, we open perspectives and propose theoretical issues left to
  explore.