The word stream as used in this report includes such lotic environments as brooks, creeks, and rivers. Typically, a stream is a much more complex body of water than a lake. In physical characteristics, a stream may be long or short; torrential or slow; steep or flat; slow-growing or fast-growing in size; rocky or soft-bottomed; clear or turbid; stable or widely fluctuating in flow volume; varying widely in temperature both seasonally and from point to point along the course; free-flowing or cluttered with obstacles or debris; sparsely or heavily pooled; relatively unaltered or dammed, channeled, or diverted; and clean or polluted with industrial, agricultural, and municipal wastes. Biologically, a stream may be fertile or infertile, poor or rich in vertebrate and invertebrate fauna, and weedless or choked with vegetation. And politically, a stream usually passes through a variety of private, commercial, and public jurisdictions including intercity, interstate, and international. It also may serve many recreational, agricultural, industrial, and municipal purposes. Often a single stream possesses several or more of the physical, biological, and political attributes mentioned.
Social and legal problems. Laws pertaining to streams and rivers date back to the early civilizations of Assyria and Babylonia (Bardach, 1964). The ancient water laws of Rome shaped the later water laws of Europe and the United States, and formed the basis of riparian doctrine. It would be erroneous to infer, however, that water laws have some considerable consistency from nation to nation or region to region. The laws today vary greatly between areas with different climatic and hydrological conditions that affect the supply of surface water. In general, the water laws are less restrictive in regions where rainfall and surface waters are abundant than in regions where water is scarce. In the latter regions, the fisheries in streams may hold lower priorities than agricultural or industrial uses.
A fishery manager may be confronted with problems stemming from the classification of a stream as public or private, navigable or non-navigable, wild or multipurpose, and managed or unmanaged. Political boundaries along a stream seldom coincide with ecological boundaries, and multiple ownerships and jurisdictions may thwart ecologically based management plans. Achieving a unanimous or majority approval of a broad management plan for a stream system is difficult. And, obtaining approval for the reclamation of a stream with toxic chemicals is even more difficult. Pintler and Johnson (1958) stressed the need for extensive and sound public relations in executing the reclamation of streams.
Physical and biological problems. Huet (1959) pointed out that little study had been given to streams in contrast with the much larger amount of research done on ponds, lakes, and marshes. Moreover, he demonstrated the complexity of a stream by defining its biological zones, each with its own distinctive fish fauna. He described the fish zones for streams in Western Europe, between source and mouth, as: 1) trout; 2) Arctic grayling; 3) barbel; and 4) bream. Salmonids predominate in zones 1 and 2, and cyprinids predominate in zones 3 and 4. In the United States, Lennon and Parker (1960) plotted the vertical distribution of 15 fishes in the upper 37 km (23 mi) of a mountain stream lying between 335 to 1,615 m (1,100 to 5,300 ft). The brook trout was the sole species present at elevations above 1,372 m (4,500 ft); rainbow trout and brook trout predominated over longnose dace and banded sculpin between 914 and 1,372 m (3,000 and 4,500 ft) elevation and nine of the 15 species occurred only at elevations under 762 m (2,500 ft), with cyprinids predominating. A problem or pest fish, therefore, may inhabit only a certain section of a stream, encompassing one or two of the biological zones described by Huet (1959), and it is necessary to determine the upstream and downstream limits of a pest's biological zone before initiating control measures.
Barrier falls and natural dams may restrict or enlarge the distribution of a problem fish. Where such obvious factors are absent, other factors that affect the distribution of problem fish have to be found. In their study of the stoneroller, a fish that competes harmfully with rainbow trout for spawning grounds in streams of Great Smoky Mountains National Park, Lennon and Parker (1960) learned that an average stream gradient of 4.4 percent (43.5 m/km or 230 ft/mi) is the upper limit for the species. Thus, the stoneroller is absent in torrential headwaters with an average gradient of 8.8 percent.
Man-made problems. The distribution and abundance of desirable and undesirable fish populations in streams also may be affected directly or indirectly by man's activities. Denuding watersheds by logging may raise water temperature and degrade water quality. Agricultural practices may overly enrich streams with fertilizers and manures or poison them with pesticides. Water courses may be channeled, diverted, or impounded. Erosion may change the character of stream bottoms and the composition of aquatic life. In addition, selective exploitation of game fishes by anglers may alter the relationships between predator and forage species.
The abundant and widespread stocking of transplant and exotic game, food, and forage fishes in streams has altered greatly the composition of populations. The introduction of Pacific and Atlantic salmons and stream trouts into new waters often is beneficial, but some transplants are mistakes because of harmful influences on native fishes. The carp is an undesirable exotic in many streams in the United States and the target of various control measures. Even the goldfish exists in pest proportions in some streams. Cyprinids, suckers, sunfishes, percids, and cottids often are subjected to control efforts. Whereas conservation agencies now attempt to curtail unwise or illegal stocking of fish, great damage already has been done. The fish populations in most of the streams and rivers in the United States have been altered drastically by introduction of transplants and exotics.
Assessment of fish problems in streams. Prior to mid-twentieth century, the management of sport fisheries in streams consisted largely of:
1) Regulating fishing seasons by law, with closed seasons coinciding with the spawning seasons of preferred game fishes.
2) Regulating the rate of game fish harvest by law, with minimum legal lengths and daily bag limits fixed.
3) Stocking game fishes. Stocking of trout at the turn of the century included mostly fry and fingerling fish, but legal-length or adult fish were more often stocked after the mid-1930's. The stocking of species other than trout in streams has been greatly reduced in recent decades.
3) Stream improvements, consisting of the removal of obsolete mill or beaver dams to facilitate migrations; the installation of wing dams, current deflectors, riprap banks, and streamside vegetation to increase cover and food; and the construction of spawning beds or spawning channels. Such improvements were made on relatively few streams -- salmonid streams for the most part -- because of high costs.
4) Fisherman census to determine fishing intensity on selected waters and the rate of catch of game fish. Again, relatively few streams were censused because of the expense.
Toxicants offer the only practical means for controlling fish in streams. Nets, traps, electrofishing, and explosives have been employed to reduce fish populations locally, but they are too laborious, expensive, and inefficient for large-scale reclamation. A toxicant in appropriate solution or suspension, properly applied to obtain thorough dispersion in a moving bolt of water, can reach all fish residing in cascades, riffles, pools, at the sides, on the bottom, and in upper layers of water in streams. There may even be a desirable percolation of toxicant-laden water through some types of bottom substrates, as observed by Meehan and Sheridan (1966) when they applied toxaphene to an Alaskan salmon stream to control the coastrange sculpin. They concluded that the toxicant may have penetrated as far as 18 cm (7 in) into the gravel stream bed.
In the United States, the events leading to the reclamation of a stream are three=fold. First, the fishing public or a sportsmen's organization complain to authorities that the quality of fishing in a given stream has declined. Second, the conservation agency studies the stream and its sport fishery potentials and recommends corrective action which may include the removal of existing fish and replacement with desired species. Third, the fishermen involved and the conservation agency cooperate to obtain approvals from landowners, to prepare the stream for the project, to purchase the fish toxicant, to apply the toxicant, and to pick up and process the dead fish. Satisfying the legal and social requirements may take months of effort, and indeed, some requirements may prove to be insurmountable.
The reclamation of lakes with toxicants had become a common practice in fishery management before serious attention was given to reclaiming streams. This was due in part to the fact that lakes were easier to survey and treat with the relatively imperfect materials and methods of the 1930's and 1940's, and there were numerous lakes deserving attention of the limited numbers of fishery biologists. Second, imbalances in fish populations were more difficult to assess in streams than in lakes until electrofishing gear came into wide use in the 1950's. Prior to that time, the use of nets or cresol in streams was restricted mostly to sampling fish for taxonomic purposes rather than for estimates of populations (Embody, 1940). Third, until 1956, there was no known way of detoxifying a fish eradicant in water (Lawrence, 1956 and Jackson, 1957). A toxicant in a lake is relatively confined until degraded by natural processes, but a toxicant in a stream may travel many kilometers, traverse many jurisdictions, and confront many users of the water before deactivation occurs. Thus, control of a toxicant in a stream by detoxification, impoundment of treated water, or by dilution was a limiting consideration. Fourth, the early fish toxicants tended to repel fish, driving them ahead of toxic water into springs or tributaries. Lastly, the relationships of concentration of toxicant to duration of exposure necessary for an effective, uninterrupted, and sustained dose in flowing water, maintaining the dose over distances of kilometers in the face of increasing, diluting volume from tributaries were not well understood until the 1950's.
An early motivation for the removal of fish from streams was supplied by Davis (1934) who advised from the point of view of disease control that no fish of any kind be allowed in a spring from which a hatchery obtains its water supply. Recognizing the potential for disease, the Department of Fisheries of Canada arranged in 1936 to eradicate fish from a 40-km (25-mi) stream system that supplied water to a new hatchery in Nova Scotia (M'Gonigle and Smith, 1938). Powdered derris root, containing 5 percent of rotenone, was applied to the mainstream and tributaries by established sections that were blocked upstream and downstream by nets, and to a 2.4-ha (6-a) lake. The treatment was accomplished in August when flow was low and water temperature high, and some sections were re-treated to ensure the kill of all fish. Fish in live-cages demonstrated that the lake remained toxic for 1 month. The fish removed from the system included brook trout, parr Atlantic salmon, and a few cyprinids. The eradication was believed to be complete.
Because of an outbreak of furunculosis (Aeromonas salmonicida) at a Federal trout hatchery in New Hampshire, Hoover and Morrill (1938) applied sacks of powdered rotenone in two small streams that serve as the hatchery water supply. The objective was to eliminate brook trout and slimy sculpin as part of the disease control program. Five and one-half km (3.5 mi) of stream were treated in sections blocked off with nets, and the eradication was considered successful.
In another disease control effort, Hagen (1940) used liquid chlorine and powdered rotenone to remove brook trout and dace from a spring-fed pond and stream that provided water to a Federal trout hatchery in Wyoming. The chlorine at a concentration of 3 mg/l (ppm) was not effective for a total kill because it repelled fish and failed to penetrate spring flows on the bottom of the pond. A subsequent application of rotenone at 2 mg/l (ppm), sprayed as a slurry, killed all trout and most of the dace in the pond and short stream.
In 1946, Siegler and Pillsbury (1946) stated that the degree of success in lake reclamation depends largely on how thoroughly inlet waters are poisoned. They sprayed each inlet water freely with toxicant, but urged further research be done on the use of rotenone in streams.
In an evaluation of reclamation work in Canada, Smith (1950) stated that calcium hypochlorite, copper sulfate, rotenone, and a mixture of rotenone and toxaphene were being used as fish toxicants, but that rotenone was the most nearly adequate. The objective of most projects was to improve trout fishing, and target species often included carp and other cyprinids, suckers, and yellow perch. He noted that complete kills of fish are impossible in spring-fed streams unless sufficient toxicant is maintained in spring holes and seeps for an adequate period of time.
Miller (1950) discussed the need and use of fish toxicants and pointed out that the literature on reclamations lacks data on the completeness of kills and on the rapidity of re-infestation by rough fish. He added that complete kills are very rare, and re-infestation is usually rapid unless barriers are installed.
Although the pace of pond and lake reclamation was accelerated in the late 1940's and 1950's, with attendant treatment of inlet streams, there was little done on methods and materials for reclamation of streams until the late 1950's.
The theory and methods of rough fish control were reviewed by Meyer (1963). Whereas the elimination or reduction in rough fish is an important tool in improving fishing, he discussed criteria for determining the need for control and suggests a pre-treatment procedure. Rotenone is the most widely used, but other fish toxicants including toxaphene, endrin, cresol, Aqualin, and copper sulfate are employed. He recognized that a complete kill of rough fish almost never is possible, but urged caution to avoid risking the extermination of a native species.
Meyer also cited the need for evaluation of stream reclamations. Among the 17 streams in California that had been treated to suppress rough fish, the recovery of undesirable species has been rapid and benefits have been questionable and temporary.
A summary of the many lake and stream reclamation projects in Michigan between 1947 and 1963 was compiled by Scott (1961). Emulsifiable rotenone was the toxicant used most often against rough fish inhabiting gamefish waters.
The U.S. Bureau of Sport Fisheries and Wildlife (1957) reported that in the first 5 years of its Federal Aid in Fish Restoration Act, 1952-1956, more than 1,600 km (1,000 mi) of streams were reclaimed. In its 10-year review of the program, the Bureau (1962) reported that biologists in 34 States had used chemicals to restore good fishing in about 4,000 km (2,500 mi) of streams and more than 90,000 ha (225,000 a) of lakes and ponds.
The Sport Fishing Institute reviewed fish conservation highlights for the years 1963–1967 and indicated that 9,320 km (5,825 mi) of streams in the United States and 80 km (50 mi) of streams in Canada, excluding sea lamprey streams, have been reclaimed with toxicants, principally rotenone and toxaphene (Stroud and Martin, 1968). The restoration of sport fisheries, frequently salmonids, in the treated waters is regarded as good fishery management.
The evolution of the principles and methods for reclamation of streams is traced by a review of selected projects accomplished within the past two decades.
1) Berry and Larkin (1954) undertook to compare the efficiency of poisoning versus operation of fish traps to control undesirable fish. They subjected Fish-Tox, a mixture of rotenone and toxaphene to 11 field trials in streams in British Columbia to determine effective concentrations and the factors influencing success. Sections of selected streams were blocked off with seines or metal screens and treated with 1 to 20 mg/l (ppm) of toxicant for periods of 14 to 90 minutes. The chemical was applied either as a paste in flow=through bags or as a “soup”. The largest stream flow was 127 m³/min (75 cfs), and water temperatures ranged from 48 to 65° C.
The target fishes in the field trials included carp, northern squawfish, other minnows, and prickly sculpin. One and 2 mg/l (ppm) of Fish-Tox failed to produce complete kills in exposures of only 15 to 20 minutes. Concentrations above 5 mg/l (ppm) were needed to kill all carp. Poison application stations had to be relatively close together -- at intervals of 0.4 km (0.25 mi) -- on slow moving streams, but they could be up to 1.6 km (1 mi) apart on fast streams. The slow streams were characterized by aquatic vegetation and silt bottoms that adsorb the toxicant and by an abundance of aquatic invertebrates that remove the poison from solution. A comparison of costs, however, favored poisoning over fish traps for the removal of coarse fish.
2) A study on the manipulation of fish populations in streams was initiated in Kentucky in 1952 because information on the dynamics of warm-water fishes was insufficient for management or control (Charles, 1958). In general the sport fishing streams were being altered by erosion, siltation, and pollution, and excessive numbers of rough fish including carp, suckers, bullheads, and freshwater drum were present. Following extensive surveys of fish populations, two streams were treated with powdered rotenone.
Seventy-four km (46 mi) of the North Fork River were reclaimed with powdered rotenone at 735 mg/m³ (2 lb/a-ft) or more over a period of 2 weeks in 1952. Three to 11 km (2 to 7 mi) of stream were treated on alternative days, and the toxicant was premixed into a slurry for distribution by hand or spray apparatus by workers on foot or in boats. Barriers of 2.5-cm (1-in) wire mesh were installed in the stream at the end of each day to prevent upstream migration of fish into treated zones. Every other day was spent in collecting and burying dead fish. Subsequent sampling indicated a nearly complete kill consisting of approximately 16,818 kg (37,000 lb) of fish.
A 19-km (12-mi) section of Whippoorwill Creek was treated in 1 day with 8.9 kg (8 lb) of powdered rotenone per surface hectare. High water temperatures and flushing rains contributed to rapid removal of the toxicant in North Fork River and Whippoorwill Creek, so that the streams could be restocked with game fish and forage fish within a few days.
Selected stations in North Fork River and Whippoorwill Creek were sampled each year for 5 years following the reclamations. Rough fish began entering experimental sections during the first winter and spring following treatment. Gizzard shad, carp, and white sucker, for example, spread quickly throughout the 74-m (46-mi) section of North Fork River. Other species re-entered slowly, and some never re-entered. The trend, however, was toward re-establishment of the original population composition. The investigator's conclusion was that population manipulation by toxicants, unless accompanied by environmental improvement, holds little promise as a management tool in warm-water streams in Kentucky.
3) Zilliox and Pfeiffer (1956) demonstrated the importance of barrier dams in preventing re-infestation of treated waters by rough fish. They succeeded in eliminating non-native yellow perch and restoring brook trout in a chain of cold, infertile ponds and streams in the Adirondack Mountains in New York. Powdered rotenone was applied to 14 ponds and 34 km (21.3 mi) of streams, and the pond waters remained toxic for 30 to 50 days. A major obstacle to good distribution of the toxicant was the presence of beaver dams. The beavers and dams were removed, and the impoundments, consisting of “floating-bog” type of environment, were drained. But, hidden underground channels made by beaver and muskrat and watered by springs and seeps may have permitted survival of some fish or fish spawn. Fish-proof barriers were installed in all connecting waters that could serve as sources of re-infestation by rough fish.
The investigators also made some very significant observations during post-treatment assessments of success. They learned that intensive gill netting during the same year or in the following year after a reclamation cannot be relied on to indicate a complete kill of fish. They concluded, however, that netting 2 years after reclamation is perhaps a valid measure because surviving species may be sufficiently abundant to be caught. For example, their netting during the first year after reclamation disclosed the presence of creek chub and common shiner, and netting during the second year caught white sucker and brown bullhead. However, no yellow perch, the principal target species, was taken in post-treatment netting, and barrier dams were effective in preventing re-invasion from untreated waters.
The restoration of brook trout was considered a success, due principally to the control of yellow perch.
4) In California, Pintler and Johnson (1958) accomplished the first large-scale reclamation of a watershed. They treated 33 tributaries and the mainstream in the Russian River drainage, totalling 458 km (286 mi), over a 3-year period, 1952 to 1954. Flow volumes ranged from a few to more than 1,000 m³/min. The object was to reduce the numbers of rough or competing species that made up almost 100 percent of the fish population and to restore rainbow trout (anadromous steelhead) in the watershed.
Powdered cubé root containing 2 to 5 percent of rotenone in mechanically mixed slurries was dripped or sprayed into the streams. “Mud balls” of cubé powder were thrown into some pools to penetrate deeper water. It was very difficult to calculate the quantities of toxicant needed because of the length of streams involved, the poor access to some waters, and the widely fluctuating flows in some streams. Thus, a decision was made to use quantities of toxicant many times the amounts estimated to be necessary to kill fish. A treatment once started on a stream was continued without lapses to prevent re-entry of rough fishes. Immediate assessments of kill were made on treated waters, and if there were any doubt regarding the potency of toxicant and kill of fish, the concentration was increased or the questionable area was re-treated. The reclamation was further complicated because some of the cubé powder had lost part of its strength.
Pintler and Johnson (1958) added observations on the effects of cubé powder on personnel. During the reclamation, some personnel were exposed to the toxicant more or less continuously for as much as 3 weeks. Headache, sore throat, and other cold symptoms were the primary complaints. Sores developed on mucous membranes, eyes were irritated severely, and there were eczema-like rashes on skin. Some personnel suffered loss of appetite and ability to taste. The authors concluded that increased safety and protection of the health of personnel should warrant the somewhat higher cost of rotenone in the new wettable powder or emulsifiable formulations.
The estimated total kill of fish was 86 metric tons (95 tons) including many Sacramento sucker, carp, Sacramento squawfish, green sunfish, and other rough species. Post=treatment sampling with seines, gill nets, electrofishing, and spot-poisoning with rotenone disclosed no living fish in treated areas. However, re-entry of fish was expected from several sources including untreated headwaters, water-power diversions, numerous farm ponds within the drainage, and accidental or deliberate stocking by anglers and private citizens.
The populations of anadromous rainbow trout (steelhead) in the streams increased about 13-fold during the year following reclamation. The treatment of the streams had cost approximately 12.5 US dollars/km ($20/mi).
5) Mullan (1956) compared the results of reclaiming a good, a marginal, and a poor trout stream with rotenone in Massachusetts in 1955. He concluded that the good and the marginal streams were able to support more trout following reclamation, but results were negative in the poor stream. Although complete kills of rough fish did not occur in any stream, no species regained its former abundance within 1 year after reclamation.
6) The contribution of funds to states under the Federal Aid in Fish Restoration Act in the mid-1950's stimulated some very ambitious reclamation programs on streams and rivers, particularly in conjunction with the establishment of new reservoirs. The objective of pre-impoundment reclamations is to eliminate or reduce populations of rough fish from the waters upstream from the dam site so that stocked game fishes may be free from competition or predation.
Thoreson (1956) described a pre-impoundment reclamation on the Marias River in Montana. The Tiber Dam was scheduled for completion and closure in 1955, so fishery surveys were made in 1954 and 1955 to determine the upstream limits of distribution for carp, goldeye, and other rough fishes. During a 2-month period in the summer of 1955, 36 metric tons of Fish-Tox, a mixture of rotenone and toxaphene, were applied to more than 960 km (600 mi) of river, streams, and sloughs upstream from Tiber Dam. Major streams were treated by sloshing bags of toxicant in the water; backwaters, sloughs, and large pools were treated by an airplane-mounted duster unit; and small streams were treated by spraying with backpack pumps or by dragging sacks of toxicant in them. It was a prodigious task to apply the toxicant and to maintain continuity in the treatment.
In his appraisal of the project, Thoreson (1956) noted that planning and surveys were inadequate for a reclamation of this magnitude. He recommended that personnel to be involved in the treatment phase also be involved in the planning and survey phase in order to become acquainted with the waters and terrain. He called for research on methods of applying toxicants, on rapid, field-type analytical methods for measuring the concentration of a toxicant in water, and on the pattern and dispersion of toxic material in flowing water. And, he stated that the aerial application was not as effective as desired because the dust formulation of Fish-Tox clogged in dispensers and acted slowly in the water.
A year later, Thoreson (1957) reported that some carp, minnows, and suckers had survived in relatively isolated areas, and they were subjected again to rotenone. The new Tiber Reservoir and upstream waters were stocked with 7.5 million rainbow trout with good promise of a successful sport fishery. The Montana Fish and Game Commission (1958) summarized the results of studies during 3 years following the reclamation by stating that the principal target species, carp and goldeye, had been eliminated from the drainage and that trout fishing was excellent. The Commission also attested the value of installing low dams as fish barriers on streams to inhibit re-invasion by rough fish.
7) The prospects of establishing a trout fishery in a new reservoir in the Deep South led to the reclamation of a portion of the Little Tennessee River (Wilkins, 1957). Plans were made to reduce the abundant gizzard shad, carp and other cyprinids, suckers and smallmouth buffalo, sunfishes, and freshwater drum in the 16 km (10 mi) of main river and 42 km (26 mi) of tributaries that were to be inundated in whole or in part by the new reservoir.
The biologists arranged to treat the main river with a liquid formulation of synergized rotenone at a time when the flow from Calderwood Dam upstream was shut down completely and just before the gates were closed at the new Chilhowee Dam downstream. The closure of gates in the new dam would prevent the toxicant from killing fish downstream, and the rapid filling of the reservoir would dilute the toxicant beyond all risks. The time available for dispersal of rotenone in the largely dewatered river bed was 14 hours. Because of the very limited time available for the treatment, sodium fluorescein dye was added to the toxicant to enable the applicators to trace the paths and completeness of distribution. A large helicopter was employed to dispense the liquid toxicant through a hose into the river, and a small airplane with spray apparatus was used to reach more remote waters. The 2,081 liters (550 gal) of toxicant applied resulted in a concentration of at least 1 mg/l (ppm).
The results were considered good, and the kill was estimated at 95 percent of the fish in the treated waters. Ninety-six percent of the fish checked by biologists were undesirable species. As the new reservoir filled, it was stocked with fingerling rainbow trout, and a growth rate of 2.5 cm (1 in) per month throughout the year was predicted.
8) Refinements in stream reclamation techniques were made by Lennon and Parker (1959) during their investigations on trout streams in Great Smoky Mountains National Park in Tennessee and North Carolina in 1957. They used malachite green dye with comparator standards and a salt-resistivity technique to measure the velocity, stretchout, and dilution of a toxicant as it moved downstream over distances of 7.2 to 23.4 km (4.5 to 14.6 mi).
The dye technique involved adding a known and strong concentration of malachite green at a constant rate to a stream at a selected site for 15 minutes. By means of a comparator, the concentrations of dye and the time required for the bolt of dye to reach and to pass a selected site downstream were measured. Concurrently, trials were made with blocks of cattle salt and a resistivity meter. For example, salt was added to a stream at the rate of 12.2 mg/l (ppm) for 15 minutes; the bolt of salt water reached a check point 5.6 km (3.5 mi) downstream in 2 hours and 34 minutes; but the bolt had become stretched out and diluted to a maximum of 2 mg/l (ppm) and required more than 2 hours to pass the check point.
Whereas the dye-comparator and the salt-resistivity techniques gave equally reliable data, the salt-resistivity technique won favor because it was easier to use over long distances, in turbid water, and at night (Lennon, 1959). On the basis of salt-resistivity readings along the course of a stream, stations were located for superimposing strengthening quantities to a bolt of toxicant progressing downstream.
The experiments with dye, salt, and rotenone in small streams demonstrated that it is difficult at best, but critically important, to maintain sufficient concentration of toxicant for sufficient time to kill target fish at any point in the stream. The data in literature on 24- or 96-hour LC50's of rotenone to various species of fish are of little help. Streamside bioassays were then conducted to define the minimum durations of exposure necessary for selected concentrations to toxicant to kill target fishes. In one soft-water stream, rainbow trout and longnose dace had to be exposed to 0.5 mg/l (ppm) of rotenone for at least 1 hour at 14.4° C (58.8° F) to ensure a complete kill. Bioassays on another stream, demonstrated that carp required 6 hours of exposure to 0.1 mg/l (ppm) of rotenone, or 4 hours of exposure to 3.2 mg/l (ppm) for complete kills at 17.8° C (64° F). Goldfish, however, survived as much as 6-hour exposures to 4 mg/l (ppm). Other species that exhibited resistance to relatively high concentrations of toxicant in long exposures were blue catfish and brown bullhead.
One experiment involved the reclamation of a small trout stream 7.2 km (4.5 mi) long with Pro-Noxfish, a synergized formulation containing 2.5 percent of rotenone and 2.5 percent of sulfoxide. Strengthening quantities of toxicant were superimposed on the bolt at three points downstream to maintain a dose of at least 1 mg/l (ppm) for 1 hour. When the bolt passed the final check point, it had stretched out to 2 hours and 55 minutes long.
The only fish present in the stream were rainbow trout, brook trout, and longnose dace, and the toxicant eliminated all yearling and older fish. But, because the experiment was made in early May, there was a possibility that fertilized eggs or sac fry of some fish might escape poisoning. This indeed happened because subsequent surveys with electrofishing gear turned up a few young-of-the-year rainbow trout and longnose dace. There is a risk, therefore, in reclaiming streams during the reproductive season of target species.
Another element of the experiment was detoxification of the rotenone with potassium permanganate, based on Lawrence's (1956) findings in ponds. Two-and-one-half mg/l (ppm) of permanganate were superimposed on the bolt of rotenone for 2 hours and 55 minutes as it passed the final check point. A series of cascades aided in mixing the chemicals, and no fish were killed below that point. The effectiveness of KMnO4 for rapid detoxification of rotenone in a stream was thereby demonstrated.
9) The salt-resistivity technique and streamside bioassays were put to a large test in the reclamation of Abrams Creek in Great Smoky Mountains National Park in June 1957 (Lennon and Parker, 1959). A 23.4-km (14.6-mi) section of the stream, extending downstream from a barrier falls to its confluence with the Little Tennessee River, was treated with Pro-Noxfish in conjunction with the State of Tennessee's reclamation of the Chilhowee Reservoir site (Wilkins, 1957). A requirement established early in the planning phase called for the bolts of toxicant in Abrams Creek and the Little Tennessee River to arrive at the confluence at 1200 hours on June 9 in order that rough fish would have no sanctuary from the poison in either stream.
Most of Abrams Creek and the tributaries were inaccessible except on foot over primitive, mountain trails. Thus, stations for introducing or fortifying the toxicant were few and far between, and appropriate quantities of rotenone had to be backpacked in and cached ahead of time. Special efforts were made, therefore, in preceding weeks to measure volumes of flow at selected check points and to measure the velocity and other characteristics of flow by means of the salt-resistivity technique. Known volumes of salt were administered to the mainstream and tributaries for various lengths of time, and residual concentrations of salt and time-lengths of bolts were measured with a resistivity meter at check points several kilometers downstream. The measurements were plotted against volumes of flow, and they indicated that a bolt of toxicant would require 20 or more hours to traverse the distance between the barrier falls and the confluence with the Little Tennessee River.
The volume of flow in Abrams Creek at a key check point during the night and early morning of June 8 was 156 m³/min (90 cfs). Plotted data indicated that rotenone would require exactly 24 hours to traverse the 23.4 km (14.6 mi). The introduction of toxicant was started at Abrams Falls at 1220 hours on June 8, about 20 minutes later than planned. The bolt, however, arrived at the confluence of Abrams Creek with the Little Tennessee River at 1200 hours on June 9, exactly on time.
Forty-six species of fish were present in Abrams Creek, including gizzard shad, carp, stoneroller, suckers and redhorse, yellow and brown bullhead, rock bass, and freshwater drum. The streamside bioassays and stream flow characteristics indicated that rotenone should be introduced at 5 mg/l (ppm) for the first hour and at 1 mg/l (ppm) for 5 hours thereafter. The water temperatures ranged from 18.9 to 21.7° C (66 to 71° F). The fortifying stations downstream were operated to ensure that any fish in the stream would be exposed to at least 1 mg/l (ppm) of toxicant for at least 6 hours.
The results of the reclamation were excellent. All fish apparently were eliminated. The streams were restocked with rainbow trout, and they spawned abundantly during the following spring. The good fishing during the summer of 1958 was reflected in a rate of catch that exceeded one fish per hour of angling effort. During the first year after reclamation, no fish other than trout were observed in the stream by biologists or anglers.
The reclamation of the Little Tennessee River at the Chilhowee Reservoir site was not complete, however, and during the fall of 1958 -- about 16 months after the reclamation -- 11 species of rough fish were observed moving out of the new reservoir into Abrams Creek. The error in not erecting a fishproof barrier at the mouth of Abrams was then appreciated.
10) Keating (1961) reported a novel attempt to control the abundant and predaceous northern squawfish in Cascade Reservoir in Idaho in 1958 and 1959. Because the reservoir does not lend itself to reclamation, Keating poisoned spawning-migrant adults and emerging fry in 28.8 km (18 mi) of the North Fork of the Payette River upstream from the reservoir. A formulation of liquid rotenone was dripped into the river at 1 mg/l (ppm) for 3 hours at two sites to kill the majority of adults at or near the peak of migration when water temperatures ranged from 18.9 to 23.3° C (66 to 74° F). Cognizant that some fish may have already spawned, Keating used 0.5 mg/l (ppm) of rotenone a little later to kill emerging squawfish fry. He estimated that 250,000 adult squawfish were killed in each of the 2 years, comprising at least 80 percent of the total number of spawners from Cascade Reservoir. Large numbers of fry, but relatively few trout, whitefish, or suckers, were killed in the late-May treatments.
Based on this success, Keating proposed that spawning squawfish and fry in the river be treated with rotenone for 4 or 5 consecutive years, or until all immature squawfish in the reservoir had matured and entered the river. Control of the squawfish was deemed worth the estimated cost of 1,100 US dollars per year.
11) An investigation was initiated in Texas during 1957 to develop a device that would meter a constant or near constant amount of a toxicant into a flowing stream for a selective kill of undesirable fish (Lowman, 1958). Basically, the device permits liquid toxicant to enter the stream by gravity flow from a tank in which a constant level of toxicant is maintained by means of a float-type control. The device and modifications of it were tested on two small streams in 1958 (Lowman, 1959). The objective was to meter rotenone into the water for many hours to achieve a selective kill of gizzard shad.
In the first stream, a concentration of 0.10 mg/l (ppm) of the Pro-Noxfish formulation of rotenone was maintained for 42 hours. Gizzard shad, suckers, bullheads, and sunfishes were killed, but some minnows, channel catfish, and largemouth bass survived. Concentrations of 0.08 to 0.12 mg/l (ppm) of the toxicant for 51 hours in the second stream killed both game and rough fish with no evidence of selectivity. But, there was no kill of any fish beyond 0.8 km (0.5 mi) from the point of introduction of rotenone. Lowman concluded that more development work is needed on metering devices and on achieving and maintaining selective-kill concentrations of a toxicant in flowing water. He also suggested that a study be made to determine if the Pro-Noxfish formulation has a general tendancy to settle to the bottom of a stream instead of dispersing through the flow.
12) Gaffney (1959) attempted a partial reclamation of that portion of the Clark Fork River in Montana that was to be inundated by a new reservoir. The objective was to reduce the numbers of rough fish that might limit the establishment of a rainbow trout fishery in the reservoir. Elimination of the rough fish was preferable, but the great size of the drainage area and streams made such an attempt impractical at the time.
Pre-impoundment surveys of the fish populations began in 1955, and 60.8 km (38 mi) of the Clark Fork River and immediate tributaries were treated with liquid rotenone and a mixture of powdered rotenone and toxaphene in the summer of 1958. On the basis of flow at 2,720 m³/min (1,600 cfs) and water temperatures ranging from 15.5 to 17.2° C (60 to 63° F), crews in boats and in a helicopter applied the toxicants at 0.4-km (0.25-mi) intervals to achieve a concentration of 1.25 mg/l (ppm) in the water. Six crews working simultaneously, but not progressively, downstream required 13 hours to complete the treatment. Because some pools in the main river were as much as 22.5 m (75 ft) deep, weighted sacks of rotenone-toxaphene powder were dynamited in the deepest pools in an attempt to disperse the toxicant.
A team of biologists and two skin divers assessed the kill of fish. The divers, however, encountered poor visibility, and the value of their observations was limited. Large numbers of mountain whitefish, northern squawfish, largescale sucker, redside shiner, and smaller numbers of rainbow trout, cutthroat trout, and peamouth were killed. The team also observed that fish moved downstream ahead of the toxicant.
In the management follow-up, the Montana Fish and Game Department (1962) reported that three million fingerling rainbow trout were stocked in the treated waters between 1958 and 1962. Surveys of the fish populations demonstrated that rainbow trout comprised 80 percent of the total population in 1959, and sport fishing was good. But, in 1961, rainbow trout comprised only 10 percent, and indications in 1962 were that the species was declining to about 5 percent of the total population. The Department added that the rapid increase of rough fish and the decline of rainbow trout within a few years casts doubt on the value of partial chemical treatment of large rivers being impounded for reservoirs.
13) The importance of temperature during reclamation was appreciated by Huntington and Jester (1958) when they treated a portion of Negrito Creek in New Mexico with liquid rotenone. Their objective was to reduce large populations of suckers to make the stream more suitable for rainbow trout. Their drip systems for dispensing liquid rotenone froze during a cold night, and the toxicant failed to give a complete kill at water temperatures as low as 1.1° C (34° F).
Huntington and Jester recommended that the stream be re-treated when sucker populations again were large, that re-treatment be done when water temperatures are warm, and that stations for dispensing rotenone be located at intervals less than 8 km (5 mi).
14) The North Fork of Fishing Creek in West Virginia was chosen in 1958 for a test of the effectiveness of reclaiming and restocking a warm-water stream (Robinson, 1961). Crews equipped with backpack sprayers treated 56 km (35 mi) of mainstream and tributaries above a barrier dam with a synergized formulation of liquid rotenone in September. The kill included such target species as creek chub, white sucker, and rock bass. The water was detoxified with potassium permanganate as it spilled over the barrier dam.
The stream system was restocked with smallmouth bass, 10 to 20 cm (4 to 8 in) long, 1 month later. Surveys with seines during the next 6 months found only stocked bass, but spot-treatments with rotenone about 1 year after the reclamation disclosed that minnows, white sucker, and centrarchids had survived. Two years after the reclamation, the numbers of rough fish showed a marked increase. Whereas the growth and reproduction of smallmouth were good at first, declines were evident after 2 years.
Robinson concluded that a complete kill of rough fish is impossible unless elaborate measures described by Lennon and Parker (1959) are taken. He added that a greater degree of success might have been achieved if larger numbers of smallmouth bass were stocked immediately following the reclamation.
Another test of stream reclamation in West Virginia involved two marginal trout streams that possess barriers against re-invasion (Robinson and Sumner, 1964). The small streams had been managed on a put-and-take basis with catchable size trout prior to reclamation with no detectable carry-over of the stocked fish. Following applications of rotenone to reduce the number of rough fish, brown trout were established in one stream and brook trout in the other. An annual stocking of fingerling trout in addition to natural reproduction helps support an intense sport fishery at a dollar cost significantly less than that of the pre-reclamation stocking. The growth and quality of trout are improved, and the authors concluded that reclamation of such West Virginia trout waters is practical and worthwhile.
15) The art of controlling undesirable fish in streams with chemicals advanced greatly with the development of materials and methods for selective poisoning of sea lamprey larvae in tributaries to the upper Great Lakes. Bardach (1964) called the lamprey-eradication project the largest aquatic control measure ever attempted. The invasion of the sea lamprey into the upper Great Lakes and the disastrous effects on native fishes were described in detail by Applegate and Moffett (1955), Howell (1966), and Baldwin (1968).
Limited attempts to capture and destroy spawning-migrant lampreys in streams with mechanical and electrical weirs were initiated in 1950 and continued until 1960. In the meantime, intensive research to find a selective toxicant for lamprey larvae was prosecuted by the U.S. Fish and Wildlife Service and the Great Lakes Fishery Commission. The breakthrough came in 1957 with the finding that 3-trifluormethyl-4-nitrophenol (TFM) kills larval lampreys at concentrations that are relatively harmless to game fish, wildlife, and livestock. Field trials of TFM in four lamprey-infested streams in 1958 confirmed that it possesses an acceptable degree of selectivity against larvae in the presence of bony fishes.
As lamprey control with TFM evolved, several notable improvements in efficiency and economy were made (Schneider, 1969). Three portable feeders for metering precise amounts of lampricide into streams were perfected (Anderson, 1962). A small unit consists of a carburetor float and needle valve assembly of the type used in oil burners, and it discharges toxicant into the stream by gravity flow. A second unit is composed of an automotive-type fuel pump powered by a 12-volt wet-cell battery, and it yields excellent results in streams with 1.7 to 85 m³/min (1 to 5 cfs) of flow. A larger unit has a centrifugal or diaphragm pump driven by a 12-volt, battery-powered motor, and it is capable of dispensing 0.95 to 228 1 (0.25 to 60 gal) of lampricide per hour with accuracy.
The critically important on-site bioassays of TFM at target streams were implemented by equipping a small house trailer as a portable bioassay laboratory (Howell and Marquette, 1962). An auxiliary trailer unit supplies essential electric power by means of a propane=driven generator. The TFM is bioassayed against larval lampreys and rainbow trout in water from the target stream. The results indicate the safe and effective concentration to be applied to the stream.
A method to estimate the minimum concentration of TFM lethal to sea lamprey and the maximum that will not kill fish was devised by relating the concentrations to the alkalinity and conductivity of receiving waters (Kanayama, 1963). The estimates are not substitutes for on-site bioassays, but they enable efficient planning of bioassays by delineating the range of concentrations to be tested.
The continuing research on means and methods for control of sea lamprey soon produced other benefits. Investigators found that the molluscicide, Bayluscide (5,2‘-dichloro-4’-nitro-salicylanilide), synergizes TFM (Howell et al., 1964). Mixtures containing 2 percent of the molluscicide do not detract from the selectivity of TFM against lampreys, but do reduce treatment costs by approximately 50 percent. Later, granular formulations of Bayluscide were found to be effective for treating lamprey larvae in estuaries and alluvial fans and on the lake bottom near mouths of rivers (Tibbles, Lamsa, and Johnson, 1969).
The enormity of the sea lamprey problem is indicated by the size of the upper Great Lakes -- Lakes Superior, Huron, and Michigan. Together they have an area of 168,946 km² (65,230 sq mi) and a shoreline of 12,570 km (7,856 mi). Surveys made with electrofishing gear in the 1950's disclosed that about 10 percent of the 3,000 tributary streams -- ranging from small brooks to large rivers -- contained sea lamprey larvae. These larvae, which live 4 to 7 years as ammocoetes in streams, are the targets of the control efforts with TFM. Because Lake Superior was the last of the lakes to be invaded by the sea lamprey, some natural stocks of food and game fishes remained in 1958. It was to protect these stocks that the initial control efforts were applied in tributaries to Lake Superior. By 1961, the principal lamprey streams had been treated once, and the incidence of lamprey wounds on lake trout had declined sharply (Baldwin, 1968).
Re-infestation of streams occurred as adult lampreys moved from Lake Superior into tributaries to spawn after 12 to 20 months of parasitic life. The second round of treatments in tributaries began in 1962 and was completed in 1966 with the result that assessment catches of spawning lampreys declined to one-tenth of pre-control numbers. Stocked and native populations of lake trout showed increased survival and production. Since 1966, some streams have been re-treated for the third time or more, and special efforts have been made to eliminate larval lampreys in the alluvial fans of some major rivers (Tibbles, Lamsa, and Johnson, 1969). The objective on Lake Superior is to hold lampreys down to 5 percent of pre-control numbers in order to obtain an 85 percent of “normal” production of lake trout, worth an estimated 2 million US dollars per year. The annual cost of lamprey control to reach these objectives is about $400,000.
Control efforts were extended to lamprey-infested tributaries of Lake Michigan in 1960, and the first round of treatments was completed in 1966. The reduction in numbers of sea lamprey is reflected in a multi-fold increase in commercial catch of whitefish. Also, the angler catch of rainbow trout (steelhead) in Lake Michigan was the highest ever in 1967.
The treatment of tributaries to Lake Huron began in 1960, but was interrupted because of a lack of funds. The control program was reactivated in 1966, and the first round of treatments is scheduled for completion in 1970. Continuing surveys during the 1960's showed that there are at least 12 lamprey streams on Lake Erie and 22 on Lake Ontario, and control efforts may be extended to them in the 1970's.
In its Annual Report for 1968, the Great Lakes Fishery Commission (1969) provided some evaluations on sea lamprey control in the upper Great Lakes. The chemical treatments in Lake Superior streams were not in time to prevent the failure of lake trout reproduction in the inshore waters of Lake Superior where mature fish had been reduced drastically by lampreys. But, the survival of stocked and native lake trout improved after 1962, and natural reproduction resumed in 1965. The annual stocking of yearling lake trout that began with 987,000 fish in 1958 increased to more than 2 million fish in 1963 and to more than 3 million fish in 1968. Assessment efforts in Michigan waters of Lake Superior indicate that there were 39 marketable lake trout caught per 3,048 m (10,000 ft) of large mesh gill net in 1962 in contrast with 245 trout taken in 1968. Other recent plantings of salmonids in Lake Superior have included coho and chinook salmon, and rainbow, brown, and brook trout; and these species are supporting an increasingly prosperous sport fishery.
The residual population of sea lamprey in Lake Superior, as determined by weirs on 16 index streams, has ranged from 7 to 23 percent of pre-control level. The objective of reducing lampreys to 5 percent of pre-control level has not been attained yet, and a need for continuing control is indicated. A similar situation exists in Lake Michigan; lamprey numbers have been reduced and the survival of food and sport fishes has improved greatly. High costs, however, have prevented as thorough an assessment of control efforts on Lake Michigan as that made on Lake Superior.
16) State-wide rough fish control in New Mexico in 1961 involved Negrito Creek to remove suckers (Regan, 1961), the upper watershed of Lake Maloya to eliminate suckers and other rough species (Harrison, 1961), and Bonito Creek to control the carp, goldfish, and other minnows that were possibly present (Little, 1961). Applications of rotenone resulted in large kills of fish, and each of the reclaimed waters was stocked successfully with brown trout, rainbow trout, or cutthroat trout.
Later, nearly 160 km (100 mi) of the San Juan River drainage were treated with rotenone to control flannelmouth sucker and bonytail (Olson, 1962). Drip stations were located 8 to 28.8 km (5 to 18 mi) apart on the streams, with the upper stations dispensing 2 mg/l, (ppm) for 48 hours, lower stations dispensing 1 mg/l (ppm) and still lower stations applying 0.5 mg/l (ppm). Water temperatures ranged from 13.3 to 19.4° C (56 to 67° F) and the treatment was considered very successful. The river system was restocked with salmonids.
A treatment of 104 km (65 mi) of streams in the Ute Lake watershed in 1962 with powdered and liquid formulations of rotenone killed many rough fish, but fishery managers considered it unlikely that any species was eradicated (Jester, 1963).
17) The increasing use of toxicants in streams prompted studies on the resilience of fish populations and the relative efficiency of rotenone and electrofishing in taking fish in small streams in Pennsylvania (Boccardy and Cooper, 1961 and 1963). A concentration of at least 1 mg/l (ppm) of rotenone was applied twice to 5.6 km (3.5 mi) of a small stream to eradicate brown trout, white sucker, northern hog sucker, stoneroller, creek chub, longnose dace and blacknose dace. The objective was to restore native brook trout. No living fish could be seen after the first treatment, but electrofishing disclosed the presence of many target fish. The second treatment with rotenone 1 year later turned up young-of-the-year brown trout as well as adults at about 66 percent of their former abundance. Stoneroller had increased greatly in abundance in the year, and 23 species were taken in contrast with 16 species during the first treatment.
In a series of comparative trials, rotenone proved to be a more effective sampling tool than electrofishing in the soft-water streams. The bolts of rotenone were color tagged with sodium fluorescein and detoxified, when necessary, with potassium permanganate. The investigators concluded that fish in streams may be eliminated with rotenone providing that extraordinary precautions are taken to treat the waters thoroughly. They noted that surviving fish populations recover in numbers very rapidly.
18) One of the most noteworthy reclamations of the early 1960's was that on the Green River in Utah and Wyoming. Before the Flaming Gorge Dam in Utah was closed in November 1962, approximately 720 km (450 mi) of the Green River and tributaries were treated with 81,681 1 (21,495 gal) of 5-percent rotenone. The objective was to suppress non-game fish populations to allow stocked rainbow trout to thrive in the new 16,800-ha (42,000-a) reservoir and connecting waters. Because of the magnitude of the reclamation, comprehensive pre-treatment surveys were made in Utah (McDonald and Dotson, 1959) and Wyoming (Bosley, 1960) over a period of several years.
Near Flaming Gorge, the river flows vary from about 20,400 m³/min (12,000 cfs) during spring runoff to average lows of about 765 m³/min (450 cfs). Water temperatures in the project area range from 0 to 23.9° C (32 to 75° F). Suckers, chubs, and carp comprised the bulk of the fish populations in the project area, but good trout habitat was found on the upper reaches of the river in Wyoming. Based on experience with the rapid rise of carp to overwhelming dominance in new reservoirs in Wyoming, this species was the principal target in the Green River project. A controversy arose prior to the reclamation, however, because some biologists and a scientific society believed that the survival of a few species of fish indigenous to the Colorado River drainage would be threatened. These species are Colorado squawfish, humpback sucker, and humpback chub. The proponents of reclamation contended, however, that these species inhabit swift water environments, and their preferred habitat would be altered drastically by the impoundment. Also, state biologists in Utah, Wyoming, and Colorado found in pre-treatment surveys that the fish in question are well established in tributaries outside the project area.
Many experiments and field trials were performed during the year or two prior to the reclamation on formulations of rotenone, on apparatus for dispensing the toxicant uniformly over a period of hours, and on the effects of the toxicant on target fishes (Binns et al., 1963). The final plan called for treatment stations located 16 km (10 mi) apart on the river system at which the liquid formulation of rotenone (5 percent active ingredient) would be introduced at 5 mg/l (ppm) for a period of 6 to 7 hours. Barrels of rotenone (209-1 [55-gal] capacity) were coupled together with 2.5-cm (1-in) yokes and located at highest points on the river bank. The toxicant was delivered to the river by gravity flow through 2.5-cm (1-in) industrial hose. This hose was attached to a support wire that sloped from the high point of barrel location across the river to the opposite bank. Steel fence posts kept the wire and hose well above the surface of the river. Because the main channel ranged from 60 to 90 m (200 to 300 ft) in width, control valves were coupled between hose sections to permit discharge of toxicant into the water at two or three points. The system worked well. The stations which were located 16 km (10 mi) apart were put into operation on September 4, 1962 at consecutive 3-hour intervals proceeding downstream. The reclamation was completed on September 7, 1962. Many sloughs offside or discontinuous with the main channel were treated by means of helicopter. Heavy use was made of airboats and helicopters to ferry supplies and to monitor the progress of reclamation.
A detoxification station was established at a bridge 49.6 km (31 mi) below Flaming Gorge Dam, and crystalline potassium permanganate was dispensed to the river through electric-powered pellet feeders at 2.3 mg/l (ppm) of permanganate per 1 mg/l (ppm) of rotenone. Live-cages containing several species of fish were located downstream to evaluate the detoxification.
Some problems were encountered in the detoxification process. The crystals of potassium permanganate were too small to feed well through the pellet dispensers. The rate of decay of rotenone in the various segments of the treated area was not consistent, and varying concentrations of rotenone passing the detoxification station were difficult to monitor and treat appropriately. Also, the volume of flow in the Green River rapidly decreased during the treatment period, and difficulties with the time lags were experienced in correlating both the application and the detoxification of rotenone with data from stream gauging stations. As a result of the problems, some toxicant reached Dinosaur National Monument 25.6 km (16 mi) downstream from the detoxification station and also Echo Park 28.8 km (18 mi) further downstream. Some fish were killed on September 13 to 15, but many live fish of various species were sampled later and no species was eliminated.
The kill of fish downstream in Dinosaur National Monument precipitated protests by the American Society of Ichthyologists and Herpetologists against the mass eradication of fish from large bodies of water (Hubbs, C.L. 1963). As a consequence, the Secretary of the U.S. Department of the Interior reviewed the Green River project and issued the following directives on any future reclamations where Federal funds are involved:
1) That adequate research be undertaken on the effects of rotenone, potassium permanganate, or other fish controlling agents under varying environmental conditions before additional reclamations are accomplished.
2) That whenever a reclamation may pose a danger to a unique species, the potential loss must be a dominant consideration in evaluating the advisability of the project.
3) And that future projects are to be reviewed by disinterested parties to assure that experimental work is taken into consideration and possible deleterious effects are avoided.
Moreover, the Secretary ordered the National Park Service and the U.S. Bureau of Sport Fisheries and Wildlife to study the impairment of fish populations in Dinosaur National Park. Fortunately, pre-treatment data were available for comparison with post=treatment sampling, and the study showed only minor shifts in population numbers and no eliminations of unique species. A comprehensive report on the fishes of the Green River below the Flaming Gorge Dam and the ecological changes in the river caused by the impoundment was made by Vanicek (1967). A companion study on the macroinvertebrates downstream from Flaming Gorge was made by Pearson (1967).
Flaming Gorge Dam was closed in November 1962, 2 months after the rotenone treatment. The new reservoir was stocked heavily with fingerling rainbow trout, including 3.3 million fish in 1963 and 3.9 million in 1964 (Eiserman et al., 1965). Some kokanee salmon (lacustrine sockeye salmon) also were stocked in the reservoir, and some brown trout in addition to rainbow trout in the river system. Growth and survival of trout were good. The catch of rainbow trout in Flaming Gorge Reservoir in 1964 amounted to 676,686 fish. In 1965, a creel census on the reservoir showed that 191,619 anglers caught 800,657 rainbow trout weighing nearly 189,875 kg (421,944 lb). The stocking of fingerling rainbow trout was reduced to 1.7 million fish in 1966, but 266,473 anglers caught 793,665 rainbows weighing 349,564 kg (776,808 lb). The development of a trout fishery was considered a success.
Intensive post-treatment sampling in Flaming Gorge Reservoir and upstream was accomplished from late 1962 through 1964 (Binns, 1967). The investigators found that brown trout, mountain whitefish, redside shiner, speckled dace, fathead minnow, and flannelmouth sucker had survived the treatment and regained their pre-treatment distribution within the 2 years. The cyprinids again were abundant. Utah chub and carp were found only in scattered places. The biologists tend to attribute the survival of these species to sloughs that were overlooked during the treatment process. In contrast, Colorado squawfish, humpback sucker, channel catfish, and black bullhead did not appear after the treatment because the only source of recruitment was from below Flaming Gorge Dam. The survey also disclosed that aquatic invertebrates had been reduced drastically by the rotenone treatment, but recoveries of populations occured within months to 1 year. Some changes, however, were noted in the composition of populations.
By 1970, the Utah chub became a problem in Flaming Gorge Reservoir. It is abundant; it is competitive with game fish; and it interferes with sport fishig for salmonids. Fishery managers are looking into possibilities for selective kill or capture of seasonal concentrations of the Utah chub in littoral zones of the reservoir.
19) Fish and Ratledge (1963) published recommendations for the management of reclaimed trout streams, and based them on the results of treating Fires Creek in Nantahala National Forest, North Carolina, with rotenone. A barrier was installed in the creek to prevent re-infestation, and rough and competitive fishes were eliminated. The establishment of a rainbow trout fishery was considered an outstanding success, and fishing pressure increased by 48 percent in the year following reclamation. Among the recommendations, the investigators emphasized proper and adequate post-treatment stocking to obtain at least one good natural spawning, and they advised that potential spawning fish be protected from over=exploitation.
20) An ichthyologist's point of view on the use of rotenone to improve sport fishing appeared in 1963 (Hubbs, C. 1963). The great increase in reclamations was noted, with some projects resulting in long-term improvements in fisheries and other projects failing. The failures indicated unsolved problems in reclamation techniques. Taking the Concho River in Texas as an example, the author observed that treatment with rotenone drastically altered the ecological interactions and had selective effects in that some organisms were damaged and others were not. Gamefish food organisms were harmed, but typical carp and sucker foods were not. He proposed, too, that partial reclamations of waters with rotenone may lead to poison-resistant strains of rough fish. He also criticized claims of successful reclamations when rough fish are reduced by 50 to 95 percent because the fecundity of most rough fish permits immediate recovery. Moreover, he contended that fishery biologists fail to prove that rough fish are causing problems in sport fisheries.
21) The S.B. Pennick & Company (1963), processors of rotenone, advised on the use of Noxfish and Pro-Noxfish in streams. The products degrade while moving downstream, and the rate of degradation depends on concentration, water temperature, aeration, alkalinity and sunlight. Chlorine or potassium permanganate can be added to the water to detoxify rotenone. The need for uniform distribution of the toxicant in streams and lakes was stressed.
Price and Haus (1963) of the S.B. Pennick & Company described some aids for stream reclamation, including a liquid metering device for attachment to drums of rotenone.
22) The State of North Dakota has been very active in reclaiming streams and lakes. In his summary of 8 years of reclamation, Henegar (1964) reported good results in controlling carp, white sucker, black bullhead, and yellow perch in 160 km (100 mi) of the James River with Fish-Tox, a mixture of rotenone and toxaphene, at 1 mg/l (ppm).
23) Hooper et al. (1964) made a detailed evaluation of lake and stream reclamation in the United States and Canada. They found that fish eradication is usually incomplete even under the best of conditions, but stated that reclamations that result in improved fishing for 3 years or more are worth the investment. They recommended that treatments be repeated every few years on waters under intensive management and heavy fishing pressure. In respect to streams, they noted that reclamation of flowing waters has had limited success as compared to lakes because re-invasion by undesirable fish is often rapid.
24) Irrigation dams and diversions made the reclamation of 184 km (115 mi) of the Niobrara River in Nebraska difficult despite preliminary tracing of volume and flow patterns with dye (Schoenecker and Petersen, 1965). However, great quantities of carp and freshwater drum were killed, and the river was restocked successfully with rainbow trout, brown trout, and centrarchids.
25) Water temperature limited the success of two major stream reclamations in Oregon. Herrig (1964) attempted to eradicate goldfish, northern squawfish, suckers, and brown bullhead in 139 km (87 mi) of Lower Crooked River and tributaries. The water temperature was 5.6° C (42° F), and several applications of rotenone at 2 mg/l (ppm) were made. Even then, some rough fish survived.
Bisbee (1968) treated 26.4 km (16.5 mi) of the Lower Owyhee River with rotenone in March to take advantage of low water flow, but the water temperatures were less than favorable at 7 to 13° C (46 to 56° F). Also, the helicopter employed to spray the liquid rotenone encountered poor weather, and numerous obstacles along the river hampered complete coverage by the aircraft. The toxicant failed to penetrate into depths of pools, and it degraded rapidly. A large kill of rough fish was obtained, but some of each species survived throughout the course of the treated area. The investigator suggested periodic re-treatment of the river to maintain a good trout fishery.
In contrast with the above projects, Swan (1965) reported success in reclaiming 62.4 km (39 mi) of Big Fall Creek and tributaries in Oregon. Twenty-four drip stations were established to introduce rotenone at 2 mg/l (ppm), and some stations were operated as long as 24 hours. A detoxification unit at the Fall Creek Dam applied potassium permanganate at 2 mg/l (ppm) for 84 hours. A machine prepared a slurry of the permanganate to be metered into the stream, thus eliminating some of the problems frequently encountered in applying permanganate crystals.
Swan (1965) added that the extent of fish kill was assessed by observers in SCUBA gear and by sampling with rotenone. No surviving fish were found. Extraordinary growths of algae occurred within 10 days after the reclamation, and they were attributed to the release of nutrients from dead fish and the reduction of insect life. Coho salmon, chinook salmon, and rainbow trout were stocked successfully in the stream.
26) The growth of stocked brown trout was phenomenal following the reclamation of Pine Creek, a small stream in Tennessee (Little, 1966). Goldfish, stoneroller, northern hog sucker, green sunfish, rock bass, and largemouth bass were target species, and their numbers were greatly reduced by rotenone at 5 mg/l (ppm) for 1 hour followed by 1 mg/l (ppm) for 5 hours. A successful brown trout fishery was established, but within 16 months 12 of the 17 species of rough fish had reappeared. The author concluded that barriers against re-invasion by rough fish should be present on reclaimed waters and that gamefish stocking should be arranged to obtain the earliest possible natural reproduction in quantity.
27) Meehan and Sheridan (1966) experimented with means to control the coastrange sculpin that preys heavily on fry of pink salmon in some Alaska streams. An application of toxaphene at 1.5 mg/l (ppm) for 18.5 hours in a small stream eradicated the sculpin, but temporarily eliminated insect fauna and reduced numbers of clams and snails. The penetration of toxaphene into the gravel stream bed was monitored by tagging the toxicant with sodium fluorescein and sampling it through standpipes located in the stream. Penetrations of at least 17.5 cm (7 in) were observed.
28) Antimycin (Fintrol) was registered as a fish toxicant in the United States and Canada in 1966 with successful applications in small streams in Wisconsin contributing to the registration (Gilderhus, Berger, and Lennon, 1969). A pre-impoundment treatment of Sidie Hollow Creek with 5 μg/l (ppb) of antimycin for 5 hours exterminated 11 species of fish within 12 hours at 10 to 18° C. The bolt of toxicant moving downstream did not repel fish, and no fish was observed fleeing ahead of the bolt.
An application of 7.5 μg/l (ppb) for 8 hours in Mullet River, characterized by a slow flow of very hard water, killed 8 metric tons of carp in 3 km (2 mi) of stream within 72 hours at 20° C. All rough fish except carp were eliminated, but less than 1 percent of the carp survived the treatment. The concentration of toxicant was perhaps marginal, and the loss of toxicant between application stations may have been greater than anticipated.
The effectiveness of antimycin in cold water was demonstrated in 9.6 km (6 mi) of Rathbone Creek. The water temperature was 5° C, but 7.5 μg/l (ppb) of toxicant applied for 8 hours killed 99 percent of the target fishes. Two mg/l (ppm) of potassium permanganate quickly detoxified the antimycin at the spillway of Cataract Dam.
A 1.6 km (1 mi) portion of Waterloo Creek offered an opportunity to evaluate the performance of antimycin against carp and white sucker in warm water (25° C) that contains a heavy load of effluent from a municipal sewage treatment plant. The toxicant at 10 μg/l (ppb) for 5 hours killed fish rapidly, and no survivors were found in-post-treatment surveys. The dye, rhodamine-B, was useful in plotting the flow characteristics of the stream.
The results of the above-mentioned trials contributed to the development of a liquid formulation of antimycin (Fintrol-Concentrate) for reclamation of streams. Among the advantages of the toxicant in streams are its effectiveness on carp and suckers, the lack of color and odor in the water, and the non-repellency to fish.
29) The possibilities of poisoning carp under ice were investigated by Hacker (1969), and resulted in a major application of Fintrol-Concentrate in 181 km (113 mi) of the Beaver Dam River system. The toxicant was diluted sufficiently with ethanol to prevent freezing in drip systems, and applied at 10 μg/l (ppb) for 24 hours in water that ranged from 0 to 4° C. The treatment was accomplished in mid-winter when widespread marshes along the river were frozen and inhospitable to fish. Surveys with nets and electrofishing gear after ice=out took no live fish, and the investigator estimated that the toxicant had killed 113 metric tons of carp and bigmouth buffalo.
Hacker (1969) concluded that the use of antimycin (Fintrol) during cold-water periods, or under the ice, holds much promise for control of carp and other rough fish in streams. Whereas a higher concentration of toxicant is needed than in warm water, the restriction of target fish to the mainstream by frozen marshes actually reduces the quantity of water to be treated. He found that access to the river across frozen marshes with all-terrain vehicles was easy because the ice ranged from 60 to 65 cm (24 to 26 in) in thickness.
31) A new and selective toxicant for squawfishes has been developed by MacPhee and Ruelle (1969). Its targets are the northern and Umpqua squawfishes that prey heavily on the young of Pacific salmons in fresh waters of western North America. The compound, called Squoxin, is said to be 3 to 100 times more toxic to squawfishes than to salmonids, and concentrations of 0.1 mg/l (ppm) or less are adequate for selective control. The toxicant is slow-working and short-lived in water, and is non-repellent to target fishes. If the compound lives up to its present promise, the management of salmon will be advanced.
The basic difficulties of reclaiming streams naturally led to the preparation of manuals on techniques. Much remains to be done, however, to perfect the techniques, to develop formulations of toxicants well suited to use in flowing water, and to draw on the many and wide experiences of biologists engaged in stream reclamations. Thus, more comprehensive and explicit manuals, properly published, are needed.
1) McCoy and Ratledge (1967) compiled a manual for reclamation of mountain trout streams in North Carolina. It is a mimeographed paper produced by the North Carolina Wildlife Resources Commission in limited distribution. The authors admit that many of the streams never can be restored to their original status, but they can be improved as trout habitat by removing the undesirable fishes. They emphasize that the need for reclamation be based on: 1) the extent of infestation with undesirable species that compete with trout for food or space; 2) the catch of trout per unit effort as compared with that of a good trout stream in the immediate area; and 3) the anticipated increase in fishing pressure and amount of stocking necessary to sustain it.
McCoy and Ratledge (1967) discuss in detail the following elements of pre-reclamation investigations: 1) the need for adequate assessment of fish populations in the project area; 2) the mechanics of time-motion studies of stream flow and establishment of the rate of longitudinal mixing; 3) determination of linear relationship of specific conductance and known salt concentrations at a constant temperature; 4) rotenone toxicity studies to determine concentration of toxicant and duration of exposure required to kill target species; and 5) detoxification of rotenone with potassium permanganate. They also present a plan of operation that includes personnel, equipment, and supplies.
Although directed toward the reclamation of small, soft-water trout streams, the manual by McCoy and Ratledge is a good start in filling a basic need, and it is unfortunate that it did not receive more adequate publication and better distribution.
2) Fernholz and Slifer (1967) outlined procedures for the chemical reclamation of soft=water trout streams in Wisconsin. The 15 steps involved in a stream reclamation project are:
A) Conduct a biological survey to determine need for project
B) Conduct a public relations program to win general support for the reclamation project
C) Obtain written approvals from riparian landowners, sportmen clubs, civic groups, and the public service commission as necessary
D) Set a tentative date for the project
E) Prepare an outline map of the project area including all access points, the stream, tributaries, backwaters, and toxicant stations
F) Establish stations for application of toxicant approximately 1.6 km (1 mi) apart
G) Measure volume of stream flow at each station
H) Establish a bench mark at every fifth station at the time the measurements of flow are made
I) Conduct salt-resistivity tests at each station, above and below the point of salt introduction
J) Calculate toxicant requirements for each station from reference table in the manual
K) Calculate the volume of any impoundments present and the amounts of toxicant needed
L) Establish a detoxification station if required
M) Establish stations for live-cages of test fish in treated area and untreated areas downstream
N) List manpower and equipment needs
O) Establish post-treatment survey procedures
Reference tables and formulae for calculating the characteristics of stream flow and the quantities of antimycin or rotenone needed are included in the manual. This manual, like the one by McCoy and Ratledge, is an advance in the complicated process of stream reclamation. It is regretable that it, too, did not receive more formal publication and wider distribution.
3) The Fish Control Laboratory of the U.S. Bureau of Sport Fisheries and Wildlife is doing research on materials and methods pertaining to the reclamation of streams (Gilderhus, 1970a). One study is focused on the minimum lethal concentrations and exposures to antimycin and rotenone that are required to kill selected target fishes. Another study involves the refinement of techniques for employing fluorescent dyes to trace stream flows and the dispersion of toxicants in streams (Gilderhus, 1970b). The Laboratory plans to incorporate the results of these and other studies into a manual on stream reclamation.
4) Research on stream reclamation also is in progress at the New York Fish Laboratory (Loeb and Engstrom-Heg, 1970). The studies involve 1) the detoxification of rotenone in flowing water with potassium permanganate followed by the instantaneous neutralization of the permanganate by tannic acid; 2) multiple chemical systems for reclamation, such as antimycin-potassium permanganate-sodium thiosulfate; and 3) utilizing measurements of chlorine demand of the stream to better calculate the concentrations of potassium permanganate needed for deactivation of a fish toxicant.
Large investments of funds and effort have been devoted to the reclamation of streams within the past two decades. In addition to the projects cited in this report, numerous small streams and inlets are treated with toxicants each year incidental to the reclamation of ponds, lakes, and reservoirs. Frequently, the details on treatment of tributary streams are lacking in reports made on the projects. One can assume in many instances that the importance of adequate stream treatment is under-estimated with the result that surviving fishes in the stream serve as sources of re-infestation of the lakes.
There are few unqualified successes in stream reclamation, but this is not surprising when one considers the great diversity of streams, the difficulties of attaining uniform dispersal of a toxicant throughout the confines of a stream, the problems of maintaining sufficient concentration of a toxicant and duration of exposure over long distances, and the lack of toxicants that are formulated specifically for use in streams. The art of stream reclamation is evolving more slowly than that of lake reclamation because of the greater complexities. The worth of stream reclamation as a fishery management tool is recognized, however, and its use will increase as methods and toxicants are improved. The research-based progress made in controlling sea lamprey larvae with a selective toxicant in streams tributary to the Great Lakes has had great influence in furthering the art of stream reclamation.
