Modern electrofishing pulsator (control box) manufacturers as a whole produce a variety of direct current pulse shapes. Which of these are more effective or more efficient for fish capture? We have noticed some differences in fish reaction thresholds and overall behavior when exposed to different pulse. For years, I have wanted to compare various pulses under controlled laboratory conditions. The challenge has been acquiring a suitable power supply that can produce the desired pulse shapes. That opportunity recently became available.
Smith-Root, Inc. (SRI) has a new headquarters building in Vancouver, Washington, USA which has a laboratory area for conducting controlled experiments, and they allowed me to use their facilities for research. I used the opportunity to evaluate immobilization thresholds for four pulse shapes in a small pilot study.
Most backpack electrofishers and some boat electrofishers built in the U.S. produce a rectangular pulse shape. The pulse has a vertical rise from the voltage baseline, a flat top, and a vertical fall back to the baseline. The pulse corners are squared off, so these are also referred to as square pulses. Exponential decay is another pulse shape; this is sometimes called a capacitor discharge pulse with a vertical rise and a curved, slower fall. Some boat electrofishers produce pulses in the shape of the upper half of a sine wave (herein called a parabolic pulse shape). Parabolic pulses have an intermediate speed rise and an intermediate speed fall, with a rounded top. A fourth pulse type has a saw-tooth shape with spaces between pulses. It resembles that of a skip tooth scroll saw blade rather than the saw-tooth pattern typically produced by electronic signal generators. The saw-tooth shape produced for this study had a vertical rise with a slower, linear fall.
Let me provide some examples of electrofisher models which produce these pulse shapes. Rectangular pulses are generated by the ETS Electrofishing Systems ABP-3 and ABP-2 backpack units and the MBS boat unit; by the Midwest Lake Electrofishing Systems Infinity and Infinity HC-80 boat units and the Xstream backpack unit; and by the SRI LR-24 and LR-20B backpack units. Parabolic pulses are generated by the SRI Type VI-A and GPP series boat units when the control knobs (the pulse width in ms for the Type VI-A and the Percent of Range in percent for the GPPs) are turned fully clockwise. Those SRI boat units can generate the saw-tooth pattern used herein if the control knobs are turned counter-clockwise such that less than a quarter sine wave is produced. Exponential decay pulses are generated by backpack and boat units produced by Hans Grassl in Germany. This list is not intended to be comprehensive, and the intent is not to promote or to eliminate any manufacturer or model. These are just pulse shapes which I have seen or know about from these electrofishing units.
From experience with boat electrofishers in electrofishing classes, Alan Temple and I have noticed that a rectangular pulse requires less applied voltage and power to capture fish than does a parabolic pulse. Also, we have observed fish response when using an exponential decay waveform from a boat electrofisher. Various factors in the field can affect fishing success, so we were eager to compare these waveforms in the lab to reduce or eliminate other variables. SRI engineer Lee Carstensen was gracious enough to set up and operate a waveform generator to produce the desired four waveforms for this comparison study.
One output of the study to be compared was the peak voltage gradient (V/cm) required to immobilize fish. Voltage gradient is directly related to the voltage applied to electrodes. Another output to be compared was the peak power density (µW/cm3) needed to immobilize fish. Power density equals voltage gradient squared times ambient water conductivity; it is directly related to the power applied to electrodes. Average power density equals peak power density times the pulse duty cycle (percent on time), expressed as a proportion. Average power density, which is directly related to average applied power, relates to the power demand on an electrofishing generator or battery power supply, so both peak and average power density relate to electrofishing. Threshold peak power density correlates to applied peak power and is a measure of electrofishing effectiveness. Threshold average power density relates to the power demand from the power source, to the energy per pulse and to the electrical efficiency of fish immobilization (and presumably to the energy needed for fish capture).
Laboratory setup at SRI headquarters for the trials to compare threshold voltage gradient and power density values for goldfish. Note the electronic equipment to produce the desired waveforms and to measure the threshold voltages. Fish reactions to the electrical treatments were documented with digital video recordings. The test tank was a 10-gallon aquarium fitted with stainless steel plate electrodes on each end. The tank inside dimensions were 50 cm long x 25 cm wide. There were 48 cm between the plate electrodes, except for a couple of trials in which they were moved to 27 cm apart, and the water depth was 15 cm. Ambient conductivity was 239 µS/cm at 19.0 C. Prior to the trials, a uniform electrical field was confirmed with a Velleman HPS140i pocket oscilloscope attached to a voltage gradient probe. Pulse shapes and voltages were confirmed with a Tektronix TDS 3032B digital bench oscilloscope and with two portable oscilloscopes, a Velleman pocket oscilloscope and a UNI-T UT81B scopemeter. The response of each fish was recorded with a video camera, and each fish was measured for total length to the nearest mm.
Exponential decay pulse shape and characteristics as measured by the Tektronix bench oscilloscope.
UNI-T UT81B scopemeter traces of the four pulse shapes. From top left, clockwise: rectangular, saw tooth, parabolic and exponential decay. Pulse heights shown here are not threshold values; these traces merely show the pulse shapes. Frequency was 60 pulses per second for all waveforms. Duty cycle values were 24, 12, 16 and 5-6% for the rectangular, saw tooth, parabolic and exponential decay pulse shapes, respectively. One meter indicated 5% duty cycle for the exponential decay pulse whereas another meter indicated a 6% duty cycle or something in between. Therefore, a duty cycle value of 5.5% was used for calculation of average power density from peak power density.
Two size groups of goldfish were exposed to the electrical treatments. Smaller goldfish (57 mm average, 50-61 mm total length range) were exposed to all four pulse shapes. Larger goldfish (76 mm average, 67-92 mm total length range) were exposed to all but the saw-tooth treatment. Each fish was exposed to only one treatment. The peak voltage applied at threshold for immobilization (the minimum voltage required to achieve complete immobility) was determined for each fish. Voltage gradient was calculated as peak voltage applied divided by 48 cm between the plate electrodes, or by 27 cm in a couple of cases when the maximum voltage would not produce immobilization with a 48-cm electrode separation distance.
For the smaller goldfish, the least peak voltage gradient for immobilization was required for the rectangular pulse shape and the highest peak voltage gradient was required for the exponential decay pulse shape. The pattern was the same for the larger goldfish, and the absolute voltage gradient values were lower for the larger fish, as expected. For the small goldfish, peak voltage gradient ratios of parabolic to rectangular and of exponential decay to rectangular were 1.39 and 1.62, respectively. The latter value represented the largest ratio; other ratios were lower. Overall, the rectangular and saw-tooth pulses were relatively effective in terms of peak voltage required to produce immobilization.
The pattern for threshold values of peak power density was basically the same as for the corresponding values of peak voltage gradient. However, the relative differences were magnified because power density is a function of voltage gradient squared. For the small goldfish, peak power density ratios of parabolic to rectangular and of exponential decay to rectangular were 1.95 and 2.68, respectively.
Average power density directly relates to energy per pulse and to power demand on an electrofisher generator or battery. Average power density required for immobilization of small goldfish was lowest for the exponential decay and saw-tooth pulse shapes and highest for the parabolic pulse. In situations where an electrofisher is operating at or near maximum capacity, such as in high water conductivity, the saw-tooth and exponential decay pulse shapes may allow successful operation whereas the others may not. Both the saw-tooth and exponential decay pulses have a vertical rise and a slower fall, and they have the lower duty cycles (percent on time) of the four pulse shapes. The parabolic pulse, with its intermediate speed of rise and fall, requires the most energy per pulse and overall power demand even though its duty cycle is not much higher (16 vs 12%) than for the saw-tooth pulse. The rectangular pulse has a vertical rise and fall; it requires an intermediate amount of energy per pulse due mainly to its highest duty cycle (24%).
The pattern of relative average power density thresholds differed somewhat for the larger goldfish. The exponential decay pulse shape produced the lowest threshold, as for the smaller fish. However, the parabolic pulse threshold was slightly lower than the rectangular pulse threshold.
Collectively, these results indicate the relative effectiveness for immobilization of four direct current pulse shapes. The rectangular pulse required the least peak voltage gradient and power density to immobilize goldfish under these experimental conditions. This matches some empirical evidence from actual field threshold trials in which less peak power was required to capture fish with rectangular versus more parabolic pulse shapes. The saw-tooth pulse was next most effective in terms of voltage gradient required for immobilization. However, the rectangular pulses in this study were associated with the highest duty cycles, so average power density required was less for exponential decay and for saw-tooth pulses, both of which had lower duty cycles compared to rectangular pulses. The most efficient pulse shapes, saw-tooth and exponential decay, each have a fast rise and a slower fall. The next efficient pulse shape for the small fish, rectangular, has a fast rise and a fast fall. The least efficient pulse shape for the smaller fish was the parabolic with its slower rise. It appears that effective and efficient pulse shapes for producing immobilization are associated with fast rises. The saw-tooth pulse required relatively low voltage and almost the lowest average power density for immobilization, an impressive combination of desired characteristics.
This was a preliminary study conducted quickly and with very few fish. The intent was to serve as a pilot project so that further testing could be conducted with more fish. Despite the limitations of time and fish for these preliminary trials, there were significant differences among the pulse shape treatments. One-way ANOVA P-values for all comparisons were 0.01 or less except for large goldfish, mean power density where the P-value was 0.016.
The reader should keep in mind that immobilization is not the only factor associated with fish capture in the wild. Taxis, or attraction to the anode, can be very important in waters of high turbidity and in areas of extensive structure along the shoreline or in deep water where fish may be immobilized beyond the range of the netter to capture them. Another aspect to be considered is the effect of various pulse shapes on injury or stress to the fish. We did not attempt to assess fish injury or other damage in this study, but most of the fish recovered normal activity soon after the electrical treatment was terminated. The video recordings indicated little or no taxis at 60 Hz for the saw-tooth, parabolic and exponential decay pulse shapes. The rectangular pulse did appear to produce some taxis to the anode. Fish recovery was variable for all pulse shapes; some fish recovered almost immediately, some took several seconds, and some appeared narcotized until removed from the test tank.
Let me say thanks to Carl Burger for coordinating the research trip, to Jackson Gross for getting the fish, to both Jackson Gross and Yale Smith for setting up the fish tanks, to Phong Nguyen for synchronizing the fish and videos, to everyone else who helped, and especially to Lee Carstensen for taking the time to set up and operate the power supply and waveform generator. My hope is that it provided useful information for electrofisher design and operation.