Spatial AHM

1
Spatial AHM
2
Overview of Spatial AHM

• Rationale for spatially-stratified AHM
• Biological stratification vs. operational
  stratification
• Issues
• Data demands
• Management complexity
• Connection to other species
• Computation

3
Breeding and Harvest Areas
4
Rationale for spatial AHM

• BIOLOGICAL
• Distinct gradient W-E in abundance, environmental
  factors
• Mallards!
• Perceived differences in tolerance to harvest
• Operational
• Existing political/ administrative units

5
Unifying themes

• Common resource goal
• Sustainable harvest
• NAWMP
• International Harvest Strategy

6
Dis-unifying themes

• Different resource goals?
• Notions of equity/ parity
• Different levels of belief in causative factors

7
Accomodation

• Common goals objective function
• Equity/ parity constraints
• Disparate beliefs alternative models

8
Issues (I)- Data demands

• Multiple strata information about spatial
  aspects of populations
• Multiple strata less precision for same about of
  survey effort

9
Issues (II)- Management complexity

• More harvest units
• Better ability to allocate harvest
• Better (maybe) control over impacts on specific
  stocks
• More harvest units
• More data demands
• More regulatory complexity
• Overlapping impact
• Non-unique (tied) decisions

10
Issues (III)- Other species

• Mallards
• Strong likelihood of mallard impact through model
• Mallard dynamics affect black duck policy
• Mallard dynamics affected by mallard harvest
• Strategies
• Harvest of mallards for black duck objective
• Harvest of black ducks conditioned on optimal
  mallard strategy (or vice versa)
• Joint bd-mall harvest for bd objective
• Joint bd-mall harvest for bdmall objective

11
Issues (IV)- Computation

• Adding species, space EXPONENTIALLY adds
  dimensionality to DP models
• Dramatic increase in time to find stationary
  policies if even possible
• Alternative near-optimal approaches may be
  necessary

12
Summary

• Adding spatial aspects increases biological
  realism but there are costs
• Management control may be achievable at a
  somewhat finer scale but there are limits and
  costs
• Incorporating other species is needed but further
  adds to complexity
• Computation of optimal policies will be challeng

13
The real question

• We can build it will you use it?