5. Risk Assessment of a Proposed Introduction of Pacific Salmon
73
Application of return rates and stray rates to cohorts rather than return years
reduces variation in the number of strays occurring in any year, because more than
one cohort contribute to each spawning run. Thus the model output is relatively
insensitive to the variance in cohort-specific return and stray rates. This is true to a
greater extent for Chinook and Steelhead, because in those species cohorts return
over several years, versus 2 years in Coho.
The discrete finite PDF for each species defines the distribution of strays
among streams and limits the number of streams to which fish can stray. That limit
is the maximum number of streams observed to have received strays from a single
natal stream. Missed strays and streams are unlikely, however, to seriously bias
the estimates of maximum number of strays per stream, because the number of
missed strays and the number of strays per missed stream are both probably very
low. The effect of missed streams on model output is offset by the marginal
reduction in stray rate attributable to the strays missed in those streams.
The set of values used to approximate the variance of the distribution for each
variable includes an unknown amount of measurement error, which biases upward
the variance of the model output, thereby broadening the stray density CDFs.
The proposed experimental introduction of Pacific salmonids is unlikely to
result in large numbers of stocked fish entering streams other than the Musconetcong River. The short-term nature of the experimental program and the expected
small number of strays per stream make it unlikely that stocked fish could seriously disrupt the streams into which they might stray. If any significant longterm effects arose from the experimental program, they would probably be a
consequence of successful reproduction and establishment of a self-sustaining
population.
The probability of establishment of a self-sustaining population by the limited
experimental stocking program is very low but not zero. Spawning would depend
entirely on strays—any fish returning to the carefully monitored Musconetcong
River would be trapped and removed. The numbers of strays presented above are
for entire years and streams. Successful spawning depends on the co-occurrence
in time and space of heterosexual pairs, which may not happen with small numbers of strays returning to streams over a protracted period. If successful spawning
were to occur, the eggs, larvae, and juveniles would be exposed to the suite of
mortality factors that hatchery propagation attempts to minimize or eliminate.
Wild smolts would also be more likely to migrate at a time of the year when
environmental conditions in the middle and lower Delaware River estuary are
hostile. These factors would all tend to diminish the chances that the number of
fish completing the life cycle would be sufficient to establish a viable population.
However, unlike the situation with stocked fish in which fish that return to their
natal stream would be trapped and removed, all the wild fish that did return would
be potential spawners, including the vast majority of them that would likely return
to their natal stream.
A quantitative analysis to evaluate the likelihood of establishing a selfsustaining population was beyond the scope of the draft EIS, and it is beyond the
scope of this chapter. If establishment of a self-sustaining population is unaccept-
73
Application of return rates and stray rates to cohorts rather than return years
reduces variation in the number of strays occurring in any year, because more than
one cohort contribute to each spawning run. Thus the model output is relatively
insensitive to the variance in cohort-specific return and stray rates. This is true to a
greater extent for Chinook and Steelhead, because in those species cohorts return
over several years, versus 2 years in Coho.
The discrete finite PDF for each species defines the distribution of strays
among streams and limits the number of streams to which fish can stray. That limit
is the maximum number of streams observed to have received strays from a single
natal stream. Missed strays and streams are unlikely, however, to seriously bias
the estimates of maximum number of strays per stream, because the number of
missed strays and the number of strays per missed stream are both probably very
low. The effect of missed streams on model output is offset by the marginal
reduction in stray rate attributable to the strays missed in those streams.
The set of values used to approximate the variance of the distribution for each
variable includes an unknown amount of measurement error, which biases upward
the variance of the model output, thereby broadening the stray density CDFs.
The proposed experimental introduction of Pacific salmonids is unlikely to
result in large numbers of stocked fish entering streams other than the Musconetcong River. The short-term nature of the experimental program and the expected
small number of strays per stream make it unlikely that stocked fish could seriously disrupt the streams into which they might stray. If any significant longterm effects arose from the experimental program, they would probably be a
consequence of successful reproduction and establishment of a self-sustaining
population.
The probability of establishment of a self-sustaining population by the limited
experimental stocking program is very low but not zero. Spawning would depend
entirely on strays—any fish returning to the carefully monitored Musconetcong
River would be trapped and removed. The numbers of strays presented above are
for entire years and streams. Successful spawning depends on the co-occurrence
in time and space of heterosexual pairs, which may not happen with small numbers of strays returning to streams over a protracted period. If successful spawning
were to occur, the eggs, larvae, and juveniles would be exposed to the suite of
mortality factors that hatchery propagation attempts to minimize or eliminate.
Wild smolts would also be more likely to migrate at a time of the year when
environmental conditions in the middle and lower Delaware River estuary are
hostile. These factors would all tend to diminish the chances that the number of
fish completing the life cycle would be sufficient to establish a viable population.
However, unlike the situation with stocked fish in which fish that return to their
natal stream would be trapped and removed, all the wild fish that did return would
be potential spawners, including the vast majority of them that would likely return
to their natal stream.
A quantitative analysis to evaluate the likelihood of establishing a selfsustaining population was beyond the scope of the draft EIS, and it is beyond the
scope of this chapter. If establishment of a self-sustaining population is unaccept-