AN EXPERIMENTAL INVESTIGATION OF THE ELECTRIC FISH SENSORY AND D-ARL"A PROCESSING SYSTEMS DESIGNED TOWARD DEVELOPING PHYWOICAL ANALOGS FOR @Li,4iRDI@"ARE PROT OTYPES H.A-V-iNG SIMILAR CAPABILITIES 0 2 Octobe.- 1973 Experiments Needed to Assess Sensitivity, Range and Effective- ness of the Electric Fislies to Detect Objects and Communicate- Underwater We consider it necessary to perform a number of experiments de- signed to quantify some of the electric fish properties. Four different species of el ectr-Lc f-'.sl.ics have been considered for these experiments be- 2 cause of their basically different @;ysteni:@ used as transmitters and electro- receptors. These fishes are: a. Gymnarchus niloticus 'th -AnAfricanweaklyfresh-waterelectricfish.-wi amedium.-- fixed frequency(260 to 300 Hz) d a efor'm.'' Frequency an compos2ite wav does not change .vith ter--L-1@n-&-ature. It has about seven kinds -of electro- receptors plus displacement,, acoustic and chemical sensors located on or near'the skin. Electroreceptors are located on the whole body but are. more numerous near and on the head and near its roi.,ited W1. Countries or origin Sudan, Nigeria and Ivory Coast and the two Congos2 of Africa. It can grow to a maximum size of 5 feet and has'a life span of about 40 years. The electric organ is located caudally occupying about 1/2 to 2/3 of the fish le ngth. Electric organ is derived from modified muscle spindels. Difficult to obtain and to keep alive. Has strange parasitic diseases affecting the spinal cord. The fish is blind and has only vestigial eyes. 3b. Sternarchus albifrons A South-American weakly fresh-water electric fish with a high, relatively fixed frequency (700-800 kHz) and a composite Nvaveform. The rate of dischar-e is temperature dependent at a rate of between 40 to 60 Hz er de,,r p ree centigrade. Has at least three kinds of electro- receptors plus displacement, acoustic and chemical sensors located on or near the skin. Electroreceptors are located on the whole body and. preferentially on or around the head. Counties of origin: Brazil., Columbia., Venezuela, Guyanas, Argentina.. Bolivia, Ecuador, Peru 2 in South-America and some parts of Central Amer- ca. Can grow to a maximum size of one foot. Life span is at least 10 years. The electric organ is located candally occupying about 2/3 of the fishs' body length. The electric organ is derived from modified -It is a hardy species, easy to maintain and easy to nervous tissue. procure. This fish is also bli2nd and has vestigial eyes. c. Gymno,'.us carapo A South-American weakly fresh water electric fish with a mediurii low variable frequency (30 to 150 Hz) and a composite wave- form. Has multi electro and sensory receptors located on or near the skin. Electroreceptors are located on the whole body and preferenti- ally on or near the head. Counti2es of origin same as for Sternarchus albifrons. Can grow to a maximum size of one and a half feet. Life span is several years. The electric organ is located caudally occupying about 1/2 of the fishs I body length. The electric organ is derived from modified muscle spindels. It is a hardy species but is not as easy to procure as Sternarchus. d. Gna5thonemus petersii An- African weakly fresh-Nvater electric fish with a low variable pulse-form repetition rate signal (5 to 170 pps). Has multiple ampullary and tuberous electroreceptors and sensory receptors located on the body 2 and preferentially on or near the head. Counties of origin located in the subtropical, tropical and equatorial Africa. Can grow to a maximum size of one foot. Life span may be several years. The electric organ is located in the tail and is derivedfrom modified muscle tissue. (It is relatively easy to procure, but is very difficult to maintain it alive for loncrer periods of time2 in captivity.) For all experiments we will use a fibero'lass water tank of 12 R diameter and 4 ft height. Experiment #1 (B) This experiment is designed to establish the ability of an electric fish of tl-.o species mentioned under (a) and (b) to detect metallic or non- metallic objects having different masses and introduced in the 2 water tank at different distances from the fish. The fish will be held in a position A by a vertical nylon net. The nylon-net will be raised and the object intro- duced at position B. The time until the fish will detect the object will be noted by observina the fish which may retract or advance in the direction of the 61.)j ect, depending on the _composition and mass of the object and on 2 the species of the fish. This experiment will be repeated after lining the interior of the tank with aluminum foil (see Fig. 1). The objects con- sidered are: iron, stainless steel (non-magnetic)., wood, plexiglass, each in different dimensions (like 10 cm x 2 cm; 5 cm x 1 cm; 2 cm x 0. 5 cm; etc). The objects will be fixed to a nylon thread and introduced vertically into the tank. 0 Experiment #2 (B) This experiment is designed to assess the capability of,electric fishes to use their naviaation system to avoid obstacles like fine nylon 3 Electric Fish Position "All Nylon Net (Can be moved vertically) Water Tank 12 ft 6 and 4 ft high made of f-:berglass 2 Object to be I.-itroduced Position "B" Fig. 1. The reaction time of the fish will be plotted against different masses of the same material type of object and for the same mass of dif - ferent kind of object materials. A short film 7 may be made to illustrate the reaction of the fish. 4 thread or fine wire (aluminum@, copper cannot be used because copper even in minute quantities will kill these fishes. Again we will use fishes of the species mentioned under (a) and (b). Gymnarchus niloticus is an air breather and cannot be confined in a tube but it can be used in experi- ments with free swimming fishes. It is also a small goldfish eater. This fish w2ill be put in the tank at the point "All (see Fig. 2) behind a nylon net. A double net will divide the tank. A goldfish will be put in a nylon net bag and introduced in the tank at point "B" after the nylon net holding the electric fish has been removed. The reaction of the fish and the avoidance of the net will be observed and filmed. The fish species type (b) Stern- archus albifrons is2 not a fish eater but usually reacts with an escape to a metallic object. In this case a metallic object will be used to force the fish to cross the double net. Its avoidance of the obstacles will be noted and filmed. Both fishes are bl.ind. The experirrent will be repeated usincr a grounded aluminum foil along the inner wall of the water tank. Experiment Tlr3 (B) This experiment is designed to demonstrate the ability of electric fish to detect a magnetostatic field. The arrangement will be similar with experiment fl but instead of using an object in the water we will use a permanent magnet outside the tank. Magnets of 10 kG, 1 kG, and 500 G, will be used at distances of 3 ft from the tank wall or near the tank0 wall at Position "B" or IC''. The fish will be located at position "All behind a nylon net (see Fig. 3). The magnetic field in the tank will be measured with a Hall-effect-probe and a Gaussmeter. The reaction of the fish to the maanetic field will be noted and filmed. The threshold of detection of 5 Electric Fish Position "All Nylon Net (Can be removed) Eventual direction of fish movina toward the Water Tank t@, erclass 12ft x 4 ft prey at "B2l' Nylon Nets Nylon Ba. LT Gold Fish Position "B" Position of Wire Nets Figure 2. 0 6 the magnetic field by different fishes of the same species and of the average of different species will be plotted one a-ainst each other. The magnet will be also moved from I'D" to "El' and the reaction of the fish observed. lectric Fish Position "Ati Nylon Ne+. 2 Fiberglass Water T@Lnk 12 ft x 4 ft high Direction of Movement PO; Permanent Magnet Position "C' Figure 3. The sensitivity of electroreceptors will be calculated, counting 4 the number of electroreceptors per square cm and plotting it against the minimum magnetic field gradient that could be detected by the fish. 7 Experiment #4 (B) This experiment is designed to demonstrate the ability of electric fishes to detect electrostatic fields. The arrancement for this experiment will be similar to the preceding experiment, #3, but instead of a permanent magnet we will use a sphere charged electrostatically to 100, 200t 400t ' and 1000 ESU a2nd isolated on a teflon and plexiglass support. The gradient of the field generated by the charged sphere in the tank will be calculated. The threshold of reaction of the electric fishes (different fishes of the same species and different species) will be established (see Fig. 4). Electric Fish Position "All Nylon 14et 2 Fiberglass Water Tank 12 ft @ x 4 ft hiah "D"o 0 Direction Of Movement tro statically Position Charaed Sphere t> 4 Position 'IC" Figure 4. 8 The threshold of detection of the electrostatic field (stationary or ving) of different fishes of the same species aa inst each other and the MO a average of different species against each other will be plotted. The sen- sitivity of the electroreceptors will be calculated, counting the number of .electroreceptors per square cm2 and plotting it against the minimum electro- static field aradient that could be detected by the fish. Experiment #5 (B) This experiment is designed to assess the ability of electric fish to detect DC, AC repetitive signals, square wave and transient signals. The fish will be positioned behind a nylon net in the tank. Carbon elect- rodes will be put in th2e tank at the opposite end of the tank at one feet apart (see Fig. 5). The four experiments to be performed are as follows: a. DC signals will be applied with a telegraph key monitored by an oscilloscope and attenuated by p@tentimeter and furnished by DC bat- teries. A resistance in series will limit the current (see Fig. 6). The reaction and threshold detection of the electric fis2h of the applied current/ and voltage will be noted. The gradient of the current will be calculated and a measurement of the voltacre at the position where the fish will be put -will be made prior to putting the fish in the tank with two carbon electrodes one foot apart. The threshold detection of the different fishes of the same species will be plotted one against another and the average of differe 2 species will be plotted also each against another. b. The same procedure will be used to find the threshold of de- tection of AC sinusoidal current for frequencies of 5, 10, 20, 50, 1002 200., 5002 1000@ 20002 5000, and 10, 000 Hz (see Fig. 7). The signals will be ap- plied with nonpolarizable Aa, Aacl electrodes and1 measured with the same kind of electrodes. Also frequencies close to the fishes own frequency will be used. 9 c. Square waves of same repetition rate as the sinusoidal currents will be used and applied with Ag, AgCl electrodes. d. The transients will be applied with the help of a network using a pulse transformer., resistors and a charged capacitor (see Fig. 8). The values NNill be calculated and the resulting waveforms will be measured with an oscilloscope. These signals will be applied a2lso with Ag, AgCl electrodes. The reaction and threshold of detection of the signals will be noted and plotted the same way as for DC current. Fiberglass Water Electrodes for measurement Tank of applied current attenuated Fish 2 ft x 4 ft high by the water in the tank NylonNe 1 ft 4 12 ft 12 ft 1 f Electrodes for Applying Different Currents and Signals Figure 5. 10 Shie Id Shielded Cable p 2 Carbon Electrodes 1 ft apart 1 cm free ends Oscilloscope R '7 B Battery (optimum v2oltage will be determined by experi- ments but tentatively set between 2 and 6 volts) pll P23' P3 Potentiometers calculated to have constant impedance together with RI = resistor to correspond to the im@- pedance of the carbon electrodes in water measured with the AC liquid impedance bridge C and R2 = capacitor and resistor to4 suppress sparks when the key is manipulated k = key (telegraph type) Fig. 6. DC signal system. Transmitting Electrodes in the Water Receiving Electrodes in the Water Viaveform Impedance Ag, AgCl Ag$ AgCl Generator Adapter 0 Oscilloscope Oscilloscope Fig. 7. Set-up for sinusoidal AC and square waves. - --------- -- Shield To battery for char,,:-,inc, R j Shielded Cable the capacitor --c T Ag, AgCl Electrode. in IVater JC c3 Oscilloscope c High capacity tantalitic capacitor (val2ue to be experi- mentally established according to the im-edance of the electrodes in the water) CR,;CR = Spark suppressing networks 2 3 4 RJY R2 =Varial,,!,-@ resistors to adapt the circuit to the Ag-AgCl electro@u'es in the water i = Jack for peak current measurement (with a resistor in series and an oscilloscope in parallel) T0 Pulse transformer Fig. 8. Transient generating set-up. 13 Experiments #I(P), 3(P), 4(p), and 5(P) These experiments will be similar to the experiments l(B),, 3(B)'4(B) z 5B but instead of using free swimming fishes we will use fishes of species mentioned in 'IC" and ID" on page 2 like Gymnotus carapo and Gnothonemus petersii or equivalents restrained in a plexiglass 2tube with holes and two c4,@inless steel electrodes at the end (see Fig. 9). Instead of observing the reac@L.'ion of the electric fish to the different stimuli. we will monitor on the oscilloscope and frequency counter the charging rate amplitude and wave- form of the fish response to the stimuli. The latency and habituation win be noted. This time the graphs will show2 the relation between stimulus and response and also the threshold of reaction. With this species of fish and system a better quantification is possible of the ability of fish to detect ob- jects, magnetic fields., electrostatic fields, DC currents, AC repetitive sinusoidal or squarewave signals and transient signals. Plexiglass tube 2 with fish in it and electrodes at the end Fiberglass Water Tank 2 x 4 ft high Amplifier 'Oscilloscope Frequency odes for Experiment T:va- P, Center Object to be detected, E.%-perinient f l(p) Perminent mn-,--,net, Electrostatically cliarf-ed sphere, Experiment 76,--'3(P) Experiment # 4(P) Figure 9. 14 Experiment #6 This experiment is designed to relate the ability of electric fishes to communicate underwater and the distance at which they can manage it (see Fig. 10). The natural noise and the signal strength of the fish will be measured. The original signal will be recorded with electrodes on a magnetic tape recorder and monitored on an oscilloscope. This signal wil2l be played back with the aid of two carbon electrodes in the tank at a place 'If 11 at the distance 11111 from the fish situated at the point llg" and confined there with a nylon net. The reaction of the fish to its own signal will be observed and eventually filmed. The signal will be subsequently attenuated to precalculated attenuations corresponding to distances of 100 ftp 500 2fti 1000 ft and 5000 ft. Jf the -.@i-sh will not react to a certain attenuation a variable attenuation in between the fixed points will be used to assess the distance at which the fish can detect its own signal. Also later noise will be introduced in the form of white noise and its effect on the ability of the fish to detect its own signal will be observed. At the end graphs will be pl2otted for the signal to noise ratio versus distance of signal detection. The field generated by the electric fish will be calculated and plotted on a graph. The experiments #IB, 2B. and 3B will be made in the fiberglass tank as it is and with aluminum foil set around the inner wall of the tank and grounded. The results with or without the grounded aluminum-foil will be plotted on graphs an8d eventually graphs will be plotted for the differ- ences in results with and without grounded aluminum foil. 15 Receivincr Electrodes 0 Ag, AgCl Nylon Net Oscilloscope Amplifier Fiberziass Water Tank 12 ft x 4 ft 2high Transmittine, Electrodes Ag, AgCl Oscilloscope impedance Amplifier nl Tape Recorder Network Electric Fish Position G Nylon Net Fiberglass Water Tank 1 12 ft x 4 ft high Transmitting Electrodes Ag2 AgCl, Position E. Oscilloscope Impedance Amplifier FY, Tape Recorder Network Fic,,ure 10. 16 Experiment #7 In this experiment the electromagnetic field around some electric fishes (the ones which can be confined in a tube) will be measured and compared with values resulting from the calculated field. 0 ol Electric Fish g, AgCl Electrodes 0 9 Electrodes Ag, AaCl Moved in the Direction Oscilloscope Amplifier to the Center '@7 lamplifier Oscilloscope Figure 11. 17