Flyways.us

The establishment of duck hunting regulations involves such things as the number of days the season will be open, how many ducks to allow in the bag limit, how many mallards and other species to allow to be taken each day and what beginning and ending dates are suitable. In the past, those decisions have involved politics, the opinions of professional migratory bird managers and others, and the social need to "do something" when a population level is low. All those factors are affected by the status of waterfowl populations and the understanding of the relationships among populations, regulations and harvest

Before 1995, decisions were made based on discussions that sounded something like this: "I think the bag limit should be lower than last year's because ducks are down." "I think an increase in season length by seven days is too many." "Ducks are up so we can liberalize the bag limit but not season length." "Ducks are up, so let us hunt more days - we will leave the bag limit alone." "The data shows that mallard survival rates have not changed over the years, so give our hunters 60 days to hunt and leave the bag limit at five and let's go to work on habitat." "Close the duck season."

In many ways, the system worked. Population levels and the associated hunting opportunities have been maintained in spite of greatly varying habitat conditions. Nevertheless, there has been much uncertainty about the effect of regulations on harvest and, in turn, on duck populations. To help resolve that uncertainty, a process called Adaptive Harvest Management (AHM) was proposed in the early 1990s. This is a specific implementation of a broader process called Adaptive Resource Management. (learn more)

AHM provides a more objective, better-informed and less contentious decision-making process; an explicitly defined role for monitoring programs; and a formal and coherent framework for addressing controversial harvest management issues. The system includes explicit statements about the uncertainty of the information included. For example, the exact size of the mallard population and the exact harvest rate are not known. ARM also includes a "learning" objective - the system is constructed so that the rate of learning is identified and all information learned is explicitly incorporated into the next round of decision-making. An environment is achieved in which achievement of management objectives can be measured.

That computers and models are an important tool in AHM should surprise no one. AHM is data hungry: Pond counts must be related to bird counts; harvest data must be related to band and band reporting data and to bird counts; biological parameters such as survival and production rates must be related to bird counts and harvest; harvest characteristics must be related to regulations; and many years' data from many regions of the continent must be considered.

While it would be ideal if there was agreement about the nature of those relationships by all parties, agreement is not necessary for AHM to succeed. That is where the "learning" component comes in. Several different relationships can be described by models that use much of the same data. Currently there are four AHM models encompassing the broad theories associated with relationships between mortality caused by hunting and mortality caused by other factors. Further, two different theories about mallard reproduction are included.

If at least one of the models accurately represents the "real" world, all four can be run and let the data and assumptions produce a prediction about, for example, mallard population size. The accuracy of the four predictions can be measured and, eventually, scientists can learn which hypothesis is correct. In addition, an evaluation of how well the models depict the real world can occur and needed modifications can be made to them.

Since 1997 and into 2000, the models have incorporated four alternatives for hunting seasons labeled according to content: very restrictive, restrictive, moderate and liberal. Each regulatory package contains a set of frameworks for each of the four flyways (for example, the liberal alternative contains a 60 day duck season for the Mississippi Flyway and 74 days for the Central). Since the best data is available for mallards and they are one of the most heavily harvested of all duck species, each regulatory package has a target harvest rate (the percent of the population harvested) for mallards associated with it. When the models are run, the results are weighted by several factors and an "optimal" harvest alternative is identified. Once the computer has completed its tasks, it is still up to people (wildlife agency administrators, biologists and the public) to select a regulatory package for the current year. However, they now can do this with a great deal more confidence and considerably less controversy than in the years before AHM.

Each year, current survey data is included with historical data and included in new model runs. The predictions of the models are compared to the results of surveys and adjustments to some model parameters can be made. Thus, waterfowl managers learn a little more about the complex relationships of regulations, harvest and duck populations.

Adaptive Harvest Management for ducks is in its infancy. Including other duck species will take years, but some interim models for pintails and canvasbacks have been implemented, allowing an objective approach to establishing hunting regulations and avoiding knee-jerk reactions to changes in harvest or survey results. In 2000, new models were available to consider an eastern mallard population and its interaction with a much larger mid-continent population which largely supplies mallards for most of the United States.

The alternative to AHM is a return to days of setting duck hunting regulations in a more contentious atmosphere and one driven less on science than on opinion. It is the view of the CFC that this is not a good alternative and support for AHM remains high. (learn more about AHM - link to FWS report).