- Models were developed that described cohort strength of female
Dover sole (Microstomus pacificus) and female English sole (Parophrys
vetulus) in stocks off the Columbia River as functions of environmental
variables. The environmental variables considered were those
that could hypothetically influence spawning success or survival.
They included a spawning power index, monthly mean measurements of
different oceanographic and Columbia River factors, and short-term
measurements of weather variability. Cohort strength was calculated
by using Pope's cohort analysis.
For Dover sole, only upwelling in June and July, which may regulate
food availability following yolk sac absorption, and offshore
divergence the next December and January, which may regulate the
location of larval settling, explained significant cohort strength
variation. A model incorporating these two variables explained 64.7
percent of cohort variation. Although the Dover sole stocks were
clumped around the Columbia River, it appeared that Columbia River
factors had less influence on recruitment than did these two oceanographic
factors. Spawning power was also not significantly related
to cohort strength over the range observed.
For English sole, the factors that were associated with the
great success of the 1961 cohort were apparently unique to that
cohort; thus, they could not be examined statistically. From data
examined here, however, it appeared that unique weather variability,
characterized by high storm frequency but low average wind speed, may
have been partly responsible. For the other cohorts, variation in
early fall (prespawning) values of upwelling, barometric pressure,
and sea surface temperature explained 72.8 percent, 83.8 percent, and
49.0 percent, respectively, of the cohort variation. During upwelling
these three factors are associated with bottom temperature; therefore,
it was proposed that bottom temperature variation, which may regulate
time of spawning or egg condition, was the factor most directly linked
to English sole cohort strength. Colder bottom temperatures were
associated with stronger cohorts.
Principal components of the environmental variables were derived
and regressed against recruitment. The first two principal components
explained 65.9 percent of the variation in Dover sole recruitment.
For English sole, the third principal component explained 76.9 percent.
Because many environmental variables probably influence recruitment,
and because every environmental variable considered contributes
to each principal component, models incorporating principal
components could predict recruitment more accurately than models
which include only one or two environmental variables.