| Summary: | Eight scale characters of known hatchery and wild coho salmon
(Oncorhynchus kisutch) were compared, and a linear discriminant
function was used to determine if hatchery and wild adult coho
salmon could be correctly identified by their scales. Eighty-two
percent of the hatchery and 89% of the wild fish were correctly
identified. Based on analysis of scales from adult salmon of
unknown origin (hatchery or wild) and the estimated catch of
hatchery coho (marked) taken by the Oregon sport fishery,
concluded that 74.9% of the fish caught in the ocean from mid-June
to mid-September 1977 had been released as smelts from hatcheries.
Percentages of hatchery fish in the catch ranged from 85% near the
mouth of the Columbia River to 61% at Winchester Bay on the southern
Oregon coast. Fisheries on the south and central Oregon coast may
have fished on higher percentages of wild coho salmon later in the
season, probably because wild fish from coastal streams remained
off of these ports while most fish destined for Columbia River
hatcheries had already migrated northward.
Scales from coho salmon were used to determine if location on
the body from which they were taken would affect the values of five
scale characters. Scales obtained from within a relatively small
area above the lateral line between the dorsal and adipose fins
differed widely in total radius, in radius of the freshwater zone,
and in number of circuli in the freshwater zone. Scales taken
farther above the lateral line had significantly lower values for
all five characters observed. I conclude that substantial error
can be introduced into interpretation of scale data if care is not
taken to insure that scales from each fish come from precisely the
same area of the body. A well chosen scrape sample yielded a
result as satisfactory as that of a "preferred" or "key" scale.
To investigate the relationship between streamflow and abundance
of coho salmon, I correlated flow from several Oregon coastal
rivers with catch of coho salmon from these rivers and with catch
from the Oregon commercial troll fishery. I found a highly significant
relationship between total streamflows during the freshwater
residency of the fish for five Oregon coastal rivers combined and
pounds of adult coho salmon caught by the Oregon commercial troll
fishery from 1942 to 1962. There is also a significant relationship
between total combined annual (January-December) flows for these
rivers and a catch 2 years later. Conversely, I found a poor
relationship between the lowest 60 consecutive days of summer flow
and 2 two years later. I also found significant relationships
between annual flows and catch in Tillamook Bay from 1934 to 1946.
Only on the Siletz River from 1927 to 1940 do find a significant
relationship between summer flows and catch. Higher flows during
the freshwater stages of coho salmon probably provide more habitat
and better conditions for growth as well as lessen susceptibility
of fry and smelts to predation. I concluded that the relationships
I found should probably not be used now to predict abundance of
wild coho salmon because of (l) the unknown interaction between
wild and hatchery fish, and (2) the preponderance of hatchery fish
in the catch. === Graduation date: 1979
|
|---|
