simulation of landowner behavior and potential forest policies at a large scale in coastal oregon
DESCRIPTION
Simulation of landowner behavior and potential forest policies at a large scale in Coastal Oregon. Pete Bettinger Warnell School of Forest Resources University of Georgia. Marie Lennette Department of Forest Resources Oregon State University. Overall Simulation Process. Tools and data - PowerPoint PPT PresentationTRANSCRIPT
Simulation of landowner behavior and potential forestpolicies at a large scale in Coastal Oregon Pete Bettinger
Warnell School of Forest ResourcesUniversity of Georgia
Marie LennetteDepartment of Forest Resources
Oregon State University
Existing forest
inventories
Managementintentions
Prices andcosts
GIS databases
Land usepattern
Land usechange
Stand structure
projections
LAMPSResponsemodels
Timber volumeand value
Habitat conditionfor focalspecies
Successional stages
Recreationopportunities
Aquatic habitat /watershedpotential
Landslide / debris flow
Employmentand income
Synthesis of effects of alternative
managementscenarios
Tools and data
for policy
analysis
Policy guidance
Overall Simulation Process
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
State-managed forest landPrivate industrial forest landOther ownership classes
Studyarea
Landscape Modeled
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Basic simulation units
Management units
Harvest blocks
Interior habitat blocks
Land ownerships
Landscape
Hierarchical Spatial framework
Softwood
Hardwood
Mixed forest
Management unit boundary
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Basic simulation units
Management units
Harvest blocks
Interior habitat blocks
Land ownerships
Landscape
Hierarchical Spatial framework
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Seeks to emulate management behavior of four major landowner groups represented in the Oregon Coast Range: federal, state, forest industry, non-industrial private.
Facilitates an evaluation of forest policies across large areas, and over a long time frame (100 years).
Merges strategic and tactical planning processes.
Formulates problems as "Model I."
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
LAMPS simulation model
Interface: VB 6.0
Schedulingmodel: C
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
LAMPS simulation model
Four types of scheduling processes can be used.
FederalStateForest IndustryNon-industrial private
The processes include spatial harvest scheduling aspects, yet theoverall goal is to emulate management behavior across a broad scale.
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
“Federal scheduling process”
Monte Carlo approach to scheduling clearcuts.No clear objective function to optimize. Constraints:
A maximum percentage of “matrix” land area can be clearcut in any one 5-year time period.
A minimum amount of “older forest” within a watershed must be present before any clearcuts can be scheduled in that watershed.
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
“State scheduling process”
Monte Carlo approach to scheduling clearcuts.Objective is measured by the achievement of maximum even-flow of timber harvest volume, using binary search.
Constraints:
Achievement of structural conditions by State management District.
Maintain a distribution of sizes of Interior Habitat Area (IHA) patches (older forest conditions spatially connected).
Unit restriction adjacency.
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
“Forest Industry scheduling process”
Objective can be measured by the achievement of maximum even-flow of timber harvest volume, using binary search, or target harvest levels can be set by the user as goals to achieve.
Constraints:
Attempt to emulate a historical clearcut size distribution by “blocking” management units together for harvest using a dynamic process (block patterns are not fixed).
Area restriction adjacency.
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
“Non-Industrial Private scheduling process”
Harvest probability approach to scheduling clearcuts.No clear objective function to maximize, although target harvest levels can be set by the user as goals to achieve.
Constraints:
Attempt to emulate a historical clearcut size distribution.
Area restriction adjacency.
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Riparian policies:
1) No harvest within Oregon Forest Practices Act buffers.2) Allow partial cutting within Oregon Forest Practices Act buffers, to the extent allowed by the Act.3) Choice #2, yet no harvest of hardwoods within 100 feet of a stream.4) Choice #1, and no harvest of hardwoods within 100 feet of a stream.
No harvest in FPA buffers
No harvest of other hardwoods
within 100 ft
No harvest in FPA buffers
Harvest of other hardwoods
within 100 ft
Harvest in FPA buffers
Harvest of other hardwoods
within 100 ft
Harvest in FPA buffers
No harvest of other hardwoods
within 100 ft
1 2
34
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Leave tree strategies:
1) Leave Oregon Forest Practices Act minimums (2 small TPA).2) Leave 2 large TPA.3) Leave 5 TPA, according to State Forest Plan guidelines.4) Leave 14 TPA according to State Forest Plan guidelines.5) Leave clumps of trees (BSUs, as a percentage of land area) uncut.
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Transitionprobabilities:
After clearcutting, each BSU within a clearcut unit has a probabilityof being regenerated as one of four types: open, hardwood, mixed, or conifer.
Probability = f (previous forest type, owner, ecoregion, distance from stream)
LAMPS simulation model
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Simulation notes:
Projections are for 100 years, with 5-year time steps.
All ownerships must be classified as either federal, state, forest industry, or non-industrial private.
Classification of owners and owner policies is flexible.
Forested area in the landscape modeled: 1,457,602 acres
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
LAMPS simulation model
1. Base Case. Simulations tailored for the Coast Range of Oregon.
2. Extended green-up requirements, from 5 years (base case) to 10, 15, and 20 years.
3. Require leaving 10% or 20% of clearcut area in un-cut clumps. This is modeled at the BSU level.
4. Require, through a riparian policy, that no harvesting occurs within the Oregon Forest Practices Act buffers, and that no hardwoods are cut within 100 feet of the stream system.
5. Change the transition probabilities: Increase them 10% towards the successful establishment of conifers; decrease them 10% away from the successful establishment of conifers.
Policies modeled
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Basic simulation units
Management units
Harvest blocks
Interior habitat blocks
Land ownerships
Landscape
Results
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Basic simulation units
Management units
Harvest blocks
Interior habitat blocks
Land ownerships
Landscape
Results
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Results
Base case Extended green-up requirements
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Results
Base case Leave tree clump requirements
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Results
Base case Riparian policy
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Results
Base case Changes in transition probabilities
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Results
Average difference in periodic volume from the base case
10-year green-up15-year green-up20-year green-up
10% leave tree clumping20% leave tree clumping
Riparian policy
Lower conifer transition probabilitiesHigher conifer transition probabilities
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
-19.4%-26.9%-27.9%
-7.2%-14.0%
-6.9%
-6.1%+0.1%
LAMPS uses a hierarchical spatial structure to allow the modeling of large-scale, long-term forest policies for landscapes withmultiple landowners.
The framework is, of course, a simplification of a complex systemof forest management.
Recognition of management behavior, however, increases the credibility and realism of the simulations.
Discussion
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Neither the even-flow assumptions nor the constraints examined in these policies are Law.
The even-flow objective for forest industry landowners was obtained from evidence of recent behavior of the industrial group as a whole.
In the past, industrial landowners in Oregon had the ability to utilizefederal timber sales to buffer changes in the market. Since federal timber sales have declined dramatically in the past decade, it remains to be seen whether these landowners will continue to harvest a relatively even amount of timber per year.
Discussion
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
We have simulated a higher amount of volume harvested wheneven-flow objective is relaxed.
Most simulations show a build-up of standing inventory after the"bottleneck" time period that constrains achievement of even-flow.
Discussion
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
The LAMPS simulation model is a spatial simulation model of landowner management behavior, and is associated with a quantitative projection of stand structures, which as nested ina hierarchical spatial framework.
The model was designed to help policy makers and land managers “think through” forest policies before implementing them.
Summary
Center for Forest Business, Warnell School of Forest Resources, University of Georgia
Funding for the development of LAMPS was provided by:
USDA Forest Service, Pacific Northwest Research Station
College of Forestry, Oregon State University
Oregon Department of Forestry
Support for this modeling effort provided by:
College of Forestry, Oregon State University
Warnell School of Forest Resources, University of Georgia
Acknowledgements
Center for Forest Business, Warnell School of Forest Resources, University of Georgia