Page 135

ANOTHER YELLOW PERCH POPULATION DECLINE IN THE 1990s:
THE LAKE ST. PIERRE CASE STUDY, ST. LAWRENCE RIVER, QUEBEC
Yves Mailhot, Faune et Parcs Quebec, 1579, boul. Louis-Frechette, Nicolet
(Quebec), Canada J3T 2A5,
yves.mailhoti)vfapag.gouv.qc.ca and Pierre Dumont, Faune et Parcs Qudbec,
201, Place Charles-Lemoyne, 2e etage
Longueuil (Qudbec), Canada J4K 2T5
Introduction. Yellow perch sport and commercial
catch in Lake St. Pierre, a shallow 36 000 ha
enlargement of the St. Lawrence River, increased
between the late 1970s and the 1990s. Total annual
catch was estimated at about 300 t in 1986 (declared
commercial fishing, 220 t; sport fishing, 80 t). This
perch population is heavily exploited: growth rates
and fecundity are high; age at sexual maturity is low;
annual total mortality has been consistently high
since 1978 and was estimated at 77 % (natural
mortality: 25 %) between 1986 and 1991 while it was
only 35 % at the same period in Lake St. Louis,
another freshwater enlargement of the river 100 kan
upstream where there is no commercial fishery
(Dumont, 1996). In the mid 1990s, Gudnette et al.
(1994) predicted that population failure would
happen if consecutive low year-classes were
produced without reducing fishing effort and catch,
using Thompson and Bell yield per recruit model and
virtual population analyses based on data from yearly
commercial catch sampling.
Afin .
86 88 90 92 94 96 98 00 02
Figure 1. Declared annual commercial catch (t) of
yellow perch in Lake St. Pierre between 1986 and
2002.
Between 1994 and 1996, the total annual declared
commercial catch decreased from 262 t to 140 t
(Figure 1). In 1996, a management plan was designed
to reduce by 12 % the total annual mortality rate. A
165 mm minimum legal length was imposed for sport
and commercial fisheries; sport fishermen were
allowed a 50 fish bag limit and could no longer sell
their catch; commercial fishermen were limited to
only one helper. In 1997 and 1998, the annual
declared commercial catch decreased again to 67 t
and 80 t. Over and under declarations naturally occur
in voluntary commercial catch reports but we think
that the importance of the decrease reported between
1994 and 1998, combined to the contents of the
discussions we had with the fishermen during this
period, indicate the correct order of magnitude of the
population failure since no management changes
occurred. After 1999, they can no longer represent
the abundance of the resource since the fishing effort
was reduced. In order to evaluate the efficiency of the
management plan, we realized a four consecutive
years (1997-2000) survey of the perch population to
determine if the total annual mortality rate had
changed.
Methods. Since over 8 years old perch proved to be
rare in the past studies, we designed our protocol to
sample adequately the most abundant segment of the
population in order to produce sex-separated catch
curves between ages 2 and 7. We used the results of a
1995 experimental gill-net sampling of the Lake St.
Pierre where a set of two experimental multimesh
gill-nets (mesh sizes 1, 1.5, 2, 2.5, 3, 4, 5 and 6
inches stretched mesh) had been fishing for 24 hrs at
94 stations. In order to optimize the results for older
than age 2 perch, we excluded the I inch mesh size
and kept 1.5 to 3 inch mesh. We then choose the
stations with the most significant perch catch per
effort well spread all over the perch habitat after we
made sure the length structure of the perch from the
selected stations and the rest of them was the same.
The final length of the mesh panels (50 ft) was
chosen according to the results of simulations to
determine the number of perch needed to sample at a
95 % probability level all the present age groups and
to reduce the confidence interval of the distribution
of the simulated mortality rates.
The sampling took place at 27 stations. The
samples were corrected for gill-net size selectivity
and encounter probability as proposed by Kraft and
Johnson (1992) and Spangler and Collins (1992) in
order to represent more accurately the size and age
structure of the perch population. The total annual
mortality was estimated by Ricker's catch curve and
Robson and Chapman methods with the corrected
numbers of perch at age for males and females. The
samples for both sides of the lake were grouped to
allow the temporal comparison with the 1986-1991
period.
Results. Table I presents the total instantaneous
mortality rates (Z), calculated for each sex and year
by both methods as well as a mean Z for the four
years period, included to minimize the ever-present
impact of the year-class strength variation on the
estimation of this parameter. The overall global mean
Z (1.402) corresponds to a total annual mortality rate
(A) of 75 % which cannot be considered as different
from the 77 % 1986-91 period mortality rate.
Discussion. Our results show that the 1996
management plan failed to reduce the total annual
mortality rate over the whole period studied. The year
2000 lower values could theoretically result from
135
vuu
250
200
150
100
50
0