| Summary: | Including ecological objectives within strategic planning on forested lands is important
because timber harvesting can reduce the value of these objectives. Harvesting is a
valuable source of economic revenue, but changes the age-class structure of the forest,
often significantly reducing the amount of late-seral stands. Late-seral stands help meet a
wide variety of objectives, such as biodiversity, water quality, and, recreation. From an
economic perspective, proper management of late-seral stands often is necessary to
acquire "social license" for continued harvesting operations.
Most forested lands in North America are managed under some form of multiple-use, and
there are many decision support tools available that integrate timber harvesting and serai
objectives. However, due to the conflicting requirements of harvesting and some
ecological objectives, there is growing evidence that some form of zoning may be a more
efficient land-use method than multiple-use. I investigated questions of how best to
define, distribute and maintain objectives requiring intact forest stands, focusing on the
creation and use of decision support systems for zoning.
I first demonstrate a decision support system for landscape-level zoning that uses site
attributes to create large (>5,000 ha) static zones. The Zone Allocation Model (ZAM)
uses the Simulated Annealing algorithm to allocate areas into zones defined around the
intensity of harvesting: Old Growth zone, Habitat zone, and Timber zone. Important
ecological criteria, such as ecological representation, size, and shape of "reserves" in the
Old Growth zone, are optimized relative to criteria that influence economic returns, such
as site productivity and ownership, in the Timber zone. On a 1.2 million hectare landbase
from coastal British Columbia, the Z AM model found solutions within 1.7% of the
calculated optimum level. I then demonstrate a decision support system for small-scale
zoning that uses stand attributes to reserve serai patches. The Saltus model uses
simulation algorithms to create dynamic zones that move around on the landbase as
disturbance creates the need to replace previously reserved stands. Saltus is
demonstrated on a 139,966 hectare landbase from coastal British Columbia, as well as on
computer generated landbases. The small-scale zoning method is shown to separate
reserve and harvesting objectives, increasing operational flexibility.
Planning approaches to older serai stands always involves some definition of these stands
and discrimination from younger stands less able to sustain late serai objectives. I
address how serai stages are defined within the context of harvest scheduling models, and
introduce fuzzy sets as a method of modelling serai constraints. The Serai Constraint
Harvest Scheduler (SCHS) is used to compare the effects of traditional and fuzzy serai
constraint methods for calculating sustainable harvest volumes for strategic planning.
When compared with traditional serai definitions at the stand level, fuzzy definition
methods are shown to better correspond with how stands develop in a serai trajectory. At
the landscape level, the fuzzy serai definition produced a smoother accumulation of serai
lands, resulting in a more uniform level of constraint and harvest availability. All planning occurs in the face of uncertainty, but different approaches to projecting
harvest are differentially responsive to this uncertainty. As a final task, I created a
robustness test to evaluate the sustainability of projected harvest volumes. Most strategic
planning tools use a harvest schedule to illustrate that at least one projected sequence of
harvests can sustain the projected harvest volume. The robustness test adds a second
dimension to sustainability by measuring the amount of change the harvest schedule can
withstand during the implementation process while still maintaining the projected volume
flow. Results using both simulation and optimization models indicate the projections
have very little robustness when using a maximum sustainable volume objective.
Reducing the target volume increases robustness, but only after a large cost to timber
production. Matching the level of uncertainty in the planning environment with a
corresponding level of robustness in projections is an important factor in creating
sustainable forest management plans.
|
|---|
