A Review of Methods to Reduce Bird Predation
on Land-Based Fish Farms
Sherrene D. Kevan
October 1992

We wish to thank the following individuals and organizations for their assistance: G. Lutz (Louisiana State University Agricultural Center and Cooperative Extension Service); D. Mott (Denver Wildlife Research Center, Mississippi Research Station); D. V. Weseloh (Canadian Wildlife Service - Burlington); I. Price (Dept. Fisheries and Oceans - Ottawa); A. Holgate (Institute of Terrestrial Ecology, England); A. Kerksieck (Fish Farming Experimental Laboratory, Arkansas); P. Gorenzal (University of Davis, California); B. Robinson (Canadian Wildlife Service - Ottawa); V. Thomas (University of Guelph); U.S. Fish and Wildlife Reference Service (Bethesda, Maryland).

We would also like to extend our gratitude to J. Carreiro, B. Robinson, D. V. Weseloh and H. Blokpoel for their useful comments. This project was funded by the Canadian Wildlife Service, Environment Canada, Nepean, Ontario, to which I am grateful for permission to publish this document. Thanks also to Richard Moccia and the Aquaculture Extension Centre for support to copy and distribute the report to fish farmers in Ontario.
1.1 Effects of predation by birds

Predation by birds on commercial fish farms is an on-going problem that is encountered world-wide (Kevan & Weseloh 1992; Rugeggeberg & Booth 1989; EIFAC 1988; Kiskaroly & Tafro 1987; Salmon et al. 1986; Ashkenazi 1985; Cottam and Uhler 1936).

Piscivorous birds like the Great Blue Heron (Ardea herodius), Belted Kingfisher (Ceryle alcyor) and Osprey (Pandion hatiaetus) are naturally attracted to fish farms where open ponds, tanks and raceways stocked with fish, provide feeding opportunities. However, birds can directly cause extensive damage by consuming and wounding fish. Indirectly, damage from bird predation can result when the feeding or spawning habits of fish are disturbed (EIFAC 1988). Birds can also act as hosts to parasites and therefore transport infectious fish diseases between or on to fish farms (Northern Aquaculture 1991; Stickley 1990).

Although there are many deterrent devices commercially available (e.g., automatic exploders, pop-up scarecrows), some devices are not effective for large farm operations (Hanebrink & Byrd 1989). Deterrents such as complete enclosures may be cost-effective for smaller facilities but not for larger ones. Fish farmers are continuously searching for methods and devices to control and deter piscivorous birds from aquaculture operations. For them, the preferred deterrent should be cost-effective and applicable (Parkhurst et al. 1987).

With an increase of habitat loss for piscivorous birds, along with an increase in new commercial fisheries operations, there is a likelihood of more predation on fish farms (Parsons et al. 1990; EIFAC 1988). Therefore, a review on preventive and control methods used to deter birds from freshwater fish farms is timely and warranted.

1.2 Purpose

The purpose of this literature review is to address the problems of bird predators on freshwater fish farms worldwide. The main objectives of the review are:

1. To gather and review information on past and present deterrent devices and methods used to control birds
2. To assess the advantages, disadvantages and the use of deterrents and,
3. To make recommendations on those deterrents that may be applicable for use on aquaculture facilities in Ontario.
2. 1 Automatic Exploders
Automatic exploders are small "cannon-like" devices which run on propane or acetylene (Salmon et al. 1986; Booth 1983). The devices emit a loud, explosive blast which can be adjusted to explode at different time intervals. They are also mobile, and one cannon may be used to cover a farm site area of 1.3-2 ha (3-5 acres), if combined with other deterrent devices (Salmon et a/. 1986). This device is very useful as a temporary control measure for scaring away migrant birds (Salmon et a/. 1986).

It is best to begin frightening/scaring methods before bird predators appear. Once the birds start feeding, their habits become difficult to break (Salmon et a/. 1986). Foraging birds attract more birds.

Using a combination of the automatic exploders with other deterrent devices (e.g., netting) is common practice (Salmon et a/. 1986; Booth 1983). If birds become accustomed to the devices then changing methods more often is required (Littauer 1990a). Exploders should be moved every 2-3 days to a different location on the farm site. An exploding device which has a timer and a rotary mount that changes the interval and the area to which the sound is directed can be more effective (Littauer 1990a; Nomsen 1989). If the exploders are elevated, interference of sound dispersal by other equipment, buildings etc., is reduced (Salmon et al. 1986; Booth 1983). Two exploders stacked one on top of the other on a rotary bar, facing in opposite directions have also been successfully used (Nomsen 1989).

Nomsen (1989) found that the noise of the blast can be amplified by directing the firing of the cannon through a hole in one end of a small steel drum from which the other end has been removed.
The use of automatic exploders requires the device to be used along with other control methods (e.g., netting) (EIFAC 1988; Salmon et al. 1986).

Most automatic exploders are effective only temporarily in scaring away birds (Coniff 1991; EIFAC 1988; Moerbeek et al. 1987; Ostergaard 1981; Spanier 1980; Lagler 1939). In some cases birds are frightened off only to other neighbouring ponds (Martin 1982; Brown & Gratzek 1980). Birds have been known to become accustomed to the noise (Moerbeek et al. 1987; Salmon et al. 1986; Meyer 1981; Brown & Gratzek 1980) especially if short time intervals between explosions are used.

Loud noise emitted by the exploders can agitate spawning catfish (Salmon et a/. 1986) and may also disturb businesses and homes neighbouring the fish farm (EIFAC 1988).

2.2 Lights
Strobe lights (similar to those on aircraft), revolving light units, and amber barricade lights, are useful especially for deterring night-feeding birds such as herons (Littauer 1990a; Nomsen 1989). The light produces a blinding effect which causes the birds to become confused and restricts the birds ability to catch fish (Salmon et a/. 1986).

Covering unprotected areas of a fish farm using reflective tape, along with increasing the number of lights (Salmon et a/. 1986; Booth 1983) has successfully deterred herons.

Revolving/moving lights are most effective when placed on the banks of fish ponds or raceway walls (Salmon et al. 1986).

There have been mixed results from using lights. Birds may become accustomed to lights quickly (Littauer 1990a; Salmon et al. 1986; Meyer 1981). Black-crowned Night Herons (Nycticorax nycticorax) have been shown to avoid the bright glare by landing with their backs to the lights. Other deterrent methods must be incorporated when using lights.

2.3 Alarm/Distress Calls
Distress calls have been used successfully for discouraging migratory birds from farms in rural and urban situations (Salmon et al. 1986; Booth 1983; Busnell & Giban 1968). Some studies have shown that this method can be particularly effective for deterring some species of birds (e.g., gulls, night herons) for many months (EIFAC 1988).

In Haifa, Israel more than 80% of visiting Night Herons were frightened off trout ponds when recorded distress calls of juvenile and adult (combined) night herons were broadcasted (Spanier 1980).

Taped recordings of alarm/distress calls of birds are broadcast over the area of the farm in order to repel or frighten other predatory birds (Salmon et al. 1986; Busnell & Giban). Playing the calls at different times using a timing device with random repeats of calls may be necessary. It is best to broadcast alarm/distress calls as soon as birds arrive on farm sites (EIFAC 1988; Salmon et al. 1986). When using a stationary unit. louder volume achieved better results (Booth 1983).

Mobile broadcasting units can be more effective than stationary ones (Booth 1983).

Distress calls broadcast from speakers that are mounted on a vehicle can work effectively and also require limited manpower (Littauer 1990a,b).
The alarm distress calls must be re-enforced with other deterrent methods (Salmon et al. 1986; Busnell & Giban 1968). The calls may loose their effectiveness over time as the reaction to a call varies with the location, species of bird, the size of the fish farm site and the time of year (EIFAC 1988).

Other sounds produced electronically (human voice, automobile noise, etc.) are not as effective as alarm/distress calls of birds (Booth 1983).

2.4 Pyrotechnic Dispersal Devices

2.4.1. Noise bombs, Whistlers and Rockets
The pyrotechnic dispersal devices (PDD) include noise, claw, bird, racket and whistle bombs, bird whistlers, and noise rockets (Salmon et al. 1986; Booth 1983). Rope-firecrackers are another type of PDD (discussed separately) are useful for harassment of birds.

Some PDD use 12-gauge exploding shells from single-barrel, open-bore shotguns (Booth 1983). The range of the exploding shells is approximately 45-90 m (50-100 yards) (Salmon et al. 1986). Other PDD are fired from 15 mm or 17 mm pistols and have shorter ranges (33-69 m or 35- 75 yd) than the 12-gauge shotguns. Most PDD are commercially and readily available.

For best results the projectile should be shot directly into the air above the birds. Cleaning and checking the barrel after firing a gun should be done systematically to remove any debris. The use of PDD requires a permit from government agencies such as the Canadian Wildlife Service, and it is best to check with authorities regarding the discharge of firearms in local areas.
The use of PDD requires manpower which may not always be available (Rueggeberg & Booth 1989). Continuous cleaning and checking of the barrel after each round fired is necessary to insure that corrosion and/or obstructions do not remain (Salmon et al. 1986; Booth 1983). Injury to fish farm personnel could occur if the safety procedures are not carried out.

Pyrotechnic devices have proven to be ineffective for control of cormorants (Moerbeek et al. 1987). Godin (1986) indicated that whistle and racket bombs could be ineffective against scaring birds if the bombs were used frequently.

2.4.2 Rope-firecrackers
Rope-firecrackers, also known as agricultural explosive devices, are inexpensive, commercially available, and require little manpower (Booth 1983).

Fuses of the firecrackers are inserted through a 8 or 9.5 mm (5/16 or 3/8 inch) cotton rope. As the rope burns, the fuses ignite causing a series of loud explosions. It is best to place rope-firecrackers over something which will help to prevent a fire, such as a steel drum (Booth 1983). The burning speed and test time of a section of rope, along with weather conditions, should be checked before putting these deterrents onto the farm site. The burning speed of the rope can be increased by soaking it in a saltpetre solution (3 ounces/quart or 84 g/.95 L of water) and allowing the rope to dry (Booth 1983).
Using rope-firecrackers with other deterrents is necessary for scaring away birds (Salmon et al. 1986; Salmon & Conte 1981).

Weather conditions (i.e., wind, humidity) can affect the burning speed of the rope. There is also the danger of creating a fire hazard and preventive measures for fires should be in place before using the devices.

Creating loud noises around livestock and human habitation close to the farm site may be cause for concern (EIFAC 1988; Salmon et a/. 1986).

2.5 Water Spray Devices
A rotator device for protecting circular fish ponds was developed by the Swedish Salmon Research Institute in Alvkarleby (Svensson 1976).

The rotator, which is made of galvanized steel pipes and other fittings, is powered by the gravity flow of water at approximately 75 L/min (for each rotator). The arms of the rotator, which is shaped like a cross, can be extended to the pond/tank dimensions. When in use, the arms of the rotator sprays a "mist-like" cover of water over the tanks/ponds providing shade and possibly oxygenation of the water (Svensson 1976). The spray also helps to diffuse direct sunlight and reduce the amount of light penetration into the tanks to protect fish. Most importantly, the spray of water reduces visibility and prevents birds from seeing fish (Littauer 1990a; Mills 1985). It is necessary to obtain sufficient water pressure and to operate sprinklers on an on:off cycle to be effective (Salmon et al. 1986).
The water spray device may not be applicable for deterring wading or diving birds, and there is no available literature to support this use. It was designed mainly to deter gulls and terns (Svensson 1976) from circular tanks and may also be suitable for smaller ponds. Its application for raceways requires further research.

This method mayor may not interrupt fish farm operations (i.e., feeding or spawning) and therefore timing of the spray activation should be considered.

2.6 Visual Deterrents
The range of visual deterrent devices includes scarecrows, bird silhouettes and models (e.g. owls, raptors, foxes), model aircraft, whirling novelties, tethered balloons, and small aircraft such as helicopters and ultralights (Moerbeek et al. 1987; Booth 1983). Other types of visual deterrents include dogs and human activity. Most visual deterrent devices are effective when they have been put out on the farm site before the birds appear (Nomsen 1989).

Scarecrows: Scarecrows with a pop-up feature and that fire a double propane exploder, have shown some success for deterring birds from fish farm sites (Littauer 1990a; Nomsen 1989). Several scarecrows may be necessary to deter birds. Putting scarecrows in isolated areas of the farm can be effective for shorter periods when carrying out intensive scaring and harassment methods (Littauer 1990b).

Reflectors, streamers. whirlers and silhouet1es: Streamers, reflectors, and spriolum whirlers [metal strips commonly used in advertisement for business (e.g., gas stations)], and hawk and bird silhouette kites, have some short-term success when used with other devices (Littauer 1990a; Nomsen 1989; Martin 1982).

Small aircraft: Helicopters, airplanes and radio-controlled aircraft have all been used to "haze" birds away from ponds, tanks and raceways (Coniff 1991; Mott 1991; Litthauer 1990a; Moerbeek et al. 1987). A radio-controlled aircraft can be used to haze birds when they are attempting to land on a farm site (Littauer 1990a). An advantage to using the hazing method tc scare birds is that the structure of a pond does not have to be changed (Bradford et al. 1991).

A helicopter flown 30 and 150 m above roosting sites, along with firing of pyrotechnics reduced the number of cormorants at those sites after harassment (Mott 1991).

Human activity: Human activity around the farm site has been known to reduce heron visitation (Kevan & Weseloh 1992; Drauland and Van Vessem 1985). Fish farmers working the farm 3-5 times daily decreased the number of herons around the site, and also increased the delay in return of heron visitation (Draulans and Van Vessem 1985). An increase in human activity patrolling during dawn and dusk hours of the day was effective in discouraging birds, such as herons which like to feed at these times (Brown & Graztek 1980).

Dogs: Dogs have been used successfully on some farm sites in Ontario (Kevan & Weseloh 1992) and elsewhere in North America (Ruggeberg & Booth 1989; Brown & Graztek 1980). Dogs that are trained to chase away birds immediately as they land, appear to be the most effective. Putting dogs on a slip-wire or, tethered dogs moved periodically to a bird problem location works well also.
The use of helicopters, airplanes and radio-controlled aircraft have limited success (Moerbeek et al. 1987). There is no longterm effectiveness because birds are frightened from the ponds for a short time and then return after the aircraft has left the site, especially when there are no other ponds or suitable wetland areas for the birds to go to (Bradford et al. 1991.) Weather and expense also decreases the effectiveness of scaring off birds (Mott 1991; Littauer 1990a). Many trips are needed to keep birds off ponds. Manpower, fuel, maintenance and machinery costs may not make it economically feasible to employ this method. An estimated cost for hazing was calculated to be at $251.00 (U.S.) per ha-m annually (Bradford et al. 1991).

Human activity to cause disturbance is not the most effective method for scaring birds. If the birds acquire the habit of landing then it becomes harder to scare them off. The more birds there are (as in the case of herons) the less they are likely to be affected by human disturbance. Drauland and Van Vessem (1985) found that the success rate of herons catching fish is highest directly after arrival at the farm. Therefore, it is best to have preventive control measures already out on the farm site that would keep birds from landing. Human activity would not be effective in this respect.

The effectiveness of using streamers, reflective tape, whirlers and silhouettes to scare birds has not be determined (Littauer 1990a). Weather conditions such as high winds, can affect the usefulness of these devices.

2.7 Shooting
Shooting problem birds on fish farms was once a common deterrent method conducted worldwide (EIFAC 1988). However, at present, most bird species are protected by law in Europe and North America and most of these countries require a permit or licence to eradicate problematic birds.

The short term advantages to shooting and/or killing birds is that this control method is immediately effective and may give the fish farmer some self-gratification (Williams 1992; Pirque 1990; EIFAC 1988).

Shooting can work effectively if combined with other deterrent devices (Nomsen 1989). Repeated firings keep birds flying away from the farm site. Shooting at a 30 degree arc from the end of the muzzle is recommended, and the shooter should never fire directly into the sky (Nomse 1989). A .22 rimfire rifle was noted as working as well as or sometimes better than a shotgun.

A controlled shooting has been suggested for some countries like Yugoslavia and German (Kiskaroly & Tafro 1987). It would take place annually for one month in order to keep fish farmer from completely eliminating some species of bird predators. Supplemental killing of a limite number of birds has also been advised (Littauer 1990a). When the fish farmer has found that the birds have become accustomed to the devices and techniques used to scare them it may be necessary to kill some birds to instill fear in other birds. However, a permit to kill is required, and also verification from the fish farmer that his/her methods to scare the birds have failed.
Although killing birds is illegal without obtaining a permit, this has not stopped fish farmers (Kevan & Weseloh 1992; Williams 1992; Coniff 1991; EIFAC 1988; Ross 1988; Meyer 1981). Estimates of 3600 to 5000 Great Blue Herons per year were shot on fish farms in England, and close to 400 Great Blue Herons annually in Ontario, Canada.

Studies carried out in Europe on herons found that shooting them had little impact on fishing success and heron visitation (see EIFAC 1988 for review). Because birds have the highest success rate at catching fish right after landing, birds need to be shot directly after arrival, and this may not always be possible.

Killing herons at fish farms may have a detrimental effect on their populations (Meyer 1981). Juvenile herons have been recorded as spending more time on fish farms because they are less adept at catching fish. This may lead to farmers killing more younger birds, resulting in a less than stable bird population. A severe winter, and killing, may cause a heron population crash (Meyer 1981).

Discharging firearms around farm sites may have legal and safety ramifications, therefore, local laws and regulations should be checked before farmers consider this method (Salmon et al. 1986).

2.8 Physical Barriers

2.8.1 Complete enclosures
Complete enclosures using screening mesh and/or netting on tanks, ponds and raceways is considered the most effective way to deter all fish-eating birds (EIFAC 1988; Salmon et al. 1985; Martin 1982; Cottam & Uhler 1936).

Mesh panels that were put on top of raceways have been effective in hatchery operations in the United States for temporary or seasonal use (Salmon et al. 1986; Salmon & Conte 1981).

Results from a survey conducted by Parkhurst et al. (1987) to hatchery managers in the U.S. found that top screens were one of the most effective control methods used to deter all fish-eating birds. Secure-fitting top nets with the appropriate mesh size can successfully keep herons out of tanks (Carass 1990; Ueckerman 1981). In the Netherlands, complete enclosure using 3-5 cm mesh attached to frames or overhead wires was successful at deterring diving ducks on ponds (EIFAC 1988).

Light-weight netting used to cover tanks, ponds and raceways is readily available in various mesh sizes. Ponds that were protected with netting that was firmly secured on the sides and to the ground keep herons and other birds from gaining access by going underneath the net. Stringing the nets above the water surface of the pond, as high as possible, is generally advised. Salmon et al. (1986) suggest using 3-5 cm (1 1/2 - 2 in) mesh netting that is attached to a frame, or secured to overhead wires.

Various materials for construction of cover (top) screens has been used. Wood slat and tar paper, chicken wire, cloth, and mesh or netting are common (EIFAC 1988; Salmon & Conte 1981; Lagler 1939). The size of the mesh should correlate with the species of predator that is to be deterred.

For the protection of fry and fingerlings stocked in small ponds, creating a cage-like enclosure with a gate, high enough to work in easily, can be effective against all bird predators, especially for Great Blue Herons (Meyer 1981). Danish fish farms use this control method (EIFAC 1 988).
The cost of complete enclosure may be more expensive than other deterrent methods and could be viewed as a disadvantage at first. However, initial costs may be recovered from reduced losses of fish, and from reduced costs from using other less efficient control measures (EIFAC 1988; Salmon et al. 1986; Meyer 1981; Salmon & Conte 1981).

Using cover or top screens on ponds and raceways can interfere with farm maintenance such as feeding and cleaning. Therefore seasonal or temporary use has been suggested (Salmon et al. 1986; Lagler 1939).

2.8.2 Partial Enclosures
Partial enclosures are more advantageous to use on large fish farm operations (EIFAC 1988; Ostergaard 1981; Lagler 1939). Types of partial enclosures include using lines, overhead wires, screening and netting.

Screening and/or fencing ponds, tanks and raceways can provide protection from herons (and other wading birds) (Avault 1988; Bloomfield 1981; Morrison 1975; Naggiar 1974; Pough 1936, 1940, 1949) as well as kingfishers (Melaceryle alcyon) (Slayer & Lagler 1946; Louis 1936). A twine grid that was used at 5 ft. intervals was effective against heron predation during the day and night (Martin 1982b). Strands of polypropylene line that were stretched between posts 20- 35 cm above the ground prevented herons from feeding on the banks of ponds (Meyer 1981). Using a chain of white polyethylene floats along the edges of the pond with a length of twine posted around the perimeter of the pond was even more effective against Great Blue Herons because the herons would not go over the floats or feed in between them (Meyer 1981).

For deterring gulls, Ostergaard (1981) used 50 lb. monofilament fishing line that was stretched lengthwise on the sides of raceways. The routine farm duties of feeding, sampling and cleaning could be carried out without removal of the lines.

Overhead nylon lines, made into 20 x 20 m squares and fixed to metal poles, were effective in preventing large groups of cormorants from visiting a fish farm (Moerbeek et al. 1987).

One method of partial enclosure for ponds would be to erect a 1 m (3 ft.) high fence in .75- 1 m (2-3 ft.) of water. Putting a net with small mesh size around the pond can keep small fish from entering the shallow area (Salmon et al. 1986). For raceways, the fences should be high enough to prevent birds, such as herons, from feeding from the walls. The spacing of overhead wires should correspond with the birds habits. The following guidelines for spacing of overhead wires was suggested by Salmon et al. (1986): 126 cm (4 ft.) for gulls; 63 cm (2 ft.) for mergansers; and 31 cm (1 ft.) for Great Blue Herons or Fosters Terns. Placing netting along the sides of a pond where overhead wires are used, provides protection against birds maintaining access on the sides.
Partial enclosure is less expensive and uses less materials, but this method cannot exclude all fish-eating birds from predation (Salmon et al. 1986; Salmon & Conte 1981). Partial Screening was found to be useless in deterring kingfishers at many European fish farms (EIFAC 1988). Overhead lines, successful at deterring large groups of cormorants, did not deter single birds which were able to land or go in between the wires and prey on fish (Moerbeek et al. 1987). However, several factors may have contributed to this method being unsuccessful. Not all the ponds were covered during the observations, and deterrents were not set up before the birds had established a pattern of landing and feeding.

Using partial enclosure with screening may interrupt routine maintenance on the farm if lines are constructed in a way that keeps wading birds from walking under the wires (EIFAC 1988).

Aesthetic reasons may also have to be considered in some cases, as partial screening can be unsightly (EIFAC 1988).

2.9 Holding Ponds and Raceway Designs

The number, size, density and design of holding ponds and raceways may be a factor that determines where fish-eating birds feed (Parkhurst et al. 1987; Salmon et al. 1986; Meyer 1981).

Keeping water levels in ponds and raceways at different depths to reduce predation by birds has been suggested since the early 1930's and is still considered as an acceptable method at present (EIFAC 1988; Meyer 1981; Naggier 1974; Neff & Meanly 1954; Pough 1941; Johnson 1937). Keeping pond water at least 1 m deep, or 14 inches down from the top of the pond embankment can also deter wading birds (EIFAC 1988; Meyer 1981). Water levels kept at 1 m below the top of vertical walls of raceways has been suggested for keeping wading birds such as herons, from feeding (EIFAC 1988).

Great Blue Herons have been attracted to ponds that contained water levels that were near the top of the embankment. The herons were able to gain greater visibility for finding fish this way.

Ponds that were constructed with steep (approximately 70 degrees) or vertical banks, successfully deterred herons in Britain (Meyer 1981) and is also recommended for ponds in Canada (OMAF 1991) and the U.S. (Salmon et al. 1986). However, it may not be cost-effective to construct deep slopes, especially for larger ponds. Steep banks may be subject to erosion, therefore needing continuous repair and restructuring. Addition of clay to the soil may reduce erosion effects. Other suggestions are to add lime to the soil or use geotextiles with a granular fill cover to reduce erosion (Bradford et al. 1991). Less expensive slope stabilization would require using a floating windbreak device to reduce the amount of wave action in the ponds.

Salmon et al. (1986) recommended that rectangular ponds be constructed because the birds would frighten more easily because of their being closer to "the proximity to shore".

2.10 Location of Farm Sites

The habitat that makes up, and surrounds the fish farm can determine the species of birds that may later on become a predator problem (Parkhurst et al. 1987). Carefully selecting a site that is located away from a major route of migration of fish-eating birds, and also one that is located way from heronries and breeding colonies may have less predation problems (EIFAC 1988; Parkhurst et al. 1987; Salmon et al. 1986; Meyer 1981; Salmon & Conte 1981). Meyer (1981) found that predation on fish by herons was largely because of farms that were close to heronries.

A farm site in Lelystad, Netherlands, consisting of 218 ha of fish ponds, had severe problems with cormorants. The farm was located 13 km away from a colony of 4500 breeding pairs (Moerbeek et al. 1987). Kingfisher predation at fish hatcheries that were located near the Great Lakes (in the USA) was higher compared to those hatcheries that were further inland (Lagler 1939).

2.11 Buffer Populations

Mixing valuable large fish stock (i.e., spawners) with smaller fish (e.g. mosquito fish (Gambusia affinis) in ponds may reduce losses to bird predators (EIFAC 1988; Lagler 1939). Including another species in fish farm tanks can distract the predator and reduce the loss of individuals from the valuable fish stocks. Jurek (1974) suggested establishing a buffer population of mosquito fish near a hatchery in California to use as prey for piscivorous birds.

Studies carried out on warm-water fish farm dams in New South Wales, Australia (Barlow & Bock 1984) found that predation by cormorants (Phalarcocorax spp.) was reduced when dams were stocked with abundant populations of crayfish (Cherax destructor) along with low fish stocking rates.

Building "buffer ponds" near a fish farm site stocked with frogs and toads has also been suggested (Lagler 1939) but effectiveness of this method has not been proven nor reported in the literature.

2.12 Stocking Density and Management Practices

Juvenile Great Blue Herons have been observed in areas of fish farms that have a "high density of active fry" (Meyer 1981). It is advisable to keep smaller fry (4 oz. to 9 oz.) in areas of the farm site that are not located near the approach route of birds nor in areas that are remote and rarely disturbed. These precautions help to reduce predation (EIFAC 1988; Meyer 1981). Fish stocked in ponds that are nearer to buildings, etc. where there is human disturbance, can deter most piscivorous birds (Salmon et al. 1986; Salmon & Conte 1981).

Hatchery managers for the U.S. Department of Fish and Wildlife report that delaying the transfer of fry from inside hatchery facilities to outdoor ponds and raceways can effectively reduce the predation and temptation to birds (Parkhurst et al. 1987). Moerbeek et al. (1987) found that predation by cormorants was reduced when carp fry were put into open ponds near the end of the month of June.

2.13 Farm Site Maintenance

Stopping birds from nesting, roosting, landing and perching has been tried with some success (Schram et al. 1987; Mills 1985). Vegetation that encourages birds to perch or nest near the farm ponds and raceways may need to be removed (EIFAC 1988; Brown & Gratzek 1980). Lagler (1939) reported that removing perches (i.e., tree branches) reduced predation by kingfishers.

Removing and disposing of dead fish, while maintaining netting and screens in good working order is also suggested (EIFAC 1988).
Choosing the proper control method and deterrent device is dependant upon many factors associated with running a fish farm operation. The type of deterrent that is used to scare birds often depends upon the species of predator involved.

The location and constructional design of fish ponds, raceways and tanks may increase or decrease the incidence of predation. Heavy losses of fish can result when birds have easy access to fry and fingerlings in isolated or less disturbed areas of the farm. Choosing farm sites on migratory routes or flyways of piscivorous birds may also increase visitations and/or predation.

Common deterrent devices such as visuals, water spray, automatic exploders, pyrotechnic's, lights, and alarm/distress calls have variable and limited uses. A combination of these devices is necessary. Success in frightening birds away depends upon the number of devices used, how and where they are administered, and if their use precedes the establishment of the birds feeding habits.

Shooting is still a common method to get rid of problematic birds although most of the killing is carried out illegally. Concerns for humane methods of controlling birds, as well as avoiding legal liabilities make this method less desirable.

The best device that is available to fish farmers is complete enclosures. This is the one method of control that deters all piscivorous birds from predating on fish. The only problem with this device is that it is costly to implement at first. However, the savings from reduced fish losses, and from future purchases of ineffective equipment can help the farmer to regain expenses.

If complete enclosures are not feasible, the use of partial enclosures with netting of ponds and raceways is the next best choice of deterrent. There are many different kinds of partial enclosures and their effectiveness can depend upon the species being deterred. Partial enclosures do not exclude all fish-eating birds.

Creating buffer populations for predators may have some relevance but this method would have limited applications because of costs, space and possible disease factors.

Stocking fish at different times of the year or keeping active fry in areas of the farm where there is more disturbance are successful practices. Removal of vegetation around ponds and raceways can stop birds from perching, nesting and landing.

Methods used by fish farmers to control piscivorous birds depends upon the cost, availability, and effectiveness. The size of the farm and its susceptibility to predators may also be a factor for choosing a particular deterrent. Therefore, based upon the methods and deterrent devices that have been reviewed, it is recommended that the following practices be implemented for use in Ontario:

1. Complete Enclosures
Although this may be one of the most expensive ways to deter birds, it is also the most effective. Materials are readily and commercially available.

2. Partial Enclosures
This is the next best method. Partial enclosures have successfully been used to deter gulls and wading birds (e.g., herons). For large operations this method may be more cost-effective.

3. Stocking Times and Stocking Density
Changing the time and the amount of fish stocked in ponds may help the fish farmer to avoid any predator problems. Knowing about the birds habits (i.e, migration period; breeding) can help the farmer by stocking valuable fry when the probability of visitation by predators is lower. Keeping fish indoors as long as possible, and stocking at the largest size practical, are the two obvious conclusions.
1. Ashkenazi, S., 1985. Coexistence of herons and fisheries. South African Journal of Science, 81 (11):694.
2. Avault, J. W. Jr., 1988. Birds as predators and pests. Aquaculture Magazine, Jan./Feb., pp. 64- 65.
3. Barlow, C.G. and Bock, K., 1984. Predation of fish in farm dams by Cormorants, Phalacrocorax spp. Australian Wildlife Research, 11 :559-66.
4. Bloomfield, C., 1981. 99 percent heron-proof. Salmonid, Sept.-Oct. pp. 16-17.
5. Booth. T. W., 1983. Bird dispersal techniques. Institute of Agriculture and Natural Resources, University of Nebraska, Lincoln, Nebraska, 5 pp.
6. Bradford, D.F., Smith, L.A., Drezner, D.S., and Shoemaker, D.J., 1991. Minimizing contamination hazards to waterbirds using agricultural drainage evaporation ponds. Environmental Management 15(6):785-795.
7. Brown, E.E. and Gratzek, J.B., 1980. Fish farming handbook. AVI Publishing Company, Inc., Westport, Connecticut, U.S.A. 391 pp.
8. Busnell, R.G., and Giban, J., 1968. Prospective considerations concerning bioacoustics in relation to bird-scaring techniques. In: The problems of birds as pests, Eds. Murton, R.K. and Wright, E.M., Academic Press, London, England. pp. 17-28.
9. Carass, D., 1990. "Beak-prints" help in war against aerial invaders. Fish Farmer 13(6):46-47.
10. Conniff, R., 1991. Why catfish farmers want to throttle the crow of the sea. Smithsonian, July, pp. 44-55.
11. Cottam, C., and Uhler, F.M., 1936. The role of fish-eating birds. Prog. Fish Cult. 14:1-14.
12. Draulans, D. and Van Vessem, J., 1985. The effect of disturbance on nocturnal abundance and behaviour of grey herons (Ardea cinerea) at a fish-farm in winter. Journal of Applied Ecology, 22:19-27.
13. EIFAC (European Inland Fisheries Advisory Committee), 1988. Report of the EIFAC Working Party on prevention and control of bird predation in aquaculture and fisheries operations. EIFAC Tech. Paper 51.79 pp.
14. Godin, A.J., 1986. Birds at airports. Cooperative Extension Service, Institute of Agriculture and Natural Resources, University of Nebraska, Lincoln, Nebraska. pp. 95-97.
15. Hanebrink, E.L. and Byrd, W., 1989. Predatory birds in relation to aquaculture farming. Aquaculture Magazine, 15(2):47-51.
16. Johnson, O.H., 1937. Preventing depredations of fish-eating birds. Prog. Fish Cult. 28:38.
17. Jurek, R.M. 1974. American River green heron study, 1974. State of California, Department of Fish and Game. Job 111-5.2. 19 pp.
18. Kevan, S.D. and Weseloh, D. V., 1992. A survey on bird predation at Ontario Trout Farms. Ontario Aquaculture Association Newsletter, Spring Edition, Ontario, Canada. 6 pp.
19. Kiskaroly, M. and Tafro, A., 1987. Some methods of reducing losses caused by fish-eating birds. Veterinarski Glasnik 41 (1): 69- 73.
20. Lagler, K.F., 1939. The control of fish predators at hatcheries and rearing stations. The Journal of Wildlife Management 3(3):169-179.
21. Littauer, G.A., 1990(a). Avian Predators. Frightening techniques for reducing bird damage at aquaculture facilities. SRAC (Southern Regional Aquaculture Center) Publication No. 401, Louisiana Cooperative Extension Service, Louisiana State University Agricultural Center, Louisiana, 4 pp.
22. Littauer, G.A., 1990(b). Control of bird predation at aquaculture facilities. Strategies and cost estimates. SRAC (Southern Regional Aquaculture Center) Publication No. 402, Louisiana Cooperative Extension Service, Louisiana State University Agricultural Center, Louisiana. 4 pp.
23. Louis, W.A., 1936. Herons, bitterns, and kingfishers. Prog. Fish Cult. 17: 16.
24. Martin, M., 1982(a). Impact of Predators on fish farming. Part 1. Aquaculture Magazine Jan-Feb., pp. 36-37.
25. Martin. M. 1982(b). Impact of predators on fish farming. Part 2. Aquaculture Magazine 8(3):44- 45.
26. Meyer, J., 1981. Room for bird and fish. RSPB's survey of heron damage. Fish Farmer, 4:23-26.
27. Mills, D.H., 1985. Predator control. In: Angling and wildlife in fresh waters. Proceedings of the Scottish Freshwater Group and the British Ecological Society, Eds. Maitland, P .S. and A.K. Turner, Institute of Terrestrial Ecology, Glasgow, Scotland. pp. 53-56.
28. Moerbeek, D.J., van Dobben, W.H., Osieck, E.R., Boere, G.C. and Bungenber de Jong, C.M., 1987. Cormorant Damage Prevention at a Fish Farm in the Netherlands. Biological Conservation 39:23-38. 29. Morrison, K., 1975. War on birds. Defenders of Wildlife 50(1):17-19.
30. Mott, D.F., Andrews, K.J., and Littauer, G.A., 1991. An evaluation of roost dispersal for reducing cormorant activity on catfish ponds. Proc. Eastern Wildl. Damage Control, Ithaca, N.Y., U.S.A (in press).
31. Naggier, M., 1974. Man vs. birds. Florida Wildlife 27(12):2-5.
32. Nomsen, D.E., 1989. Preventing waterfowl crop damage. In: Knittle, C. and Parker, R.D. (eds.), Waterfowl, ripening grain damage and control methods. U.S. Fish and Wildlife Service, FS 837. 4 pp.
33. Northern Aquaculture, 1991. Virus found in bird droppings: Fish farm predators carry troublesome virus. Northern Aquaculture Vol. 7(3):12.
34. OMAF (Ontario Ministry of Agriculture and Food), Publication 515. Farm Ponds. Information Branch, Legislative Buildings, Toronto, Ontario. 43 pp.
35. Ostergaard, D.E., 1981. Use of monofilament fishing line as a gull control. Prog. fish-Cult. 43(3):134.
36. Parkhurst, J.A., Brooks, R.P., and Arnold, D.E., 1987. A survey of wildlife-depredation and control techniques at fish-rearing facilities. Wild. Soc. Bull. 15:386-394.
37. Parsons, J. L., Hiscock, E.H.J., and Hicklin, P. W., 1990. Reduction of losses of cultured mussels to sea ducks. ERDA report No. 17. Nova Scotia Dept. of Fisheries, Industrial Development Division, Halifax, Nova Scotia, Canada. 69 pp.
38. Pirquet, K.T., 1990. Predator. Canadian Aquaculture, May/June, 50-57.
39. Pough, R.H., 1936. The Audubon Society urges bird-fish protection. Prog. Fish Cult. 20: 15-16.
40. Pough, R.H., 1940. Blue herons can't read. Bird Lore 42(6):507-516.
41. Pough, R.H., 1941. The fish-eating bird problem at the fish hatcheries of the northeast. Trans. N.A. Wildlife Conference 5:203-206.
42. Pough, R.H., 1949. Whose fish is it. Audubon 51 (2):90-96.
43. Ross, A., 1988. Controlling nature's predators on fish farms. Report for the Marine conservation Society, Ross-on-Wye, Scotland. 95 pp.
44. Robbins, C.S., Bruun, B., and Zim, H. S., 1983. Birds of North America: A guide to field identification. Golden Press, New York, N.Y., U.S.A. 360 pp.
45. Salmon, T.P., F.S. Conte, 1981. Control of bird damage at aquaculture facilities. Wildlife Mgmt. Leaf. No. 475, 11 pp.
46. Salmon, T.P., Conte, F.S., and Gorenzel, W.P., 1986. Bird Damage at Aquaculture Facilities. Institute of Agriculture and Natural Resources, University of Nebraska, Lincoln, Nebraska. 9 pp.
47. Sayler, J.C., II, and Lagler, K.F., 1946. The eastern belted kingfisher, Megacery/e a/cyan a/cyan (Linnaeus), in relation to fish management. Amer. Fish Soc. Trans. 76:97-117.
48. Schramm, H. L. Jr., Collopy, M.W., and Okrah, E.A., 1987. Potential problems of bird predation for fish culture in Florida. Prog. Fish Cult. 49(1 ):44-49.
49. Spanier, E., 1980. The use of distress calls to repel night herons (Nycticarax nycticarax) from fish ponds. Journal of Applied Ecology, 17:284-7.
50. Stickley, A.R. Jr., 1990. Avian predators on southern aquaculture. SRAC (Southern Regional Aquaculture Center) Publication No. 2465, Louisiana Cooperative Extension Service, Louisiana State University Agricultural Center, Louisiana. 8 pp.
51. Svensson, K.M., 1976. Rotator for protecting circular fish ponds against predatory birds. Prog. Fish Cult. 38(3): 152-154.
52. Ueckermann, E., 1981. Technical measures to protect fish ponds and fish farms against the heron. Zeitschrift Fuer Jagdwissenschaft 27(4):271-282.
53. Williams, T., 1992. Killer fish farms. Audubon March-April, pp. 15-22.
University of Guelph
50 Stone Road East
Guelph, Ontario, N1G 2W1