TPWD 1960 F-7-R-8 #569: Job Completion Report: Experimental Control of Undesirable Fish Species in Lake Diversion, Project F-7-R-8

JOb Completion Report State of TEXAS Project No. F-7nR-8 . Name: Fisheries Investigations and Surveys of the waters of Region 1-B. JOb No. E-2 Title: Experimental Control of Undesirable - Fish Species in Lake Diversion. Period covered: January 12 1960 - Denember 31, 1960 OBJECTIVES To determine the practical application and effectiveness of methods developed under Job Eel (Experimental Control.of Undesirable Fish Species). Specifically, Objectives of the work covered by this report were to determine the effects of the selective~kill treatment daring March, 1957. TECHNIQUES Beginning nine months prior to the selective—kill treatment of Lake Diversion, monthly gill net and seine cellections were made. Six netting stations were selected from different areas of the lake, ranging in depth from three to thirty feet. These stations were each netted with approximately 225 feet of gill net ranging in mesh size from one to three inches. All fish taken in these gill nets were weighed, measured, and sexed. Game fish stomachs were Opened and inspected for food_contents. Seine samples were collected with 20-foot minnow seines and were preserved for laboratory identification and counting. These techniques, which were established in 1956, have been standard throughout the extent of Job E-2. BACKGROUND INFORMATION- For background information concerning this report, readers are referred to the following job completion reports: _ - ' ' ' F—7—R—h, Job E-2‘ ' F-7uR-5,'J0bS'E—2 and E—l F-7—R~6, Job E-2 FeT—R-7, Job E-2 ' The basic inventory and survey of Lake Diversion (June 1953 — May 195k) showed a fish population consisting of 73 percent rough fish. Thirty percent of the popuw lation was composed of gizzard shad alone. Freshwater drum, although constituting a small percentage of the gill net catch, were very abundant as evidenced by a rote~ none check in a small bay in 1953. This ten acre bay was treated with enough rote~ none to effect a total kill, and all fish killed were collected. 0f the h,136 fish killed, 2,601 or 63 percent were freshwater drum. ' In July 1956, monthly net and seine collections were begun. Netting stations -2- were chosen and techniques were established from which comparative data could be col« lected before and after a selective rotenone kill. In March 1957, the lake was treated with 10,320 pounds of five percent rotenone powder which was distributed by a large unit mounted on a barge, and smaller boat-mounted unit. Immediately following the kill, shoreline counts and estimates of total numbers of fishes killed were made. An esti- mated minimum of 185 tons of rough fish, consisting mostly of shad and drum, were killed. Since game fish species constituted only a small percentage of the total number killed, the treatment was considered successful. Continued netting and seining FINDINGS Fish Collections Ten gill net collections, made during this segment, produced a total of 1,153 fish. Monthly collections produced an average of 115 fish per trip, and ranged from in. fish in February to 162 fish in September. During the hot months of July and August and the cold months of October, November, January and February, the catch was less than during the remaining months. Monthly totals of fish taken by gill nets are given in Table l. The catch of gizzard shad showed an unexpected decrease of 2.59 percent over last year. Longnose gar, shortnose gar, and carp also decreased, however smallmouth buffalo and river carpsucker increased slightly. The total percentage of rough fish was 81.97 percent, a decrease of 2.75 percent over last year. Game fish, except for white crappie, all showed from slight to good increases, with white bass having the best gain (5.h5 to 7.90 percent). Table 2 gives percent- age cOmposition, sex ratios, and average weights by sex. Percentage composition by'weight changed slightly during this segment. Longnose gar, smallmouth buffalo, and carp decreased while gizzard shad, river carpsucker, and freshwater drum increased slightly. Game fish all showed increases in percent of total weight. The total increase for all game fish c0mbined was about four percent above last_year's total. Gizzard shad, redhorse shiners, bullhead minnows, mosquitofish, and bluegills were the five more abundant species taken by seine. Common in seine samples were river carpsuckers, plains shiners, sand shiners, Red River pupfish, Red River shiners, white bass, white crappie, and spotted sunfish. The 22 other species found in our samples were uncommon or rare, and none comprised more than one percent of the total number taken. Table 3 gives the number and percent of each species taken by seine. Table h is a checklist and gives the total number of fish taken by both _gill nets and seines. Very few diseased or abnormal fish were taken during this segment. An oc- casional channel catfish had small leeches around its mouth or fins. Except for white crappie, largemouth bass, and channel catfish being parasitized'by body- cavity nematodes, only three other abnormal fish were caught. One longnose gar had no left eye. This fish's "KT factor was normal, so its handicap apparently did not E E.» B restrict its feeding. One gissard shad had a portion of its caudal peduncle missing, and one river carpsucker had a deformed tail. Both of these fish also had normal "K” factors. Food Habits Food habits data from some of the game fish collected on Job B~l5 were included in the compilation of the food habits chart included in this report {Table 5). irty different food items were noted, but only four were common. Gissard shad, unidentifi- able fish remains, mayfly nymphs, and grasshoppers were the items most often found. Channel catfish prayed mostly upon fish, mayfly nymphs and grasshoppers. Crayfish, fly larvae, plant material, and grain were also fairly common. Only one flathead catfish had food in its stomach; it had eaten a giszard shad. White bass preferred gizaard shad, mayfly nymphs, and unidentifiable fish. The stomach of one white bass examined contained a small rock, and nothing else. Both largemouth bass and white crappie fed mostly upon gimsard shed and other fish. Table 5 gives the food items of each species of game fish, the frequency of occurrence, and the total number identified, as well as the frequency of occurrence and total number identified of each food item from all game fish combined. Coefficient of Condition In general, "K” factors were very near the same as last year. Male freshwater drum, however, dropped sharply from 2.9 last year to 2.3 this year. The average "KN factor of female drum also dropped, bot to a lesser extent (from 2.9 to 2.7). Considering males and females separately, there were nine decreases and seven in» creases in ”K" factors,'while ten remained the same as last year. Table 6 gives the distribution of "K” factors by sea for each species. Table 7 gives a comparison of "K? factors by sex from 1956 to l96d. SsssainD q eve if} essst.andw8psrsis; Secesss Seine samples indicated good spawns of gizzard shad, white bass, carp and river carpsucker. The number of small channel catfish taken in August indicates a fair spawn of this species. Shad began spawning in late May or early June, and by later June, 89.#7 percent of the adult females taken had spawned. Buffalo, carpsucker, channel catfish, and white bass also began spawning in late May or early June, but crappie and carp spawned in July and Angoet. firemendous spawns of longnose gar occurred on the 25th and 27th of May, when hundreds of these fish deposited countless thousands of eggs along two stretches of rocky shoreline on the south side of the lake. It was interesting to note that smallmonth buffalo taken in the spawning area were found to have been feeding on the gar eggs. All fish taken in gill nets were opened and the stage of gonadal development was recorded. Subeadult fish were recorded as ”GM“ or "HF“ (undeveloped males or undevel- oped females). Fish in which the gonads contained immature eggs were classified as immature = "IM" or ”IF”. Fish ready to spawn were recorded as ripe w ”RM” or "RF", and fish taken shortly after they had spawned were classed as spent e "SM” or "SF". Physical Characteristics Physical conditions recorded at the time of each collection included water and -1... air temperatures, wind speed and direction, and weather conditions. Surface water temperatures ranged from a low of #1 on February 26, to a high of 95 on July 27. Air temperatures ranged from 28 to 105 on the same dates. water clarity at Lake Diversion varied with conditions. Following high winds, the lake was quite turbid, but it usually cleared up within two or three days after the wind subsided. Table 8 gives the physical conditions on collection dates. DISCUSSION When Lake Diversion was treated, the use of rotenone powder as a selective fish toxicant was relatively new. Methods and machinery for the distribution of rotenone powder were not as efficient as they are at present. Bigger machines capable of mixing and dispensing 2000 pounds of powder per hour are now available, and these machines can easily be operated by'a crew of three men. They are lightweight, portable, and allow a minimum of contact of crew with the powdered chemical. Efficient treating methods developed since 1957, have cut the cost of treating with powder, and have increased its effectiveness. When considering the results of the Lake Diversion kill, 'with thoughts of comparing cost with benefit gained, it must be kept in mind that new methods, less expensive chemical, and greater capacity of machines now in use would permit a better and more economical treatment than was accomplished at that time. It must also be remembered that there were two sources of contamination from which shed and drum could get back into the lake. The treatment of these sources would have prolonged the period of time it took shed and drum to increase to their original numbers. The Lake Diversion jdb was primarily experimental, and was completed in order to determine relative changes in the fish population. The effect of the treatment on fishing success has not been studied.' Creel censuses.have never been made.at Lake Diversion, and accurate data concerning changes in fishing success following the kill are not available. However, reports from cabin owners, concessionaires, and fisher— men state that the fishing did improve after the rotenone treatment. These reports are in accordance with the data collected during the year following the kill. The removal of thousands of pounds of shed and drum (basic food items of-game fish in Diversion) should naturally have made the game fish easier to catch. An unexpected benefit to fishermen was the apparent removal of mayfly nymphs'by the rotenone. This left game fish without two of their three most important food items and forced them to spend more time searching for food. Thus, fishing was improved. Since Job E-2 will be terminated this segment, and this will be the final report on this jab, the following discussion will include a brief comparison of data cola lected during all of the five segments that this jab has been in effect. For cone venience the discussion will be divided under six headings. The Effect of the Selective Rotenone Kill on the Percentage Composition of the Fish Population_ ' . The relative abundance of gizzard shad prior to the treatment was 36.6 percent as determined from gill netting during the nine months before the treatment. During the segment following the treatment, the figure was down to 6.8 percent, or approximately an 80 percent reduction of the shed. The following year the shed increased to 22.52 percent, or up to about 60 percent of the original number. In 1959, they increased m5... an additional 9.12 percent or up to about 85 percent of the original number. In 1960, the percentage of shad dropped to 29.05 percent. Thus, it took about three years for shad to repopulate the lake. Except for shad and drum, other rough fish were not greatly affected by the selective kill, and while their relative abundance increased, this was largely mathe- matical and not an actual increase in numbers. The same thing was true of game fish. Their relative abundance increased somewhat, but again this was mostly a mathematical increase rather_than an actual increase in numbers. Table 9 gives percentage composin tion grom 1956 to 1960. Table 10 gives the percentage composition by weight from l956 to 19 O. The Effect of the Selective Rotenone Kill on the Physical Condition of the Fish Pepulation The greatest effect of the kill, other than changing the relative abundance of species, was the effect upon the physical condition of the fish. It would seem that with less food available following the kill, game fish would lose weight, but such was not the case. Following the kill the "K? factors of all species of game fish in“ creased. Increases in "K” factors of white bass, largemouth bass, and crappie were _ exceptionally greater. Drum and shed which survived the kill also showed increases in "K". Less crowded conditions or changed food habits prdbably accounted for the increases. As "K" factors increased, average weights increased. Fish were bigger and in better condition and they remained this way until shad began to increase. There was an inverse relationship between ”K" factors and average weights and the shed population. Tables 7, 9 and ll clearly show this relationship. A large shad populam tion is apparently not only detrimental to fishing, due to the abundance of food avail- able, but also the physical condition of game fish. The Recovery Capacity of Shad and the Changes in the Fish Population due to Reinfestation by Shad Tremendous reproductive potential enables shad to spawn off thousands of fry each summer. Predation on shed is greatest during the period when young shad are schooling in shallow water, but many survive this period and grow large enough to prevent any'butr the larger predators from eating them. This is evidenced from the increase in shad from 1957 to 1958 when they increased from 7 to 23 percent of the population. Increases in shad during 1958 to 1960 raised the population to about 30 percent. At this point game fish were nearing the conditions, both in numbers and in physical condition that were prevalent before the treatment. The factors which control the shed population are not known, but something causes shed in Lake Diversion to "level off" at about 30 percent. Competition for food and space may be limiting factors. Predation may exert a measure of control on the shed population, however, predation alone cannot be too important, otherwise it would seem that a shed increase from 7 to 30 percent within three years would have been impossible. The sources of recontamination of Diversion (seven miles of river above the lake and 52 miles of main irrigation canal below the lake) prObably contributed to this rapid increase of shad. It should be noted that after almost four years have passed since the treatment, and as shad have become abundant again, the game fish population is still in better condition than it was before the selective kill, and the benefits of the kill have not entirely diminished. The Inaccurac_ of Gill_Nettin;_in Determining the” Relative Abundance of Freshwater Drum -6- Unfortunately, the inability of gill netting to show the relative abundance of drum has left a blank space in our data. Although an estimated 100 tons of drum were killed, the percentage of drum taken in gill nets following the kill did not drop _ appreciably. However, since not more than 1.50 percent drum were taken by nets during any segment before or after the treatment, gill netting is Obviously totally unreliable for establishing the relative abundance of this species using the netting techniques employed on this job. It would have been interesting to compare the changes in the drum population following the treatment with changes in the population of other species, but this was not possible. Due to special conditions occurring in the upper reaches of Lake Diversion last summer, large numbers of drum were taken on two Jab Bnl5 gill net collections. On the basis of these two oceasions, it is believed that drum are again abundant in Lake Diversion. The Effect of the Selective Kill on Water Quality and Aguatic Organisms Lake Diversion is ordinarily moderately clear, but the rotenone treatment resulted in increased clarity, especially in shallow areas. This "clearing up? of the water was a normal occurrence that usually follows with the removal of shad, drum, carp, and other bottomwfeeding species. It is also possible that many zooplankters were killed which could have contributed to the increased water clarity. As previously stated, the rotenone affected the bottom fauna. Mayfly nymphs were apparently greatly rem duced, and bottom sampling would prdbably have shown that other bottom organisms such as damselfly larvae, dragonfly larvae, and chironemid larvae were also affected. The Length of Time that the Benefits of the Rotenone Treatment Lasted The extent that fishing was improved by the selective kill, and how long good fishing continued cannot be accurately stated because a creel census was not made, however, fishing did improve for a while. Continued good fishing could probably be attained only by repeated treatments at intervals of two or three years depending on conditions. This would periodically eliminate most of the young shad, and create more space for game fish. Treatment of lakes as early in the fall as conditions would per— mit should give the best results, because the small shad spawned during the previous summer which are the only shad small enough to be eaten by the average sized game fish, and the ones most susceptible to the effects of rotenone would be eliminated. Assuming that a kill could be done in October, it would remove small shed for a period of at least seven months, or until surviving shad spawned during the next summer. Spring treatment, on the other hand, would eliminate small shad for a period of only three to four months (from the treatment date until about June), and after June, shad would again be present as an abundant source of food for game species. Prepared by Lonnie J. Peters Approved'by 2 : :siLskizveyLr.L4’é*1¢“{i::f Assistant Project Leader Director Inland Fisheries Division Date April 6, 1961 Table 1. Monthly gill net collections, Lake Diversion, 1960 77,7:387_MB:_73777 iw-1_ June .. _u: Species § Fish Percent- Fish Percent ' _- . Otal H081 Longnose gar 3.12 Shortnose gar m . 1.88 Gizzard shad "g a no.62 Smallmouth buffalo E g 22.50 'Bigmouth buffalo 8 E 0.63 River carpsucker ,3 .2 13.12 Channel catfish 5.; g 3.12 Black bullhead Si no 0 Flathead catfish 8 E, 0.63 White bass 0 a 3.12 Largemouthfibass :3 '5 0.63 White crappie To 50 Freshwater drum 0 Carp 1' _d 3 l3 II-I oo 7 ' —-7 7 Totals ' Species Percent Percent Percent Fish Percent of t_ota1 of total of total of total of total ___ of total_ Longnose gar 1.73 Shortness gar 0.52 Gizzard shad 29.05 Smallmouth buffalo 22.21 Bigmouth.buffalo 0.h3 River carpsucker 22.12 Channel catfish 2.60 Black bullhead 0.09 Flathead catfish 0.60 White bass 7.90 Largemouth bass 0.69 White crappie 6.2M Freshwater drum 1.05 #077 ”81“ Total I - 100 00 101 I 100...; 100.00 100: 00 - 100:00 11,3 . 100.00 Table 2. Results of gill netting, Lake Diversion, 1960. 7" 77Fish77'”77 77weight “or fish 7 7 7 Males 7" 7""”'-' 7 7"77 Females7 collected collected _ _ . Numbers Weights 77' Numbers Weights Species 7 Number Percent Weights Percent 7 Fish Percent Pounds Average Fish Percen- Pounds Average _ total counds‘weight e . . tounds .0und: Longnose gar Shortness gar 1.9’ Gizzard shad 0-73 Smallmouth buffalo 2.92 Bigmouth buffalo h.6( River carpsucker 1:52 Carp 3.9( Channel catfish 2o3f Black bullhead - Flathead catfish 3.7( White bass 0.81 Largemouth bass 2.2' White crappie 0°55 Freshwater drum 3.21 -9... Table 3. Results of seining, Lake Diversion, 1960. Species Number seined Percent of total Shortness gar 3 .01 Longnose gar .01 Gizzard shad 30.51 Smallmeuth buffalo .29 River carpsucker 1.35 Carp .07 Golden shiner . .01 Suckermeuth.minnow .58 Plains shiner 1.12 Sharpness shiner .58 Red River shiner " 1.01 ArkanSas River shiner .30 Redhorse shiner 18.58 Sand shiner 1.38 lMimic shiner .23 Ghost shiner .Oh Silvery minnow .01 Plains minnow '.28 Bullhead minnow 9.66 Channel catfish .22 Black bullhead .16 Plains killifish .05 Red River pupfish 1.30 .Mosquito fish 13.81 White bass 2.63 Black base .51 Green sunfish .12 Spotted sunfish 2.51 Redear sunfish .48 .Bluegill sunfish 8.60 Orangespetted sunfish .89 Longear sunfish .17 White crappie 1.19 Legperch .hT Freshwater drum .87 Totals 18,768 100.00 1101 Table h. A checklist and total number 0f fishes taken by gill nets and seines from January 1, 1960 a December 31, 1960 _Common name Scientific Name Shortness gar Lepisosteus platostomus Longnose gar . L. osseus ' ' Gizzard shad fibrosoma cepedianum Bigmouth'buffalo Ictiobus oyprinellus Smallmouth buffalo ;, bubglus River carpsucker Carpigdesc carpig Carp Ciprinus oarpio Golden shiner Notemigonus crysoleucas Suckermouth minnow PhenaCObius mirabilis ' Plains shiner fl otrgpw percobromus Sharpnose shiner .fl. oxxrhznchus Red River shiner N bairdi Arkansas River shiner afi. girardi Bedhorse shiner 'fi. lutrensis. Sand shiner E} stramineus Mimic shiner ' E.voluce11us Ghost shiner y; buchanani " Silvery minnow Hybognathus nuchalis Plains minnow H. placita Bullhead minnow W' Milan Channel catfish Ictalurus punctatus Black bullhead , ‘E. males Flathead catfish Pilodietus olivaris Plains killifish Fundulus kansse Red River pupfish Cxprinodon rubrofluviatilis Mosquito fish Gambusia affinis White bass Roccus chrxsops Largemouth bass -lMicropterus salmoides Green sunfish Lepomis cyanellus Spotted sunfish is punctatus Bedear sunfish lo mierolophus Bluegill sunfish £9 ma acrochirus orangespotted sunfish . g, h umilis Longear sunfish ' in megalotis White crappie Pomoxis annularis LogperCh Percina caprodes Freshwater drum é Aplodinotus grunniens Totals Number“ 22 6, 062 311 509 109 209 110