5.  CHEMICALS EMPLOYED AS FISH TOXICANTS

5.1  Introduction to Fish Toxicants

The chemicals tested or employed as toxicants for fish and other aquatic life can be listed roughly under four categories, according to principal uses in fishery management:

Those chemicals that have had at least some documented use in the field are described in greater detail in the following pages. Few have been investigated exhaustively in terms of 1) toxicity to fish and higher vertebrates; 2) efficacy on target organisms in waters of various qualities; 3) residues in fish and other foods of aquatic origin; and 4) residues in the environment. Whereas there are data on the acute toxicity of some of the chemicals to fish and other aquatic life, information on chronic toxicity and effects on reproduction of aquatic animals often is lacking. Also, only a few of the chemicals are registered by regulatory agencies for use as fishery tools.^* Lennon (1967), discussing the requirements for clearance and registration of fishery chemicals in the United States, pointed out that some pesticide and public health authorities in the States and in Canadian Provinces are adopting strict standards governing the use of chemicals in water. For example, the Secretary of the U.S. Department of the Interior announced a new policy on June 18, 1970, that prohibits use of endrin and toxaphene on Federal lands. He further stated that the following chemicals are included on a Restricted List, and are to be used only in small applications if other systems will not work: benzene hexachloride, cyanide compounds, ethyl parathion, Guthion, methyl parathion, and Thiodan.

All toxicants mentioned in this report should be considered as generally hazardous and as contaminants in the environment. All use of them should be made as specific as possible to target organisms, and detailed preliminary and follow-up studies should be accomplished. A checklist of preliminary observations for reclamation of standing and flowing waters was presented by Lennon and Berger (1970). Moreover, mention or discussion of a fish toxicant in this report is not to be considered as an endorsement or recommendation regarding its safety and efficacy.

5.2  Technical Data on Fish Toxicants

The technical data were obtained largely from three sources -- “The Merck index”, edited by Stecher, Windholz, and Leahy (1968); “Handbook of toxicity of pesticides to wildlife”, by Tucker and Crabtree (1970); and “Water quality criteria”, edited by J.E. McKee and H.W. Wolf (1963). The toxicity ratings were based on Appendix 9.2 Combined Tabulation of Toxicity Classes.

 * The U.S. Food and Drug Administration requires that all chemical tools be cleared and registered for specific uses against designated species under precisely defined conditions. Chemicals not registered are illegal for use in the U.S.A.

Ramachandran (1960; 1962) injected anhydrous ammonia at 12 to 18 mg/l (ppm) in fish cultural ponds in India to kill noxious, submersed weeds, to repel or kill fish, and to fertilize the ponds. He added that killed fish are safe to eat, that prawns and frogs in the immediate treatment area are killed, that the ammonia is non-persistent in water, and that the treated water is non-toxic to mammals. Applications of 13 to 40 mg/l (ppm) of the compound in Texas lakes produced selective to total kills of fish, decimation of phyto- and zooplankton, profound changes in water chemistry, but no persistent residues (Klussmann, Champ, and Lock, 1969).

Antimycin is an antibiotic produced in cultures of Streptomyces. Its chemical nature and structure were described by Strong (1956) and Van Tamelen et al. (1961). It was patented as a fish toxicant by Strong and Derse (1964). The first formulation of Fintrol, consisting of antimycin coated on sand grains, was registered in the United States and Canada in 1966.

The effects of antimycin on fish and other aquatic animals in the laboratory were reported by Walker, Lennon, and Berger (1964) and by Berger, Lennon, and Hogan (1969). Experimental applications of the toxicant to ponds, lakes, and streams were conducted by Loeb (1964), Powers and Bowes (1967), Gilderhus, Berger, and Lennon (1969), Callaham (1969), and Lennon and Berger (1970). The performance of antimycin against 49 species of marine fish was investigated by Finucane (1969) in a salt-water impoundment in Florida. Studies on the toxicity of the compound to birds and mammals confirmed its relative safety as a piscicide (Herr, Greselin, and Chappel, 1967 and Vezina, 1967).

The potentialities of antimycin as a selective toxicant were exploited by Radonski (1967) against yellow perch, by Avault and Radonski (1967) and Burress and Luhning (1969a) against undesirable scale-fishes in channel catfish ponds, and by Stinauer (1968) against gizzard shad, carp, white sucker, white crappie, and longear sunfish in largemouth bass ponds. Burress and Luhning (1969b) and Moe (1970) described the use of antimycin for selective thinning of sunfish populations in ponds, and observed that repeated exposures of fish to sublethal doses of antimycin did not affect their reproduction.

The non-repellency of antimycin to fish enables effective control of target fish in littoral zones or in the epilimnion of thermally stratified lakes without harm to non=target fish in open water or in deep strata (Gilderhus, Berger, and Lennon, 1969 and Lennon and Berger, 1970). The advantage of non-repellency also was exploited in Oregon to eradicate problem fish from gold-dredge ponds that are interconnected underground (Sayre, 1969).

Antimycin is pH sensitive and degrades within a few hours at pH's of 8.5 and over. In waters with great diurnal variations in pH, reclamations should be scheduled in the early morning to ensure that target fish get a lethal exposure before the pH rises and deactivates the toxicant. In soft, acid waters, antimycin usually degrades to harmless components within 7 to 10 days. The compound, however, is deactivated quickly and easily with potassium permanganate (Gilderhus, Berger, and Lennon, 1969 and Loeb and Engstrom-Heg, 1970).

The formulations of antimycin on sand grains are dispensed to the water from aircraft, from boats, or by persons on foot. Hand-operated or powered-seed spreaders can be used to discharge a wide swath of toxicant easily and uniformly (Lennon, Berger, and Gilderhus, 1967), or the toxicant can be discharged into the wash of a power boat. The Fintrol-5 formulation releases its antimycin into the water within the first 1.5 m (5 ft) of depth as the sand grains sink to the bottom. The Fintrol-15 formulation releases its antimycin within the first 4.6 m (15 ft) of depth (Lennon and Berger, 1970). These sand formulations are especially useful in treating marshes and water areas that are choked with emergent and non-emergent aquatic vegetation. The sand grains bounce off vegetation and penetrate weed-choked water where circulation is greatly inhibited. Liquid toxicants in such situations usually yield very poor results because they cling to or dry on emergent vegetation and fail to penetrate weed-choked water.

Fintrol-Concentrate is a liquid formulation of antimycin that can be administered to streams by means of drip systems or applied in ponds and lakes as other liquid toxicants are. It, like other liquid toxicants, however, should not be sprayed from aircraft, because the volatile solvents often are lost in the air, with the result that insoluble toxicant rests on the surface film of water.

A cake formulation of antimycin is in development for use in streams (Lennon, 1970a). Preliminary tests indicate that a cake may be suspended in a stream to dissolve and release its antimycin at a uniform rate over a certain period of hours. The cake would eliminate the many problems encountered in extended operation of drip equipment and allow savings of manpower.

With regard to the increasing use of antimycin in streams, Marking (1969b) demonstrated that the toxicant is compatible with the fluorescent dyes, rhodamine B and fluorescein sodium, that can be employed to trace stream flows or the movement of a bolt of toxicant downstream. And, because there are situations where fishery managers may wish to treat tributaries with antimycin and the receiving lake with rotenone, Howland (1969) investigated the interaction of the two toxicants. He concluded on the basis of tests against rainbow trout and bluegill that the toxicants are compatible, and no nullifying interaction occurs upon admixture.

The toxicity of Aqualin to fish was reviewed by St. Amant, Johnson, and Whalls (1964). Following successful tests against goldfish in the laboratory, the compound was applied at 1 to 3 mg/l (ppm) in several small lakes in California, including one ice-bound lake. Goldfish and other fishes were eliminated in each case, and the investigators recommended further studies on Aqualin. They pointed out, however, that the compound is lacrimatory and toxic, and that it must be kept in tightly closed containers and be injected under water by a closed pumping system.

The highly toxic activity of Bayluscide against larval sea lamprey and rainbow trout was reported by Howell et al. (1964). More importantly, they discovered that the compound synergizes 3-trifluormethyl-4-nitrophenol as a selective toxicant for lamprey larvae. Marking and Hogan (1967) reviewed literature on Bayluscide toxicity to plankton, fish, and mammals when used as a molluscicide, and showed that the compound was highly and generally toxic to 17 species of game and rough fish in laboratory bioassays. Small-scale trials in ponds indicate that Bayluscide is rapidly toxic to bullheads, but the toxicant itself is subject to very rapid degradation in the high pH waters where bullheads are often abundant and undesirable. The problems of formulation, residues in water and fish, and product cost must be solved before the toxicant can be registered and accepted in fishery use. In the meantime, a heavy-granule formulation of Bayluscide is used increasingly in sampling deep or turbid streams in Canada for sea lamprey larvae and in eliminating populations of lamprey larvae in estuarine or lacustrine areas of the Great Lakes (Tibbles, Lamsa, and Johnson, 1969).

Huston (1955 and 1956) experimented with calcium carbide as a selective control for carp. Pellets of the compound were coated with beef tallow, paraffin, liquid plastic, or placed in gelatin capsules to make them waterproof and attractive to carp. Following ingestion of the pellets by fish, the coating material is digested, and the carbide reacts with liquid in the gut to form a large quantity of acetylene gas. The inflation of the gut leads to death of the fish. Some dead carp, bullhead, and crappies were picked up during a field test of tallow-coated pellets, but evidence of carbide action was inconclusive.

Chlorine in liquid, gaseous, or powdered forms has been employed for sterilization of fish hatchery facilities since at least the mid-1930's (Connell, 1939). An early large-scale use involved disinfecting and disinfesting a series of earthen ponds, connecting ditches, and a section of a natural stream to control a bacterial disease problem at a trout hatchery in New Hampshire (Davis, 1938). Panikkar (1960) recommended calcium hypochlorite for eradication of fish and tadpoles in partially drained fish ponds. A more comprehensive study on chlorine as a fish toxicant was conducted by Jackson (1962) in the laboratory and in the field. He noted that chlorine must be applied in amounts sufficient to meet the chlorine demand of the water, plus the lethal dosage for the species to be controlled; the toxicant is non-persistent in water; and its greatest potential appears to be in certain partial reclamations or in reclamation of water supply reservoirs where other toxicants may be forbidden.

Personal communications made in connection with this report indicate that chlorine is a preferred fish toxicant and disinfectant at some fish hatcheries and fish farms where rapid control and quick return to production are desired. The ease of neutralizing chlorine with sodium thiosulfate is considered an additional advantage.

Titcomb (1914) was probably the first to apply copper sulfate for the purpose of killing entire fish populations. Use of the compound as a fish toxicant continued, but complete kills of fish were not always obtained, and frequent side effects included decimation of phytoplankton, zooplankton, insect larvae, and mollusks (Smith, 1935 and 1940). These deficiences, plus the introduction of rotenone and other fish toxicants, caused the use of copper sulfate in fish control to decline.

Allison (1964) controlled bluegill reproduction in an Ohio farm pond by dropping crystals of copper sulfate into nests containing eggs or fry, but Beyerle and Williams (1967) had little or no success with the same technique in eight lakes in Michigan. The compound yielded successful selective control of shad, suckers, and bullheads in 10 lakes in Texas where total alkalinity was under 100 mg/l (ppm) (Toole, 1968). Probably the State of Virginia has made greater use of copper sulfate in fishery management in recent years than have other states (Stroud and Martin, 1968). Based on 10 years of tests in the laboratory and field, the toxicant is applied to ponds of 0.4 to 60.7 ha (1 to 150 a) for the selective control of rough fish.

Croton seed powder is the residue after croton oil is expressed from croton seed (Croton tiglium L.). The powder is one of the fish poisons that have been used in China for many years to eliminate predators from carp nursery ponds (Hora and Pillay, 1962).

Some South American Indians use the leaves of Clibadium sylvestre as a fish toxicant. Simple aqueous extracts of the leaves were rapidly and extremely toxic to guppy and goldfish in laboratory tests (Quilliam and Stables, 1968). The typical response of fish was violent activity, followed by loss of coordination, paralysis, and death.

Srivastava and Konar (1966) conducted laboratory bioassays of DDVP against fish and insects, and concluded that the compound is a promising selective toxicant for predaceous fishes and insects, and competitor fishes in fish culture ponds in India. The lethal doses for fish are much higher than those for aquatic insects. Sreenivasan and Swaminathan (1967) recommended the compound as a selective toxicant for fish and as an effective toxicant for tadpoles in fish ponds. Comparative trials by Konar (1969) demonstrated that DDVP is superior to phosphamidon because it is more efficacious against fish, it is not adversely affected by turbidity, and it degrades more rapidly.

Hoff and Westman (1965) tested a 3:2 mixture of Dibrom and malathion (active ingredients) at 0.1 mg/l (ppm) in soft-water ponds in New Jersey, and reported that white perch, chain pickerel, bluegill, pumpkinseed, and possibly other sunfishes were controlled selectively without undue damage to largemouth bass. Meyer (1966) reported one experimental treatment in California which resulted in a 30-percent kill of abundant green sunfish and bluegill. Limited tests elsewhere in hard water failed to demonstrate the selective toxicity of the chemical mixture.

Endrin was reported by Henderson, Pickering, and Tarzwell (1959) to be by far the most toxic of the insecticides to all species of fish and the most toxic chemical that had been tested in their laboratory. The most extensive use of endrin as a fish toxicant appears to have been in Malaya where Soong and Merican (1958) removed all fish from 108 mining pools and fish ponds prior to restocking. Iyatomi et al. (1958) experimented with three formulations of endrin against fish in paddyfields in Japan and found that toxicity may persist for more than a month. Treatment of a small lake in Michigan with 0.008 mg/l (ppm) was only partially successful, and further use of endrin was not recommended (Hooper et al., 1964).

More recently, Bhimachar and Tripathi (1967) reported that endrin at 0.1 mg/l (ppm) has been used to kill predatory and weed fishes in carp nursery ponds in India. Large amounts of charcoal served well to detoxify the treated water.

Some commercial fish farmers report informally that Guthion is very effective for selective removal of centrarchids from bait minnow ponds, but it is regarded generally as unsuited for such use in catfish ponds. Meyer (1965) applied Guthion experimentally in ponds in Arkansas and obtained kills of green sunfish and other undesirable species without harm to channel catfish. He pointed out that water temperature and water quality have little effect on the performance of Guthion; the compound has more potential for fish control than malathion, parathion, or trithion; but Guthion-killed fish cannot be eaten.

Natives in the Lower Amazon Basin in Brazil have used the leaves of the small herb, Ichthyothere terminalis, as a fish poison for a long time (Cascon et al., 1965). The leaves are incorporated into baits prepared with locusts or manioc flour, and the baits are thrown into the water to be swallowed by fish. This use differs from that of rotenone=bearing plants where the toxicant is applied directly to the water.

The active ingredients in the herb leaves are ichthyothereol and ichthyothereol acetate. Cascon et al. (1965), reporting on the isolation and identification of these compounds, stated that the guppy reacts promptly to minute quantities of the poison with extreme agitation and dies after a few minutes.

Lime in its various forms has been employed to control unwanted organisms in fish cultural ponds for many years. Schäperclaus (1933) recommended the use of caustic lime to control parasites and disease-producing organisms in drained ponds, and Markevich (1951) recommended lime-milk applications at 2,500 kg/ha (2,230 lb/a) to sterilize dewatered ponds that had contained diseased fish. Prather et al. (1953) pointed out that hydrated lime used as a disinfectant in ponds would kill undesirable fish, and Hora and Pillay (1962) reported that quicklime often is applied to carp nursery ponds in China to eliminate noxious organisms including fish. More recently, Sidthimunka and Choapaknam (1968) indicated that predators in prawn-culture ponds are controlled with lime.

There is extensive literature on the biological activity of malathion, including its effects on fish and aquatic invertebrates (Walker, 1969). There is a 1,000-fold range in toxicity from a few parts per billion to several parts per million between fish species depending on exposure, temperature, pH, and water hardness. Some private fish farmers make use of this differential toxicity to control predaceous or competitor fishes in production ponds. Al-Hamed (1967) found, for example, that wild fishes could be eliminated from ponds in Iraq without harming the cultured carp. Undesirable sunfishes can be removed selectively from minnow ponds by applying 0.5 mg/l (ppm) of malathion when water temperatures are between 4.4 and 26.7° C (U.S. Bureau of Sport Fisheries and Wildlife, 1970b).

The possibility remains that malathion may have some registerable uses in fish culture if residue tolerances in water and fishery products are established.

Srivastava and Konar (1965) conducted bioassays of phosphamidon against the rohu, and such predatory fishes as cuchia, koravai murrel, nandus, khalisa, climbing fish, and tengra. They concluded that predatory fishes and predatory insects could be eradicated without harm to the carp.

Research on fish toxicants in Russia resulted in the development of polychlorpinene (Burmakin, 1965). The compound is a chlorinated turpentine, resembling toxaphene in some respects. By 1963, 118 lakes in Russia had been treated with the compound at 0.05 to 0.20 mg/l (ppm) to control rough fish, and some tests had been initiated in small lakes in Germany (Schäperclaus, 1963). By 1965, 241 lakes in Russia, totalling 17,000 ha (42,000 a), had been reclaimed with polychlorpinene (Burmakin, 1967). The toxicant persisted up to 1.5 years in lakes in northern Russia, and degradation in water depends on concentration, water temperature, alkalinity, depth, and the extent of water mixing. Small, shallow lakes are preferred for treatment, and intensive management for production of food fishes usually follows the reclamations.

Bizyaev, Antimov, and Moskalev (1965) noted that the search for fish toxicants is continuing in Russia. They added that polychlorpinene has disadvantages such as non=specificity to fish, long persistence in water, and safety problems when used near human populations. In contrast with polychlorpinene, the investigators are seeking toxicants that are highly toxic to fish, harmless to warm-blooded animals, and quickly degradeable.

The rotenone-bearing roots of many plants of the family Leguminosae have been used for many centuries to stun and kill fish by primitive peoples in different parts of the world (Leonard, 1939). Powdered derris root, containing about 4 percent of rotenone, was used first as a fish toxicant in Michigan in 1934, and gained rapidly in favor for reclamation of ponds and lakes. Different uses, different techniques for application, and new formulations soon evolved. Davis (1940) and Greenbank (1941) demonstrated that warm-water fish could be controlled by treating the epilimnion of trout lakes with little harm to the trout. Wales (1942) controlled carp by poisoning coves in a lake where they were spawning. Surber (1948) found that emulsified rotenone is superior to the powdered form. By 1949, 34 states and several Canadian provinces were using rotenone routinely in reclamation projects (Solman, 1950).

When the use of rotenone in fishery management was initiated, a concentration of about 0.5 mg/l (ppm) was advocated, and under ideal conditions it was adequate. However, failures often occurred under conditions that were less than ideal, with the result that application rates gradually were increased to whatever levels experience dictated. In general, there has been a tendency to calculate the amount needed, and to add a certain excess to provide a margin of safety. Depending upon the amount of water to be treated and other conditions involved, the excess might amount to as much as two or three times the “normal” dose.

Instability of product and inconsistency of results were early problems with rotenone. Moorman and Ruhr (1951) pointed out that deterioration in strength of stored rotenone could contribute to failure of reclamations. Almquist (1959) noted that the toxicity of rotenone is decreased by exposure to light, heat, oxygen, alkalinity, and turbidity. Pintler and Johnson (1958) found that the rotenone content of cubé powder ranged from 2 to 5 percent. Manufactured formulations that have guaranteed content of rotenone, however, became available in the 1950's.

Rotenone formulations continued to evolve, and Shannon (1969) tested nine, commercially available formulations for toxicity and detoxification. They included one wettable powder and eight emulsions. Some of the latter formulations contain 5 percent or more of rotenone; others contain 2.5 percent of rotenone plus synergists; and some are homogenized for enhanced performance in special situations. These toxicants are effective against fresh-water and marine fishes. The liquid formulations, however, are malodorous because of solvents or carriers, and they obviously repel fish. Great care must be taken, therefore, to deny target fishes any avenue of escape during reclamations.

The commercially available formulations of rotenone can be detoxified largely or completely with potassium permanganate or chlorine. Moreover, experimental work has demonstrated that safe and palatable drinking water can be produced by water-treatment facilities that have adequate activated-charcoal capacity (Cohen et al., 1961 and Bonn and Holbert, 1961).

A review of techniques and equipment for reclaiming waters with rotenone was made by Hooper (1955). He cited the difficulties of getting good distribution of the toxicant into water more than 4.7 to 6.3 m (15 to 20 ft) in depth unless the chemical is pumped down through a weighted hose or advantage is taken of the autumnal overturn of lakes to disperse the chemical throughout all depths. Bassett (1956) did an economic evaluation of several formulations of rotenone, and demonstrated that some disperse more readily than others into depths. Turner (1959) analyzed the results of treating 56 farm ponds in Kentucky with four formulations of rotenone. Liquid preparations were sprayed over the surface of the ponds by powered pumps, and 1 mg/l (ppm) of toxicant was sufficient for kills of fish. He observed that effective penetration of the formulations ranged from 1.5 to 2.4 m (5 to 8 ft) in depth during summer and autumn treatments. Complete kills occurred in 97 percent of the 24 ponds treated in summer when water over 2.4 m (8 ft) in depth was stagnant and deficient in dissolved oxygen. In the autumn, however, complete kills occurred in only 72 percent of 32 ponds, and Turner (1959) assumed that fish fled from the rotenone into deep water that was now homothermous and well oxygenated. He concluded that the treatment of Kentucky farm ponds over 2.1 m (7 ft) in depth should be restricted to the summer months, June through mid-September.

In reviewing the use of fish toxicants in the Province of Quebec, Canada, Prévost (1960) noted that the rehabilitation of lakes through poisoning fish is the best tool at the fish manager's disposal. He listed rotenone as the best of the toxicants available, but advised managers to consider the thermocline, weed beds, floating islands, springs, water temperature, turbidity, and alkalinity during the treatment of lakes, and to consider beaver dams, marshes, isolated pools, and upstream and downstream sources of re-infestation during the treatment of streams. He cited some lakes as being free of rough fish 10 years after reclamation and producing spectacular trout fishing. Stroud and Martin (1968) summarized the treatments of lakes and streams in the United States and Canada, observing that rotenone is the toxicant most used in the United States. Kinney (1968) advised on the use of rotenone to obtain complete, partial, or selective kills of target fish. Howland (1969) showed that rotenone and antimycin are compatible in situations where the two might be used simultaneously in the reclamation of a lake-stream system.

The mode of action of rotenone in fish has been studied by a number of investigators. Hamilton (1941) reported that rotenone is a respiratory poison in fish and acts by vaso=constriction of the gill capillaries. Oberg (1967) found that rotenone is a powerful inhibitor of the respiratory chain in fish with the site of action located in the flavoprotein region of the chain. The specialized structure of gills favors entrance of rotenone into the blood for transport to vital organs for inhibition of respiration. The action is reversible, and Bouck and Ball (1965) demonstrated that methylene blue can be used to revive warm-water fish that have been poisoned by rotenone.

The effects of rotenone on aquatic invertebrates were reviewed by Taube, Fukano, and Hooper (1954), Almquist (1959), Wollitz (1962), and Binns (1967). Mallards and pheasants have oral LD50's in excess of 1,000 mg/kg (Tucker and Crabtree, 1970), whereas the oral LD50's for certain mammals are: 60 mg/kg for guinea pigs, but 1.5 g/kg for rabbits and 3 g/kg for dogs (Cohen et al., 1960). Tilemans and Dormal (1952) reported that the oral LD50 of rotenone for man is 2,850 mg/kg.

Saponins are water-soluble glycosides that occur in 75 or more families of plants. They are foaming agents with a history of uses in washing silk, wool, and cotton fabrics; in preparing sparkling wines and effervescent water; and as components in expectorant medicines. From 300 to 400 species of saponin-bearing plants, including such forms as azalea, camellia, rhododendron, and heath, have been known since time immemorial as “fishing plants” in Asia for collection of fish in ponds, rivers, and marine estuaries (Tang, 1961 and Noller, 1965). Tea-seed cake is a common form of the fish toxicant, and it is the saponin-bearing residue remaining after the oil is expressed from the seeds of camellia. The cake contains 10 to 13 percent of saponins.

Soong and Merican (1958) cited uses of tea-seed cake to eradicate fish from nursery ponds in Malaya and China. Tang (1961) stated that it is customary for Chinese fish farmers to employ tea-seed cake for control of undesirable fish in ponds before stocking. He also conducted successful experiments in Taiwan in controlling predaceous fish in shrimp ponds with powdered saponins and crumbled tea-seed cake. Ryther (1968) indicated that tea-seed cake is used routinely for this purpose in the Singapore area.

In the United States, Lunz and Bearden (1963) experimented with low concentrations of saponins in a small shrimp pond in South Carolina, and most species of fish present were killed. They concluded, however, that tea-seed cake is so expensive here that it cannot be used economically as a fish toxicant.

On the other hand, a search in Russia for ichthyocides that are highly toxic to fish, harmless to warm-blooded animals and man, and rapidly degradeable has focused attention on saponins (Bizyaev, Antimov, and Moskalev, 1965). They confirmed the toxicity of glycosides in azalea flowers to fish and other aquatic life, but concluded that the chief source of saponins in the Soviet Union appears to be the sugar beet. The saponins occur mostly in the surface layer, rootlets, and tail of the beet. Concentrates of saponins are obtained by centrifuging the foam on beet-pressing water, resulting from the pressing of beets into briquettes. The liquid concentrate is toxic to fish at 0.2 ml/1 (ppm) whereas the dry saponins are toxic at 2 mg/l (ppm). The toxic action on fish occurs within 20 to 24 hours, and degradation of the saponins is complete within 7 to 10 days. The authors attest that the saponins from sugar beets are the most effective and acceptable means for ridding inland waters of nuisance fishes.

The use of sodium cyanide as a fish toxicant was introduced by Bridges (1958) as a result of his experiments in laboratory aquaria and farm ponds in Illinois. He reported that 1 mg/l (ppm) of sodium cyanide, easily applied as 28-gram Cyaneggs, kills various species of warm-water fishes in a variety of temperature and pH conditions at a cost of 0.55 US dollars/1,000m³. Within minutes, fish begin coming to the surface, and desirable species can be removed to fresh water for complete recovery. Water in small farm ponds remained toxic for about 4 days.

Lewis and Tarrant (1960) continued experiments in Illinois to demonstrate the effectiveness of sodium cyanide as a collecting tool and general toxicant. They recommended the compound for preparing rearing and brood ponds. In relation to these experiments, Leland (1964) studied the loss of cyanide from the water, soil, and fishes. He determined that residues of cyanide are no problem in bottom muds; the toxicant disappears from water within 4 to 20 days, with dissipation slower in cold waters or under ice; and cyanide residues in live fish drop to less than 0.1 mg/kg after 24 hours in fresh water.

Multiple uses of sodium cyanide were exhibited in Nebraska by Miller and Madsen (1964) including the live removal of northern pike from nursery ponds, the salvage of fish from irrigation canals, the sampling of fish populations in lakes and streams, and the eradication of undesirable fish from lakes. The toxicant was used with variable success in South Dakota for simultaneous removal of live walleyes from rearing ponds and reclamation of the ponds for new stocking in the following year (Hanten, 1966). Whitley (1967) reviewed the effects of sodium cyanide on fish in Missouri, noting that the compound is toxic at all temperatures, but more rapidly so in warm water. At present, the major use of cyanide is centered in the lower Mississippi River Valley where fish farmers apply many thousands of kilograms in fish cultural ponds to eliminate competing fishes and predaceous invertebrates (Prewitt, 1970). Fish-eating snakes and birds are killed sometimes by eating fish treated with high concentrations (10 mg/l [ppm] or more) of cyanide.

Pellets of sodium hydroxide have been dropped into the nests of problem sunfishes to kill eggs and fry (Jackson, 1956). The control is limited, however, to waters where nests can be located and treated easily with reasonable expenditure of time and effort.

The LC50 of sodium pentachlorophenate to fingerling channel catfish is 0.46 mg/l (ppm) (Clemens and Sneed, 1959). Walker (1969) stated that concentrations as low as 0.06 mg/l (ppm) are lethal to fish under laboratory conditions and that piscicidal activity varies with temperature, pH, and other factors. A private fish farmer informed us that sodium pentachlorophenate effectively removed fish from his ponds, but did not kill tadpoles or snails. He discontinued use of the compound upon discovery that its residues were detrimental to fish during early developmental stages, causing excessive mortalities and teratogenesis, especially in goldfish.

Westman and Hunter (1956) made an experimental application of sodium sulfite at 168 mg/l (ppm) to a small pond in New Jersey to salvage certain fishes and reduce numbers of others. They concluded that salvage operations would be practical in small areas, but that the compound is too expensive for large waters. The sodium sulfite lowers the concentration of dissolved oxygen in the water very quickly, and fish suffocate. The water is non-toxic, and the dissolved oxygen is restored rapidly. Affected fish are salvaged by removing them to fresh water. Species with inferior mouths have difficulty gulping air at the surface, causing them to be more susceptible to suffocation.

Grice (1961) reported that 100 mg/l (ppm) of sodium sulfite followed by 50 mg/l (ppm) about 22 hours later were not effective in incapacitating fingerling walleye for removal from a pond in Massachusetts. More recently, Vanderhorst and Lewis (1969) used cobalt chloride to catalyze sodium sulfite, and concluded that the combination has promise for selective removal of fish, particularly channel catfish.

A search at the University of Idaho for a selective toxicant for squawfishes, problem predators on salmonids, resulted in the patenting of Squoxin in 1968 (MacPhee and Ruelle, 1968). In a subsequent report, MacPhee and Ruelle (1969) demonstrated a 10- to 17-fold margin between the minimum LC100 for two species of squawfish and the maximum LCO for the least tolerant salmonids. Marking (1969a) listed 96-hour LC50's of 0.182 to 0.779 mg/l (ppm) for nine species of fish at 12° C.

A summary in the Commercial Fisheries Review (Anon., 1970) classes Squoxin as an exceptional selective toxicant for squawfishes. The principal advantages of Squoxin are: squawfish-killing concentrations have no effects on aquatic invertebrates, mammals or humans; the toxicant is short-lived in water; and it is non-repellent to squawfish.

An intensive research effort by the U.S. Fish and Wildlife Service resulted in the development of TFM as a selective toxicant for larvae of sea lamprey (Applegate et al., 1961). The compound proved to be a practical and safe control for sea lamprey in the upper Great Lakes. Its toxicity is influenced greatly by water hardness and pH, and it is most effective on larval lampreys in soft, acid waters in late fall, winter, and early spring.

Because TFM is relatively expensive, research continues on more economical means of controlling sea lamprey. Howell et al. (1964) discovered that a 2-percent addition of Bayluscide synergizes TFM, retains selectivity to lampreys, and reduces treatment costs by approximately 50 percent. The liquid mixture of TFM and Bayluscide is metered precisely into streams to kill lamprey larvae residing in bottom muds, and there is little effect on other aquatic organisms and game fishes. The status of control efforts in the upper Great Lakes area and the improvements in sport and commercial fishing were reviewed by Baldwin (1968).

Thanite was tested at 0.7 to 1.5 mg/l (ppm) in ponds in Illinois to determine its potential for live removal of fish and for selective and total kills of fish (Lewis, 1968). The compound at first has an anesthetizing effect on fish, allowing desirable species to be collected easily at the surface during this stage. Their recovery in fresh water is rapid and complete. The live removal of adult largemouth bass from the ponds was highly successful. Lewis (1968) also observed that Thanite can be applied as a selective toxicant against centrarchids in the presence of cyprinids and ictalurids. Working in conjunction with Lewis, Leland (1964) demonstrated that the blood of fish killed by Thanite contains cyanide, but that there is a rapid loss of cyanide from the blood of exposed fish held in fresh water.

Additional studies on Thanite as a fishery tool are in progress at the Fish Control Laboratories, U.S. Bureau of Sport Fisheries and Wildlife, La Crosse, Wisconsin and Warm Springs, Georgia.

The toxicology of Thiodan in fish and aquatic invertebrates was studied by Schoettger (1970) to determine potentials of the compound as a fish toxicant. He also reviewed the literature on the nature of Thiodan and on its performance as a fish toxicant in limited tests dating from the early 1960's. He found that fish were at least seven times more susceptible than invertebrates to Thiodan; exposures of 2 hours to 50 mg/l (ppm) of Thiodan were not toxic to fertilized eggs of rainbow trout; Thiodan has little value as a selective toxicant against carp and suckers; the compound may be a good general fish toxicant under certain conditions; and residues of the toxicant occur in the skin and muscles of fish exposed to acute and multiple subacute concentrations.

The future of Thiodan as a fishery tool remains in doubt. It is a chlorinated hydrocarbon pesticide; and it probably would compete poorly in the registration process with non-persistent, more selective fish controls.

Tobacco wastes are added to milkfish ponds as fertilizer in Southeast Asia and have the advantage that the nicotine kills aquatic insects (Bardach, 1968). Personal communications disclose that tobacco wastes are used at about one ton per acre in ponds in Taiwan. The combination of nicotine from the tobacco and oxygen-depletion resulting from the decomposition of the plant acts to poison and suffocate unwanted fish, fish parasites, and possibly bacteria. Tobacco dust at 12 to 15 kg/ha of nicotine eliminates fish, snails, and polychaete worms. The tobacco dust also serves as a direct fertilizer and accelerates the fertilizing action of rice straw, rice bran, and other plant materials.

In India, Konar (1970) suggested that nicotine may be very useful as a fish-collecting aid and toxicant. Rohu exposed to 3.2 mg/l (ppm) of nicotine and punti exposed to 5.0 mg/l (ppm) surfaced within 5 to 10 minutes and recovered within 2 to 4 minutes in fresh water. Some fish remaining in the solutions of nicotine exhibited signs of acute poisoning and perished, but others showed no symptoms of poisoning. The low concentrations of nicotine tested were less toxic to aquatic insects than to the fish.

Toxaphene consists of a mixture of polychloro bicyclic terpenes with a predominance of chlorinated camphene. At highest purity, it contains 67 to 69 percent of chlorine. It has been used widely against a variety of insect pests on agricultural crops (Gebhards, 1960). Surber (1948) was perhaps the first to test toxaphene against fish, and he observed that 0.04 mg/l (ppm) of toxaphene killed all fish in a small pond. Noting that toxaphene was much more toxic than rotenone to fish, Tarzwell (1950) suggested that the compound may be useful in fish management. The first major field trials of toxaphene as a fish toxicant were conducted by Hemphill (1954) in two Arizona lakes in 1951. A concentration of 0.1 mg/l (ppm) eliminated the rough fish, including carp, in one lake and greatly reduced their numbers in another. He added that the killing action of toxaphene was slow in comparison with rotenone and extended over a period of days. The insect life in the lakes was severely affected, but not eliminated.

Tanner and Hayes (1955), evaluating toxaphene as a fish toxicant in Colorado, indicated that a lake may be treated effectively with the compound for about $0.10/1,000 m³ as compared with $0.77/1,000 m³ with rotenone. Admitting that toxaphene is attractive from the standpoint of economy, they advised that it is an extremely powerful poison of greater toxicity to warm-blooded animals than rotenone and requires greater precautions in handling. They concluded that toxaphene may persist for at least 7 months at toxic level in a lake at pH 8.0 or higher.

In Michigan, Hooper and Grzenda (1957) demonstrated that toxaphene is more toxic to fish in hard water than in soft water, and more toxic in warm water than in cold water. Although toxaphene at 0.1 mg/l (ppm) gave good results against fish, the lakes remained toxic to fish for periods of 2 to 10 months. Bottom invertebrates are killed in large numbers, but they quickly reappear in abundance.

The observation that 5 μg/l (ppb) of toxaphene in hard water killed small fish, but left large bluegill and largemouth bass unharmed, prompted Fukano and Hooper (1958) to suggest that the compound has potential as a selective poison. Stringer and McMynn (1958) applied the compound at 0.01 to 0.10 mg/l (ppm) in eight alkaline lakes in British Columbia, and eliminated all fish and amphipods. They noted that toxaphene is an effective and economical fish toxicant, but the lakes were still toxic to fish 9 months after treatment. In a follow-up study, Stringer and McMynn (1960) discussed methods for dispensing toxaphene, the killing time for fish, the lower lethal concentrations for a number of fish species, and factors influencing degradation. They pointed out that small concentrations of toxaphene applied to control cyprinids and cottids in deep, clear, stratified lakes in British Columbia may persist at toxic level for 2 years. On the other hand, detoxification proceeds so rapidly in some turbid lakes that relatively high concentrations produced only partial kills of fish.

In Iowa, tests of toxaphene against fish in the laboratory and field were encouraging (Rose, 1958). Carp and bullheads required over 25 μg/l (ppb) for kills in cold, clear water, whereas 200 μg/l (ppb) were needed against the same species in highly turbid water. Silt was suspected of having a direct detoxifying effect.

The results of 4 years of reclamation efforts with toxaphene in Nebraska lakes were reviewed by McCarraher and Dean (1959). They found that at least 0.5 mg/l (ppm) of toxaphene was required for complete kills of fish in Sand Hill lakes having moderate alkalinity, high turbidity, and pH 8.5 to 9.5. They recorded serious problems, however, that arose during aerial applications of the toxicant. An aerial application of 0.61 mg/l (ppm) of toxaphene in one lake killed every wild duck, but carp and bullheads survived. A similar application of 0.52 mg/l (ppm) in another lake killed all fish, but also killed 33 percent of the mallards and 29 percent of the gadwalls, but less than 10 percent of the gulls and grebes present in the treated area. Each of the aerial applications of toxaphene was accompanied by losses of waterfowl ranging from 15 to 100 percent. Dead mammals possibly associated with the operations included raccoon, dog, skunk, and cow. In contrast, there were few mortalities of birds when toxaphene was sprayed on the water from a boat.

Gebhards (1960) documented the increasing use of toxaphene in states and provinces of western North America. He also discussed the toxicity of toxaphene to humans, livestock, waterfowl, fish, and aquatic invertebrates, and stated that the factors increasing the rate of detoxification of toxaphene are sunlight, high concentration of dissolved oxygen, high temperature, water circulation, and turbulence. Kallman, Cope, and Navarre (1962) demonstrated that aquatic vegetation in a treated lake accumulated high concentrations of toxaphene and that rainbow trout and black bullhead concentrated the toxicant within their bodies. Hunt and Keith (1963) discussed the biological magnification of toxaphene residues that results in death of birds. Following the treatment of Big Bear Lake in California, Johnson (1966) recommended that toxaphene not be used as a fish poison anywhere in the state. Terriere et al. (1966) observed the persistence of toxaphene in Oregon lakes up to 6 years, with residues accumulating up to 14 mg/l (ppm) in rainbow trout and 17 mg/l (ppm) in aquatic plants. Other reviews on the performance and persistence of toxaphene were made by Nehring (1964), Johnson, Lee, and Spyridakis (1966), Henegar (1966), and Moyle (1968).

A survey in 1966 indicated that toxaphene ranked second to rotenone as a fish toxicant in the United States, but ranked first in Canada (Stroud and Martin, 1968). The limited use of the toxicant against fish in Germany was described by Anwand (1968b). Applications of the compound as a fish toxicant declined rapidly in the United States in the late 1960's, however, due in part to a ban imposed by the U.S. Department of the Interior in 1963 (Dykstra and Lennon, 1966). This ban was prompted by the persistence of toxaphene in water, its high toxicity to invertebrates and vertebrates, especially waterfowl, and the accumulation of residues in plants and animals. Further use of toxaphene as a fish toxicant in Federal projects or federally aided projects was forbidden. Walker (1969) observed that toxaphene has been one of the most extensively misused fish toxicants in the United States and Canada.