*::$tº | **, \ . g MORTALITY, GROWTH, AND YIELD OF CHANNEL CATFISH IN SAGINAW BAY, LAKE HURON by Raymond J. Haak A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Natural Resources School of Natural Resources The University of Michigan 1987 Thesis Committee: Associate Professor James S. Diana, Chairman Adjunct Professor William C. Latta Ex-officio examiner Richard D. Clark, Jr. ABSTRACT In Saginaw Bay, Lake Huron a higher catch per unit effort of Channel catfish Ictalurus punctatus in recent years has prompted commercial fishermen to request the Michigan Department of Natural Resources to allow them to increase their fishing effort. Previous studies on the potential for increasing commercial harvest were incomplete or are dated, so a new study was needed. Of 2,460 channel catfish tagged in the Wild Fowl Bay area of Saginaw Bay in 1985, 277 tags were returned by sport and commercial fishermen. A catch curve constructed using data collected by sampling catfish from commercial trap nets in three management grids in the bay revealed that instantaneous total mortality rate for catfish in Saginaw Bay was 0.45, down from 0.67 in 1971 and 1981. An instantaneous fishing mortality rate of 0.26 and an instantaneous natural mortality rate of 0.19 were determined by partitioning the instantaneous total mortality rate using tag returns. A commercial reporting rate of 45% was determined by sampling the commercial catch for unreported tagged fish. Comparison of mean backcalculated lengths at age with previous studies revealed a decline in growth rate of catfish. A dynamic pool model indicated commercial fishing effort could be increased to obtain greater yields. Greatest increases in yield could be achieved by increasing both the commercial size limit and effort. Sensitivity analysis indicated that instantaneous total mortality rate, commercial reporting rate, and von Bertalanffy growth parameters were most important in determining the levels of commercial fishing effort and commercial size limit to maxmize yield. Increasing the commercial size limit to 406.4 mm (16 in) was recommended to increase total, commercial, and sport yield per 1000 recruits, 2.6%, 0.6%, and 9.1%, respectively. iii ACKNOWLEDGEMENTS This thesis is a result of work sponsored by the Michigan Sea Grant College Program under grant No. NA-80AA-D-0072, project No. R/GLF-17 from the National Sea Grant College Program, National Oceanic and Atmospheric Administration (NOAA), United States Department of Commerce, and funds from the State of Michigan. Michigan Sea Grant also publicized the study. The Michigan Department of Natural Resources (MDNR) provided equipment. Colleen Pickett, David Rivard, Frank Shivley, and Paul Longfellow aided in field work. Ken, Kenny, and Brian Dutcher provided equipment and assistance. Forrest, Tod, and Connie Williams, Dennis Root, Randy Carter, and others at Bay Port Fish Company were helpful in collecting data and kept me stimulated in my research. They also made my three summers in the Thumb enjoyable. Bay Port Fish Company provided equipment and working space. Robert Lorantas provided information regarding methods he used in his study, insightful comments, and encouragement. Sam Morgan, Richard Beardsley, Warren Beers, and Sandy and Louie Whyte provided space and time on their boats so that I could measure fish and/or collect spines from fish. The Pigeon Conservation Club, Sebewaing Sportsmen's Club, and Bay Port Fish Company each provided $100 for the lottery prizes. Ed Dootz provided laboratory space, in the Dental Materials Department of the Dental School at the University of Michigan, as well as use of a dental lathe and blades for sectioning spines. Don Nelson and John Weber of the MDNR provided commercial catch data and information on commercial fishermen. Tony Frank and Scott Nelson of the U. S. Fish and Wildlife Service provided commercial effort data. All the members of my committee, Drs. Richard D. Clark, Jr., James S. Diana, and W. Carl Latta deserve thanks as without them this project would not have been possible. Drs. Latta and Clark helped with all aspects of this study. Their comments were indispensable in shaping this thesis. A very special thank you to Dr. Diana, chairman of my committee and academic advisor, for his guidance and support. He made my education and research at the University of Michigan possible. Finally, I would like to thank my parents and sister for their support; without them I would have never completed college. TABLE OF CONTENTS ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . iv LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Age Structure and Size Composition of Catch . . . . . . . . . . . . 5 Growth Analysis . . . . . . . . . G - e º tº º G & © tº G & © tº ſº tº 8 Tagging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Mortality Rate Estimation . . . . . . . . . . . . . . . . . . . . . . . 12 Yield Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Age Structure and Size Composition of Catch . . . . . . . . . . . . 17 Growth Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . e 17 Tagging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Mortality Rate Estimation . . . . . . . . . . . . . . . . . . . . . . . 24 Yield Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . 42 LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 vi Table LIST OF TABLES Number of fish in samples from commercial trap nets in four management grids in Saginaw Bay, 1985. . . . . . . . . . Number of channel catfish tagged each month in Saginaw Bay, 1985. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Number of tags returned by sport and commercial fishermen in 1985 and 1986. . . . . . . . . . . . . . . . . . . . Average percent change from existing conditions in FC (max) and , (max) with corresponding yields when input parameters of yield model are varied 3-10%. . . . . . . . Instantaneous total mortality rates (with 95% confidence intervals when calculated) for channel catfish from Saginaw Bay and other selected locations. . . . . . . . . 6. Mean backcalculated total length (mm) at age for channel catfish from Saginaw Bay and other selected areas. . . . . . . Tag reporting rates of sport fishermen for various species. . . . Page 23 23 31 33 vii LIST OF FIGURES Figure Page 1. Catch per unit effort of channel catfish by commercial fishermen in Saginaw Bay, 1971-85 (see Hile (1962) for detailed information on effort units). . . . . . . . . . . . . . e 2 2. Management grids in Saginaw Bay where commercial trap nets were sampled in 1985. . . . . . . . . . . . . . . . . . . 6 3. Sites where channel catfish were tagged in the Wild Fowl Bay area of Saginaw Bay, 1985. Locations are marked by number of fish tagged at each site. . . . . . . . . . . . . . . 10 4. Length frequency distributions for channel catfish sampled from commercial trap nets in four management grids in Saginaw Bay, 1985. . . . . . . . . . . . . . . . . . . . . . . 18 5. Age frequency distributions for channel catfish sampled from commercial trap nets in four management grids in Saginaw Bay, 1985. . . . . . . . . . . . . . . . . . . . . . . 19 6. Comparison between lengths at age determined from the von Bertalanffy equation and actual lengths measured from channel catfish from Saginaw Bay, 1985. . . . . . . . © 20 7. Locations where tagged channel catfish were recaptured and reported by sport and commercial fishermen in 1985 and 1986. . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 8. Catch curve constructed using combined data from management grids 1507, 1509, and 1606 of Saginaw Bay, 1985. Regression line fitted to ages 6 to 14 of the descending limb is shown. . . . . . . . . . . . . . . . . . . . 25 viii Figure Page 9. Yield isopleth diagram for commercial yield per 1000 recruits of channel catfish from Saginaw Bay, 1985. Instantaneous natural and sport fishing mortality rates were 0.19 and 0.05, respectively. The commercial fishery was operating at point x in 1985. . . . . . . . . . . . . . . . 27 10. Commercial and sport yield for channel catfish from Saginaw Bay as a function of commercial size limit. Instantaneous commercial fishing mortality rate is 0.21. . . . 28 11. Commercial and sport yield for channel catfish from Saginaw Bay as a function of commercial fishing mortality rate. Instantaneous commercial fishing mortality rate is 0.21. . . . . . . . . . . . . . . . . . . . . . 30 INTRODUCTION Saginaw Bay is a relatively shallow, productive bay on the western side of Lake Huron. The bay supports extensive sport and commercial fishing. The channel catfish lctalurus punctatus is an important species to the fisheries of Saginaw Bay. It usually ranks first or second in weight caught annually by commercial fishermen and has ranked as high as second in numbers of fish caught annually by sport fishermen. In recent years catch per unit effort (CPUE) of channel catfish by commercial fishermen has increased (Fig. 1). Sport fishermen have also been catching increasing numbers of channel catfish outside the bay (John Weber, Michigan Department of Natural Resources, personal communication). An earlier study on catfish in Saginaw Bay showed a decrease in the growth rate of catfish in the bay (Lorantas 1982). Increasing CPUE, increasing catch outside the bay, and decreasing growth rate all imply the population of channel catfish in the bay may be increasing. For the past several years commercial fishermen have requested the Michigan Department of Natural Resources (MDNR) to allow them to increase their effort on channel catfish. If the population of catfish is indeed increasing effort could be increased without a harmful effect on the population. Two previous studies were conducted on the catfish fisheries of Saginaw Bay to determine if commercial harvest could be increased. The first in 1971 determined that a decrease in effort and an increase in the commercial size limit would maintain the current harvest while allowing 250 tº -O- Trap Net 200 — -A- Seine – H. Set Hooks 150 – ſy tº- ry ry ry rv rv 100 – 50 – A O | | | | | | | | | | | | | | | 1971 1975 1980 1985 Year Figure 1. Catch per unit effort of channel catfish by commercial fishermen in Saginaw Bay, 1971-85 (see Hile (1962) for detailed information on effort units). the reproductive population to increase (Eshenroder and Haas 1974). The authors believed the larger reproductive population would stablize recruitment between years. Eshenroder and Haas (1974) estimated that a harvest of 136,079 kg (300,000 lbs) would be possible by 1980 if the commercial size limit was increased from 381 mm (15 inches) to 432 mm (17 inches). This size limit was recommended because most of the fish they examined less than 432 mm were immature. Even though no regulation change was made the annual harvest easily exceeded 136,079 kg by 1976 and has increased since without any apparent damage to the population. The second study in 1981 came to similar conclusions but found that the management actions necessary to obtain the maximum harvest depended on the level of the sport harvest and the natural mortality rate used in the yield analysis (Lorantas 1982). The only estimate of the sport harvest was from a mail survey administered by the MDNR. This estimate was likely inaccurate, as has been shown with other mail surveys (Rybicki and Keller 1978), so the estimate of yield may have been in error. In addition, both studies assumed that the natural mortality rate of catfish in Saginaw Bay was the same as the natural mortality rate of catfish in Lake St. Clair. However, the applicability of this natural mortality rate to the catfish population of Saginaw Bay was unknown. Thus, a new study, directed at more accurate estimates of fishing and natural mortality rates for the channel catfish population of Saginaw Bay, was deemed necessary. The objectives of this study were: (1) to determine the instantaneous total mortality rate for the channel catfish population in Saginaw Bay, (2) to determine the instantaneous fishing and natural mortality rates for the catfish population, and (3) to determine the effects of various management actions on yield. METHODS Age Structure and Size Composition of the Catch Data for determining age and length frequency distributions were collected by sampling commercial trap nets in four management grids in Saginaw Bay (Fig. 2). Total lengths of catfish caught in commercial trap nets were measured to the nearest millimeter (Table 1). The left or right pectoral spine of five fish per 25 mm length interval was collected from fish measured for length frequency calculations or from fish caught in the same area. Spines were sectioned at the distal end of the basal groove (Sneed 1951) using a dental lathe with two aluminum oxide blades separated by a 0.52 mm spacer. The spine sections were mounted onto acetate slides with cyanoacrylate glue. The sections were magnified using a microfiche projector allowing annuli to be counted and annular radii to be measured. Radii were measured from the approximated center to the distal anterior point. More detailed information on spine preparation and aging is given by Lorantas (1982). A BASIC computer program was written to convert measured lengths into ages. The program first sorted lengths of aged fish into 25.4 mm length intervals. For each length interval the proportion of fish in each age class was calculated. Lengths of catfish measured from commercial trap nets were then converted to ages by multiplying the number of fish measured in a specific length interval by the proportion of fish in each age class of that length interval. The total number of fish at each age was then calculated by summing the number of fish in each age class for all lengths. Point Lookouts, - - ~ *-NJ - —ſ 1509 J- 1507 | sº Port & 1608 / SAGINAW BAY H O 10km Figure 2. Management grids in Saginaw Bay where commercial trap nets were sampled in 1985. Table 1. Number of fish in samples from commercial trap nets in four management grids in Saginaw Bay, 1985. Date Grid Number of Number of - fish measured spines collected August 2 1507 408 103 June 5 & October 31 1509 868 150 October 22 1606 650° 106 November 13 1606 O 52b June 24 & October 30 1608 202 127C * Only a portion of the fish in some nets were measured. *Additional spines were collected from the same area of this grid. . ° Most of the spines collected from fish caught in this grid were not from the fish measured for length frequency data. Growth Analysis Backcalculated lengths at each age were determined for 237 fish to calculate parameters of the von Bertalanffy equation and compare growth with previous studies. Annular measurements were transformed into lengths using a modification of the direct proportion method (Bagenal and Tesch 1978): L = (S/S) (Le- a) + a, where L = total length of fish when annulus iformed, Le = total length of fish at capture, S = spine radius at annulus i, S = total spine radius, a = intercept of regression equation when Le regressed on S for all fish. Parameters for the von Bertalanffy equation were calculated from mean backcalculated lengths at age using the method of Rafail (1973). This method involves determining K as the slope of the regression equation when the natural logarithm of the annual growth increment per unit of age is regressed on age. The other parameters are then calculated using a set of equations derived from the von Bertalanffy equation. A length-weight equation was calculated from lengths and weights of catfish sampled from the commercial catch. The length-weight equation was linearized using a logarithmic transformation: Ln W = Ln a + b Ln L, where W = weight measured to the nearest 0.001 kilograms, L = total length in millimeters, Ln a = y-intercept, b = slope. Tagging Channel catfish were obtained for tagging using small trap nets set near shore and to depths of about 3.5 meters in the Wild Fowl Bay area of Saginaw Bay (Fig. 3). Tagging occurred from May through November 1985. From May through September, only catfish greater than the legal commercial size limit of 381 mm (15 inches) were tagged. After September, some fish between 355.6 mm (14 inches) and 381mm were tagged. Floy FD-68B anchor tags, encoded with a unique number and the address for the MDNR office in Mt. Clemens, Michigan, were implanted slightly below the dorsal fin on the left side of the catfish. Each fish was measured prior to release. In addition, the adipose fin was clipped so that a tag loss rate could be determined by later examining fish in the commercial catch. Tags were recovered by sport and commercial fishermen. A lottery with a $200 prize for sport fishermen and a $100 prize for commercial * 100 – O | ! lſ T | | I | I | I ſ ſ | | | ſ | 300 400 500 600 700 800 Commercial Size Limit (mm) Figure 10. Commercial and sport yield for channel catfish from Saginaw Bay as a function of commercial size limit. Instantaneous commercial and sport fishing mortality rates are 0.21 and 0.05, respectively. 28 maximized when no commercial fishing existed. If size limit was increased from 381 mm to 511 mm, commercial yield would decrease 5.0% and total yield would increase 7.4%, while sport yield would increase 48.8% (Fig. 10). An increase in the size limit to 416 mm would increase total yield 3.4%, commercial yield 0.7%, and sport yield 12.5%. At the current commercial size limit (381 mm), the predicted commercial and total yield per 1000 recruits was maximized at a FC of 0.42 and 0.22, respectively. If FC was increased to 0.42, commercial yield would increase by 6.8% while total and sport yield would decrease by 3.3% and 37.0%, respectively (Fig. 11). Increasing FC to 0.22 would increase total yield per 1000 recruits by 0.02% and increase commercial yield by 0.9%. Sport yield would decrease by 2.9%. In the sensitivity analysis, variations in commercial reporting rate, instantaneous total mortality rate, and von Bertalanffy parameters had a large impact on FC (max) and , (max) (Table 4). Variations in sport reporting rate did not affect FC (max) and , (max) for commercial yield because of the way in which the natural mortality rate was calculated. Variations in tag loss rate and length at recruitment had a small effect on FC (max) and I (max). Thus, the accuracy of instantaneous total mortality rate, commercia reporting rate, and von Bertalanffy parameters determine reliability of this yield analysis. Obviously, improving the accuracy of these factors would increase confidence in the yield predictions. 29 m Commercial Sport O I | ſ T I I | T I | | | ſ ſ T | I T I ſ 0.0 0.5 1.0 1.5 2.0 Instantaneous Commercial Fishing Mortality Rate Figure 11. Commercial and sport yield, for channel catfish from Saginaw Bay, as a function of commercial fishing mortality rate. Commercial size limit is 381 mm. 30 Table 4. Average percent change from existing conditions in FC (max) and I (max) with corresponding yields when input parameters of yield model are varied 3-10%. Parameter FC (max) Yield | (max) Yield Instantaneous Total Mortality Rate 51.2 15.2 11.7 22.7 Commercial Reporting Rate 27.9 10.1 - 10.4 18.6 Sport Reporting Rate O O O O von Bertalanffy Growth Parameters 69.8 24.0 15.5 32.5 Tag Loss Rate 1.2 0.4 1.1 0.7 Length at - ---- Recruitment 1.2 14.3 0.9 14.3 31 DISCUSSION Commercial yield per 1000 recruits of channel catfish could be increased by increasing commercial fishing effort. However, management objectives must be carefully considered if such a change were allowed, as increased effort would decrease sport yield. Greatest increases in commercial yield would be achieved with increases in both effort and commercial size limit. Previous studies (Eshenroder and Haas 1974, Lorantas 1982) found that commercial yield could be maintained or increased if effort was reduced; however, my results indicate that only sport yield would be increased by this sort of change. Reasons for the differences between these studies include higher natural mortality rate, slower growth rate, and lower fishing mortality rate found in the present study. The instantaneous total mortality rate of 0.45 was lower than the total mortality rate calculated by Eshenroder and Haas (1974) and Lorantas (1982). The total mortality rate is low compared to total mortality rates for channel catfish from other studies (Table 5). One possible reason for the decline in the instantaneous total mortality rate since 1981 is that during the years affecting the studies in 1971 and 1981 recruitment was probably increasing. Thus, the assumption of constant recruitment for the catch curve analysis was violated. This increased recruitment would cause the total mortality rates calculated to be too high. By the time I did this study recruitment may have stabilized so that the true instantaneous total mortality rate could be calculated from a catch curve. 32 Table 5. Instantaneous total mortality rates (with 95% confidence intervals when calculated) for channel catfish from Saginaw Bay and other selected locations. Location Year(s) of Instantaneous total study mortality rate Saginaw Bay - 1985 0.45 (+0.21) Saginaw Bay * 1981 - 0.67 (+0.27) Saginaw Bay” 1971 0.67 Des Moines River, lowa 9 1966–69 0.64 Upper Mississippi River, Iowa" 1977-79 0.94 Lake Sharpe, South Dakota º 1945-56 0.37 Rivers of Sacramento Valley, California' 1955–59 0.82 * Lorantas (1982) * Eshenroder and Haas (1974) ° Mayhew (1972) cited by Lorantas (1982) ° Pitlow and Bonneau (1979) cited by Lorantas (1982) e Elrod (1974) "McCammon and LaFaunce (1961) 33 The natural mortality rate of 0.19 which I calculated for the bay was higher than the assumed rate of 0.1 used by Eshenroder and Haas (1974) and Lorantas (1982). The natural mortality rate calculated in this study lies within the range of published values. McCammon and LaFaunce (1961) calculated a natural mortality rate of 0.38 for channel catfish in the rivers of the Sacramento Valley and the natural mortality rate calculated for an unexploited stock of channel catfish in Lake St. Clair was 0.105 (Great Lakes Fishery Commission, Lake Erie Committee Report for 1971 cited by Eshenroder and Haas 1974). If Lorantas (1982) had used this higher natural mortality rate his conclusions would have been similar to mine. At the 381 mm commercial size limit, commercial yield would have been maximized at a FC greater than 1.00 instead of the FC of about 0.25 calculated by Lorantas. Backcalculated total lengths at age for the three Saginaw Bay studies revealed a decrease in growth rate of channel catfish since 1971 (Table 6). The growth rate in this study was also slower than for catfish from many other areas (Table 6). Two factors could explain the decline in growth rate of the catfish. First, channel catfish may compete with other bottom feeders and predatory species utilizing the same habitat (Scott and Crossman 1979). Interspecific competition has likely increased as walleye Stizostedion vitreum and salmonid populations in Lake Huron have increased through stocking or natural reproduction (Great Lakes Fishery Commission 1985). In addition, perch production has been good for the past several years (Great Lakes Fishery Commission 1985). Second, intraspecific competition has probably intensified as the population of catfish increased. 34 Table 6. Mean backcalculated total length (mm) at age for channel catfish from Saginaw Bay and other selected areas. Location Year(s) of Sample Aqe study size 1 2 3 4 5 6 7 8 9 1 0 1 1 12 13 14 15 Saginaw Bay, Lake Huron 1985 237 42 111 168 233 283 328 376 420 473 509 533 566 592 630 685 +2 + 4 + 4 +5 +5 +6 3-7 4-8 +12 + 1 3 +21 +27 +33 + 42 +694 Saginaw Bay, Lake Huron * 1981 91 6 54 132 198 256 31 0 358 420 469 507 546 594 604 612 +1 + 2 +2 + 2 + 3 + 3 +5 +6 +6 +8 +14 +20 +76 Saginaw Bay, Lake Huron b 1971 253 69 157 254 335 404 490 561 569 589 607 630 Lake Erie, Michigan waters * 1971 495 170 221 264 305 340 373 386 417 452 505 549 Lake St. Clair" 1969-71 507 76 208 226 272 305 383 434 485 531 564 604 St. Lawrence River * 1975 28 1 19 164 204 239 272 302 330 353 377 400 432 455 Lake Sharpe, South Dakota" 1945-56 535 46 124 196 256 31 2 381 442 490 546 617 645 640 676 Santee-Cooper reservior system, South Carolina º 1959 210 86 185 284 368 442 531 602 665 726 772 807 853 917 904 Lake of the Ozarks, Missouri 1951 434 53 1 09 155 196 234 264 292 330 Farm ponds, Central Texas 9 1972 82 178 333 429 516 * Lorantas (1982) b Eshenroder and Haas (1974) ° Maguin and Fradette (1975) d Elrod (1974) 9 Stevens (1959) ! Marzolf (1955) 9 Prentice and Whiteside (1974) 3. The increase in catfish population is indicated by several factors, including older fish appearing in the commercial catch (Table 6) and increasing catch per unit effort (Fig. 1). Variations in the instantaneous total mortality rate, commercial reporting rate, and von Bertalanffy growth parameters were most important in affecting FC (max) and , (max) values. Instantaneous total mortality rate and growth rate seem reasonable when compared with results from earlier studies. On the other hand, no previous data exists for comparison of reporting of tags from channel catfish caught in Michigan by commercial fishermen. Thus, improving estimates of commercial reporting rate would increase confidence in predictions from the yield model. Sport reporting rate was not important in determining FC (max) and , (max) for commercial yield because of the way in which the natural mortality rate was calculated. If an independent estimate of natural mortality rate was available, sport reporting rate may have been important in determining FC (max) and , (max) for commercial yield. The effect on these values would probably be less than the effect of the three factors above. All assumptions necessary for the catch curve analysis (Ricker 1975) and the yield per recruit model (Clark and Huang 1983) were probably violated to some extent, but I feel that these deviations did not significantly affect results. The assumption of constant mortality rates would probably cause the most problems. Constancy of the total mortality rate with time is unknown. If the total mortality rates calculated by 36 Lorantas (1982) and l are both correct, then the mortality rate has not been constant. Examination of the commercial catch indicates that from 1975 to 1980 the catch increased steadily without any trend in effort. Specific reasons for the increased catch are unknown but may include decreased natural mortality. The environment in Saginaw Bay is continually changing as water quality, water levels, and species composition change. This changing environment might affect natural mortality. If so, the natural mortality rate is probably never constant. Thus, some of the age classes used to determine the total mortality rate may have experienced higher mortality rates than others. Adjusting for a decline in the natural mortality rate would mean a further reduction in the total mortality rate. A lower natural mortality rate would favor increasing the commercial size limit to maximize yield. Yield per 1000 recruits at all levels of fishing effort and Commercial size limit would also increase. Another assumption for the yield per recruit model which appearedto be violated was that of constant growth rate. Examination of backcalculated length at age data for the three studies on catfish in Saginaw Bay (Table 6) revealed that the growth rate has slowed. Although the sensitivity analysis revealed that variations in von Bertalanffy growth parameters were important in determining FC (max) and , (max) values, I redid the yield analysis using von Bertalanffy growth parameters calculated by Lorantas (1982) and found only slight changes in model predictions. Thus, the change in growth rate that has occurred since 1981 should not affect the reliability of model predictions. 37 Accuracy of tag reporting rate by all commercial fishermen is contingent upon two assumptions. The first is that reporting by fishermen associated with Bay Port Fish Company did not differ significantly from reporting by other fishermen in the bay. The second is that the sample of fish examined from the commercial catch was of adequate size. The first assumption is likely to be violated somewhat because fishermen associated with Bay Port Fish Company were more informed of the project and seemed to be more watchful for tagged catfish. Catfish caught by Bay Port Fish Company were handled more frequently and held in indoor tanks so that tagged fish were more likely to be detected than at other locations. Fishermen in other areas used ponds or floating cages to hold catfish. Thus, the tag reporting rate may actually be too high. However, a majority of all catfish caught in Saginaw Bay were handled by Bay Port Fish Company, so variability in reporting among commercial fishermen should not be a large problem. The second bias (sample size) could be a significant problem. Even though 2,872 fish were examined, this amounted to only 1% of the combined commercial catch in the study area for 1985 and 1986. In addition, some of the daily sample sizes were small (range 20-375), and overly small sample sizes would make the reporting rate estimate conservative. I feel that if the reporting rate was in error, it most likely was too low. If so, harvesting at larger sizes would be favored to maximize yield. No estimate was made of the reporting rate of tags by sport fishermen. The lottery was conducted because I hoped this would entice a majority of fishermen catching tagged fish to return the tags. Other studies have found 38 that people will keep tags as momentos even though a reward is offered (Rawstron 1971). I know of at least two people who wanted to keep the tags (both did report the catch though) and I suspect other people, especially if they did not know of the reward, kept tags. Other studies examining reporting rate of tags taken from various fish species by sport fishermen have found reporting rate to range from 15% to 66% (Table 7). Thus, the assumed sport reporting rate of 45% seems reasonable. One reason for the differences seen in the length and age frequency distributions among grids may be the habitats where the nets were set. Nets on the western side of the bay were set in deeper water several miles offshore. In grid 1509, nets were set a short distance offshore of the islands bordering Wild Fowl Bay (Fig. 3). Wild Fowl Bay is used extensively for spawning by catfish; thus, large fish have only a short distance to travel to get to the nets. Nets set in grid 1608 were much smaller than nets set in other grids and set so the lead abutted the shore. They were also set within an area of good spawning habitat. Thus, large fish were probably very abundant near these nets, and this may explain the unusual frequency distribution. The age and length frequency distributions may not have been entirely representative of the true frequency distributions in the various grids. Many of the nets sampled contained an extremely large number of small fish, so that measuring every fish would have taken an inordinate amount of time. Thus, most fish less than about 250 mm were not measured. The modes of the age and length frequency distributions and the frequencies greater than 39 the modes should not be affected by this selection. The distributions for grid 1606 may have been affected by this, because only a subsample of all the fish in Some nets was measured. 40 Table 7. Tag reporting rates of sport fishermen for various species. Location Species Year of Reporting Study Rate Folsom Lake, White Catfish 1962–68 0.61 California * (Ictalurus catus) Largemouth bass 0.54 (Micropterus salmiodes) Bluegill 0.16 (Lepomis macrochirus) Merle Collins Reservoir, California” Largemouth bass 1966-71 0.66 Sacramento-San Joaquin river system, California * Striped bass 1957-67 0.39 (Morone saxatilis) West Point Reservoir, Alabama-Georgia * Largemouth bass 1976–79 0.34° 0.60' a Rawstron (1971) b Rawstron (1972) ° Chadwick (1968) d Folmar et al. (1979) ° No reward offered. "Five dollar reward offered. 4 | RECOMMENDATIONS In recent years conflicts between commercial and sport fishermen in Saginaw Bay have been increasing. Sport fishermen have become angered over the use of trap nets in the bay, because they entangle fishing lines and obstruct fishing in certain areas. Sport fishermen have also blamed the perceived lack of large yellow perch Perca flavescens and poor walleye catches on commercial fishing. Although competition for large yellow perch is minimal and allegations against walleye catches is false, a regulation change increasing commercial yield at the expense of sport yield would likely be opposed by sport fishermen. Increasing the commercial size limit for channel catfish is the only way to increase both the commercial and sport yield of catfish. Increasing the size limit to 406.4 mm (16 inches) would increase total, commercial, and sport yield per 1000 recruits, 2.6%, 0.6%, and 9.1%, respectively. Yield was actually predicted to reach a maximum at 416 mm (16.4 inches). However, for management purposes the Smaller size limit is more convenient because it is a whole number when converted to inches and corresponds to a 1 inch (25.4 mm) increase in the current size limit (381 mm). A greater increase in total yield could be obtained by increasing the commercial size limit to 432 mm and doubling the commercial fishing effort. Then the predicted increase in total, and commercial yield per 1000 recruits, would be 5.3% and 12.0%, respectively. However, sport yield would decrease 16.8%. Other types of gear besides trap nets, such as seines, could be used to increase commercial fishing effort without increasing conflicts with sport fishermen. 42 Increasing the commercial size limit would increase the biomass of channel catfish in Saginaw Bay, which would cause some prey items to be more heavily utilized. Species interactions within the bay are poorly understood so the exact effect of an. increased catfish biomass cannot be determined. However, interspecific and intraspecific competition may increase. If fishing effort was also increased, any effect resulting from the increased size limit would be reduced because increasing fishing effort would reduce the biomass of catfish. Management objectives must be considered before any changes are made because of the decrease in sport yield that would occur with an increase in commercial effort and the fact that sport yield is maximized when commercial fishing is prohibited. 43 LITERATURE CITED Bagenal, T. B., and F. W. Tesch. 1978. Age and Growth. Pages 101-136 in T. B. Bagenal, editor. Methods for assessment of fish production in fresh waters, 3rd edition. Blackwell Scientific Publications Ltd., London. Chadwick, H. K. 1968. Mortality rates in the California striped bass population. California Fish and Game 54:228-246. Clark, R. D., Jr., and B. Huang. 1983. The dynamics of competition between sport and commercial fishing: effects on rehabilitation of lake trout in Lake Michigan. Michigan Department of Natural Resources, Fisheries Research Report No. 1909, Ann Arbor. Elrod, J. H. 1974. Abundance, growth, survival, and maturation of channel catfish in Lake Sharpe, South Dakota. Transactions of the American Fisheries Society 103:53-58. Eshenroder, R. L., and R. C. Haas. 1974. Status of selected stocks in Lake Huron and recommendations for commercial harvest. Technical Report 73-32. Pages 151-206 in J. D. Bails and M. H. Patriarche. Status of selected fish stocks in Michigan's Great Lake's waters and recommendations for commercial harvest. Everhart, W. H., and W. D. Youngs. 1981. Principles of fishery science. Cornell University Press, Ithaca, New York. 44 Folmar, H. G., W. D. Davies, and W. L. Shelton. 1979. Factors affecting estimates of fishing mortality of largemouth bass in a southeastern reservoir. Proceedings of the Southeastern Association of Fish and Wildlife Agencies 33:402-407. Great Lakes Fishery Commission. 1985. Annual report for the year 1983. Great Lakes Fishery Commission, Ann Arbor, Michigan. Hale, M. M., J. E. Crumpton, and D. J. Renfro. 1983. Tag retention and survival of Floy-tagged and fin-clipped white catfish and channel catfish in hatchery ponds. Proceedings of the Southeastern Association of Fish and Wildlife Agencies 37:385-390. Hile, R. 1962. Collection and analysis of commercial fishery statistics in the Great Lakes. Great Lakes Fishery Commission, Technical Report No. 5. Ann Arbor, Michigan. Lorantas, R. 1982. Assessment of the channel catfish fishery of Saginaw Bay, Lake Huron. Michigan Department of Natural Resources, Fisheries Research Report No. 1908, Ann Arbor. Maguin, E., and G. Fradette. 1975. Growth of the channel catfish (Ictalurus punctatus) in the St. Lawrence River near Quebec. Journal of the Fisheries Research Board of Canada 32:1867-1870. Marzolf, R. C. 1955. Use of pectoral spines and vertebrae for determining age and rate of growth of the channel catfish. The Journal of Wildlife Management 19:243-249. 45 Mayhew, J. 1972. Some biological characteristics of a channel catfish population in the lower Des Moines River with an evaluation of potential commercial harvest. Iowa Conservation Commission Fisheries Section, Technical Series 72-2, Des Moines. McCammon, G. W., and D. A. LaFaunce. 1961. Mortality rates and movement in the channel catfish population of the Sacramento Valley. California Fish and Game 47:5-26. Pitlow, J., Jr., and D. Bonneau. 1979. Assessment of change in commercial size limit of channel catfish. Iowa Conservation Commission Fisheries Section, Final Report Commercial Fisheries Investigations, - Des Moines. Prentice, J. A., and B. G. Whiteside. 1974. Validation of aging techniques for largemouth bass and channel catfish in central Texas farm ponds. Proceedings of the Annual Conference of the Southeastern Association of Game and Fish Commissioners 28:414–428. Rafail, S. Z. 1973. A simple and precise method for fitting a von Bertalanffy growth curve. Marine Biology 19:354-358. Rawstron, R. R. 1971. Nonreporting of tagged white catfish, largemouth bass, and bluegills by anglers at Folsom Lake, California. California Fish and Game 57:246-252. Rawstron, R. R. 1972. Nonreporting of tagged largemouth bass 1966-1969. California Fish and Game 58:145-147 Ricker, W. E. 1975. Computation and interpretation of biological statistics of fish populations. Bulletin of the Fisheries Research Board of Canada No. 191. 46 Rybicki, R. W., and M. Keller. 1978. The lake trout resource in Michigan waters of Lake Michigan, 1970-1976. Michigan Department of Natural Resources, Fisheries Research Report No. 1863, Ann Arbor. Scott, W. B., and E. J. Crossman. 1979. Freshwater fishes of Canada. Bulletin of the Fisheries Research Board of Canada No. 184. Sneed, K. E. 1951. A method for calculating growth of channel catfish, Ictalurus lacustris punctatus. Transactions of the American Fisheries Society 80:174-183. Stevens, R. E. 1959. The white and channel catfishes of the Santee-Cooper Reservoir and tailrace sanctuary. Proceedings of the Annual Conference of the Southeastern Association of Game and Fish Commissioners 13:203-219. 47 UNIVERSITY OF MICHIGAN ||||||||||||||| 3 9015 01263. 31.6% THE UNIVERSITY OF MICHIGAN DATE DUE