You are on page 1of 15

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233077195

Effects of Pulsed and Continuous DC Electrofishing on Juvenile Rainbow Trout

Article  in  North American Journal of Fisheries Management · November 1998


DOI: 10.1577/1548-8675(1998)018<0905:EOPACD>2.0.CO;2

CITATIONS READS
61 381

3 authors, including:

Andrew J. Paul
Government of Alberta
28 PUBLICATIONS   1,456 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Andrew J. Paul on 08 October 2015.

The user has requested enhancement of the downloaded file.


North American Journal of Fisheries Management 18:905–918, 1998
q Copyright by the American Fisheries Society 1998

Effects of Pulsed and Continuous DC Electrofishing on


Juvenile Rainbow Trout
BARBARA J. AINSLIE, JOHN R. POST,* AND ANDREW J. PAUL
Division of Ecology, Department of Biological Sciences
University of Calgary, Calgary, Alberta, T2N 1N4, Canada

Abstract.—Three hundred fifty juvenile rainbow trout Oncorhynchus mykiss, reared under captive
conditions, were exposed to 300-V continuous DC or low-frequency (30-Hz) pulsed DC (PDC)
for one or three electroshocking passes to evaluate the effects of electroshock on mortality, injury,
and growth over 147 d. Mortality was negligible (;1%) in all treatments. Injury rates varied from
15% to 39%, with PDC causing a greater number (but typically less severe) injuries than analogous
DC sampling. Multiple-pass sampling designs caused more spinal injuries than single-pass designs.
Longer (heavier) fish sustained more spinal injuries. Electroshocking reduced mean growth rates,
but there were no statistically significant differences in growth between treatment groups. Growth
in length was significantly reduced with increasing severity of injuries. Thus, it appears that growth
was not directly impaired by electroshocking but rather by the occurrence of spinal injury, the
severity of which was directly proportional to the magnitude of the growth depression. Extrapo-
lation of the experimental data to field studies in which 20% or less of the population is sampled
suggested reductions of 3% or less in mean population growth with DC or low-frequency PDC
electroshocking.

Electroshocking is commonly used by fisheries fishing (Fredenberg 1992; Snyder 1992, 1995;
biologists to sample freshwater fish populations, Sharber et al. 1994; Dalbey et al. 1996) because
particularly in lotic systems, where many other lower spinal injury rates, particularly in salmonids,
techniques are not viable (Bohlin et al. 1989). occur with these waveforms (Fredenberg 1992;
However, studies have shown that high proportions Taube 1992; McMichael 1993; Sharber et al. 1994;
of sampled fish sustain skeletal injuries (Hauck Dalbey et al. 1996). However, the potential cu-
1949; Sharber and Carothers 1988; Fredenberg mulative effects of commonly used multiple-pass
1992; McMichael 1993; Hollender and Carline depletion or mark–recapture population estimation
1994; Sharber et al. 1994; Dalbey et al. 1996; techniques on the sampled individuals are not well
Thompson et al. 1997a) and external brands (Hor- understood. Such electrofishing techniques alter
ak and Klein 1967; Fredenberg 1992) as a result short-term behavior (Mesa and Schreck 1989),
of electrofishing. Injuries from electrical shock in cause spinal injuries (McMichael 1993), reduce
fish can also occur with no visible external signs growth (Gatz et al. 1986; Dwyer and White 1997;
(Reynolds 1996). These injuries include muscle, Thompson et al. 1997b), decrease body condition
nerve, and tissue damage (Hauck 1949; Mc- (Thompson et al. 1997b), and increase short-term
Michael 1993; Hollender and Carline 1994; mortality (Pratt 1955) among individual salmonids
Thompson et al. 1997a), physiological and behav- compared with control fish.
ioral disturbances (Hauck 1949; Schreck et al. Large (.300 mm) adult rainbow trout Onco-
1976; Mesa and Schreck 1989; Mitton and rhynchus mykiss sustain high injury rates from di-
McDonald 1994a), reduced swim stamina (Horak rect current electrofishing (Sharber and Carothers
and Klein 1967; Mitton and McDonald 1994b), 1988; Sharber et al. 1994; Thompson et al. 1997a);
and reduced fertility (Marriott 1973; Muth and however, the effects on smaller juvenile fish are
Ruppert 1996). less well known. Smaller fish intercept a smaller
The incidence and severity of electrofishing head-to-tail potential than larger fish (Sharber and
damage is partly related to the type of equipment Carothers 1988) and may therefore be subject to
chosen and the waveform produced (Sharber and lower injury rates (e.g., Hollender and Carline
Carothers 1988; McMichael 1993; Sharber et al. 1994; Dalbey et al. 1996; Thompson et al. 1997a;
1994; Dalbey et al. 1996; Dwyer and White 1997). but see Jesien and Hocutt 1990). The long-term
Continuous DC or low-frequency (#30-Hz) pulsed effects of electroshocking damage on juvenile fish
DC (PDC) have been recommended for electro- growth also need to be quantified. Taube (1992)
and Dalbey et al. (1996) found that large injured
* Corresponding author: jrpost@ucalgary.ca rainbow trout adults experienced impaired long-
905
906 AINSLIE ET AL.

term (182 d and 335 d, respectively) growth rates pulsed waveform (ordinary multimeters are not de-
after electroshocking, though long-term survival signed to measure peak values for PDC; Jesien and
was not affected in either study. Hocutt 1990). We measured voltage gradient start-
We compared the effects of single-pass and tri- ing from the anode ring and moving at 0.10-m
ple-pass sampling designs and low-frequency (30- intervals directly in front of the anode. The cath-
Hz) PDC and DC waveforms on different groups ode was behind the anode and was directly op-
of juvenile rainbow trout electroshocked in a posite the voltage gradient meter. Maximum volt-
hatchery raceway. Our study addressed the effects age gradient, which occurs at very close proximity
of electroshock on (1) acute mortality, (2) spinal to the anode, could not be interpreted with our
injury and, (3) long-term growth. We also esti- methods (Kolz 1993). The CRO was used to de-
mated population-level growth effects of electro- scribe the output waveform for both PDC and DC.
shock by extrapolating from observed individual This was done with the voltage gradient probe po-
effects. sitioned 1 m directly in front of the anode ring,
with the cathode situated behind the anode, as in
Methods the electrofishing design. The two waveforms pro-
Three hundred and fifty rainbow trout (Mount duced by the electrofishing unit were traced from
Lassen Livingston strain; mean fork length, FL 6 the CRO onto graph paper by hand.
SD 5 13.34 6 1.56 cm, mean mass 5 27.06 6 At the start of each electrofishing pass, all fish
9.84 g) were randomly assigned to one of five were released into the raceway and were allowed
treatment groups: (1) single-pass continuous direct to acclimate for 15 min. During each pass, one
current (DC*1), (2) single-pass pulsed direct cur- person operated the electrofishing unit, and anoth-
rent (PDC*1), (3) triple-pass DC (DC*3), (4) triple- er dipnetted shocked fish. Fish were removed after
pass PDC (PDC*3) and, (5) control (handled only). they showed narcosis in response to electroshock-
Each experimental treatment was applied with a ing. Limits on the rate at which fish could be dip-
Smith-Root model 15-C Programmable Output netted and removed may have caused some fish to
Waveform backpack electrofishing unit with a gas- remain in the electric field long enough to expe-
oline-powered Honda EX350 generator. The gen- rience tetany. An electroshocking pass was con-
erator outputs alternating current (AC) at a fre- sidered complete only after all 70 fish in a treat-
quency of 60 Hz to the electrofishing unit; the ment group were transferred from the raceway to
electrofishing unit was set at 300 V to output either a recovery vessel. All electroshocking was com-
PDC (frequency of 30 Hz and pulse width of 4 pleted within 1 d. The triple-pass treatment groups
ms) or DC. The anode was a 0.28-m-diameter alu- were allowed approximately 3 h to recover in sep-
minum loop (10-mm tubing, outside diameter), arate holding pens between sequential shocking
and the cathode was a 3.02-m cable (1.78 m was passes.
bared and 1.24 m insulated; the bared cable had a After the application of the experimental treat-
diameter of 5 mm). Electrofishing was conducted ments, individual fish were netted, anaesthetized
on September 15, 1995, in a barricaded concrete with tricaine methanesulfonate (MS-222, 0.03 g/
raceway. The barricaded area was 6.22 m long, L), weighed to the nearest 0.1 g, measured for fork
2.45 m wide, and had a water depth of 0.18 m. length to the nearest millimeter, and individually
Specific conductivity was 285 mS/cm, and water marked with a passive integrated transponder
temperature was 10.58C. (PIT) tag (Avid Manufacturing, Norco, Califor-
Field intensity and output waveform of the elec- nia). All visually apparent brand marks were also
trofishing unit were measured on March 12, 1997, documented. After processing, each treatment
in the same raceway under similar conditions of group was transferred into a separate mesh holding
water depth (0.18 m), water temperature (128C), container (1.81 3 1.02 3 1.00 m) within a cement
and specific conductivity (284 mS/cm). Voltage raceway and held for 72 h to assess short-term
gradient was measured with a voltage gradient mortality.
probe consisting of two insulated wires separated Following the 72-h holding period, equal num-
by a distance of 10 mm and with approximately 2 bers of rainbow trout from each of the five treat-
mm of bared wire exposed at the tips (Kolz 1993). ment groups were randomly assigned to each of
The voltage gradient probe was attached, at sep- six tanks where they were held for 147 d. Tanks
arate times, to both a digital voltage meter and a were assigned a number of fish approximately pro-
cathode-ray oscilloscope (CRO). The CRO was portional to their volume. There were four rect-
used to confirm voltage gradient readings for the angular tanks (one tank, 3.18 3 1.52 3 0.25 m
ELECTROFISHING JUVENILE RAINBOW TROUT 907

with 90 fish; one tank, 1.37 3 1.52 3 0.43 m with and single and multiple passes were compared be-
64 fish; and two tanks, 2.08 3 0.55 3 0.38 m with tween DC and PDC treatments. Functional rela-
39 fish each) and two circular tanks (1.22-m di- tionships between the number of malformed ver-
ameter, 0.50 m deep with 59 fish each). tebrae and the initial length and mass of treatment
Fish were fed once daily with a type and quantity fish were assessed with model-1 linear regressions
of food designed to optimize growth as a function (Sokal and Rohlf 1995). The hypothesis that the
of the previous week’s water temperature and the pattern in position of injured vertebrae was not
average size of the fish in each tank. Feeding significantly different from uniform was assessed
schedules were updated weekly. A 10-h light : 14-h with the G-test for goodness of fit of single-clas-
dark cycle was maintained throughout the exper- sification data (Sokal and Rohlf 1995). The hy-
iment. Tanks were checked daily for dead fish. pothesis that the pattern in position of injured ver-
The fork length (nearest mm) and mass (nearest tebrae did not differ among treatments was eval-
0.1 g) of all individuals, as identified by their PIT uated with the Kolmogorov–Smirnov two-sample
tag, were recorded 14 times throughout the ex- test, as modified for large sample sizes (Sokal and
periment. Individuals were lightly anaesthetized Rohlf 1995). Numbers of vertebrae damaged per
(MS-222, 0.03 g/L) before handling. This sam- fish were compared between branded and non-
pling was done at weekly intervals for the first 7 branded injured fish with a two-sample t-test (So-
weeks of the experiment, at biweekly intervals for kal and Rohlf 1995).
the next 8 weeks, and at triweekly intervals for Growth analysis.—Standardized growth in
the last 6 weeks. At the conclusion of the exper- length and mass was calculated for each treatment–
iment, all fish were killed with MS-222 and frozen tank combination for the 13 sampling intervals.
for X-ray analysis. Growth for each treatment–tank over a time in-
X-ray assessment.—All fish were X-rayed to de- terval was calculated as the mean from individual
termine the number and location of spinal defor- fish within that treatment–tank. Standardized
mities. Fish were partially thawed, then X-rayed growth for each treatment–tank was then calcu-
(right lateral view) at a focal distance of 91.4 cm lated by subtracting the mean growth of control
with a Philips Medical System X-ray machine fish from the mean growth of treatment fish. The
(1.0–4.0 mA, 2.5–10.0 ms, 40-kV potential). The resulting data set is a time series of standardized
X rays were taken with 35.6 3 43.2-cm Fuji growth for six replicates of each of four treatment
SHRS-30 X-ray plates. There were approximately groups. A negative standardized growth indicated
six fish radiographed per X-ray plate. the electroshocked fish grew at a slower rate than
The X-ray photographs were analyzed by a sin- the control fish. This method of assessing growth
gle reviewer, and the locations of all vertebral ab- avoids the potential influence of tank effects by
normalities were identified by number. The first scaling observed growth rates of treated fish to
well-defined vertebra behind the skull was des- untreated fish raised in the same tank under iden-
ignated as vertebra 1. The treatment groups to tical conditions. These data were analyzed with
which the fish belonged were unknown to the re- repeated-measures analysis of variance (ANOVA)
viewer during this analysis. X rays were of good to test the hypotheses that waveform, frequency of
quality and all were readable because they were shocking, and their interactions affected standard-
taken from fish that had grown through the 147-d ized growth rate.
experiment (mean FL 6 SE 5 22.47 6 1.15 cm). The hypothesis that growth rate was modified
It was not possible to rate injury severity by tra- by the severity of injury, rather than simply by
ditional methods (e.g., Hollender and Carline electrofishing, was tested by regressing the number
1994: Table 2; Dalbey et al. 1996: Table 1); rather, of vertebrae damaged for an individual fish in the
injury severity was quantified by the number of experiment against its standardized growth over
malformed vertebrae per fish. the 147 d. Growth was assessed by using length
We compared frequency distributions of the and mass. These measures were standardized by
number of malformed vertebrae per injured fish subtracting the mean growth of control fish from
per treatment between pairs of treatment groups the mean growth of treatment fish in the same tank.
with the Kolmogorov–Smirnov two-sample test, as We used t-tests to test the regression hypotheses
modified for large sample sizes (Sokal and Rohlf that (1) severity of spinal injury affects growth, as
1995). Treatment groups were each compared with evidenced by a nonzero slope and (2) electro-
the control group, waveforms were compared be- shocking itself directly affects growth, as evi-
tween single-pass and multiple-pass treatments, denced by a nonzero intercept.
908 AINSLIE ET AL.

Population-level growth effects.—The data from When a stream population assessment is done,
the laboratory experiment were used to model the typically only a small proportion of the total
effects of six commonly used field electroshocking stream will be electroshocked (e.g., Schill and Be-
protocols on population growth rates. To accom- land 1995); thus, our model results for the sampled
plish this, a weighted-average model, similar to section were extrapolated to the population of the
that used by Schill and Beland (1995) was pro- entire stream. When extrapolating, we assumed
posed. The factors considered in the model were that if x% of the whole stream population was
(1) waveform (DC versus 30-Hz PDC), and (2) electrofished, then it would experience a decrease
potential number of exposures to electrofishing in in growth equal to x% of the growth alteration for
each of three different abundance estimates (sin- the sampled section. We modeled x 5 5% and x
gle-pass removal, triple-pass depletion removal, 5 20%.
and capture–mark–recapture with three passes).
For each waveform, we estimated the standard- Results
ized growth of fish as the mean daily growth rate Electroshocker Output
of fish electroshocked a particular number of times Voltage gradients for both DC and PDC were
divided by the mean daily growth rate of control similar and ranged from 1 V/cm near the anode to
fish. Mean daily growth rates, as a function of less than 0.1 V/cm at distances greater than 0.5 m
electroshocking protocol, were determined from in front of the anode (Figure 1). The CRO trace
the total growth in mass over the 147-d period, for the PDC waveform showed that the frequency
averaged over the six tank replicates. Growth rates of pulses was 30 Hz (i.e., wavelength 5 33 ms)
for two electrofishing passes were estimated by with a pulse width of approximately 4 ms (Figure
taking the arithmetic average of the single-pass 1). However, the pulsed waveform was not square
and triple-pass growth rates for a particular wave- but showed a leading spike of about 1 ms duration
form. The change in growth within the sampled (Figure 1). The DC waveform showed a pattern of
stream section was modeled as the percentage rippled DC with an embedded pulse having a fre-
change of observed standardized growth from the quency of approximately 120 Hz and pulse width
ideal case of standardized growth, in which elec- of 3–4 ms (Figure 1).
troshocked and control fish grew at the same rate.
For multiple-pass removal treatments, we assumed Mortality
that a proportion of the population equal to the Mortality was negligible in all treatments over
capture efficiency of the pass was removed after the 147 d of the experiment. No fish died in the
each pass. PDC*1, DC*1, and control groups, and only one
In our population extrapolation, we assumed fish died in each of the DC*3 and PDC*3 treatment
that (1) when using a capture–mark–recapture pro- groups. Therefore, the mortality rate was only
tocol, the captured fish are returned to the sampling about 1% in each of those groups.
area after each electrofishing pass, whereas fish
captured under a depletion–removal protocol will Spinal Injury
only be returned to the sampling area once all elec- The injury rate was greater for triple-pass than
trofishing is completed, (2) capture efficiency is for single-pass electroshocking and for PDC than
identical for both waveform types, (3) all fish in for DC designs (Figure 2). Of the electrofishing
an area are exposed to the electric current during treatments, PDC*3 had the highest injury rate
an electroshocking pass, (4) all fish exposed to a (39%), and DC*1 had the lowest (15%).
particular number of electroshocking passes will Electrofishing had a significant affect on the fre-
experience an alteration in their growth pattern quency of injury intensity in three out of four treat-
consistent with that number of passes, (5) growth ment groups. Frequency distributions of vertebral
rates experienced by domestic rainbow trout in a malformities differed significantly between un-
hatchery are indicative of the growth of wild trout treated and all electrofished groups, except the
in nature, (6) the probability of capture decreases DC*1 group (Figure 3; Table 1). Sampling design
during sequential electroshocking passes (Schnute significantly affected the frequency of injury in-
1983; Riley and Fausch 1992) from an estimated tensity for the DC treatments only. Waveform did
50% on the first pass to 30% on the second pass not significantly affect the frequency of injury in-
and 10% on the third and, (7) there is no immi- tensity for either sampling design. The injuries
gration into or emigration out of the electrofished caused by DC tended to be more severe than the
area of stream during the sampling procedure. analogous PDC treatments as the mean numbers
ELECTROFISHING JUVENILE RAINBOW TROUT 909

FIGURE 1.—Voltage gradient measured with a cathode-ray oscilloscope or digital multimeter at given distances in
front of the anode for either DC or pulsed DC (PDC). Insets are cathode-ray oscilloscope tracings of the pulsed and
continuous waveforms measured with a voltage gradient probe at 1 m in front of the anode (power was initiated at time
5 5 ms).

of vertebral malformities per malformed fish Vertebral abnormalities were detected over most
(6SE) were greater: PDC*1 5 9.0 6 0.8, DC*1 of the spine and were indistinguishable from nat-
5 10.6 6 2.8, PDC*3 5 11.2 6 1.0, DC*3 5 11.9 ural abnormalities. The majority (;60%) of all
6 1.1. abnormalities were anterior to the dorsal fin (Fig-
The number of malformed vertebrae was posi- ure 5). Vertebra 21 was injured most often, with
tively related to the initial fork length of the fish 52% of all injured electroshocked fish sustaining
in the PDC treatment groups but not related to an injury in this position. Few injuries were de-
length in the DC treatments (Figure 4). The num- tected in either the far anterior or posterior regions
ber of vertebrae damaged was also related to the of the spine.
initial mass of the treatment fish in the PDC*1 (r 2 Electroshocking had a selective effect on the
5 0.35, N 5 23, P 5 0.003), PDC*3 (r 2 5 0.16, position of spinal damage in three out of four treat-
N 5 26, P 5 0.045) and DC*3 (r 2 5 0.19, N 5 ment groups, with certain vertebrae sustaining
26, P 5 0.028) treatments but unrelated to initial higher injury rates than others. Frequency distri-
mass in the DC*1 (r 2 5 0.10, N 5 10, P 5 0.373) butions of injury position for the control and DC*1
treatment. groups did not significantly differ from an even
910 AINSLIE ET AL.

FIGURE 2.—Proportion of rainbow trout with at least


one damaged vertebra in each of five electrofishing treat-
ment groups. Sample sizes were as follows: control 5 69;
DC, one pass (DC*1) 5 69; pulsed DC, one pass (PDC*1)
5 71; DC, three passes (DC*3) 5 71; and pulsed DC,
three passes (PDC*3) 5 67.

distribution (Gcontrol 5 63.94, GDC*1 5 77.93; P .


0.05), whereas the other electrofished groups had
uneven distributions of injuries (GPDC*1 5 283.99,
GPDC*3 5 209.92, GDC*35 443.59, P # 0.05; Figure
5). Frequency distributions of injury position for
DC*3 and PDC*1 treatments did not differ sig-
FIGURE 3.—Frequencies of spinal injuries to rainbow
nificantly from one another(D 5 0.108, P . 0.05), trout for each of five electrofishing treatment groups: con-
whereas PDC*3 and DC*3 treatments differed sig- trol; DC, one pass (DC*1); pulsed DC, one pass (PDC*1);
nificantly (D 5 0.154, P # 0.05), as did PDC*1 DC, three passes (DC*3); and pulsed DC, three passes
and PDC*3 (D 5 0.155, P # 0.05). (PDC*3). Injury frequency is expressed as the number of
malformed vertebrae per fish.
External Signs of Injury
External signs of injury (brands) were indicators
of internal spinal damage. Of the 85 injured fish,
22 (26%) had visible brands, whereas no uninjured TABLE 1.—Comparisons between frequency distribu-
tions of electroshocking injuries for selected pairs of treat-
fish were branded. Branded fish also had signifi-
ment groups, determined by the Kolmogorov–Smirnov
cantly more vertebrae damaged per fish (13.0 6 two-sample test, as modified for large sample sizes. Sig-
1.2) than nonbranded injured fish (9.9 6 0.6; t 5 nificantly different distributions (D) are indicated by an
2.3487, N 5 22,63, P # 0.05). Brands were found asterisk. Treatment groups are as follows: DC, one pass 5
to be temporary indicators of electrofishing-related DC*1; DC, three passes 5 DC*3; pulsed DC, one pass 5
spinal injury because only one brand (of 22) re- PDC*1; pulsed DC, three passes 5 PDC*3. Parenthetical
mained visible 1 week after the initial electro- values are sample sizes.
shocking. A comparison of approximate brand po- Comparison D P
sitions with the frequency distribution of vertebral
Electroshocking effect
malformities for branded fish indicated that brands Control (69) versus PDC*1 (71) 0.280* 0.009
were usually posterior to the position of spinal Control (69) versus DC*1 (69) 0.101 .0.9
injury (Figure 6). Control (69) versus PDC*3 (67) 0.344* 0.0007
Control (69) versus DC*3 (71) 0.337* 0.0007
Electroshocking Effects on Growth Waveform effect
Standardized rates of growth in length were not PDC*1 (71) versus DC*1 (69) 0.180 0.21
PDC*3 (67) versus DC*3 (71) 0.047 .0.9
significantly affected by electroshocking wave-
form, number of shocking passes, or their inter- Sampling-design effect
action (repeated-measures ANOVA: F 5 0.33, df PDC*1 (71) versus PDC*3 (67) 0.103 .0.9
DC*1 (69) versus DC*3 (71) 0.264* 0.015
5 1, P 5 0.858; F 5 3.749, df 5 1, P 5 0.067;
ELECTROFISHING JUVENILE RAINBOW TROUT 911

FIGURE 4.—Number of malformed vertebrae versus initial fork length of rainbow trout for each of four electrofishing
treatment groups: DC, one pass (DC*1); pulsed DC, one pass (PDC*1); DC, three passes (DC*3); and pulsed DC, three
passes (PDC*3). Also shown are regression lines with slopes significantly different from zero.

F 5 0.339, df 5 1, P 5 0.567, respectively; Figure was significantly less for electroshocked fish than
7). Standardized rates of growth in mass were also for the control fish.
not significantly affected by electroshocking
waveform, number of shocking passes, or their in- Population-Level Growth Effects
teraction (repeated-measures ANOVA: F 5 0.123, Electrofishing a section of stream can be ex-
df 5 1, P 5 0.730; F 5 3.668, df 5 1, P 5 0.070; pected to change the natural growth rates of the
F 5 0.338, df 5 1, P 5 0.567, respectively; Figure juvenile salmonids in that section of stream by
7). However, mean standardized growth rates in 11.54% to 214.29%, depending on the sampling
length and mass of the triple-pass treatments tend- protocol chosen (Table 2). Obviously, the mag-
ed to be lower than those of the single-pass treat- nitude of the electrofishing-associated growth al-
ments (Figure 7), and sampling design was only teration was lower at the whole-stream population
marginally nonsignificant for both length and level. In our model, the largest population level
mass. growth reduction was caused when 20% of the
Standardized growth in length was inversely re- juvenile rainbow trout population was electro-
lated to the number of injured vertebrae in exper- shocked (22.86%), and the smallest change was
imental fish (Figure 8). Growth in mass was un- caused when 5% of the population was sampled
related to the number of vertebrae damaged but (10.08%). For both DC and 30-Hz PDC wave-
912 AINSLIE ET AL.

FIGURE 5.—Frequencies of injuries to vertebrae of rainbow trout subjected to five electrofishing treatments: control;
DC, one pass (DC*1); pulsed DC, one pass (PDC*1); DC, three passes (DC*3); and pulsed DC, three passes (PDC*3).
The locations of abnormal vertebrae were recorded from X rays of all fish. Vertebrae were numbered sequentially,
beginning with the first clearly defined vertebra behind the skull. Arrows indicate the middle of the dorsal and anal
fins, respectively.

forms, the smallest alterations in natural growth not important in our experiment as we observed
rate occurred with a single-pass sampling design; negligible short- and long-term (147-d) mortality.
a capture–mark–recapture design with three passes High short-term survivorship (#20 d; $90%) is
caused the largest magnitude alterations. commonly reported in electroshocking studies on
hatchery trout (Pratt 1955; Horak and Klein 1967;
Discussion Maxfield et al. 1971; Hudy 1985; McMichael
Mortality 1993; Mitton and McDonald 1994a) and wild trout
Factors thought to be important in causing elec- (Dalbey et al. 1996; Habera et al. 1996). Longer-
trofishing deaths, such as respiratory failure, hem- term studies (7–33 months) have reported some-
orrhaging, internal injuries, and trauma (Hauck what lower survival of hatchery trout in the lab-
1949; Schreck et al. 1976; Lamarque 1990), were oratory (65–93%; Maxfield et al. 1971; Taube
ELECTROFISHING JUVENILE RAINBOW TROUT 913

FIGURE 6.—Comparison of the frequency distributions of brand positions (line) and injury positions (bars) for all
branded, electroshocked fish (combined over treatment groups, N 5 22). Arrows indicate the mid-points of the dorsal
and anal fins, respectively.

1992) and of wild trout in a natural pond (54– this hypothesis. Other studies have also reported
60%; Dalbey et al. 1996). Despite these studies, increases in both injury occurrence and severity
it is not clear whether natural mortality rates are with size (Hollender and Carline 1994; Dalbey et
compensatory or cumulative with electrofishing- al. 1996; Thompson et al. 1997a), but no studies
related mortality rates. have been conducted to specifically address this
hypothesis.
Spinal Injury In our study, most injuries were anterior to the
The rate of injury we observed (15–39%) was dorsal fin, which contrasts other studies that found
similar to that reported for experiments that used that most injuries occurred posterior to the dorsal
similar waveforms and pulse frequencies (Taube fin (Sharber and Carothers 1988; Fredenberg 1992;
1992; McMichael 1993; Sharber et al. 1994; Dal- Hollender and Carline 1994). The reason for this
bey et al. 1996). In our study, DC treatments difference may be due to the smaller size of our
caused lower injury rates than 30-Hz PDC treat- experimental fish or to some unknown effect of
ments, which, in turn, were lower than injury rates environmental conditions during shocking. How-
caused by higher pulse frequencies (Sharber and ever, our results and those of others (Sharber and
Carothers 1988; Fredenberg 1992; Taube 1992; Carothers 1988; Fredenberg 1992) have found few
McMichael 1993; Sharber et al. 1994; Dalbey et damaged vertebrae in either the far posterior or far
al. 1996; Thompson et al. 1997a). When other fac- anterior regions of the spine. This is expected be-
tors are held constant, injury rates tend to increase cause spinal injury tends to occur where the white
with pulse frequency (Taube 1992; McMichael muscle tissue on both sides of the spinal column
1993; Sharber et al. 1994). simultaneously excites; the spinal column com-
Injury rates were greater for triple-pass than for pensates for this stress by compressing, fracturing
single-pass electrofishing. In the DC treatments, or misaligning (Lamarque 1990). The central re-
22% more fish were injured with triple-pass sam- gion of the body contains more muscle mass than
pling than with single-pass sampling, which was the far anterior or posterior spine regions (Lam-
9–27% higher than has been reported in single- arque 1990) and is therefore more susceptible to
pass electrofishing with this waveform (Freden- damage.
berg 1992; Taube 1992; McMichael 1993; Dalbey We may have underestimated the frequency and
et al. 1996). In the PDC treatments, triple-pass severity of spinal injuries because we only used
sampling only caused a 6% increase in injury rate one X-ray view and may have missed some injuries
over single-pass sampling. (e.g., Thompson et al. 1997a). On the other hand,
Injury rates were directly related to initial fork spinal injury rates may have been overestimated
length for both single-pass and triple-pass PDC as we found spinal malformities in unshocked con-
sampling and were directly related to initial mass trol fish that were impossible to distinguish from
for single-pass PDC and triple-pass DC sampling. electroshock-related injuries. The 4% malformity
Given that we had a small range of initial fish sizes, rate in the unshocked control group, an estimate
our experiment was not designed to adequately test of the natural malformity rate, was similar to mal-
914 AINSLIE ET AL.

FIGURE 7.—Standardized growth of rainbow trout by length (mm/d) and mass (g/d), for four electroshocked treatment
groups over 147 d. The treatment groups were DC, one pass (DC*1); pulsed DC, one pass (PDC*1); DC, three passes
(DC*3); and pulsed DC, three passes (PDC*3).

formity rates reported for other unshocked salmo- injury. Brands are thought to be caused by erratic
nids (McCrimmon and Bidgood 1965; Gill and stimulation of chromatophores when nerve fibers
Fisk 1966). are damaged by vertebral dislocation (Lamarque
The presence of brands was related to the pres- 1990; Fredenberg 1992) or by hemorrhaging of
ence and severity of spinal damage, but lack of damaged capillaries near the skin (Reynolds
branding did not indicate a lack of spinal injury. 1996).
Even though brands accounted for only 25% of all
injured electroshocked fish, all branded fish sus- Growth
tained injuries and were injured the most severely Triple-pass sampling did not significantly re-
overall. Fredenberg (1992) also found that brand duce growth rates of electroshocked fish when
presence was related to more serious spinal inju- compared with single-pass sampling. However,
ries. However, brands remained visible for a very our inability to detect a difference in growth rates
short period (#1 week postelectroshock) and are between triple-pass and single-pass sampling may
thus not reliable indicators of past electroshocking have been because treatment groups contained dif-
ELECTROFISHING JUVENILE RAINBOW TROUT 915

FIGURE 8.—Relationships between the number of vertebral malformities and standardized growth in length and mass.
Growth rates were standardized to the mean growth of control fish in the same tank over the 147-d experiment. The
sample size in all regressions is 277 fish. Fitted regressions and zero growth lines are shown. Intercepts (a) and slopes
(b) are reported for all plots; those significantly different from zero are marked with an asterisk, and those not signifi-
cantly different are marked by ‘‘NS.’’

ferent proportions of injured fish, which grew dif- after 335 d than did uninjured and slightly injured
ferently from uninjured and control fish. We ex- fish. Previously injured fish recaptured with elec-
pected a priori that electrofishing-related injuries troshock have likely grown more slowly than other
would decrease growth rates because (1) injured individuals within the general population.
fish may need to use (potentially substantial) por-
tions of their energy budget to repair damaged tis- Population-Level Growth Effects
sues and bones (Gatz et al. 1986; Gatz and Adams The population-level effect of electrofishing de-
1987), (2) recovery from electroshock causes a pends on the type of abundance estimate, the num-
short-term decrease in feeding behaviors (Bouck ber of sampling passes, and the proportion of the
and Ball 1966; Mesa and Schreck 1989), and (3) population sampled. If one-fifth or less of the pop-
the heavy calcification and fusion of injured ver- ulation is sampled according to the techniques
tebrae during healing (Sharber and Carothers used in this study, the growth effects would prob-
1988; Fredenberg 1992) may cause abnormal ably be undetectable in the field. However, these
swimming behavior, which could interfere with effects may confound long-term monitoring stud-
foraging effectiveness and predator avoidance. ies that use electroshock as a sampling tool be-
In our study, fish with a large number of dam- cause the sampling method is creating growth al-
aged vertebrae gained less length than did less terations in individuals that, over time, could ac-
severely injured fish. Dalbey et al. (1996) also re- cumulate and affect the population as a whole. The
ported that moderate to severely injured fish (i.e., only long-term (.1 year) studies examining the
fish which sustained spinal misalignment or frac- effects of electroshocking on salmonid populations
ture) had significantly lower growth and condition found that salmonid abundance was stable or had
916 AINSLIE ET AL.

TABLE 2.—Mean growth effects expected within an Management Implications


electroshocked stream section and when various propor-
tions of the total juvenile salmonid population are elec- Electrofishing waveform and sampling design
troshocked. Growth is expressed as the percentage change (number of passes) should be taken into account
in mass (g/d) expected with a particular waveform and when designing studies of juvenile stream salmo-
type of abundance estimate. In the sampled section, all fish nids. Our results support those of others (Freden-
are electroshocked; whereas in the two extrapolations to berg 1992; Snyder 1992, 1995; Sharber et al. 1994;
whole populations, only 20% or 5% of the population is
Dalbey et al. 1996) who have recommended the
exposed to electroshock.
use of low-frequency pulsed DC or continuous DC
Sample Sample to reduce injury rates. However, injury rates and
Sampled 20% of 5% of growth alterations increased with the number of
Abundance estimate section population population
passes. Investigators must balance the need for in-
Continuous waveform
creasing the number of passes to improve the pre-
Single pass 1.54% 0.31% 0.08%
Depletion–removal with
cision of population estimates against the impacts
3 passes 25.19% 21.04% 20.26% created by improving those estimates. Growth ef-
Capture–mark–recapture fects may not be detectable in the field when sam-
with 3 passes 214.29% 22.86% 20.71% pling juvenile rainbow trout with the methods used
Low-frequency pulsed waveform in this study and when one-fifth or less of a pop-
Single pass 23.34% 20.67% 20.17% ulation is sampled but could be a factor when re-
Depletion–removal with
3 passes 27.77% 21.55% 20.39%
peated electroshockings are used to monitor long-
Capture–mark–recapture term growth trends. Some fraction of detected
with 3 passes 213.77% 22.75% 20.69% growth alterations will probably have been caused
by the sampling procedure. With these consider-
ations in mind, electrofishing with continuous DC
or low-frequency pulsed DC will remain a valuable
increased after 4 years (Habera et al. 1996) or 8 tool for sampling juvenile salmonids in streams.
years (Kocovsky et al. 1997) of annual electro-
shock sampling. Both studies reported relatively Acknowledgments
low (3–26%) injury rates. These results suggest
Funding for this study was provided through the
that these stream salmonid populations were not
Alberta Fisheries Management Enhancement Pro-
adversely affected by multiple years of electro-
gram Trust Fund. We thank the Sam Livingston
shock sampling, although the effect of electrofish-
Fish Hatchery (Alberta Department of Environ-
ing on growth was not assessed in either study
mental Protection, Natural Resources Service) for
(Habera et al. 1996; Kocovsky et al. 1997).
providing additional funds, fish, experimental fa-
Direct extrapolation of our results to wild salmo-
cilities, and technical expertise. The University of
nids in natural streams must be made cautiously.
Calgary Health Sciences Center provided the X-
Low mortality rates of hatchery rainbow trout after
ray facilities and staff for the X-ray analysis. J.
electroshocking may be due to the favorable living
Doll, D. DePape, H. Finch, T. Hamor, M. Lutz, S.
environment of hatchery fish (Taube 1992;
Shima, and W. Schrenk all contributed valuable
McMichael 1993). Cultured fish do not have to
assistance.
endure the stress of variable environmental con-
ditions that wild fish experience but, rather, are References
provided a nutritionally balanced food at a ration
selected to optimize growth. Conversely, densities Bohlin, T., S. Hamrin, T. G. Heggberget, G. Rasmussen,
of cultured fish, typically much higher than in na- and S. J. Saltveit. 1989. Electrofishing—theory and
practice with special emphasis on salmonids. Hy-
ture, result in higher levels of aggression in the drobiologia 173:9–43.
hatchery (Wootton 1994). Aggression in the hatch- Bouck, G. R., and R. C. Ball. 1966. Influence of capture
ery may therefore amplify the growth performance methods on blood characteristics and mortality in
differences among injured and uninjured individ- the rainbow trout (Salmo gairdneri ). Transactions
uals. Our experiments were also carried out on of the American Fisheries Society 95:170–176.
juvenile rainbow trout of a maximum size of 170 Dalbey, S. R., T. E. McMahon, and W. Fredenberg. 1996.
Effect of electrofishing pulse shape and electrofish-
mm FL. Larger fish of a different species or of ing-induced spinal injury on long-term growth and
wild stock may experience different injury rates survival of wild rainbow trout. North American
and may not experience the same rates of growth Journal of Fisheries Management 16:560–569.
reported in this experiment. Dwyer, W. P., and R. G. White. 1997. Effect of electro-
ELECTROFISHING JUVENILE RAINBOW TROUT 917

shock on juvenile Arctic grayling and Yellowstone Survival, growth, and fecundity of hatchery-reared
cutthroat trout growth, 100 days after treatment. rainbow trout after exposure to pulsating direct cur-
North American Journal of Fisheries Management rent. Transactions of the American Fisheries Society
17:174–177. 100:546–552.
Fredenberg, W. 1992. Evaluation of electrofishing- McCrimmon, H. R., and B. Bidgood. 1965. Abnormal
induced spinal injuries resulting from field electro- vertebrae in the rainbow trout with particular ref-
fishing surveys in Montana. Montana Department erence to electrofishing. Transactions of the Amer-
of Fish, Wildlife, and Parks, Report, Helena. ican Fisheries Society 94:84–88.
Gatz, A. J., Jr., and S. M. Adams. 1987. Effects of McMichael, G. A. 1993. Examination of electrofishing
repeated electroshocking on growth of bluegill 3 injury and short-term mortality in hatchery rainbow
green sunfish hybrids. North American Journal of trout. North American Journal of Fisheries Man-
Fisheries Management 7:448–450. agement 13:229–233.
Gatz, A. J., Jr., J. M. Loar, and G. F. Cada. 1986. Effects Mesa, M. G., and C. B. Schreck. 1989. Electrofishing
of repeated electroshocking on instantaneous mark–recapture and depletion methodologies evoke
growth of trout. North American Journal of Fish- behavioral and physiological changes in cutthroat
eries Management 6:176–182. trout. Transactions of the American Fisheries So-
Gill, C. D., and D. M. Fisk. 1966. Vertebral abnormal- ciety 118:644–658.
ities in sockeye, pink, and chum salmon. Transac- Mitton, C. J. A., and D. G. McDonald. 1994a. Conse-
tions of the American Fisheries Society 95:177– quences of pulsed DC electrofishing and air expo-
182. sure to rainbow trout (Oncorhynchus mykiss). Ca-
Habera, J. W., R. J. Strange, B. D. Carter, and S. E. nadian Journal of Fisheries and Aquatic Sciences
Moore. 1996. Short-term mortality and injury of 51:1791–1798.
rainbow trout caused by three-pass AC electrofish- Mitton, C. J. A., and D. G. McDonald. 1994b. Effects
ing in a southern Appalachian stream. North Amer- of electroshock, air exposure, and forced exercise
ican Journal of Fisheries Management 16:192–200. on swim performance in rainbow trout (Oncorhyn-
Hauck, F. R. 1949. Some harmful effects of the electric chus mykiss). Canadian Journal of Fisheries and
shocker on large rainbow trout. Transactions of the Aquatic Sciences 51:1799–1803.
American Fisheries Society 77:61–64. Muth, R. T., and J. B. Ruppert. 1996. Effects of two
Hollender, B. A., and R. F. Carline. 1994. Injury to wild electrofishing currents on captive ripe razorback
brook trout by backpack electrofishing. North suckers and subsequent egg-hatching success. North
American Journal of Fisheries Management 14:
American Journal of Fisheries Management 16:
643–649.
473–476.
Horak, D. L., and W. D. Klein. 1967. Influence of cap-
Pratt, V. S. 1955. Fish mortality caused by electrical
ture methods on fishing success, stamina, and mor-
shockers. Transactions of the American Fisheries
tality of rainbow trout (Salmo gairdneri ) in Colo-
Society 84:93–96.
rado. Transactions of the American Fisheries So-
Reynolds, J. B. 1996. Electrofishing. Pages 221–253 in
ciety 96:220–222.
B. R. Murphy and D. W. Willis, editors. Fisheries
Hudy, M. 1985. Rainbow trout and brook trout mortality
techniques, 2nd edition. American Fisheries Soci-
from high voltage AC electrofishing in a controlled
ety, Bethesda, Maryland.
environment. North American Journal of Fisheries
Riley, S. C., and K. D. Fausch. 1992. Underestimation
Management 5:475–479.
Jesien, R., and R. Hocutt. 1990. Method for evaluating of trout population size by maximum-likelihood re-
fish response to electric fields. Pages 10–18 in I. G. moval estimates in small streams. North American
Cowx, editor. Developments in electric fishing. Journal of Fisheries Management 12:768–776.
Fishing News Books, Oxford, UK. Schill, D. J., and K. F. Beland. 1995. Electrofishing
Kocovsky, P. M., C. Gowan, K. D. Fausch, and S. C. injury studies: a call for a population perspective.
Riley. 1997. Spinal injury rates in three wild trout Fisheries 20(6):28–29.
populations in Colorado after eight years of back- Schnute, J. 1983. A new approach to estimating pop-
pack electrofishing. North American Journal of ulations by the removal method. Canadian Journal
Fisheries Management 17:308–313. of Fisheries and Aquatic Sciences 40:2153–2169.
Kolz, A. L. 1993. In-water electrical measurements for Schreck, C. B., R. A. Whaley, M. L. Bass, O. E. Maugh-
evaluating electrofishing systems. U.S. Fish and an, and M. Solazzi. 1976. Physiological responses
Wildlife Service Biological Report 11. of rainbow trout (Salmo gairdneri ) to electroshock.
Lamarque, P. 1990. Electrophysiology of fish in electric Journal of the Fisheries Research Board of Canada
fields. Pages 4–33 in I. G. Cowx and P. Lamarque, 33:76–84.
editors. Fishing with electricity: applications in Sharber, N. G., and S. W. Carothers. 1988. Influence of
freshwater fisheries management. Fishing News electrofishing pulse shape on spinal injuries in adult
Books, Oxford, UK. rainbow trout. North American Journal of Fisheries
Marriott, R. A. 1973. Effects of electric shocking on Management 8:117–122.
fertility of mature pink salmon. Progressive Fish- Sharber, N. G., S. W. Carothers, J. P. Sharber, J. C. de
Culturist 35:191–194. Vos, Jr., and D. A. House. 1994. Reducing
Maxfield, G. H., R. H. Lander, and K. L. Liscom. 1971. electrofishing-induced injury of rainbow trout. North
918 AINSLIE ET AL.

American Journal of Fisheries Management 14: Thompson, K. G., E. P. Bergersen, and R. B. Nehring.
340–346. 1992a. Injuries to brown trout and rainbow trout
Snyder, D. E. 1992. Impacts of electrofishing on fish. induced by capture with pulsed direct current. North
Final Report of Colorado State University Larval American Journal of Fisheries Management 17:
Fish Laboratory to U.S. Bureau of Reclamation, Salt 141–153.
Lake City, Utah, and to Glen Canyon Environmental Thompson, K. G., E. P. Bergersen, R. B. Nehring, and
Studies, Aquatic Coordination Team, Flagstaff, Ar- D. C. Bowden. 1992b. Long-term effects of elec-
izona. trofishing on growth and body condition of brown
Snyder, D. E. 1995. Impacts of electrofishing on fish. trout and rainbow trout. North American Journal of
Fisheries 20(1):26–27. Fisheries Management 17:154–159.
Sokal, R. R., and F. J. Rohlf. 1995. Biometry: the prin- Wootton, R. J. 1994. Ecology of teleost fishes. Chapman
ciples and practice of statistics in biological re- and Hall, New York.
search, 3rd edition. Freeman, New York.
Taube, T. T. 1992. Injury, survival and growth of rain-
bow trout captured by electrofishing. Master’s the- Received July 7, 1997
sis. University of Alaska, Fairbanks. Accepted March 13, 1998

View publication stats

You might also like