Research
PNW
Funded Projects
2001
2000
1999
1998
1997
1996
1995
|
PNW
Project Overview 1995
Land
Use Impacts on Salmon Abundance
Principal
Investigators:
Dr. William Jaeger, Williams College, Department of Economics
Dr.
David Fluharty, University of Washington, School of Marine Affairs
Awarded: $50,000
Project
Description
The
impact of changes in land cover and land use on spawning habitat and anadromous
fish abundance is one of the central issues and sources of conflict both
among different commodity production industries (timber, agriculture,
salmon), and between commodity production and the maintenance of
ecological values on the Olympic Peninsula. Despite the importance of
understanding the causal relationship between land use and salmon
abundance, there is currently no broadly applicable empirical estimate of
the effect of changes in land use on salmon abundance. The goal of this
study is to provide an empirical basis of impact of changes in land use on
anadromous fish abundance using cross-sectional statistical analysis of
salmon abundance, riverine characteristics, riparian border
characteristics, and watershed characteristics.
This
analysis would draw on satellite image/GIS data combined with salmon
abundance data from Oregon, Washington, British Columbia and Alaska. The
results would then be applied to two watersheds on the Olympic Peninsula
for the purpose of assessing: a) the relative impact of land use versus
other factors on salmon abundance, b) the expected impact of restoration
activities, and c) estimates of the "potential" salmon
abundance. The goal is to provide communities and local decision makers
with a stronger basis on which to evaluate current and future
alternatives, and to assess the benefits and costs of alternative courses
of action.
Project
Status:
Final report received
June 4, 1999
Final
Report Abstract:
Wild salmon abundance data were collected for over 600 rivers between
Alaska and Northern California. Data
on forest condition were compiled based on satellite image/GIS databases.
In the end, only 63 watersheds had sufficiently complete data to be
evaluated. The statistical inference was based on multivariate regression
analysis of cross sectional data as a ‘space for time’ alternative to
long-term studies. Independent
variables included watershed specific forest area and land areas by forest
classification and by geological classification, as well as longitude of
each watershed’s river mouth.
The
results indicate a statistically significant positive association between
undisturbed forest and salmon abundance, with coefficients of 0.60 and
0.67 salmon per hectare based on regression analysis involving two
different databases. Geological
characteristics of the watershed were also quite significant.
Salmon abundance was higher in watersheds with Mesozoic accreted
terrains as compared to the other two types, granitic and sedimentary.
Higher salmon abundance was also shown to be associated with higher
latitude.
|
