The long-term
objective of this research is to identify potential trade-offs between
forest management strategies and acceptable bounds of aquatic stream
habitat conditions in coastal British Columbia. The question is not
only how much we harvest in these coastal watersheds, but how and where
we log. Cumulative effects of forest harvesting over many years are
considered one of the major threats to salmon populations. Productive
freshwater habitat can play a critical role in aiding salmon populations
during periods of unfavourable ocean conditions. Preservation of aquatic
habitat is thus a critical link in preserving native salmon stocks.
I have
developed an integrated modelling framework to qualitatively contrast
effects of forest management on channel morphology and fish habitat,
using the example of coho salmon (Oncorhynchus kisutch), in small and
intermediate streams. I attempt to represent the major coastal watershed
processes and forest management activities. The model uses operationally
available ecological information about forest stand dynamics. It simulates
storm peakflow events that drive the system. Debris slides act as the
input mechanism of coarse sediment into the channel network and bedload
transport in channels is simulated. The recruitment of large wood into
channels from hillslopes, riparian zones, and upstream channels, and
the dynamics of log jams as critical elements for channel morphology
and aquatic habitat in coastal streams are also simulated. Changes in
channel morphology are tracked and coho salmon habitat capability of
the channels is rated. Forest harvesting is simulated to produce diverse
cutting patterns across the landscape in terms of harvest volume, spatial
and temporal pattern, cutblock size, yarding system and road network.
I applied the pixel-based, stochastic modelling framework to a watershed
on Vancouver Island. I use this heuristic simulation as an experimental
platform to explore alternative management strategies by addressing
'what if' questions.
The modelling
framework produces expected trends in regard to log jam numbers, bedload
yield, and coho salmon habitat capability rating. In the absence of
riparian buffers log jam numbers decrease with increasing harvest volume.
Bedload yield increases with increasing debris slide rates and decreasing
log jam numbers. Coho salmon habitat capability rating tends to decrease
with decreasing log jam numbers. Thus, if we want to create a forest
sustainable in terms of productive fish stream habitat, forest resource
managers have to plan to provide large wood from headwater streams through
rivers.
More work
must be done in subsequent research to better parameterize this modelling
framework. Overall, the system dampens most effects of parameter changes.
Results are very sensitive to assumptions about the storm threshold,
initial average stand age, frequencies of channel morphology states,
maximum channel depth, log jam parameters and the maximum passable channel
gradient for coho salmon. Due to the steepness of the channel network
in the watershed, channels were degraded rapidly and the habitat capability
rating for coho salmon was low.
forest planning, simulation, decision
support tool, habitat supply model, land use practice, harvest scheduling,
riparian buffer, aquatic resources, fish habitat, channel morphology,
stream morphology, coho salmon, hydrology, debris flow, large woody
debris, LWD, large wood dynamics, ForestWerks Model, ForestWerksModel,
Forest Werks Model, ForestWorks Model, ForestWorksModel, Forest Works
Model, fish-forestry interactions, forestry-fish interactions, fish
forestry interactions, forestry fish interactions