ODFW's Oregon Plan Habitat
for Coastal Basins
Oregon Plan Habitat Surveys are designed to assess all
streams within the range of coho salmon.
The target population of streams were contained within watersheds of
western Oregon draining into the Pacific Ocean south of the Columbia River. The area encompassed two Evolutionary
Significant Units (ESU’s) for coho salmon: the Oregon Coastal ESU and the
Southern Oregon/Northern California ESU. The Oregon Department of Fish and
Wildlife (ODFW) has further divided the Oregon Coastal ESU into three Gene
Conservation Areas (GCA) for coho salmon based on studies of genetic variation
and life history traits (Kostow 1995).
For fishery management the North Coast GCA was further split into a North
Coast and a Mid Coast GCA (Bodenmiller et al. 1997). These five GCAs are used as the basis for
monitoring coho habitat in Oregon coastal streams (Figure 1).
target populations of streams for the study were based upon a hydrography data
layer developed by the USGS at the 1:100,000 scale. Streams upstream of large dams that blocked
anadromous fish passage were removed from the selection frame. A random tessellation stratified (RTS) design
(Stevens 1997) was used to select potential sample site locations within the
population of stream segments. Stevens
and Olsen (1999) described the RTS survey design as applied to the integrated
monitoring of habitat, adult spawners, and juvenile salmonids for the ODFW. The advantage of the RTS selection protocol
was the selection of sites spread randomly across the landscape, better
representing habitat conditions within a GCA, and reducing overall sample
variance. In all GCAs surveyed, samples
were weighted to provide an equal number of sample sites (50).
sample sites were not surveyed. The
primary reason for not surveying a site was denial of access from
landowners. Additional sites were dropped
because they were small (<0.6 km2 catchment area), tidally
influenced, or a result of errors in the selection coverage (Table
The overall rate of access denial was higher in 2000
(12.5%) than 1999 (6%), and continued to encompass a large percent (45%) of
private non-industrial sites. As in
previous survey seasons these unsurveyed sites contribute to a bias in the final
dataset. Historically, private
non-industrial lands have had the lowest habitat quality (Thom et al.
1999). Given the lower quality habitat
that was observed on private non-industrial lands in the past, and the high
percentage of these sites that have been unsurveyed between 1998 and 2000, all
findings provide a biased estimate of conditions for private non-industrial
ownership as well as the coast as a whole.
Table 1: Summary table of surveyed and not surveyed sites for 2000
*Target sites are defined as sites selected in the annual sample draw that
**Non-Target sites are sites selected in the annual sample
draw that were incorrectly identified on the base coverage and are not
surveyable. These sites include sites located in tidal areas, on small streams
(upstream catchment area of <0.6 km2 ) or are the result of an
error on the GIS coverage.
***P/A = Presence/Absence surveys using
Channel habitat and riparian surveys were conducted as described by Moore et
al. (1997) with some modifications. Modifications to the survey methods
included: survey of stream lengths of only 500-1000 m and measurement of all
habitat unit lengths and widths (as opposed to estimation). Ten percent of the
sites were resampled with a separate two-person crew. Repeat surveys were a
randomly selected sub-sample from each geographic area and survey crew. The
repeat surveys were intended to measure within-season habitat variation and
differences in estimates between survey crews.
Fish presence/absence surveys using electofishing were conducted at habitat
sites outside of known coho salmon distribution in all GCAs. A complete
description of the methods used is contained in ODFW (1998). A coordinated but
separate project with