TPWD 1956 F-7-R-4 #239: Inventory of Fish Species Present in Buffalo Lake

SEGMENT REPORT State of TEXAS Project No.__flF7Rh. Name: Eigheries Investigations and Surveys of the Waters of Region _l_n_in_ l-B, iim__m__n__n_n___nlnn__nl_n__i__ . ..... n. Job No, .n. B~9 Title: Enventorymgf Fish Species Present in Buffalo Lake (Contine m ued from ELF”R”3L “w PeriOd comma: {332i 1956 Ediflhdh fiXfianflw ........... m. ...... M ............ m. M. ABSTRAQE: An estimated minimum of 200 tons of fish, mostly gizzard shed and carp, has been removed from Buffalo Lake by experimental selective-kill and spot treatments, seining and gill netting, since May 1956, Details of the treatments and a discussion of factors involved are given. Inventory has been taken by gill net and seine collections before and after treatments in order to compile comparative data which may indicate possible effects of treatments, A total of 2,108 fishes was collected by gill nets frem which data concern» ing food habits, sexual development, spawning success, coefficient of condition and pathological conditions was recorded, and as far as pessible, reduced to tabular form. OBJECTIVES: To determine the fish species present and their relative abundance, as well as to determine the ecological factors influencing their distribution. To discover both the immediate and progressive changes which may occur in the fish pepulation as the result cf experimental selectivenkill treatments under JOb E—l. ACKNOWLEDGMENTS: Dro Welter Dalquest, whose employment with the Texas Game and Fish Commission terminated in October 1956, conducted laboratory work and compiled field collected data for the job during the first five monthso He has also checked the manuscript of this reporto Dr, Carl Gray, Soils Scientist of Midwestern University and Wichita County Water Improvement Districts, made all of the chemical water analyses during the study, ' TECHNIQUES: The work on this job started during the preceding segment period (June 1955 through may 1956) and continued through the present segment period, An attempt was made to continue monthly collections so that data comparable to the preceding segment could be obtained“ However, high winds, frozen surface water and other conditions beyond our control prevented collections during the months of July, September and November 1956, and February and March 1957. Approximately 1200 feet of gill net, ranging in size from 1 inch to 3 inch (square mesh) were set on each visit. 'Each fish taken in these nets was measured, weighed, and the stage of gonadal development record- edo In addition, filled stomachs of predacious species were saved in formalin, as well as ripe ovaries, pathological tissues and parasites. Samples of forage fishes were obtained with the use of small—mesh seines, and the complete collections preserved in formalin, In the laboratory, formalin-preserved materials were identified, examined and the data recorded, In addition to collections made with gill nets and minnow seines, supplemental data was obtained by the use of a 1200 foot drag seineo Although the use of this seine was limited by stumps, rocks, trotline stakes and other obstructions, it produced some interesting results and removed large quantities of undesirable fish, Physical data, including air temperature, waterwsurface temperature and turbid- ity were recorded at the lake. Samples of lake water were saved and the pH and dissolved solids present were determined, Ecological data pertaining to the lake were obtained in the usual ways and detailed notes were taken, Data obtained during the last segment (Fw7—R-3) is recorded in two separate reportsa An Inventory of the Species Present in Buffalo Lake is reported as JOb B99, and Laborato and Statistical Analyses of Materials and Data Collected in the Field is reported as Jeb B— a For the sake of convenience and simplicity, however, all the data obtained from Buffalo Lake during this segment (F-7—R—h) is included in this report. Information reported in this manner should make it much easier to detect any possible results of experimental treatments under JOb E-l, as Well as to permit the interpreta- tion of the value of such treatments, FISH COLLECTIONS: .A total of 2,108 fishes was collected by gill nets from Buffalo Lake during the segment period (Segment h) from JUne 1, 1956 to may 31, 19579 Percentage composition, sex ratios and average weights by sex of fishes collected by gill nets for Segment h are given in Table 2, Identical type information for the preceding segment (June 1, 1955 to may 31, 1956) is given in Table 1 for comparisono Percentage composition by weight and mean weights of fishes collected by gill nets during 1955—1956 and 1956-1957 is given in Tables 3 and h respectively, The total number of forage fishes taken from Buffalo Lake during the present segment with small mesh minnow seines is given in Table 50 The total number of fishes taken during both segment periods is given in Table 6, Not included in Table 6 is the number of fishes taken by the 1200 foot nylon drag seine, EXPERIMENTAL SELECTIVEmKILL TREATMENTS; Following 11 months of inventory and study, Buffalo Lake was treated in may 1956, for an experimental selectivemkill of gizzard shad and, incidentally, carp and goldfish, Details of this treatment and the immediate results is given in the report for Job E—l, F—TmRe-u3o A concentration of .35 pounds of 5% rotenone to the acre-foot of water (.13 pop.ms) was distributed when the temperature of the lake was 18°C (65¢h0F)o The shore— line count which was made on the third day following treatment estimated the total fish killed as follows: Total weight of EurOpean carp killed, 63,365 lbs. 31 tons Total weight of gizzard shad killed, 67,920 lbs. 3H tons “ Total weight of carp and shad killed, 131,285 lbs. 65.5 tons Percent of fish killed, other than carp and shad. 2.2 percent Percent of game fish killed. 1.8 percent removed by spectators and bait dealers on the first'mmadays,or the large numbers of fishes that floated in to shore after the shoreline count was made. The count was made on the third day, and fish (mostly shad) were still drifting in when the fisheries crew left the lake, a week later. The concessionaire, who was in charge of the crew that finished the cleanup, estimated that a minimum of 100 tons of fishes were killed in the treatment. ‘ On April 23, 1957, Buffalo Lake was treated again for a selective—kill on shad. Rotenone powder (5%) was used at a concentration of .3 lbs/acre foot (.1 p.p.m.), when the lake temperature was 56°F. Excellent selectivity on shad resulted. The number of carp killed was considerably less than during the past treatment, but there was a great reduction in the number of game fishes killed also.~ Within an hour following treatment, large numbers of shad began to work the surface in a moribund condition. After death, most of the shed (and some of the carp) sank to the bottom where they apparently decomw posed without rising to the surface. Since most of the fish that were killed by the treatment did not float before the fisheries crew left the lake, a total kill estimate was not possible. EXPERIMENTAL SPOT TREATMENTS On June 8, 1957,* a concentration of spawning carp and goldfish was located at , the upper end of Buffalo Lake. This Spawning activity was stimulated by recent rains which caused the lake to rise over terrestrial weeds that had grown in the dry lake bottom. Both carp and goldfish utilized this green vegetation for spawning, but they seemed to prefer dry tumbleweeds that had blown into the spawning area from surrounding fields. Although it was not neticed until after treatment, there were almost as many shad in the spawning area as carp and goldfish. Rotenone was distributed at a concentration of approximately 1% lb./acre foot of water in the spawning area only. Since most of the water was too shallow to float a loaded boat, rotenone was broadcast by hand. The spawning fish paid no attention to workers distributing the toxicant. After the workers passed over an area, Spawning activities ceased immediately, and all fish except shad remained calm from 10 to 15 minutes before surfacing and gasping for air. Shad surfaced and died immediately where- as carp and goldfish continued to die for about 4 hours. ' The area was checked early the following morning. Complete kills were apparently affected on the very active parent fishes, as well as fry from earlier Spawns, and eggs that were in advance stages of develOpment. Microscopic examination of freshly spawned eggs revealed that they were not affected.by the treatment, however, and the retenone either dissipated very rapidly, or was removed to the lake proper by wind currents be- fore the eggs develOped to the stage at which rotenone would be toxic. The feasibility of re-treating the area 2 days after the first treatment was being considered in an * This work was done shortly after the termination date for the segment covered by’ this report. It is discussed here, however, so that this information can be considered, and perhaps utilized, by fisheries workers one year sooner than if discussion was postponed until termination of next segment. attempt to kill all of the eggs and fry, when another large concentration of carp and goldfish began moving back into the treated area for spawning. The next morning, the area was congested with carp and goldfish, very actively engaged in Spawning amongst the bloated and stinking corpses of the thousands of fishes that had been killed during the first treatment, and had not been blown to the banks or completely out of the spawning location. The area was treated again at the same concentration so that the parent fishes would be killed before they deposited too many fresh eggs that would not be affected by the rotenone, and to kill fry and eggs in advance stages of development from the past Spawn. An attempt was made the following morning to make a total-kill estimate. Shift- ing winds had formed windrows of dead fish along the banks and small islands. The thick- ly weeded areas were so congested with dead fish that many of them could not find room to surface. Many had been blown out of the area and formed windrows along the southeast shoreline for a mile or so. None of the eggs from the previous spawn were found to be alive. No fry were found by seining with small mesh nylon seine. It was quickly appar- ent that spot treating large concentrations of Spawning fish could possibly be the most effective, economical, and certainly the most selective method of controlling carp and goldfish in Buffalo Lake. Condensed information concerning this treatment and the esti~ mated total-kill by shoreline count is given as follows: Number of treatments - 2 Amount of rotenone used per treatment — 160 lbs. Total cost of rotenone used - $112.00 Surface area treated - approximately 10h acres. Shoreline considered for total-kill estimate - 10,000 feet. \Species Number Total Pounds Tons Carp 22,000 28,600 iu.3 Goldfish h,200 5,0u0 2.5 Shad 20,800 12,h80 6.2 Channel cat 700 Black base 200 Total 57,900 M6,120 23.0 Practically all the bass and channel catfish were fingerlings or young of the year fish. Very few were yearlings. Evidence of heavy feeding was obvious in all cat- fish. Examination of their stomachs disclosed they were gorged with eggs, probably those of carp and goldfish, which explained their presence in the warm, muddy spawning waters. Since gonads of the shad that were killed were in variable stages of develOpment, it is doubtful that they were in the spawning area for the purpose of Spawning. They, too, may have been feeding on carp and goldfish eggs. Immediately following the first Spot treatment in the spawning area, observa- tions were made in the Tierra Blanca Creek above Buffalo Lake. At this time, the creek was almost back to its normal stream—flow level following the recent rise. Rocks along the creek banks were coated with millions of dried fish eggs, presumably eggs of shad zhich were spawned during the rise and stranded as flood water receded. Since a great number of goldfish was observed in the creek, apparently spawning, the creek was roten— oned for about 3/h mile above the lake. In addition to the Spawning goldfish, numerous shad (mostly ripe males), minnows and small channel catfish were killed. All stomachs of channel catfish were filled with fish eggs, apparently those of goldfish. Since the relative abundance of rough fish species to channel catfish and minnows was not great enough to justify further treatment, the remainder of the creek was left unmolested. FINDINGS: WATER QUALITY Table 7 gives the chemical nature of water at Buffalo Lake from June 1955 to May 1957. Other than a slight increase in total hardness, there are no significant chan- ges in water quality. PHYSICAL CHARACTERISTICS Physical conditions.of Buffalo Lake water, when fish samples were taken, are given in Table 8. All of the turbidity recordings in this table are not accepted as valid of fish kills at Buffalo Lake. Sport fishermen report catches of black bass on artificial lures, which was not done before the water cleared. Errors in turbidity are explained by the fact that turbidities were measured with a secchi disk which may vary according “o the amount of sunlight at the time of reading, and difference in vision of persons .aking the reading. Therefore, this method of determining turbidity has been discontinued. In the future, all turbidities will be taken by the same person, using a standardized can- dle in the Jackson turbidemeter. FOOD HABITS Food remains were feund in stomachs of 15 specimens, 11 of which were channel catfish. Table 9 presents results of analyses made on food remains found in stomachs of channel catfish. Conspicious by their absence in catfish stomachs this segment are remains of shad. They could have been present, however, because unidentifiable fish remains were found on numerous occasions. Since all predacious species are presumed to eat other fishes, stomach contents were not recorded unless identification was reasonably certain. ' = Only one flathead catfish was found to have food in its stomach. It contained two sunfishes about 2% inches long. The only crappie collected with identifiable food remains contained a shad, four inches long. One large black bass contained a A% inch shad. SEXUAL DEVELOPMENT AND SPAWNING SUCCESS Larger fishes were opened in the field and their gonads examined and the stage qf development recorded. If the gonads were of medium-size or smaller, and were poorly 4‘ develOped, they were recorded as ”immature.” If they were large and well~developed, obviously approaching spawning condition, they were termed "ripe". In those few inn stances where a fish was captured shortly after Spawning, it was called ”spent". The latter condition is difficult to determine in males, but easier in females. Immature individuals of the larger fishes were taken in seine drags and mea- sured and counted. In addition, notes were made of schools of fry seen, young fishes found in the stomachs of predacious fishes, etc. When ovaries contained large eggs, nearly ready to be spawned, the two ovaries of such a fish were carefully removed, labeled and preserved in formalin. In the lab— oratory, the ovaries were carefully cleaned of excess tissues and weighed to the near- est one-tenth of a gram. Then a small quantity, roughly a gram, was snipped from one ovary and weighed on a chemical balance to the nearest one—one hundredth of a gram. The eggs in the small portion were then counted. An average of 3 counts was determined and the total number of eggs present in the two ovaries estimated (number of eggs counted times weight of both ovaries divided by the weight of the small section). Table 10 gives the total number of eggs found in ovaries of 13 ripe females. Dorosoma cepedianum (Gizzard shad; Ripe shad were found in mideApril, but the major Spawning season for shad in Buffalo Lake is_late May through June. Ybung of the year were common in July and be- came abundant in August. Off-season spawns of shad are quite successful. Both SXperimental selective-kills at Buffalo Lake greatly reduced the shad population. This reduction is quite apparent in seine collections following the kills, and tO‘a lesser extent in gill net collections (See Table 1h). However, shad have the ability to replenish themselves with amazing rapidity. Carpiodes carpio (river carpsucker) Very little information on the spawning of carpsucker was found this segment. Ripe males and females were taken from April to September, but none of the gonads of the 82 specimens taken were in a_spent condition. No young-of—the-year were taken in seine samples. Carpsucker are not numerous and do not constitute a problem in Buffalo Lake. The few that are present are in excellent condition with "K" factors up to A. Efforts will be made in the future to determine reasons for the scarcity of this species. This information may be useful in setting up biological controls in other lakes where they are abundant. Cyprinus carpio W597— Most female carp became ripe in April, but deferred spawning until late May and June. A few ripe females were taken in December, and ripe males were taken in collections every month except January. Carassius auratus igoldfishi Gonadal conditions of goldfish are quite similar to those of carp. Overlap of .pawning seasons, lack of vegetation on which to spawn, low water levels and scarcity of spawning areas, all of which force carp and goldfish to use the same Spawning areas, are reSponsible for the great number of carp—goldfish crosses in Buffalo Lake. Carassius X Cyprinus (carp-goldfish hybred) Apparently, very few of these generic crosses reproduce successfully. Only two of the 58 males, and two of the 16 females taken during this segment were ripe. Ictalurus punctatus (channel catfishi Spawning started in May, but the main spawn of this species occurs in June and continues throughout the summer. Ictalurus melas (black bullhead catfish) Spawning apparently starts in June and continues throughout the summer. Pylodictus olivaris (flathead catfish) Two of the three females taken in June were ripe. The other one was spent. Roccus chrysops (white bassi No white bass were taken. Fishermen'report catching this introduced species occasionally, however, there were no indications of reproduction during this segment. Micro terns salmoides black bassi Female base were ripe in May, but most of the spawning occurred in June. Pomoxis annularis A (white crappiei Most females taken in May are ripe. The major spawn occurs in June. COEFFICIENT OF CONDITION ”K" factors were worked out for all of the larger fishes taken in gill nets. Distribution of ”K" factors for fishes in Buffalo Lake is given in Table 11. All of the fishes taken are not listed in Table 11 because immature animals less than 50 mm. were rejected, and a few were recorded erroneously, resulting in ridiculous "K" factors. Interesting to note is the general increase in the distribution of "K" factors for nearly all species during the segment ending in 1957, as compared to the one ending in 1956. Also interesting is the agreement between the increase in "K" factors shown in Table ll and the increase in average weights of species shown in Tables 1, 2, 3, and A. An attempt was made to use monthly variations of "K” factor distribution as an indication of any possible effects of selective-kill treatments. This information was worked out for shad only and is presented in Tables 12 and 13. According to information presented in this manner, "K" factors for shad follow a general pattern over a period of a year - increasing during the summer and decreasing during the winter. Apparent also, is that the greatest average "K" factors for the segment preceding the selective-kill treatment are lower than the lowest average "K” factors for the segment following treat- ment. Immediately, one would conclude that the obvious explanation for this increase is the great reduction in numbers of shad, thereby providing additional space and food, and improving habitat for those that remain. In search for other explanations for increases in average "K", Table 1A was made to show monthly average lengths and average weights of shad for a 12 month period before and after the first treatment. According to Table 1h, the first treatment killed a great portion of larger and older shad, thereby decreasing the average weight and length of both sexes (compare June 1955 to June 1956). Since the treatment was made prior to any major spawn, these figures are not affected by young-of-the—year fish. Interesting to note, however, is that the average shad in April 1957 (ll months after treatment), is considerably larger and heavier than those in April 1956.' Apparently, more male shad were killed than females because 2A.2% of the collection after treatment were males as compared to 44.6% males before. PATHOLOGICAL CONDITIONS All fishes taken during the study were examined for evidence of disease, para- sites, or other abnormalities. The fungus, Saprolegnea parasitica, was present on only a few minnows. In all instances, the infected fishes had escaped from a hook or had otherwise been injured. Five crappie were found to be infected with the hard, encrusted "tail rot" fungus. Encrustation on one was so large that only the infected portion of the body was caught in the gill net. Infected areas were found on all fins _ not restricted to the caudal fin. Gonads of several goldfish were greatly enlarged and filled with water. It appeared to be an enormous hydrocele which gave the entire fish a round appearance. This condition greatly increased the "K" factor in the infected fish. Blindness in large shad was common, but it had no apparent effect on their health. Eyes and part of the head of infected fishes were covered with what appeared to be ossified mucous. DISCUSSION Since the theory and practice of ”selective-killing" is still in its infancy, much work is necessary before it can be accepted as a desirable tool in fisheries management. Great caution must be exercised before attempting to erradicate any part— icular Species of fish from a body of water, even if it were possible. Therefore, the purpose of the work discussed in this report is to attempt to effect selective-kills on undesirable species as a control measure only, and to collect continuous data from which to determine the final effects of the treatments. The weight of fishes removed from Buffalo Lake, mostly shad and carp, since May 1955, is very conservatively estimated at 200 tons. This figure includes all fishes removed by selective-kill treatments, spot—treatments during Spawning periods, seining and gill netting. This report presents data which represents results of this fish removal based on ll months of inventory prior to the first treatment and 12 months of inventory following the first treatment. Since approximately one-half the total kill estimate has been removed since April 1957 (one month prior to termination date of this segment), final results of this work cannot be expected to appear in data of this report. However, inventory has been approved for continuation into the next project year. The next report on this work will present data from three consecutive segments and should indicate significant effects of the treatments - if they exist. SELECTIVE-KILL TREATMENTS Results of selective—kill treatments at Buffalo Lake do not compare with simi- lar treatments in the laboratory. Complete shad kills have been obtained under labora- tory conditions without effecting game fish. This has not been accomplished at Buffalo Lake ~ nor was it eXpected. Although results of both selectivewkill treatments are gratifying, several fac— tors may be responsible for not achieving even greater success. Certainly, there were errors in water measurements. Water volume of Buffalo Lake was estimated from a height gage reading and capacity curve charts which were made from pre—impoundment survey maps. A reasonable figure was subtracted from the capacity shown on the chart to allow for siltation. Although there may have been errors in water measurements, they were insignw ificant and are not considered as very great influencing factors in the Buffalo Lake treatment. Lack of depth penetration, thereby permitting many shad in deeper parts of the lake to avoid contact with rotenone is suspected. The wide dispersal patterns of the treating barge, necessitated by high velocity winds during both treatments, caused excessive concentration of toxicant in some areas and perhaps none in others. If the wind, waves and water currents failed to dilute and spread the rotenone as heped, many surviving shad failed to come into contact with the toxicant, and the small percentage of game fishes killed may have resulted from being directly in the path of the barge and, hence in areas of higher concentration. Since the rotenone may dissipate before coming in contact with some of the shad, perhaps two treatments with a smaller concen- tration, and at least 2% hours apart, would be more effective than only one treatment with the original concentration. There is little doubt that reducing the concentration and sustaining the toxicity for an additional 2A hours during the first treatment would have been far more effective, and perhaps eliminated the necessity of the second treat- ment eleven months later. Yet to be decided is the preper season and water temperature for treating in order to obtain optimum selectivity. Since youngwofuthe-year fishes are most susceptible 10. to the effects of rotenone, treatment after Spawning season will destroy countless num- bers of young Shad. But, young—of-the-year game fishes will also be killed — especially channel catfish and white bass. Experiments in the lab and on small ponds have proven that greater selectivity on shed (and drum) is obtained with rotenone treatments at temperatures between 55°F and 60°F. Treatment in the early Spring, as water temperatures are rising to 60°F may kill ripe shad before the Opportunity to spawn is permitted, whereas, treatment in the fall, as temperatures drOp below 60°F, may kill off both young- ofwthe~year as well as the parent fishes. According to "K" factor data (see Tables 12 and 13), Shad are in poorer condition during winter months than any other time, and pos— sibly, more susceptible to effects of rotenone; whereas, treatment following the summer months would find shed in their peak of condition. As previously stated, much work and study must be done before ”selective-killing" can be proven and accepted as a desirable tool in fisheries management. Only until the uncertainties and problems mentioned above, and many others, are solved and proven to workable conclusions, can this method be safely and wisely used. Admittedly, the statement concerning shad feeding on carp eggs is far—fetched and contrary to all available food habit studies on shad. However, non—Spawning shad have been found in carp spawning areas in lakes other than Buffalo Lake, and their pre- sence has not been explained. Tables 1 and 2 Show that carp increased from 11.1 percent of the total collection before the first treatment, which occurred before the major carp Spawn, to 20.3 percent of the collection following. This'great increase is not shown for the other species of fish. Therefore, ridiculous as the idea may seem, efforts to conform or deny it Should be made. Until this question is answered, all efforts to con- trol shad Should be accompanied by equal efforts to control carp. Otherwise, one of the natural biological controls on carp may be removed, and carp may become a greater prob- lem than the carp and shed combination. Buffalo Lake is an ideal body of water for eXperimental control of undesirable species. There are no lakes on the tributary streams and watershed which may recontam- inate the treated lake by overflowing during floods. Therefore, continuous inventory and study should indicate results of any experimental work that may alter the fish population. Authorization has been approved to continue this work until at least May 31, 1958. 4. Prepared by: Lee D. Lewis Approved by: 129)». \_jZHaJZA_s Project Leader Chief Aquatic Biologist Date: Se tember 23, 1957