THE EFFECTS AND FATE OF LAMPRICIDE > _. ITEM: 3 - TRIFLUORMETHYL - 4 - NITRGPHENOI.) ’ IN MODEL STREAM COMMUNITIES Dissertation for the Degree of Ph. D. ' ' MICHIGAN STATE UNIVERSITY ALAN WALTER MAKI 1974 LIB? "‘RY Michy 3mm '9, I‘TIVCH‘I‘Y This is to certify that the thesis entitled The Effects and Fate of Lampricide (TFM: 3-Trifluormethyl-h-Nitrophenol) In Model Stream Communities presented by Alan W. Maki has been accepted towards fulfillment of the requirements for Ph.D. degree in Fisheries and Wildlife Ogd/Kf/m/Eggdm/ Maj rpr essor Date jl 2/714 0-7639 QCLOS ‘-‘/F' J ABSTRACT THE EFFECTS AND FATE OF LAMPRICIDE (TFM: 3-TRIFLUORMETHYL-4-NITROPHENOL) IN MODEL STREAM COMMUNITIES BY Alan Walter Maki The effects of lampricide (TFM: 3-trifluormethyl- 4-nitrophenol) on the structure and function of benthic communities were studied in a series of six replicated in— door model streams. The systems were considered as models of a typical woodland stream each consisting of a 13 foot pool section and a 13 foot riffle section. Artificial il- lumination was provided for each stream at a uniform sur- face intensity of 900: 50 ft—ca. A specially developed in situ stream respirometry system was used for measurements of net primary production and community respiration in pool and riffle communities. Pre-treatment levels of gross primary production varied from 10.7 mgOzhr-lm_2 to 79.0 mgOzhr-lm—2 through the course of the year and were suppressed by 25-50% during exposure to 9.0 mg/l TFM. Community respiration varied from 10.5 to 36.2 mgOzhr—lm-2 and was increased 3-50% by TFM treat- ment. Calculated P/R ratios proved to be sensitive indi- cators of the influence of the toxicant. Alan Walter Maki No effects due to the lampricide exposure could be demonstrated on periphyton community structure including measurements of diatom cell density, diversity, equitabile ity, biomass or species composition of filamentous green algae. A statistically significant decrease in the total number of macroinvertebrates per unit area was demonstrated following treatment. Complete recovery to pre—treatment densities was observed within 2 to 3 months primarily through drift input. No treatment effect was determined from measure- ments of diversity, equitability, number of taxa or produc- tion rates of the macroinvertebrate community. A four-fold increase in drift rates of TFM-susceptible species was ob— served in the experimental channels during treatment. Mean growth rates of experimental and control popu- lations of young of the year brown trout, §§lmg trutta 1 day-1 varied from 1.709 to 3.676 mg gm_ in replicated ex— periments and no significant difference due to the lampri- cide treatment was demonstrated. The uptake of TFM residues employing l4C-TFM was demonstrated to be basically an adsorption process. A two component curve consisting of a rapid initial uptake rate during the first two hours followed by a reduced linear rate for the remainder of the 24 hour exposure best describes the uptake curve for all 20 plant and animal species examined. Alan Walter Maki A“. ‘. 2:?ations than those species with hard chitinized or .fiiiééreous exoskeletons. Rates of loss of TFM accumulations correlated with water current and substrate associa- The mean half-life for riffle species was 17.8 .9 and pool-dwelling species had a mean half-life of Zv“ 0‘ vws . ' "tam-om 4,- THE EFFECTS AND FATE OF LAMPRICIDE (TFM: 3-TRIFLUORMETHYL-4-NITROPHENOL) IN MODEL STREAM COMMUNITIES BY Alan Walter Maki A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 1974 ACKNOWLEDGMENTS I wish to thank the following for their advice and assistance during my doctoral program: Mr. Les Geissel for his assistance in construction and maintenance of the model streams and assistance in all phases of field and laboratory work. Mr. Leo Mrozinski for his help in fish and lamprey collections and for providing the facilities of the Paris Fish Hatchery for this investigation. Dr. Kenneth W. Cummins for invaluable consultation throughout all phases of this study and for offering his time and that of his staff to aid with data interpretation and taxonomic problems. Drs. Niles R. Kevern and Robert C. Ball for consul- tation and encouragement, and for serving as members of my guidance committee. A special thanks to Dr. Howard E. Johnson for offer- ing me the opportunity to pursue this program and for many hours of invaluable advice and consultation throughout the last three years. Finally, I wish to thank my wife, Ann, for her moral support, patience and good humor during the last ii than they probably realize. I J"‘ . ‘ - kip“ This research was supported through Research Con- Pt“ u:‘E‘Mo. 14-16-0008-651 from the Bureau of Sport Fisheries fig administered by Dr. Joseph B. Hunn and through the irt. {61:4 . , f‘ .I.‘ chigan State University Agricultural Experiment Station. TABLE OF CONTENTS - I —a.. Page Hg, {94;fl ‘I‘ mgr or TABLES . . . . . . . . . . . . vi _tfi§§sm or FIGURES . . . . . . . . . . . . ix u ‘4'," s 7 us: ; VfiTBRATURE REVIEW . . . . . . . . . . . 1 5 A.. History of the Sea Lamprey in the Great I ? ‘C Lakes . . . . , . . . . . 1 ' ",'. The Development of TFM . . . . . . 2 ' »~.$§4 Chemical and Physical Properties . . . . 6 E ' Environmental Safety . . . . . . . . 8 ‘ Scope of Use of TFM . . . . . . . . . 11 - ‘infl’RODUCTION . . . . . . . . . . . . . Experimental Approach . . . . . . . . The Model Streams . . . . Determination of Treatment Concentration . . Model Stream Treatments . . . . . . Photosynthesis-Respiration Chamber . . . . Periphyton Sampling . . . . . . . . . ; Macroinvertebrates . . . . . . . . . «Fish Production . . . . . . . . . T Kinetics of TFM Residues . . . . . . . flCommunity Metabolism . . . . . . »Periphyton Community Structure . . . . . Macroinvertebrates . . . . . . . . . Page '3 - Kinetics of TFM in the Stream Community . . . 109 w Uptake of TFM Residues . . . . . . . . 109 Elimination of TFM Residues . . . . . . 118 I u o a O I o o o o a a I a 0‘ 124 ': Community Metabolism . . . . . . . . 124 Periphyton Community Structure . . . . . . 128 Macroinvertebrate Community Structure . . . 131 _‘ A I. .14 Fish Production . . . . . 138 » Kinetics of TFM in the Stream Community . . . 139 2: o o o o I o a a I c o o o I I 144 mug-1;; LUSIONS AND RECOMMENDATIONS . . . . . . . 147 T 3‘ . . I . . . U . . . U I . . . 150 ‘ viz: EST 01‘ REFERENCES . . . . . . . . . . . . 154 ":3 alum. :iui'; Table l. —..-_W- I }I A“ I i IMI‘ INN LIST OF TABLES Page Annual amounts of TFM applied to each of the upper Great Lakes' tributaries 1958— 1972 and the total costs of U. S. and Canadian programs for sea lamprey control and research . . . 12 Water quality summary for model streams June- December, 1973 . . . . . . . . . . 21 Taxonomic listing of flora and fauna of indoor model streams, Paris, Michigan 1973—1974 . . 24 Time-to- death of larval lampreys, Ichthyomyzon unicuspis, and rainbow trout, Salmo gairdnerii, tested in several concentrations of field grade TFM. Figures represent total number of mortalities from two replicated tests furing each time interval, a total of 8 individuals tested.per concentration . . . . . . . . 26 Sample calculations for a higly productive riffle community in Stream B on August 21—22 . 46 Primary production, community respiration and ; P/R ratios for Stream B, August 21 through 24. I Communities exposed to 9. 0 mg/l TFM on Aug- ust 23 . . . . . . . . . . . . 48 Results of Students' t-test of significant difference between annual mean values of commu- nity metabolism and standing crop data from control and treated dates . . . . . . . 56 Structure of diatom communities existing in Model Stream A control and experimental channels . . . . . . . . . . . . . 66 Structure of diatom communities existing in Model Stream B control and experimental channels . . . . . . . . . . . . . 67 vi . Table 10. 11. 12. l3. 14. 15. 16. 17. 18. 19. Page Approximate percent composition of the fila- mentous green algal flora of Stream A experi- mental and control channels determined from wet mounts of preserved material . . . . . 69 Total number of taxa collected from each stream community throughout the sample year . 73 Shannon diversity indices and equitability in terms of numbers of individuals from each community during each season and annual means . . . . . . . . . . . . . . 75 Community structure of benthic macroinverte- brates from Stream A control and experimental channels . . . . . . . . . . . . . 78 A comparison of standing crop, production and cohort turnover ratios for two Species of macroinvertebrates from control and experi— mental channels of Stream A . . . . . . 89 Community structure of benthic invertebrates in Stream B control and experimental chan- nels . . . . . . . . . . . . . 91 Comparison of standing crop, production and cohort turnover ratios for two species of in- vertebrates from control and experimental channels of model Stream B . . . . . . . 96 Results of Students' t—tests of significance of difference between pre— and post- treatment community structure of benthic macroinverte- brates . . . . . . . . . . . . 97 Comparison of standing crOp, production and turnover ratios for Gammarus pseudolimnaeus from control and experimental pool channels of model Streams A and B . . . . . . . 99 Community structure of benthic macroinverte— brates from Stream C control and experimental riffles . . . . . . . . . . . . 101 Page A comparison of standing crop, production and cohort turnover ratios for two species of mac- roinvertebrates from control and experimental riffles of Stream C . . . . . . . . . 103 ‘54fl4 Growth and production of young-of—the-year ‘ brown trout, Salmo trutta in experimental and control channels of Streams B and C following an exposure to 9. 0 mg/l TFM . . . . . . 105 L',I 22. A comparison of individual growth rates of “ , brown trout, Salmo trutta, between control and experimental channels of Model Streams B and C . . . . . . . . . . . . . . . 108 ' 23. Accumulation of total TFM residue in several plant and animal components exposed for 24 hrs. to a recirculating concentration of 9. 0 mg/l TFM . . . . . . . . . . . . . 116 24. Half—life and rate of loss of total TFM residue from several plant and animal components fol— lowing a 24 hr. exposure to 9.0 mg/l TFM . . 122 #‘wm‘tHJ‘J- '10.: .. 425. A comparison of community metabolism measure- ‘ ments obtained in the present study with esti- mates from other laboratory and natural systems . . . . . . . . . . . . . 127 -41, v viii LIST OF FIGURES Figure Page 1. Numbers of sea lamprey taken at electri- cal barriers operated in 8 tributaries of Lake Superior (1958-1972) . . . . . 14 2. Diagrammatic representation of photo— synthesis-respiration (P/R) chamber . . 28 3. Cross- -section diagram of stream periphyton sampler . . . . . . . 34 4. Cumulative number of taxa obtained with increasing number of samples demonstrat- ing inflection point . . . . . . . 37 5. Community metabolism in Stream B experi- mental riffle during August 21- 24 . . 49 6. Community metabolism in Stream B experi- mental pool during August 21-24 . . . . 50 7. Rates of gross primary production recorded throughout 1973 under an illumination of 900 ft.-ca . . . . . . . . . . . 52 8. Rates of community respiration observed in riffle and pool areas throughout 1973 . 54 9. Pre- and post—treatment mean P/R ratios for riffle communities compared with ratios obtained under exposure to 9.0 mg/l TFM . . . . . . . . . . . 59 10. Diatom cell densities in Stream A . . . 61 ll. Diatom cell densities in Stream B . . . 63 12. Biomass as ash—free dry weight of fila- mentous green algae from control and ex- perimental channels of Stream A . . . 70 ix Page Number of macroinvertebrate individuals per unit area of substrate in Stream A riffles and pools . . . . . . . . 81 Standing crop as ash-gree dry weight of macroinvertebrates from Stream A . . . 82 Simultaneous drift rates from Stream A experimental and control riffle channels during treatment with 9.0 mg/l TFM . . . 84 Simultaneous drift rates from Stream A experimental and control pool channels during treatment with 9.0 mg/l TFM . . . 87 Simultaneous drift rates from Stream B experimental and control riffle channels during treatment with 9.0 mg/l TFM . . . 94 Uptake of 14C-TFM expressed as loss from water. Mean 3 l 5.0. . . . . . , , 110 The rate of uptake of 14C- -TFM by the amphipod, G. pseudolimnaeus during a 24 hour exposure to 9. 0 mg/l . . The rate of elimination of 14 C-TFM by the amphipod, G. pseudolimnaeus follow- ing a 24 hour exposure to 9. 0 mg/l TFM . LITERATURE REVIEW History of the Sea Lamprey in the Great Lakes The history of the expansion of the sea lamprey, Petromyzon marinus Linnaeus through the Great Lakes has been_\~ well documented (Hubbs and Pope, 1937; Shetter, 1949). Prior to 1921, the sea lamprey had been found only in Lake Ontario and its tributaries where it is apparently native (Dymond, 1922). Their attacks on local fish species there were noted at an early date (Gage, 1893; Surface, 1898, 1899). Until relatively recent times the expansion of the sea lamprey through the remaining Great Lakes was halted by the natural barrier imposed by the Niagara Falls. It is gen- erally accepted that the completion of the Welland Canal from Lake Ontario to Lake Erie in 1829 opened the route for the movement of the sea lamprey into the upper lakes. The construction of the Trent Waterway connecting Lake Ontario with the Georgian Bay of Lake Huron in 1918 opened another possible route of introduction (Applegate, 1950). During the next 25 years the record of the expansion through Lake Erie, Huron, Michigan, and Superior, in that order, was ascertained from singular fisherman's encounters and l I : I documentation of spawning runs occurring in the tributari The expansion was considered complete in 1948 with the first report of a sea lamprey from the Minnesota waters of Lake Superior. In 1947 and 1948 an inventory by the Michigan Department of Natural Resources indicated the presence of migrating sea lampreys or spawning activity in 92 Michigan streams in the drainages of Lakes Erie, Huron, Michigan and Superior. Lamprey predation appears to be the cause of the de- cline of the lake trout fishery in Great Lakes waters through the 1940's and 1950's, and is probably the factor that is presently limiting abundance of spawning stocks in Minnesota, Wisconsin, and Michigan waters. This is indicated from an examination of the relation of wounding rates to length of trout (Pycha, 1970). The wounding rate for trout in the early 1960's averaged zero at about 13 inches, 20% at 20.5 inches and 30% at 22.5 inches. Very few fish survived past the length at which 25—30% of the fish were wounded which resulted in a very low spawning population. Summary data through 1972 have also shown the relationship of scarring frequency to total length is found in other species in- cluding Chinook and coho salmon from Lake Huron and Michi- gan (Pycha, 1972). The Development of TFM The problem of controlling the sea lamprey in the Great Lakes has received considerable attention since their impact on the commercial and sport fishery became apparent (Applegate and Moffet, 1955). Electromechanical wiers, traps and electrical fences were initially employed to block or destroy migrating adult sea lampreys. These devices, when installed in all known major spawning streams, provided an effective method of reducing the numbers of sea lampreys in each lake basin (Applegate, Smith and Nielsen, 1952; Erkkila, Smith and McLain, 1965). Unfortunately, problems arising with the cleaning and maintenance of these structures particularly during spring runoff make these methods economically undesirable. Also, a control program based on the prevention of spawning has no effect on the substantial reserves of ammocoete larvae already existing in upstream areas. Since the larvae may spend as many as 5—15 years in the bottom muds of a stream before transforming (Howell, person. comm.), there exists the potential for continued production of high num- bers of adult parasitic lampreys for several years after the installation of a 100% effective wier. In light of these problems, research was initiated in the mid-1950's to find a more direct and immediate method for control of the sea lamprey. A detailed investigation of the life history of sea lamprey in Michigan revealed that the mean length of actual parasitic existence in the open waters of the lakes is only about 17 months. After this parasitic stage, the sexually mature adults begin upstream spawning migrations in the winter months with the peak run occurring in late May and June. Spawning occurs in shallow riffles over a gravel and small rubble substrate with hatch- ing approximately 20 days later. The newly hatched individu— als (mean length approx. 8.3 mm) then burrow into soft silt banks and exist in relatively dense colonies with as many as 15 individuals per square foot in these backwaters and areas of low flow. The most vulnerable period in the life history of the lamprey appears to be thrs long, non—parasitic larval stage and the most direct attack would be the intro- duction of a toxicant to eliminate these larvae while they are concentrated in relatively high densities and relatively accessible areas of river bottom. The use of broad spectrum fish toxicants was undesirable because many of the lamprey spawning streams support substantial populations of other game fishes. Therefore it was decided to develop chemicals acutely toxic to larval lampreys at extremely low concentra- tions and non-toxic, at the same concentrations, to other fish species inhabiting the same natural environment (Apple- gate, et. al., 1957). The initial step in achieving this objective was the preliminary screening of a large and di- verse series of predominantly organic chemicals. Static bioassays were employed as the test procedure and were de— signed to yield relative toxicity data at low levels within short time periods. The screening tests revealed some compounds, at par- ticular concentrations, to be more toxic to larval lampreys than to fishes. Among these substances was a group of mononitro-phenols consisting of several derivatives contain- ing various amounts of the halogens, bromine, chlorine and fluorine. The differential toxicities of these mononitro- phenols covered a sufficiently broad range of concentrations to permit their regulated application in natural streams to destroy lamprey larvae without damage to other fish. Re- quirements for a short treatment period were met by these compounds and at minimum effective concentrations of each of the candidate chemicals, all of the lamprey were killed in 16 hours. Two of the compounds killed all of the lampreys in less than 45 minutes without apparent harm to the game fish which remained exposed to these concentrations for 24 hours. Some of the compounds were also tested against sev- eral fish species in a running water raceway in which natural stream flow conditions were simulated. In these cases aque- ous solutions of the sodium salts of each of these phenols were applied under conditions where continuous flow required their regulated metering into the stream during the entire test treatment period. Under these conditions, the toxic effects of the substances upon fishes seemed to be consider- ably less than the effects observed under laboratory-jar test conditions. Exposures to concentrations four times the dose needed to kill 100% of the larval lampreys did not cause any evident harm to 12 different fish species present in the raceways (Applegate et. al., 1958). Although all of these compounds are more toxic to lampreys than to most other aquatic organisms, certain ones are more desirable for practical applications because of their physical and chemical properties, ease of handling in the field, effectiveness at low concentrations, and cost. Of the compounds tested, TFM (3—trif1uormethyl-4-nitrophenol) most closely met these requirements and was selected for development for field use (Applegate et. al., 1961). Chemical and Physical Properties TFM exists as a crystalline solid at room tempera- ture, yellow to orange in color in its pure form, and light yellow-brown in technical grade preparations. It has a melt- ing point of 760C, molecular weight of 207 and ionization constant of 4.4 X 10-7 (DeBrouwer, 1930). It is sparingly soluble in water (0.498 gms TFM/100 gms HOH at 24.50C) but highly soluble in most organic solvents. Aqueous solutions of TFM are acidic (pK 6.07 i 0.03) and form phenolate salts in the presence of alkalis. Phenolates of the alkali metals are weak bases, the free phenol is colorless in acid solution but deep yellow in base solution. The compound is highly stable and resis- tant to biochemical degradation. It is not detoxified by 1 any known natural process (Applegate et. al., 1958). Daw- son, (1971) concluded that the toxicity of TFM to fish is not significantly reduced by short-term exposures to sun- light as an ultraviolet source. TFM is selectively toxic among larval lampreys and fishes not only in the form of its free phenol but also as a sodium salt (Applegate et. al., 1958). The solubility of the sodium phenolate in water is much greater than that of the free phenol. Since the larvicide is best applied in the field as a water-miscible liquid formulation, only the phenolates are practical because their greater solubilities permit the preparation of more highly concentrated stock solutions. Certain of the mixed amine salts of TFM are as effective in their biological properties as the sodium salts. The amine salts have more desirable physical proper- ties than the sodium salts, permitting commercial prepara— tions of greater stock strength with better stability of the solutions when stored at low temperatures. Commercial formu- lations of the amine salts are used almost exclusively in the field treatments (Applegate et. al., 1961). The TFM used as a lampricide in the United States is manufactured by the Farbewerke Hoechst Ag in Frankfurt, Germany, for the Great Lakes Fishery Commission. Technical and field grade formulations have the following composition: Technical Grade: An aqueous formulation containing 82.4% TFM, used to formulate the field grade TFM (35%). Field Grade: The field grade TFM (35%) is formulated with DMF (N,N-dimethyl formamide) in a highly basic solution. Environmental Safety Early field tests to determine the toxicity of TFM under natural stream conditions were conducted in the Mos- quito, Silver, and Pancake Rivers in Michigan (Applegate et. al., 1961). Each application concentration was based on pretreatment toxicity tests conducted with water from the test stream. An application of 5.5 mg/l of TFM for 9 hours killed all lamprey larvae within 8 hours in the Mosquito River, and a concentration of 2.8 mg/l applied for 14 hours killed all lamprey larvae in the Silver River (U. S. Bur. of Comm. Fisheries, 1958). A range of 3 to 10 mg/l had been established for lamprey control in labora— tory tests (Applegate and King, 1962). A typical stream treatment involves conducting toxic- ity tests with actual stream water to determine the effective concentrations of the toxicant for each particular water quality. The term minimum lethal concentration is used to refer to the lowest concentration of TFM that kills 100% of the lamprey larvae in 9 hours or less, and the term maximum allowable concentration is used to refer to the highest con— centration of TFM that does not kill more than 25% of the test fish in 18 to 24 hours. These two concentrations then define the concentration and time limits within which the stream will be treated. The proper concentration of field grade TFM is then metered into the stream with calibrated pumps and the concentration is fortified at predetermined 'downstream locations to correct for dilution by tributaries and springs. An experimental study was conducted in five streams tributary to Lake Superior and four tributary to Lake Michi- gan (Torblaa, 1968). Samples of the bottom fauna before and after treatment revealed that most groups of aquatic organisms were not adversely affected by exposure to lamprey larvicide. The total number of insects was less one week after treatment than before treatment, increased somewhat by 6 weeks after treatment and had returned to pretreatment levels one year after treatment. However, the sample sizes were small and the results may more accurately reflect in- herent variability in substrate types than the effect of TFM. A similar experimental design was employed by Haas (1970) to evaluate the effect of TFM on bottom fauna. In- vertebrate community structure was unaffected by exposure to 1 mg/l but a treatment at 4 mg/l reduced the densities Of all but two species examined. Small sample sizes and general variability in substrate type prevented an exact determination of the effects of TFM. Exposure to a concentration of 10 mg/l TFM was fatal to 100% of exposed hydra, turbellarians, and blackfly 10 larvae, 99% of burrowing mayflies 89% of leeches and 50% of clams tested in static toxicity tests (Smith, 1967). Similar methods were used by Marking (1971) to determine the toxicity of TFM to two genera of aquatic invertebrates, snails (Physa sp.) and scuds (Gammarus sp.). The 96-hour LC50 for purified TFM was 1.29 mg/l for snails and 1.80 mg/l for scuds at 120C. TFM was tested by Marking (1971) for its toxicity to 11 species of cold and warm water fish. In standard static toxicity tests with all species, the LCSO'sranged from about 1 to 6 mg/l of purified TFM. The toxicity of TFM was reduced by as much as a factor of 10 in hard water and by as much 59 at pH 9.5. In another investigation Dawson (1971) tested TFM against fingerling green sunfish in waters of various hardness buffered to pH's of 6.5, 7.5, 8.0 and 9.0. The compound was more toxic in softer waters than in harder water, especially at high pH values. The toxicity of TFM is related to the amount of free phenol present since when TFM is exposed to low pH (6.8), it exists as a free phenol, but when exposed to a high pH, most of the TFM exists as a phenolate ion (Lennon, 1971). The relationship of toxicity and hardness was investigated by Johnson (1970), demonstrating that increased concentra- tions of divalent metal ions decrease the toxic action of TFM. Considering the wide range of hardness and pH that 11 may be encountered in Great Lakes tributaries, concentra— tions ofll to 17 mg/l have proven lethal to exposed lamprey larvae (Dykstra and Lennon, 1966). An extensive review of the literature concerning safety and application of TFM was published by Schnick (1972). Scope of Use of TFM TFM was registered in the United States and Canada for use in lamprey control, based on research data developed through the U. S. Fish and Wildlife Service Laboratory at Hammond Bay, Michigan, and several experimental field trials in the early 1960's. Following applications of the toxicant to streams harboring known populations of larval lampreys, populations in the lake were reduced by approximately 80% by 1962 and 90% by 1966. Treatment schedules were expanded to tribu- taries of Lakes Michigan and Huron with similar sharp de- clines in lamprey populations. The magnitude of this expan- sion is represented in Table 1. This summary of annual amounts of TFM applied to each of the upper Great Lakes' tributaries since 1958 and the total costs of U. S. and Canadian programs for sea lamprey control and research was tabulated from Annual Reports of the Great Lakes Fishery Commission (1960-1972). Through 1972, a total of 1,476,407 lbs. of TFM, expressed as active ingredient had been ap- plied to Great Lakes tributaries at a total program cost in 12 Iakes' tributaries 1958-1972 and the total costs of U.S. and Canadian programs for sea lamprey control and re- search. S 3.)}- Amount of TFM Used Lamprey (lbs. active ingredient) Research & Control , Lake Lake Lake Lake Program Costs Superior Michigan Huron Ontario U.S. & Canada* 6,576 3?}1959 24,977 npj’isao 105,951 1,750 5,164 2,580,977 5“{4961 17,920 24,689 37,651 2,723,428 "?;a§62 64,743 15,173 3,114 2,499,533 .I L 71,529. ' 25,125 2,412,927 .1964 77,345 100,585 2,607,939 ‘oylées _ 36,458 62,843 2,571,294 E3€1§65I 32,599 16,819 30,963 2,620,298 28,409 31,124 40,101 2,801,490 7,594 64,706 12,118 2,659,370 24,960 42,255 23,711 2,662,199 45,745 23,609 49,099 3,167,737 40,030 37,134 63,523 18,319 3,369,912 31,798 65,224 51,004 13,970 4,752,622 ;.f616;§34 511,036 316.448 32.289 1,476,407 $37,429,726 n'ada, Department of Enviornment Canada, and U.S. Bureau ‘ I: heries and Wildlife. r l3 excess of $37.4 million. The impact of these treatments on spawning populations of sea lamprey in Lake Superior is rep- resented in Figure l, where the annual amounts of lampricide applied are plotted against the number of sea lamprey re- covered from electrical barriers operated on eight tribu- taries of Lake Superior from 1958 to 1972. The graph ac- curately reflects the current status of lamprey stocks in the lake. Due to the non—persistent nature of TFM stream treatments, this method of control constitutes only a short- term stress on the population and must be repeated within particular tributaries every 3-5 years. Present plans of the Great Lakes Fishery Commission call for continued TFM treatment at these relatively high levels in order to main- tain the present level of control. 14 Figure l.--Numbers of sea lamprey taken at electrical barriers operated in 8 tributaries of Lake Superior and the total amount of TFM applied to all tributaries of Lake Superior (1958-1972). or. ON. 0? as 980: 5.“: ho ucsoE< o? 8e FEES En.» II Amoco: .. FEES so .02 .8 6w .OOF INTRODUCTION The United States Environmental Protection Agency cancelled registration of TFM originally on December 31, 1970, stating that "All uses of this pesticide will be can- celled for use in water on December 31, 1970, unless toler- ance(s) is obtained in water and fish." However, the reg- istration was extended on a yearly basis to allow time "to obtain additional data on methodology and toxicology to support a petition which will be submitted proposing a permanent tolerance for TFM." To satisfy these objectives, the Great Lakes Fishery Commission obtained funding from the United States and Canada to support approximately $600,000 worth of registration oriented research on TFM. Several investigations are now under way in various con- tracted institutions and agencies to generate the required data. The present research was conducted with support from the U. S. Bureau of Sport Fisheries and Wildlife, Fish Con- trol Laboratory, LaCrosse, Wisconsin, as a section of this re-registration oriented research program. The specific ob— jectives of this investigation were to examine the effects Of? a TFM treatment on the structure and function of benthic Pelfliphyton and macroinvertebrate communities existing in 16 17 model stream environments and to characterize the uptake and elimination rates of lampricide residues in several biotic components of these streams. Experimental Approach The effects of a particular environmental pollutant are generally assessed on a single organism existing in iso- lated aquaria or toxicity test chambers for set periods of time. An animal exists in nature as an integral part of the ecosystem both exerting effects and being affected by the interactions of the natural community. Rarely have investi- gators taken into account the influence of the interactions of community structure when assessing the effects of a toxi- cant on an animal (Ellis, 1968; Seim, 1970; Williams, 1969). This community approach was used to examine the toxic influence of larval lampricide (TFM,-3-trif1uormethyl- 4-nitrophenol) on production and metabolism of replicated model stream communities. By imposing constraints on the physical and biological complexity of the model stream popu- lations, the description of the community interactions under the influence of the toxicant are simplified. Knowledge of the numbers and kinds of species in the model streams and description of trophic level associations permits an accur- ate assessment of the impact of TFM on the productive capac- ity of the stream ecosystem. A series of replicated model streams allows for the examination, under controlled 18 conditions, of the effects of incremental changes in an en- vironmental factor on ecological systems of convenient com- plexity and allows for the development of relevant predictions of the outcome of the effects of human activities on the biota of the system (Warren and Davis, 1971). Artificial streams as models of small natural streams employing constraints on the complexity of communities have been used in this way by several investigators to observe interacting individuals and populations leading to a descrip- tion of the operation of the community and existing trophic levels (Kevern and Ball, 1965; McIntire and Phinney, 1965; Davis and Warren, 1968). This investigation was designed to simulate an en- vironment where the toxicant is introduced into a stream allowing for colonization of the treated section by drift of upstream-dwelling organisms. Lamprey control rarely re— quires that TFM be introduced in the head waters of a stream since natural and manmade barriers effectively prevent adult lamprey migration and spawning in these upstream areas. MATERIALS AND METHODS The Model Streams The site selected for the construction of the model streams was the Department of Natural Resources fish hatch- ery located in Mecosta Co., Paris, Michigan. The area has an abundance of small spring fed streams and offers a wide diversity of insect fauna for colonization of the model streams and a source of organisms for planned toxicity tests. The hatchery is located adjacent to Cheney Creek, a small stream of about .098 m3/sec., arising from several springs approximately 1.45 km to the west. Paired cement rearing channels inside the hatchery were used as model streams. Each channel was 3.96 m long by 0.61 m wide and drained through a pair of 5.08 cm pipes into another downstream channel of equal dimensions. The replication possibilities offered by 13 of these channels afforded an excellent oppor- tunity for conversion to model stream systems. A cement head box and dam located about 320 m up- stream on Cheney Creek maintained a constant water head which delivered water through a 30.58 cm underground pipe to the hatchery building. A steel rack and 4.76 mm mesh screen prevented most sticks and debris from clogging the 19 20 pipe but allowed for the passage of drifting insect larvae and particulate matter. Water was supplied to each pair of channels through four 3.17 cm iron pipes; water pressure was maintained by the change in elevation between channels and head box which created an 2.44 m head. The flow rate through each pair of channels remained constant throughout 1972 and 1973 at 200 l/min. Water supplied to each channel passed through the length of the model stream only once and was collected into common floor drains and discharged through a 25.40cm pipe back into Cheney Creek downstream from the hatchery building. The standing volume of a single stream, including both riffle and pool sections, was approximately 550 liters. At a flow rate of 100 liters per minute per channel, the time for total water renewal rate was about 5.5 minutes. The chemical characteristics of the water in the model streams are summarized in Table 2. Replicate water samples were taken during summer, fall and winter periods, preserved with HgCl and returned to the Water Chemistry Laboratory on campus for chemical analysis. The data show that the water quality remained relatively constant through- out the year with an average alkalinity of 179 mg/l CaCO3, pH of 7.79, and average hardness of 211 mg/l. Water tem- perature reflected the spring origin of the water, never EXCeeding 130C during summer months and dropping to 50C in early December. 21 TABLE 2.--Water quality summary for model streams June- December, 1973. ' Characteristic 37mgfifie EIlSeptember Igngzzember Alkalinity (mg/l CaCO3) 179 182 176 Ammonia (mg/l N) 0.07 --— -—- Calcium (mg/l CaCO3) 144 152 160 Carbon, Total (mg/l C) 38 38 42 Carbon, Organic 1 0 3 Chloride (mg/l Cl) 4.0 2.4 2.2 Specific conductance (micromhos at 25°C) 430 360 400 Hardness (mg/l CaCO3) 214 214 204 VIron (mg/l Fe) 0.1 0.2 0.2 Nitrate (mg/l N) 1.05 1.32 1.57 Nitrite (ug/l N) 4.0 4.0 5.0 pH 7.92 7.45 8.0 Phosphorus, Total (mg/l P) 0.01 0.01 0.01 Sodium (mg/l Na) 2.0 3.0 2.3 Solids, Total (mg/l) 240 226 227 Sulfate (mg/1 $04) 14 16 15 I I 22 Six hatchery channels were utilized for the model stream communities providing for three complete systems each consisting of one control channel and an adjacent experi- mental channel. The upstream channels were designated pool communities and maintained at a25.40 cm water depth. Coloni- zation of these pool areas was begun in the fall of 1972 through input of organic matter and drifting insect larvae entering through the head box on Cheney Creek. Lighting was supplied to the model streams in late April, 1973, in the form of 24 banks of four high intensity cool-white fluorescent lamps. The lamps were attached to sheets of .95 plywood painted white and supported by a superstructure of structural steel. The plywood sheets were attached by chains allowing for raising or lowering the lamps to adjust the surfact intensity of light. All lamps were placed to provide an intensity of 900 i 50 ft-ca. at the water surface. This intensity was selected to allow for the colonization and growth of filamentous green algae on the substrate of the riffle communities. The illumination intensity was measured with a Weston Model 756 Sunlight Illumination Meter. All lights were wired through a relay box and electrical timer; the photoperiod was set to conform to natural day length and adjusted for seasonal variations. In early May, natural stream substrate with associa- ted flora and fauna was removed from Buckhorn Creek, a 23 similar spring-fed stream .40 km north of the hatchery andtrucked to the model streams. The substrate, consisting of gravel and small rubble up to 10 cm diameter was added to each riffle section providing for a depth of 15.24 to 20.32 cm of substrate. Care was taken to insure a uniform distribution of particle sizes between streams. An eight week period was allotted for stabilization and growth of these riffle communities before experimentation was initia- ted in late June. By that time the streams had developed rich populations of diatoms and filamentous greens. Insect activity was evidenced by feeding of the species present and casebuilding behavior of the numerous species of Trichoptera. A complete listing of the flora and fauna that developed within the model system is presented in Table 3. Determination of Treatment Concentration A series of replicated toxicity tests were conducted with 7.62 to 10.16 cm rainbow trout obtained from hatchery stock at the Baldwin Rearing Station and larval silver 1am- prey, Ichthyomyzon unicuspis, collected by electroshocking mud banks of the South Branch of the Tobacco River at Clare, Mighigan. All specimens were brought to the lab and held in hatchery channels. The standard methods for streamside toxicity tests of Howell and Marquette (1962) were used to develop toxicity data for lampreys and to establish the level of protection for rainbow trout. Static toxicity 24 TABLE 3.--Taxonomic listing of flora and fauna of indoor model streams, Paris, Michigan 1973-1974. baacroinvertebrates Filamentous Green Algae Diatoms Gammarus pseudolimnaeus Asellus militaris Plecgptera Paracapnia sp. Amphinemura varshava Leuctra tenuis Phasganophora sp. Acroneuria lycorias Ephemeroptera Ephemerella lata Paraleptophlebia mollis Baetis sp. EpichOptera Hydropsyche sp. Cheumatopsghe sp . Goera calcarata Glossosoma sp. Neophylax sp. Limnephilus sp. Brachycentrus americanus Chimarra obscura Ochrotrichia sp. Diptera Simulium sp. Antocha sp. Chrysops sp. Prodiamesa sp. Diamesa sp. Cricotopus sp. Eukiefferiella sp. Tanytarsus sp. Chironomus sp. fli§pellaneous Sialis sp. Stenelmis sp. Hydracarina sp. Physa sp. Lymnaea sp. Annelida \ Stigeoclonium tenue Spirogyra sp. Cladophgra sp. Rhizoclonium sp. Tabellaria sp. Cocconeis sp. Diatoma sp. Achnanthes sp. Nitzschia sp. Meridion sp. Fragilaria sp. Synedra sp. Navicula sp. Eunotia sp. Pinnularia sp. Gyrosigma sp. 25 tests were conducted in 25 liter glass aquaria placed in the hatchery channels. Ten liters of water were added and the proper concentration of field grade TFM was added. The ma- 'terial used was formulated in March, 1973, was labelled as 36% active ingredient and had a specific gravity of 1.212. Two animals of each species were tested in each aquarium ‘with a single replicate for each test concentration. Water temperature was 580C and pH 8.0. The entire test was repli— cated once and the results are presented in Table 4. Analy- sis of these data show a 10 hour Maximum Allowable Concen- tration (MAC ) for rainbow trout to lie between 11.0 and 25 13.0 mg/l and a Minimum Lethal Concentration (MLC for 100) larval lamprey to lie between 3.0 and 3.5 mg/l. On the basis of these calculations, Cheney Creek water would be treated at 9.0 mg/l for lamprey control in actual field con- ditions. This concentration was determined as the level to be used in all tests and experiments to follow. Model Stream Treatments Stream treatment equipment was borrowed from the Bur. of Sport Fisheries and Wildlife Field Station at Ludington, Michigan. A 1:10 dilution of 1973 Field Grade TFM was made With water and pumped into the experimental channel of the IModel streams with an automotive fuel pump powered by a 12 \Rblt battery at a flow rate of 28 ml/minute. Concentrations VWEre continuously monitored by comparison with a standard 26 TABLE 4.--Time-to-death of larval lampreys, Ichthyomyzon unicuspis, and rainbow trout, Salmo gairdnerii, tested in several concentrations of field grade TFM. Figures represent total number of mortali- ties from two replicated tests during each time interval, a total of 8 individuals tested per concentration. Concentration Elggsed Larval Lamprey Rainbow Trout (mg/l) Hours 2 4 6 10 20 2 4 6 10 20 3.0 O 0 O 2 2 0 O 0 0 O 3.5 O 0 0 8 O O 0 0 0 5.5 l 4 8 O O 0 0 0 7.5 0 8 0 O 0 0 0 9.0 2 8 O 0 0 0 4 11.0 0 8 0 0 0 2 8 13.0 0 8 O 0 0 8 27 curve develoPed for the Klett-Summerson Colorimeter based on the yellow color of TFM in an alkaline solution (Smith et. al., 1960). The experimental channel of model stream .A.was treated on September 5 at 9.0 mg/l for a period of 14 Ihours. The experimental channel of stream B was treated on November 1 at the same level also for 14 hours. Stream system C was reserved for a recirculating dose of l4C-TFM to examine residue dynamics in various plant and animal com- ponents. Several experiments were conducted simultaneously in these model systems to examine and describe the impact of the toxicant on resident flora and fauna; these will be described in detail in the sections to follow. Photosynthesis-Respiration Chamber Rates of photosynthesis and respiration by intact benthic communities were monitored by the use of a modified photosynthesis-respiration (P/R) chamber as described by Cummins and King (1972 personal communication). The cham- ber consists of a circular plexiglass tray placed in the hatchery channels prior to colonization and addition of Substrate and a tight-fitting circular top section which fits down around the buried bottom section sealing off the enclosed community (Figure 2). The bottoms were constructed Of a 5.08 cm section of .63 cmt plexiglass tubing of 27.94 Culinternal diameter bonded to a flat 30.48 cm square of ~63 cmt plexiglass which served as the bottom of the chamber. 28 Figure 2.--Diagrammatic representation of photosynthesis- respiration (P/R) chamber. 29 =11 l I I I I I l l I I I 33> 66.6583 ConEnno Em:0bcuo_2 8.3.5 at...» :. 33> sch \N A n \\\ \ U”/ \“\x\\ /////// 3 // S E i O :_ E/ \S 64 be / ”///II.H.\\\\\“\ 30 Thirty-six of these bottoms were placed in the control and experimental channels of Streams A, B, and C prior to coloni- zation and addition of substrate. Top sections were simi- larly constructed fo a 7.62 cm section of tubing of 29.21 cm internal diameter also bonded to a 30.48 cm square of plexiglass sheet serving as the lid of the chamber. The lid was provided with a centrally located 3.81 cm port to accommodate a Yellow Springs Instruments Model 5420A Self- stirring B.O.D. Oxygen/Temperature Probe. The probe was connected to a YSI Model 54 Oxygen Meter designed to read directly in parts per million and compensated for tempera- ture effects on both membrane permeability and oxygen solu- bility in water. A continuous record of oxygen concentration was provided by a Sargent-Welch Model SR Strip Chart Re- corder. Prior to an experiment, a pair of 1.27 x 17.78 cm rubber bands were stretched around the outside of a chamber bottom and covered with a light application of silicone stopcock grease. A chamber top was fitted down over the outside of the bottom counterpart and a water-tight seal was insured. Care was taken to evacuate all air bubbles through the center part prior to insertion and sealing of the oxygen sensor. The height, and consequently the volume, of the chamber was adjusted to equal the water depth of the stream outside the chamber by raising or lowering the top section. This was done to insure that the temperature 31 inside the chamber would remain equal to the ambient stream temperaturethereby'avoiding the effects of increased tempera- ture on the metabolic rates of the enclosed organisms during the experiment. The chamber volumes varied from 3500 ml in the riffles to 4500 ml in the pools. The stirring rod of the BOD probe provided a continuous circular current within the confines of the P—R chamber to simulate natural stream flow. An experiment with a single chamber was typically of 5 1/2 to 6 days duration to allow for two days of pre- treatment data, one complete day under exposure to 9.0 mg/l TFM, and two days of post-exposure data. Each chamber was opened, allowed to rinse with stream water and refilled at least every 24 hours to eliminate metabolic effects due to build-up of concentrations of ammonia or other waste prod- ucts from the enclosed community. Several experiments particularly in the strongly heterotrophic pool communities required rinsing and refilling at intervals of 5 to 6 hours since oxygen concentrations would drop below 5 mg/l. Also during the summer months of peak standing biomass of 923g; ophora and Stigeoclonium tenue in the riffle communities, supersaturation with oxygen and the formation of oxygen bubbles in the chamber during photosynthesis required more frequent rinsing and refilling of the chamber. Measurements of net photosynthesis were made during the daylight hours under constant 900 ft-ca. illumination 32 and measurements of community respiration were made on the same community after lights out (sunset). Two complete systems were run simultaneously, one in the autotrophic riffle communities and one in the heterotrophic pools. On the morning of the third day, after rinsing and refilling with stream water, each chamber was injected via a syringe through the rubber stopper with a pre-calculated dose of field-grade TFM to bring the concentration inside the cham- ber to 9.0 mg/l. The exposure continued for 24 hours to assess the effects of TFM on the entire daily photosynthetic and respiration rates. The following two mornings the cham- ber was rinsed and refilled with clear stream water for a two-day measurement of post-exposure metabolic rates. All rates were then calculated and compared from the calibrated strip-chart recorder paper. Periphyton Sampling The species composition of the periphyton community was sampled by scrubbing the rubble in a P-R chamber and rinsing into 1 liter of water. Four 20 ml subsamples were taken, two for determination of diatom species composition and two for determination of biomass as ash-free dry weight of organic material. The diatoms were cleaned and mounted following a modification of Van der Werff (1955): The sample was placed into a 100 ml Berzelus tall form beaker and 25 m1 of 30% hydrogen peroxide was added. This was allowed to 33 oxidize for 1 hour prior to the addition of l microspatual of potassium dichromate. After termination of the reaction, the beaker was filled with distilled water and allowed to settle overnight. The liquid was decanted, rinsed once more and allowed to settle for about 12 hours. The liquid was decanted to a 5 m1 volume and the entire sample was stored in a 20 ml scintillation vial. A 0.5 ml subsample was placed on a heated coverslip, allowed to dry and inverted over a warmed slide with a drop of Hyrax mounting medium. Generic composition was determined by counting individuals in three transverse cross-sections of the slide under oil. Variation in early counts was assumed to be due to varia- bility in subsampling a 20 ml volume from the original 1 liter sample, so a modified procedure was employed. A periphyton sampler was developed by Cummins (1972 personal communication) which consistently removes a 1 cm2 area of substrate through the action of a nylon spatula. The sample is forced up through a length of tygon tubing and collected in a 60 m1 centrifuge tube by blowing a column of water through the apparatus with the mouth (Figure 3). This sampler was employed in all three model streams from August to December to characterize the generic composition of the diatom communities by scraping and combining three 2 5 cm samples from each channel on a biweekly basis. The entire sample was then passed through the procedure outlined 34 Figure 3.--Cross-sectional diagram of stream periphyton sampler. 35 Sample Collecting Tu be V/ I _— \\\\\\\\\‘\\\\\\\\\X\\\\\\\}// Metal Drive /Shaft ‘\\\ \ Water Flow ii Rubber Washer %X\\\\\\\\\\\\_L\\\\\\\\ b Y Rotating Teflon Spatula 36 above and it appeared that this technique yielded more con- sistently reproducible results. Macroinvertebrates The structure of the benthic macroinvertebrate com- munity was determined from sample cores removed from both pool and riffle areas. 11 45.72 cm length of 10.16 cm di- ameter plastic pipe was fitted with a rubber gasket on the bottom lip. This was then pushed down through the substrate to form a water-tight seal with the flat cement bottom of the model stream channels. The rubble and gravel substrate with associated flora and fauna was removed from the center of the core and the remaining water and fine matter was siphoned into a 0.5 mm mesh Nytex bag. The sample was rinsed of fine silt, filamentous green algae and all insects were picked live. All samples were preserved in 10% formalin and the substrate returned to the sample area. Samples were similarly removed from the pool areas employing a 25.40 cm section of 5.08 cm diameter thin gauge metal pipe. By placing the hand over the tOp of the corer, the core was held in the tube by vacuum pressure and was transferred to a 0.5 mm mesh Nytex bag and treated as the riffle samples. A graph was constructed of number species vs. number of samples and the inflection point of increase in new spe- cies generally occurred at three samples (Figure 4). Based 37 (a) O A. [0 0| .6 .5 N 9 9 ° Cumulative No. of Species 0| 1 2 3 4 5 6 No. of Samples Figure 4.--Cumulative number of taxa obtained with increasing number of samples demonstrating inflection point. 38 on this data, three samples were selected as the number re- moved from a pool or riffle area at each sampling interval. All samples were identified as near to species as possible and recorded on individual data sheets. A Digital PDP 1140 computer program was written to calculate the Shannon-Weiner diversity index, as described by Wilhm and Dorris (1968), employing sample data to estimate the actual population diversity. The formula is: s d : _i:l i: 1092 2; where N = the total number of individuals; ni = the number of individuals per taxa; s = the number of taxa; and d = an estimate of population diversity. The values of this index range from 0 to + infinity, however, they rarely exceed nine and usually fall between three and four (Wilhm, 1970). The three samples taken during each sampling period should yield an accurate estimate of the true population diversity (Wilhm and Dorris, 1968; Warren, 1971). This particular index was employed for the following general advantages: It is completely independent of sample size and accounts for relative abundance therefore is less affected by the absence of rare species which may be missed in sampling. It has no dimensions and is therefore independent of samp- ling units, i.e., biomass or numbers may be used to obtain similar values. 39 The index of species equitability, E, was calculated after MacArthur (1965). If all species are equally abundant, d = 10925 and s = 2d. The ratio 2 d/s, is‘a measure of Spe- cies equitability or the relative evenness of distribution. Values range from a minimum of 0 to a maximum of 1. After identification and.counting of samples, esti- mates of biomass as ash-free dry weight of organic material were obtained on the total macroinvertebrate community and the filamentous green algae removed from the riffle areas with the 10.16 cm sample corer for Stream A. Samples were dried in a ventilated oven at 500C for 48 hrs. prior to ash- ing in a muffle furnace at 5000C for 1 hour. The difference between pre- and post-muffle furnace weights represents biomass of organic material with no correction made for loss of water of hydration. Estimates of production of two species of macroin- vertebrates, a stonefly, Amphinemura varsharva and the amphi— pod, Gammarus pseudolimnaeus, were made using the Hamilton (1969) modification of the Hynes and Coleman (1968) techni- que. This method was used for its rapid deployment and relative accuracy (Waters and Crawford, 1973). .It avoids the drawback of all other methods for production estima- tion, that of having to assign each individual of a particu- lar population to a specific age class. 40 Fish Production Two experiments were conducted to examine the effects of a 24 hr. exposure to 9.0 mg/l TFM on the growth rates and production of young-of-the-year brown trout, Salmo trutta. Sixty fish ranging in size from 60 to 100 mm were electro- shocked from Buckhorn Creek in Paris, Michigan, on October 5 and held in a hatchery trough. Individuals were anesthetized with MS-222 and each fish was differentially fin clipped and total length, fork length and weight were recorded. Twelve fish were added to each control and experimental pool areas of Streams B and C 5 days prior to treatment. Growth was then determined from control and experimental pool fish in Streams B and C after 43 and 35 days respectively. Fish were collected from each channel by dip-net, identified by fin clips, measured and weighed. Kinetics of TFM Residues Several experiments were conducted to examine the dynamics of uptake and elimination of TFM in 20 plant and animal components of the model streams using uniformly ring- labelled TFM obtained through the Fish Control Laboratory at LaCrosse, Wisconsin, and synthesized by Mallinckrodt Chemical Co. of St. Louis, Mo. It was supplied in sealed glass ampules containing 0.125 mCi with specific activity (If 3.66 mCi/mM, dissolved in benzene and stored under nitro- gfin. For all experiments, the benzene was evaporated under 41 a stream of nitrogen, the l4C-TFM resuspended in acetone and an isotope dilution was made with 95% pure analytical grade TFM (Aldrich Chemical Co.). The total concentration of labelled and unlabelled TFM in the treatment standard was then 21.6 mg/ml acetone. Dilution by this method yields an approximate activity in the water at time zero of 15.0 x 106 to 20.0 x 106 cpm per liter. Due to the hazards encountered in releasing large amounts of radioactive materials into the aquatic environ- ment, all experiments concerning exposure and uptake of 14C-TFM were conducted in completely closed, recirculating systems. The buried P—R chamber bottoms enclosing unexposed communities in both pool and riffle areas served as enclos- ures for the l4C-TFM experiments. The community was enclosed by placing a 10.16 cm section of 30.48 cm ID. tubing over the chamber bottom and sealing with rubber bands similarly to the respirometry experiments. The chamber had no cover to permit access to sampling of enclosed plants and animals and the height of the top was adjusted to protrude about 1.27 cm above the water surface to insure no exchange of water. Current was simulated inside the chamber through the use of a variable speed laboratory stirring motor. The depth of water was measured, the volume calculated (between 3 and 4.5 liters) and treated with the above isotOpe dilu- 14 tion of C-TFM to bring the total concentration of TFM to 9.0 mg/l. Samples of enclosed flora and fauna were then 42 removed at approximately 1, 3, 5, 8, 14, 20, and 24 hour intervals with forceps thoroughly rinsed in clear water, placed in labelled vials and immediately frozen. Five ml samples of water were also removed at the same time inter- vals and immediately frozen. At the termination of the 24-hour exposure period, the 14 C-labelled water was siphoned into a carboy, the cham- ber top gently lifted, and the community was exposed to the TFM-free water of the model stream channel again. Continual sampling of this community then developed data on retention times and elimination rates of TFM for the various compon- ents of interest. This treatment procedure was followed on a bi-weekly basis from July to December, each time employ— ing a new, previously unexposed community. Occasionally, if the species of interest was not present within the commu- nity in sufficient numbers to permit sampling without deple- tion, that species was stocked into the chamber just prior to the treatment. This was particularly true in the case of the aquatic macrophytes which did not naturally occur in the model streams. Six to eight components were sampled during a particular experiment and three to five replicates of each of the 20 components were tested during the year. Samples were kept frozen generally for a one week period prior to analysis. All plant, animal and sediment samples were dried at 500C for 48 hrs., weight adjusted to 43 100-150 mg and combusted to 14CO and water in a semi- 2 automated Nuclear Chicago combustion apparatus. The 14CO2 was taken up in 10 ml of monoethanolamine-methyl cellosolve, 1:2 (V/V). A 2 ml portion of this was radioassayed with a dual channel Nuclear Chicago Unilux I (Model 6850) liquid scintillation spectrometer in 15 ml of a toluene-methyl cellosolve, 2:1 (V/V), fluor mixture. A 2 m1 aliquot of each water sample was radioassayed in 15 ml of a toluene- Triton x-100 fluor, 2:1, (V/V). Periodic efficiency curves were established for the instrument employing a series of internally quenched stand- ards and all sample counts were converted to actual distin- tegrations per minute and uCi values employing the channels ratio and efficiency curve. The entire experimental channel of model stream C was treated with a recirculating dose of l4C-TFM on Novem- ber 13. Water flow through the channel was blocked by in- serting rubber stoppers in the spouts and drains. Recircu- lation of water was provided by placing a large diaphragm pump in the downstream end of the riffle area and discharg- ing into a 7.92 m length of 5.08 cm plastic pipe which ran back up to the beginning of the pool area. The pump volume was sufficient to pump slightly more than the 100 l/minute normal stream flow rate and air was constantly sucked into the pipe providing for reaeration during the day. The water temperature rose 30C during the treatment and was 44 considered nominal. Stream volume was calculated to be 500 to 525 liters and was treated with an isotope dilution of 1.0 mCi of 14C-TFM and 4.500 gms of analytical grade TFM all in 10 ml acetone. This dilution gave approXimately _1liter-l at time zero and an actual 2.6 x 106 counts min. concentration of 8.68 mg/l labelled and unlabelled TFM. Following the 24 hour exposure period, the discharge from the pump was directed into a 984 liter metal container and water flow was immediately begun in the channel. In this way discharge of radiolabelled water from the channel was minimized. Sampling of plant and animal components for de- termination of uptake and elimination rates was done similarly to the recirculating chamber experiments. The recovery of l4C-TFM from the large volumes of water was facilitated through the use of a non-ionic poly- meric adsorbent (Lech, 1971). The water was acidified to pH 4.0 with 0.1 N HCl and passed through a 7.62 x 60.96 cm column of Amberlite XAD-Z resin. The water effluent from the column was periodically radioassayed and no activity was observed. The column was washed with two volumes of distilled water and the 14C-TFM was eluted with one volume of methanol. The methanol was removed in a rotary evapora- tor and the compound was resuspended in acetone. RESULTS Community Metabolism An example of the calculations for community metabo- lism data during a 24 hour period are given in Table 5. The values for dissolved oxygen were taken directly from the con- tinuous record of oxygen concentrations provided by the strip chart recorder. Negative values for oxygen production frequently were obtained in the strongly heterotrophic pool communities. For example, an oxygen concentration of 9.8 mg/l at sunrise may drOp to 6.4 mg/l during the course of the day under full illumination resulting in a negative value for gross primary production. The rate of loss was calculated and compared with the rate of oxygen depletion for that same community during night hours. If this night- time rate was greater than the daytime rate of loss, as was generally the case during the summer months, the difference between these day and night rates was interpreted to be due to the oxygen production by diatoms and filamentous green algae on the surface of the pool community and was recorded as daytime gross production. However, during the exposure to 9.0 mg/l TFM, there was generally no detectable differ- ence between rates of oxygen depletion in the pools during 45 46 TABLE 5.--Sample calculations for a highly productive riffle community in Stream B on August 21-22. (After Whitworth and Lane, 1965.) Dissolved Time of Day Oxygen (mg/1) Time Intervals Sunset (Day 1) 16.0 11 hours of dark Sunrise 10.8 13 hours of light Sunset (Day 2) 15.4 Calculation of community respiration: _ sunset (1) - sunrise Resp1rat1on/hr. — time interval (hrs.) 1 16.0 - 10.8 = .433 mgO liter-1hr- 12 2 Gross community respiration .433 x 24 = 10.39 mgOzliter—lday-l Calculation of community oxygen production: Daytime oxygen increase = sunset (2) - sunrise 15.4 - 7.0 8.4 mg/liter Daytime respiration - respiration/hr. x day length = .433 x 13 hrs. = 5.63 mg/liter Gross primary production oxygen increase + daytime respiration = 8.4 + 5.63 = 14.03 mgozliter’lday'l 47 day-time and night. In these instances the gross produc- tion was recorded as zero and the rate of oxygen depletion was recorded entirely as respiration. In these instances the P/R ratio was also recorded as zero. The summary calculations of primary production, com- munity respiration and P/R ratios for a 5 day experiment with a particular chamber located in Stream B riffle com- munity are shown in Table 6, and presented graphically in Figures 5 and 6. These figures for late August are among the highest values for gross primary production achieved in the model streams during the year. The high production values of .950 gm 02 111'.2 day-1 are probably due to the large standing crop of filamentous green algae, primarily Cladophora sp. and Stigeoclonium sp. that were obvious dur- ing that time of year. The respiration values for the riffles remained relatively constant throughout the year at about .630 gm 0 m"2 day-1. 2 The characteristic influence of the 9.0 mg/l expos- ure to TFM is represented in the respirometry data for August 23. Under these conditions the rate of gross primary production has dropped to a low of 0.815 gm 0 m.-2 dayfl and 2 the respiration rate of the enclosed community shows an in- 2 l crease to .640 gm 0 m7 day- . The depressed evolution of 2 oxygen by the algal community reflects the algi-static ef- fect of TFM as demonstrated in laboratory toxicity tests 48 TABLE 6.-—Primary production, community respiration and P/R ratios for Stream B, August 21 through 24. Communities exposed to 9.0 mg/l TFM on August 23. 252:3: its??? site"s-1, .4. Stream B Riffle Community 21 Aug .939 .667 1.41 22 Aug .961 .635 1.51 23 Aug (exposed) .815 .640 1.27 24 Aug 1.070 .625 1.71 Stream B Pool Community 21 Aug .135 .560 0.24 22 Aug .139 .547 0.25 23 Aug (exposed) 0.0 .779 0.0 24 Aug .108 .651 0.17 49 .vNIHN umsmsé wamwflu Hmucmeflnmmxm m Emmuum ca EmHHonmumE huchEEOUII.m musmfim m=< «a mm «a 92 a cozmtfimom — ON LEV I 1 0* AC/ camon . Decca “co—:39... .E\«E\ 0 9: Do core—69a EmEta $20 on cow 50 .wmlam umsmsm mcflusp Hoom Hmucmfiflummxm m Emmuum CH Emflaonmuwe >uac:EEounl.m magmam 92 Va nu actuated >._mE_..n_ $90 Bacon. / «N 9.4. PN Tr comma 20859.... cozmtfimom >:c:EEoo 1 0.. .. £~§~o me 00 L L O? 51 (Maki et. al., 1974) and the increased respiration rates are believed to be primarily due to the response of the macroin- vertebrates to the sublethal level of TFM. The increased respiration rate of invertebrates is recognized as a meta- bolic response to stress agents (Whitley and Sikora, 1970). Rates of gross primary production, community respira- tion and P/R values obtained throughout the year are listed in the Appendix and presented graphically in Figures 7 and 8. The values for gross primary production in the riffles ranged from approximately 58.2 mg 0 hr-lm-2 to 79.0 mg 0 2 2 hrmlm-2 during the course of the experiments. Highest rates were recorded during the late summer and late fall periods which presumably were brought about by the increase in standing crop biomass of filamentous greens and attached diatoms respectively. The lowest production rates were recorded between these-periods during the time when the com- munities were in a state of change, i.e., just developing green algae or greens sloughing off and diatoms becoming more obvious. The effects of the 9.0 mg/l exposure of TFM on gross production rates also shown in Figure 7. For each experi- ment throughout the year, the rate of oxygen evolution was decreased. The amount of this departure presumably reflects the Species composition of the primary producers. The gross production rate was depressed by about 5-10% during summer 52 Figure 7.--Rates of gross primary production recorded throughout 1973 under an illumination of 900 ft.-ca. 53 .2. >02 ~00 “new m3< c2. “co-Snot. mat-5 . .rou Econ 11:. «co—5mg... « loo mots—E 54 70' Riffle Community Respiration Treated a“) ——' “~\\‘\\//////// / Controlsfl/ N 1 5. O a: Eso Pool Community ReSpiration 6 Treated 20 Controls 1 Jun Jul Aug Sept Cot Nov Figure 8.--Rates of community respiration observed in riffle and pool areas throughout 1973. 55 months when the producer community was dominated by filamen- tous greens, however, in the later fall months when the producer community was composed primarily of attached dia- toms, the depression of oxygen production averaged 25-50%. In laboratory toxicity tests the diatoms were found to be more susceptible to TFM (EC 50:1.5-3.0 mg/l) than the green algae (EC 50 5.0-7.9 mg/l) (Maki et. al., 1974). The rates of gross primary production of the fall diatoms communities were obviously much more influenced by the TFM exposure than were the production rates of the summer filamentous green communities. The rates of gross primary production in the pretreatment controls were compared with the during treatment rates and found to be significantly different at the 0.01 level (Table 7). The strongly heterotrophic nature of the pool com- munities throughout the year are shown in Figure 7. Values for gross primary production averaged 10.7 mgO hr-lm-2 to 2 22.5 mg02hrmlm-2 during the summer months and were primarily due to a low density of filamentous greens covering the sur- face of the organic detritus in the pools. As in the riffle communities, these populations disappeared from the model streams in early fall and positive values for oxygen produc- tion in the pools were not detected beyond this date. The response of these communities to the 9.0 mg/l exposure to TFM was also similar, showing a depression of oxygen evolu- tion in all experiments. The depressed rate compared with 56 .H. mm.mH mm.ma EsacoHoommHum .Hv mm.~o oo.om muonaocmau .Hv ¢¢.ma HH.HN muwmouflmm “wocmuudouo mo mocmswmum ucmouwm m.ov mm.na Hm.mH A.03 amp mmum cmmv mmmaoflm Hmuoe .Hv moa x mm.a moa x mm.a suamcmo soumao m Emwuum .H. moa x NH.N mos x nm.a shamans soumao a smmuum N.o. mm.o ucwEucouplumom Ho.ov mH.o mm.o omomm m\m Hoom CoH.V mMoH UCQEVNCmUHJIHWOAH Fo.ov mn.o mm.a oaumm m\n onme o.Hv oo.om ucmEummuumumom m.ov m¢.om mm.mm Amie any: omev coflumnflmmmm Hoom o.Hv av.mm ucmEummuumumom ~.ov ve.sm mo.mm Amus any: cost coHumuHmmwm mammam o.Hv vw.HN ummfiummHUlumom No.ov vm.aa mv.mm Amie In: omev cofiuoso loam umEflum mmouo Hoom o.Hv ov.ao N bcwsumouulumom Ho.ov mm.mv no.6m Axle Hun: ooEV cofluojooum n mumfiaum mmouu mmwmflm zms H\ms o.m Hepucoo Amy mUCMUAMHcmflm ou owmomxm ucmEummuBnmum xuummonm muflcsesou mo Hm>mq mcmoz .mquU cmummuu cam Hon» Icoo Eoum dump mono mcflpcmum paw Emflaonmumfi huficsEEou mo mosam> :mmE Hmsccm com3pmn mocmummwflo unmofiMHcmwm mo ummuuu .mucmosum mo muasmmmll.n mamde 57 the pre-treatment control data was found to be highly sig- nificant at the 0.02 level (Table 7). The temporary nature of the effects of TFM was dem- onstrated by the post-treatment experiments. Values for gross primary production in both the riffle and pool commu- nities returned to pre-treatment rates within 24-36 hours post-exposure. No significant difference was found between these pre- and post-treatment rates (Table 7). The influence of the TFM exposure on pool and riffle community respiration values as measured by the P/R chamber is shown in Figure 8. The pre-treatment respiration values for the riffle communities ranged from 20.3 to 32.4 mg02hr-1 m-2 and increased to 22.0 to 60.7 mgogzhr-lm-2 under the ex- posure to 9.0 mg/l TFM. This represents an increase of 3 to 50% due to the toxicant. Similar results were obtained in the pool communities where the high density of burrowing midge larvae was primarily responsible for the increased respiration values obtained in the presence of TFM. Normal values for respiration in the pools ranged from 10.5 to 36.2 mg02hr-lm-2 and were increased to 15.5 to 73.5 mgOZhr-lm-2 by the exposure to TFM. Respiration results were generally more variable between the sampling dates presumably due to emergence of species during the year and unequal distribu- tion of organisms between the chambers. The difference in mean riffle and pool respiration rates were significant at 58 only the 0.2 level (Table 7). Post-treatment experiments indicated a rapid recovery and return to pre-treatment respiration rates within 24—36 hours. No statistical dif- ference was found between pre—treatment and post-treatment rates (Table 7). The most sensitive indicator of the effects of TFM on community metabolism was the comparison of P/R ratios obtained on a pre- and post-treatment basis with those ob- tained under exposure to the 9.0 mg/l in the riffles (Fig- ure 9). Since the action of TFM is to decrease oxygen production values and increase respiration values, the P/R ratio will naturally be subject to large changes in the presence of the toxicant. The mean ratios of pre- and post-treatment experiments varied from 1.06 to 1.78 while those ratios obtained under exposure to TFM varied from 0.5 to 1.27. In most experiments, the P/R ratio for the pool communities was reduced to zero and averaged approximately a 50% decrease in the remaining experiments. Analysis of the differences between pre-and post-treatment means and the during treatment means for both the riffle and pool com- munities proved the differences significant at the 0.01 level (Table 7). Periphyton Community Structure Densities of dominant genera of diatoms observed in samples removed from experimental and control channels of 59 .zma H\ma o.m ou musmomxm umpss posflmuno moflumu £ua3 pmummfioo mmwuassfifioo mammau How mofiymu m\m smmE usmEummHUIumom cam Imumll.m musmflm ‘ d u d «c9500.; 9:50 .md 10.9 03mm mk— lfi 5;. mcmoE “co—5mm: - amen. $-05 60 Streams A and B are represented in Figures 10 and 11. Cell counts represent the total number of frustrules observed in three horizontal transects of a slide containing an aliquot of cells from a sample area of 5cm2. Correction factors were applied to compensate for the area of slide counted, aliquot of cells and finally to normalize the cell count to a square meter basis. The data for both Streams A and B are quite similar, the general trend representing a drOp in total number of in- dividuals through the summer to a minimum density of about 9 cells/m2 in the early fall months. A period of 1.20 x 10 relatively rapid growth follows in late fall and early winter to reach a density of approximately 3.0 x 109 cells/m2 by December. There is no significant difference in the total number of individuals between control and experimental chan- nels. However, a slight decrease in the total density during the week immediately following the dose in the experimental channels may indicate a toxic effect on the relatively TFM- sensitive diatom community. As mentioned earlier, the 96 hr. LC50 for TFM in laboratory cultures of diatoms lies in the vicinity of 1.0 to 3.0 mg/l and it may be possible that the 9.0 mg/l exposure for 24 hours exerts a toxic influence on the exposed diatoms. This drop in total number of indi- viduals occurring in only the experimental channels during the week following treatment, although not statistically 61 Figure lO.--Diatom cell densities in Stream A. 62 8o. v. .626 too? name ammo ammm 62 mm as: .32.“ a.-. «costume... . /\ Jfi 32:20 .9350 .2550 .macoEtmaxm 0.0 63 Figure ll.—-Diatom cell densities in Stream B. 64 00mg mp >01 0 >01 0 ~00 Fm «Q00 or 500 m ma< mm .0520 1 .9300 .10 _. _ WA .0520 tiON E08300... _0~:0E_..0nxm 65 significant may reflect a temporary change in community structure brought about by exposure to TFM. At any rate, the effects are temporary and growth rapidly ensues to parallel or actually exceed densities found in the unex- posed control channels (Figure 1l). A t-test of the mean cell densities in control and experimental channels re- vealed no significant difference between these communities (Table 7) . The generic composition of the diatom flora of the riffle communities in Streams A and B were analyzed with the Shannon-Wiener diversity index (Wilhm and Dorris, 1968) and MacArthur's (1965) method for species equitability (Tables 8 and 9). Diversity values ranged from a low of 1.36 in the early fall to a high of 2.64 in early winter samples. Values for species equitability varied as the reciprocal of divers- ity with highest equitability values recorded in the early fall community indicating that the population existing at that time consisted of low densities of few different genera. The experimental channel of Stream A was treated with 9.0 mg/l TFM on September 4 and Stream B treated at the same level on November 1. Tables 8 and 9 demonstrate no signifi- cant difference in the diversity or equitability values cal- culated from samples removed from both control and experi- mental channels of each stream following the treatments. The standing crop of filamentous green algae removed from Stream A riffle communities with the use of the 10.16 cm 66V mm.mflhm.h 0H mvo.Hwh.o mHh. mb~.HmN.N 0H.N HH 0 v n m 000000 00 umnfisz huh. «h. mmm. was. huh. wuflaflnmuflsqm ww.~ mH.N mo.N HN.N om.N huflmumbflo amusmeflummxm d Emmuum mm.HHmm.h m NHH.Hom.o ham. mmm.Hmm.H . mo.m ca 0 m h h mumsmm mo Hmnfiaz Ham. mmm. nmm. one. man. wuHHHnmuwswm mm.~ mm.H mm.H HH.N mm.m auwmu0>flo Houucou d Em0uum .0.m chmmz Hmsmna >oz pm 000 ms ummm ma mam m~ mash hm sash as mama Houusoo fl Emmuum .mamccmsu amusmafiummxm 0:0 H0002 CH mcflumflxw mmfluwcdaaoo Eoumwp mo 0usuunuuml|.m mamda 0H.NHBH.m mmo.H¢mh. NNm.Hmm.H OH OH 0 Han. mmh. mom. vm.m mm.H hm.H 000008000mxm m E0000m 67 0H.NHHv.m moa.Hmmh. HmN.Hmm.H 0H HH 0 Hmm. mom. Hmm. mH.N mm.H mm.H HOHUCOU m fimwhfim .0.m Hicmwz H0sm0¢ >oz m 000 0m 0000 00 00000m 00 000802 000000000900 >00000>00 000000 00 000802 000000000500 >00000>HQ .ma000000 H0000EH00mx0 000 0000000 m E000um 00002 00 @00000x0 00000000000 500000 mo 00000000mll.m mamfia 68 corer was analyzed in detail to determine any differences that may be due to the lampricide treatment. The samples were homogenized by hand shaking in the sample vial, two wet mounts of an aliquot of this homogenized material were made, and 15 random fields were examined to determine rela- tive percent frequency of occurrence of the three dominant species. Table 10 gives the results for Spirogyra, Clado— phora, and Stigeoclonium from experimental and control channels of Stream A. The major trend evident from the samples is a gradual succession throughout the year from a Spirogyra dominant community in early summer to a Cladophora dominant community in the early winter. The relative fre- quency of occurrence was not significantly different between control and experimental channels (Table 7). The biomass of the filamentous green algal community is represented in Figure 12. The ash-free dry weights ex— pressed as grams organic matter/m2 are graphed for both ex- perimental and control channels of Stream A. The peak bio- mass occurred in both channels during late August with a continuous decline in biomass occurring through the December samples. Analysis of these biomass figures demonstrated no statistical difference between control and exposed popu- lations that could be caused by the lampricide (Table 7). This presumably could be predicted from the relative toxicity of TFM to green algae since most species have a 96 hr. LC50 69 o0 o0 om on 00 ON 000 m0 om ow om om o0 00 >02 m OH O om CO0 o o 000 00 00 o om om 00 o0 0000 00 om cm on oo 00 om 0000 m 000800009 0 800000 mm om oo oo m 00 0000 m om ow om om O0 00 000 mm om om ow ow ow om 0:0 0 o o om om om om 0000 00 0000000 000008000000 0000000 000008000000 0000000 000008000000 0000 8000000000000 0000mo000u 0000000mw .00000008 00>00000m 00 000008 003 8000 0000800000 00000000 0000000 000 000008000000 0 800000 00 00000 00000 00000 00000080000 000 00 00000000800 0000000 00080000mmdll.o0 mnmfla 70 Figure lZ.--Biomass as ash-free dry weight of filamentous green algae from control and experimental channels of Stream A. 71 80 3 >02 0 .oo 9 80 00 30 o 80¢ 03 00 an: 33 «00.00000... 00—. m\,\ I w .2520 n .90000 00“ NE\ Em .000000 .0000E_._00xm do» 72 of about 6-8 mg/l TFM, a 24 hr. exposure to 9.0 mg/l would be expected to have minimal effects. Macroinvertebrates There were 38 different taxa representing 14 orders of macroinvertebrates sampled from the total model stream systems. Substrate associations and current differences between riffle and pool communities played a key role in the distribution of species. The pool areas, with soft particulate organic matter and flow rate of about 0.5 ft./ sec. were characterized by a heavy density of midge larvae, annelid worms, gastropod and pelecypod molluscs. Through- out the summer months the dominant member of the pool com- munities was the midge, Tanytarsus sp. often accounting for more than 90% of numbers and total biomass. This species emerged in mid-September and temporarily left the pool com- munities seemingly depauperate of fauna, however, the next generation soon grew to sufficient size to be retained by the mesh of the sample rinsing net and Tanytarsus sp. con- tinned dominant in numbers throughout the late fall and early winter. The riffle communities supported a much greater number of taxa due to the more diversified habitat of the gravel and rubble substrate (Table 11). No single species was particularly dominant as in the pool communities, in- stead, five or six species were generally common to all 73 TABLE ll.--T0tal number of taxa collected from each stream community throughout the sample year. Summer Fall Winter Total E erimental ( Riffle 26 28 31 34 XP ( Pool 8 11 12 14 Stream A 'Control ( Riffle 27 27 30 32 ( Pool 7 10 ll 13 . ( Riffle 28 30 33 35 Experimental ( Pool 9 ll 10 14 Stream B ( Riffle 28 27 29 32 C°ntr°l ( Pool 10 11 12 12 .. ( Riffle 23 27 26 29 Experimental ( Pool* __ __ __ __ Stream C ( ff 27 29 28 30 Ri 1e Control ( Pool* __ __ __ __ *Used for fish production experiments - no samples taken. 74 three streams each sharing equal prominence among samples. The amphipod, Gammarus pesudolimnaeus and a Nemourid stone- fly, Amphinemura varsharva were common in all samples taken. Also abundant in most of the samples were the small mayfly, Baetis sp., isopods, Asellus militarus, trichopterans Glossosoma sp., Cheumatgpsyche sp., Brachycentrus sp. and diamesian and orthoclad chironomids. The variation between pool and riffle macroinverte- brate populations is clearly shown in Table 12 where the Shannon diversity index values and equitability are presented for the three streams on a seasonal basis. In all instances, the calculated values for diversity of the riffle communities exceeds the values calculated for the corresponding pool com- munities generally by 2 to 3 times. This result could be predicted from the Species composition information since the pool communities average only about 14 taxa and the riffle areas averaged about 31 different taxa present, a higher value for the diversity index should be obtained from the riffle communities. The range of diversity values for rif- fle communities varies from 2.16 to 3.49 while the range of diversity for the pool communities is from 0.48 to 2.05. A comparison of the standard deviations of the annual means demonstrates that no significant difference exists between diversity values obtained from control and experimental channels of the same stream system and indeed only slight differences exist among similar communities of all three 75 .00000 0000800 008800 02* 000.Hv N0.0 000.Hv 00.m 00.0 00.m 00.0 0N.m « 000000 0000000 0 800000 Amo.Hv 00.0 A0N.HV mm.m 00.0 00.N m0.o mm.m 0 000000 0000080000xm A0N.Hv Nmuo 000.HV mm.0 mm.0 00.0 N0.0 hmfl0 Nmuo 00.0 0000 0000000 000.HV 0m 0 A0N.Hv 00 N 00 0 MN m 00.0 00 N m0 0 00.N 000000 0 800000 A0N.HV 00.0 Amv.Hv mm.0 0m.o 00.0 Nn.o 00.0 mm.o mo.N 0000 0000080000xm A00.Hv mm.o 000.HV oo.m 00.0 mm.N mm.o mm.m 00.0 0m.m 000000 .000HH0 00.0 Amm.HV mvu0 mvflo hm.o 50.0 00.N mm.o m0.o 0000 0000000 000 My 00.0 AMN.HV mm N N0 0 om.N Nm.0 0h.N 00.0 00.N 000000 0 800000 000.“ Nm.0 Amm.HV 50.0 0m.0 Nh.0 00.0 00.0 00.0 00.0 0000 0000080H0mxm A00.HV Nm.0 0000.Hv m0.N 00.0 00.N 00.0 m0.N 50.0 00.N 000000 0 :0. 0 p m 0. 0 p .0.0 0 H 0002 000000 000003 0000 008800 .00008 000000 000 000000 0000 000000 000008800 0000 8000 000000>0000 00 0000800 00 08000 00 000000000000 000 0000000 >00000>00 0000000||.m0 mqmflUCH mo Hmnfisz ¢v.HH vo.m hm.NHmm.NH OHH.HHm>.m mh.NHbo.ma mmflommm mo Hwnfidz Hma.oH mm.o MHH.oHNm.o va.0va.o Nva.oflm¢.o hmflaflnmuflsvm omm.oH bw.a vma.owmm.m mmm.OHmv.H hMN.onm.N huflmuw>flo ucmfipmmuulumom omHhH mvaom vmmmHmmvNH wwwmflaowmm Nmmaflmmooa NE\mHMDUfl>HUCH mo Hmnfisz mo.HH m¢.v Ah.mHmm.NH mam.onm.v Nh.HHoH.ma mmflowmm MO Hwnfisz mmo.oH av.o mno.low.o Nvo.oHnm.o hmo.onm.o muflafinmuflsvm hmH.oHlo.o mam.onm.N omo.onm.o «5mm.oflmm.m muflmum>fla #CQEHMGHfllmhfl Hoom wawwflm Hoom mammwm Hmucmfiflummxm m Emmuum Houucoo m Ewmuum .mawccmno Hmucwfiflnmmxm cam How» 1:00 ¢ Emmuum Eouw mmumunwuum>cflouome owsucmn mo wasposuum >UAGDEEOUII.MH mamma 79 by the natural emergence of resident species. The magnitude of this reduction in number of individuals is greater in the pool communities than the riffles due to the extensive emer- gence of Tanytarsus sp. from these pools. Analysis of sample data on the basis of mean numbers may tend to mask any short-term effects present in the ex- perimental channel. A standard t-test of the significance of difference between pre- and post-treatment community structure of the macroinvertebrates was conducted by combin- ing all the data from samples removed up to 2 weeks post- treatment and comparing with pre-treatment levels. The reduction in fauna was found to be significant at the 0.05 level (Table 17) and more importantly, the reduction in fauna in the adjacent control channel was not significantly different for the same time period. However, this reduction in total number of individuals did not remain significant throughout the sample period. When the levels of pre-treat- ment abundance are compared with samples removed at 2 and 3 months following treatment, no difference can be shown be- tween the samples (Table 17). The reduction in total number of individuals in the pool communities was also compared between pre- and post- treatment samples employing the standard t-test. In this instance, the number of individuals lost from both control and experimental communities between pre-treatment and 2-week 80 post-treatment samples was significant at the 0.05 level (Figure 13). The post-treatment mean for the control channel was lower than that for the experimental channel indicating that the significant reduction was due to the mass emergence of Tanytarsus sp. occurring at that time. Figure 13 also represents the reduction in fauna between pre- and post- treatment samples from the riffle area of model Stream A. From the graph, one can see that the drop in total number of individuals in the experimental riffle is much sharper than in the control riffle but that recovery occurs within approximately 2 months due to input from drift and the re- cruitment of new individuals into the population. The de- gree of recovery increases the population density to above pre-treatment levels and exceeds the number of individuals in the control channel. The biomass, expressed as ash-free weight of the entire macroinvertebrate community of Stream A was deter- mined by ashing a series of samples from Stream A after the individuals had been sorted and identified. Figure 14 rep- resents the results flnmnexperimental and control channels of model Stream A and indicates that the total macroinverte- brate biomass ranged from approximately 3.0 gms/m2 in late summer to a peak of 19.0 gms/m2 in March. Results agree relatively closely with the actual numbers of individuals shown in Figure 14 with the exception that the increase in 81 : Riffles ‘57 Treatment Communities Experimental 01 E o g _ Control 2 40l 120r Experimental Pools . Communities "E o 80} 0 Control N .- \ 2 40¢ U 8/23 972 976 97161011161179 12713 3712 Figure 13.--Number of macroinvertebrate individuals per unit area of substrate in Stream A riffles and pools. 82 20' Stream A Riffles 15- Experimental "51 i Treatment \\ I «n E : 3 5 Control 2 g: 1 l . l . 1 o B 2' 1’10 Stream A Pools 0 g 3 EXperimental '5 . < 6 COHtI’Ol Treatment 4i- 2, nil A89 sép oét Nov Dec Mar Figure l4.--Standing crOp as ash-free dry weight of macroinvertebrates from Stream A. 83 number of individuals in the experimental riffle community does not result in an increase in total biomass to exceed control level. Drift.--The most striking effect occurring in the macroinvertebrate community of Stream A during lampricide treatment was the significant increase in drift pulse re- corded immediately after the initiation of treatment (Fig- ure 15). A simultaneous record of the number of individuals drifting from both experimental and control channels was obtained by covering the stream outlets with a .186m2 sec— tion of 0.5 mm nylon screen. Due to blockage by drifting particulate organic matter and insects, the nets were fished for 15 minute intervals every three hours and once every hour at the onset of treatment. A full 24 hour record of drifting individuals obtained on the day immediately prior to treatment shows both experimental and control channels have similar drift rates of approximately 1 to 5 individuals per sampling period. Curiously no significant increase in drifting individuals was recorded after sunset, the period of highest drift rates in natural streams. In samples taken immediately after the initiation of treatment, the drift numbers rise to a peak of 32 individuals per 15 minutes in the experimental channel while simultaneous samples taken in the adjacent control channel indicate no increase in drifting individuals over the previous days' estimates. The 84 Figure 15.--Simultaneous drift rates from Stream A experi- mental and control riffle channels during treat- ment with 9.0 mg/l TFM. 85 v .53 a .33 ppm” .Eaa coma .Emaaloomm .518 .83 . 5mew l} \ on? .9950 comma «contact. 2:: 5:05:83 < Eoobm ._.. f .. 835E 9 ha 500532 .13 an 86 drift rate in the experimental channel remains higher than the control channel throughout the entire treatment period. The initial sharp increase in drift rates is the result of only a few of the more TFM susceptible species and only three genera constitute approximately 80% of the total num- ber of the drift peak. These three genera are a trichopteran, Trentonius distinctus, the amphipod, g. pseudolimnaeus and a mayfly Baetis sp. The caddisfly is the single most abundant species in all drift measurements taken during treatment but was never found as a prominent component of drift in pre- treatment or control samples. Therefore, its presence and numerical abundance in the drift samples taken during treat- ment are brought about by exposure to the toxicant. Toxicity data developed for this species employing continuous flow proportional diluters indicates that the 96-hour LC50 for this species is approximately 1 to 2 mg/l TFM making this caddisfly one of the most susceptible of all invertebrate species tested. Identical procedures were employed to characterize the effects of the toxicant on drift of macroinvertebrates leaving the pool areas. Similar results were obtained for these communities indicating a significantly higher drift rate leaving the experimental pool area during treatment with TFM (Figure 16). In this instance, the drift was com- posed of one species, the amphipod E. pseudolimnaeus. Since 87 .Zme H\mE o.m sufl3 ucmfiummuu maaunp mamccmno Hoom Houucoo pom HmucoEHHmmxm fl Emmnum Scum mmumu umaup msomcmuasfiflmll.ma musmwm 85 .248 -85 4 Emma . Ba .Ea .83 5.8 .‘l d u in “ canon _ood .9580 3:0anth lo. 335E 00 too ton—:32 .fir _ooo fiEoEtooxo < Emobm as 88 the water leaving the pool sections had very little particu- late organic matter and since drift rates out of the pools were significantly lower than riffle areas, the nets were fished for a total of 60 minutes at each sampling period without problems due to clogging. Pre-treatment drift rates average 2 to 6 individuals per hour and the exposure to TFM increases this rate to approximately 12 to 19 individuals per hour. Production.--The production figures for two species, A. varsharva and g. pseudolimnaeus in the control and experi- mental channels of Stream A are given in Table 14. The standing crOp of the stonefly was significantly lower in the experimental channel than in the control channel throughout the entire sampling period. Average biomass throughout the year in the experimental channel was 2.77 gms/m2 compared to 4.13 gms/m2 in the control channel and calculated values for production varied correspondingly from 13.63 gms/m2 in the experimental area to 17.89 gms/m2 in the control. Co- hort turnover ratios or the ratio of production to standing crop were 4.91 in the experimental and 4.33 in the control indicating that a proportionately greater amount of produc- tion per unit biomass occurred in the experimental channel than in the control. Obviously, no negative or adverse ef- fect on the production of A. varsharva could be attributed 89 oawmem Hmucmeflummxm < Emmnum oHMMHm Houpcou d Emouum nv.v mm.vma mm.hm mH.v Nv.mm vu.ma Acoumznwmomv Hmuoe mv.v nm.mma mo.om ma.v mm.mm Ho.oa ucmEumwuunumom Hm.v mo.mw v©.mH om.v Ha.vma mo.mm ucwEummuuloum mswmcEHHOpsumm maumEEmw Hm.v mo.ma hh.m mm.v mm.na ma.v Acoumzlummmv Hmuoe m©.v NH.®H mv.m mv.m mh.ma mv.v ucmfiummuulumom mm.m mm.a No.0 on.m mn.m mm.a ucmEumouulmum m>umnmum> MHDEmcfiszd 3%” .Ngmpmfiw swim”. Emu. .N...........Ww.m..... “Magnum”. uuonoo . . uuozoo . . . .¢ Emouum mo mamccmno HmucmEfiHmme pom Houucoo Scum moumun Iwuuo>cflouoms mo mwfloomm OBu Mom moflumu uo>ocusu unocoo new coauospoum .mouo mcflpcoum mo COmAHmmSoo <||.va mqmde 90 to the TFM treatment since growth and production in the ex- posed community exceeded that recorded for the unexposed control. Similar results were also obtained for the amphipod, g. pseudolimnaeus. Production values of 124.5 gm/m2 and 82.42 gm/m2 were recorded for standing crops of 27.8 mg/m2 and 19.74 gm/m2 in experimental and control communities respectively. The calculated cohort turnover ratios were then 4.47 for the experimental channel and 4.18 for the control indicating a greater amount of production per unit biomass occurred in the experimental channel. Therefore, the exposure to TFM has no negative effect on production of g. pseudolimnaeus and indeed, may have a stimulatory effect causing increased growth. Stream B Community Structure.--A replicated treatment employ- ing the experimental and control channels of Stream B was carried out on November 1. The mean values of the various measurements of community structure are represented in Table 15. The pre-treatment numbers represent means of samples taken from 24 August to 1 November and the post-treatment samples cover the period from treatment to 12 March. The mean values were not significantly different in the experi- mental community treated with TFM. In all communities the mean number of individuals per square meter increased in 91 .cmmE may no Houum Unmpsmum« mmmmfimmvma vmmmHthoH 5mm.ova.m cmv.HHmm.MH Hom.oHHm.o HHH.onm.o ¢m¢.onm.H mna.ono.m mmamflvmmwa momNHothH mov.loH.m vw.HHHm.mH mHH.oH¢h.o hoa.oflmm.o wha.onm.H Hmm.oflmo.m hoovflmmoam HNNNHmNHHH mm.HHmm.m mm.NH>h.ma MAN.OHmm.o who.onm.o mnm.onm.a vmm.onm.N Havmflmmmma mamaflomooa Hhv.o mmo.o vmm.o Hmm.¢ omH.HHmm.NH Hmw.o vmo.ova.o me.a «th.OHbm.N ~e\.ecH mo nonesz mwflommm mo Hwnfisz sonannmonswm >uamum>flo ucmfiummuulumom ms\.ch mo monasz mmfiommm mo Honfidz sufiannmuflswm >uwmum>fio ucmfiumouplmum Hoom mammam Hmucmfiwummxm m Emmuum doom mammwm Houucou m Emmuum .mamccmno HmuamEflHomxw Ucm Honucoo m Emmuum ma monounmuum>cH omnucmn mo musuosupm muficoafiooll.ma mamfla 92 the post-treatment samples which is the exact opposite effect observed in Stream A where the mean number of individuals was observed to decrease in all post—treatment samples. Be- cause the effect was observed in both control and experimental channels, the increase in individuals was probably a function of the natural population dynamics rather than effects due to the toxicant. The apparent discrepancy in results obtained between Stream A and B is best explained by considering the differ- ence in timing of the two treatments. In Stream A, the pre- treatment samples were taken during the summer months when the streams had a greater number of individuals present. The post-treatment samples were taken during early fall when the summer and fall-emerging insects had left the system hence showing a slight decrease in total density of individu- als in the streams. The pre-treatment samples for Stream B were taken during this time of low density, however, the post-treatment samples were taken throughout the early winter months when the following year"sgeneration had grown to suf- ficient size to be retained in the sampling apparatus. Val- ues for diversity, equitability and total number of taxa were consistent between pre- and post-treatment samples Showing no significant variation due to the toxicant or natural events occurring within the stream. Since comparison of the data on the basis of seasonal means may obscure immediate and short-term effects of the 93 toxicant on the macroinvertebrate community, the samples taken immediately before treatment and those taken two weeks post-treatment were also compared. In this instance, as in Stream A, the samples taken two weeks following treatment were found to contain significantly fewer individuals than those samples taken immediately prior to treatment. The difference in number of individuals in the control channels for the same time period showed no significant decrease and actually demonstrated an increase in total numbers attributed to random sample variation (Table 17). E£i£3.—-Similar to Stream A, an immediate increase in the number of individuals drifting from the experimental riffle channel was demonstrated following the initiation of treatment with 9.0 mg/l TFM (Figure 17). A simultaneous record of the number of individuals composing the drift in both control and experimental channels demonstrates that approximately four to six individuals are caught in the drift nets during a 10 minute period. However, immediately after the initiation of lampricide treatment, the drift rate reached about 33 individuals per 10 minute interval in the exposed channel and remained at one to five individuals throughout the entire time period in the adjacent control channel. During the entire treatment period the drift rate remained approximately-eight individuals higher than simul- taneous measurements of drift in the control. In this .zma H\mE o.m cua3 ucmfiummuu mCHHDU mamccmno mammau Houucoo pom HmucmEflHmme m Emmuum Scum mmpmu “mayo msomcmuHDEHmll.hH musmflm cam. . m .88 - . .85. . .36. . as. _ .2200 i0.. _ canon . “c0839.... .4, So «.29. Low 3352.. OF too .3252 .0555 3295.898 -00 w m 535m .18 95 instance, as in Stream A, no correlation of drift rate was observed with the onset of darkness. Trentonius distinctus and Baetis sp. made up 85% of the drift. No macroinverte- brate drift measurements were made on the pool communities because the ends of the pool areas had been screened to contain trout that had been stocked in these sections. Production.--The production of the Nemourid stonefly, A. varsharva and the amphipod g. pseudolimnaeus were analyzed in detail (Table 16). The standing crop of the stonefly in the experimental riffle area was 4.19 gms/m2 compared to 3.19 gms/m2 in the control riffles. Production figures were also comparably higher in the experimental riffle at 16.22 gm/m2 compared to 13.27 gmx/mz. However, the turnover ratio of production to biomass was 3.87 in the experimental area and 4.15 in the control indicating that the control commun- ity produced more on a per unit biomass basis than did the exposed community. However, the magnitude of difference in turnover ratios, 0.28, is well within the error tolerance for this method of production estimate and does not neces- sarily represent a real difference. Similar results were obtained for the amphipod, g, pseudolimnaeus with a standing crOp of 19.95 gms/m2 in the . 2 . experimental area compared to 8.90 gms/m 1n the control. The higher biomass in the experimental channel accounted . 2 . for 84.68 gms/m2 of production compared to 38.17 gms/m in 96 v~.e mo.vm mm.ma mm.¢ ha.mm om.m xenma um: NH -msma ummm me Hmuoe ea.v om.em mm.mH sm.m 8H.Nm mo.m ucmsummnuuumom mm.v sv.ama mo.mm no.8 mm.vv Hm.oa pawsummnuumum mswmcaflaopsumm moHMEEmo em.m mm.oH ma.v mH.e nm.ma mH.m lemma um: NH umnma ummm my Hmuoe we.m mn.e mo.m 88.8 no.8 om.a ucmsummuuuumom mm.m vv.m om.m mm.m mv.e me.m ucmfiummnuumnm m>umnmum> musfimcflnmam ofiumm oflumm Hw>ocusa ANE\u3 umz Boo WNE\uM uoB Evy um>ocuze Amfi\u3 um3 Emv mm”\umgwonEmv uuonou cofiuospoum 0H0 :Hpcmum HHOSOU COAUUSUOHQ U .U um mammam Hmucmsflummxm m Emmuum mamwflm Houucoo m Emouum .m soouum H0008 mo mawccmno HmucoEwummxo pom Houucoo Bonn monounmuuo>cw mo moflommm 03“ How mofiumu uo>ocusu uuozoo can cowuoscoum .mouo mcflocmum mo :OmAHMQEOUII.mH mgmée 97 o.Hv hmo.m Hom.o mmumunwuum>cflouome Hmuou mo unmflm3 who mmHMIzmd Hoom d Emouum o.Hv Oma.w mmm.> monounwuum>cH0HUME Hmuou mo unmflm3 who omHMIzmd wamuflm a Eomuum o.Hv en.mv m~.ov HmucmeHummxm o.Hv Ho.Hs om.vm Honucoo ucmEumwquumom mxmmz N I mHOOm m smouum o.av vv.om mm.mm unmeummuunumom mxmm3 m mammam Houucoo m smmnum o.Hv mm.oHH m.moH usmEummquumoo mgucoe m mo.ov hm.m> m.moa ucmfiummuuuumom mxmwz m oawwflm Hmucmeflummxm m Emmuum mo.ov mm.vn vo.moH Hmucmsflnmmxm mo.ov oo.mm so.mm Houucoo ucmEummHUIumom mxmmz N I mHOOm m Emwuum o.Hv mm.am so.ooH ucmsummnuuumoo mnucoe m mo.ov 00.80 so.ooH ucmsummnuuumom mxmm3 m mamsz>H©cH mo HmnESZ Hmuoe Hmucwaflummxm 4 Emouum o Amy mocmoamacmam She H\ E HOHHCOU vmumme . . . o.m ou pomomxm ucofiumoquwum mo Hw>oq memo: ousuosuum wuflcsafiou .mmumunouum>cflouoma canucmn mo ousuonnum >uficzfieoo usofiumoHqumom pom Iona coo3umn mucoquMHp mo oocmowmasmflm mo mummqu .wucopaum mo muasmomII.ha mqmde 98 the control community. Similar turnover ratios of 4.29 and 4.24 were obtained for the control and experimental areas indicating that essentially no difference exists between the two communities when production is expressed on a per unit biomass basis. Based on these results, 9.0 mg/l TFM is shown to have neither a stimulatory nor inhibitory effect on the long-term production of these two species of macro- invertebrates. Since the amphipod, E. pseudolimnaeus was common in all core samples taken from control and experimental pool communities of both Streams A and B, an additional production estimate was calculated for these channels to determine if an effect due to the lampricide exposure was present. Pro— duction figures calculated for these four communities ranged from 46.44 gms/m2 to 57.34 gms/m2 and biomass estimates ranged from 9.70 gms/m2 to 13.86 gms/m2 (Table 18). Differ- ences in the total turnover ratio for control and experimental channels of Stream B were 4.28 and 4.14 respectively, indi- cating no significant difference existed between these two communities when production was expressed on a per unit bio- nwss basis. Differences were somewhat more pronounced between the control and experimental channels of Stream A where the turnover ratios were 4.53 and 5.13 respectively. This indi- cates that although there was a lower standing crop in the experimental area than in the control, the individuals present 99 va.v vm.hm om.MH mm.v mo.mv MN.HH AnohmZIummmv HmuOB mm.v mm.am mo.ma hH.v ma.mm vv.m ucweummuulumom mm.m ma.ah mm.mH av.v hm.om mm.mH ucmEummquonm Hoom Hmucwfiflummxm m Emmuum Hoom Houucoo m Emmuum mH.m mw.mv 05.0 mm.v vv.wv om.oa Acouszummmv Hmuoe vv.m mv.mm no.h ©m.m mm.mm om.m ucmEuomHqumom nv.v ov.mm mm.0m mm.m mm.oo om.mH ucmEummqumum Hoom Hmucwefluomxm 4 Emmuum Hoom Houucou m Emwuum oflumm ANE\u3 umB meme ANE\u3 uoz mfimv oflumm ANE\u3 umz mEmv ANE\u3 uo3 mfimv um>ocuse cofluospoum mono mcflpcmum Hm>ocusa cofluospoum mono mcflocmum mammcEHHOUsomm mDHmEEMU .m com d mEmmnum Hmooe mo mamccmco Hoom Hmucweflnmmxm pom Houucoo Eoum mommcfiflaoosmmm msuofifimo How moflumu umDOCMDu cam cofluospOMQ .mouo mcflpcmum mo cemwummfiooll.ma mqmme 100 in the experimental channel produced a greater amount of biomass per unit weight of standing crop than did the con— trol community over the same time period. This increase in turnover ratio is more likely due to the variability of samp— ling technique and inherent error of the production estima- tion technique rather than to any real difference brought about by the exposure to TFM. Stream C Community Structure.--The primary purpose of Stream C was for description of the dynamics of TFM residues in a stream system. The treatment was conducted with l4C-TFM and exposure was done under recirculating conditions. Drifting individuals were not lost from the system, and therefore only minimal sampling was done to describe the effects of the toxicant on community structure. No samples were taken from the pool communities of Stream C since young brown trout were held in these communities for a replicate experi- ment concerning the effects of the toxicant on fish produc- tion. The pre- and post-treatment effects of the recircu- lating treatment with l4C-TFM on community diversity, equit- ability, number of taxa and number of individuals per square meter as shown in Table 19. Again in this instance, as with Streams A and B the standard error of the mean for the pre- treatment samples overlaps the standard error for the 101 TABLE l9.--Community structure of benthic macroinvertebrates from Stream C control and experimental riffles. Stream C Control Stream C Experimental Riffle Community Riffle Community Pre—treatment , , + + D1vers1ty 3.23-0.268* 2.95—0.154 . . . + + Equitability 0.66-0.050 0.65—0.081 . + + Number of SpeCies 14.33-3.06 12.30-l.73 . . 2 + + Number of Ind1v1duals/m 9785-1484 9.98-1145 Post-treatment , , + + Diver51ty 3.14-O.562 3.35-0.221 . , , + + Equitability 0.62-0.047 0.74-0.054 . + + Number of Spec1es 13.83-2.l3 13.38-l.94 . . 2 + + Number of Ind1v1duals/m 9377-1322 9308-1452 *Standard error of the mean 102 post-treatment samples indicating that for all samples taken, essentially no difference can be detected due to the lampri- cide exposure. A slight rise was indicated in the mean number of individuals per square meter in the experimental channel following treatment but the increase is within the range of sampling error and does not indicate a real dif- ference. Production.--The production of the Nemourid stonefly, A. varsharva and amphipod, g. pseudolimnaeus were analyzed in detail to determine if any difference between control and experimental channel populations could be attributed to the effects of TFM. Both species were common in samples taken from Stream C but their density was significantly lower than in Streams A and B. Standing crop figures were 0.66 gms/m2 and 0.63 gms/m2 for the stonefly in Stream C experimental and control channels respectively. The production value for the control area was 2.32 gms/m2 with a turnover ratio of 3.65 and was 2.30 gms/m2 with a turnover ratio of 3.47 for the experimental area indicating no difference between the two communities with respect to the effects of the toxicant (Table 20). The biomass and corresponding production figures for E. pseudolimnaeus were also significantly lower in both con- trol and experimental channels of Model Stream C than in Streams A or B. The standing crop in Stream C experimental 103 no.v mm.mm mm.h ma.v mv.vm mm.m ASUHMZIummmV HmuOB mm.m bv.®m mm.m ©N.¢ ma.mm mm.® ucmfiummuulumom mm.m mH.© mm.a mm.m mv.mm mh.ma ucmfiummHUImMm mommcEHHO©3wmm mDHmEEmU hv.m om.m mo.o mo.m mm.m m©.o Abouszummmv Hmuoe vv.m mm.o mm.o av.m oa.m H®.o ucmEummquumom mm.m HN.H no.0 mm.m mm.H H>.o ucmfiummuunwum m>umnmum> MADEmcHLmEd oflumm oflumm Hm>OCHDB ANE\u3 um3 mfibv AME\u3muw3 meme Hm>OCHDB ANE\93 uo3 mEmV Am5\u3mum3 mEmv uuonou cofluospoum Ono capcmum uuonou cofluospoum 0M0 aflocmum meMHm Hmucmeflummxm U Emouum mammam Honucoo U Emmuum .U Emmuum mo mmHMMHM Hmucmfiflummxm 0cm Houucoo Eoum mmumunmuno>cHOHome mo mwflommm O3» MOM moflumu uo>0cusu uuozoo pom coHpospoum .mouo mafipcmum Mo comflnmmfiou «II.ON mqmfie 104 channel was 7.33 gms/m2 with a production of 29.85 gms/m2 while the standing crop in Stream C control channel was 8.23 gmx/m2 with a standing crop of 34.48 gms/mz. The correspond- ing turnover ratios for experimental and control communities were 4.13 and 4.07 respectively which was again well within the error of the method and represents no essential differ- ence in production of g. pseudolimnaeus between control and exposed populations. Fish Production The results of the growth and production experiments with young-of-the-year brown trout, Salmo trutta in experi- mental and control pool areas of model Streams B and C are given in Table 21. The growth test in Stream B covered 43 days and the fish were initially stocked at a rate of 25.55 gms/m2 in the control pool and 25.95 gms/m2 in the adjacent experimental pool. A 1.27 cm mesh seine was suspended over these communities to insure that the fish would not escape. Also, the hatchery raceway screens were put in place at the end of the pool sections to prohibit emigration downstream into the riffle communities. During the treatment, these young of the year fish were visibly affected by the toxicant. After approximately 8 hours of exposure the individual fish were observed to move from their protective and secretive habitats in the shadows of the pool to open, unprotective areas. Their 105 .mofluflamuuoe mcflosaoxo muo>fl>n5m mo mflmmn co omumHsono cofluosooumxx .musmomxm Shh OCHHDU cmflm OED mo swoop 0» wow mmOH unmwm3« ona.m mom.H oo.mH mH.eH mxmp mm Hmucmsnumoxm o ammuum who.m meo.m oo.em mm.mm msmu mm Honucoo o emmnum mms.a «.Hmm.a moH.ou Hma.ou 4mm.mm mm.mm msmo me Hmocmsflummxm m smmnum moe.a enm.H mv.em mm.mm memo me Houucoo m smmnum s A mo\Em\mEv ANE\Emv ANE\EmV ANE\EbV ucmfiflummxm wumm £u~3OHU COHUUDUOHQ mmflEOHm mmMEOHm O Efimvaw‘H HTCCMEU Smeum cows . ucoaumoquumom usoEummquoum m .Zme H\oE o.m ou musmomxm cm mcflonaom U can m mEowuum mo maoccmno Houucoo pom Hmucmefluomxo cfl muusuu OEHmm .usouu c3oun umoonnuImoImcsow mo cofluosooum pom £u3ou0II.Hm mqmée 106 behavior was lethargic and they were not visibly affected by movements of the observer. An increase in gill ventila- tion rates was observed for these individuals but the leth- argic state and lack of fright response remained throughout the exposure. The two smallest individuals of 6.0 cm and 6.9 cm standard lengths and weights of 1.78 gm and 2.00 gm respectively died during the treatment presumably as a direct effect of the toxicant. The survivors had returned to normal secretive behavior patterns and normal gill ven- tilation rates by the following morning. At the termination of the 43 day growth period the control pool had a post-treatment biomass of 27.43 gm/m2 indicating a gross production of 1.89 gm/m2 of fish. In the experimental pool, the loss of the two individuals due to mortality caused a net decline in post-treatment bio- mass to 25.83 gm/m2 and a negative gross production of -0.121 gm/mz. Results for the experimental pool fish were also calculated by subtracting the weight of the dead fish from the initial mean biomass in order to examine the growth rate of the survivors without the negative effect of the mortality losses. Calculating by this method yielded a gross production of 1.851 gms/m2 for the survivors. Due to the inherent variation in growth between in- dividual fish these figures for mean production per unit area of stream bottom mean little for a comparative test. 107 For this reason the rate or amount of production per gram of fish flesh standing crop was calculated for experimental and control fish. The mean growth rate calculated by this method for the control fish was 1.709 mg gm-l day 1 compared 1 day-1 in the experimental to a net decline of -0.108 mg gm- pool. However, if the growth rate is calculated on the basis of the survivors alone, neglecting the two mortalities, a growth rate of 1.795 mg gm.l day-1 is obtained. A test of the null hypothesis that the individual growth rates for survivors in the experimental pool were equal to growth rates obtained in the control pool indicated that the growth rates between communities were equal with 99% confidence (Table 22). Similar results were obtained with the recirculating treatment of Stream C (Table 21). Fish were stocked into the control pool at a rate of 23.93 gm/m2 and into the ex- perimental pool at 17.15 gm/mz. During the recirculating treatment with l4C-TFM behavioral changes in the fish were also observed, however, no mortalities were recorded and all fish were recovered at the termination of the 35 day experiment. At this time, the post-treatment biomass in the control community had increased to 27.00 gms/m2 indicat- ing a gross production of 3.078 gm/mz. The experimental fish biomass had risen to 19.06 gm/m2 with a gross production of 1.903 gm/mz. Calculation of growth rates on a per gram of standing crop basis indicates a rate of 3.676 mg gm-1 day-1 in the control pools and a rate of 3.170 mg gm.l day-l 108 TABLE 22.--A comparison of individual growth rates of brown trout, Salmo trutta, between control and experimental channels of Model Streams B and C. Mean Growth Stream B Stream C Rates -1 -1 -1 -1 Control X1 = 1.709 mg gm day X3 = 3.676 mg gm day . -l -1 -l -1 Experimental X2 = 1.795 mg gm day X4 = 3.170 mg gm day Statistical fiypotheses HO: X1 = X2 HO: X3 = X4 H1: Xl # X2 H1: Xl # X2 t = 0.3274 t = 0.7404 t(.01,20) = 2.845 t(.01,20) = 2.845 Conclusion: Accept HO: X = X and X = X 1 2 3 4 109 in the experimental pool. A statistical comparison of the individual growth rates for control and experimental fish indicates that no difference exists between these communities with 99% significance (Table 22). Kinetics of TFM in the Stream Community Uptake of TFM Residues The uptake, bioconcentration and elimination rate of TFM was determined for 20 separate plant and animal com- ponents in nine replicated experiments employing l4C-TFM in recirculating systems. All exposures were for 24 hours with an initial TFM concentration of 9.0 mg/l. Sample data for each species was averaged between the three to five repli- cates to add statistical significance to the mathematical descriptions of residue dynamics. The uptake rate of TFM by all components of the stream community is expressed in Figure 18 by the reciprocal loss of TFM activity from water over the 24 hour exposure period. There is a rapid initial drop in concentration to about 8.0 mg/l within the first two hours following initia- tion of treatment. The slope of the loss of activity curve then levels off to about 6.9 mg/l after the entire 24 hour exposure period. The graph accurately reflects the reciprocal behav- ior of the toxicant in the biotic components of the stream. An examination of the TFM concentrations in most of the 110 Figure 18.--Uptake of l4C-TFM expressed as loss from water. Mean i 1 S.D. 111 #N m. 495 95... A. 358.: . Od 112 species and components demonstrated a rapid uptake within the first 2 hours. This initial uptake would then corres- pond to the rapid loss of activity in the stream water occur- ring during the same time period. Following this initial surge, the uptake rates level off and increase at a linear rate throughout the duration of the exposure. Since there were generally only two samples taken during this initial rapid uptake component, accurate mathematical description of this portion of the curve was not possible. The mathematical expression employed for the final characterization of uptake rate was a simple linear regres- sion of ug TFM/gm tissue on a dry weight basis against the exposure time in hours. Calculation of the sample data by this method yields a Y-intercept value which is an accurate estimate of the tissue concentration reached at the end of the two hour initial rapid uptake period. The linear portion of the uptake curve is then accurately described by the re- gression equation of the form: Y = a + b (X) where: Y = concentration of total TFM residue in the organ- ism expressed as ug per gm dry wt. a = the y-intercept of regression b = rate of uptake or slope of the regression X = exposure time in hours. The regression intercept, regression coefficient, sample standard deviation of the regression coefficient Sb’ 113 and confidence intervals of the slope were calculated ac- cording to Steele and Torrie (1960). An example of these statistics for the amphipod, g. pseudolimnaeus is shown in Figure 19. The data are typical of the uptake dynamics of TFM in all of the biotic components examined. The two earl- iest data points taken at l and 3 hours indicate that the amphipod has a total body residue of 39.5 i 5.6 ug/gm and 78.9 i 21.7 ug/gm at these time intervals respectively. A simple linear regression of all of the data points through- out the 24 hour exposure indicates that the Y-intercept is 54.6 ug/gm with a positive slope of 4.77. The calculation of the Y-intercept value thus yields an accurate estimate of the total TFM residue at the termination of the 2 hour initial rapid uptake period and the equation for the straight line allows for accurate prediction and descrip- tion of the secondary or reduced linear phase of uptake. The calculated value of the slope indicates that during this secondary phase, the amphipod concentrates a body resi- due of TFM at approximately 4.77 ug/gm body weight per hour and 95% confidence intervals for this slope was £1.126. The relatively high correlation coefficient of .94 indicates that the assumption of linearity in the relationship is valid. The equation can be employed to estimate the total residue concentration at any time during the 24 hour ex- posure by substituting the time of exposure for the value of X in the equation. 114 2N3(,- -F Gammarus A/ Isol .. _. E Q 8’ 100r , Reduced phase: I ‘ Y: 54.59 + 4.77(X) ’ 331.126 50¢ ,/ Initial "-940 phase I I I I , L 4 l 1 1 1 4 4 24 Time (In-s.) Figure l9.-—The rate of uptake of l4C-TFM by the amphipod, g. pseudolimnaeus during a 24 hour exposure to 9.0 mg/l TFM. 115 Table 23 lists the critical values of linear regres- sions for the remaining 19 plant and animal components exam- ined during the year. Calculated values for the Y-intercept which estimate the total body burden at the end of the 2 hour initial rapid phase of uptake range from 1.603 ug/gm for young crayfish, Orconectes propinquus to a high of 221.994 ug/gm for annelid worms. Rates of uptake for animal compon- ents during the secondary linear phase range from 0.344 pg gm-l hr.l for the crayfish to 17.899 ug gm.l hr.l for the caddisfly larvae, Brachycentrus americanus. The general trend evident from these comparisons is that the rate of up- take and therefore the ultimate total body burden of TFM is directly related to the permeability of the test species' integument. Those species with relatively rigid sclerotin- ized exoskeletons or those with calcareous shells such as the clams and snails concentrate TFM at much slower rates and consequently the total TFM concentration at the termina- tion of the 24 hour exposure period in these species is sig- nificantly lower than those species with relatively soft, membranous integuments. Large molluscs of the genus Anodonta were separated into three components and each was analyzed separately as part of an experiment to evaluate the mollusc as an indicator of environmental concentrations of TFM in a treated stream. Although many more clams were exposed and analyzed than for most of the other species, the span of the 116 TABLE 23.--Accumu1ation of total TFM residue in several plant and animal components exposed for 24 hrs. to a recirculating concentra- tion of 9.0 mg/l TFM. Rate of 95% . . Species Y-Intercept Uptake Confidence Correlation (Slope) Intervals Coeff1c1ent of Slope Gammarus pseudolimnaeus 54.597 4.774 i 1.126 .940 Asellus militaris 43.193 4.509 f 0.144 .981 Orconectes pgppinquus 1.603 0.344 f 0.144 .981 Anodonta sp. Foot 18.297 10.899 i 7.774 .824 Anodonta sp. Gills 45.115 7.831 i 6.570 .881 Anodonta Sp. Viscera 32.397 9.054 i 7.044 .809 Pisidium sp. 34.966 3.634 i 1.395 .986 PAy§A_sp. 16.989 2.729 i 0.447 .994 Hexagenia sp. 39.343 1.965 f 1.673 .793 Glossosoma sp. 114.543 7.976 i 2.307 .976 Limnephilus sp. 95.053 3.320 t 4.819 .987 Brachycentrus americanus 129.779 17.899 :10.327 .921 Brachycentrus cases only 189.258 27.939 :19.529 .987 Annelid worms 221.994 9.663 f 7.906 .868 Cladophora and + Stigeoclonium (p001) 39.429 0.891 - 0.806 .969 Stigeoclonium (riffle) 43.951 2.587 f 1.242 .838 Ceratophyllum demersum 54.089 2.327 1 0.312 .964 Elodea canadensis 21.029 1.249 i 1.812 .991 Moss 27.282 0.482 f 0.551 .988 Aufwuchs (5 cm2) 0.202 0.252 f 0.173 .973 (wet weight) Leaf discs 43.609 1.784 f 0.563 .903 (20 mm dia.) Sediment 14.476 0.866 f 0.771 .865 117 confidence limits on the uptake rates for the three compon- ents indicates the variability encountered between the in- dividuals. This variation in residue concentrations was assumed to be explained by the variations in activity levels of the different clams during the exposure. Some individu- a;s were observed to be actively siphoning, others moving about with foot extended and still others tightly clamped shut with minimal siphoning activity during the period of exposure. These behavioral differences add great variability to the total residue concentrations and make the clam an un- reliable indicator of environmental levels of TFM. Brachycentrus americanus and its vegetative cases were analyzed separately to examine variations in total residues. After the samples were thawed the animals and cases were separated, dried and combusted as individual samples. The cases consistently produced high residue levels as did the animals. It is speculated that this high level in the case may be due to the microflora and microfauna as- sociated with the case. However, 20 mm diameter leaf discs did not concentrate the extremely high residue levels ob- served for Brachycentrus cases. Since they had entered the stream during the previous fall, they were relatively well processed and probably did not support comparable commu- nities of microflora. Several plant components of the model streams were also examined and concentration ranges at the end of the 118 initial rapid uptake period ranged from 21.029 pg/gm for Elodea canadensis to 54.089 ug/gm for Ceratophyllum demersum in the pool communities. Rates of uptake during the second- ary linear phase were generally slower than animal compon- 1 -1 ents ranging from 0.482 pg gm- hr for stream moss to 1 -1 hr 2.587 ug gm- for the filamentous green algae Cladophora and Stigeoclonium growing together. Uptake and accumulation by the combined Aufwuchs community was examined employing the periphyton sampler used for species composition determe inations. A total of 5 cm2 were removed from an exposed rock surface and concentrated onto a preweighed 25 mm dia. Millipore filter. The entire filter and sample were then combusted for analysis of total residue. Due to the varia- tion in dry weight figures with such low total biomass the residue accumulations are expressed on a wet weight basis after air drying at room temperature for approximately 1 hour. Elimination of TFM Residues Numerous transformations were also attempted with elimination rate data for several representative components and it was determined that the best description of these data was obtained by a semi-log transformation of the time in hours. The actual concentrations of TFM as determined by radioassay for each component were regressed against the log time in hours. The majority of the total body burden 119 was eliminated rapidly, within the first day after exposure for most components. However, detectable residue levels were present in most of these components for 2 to 4 weeks following exposure. The use of the log time axis thus per— mitted an accurate description of the elimination curve with a linear equation. The equation for this relationship is of the general form: Y = a + (-b) (log X) where: Y concentration of total TFM residue in the organ- ism expressed as ug per gm dry weight a = the Y-intercept of regression or initial concen- tration in tissue at the initiation of the elim- ination period b = rate of loss or lepe of regression log X = log of time in hours. A graph of these data for a typical macroinverte- brate, g. pseudolimnaeus is shown in Figure 20. The value for the Y-intercept in the equation is 169.14 ug/gm and rep- resents the initial concentration in the tissues of the amphipod following the 24 hour exposure period. The rate of loss indicated by the negative slope is 62.41 ug gm.-1 [1og (hr)]-'1 with 95% confidence intervals of :0.253. The relatively high correlation coefficient of .97 indicates that the linear relationship is an accurate description of the rate of loss data. Calculation of the half-life, the time until 50% of the initial TFM residue is eliminated, was 120 2007' Gammarus 1501 T Y=169.14-(62.41)IogX 0.253 {:2 i T r=.970 :3 100+ ‘\3 sol T 1/2 =26 hrs. )- p— h- F- )- D h r 1 ' ' "717140 100 Time (hrs.) Figure 20.-—The rate of elimination of l4C-TFM by the amphipod, g. pseudolimnaeus following a 24 hr. exposure to 9.0 mg/l TFM. 121 done by substituting 50% of the initial body burden as ug/gm into the equation for Y and solving for X in hours. The half-life solution was checked for accuracy with the graph of the actual data. The half-life for g. pseudolimnaeus was found to be 26 hours. The critical values for elimination data for the re- maining plant and animal componentsare:listed in Table 24. The half-life data of the animal components range from 6.52 hours for the crayfish, Orconectes to a high of 38.3 hours for Hexagenia. The half-life figures for the isopod, Asellus militaris and annelid worms were 194. hrs. and 3808 hrs. re- spectively. It is probable that the longer half-lives for these species reflect continued residue concentration from their close association and burrowing habits into the organic matter of the pool bottoms. A strong correlation exists between the half-life data and the substrate associations of the test species, i.e., those species closely associated with the organic matter and sediment of the pool communities had significantly longer half-lives than those associated with the gravel and rubble substrate of the riffle communities. The mean half- life calculated for pool species was approximately 140 hours while the mean for the riffle-dwelling species was only 17.8 hours. A comparison of the filamentous green algae, Cladophora and Stigeoclonium sampled from pool and riffle 122 TABLE 24.--Half*11fe and rate of loss of total 18M residue from several Plant and an.mn3 comporonts following a 24 hr. wxrosnre to 9.0 mg/l TFM. Initial Rate 95% . Concentration Half-life of Confidencc Correlation Spec1es . . . . 1n T1ssue (hours) Loss Intervals Coeff1c1vnt (ug/gm) (slopw) of slope + Gammarus E§EfigjliFULii§i 169.142 26.1 - (7.409 -O.253 .970 . . . + Asellus {3135;gj;3 H w;l) 151.382 194.0 - n 4r; -9.057 .984 + Orconectes EISLLESPA; 9.856 6.5 - 5.933 -0.211 .885 + Anodonta Sp. V00: 279.776 26.9 ~ 9.687 -0.244 .600 + Anodonta sp. Gills 232.944 22.9 - 14.0I5 -0.296 .670 + Anoggfl£g_sp. V)? "m 249.272 22.6 — 11.538 -U.304 .644 . . . .9. Pisidium sp. 122.182 7.1 - 29.881 -O.844 .931 + Physa sp. 82.479 23.2 - 15.703 -0.733 .893 + Hexaggfli§.sp. (pool‘ 86.481 38.3 - 27..58 -1.851 .977 Glossosoma_sy. 305.410 11.3 ~11. I63 -0.999 .926 _ + Limncphilus 5p. 174.680 14.0 — 46.506 -0.699 .937 + Brachyccglrni 2231:. WY; 558.361 19.? ~IL4.03O -2.486 .957 + Brachyccntgus can; uni} 859.536 23.7 - 90.631 - .925 .965 + Annelid worms 452.912 3808.0 w 57.534 -2.743 .899 Cladthora and + Stigeoclonilm {nwwli 60.876 66.1 — 11.168 -O.878 .977 CladOphorfl and , . + . Stigeocloujum 'riffle) 105.088 25.3 - 11.446 —0.884 .986 + . Ceratophylldm gogggium 129.848 440.0 — 21.144 -1.869 .945 + Elodea canadensiu 50.976 52.4 - 10 236 -O.989 .961 _ + . Moss 38.568 13.2 - «.5’5 -0.302 .947 ,) + Aufwuchs (5 um‘) 6.250 24.5 - 2.0u9 —0.034 .925 (wet wt.) + Leaf discs 86.429 11.6 - 21.322 —3.073 .997 (20 mm. dia.) Sediment 35.254 170.0 m 6.230 -1.895 .898 123 environments demonstrates this difference in half-lives between the two areas. The half-life calculated from data obtained by exposing and following the algae in the pool area is 66.1 hr-.while the half-life calculated for the same species in the riffle area is only 25.6 hr. These relatively rapid half-lives of residue elim- ination and substrate associations indicate that the uptake of TFM is largely an absorption phenomenum on the integument of the animal. The compound is relatively polar and water soluble and for this reason does not appear to be tightly bound in any significant concentrations and is rapidly elim- inated upon exposure to TFM-free water. DISCUSSION Measurements of changes in the metabolic rate of stream and river communities have generally been used to characterize changes in the stream community brought about by nutrient inputs and cultural eutrophication (Odum, 1956; Cole, 1973). These changes are generally of a long-term nature and involve comparisons with similar upstream stations for control data. Several workers have employed laboratory model streams of various degrees of complexity to describe the metabolic responses of the benthic community to nutrient input or changes in the physical environment such as tempera- ture or light intensity (Odum and Hoskins, 1957; Beyers, 1963; McIntire et. al., 1964; Kevern, 1962; McIntire and Phinney, 1965). However, very few investigations into the effects of toxicants on community metabolism of stream sys— tems exists outside of work done at Oregon State with Kraft mill effluent (Williams, 1969). Community Metabolism The primary purpose of the community metabolism measurements in this study was to determine if respirometry measurements would provide a short-term assessment of the influence of TFM lampricide on the stream community. Since 124 125 the compound has been shown to be strongly algi-static at relatively low exposure levels (Maki et. al., 1974), it was hypothesized that chronic effects on algal metabolism may be produced during a short-term exposure. This hypothesis was also supported by the work of Huang and Gloyna (1968); who tested the effects of numerous phenolic compounds on photosynthetic oxygen production using a direct manometric technique, they found that halogenophenols exhibited greater destructive effects than the unsubstituted phenol and that introduction of a nitro group to phenol greatly decreased the chlorophyll content of exposed cultures. They found that the o-,m-, and p-nitrophenols reduced the rate of photosynthetic oxygen production by as much as 70% at relatively high exposure levels of 50 to 100 mg/l. The values for gross primary production in the riffle communities of this study ranged from 58.2 mgO hr“1 2 m-2 to 79.0 mgOzhr-'1m-2 during the course of the experiments which is considerably lower than the 189 to 434 mgOzhr-lm-2 reported by McIntire and Phinney (1965) (Table 25). How- ever, greater illumination intensities, warmer temperatures and extremely high standing crOp figures for the primary producers in those streams makes comparisons invalid. Reese (1966), obtained measurements of primary production in non- enriched sections of a shallow woodland stream ranging from 1 52 to 200 mgOZhr- m-2. Since the intent of this study was 126 to model a small woodland stream, favorable comparisons can be made. Under an exposure to 9.0 mg/l TFM the gross pri- mary production dropped from 5 to 50% of the pre-treatment rates in the riffle communities and was reduced to non- detectable levels in the pool areas. Pre-treatment rates of community respiration ranged - - F“ from 20.3 to 33.1 mgOzhr 1m 2 in the riffle areas and from * 10.2 to 54.0 mgO hrulm-2 in the pool areas. These values 2 are again lower than the range of 104 to 171 mgOZhr-lm-2 i reported by McIntire and Phinney also probably due to the lower standing crop of primary producers in this study. The values compare favorably with the range of 16 to 125 mgOZhr-nlm-2 determined by Williams (1969) and Reese (1966) for small woodland streams. Exposure of the communities in this investigation to 9.0 mg/l TFM consistently caused an increase in respiration rates of from 5 to 40% in the riffle areas to a range of 10 to 50% in the heterotrophic pool areas. This is undoubtedly a measure of the increased respiration rates of the enclosed macroinvertebrate commu- nity as a response to the sublethal stress of the toxicant exposure. The P/R ratios calculated for the riffle communities in this study ranged from 0.92 to 2.07 and from 0.0 to 0.68 in the pool communities. The P/R ratio is an excellent measure of the degree of heterotrophy of a community, 127 m.m see ooo mmoeoom seesaw lemmec sooo 0m.0Ivm.0 0Haumnv m0mI0.0HH mEOQMHxO .xoonu coumaoxm Aoomav uwcwnmmasmm m.mum.a HnHIvOH vMVImmH Emonum muoumnoomH ”momav wmccflzm ocm onflucHoz 0.6I0H.H mmalma mmvnmn Emmnum SpoumnoomH Amomav mEmHHHHZ no.0 0NH mom mm.0 m.vm mm Emouum womawooz Aoomav mmoom mm.o 00.0w vo.Hm maoom memo: ucmEumonqumom 0m.H av.mm 08.H0 mmamwem memo: ucmEumoquumom ma.0 0v.0m vw.aa maoom memo: pcoEumonu msHHDQ 06.0 em.VM mm.0v moawmem memo: ucmsumouu 0cflu50 mm .0 ©m.NN wv.©N maoom mcmwz ucmfiumwuulmum 0m.H m0.0m 00.00 moamwflm memo: umeuwmqumnm msmwuum Hopoz >050m mesa m m E n n oeoom m\d Ami mu g m H- n SHHMQ 2 Hanna mom COeuosooum coflumooq coHpmuHU . >uHCSEEou mmouw .mEmpm>m Honsumc 0:0 wuoumuoan umnuo Scum mouoEHumm LDHB >050m ucmmmnm on» ca posemuno mucmEonomomE Emflaonmuoe >uHCDEEoo mo conflummsoo ¢II.mN mqm<9 128 generally the lower the ratios, the more strongly hetero- trophic conditions as shown by the calculated ratios for the pool areas. The data again compare favorably with Reese (1966) where P/R ratios ranged from 0.16 to 0.85 and averages slightly lower than the P/R ratio of 2.5 calculated by McIntire and Phinney (1965). The P/R ratio was found to be the most sensitive indicator of the toxic influence of TFM where values calculated for communities under exposure to the toxicant dropped from 0.5 to 1.0 compared to pre- treatment ratios. In most instances the effects of the lampricide were sufficient to drive gross production in the pool communities to unedetectable levels which then pro- duced a P/R ratio of zero. The temporary nature of the toxic effects of TFM is demonstrated in the post-treatment data and should be emphasized. In all cases examined in both pool and riffle areas, the gross production, respiration, and P/R ratios measured within one or two days following the day of ex- posure had returned to pre-treatment levels with no signifi- cant differences detectable. Periphyton Community Structure The toxicity of TFM to various algal groups can be generally correlated with taxonomic divisions at the ordinal level. The values for 96 hr. EC-50's range from a mean of 2.2 mg/l for diatom species (Bacillariophyceae) to a mean of 129 4.8 mg/l for green algae (Chlorophyta) and a mean of 7.9 mg/l for blue-greens (Cyanophyta) (Maki, et. al., 1974). The values do not actually represent algicidal levels of the toxicant for, when these exposed cultures were filtered free cf TFM and resuspended in lampricide-free growth media, growth ensued at normal control rates. Similar results were obtained with cultures exposed to concentrations as high as 30 mg/l. Therefore, the compound appears to exert an algi-static effect with most pronounced influence on the diatom species. Several experiments were conducted during the course of the model stream treatments to determine if an effect on the species composition of the periphyton community could be demonstrated from a 24 hour exposure to 9.0 mg/l TFM. Seasonal densities in the standing crops of the diatoms in Streams A and B varied from a low of approximately 1.20 x 109 cells/m2 during the early fall months to peaks of ap-. 9 cells/m2 during the early summer and proximately 3.0 x 10 early winter periods. With these strong variations, only through comparison with periphyton community structure in the adjacent control channel could any effects due to the TFM treatment be identified. Results between control and experimental riffle channels of Streams A and B agreed satisfactorily and the agreement is believed due to the accuracy and reproducibility of the special periphyton 130 sampler employed in the sampling of these communities. A , statistical analysis of the species composition data for pre- and post-treatment sampling dates indicates that no difference in relative densities of the total diatom commu- nity can be attributed to the lampricide exposure. An analysis of the community structure of the diatoms in Streams A and B also demonstrated no difference between the generic diversity, equitability and total number of genera present in control and experimental channels on a pre- and post-treatment basis (Tables 8 and 9). Since no effects on the structure of the diatom com- munities were demonstrated due to exposure to TFM, and since the diatoms appear to be the most sensitive of the algal groupings, predictably minimal effects due to the toxicant should be demonstrated on the filamentous green algal commu- nity. This hypothesis was tested on the green algal assemb- lage of Stream A by sampling to measure the standing crop as ash-free dry weight and a determination of the relative percent frequency of occurrence of the Species present. The biomass of the total green algal community reached a peak in late summer at 38 gms/m2 and dropped continuously to a low of about 3 gms/m2 in mid—December. Control and experi- Inental riffle populations dropped at approximately the same rates and no difference could be attributed to the TFM ex- posure of September 4. Similarly, there were no effects on 131 the relative frequency of occurrence of the three dominant greens, Spirogyra, Cladophora and Stigeoclonium although a gradual change in succession from a Spirogyra-dominant to a Cladophora-dominant community was observed. Since this succession also occurred in the adjacent control channel, the influence of TFM was insignificant. Macroinvertebrate Community Structure The taxonomic composition and general community structure of the macroinvertebrates in all three model stream systems was directly related to flow rate and substrate type differences between the pool and riffle communities. Numbers .of taxa were highest in riffle areas ranging from 23 to 25 taxa and from 7 to 14 taxa in the pool areas with abundance, measured as total number of individuals per unit area of stream bottom, measuring 2 to 3 times higher in the pools than riffles.- This decrease in total number of taxa and increase in number of individuals produced equal differences in the calculated values of the Shannon diversity index for pre-treatment samples, generally 2 to 3 times lower in the pool areas than the riffles. Several concurrent measurements on the community structure and population dynamics of the macroinvertebrates existing intjuamodel streams were made in the attempt to accurately characterize the effects of a lampricide treat- ment on the exposed communities. Much effort was expended 132 to accurately measure the species diversity in all riffle and pool communities on a pre- and post—treatment basis gen- erally since diversity indices have been proposed as sensi- tive indicators of environmental change in aquatic systems (Wilhm and Dorris, 1968; Hooper, 1969). The theoretical major regulators of species diversity are stability, pre- dictability, and rigor of the environment (Paulson and Cul- ver, 1969). Communities with high Species diversity are characterized as highly predictable with low temporal varia- bility (Slobodkin and Sanders, 1969). If TFM were to exert a selective toxicity against certain species or species com- plex and eliminate them from the model streams a comparison of diversity indices between pre- and post-treatment samples should reflect the difference in community structure. Cal- culated values for diversity in all three streams indicated no variation, either increase or decrease, that could be correlated with the lampricide exposure. It appears that most severe effects of an introduc- tion of a toxicant to a stream system is the immediate re- duction in macroinvertebrate density. Reductions in the bottom-living insects ranging from 100% to 10% have been re- ported following forest spraying operations (Ide, 1968). The insect survivors are mainly tolerant egg, pupal or dia- pausing larval stages (Hynes and Williams, 1962). The most sensitive indicator of macroinvertebrate community change 133 in this study and in the two previous field studies of the effects of TFM (Torblaa, 1968; Haas, 1971) was the decline in total number of individuals per unit area of stream bottom. Although lack of replications and variability in substrate types sampled add a high degree of sampling error to the re- sults Torblaa (1968) determined that approximately 77% of the invertebrate taxa present were reduced in abundance one week following treatment. Haas (1971) demonstrated similar effects with an approximate reduction of 90% of invertebrate taxa examined. An approximate reduction of 25 to 30% in density of the total macroinvertebrate population occurred in the riffle areas of both Streams A and B following exposure of these communities to 9.0 mg/l TFM as determined from samples taken 2 weeks after treatment. Although reductions in the macro- invertebrate fauna were also observed in Stream A control riffle, the drop was more severe in the experimental channel and was determined significant at 95% confidence. A similar drop did not occur in the control channel of Stream B but an increase was actually recorded lending additional signifi- cance to the decrease observed following treatment of the experimental riffle of Stream B. . A decrease in the total number of individuals in the pool fauna was also observed in post-treatment samples of Stream A but since an equal decrease occurred simultaneously 134 in the control channel no significance due to the lampricide was implied. The drop in total individuals was explained by the natural hatching and emergence of the resident popu- lation of the midge Tanytarsus which occurred during the two weeks following treatment. In the post-treatment samples of Stream B pool areas taken 2 weeks following the November 1 treatment, an actual increase in total number of individuals was demonstrated due to the growth of early instar midge larvae to sufficient size to be retained by the 0.5 mm samp- ling net. A sharp increase in the frequency of drifting indi— viduals has been reported by several investigators as an im- mediate response to the introduction of a toxicant to a stream system. Hoffman and Surber (1945) reported signifi- cant increases in the drift rates of immature Ephemeroptera and Trichoptera following DDT spraying in West Virginia. Similarly, Hoffman and Drooz (1953) demonstrated the sharp decline in abundance of Trichoptera larvae as a result of drift following a DDT application. Immediate increases in the drift rates of the total benthic fauna have been demon- strated as a response due to the toxicants rotenone and Zectran (Binns, 1967; Gibson and Chapman, 1972). Similar increases in drift rates of the macroin- vertebrate fauna were demonstrated within 15 minutes follow- .ing the introduction of TFM to model Streams A and B. 135 Increases of 5 times the number of individuals taken in si- multaneous samples of the adjacent control channel were re- corded ill the experimental riffle. However, only three of the more TFM -susceptible species made up approximately 80% of this initial drift pulse, Trentonius distinctus, Gammarus pseudolimnaeus and Baetis sp. and the rate rapidly dropped to within 2 times the control channel drift rate within 3 hours following initiation of treatment. The rate then re- mained at approximately 1.5 to 2 times the control rate throughout the remainder of the treatments. This increase in drift pulse during the treatment is proposed as a major route of egress of the macroinvertebrate fauna which is re- flected in the observed decrease in density of post-treat- ment fauna. This factor along with direct toxicant-induced mortality of the more susceptible species are the major causative agents bringing about the reduction of fauna. Waters (1964) demonstrated that drift of benthic fauna is also a major mechanism in bringing about the re- colonization of artificially disturbed sections of stream substrate and that this recolonization can occur in rela- tively short time intervals. He found that the mayfly EEEEEE sp. returned to 100% of its former density within 4 days fol- lowing artificial removal. Torblaa (1968) reported that in- sect abundance exceeded pretreatment densities within six weeks following TFM treatment. Similar results reported in this study from post-treatment samples indicate that drift probably in combination with growth of young instars into 136 effective sampling size brought about the full recovery of stream populations within 3 months following treatment. Production estimates were made for two species of macroinvertebrates common to all three streams, Amphinemura varsharva and Gammarus pseudolimnaeus in an attempt to de— termine if a long-term chronic effect of the TFM treatment could be demonstrated on these species. The Hynes and Cole- amn (1968) technique was used for its relative simplicity and accuracy. Waters (1973) has demonstrated that results obtained with this method are slightly higher but within about 15% of estimates obtained by removal-summation, in- stantaneous growth and the Allen curve methods. Although one of the implied assumptions is that the species be uni- voltine which applies to the stonefly but not to the amphi- pod, the data were calculated for both species from control and experimental channels to gain comparative figures for the two populations. In any case, the method should yield accurate estimates of standing crops for both species. The data demonstrated that the populations were com- parable between control and experimental channels of the same stream system but that larger differences in standing crOps, and therefore total production, existed between Streams A, B, and C. The standing crop figures for Amphif nemura varsharva varied from a low of 0.66 gms/m2 in 137 Stream C to a high of 4.13 gms/m2 in Stream A with produc- tion values of 2.30 gms/m2 and 17.89 ng/m2 respectively. The figures do not represent entire annual production since the estimates were calculated from sample data taken between September when the insects were 1 and 2 mm in length through March when they had attained a total length of 8 and 9 mm. Therefore, the figures do not estimate entire annual produc- tion and should more accurately be termed cohort production over the sampling interval. Calculated results for the amphipod show wide varia- tions between Streams A, B, and C as a result of their multi- voltine nature and the presence of staggered generations present during each sampling period. Standing crOps vary from 7.33 gm/m2 in Stream C to 27.85 mgs/m2 in Stream A resulting in production figures of 29.85 gms/m2 and 124.55 gms/m2 respectively. Calculation of turnover ratios which reflect the quality of production on a per unit biomass basis showed no consistent pattern between control and experimental channels. The ratio was higher than controls in the experimental chan- nels. The ratio was higher than controls in the experimental channels of Stream A and lower than controls in Stream B and C. However, variations were slight, generally of the order of 0.25, and undoubtedly lie within the experimental error of the technique. 138 Therefore, similarly totflmacommunity metabolism data, the temporary nature of the effects due to TFM treatment should be emphasized. The most severe effect demonstrated on the macroinvertebrate community was the temporary reduc- tion to 25 or 30% of the pre-treatment density observed in the experimental channels. Effects were of a short-term nature and no significant alteration of the environment oc- curred to prevent relatively rapid recolonization of the treated channels through drift input. Fish Production Fish production in various aquatic systems has been correlated with standing crop of benthic macroinvertebrates in numerous investigations (Allen, 1951; Ellis and Gowing, 1957; Davis and Warren, 1965). Differential responses in production rates of fish populations exposed to toxicants have been reported as a reflection of variation in standing crops of food organisms. Elson (1967) demonstrated signifi- cant decreases in production of young salmon in the Miramichi River following DDT application. The reduction in standing crop and production was attributed to sharp declines in fish food organisms and direct toxicant-induced mortalities of the fish. Exposure of the tropical fish Cichlasoma bimacus latum to pentachlorophenol caused an increased food consump- tion which increased production to equal that of unexposed 139 controls (Chapman, 1965). Assessment of the effects of a toxicant on a fish population by estimations of growth pre- sent an excellent opportunity to examine the total impact of the toxicant on the fish's behavioral and physiological responses because growth and production reflect an integra- tion of these effects. The effects of TFM treatment on the short-term pro- duction of control and experimental populations of young of the year brown trout, Salmo trutta were determined during the treatments of Streams B and C and there was no signifi- cant difference between mean growth rates of the control and experimental fish for either stream. Kinetics of TFM in the Stream Community Several investigations regarding the ultimate fate of TFM residues in the aquatic environment have been con- ducted. Most of the early investigations suffered from the lack of good quantitative methodology to detect lampricide residues from field and laboratory samples. A streamside bioassay unit was described by Howell and Marquette (1962) employing colorimetric methods to determine concentrations in natural stream waters. The method is useful for deter- mining the quantity of TFM present in natural waters which possess conflicting background colors (Smith et. al., 1960; Ebel, 1962). These colorimetric measurements, however, are limited to solutions containing about 0.1 mg/l or greater 140 from concentrated samples. Compounds other than the lampri- cide also may be concentrated to such a degree that the re- sulting background color obscures the determination (Shapiro 1957, 1958). Daniels et. al., (1965) investigated several analytical techniques including adsorption onto activated carbon, ion exchange resins and activated carbon. These methods were generally considered ineffective for detec- tion of TFM residues from samples of fish, bottom sediments and water (Billy et. al., 1965; Smith, 1966). A satis- factory gas chromatographic procedure has since been de- veloped (Allen and Sills, 1971) which is capable of detecting TFM residues concentrations in fish and other biological tissues of 0.01 ug/gm. Several investigations have been conducted employ- ing these various methods to characterize the environmental fate of TFM residues. Kempe (1973) employed model systems in 500 cc Erlenmeyer flasks with water and natural lake sediments to determine the rate of loss of TFM. He demon- strated the progressive decline of TFM to undetectable levels within 4 weeks following initial exposure; and the systems became nontoxic to see lamprey and goldfish. He concluded that TFM is degraded to nontoxic metabolites by microorganisms that live in sediment water systems. The fate of TFM in rainbow trout was examined by Lech and C051 trini (1972) during studies of the 18 2339 and in vitro metabolism of the toxicant. TFM was found to be reduced i3 vitro by nitroreductase to 3-trifluormethyl-4-amin0phenol. 141 The formation of TFM glucuronide was demonstrated in vitro utilizing liver extracts and the major metabolite of the compound in vivo also appeared to be TFM glucuronide. l4C-TFM and liquid The use of uniformly labelled scintillation techniques greatly simplified the analysis ' and quantification of lampricide residues in the several components examined during the course of this study. The uptake and accumulation rates are unlike those reported for chlorinated hydrocarbon pesticides (Rudd, 1964; Woodwell, 1967; Reinert, 1972) and the differences are believed due to the relative polarity and water solubility of TFM com- pared to these compounds. The graph of the uptake rates of TFM were similar for all plant and animal components ex- amined and can best be described as consisting of two sepa- rate components, the first being a rapid initial rate probably corresponding to simple adsorption of the polar compound on the surface of the test species and the second or reduced phase of uptake corresponding to incorporation of the compound into internal tissues. The uptake rates were analyzed for 24 hr. periods in all experiments to allow for comparisons with actual field treatments of 9.0 mg/l. Although this period is longer than most stream treatments which are generally of the order of 14 to 18 hours, the experiments were standardized at 24 hours to allow a com- plete day and night cycle under exposure. 142 Rates of uptake and therefore the total body burden of TFM residues at the termination of the 24 hour exposure were strongly correlated with the epidermal structure of the animal components. It was demonstrated that those animals with relatively well chitinized or calcareous exoskeletons adsorbed lower quantities of TFM during both the initial and secondary phases of uptake. This type of behavior would then lend support to the hypothesis that the uptake of the lampricide is primarily an adsorption phenomena since those animals with relatively soft and water permeable integuments accumulated extremely high residues compared to the relae. tively impermeable crayfish, clams and snails. These data also correlate well with the acute toxicity data for the compound. The 96 hour LC50 values of the compound are simi- larly correlated with the epidermal structure of the test species, i.e., those with the relatively soft exoskeletons have lower LC50 values than do species with rigid integuments exposed under identical conditions (Maki, unpublished data). Additional support for the explanation of uptake and accumulation on the basis of adsorption is offered from ex- amination of the elimination rates of the compound from theseauflmmfl.components. Significant reductions in total body burdens begin immediately following the termination of exposure and introduction of clear stream water. Values for half-life of residue concentrations range from 6.5 to 38 hrs. and are also correlated with the epidermal structure of the 143 test species. Those species with relatively rigid exoskele- tons, offering little possibility for adsorption have shorter half-lives than do the species with softer integuments. These rapid rates of loss and half-lives of a matter of hours also suggest that TFM is rather loosely bound or adsorbed on the surface of the animals and rapidly becomes water soluble upon termination of exposure. The amount of water passing over the surface of the test animal also appears to have an influence on the elimina- tion rate. A comparison of half-lives between riffle—dwell- ing and pool-dwelling species demonstrates means of 17.8 hrs.and 140 hrs.respectively. The longer half-lives deter- mined for annelid worms and those species in close associa- tion with the rich organic substrate of the pool communities does not rule out the possibility of metabolism and bio- transformation of TFM residues by associated microflora " and microfauna. Ancillary experiments with solvent extracts of sediment cores and subsequent analysis by thin-layer radiochromatography in various solvent systems indicates that TFM residues become tightly bound to a component of the sediment and that this is a relatively large and polar mole- cule. However, results are inconclusive at this time and further description would be speculation. SUMMARY 1. A series of six replicated model streams were established in indoor hatchery channels, stocked with natural stream substrate and associated fauna, provided with artifi- cial illumination and allowed to colonize for 2 to 6 months prior to experimentation. 2. lg §1£2 measurements of net primary production and community respiration were made on these streams em- ploying a specially developed respirometry system. The ef- fects of a 24 hr. exposure to 9.0 mg/l of TFM indicated a 25-50% suppression in rates of gross primary production and a 3-50% increase in rates of community respiration in riffle and pool areas. Post-treatment measurements of community metabolism indicated rates had returned to pre-treatment levels within 1 or 2 days following exposure. 3. Description of periphyton community structure was determined from pre- and post-treatment samples taken from the model streams. No permanent effects on diatom cell density, diversity, equitability, biomass of fila- mentous green algae or species composition of the green algae could be demonstrated from a 14 hr. exposure to 9.0 mg/l TFM. 144 145 4. Minimal effects due to the lampricide were detected in the community structure of benthic macroin- vertebrates. A significant decrease in the total number of individuals was detected from samples removed up to 2 weeks following exposure. However, this difference was of a temporary nature and exposed communities had increased to pre-treatment densities within 2 to 3 months following exposure. No effects due to the lampricide could be demon- strated on diversity, equitability, number of taxa or pro- duction rates of macroinvertebrates. 5. Increases of up to 4 times the control rate of macroinvertebrate drift were demonstrated in experimental channels during exposure to the lampricide. 6. No significant differences were detected in mean growth rates of young of the year brown trout, Salmo trutta between control and experimental populations exposed to 9.0 mg/l of TFM. 7. The uptake and accumulation of TFM residues by several plant and animal components was described as basi- cally an adsorption phenomena. The two component uptake curve consists of a rapid initial phase followed by a re- duced linear phase. Macroinvertebrate species with soft, relatively permeable integuments accumulated higher body burdens than those species with hard chitinized or calcare- ous exoskeletons. 146 8. Elimination of accumulated TFM concentrations occurred rapidly in the plant and animal components exam- ined. Riffle-dwelling species had a mean half-life of 17.8 hrs. and pool-dwelling species had a mean half-life of 140 hrs. indicating a strong correlation between rates of loss and water current-substrate associations. CONCLUS IONS AND RECOMMENDATIONS The underlying theme to the investigations conducted with typical woodland stream flora and fauna during the course of this study must stress the temporary nature of the effects of TFM. No permanent deleterious effect on stream metabolism, production or taxonomic composition could be demonstrated by the methods employed in this research. Recommendations: In consideration of the above, the following recommendations concerning the continued use of TFM are indicated. The effects of the toxicant on long-term genetic changes of macroinvertebrate taxa and direct effects on the composition of the microbial community were not monitored and remain a possibility for continued research. Mortalities of experimental fish during treatment indicate the further need for toxicity information concerning fish. Toxicant concentrations may appear sublethal in simple toxicity tests but exposed fish in the natural stream may accumulate lethal levels from ingestion of invertebrate individuals and associated residue concentrations in addition to residues obtained from the water. The wide range of toxicity values influenced by water quality 147 148 variations, in particular hardness and pH, indicate that levels for protection of aquatic species must consider local water quality. Also, treatments must not include head water reaches of stream systems thereby insuring a continued source of drifting species to re-populate down- stream treated sections. Assuming field treatments with TFM continue to be preceded by a careful series of bio- assays to determine minimally effective concentrations of the toxicant for each watershed, minimal deleterious effects on stream flora and fauna seems assured. However, chemical treatment for pest control is probably never the ultimate method of eradication. This point is underlined from a consideration of the present methods of sea lamprey control with TFM. The compound has been extensively used throughout the Greak Lakes drainage for approximately 14 years with results indicating signifi- cant suppression of lamprey stocks rather than total eradi- cation. The continual potential exists for the rapid increases in lamprey populations should the annual stress of TFM treatments be eliminated. The use of TFM is highly efficient in terms of the numbers of individual lamprey destroyed per treatment when the densities of lamprey pOpulations are high but becomes progressively less effective on a per treatment basis when the populations are low (Hanson, 1970). The continued existence of this residual spawning pOpulations in the lakes is sufficient 149 to provide for a continual annual crop of larval lampreys in successive years following a stream treatment. For this reason, major lamprey spawning streams must be re- treated every 3 to 5 years for the foreseeable future. The need for an additional ancillary or integrated approach to the control of lamprey populations is evident. However, at present there is no substitute that combines the relative F“7 effectiveness and economy of TFM. Research programs are currently underway in both Canada and the United States to develop an integrated method of control. The use of pathogens, parasites, predators, and non-parasitic competi- t1? tor releases, combined with the release of sterile males all hold promise toward the development of a future integrated control program for sea lamprey. APPENDIX 150 151 TABLE A-l.--Daily values of community metabolism demonstrating effects of a 9.0 mg/l TFM exposure. Gross Primary Production Respiration Date P R Ratio (ngZ/mZ/day) (mg/hr/mz) (ngZ/mZ/day) (mg/hr/mz) / July 10-14 RIFFLE (14 hr. daylength) 7/10 1.008 77.5 .488 20.3 2.07 7/11 treated .684 52.6 .636 26.3 1.08 7/12 treated .469 36.1 .636 26.5 0.74 7/13 .540 41.5 .449 18.7 1.20 7/14 .817 62.8 .459 19.1 1.78 July 24-27 RIFFLE (13 hr. daylength) 7/24 .775 59.6 .603 25.1 1.29 7/25 .762 58.6 .620 25.8 1.23 7/26 treated .739 56.8 .752 31.3 0.98 7/27 .832 64.0 .643 26.8 1.29 July 24-27 POOL (13 hr. daylength) 7/24 .201 15.5 .752 31.3 0.27 7/25 .210 16.2 .686 28.6 0.31 7/26 treated .0 0 .811 33.8 0 7/27 .188 14.5 .727 30.3 0.26 August 21—24 RIFFLE (13 hr. daylength) 8/21 .939 72.2 .667 27.8 1.41 8/22 .961 73.9 .635 26.5 1.51 8/23 treated .815 62.7 .640 26.7 1.27 8/24 1.070 82.3 .625 26.0 1.71 August 21-24 POOL (13 hr. daylength) 8/21 .135 10.4 .560 23.3 0.24 8/22 .139 10.7 .547 22.8 0.25 8/23 treated .0 0 .779 32.5 0 8/24 .108 8.3 .651 27.1 0.17 September 3-6 POOL (12 hr. daylength) 9/3 .271 22.6 .400 16.7 0.68 9/4 .257 21.4 .388 16.2 0.66 9/5 treated .200 16.7 .565 23.5 0.35 9/6 .107 8.9 .555 23.1 0.19 152 TABLE A-l.—-(Continued). Gross Primary Production Respiration . Date P R Ra (gm02/m2/day) (mg/hr/mz) (gmoZ/mz/day) (mg/hr/mz) / tlo September 18-21 RIFFLE (12 hr. daylength) 9/18 .743 61.9 .662 27.6 1.12 9/19 .747 62.3 .637 26.5 1.17 9/20 treated .680 56.7 .733 30.5 0.93 9/21 .789 65.8 .669 27.9 1.18 September 18-21 POOL (12 hr. daylength) 9/18 .311 25.9 .616 25.7 0.50 9/19 .262 21.8 .586 24.4 0.45 9/20 treated .194 16.2 .808 33.7 0.24 9/21 0.78 6.5 .483 20.1 0.16 October 14-19 RIFFLE (12 hr. daylength) 10/14 .679 56.6 .566 23.6 1.20 10/15 .720 60.0 .682 28.4 1.06 10/16' .778 64.8 1.091 45.5 0.71 10/17 .405 33.8 .249 10.4 . 1.63 10/18 .301 25.1 .241 10.0 1.25 10/19 .479 39.9 .364 15.2 1.32 October 15-18 POOL (12 hr. daylength) 10/15 0 0 .261 10.9 0 10/16 0 0 .245 10.2 0 10/17 treated 0 0 .372 15.5 0 10/18 0 0 .298 12.4 0 October 29—November 1 RIFFLE (11 hr. daylength) 10/29 .755 68.6 .506 21.1 1.49 10/30 .741 67.4 .519 21.6 1.43 10/31 treated .262 23.8 .527 22.0 0.50 11/1 .664 27.7 .407 19.6 1.41 October 29-November 1 POOL (11 hr. daylength) 10/29 0 0 .461 19.2 0 10/30 0 0 .434 18.1 0 10/31 treated 0 0 .489 20.4 0 11/1 0 0 .176 7.3 0 153 TABLE A-l.--(Continued). Gross Primary Production Respiration P R Rat' Date (ngZ/mZ/day) (mg/hr/m2) (ngz/mZ/day) (mg/hr/mz) / 1° November 13-16 POOL (11 hr. daylength) 11/13 0 0 .479 20.0 0 11/14 treated 0 0 .468 19.5 0 11/15 0 0 .480 20.0 0 November 26-30 RIFFLE (10.5 hr. daylength) 11/26 .776 32.3 11/27 .742 71.4 .781 32.5 0.95 11/28 .733 70.9 .794 33.1 0.92 11/29 treated .608 57.9 1.457 60.7 0.42 11/30 .917 87.3 .702 29.3 1.31 November 26-30 POOL (10.5 hr. daylength) 11/26 0 0 .799 33.3 0 11/27 0 0 .868 36.2 0.61 11/28 0 0 1.296 54.0 0.50 11/29 treated 0 0 1.763 73.5 0 11/30 0 0 1.102 45.9 0.67 LIST OF REFERENCES 154 LIST OF REFERENCES Allen, K. R. 1951. The Horokiwi Stream: A study of a trout population. Fish. Bull. N.Z. 10: 1-238. Allen, J. L. and J. B. Sills. 1971. Gas chromatographic determination of TFM in fish. In: Progress re- port for 1971 on registration-oriented research of TFM used in control of the sea lamprey. Fish Control Laboratory, La Crosse, Wisc. Applegate, V. C. 1950. Natural history of the sea lamprey (Petromyzon marinus) in Michigan. U.S. Dept. of Int. Spec. Sci. Rept.-Fisheries No. 55. 237 pp. Applegate, V. C., B. R. Smith, W. L. Nielsen. 1952. Use of electricity in the control of sea lampreys: Electro-mechanical wiers and traps and electrical barriers. U.S. Fish and Wildlife Serv., Spec. Sci. Rept. Fisheries, No. 92. 52 pp. Applegate, V. C. and J. W. Moffet. 1955. The sea lamprey. Scientific American, 192: 36-41. Applegate, V. C., J. H. Howell, A. E. Hall, and M. A. Smith. 1957. Toxicity of 4,346 chemicals to larval lampreys and fishes. U.S. Fish and Wildlife Serv., Spec. Sci. Rept. Fish. No. 207. 157 pp. . 1958. Use of mononitrophenols containing halo- gens as selective sea lamprey larvicides. Science 127: 336-338. Applegate, V. C., J. H. Howell, J. W. Moffett, B. G. H. Johnson, M. A. Smith. 1961. Use of TFM as a selective sea lamprey larvicide. Great Lakes Fishery Commission, Tech. Rept. No. l. 35 pp. Applegate, V. C. and E. L. King, Jr. 1962. Comparative toxicity of 3-trifluormethyl-4-nitrophenol (TFM) to larval lampreys and eleven species of fishes. Trans. Amer. Fish. Soc. 91(4): 342-345. Baumgardner, R. K. 1966. Oxygen balance in a stream re- ceiving oil refinery effluent. Ph.D. Thesis. Oklahoma State University. 42 pp. 155 156 Beyers, R. J. 1963. The metabolism of twelve aquatic lab- oratory microecosystems. Ecological Monographs 33: 281-303. Billy, T. J., S. L. Daniels, L. L. Kempe, and A. M. Beeton. 1965. Field application of methods for recovery of the selective lampricide, 3-trifluormethyl-4- nitrophenol. Univ. of Mich., Great Lakes Research Division, Publication No. 13: 17-24. Binns, N. A. 1967. Effects of rotenone treatment on the fauna of the Green River, Wyoming. Wyoming Fish and Game Commission, Fishing Tech. Bull. No. l. 114 pp. Chapman, G. A. 1965. Effects of sub-lethal levels of pentachlorophenol on the growth and metabolism of a cichlid fish. M.S. Thesis. Oregon State Uni- versity, Corvallis. 77 pp. Cole, R. A. 1973. Stream community response to nutrient enrichment. Jour. Wat. Poll. Cont. Fed. 45(9): 1874-1888. Cummins, K. W. 1972. Personal communication. Cummins, K. W. and D. King. 1972. Personal communication. Daniels, S. L., L. L. Kempe, T. J. Billy, A. M. Beeton. 1965. Detection and measurement of organic lam- pricide residues. Great Lakes Fishery Commission, Tech. Rept. No. 9. 18 pp. Davis, G. E. and C. E. Warren. 1965. Trophic relations of a sculpin in laboratory stream communities. Jour. Wildl. Mgmt. 29: 846-871. . 1968. Estimation of food consumption rates. In: Methods for Assessment of Fish Production in Fresh Waters, I.B.P. Handb. No. 3, ed. W. E. Ricker. Oxford: Blackwell. 313 pp. Dawson, V. K. 1971. Photolytic sensitivity of TFM. in: Progress report for 1971 on registration-oriented research on TFM used in control of the sea lamprey. Fish Control Laboratory, LaCrosse, Wisc. DeBrouwer, S. 1930. Sur l'acid orthotrifluortoluique et 1e nitrotrifluorcresol 1-3-6. Bulletin Soc. Chim. Belg., 39: 298-308. 157 Dykstra, W. W. and R. E. Lennon. 1966. The role of chemi- Dymond, cals for the control of vertebrate pests, p. 29-34. In: E. F. Knipling (Chrm.), Pest control by chemi- cal, biological, genetic, and physical means. A Symposium. U.S. Dept. Agric., ARS 33-110. J. R. 1922. A provisional list of the fishes of Lake Erie. Publ. Ont. Fish. Res. Lab., 4: 55-74. Ebel, W. J. 1962. A photoelectric amplifier as a dye de- Ellis, Ellis, Elson, tector. Great Lakes Fishery Commission, Tech. Rept. No. 4. pp. 19-26. R. H. 1968. Effects of Kraft pulp mill effluent on the production and food relations of juvenile chinook salmon in laboratory streams. M.S. Thesis. Oregon State University, Corvallis. 55 pp. R. J. and H. Gowing. 1957. Relationship between food supply and condition of wild brown trout, Salmo trutta Linnaeus, in a Michigan stream. Limnol. and Oceanogr. 2(4): 299-308. P. F. 1967. Effects on wild young salmon of spray- ing DDT over New Brunswick forests. Jour. Fish. Res. Ed. Canada 24(4): 731-767. {Erkkila, L. F., B. R. Smith, and A. L. McLain. 1956. Sea lamprey control on the Great Lakes - 1953 and 1954. U.S. Fish and Wildl. Serv., Spec. Sci. Rept.-Fisher- ies No. 175. 27 pp. Gage, S. H. 1893. The lake and brook lampreys of New York, Gibson, especially those of Cayuga and Senaca Lakes. In: Wilder Quarter Century Book. Ithaca. pp. 421-493. H. R. and D. W. Chapman. 1972. Effects of Zectran insecticide on aquatic organisms in Bear Creek, Idaho. Trans. Amer. Fish. Soc. 101(2): 330-344. Great Lakes Fishery Commission. 1960-1972. Annual reports of Great Lakes Fishery Commission, Ann Arbor, Michigan. Haas, R. C. 1970. The effects of lamprey larvicide on the bottom fauna and periphyton of the Chocolay River, Marquette County, Michigan. Michigan Dept. Natural Resources, Research and Development Rept., No. 200: 1-66. 158 Hamilton, A. L. 1969. On estimating annual production. Limnol. and Oceanogr. 14(5): 771-782. Hanson, L. H. 1970. Prospects for the integrated control of the sea lamprey. Great Lakes Fishery Commission, Interim Meeting, Ann Arbor, Mich., Dec. 1-2, 1970. Hoffman, C. H. and E. W. Surber. 1945. Effect of an aerial application of wettable DDT on fish and fish food organisms in Back Creek, W. Va. Trans. Amer. Fish. Soc. 75: 48-58. Hoffman, C. H. and A. T. Drooz. 1953. Effects of C-47 air- plane application of DDT on fish food organisms in two Pennsylvania watersheds. Am. Midland Naturalist 50(1): 172-188. Hooper, F. F. 1969. Eutrophication indices and their rela- tion to other indices of ecosystem change. In: Eutrophication: Causes, Consequences, Correctives. National Acad. Sci., Washington, D. C. (1969). Howell, J. H. 1970. Personal communication. Howell, J. H. and W. M. Marquette. 1962. Use of mobile bioassay equipment in the chemical control of sea lamprey. U.S. Fish and Wildl. Serv., Spec. Sci. Rept.-Fisheries. No. 418. pp. 1-9. Huang, Ju-Chang and E. F. Gloyna. 1968. Effect of organic compounds on photosynthetic oxygenation I. Chloro- phyll destruction and suppression of photosynthetic oxygen production. Wat. Res. 2: 347-366. Hubbs, C. L. and T. E. B. Pope. 1937. The spread of the sea lamprey through the Great Lakes. Trans. Amer. Fish. Soc. 66: 172-176. Hynes, H. B. N. and T. R. Williams. 1962. The effect of DDT on the fauna of a Central African stream. Ann. Trop. Med. Paristol. 56(1): 78-91. Hynes, H. B. N. and M. J. Coleman. 1968. A simple method of assessing the annual production of stream benthos. Limnol. and Oceanogr. 13: 569-573. Ide, F. P. 1968. Effects of pesticides (toxicants) on stream life. In: Developments in Industrial Micro- biology. Amer. Inst. of Biol. Sci. Washing, D. C., Vol. 9, pp. 132-145. 159 Johnson, B. G. H. 1970. Water quality and biological activ- -ity of lampricide, Appendix 20, p. 90-93. In: Great Lakes Fishery Commission, Report of annual meeting, St. Paul, Minn., June 16-18, 1970. Kempe, L. L. 1973. Microbial degradation of the lamprey larvicide 3-trifluormethyl-4-nitrophenol in sediment- water systems. Great Lakes Fishery Commission, Tech. Rept. No. 18. pp. 1-16. Kevern, N. R. 1962. Primary productivity and energy rela- tionship in artificial streams. Ph.D. Thesis. Michigan State Univ., 132 pp. Kevern, N. R. and R. C. Ball. 1965. Primary productivity and energy relationships in artificial streams. Limnol. and Oceanogr. 10: 74-87. Lech, J. J. and N. V. Costrini. 1972. In vitro and in vivo metabolism of 3-trifluormethyl-4-nitrophenol (TFM) in rainbow trout. Comp. and Gen. Pharmac. 3(10): 160-166. Lennon, R. E. 1971. Research on lampricides. Presented at the annual meeting of the Great Lakes Fishery Commission, Toronto, Ontario, June 23-25, 1971._ MacArthur, R. H. 1965. Patterns of species diversity. Biol. Rev. Cambridge Phil. Soc. (G.B.), 40: 510. Maki, A. W., L. W. Geissel, and H. E. Johnson. 1974. Tox- icity of TFM (lampricide) to 10 species of algae. U.S. Bur. or Sport Fisheries and Wildlife. In- vestigations in Fish Control (in review). Marking, L. L. 1971. Toxicity of TFM to fish. In: Prog- ress report for 1971 on registration-oriented re- search on TFM used in control of the sea lamprey. Fish Control Laboratory, LaCrosse, Wisc. McIntire, C. D., R. L. Garrison, H. K. Phinney and C. E. Warren. 1964. Primary production in laboratory streams. Limnol. and Oceanogr. 9: 92-102. McIntire, C. D. and H. K. Phinney. 1965. Laboratory stud- ies of periphyton production and community metabolism in lotic environments. Ecol. Monog. 35: 237-258. Odum, H. T. 1956. Primary production in flowing waters. Limnol. and Oceanogr. 1: 102-117. 160 Odum, H. T. and C. M. Hoskins. 1957. Metabolism of a laboratory microcosm. In: Publications of the Institute of Marine Science, Vol. 5, University of Texas. pp. 16-46. Paulson, T. L. and D. C. Culver. 1969. Diversity in terrestrial cave communities. Ecology 50: 153. Pycha, R. L. 1970. Lake Trout mortality in relation to lamprey predation. U.S. Bur. Sport Fish. and Wildl. Mimeo. 8 pp. Pycha, R. L. 1972. Sea lamprey wounding rates on lake trout, U.S. Waters of Lake Superior, Sept. 1972. Great Lakes Fishery Commission. Interim meeting, Ann Arbor, Mich., Dec. 5-6, 1972. pp. 1-6. Reese, H. W. 1966. Physiological ecology and structure of benthic communities in a woodland stream. Ph.D. Thesis. Oregon State University, Corvallis. 134 pp. Reinert, R. E. 1972. Accumulation of dieldrin in an alga (Scenedesmus obliquus), Daphnia magna and the guppy (Poecilia reticulata). Jour. Fish. Res. Ed. Canada 29: 1413-1418. Rudd, R. F. 1964. Pesticides and the living landscape. Univ. of Wisconsin Press, Madison, Wisc. 320 pp. Schnick, R. A. 1972. A review of the literature on TFM (3-trif1uormethy1-4-nitrophenol) as a lamprey larvicide. U.S. Bur. of Sport Fisheries and Wildl. Investigations in Fish Control, No. 44. 31 pp. Seim, W. K. 1970. Influence of biologically stabilized Kraft mill effluent on the food relations and production of juvenile chinook salmon in labora- tory streams. M.S. Thesis. Oregon State Univers- ity, Corvallis. 55 pp. Shapiro, J. 1957. Chemical and biological studies on the yellow organic acids of lake water. Limnol. and Oceanogr. 2: 161-179. 1958. Yellow acid-cation complexes in lake water. Science 127: 702-704. Shetter, D. S. 1949. A brief history of the sea lamprey problem in Michigan waters. Trans. Amer. Fish. Soc. 76: 160-176. 161 Slobodkin, L. B. and H. L. Sanders. 1969. On the contribu- tion of environmental predictability to species diversity, pp. 82-95. In: Diversity and stability in ecological systems. Brookhaven National Labora— tory, Upton, New York. Smith, B. R. 1966. Lamprey control and research in the United States, p. 31-45. In: Great Lakes Fishery Commission, Annual Report 1965, Ann Arbor, Michigan. Smith, A. J. 1967. The effect of the larval lampricide, TFM, on selected aquatic invertebrates. Trans. Amer. Fish. Soc. 96: 410-413. Smith, M. A., V. C. Applegate, B. G. H. Johnson. 1960. Colorimetric determination of halogenated nitro- phenols added to streams as sea lamprey larvicides. Anal. Chem. 32: 1670-1676. Steele, R. G. D. and J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill Book Co., Inc. 481 pp. Surface, H. A. 1897. The lampreys of central New York. Bull. U.S. Fish. Comm., 1897 (1898) 17: 209-215. Torblaa, R. L. 1968. Effects of larval lampricides on invertebrates in streams. U.S. Fish and Wildl. Serv., Spec. Sci. Rept.-Fisheries, No. 572. 13 pp. U. 8. Bureau of Commercial Fisheries. 1958. Lamprey con- trol and research in the United States, p. 34-50. In: Great Lakes Fishery Commission, Annual Report 1958, Ann Arbor, Michigan. Warren, C. E. 1971. Biology and water pollution control. Philadelphia: Saunders. 434 pp. Warren, C. E. and G. E. Davis. 1971. Laboratory stream research: Objectives, possibilities and constraints. Ann. Rev. of Ecolog. Systemat. 2: 111-144. ‘ Waters, T. F. 1964. Recolonization of denuded stream bottom areas by drift. Trans. Amer. Fisheries Soc. 93: 311-315. Waters, T. F. and G. W. Crawford. 1973. Annual production of a stream mayfly population: A comparison of methods. Limnol. and Oceanogr. 18(2): 286-296. 162 Whitley, L. S. and R. A. Sikora. 1970. The effect of three common pollutants on the respiration rate of tubificid worms. Jour. Wat. Poll. Cont. Fed. 42: 57-66. Whitworth, W. R. and T. H. Lane. 1969. Effects of toxicants on community metabolism in pools. Limnol. and Ocean- ogr. 14: 53-58. Wilhm, J. L. 1970. Some aspects of structure and function of benthic macroinvertebrate populations in a spring. Amer. Midland Naturalist 84: 20. Wilhm, J. L. and T. C. Dorris. 1968. Biological paramenters for water quality control. Bioscience 18(6): 477- 481. Williams, H. 1969. Effect of Kraft mill effluent on struc- ture and function of periphyton communities. Ph.D. Thesis. Oregon State University, Corvallis. 117 pp. Woodwell, G. M. 1967. Toxic substances in ecological cycles. Sci. Amer. 216(3): 24-31. "I11111111111111.7113