A REPORT ON FISH COUNTING INSTALLATIONS IN IRELAND
COMPTE RENDU DU FONCTIONNEMENT D'INSTALLATIONS POUR LE DENOMBREMENT DU POISSON EN IRELANDE

by/par

C.J. McGrath
Department of Agriculture and Fisheries, Dublin
Ireland/Irlande

ABSTRACT

Sixteen of the twenty-two fish-counting installations in Ireland are electronic tube counters, which give highly accurate counts of the numbers of fish moving upstream. Precautions have to be taken, however, to avoid the blocking of the tube and details of a trash rack are given. An electro-mechanical fish counter, mounted in a fish pass is also accurate. Experimental weir-trap fish counters rely on visual or individual counting although in one such installation an electronic system is being tried.

RESUME

Sur 22 installations de comptage de poisson en Irlande, 16 sont des compteurs électroniques tubulaires, qui permettent un dénombrement extrêmement précis du poisson se déplaçant vers l'amont. Il convient toutefois de prendre des précautions pour éviter de bloquer le tube et l'on donne le détail d'un tamis à débris. Un compteur électro-mécanique, monté sur une passe à poissons, donne également des résultats exacts. Les compteurs à poissons expérimentaux, montés sur des barrages ou des pièges, utilisent des comptages visuels ou individuels; un système électronique est également expérimenté.

CONTENTS

1. INTRODUCTION

2. ELECTRONIC FISH COUNTERS

3. ELECTRO-MECHANICAL FISH COUNTER

4. WEIR TRAP FISH COUNTERS

4.1 Galway Sluice Barrage Fish Trap
4.2 Mill Race Trapping Installation
4.3 Salmon Leap Fish Counting Installation
4.4 The P.E.T. Fish Trap

5. GLENTIES HATCHERY SMOLT TRAP

REFERENCES

1. INTRODUCTION

There are 22 fish-counting installations under the jurisdiction of the Fisheries Division of the Department of Agriculture and Fisheries, Dublin. Sixteen of these are electronic fish counters including an experimental unit; one is an electro-mechanical fish counter and the remaining five are weir-trap fish counters, of which one is experimental in nature and depends on electricity for its operation. The location of each of these installations is indicated on Figure 1.

2. ELECTRONIC FISH COUNTERS

The unit employed at every site in Ireland is that designed and manufactured by Mr. P.J. Sharkey who is retained by the Department of Agriculture and Fisheries on a part- time consultancy basis to further the application of electricity to fishery management and development which includes the development of electronic fish counters.

Mr. Sharkey has been provided by the Department with a field laboratory, suitably equipped, at the Department's salmon hatchery at Glenties, Co. Donegal. His experimental work with live fish, both in the laboratory and under natural river flow conditions, is carried on here with the assistance of officers of the Department under the overall direction of the author. Prototypes are first tested here and developed to the state of practical application.

The complete electronic fish-counting unit comprises the underwater fitment, through which the fish being counted pass, the electronic instrument and the power source. The under- water fitment can be a tube completely submerged in the water or a flume at the water surface. Recent applications to existing weir structures dispense with a special flume arrangement and instead a portion of the surface of the weir is suitably prepared for the purpose and fitted with the electrodes.

The counting tubes have been constructed in the past from a variety of materials including rigid plastic and marine plywood but the Department has now adopted, for general use in installations sponsored by it, a standardized tube 1.600 m in length and 0.460 m in diameter. This is made from glass fibre reinforced polyster resin cast on a special mould to which three circumferential strip electrodes of stainless steel have first been fitted 0.460 m apart. The electrodes are 25.400 mm wide by 2 mm and are in this manner embedded in and are flush with the finished side wall of the tube so that the internal surface of the tube is smooth throughout. The upstream end of the tube is moulded to provide a standard bell mouth entrance to assist in producing streamline flow of water through the tube which is a very important requirement for its successful operation.

Despite the limitation inherent in this device, as regards freedom and speed of fish movement, in the opinion of the writer the tube arrangement is to be preferred to all others for the underwater fitment. This is provided that conditions are suitable otherwise and that suitable precautions can be taken, as described later, to ensure the desirable water- flow conditions through it. It is possible by such an arrangement to eliminate one cause of instability arising from fluctuations of the water mass enclosed by the electrodes and the variations in resistance arising therefrom.

There are situations, however, where conditions are such that a tube cannot be satisfactorily employed and instead, a flume arrangement has to be used. This has been so in the Borland-type fish passes built in Ireland where flumes are fastened to and project from the overfall gate at the upstream entrance to the fish pass. These flumes are usually 1 m wide, 1.500 m in length and 0.500 m deep. They have been made initially from marine plywood alone but more recent units have been made of marine plywood coated with glassfibre reinforced plastic.

The stainless-steel strip electrodes, such as already described for the tube, are fastened to the floor of the flume. The depth of flume specified is to ensure that, in the conditions likely to obtain in reservoirs and especially with wave conditions created by the wind, the electrodes will not be bared but that there will be at all times a minimum depth of 0.360 m of water over the electrode and thereby variations in the resistance of the water mass between electrodes kept within tolerable limits.

In more recent applications, a counting pad has been provided in the apron of weirs. A similar electrode arrangement to that described has been installed at the counting site selected near or at the crest of the weir. The surface of the weir is excavated to receive the electrodes which are fastened to the fibreglass filling put into the excavations for this purpose. In such situations, the nature of flow across the weir is such that, provided turbulent water conditions are absent, a minimum depth of 0.150 m of water across the electrodes is tolerable.

In operation, the two bodies of water between each pair of electrodes are made electrically sensitive by a high-frequency low voltage supplied by an oscillator housed in the control unit. Each constitutes an electrical resistance which forms part of a modified Maxwell Bridge circuit operated asymmetrically with the ratio arms in the counter instrument. The entry of a fish into any one of these bodies of water causes a change in the resistance of it and this alteration results in a voltage change in the ratio arms which is amplified to operate a tally counter. Special provision is made so that the counter is uni-directional in the upstream sense and fish going downstream through the tube are not recorded. Provision is also made to prevent the register of spurious counts caused by fish delaying and moving to and fro in the underwater fitment.

It is also possible to arrange that the counter will discriminate between two or three different fish-weight categories and record accordingly.

Any change in supply voltage, ambient temperature or water conductivity can upset the predetermined balance of the instrument and an automatic error sensing and correcting system is incorporated to maintain the proper balance of the instrument at all times.

These fish-counting units are almost invariably located in remote areas where voltage variations in the power supply can be expected as likewise power failures. The unit is designed, therefore, to operate from a 12 volt 60 ampere-hour nickel cadmium accumulator maintained at full charge by a conventional trickle-charging unit for accumulators supplied by the mains supply. In this way, both these factors can be eliminated as possible causes of failure in the operation of the instrument.

A recent refinement in the instrument has been the incorporation of an arrangement which ensures that only signals arising from the movement of fish upstream is recorded on recording devices. These usually consist of recording milliameters by means of which a time of count can be noted. Information about the pattern of fish movement throughout the 24-hour period over three months acquired in this way is shown in Figure 2.

The maximum rate of count by this instrument is claimed by Mr. Sharkey to be 10 fish per second but experience to date would suggest that, in practice, such a rate of movement is not practicable in the tube arrangement. On 21 June 1967, the number of fish attempting to pass through the counting tube of the fish counter at Galway Sluice Barrage became so great at one period of the day that the tube became blocked with fish for a short time and those held up jumped across the grill barrier into which the tube is fitted. Unfortunately, it was not possible to record the rate of count on this occasion but the maximum number counted in one day in this counter in this year was 1 420 on 9 July.

As well as a time of count recorder, it is possible to integrate other electrical apparatus with the counting instrument so that other pertinent information can be recorded, such as water level and meteorological data. Arrangements can be made to record this information on tape which can be subjected to subsequent computer analysis. However, no such arrangements are in operation in Ireland.

The accuracy of this type counter under conditions experienced in Ireland was originally established in 1958 when the prototype was installed in a Borland fish lock in a hydro-electric dam. In a prolonged series of tests 98–100 percent agreement was recorded between visual and instrument counts when fish passed through the counting flume individually. The accuracy dropped when large numbers of fish were passing through together but even under conditions where several hundred fish passed in an hour, the instrument count did not fall below 93 percent of the visual count.

Since 1958 Mr. Sharkey has carried out modifications to the design of the instrument which have significantly improved the speed of counting which is critical when heavy runs of fish occur but the overriding factor continues to be the pattern of arrival of fish at the counting tube.

Mr. D. O'Leary of the Electricity Supply Board of Ireland has comparative figures for two of these counters in the Shannon system. The Ardnacrusha counter at the head of a Borland fish lock can drop to 90 percent accuracy when large numbers of salmon come up in one lift; while the Parteen counter, which is located at the head of a pool overfall-type fish pass, consistently gives a 95–98 percent accuracy.

Practical experience in the design of such installations in Ireland and their subsequent supervision and maintenance has demonstrated the need for certain special arrangements which should be incorporated or provided for in existing and all future installations to ensure continuous trouble-free operation.

The counting tube or flume should be sited and aligned so as to be centrally located in the probable path of the fish when moving upstream.

The velocity of water flowing through the tube should be from 0.910 to 1.520 m/s to ensure that fish will not delay in entering the tube and, having entered, will be compelled to pass quickly through and will not loiter there moving to and fro.

A minimum depth of 0.150 m should be provided at all times over the electrodes and when this cannot be assured positive arrangements should be made to “switch off” the instrument automatically when the depth of water over the electrodes becomes less than 0.150 m.

Streamline conditions of water flow should be provided in the tube and, in particular, there should be an absence of air bubbles. Where there is an overfall of water upstream of the counting tube site, such as can occur at the notch of a fish pass, the tube should be located at a distance downstream from it where the turbulence and air-entrained waters generated have become eliminated.

The external electrical connexions of the tube or flume should be located on the under-side of the unit and thereby protected from damage by debris washdown or cleaning operations for its removal. The extension of the leads from the electrodes should be located under-ground as experience has shown that when suspended overhead, they can be affected adversely by electrical storms.

It should be possible to remove the underwater fitment easily for inspection and maintenance when necessary.

It is desirable that each installation be inspected for a brief period daily and the check count made to ensure that the electronic equipment is operating correctly and that there is no other malfunction. It is surprising the number of small things which can happen that would interfere with the proper working of the instrument which can easily and quickly be put to right if discovered in time. These, in the majority of cases, arise from factors which have nothing to do with the basic design of the installation as such, such as strands of weeds floating downstream becoming caught in the counting unit.

The electronic instrument should be housed in a building or hut fitted with a heating element capable of maintaining the ambient temperature at 20°C at which the instrument can best operate satisfactorily.

The installation should be inspected once a month by a person who has received a basic course in the operation and maintenance of the instrument. This person should be equipped with a device stimulating the resistance of the underwater fitment by means of which, if there is any malfunction, he can establish whether this is located in the instrument or in the underwater fitment and act accordingly.

It would be an ideal arrangement where one instrument only is being supervised and maintained to have spare units, both for the underwater fitment and for the electronic instrument. This is an essential arrangement where the supervision and maintenance of a number of units is involved as in the case of any malfunction, the item requiring attention can be replaced and returned to the manufacturer for proper servicing.

The underwater fitment should be protected against floating debris and an arrangement found satisfactory in Ireland has been the provision of a trash rack semi-circular in plan with vertical bar spaced 0.300 m apart and projecting below the water surface, but stopping 0.300 m from the bed of the channel (Figure 3). Care must be taken in the siting of this unit where it will not give rise to turbulence or air-entrained flow through the counting unit.

It is apparent from the experience gained to date in the operation of this type of fish- counting installation that one unit is capable of coping with the possible maximum rate of fish movement likely to occur in the majority of salmon rivers in Ireland but more than one unit may be required for the rivers in which there is a larger run and especially should this occur over a short period of time.

Furthermore, these counting installations are most useful if they can be located as near to the head of tidal waters in a river as is possible.

In view, therefore, of the arrangements which must be made at such locations to prevent the movement of fish upstream, except through the comparatively small passageway in which the counting unit is located, considerable expense is entailed in making suitable arrangements, unless there is in the river channel already a natural feature; such as a waterfall or an artificial one; such as a mill weir or hydro-electric dam to which such an arrangement can be fitted.

Mr. Sharkey's work on electronic fish counting has been directed, in the main, in recent years toward the devising of a solution to this particular problem. One method developed for use in low weirs requires that the counting pad arrangement be incorporated direct in portion of the surface of the apron of the weir near the crest which has been prepared for this purpose, as already described above in the comments on flumes. The same electronic instrument is employed for the actual counting and the principle of operation is that already described. However, part of the weir structure not occupied by the counting pad is made a barrier to fish movement by the creation of an electrified zone which prohibits fish movement across it. Two electrodes consisting of galvanized GB tubes 50.800 mm in diameter are fastened to the apron of the weir, 0.510 m apart and energized with 15–30 Volts AC at 400–500 Hz.

The writer has no information about the efficacy of this arrangment at water depth greater than 0.510 m. However, this arrangement was first employed successfully by the writer in 1968 to block salmon movement through Glencullin Bridge, Co. Mayo. Water flow conditions of high velocity obtained there and the barrier was effective up to 0.510 m in water depth; the maximum occurring during the period of operation.

The other method which has been devised by Mr. Sharkey and which he has named the Delta Vee Wide Gap electronic fish counter is undergoing field trials at the Department's salmon hatchery at Glenties where the prototype has been built and is undergoing modifications in the light of practical experience in operation.

This system employes the phenomenon that when a fish exerts itself vigorously, it generates a voltage potential. This is in the milli-Volt range which is further attenuated in the surrounding water. Mr. Sharkey, however, has been successful in developing an instrument which can record this electrical discharge. By placing stainless-steel strip electrodes at or near the weir crest, 0.559 m apart and extending for the full width of the weir structure, the signal generated by the fish in passing can be picked up by the electrodes and translated into an impulse to actuate a recording unit. The field trials have demonstrated that care is needed in the installation and, in particular, special precautions have to be taken to screen out background “noises”, such as generated by stray electric currents, among which are leakages from the national electricity grid network. This is being provided for by incorporating a shunt screen arrangement in the weir surface beneath the electrodes consisting of a weld mesh grid which is suitably earthed. The stainless-steel strip electrodes are fastened to fibreglass reinforced polyester resin mountings. These are constructed by casting dowel shaped box-outs in the weir surface which are subsequently filled with the plastic filling.

The passage of fish downstream will not be recorded and the problem arising from lack of depth of water across the electrodes turbulence and air-entrained water does not arise in the operation of this system.

Indications to date are that the length of weir which can be covered by this system varies with the conductivity of the water crossing the weir and can range from 30 m at 100 micromhos per cm³ to 6 m at 500 micromhos.

No attempt has been made as yet to calibrate this system and, while it is true to say that the results to date are encouraging, it would be premature to give an unqualified approval of it as sufficient experience in the operation of it has not been gained as yet to categorically state that all problems in relation to it have been solved.

The rate of count that will be achieved by this instrument when fully operational will be 10 fish per second and all indications point to such a rate of count being achieved in view of the nature of the counting system, and the possibility of a number of fish to move upstream close to one another.

3. ELECTRO-MECHANICAL FISH COUNTER

One unit which comes under this category has been installed in the submerged orifice type fish pass which has been constructed at the Cathaleen's Fall Power Station on the River Erne at Ballyshannon, Co. Donegal. This instrument has been devised and developed by Mr. P.A. Jackson, Hydrometric Engineer at the power station and the general arrangement is as set out in Figure 4.

The unit consists of inscales screen unit and recording instrument. The inscales are in the form of a truncated pyramid 4 foot long which is placed over the upstream opening in the submerged orifice pipe and by means of which the fish passing through the orifice are directed to the screen unit. The screen unit has an opening 0.300 m by 0.230 m in which there are vertical nylon monofils about 1 mm in diameter and spaced 38 mm apart. The cords extend up out of the water to the counting unit where they are fastened through a series of springs to the structural frame. Movement of the cord causes two metal strips to make contact causing an electric current to flow which operates a numerical register and/or a graphical recorder. The passage of a fish through the screen of cords spreads them apart which pulls down the metal contacts together.

The accuracy of this instrument was checked in 1958 when 3 029 fish were released from a trap in the fish pass downstream of the counter and 3 068 fish were recorded by this counter, indicating an error of +1.3 percent in the counter record.

This counting instrument suitably modified to suit the location was utilized also at the upstream entry to a pool overfall type fish pass where it was found to operate satisfactorily after a method had been devised of shielding it from floating debris.

4. WEIR TRAP FISH COUNTERS

4.1 Galway Sluice Barrage Fish Trap

This unit was designed by the writer and was provided originally to capture smolts passing through the by-wash in the sluice barrage which has been built across the River Corrib at Galway. The sluice barrage is semi-circular in plan and its primary function is to provide a means of discharging flood flow in the river to prevent the flooding of land upstream. Another function is to provide a head of water to supply power to a number of milling concerns fed by canals which continue downstream on each bank of the river channel.

In one of these canals a wire mesh smolt screen is provided at the appropriate time of the year to prevent smolts continuing on down the canal. The smolt screen is placed at an angle projecting downstream at the end of which the by-wash is located.

A horizontal wooden grill, made up of battens ½ inch apart, leads from the sill of the by-wash to a flume. The flume, in turn, leads to a tank in which all fish diverted by the smolt screen through the by-wash and which are sieved out of the water flow by the horizontal grid and diverted into the pipe through the flume, are held until they are counted. A flexible elbow pipe has been provided on the river side of this tank. Fish can be emptied from the tank by means of this pipe and the rate of emptying controlled by lowering and raising it as desired.

The tank is emptied each day in the course of the annual smolt run and the record of smolt movement obtained in this way by visual count is as shown in Figure 5.

Other species of fish are also captured in this trap and a typical record is set out in Table I.

While it can be said that the efficiency of this fish trap is 100 percent, this is in respect only of the fish diverted to it by the smolt fence. It does not represent the total downstream migration of fish in this river. It is possible for fish to pass downstream by the other canal on the right bank and through the gates of the sluice barrage if they should be operated at the time of migration so that there is a discharge of water across them.

4.2 Mill Race Trapping Installation

This is one of the trapping installations built by the Salmon Research Trust of Ireland in accordance with the design of the writer. It consists of a fish fence across one of the river channels leading from Lough Feagh to Lough Furnace in Co. Mayo. At the upstream end of the fish fence there is a conventional trapping arrangement consisting of a chamber with inscales and upstream heck. It is fitted with a brailing device.

At the downstream end of the fish fence there is a raceway constructed on the left bank leading to the smolt trapping installation. This consists of a wooden grill at an inclination of 5°. The water flow escapes back direct to the river channel but the smolts and other fish are retained in a holding pool into which the grill discharges.

The efficiency of this installation is believed to be 100 percent and, according to the Annual Report No. 16 of The Salmon Research Trust of Ireland Incorporated for 1971, the number of fish caught and counted in this installation was as tabulated below.

4.3 Salmon Leap Fish-Counting Installation

This is the second fish-trapping installation constructed by The Salmon Research Trust of Ireland Incorporated in accordance with the design prepared by the writer. It is constructed in the second channel connecting Lough Feagh with Lough Furnace in Co. Mayo and is described in detail by McGrath (see p.447).

A concrete weir incorporating two sluice gates was constructed at a waterfall in a rock gorge. Grills have been provided at a gentle slope leading to the weir crest on the upstream side and leading from the weir crest on the downstream side. By opening the gates to the desired width the major portion of the flow in the channel can be by-passed under the weir; sufficient flow only being allowed to cross it to convey fish to the flume at the downstream end of the grill on this side of the weir. This flume leads into a smolt trap which is housed at the right bank. A brailing basket is provided so that the smolts and other fish entering the tank can be examined and counted visually before being released back into the river channel.

A Denil fish pass is constructed leading fish from below the weir to an adult fish trap upstream of the weir. This is also housed and fitted with a brailing device by means of which the fish trapped can be examined visually and counted before release. The number of fish captured in this installation in 1971 is shown in Table II.

4.4 The P.E.T. Fish Trap

This is a fish-trapping installation being developed by Mr. Sharkey, the prototype of which has been built in the river channel adjoining the Department's salmon hatchery at Glenties, Co. Donegal. Mr. Sharkey has named this the P.E.T. Fish Trap which is an abbreviation for Programmed Electronic Trap. It has been designed for the capture of downstream migrant fish, more particularly eels, but tests conducted with smolts which have been captured otherwise and released upstream of this particular trapping arrangement gave a 65 percent efficiency. It has also been the experience that salmon moving upstream which come into the zone of influence of this trap are likewise captured by it.

The unit constructed to date comprises a conventional fyke net mounted between two narrow walls constructed on the bed of the stream to support the fyke net arrangement and to permit the servicing of the net when in operation. A footbridge leads from the river bank to the wall supporting the net for this purpose.

An electrode array, consisting of galvanized GB tubing 50.800 mm diameter placed 1.220 m apart lead directly upstream from the mid-point of the opening leading to the net. The vertical electrode array is supplemented by a longitudinal electrode made up of similar tubing which is laid along the river bed at the feet of the vertical electrodes.

Parallel to the central electrodes, and 2.440 m from it on each side, there is laid a longitudinal electrode likewise made of the same type of tubing. These are the cathodes as opposed to the central anodes.

The array is energized for the capture of smolts with full wave rectified DC at 60 Volts and for eels smoothed DC is employed at 24 Volts. The experience to date has shown that this is best provided from a battery source with a smoothing capacitor incorporated in the circuit. Any trace of ripple renders it unsuitable for the capture of eels as they become tetanized and are washed downstream by the river flow and are not induced to swim involuntarily into the trapping arrangement.

This installation is still undergoing field tests under natural conditions and, while the results to date are encouraging, it is premature to make any authoritative announcement on the efficacy of the arrangement and its suitability for general application.

5. GLENTIES HATCHERY SMOLT TRAP

This installation consists of a horizontal grid made up of wooden battens a ½ inch apart and extending across the full width of the river channel at a low rock waterfall. The grid is an extension of the bed of the river channel on the upstream side of the waterfall and projects downstream from it. The flow of the river passes through the grid and fish moving downstream are thereby sieved out of the waterflow and are washed into a flume at the end of the grid and extending for the full length of it across the river and discharging into a fish-holding tank at the right bank. This unit, however, is operational under low river discharges only and, at high flood flows, it becomes drowned out and its usefulness as a fish- trapping device is limited accordingly because of this factor.

In conclusion of this account of fish-counting installations in operation in Ireland and desirable safeguards which should be incorporated as demonstrated by practical experience of their operation, it may perhaps be appropriate to emphasize the fact that, in general, before a fish-counting unit can be installed at a site, it is invariably necessary to make alterations to the existing structures or to provide new ones in order to make fish counting possible. The efficacy of fish-counting installations at a particular site, moreover, is dependant to a large extent on the skill and ingenuity displayed in making the site suitable for the particular purpose. The magnitude of the overall cost in mounting fish-counting installations is determined, in many cases, by the cost of the execution of the necessary ancillary arrangements and not that of the actual fish-counting instrumentation whatever type is employed.

Figure 1 Fish counting installations in Ireland (see Appendix for further details of the installations)

Figure 2 Pattern of fish movement over 24 h at Islandbridge Weir on the River Liffey, Co. Dublin in 1968

Figure 3 Details of construction of trash rack

Figure 4 The Jackson fish counter

Figure 5 Number of fish passing through the Galway Sluice Barrage fish trap in 1963 as determined by visual counting

APPENDIX