Following is a summary of the Black Duck Adaptive Harvest Management Working Group workshop held on November 17, 2000 in Sandusky, Ohio. The workshop was held in conjunction with the Black Duck Joint Venture Technical Committee meeting and was graciously hosted by the Ohio Division of Wildlife. The intent of this summary is to provide an overview of the workshop along with key discussions and decisions. This summary does not include minutes of all of the activities and discussion that took place.

The agenda for the workshop is included in Appendix A. Appendix B includes additional references. This summary is posted on the Black Duck AHM Working Group web site at http://coopunit.forestry.uga.edu/blackduck/ . Components of the summary included in the appendices may also be included other places on the web site and updated over time.
 
 
 

Welcome and Introductions

Michael J. Conroy with the Georgia Cooperative Wildlife Research Unit opened the meeting and welcomed all participants. David J. Case of D.J.Case and Associates served as meeting facilitator.

Introductions were made. Meeting attendees included:
 
 
 

Name	Representing	email
Fred Johnson	USFWS- OMBM	fred_a_johnson@fws.gov
Bob Blohm	USFWS- OMBM	robert_blohm@fws.gov
Steve Wendt	CWS- National	steve.wendt@ec.gc.ca
Ken Ross	CWS- Ontario	ken.ross@ec.gc.ca
Myrtle Bateman	CWS- Atlantic	myrtle.bateman@ec.gc.ca
Jerry Serie	USFWS- Migratory Birds	Jerry_Serie@fws.gov
Graham Smith	USFWS- Migratory Birds	Graham_Smith@fws.gov
Bill May	Ohio Division of Wildlife	bill.may@dnr.state.oh.us
Steve Barry	Ohio Division of Wildlife	steve.barry@dnr.state.oh.us
George Haas	USFWS- R5	George_Haas@fws.gov
Brigitte Collins	CWS- Ontario	Brigitte.Collins@ec.gc.ca
Mark Shieldcastle	Ohio Division of Wildlife	Mark. Shieldcastle@dnr.state.oh.us
H W Heusmann	Mass. Div. Fish & Wildl.	H.heusmann@state.ma.us
Melody Miller	Indiana DNR	mmiller@bluemarble.net
Daniel Bordage	CWS- Québec	Daniel.Bordage@ec.gc.ca
Nathan Zimpfer	Univ. Georgia	nathanz@mac.com
Christopher Fonnesbeck	Univ. Georgia	chrisf@fonnesbeck.net
Dave Case	USFWS	dave@djcase.com
Michael Conroy	USGS/ Univ. Georgia	conroy@smokey.forestry.uga.edu

 
 

Workshop Objectives

The objectives for the workshop were to:

Update Black Duck AHM Working Group members on progress by the University of Georgia group since the June meeting;

Provide Working Group members with some hands-on experience using the models; and

Get feedback from the Working Group.

Mike Conroy gave a brief overview on the project for new people-- how we came to be where we are today and how we got to this meeting. The group revisted the fact that we are working toward an international harvest strategy, but that no decision has been made that such an approach will be implemented.

Mike reviewed the single-population black duck model. The basis for this model isi laid out in detail in the report by Conroy and Miller (download). Basically, the model starts with breeding populations for black ducks and mallards, and specifies production rates under 2 competing hypotheses: (a) competition with mallards , (b) no competition with mallards. For each of the fall populations projected, the model then specifies a response to harvest management under 2 competing harvest hypotheses: (a) compensation through density dependent mortality, and (b) no compensation (density-independent mortality). This results in 4 projected breeding populations for the next year; under adaptation the weighted average of these is used, with the weights being based on the relative agreement of the models to data. The weights are modified by comparison of each model's prediction with the breeding population estimates the following year; models that come closer to the observed BPOP receive more weight in decision making in subsequent years.  A major shortcoming of this model is that it doesn't incorporate the fact that there may be more than 1 stock of black ducks.

Opened for questions. Most questions were to clarify the data sets and procedures used to obtain estimates for the model.
 
 

Schematic Overview of Spatial Black Duck Model

This presentation revolved around proposed, preliminary delineation of populations based on band recoveries and other data. Refer to accompanying text and maps and other figures.

Mike went through the Proposed Breeding Areas, Harvest Areas, and Wintering Areas (download). Several questions followed:
 

Q:What about birds that winter in Ontario?  How are they accounted for?
A: This is an important point. At present these birds are treated as if they winter in the U.S.   We may need to consider re-drawing the boundaries of the wintering areas to account for this.   In general we may have to redraw some of the boundaries between the proposed breeding areas to better match administrative units.

Q:Seems like we may need to split the Atlantic Flyway wintering area into a north and south unit
A: This is one possible approach.  Another is to allow birds to first  move to the Mississippi Flyway and then to the southern Atlantic Flyway; at present birds starting in the western breeding area only move either to the Mississippi Flyway or the  Atlantic Flyway.

Q: Did you only use direct recoveries
A: Yes, for the estimation of movement and harvest rates. For fidelity rates we examined using indirect recoveries.

Reviewed annual life cycle

Q: Fred asked question re integration of BD with E Mallards.

Data Sets Used to Parameterize Spatial Model

Mike describe some data bases and analyses used to parameterize the spatial black duck model.  We used re-analysis of age ratio data based only on Canadian harvest, corrected for differential vulnerability using direct recoveries of adults and juveniles in Canada, to estimated production rates in the 2 breeding areas. Analysis provided preliminary estimates of production rates (based only on 5 years of data in the 1990s) suggesting that production is higher in the eastern area (B1) than in the western area (B2). Thus, a new intercept term was added to the production model to account for this apparent geographic difference in average reproduction rates. Otherwise the model is the same as under the single population model.

Examined distribution from breeding areas / northern harvest areas (B1/H1N and B2/H2N) to southern harvest areas (H1S, H2S) using direct band recoveries. Analysis suggested greater tendency (up to 0.5 or higher) for B2 birds to move to H1S (s. Atlantic Flyway) than for B1 to go to H2S. Estimates are now encoded in transition matrices in the spatial version of ASDP but are subject to revision following more detailed analyses.

Attempted to estimate fidelity rates (defined as probability that a bird breeding in a specific area returns to breed in that area the following year), essential components of the spatial model, using combination of direct and indirect recoveries. So the approach has limited success (estimates are not believable) and for now we are simply assuming 0.9 fidelity rates for all areas, ages, and sexes. This is a high priority to get right.

A: This is an intriguing possibility and bears looking into.
 

Analysis of US and Canadian Harvest Regulations for Use in AHM modeling

The intent of the analysis was to use annual regulations data (i.e. season length, and bag limits) and create distributions of realized harvest rates under the 3 regulatory packages of restrictive, moderate, and liberal for the US and Canada (recognizing that these mean different things in the US and Canada). The resulting distributions were to be used as distributions from which ASDP would randomly choose a harvest rate, adding partial controllability to our model. At the moment we have modeled our analysis after the Francis et al (1998, Journal of Wildlife Management  62:1544-1557.) analysis, by placing groups of years into a particular regulations package.

The data is very noisy, meaning that at the moment no clear indication that we can associate a particular range of harvest rates with a particular regulations package (i.e. a harvest rate of 0.10 could have resulted from any of the regulations packages). The suggestion was made that the harvest may be confounded with changes in the number of hunters over time. Thus, a liberal harvest today and a liberal harvest 15 years ago could produce vastly different harvest rates. In a subsequent analysis we will attempt to remove this extraneous variation through an analysis of covariance and re-estimate the distributions.

Download presentation in PDF format.
 

Introduced exercise for participants to choose a regulations package
 
 

Interactive harvest simulator:

The simulator is available in both Microsoft Excel and Quattro pro. It is preferable to have at least MS Office 98, and Quattro Pro version 8 or later. The simulator was based on the observed distribution of harvest rates in the U.S. and Canada over 3 historical periods, corresponding to "liberal" , "moderate", and "restrictive" season lengths and bag limits for black ducks.

Participants were asked to use the harvest simulator to attempt to solve each of the following hypothetical harvest management problems:

    1. Select regulatory packages so as to assure that harvest rate is equal in U.S. and Canada

    2. Select regulatory packages to assure that harvest is equal among all 4 harvest areas.

    3. Select regulatory packages to assure that AF harvest rate > MF harvest rate.

Each group of 3-4 participants was asked to keep track of how many years the result turned out "right" (that is, the achieved harvest rate met the stated goal, e.g., equal between the U.S. and Canada under #1) and to repeat the simulation 20 times (100 years total). The results were highly variable, both between groups for the same question, and within group among the questions. The exercise reinforced what we already suspected: that regulations result in highly variable achieved harvest rates, making fine-tuning of harvest via regulations extremely difficult.

Results of Application of Spatial AHM Model for Obtaining Optimal Harvest
 

Mike presented the results of some preliminary analyses comparing decision making under 4 approaches:

Single, uniform harvest decision

Harvest decision delineated E-W but same across international border within E or W

Harvest decision delineated N-S (i.e., Canada and US) but same E-W

Harvest decision independent among 4 harvest areas.

Java Simulation Applet for Spatial Model

A prototype of the Java simulation applet for the spatial model was previewed by Chris Fonnesbeck. The revised interface allows users to set area and population-specific harvest policy and model specific weights. The Java code must still be reviewed for consistency with the ASDP output. This will be available on the website as soon as possible. Note that the general applets not directly involving the blackduck project (i.e. density dependence, compensation, Bayesian model weighting) will be formally moved out of the blackduck website, and onto a more general site pertaining to population modelling.
 
 

Next Steps in AHM Development

Mike Conroy gave an overview of the priorities of the University of Georgia team over the next few months. They include:

Make sure ASDP and Applets ares doing what they are supposed to be doing - debug.

Analytical issues: better estimation estimation of the production function, updating the production models, introducing environmental conditions, resolve issues about overlap between breeding populations surveys (need common metric), harvest distribution/parameterization, work on survival functions, tons of work on the spatial models

Wrestling with issues of spatial delineation of populations - spatial separation of decision-making

Objective function needs to be worked on.

Have to deal with mallards

Communications

The Working Group discussed communications and agreed that the primary focus should be on internal audiences. Specific action items include:

Flyway Council Updates. The Univ. of Georgia team will try to attend the Flyway Council meetings in February (depending on class/council meeting schedules);

Q&A Summary. Dave Case will work with Mike Conroy to develop a draft Q&A on the effort for use primarily with internal audiences. The draft will be distributed to the Working Group for review. A final will be ready for distribution at the Flyway meetings;

Private section of website. It was agreed that items should be moved to the public section ASAP - the private section should be used primarily for draft material awaiting review and final approval by the Working Group. It terms of passing out the access codes to the private section, it was agreed that we shouldn't encourage use of the private section, but if someone really wanted access it was not that big of a deal; and,

Applets. Those applets dealing with "general" modeling will be moved to a generic "learning tools" site and linked to the black duck site. The black duck specific applets will be fined-tuned and posted for review by the Working Group over the next couple of months.

The next Working Group meeting is scheduled for the week of June 11-15, 2001 in Quebec City, Quebec. The specific dates and meeting length will be determined based on feedback from those not attending the workshop and on progress over the next couple of months.
 

Appendix A
 

Agenda
 

WORKSHOP ON ADAPTIVE HARVEST MANAGEMENT OF AMERICAN BLACK DUCKS
 

0815 --Welcome and Introductions
 

0830 -- Meeting objectives (Dave Case)
 

0845-0915 -- Review of single-population AHM model for black ducks (M. Conroy)
 

0915-0945 -- Schematic overview of spatial black duck model (M. Conroy)
 

1000-1015 -- Break

1015-1045 -- Data sets used to parameterize spatial model (M. Conroy)
 

1045-1115 -- Analysis of US and Canadian harvest regulations for use in AHM modeling (N. Zimpfer)
 

1115- 1200 -- Discussion(facilitated by Dave Case)
 

1200-1300 -- Lunch
 

1300-1400 -- Results of application of spatial AHM model for obtaining optimal harvest policies (M. Conroy)

1400-1500 -- Java simulation applet for spatial model (explanation, hands on demo: C. Fonnesbeck)

1500- 1700 -- Discussion and feedback from group (facilitated by Dave Case)

Appendix B
 

Project homepage: http://coopunit.forestry..uga.edu/blackduck/
 

Project reports and presentations: http://coopunit.forestry.uga.edu/blackduck/papers.html

Working Group meetings: http://coopunit.forestry.uga.edu/blackduck/meetings.html
 

More information on AHM: http://www.consecol.org/Journal/vol3/iss1/art8/