Box 1.

Procedures of annual adaptive harvest management of the pink-footed goose population in a nutshell

The development of an adaptive harvest management (AHM) strategy requires the specification of four elements: (a) a set of alternative population models, which bound the uncertainty about effects of harvest and other environmental factors, (b) a set of probabilities describing the relative credibility of the alternative models, (c) a set of alternative harvest quotas from which to choose and (d) an objective function, by which alternative harvest strategies can be evaluated. An optimal management strategy prescribes a harvest quota for each and every possible set of model probabilities, and for population abundance and environmental conditions that may be observed at the time a decision is made

Nine models of pink-footed goose dynamics describe competing hypotheses about how reproductive and survival rates might vary over time. The models focus on whether spring temperature and density dependence influence survival and/or reproduction. Bayesian probabilities are used to express the relative ability of each model to accurately predict the changes in population size that actually occur, and they are updated each year using monitoring information. In the figure below are the time sequences of the aggregate probabilities on models that incorporate (A) density-dependent survival, (B) density-dependent reproduction and (C) days above freezing in May in Svalbard in the reproductive and survival processes

The four elements of AHM (models, model probabilities, alternative quotas and objective function) are used each year to calculate an optimal harvest strategy designed to maintain the population near the goal of 60 000. The optimal harvest strategy is a large lookup table that is difficult to display graphically. Below is a simplified representation of the strategy for model probabilities in 2016, in which a series of yes–no questions are asked (yes is the left branch; no is the right branch) about the abundance of adults and young (A and Y in thousands, respectively) and the number of days above freezing in May in Svalbard (DAYS). The approximate harvest quota (in thousands, to the nearest 2.5) is given at the ends of the branches