DYNAMICS OF THE FISH PQPU‘LMIONS‘ m .A WARM—WATER STREAM 0" “v ‘—...' wrv“- «-- a. .- I'C-oa.|....’.l!". v mm {or flu ;me' 34.175. ' . MICE-1m sum UNIVERSITY - 7‘ Kenneth Jack Linton if.1964- ‘ a II in um; mleyugmu w: m 1! w I21"! 111"th ll LIBRARY Michigan State University 564—9411 W 2566‘ «WV/“(£11 \+, ‘- 1 2 . ”‘5; ’9‘ "$5". ‘ @711."‘<3 . smug 1 7“. . O n .9- . &%m 2 316:}; ”1" WW "APR 1 51999 "" 416 2081 5:5»(‘ 0350015 ABSTRACT DYNAMICS OF THE FISH POPULATIONS IN A WARM-WATER STREAM by Kenneth Jack Linton The fish in the Red Cedar River, a warm-water stream in central Michigan, were investigated to determine what differences existed between the major fish populations in five arbitrarily selected zones comprising 30 miles of the main stream. The characteristics studied included growth rates, standing crop in numbers and biomass, and net production in the Ivlev sense. A limited study of fish movement revealed that a portion of the rock bass population maintains a small range of movement. The growth study was based on the results of scale reading and back-calculation of the total length at annulus formation for about 3100 fish including 13 species. Significant differences in the rates of growth for the rock bass and hog suckers in the five zones were demonstrated, with growth being generally faster in the three upstream zones. No significant differences in the growth rates of white suckers and redhorse were found. The standing crops of fishes were estimated by the Petersen method employing an electrofishing technique, after a comparison of several methods indicated that this was the most practical means for the present study. A source of bias in the Petersen method when applied to small samples is suggested as being the effect of estimating a continuous distribution with a discrete distribution. Population estimates were made for each of the five zones during four sampling Kenneth Jack Linton periods over a span of three years. Significant heterogeneity was demonstrated in the number of fish per acre for the five zones, but the differences in the biomass estimates were not statistically signifi— cant. Zone III, in the middle of the study section, most consistently exhibited the largest standing crop of fishes, while Zone I, further- most downstream, supported the smallest. A downward trend in the abundance of all species was noted for Zone I. The mean estimate of the standing crop of fishes (other than minnows, darters, and mud- minnows) was 101 lbs/acre with the individual estimates ranging from 18 to 3381bs/acre. The rates of net production were estimated using the technique of Ricker. Estimates of the instantaneous mortality rates were ob- tained by the Chapman- Robson method and these were used to back- calculate each population estimate to the beginning of the growing season. The mean total net production for all of the major species in the river was estimated to be 13.43 mg dry weight/mZ/day in the 180 day growing season with a range of 7. 37 to 18.68 mg dry weight/mz/day. Zone III was estimated to have the highest rate of net production of I fishes, but it was determined in a previous study to have the second lowest rate of primary production. This apparent discrepancy was attributed to the fact that Zone 111 receives the greatest supply of allochthonous nutrients and supports the largest production of heter- trophic aufwuchs . DYNAMICS OF THE FISH POPULATIONS IN A WARM-WATER STREAM BY Kenneth Jack Linton A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1964 ACKNOWLEDGMENTS Iwish to express my sincere gratitude to Dr. Robert C. Ball, whose guidance and encouragement in the conduct of my research and preparation of the manuscript was invaluable. I am grateful to Drs. Eugene W. Roelofs and Phillip J. Clark for their suggestions and criticisms during the course of the project. I also wish to thank my fellow graduate students for their generous aid in the field collections and particularly Mfssrs. D. L. King and R. L. Vannote for their suggestions and stimulating dis- cussions. Collections were made and the analyses performed while the author was employed under a National Institutes of Health Grant administered by Drs. R. C. Ball and Don W. Hayne (K-24[CZ]) The manuscript was prepared during the tenure of a Predoctoral Research Fellowship (WP-15, 866-02) sponsored by the Division of Water Supply and Pollution Control of the United States Public Health Service. * >i< * >}< >i< * >I< >’.< >l< >l< >i< >l< * ii TABLE OF CONTENTS Page INTRODUCTION 0 O O 0 O O O O O O O O O O O O O O O O O O O O O 1 DESCRIPTIONOFTHERIVER................. 3 METHODS O O O I O O O O O O O O O O O O O O C O O O O O O O O O 12 RESULTS. 0 g o o o o o o o o o o o o o o o o o o o o o o o o o o 22 Movement StUdy O O O O O O O O O O O O O O O O O I O O O O 23 Grthh Study 0 C O O O O O O O O O O O O O C O C . 0 O C O 28 Length-Weight Relationships . . . . . . . . . . . . . . . 33 Population and Biomass Estimates . . . . . . . . . . . . 33 ProductionEstimates................... 46 DISCUSSION 0 O C O O O O O O 0 O O O O O O O O O O O O O O O O O 50 SUMMARY. 0 O I O O O O O O I O O C O O O O O O O 0 C C O O C O 60 LITERATURE CITED 0 O O O O I O O O O 0 O O O O O O O O O O O 63 APPENDICES O O O O O O 0 O O O O O O O O O O O O O O O O O O O 67 iii LIST OF TABLES TABLE Page 1. Monthly maximum and minimum discharge of the Red Cedar River recorded at Farm Lane Bridge for the yearsofl958throughl96l............... 8 2. List of dates and locations of fish collections with the length of station sampled and the mean width of the zones 0 C O O O O O O O O O O O O O O O O O O O O O O O O 14 3. List of constants for the regression of total length on scale radius used in estimating growth for thirteen speciesoffish........,............. 19 4. Summary of the recovery of marked fish in the summer Of 1961. O I O O O O O O O O O O O O O O O O O O I 24 5. Summary for all collections of the direction and mean distance moved by recovered fish tagged in the sum- mer Of 1961 O O O O I 0 O O O O O O O O O O O O O O O O O 25 6. Direction and mean distance moved by tagged rock bass by month of recapture in 1961. . . . . . . . . . . 27 7. Results of an analysis of variance for the significance of observed differences in the mean length at age II of rock bass in each of the five zones for the years 1958 throughl96l....................... Z9 8. Summary of pairs of zones which showed a significant difference with respect to length of rock bass at age II 29 9. Results of an analysis of variance for the significance of observed differences in the mean length at age II of white suckers in each of the five zones for the years l958throughl96l.................... 3O 10. Results of an analysis of variance for the significance of observed differences in the mean length at age II of hog suckers in each of the five zones for the years l958throughl961 ..... .......... 30 iv LIST OF TABLES - Continued TABLE 11. 12.. 13. 14. 15. 16. 17. 18. 19. Page Summary of pairs of zones which showed a significant difference with respect to length of hog suckers at age 11 O O O O O O O O O O O O O O I O O O O O O O O O O O 0 Results of an analysis of variance for the significance of observed differences in the mean length at age 11 of redhorse in each of the five zones for the years 1958 through1961....................... Actual and calculated weights of three rock bass using three length-weight regression formulas . . . . . . . . Observed values of the constants log c and_r_1 for length- weight regression formulas for several species of fish in the Red cedar River. 0 O O O O O O O O O O O O O O O 0 Summary of population and biomass estimates of fish in each zone of the Red Cedar River during 1959, summer of 1960, fall of 1960, and summer of 1961 collection periods in terms of numbers per acre, num- bers per mile, and pounds per acre. . . . . . . . . . . Comparison of the numbers and weights of fish esti- mated by the Petersen method with the numbers and weights of fish recovered after poisoning the same area withrotenone....................... Population and biomass estimates by the Petersen and Bailey formulae for several hypothetical populations under various conditions of capture efficiency . . . . . Comparison of the estimated fish biomass in three taxo- nomic categOries and the total biomass for the Red Cedar River and that reported for eight other streams. Distribution and estimated abundance of the various sucker species in number per mile of stream for the Red Cedar River in 1959, 1960 summer and fall, and 1961.....oooooooooooooooooo..... 31 31 34 34 35 37 41 43 45 LIST OF TABLES - Continued TABLE 20. 21. 22. Estimated dry weight production for several species of fish in the Red Cedar River (units are milligrams persquaremeterperday). . . . . . . . . . . . . . . Estimated dry weight production for the two major families of fishes considered in the Red Cedar River study (units are milligrams per square meter per day) Summary of the estimates of population number, bio- mass, and production of fish compared with several other physical and biological characteristics of the five zones of the Red Cedar River . . . . . . . . . . . vi Page 47 48 52 LIST OF FIGURES FIGURE Page 1. Map of the Red Cedar River watershed showing the locationofstudyzones................. 5 2. Monthly mean discharge at Farm Lane Bridge for the yearsl958through1961................ 9 vii LIST OF APPENDIC ES APPENDIX Page A. List of fishes encountered in the Red Cedar River during the course of this study. . . . . . . . . . . 68 B. Mean back-calculated lengths at annulus form- ation for thirteen species of fish in the five zones 69 viii INTRODUCTION In 1958, a study of the fixation and utilization of solar energy was commenced on the Red Cedar River, a warm-water stream. For this study, the river was arbitrarily divided into five zones which were believed to represent somewhat distinct ecological communities. The original plan called for five major phases of study: (1) fish sampl- ing in each of the zones; (2) measurement of primary production on artificial substrates; (3) determination of the level and source of nutrients in the stream; (4) measurement of the composition and amount of suspended materials; and (5) estimating the production rates of bottom faunal material. Emphasis in the early part of the study was placed on physical and chemical factors and on primary production in the several zones. Results of the chemical and hydrological investigations were presented by Vannote (1961, M. S. thesis). Grzenda (1960, Ph. D. thesis) has reported the results of the study of primary production on artificial substrates and nutrient utilization. Rawstron (1960,» M. S. thesis) provides further information on the growth of periphyton on artificial substrates. In the following years, emphasis was shifted to the utili- zation and exchange of energy among consumers. King (1961, M. S. thesis) provides the results of the early investigations of the distribution and biomass of benthic organisms. The intention and thesis of this study may be stated: there are recognizable and demonstrable differences in the fish community structure of the five zones of the Red Cedar River. The data and observ- ations in this thesis are the results of the first four years of study of the characteristics of the major fish populations. DESCRIPTION OF THE RIVER The study was conducted on the Red Cedar River, a warm- water stream in the south-central portion of the Lower Peninsula of Michigan. The river arises in Cedar Lake, Marion Township, Livingston County (TlN, R3E) and flows in a northwesterly direction about 19 miles through Livingston County and about 29 miles through Ingham County, reaching its confluence with the Grand River within the city of Lansing. The Red Cedar River receives the waters of twelve major tributaries, the largest being Sycamore Creek, and drains a total area of about 472 square miles (Figure 1). The upper portion of the river flows through grain and dairy farm lands and is used in part for irrigation of crops and watering of livestock. The channel has been dredged and straightened throughout much of this portion as part of a program of land reclamation. In the lower part of the drainage, it flows through farm lands and extensive urban areas where it is used for domestic, industrial, and recreational purposes. There are two major impoundments on the main stream. The largest of these is at Williamston and is used to provide power for a private frozen food storage plant. It is a concrete and wooden structure maintaining a 13-foot head of water. The other dam is on the Michigan State University campus in East Lansing. It is a concrete and rock dam which maintains a water supply for the steam generating power plant on the university campus. At least two other small dams (maximum two-foot head) are found on the main stream. Further discussion of the history and characteristics of the dams on the Red Cedar River may be found in Brehmer (M. S. thesis, 1956). The study section is about 30 miles long and extends from the Farm Lane bridge on the Michigan State University campus in East Lansing to the VanBuren Road bridge about one mile upstream from Fowlerville. The Sycamore Creek drainage is not included in this Figure 1. Map of the Red Cedar River watershed showing the location of study zones. 9.4... demo «I x W. V 0 1 o a V V .8. N, N O .N. 0 416 x 3 a V I. .8 8 0 Z 0 8 3 U . O 3 M H .— >¢hmmu3 O a. s . mm. s . is... o 3’ 1 S 4. O O V 8 M 0 3 40¢ a @019. 2W n o oz_mz<... ”v.44 Nouws‘x O -.¢aue Si... 05w .4. m ww.=_2 . I O e. no 00 C 00000 oz._.w.z.<.J 01 system, so the drainage for the study section is about 355 square miles. Table 1 and Figure 2 include information on the discharge for the years 1958 through 1961. These data were furnished by the United States Geological Survey, which maintains a data record- ing station at the Farm Lane bridge. The study section has been divided into five zones for this investigation. Zone I includes that three and one-half mile portion of the stream from the Farm Lane bridge to Okemos Road (Figure 1). The water here is slow-moving and drops silt and detritus on the mud bottom. The river is influenced by the dam mentioned above, which i is located about one-fourth mile below the Farm Lane bridge. The bottom at the upstream end of this zone is largely sand and rubble. Zone II (eight and one-half miles long) consists of that part from the Okemos Road bridge to the Zimmer Road bridge below Williamston. It is the cleanest of the five zones, having a sand bottom in a large portion and gravel and boulders in most of the remaining portion. There are no urban areas in this part of the river except for the influence of Okemos at the extreme lower end. The number of riffle-pool combinations is higher in the part from the bridge at US 16 to Zimmer Road than in any other part of the river (Robin Vannote, personal communication). Below US 16, the water is slower and deeper with fewer riffles. Zone III extends from Zimmer Road bridge to the dam at Williamston, two and one-half miles upstream. The bottom varies from silt to sand and cobbles, with some detritus. It is strongly influenced by the Williamston urban area, particularly by the Williamston sewage disposal plant. Raw sewage may be seen seeping from the banks and draining into the river in the urban area proper. There are extensive beds of rooted higher aquatic plants immediately below the dam. Table 1. Monthly maximum and minimum discharge of the Red Cedar River recorded at Farm Lane Bridge for the years of 1958 through 1961.* (87:) 1958 1959 1960 1961 Month Max. Min. Max. Min. Max. Min. Max. Min. January 260 94 60 41 1150 168 60 13 February 780 71 264 47 867 143 118 14 March 800 148 1810 258 3530 118 413 109 April 292 121 1540 163 3240 351 1090 90 May 110 49 475 115 505 208 550 72 June 94 41 343 47 872 104 142 36 July . 546 47 339 11 109 30 60 16 August 81 28 272 60 88 17 202 25 September 49 11 186 30 72 14 199 50 October 71 36 2210 94 82 16 104 36 November 81 45 950 209 75 21 173 62 December 60 38 775 180 60 16 189 60 5:: Data provided by the United States Geological Survey Field Office, Lansing, Michigan Figure 2. Monthly‘mean discharge).< at Farm Lane bridge for the years 1958 through 1961. >(< Data provided by the United States Geological Survey Field Office, Lansing, Michigan. 10 £qu .>oZ .uoO .umom .ms< 3.3. 05;. .332 £96. £32 .noh .cmh \ .J / . a .room \ . / . .. \I. \ , / .. «.11.. I, . . OOH"V III/\\ / // .\ s ./ / .\ \ :oom . z \ s / , . a .63 . I ,. ,\ If x .>.. L :8» $2 < z I 32 I.I.I.I / 2 50% $2 IIIIII I 3.2 i < .32. :82 ('n- '393) sfixeqosrq 11 Zone IV includes the largest impoundment in the system and extends from the dam at Williamston to the bridge at Dietz Road. The reservoir proper comprises approximately half of the four mile length, but the influence of the reservoir extends throughout the zone. The reservoir is confined to a narrow basin, but the flow is very slow. For the fish study, the reservoir has been disregarded due to sampling problems and other considerations. Therefore, the results discussed refer only to the upstream portion of this zone, arbitrarily that portion where the river is no more than 100 feet wide. The mean width included in the calculations is also based only on this upstream portion. Further description of the impoundment may be found in Brehmer (ibid. ). Zone V represents the remaining 12 miles of the study section from the Dietz Road bridge to the bridge at VanBuren Road. It is influenced by the Webberville urban area. Probably the greatest influence is the industrial waste of the metal plating plant in Fowler— ville. A large portion of this zone has been dredged and straightened and flows through farm land. The bottom is largely silt and mud. The depth and width of the dredged portions are, of course, quite uniform and the flow is slow. METHODS 12 13 The fish were collected with a 220-volt Homelite direct-current generator which was placed in an eight-foot wooden boat. The hand- held positive electrodes consisted of coiled copper tubing mounted on six-foot wooden handles and were opposed by a metal negative electrode plate in the bottom of the boat. The stunned fish were re- trieved at the positive electrodes with dip nets having a one-fourth- inch mesh woven nylon bag or a graded mesh (one inch to one-eighth inch) cotton bag. The stations (Table 2) were randomly selected within the limits of accessability. In 1958, the collections were made without the use of block nets since only growth data were being collected. In subsequent years, nets were placed in the river in such a manner as to prevent the movement of fish into or out of the station. The nets were sus- pended on metal fence posts and secured to the bottom of the river with rocks, logs, and fence posts. The block nets consisted of one- fourth-inch woven mesh nylon nets (one net 25 feet by 4 feet and two nets each 50 feet by 6 feet) or three-eighths-inch bar mesh cotton net (one net 75 feet by 4 feet). The nets were placed in the river with as little disturbance of the fish as possible. In 1959 and 1960, the procedures were similar. The crew started at the downstream net and shocked upstream, fin- clipping all fish and releasing them at the approximate place of capture, recording only the number of fish of each species. On the second and subsequent runs, the fish were placed in a metal tub in the boat. About half-way through the station and again at the end, the fish would be weighed to the nearest gram and measured to the nearest millimeter. Scale samples were taken from the region near the tip of the compressed pectoral fin, recaptures recorded, and all fish were released. In 1960, the rock bass were tagged prior to release. Table 2. List of dates and locations of fish collections with the length of station sampled and the mean width of the zones. Zone Station Length Collection Dates Mean Widg— I 79. 7 feet .A - 7/30/58 — 8/18/58 B - 8/13/58 - 8/18/58 528‘ 9/18/59 716' 7/18/60 716' 9/12/60 300' 6/26/61 11 65. 5 feet .A - 7/30/58 528' 8/6/59 300. 6/21/61 300' 7/11/61 300' 9/12/61 B - 8/19/58 528' 9/17/59 612' 7/15/60 612' 9/13/60 c 300' 6/28/61 III 62.4 feet .A - 7/31/58 722' 7/13/60 722' 9/14/60 B - 8/14/58 528' 9/14/59 300' 7/5/61 300' 9/20/61 c 300' 6/30/61 300' 7/18/61 IV 60. 9 feet .A - 8/14/58 - 8/15/58 B - 8/8/58 - 8/19/58 528' 9/10/59 910' 7/11/60 910' 9/15/60 300' 7/7/61 300' 9/19/61 Continued 15 Table 2 - Continued Zone Station Length Collection Dates Mean Width V 34. 3 feet .A - 8/7/58 820' 7/6/60 820' 9/16/60 B 528' 9/9/59 C - 8/1/58 D - 8 /7 /58 300' 7/19/61 300' 8/28/61 E 300' 7/10/61 16 In 1961, the procedures were altered somewhat. Six live-boxes were constructed prior to commencing the summer operation. These consisted of three-fourths inch diameter steel conduit welded frames with one-fourth inch woven mesh nylon bags laced to the frames. They were designed so that three boxes would nestle for easy storage. The largest were 28 inches by 28 inches by 40 inches and the smallest were 20 inches by 20 inches by 32 inches. The block nets were placed in the stream as before. Then if the conditions warranted it, one one-inch bar mesh nylon net (50 feet by 4 feet) was placed in the river about 15 feet upstream from each of the block nets to catch detritus. The live-boxes’were then placed in the station at approximately 50-foot intervals. The crew shocked up- stream, putting all fish in the live-boxes. At the completion of a run, growth data were collected at each live-box and the fish were tagged and released. On subsequent runs, recaptures were recorded and released and growth data were collected on the remaining fish. In 1961, an attempt was made to estimate the populations by the DeLury method (Ricker, 1958), retaining all fish in the live-boxes after the first run. However, it was found that this method was not feasible due to the low efficiency of collection and to the limited number of runs it was possible to make during daylight hours (maximum of four runs). The nets could not be left overnight due to the increased pressure with accumulation of detritus and to the possibility of theft. An attempt was also made to estimate the populations by the Schnabel method (Ricker, ibid. ). Again, the number of fish of any species on any run was so low that nothing was gained. It was decided that the Petersen method (Ricker, ibid.) was most useful. In previous years, the following formula was used to estimate the populations: M9 T: R 17 where T equals the estimate of the population in the station, M equals the number of fish marked on the first run, C equals the number of fish captured on the subsequent runs, and R equals the number of recaptures. If the value of either C or R were zero, then the equation used was the same as that used for all of the 1961 estimates, which is the Bailey equation: T = M—gsrn where the symbols are as previously explained. If the value of M was zero, as sometimes occurred, a one was substituted. In 1960, the rock bass were tagged with plastic flutter tags attached by means of a small nylon thread. The tags were applied with a curved surgical needle just anterior or posterior to the dorsal fin. Of the 225 tags applied, only five were recovered and the fish were in extremely poor condition. It is suspected that the tags were pulled out when they became entangled in the brush where the rock bass were found. In 1961, Monel metal band tags (Style 4-1242M, National Band and Tag Company) were used as jaw tags for several species of fish, including suckers. Three sizes of bands were used: Nos. 5, 7, and 8. It was found that the smaller size was more difficult to apply with- out damage to the mandible and offered no apparent advantage over the largest size used. Even the smallest of the tags could not be used with fish such as the bluegill due to the size of the plier nose. The No. 8 bands were found to be satisfactory when used on larger- mouthed fish such as the rock bass if the fish was about nine centi- meters long or longer. With the band tag, a larger percentage were recovered and the fish were in better condition. For the growth study, impressions were made of the scales on small squares of acetate without heat, as described by Smith (1954). 18 These were then projected at a magnification of about 22 diameters with a Bausch 81 Lomb Tri-Simplex Micro-projector. Annuli were observed and marked off on ruled data cards. The total length-scale radius regression line was calculated for each species (Table 3) using a minimum of 150 fish except for those species for which fewer were available, in which case all data were used. Back-calculations were made according to the method of Hile (1941) and others, assuming a linear relationship. Initially, it was planned to compute the scale length-body length regression line for each sample of each species. This was done, but it was found that the intercept values varied quite widely. Analysis of covariance showed that the slopes of the lines were nearly always significantly different. A scatter diagram was constructed using the data from a well-distributed sample of hog suckers. This showed that the reason for the difficulties was a tendency for the scale diameter to vary quite widely for any given length of fish. This was especially true of the suckers, with the minimum effect occurring in the rock bass. Those samples which contained fish of a limited size range would, therefore, show small slopes and elevated intercept values. Consequently, it was decided that a common intercept value for each species would be more realistic and would be conservative with respect to statistical comparisons between the zones. Biomass estimates were made by converting the estimated number of fish in each mile of stream to number per acre using the mean width of the zones as they appear in Table 2. These were then multiplied by the mean weight of the fish in the sample from which the estimate was made. Grams per acre were then converted to pounds per acre using the conversion factor of 2. 205 pounds per kilogram. The net production (in the Ivlev sense) of centrarchid and cato- stomid fishes was estimated with the technique of Ricker (1958) as 19 >3 Table 3. List of constants for the regression of total length on scale radius used in estimating growth for thirteen species of fish. Species a (cm.) b White sucker 7.52 2.705 Northern hogsucker 5.62 2. 362 Spotted sucker 5.05 2. 031 Redhorse 3.77 2.044 Rock bass 2. 20 1.686 Warmouth 4.77 1. 533 Green sunfish 0.62 2.691 Pumpkinseed 3. 25 l. 526 Bluegill 1. 58 2. 225 Longear sunfish 1. 20 2.000 Smallmouth bass 5. 18 2. 736 Largemouth bass 1. 32 3.112 Black crappie 4.90 1.814 >:< The total length of a fish at annulus formation may be estimated by the formula: Total length = a + b (22) (scale radius) where a and b are the appropriate constants selected from the above table, 22 is the magnification of the image, and the scale radius is measured in centimeters from the focus to the given annulus. 20 employed by Gerking (1962), with the following exceptions. The sur- vival rates (rs) were obtained in the manner proposed by Robson and Chapman (1961). These were calculated for each species of fish and converted to the instantaneous mortality rates (i), which were then assumed to be constant over all age groups and throughout the year. The mean back-calculated lengths at annulus formation were con- verted to the initial mean weights for individuals using the length- weight relationships in Table 14. For those species not appearing in Table 14, the weights were estimated from log-log plots of lengths versus weights. The initial weights for the age 0 fish were assumed to be 10'3 grams. These were arbitrarily Selected on the basis of the average volume of eggs reported in the Handbook of Biological Data (1956) and their approximate density. ‘ The common logarithm of the initial weight was then plotted on the age. The slope of this line at the midpoint between two consecutive ages multiplied by 2. 303 constituted the estimate of the instantaneous growth in weight for that year. May 1 was assumed to be the date of annulus formation and the initial date for the growing season. This was selected on the basis of recommendations made by Beckrnan (1943) for this part of Michigan. The water temperature of the Red Cedar River on that date agrees very well with the temperature suggested by Beckman as being conducive to annulus formation (about 53°F.). The end of the growing season was taken to be 180 days after May 1, which coincides with the time that the temperature of the Red Cedar River again approximates that at May 1. Each population estimate was back-calculated to the initial date (May 1) using the estimate of instantaneous mortality rate for that species. The age distribution within the mean estimates of initial population number were taken to be those distributions which were fixed 21 by the maximum observed age of a fish of that species collected from that zone, the Robson-Chapman survival rate estimate, and the mean estimate of initial population number. These were calcu- lated by trial and error. Because the young-of-the-year fish were not vulnerable to the capture method used, it was assumed that the estimates did not include them. Consequently, the initial population number of age 0 fish was estimated from the number of age I fish and the instantaneous mortality rate. Samples of fish including 4 rock bass, 3 white suckers, and 3 hogsuckers were weighed on a Mettler electrical balance (Model H4) and subsequently dried to constant weight at 55°C. The average percent dry weight was then used to convert the production estimates to dry weight units. RESULTS 22 23 Movement Study In the summer of 1960, approximately 225 rock bass were marked with flutter tags and released during the population estimate procedures. Of these, only five tags were recovered. The fish bear- ing the recovered tags had shown no movement from the original release point. During the June-July collection period in 1961, several species of fish were tagged (Table 4). The results, however, must be viewed with caution due to the method of recapture. A strong bias is intro- duced by attempting recapture only within about three or four hundred feet of each release site. Funk (1957) points out that this bias results from the decreased likelihood of recapturing strays, which is quite obvious. However, Gerking (1959) points out the less obvious fact that relying on fisherman returns causes a bias toward recording recapture of the stray fish, which is to be avoided for the purposes of this study. It was intended to test the hypothesis that a significant portion of the population does stay in a limited space. Table 5 was included to show the direction and mean distance moved of the recaptures for the most successfully tagged species, the rock bass, and for the few other returns that occurred. Of course, the data only indicate the apparent movement in that complex and extensive movement could have preceded recapture. It is suspected that such movement did occur more than these data indicate. I observed that a few tagged fish occurred in the vicinity of the downstream block net at the end of the day. Presumably, these fish may have drifted down- stream as a result of the stress due to tagging after the block nets were removed. Early attempts at tag recovery (less than two weeks after tagging) were not generally successful. The later attempts seemed to show that the fish had returned to the release point. 24 Table 4. Summary of the recovery of marked fish in the summer of 1961. Number Number Percent Species Marked Recaptured Recaptured Rock bass 166 36 22 Smallmouth 9 2 22 Bullhead 51 5 10 White sucker 23 1 4 Hog sucker 61 3 5 Redhorse 24 1 4 Northern pike 2 l 50 Spotted sucker 6 0 0 Carp 7 0 0 Grass pickerel 1 0 0 Green sunfish 3 0 0 Bluegill l 0 0 25 Table 5. Summary for all collections of the direction and mean distance moved by recovered fish tagged in the summer of 1961. M (I Upstream Downstream Complex* None Species 96 d 70 d ‘70 d % (1 Rock bass 56 90' 30 241' 3 300' 11 0' Smallmouth 100 350' Bullhead 80 1470' 20 0' White sucker 100 50' Hog sucker 67 168' 33 0' Redhorse 100 150' Northern pike 100 5200' :1. Includes those fish captured more than once mean maximum distance moved. and only indicates the 26 From Table 6, it appears that the emphasis shifts from an early upstream movement to a downstream movement. The October recap- tures were made by another group without attempting to recover tags upstream from the release site. The effects of spawning, which should have been well under way for the rock bass prior to the dates of original marking and releasing, are not known. I have concluded that there is at least a large segment of the rock bass population which maintains a small range of movement. This observation agrees with those of Funk (1957), Gerking (1953), and Scott (1949). The other species did not appear in our recaptures with sufficient frequency to warrant any conclusions. This may have been due to excessive movement, or, especially in the case of suckers, loss of the tag or mortality due to the tagging procedure. It was noted that in the case of the suckers it was necessary to apply the tag directly in the middle of the anterior portion of the mandible. This often re- sulted in severing an artery in the lower jaw. Many such fish recovered at least temporarily, but further irritation by the tag could have resulted in extensive delayed mortality. It is also suspected that feed— ing would be difficult for suckers so tagged. Gerking (1959) lists the hog sucker as one species which is known to have restricted movement, but does not include the white sucker found in this region (he includes the western white sucker). Funk (1957) did not include the white sucker in his study. Ball (1947) suggests that the yellow bullhead has a restricted movement pattern in lakes. Gerking (1959) includes this species in the list of fish with restricted movement, and Funk (1957) places it in his "sedentary" classification. Other fishes included in the Red Cedar River study which are believed to have a restricted movement are: bluegill, redhorse, green sunfish, smallmouth bass, and pumpkinseed (Gerking, 1959). 27 48306.36 on Summon “503$? 9203950on 2m wawpsaocfi po>o§ 00:3me G832 *uwvw .ooamumfip Gama UBEUOmmm a: @3632: maofiQ/nmmno mo quESZ %* .uoom GM po>o§ oodmumflp 532 * 02 v .omm o3 u RR 03 a: .mo on: Am Lw .wssopSHOmn/w o o E H 2 N «L w 6:02 OCH a. 5mm 2. m 1.: Hm m .0w 2 2v 63 EmonumGQ/OQ o o m... m .3 6m 6 .S No 2 .2: 88332 65 a p as G p 65 G p 65 in: RU 36:85.02 mo 2:8 HonouoO Meagan—mom umnms< 3.9m. onnumwoom «0 £302 Jog a“ spammoon m0 #305 >3 mmmo, xoos powwow >3 p.355 mocmpmflo smog paw \ 663685 .6 633. 28 Growth Study Mean back-calculated total lengths in centimeters for each age, zone, and species were computed (Appendix B). For statistical pur- poses, the mean back-calculated length for each age group at age 11 was used because the sample size for age group II was larger than that for older groups. Age I was rejected due to the possibility of strong effects of the time during the previous summer when the eggs hatched. The use of the mean back-calculated length at age II for each of the age-groups also gave a better estimate of internal variation for the several years of the study, thereby increasing the sensitivity of the tests. One-way analyses of variance, Model I (Dixon and Massey, 1957), were used to examine the significance of the observed differences in the mean back-calculated lengths at age II for the rock bass, white suckers, hog suckers, and redhorse for the four years of the study. The salient features of these analyses are presented in Tables 7, 9, 11, and 12 as recommended by Dixon and Massey (1957). In the case of the rock bass and hog suckers, differences were observed to be significant at least at the 5% level. Therefore, further investigation of the nature of the observed differences was undertaken by means of the Tukey multiple-range test (Snedecor, 1956) with the modification of Kramer (1956) and the modification of Hartley (Snedecor, ibid.). The results of these tests are presented in Tables 8 and 10. For those pairs of zones where the observed growth was significantly different at the 5% level, an X appears in the table. It may be seen from Table 7 that the rock bass showed very significantly different mean lengths. The individual comparisons re- vealed that the heterogeneity was due to the differences shown in Table 8. The data indicate that the growth of rock bass is faster, 29 Table 7. Results of an analysis of variance for the significance of observed differences in the mean length at age II of rock bass in each of the five zones for the years of 1958 through 1961.. Source Sum of Squares df Mean Square F Means 8.185 4 2.046 9.789 Within 9. 598 46 0. 209 Total 17. 783 50 F3995 (4, 40) = 6.30 Table 8. Summary of pairs of zones which showed a significant dif— ference with respect to length of rock bass at age II. Zone I II III I\f II 0* IHI X’ IX Zone IV X' 0 0 V X' 0 0 O >1< X denotes significant difference at 5% level. 0 denotes no significant difference at 5% level. 30 Table 9. Results of an analysis of variance for the significance of observed differences in the mean length at age II of hog suckers in each of the five zones for the years 1958 through 1961. Source Sum of Squares df Mean Square F Means 54. 296 3 18.099 18.646 Within 21.354 22 0.971 Total 75. 650 25 (3, 20) = 9. 20 F, 9995 Table 10. Summary of pairs of zones which showed a significant dif- ference with respect to length of hog suckers at age II. Zone I II III IV 11 0>i<>€< III 0 0 Zone IV * * * V X X X * :5: No observations from Zone IV. ** X denotes significant difference at 5% level. 0 denotes no signifcant difference at 5% level. 31 Table 11. Results of an analysis of variance for the significance of observed differences in the mean length at age II of white suckers in each of the five zones for the years 1958 through 1961. Source Sum of Squares df Mean Square F Means 20.978 3 6.993 1.988 Within 87.954 25 3. 518 Total 108.932 28 F1759» 24) = 1.46 12.900, 24): 2.33 Table 12. Results of an analysis of variance for the significance of observed differences in the mean length at age II of red- horse in each of the five zones for the years 1958 through 1961. Source Sum of Squares df Mean Square F Means 3.570 2 1.785 1.728 Within 27. 902 27 l. 033 Total 31. 472 29 F.75(2, 24)=1.47 F.90(2, 24) = 2.54 32 at least through age 11, in the three upstream zones than it is in zones I and II. The best growth apparently occurred in zone III, likely the richest zone with respect to allochthonous nutrients, while the poor- est growth apparently occurred in zone I. The hog suckers followed approximately the same pattern, but the difference in their growth above and below the darn between zones III and IV is so pronounced as to appear in error. From Tables 9 and 10, it may be seen that with respect to growth, the three zones below the dam (I, II, III) represent a homogeneous subset. The records for zone V result from the observation of only six back-calculated lengths and must be treated with caution. No significant difference in growth in the five zones for the white suckers (Table 11) and redhorse (Table 12) were revealed by the above methods. It is obvious that the variation in lengths of age 11 fish for the four years of the study was great. The Red Cedar rock bass growth is comparable to the growth in the poorer year--classes reported by Scott (1949) for the Tippecanoe River. It is a-l-se slightly lower than that reported by Brown (1960) for Massie Creek and the Little Miami River, both in Ohio. The growth of river-inhabiting rock bass in Kentucky reported by Tompkins and Carter (1951) far outstrips the growth of Red Cedar River rock bass or any other reported growth considered in this paper. At age I, the Kentucky rock bass were more than twice the length of the age I Red Cedar River rock bass. At age VI, the Red Cedar fish had reached approximately the same length as that attained by Kentucky fish at age IV. A cursory inspection of these results seems to substantiate the claim that, within this species, the northern populations are made up of slower-growing individuals than the southern populations. However, it became obvious during the course of this study that the direct com- parison of calculated lengths of fish from one area with those of another 33 area is of little value when based on small, localized samples. The observed growth of fish in the five zones of the Red Cedar River varied enough so that sampling from one station in the river could not be considered even a good representation of this river. The use of these data to extrapolate to all Michigan fishes would be wholly unjustified. Length- Weight Relationships The length-weight relationships for rock bass and suckers were calculated by the method described by Rounsefell and Everhart (1953). The data from September collections were pooled over the several years of the study for these purposes. An analysis of covariance revealed a significant difference in the regression lines for two successive years, 1959 and 1960, for rock bass in the same locality, which indicates that, for a given total length, the fish were slightly heavier in one year than in the other. It appears that the young fish were heavier in 1960 than in 1959. Further investigation showed that, although the difference was statistically significant, it was so small as to be unimportant from the standpoint of weight estimation. A comparison of the actual weights of three rock bass and the weights calculated with the 1959 and 1960 formulae as well as the pooled formula is presented in Table 13. When- ever necessary, the weights of fish were calculated with the appropriate mathematical regression formulae as they appear in Table 14, with lengths in centimeters and weights in grams. Population and Biomass Estimates The results of the estimation procedures appear in Table 15. The estimates for 1961 in zones 11, III, and V are the mean values for the two stations sampled in each case. The population sizes are 34 Table 13. Actual and calculated weights of three rock bass using three length-weight regression formulas. Weight 1959 1960 Pooled Total Length Actual Formula Formula Formula 6.2 cm 5gm 4.6 gm 5.1 gm 5.0 gm 12.4 38 37.1 37.6 37.2 19.0 133 128 124 129 Table 14. Observed values of the constants log c and _r_1 for length- weight regression* formulas for several species of fish. _S_pecies log c* n):< Sample Size Hog sucker -2.054 3.047 62 Redhorse -2.071 3.059 62 White sucker -2.212 3.158 62 Rock bass Zone I -1.6107 2.9090 230 Zone 11 -2. 1426 3. 3825 205 Zone III -1.8204 3.0768 110 Zone IV -1. 5828 2.8649 91 Zone V -1.6407 2.9459 61 :{c The weight of a fish in grams may be estimated from the total length using the regression formula: common log of weight = log c + 2 (common log of length) where the weight is expressed in grams, log c and _r_1_ are the appropriate constants selected from the table above, and the length is expressed in centimeters. 35 0HHE A030 00m 000.832 0000 n0nH 000509 0H ouum 00m 0032.992 +4 44 4 wH mp cum H HH $4 w Hym HHVH 0 Hum HHH o o o > wHH 5mm 022 mm mm Nwm c 00 m5. 0H HHH Hle m HHV mom >H mNH mom mmmw NM NHN No3 Fm ow Hmo oo hmH owHH 0H mHuH mmoH HHH we :0 mmww o wH HHVH H m hm om on momH NM Hmw Humwm HH ON mmv. oomv H 0 mm M w ow H. HmH Eme H: mom onN H 0HooH on mmo hmwm hm ohm oHVHH c on. wmm om omH Hume m0 wHH4 mmvH > hum oomH onoH m4 HHH. mvmm o no mwo HHH 9.: AonH Hym mow HHom >H mHV 00¢ 00mm m we NHm H» Hm HoH 0 mm omw pm mom NwNN. HHH me was“: vwva wH 3. Hub 0 0H4 mom mm 0H; 000m om omw hmmo HH on Hum momm NH va momH H mm 340 NH 00H :1: mm mum oHHm H HoooH ow. $0M wooH H Hum mNN H c on mm me, New mH HHVH mwm > $4 mHm onN mH Hun: vomH 5 cm com > o \qu mH H40 coo >H mo com HEN m 0H4 Hmm NH mm omw om 0o >00 mm 00H wwwH HHH co HwH... Hmwm m o 00 m MN m: mm moH moNH om . wwm owmm HH mo Awmm comm H4 mm on pH on. >3. H m NH. 34 mHHu :34 H 0922 Hp cwm ooHH mH H3 com m HVH 00 pm mm me on NmH wmm > FHH mmm mmom me T: 0.6,: 0H mo con Hum hmH oHoH oH wvH mooH >H wmm 022 Not. mHN hmH quVH c Q: ow» Ho mmm Chem mm com , wohm HHH NmH wHo owNN. H. E. oHo NM 2: ovw we HmN ooaH NHV vwv ovmm HH ow owe Choc w mm 00m 2 mm Ohm me m: cm: mm +4No¢ owwm H omoH 40002 44.02 40002 44.02 400.02 44.02 40002 44.02 40002 44.02 000N .00? 01308 00050 HH< 0000£HHde 000x000 H0008 00.00. vHoom .0200 .00 000000 H000 .0HHE .00 00030000 .0n00 u0nH 0900—80: Ho 05.00 CH 0HuoH00nH 00300200 HooH Ho 00000.30 0:0 .oooH Ho :3 6qu Ho 0085.90 6qu 0H» wGHusHu 003% 00000 H002 02H“ Ho 0000 £000 0H £0HH Ho 00008300 00.0505 H000 00330000 Ho Innmggm .mH 030R. 36 expressed as number per acre and number per river mile and the standing crop of biomass is expressed as pounds per acre. An evaluation of the estimation procedures was attempted in August of 1962. by conducting population estimates using the electro- fishing technique (with the Petersen method) and subsequently poison- ing the fish with rotenone. The site (in Zone II) selected for the study included four pools and was bounded at either end by shallow riffles. The station was 530 feet long and comprised 0. 59 acres. Block nets were employed both times in the manner described for the previous population estimates. The combination of low stream discharge (40 to 50 cfs) and low turbidity provided excellent conditions for the operations. Rainfall caused a delay in the poisoning, so that a week elapsed between the two estimates, which made possible the movement of some fishes into and out of the study section. A comparison of the results of the two estimates is presented in Table 16. The data from the poisoning are used as a base for percent error in each case. Friedman two-way analyses of variance (Siegel, 1956) were em- ployed to investigate the significance of the observed differences in the biomass estimates and in the population estimates for the five zones. No significant differences were found between the zones in the case of the total biomass, but the population estimates (number of fish per acre) were found to be significantly different at the 10% level. This suggests that the numbers of fish per acre and the age distributions may be different in the five zones and compensatory in such a manner that the biomass per acre is similar. That is, a large proportion of young fish in one zone would tend to yield a biomass estimate compar- able to one for a zone having a smaller number of fish, but with a larger proportion of older ones. Another interpretation is possible. The biomass estimate is the number of fish per acre multiplied by the mean weight of fish in the sample from which the population estimate 37 Table 16. Comparison of the numbers and weights of fish estimated by the Petersen method with the numbers and weights of fish recovered after poisoning the same area with rotenone. Rotenone Sample Petersen Estimate Percent Error Species Number Wei&ht* Number Weight* Number WeighL Rock bass 204 13.38 252 12.67 +24 -5 Smallmouth 116 10. 99 72 5. 58 -38 -49 White sucker 194 13.57 208 10.33 +7 -24 Hog sucker 141 18.84 165 10.85 +17 -42 Redhorse 87 10. 28 26 2. 88 -70 -72 Total suckers 422 42.69 399 24.06 -5 -44 Pounds per ac re . 38 was made. This results in a multiplicative operation on the sample variances. Therefore, the biomass estimates would have a much greater "within variance" than the population estimates have, which could obscure the differences between the zones in the case of the biomass estimates. That is, if repeated population and biomass esti- mates were run on the same population of fish, the biomass estimates would be more variable than the population estimates. The non-parametric tests were selected for the following reasons. The testing of parametric analyses of variance requires certain assumptions which these population and biomass estimates fail to satisfy. These assumptions are: (1) homogeneity of variance, (2) normality of distribution, and (3) lack of interaction. The hetero- geneity of variances was tested with Bartlett's test (Dixon and Massey, 1957) in the case of the biomass estimates and was found to be signifi- cant at the 0. 0005 level, which indicates that the assumption of homo- geneity of variances is not valid. The assumption of normality of distribution of the biomass esti- mates may be shown to be unjustified in the following manner. The distribution of individual weights of fish in a population is strongly skewed. The distribution of the mean weights in small samples, though having a tendency toward normality of distribution, will then be skewed. Consequently, the distribution of the biomass estimates, which are based on the mean weights of the fish in the samples, will also be skewed. It may be seen in Table 16 that although the population estimates both over- and underestimate the results from the poisoning, the biomass estimates all fall below the biomass recovered after poisoning. This substantiates the claim that the biomass estimates are drawn from a skewed distribution. The third assumption, that of linearity, is important when it may not be tested, which is the case when only one replication appears in 39 each cell of the matrix of observations. If replications are used, they must consist of the estimates obtained in the various sampling periods. But if these estimates from the various sampling periods constitute the replications, the estimate of internal variation (or "within variance") is so large that it obscures the difference that may exist between zones. If the parametric three-way analysis of variance is used, the main effects of the sampling periods may be tested, but it is then necessary to assume that no interactions exist. This is probably not true with respect to the relationship between the various habitats and the abundance of the fish species present. It is generally accepted that certain environments are more favorable for a given species of organism. This effect has been reduced by using the total biomass of suckers instead of the biomass of individual species, but the same consideration applies, to a lesser extent, to taxonomic families of organisms. Hence, the assumption of no interaction is invalid. Cooper and Lagler (1956) have shown that the use of total numbers over all age groups in the Petersen estimate tends to under- estimate the total population size. They suggest that the sum of the age-group estimates is less subject to the bias found in the other method. This is attributed to the fact that the efficiency of capture by any known method is dependent on the size of the individuals or some function of the size of the individuals. I suggest that a reason for this effect, at least in the estimates based on small samples, may be that there exists a minimum proportion of the population (here used in a statistical sense) which must be captured in order to provide an un- biased estimate of the population. Though the underlying distribution of estimates is continuous, i. e. , one may estimate that a part of a fish occurs within a given area, the number of fish handled follows a dis- crete distribution. Assuming that no correction factors were used 40 (i. e. , C + 1 or R + 1), then at least one recapture must be recorded, or the population estimate will be zero or infinity, if “ MC T - T where: T 2 estimate of population number, T, M= number of marked fish, C. and R ll number of fish captured on census run, number of recaptures. Since the number of recaptures follows a discrete distribution, the minimum value (other than zero) will be one. If a one is obtained, then the product MC will equal or exceed the total population number only if the maximum of the two values equals or exceeds the square root of the total population number. The product will equal the total population number if M = C = ’J—T—I , in which case the efficiency is equal to NIT/T. It was found that M and C were not extremely divergent within the estimates in the experimental data reported in this paper. This tends to support the claim that a real and somewhat consistent efficiency of capture exists. In order to have an unbiased estimate of a parameter, it is necessary that the expected value of the statistic (T in this case) agrees With the parameter. This will be realized in the Petersen estimate only if the efficiency is sufficiently great to ensure that R is large enough so that the effect of the discrete measure- ment is negligible. Under good conditions, the efficiency of collection for small rock bass is about 10%. Table 17 has been included to show the effects of selected efficiencies on population estimates in hypothetical populations. It may be seen that the bias is important at realistic levels of esti- mation. It is suggested that the minimum value of R which should be accepted for calculation of a valid estimate of T be 5. If this condition 41 Table 17. Population estimates by the Petersen and Bailey formulae for several hypothetical populations under various conditions of capture efficiency. E* ' .. Percent A Percent T* (percent) M(=C)>l< R96 T1* Error Tf‘ Error 50 5 3 1 6 88 9 82 10 5 1 15 70 25 50 15 8 1 36 28 64 28 25 13 3 46 8 56 12 100 5 5 1 15 85 25 75 10 10 l 55 45 100 0 15 15 2 80 20 113 13 25 25 6 93 7 104 4 1000 5 50 3 638 36 833 17 10 100 10 918 8 1000 0 15 150 23 944 6 978 2 25 250 63 980 2 992 l >rExplanation of symbols: T = number of fish in the population E = hypothetical efficiency of capture M = number of fish marked C = number of fish captured on census run (includes recaptures and non- recaptures) T1 = estimated number of fish in the population using the Bailey equation: ‘ __ M (C+1) - T1 ‘ R+1 T2: estimated number of fish in the population using the Petersen equation: .. MC T2: R 42 holds, it is also apparent that the bias introduced by the correction factor is unnecessary and the factor may be eliminated. The same argument may be applied to the estimation of age- classes because the term "population" was used in a statistical sense. Furthermore, it is contended that this condition exists with respect to the age-classes even though the data are lumped over all ages. That is, even though the value of R exceeds 5 in the overall estimate, if those portions of R which correspond to the ages for which E is small are less than 5, the estimate will be biased. A means of reduc- ing the effect of this bias is to increase T to the point where the collection efficiency results in an R value greater than 5. This, of course, is done by increasing the sampling area. A comparison of the biomass of fishes in the Red Cedar River with the biomass reported for several other rivers is presented in Table 18. The species of fish making up the categories entitled "suckers" and "bullheads" vary with location, but are all included in the families Catostomidae and Ictaluridae. The total biomass reported by the several authors has been adjusted to comparable values by sub- tracting the biomass of those groups not considered in the present study, i.e., minnows, darters, mudminnows, etc. Larimore (1961) provides a discussion of the difficulties encount- ered in the comparison of biomass estimates and measurements. Probably the most important consideration should be the methods em- ployed by the authors. Charles (1957) used rotenone in his sampling, though he occasionally preceded it with electric seine sampling. Larimore, working on a small stream, was able to pump most of the flow of the river onto adjacent fields. He followed this with the em- ployment of an electric seine and subsequent poisoning with rotenone. Gerking depended on seining short sections with minnow seines in an attempt to remove all of the fish. A subjective self-evaluation is 43 2 E m Si S. 88 S > 02 H8 0 33 we a: 2 >0 $0 a. 2 0.2 3 2.: mm B 0: E a 0E 04 $0; 0... 0 3 a: 2 $3 2 8i om H 3030 0000003 HoH A0m000>0v 003% 00000 00m :02 6.2.308 S E m 00$ 00 0 .8 0 .o .. 035 View 0.320 2.02 .3228 2 E N .o :t S 3: N - 0620 0588020 :02 .3028 03 2 3 S 2.3 40 8v o - 0.80 582 :02 6.86330 mo 83 mm a: 2 E m .. 035 52:2. 842 0:289 2 8v 0 £8 E E To - 0320 88:0 $42 05.289 2 so o 83 o E To - .830 833823 02. A: m. 88 EM 0: o N om 83 2 Si 3. 8V o 0 842 05089 0mm A0w0h0>0v 005.00 00Hm 00.qu em 8V o 38 MN 3: m m 3 30 m 88 $ 3: 2 2 $1: 0:239 mm H0m0n0>0v 00000 30m 30m 0\0o.H $3 0N0nHH Hokey 0N0QH Hem; 0\0nHH 003000 0000000000 H0008 0000rHHHsm 000x35 0000. 0300 000 500.05 00050HmH 0500.30 00000 Ems .000 0050000 005 000 0030 00000 00% 000 no“ 000505 H0000 000 000 00H00m0u00 5500000» 00.30. 5 000503. 000 00005300 00» «0 000300500 .wH 030B 44 included in his paper. Therefore, the results of all of the above authors may be looked upon as the minima for biomass in the section being studied. Extrapolation from the study area to the entire river is, of course, dependent on the variation introduced by subjectivity or randomness in site selection. Petersen estimates, on the other hand, may be higher or lower than the actual population and are sub- ject to the bias discussed above. Of the streams considered in Table 18, it may be seen that the average total biomass of the Red Cedar River was exceeded by that of only two streams: North Fork in Kentucky and Little Blue River in Indiana. Larimore (1961) quoted the results of total biomass esti- mates by several other workers, but the contribution of the various species were not reported. Assuming the proportion of the total bio- mass which consists of minnows and darters is the same in the Red Cedar River as that reported by Larimore (1961), the total biomass for the present study would be multiplied by 1. 74, or the mean total biomass would be 176 lbs/acre. Of those twenty-five streams reported by Larimore which are not considered in Table 18, eight were esti- mated to have a total biomass exceeding 176 lbs/acre. The distribution and abundance of the components of the sucker group is reported in Table 19 in terms of number per mile of stream. Of particular interest is the observation that the average number of white suckers per mile of stream is fairly consistent for the five zones. Therefore, the apparent increase of importance (percent of total) in succeeding upstream zones is primarily a function of the decrease in numbers of the other suckers. The hog suckers, on the other hand, appear with greater frequency in the lower parts of the river and reach their greatest abundance in Zone II. The habitat in Zone II agrees with those which Scott (1954) and Hubbs and Lagler (1955) report as 00N $0 00: NSN 1: H 880. o o o o 00 000 NN N0... 00 MS 8880080 o o 0N 03 N MN 0 N H 2 88:8 03800 0 «N o N 2 i: 00 00.: mm 00m 88:0 000 080082 N0 NpN 3 00m 0N 2... 2 mmm 0 mm .8800 8:03 00oMHmecoflo0200 :0 00w 0002 3: $0 00: 82 $2 880. o o o o N... 02 0 m: o o 8080000 o o 0 00 0 N0 0 o o o 808:0 03800 0 o o o NN EN .00 2: 2: $2 8028 000 080802 2: :1 No 0: N0 30 m 2. o o 800:... 8:03 0000 0N0 00: 02. $00 $3 880. o o o 0 mm 03 ON 2.: 00 002 808000.0 0 o 0 03 N 2 o 0 A 2 88:... 03800 N0 E H 0 0... 0mN 2. 3.2. A NN 88:... 000 080082 00 $0 2. 003 m NN S 0N0 0 H: 880... 8:03 :00 $2 5. 4 N0N 00 30 2.2 N0 880. o o o 0 N0 0mN 0N N00 2. 0N 8880090 o o 3. mm o o o o S 0 .8808 08800 N 0 o o 00 EN 00 000 S 0 .8808 000 080082 00 $0 0N NH 0 NN 2 N: 00 0N 88:... 82.3 008550 com; 02 0:: OSN 002 ON: 880. o o o o 3 00m: 0N 0N0 3. EN0 380080 o o 00 000 H 0N o o 0 ON .883 08800 0 2 o o H mm 0... 0N0 3 SN .8808 000 080082 2. 81 N0 NNv 0N 0N0 0N 000 Z 02 88:... 8:03 00m; 00 > .0 >0 0... H: .0 0 00 H @GON OGON QGON QGON OGON .32 00... .200 0.8 88:80 002 .003 fi 830 8080 08.0 20. .80 0838 Ho 0220 00m 0008.9: GM 00000Q0 00070.90 0.3030.) 000 m0 000005.50 00008300 0G0 cofldQQu—mflm .oH 030B 46 being suitable for hog suckers in this region. Redhorse did not appear in any collection made on the river upstream from the dam at Williamston. The most abundant fish (other than minnows and darters) in the Red Cedar River is the rock bass. It may be seen (Table 15) that there is a general tendency for the numbers of rock bass to diminish in an upstream direction. Zone 11 most consistently displays the highest number per acre and the largest standing crop of rock bass. Bullheads were widely distributed, but their contribution to the total biomass was generally small and displayed no consistent rank in the zones. Production E stimate 3 Net production was estimated for the centrarchid and catostomid fishes in the five zones. The figures represent net production in the Ivlev sense, i. e. , the total elaboration of biomass, whether or not it reaches the end of the period for which the estimate was made. The results, in dry weight units, are presented in Table 20 for all the species considered. Dry weight was found to equal 0. 242 x wet weight for these species. Table 21 includes these same estimates separated into the contributions made by each of the major groups. The Robson-Chapman estimate presupposes that the recruitment is similar in subsequent years, so that there is no effect of strong or weak year-classes. For that reason, the survival estimates were made by combining the data for the 'four years of the study. Although this will reduce the year-class effect, it precludes the possibility of replicating the estimates. Consequently, only one mortality estimate may be made for each species. This was assumed to be constant, though it is likely that a considerable error is introduced into the production estimates. It is believed that the mortality is greater in * Table 20. Estimated dry weight production for several species of fish in the Red Cedar River (units are milligrams dry weight per square meter per day). Zone Species I II III IV V White sucker 0.66 1.35 2.90 6.86 4.40 Northern hog sucker 1.03 6. 22 5.15 0.18 0. 39 Redhorse 3.06 2.33 5.43 0.00 0.00 Spotted sucker 0.02 0.03 0.96 4.42 0.00 Rock bass 2.92 3.89 2.86 1.60 1.75 Smallmouth bass 0.07 4.47 0.55 0.16 0.06 Green sunfish 0.00 0.14 0.06 0.06 0.00+ Bluegill 0.00+ 0.00 0. 31 0.00+ 0.04 Longear sunfish 0.00 0.00 0.00+ 0.00 0.00 Black crappie 0.00 0.00 0.00 0.18 0.00 Pumpkinseed 0.05 0.01 0.43 1.33 0.72 Largemouth bass 0.01 0.00 0.00 0.00+ 0.00 Warmouth 0.00+ 0.00+ 0.00 0.01 0.00 Total 7.83 18.45 18.65 14.81 7.36 >:< Wet weight = 4.13 x dry weight 48 :3 Table 21. Estimated dry weight production for the two major families of fishes considered in the Red Cedar River (units are milli- grams dry weight per square meter per day). Zone Family I II III IV V Mean Centrarchidae 3.06 8.51 4.23 3.36 2.58 4.35 Catostomidae 4.76 9.93 14.45 11.46 4.79 9.08 Total 7.82 18.44 18.68 14.82 7.37 13.43 * Wet weight = 4.13 x dry weight 49 the early periods of life. This implies that the production estimates would probably be increased if the true mortality could be applied. An error of unknown extent is introduced by the back—calcu- lation of the population estimates to May 1 using the estimate of instantaneous mortality rate. However, I believe that this error is small in relation to the error component in the variance of the Petersen estimate of population size. The effect in this case is likely to be one of reducing the estimate. This would result from using an annual mortality rate for that period from May 1 to some time during the summer or early fall. The actual mortality during this period is probably lower than that during the more harsh seasons of the year, which would also be reflected in the estimate of annual mortality. The error component in the estimate of instantaneous growth rate for individuals is small and the bias, if any, is probably quite small. The average biomass which represents an intermediate step in these calculations is not comparable to the average biomass estimates reported above, although the two are not independent in the sense that some of the same basic data were used in both calculations. Those reported above are arithmetic means of a set of simple estimates, while those which appear in the production estimate calculations are a geometric mean of the biomass over the year. Since only one estimate of the net production of each species in each zone is obtained, and because of the bias in the estimates, they are not amenable to a rigorous statistical analysis. In fact, due to the bias, even descriptive confidence intervals are meaningless. Therefore, these estimates may be considered as minimum absolute estimates of the true net production for the species considered. DISCUSSION 50 51 This study indicates that the Red Cedar River fish populations do not comprise one homogeneous community, but rather make up several incompletely separated communities in which the populations differ both in relative quantity and in certain qualitative characteristics, such as growth rate of individuals and productivity of the populations. Table 22 is a summary of the salient features of the fish populations and certain physical and biological characteristics of the study zones. Near the upstream end of Zone V, the river receives the effluent from a metal plating plant and organic enrichment from the Fowlerville urban area. The major part of this study was confined to the lower reaches of the zone where the fish populations are recovering from the effects of these effluents. Those sampling stations which fell further upstream, nearer the plating plant and sewage outfall , supported extremely low populations. In the spring of 1961, an extensive fish kill occurred in Zone V as a result of an accidental release of toxic materials from the metal plating plant. The influence of this kill could be recognized in the decreased fish populations as far downstream as Stow Road, about 6 miles downstream from the plating plant, in the summer sampling period. Zone V includes the widest variety of major habitat types of any of the study zones. They range from the slow-flowing, deeper, silty, dredged portions which make up the greater part of the zone to the shallow, gravel-bottom riffles and runs found in the vicinity of Webberville Road and Highway M47. The low gradient and subsequently slow flow of the river in most of this zone coupled with its small width (mean width: 34. 3 feet) are conducive to the retention and accumulation of fallen trees and brush. These, with the overhanging brush on the banks, provide excellent cover for larger rock bass and northern pike. Regardless of the organic enrichment from the Fowlerville area and the variety of the habitat types, fish production is low (see Table 22) Emofi .m .2 .28: 89:55 82a 33* 339: .Q Am 4.3: mag Soc 83* * om .N. m .mp hon: H.308 05$ “Ham com muoxodm o.¢mm Nm¢H 2.2 No.0 > mu; oJN ANN moon Mood - - - .zw...a..z ..... m. ”0.? ..... mwmw ..... 2 ME. ............................................................. £1: wfiv For mnvxodm ed; mmma mwmm cm; .5 co; 0.“: mwmfi moon— Moom - - - flaw» ..... m ”min: ..... mmww ..... z mam. ............................................................. mafia; mow. mow: muoxodm wdon moo woam «A»; HHH owd msvm XXL ammo. xoom w .............................................................................................. mvég o .2: 05mm. 130R. 2.5 N.w¢ Nmom muoxodm ~4on om: Twmm SUN 2 ow.m mom hmow ammo, xoom - - - Wade ...... o. ”.3. ..... wmmm ..... H wwmw- - - 41. .......................................................... 33¢ m.m~ om: muoxosm w.w~m wow “wow wo.m H Nod o.om wmqm mmma Moom 3.36 $90M\mo.5 032 mowoomm A>mp\~8\mfiv 33o Mom Flog NV Aoda\.ub oGoN \~E\wd£ mmmEoMm 1\ .02 3nd :033 ~S\Hmo Emv .Hd0>\nu3 unoflomhu coflodp . ucofifipom cofiosponm owpgomwm a. swarm new nounm uwcmmhofi *1? datum * 333% .3600 pad 05 mo moaou 03m 65 mo mofimwnouomuzfiu HmofimoHoE. paw €03.39 noguo Hoho>om £33 woummfiou 5mm mo coflosponm paw .mmmgofi .Hoorgn aoflmgmom mo woumESmo 65 mo >Hd§§w .mm 3an 53 and the species composition is rather limited. The dominant cato- stomid in Zone V is the white sucker, although it is less abundant here than in three of the four downstream stations (Table 19). This results from the fact that neither the redhorse nor the spotted sucker are found here and the hog sucker is present in only limited numbers. The absence of the redhorse from this Lzone and from Zone IV is not readily explained. It seems unlikely that the dam between Zones III and IV would provide a completely effective barrier for so many years against the invasion of the upper parts of the system. It is quite likely, though not necessary, that redhorse were present in the stream prior to the construction of the dam. The spotted sucker seems to prefer the more lacustrine type of environment found in the reservoir proper, which is the only part of the river where it is abundant. The habitats described above favor the rock bass, which is quite abundant in the lower part of the zone. But it appears that the domestic and industrial pollution from Fowlerville exerts a strong depressing effect on their abundance in the upper portion of this zone. The mean of the weight of individual rock bass is higher in this zone than in any of the other zones. This reflects the age distribution of the population more than an increased growth rate of individuals. Zone V supports a larger proportion of older rock bass. This could be due to any one or a combination of the following reasons. The habitat provides relatively more extensive cover for older fish than is found in the other zones. This situation could also be indicative of a senile and degenerat- ing population with poor recruitment. Or it might represent the results of sampling from an unexploited population which is compared with the rather heavily fished populations in the downstream zones. A similar age distribution might result from a high vulnerability of young fish to some predator with a reduction in vulnerability as age increases. Northern pike are quite common in this zone. An age distribution 54 similar to this one would be observed when sampling the preferred spawning areas of a population during the spawning period. However, the latter does not appear likely in view of the fact that rock bass spawn in the late spring and early summer, whereas these data were collected from June through September. It seems likely that the ob- served effect results from all of the above factors except for that of sampling from spawning grounds. Zone IV does not receive directly any extensive sewage or in- dustrial pollution throughout most of its length. The extreme lower end of this zone lies within the city of Williamston, from which it receives a limited amount of organic pollution, but the outfall of the Williamston sewage treatment plant enters the stream in Zone III about one mile below the end of Zone IV. Squaw Creek and Doan Creek both empty into the main stream in this zone. Neither of these tributaries drain urban or industrial areas, although the runoff from this intensively cultivated area would contain some agricultural chemicals. The reservoir proper comprises approximately half of Zone IV and its influence extends throughout the length. The flow is very slow and the bottom consists largely of mud and silt with occasional shallow sand flats in the upper half mile. The reservoir occupies a rather narrow and sinuous basin with a maximum depth of approximately 10 feet. Only the upper half (about two miles) of this zone was sampled for the fish study. The north side of the channel is quite steep through much of this portion, providing a run four to five feet deep immediately adjacent to the bank. Numerous logs with accumulated debris furnish cover for the larger pike and rock bass, while the deeper runs and pools are occupied by white suckers and spotted suckers. Extensive beds of several macrophytes (primarily Sagittaria sp.) offer protection for young fish of most of the species present. 55 In Zone IV, the dominant catostomid is the white sucker, which reaches its peak abundance here (see Table 19). The spotted sucker is also most abundant in this zone. The production of rock bass is lower in Zone IV than in any other, though the total production by centrarchids occupies a median position for the five zones (Table 21). This is due primarily to the presence of a large population of pumpkinseeds, which contributed 40% of the total centrarchid production (see Table 20). Pumpkinseeds, like the white sucker and spotted sucker, reach their greatest population density in this zone. It is believed that these species are numerous here due to the more lacustrine nature of the habitat. The scarcity of hog suckers may be attributed to a lack of the shallow sand flats and riffles that they occupy in other zones. On the other hand, the presence of largemouth bass and black crappies is a further indication of the lentic nature of the environment in this zone. Zone III receives the effluent of the sewage treatment plant in Williamston as well as direct seepage from the septic tanks and tile fields of the residences adjacent to the river. The latter problem is in the process of being corrected by extending the sewer facilities of the city. The sewage disposal plant applies excellent primary treatment so that the effect on the river is one of organic enrichment rather than foul pollution. The result is an extensive production of heterotrophic aufwuchs (King, 1964, Ph. D. thesis). Deer Creek drains into the main stream near the upper end of Zone III and contributes greatly to the turbidity of the river during periods of high water. The nutrients intro- duced into the main stream from this source remain low except during heavy rainfall and rapid runoff. However, the rise in nutrient levels during rapid runoff accompanies high water conditions when these nutrients are not efficiently utilized by the biota (Brehmer, 1958, Ph. D. thesis). Consequently, the main source of allochthonous nutrients is supplied by the sewage treatment plant, which empties about 2. 5 metric tons of phosphorus into the river annually (Brehmer, ibid. ). 56 The gradient in Zone III is slight and the bottom is relatively uniform with occasional deep pools. The bottom materials consist mainly of sand and cobbles with some silt and detritus. A few riffles are present near the lower end of the zone. Concealment for fishes is provided by extensive beds of rooted macrophytes (Vallisneria and A Sagittaria spp. ), but there are fewer accumulations of logs and other cover than in the upstream zones. The biomass and production of fish in Zone III are higher than in any other zone of the river. This may seem incongruous with the rate of primary production unless the amount of allochthonous nutrients is considered. King (1964, Ph. D. thesis)~describes the unusual con- ditions found here. He estimates that the production of heterotrophic aufwuchs in this zone is 885 g-cal/mz/day, which is more than 3. 5 times as much production as occurs in Zone V, which most nearly equals it. In Zone III, redhorse are the most important species with respect to the standing crop of biomass and rate of production. This zone seems to provide the most nearly optimum conditions for catostomids to be found in, the Red Cedar River; they reach their greatest abundance and production here. This situation is especially true for the redhorse and hog suckers, although white suckers may also be found here in large numbers. The standing crop and production of centrarchids are high in this zone, but are not as great here as in Zone 11. This could be due to the reduced availability of suitable cover, which prevents them from making full use of the abundant food. Rock bass are the dominant centrarchid and the pumpkinseed, though present, is relegated to a minor position. Zone II is the cleanest of the five study zones. It receives almost no organic or industrial pollution except for the influence of Okemos in the last half mile before entering Zone I. The bottom consists of sand, 57 boulders, and gravel with a large number of riffle-pool combinations. Extensive sand flats occur in the lower end of the zone, but the gradient is high and provides a current sufficiently swift to prevent much silt deposition. Valisneria sp. appears to be invading the area from Zone III and may be 4found in thick beds through much of the upstream end of the zone. Runoff from the construction sites of a major highway in the southern portion of the Red Cedar River drainage system in the summer of 1961 introduced a considerable amount of sand and clay into the main stream. This has most noticeably affected Zone II where it is responsible for the filling of many pools. In the subse- quent low water years, this material has been somewhat stabilized and may represent a permanent alteration of the environment. The dominant catostomid in Zone II is the hog sucker, which is abundant throughout the zone. Redhorse and white suckers are present in fewer numbers, but contribute a significant part of the biomass. The greatest production of centrarchids occurs here, representing the highest proportion of the total production to be found in any part of the river. Of particular importance is the status of the smallmouth bass. This is the only zone where its production exceeds that of the rock bass; indeed, the only zone where it contributes a significant part of the total production. Vannote (1963, Ph. D. thesis) presents an excellent treatise on the ecology of the smallmouth bass in a part of Zone II. He notes that in the recent past the Red Cedar River was considered to be an excellent smallmouth bass stream. However, the stream has now reached a stage of degradation where only this part of the river maintains a semblance of its former stature. He also indicates the importance of a high density of riffle-pool combinations for the main- tenance of a highly productive smallmouth bass population. It is likely that the destruction of some of the pools in this zone by the transport of sand and clay from the highway construction site to the main stream 58 has resulted in permanent damage to the smallmouth bass population. However, this zone still represents the most desirable part of the river from the standpoint of the sport fisherman. Zone I receives organic and some small industrial pollution from Okemos at the upper end as well as seepage and raw sewage from the urban development near and in East Lansing at the lower end. The high gradient (Table 22) is obscured by the presence of the dam just below the downstream boundary of the zone. Since the river is wide in this section and the effective gradient is low, the current is slow. In the upstream end of the zone, the bottom is largely sand and rubble which gives way to mud and silt in the middle portion. The figure for inorganic sedimentation rate for Zone I in Table 22 is likely to be an underestimate due to sampling problems (Darrell King, personal communication). In a river system, the transport of nutrients and other materials is unilateral, with some few exceptions, though the rate of transport may be very low in the case of some biologically utilized materials. It would seem likely, then, that the cumulative effects of all the allochthonous energy would eventually be evident in that zone which is furthest downstream. However, it is necessary to consider that any toxic materials added to any part of the stream would also eventually reach the downstream zone provided they were not degraded or removed biologically or became unavailable in the bottom sediments. If the cumulative effects described above result in an already marginal water quality, then the addition of even a slight amount of pollution may be sufficient to make the waters unsuitable for fish populations . This appears to be the case in Zone I of the Red Cedar River. It may be seen from Table 22 that the primary production in this zone is the lowest in the river. King (1964, Ph. D. thesis) reports that insect production is also lower in Zone I than in any other part of the river. 59 It may be expected that heterotrophic production would offset the low primary production, but he shows that this is not the case. Therefore, it is not surprising to find that the production of fish is also very low in this zone. Early in the study, the biotic community included fairly large populations of rock bass and several species of suckers. Later (in 1961) a sharp reduction in the rock bass population was observed as well as a noticeable downward trend in the sucker populations. Subsequent observations indicate that this trend has not reversed. The natural eutrophication of a river proceeds in an upstream direction. It is believed that in the Red Cedar River, this eutrophication has been accelerated by the activities of man and is resulting in the rapid degradation of the stream as a productive community. Continued urban development in the watershed seems certain, but additional im- pairment of the water quality is likely to render the river unsuitable for the support of fish populations. SUMMARY 60 61 1. The fish populations in the Red Cedar River, a warm-water stream in central Michigan, were investigated to determine any dif- ferences which might exist between five selected zones in a 30-mile study section of the river. 2. A movement study involving tagged fish demonstrated that at least a large portion of the rock bass population exhibits limited movement. 3. Comparative studies included growth rates, standing crop in numbers and weights, and production of the major fish populations in the five zones. 4. Age determinations were made for approximately 3100 fish from scale samples using the acetate impression method. 5. A significant difference in the growth rates of rock bass and hog suckers for the five zones was demonstrated. Both species showed better growth in the upstream zones. No significant differences in the growth rates of the white suckers and redhorse were demonstrated. 6. Population numbers and biomass were estimated by the Petersen method and the mean weight of fish in the samples using an electro-fishing technique. 7. Non-parametric statistical analyses indicated a significant difference in the numbers of fish per acre but not in the standing crops of biomass. The mean standing crop of biomass for the five zones was estimated to be 101 lbs. /acre with the individual estimates ranging from 181bs../acre to 338 lbs./acre. 8. A source of bias in the Petersen estimate using small samples was shown to be the effect of estimating a continuous distribution with a disc rete distribution. 62 9. Net production of the major fish populations was estimated by the technique of Ricker with some modifications. 10. The total mean net production of the major fishes in the Red Cedar River was estimated to be 13.43 mg dry weight/mz/day in the growing season with a range of 7. 37 mg dry weight/mz/day to 18.68 mg dry weight/mz/day. LITERATURE CITED 63 64 Ball, R. C. 1947. A tagging experiment on the fish population of Third Sister Lake, Michigan. Trans. Am. Fisheries Soc., 74: 360-3690 Beckman, W. C. 1943. Annulus formation on the scales of certain Michigan game fishes. Papers Mich. Acad. Sci. A. L., 28: 281-312. Brehmer, M. L. 1956. A biological and chemical survey of the Red Cedar River in the vicinity of Williamston, Michigan. Unpub. Master's thesis, Mich. State Univ. Lib., 51 pp. . 1958. A study of nutrient accrual, uptake, and regeneration as related to primary production in a warm-water stream. Unpub. Doctor's thesis, Mich. State Univ. Lib., 97 pp. Brown, E. H. 1960. Little Miami River headwater-stream investi- gations. Ohio Dept. Nat. Res., 143 pp. Charles, J. R. 1957. Final report on population manipulation studies in three Kentucky streams. Fish. Bull. 22, Kentucky Dept. Fish and Wildl. Res. , 45 pp. Cooper, G. P. and K. F. Lagler. 1956. Appraisal of methods of fish population study - Part III: the measurement of fish population size. Trans. let N. A. Wildl. Conf., 281-297. Dixon, W. J. and F. J. Massey, Jr. 1957. Introduction to statistical analysis. 2d Ed. McGraw-Hill Book Co., Inc., New York, 488 pp. Funk, J. L. 1957. Movement of stream fishes in Missouri. Trans. Am. Fisheries Soc., 85: 39-57. Gerking, S. D. 1953. Evidence for the concept of home range and territory in stream fishes. Ecology, 34: 347-365. . 1959. The restricted movement of fish populations. Biol. Reviews, 34: 221-242. . 1962. Production and food utilization in a population of bluegill sunfish. Ecol. Monogr., 32: 31-78. 65 Grzenda, A. R. 1960. Primary production, energetics and nutrient utilization in a warm-water stream. Unpub. Doctor's thesis, Mich. State Univ. Lib., 99 pp. Handbook of biological data. 1956. W. B. Saunders Co., Philadelphia. 584 pp. Hile, R. O. 1941. Age and growth of the rock bass, Ambloplites rupestris (Rafinesque), in Nebish Lake, Wisconsin. Trans. Wisconsin Acad. Sci. A. L., 33: 189-337. Hubbs, C. L. and K. F. Lagler. 1958. Fishes of the Great Lakes region. (Rev. ed.) Bull. No. 26, Cranbrook Inst. Sci. 213 pp. King, D. L. 1962. Distribution and biomass studies of the aquatic invertebrates of a warm-water stream. Unpub. Master's thesis, Mich. State Univ. Lib., 96 pp. . 1964. An ecological and pollution-related study of a warm- water stream. Unpub. Doctor's thesis, Mich. State Univ. Lib., 154 pp. Kramer, C. Y. 1956. Extension of multiple range tests to group means with unequal numbers of replications. Biometrics, 12: 307'3100 Larimore, R. W. 1961. Fish population and electrofishing success in a warm-water stream. J. Wildl. Mgmt., 25: 1-12. Rawstron, R. R. 1961. A study of primaryproduction on artificial substrates in a riffle and pool area of the Red Cedar River. Unpub. Master's thesis, Michigan State Univ. Lib. . 73 pp. Ricker, W. E. 1958. Handbook of computations for biological statistics of fish populations. Fish. Res. Ed. Canada. Bull. No. 119: 1’3000 . Robson, D. S. and D. G. Chapman. 1961. Catch curves and mortality rates. Trans. Am. Fisheries Soc., 90: 181-189. Rounsfell, G. A. and W. H. Everhart. 1953. Fishery Science: its methods and applications. John Wiley and Sons, Inc. , New York. 444 pp. 66 Scott, D. C. 1949. A study of a stream population of rock bass, Ambloplites rupestris. Invest. Indiana Lakes and Streams, 3: 169’2340 Scott,‘ W. B. 1954. Fresh-water fishes of eastern Canada. Univ. of Toronto Press, Toronto. 128 pp. Siegel, S. 1956. Nonparametric statistics for the behavioral sciences. McGraw-Hill Book Co. , Inc. , New York. 312 pp. Smith, S. H. 1954. Method of producing plastic impressions of fish scales without using heat. .Prog. Fish Culturist,‘ 16’: 75-78. Snedecor, G. W. 1956. Statistical methods. Iowa State Univ. Press, Ames, Iowa. 534 pp. Tompkins, W. A. and B. T. Carter. 1951. The growth rates of some Kentucky fishes. Kentucky Div. Game and Fish, Fish. Bull. N0. 6. 9 pp. Vannote, R. L. 1961. Chemical and hydrological investigations of the Red Cedar River watershed. Unpub. Master's thesis, Mich. State Univ. Lib. , 126 pp. . 1963. Community productivity and energy flow in an en- riched warm-water stream. Unpub. Doctor's thesis, Mich. State Univ. Lib. , 156 pp. APPENDICES 67 APPENDIX A List of names* of fishes encountered in the Red Cedar River during the course of this study. Scientific name Esocidae Esox americanus vermiculatus LeSueur E. lucius Linnaeus Cyprinidae Cyprinus ca rpio Linnaeus Catostomidae Catostomus comm e rsoni (Lac epede) Hypenteliurn ni g ricans (LeSueur) Minjtrema melanoE (Rafinesque) Moxo stoma spp. Ictaluridae Ictalurus melas (Rafinesque) _I_. natali s (LeSueur) Centrarchidae Ambloplite s rupe stri s (Rafine sque) Chaenobryttu s gum sus (Cuvie r) Lepomi s c yanellus Rafine sque 1.1. gibbo sus (Linnaeus) L. mac rochi rus Rafine sque g. me galoti s (Rafine sque) Mic ropte rus dolomieui Lac epede M. salmoide s (Lac epede) Pomoxi s nig romaculatus (LeSueur) 9,: Common name Gras 5 picke rel Northern pike Carp White sucker Northern hog sucker Spotted sucker Redhorse Black bullhead Yellow bullhead Rock bass Warmouth Green sunfish Pumpkinseed Bluegill Longear sunfish Smallmouth bass La rgemouth ba 8 8 Black c rappie American Fisheries Society. 1960. A list of common and scientific names of fishes from the United States and Canada. Special Publ. No. 2. 101 pp. 68 APPENDIX B Mean back-calculated lengths (in centimeters) at annulus formation for thirteen species of fish in the five zones of the Red Cedar River. Annulus Zone Sjecie s numbe r I II III IV V Rock bass I 3.91 4.09 4.31 4.22 4.05 (406)* (641) (290) (194) (114) II 6.61 7.13 7.80 7.45 7.32 (251) (413) (213) (129) (91) III 10.72 10.97 11.55 12.26 11.82 (103) (139) (128) (74) (67) IV 14.67 14.49, 15.19 15.99 15.40 (49) (70) (73) (25) (19) V 17.34 16.63 17.45 18.74 17.56 (11) (21) (21) (8) (11) VI 20.3 16.0 20.23 20.35 18.90 (1) (1) (4) (2) (4) VII 22.0 20.45 (1) (2) VIII 22.0 (1) White sucker I 11.02 10.55 11.09 11.08 11.21 (16) (41) (54) (114) (100) II 18.80 16.70 16.83 19.25 18.61 (4) (22) (35) (84) 7 (75) III 24.1 20.70 23.99 28.96 26.91 (1) (5) (19) (62) (37) IV 18.9 31.42 34.86 29.92 (1) (5) (21) (9) V 25.0 35. 35 30.90 (1) (2) (Z) Redhorse I 6.02 6.33 6.36 (121) (88) (149) II 10.26 11.30 11.60 (63) (63) (139) III 17.59 17.97 18.94 (26) (47) (123) IV 23.81 24.32 25.99 (13) (37) (93) V 27.72 28.71 31.25 (6) (16) (64) VI 31.00 33.35 34.90 (4) (10) (29) VII 35.9 34.90 39.35 >3 (1) (2) (11) Figures in parentheses are the numbers of fish in the samples from which the mean was calculated. 70 Appendix B - Continued Annulus , Zone Species number I II III IV V Spotted sucker I 7.55 7.95 8.20 8.38 (2) (2) (13) (38) II 11.4 14.60 15.90 (1) (7) (35) III 14.8 22.77 25.37 (1) (6) (27) IV 24.3 29.75 32.31 (1) (4) (19) V 42.7 38.09 (1) (13) VI 41.17 (6) VII 42.5 (1) Smallmouth bass I 8.79 8.55 8.80 8.60 8.60 (15) (18) (16) (3) (3) II 16.0 14.34 13.73 12.70 11.75 (1) (12) (13) (2) (2) III 21.24 18.95 18.80 16.30 (8) (12) (Z) (2) IV 33.1 23.6 (1) (1) Green sunfish I 2.7 3.17 3.36 2.48 4.10 (1) (6) (9) (5) (2) II 7.7 8.65 7.73 5.10 (1) (2) (6) (2) Bluegill I 4.40 4.23 3.05 3.7 (2) (12) (4) (1) II 7.00 8.86 9.2 (2) (10) (1) III 10.5 10.5 (1) (1) Longear sunfish I 3.35 (2) II 6.50 (2) Black crappie I 6.75 7.1 (2) (1) 11 10.75 11.4 (2) (1) 111 16.85 (2) IV 20.6 (1) Appendix B — Continued 71 Annulus Zone Specie s numbe r I II III IV V Warmouth I 6.55 6.1 6.6 (2) (1) (1) II 9.00 7. 3 9.0 (2) (1) (1) 111 10.10 10.2 (2) (1) IV 11.75 (2) Largemouth bass I 8.1 4.68 (1) (4) II 19.7 (1) Pumpkinseed I 5.36 6.4 5.42 5.40 5.55 (7) (1) (80) (70) (15) II 7.0 8.4 8.19 8.14 9.37 (1) (1) (76) (20) (3) 111 10.00 10.45 14.0 (7) (4) (1) IV 12.35 (2) Northern hog sucker I 8.43 8.74 9. 20 5.53 9.45 (114) (236) (92) (2) (10) II 13.07 13.72 15.04 18.0 17.99 (85) (163) (61) (l) (10) III 18.86 19.65 21.52 28.4 25.26 (70) (110) (47) (1) (5) IV 24.54 24.61 26. 30 31.15 (34) (31) (11) (4) V 29.02 28.50 29.03 (6) (9) (4) VI 31.95 29.25 (4) (2) VII 34.95 (2) 111383 8 40f .: (“9‘ .53": I ,~. I I "I1111111.11111111111111s