5. Risk Assessment of a Proposed Introduction of Pacific Salmon
69
tigator within a subregion of the same research area providing the data on straying
rates. Strays to a given stream were tabulated across recovery years. The straying
data used for Chinook Salmon are geographically comprehensive in that every
hatchery and all the spawning grounds available to Chinook Salmon in the Columbia River basin were surveyed. Furthermore, out-of-basin strays could be (and
were) reported in the coded wire tag database (Quinn and Fresh 1984; Quinn,
personal communication). Thus, the Chinook PDF can be considered an accurate
representation of the distribution of strays from a single natal stream in 4 brood
years.
Probability density functions for Coho Salmon and Steelhead Trout were
derived from the same studies used to derive the respective stray rate distributions.
These studies were chosen for development of the straying component of the
model because, among all the literature reviewed, they provided the most comprehensive assessments of stray distribution and abundance. In all three species,
strays entered a very small subset of the streams and rivers available to them. This
observation is consistent with results from other less comprehensive studies of
straying.
The PDF for Coho Salmon was derived from imprinting experiments conducted in Lake Michigan (Hasler and Scholz 1983). For each of four experimental
treatments, streams were ranked by the number of strays they received. The PDF
was then formed by summing the number of strays in streams with the same
ranking.
For Steelhead Trout, data from 4 return years and two natal streams on Vancouver Island (Lirette and Hooton 1988) were used to approximate the distribution of strays among streams. For each natal stream, strays were tabulated by the
stream in which they appeared, without regard to year. Two separate rankings
were then made based on the number of strays received from a single natal stream.
Counts from streams having the same ranking were summed to give the PDF. The
normalized PDFs for the three species are presented in Table 5.2.
Results and Discussion
Adult Chinook Salmon would return to the Delaware River from 3 to 10 years
after hatching of the first hatchery cohort (Fig. 5.3). Chinook stray densities were
the highest among the three species examined. Stray densities reached a maximum during years 5–7 and subsequently declined. This is a consequence of the
combined effect of five stocked cohorts reaching and then surpassing ages of peak
return (see Table 5.1). Based on the mean return and stray rates, 240 fish were
expected to return in the peak year, and 43 would be expected to stray. The
probability of 50 or fewer strays appearing in a single stream is approximately
85%. If only females are considered, there is a 90% probability of 25 or fewer
females appearing in a single stream (Fig. 5.4). Given the model assumptions,
there is a 99% probability that the proposed stocking program would produce
fewer than 100 strays in a single stream in a single year.
69
tigator within a subregion of the same research area providing the data on straying
rates. Strays to a given stream were tabulated across recovery years. The straying
data used for Chinook Salmon are geographically comprehensive in that every
hatchery and all the spawning grounds available to Chinook Salmon in the Columbia River basin were surveyed. Furthermore, out-of-basin strays could be (and
were) reported in the coded wire tag database (Quinn and Fresh 1984; Quinn,
personal communication). Thus, the Chinook PDF can be considered an accurate
representation of the distribution of strays from a single natal stream in 4 brood
years.
Probability density functions for Coho Salmon and Steelhead Trout were
derived from the same studies used to derive the respective stray rate distributions.
These studies were chosen for development of the straying component of the
model because, among all the literature reviewed, they provided the most comprehensive assessments of stray distribution and abundance. In all three species,
strays entered a very small subset of the streams and rivers available to them. This
observation is consistent with results from other less comprehensive studies of
straying.
The PDF for Coho Salmon was derived from imprinting experiments conducted in Lake Michigan (Hasler and Scholz 1983). For each of four experimental
treatments, streams were ranked by the number of strays they received. The PDF
was then formed by summing the number of strays in streams with the same
ranking.
For Steelhead Trout, data from 4 return years and two natal streams on Vancouver Island (Lirette and Hooton 1988) were used to approximate the distribution of strays among streams. For each natal stream, strays were tabulated by the
stream in which they appeared, without regard to year. Two separate rankings
were then made based on the number of strays received from a single natal stream.
Counts from streams having the same ranking were summed to give the PDF. The
normalized PDFs for the three species are presented in Table 5.2.
Results and Discussion
Adult Chinook Salmon would return to the Delaware River from 3 to 10 years
after hatching of the first hatchery cohort (Fig. 5.3). Chinook stray densities were
the highest among the three species examined. Stray densities reached a maximum during years 5–7 and subsequently declined. This is a consequence of the
combined effect of five stocked cohorts reaching and then surpassing ages of peak
return (see Table 5.1). Based on the mean return and stray rates, 240 fish were
expected to return in the peak year, and 43 would be expected to stray. The
probability of 50 or fewer strays appearing in a single stream is approximately
85%. If only females are considered, there is a 90% probability of 25 or fewer
females appearing in a single stream (Fig. 5.4). Given the model assumptions,
there is a 99% probability that the proposed stocking program would produce
fewer than 100 strays in a single stream in a single year.