203-F
North American Journal of Fisheries Management 15 :838-844, 1995
© Copyright by the American Fisheries Society 1995
Seasonal Variation in Gill-Net Sample Indexes for
Northern Pike Collected from a Glacial Prairie Lake
ROBERT M . NEUMANN1
AND DAVID W. WILLIS
Department of Wildlife and Fisheries Sciences, South Dakota State University
Brookings, South Dakota 57007, USA
Abstract .-We captured 702 northern pike Esox lucius in 416 gill-net sets on 17 monthly occasions from July 1991 to June 1993 to examine seasonal variation in sampling data in Lake
Thompson, South Dakota . Mean catch per effort (CPE) of northern pike 35 cm or longer differed
significantly among months during both years . Mean CPE was lowest in winter and peaked in
spring of both years . Size structure estimates were lowest during early spring and summer when
males dominated samples and were highest during winter when females dominated samples . Northern pike 86 cm or longer were caught primarily in winter during ice cover. Relative weight (Wr)
of male and female northern pike increased during fall, peaked in late winter, and declined during
the spawning period . A similar pattern was found for the gonadosomatic index of female northern
pike but not for males . Increases in the gonadosomatic index for males occurred entirely during
fall . Gonadosomatic index and Wr were positively correlated for female northern pike, but the
correlation was lower for males . We suspect our findings will be most applicable for interpreting
northern pike population surveys made during narrow time periods near the southern portion of
the species' range and for populations found in shallow, warmer water bodies .
Biases associated with seasonal variation in
sampling indexes such as catch per effort (CPE),
body condition, and size structure should be considered when population assessments and management decisions are based on one-time samples .
Understanding seasonal variations in these indexes
can disclose seasonal biases in fish population assessment and provide information necessary for
standardized sampling designs .
In a recent survey of conservation agencies in
the USA and Canada, gill nets were identified as
the gear most frequently used to assess length frequency and relative abundance of northern pike
Esox lucius (Fisheries Techniques Standardization
Committee 1992) . Seasonal variations in northern
pike indexes such as CPE (Casselman 1978 ; Diana
1980 ; Cook and Bergersen 1988) and body condition (Guy and Willis 1991 ; Headrick and Carline
1993) have been examined ; however, no information on seasonal variation of these indexes has
been reported for northern pike captured in gill
nets from natural prairie lakes . The objectives of
this study were to examine seasonal variations in
population indexes for northern pike collected in
gill nets from a South Dakota glacial lake and to
provide recommendations for sampling .
1 Present address : Department of Natural Resources
Management and Engineering, University of Connecticut, Storrs, Connecticut 06269, USA .
Study Site
Lake Thompson is a shallow, natural lake in
Kingsbury County in east-central South Dakota .
The lake has a surface area of 6,573 ha, a maximum depth of 7 .0 m, and a mean depth of 2 .5 m .
The lake expanded during the mid-1980s from a
4,000-ha wetland to its present size because of
unusually high precipitation during that period .
The high productivity of Lake Thompson is typical
of natural lakes in the prairie region . Lake Thompson had a morphoedaphic index (total dissolved
solids/mean depth) of 398 (Guy 1993), compared
to means of 12 and 66 for 23 natural lakes and
294 reservoirs, respectively, in North America
(Jenkins 1982) .
Methods
Northern pike were collected monthly on 17 occasions in July-October and December 1991, February-August, October, and December 1992, and
February, April, and June 1993 . Collections were
made with clear monofilament gill nets, 45 .7 m
long by 1 .8 m deep, composed of 7 .6-m panels
with bar mesh measurements of 19, 25, 38, 51, 70,
and 89 mm . Four nets were set near shore during
daylight hours for approximately 2-4 h near the
middle of each month when samples where taken .
Most nets were fished in the same area of the lake
each month ; however, we sometimes had to relocate nets to avoid anglers and problems with strong
wind and waves . Relocated nets were set in other
nearshore areas similar to where they were usually
set. Water depths at net locations ranged from
about 1 .5 to 3.0 m. Nets were reset until 50 northern pike were caught or until the midmonth sampling period elapsed. The duration of each gill-net
set was recorded to estimate CPE as catch per nethour. Captured northern pike were immediately
measured to the nearest millimeter (total length),
placed in plastic bags, packed on ice, and returned
to the laboratory at South Dakota State University,
where total body weight (nearest 1 g for fish less
than 1 kg ; nearest 5 g for fish greater than I kg),
sex, and gonad weight (nearest 0.1 g) were recorded .
Water temperature was measured on each sample date throughout the study period with a Yellow
Springs Instruments model 51B oxygen-temperature meter. Vertical temperature profiles (0 .5-m
increments) were taken at a control site in the
deepest part of the lake .
Mean CPE (number per gill-net-hour) of stocklength (S, >_35 cm) northern pike was calculated
for each sample period . Size structure was quantified with proportional stock density (PSD ; Anderson 1976) and with relative stock density of
fish of preferred length (RSD-P) and memorable
length (RSD-M). Proportional stock density is the
percentage of S-length fish that are also quality
length (Q, >_53 cm), RSD-P is the percentage of
S-length fish that are preferred length (P, >_71 cm),
and RSD-M is the percentage of S-length fish that
are memorable length (M, >86 cm) (Gabelhouse
1984 ; Willis et al . 1993). Confidence intervals
(95%) for stock density index values were determined by using the method of Gustafson (1988)
and were provided as a descriptive measure of
variability.
Relative weight (Wr = 100 x individual fish
weight/standard weight ; Wege and Anderson
1978) was calculated to index condition of northern pike . Relative weights were calculated with
Willis's (1989) standard weight (Ws) equation for
northern pike : Iog10Ws(g) = -5 .369 + 3.059
log10 TL(mm) . Two-way analysis of variance
(ANOVA) was used to test for differences in Wr
among S-Q, Q-P, and P-length northern pike
among months .
The gonadosomatic index (GSI ; 100 x gonad
weight/total body weight) was used to measure
gonad maturity . Mean GSI values were calculated
according to sex and month .
Data on CPE, Wr, and GSI were tested for normality with the Shapiro-Wilk statistic and inspection of normal probability plots (UNIVARIATE
procedure, SAS Institute 1985) . Catch-per-effort
data were transformed (log10) to better meet the
assumption of normality for ANOVA, and all statistical tests were done on transformed data . Analysis of variance (GLM procedure, SAS Institute
1985) was used to test for differences in CPE, Wr
and GSI. The least-significant-difference multiplecomparison test was used to identify differences
among individual means when the overall model
was significant (P < 0.05) . Regression analysis
(REG procedure, SAS Institute 1985) was used to
identify relations between water temperature and
sire structure indices .
Catch Rates
Results and Discussion
We captured 702 northern pike in 416 gill-net
sets . Mean CPE of S-length northern pike was significantly different among months during both
years (1991-1992, P < 0.01, F = 13 .50; 19921993, P < 0 .01, F = 5 .58) . In 1991-1992, CPE
peaked during summer and early spring (Figure 1) .
In 1992-1993, a similar spring peak in CPE was
observed, but no summer peak was detected . Mean
CPE values were lowest during winter of both sampling years.
Peaks in CPE of northern pike in Lake Thompson during spring were most probably associated
with spawning activity . Casselman (1978) and Diana (1980) also found catch rates of northern pike
caught in gill nets to be higher during spring than
summer or fall . Cook and Bergersen (1988) observed increased activity of ultrasonically tagged
northern pike that was related to spawning in April
and May in Eleven Mile Reservoir, Colorado . A
spring peak in catch rates was also observed for
northern pike captured in trap nets (Guy and Willis
1991) in a South Dakota natural lake .
We believe that concentration of smaller northern pike in aquatic macrophytes near the sampling
locations boosted CPE in summer during the first
year of our study . Northern pike have been shown
to select vegetated habitats in lakes (Chapman and
Mackay 1984 ; Cook and Bergersen 1988) . Northern pike (44-63 cm total length) tagged with ultrasonic transmitters increased their use of isolated
nearshore (<200 m from shore) patches of aquatic
macrophytes in Lake Thompson during midsummer (Neumann 1994). During the second year of
our study, aquatic macrophytes were not as abundant as in the first year, and no summer peak in
CPE was observed . Therefore, the influence of
cover such as vegetation should be considered
when one assesses CPE data or chooses standard
sampling locations. Nevertheless, a lack of a summer peak in CPE during the second year may also
have resulted from growth of the dominant cohort
to larger sizes or simply from random catch variation .
Low CPE during winter may have been associated with reduced feeding activity resulting from
decreased metabolism . However, Sammons et al .
(1994) observed higher food consumption rates by
northern pike in Lake Thompson during periods
of ice cover than during other times of the year.
Significant increases in growth (length) of northern pike in Lake Thompson have also been observed during winter (Neumann et al . 1994) . Alternatively, reductions in gill-net CPE during winter may have been related to the observed increase
in water transparency (making the nets more visible to fish) or to increased stiffness of the net in
cold water.
Size Structure
Seasonal trends in size structure of northern pike
captured in gill nets were apparent during both
years (Figure 2) . Proportional stock density, RSDP, and RSD-M were highest during winter. Size
structure was lowest during early spring and sum-
mer. Catches of M-length northern pike occurred
primarily during winter.
Proportional stock density, RSD-P, and RSD-M
were inversely related to water temperature (Figure 3) . Therefore, during coldwater periods, CPE
of northern pike captured in gill nets was low and
size structure was high . During warmwater periods, CPE increased and size decreased . Only two
of the 28 M-length northern pike captured during
this study were caught during July and August .
Seasonal differences in size structure of northern pike captured in gill nets may be attributed to
sex- and size-related differences in activity and
habitat use. Most M-length females were caught
during winter, whereas most smaller males were
caught during spawning and summer, as evidenced
by the sex ratio of captured fish (Figure 4) . Therefore, sex ratio appeared to influence size structure
of northern pike captured in gill nets . Of the 28
M-length northern pike captured during this study,
23 were captured during December and February .
These large females may be more active during
winter, seeking food to increase energy reserves
in preparation for spawning . Male northern pike
dominated gill-net samples during summer and
fall, coinciding with the period of increased testicular development. Casselman (1975) found similar trends in the sex ratio of northern pike captured
by gill nets in two regions in Ontario. In Lake
Thompson, the low catch rates of M-length northern pike during summer indicated that larger northern pike were not vulnerable to gill nets fished
during the daytime in nearshore areas . Also, large
(75-97 cm) ultrasonically tagged northern pike
(75-97 cm) moved offshore into deeper water during midday in summer, while smaller northern pike
(44-63 cm) remained near shore (Neumann 1994).
The frequency of northern pike males caught in
gill-net samples peaked during spawning . Priegel
and Krohn (1975) found that movement patterns
of males and females differed during spawning in
Gilbert Lake, Wisconsin . Males tended to move
into and out of spawning areas faster than females,
and males were the first to enter spawning areas .
The high percentage of males caught during April
in Lake Thompson may reflect greater activity of
males during spawning . Lucas (1992) also found
that acoustically tagged male northern pike were
more active than females during the spawning period in Loch Davan, Scotland .
Gonadosomatic Index and Body Condition
Mean GSI values differed significantly (P <
0.01) among months . Gonadal growth began during fall . Testicular growth was completed by the
end of fall, whereas ovarian growth continued until
the fish spawned in spring when ovarian weight
declined sharply. Thus, male GSI values were
highest during the fall, and female GSI values were
highest in spring (Figure 5) . Reduced GSI for
males during winter may have resulted from increases in somatic weight during winter, as we
explain later.
No significant differences in mean Wr among
length groups or interactions among lengths
groups and months were found ; thus, evaluation
of overall mean Wr over time was appropriate
(Murphy et al . 1991). Significant differences (P <
0.01) in mean W, values were found among months
for both males and females (Figure 6) . Both male
and female northern pike exhibited increases in Wr
beginning in fall and peaking in late winter. Guy
and Willis (1991) and Headrick and Carline (1993)
reported similar seasonal trends . The increase in
W, of males during winter and lack of increase in
GSI during the same period indicated that somatic
weight of males increased during winter. Relative
weight for both sexes declined after the fish
spawned, which we attributed to loss of both gonad
and somatic weight (Neumann 1994) . The GSI and
W, for female northern pike showed a significant
positive correlation (P < 0.01, r = 0 .68) but the
correlation was poorer for males (P < 0.01, r =
0.23), likely because of the different seasonal patterns of Wr and GSI between males and females.
However, Wr was lowest during spring and summer for both sexes.
Sampling Considerations
The CPE and size structure of northern pike
caught in gill nets varied seasonally . Catch per
effort peaked in spring of both sampling years.
Size structure of northern pike captured in gill nets
was highest during winter, when over 80% of the
M-length northern pike were caught . Size structure
was lowest during summer and during the spring
spawning period . Size structure of northern pike
captured in gill nets was influenced by the ratio of
males to females in each sample .
Understanding seasonal variations in sampling
data will enable fisheries managers to assess northern pike populations in shallow, prairie lakes more
accurately . Also, managers must understand biases
associated with sampling northern pike at different
times of the year before they design standard sampling schedules and evaluate long-term data sets .
The most reliable time for fisheries biologists to
assess northern pike populations with gill nets in
natural lakes in South Dakota appears to be during
late spring after spawning . Although the true size
structure of northern pike in Lake Thompson was
not known during our study, seasonal variability
in size structure of pike captured in gill nets stabilized during May and June, as did CPE . This
extended period may allow fisheries biologists to
sample multiple lakes over a reasonable time period without seasonal variability substantially affecting among-lake comparisons. Although Wr
was low during this time compared to its level in
other months, recognizing typical seasonal patterns in W, can help fisheries managers interpret
data that have been collected during standardized
sampling .
Because the data in this study were limited to
a single northern pike population in a South Dakota natural lake, we hesitate to assume that the
mechanisms causing these trends are the same in
other lake types from different geographic regions .
For example, in our study, northern pike longer
than 75 cm were captured primarily during winter.
Size structure was low during warmer months
when large northern pike were apparently not vulnerable to gill nets set in nearshore habitats . Large
northern pike may not inhabit nearshore areas in
natural prairie lakes during summer because they
are shallow and often devoid of aquatic macrophytes . In different lake types with deeper water
near shore and prominent annual stands of littoral
aquatic macrophytes, large northern pike may be
more likely to inhabit these areas throughout the
year, especially where water temperatures do not
restrict their habitat during summer. Although the
seasonal trends observed in our study may also
occur in other systems in different regions, the
timing and magnitude of the cycles may vary de-
pending on the local climate. We suspect that our
findings will be most applicable for northern pike
populations near the southern portion of their
range and for populations in shallow, warmer water bodies .
Acknowledgments
We thank C. Scalet, W. Duffy, and W. C. Johnson
for critical reviews of this manuscript, and S. Sammons, L. Isaak, and S . Crow for help with field
collection . Partial funding was provided by the
South Dakota Department of Game, Fish and Parks
through Federal Aid in Sport Fish Restoration, project F-15-R-1549; Electric Power Research Institute, project SFI/EPRI 91-08 ; and South Dakota
State University . This manuscript was approved
for publication by the South Dakota Agricultural
Experiment Station as journal series number 2802 .
References
Anderson, R . O . 1976 . Management of small warm water impoundments . Fisheries 1(6) :5-7, 26-27 .
Casselman, J. M. 1975 . Sex ratios of northern pike,
Esox lucius Linnaeus. Transactions of the American
Fisheries Society 104 :60-63 .
Casselman, J . M. 1978 . Effects of environmental factors
on growth, survival, activity, and exploitation of
northern pike . American Fisheries Society Special
Publication 11 :114-128 .
Chapman, C . A ., and W. C . Mackay . 1984 . Versatility
in habitat use by a top aquatic predator, Esox lucius
L . Journal of Fish Biology 25 :109-115 .
Cook, M . F, and E. P. Bergersen . 1988 . Movements,
habitat selection, and activity periods of northern
pike in Eleven Mile Reservoir, Colorado . Transac
tions of the American Fisheries Society 117 :495502 .
Diana, J . S . 1980 . Diel activity pattern and swimming
speeds of northern pike (Esox lucius) in Lac Ste .
Anne, Alberta . Canadian Journal of Fisheries and
Aquatic Sciences 37 :1454-1458 .
Fisheries Techniques Standardization Committee . 1992 .
Fish sampling and data analysis techniques used by
conservation agencies in the U .S . and Canada .
American Fisheries Society, Bethesda, Maryland .
Gabelhouse, D . W., Jr. 1984. A length categorization
system to assess fish stocks . North American Journal of Fisheries Management 4 :273-285 .
Gustafson, K . A . 1988 . Approximating confidence intervals for indices of fish population size structure .
North American Journal of Fisheries Management
8 :139-141 .
Guy, C . S . 1993 . Structure, dynamics, and movement
patterns for white crappies in South Dakota waters.
Doctoral dissertation . South Dakota State University, Brookings .
Guy, C . S ., and D . W. Willis . 1991 . Seasona l variation
in catch rate and body condition of f3ur fish species
in a South Dakota natural lake . Journal of Freshwater Ecology 6 :281-292 .
Headrick, M . R ., and R . F
. Carline . 1993 . Restricte d
summer habitat and growth of northern pike in two
Ohio impoundments . Transactions of the American
Fisheries Society 122 :228-236 .
Jenkins, R . M . 1982 . The morphoedaphic index and
reservoir fish production . Transactions of the American Fisheries Society 111 :133-140.
Lucas, M . C . 1992 . Spawning activity of male and female pike, Esox lucius L ., determined by acoustic
tracking. Canadian Journal of Zoology 70 :191-196 .
Murphy, B . R ., D. W. Willis, and T. A . Springer. 1991 .
Th e relative weight index in fisheries management :
status and needs . Fisheries 16(2) :30-38 .
Neumann, R . M . 1994 . Growth, distribution, and movement of northern pike in a South Dakota natural
lake . Doctoral dissertation . South Dakota State University, Brookings .
Neumann, R . M ., D . W. Willis, and S . M . Sammons .
1994 . Seasonal growth of northern pike (Esox lucius) in a South Dakota glacial lake . Journal of
Freshwater Ecology 9 :191-196 .
Priegel, G . R ., and D . C . Krohn . 1975 . Characteristic s
of a northern pike spawning population . Wisconsin
Department of Natural Resources, Technical Bulletin 86, Madison .
Sammons, S . M ., C . G . Scalet, and R . M . Neumann .
1994 . Seasona l and size-related changes in the diet
of northern pike from a shallow prairie lake . Journal
of Freshwater Ecology 9 :321-329 .
SAS Institute . 1985 . SAS/STAT guide for personal
computers, version 6 edition . SAS Institute, Cary,
North Carolina .
Wege, G . J ., and R . O . Anderson . 1978 . Relativ e weight
(W r ) : a new index of condition for largemouth bass .
Pages 79-91 in G . D . Novinger and J . G . Dillard,
editors . New approaches to the management of
small impoundments . American Fisheries Society,
North Central Division, Special Publication 5, Bethesda, Maryland .
Willis, D . W. 1989 . Proposed standard length-weight
equation for northern pike . North American Journal
of Fisheries Management 9 :203-208 .
Willis, D . W., B . R . Murphy, and C . S . Guy. 1993 . Stoc k
density indices : development, use, and limitations .
Reviews in Fisheries Science 1 :203-222 .