3r. 2.... 9,3..2 .. .V ‘t it in. E .. .1? 4.5.. An, 3..» .l a in :9»... 2... .11 V .1: 3.3!. . I...» . .1. 2...?! .. 2......1252‘. q {“955th .u. t t 1.! \‘ . ; S‘Létluv-ht21\)t \ 5.1.. 2|... .CQI. It!“ ‘52,. 11.; has. ‘ .I‘Ioi‘iu :IIlli “If it i ..4n\ .1. ”3.1.;iz a L- 5"}, :1 3, . .\ .«x .7 Ir in“?! k In 2, v0. .1 fl“; . . .. Ewan?» } . 1.12}! .1 as. I 31.2.; s... t o. 7y \ .3... u : g .83..» . a}; .5! 6:35.? .3 3.2 1. 3. v9.1}:2 .1; t Z ‘30:}? Pl .3. 3.133%? V 35321.1: . .. 5.. x i. .o o i\r|. g . i‘JU—‘z alt. i i352. ‘ 3503.?» I)? .14....“ 4., .Ifizu. s . r... Virilfi [it n‘t‘iviv 2“!" .1391 it.i..l..1«hur uh. vaginal... ‘ « D‘.O£bi"-vz.l$ Ivliii - . l.$é1?3.f3!1 ‘ . Elton} . £4 XIV; \r: )1. )1 .lbn. 1': I Y . {3.1.x .‘Ilfilli 5.1!: 21.51:»... ....\: I! I] vikii I! It...) u n . Huh-«Vlbg .n “3.“. 1.:I‘stl (-1145..- ‘iV-v OS 1...»? d5. 3. $33“ , , .H ASL} away? F mama 2-1 11111111111111111111111 1111111111111111‘1“‘1"T1 3 1293 01405 This is to certify that the thesis entitled ESTIMATION OF LAKE TROUT (Salvelinus namaycush) ABUNDANCE AND MORTALITY DUE TO SEA LAMPREYS (Petromyzon marinus) AND FISHING IN THE MAIN BASIN OF LAKE HURON, 1984-1993 presented‘by Shawn Paul Sitar has been accepted towards fulfillment of the requirements for Master of Science degreein Fish. & Wildl. :I DWQ (33W Major professor Date April 17, 1996 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State Unlverslty ‘ - 00--.. PLACE IN RETURN BOX to man this checkout from yam record. TO AVOID FINES return on or baton data duo. DATE DUE DATE DUE DATE DUE ESTIMATION OF LAKE TROUT (Salvelinus namaycush) ABUNDANCE AND MORTALITY DUE TO SEA LAMPREYS (Pelromyzon marinas) AND FISHING IN THE MAIN BASIN OF LAKE HURON, 1984-1993 By Shawn Paul Sitar 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 1996 ABSTRACT ESTIMATION OF LAKE TROUT (Salvelinus namaycush) ABUNDANCE AND MORTALITY DUE TO SEA LAMPEYS (Petromyzon marinas) AND FISHING IN THE MAIN BASIN OF LAKE HURON, 1984-1993 By Shawn Paul Sitar Sea lamprey (Petromyzon marinas) parasitism and overfishing have been cited as the causes of the collapse of lake trout (Salvelinus namaycush) populations in Lake Huron during the 19508. The goal of the ongoing lake trout rehabilitation program is aimed at reducing sea lamprey abundance, controlling fishing mortality, and restocking lake trout to establish self-sustaining populations. In order to rehabilitate lake trout, the magnitude of sea lamprey parasitism and fishing mortality must be determined in order to gauge progress towards the goal. With reliable estimates of lake trout deaths due to sea lampreys and fishery harvest, managers can adjust sea lamprey control programs and fishing regulations to reach rehabilitation objectives. I analyzed data on sea lamprey wounding of lake trout, from 1984-1994, to assess patterns in sea lamprey parasitism according to length of lake trout, geographic distribution, and year. Lake trout population models, calibrated by statistical catch-at-age analysis, were constructed to estimate abundance, fishery harvest, and numbers killed by sea lamprey during 1984- 1993 for the main basin of Lake Huron. Sea lamprey wounding rates on lake trout increased with length of lake trout and were higher in central Lake Huron than in the south for lake trout >533 mm. Although sea lamprey wounding of lake trout varied by year, no overall temporal trends were observed during 1984-1994 in the central and southern main basin of Lake Huron. Comparisons with northern Lake Huron were not possible because of insufficient data. Abundance of mature lake trout, an index of potential natural recruitment, was estimated to be highest in southern Lake Huron and lowest in the north. For lake trout ages most selected by sea lampreys and fishing (ages 3-10), total annual mortality rates were highest in northern Lake Huron and have exceeded the Great Lakes Fishery Commission (GLFC) target maximum total annual mortality rate of 45% in all years from 1984-1993. Total annual mortality rates in central and southern main basin of Lake Huron were below the GLF C target maximum during the same time period. Sea lamprey-induced mortality accounted for most lake trout deaths in central and southern Lake Huron, whereas commercial fishing and sea lamprey parasitism both were responsible for the high number of lake trout deaths in the north. Recreational fishing was not a significant source of lake trout mortality in the main basin of Lake Huron. The lack of success in re-establishing self-sustaining populations of lake trout in the main basin of Lake Huron was due in part to the mismatching of reproductive biomass and spawning habitat. In central and southern Lake Huron, lack of sufficient spawners and insumcient spawning habitat are possible reasons that rehabilitation has not progressed in these areas. In northern Lake Huron, where the amount of spawning habitat is greatest, excessive sea lamprey-induced and commercial fishing mortality at premature ages has limited the abundance of spawners. In order to successfully rehabilitate lake trout, total mortality rates must be reduced in northern Lake Huron. This work is dediCated to my loving wife, Kristie, my parents, Dania and Steve, and my brother, Robert. ACKNOWLEDGMENTS This study was sponsored by the Great Lakes Fishery Commission (GLFC), Michigan Department of Natural Resources (MDNR), Michigan State University (MSU), and the US. Fish and Wildlife Service (U SFWS). Fellowship support was provided by the Department of Fisheries and Wildlife, MSU. I would like to thank the biologists of the Lake Huron Technical Committee including: , Mark Ebener (COTFMA), Jim Johnson (NHDNR), Jerry McClain (U SFWS), and Lloyd Mohr (Ontario Ministry of Natural Resources) for their technical expertise and advice. Special thanks is also extended to Gavin Christie (GLFC), Rick Clark (MDNR), the crew of the R/V Chinook, MDNRuBill Cross, Clarence Cross, and J efi‘ Diemond; and to the crew at the Alpena Fishery Resources Office, USFWS--Anjie Hintz and Sheral Eakin. I thank Dr. Patrick Muzzall for his valuable insights and for serving on my graduate committee and providing suggestions and critical review of this manuscript. I also thank Dr. William Taylor for providing me the opportunity to develop and learn about fisheries science, and for his insights and editing of this manuscript. I especially thank Dr. James Bence for his support, guidance, wisdom, patience, and fiiendship while vi serving as my mentor. I am honored to have worked with such a fine role model. I am also thankful for the technical and social support from fellow graduate students including: Salvador Becerra-Mufioz, Russ Brown, Paola Ferreri, Doug Novinger, Ed Roseman, and Ted Sledge. I am also appreciative of the help and support from Julie Detwiler, Carol Graysmith, Dr. Dan Hayes, Mary Hill, Dr. Darrell King, Jane Thompson, and Julie Traver. I am very gratefiil for the guidance and support from my early mentors, Dr. Paul Haefner Jr., Dr. M. Joseph Klingensmith, and Dr. Franz Seischab at the Department of Biology, Rochester Institute of Technology. Finally, I am thankful of the support, understanding, and love from my wonderful wife, Kristie, who always bad faith in me. Without her, none of this would have been possible. vii TABLE OF CONTENTS LIST OF TABLES ........................................................................................................ x LIST OF FIGURES ..................................................................................................... xv INTRODUCTION ......................................................................................................... 1 METHODS ................................................................................................................... 8 Patterns in sea lamprey wounding ....................................................................... 8 Estimation of sea lamprey-induced mortality ..................................................... 13 Lake trout population model ............................................................................. 17 Lake trout abundance ............................................................................ 20 Natural mortality ................................................................................... 21 Fishing mortality ................................................................................... 22 Statistical.catch.—at-age analysis of the southem Lake Huron lake trout populatiai model ................................................................................................... 25 Sensitivity of the southern model to calibration data .............................. 36 Calibration of the northern and central lake trout population models ................. 36 Model projections ............................................................................................. 39 RESULTS ................................................................................................................... 41 Patterns in sea lamprey wounding ..................................................................... 41 Patterns in wounding according to length of lake trout .......................... 44 Geographic patterns in wounding rates .................................................. 47 Temporal trends in wounding rates ....................................................... 52 Estimation of sea lamprey-induced mortality ..................................................... 52 Statistical catch-at-age analysis of the southern Lake Huron lake trout population model ................................................................................................... 64 Sensitivity of the southern model to calibration data .............................. 78 Uncertainty in estimated abundance ....................................................... 9O viii Calibration of the northern and central lake trout population models ................. 91 Model output ................................................................................................... 91 Southern Lake Huron (MH-3/4/5), 1984-1993 ...................................... 91 Central Lake Huron (MB-2), 1984-1993 ............................................... 92 Northern Lake Huron (MH-l), 1984-1993 ............................................ 94 Total mortality rates in Lake Huron ...................................................... 99 Model projections ............................................................................................. 99 Southern Lake Huron (MH-3/4/5) ........................................................ 99 Central Lake Huron (MH-2) ............................................................... 101 Northern Lake Huron (MI-I-l) ............................................................ 110 Mortality trade-off: sea lamprey-induced vs. fishing mortality ........................ 117 DISCUSSION ........................................................................................................... 1 19 Role of sea lampreys in lake trout rehabilitation .............................................. 120 Survival and abundance of lake trout during 1984-1993 .................................. 121 Management trade-off: fishing vs. sea lamprey-induced mortality ................... 123 Northern Lake Huron ......................................................................... 123 Central Lake Huron ............................................................................ 124 Southern Lake Huron ......................................................................... 125 Status and potentials of lake trout rehabilitation .............................................. 126 APPENDIX- ADDITIONAL TABLES ..................................................................... 130 LIST OF REFERENCES ........................................................................................... 167 ix LIST OF TABLES Table 1. Numbers of lake trout examined for sea lamprey wounds in Michigan waters of Lake Huron in spring gill net surveys and subsampling of tribal gill net and trap net catches. Observations are stratified by lake trout length class. Data provided by the Chippewa-Ottawa Treaty Management Authority, and Michigan Department of Natural Resources. Region: MH-1= north, MH-2= central, and MH-3/4/5= south. ............................................................................................ 12 Table 2. Previously reported parameter values for estimating mortality rates of lake trout in the main basin of Lake Huron. COTFMA= Chippewa-Ottawa Treaty Fishery Management Authority, MDNR= Michigan Department of Natural Resources. 19 Table 3. Average mass-at-age of lake trout in Michigan waters of Lake Huron. Data provided by Michigan Department of Natural Resources and GP. Ferreri, Pennsylvania State University. .......................................................................... 26 Table 4. Parameters of the southern Lake Huron (MH-3/4/5) lake trout population model estimated by statistical catch-at-age analysis. .......................................... 28 Table 5. Levels for each factor in analysis of variance models used to evaluate patterns in sea lamprey wounding of lake trout in Michigan waters of Lake Huron. ........ 42 Table 6. Significance levels (attained P-value) for main effects and interactions in analysis of variance models of sea lamprey wounding rates on lake trout in Michigan waters of Lake Huron, 1984-1994. Further information on data used with these models is given in Table 5. ............................................................... 43 Table 7. Levels for each factor in analysis of variance models used to estimate mean wounds per fish when an insufficient number of observations (less than 40 lake trout) were available in Michigan waters of Lake Huron. MH-1= north, MH-2= central, and MH-3/4/5= south. .......................................................................... 58 Table 8. Sea lamprey wounding rates by length class for lake trout in central Lake Huron (NIH-2). Wounding rates expressed as mean wounds per fish. Data from Michigan Department of Natural Resources spring surveys. .............................. 6O Table 9. Sea lamprey wounding rates by length class for lake trout in southern Lake Huron (MH-3/4/5). Wounding rates expressed as mean wounds per fish. Data from Michigan Department of Natural Resources spring surveys ....................... 61 Table 10. Estimated instantaneous rates of sea lamprey-induced mortality (year") for lake trout in central Lake Huron (MB-2) during 1984-1993 .............................. 62 Table 11. Estimated instantaneous rates of sea lamprey-induced mortality (year") for lake trout in southern Lake Huron (MH-3/4/5) during 1984-1993 ..................... 63 Table 12. Estimated parameter values from catch-at-age analyses of the southern Lake Huron lake trout population model, 1984-1993. Recreational fishery parameters: qg= catchability (angler hours"), 83, .= selectivity at age a, and fa, y= fishing intensity (year"). p.’= proportionality coefficient for sea lamprey-induced mortality. Research survey parameters: q*= catchability (meters of gill net"), S*.= selectivity at age a. Population parameters: N..1924= abundance at age a in 1984, c= proportionality coefficient for natural mortality, M1= age-1 instantaneous natural mortality (year"), and t= rate of decrease in natural mortality rate (year" age"). “= parameter not estimated by catch-at-age analysis.65 Table 13. Maximum log-likelihood components from statistical catch-at-age analyses of the southern Lake Huron lake trout population model, 1984-1993. ................... 67 Table 14. Estimated parameter values from catch-at-age analysis model CAA6. Recreational fishery parameters: qR= catchability (angler hours"), SR, .= selectivity at age a, and f1; ,= fishing intensity (year"). u’= proportionality coefficient for sea lamprey-induced mortality. Research survey parameters: q*= catchability (meters of gill net"), S*.= selectivity at age a. Population parameters: N.,1934= abundance at age a in 1984, M1= age-l instantaneous natural mortality (year"), and t= rate of decrease in natural mortality rate (year" age"). = parameter not estimated by catch-at-age analysis. ......................................... 86 Table 15. Joint age-length distribution for lake trout in northern Lake Huron (MH-l). Data fi'om Michigan Department of Natural Resources annual spring gill net surveys from 1984-1994. ................................................................................ 130 Table 16. Joint age-length distribution for lake trout in central Lake Huron (MB-2). Data from Michigan Department of Natural Resources annual spring gill net surveys from 1984-1994. ................................................................................ 131 Table 17. Joint age-length distribution for lake trout in southern Lake Huron (MH- 3/4/5). Data from Michigan Department of Natural Resources annual spring gill net surveys from 1984-1994. .......................................................................... 132 Table 18. Assumed age-1 abundance (x 1000) of lake trout in the main basin of Lake Huron. Data, adjusted for migration, were based on number of yearlings and fall fingerlings (age-0) stocked. Fall fingerlings were converted to yearling- equivalents based on the assumption that 40% of fingerlings survived to the yearling stage. Sixty percent of lake trout stocked in MH-2 were assumed to migrate to MI-I-l (J. Johnson, Alpena Fisheries Research Station, Michigan Department of Natural Resources, pers. comm). ............................................ 133 Table 19. Sport harvest and effort of lake trout in Michigan waters of Lake Huron. Harvest reported in numbers of fish and effort expressed as angler hours. Data from Michigan Department of Natural Resources. .......................................... 134 Table 20. Age composition of sport fishery harvest of lake trout in Michigan waters of Lake Huron. Data, expressed as proportions at age, were from Michigan Department of Natural Resources sport harvest monitoring program. n= sample size ................................................................................................................. 135 Table 21. Canadian harvest of lake trout in southern Lake Huron (OH-3, OH—4 and OH- 5). Annual yield data from Ontario Ministry of Natural Resources. Harvest in numbers estimated by dividing yield by average mass per fish of Michigan sport harvest for each year ....................................................................................... 136 Table 22. Catch and effort of lake trout fi'om Michigan Department of Natural Resources annual spring gill net surveys in northern Lake Huron (MH-l). Effort expressed as meters of gill net per day. No data available for 1990 ................. 137 Table 23. Catch and effort of lake trout from Michigan Department of Natural Resources annual spring gill net surveys in central Lake Huron (MB-2). Effort expressed as meters of gill net per day. ........................................................... 138 ' Table 24. Catch and effort of lake trout from Michigan Department of Natural Resources annual spring gill net surveys in southern Lake Huron (MH-3/4/5). Efl‘ort expressed as meters of gill net per day. ................................................. 139 Table 25. Canadian harvest of lake trout in OH-l and OH-2 in northern and central Lake Huron. Forty percent of the harvest from zone 4-1 in district OH-l were assumed to be from the northern area. Sixty percent of lake trout harvested in zone 4-1 of OH-l, and all harvest in OH-2 were assumed to be from the central area. Annual yield data from Ontario Ministry of Natural Resources. Harvest in numbers for Canadian removals from the MH-l stock estimated by dividing reported yield by average mass per fish of tribal gill net harvest in MH-l for each year. Harvest in numbers for Canadian removals from the MH-2 stock estimated xii by dividing reported yield by average mass per fish of Michigan sport harvest for each year. ....................................................................................................... 140 Table 26. Reported tribal commercial harvest and efl‘ort of lake trout in northern Lake Huron (MH-l). Data provided by Chippewa-Ottawa Treaty Fishery Management Authority. Efi‘ort expressed as meters of large-mesh gill net targeted at lake whitefish and lake trout. ................................................................................. 141 Table 27. Parameters estimated to calibrate the northern and central lake trout population models. f c, , =commercial fishing intensity (year") in year y, u’ = proportionality coefficient for sea lamprey-induced mortality, and qR= catchability coefficient for the recreational fishery (angler hours") , p. = survival proportion for age a for cohorts before 1984 to estimate abundance in 1984 for ages>1142 Table 28. Model estimates of lake trout abundance in southern main basin of Lake Huron (MH-3/4/5) .......................................................................................... 143 Table 29. Estimates of instantaneous rates of natural mortality (M) for lake trout in main basin of Lake Huron based on statistical catch-at-age analysis of the southern Lake Huron population model. Rates were assumed constant from 1984-1993.144 Table 30. Model estimates of instantaneous rates of recreational fishing mortality (year") for lake trout in southern Lake Huron (MI-I-3/4/5) .............................. 145 Table 31. Model estimates of number of lake trout deaths (x1000) due to natural mortality in region southern Lake Huron (MH-3/4/5). .................................... 146 Table 32. Model estimates of number of lake trout deaths (x1000) due to recreational fishing mortality in southern Lake Huron (MH-3/4/5). .................................... 147 Table 33. Model estimates of number of lake trout deaths (x1000) due to sea lamprey- induced mortality in southern Lake Huron (MH-3/4/5). .................................. 148 Table 34. Model estimates of lake trout abundance in central main basin of Lake Huron (NIH-2). ......................................................................................................... 149 Table 35. Model estimates of instantaneous rates of recreational fishing mortality (year") for lake trout in central Lake Huron (MB-2) ....................................... 150 Table 36. Model estimates of instantaneous rates of commercial fishing mortality (year") for lake trout in central Lake Huron (MH-2) ................................................... 151 Table 37. Model estimates of number of lake trout deaths (x1000) due to natural mortality in central Lake Huron (MI-I-2). ........................................................ 152 xiii Table 38. Model estimates of number of lake trout deaths (x1000) due to recreational fishing mortality in central Lake Huron (MH-2). ............................................. 153 Table 39. Model estimates of number of lake trout deaths (x1000) due to commercial fishing mortality in central Lake Huron (MB-2). ............................................. 154 Table 40. Model estimates of number of lake trout deaths (x1000) due to sea lamprey- induced mortality in central Lake Huron (MB-2). ........................................... 155 Table 41. Model estimates of lake trout abundance in northern main basin of Lake Huron (MH-l) ................................................................................................ 156 Table 42. Model estimates of instantaneous rates of recreational fishing mortality (year") for lake trout in northern Lake Huron (MH-l). ................................... 157 Table 43. Model estimates of instantaneous rates of commercial fishing mortality (year") for lake trout in northern Lake Huron (MB-1). ............................................... 158 Table 44. Model estimates of instantaneous rates of sea lamprey-induced mortality (year") for lake trout in northern Lake Huron (MH-l). ................................... 159 Table 45. Model estimates of number of lake trout deaths (x1000) due to natural mortality in northern Lake Huron (MH-l). ..................................................... 160 Table 46. Model estimates of number of lake trout deaths (x1000) due to recreational fishing mortality in northern Lake Huron (MH- 1). .......................................... 161 Table 47. Model estimates of number of lake trout deaths (x1000) due to commercial fishing mortality in northern Lake Huron (MH-l). .......................................... 162 Table 48. Model estimates of number of lake trout deaths (x1000) due to sea lamprey- induced mortality in northern Lake Huron (MH—l). ........................................ 163 Table 49. Model estimates of instantaneous rates of total mortality (year") for lake trout in southern main basin Lake Huron (MH-3/4/5). ............................................. 164 Table 50. Model estimates of instantaneous rates of total mortality (year") for lake trout in central main basin Lake Huron (MB-2). ...................................................... 165 Table 51. Model estimates of instantaneous rates of total mortality (year") for lake trout in northern main basin Lake Huron (MI-I-l). ................................................... 166 xiv LIST OF FIGURES Figure 1. Lake trout commercial and recreational yield in Lake Huron fiom 1912-1992. Data from Baldwin et al. (1979) and Johnson et a1. (1995). Recreational harvest data were not available prior to 1985. ................................................................. 2 Figure 2. Statistical districts of Lake Huron (Smith et al. 1961). .................................. 10 Figure 3. Selectivity patterns of the recreational and commercial gill net fisheries in Michigan waters of Lake Huron assumed by the lake trout total allowable catch (TAC) model. ................................................................................................... 23 Figure 4. Lake trout sport harvest ports surveyed by the Michigan Department of Natural Resources in Michigan waters of Lake Huron (Rakoczy and Svoboda 1994a) .............................................................................................................. 33 Figure 5. Sea lamprey wounding patterns by length class of lake trout for central (MH- 2) and southern (MH-3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for length class, adjusted for all other effects and interactions, reported with one standard error ....................................................................... 45 Figure 6. Sea lamprey wounding of lake trout in Michigan waters of Lake Huron, 1984- 1994. (a) Central region (MH-2). (b) Southern region (MH-3/4/5). Least- square means (LSM) of square root transformed wounds per fish calculated fi'om analysis of variance with length class and year as treatment factors. Estimated means for length class, adjusted for all other effects and interactions, reported with one standard error ..................................................................................... 46 Figure 7. Geographic patterns in sea lamprey wounding of lake trout less than 636 mm in Lake Huron, Michigan for 1984-1994. Least—square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for length class and geographic region, adjusted for all other effects and interactions, reported with one standard error ....................................................................... 48 XV F11 Figure 8. Mean number of sea lampreys attached to lake trout and chinook salmon caught aboard sport fishing charter boats in Michigan waters of Lake Huron, 1989-1993. Data from Michigan Department of Natural Resources. Error bars represent two standard errors. .......................................................................... 50 Figure 9. Geographic patterns in sea lamprey wounding of lake trout larger than 533 mm in central (MH-2) and southern (MH-3/4/5) regions of Lake Huron, 1984- 1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for geographic region, adjusted for all other effects and interactions, reported with one standard error. ................... 51 Figure 10. Geographic patterns in sea lamprey wounding of 534-737 mm lake trout in central (MH-2) and southern (MH-3/4/5) regions of Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated fiom analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for geograch region, adjusted for all other effects and interactions, reported with one standard error. ................... 53 Figure 11. Sea lamprey wounding of lake trout 2 432 mm in central (MH-2) and southern (MH-3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for year, adjusted for all other effects and interactions, reported with one standard error. .................................................................................................. 54 Figure 12. Sea lamprey wounding of lake trout in central (NIH-2) and southern (MI-I- 3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for length class and year, adjusted for all other effects and interactions, reported with one standard error. ........................................................................................... 55 Figure 13. Sea lamprey wounding of lake trout 2432 mm in central (MB-2) and southern (MI-I-3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for geographic region and year, adjusted for all other effects and interactions, reported with one standard error. .................................................. 56 Figure 14. Logc-based residuals from catch-at-age analysis CAAl of the southern Lake Huron lake trout population model. (a) fishery harvest. (b) survey total CPUE.69 Figure 15. Logs-based residuals fiom catch-at-age analysis CAA3 of the southern Lake Huron lake trout population model. (a) fishery harvest. (b) survey total CPUE.71 Figure 16. Standardized residuals for fishery age composition from CAA3. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. .............................. 72 Figure 17. Standardized residuals for survey age composition from CAA3. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. .............................. 73 Figure 18. Log.-based residuals from catch-at-age analysis CAAS of the southern Lake Huron lake trout population model. (a) fishery harvest. (b) survey total CPUE.75 Figure 19. Standardized residuals for fishery age composition from CAAS. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. .............................. 76 Figure 20. Standardized residuals for survey age composition from CAAS. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. .............................. 77 Figure 21. Changes in log-likelihood components for catch-at—age model fit due to varying emphasis of fishery harvest data (M). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery efl'ort, L= total. ......................................................... 79 Figure 22. Changes in logc-likelihood components for catch-at-age model fit due to varying emphasis of survey CPUE data (12). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery effort, L= total. ......................................................... 80 Figure 23. Changes inlog.-1ike1ihood components for catch-at-age model fit due to varying emphasis of fishery age composition data (3.3). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery efi‘ort, L= total. ................................................... 81 Figure 24. Changes in log-likelihood components for catch-at-age model fit due to varying emphasis of survey age composition data (M). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery effort, L= total. ................................................... 82 Figure 25. Changes in log-likelihood components for catch-at-age model fit due to varying emphasis of fishery effort data (As). Likelihood components: Ll= fishery xvii harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery effort, L= total. ......................................................... 83 Figure 26. Age-specific instantaneous mortality rates (year") for lake trout in southern Lake Huron as estimated by statistical catch-at-age analysis models CAAS and CAA6. Mortality rates averaged from 1984-1993. M= natural mortality, FR= recreational fishing mortality, ZL= sea lamprey-induced mortality, and Z= total mortality ........................................................................................................... 87 Figure 27 . Differences in estimated age-specific instantaneous mortality rates (year") with emphasis factor for fishery age composition data (kg) set at 0.1 and 1. Mortality rates averaged from 1984-1993. M=natural mortality, FR: recreational fishing mortality, ZL= sea lamprey-induced mortality, and Z= total mortality. 89 Figure 28. Allocation of estimated lake trout deaths (ages 3-10) in the main basin of Lake Huron fi'om 1984-1993. MH-l= north, MH-2= central, and MH-3/4/5= south. ............................................................................................................... 93 Figure 29. Age-specific estimates of instantaneous mortality rates (year") for lake trout in central Lake Huron. Mortality rates averaged from 1984-1993. ................... 95 Figure 30. Temporal patterns in estimated instantaneous mortality rates (year") averaged for ages 3-10 lake trout in northern Lake Huron. ............................... 96 Figure 31. Estimates of age-specific instantaneous mortality rates (year") for lake trout in northern Lake Huron. Mortality rates averaged from 1991-1993. ................. 98 Figure 32. Model estimates of lake trout (a) abundance for ages 8+, and (b) total harvest under a total allowable catch (TAC) management scenario in southern Lake Huron from 1984-2010. Maximum total instantaneous mortality for projections was 0.59 year". Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (ZL): Current= average 2;, for 1991-1993; O.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2;, is 0. ............ 100 Figure 33. Model estimates of lake trout (a) abundance for ages 8+ and, (b) total harvest under a constant fishing mortality management scenario in southern Lake Huron from 1984-2010. Fishing mortality rates for projections were based on the average of 1991-1993 rates. Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (21,): Current= average Z, for 1991- 1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 21, is 0. ........................................................................................................... 102 Figure 34. Model estimates of ages 8+ lake trout abundance in southern Lake Huron from 1984-2010. Projections were based on a no fishing management scenario xviii according to varying levels of sea lamprey-induced mortality (ZL): Current= average 2;, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. .......................................................................... 103 Figure 35. Model estimates of ages 8+ lake trout abundance under a total allowable catch (TAC) management scenario in central Lake Huron from 1984-2010. Maximum total instantaneous mortality for projections was 0.59 year". Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (21,): Current= average 2., for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2;, is 0. ....................................... 104 Figure 36. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in central Lake Huron fiom 1984-2010. Projections were based on a total allowable catch (TAC) management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average 2;, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z], is 0. Maximum total instantaneous mortality for projections was 0.59 year". .......... 106 Figure 37 . Model estimates of ages 8+ lake trout abundance under a constant fishing mortality management scenario in central Lake Huron from 1984-2010. Fishing mortality rates for projections were based on the average of 1991-1993 rates. Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (21,): Current= average Z, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. ....................................... 107 Figure 38. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in central Lake Huron from 1984-2010. Projections were based on a constant fishing mortality management scenario according to varying levels of sea lamprey-induced mortality (ZL): Current= average ZL for 1991-1993; 0.75= 7 5% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2;, is 0. Fishing mortality rates for projections were based on the average of 1991-1993 rates. 108 Figure 39. Model estimates of ages 8+ lake trout abundance in central Lake Huron from 1984-2010. Projections were based on a zero fishing management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average 21, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z], is 0. .......................................................................... 109 Figure 40. Model estimates of ages 8+ lake trout abundance under a total allowable catch (TAC) management scenario in northern Lake Huron from 1984-2010. Maximum total instantaneous mortality for projections was 0.59 year". Projections (1994-2010) were according to varying levels of sea lamprey-induced xix mortality (21,): Current= average 2;, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. ....................................... 111 Figure 41. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in northern Lake Huron from 1984-2010. Projections were based on a total allowable catch (TAC) management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average Z], for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. Maximum total instantaneous mortality for projections was 0.59 year". .......... 112 Figure 42. Model estimates of ages 8+ lake trout abundance under a constant fishing mortality management scenario in northern Lake Huron from 1984-2010. Fishing mortality rates for projections were based on the average of 1991-1993 rates. Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (21,): Current= average Z1, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2;, is 0. ....................................... 114 Figure 43. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in northern Lake Huron fi'om 1984-2010. Projections were based on a constant fishing mortality management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average ZL for 1991-1993; 0.75= 7 5% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2;, is 0. Fishing mortality rates for projections were based on the average of 1991-1993 rates. 115 Figure 44. Model estimates of ages 8+ lake trout abundance in northern Lake Huron fi'om 1984-2010. Projections were based on a zero fishing management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average 2;, for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. .......................................................................... 116 Figure 45. Change in projected abundance of ages 8+ lake trout in the year 2010 due to decreases in fishing and sea lamprey-induced mortality rates for northern Lake Huron. Fishing mortality was based on the average of 1987-1989 rates, sea lamprey-induced mortality was based on the average of 1991-1993 rates. ....... 118 INTRODUCTION Lake trout (Salvelinus namaycush) is a long-lived species that functions as a dominant predator in the fish communities of the Great Lakes of North America (Smith 1972). Historically, lake trout populations supported important commercial and recreational fisheries in these lakes (Berst and Spangler 1973). In Lake Huron, the commercial fishery averaged annual yields of 2.4 million kg from 1912 through 1940 (Baldwin et al. 1979). In the 1940s, lake trout abundance in Lake Huron declined, and stocks collapsed in the 1950s (Figure 1; Hile 1949; Baldwin et al. 1979; Coble et al. 1990). The decline of lake trout stocks in Lake Huron has been attributed to commercial exploitation, environmental degradation, and sea lamprey (Petromyzon marinas) parasitism (Christie 1974). Sea lampreys invaded the upper Great Lakes by circumventing Niagara Falls via the Welland Canal (Lawrie 1970). Sea lampreys were first observed in Lake Huron in 1937 (Shetter 1949), and then colonized most of the lake with the highest abundance in northern waters (Lawrie 1970; Morman 1979). Although there is debate about whether the initial decline in lake trout stocks was due to fishing, sea lamprey parasitism, or a 3500 - 3000 - I Ontario Commercial I Michigan Recreational 2500 ‘ Cl Michigan Commercial Yield (x 1000 kg) N 8 O O I I I I I f 7 I I I I I I I 1 1912 1922 1932 1942 1952 1962 1972 1982 1992 Year Figure 1. Lake trout commercial and recreational yield in Lake Huron from 1912-1992. Data fiom Baldwin et al. (1979) and Johnson et al. (1995). Recreational harvest data were not available prior to 1985. combination of the two (Coble et al. 1990; Eshenroder et al. 1992), it is recognized that sea lampreys were responsible for the final demise of lake trout in Lake Huron (Berst and Spangler 1973; Coble et al. 1990; Eshenroder et al. 1995). Subsequent to the collapse of lake trout populations, a rehabilitation program was implemented with emphasis on sea lamprey suppression combined with stocking of hatchery produced lake trout, and restrictions on commercial and recreational fishing (Francis et al. 1979; Smith and Tibbles 1980; Koonce et al. 1993). Initial efforts at controlling sea lampreys were in the form of mechanical and electrical barriers that prevented upstream migration of spawning adults. Subsequently, selective chemical toxicants were used in streams to kill ammocoetes (Smith and Tibbles 1980). This efiicacious technique helped to significantly reduce sea lamprey abundance and continues to be implemented in Lake Huron tributaries (Morse et al. 1995). Stocking of lake trout in Lake Huron began in 1973 (Smith and Tibbles 1980) and continues today with current populations supported almost entirely by these hatchery fish (Johnson et al. 1995). Since the collapse of lake trout stocks, no commercial fishing for lake trout in Michigan waters has been allowed except for a tribal fishery in the northern region (Smith and Tibbles 1980). These restrictions on harvest have contributed to an increased abundance in Lake Huron, though it is less evident in the northern areas of the lake. Although some progress has been made in reducing the high mortality experienced by lake trout, sea lampreys are still one of the main factors in inhibiting the rehabilitation of lake trout (Eshenroder et al. 1995; Johnson et al. 1995). Various studies reporting the negative effects of sea lamprey on lake trout populations have been reviewed by Coble et al. (1990). Some of these studies investigated the relationship between decreasing lake trout abundance and the incidence of sea lamprey wounds and showed that they were correlated (Fry 1953; Budd et al. 1969). Wounding rates have been reported to increase with length of lake trout (Eschmeyer 1957; Farmer and Beamish 1973; Pycha and King 1975; Swink 1991), vary temporally (Pycha and King 1975; Jacobson 1985) and geographically. Hypotheses for explaining why wounding rates increase with length of lake trout include: 1) lower wounding rates on smaller hosts are due to higher lethality of sea lamprey attacks on smaller fish than larger fish (Eschmeyer 1957; Swink 1990); and 2) sea lampreys select for larger hosts (Budd and Fry 1960; Farmer and Beamish 1973; Pycha and King 1975; Cochran 1985; Swink 1991). These studies indicate that mortality caused by sea lampreys is likely to differ according to size of lake trout- which also implies that sea lamprey-induced mortality varies by age of lake trout. Sea lamprey wounds on lake trout are a record of sea lamprey attacks and an index of sea lamprey abundance (King 1980). Eshenroder and Koonce (1984) reported a protocol for quantifying and translating sea lamprey wounding data to lamprey-induced mortality rates. This procedure is dependent on an estimate of the probability of surviving a sea lamprey attack. Current estimates of this parameter for various lengths of lake trout have been reported from laboratory experiments conducted by Swink (1990). A standardized classification of wounds inflicted by sea lampreys on lake trout (King 1980) is used by most of the US. fisheries agencies in the Great Lakes, and has led to a substantial database on lake trout wounding rates. Sea lamprey-induced mortality estimates from this procedure can be used in population models to evaluate the effects of sea lampreys on lake trout abundance. The goal of lake trout restoration for Lake Huron is to re-establish self-sustaining populations that can produce a yield of 1.4 to 1.8 million kg annually (DesJardine et al. 1995). Due to the low abundance of lake trout, recent (1986-1992) annual recreational and commercial harvest of lake trout averaged 204,000 kg, which is less than 15% of the goal, and less than 10% of historic yield (Johnson et al. 1995). The success of lake trout rehabilitation has been limited by low spawner abundance and excessive mortality rates (Hatch 1983; Johnson et al. 1995). Healey (1978) reported that in order for a lake trout population to sustain itself, total annual mortality should not exceed 50%. The desired maximum for total annual mortality for lake trout restoration has been set at 45% (equal to an instantaneous rate, Z of 0.59 year") by the Great Lakes Fishery Commission (GLFC) as an attempt to increase spawner abundance (Johnson et al. 1995). The lack of progress in the rehabilitation program in northern Lake Huron has been attributed to the high abundance of sea lampreys over the past decade in conjunction with exploitation by the tribal fishery (Johnson et al. 1995; Eshenroder et al. 1987). Estimates of total annual mortality for lake trout in US. waters of Lake Huron, based on catch curves applied to data from spring assessments (1982-1992), have been reported to be greater than 70% in the north, with sea lampreys accounting for at least V2 511 CU. U111 I97 fella dEIer (Wisc “Onhf p r OleC; fiom “ 33% of annual losses of lake trout larger than 630 mm in that region (Johnson et al. 1995). However, these reports do not address the age-selective effects or the relative magnitude of sea lamprey-induced and fishing mortality. In order to rehabilitate lake trout, overall effects of sea lamprey parasitism and fishing mortality must be determined in order to gauge progress toward the goals. However, it is important to take into account the dynamics of each mortality source by understanding the age-selectivity of each mortality source in relation to temporal variations in fishing or sea lamprey abundance. It is important to assess which ages are sufi‘ering the highest mortality and how this affects spawning stock abundance. Catch curve approaches are not robust in this respect. Catch curve techniques rely on unrealistic assumptions of age-independent mortality rates, equal vulnerability to the sampling gear for ages used in the analysis, and equal recruitment for all cohorts (Ricker 1975). With reliable estimates of lake trout deaths due to sea lampreys and fishery harvest, managers can adjust lamprey control programs or fishing regulations to reach rehabilitation objectives. Stock assessments have been performed for lake trout using an age-structured, deterministic Total Allowable Catch (TAC) model in US. waters of Lake Superior (Wisconsin State/'1‘ ribal Technical Committee 1984; Ebener et al. 1989) and for parts of northern Lake Huron (Technical Fisheries Review Committee 1992). This model projects levels of allowable harvest based on estimates of sea lamprey-induced mortality from wounding data, fishing mortality, and desired maximum for total mortality. 31 YES The goal of this study was to evaluate the effects of sea lamprey parasitism and fishing on lake trout populations the main basin of Lake Huron. The specific objectives were to: 1. Analyze patterns in sea lamprey-induced mortality, as indexed by wounds, for lake trout in the main basin of Lake Huron. These results were used as a guide in accomplishing other specific objectives. 2. Estimate abundance, sea lamprey-induced, and fishing mortality for lake trout by constructing age-structured population models for the main basin of Lake Huron. 3. Evaluate changes in future spawning stock size according to decreases in sea lamprey-induced and fishing mortality. To accomplish objectives 2 and 3 of this study, lake trout population models were developed for the main basin of Lake Huron that integrated sea lamprey-induced mortality estimates from standardized wounding data (collected by the Chippewa-Ottawa Treaty Fishery Management Authority (COTFMA) and Michigan Department of Natural Resources (MDNR)) along with estimates of fishing mortality based on commercial and recreational harvest and effort data supplied by COTFMA, MDNR, and Ontario Ministry of Natural Resources. Model calibrations were performed using statistical catch-at-age approaches that used auxiliary information to estimate model parameters (Megrey 1989). Auxiliary information included fishery harvest-at-age, fishery effort, and standardized research survey indices of abundance. YE: des cah gnu S? /g. Spllllj c0ndL Huron Consist fiOmS METHODS The methods are described in five subsections, I first describe how I assessed patterns in sea lamprey wounding to determine how sea lamprey-induced mortality experienced by lake trout populations varied over time, among geographic regions, or among lake trout size categories. In part two, I then describe how the results from these analyses were used to guide the development of models to estimate wounding rates for years or geographic regions where few or no lake trout were examined. In part three, I describe the lake trout population model. The fourth part of the methods describes the calibration of the model using statistical catch-at-age analysis. In the last part, I describe simulation runs for the population models. Patterns in sea lamprey wounding I used sea lamprey wounding data for lake trout in Lake Huron, collected in spring gill net surveys from 1984-1994 by COTFMA and MDNR. These surveys were conducted from April through June at various fixed stations in Michigan waters of Lake Huron using graded-mesh multifilament, nylon gill nets that were 1.8 m deep and consisted of nine panels that were 30.5 m long with mesh sizes (stretch measure) ranging from 51 mm to 152 mm in 13 mm increments (Mema et al. 1981; Johnson and VanAmberg 1995). Wounding data were recorded using the protocol developed by King 8 9 (1980). I used only recent, potentially lethal wounds (type a, stages 1-3 (King 1980)) based on the recommendations of Eshenroder and Koonce (1984). Potentially lethal, recent wounds were characterized as wounds that have penetrated through the scales and epidermis exposing the underlying musculature (King 1980). Eshenroder and Koonce (1984) also recommended that spring wounding rates should be used because these wounding rates were correlated with catches of spawning sea lampreys at stream barriers, which was used as an index of lamprey abundance. Standardization of sea lamprey wounding data began in 1984, and I used data from 1984-1994. Wounding rates were calculated by length class of lake trout, geographic region, and year. I established four length categories (432-533, 534-635, 636-737, >737 mm) in accordance with conventions used by COTFMA, Great Lakes Fishery Commission (GLF C), and MDNR. These length classes matched those for which estimates of lethality of sea lamprey attacks were available (Greig et al. 1992). I focused on three ‘ areas in Lake Huron: northern (MH-l), central (MH-2), and southern (MH- 3/4/5)(Figure 2). These geographic regions were thought to represent discrete lake trout populations based on previous surveys (J. Johnson, Alpena Fisheries Research Station, WNR, pers. comm). Regions MH-3, NIH-4, and MH-S were pooled based on the same reasoning. Lake trout populations in these three geographic regions of the main basin of Lake Huron are exposed to different levels of fishery harvest and are reported to be exposed to differing levels of sea lamprey parasitism (Johnson et al. 1995). Low sample sizes and complete absence of data for some strata in the wounding database prevented the use of one statistical analysis to simultaneously examine the 10 F . E ‘0. )nlptPt. "as“ 3 WM... 0 PL . I l//' ' Orwnond 1.-..» . ,<..tglora‘t:;o|st',’ u Coctbum '- Q‘X. Prune Pt. 1;, MH—I "I -\.\ 68-1)(‘Puwaon Rock mom ,‘l .\~ H r on -- "Jim: ‘i\ / \.\ m" :35." ‘ ‘(lntzwmlamz . '3‘ 68-3 Adm I'D.” .N." Y” " “,«Covo I. 68-2 “‘ v.3. “”2 l‘ 0H'2 " slam. “affix: L f. .. if???" .t ................. '4 _____ 3.. - - - § . . 7 l cl... .. ~ . 68-4 Elect: """"""" \ / RM? \ 3 4. OH- ",1 MH-3 1 at, Au sure PL. ‘1 BAH-4 l ,x' j» ------- Pt. cm: a." “‘1‘... ‘ SI In M. I ' "' MH—S \ Lake Huron 801C")! 014.4 """""" 0'3".“ Boundary Richmondvlllc ----------- - International Boundary MH-s .’ ------- Grand Bond ‘1 i 0 IE 3 E if, 'OH-S Statute Mil“ Port Huron ' 30min Figure 2. Statistical districts of Lake Huron (Smith et al. 1961). 11 efi‘ects of lake trout length, geographic region, time, and their interactions on wounding rates (Table 1). Therefore, I used different subsets of the database in a suite of analyses, each aimed at evaluating one or more of these main factors. Subsets were selected so that a wide range of one or more factors could be included to provide contrast for those factors, while other factors were necessarily represented by fewer levels. This was done so that all combinations of the factors used in an analysis contained some data. A variety of subsets were analyzed so that each factor and potential interactions could be evaluated. These analyses were restricted to subsets of levels that did not include missing cells. Data were square root transformed to approximate normality (Miller 1984) based on previous indications that frequencies of sea lamprey wounds on fish were Poisson distributed (Eshenroder and Koonce 1984). Analysis of variance (ANOVA) models were constructed using subsets of the transformed data to test for effects of main factors: length class (on), geographic region (13,-), and year (61.) on sea lamprey wounding rates. The full model was: Witt: u + or; + 31 + 5k + 0430 “l“ 0‘50: + 3511: + 043501: + Silk (1) where Wm, was estimated mean wounds per fish for ith length class, jth geographic region, and kth year; aflij was the interaction of length class and geographic region, a6“. was the interaction term for length class and year, [38,-]. was the interaction of geographic region and year, and 04365;. was the interaction term for all three main factors. For some subsets of the data, one or more of the main effects and its associated interactions were not included because only one level of those factors were represented. ANOVA models were fit using the General Linear Models procedure (SAS Institute 1985). 12 Table 1. Numbers of lake trout examined for sea lamprey wounds in Michigan waters of Lake Huron in spring gill net surveys and subsampling of tribal gill net and trap net catches. Observations are stratified by lake trout length class. Data provided by the Chippewa-Ottawa Treaty Management Authority, and Michigan Department of Natural Resources. Region: MH-1= north, MH-2= central, and MI-I-3/4/5= south. Region 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 432-533 mm MH-l 457 134 380 257 143 71 63 130 202 279 78 NIH-2 181 206 181 89 44 123 31 34 87 62 61 MH—3/4/5 247 127 240 118 159 126 38 18 38 83 100 534-635 mm MH-l 171 30 53 45 14 ll 20 23 31 55 18 MH-2 61 74 80 82 15 83 23 66 81 52 82 MH—3/4/5 217 363 265 219 203 139 149 39 77 140 116 636-73 7 mm MH-l l9 0 2 4 l l 0 1 l 9 3 MH-Z 19 27 22 37 8 28 6 59 47 12 43 MH-3/4/5 359 450 241 220 244 233 281 98 135 66 137 > 73 7 mm MH-l 2 l 0 0 0 0 0 0 l 0 0 MH-Z 6 5 5 2 1 4 l 7 7 l 2 MH—3/4/5 82 149 65 72 70 45 85 72 153 68 65 13 Estimation of $3 lamprg-induced mortalig In order to estimate sea lamprey-induced mortality for all age classes in each year and geographic area, estimates of wounding rates were needed for each combination of these factors. As indicated in the previous section, attempts were made to estimate wounding rates that were not available for all levels for each of the main factors. Although the initial analyses provided information about how sea lamprey wounding rates were influenced by lake trout size and geographic location (see Patterns in sea lamprey wounding in Results section), the approach led to biased estimates of wounding rates after back-transformation. This was true even after attempts at bias correction following procedures suggested by Miller (1984). This was determined by comparisons between least-square means with original mean wounding rates, when available. Thus, results fi'om these analyses were not suitable for estimating absolute wounding rates and corresponding sea lamprey-induced mortality rates. My objectives here were first to systematically estimate mean wounding rates for specific year by length class by geographic region combinations where data were not sufficient or absent with the least amount of extrapolation. The second objective was to compute age-specific lamprey-induced mortality rates for each region and year for use in the lake trout population models. The first objective was approached by constructing another set of ANOVA models based on the information found in the analysis of patterns in sea lamprey wounding. The patterns observed were that wounding rates increased with length class of lake trout, and were higher in the central region of Lake Huron than in the south for fish >533 mm. Therefore as an example, in order to estimate a wounding 14 rate for a missing year in the central area for the 534-635 mm length class, a model can be constructed based on the relationship between the central and southern areas for all fish >533 m using the available data for all other years. Overall, this second set of ANOVA models used the available data to estimate effects of year, length class, and geographic region on mean wounds per fish. These estimated effects were then used to predict mean wounding rates for specific combinations without data. However, in northern Lake Huron there were insufficient data for lake trout >533 mm to reliably estimate wounding rates using these ANOVA models. Thus, ANOVA models constructed to estimate wounding rates were only used for central and southern Lake Huron and for the smallest length class in the north. Sea lamprey-induced mortality for lake trout >533 mm in northern Lake Huron was estimated using a different approach described later in the methods section (see Calibration of the northern and central lake trout population models). The second set of ANOVA models were constructed using untransformed mean wounds per fish as observations to estimate wounding rates for each combination of main factors in which data were absent. In these analyses, the models assumed no interactions since replicate observations were not available. Although unlikely to be strictly true, I attempted to restrict the extent to which I extrapolated across very difi‘erent size classes or distant geographic areas to minimize problems due to interactions. Mean wounds per fish for length class i, region j, and year k were calculated for each combination of main factors by: SE; for Wht the \l’eig 0118 o (21) ft 15 _ 2wijk wijk = (2) n 1th where w was number of observed wounds and n was number of fish. Mean wounds per fish were only calculated when data from 40 or more fish were available. This sample size criteria was established because of the inability to reliably estimate wounding rates such as 0.1 wounds per fish when less than 40 fish were examined. For example, the coeficient of variation for a mean of 0.1 wounds per fish and a sample size of 40 fish was about 50%. The means were weighted in the ANOVA by the inverse of the estimated variance of the mean to reduce bias from lower sample sizes. Relationships between main factors, guided by the results from the analyses of sea lamprey wounding patterns, were used to develop models to estimate wounding rates for missing cells. The basic form of the model was: Wijk: 11 + a, + 135+ 51: + 30k (3) where a; , 13,-, and fikwere as defined in equation 1. Based on review of the estimates of the variance of the means, estimated variances less than 0.0009 were set to 0.0009 when weighting was done so that observations with extremely low variance estimates did not dominate solutions. My estimates of variance did not account for some sources of variability, such as process error, therefore this procedure was implemented so that any one observation would not completely control the solution. The size- and year-specific instantaneous rate of sea lamprey-induced mortality (21,) for each geographic area of Lake Huron was estimated using Eshenroder and Koonce’s (1984) procedure: l6 — 1' Psi 21.3,]: = “In: (4) where W: was the mean wounds per fish for the ith length class in year k, and P3,.- was the probability of surviving a sea lamprey attack for the ith length class of lake trout. Assumptions in using this model included: 1) P3,,- was independent of prior attacks, and 2) Wu: was representative of the wounds accumulated over a year (see Eshenroder and Koonce (1984) for further discussion). Estimates of the probability of survival fiom a sea lamprey attack were reported by Swink (1990) based on laboratory studies using lake trout. Summarized values of the survival probabilities were: 0.35 for 432-533 mm, 0.45 for 534-635 mm, 0.55 for 636-737 mm, and 0.55 for >737 mm lake trout (Greig et al. 1992). These values were used in this study since no in situ estimates of P3,,- were available. Since this study used an age-structured model, the length-based estimates of lamprey-induced mortality had to be converted to age-specific values. This was accomplished using an age-length key (Tables 15-17, Appendix) and the equation: Zr... =2 "j ZLJ (5) J . where na‘, was the number of fish of age class a and in length class j, and Z1“j was the instantaneous rate of lamprey-induced mortality for length class j. For a specific age, this equation multiplies the proportion of fish in each length class by the appropriate rate of lamprey-induced mortality and then sums over all length classes for that age. Virtually all lake trout sampled in research surveys were of hatchery origin, thus age of fish were 17 determined from fin clip patterns. Scales were used to age unclipped lake trout (J. Johnson, Alpena Fisheries Research Station, MDNR, pers. comm). Age-length keys for each of the regional populations used in this study were tabulated fi'om unpublished data collected in spring gill net surveys of lake trout provided by the MDNR (Tables 15-17, Appendix). These keys were based on data pooled from 1984-1994. An assumption made in the above procedure was that mean wounds per fish, which are sampled in the spring of year y+ 1, were representative of attacks that occurred in year y. The length at which a lake trout suffered its attacks may be shorter than when sampled in the spring survey. Therefore, fish growing into a larger length class, which has a different P3,.- , could potentially bias the mortality estimate. However, sea lamprey attacks are most prevalent in the late summer and fall (Jacobson 1989) and this is after much of the year’s grth has occurred in lake trout (Martin and Olver 1980). Hence, the effect of the violating this assumption, though not estimated, is likely to be small. r o ulation model The model used in this study was based upon a total allowable catch (TAC) model developed for lake trout in Lake Superior (Wisconsin State/Tribal Technical Committee 1984; Ebener et al. 1989). Initial efforts at construction and parameterization of the Lake Huron models were performed by M. Ebener (COTFMA) and J. Johnson (MDNR) of the Lake Huron Technical Committee, Great Lakes Fishery Commission (GLFC). This study was initiated, in part, to complete and calibrate lake trout TAC models for Lake Huron. The major advance presented here was the use of statistical 18 catch-at-age procedures so that better parameter estimates could be obtained based on more of the available data. The lake trout TAC model integrates age-specific estimates of sea lamprey- induced, natural, and fishing mortality to estimate abundance and projections of allowable harvest. The idea underlying the model is that stocks can be managed by adjusting fishing mortality based on information on recruitment, harvest, and the other sources of mortality (i.e., sea lamprey-induced mortality). Regulation of fishing mortality can be in the form of harvest quotas or effort restrictions. Population models were constructed for each of the three regional stocks of lake trout: northern (MH-l and northwest part of OH-l), central (MB-2, most of OH-l, OH-2), and southern (MI-I-3, MI-I-4, MH-S, NIH-6, OH-3, OH-4, OH-5)(see Figure 2). The time series modeled in each area was from 1984-1993. Prior to 1984, recreational harvest data were unavailable and wounding data were not recorded following the same protocol. Model parameters, variables, and constraints that were available for use in this study are listed in Table 2. In the main basin of Lake Huron, essentially all lake trout were derived from hatchery-stockings in Michigan waters. Canada has not stocked lake trout in the main basin, and there were insignificant immigrations of fish fiom the North Channel, and Georgian Bay (L. Mohr, Lake Huron Management Unit, Ontario Ministry of Natural Resources, pers. comm). Thus, all lake trout in Canadian waters of the main basin were assumed to be immigrants from the adjacent populations in US. waters. Table 2. Previously reported parameter values for estimating mortality rates of lake trout in the main basin of Lake Huron. COTFMA= Chippewa-Ottawa Treaty Fishery Management Authority, MDNR= Michigan Department of Natural Resources. Parameter Description Source Values (age or length- (units) class) fc, , (year") Commercial fishing intensity COTFMA Proportional to effort in harvest reports fa, y(year") Recreational fishing intensity MDNR Proportional to efi‘ort in year y in creel survey reports M. (year") Natural mortality rate Rybicki 0.799(1), 0.25 (2,3), (excluding sea lamprey- (1990), 0.20 (4), 0.15 (>4) induced mortality), assumed MDNR temporally constant P3,; Probability of surviving a sea Swink (1990) 0.35 (43 2-533 mm), lamprey attack for length 0.45 (534-635 mm), classi 0.55 (636-737 mm), 0.55 (>737 mm) Sq. Commercial fishery COTFMA 0(1), 0.01 (2), 0.10 selectivity (3), 0.75 (4), 1 (5), 0.86 ( 6), 0.55 (7), 0.49 (8), 0.39 (9), 0.2 (>9) S3,. Recreational fishery MDNR 0 (1), 0.01 (2), 0.10 selectivity (3), 0.75 (4), 0.85 (5), 1 (>5) W... Mean number of sea lamprey COTFMA, From annual spring wounds per fish in length MDNR surveys class i, in year y 20 Models for northern and central Lake Huron were similar in that both areas have recreational and commercial fisheries. All lake trout harvests in statistical districts OH-l and OH-2 of Canada (Figure 2) were incorporated into the harvests of the northern and central models. Southern Lake Huron was considered to have only recreational fishing, though there was some commercial harvest of lake trout in adjacent Canadian waters. All this commercial harvest of lake trout in OH-3, OH-4, and OH-S (Figure 2) was incorporated into the sport harvest of the southern model since no accompanying biological information was available (see later section titled Statistical catch-at-age analysis of the southern Lake Huron lake trout population model). Substantial migration between geographic areas was thought to occur only between northern and central Lake Huron, with movement being unidirectional northward. The proportion of stocked fish that emigrate to the north has been approximated at 60% based on coded-wire tag results (J. Johnson, Alpena Fisheries Research Station, MDNR, pers. com). The northern and central models account for this migration by adjusting the age-1 recruitment numbers. Sixty percent of the age-1 fish in the central area were subtracted and then added to the age-1 abundance in the north. Lake trout abundance Lake trout numbers (N) at age a+1, and year y+ 1 were computed using an exponential mortality equation: 21 _ N e-zu = N e-(zL',_,+l=,.,+M.) Na+l.y+l _ my my (6) where Z was the total instantaneous mortality rate, ZL was the lamprey-induced mortality rate, F was the rate of fishing mortality, and M was the natural mortality rate excluding sea lamprey-induced mortality. Since there is no significant natural reproduction of lake trout in Lake Huron, recruitment was a direct filnction of hatchery stockings. Lake trout are stocked as yearlings and fall fingerlings, therefore age-1 abundance was equal to the numbers of stocked yearlings and the survivors of fall stocked fingerlings. Based on values used by Ebener et al. (1989), forty percent of the number of fall fingerlings stocked were assumed to survive to yearlings, thus the abundance at age-1 was the sum of the number of yearlings stocked in year y and 40% of fall fingerings stocked in year y-I (Table 18, Appendix). Natural mortality Available values of natural mortality rates (M.), excluding sea lamprey-induced mortality, for hatchery stocked lake trout ages 1-3 were reported by Rybicki (1990) in a study conducted in Grand Traverse Bay, Lake Michigan. For lake trout age-4 and older, unpublished estimates of natural mortality rates were provided by the MDNR (J. Johnson, Alpena Fisheries Research Station, MDNR, pers. comm). These values for M; are listed in Table 2. Natural mortality rates were also estimated by statistical catch-at- age analysis (CAA) of the lake trout population model using information on age-specific harvest and efl‘ort fiom the fishery and research surveys (see later in section titled 22 Statistical catch-at-age analysis of the southern Lake Huron lake trout population model). Fishing mortality The fishing mortality rate of the recreational fishery (F by) was modeled as being separable into age- and year-specific components by: Flay = SK: fky (7) where S3,, was the recreational fishery selectivity on age a, and fa, was fishing intensity which scales the overall recreational fishing mortality for year y. In the southern region, both fR, , and SRa were estimated as parameters by CAA. Prior estimates of the recreational selectivity pattern assumed it to be asymptotic because larger fish tend to be targeted by anglers (Figure 3; Table 2). In this study, I assumed that recreational selectivity was constant for ages 9+ and estimated the specific values for ages 2-8 rather than using the values in Table 2. To obtain an unique parameterization (Doubleday 1976), SR... was set to 1 for ages 9+ fish, and thus fa , was an estimate of the actual fishing mortality rate for those ages. The recreational selectivity values estimated by CAA of the southern model were used to estimate recreational fishing mortality rates in the northern and central population models. In the northern and central regions, a commercial gill net fishery exists in addition to a recreational fishery, therefore an additional fishing mortality component was added to those models with: FC,a,y = SQ: fC,y (8) 23 11 9--o--o--e--o--o--e--o”one - - o - - Recreational Commercial 0.5 - Selectivity 0.4 - 0.3 - 0.2 - 0.1 - 0 fit I I I j I I 12 3 4 5 6 7 8 9101112131415 Age Figure 3. Selectivity patterns of the recreational and commercial gill net fisheries in Michigan waters of Lake Huron assumed by the lake trout total allowable catch (TAC) model. 24 Values for commercial fishery selectivity were based on studies conducted by tribal biologists in Lake Superior (M. Ebener, COTFMA, pers. com). The selectivity pattern for this gear was dome shaped (Figure 3; Table 2). Recreational fishing intensity (fa, y) for the northern and central regions was estimated by: fa. y = QR ER. y (9) where qg was the proportionality constant (catchability coefficient), and ER , was the reported recreational fishing effort in year y in units of angler hours. Since fluctuations in recreational harvest matched the patterns in recreational effort, this procedure worked well for estimating recreational fishing intensity. Initial attempts to estimate commercial fishing intensities (ch) were approached by adjusting qc to scale the reported effort so that predicted annual harvest would be equal to observed values from harvest reports. This procedure was unsuccessful due to inconsistencies between the patterns in reported commercial effort and the patterns in reported commercial harvest. Therefore, year-specific commercial fishing intensities were estimated as parameters to match the model’s predicted harvest to observed values using equation 8. Fishing intensities for MH-3/4/5 were estimated by CAA. Fishery harvest (C.,,) for age a, in year y was calculated using the Baranov catch equation (Ricker 1975): (10) a.y a,y _ -(ZL..'y+F._y+M.) C,.y=N F [I e ] (Z +Fa.y + M,) L.a.y where F..,y = FM,y + F¢,.,y in northern and central Lake Huron. In the southern region, 25 only a recreational fishery exists so F ., y = F R, ., ,. For the northern and central area, the recreational or commercial harvest was estimated by: l _ e‘(zL.a,y+Fa,y+Ma) C = N F 11 X,a,y a,y X,e.y (ZLAy + FLy + M.) ( ) where X was either R (recreational) or C (commercial). Similarly, numbers of lake trout killed by sea lampreys (CL , u) were estimated using: (12) 1_ e-(zL.._,+l-‘..,+M.) ] CI...” = NLyZLJJ [(Z + I?“y + M.) I..a.y Biomass of the population was calculated using mass-at-age information by: By = ZNwm, (13) where By was the biomass in year y, and Nmy was the numbers at age a in year y calculated by the model, and ma was the mass at age a. The yield or biomass of the harvest was calculated in a similar fashion: Y, = z_c,_,m, (14) Average mass-at-age used in this model were based on the compilation of MDNR survey data from 1984-1994 (Table 3). A von Bertalanffy model was used to estimate average mass for missing ages (C.P. Ferreri, Pennsylvania State University, unpublished). Statistical catch-at-age analvsis of the southern Lake Huron lake trout population model Based on the availability of harvest-at-age information from MDNR creel and research surveys of Lake Huron, statistical catch-at-age (CAA) analysis was implemented to calibrate the lake trout population model. This was only performed for 26 Table 3. Average mass-at-age of lake trout in Michigan waters of Lake Huron. Data provided by Michigan Department of Natural Resources and C .P. Ferreri, Pennsylvania State University. Age North (MI-I-l) Central (MB-2) South (MH-3/4/5) 1 0.09 0.09 0.09 2 0.157 0.179 0.223 3 0.365 0.458 0.593 4 0.731 1.041 1.293 5 1.140 1.712 2.123 6 1.539 2.474 2.931 7 1.878 2.861 3.467 8 2264* 3.419 3.964 9 2610* 3928* 4.390 10 2947* 4386* 4.765 11 3276* 4816* 5.141 12 3597* 5220* 5.388 13 3910* 5599* 5.451 14 4216* 5956* 5.486 15 4514* 6290* 6.056 16 4804* 6605* 6291* 17 5088* 6900* 6453* 18 5364* 7178* 6596* 19 5634* 7439* 6722* 20 5897* 7684* 6833* >20 6154* 7914* 6930* * estimated by von Bertalanffy procedure (C.P. Ferreri, Pennsylvania State University, unpublished) all 210 ma Talc 27 the southern stock because there were insufficient data for the other regions (e.g., recreational harvest and fishery age-composition not available). The CAA approach integrates information on fishery harvest, age composition of the fishery harvest, fishery effort, survey catch per unit effort (CPUE), and age composition of the survey CPUE to estimate parameter values of the lake trout population model. Some of the reported parameter values listed in Table 2 were re-estimated by CAA. Model parameters that were estimated by the CAA analysis are listed in Table 4. In addition to estimating fishing mortality related parameters, CAA analysis was also used to assess the sensitivity of parameters used to estimate sea lamprey-induced mortality (21,) by including a proportionality coefficient (11’) to equation 4 as follows: ,- 1‘ PS.i 21.1.1; : 1* “(1.1: P (15) s.1 The proportionality coefficient would equal 1 if W11: , and P3,,- were accurate, and the assumptions used to relate these to Z), were met. Thus any such deviations from these assumptions would be indicated by the departure of the CAA estimate of p’ from unity. Natural mortality excluding sea lamprey-induced mortality (M.) was estimated using two approaches. It was possible to estimate M, in this study because recruitment was known and there were data to estimate sea lamprey-induced mortality. The first approach was based on the assumption that the relative differences in natural mortality across ages from the reported estimates (Table 2) were accurate, but the specific values may be incorrect. Hence, M. = c(M.*), where M." were the reported natural mortality rates. The value of c would equal 1 if the current vector of natural mortality rates were 28 Table 4. Parameters of the southern Lake Huron (MH-3/4/5) lake trout population model estimated by statistical catch-at-age analysis. Parameter Description Units ll, Proportionality coefficient for sea lamprey-induced unitless mortality c Proportionality coefficient for natural mortality unitless fl , Recreational fishing intensity for year y year" M. Natural mortality (excluding sea lamprey-induced year" mortality) for age-1 lake trout N;1934...N20+,1984 Initial abundance-at-age in 1984 numbers Sp~ 2 ”Sp, 3 Recreational fishery selectivity unitless S‘z ...S"‘4 , 8*5 ...S*,o+ Survey selectivity, S*5 assumed to equal 1 unitless t Rate of decrease in natural mortality (excluding sea year" age" lamprey-induced mortality) for lake trout qR Recreational fishery catchability coefficient angler hours" q" Survey catchability coefficient meters of gill net" 29 accurate. Otherwise, any variations in natural mortality fiom the current rates would be indicated by the CAA estimate of c. The second approach estimated natural mortality as a type 3 exponential survivorship function since it reasonably describes the age-specific pattern of mortality in lake trout. The equation was: M. = M1 64"" +0.1 (16) where M) was the instantaneous rate of natural mortality for age-l lake trout, r was the rate of decrease, and a was age. M, and rwere estimated by catch-at-age analysis as parameters. This procedure facilitated the solution process by allowing only two parameters to be estimated for natural mortality. The minimum natural mortality rate was set at 0.1 so that the function did not underestimate natural mortality rates for older fish. This minimum value was set just below the natural mortality rate used in the Lake Superior lake trout models, which is based on a catch curve applied to a refirge population in that lake, as described by J. Selgeby (National Biological Service, Ashland, WI) at the July 1995 Lake Superior Technical Committee Meeting (J. Bence, Michigan State University, pers. comm). Following Methot (1990), differences between model predictions and observed values were quantified using a specified error model cast in terms of a log-likelihood function. Optimum parameter values were ones that maximized the log-likelihood. The maximum likelihood solution was found numerically using a quasi-Newton search algorithm, central differencing to estimate the partial derivatives of the objective and constraint functions, and quadratic extrapolation to obtain estimates of the parameters. More specific details of the maximum likelihood approach for analyzing catch-at-age 30 data are explained by Foumier and Archibald (1982), Methot (1990), and Bence et al. (1993) The log-likelihood (L) equation was: L=L1+L2+L3+L4+L5 (17) where L, was the log-likelihood of the fit to the fishery harvest, L; was the log- likelihood associated with the survey index of abundance, L3 was the log-likelihood of the fit to the fishery age composition, L. was the log-likelihood associated with the fit to the survey age composition data, and L 5 was the log-likelihood of the fit to the fishery effort data. The individual components were: ( F — C -2‘\ ln[Cf] N 1 L = - (0.5 y > +ln[ J 18 . Z, a. [OH/E ] ( 1 K ( - - j) K r r K «21) lfl(—K—f'] L N l L = — 0.5 ’ +ln 19 2 Z; 0. [(0. 21:] ] ( ) K ( - - )/ L3 11; mm l...nk! [2,1,23,1n(p,.,)]+[zyln[ 1,1 1] (20) ' 1 L4 421.132.1313 “‘(P..y)]+[z,ln(u m 1:1...ukll] (21) 31 f f - —2‘\ E “IKE—Z] N 1 L = — +0.5 ’ . +ln[ J (22) 5 2’ OE [ow/21: :l K ( - — ,/ where C, was the model predicted fishery harvest (equation 10), C’, was the observed fishery harvest, K was the predicted survey CPUE, K ', was the observed survey CPUE, N was the total number of years of data, P4,, was the proportion-at-age of the predicted fishery harvest, P ’a, was the observed pr0portion-at-age of the fishery harvest, J, was the sample size for the fishery age composition with maximum values set to 200, na was the fishery harvest for age a, p“, was the predicted survey proportion-at-age, p ’0, was the observed proportion-at-age of the survey catch, j, was the sample size for the survey age composition, u,, was the survey CPUE for age a, E, was the predicted fishery effort, E’, was the observed fishery effort, Of was the standard error (s.e.) of the log of harvest, 6. was the s.e. of the logc of the survey CPUE, and as was the s.e. of the log of fishery effort. The predicted survey CPUE (K, y) was calculated by K... = q'SIN... (23) where q* was the survey proportionality constant, Sa“ was the survey selectivity at age a, and N4 , was the number of lake trout at age a, and in year y (from equation 6). K, was the sum of all Ka, , for year y. The predicted fishery effort (E,) was calculated using 32 By = —’ (24) where}, was the fishing intensity in year y, and q was the proportionality constant. Estimates of total recreational harvest (C ’,) of lake trout from 1984-1993 were calculated for use by the catch-at-age procedures (Table 19, Appendix). These data were fi'om MDNR creel surveys conducted at ports in Lake Huron and represents all recreational harvest of lake trout in Michigan waters of Lake Huron. For southern Lake Huron, the ports with significant harvest were Oscoda, Harrisville, Tawas, Port Austin, and Harbor Beach (Figure 4). Harvest data were not available for all ports in all years. Missing harvest data were estimated based on the ratio of the harvest in ports without data to the harvest in ports with data from the other years where data on all ports were available (Table 19, Appendix). Recreational fishery age composition information was derived from subsamples of the recreational harvest by MDNR creel clerks. These subsamples were usually collected monthly in each year from May through September. Recreational fishery age composition (P’.,,) information was only available for 1985-1988 and 1991-1992 and were not available for all months (Table 20, Appendix). Fortunately, catch-at-age analysis does not require age composition data for every year to calibrate the population model. The harvest-at-age information for each year with data available was estimated by pooling the harvest subsamples across all months by ports and estimating the proportions for each age and then multiplying these values by the total harvest. Estimates of an index of recreational fishery effort (E’,) were available for 1984- 33 Potagannissing Bay Les ($323“ \ Drummond St. Ignace . , 6% I \ ‘\ Rogers City 0 \\ ‘ Q Rockport . \ Alpena . ‘ O LAKE 1 . , 2.. \ H U R 0 N I Harrisville . \\ OSCOda . 1‘ MICHIGAN Tawas . \ 8* ‘~ Au Gres . $Q¢1, Nam“ was estimated by: Na.1984 = [N1. Isms-1)] (Dr 92 Pan-1) (26) where N, 19m.-1)was the age-1 abundance of a cohort as determined from stocking data. The “optimum” set of commercial fishing intensities (fay) and qR were those that minimized the difference in the log sum of squared residuals for total harvest (Deriso et al. 1985; Megrey 1989). The objective function (4)) was written as: 41f...1y = 1984 — 19931.4..1= Z[(Iog. C2...) — (log. C... 1]’ + ’ (27) Z[(log, cg”) - (log, CM )]2 y where C ' a, was observed commercial harvest in year y, Cc, was predicted commercial harvest, C ’3, was observed recreational harvest, and CR, was predicted recreational harvest. Reported recreational harvest from MDNR creel reports are listed in Table 19 of the Appendix. For the central region, commercial harvest was only in Canadian waters (OH-1, OH-2) and was reported as total biomass. No biological information was available fi'om the commercial harvest to estimate harvest in numbers. Thereupon, total numbers of lake trout harvested in the commercial yield each year were estimated by dividing the total annual yield by the mean mass of a recreationally harvested fish for the 38 corresponding year (Table 25, Appendix). In northern Lake Huron, there were both a tribal commercial fishery in U. S. waters and a commercial fishery in adjacent Canadian waters. Since commercial harvest was only available as total biomass, the commercial harvest portion of the objective function was expressed in terms of yield. The observed commercial harvest values used in the calibration process were scaled 20% higher than actual reported values because of suspected under-reporting of harvest by commercial fishers (M. Ebener, Chippewa- Ottawa Treaty Fishery Management Authority, pers. comm). Reported annual commercial harvest for northern Lake Huron are listed in Tables 25-26 of the Appendix. Due to the lack of sufficient sea lamprey wounding data for lake trout >533 mm in northern Lake Huron, sea lamprey-induced mortality rates for these sizes of fish were assumed to be at least equal to central Lake Huron rates. However, this assumption was likely to be conservative based on reports that sea lamprey abundance is highest in northern Lake Huron (Eshenroder at al. 1987). As an alternative, I estimated sea lamprey-induced mortality rates for lake trout >533 mm in northern Lake Huron by attempting to find the level of ZL that was consistent with harvest levels and age compositions in the surveys. I did this by estimating the parameter p’ in the objective function of the northern model. The parameter y’ was the proportionality coefficient for sea lamprey-induced mortality, which was defined in equation 15 of the methods section. This parameter scaled 2,, to allow the model predictions to match the age distribution of the survey index of abundance and the observed values for commercial harvest. This was done only for lake trout >533 mm since there were sufficient data for wounding rates for 39 the 432-533 mm length class. Thus, for the calibration of the northern Lake Huron model, the following term was added to the objective function in equation 27: L. = [2.1.2.144 1n(p.,.)]+[2.1“( 1,1 l] u, luHl l...ukl where p 2,, was the observed survey proportion-at-age a in year y and p0,, was the model’s predicted value for survey proportion-at-age, no was the fishery harvest for age a, j, was the sample size for the survey age composition, and ua was the survey CPUE for age a. Mgdel projgctions Three fishery management scenarios were run to evaluate the effects of decreasing the sea lamprey wounding rates on lake trout by model projections of abundance of ages 8 and older fish and total harvest from 1994-2010. In order to view the effects of the management scenarios in the mature portion of the population, ages 8+ were evaluated rather than total abundance of all ages. These projections were evaluated under three fishery conditions: 1) total allowable catch (TAC) with Z at the GLFC lake trout rehabilitation target maximum of 0.59 year"; 2) constant fishing mortality rate equal to average off, during 1991-1993; and 3) No fishing. The TAC plan is a management strategy that establishes harvest quotas based on estimates of mortality rates for all sources in relation to an established target maximum total mortality rate (e.g., A=0.4S, Z=0.59 year"). A quota will be possible only if natural and sea lamprey-induced mortality rates are below the established target maximum total mortality rate. 40 Age-specific natural mortality rates were assumed to be constant, as were the base sea lamprey-induced mortality and stocking numbers (recruitment), which were set equal to the average of current rates (1991-1993). Total abundance and harvest were projected for five levels of sea lamprey-induced mortality: current rates, 75%, 50%, 25%, and 0% of current rates. TAC was computed by estimating the maximum fishing intensity (fa...) that would match harvest and limit the instantaneous rate of total mortality to 0.59 year" for ages 5 and older. In northern and central Lake Huron, where there were both commercial and recreational fishing, the fishing intensity of the recreational fishery (fa, max) was estimated by the following: fgm = or fqm (28) where or was the ratio of fa, m to fc, m averaged from 1991-1993. For the north, or = 0.0095 and in the central area, or = 0.2526. RESULTS The results are reported in five subsections. First, I present the findings fi'om the analysis of sea lamprey wounding patterns on lake trout according to length of lake trout, geographic region, and year. Secondly, I discuss the ANOVA models constructed to estimate mean wounding rates for specific length class, geographic region, and year combinations where data were missing in central and southern Lake Huron. Furthermore, I report age- and year-specific rates of lamprey induced mortality for these regions. In the third part, I present results from statistical catch-at-age analysis of the southern lake trout population model. In part four, I describe results from the calibration of the northern and central Lake Huron models. Lastly, I report simulation results from the population models for northem, central, and southern Lake Huron. Panams in sea lamprey wounding ANOVA models constructed to assess patterns in wounding rates are listed in Table 5 and included as factors lake trout size, geographic region, and year. Significance levels for main effects and interactions for each model are listed in Table 6. In all models, there were significant interactions between year and geographic region, and year and length class. However, these year effects and their interactions do not seem to 41 42 Table 5. Levels for each factor in analysis of variance models used to evaluate patterns in sea lamprey wounding of lake trout in Michigan waters of Lake Huron. Factor Model Year Geographic RegionT Length Class 1 1984, 1985, 1986, 1987, central, south 432-533 mm, 534-635 mm, 1988, 1989, 1990, 1991, 636-737 mm, >737 mm 1992, 1993, 1994 2 1984, 1985, 1986, 1987, central 432-533 mm, 534-635 mm, 1988, 1989, 1990, 1991, 636-737 mm, >737 mm 1992, 1993, 1994 3 1984, 1985, 1986, 1987, south 432-533 mm, 534-635 mm, 1988, 1989, 1990, 1991, 636-737 mm, >737 mm 1992, 1993, 1994 4 1984, 1985, 1986, 1987, north, central , south 432-533 mm, 534-635 mm 1988, 1989, 1990, 1991, 1992, 1993, 1994 5 1984, 1985, 1986, 1987, central, south 534-635 mm, 636-737 mm, 1988, 1989, 1990, 1991, >737 mm 1992, 1993, 1994 6 1984, 1985, 1986, 1987, central, south 534-635 mm, 636-737 mm 1988, 1989, 1990, 1991, 1992, 1993, 1994 * North= MH-l, central= NIH-2, and south= MH-3/4/5. 43 Table 6. Significance levels (attained P-value) for main effects and interactions in analysis of variance models of sea lamprey wounding rates on lake trout in Michigan waters of Lake Huron, 1984-1994. Further information on data used with these models is given in Table 5. Main Effect Interaction Model Year (Y) Geographic Length Y x GR Y x LC GR x LC Region (GR) Class (LC) 1 0.0001 0.0148 0.0001 0.0042 0.0001 0.0527 2 0.0001 ------ 0.0001 ----- 0.0001 ------ 3 0.0001 ------ 0.0001 ----- 0.0001 ------ 4 0.0001 0.0169 0.0001 0.0001 0.0001 0.1484 5 0.0001 0.0397 0.0001 0.0132 0.0001 0.3142 6 0.0001 0.0001 0.0001 0.0093 0.0033 0.1316 44 reflect either overall or length class specific long-terrn trends. Analyses presented below suggest that the significant results were from short-tenn fluctuations in the true wounding rates from year to year. Wounding rates are presented as least-square means of square root transformed wounds per fish due to biases in back-transfonnation. However, the overall patterns in wounding rates were similar between transformed and untransformed wounding rates. Patterns in wounding according to length of lake trout In central and southern Lake Huron, wounding rates increased significantly with length class of lake trout (Table 6; Figure 5). The estimated wounding rates for the 63 6- 737 mm, and >737 mm length classes were not significantly different, possibly due to the low sample sizes for the largest length class. Northern Lake Huron (MI-I-l) was not included in this model because no fish of these sizes were collected in this region in most years. The ANOVA model for this analysis was designated as Model 1 (Table 5). Because near significant interaction between geographic region and length class was detected (Table 6), models 2 and 3 were constructed to test the effects of length class on wounding rates independent of geographic region. Model 2 contains only the central area (MH-2), and model 3 contains only the south (Table 5). For these additional models wounding rates increased significantly with length of lake trout (Table 6; Figure 6). 45 0.25 0.15 0.10 LSM of square root wounds per fish 0.05 0.00 432-533 mm 534-635 mm 636-737 mm > 737 mm Length Class Figure 5. Sea lamprey wounding patterns by length class of lake trout for central (MH- 2) and southern (MH-3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for length class, adjusted for all other effects and interactions, reported with one standard error. 46 0.40 or 1 a g 0.35 4 + g 0.30 — .5 .5 0.25 - I § §_ 0.15 - :1 0.10 - u- ” 0.05 - E E a, 0.00 . . . -| 432-533 mm 534—635 mm 636-737 mm >737 mm Length class '8 0.40 ~ b g 0.35 4 o g 0.30 4 + H 1" 8 E 0.25 ~ g z ,5 0.20 4 g 0.10 . E 0.05 - a) {—1—} A o_m I I I 1 432-533 mm 534-635 mm 636-737 mm >737 mm Length class Figure 6. Sea lamprey wounding of lake trout in Michigan waters of Lake Huron, 1984- 1994. (a) Central region (MB-2). (b) Southern region (MH-3/4/5). Least-square means (LSM) of square root transformed wounds per fish calculated fi'om analysis of variance with length class and year as treatment factors. Estimated means for length class, adjusted for all other effects and interactions, reported with one standard error. 47 Geographic patterns in wounding rates Analysis of wounding rates across all three geographic areas was only possible for the two smaller length classes of lake trout (432-533, and 534-635 mm) because few large lake trout were collected in the northern region for the 63 6-737 mm, and >737 mm length categories (see Table l). The smallest length class (432-533 mm) of lake trout had relatively low wounding rates (<0.08), and did not differ geographically, while differences in wounding rates were significantly higher in central than in southern Lake Huron for 534-635 mm lake trout (Model 4, Table 5; Figure 7). Wounding rates for the north did not differ significantly from the other two areas for the 432-533 mm length class (Figure 7). However, the results for the 534-635 mm length class in the north were biased. Further review of the 534-635 mm data revealed that most fish in this length category in northern Lake Huron were distributed towards the smaller size ranges, while in central and southern Lake Huron the observations were evenly distributed across all lengths. Consequently, the wounding rates for 534-635 mm lake trout in northern Lake Huron were not accurately represented. Thus, the only valid comparisons with northern Lake Huron were for the 432-533 mm lake trout, which did not differ geographically. Although I was not able to evaluate how sea lamprey-induced mortality rates (as indexed from wounding data) for lake trout >533 mm in northern Lake Huron compared with the other areas of the main basin, other sources of information indicated that sea lamprey abundance was highest in the north. One source of information was the observations of the number of sea lampreys attached to lake trout and chinook salmon 48 0'20 ‘ l432-533 mm I 121534-635 mm 0.18 - I 1 z 0.16 — 1 m c 8 0.14 - I 0- l {g r: 0.12 - 3 E 0.10- o 2 e 0.08 - N g. m 0.06 a E (I) -' 0.04 — 0.02 ~ 0.00 “ 1 l ”—‘—'1 North Central South (MH-1) (MH-2) (MH- 3I4l5) Geographic region Figure 7. Geographic patterns in sea lamprey wounding of lake trout less than 636 mm in Lake Huron, Michigan for 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for length class and geographic region, adjusted for all other effects and interactions, reported with one standard error. 49 (Oncorhynchus tshauytscha) caught aboard sport fishing charter boats (Rakoczy and Rogers 1991a, 1991b; Rakoczy 1992; Rakoczy and vaoda 1993, 1994b). In the main basin of Lake Huron fiom 1989-1993, the mean number of sea lampreys attached to both lake trout and chinook salmon were significantly higher in the north compared to the other regions (Figure 8). This implies that sea lamprey abundance and attack rates were highest in northern Lake Huron. Another data source that indicates that sea lamprey abundance was highest in northern waters were assessment catches of spawning phase and larval sea lampreys conducted by the Canadian Department of Fisheries and Oceans and the US. Fish and Wildlife Service. In the tributaries monitored in the main basin of Lake Huron, the highest catches of spawning phase were in the St. Mary’s, Cheboygan, and Ocqueoc Rivers which are located in northern waters. Likewise, assessment catches of sea lamprey larvae were also highest in northern waters of Lake Huron (J. Heinrich, Sea Lamprey Control, US. Fish and Wildlife Service, Marquette, MI, pers. comm). Lastly, Mormon (1979) reported that abundance of sea lamprey larvae were higher in the northern than in the southern regions of Lake Huron due to habitat preferences. Overall, there is sufficient evidence indicating that sea lamprey abundance is highest in the northern waters of Lake Huron, implying that sea lamprey-induced mortality is also likely to be highest in the north. For lake trout larger than 533 mm, wounding rates were significantly higher in the central area than in the south (Model 5, Table 5; Figure 9). Due to the predominance of extremely low sample sizes for the >737 mm length class in the central area (see Table 50 0.30 1 Lake trout I] Chinook salmon O N 01 l .0 N o l Mean number of sea lampreys attached per fish :3. 0.00 ~ 9 . . 4 North Central South Region Figure 8. Mean number of sea lampreys attached to lake trout and chinook salmon caught aboard sport fishing charter boats in Michigan waters of Lake Huron, 1989-1993. Data from Michigan Department of Natural Resources. Error bars represent two standard errors. 51 0.35 - 0.30 0.25 0.20 0.15 0.10 LSM of square root wounds per fish 0.05 0.00 Central Geographic region Figure 9. Geographic patterns in sea lamprey wounding of lake trout larger than 533 mm in central (MI-L2) and southern (MH-3/4/5) regions of Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for geographic region, adjusted for all other effects and interactions, reported with one standard error. 52 1), differences in wounding rates between central and southern Lake Huron were firrther evaluated using only the S34-635 mm and 63 6-737 mm length classes (Model 6, Table 5). For these length classes, wounding rates were found to be significantly higher in the central region than in the south (Figure 10). Temporal trends in wounding rates Overall, there were annual differences in wounding rates for lake trout in the central and southern regions of the main basin of Lake Huron. However, no obvious long-terrn temporal trends in wounding were observed from 1984-1994, although there seemed to be a cyclic pattern (model 1, Table 5; Figure 11). Peaks in wounding rates were observed in 1985, 1987, 1990, and 1993. These high wounding years were evident in lake trout >533 mm (Figure 12). Wounding rates were lowest in 1984. No temporal trends were evident in wounding rates for each of the length classes when the central and southern regions were combined (Model 1, Table 5; Figure 12), nor were there trends over time in these areas when all length classes were pooled (Model 1, Table 5; Figure 13). Again, northern Lake Huron was excluded from these analyses due to many years without data. Estimation of sea lamprey-induced mortality My objectives here were first to systematically estimate mean wounding rates for central and southern Lake Huron where data were not sufficient or absent with the least amount of extrapolation. The second objective was to compute age-specific sea F 1g] Cent Squa analy- E'Sllm r613011 53 0.30 - 0.25 - 0.20 — 0.15 ~ 0.10 - LSM square root wounds per fish 0.05 — Central (MH-2) South (MH-3/4/5) Geographic region Figure 10. Geographic patterns in sea lamprey wounding of 534-737 mm lake trout in central (MI-I-2) and southern (MH-3/4/5) regions of Lake Huron, 1984-1994. Least- square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for geographic region, adjusted for all other effects and interactions, reported with one standard error. l".- __ I Fig“, $0ch ‘001 l geOgr all 01h 54 0.35 - 0.30 - k 0.25 e 0.20 - 0.15 - 0.10 1 LSM square root wounds per fish 0.05 0.00 I I T I I I I I I I 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Year Figure 11. Sea lamprey wounding of lake trout 2 432 mm in central (MH-2) and southern (MH-3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for year, adjusted for all other effects and interactions, reported with one standard error. Fig 314 W0 and mill 55 0.60 - +432-533 mm - - 1: ~ - 534-635 mm +636-737 mm - at - >737 mm 0.50 ~ I ‘, .c 5 l 1.9:, ll ’ i -' \ I— ‘ '. \‘ l \ a 0040 A I ‘ : I 1 1 1 , I - 13 K I : \ c I \ 1 , . 3 ‘. 1 1 \ g 1 1‘ .' 1 ,- \ I '\ +- 0.30 4 -' ' ' 1 I ' : . o r \ . 1 \ I ‘ 2 ‘ ' I 'l \ .0 e .I ‘0 \I . . e .I .l. (U * - ~ m 0.20 " 5 a. _______ ;' ~ “ . , 11 2 0‘ ‘0 : . s - . v 7 (D ; ' ,' .1 . ‘. (110 " .,}’ m (100 .- . 1 . . T . . . . 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Year Figure 12. Sea lamprey wounding of lake trout in central (MH-2) and southern (NIH- 3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for length class and year, adjusted for all other effects and interactions, reported with one standard error. 56 0.45 - no - - MH-2 0,40 - —-— 1111-1-3/4/5 0.35 - 0.30 ~ 0.25 - 0.20 - 0.15 — LSM square root wounds per fish 0.10 -:' 0.05 0.00 l I I I I I I I I I I 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Year Figure 13. Sea lamprey wounding of lake trout 2432 mm in central (MH-2) and southern (MH-3/4/5) Lake Huron, 1984-1994. Least-square means (LSM) of square root transformed wounds per fish calculated from analysis of variance with length class, geographic region, and year as treatment factors. Estimated means for geographic region and year, adjusted for all other effects and interactions, reported with one standard error. 57 lamprey-induced mortality rates for the central and southern lake trout population models. The models constructed and the data points they predict are listed in Table 7. For the 432-533 mm length class, it was not possible to directly calculate mean wounds per fish for the central area in years 1990 and 1991, or for the southern area in 1990-1992 because sufficient data were lacking. Hence, mean wounds per fish were estimated for these locations based on data collected in other regions and years. Model A was constructed with year and geographic region as main factors by using the available data for all three geographic regions for the 43 2-533 mm length class with information from 1984-1994 (Table 7). Model B was constructed to predict wounding rates for the S34-635 mm length class in the central region for 1988 and 1990, and in the southern region for 1991 (Table 7). This was done using wounding rates from the other years in both years for the 534-635 mm length class. For lake trout in the 636-73 7, and >737 mm length classes, the only data that were available were for the southern region (see Table 1). Therefore, ANOVA models to predict wounding rates for these length classes in the central region were dependent on the observed differences in wounding rates among length classes in the south for estimating the length class effect, and geographic differences for fish <636 mm to estimate area effects. Wounding rates for 63 6-737 mm lake trout in central Lake Huron were estimated using the effects from lake trout >533 mm in the central and southern regions in all years (Model C, Table 7). For the >737 mm length class in the central region, there were no samples with 40 or more lake trout. Therefore, model D was constructed to project wounding rates 58 Table 7. Levels for each factor in analysis of variance models used to estimate mean wounds per fish when an insufficient number of observations (less than 40 lake trout) were available in Michigan waters of Lake Huron. MI~I-l= north, MH-2= central, and MH-3/4/5= south. Factor Model Year Geographic Length Class Data points estimated Region by model A 1984, 1985, MH-l, 432-533 mm [1990, MH-2, 432-533 mm] 1986, 1987, NIH-2, [1991, MH-2, 432-533 mm] 1988, 1989, MH-3/4/5 [1990, MH-3/4/5, 432-533 mm] 1990, 1991, [1991, MH-3/4/5, 432-533 mm] 1992, 1993, [1992, MI-l-3/4/5, 432-533 mm] 1994 B 1984, 1985, NIH-2, 534-635 mm [1988, MH-2, 534-635 mm] 1986, 1987, MI-I-3/4/5 [1990, MH-Z, 534-635 mm] 1988, 1989, [1991, MH-3/4/5, 534-635 mm] 1990, 1991, 1992, 1993, 1994 C 1984, 1985, MH-2, 534-635 mm, [1984, MH-2, 636-737 mm] 1986, 1987, MH-3/4/5 636-737 mm, [1985, NIH-2, 636-737 mm] 1988, 1989, >737 mm [1986, MH-2, 636-737 mm] 1990, 1991, [1987, MH-2, 636-737 mm] 1992, 1993, [1988, NIH-2, 636-737 mm] 1994 [1989, MH-2, 636-737 mm] [1990, NIH-2, 636-737 mm] [1993, MH-Z, 636-737 mm] D 1984, 1985, MI-1-2, 534-635 mm, [all years, MH-2, >737 mm] 1986, 1987, MH-3/4/5 >737 mm 1988, 1989, 1990, 1991, 1992, 1993, 1994 59 for this length class in relation to the 534-635 mm lake trout in the central area based on the differences in wounding rates between the 534-635 mm and the >737 mm length groups in the south (Table 7). These estimated wounding rates for >737 mm fish in the central area are unimportant in terms of model output since so few fish survive to these sizes. Never-the-less, in order to run the population model, wounding rates were needed to estimate sea lamprey-induced mortality for old lake trout; otherwise, the model could not be used to make projections for scenarios with lower mortality rates (and hence have large, older fish). For central and southern main basin of Lake Huron, mean wounds per fish for lake trout are listed by length class in Tables 8 and 9. For samples with more than 40 lake trout, raw mean wounds per fish were used, whereas mean wounds per fish were estimated by ANOVA models (Table 7) for strata in the database with observations with less than 40 fish. Age-specific lamprey-induced mortality rates, computed using equations 4 and 5, are listed in Tables 10 and 11. The only wounding data with sufficient sample sizes for northern Lake Huron were for the 432-533 mm fish and mean wounding rates ranged from 0.01 to 0.15 wounds per fish during 1984-1994. Due to the lack to of wounding data for lake trout >533 mm for northern Lake Huron, an alternative approach was used to estimate sea lamprey-induced mortality based on fitting the parameter ,u' as described in the methods section (see section titled Calibration of the northern and central lake trout population models in Methods). Estimates of sea lamprey-induced mortality for northern Lake Huron using this procedure are presented later in the results (see Calibration of the Tl H1 M 60 Table 8. Sea lamprey wounding rates by length class for lake trout in central Lake Huron (MB-2). Wounding rates expressed as mean wounds per fish. Data fi'om Michigan Department of Natural Resources spring surveys. Length Class Year 432-533 mm 534-635 mm 636-737 mm >737 mm 1984 0.00000 0.01639 0.14316' 0.19024‘ 1985 0.10194 0.41892 0.42141‘ 0.44562’ 1986 0.10497 0.18750 0.26890‘ 0.29141‘ 1987 0.03371 0.12195 0.32829‘ 0.34299‘ 1988 0.00000 0.11639' 0.27194‘ 0.29125‘ 1989 0.05691 0.25301 0.30010‘ 0.38313‘ 1990 0.01019‘ 0.25905' 0.42716‘ 0.43605‘ 1991 0.00000‘ 0.21212 0.18644 0.38853‘ 1992 0.02299 0.19753 0.36170 0.31612‘ 1993 0.08065 0.25000 0.38572‘ 0.41371‘ 1994 0.09836 0.24390 0.39535 0.39113‘ ' Estimated by analysis of variance model. 61 Table 9. Sea lamprey wounding rates by length class for lake trout in southern Lake Huron (MI-I-3/4/5). Wounding rates expressed as mean wounds per fish. Data from Michigan Department of Natural Resources spring surveys. Length Class Year 432-533 mm S34-635 mm 636-737 mm >737 mm 1984 0.00000 0.03226 0.05571 0.09756 1985 0.04724 0.22865 0.38444 0.39597 1986 0.02500 0.07170 0.24066 0.23077 1987 0.01695 0.14612 0.33636 0.58333 1988 0.01258 0.07882 0.24180 0.31429 1989 0.01587 0.17986 0.20601 0.28889 1990 0.00000‘ 0.22148 0.40569 0.42353 1991 0.00000’ 0.17455‘ 0.19388 0.38889 1992 0.03076‘ 0.18182 0.31111 0.21569 1993 0.13253 0.17857 0.36364 0.55882 1994 0.02000 0.17241 0.38686 0.41538 ‘ Estimated by analysis of variance model. 62 Table 10. Estimated instantaneous rates of sea lamprey-induced mortality (year") for lake trout in central Lake Huron (MB-2) during 1984-1993. Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 O 0 3 0.203 0.196 0.066 0.006 0.114 0.032 0.011 0.051 0.157 0.188 4 0.236 0.200 0.076 0.021 0.136 0.063 0.038 0.072 0.173 0.200 5 0.367 0.214 0.121 0.088 0.221 0.195 0.145 0.162 0.242 0.252 6 0.418 0.223 0.196 0.168 0.269 0.311 0.198 0.252 0.300 0.302 7 0.391 0.222 0.218 0.183 0.261 0.319 0.186 0.261 0.303 0.306 8 0.368 0.225 0.241 0.201 0.265 0.328 0.202 0.264 0.311 0.311 9 0.353 0.228 0.274 0.229 0.274 0.353 0.222 0.280 0.325 0.322 10 0.357 0.232 0.276 0.232 0.288 0.354 0.256 0.273 0.330 0.321 11 0.351 0.226 0.273 0.228 0.268 0.352 0.208 0.284 0.323 0.322 12 0.360 0.234 0.278 0.234 0.297 0.355 0.277 0.268 0.333 0.321 13 0.355 0.229 0.275 0.230 0.280 0.353 0.235 0.277 0.327 0.322 14 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 15 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 16 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 17 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 18 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 19 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 20 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 >20 0.365 0.238 0.281 0.238 0.314 0.357 0.318 0.259 0.339 0.320 63 Table 11. Estimated instantaneous rates of sea lamprey-induced mortality (year") for lake trout in southern Lake Huron (MH-3/4/5) during 1984-1993. Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 O 0 2 0 0 0 0 0 0 0 0 0 0 3 0.105 0.050 0.045 0.030 0.047 0.114 0.113 0.072 0.244 0.053 4 0.141 0.059 0.073 0.044 0.082 0.146 0.133 0.103 0.239 0.086 5 0.247 0.099 0.166 0.100 0.172 0.247 0.181 0.194 0.238 0.194 6 0.299 0.159 0.242 0.163 0.184 0.308 0.177 0.240 0.273 0.278 7 0.309 0.184 0.282 0.192 0.178 0.325 0.178 0.242 0.305 0.305 8 0.316 0.190 0.339 0.214 0.193 0.334 0.214 0.227 0.348 0.321 9 0.319 0.191 0.388 0.230 0.207 0.339 0.249 0.210 0.387 0.328 10 0.320 0.190 0.415 0.238 0.216 0.341 0.271 0.199 0.408 0.331 11 0.322 0.189 0.438 0.245 0.224 0.343 0.288 0.191 0.426 0.334 12 0.323 0.189 0.462 0.253 0.231 0.345 0.306 0.182 0.445 0.338 13 0.323 0.190 0.457 0.251 0.230 0.345 0.302 0.184 0.441 0.338 14 0.322 0.190 0.445 0.248 0.226 0.344 0.293 0.189 0.432 0.336 15 0.323 0.189 0.463 0.253 0.232 0.345 0.307 0.182 0.446 0.338 16 0.322 0.190 0.437 0.245 0.223 0.344 0.286 0.192 0.425 0.335 17 0.324 0.189 0.477 0.257 0.236 0.347 0.318 0.176 0.457 0.340 18 0.324 0.189 0.477 0.257 0.236 0.347 0.318 0.176 0.457 0.340 19 0.324 0.189 0.477 0.257 0.236 0.347 0.318 0.176 0.457 0.340 20 0.324 0.189 0.477 0.257 0.236 0.347 0.318 0.176 0.457 0.340 >20 0.324 0.189 0.477 0.257 0.236 0.347 0.318 0.176 0.457 0.340 64 northern and central lake trout population models in Results). Patterns in estimated sea lamprey-induced mortality were directly related to patterns in wounding rates. In general, sea lamprey-induced mortality increased with length of lake trout, and tended to be higher in the central regions than in the south. ' i .. t h- - 1: . 1 i of h o hm Lk H ron lak trout oulation model Parameters values for the southern model estimated by CAA analyses and corresponding log-likelihood components are listed in Tables 12 and 13. Several versions of the catch-at-age analysis were run based on restrictions set to particular parameters that were thought to heavily influence the calibration process. For example, the proportionality coefficient for sea lamprey-induced mortality (11’) and natural mortality proportionality coefficient (0) were either fixed as 1 or estimated by CAA analysis. In preliminary analyses, survey selectivity was fixed with values that followed an asymptotic relationship to length. This reduced the number of parameters estimated. However, for these preliminary analyses, harvest was consistently either underpredicted or overpredicted. The total log-likelihood (L) for these analyses, which ranged from - 270.48 to -299.20, indicated a poorer fit than subsequent CAA analyses. In addition, trends were observed in both predicted fishery and survey age compositions. Thus, survey selectivity values were estimated as parameters in all ensuing analyses. In CAA], parameters ,u’ and c were fixed at 1. This was designated as the baseline CAA model since this implies that I have correctly defined the relationship between sea lamprey-induced mortality and wounding data and also have correctly 65 Table 12. Estimated parameter values fiom catch-at-age analyses of the southern Lake Huron lake trout population model, 1984-1993. Recreational fishery parameters: qR= catchability (angler hours"), SR, .= selectivity at age a, and fa, ,= fishing intensity (year"). 11’: proportionality coefficient for sea lamprey-induced mortality. Research survey parameters: q*= catchability (meters of gill net"), S*.= selectivity at age a. Population parameters: N.,.9u= abundance at age a in 1984, c= proportionality coefficient for natural mortality, Mr= age-1 instantaneous natural mortality (year"), and t= rate of decrease in natural mortality rate (year" age"). ”= parameter not estimated by catch-at- age analysis. Catch-at-age model: Parameters CAAl CAA2 CAA3 CAA4 CAAS Fishery qR 1.82081 x1007 1.35120 x1007 1.32557 x1007 1.66248 x10“ 1.56302 x1007 s... 0" 0“ 0" 0" 0“ S9,2 0.000064 0.000033 0.000044 0.000022 0.000043 s...3 0.023185 0.048057 0.030697 0.040342 0.033187 3... 0.247836 0.494797 0.307270 0.421826 0.331398 5..., 0.683335 1.348836 0.833276 1.162254 0.911426 s... 0.731383 1.336467 0.880659 1.166437 0.992864 s... 0.751762 1.204770 0.854484 1.071924 0.975735 s... 0.998337 1.124553 0.996947 1.014570 0.998220 SR9. 1" 1" l” 1" 1” f. .9... 0.142549 0.091657 0.096037 0.117867 0.109834 f1L1985 0.148104 0.105439 0.103115 0.133211 0.122809 r... .9... 0.216524 0.156968 0.155994 0.196289 0.183475 1*... 9.. 0.178980 0.119329 0.129193 0.147928 0.149416 fa 1988 0.174794 0.119498 0.131549 0.144713 0.146812 fa. 1999 0.124317 0.090034 0.092790 0.109261 0.105413 13., .99. 0.208954 0.230527 0.172152 0.268877 0.214931 8., .99. 0.113067 0.084671 0.080215 0.104765 0.097112 f3, .99. 0.115043 0.075162 0.076347 0.095023 0.091895 {9, .99. 0.115995 0.035917 0.049671 0.049123 0.060918 Lamprey 11’ 1" 0.048894 1” 0.000349 1” 66 Table 12 (cont’d). Catch-at-age model: Panunemus C1041 CJUAZ (DKAS CMUM4 CMUAS Survey q* 0.001134 0.001051 0.000709 0.001225 0.000947 8*. 0“ 0" 0" 0“ 0“ 8*. 0.024729 0.026910 0.030937 0.025261 0.026284 8*. 0.180430 0.200711 0.211788 0.192967 0.208198 8*. 0.507441 0.523947 0.550043 0.512957 0.541355 8*. 1" 1" 1" 1” 1“ 8*. 0.990297 0.952700 0.996646 0.958498 1.011189 8*. 0.943763 0.813802 1.000256 0.830414 1.000370 8*. 1.206543 0.906150 1.116072 0.945295 1.122270 8*. 1.449042 0.890466 1.236359 0.958673 1.181630 8*... 2.625431 0.982813 1.968542 1.136801 1.679869 Population N219“ 426861.982 381721.967 488286.153 358963.125 434857.682 N3.1984 139172.149 112566.865 165328.627 103814.264 127660.333 N4.1984 140175.860 114038.703 174114.800 103423.934 134857.190 N61939: 119144.898 92941.040 151666.465 82863.397 117584.339 N6 1934 101295.205 81900.967 130255.278 72053946 98613.080 141.1934 42293.469 38300960 54906096 33300.385 40918.423 N8.1984 25281.895 25219.011 35039.038 21187.430 26166.310 N9. 1994 14216.299 18774.859 22715.026 15290.472 18064.430 N.9,.9.. 349.826 5026.792 989.877 2918.870 985.216 N...,.9.. 3669.679“ 3669.679“ 3669.679“ 3669.679“ 3669.679“ c 1“ 1“ 0.676613 1.114583 " M. “ t “ “ 0.666290 1' it l4 # # 1.115309 67 Table 13. Maximum loge-likelihood components from statistical catch-at-age analyses of the southern Lake Huron lake trout population model, 1984-1993. Catch-at-age model: Likelihood Component CAAl CAA2 CAA3 CAA4 CAAS Fishery harvest (L1) -4.4457 -3.9535 -2.6376 -3.3275 -2.7533 Survey CPUE (L2) -1.8072 -1.4892 -1.3296 -1.4883 -1.3362 Fishery age 7.0330 11.7392 7.5129 11.9857 9.3428 composition (L3) Survey age -169.4730 -159.7915 -162.7454 -160.4057 -162. 1344 composition (L4) Fishery effort (L5) 2.8447 1.3453 3.1637 1.5008 2.7207 Total (L= 2L.) -l65.8482 -152.1497 -156.0359 -151.7351 -154.1604 SL Cl. th pa th be Va the hill cor CA that; info Sign 68 assigned the level of natural mortality from other sources. The parameters estimated for this analysis are listed in Table 12 and log.- likelihood components are listed in Table 13. Predicted harvest was consistently below observed harvest (Figure 14a). A decreasing trend in residuals for survey total CPUE was observed (Figure 14b). Predicted total survey CPUEs were higher than observed values in 1984-1987, while they were lower in most of the later years. This analysis was based on a stringent model that assumed the current, baseline values for natural mortality (see Table 2) and sea lamprey-induced mortality were correct. However, the consistent underprediction of harvest indicates that either natural or lamprey-induced mortality was overestimated by this set of parameter values. Subsequent CAA analyses were structured to assess which one of these sources of mortality (natural or lamprey-induced) was set too high. In CAA2, c was fixed to 1, while ,u’ was estimated. This analysis produced a better model fit as indicated by the matching of model predicted harvest with observed values, and by the total log-likelihood (L) which was maximized to -152.15, and higher than the value of -l65.85 for CAA] (Table 13). I did not detect patterns in fishery harvest residuals, survey CPUE residuals, or in residuals for fishery or survey age compositions. However, p’ was estimated to be 0.0489, and if we accept the results of CAA2, the lethality of sea lamprey attacks on lake trout would be significantly lower than previously thought. I concluded that this was unrealistic based on other sources of information indicating that lethality of attacks and mortality caused by sea lamprey are significant for lake trout populations in the Great Lakes. 69 0.7 - 0.6 - 0.5 l 0.4 4 0.3 ~ 0.2 — 5.9.31: 0.1 a I Li}; 5337:}; 0.0 " 1 I A 1 I 1 1 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Ln- based residuals 9.0.0.0 O-‘NOJ-h lllJ Ln-based residuals 83.5.6.5 Atom-a 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Figure 14. Log-based residuals from catch-at-age analysis CAA] of the southern Lake Huron lake trout population model. (a) fishery harvest. (b) survey total CPUE. 70 For example, Bergstedt and Schneider (1988) compared the wounding rates on live lake trout captured in assessment gill nets to recovered dead lake trout using bottom trawls in Lake Ontario and found that nearly all (99%) of the carcasses had recent sea lamprey wounds, whereas the live fish had much lower wounding rates. They concluded that sea lamprey attack was the primary cause of death of the lake trout carcasses they collected and natural mortality other than that cause by sea lampreys was insignificant. Similar results were reported by Schneider et al. (in press) which was based on the continuation of Bergstedt and Schneider’s (1988) study. Furthermore, laboratory studies evaluating the lethality of attacks on lake trout from sea lampreys indicate that approximately 50% of attacks result in death of the host (Swink and Hanson 1989; Swink 1990). CAA 3 was used to evaluate whether adjustment of natural mortality could produce an adequate model. Parameter c was estimated while ,u’ was fixed to 1 (Table 12). The total log.-likelihood value converged at -156.04 (Table 13). There were no trends in fishery harvest or survey total CPUE residuals (Figure 15). Likewise, no patterns in residuals were observed for fishery or survey age compositions (Figures 16, 17). Parameter c was estimated to be 0.6766, indicating that natural mortality was 67.7% of baseline rates. Parameters m’ and c were both estimated in CAA4 (Table 12). The total log.- likelihood was -151.74. Since this model had an additional parameter estimated, it is not surprising that the total log-likelihood value was maximized at a value higher than the other catch-at-age analyses (Table 13). Again, no trends in residuals were observed. 71 Ln-based residuals 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Ln-based residuals 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Figure 15. Log9-based residuals from catch-at-age analysis CAA3 of the southern Lake Huron lake trout population model. (a) fishery harvest. (b) survey total CPUE. F i g 1’62 age 72 4 a x :11 .2 3 a I 3 2 x A , 01985 '5 o e 1 ‘ I X ‘ I 1986 o I E 0 i .1. '1 a: 1 1 .1987 E -1 ° x . x1988 e - - § .2 , * ‘ , 11: 311991 in _3 e 1992 .4 3 4 5 6 7 8 9+ Age 4 b ‘ .‘L’ 3 g o e 3 '5 2 . x 0 + "7, . ‘ I 4 e 1 :1: ¥ 0 X A 5 'U 0 i 21 t L J g g b 1: 0 * I x 6 E -1 I : X 7 CU ¥ lit '2 '2 ‘ A . g e 8 g '3 ‘ + 9+ _4 _ 1985 1986 1987 1988 1991 1992 Year Figure 16. Standardized residuals for fishery age composition from CAA3. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. 73 8 a m 6 Q .1984 .6 + I1985 : 1 “ r - - ‘13:: o 2 _ _ x g ‘ t + - )9: x1988 a 0 ’f ' g i I 8 ’ 9 .1989 .. _ - I g _2 1' _ . f o i f +1990 .5 _ , -1991 g .4— —1992 m .5- .1993 -8- 2 3 4 5 6 7 8 9 10+ Age 8 1 b in 6 J x as :3 , + ’2 g 4 g ' : - "' I3 3 2 ; = . ‘ . ; : A4 a 0 f .1 g 8 I i . 4 r..- "5 E 2 3‘ I ‘ t 3 ‘ I I *6 a - . - X I7 + g -4 -1 - +8 (I) -6 ~ '9 -10+ -8 1 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Figure 17. Standardized residuals for survey age composition from CAA3. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. 74 The parameter p’ was estimated to be 0.0003, 0.03% of baseline rates, while parameter c was estimated to be 1.1146. Although the results of CAA4 indicated a relatively good fit, other evidence indicates that the estimated value for ,u ’ was unrealistic (see results for CAA2) and sea lamprey-induced mortality is not trivial as these results would seem to indicate. CAAS estimated natural mortality using the second approach of fitting a type 3 survivorship function. Parameters estimated by CAAS are listed in Table 12. The total log.-likelihood was maximized to ~154. 16. No patterns in residuals were observed for fishery harvest or survey CPUE (Figure 18). Likewise, no trends were observed in the residuals for fishery age composition (Figure 19) or survey age composition (Figure 20). The instantaneous rate of natural mortality for age-1 lake trout (M1) was estimated to be 0.6663 year'1 and twas estimated to be 1.115 age'l year". In order to test whether a CAA model had a significantly better fit than the baseline model (CAAl), a likelihood ratio test was used (Seber and Wild 1989). Significant difi‘erence in total log-likelihoods was tested against the Chi—square distribution using the likelihood ratio test statistic: 2[L(é) - L(E)o )] , where L(é) was the total log-likelihood for a CAA analysis with either p ', c, both p' and c, or both M, and r estimated, while L(90) was total log-likelihood for the baseline CAA model (CAAI). Degrees of freedom were equal to the number of parameters (i.e., p ', c, M], t) estimated in L(é) minus the number of parameters estimated by L(9°). All CAA models in which the parameters [1 ’, c, M; , or 2' were estimated had significantly higher total log- 75 0.5 - 0.4 a 0.3 - 0.2 - 0.1 - 0.0 - -0.1 - -0.2 - -0.3 - -0.4 - Ln-based residuals -O.5 - 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Ln-base'd residuals 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Figure 18. Log.-based residuals from catch-at-age analysis CAAS of the southern Lake Huron lake trout population model. (a) fishery harvest. (b) survey total CPUE. its“. " 1 I l1.‘ Ill‘ Fig1 Yea: age 76 4.. m a x * § 3‘ 1985 O '0 . '5 2 " i o .1986 2 1 ' . I x ‘ 5 1 .1987 E _1 o x . x1988 I X .g _2 ' ' x1991 : a i x 5 _3 .1992 a) .4 3 4 5 6 7 8 9+ Age 4 m g 3 b ‘ 03 2 2 . 0 l4 1» x o 2 1 x : 0 § .5 u x 3 o a i i ,3, 4 x6 .. 4 x g I ’ x x :7 '2 -2 ‘ . o 08 g -3 +9+ .4 1985 1 986 1 987 1 988 1 991 1 992 Year Figure 19. Standardized residuals for fishery age composition from CAAS. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. F i g 1962 age 77 8 a . .1984 % 6 .1985 g 4 ‘ + _ _ A1986 on 2 x1987 g 8 . - i + ‘ — : x1988 3 0 3f '1 g _ g i '9 .1989 § -2 1 ; a _ 3 i +1990 g -4 q .. e -1991 5 -1992 (D -6 -1 61993 -3 9 2 3 4 5 6 7 8 9 10+ Age 8 ‘ b .3 6 - X 02 3 «I» I3 '0 4 ~ 0 - Ia . - I - ‘4 e 2 i g ‘ . : >2 X5 '3 0 s ,1 i fi i i 1 E“.- 32 x6 N g 3“ I 1 O | E _2 2‘ - - +- t g g A 07 g 4 _ I + +8 = ' -9 g -6 9 —10+ -3 J 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Figure 20. Standardized residuals for survey age composition from CAAS. (a) across years. (b) across ages. Standardized residuals= observed minus predicted proportions at age divided by estimated standard deviation. Sea Wei €311 78 likelihood values than the baseline CAA model (CAA2, P<0.00001; CAA3, P<0.0001; CAA4, P<0.00001; CAAS, P<0.0001). Furthermore, the estimation of natural mortality by the parameters M; and r in CAAS seemed to fit better than the estimation of c in CAA3 (P<0.054), although not significant at the conventional a=0.05 level. Based on the results from the likelihood ratio test and review of the parameters estimated by the various analyses; CAAS was considered to be the best model. In models CAA2 and CAA4, the estimates of ,u ’ did not realistically reflect the lethality of sea lamprey attacks (see results for CAA2). In retrospect, it appears that there was not enough contrast in wounding rates during 1984-1993 in southern Lake Huron to adequately estimate p’ (see Figures 11-13). Even a very large change in wounding rates to unrealistic levels produced little change in model fit. Moreover, CAA3 had a poorer fit than model CAAS. Based on the parameters estimated by CAAS, predicted values of southern Lake Huron fishery harvest, effort, and survey CPUE matched the observed values well. The parameters estimated were based on the assumption that each type of observed data used in the calibration process was reliable (i.e., fishery harvest, age composition, effort, and survey CPUE and age composition). This was evaluated by measuring the sensitivity of the model to each of the data sources (see below). Sensitivity of the southern model to calibration data Figures 21-25 illustrate changes in log-likelihood values according to various weightings (1;) that changed how much data source i was emphasized in the fit using catch-at-age model CAAS (see Methods). Positive changes in log-likelihood indicated Figu Varyi L5: 1 79 2 1 1.5 ~ 11 .. g 0.5 "9'“L1 +L2 .: :73 k ab L3 g 0 # ~13: —x—L4 é +L5 .I <1 -05 — ° “1"“- '1‘1 O." .J -1.5 ~ -2 6 0.1 0.5 1 5 7&1 Figure 21. Changes in log-likelihood components for catch-at-age model fit due to varying emphasis of fishery harvest data (M). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery effort, L= total. 80 0.35 1 0.25 - 0.15 . '8 "*"L1 3 0.05 +9- % A\ : +L3 é 5' i i —x—L4 3: -o.05 - +'-5 .J \ I. <1 +P ‘ . -o.15 - .. O -0.25 1 -o.35 ~ 0.1 0.5 1 5 7&2 Figure 22. Changes in log-likelihood components for catch-at-age model fit due to varying emphasis of survey CPUE data (M). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery efi‘ort, L= total. 81 30 — 20 - 1o « 1, --.--L1 8 —a—L2 % +L3 i 0 human—fl ............ % +L4 '2 ‘ +L5 21' , —o—L -10 _ -2o — ° -30 _l 0.1 0.5 1 5 7‘8 Figure 23. Changes in loge-likelihood components for catch-at-age model fit due to varying emphasis of fishery age composition data 0.3). Likelihood components: Ll= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery effort, L= total. 82 70- 601 501 40- 301 20“ .....L1 +L3 0 —:——— .———-——-— ....... 4| +L4 -10 s o +L5 +L A Ln-likelihood -30 _ -40 1 -50 s -60 - ., -70 z 0.1 0.5 1 5 914 Figure 24. Changes in logc-likelihood components for catch-at-age model fit due to varying emphasis of survey age composition data (7L4). Likelihood components: L1= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery effort, L= total. 83 8- 6.. 4- u 2 "*"U 8 —a—L2 % +1.3 5 0* ---------- -— —x—L4 E 3 ......... +L5 <1 -2— a L 0.1 0.5 1 5 A5 Figure 25. Changes in logc-likelihood components for catch-at-age model fit due to varying emphasis of fishery effort data (1.5). Likelihood components: Ll= fishery harvest, L2= survey CPUE, L3= fishery age composition, L4= survey age composition, L5= fishery efi‘ort, L= total. Co Cr 84 improvements in model fit for particular likelihood components, whereas negative values denoted worse fit. The lake trout population model was relatively insensitive to reducing or increasing the emphasis of 1.1, the emphasis factor for fishery harvest data (Figure 21). The total log-likelihood (L) did not decrease more than one unit. Similarly, altering 1»; (the emphasis factor for survey CPUE data) did not result in notable changes in overall model fit (Figure 22). However, down-weighting of 9.3 (the emphasis factor for fishery age composition data) yielded large decreases in L and L 3 (likelihood component for fishery age composition) and large increases in L4 (likelihood component for survey age composition). This indicates that model fit was strongly influenced by fishery age composition information (Figure 23). The greatest change in L resulted from the de- emphasis of 2.4 (Figure 24). Model fit was highly sensitive to survey age composition data. Reduced emphasis of 1.5 resulted in higher likelihood values for L 3 and L4 (Figure 25). Fishery effort information is usually the most questionable source of data in fishery models (Hilbom and Walters 1992). Since the fishery effort information used in CAA was based on efi‘ort targeted at all salmonines (e. g., Oncorhynchus tshawytscha, 0. kisutch, and 0. mykiss), trends in lake trout CPUE may be biased. This may be due to difi‘erences in habitat preferences or angler targeting of lake trout and other salmonines. Thus, another catch-at-age analysis was performed to explore the fit of the MH-3/4/S model without the use of any efi‘ort information (CAA6). Since there is one less component in this model without effort data, it was not directly comparable to model CAAS using the total loge-likelihoods (L). However, one can compare the individual 85 likelihood components common to both models. The parameter values estimated by the two models were similar (Table 14; see Tables 12, 13). Predicted harvest based on parameters estimated by CAA6 (L1= -5.110188) did not match observed values as well as those of CAA5 (L1 = -2.753300). The other likelihood component values for CAA6 were: L2=-1.329959, L3=10.552618, and L4=-158. 107641. The age-specific mortality rates averaged from 1984-1993 were similar between CAAS and CAA6 (Figure 26). Total mortality was slightly higher for CAA6, which is primarily due to higher recreational fishing mortality rates. Based on these results, omission of fishery effort data did not significantly improve model fit to other data sources or dramatically alter estimated mortality rates. Testing the model’s sensitivity to each data source revealed that survey and fishery age composition information were important in determining the set of parameters for optimum fit. Changing the emphasis of survey age composition data contributed the largest fluctuations in the total log-likelihood. This indicated that model predictions of lake trout abundance were heavily influenced by survey data. The research survey data were collected in a systematic and consistent manner, and were considered the most reliable data source. Since virtually all lake trout collected in surveys had fin clips, aging errors were insignificant because each cohort had a distinguishing fin clip pattern. Furthermore, identical fin clip patterns between cohorts were validated by scale analysis of age (J. Johnson, Alpena Fisheries Research Station, MDNR pers. comm). Fishery age composition data also strongly influenced model fit. However, fishery data were considered less reliable in comparison to research survey data. 86 Table 14. Estimated parameter values from catch-at-age analysis model CAA6. Recreational fishery parameters: qR= catchability (angler hours"), SK .= selectivity at age a, and fa, y= fishing intensity (year"). u’= proportionality coefficient for sea lamprey- induced mortality. Research survey parameters: q"'= catchability (meters of gill net"), S*.= selectivity at age a. Population parameters: N...934= abundance at age a in 1984, M.= age-1 instantaneous natural mortality (year"), and t= rate of decrease in natural mortality rate (year'l age'l). 1= parameter not estimated by catch-at-age analysis. Sea Lamprey and Fishery Survey Population Parameters Value Parameters Value Parameters Value q.. 1.56302 x104” q* 0.001030 .1: 1’ s.L . 0" s*. 0" N; .9... 403756.464 8... 0.000043 8*2 0.025174 N11... 118621.889 8..~ 3 0.031078 8*. 0.202199 N4, .934 126903.823 s... 0.310250 8*. 0.530940 N5, .984 110246.361 s... 5 0.887269 8*; 1" N... .9... 92275.104 8.... 0.993080 S*6 1.011411 N71911: 38053.854 8;, 1 0.975843 8*. 1.000369 N8, .944 23906748 s.L .. 0.998185 8*. 1.163516 N9. 19.. 16092.736 s“. 1“ 8*. 1.230265 1.1.0.9.... 985.310 f... .9... 0.117562 8*... 1.781492 N. .._ .9... 3669.679“ 111.1985 0.189955 M. 0.707333 r.~ .9... 0.197934 1 1.116077 fa. 1937 0.132127 11. .9... 0.086388 fn. 1989 0.098662 r... .99., 0.333537 f3, .99. 0.122551 1“., .99. 0.168678 f.~ .99. 0.072327 87 0.8 —o— MoCAAB + FIR-CAA6 + zL-CAAG + Z-CAA6 nan-M-CAAS --O--FR-CAA5--+--ZL-CAA5 ------ z-CAAs 0.7 - 0.5 - 1 . .-"'- 0.4 ~ ‘O 0.3 ~ ., ‘ \ Instantaneous Mortality Rate 0.2 3 0.1 1 Figure 26. Age-specific instantaneous mortality rates (year") for lake trout in southern Lake Huron as estimated by statistical catch-at-age analysis models CAAS and CAA6. Mortality rates averaged from 1984-1993. M= natural mortality, FR= recreational fishing mortality, ZL= sea lamprey-induced mortality, and Z= total mortality. 88 Specifically, fishery age composition data were collected in a less rigorous manner and were subject to biases associated with angler behavior. Fishery harvest and age composition data were not available for all years, and were not collected in all months for each year. In addition, age composition of fishery harvest were derived from subsamples, which may be biased due to an inconsistent sampling regime. As indicated in the methods section, some of these measurement errors were accounted for by limiting maximum sample size in a particular year to 200 fish in the log-likelihood equation for fishery age composition data (L3). Based on the considerations discussed above, model predictions of mortality rates were evaluated by de-emphasizing fishery age composition data (L3). When A; was set to 0.1, age-specific total mortality rates were lower than when 1»; was set at 1 (Figure 27). This was primarily due to reductions in natural mortality for ages 1-4 and reductions in fishing mortality for ages 5 and older. However, the proportion of lake trout killed in southern Lake Huron by sea lamprey and fishing averaged from 1984- 1993 remained roughly the same with 1.3 =0.] and 1»; =1. When 1.3 =1, fishing accounted for 2.8% of the deaths on average, while sea lamprey parasitism killed 5.9%. For 71.3 =0.1, fishing removed 2.1% of the population and sea lampreys killed 7.2% of lake trout. Average total annual abundance of lake trout in southern Lake Huron from 1984-1993 was estimated to be 1.2 million fish per year for 2;. =1 and 1.6 million fish per year for 7L3= 0.1. Overall, de-emphasizing fishery age composition data did not qualitatively change model predictions. Presumably, this is because predicted fishery age composition poorly matched the observed data--and it is those data and the estimates of their 89 0.7 l + M, 13:17.1 + F3, 139-0.1 + ZL, 13:0.1 " + Z, 13:0.1 ‘ ' * " ' M, 13:1 ’ ' . ' ' FR, 1.3:] i ..+.. ZL, 16:1 ...... LA3=1 0.6 1 30.5 « ‘1 -* "r. “ ‘ . . 0: . ,..--- P .b 1 Instantangous Mortal I- 02- '. i‘ . ..-‘."--.----O-“On-.0”... . 0-1‘ ~.-.,, I a I u o" . I H—. .0" 0 ' , . . , , . 1 ‘ 1 2 3 4 5 6 7 8 9 10 11 12+ Age Figure 27. Differences in estimated age-specific instantaneous mortality rates (year'l) with emphasis factor for fishery age composition data (2;) set at 0.1 and 1. Mortality rates averaged from 1984-1993. M=natural mortality, FR= recreational fishing mortality, ZL= sea lamprey-induced mortality, and Z= total mortality. 90 reliability that is questioned. Based on the evaluations of model sensitivity to data sources, changing the emphasis factors did not significantly alter qualitative patterns and usually did not alter quantitative estimates by large amounts. As a result of these analyses, the emphasis factors for each data source were maintained at l. Uncertainty in estimated abundance In order to evaluate the uncertainty in model estimates of abundance, the confidence bounds of parameter estimates must be determined. However, for multi- dimensional and highly non-linear problems such as the case in this study where there were 38 parameters estimated, conventional methods are often not robust (Seber and Wild 1989). Therefore, I used a one-dimensional approach aimed at a critical parameter linked to population abundance, namely recreational fishing intensity in 1993 (fig .993). I found the values (confidence bounds) of this parameter that had 5% of the total likelihood below the lower bound and had 5% of the total likelihood above the upper bound (Hilbom and Walters 1992). I calculated this 90% confidence interval using a likelihood ratio test (see Statistical catch-at-age analysis of the southern Lake Huron lake trout population model in Results section). I then evaluated the corresponding abundance values for 1993 at the limits of this confidence interval and took this as approximate confidence bounds for abundance for that year. For 1993, these bounds for abundance of ages 3+ lake trout from the southern model were 20% below and 24% above the estimated value. Thus, the model’s estimate of the mean abundance of ages 91 3+ lake trout in 1993 was 377,000 fish with a 90% confidence interval of 301,000 to 467,000 fish. This confidence interval probably underestimates uncertainty since it is conditional on the values of quantities such as sea lamprey-induced mortality, which were assumed known. Calibratipn gf the northern and central lake trout population models Year-specific commercial fishing intensities and recreational fishery catchability coefiicients for the northern and central regions estimated by the calibration procedure are listed in Table 27 of the Appendix. The central area model was successfully calibrated with the objective function minimized to match predicted commercial harvest to observed values (scaled 20% higher to account for under-reporting). The northern area model was successfirlly calibrated to both survey age composition and commercial harvest (adjusted for under-reporting). The parameter ,u’ in the northern model was estimated to be 4.06 (Table 14) indicating that sea lamprey-induced mortality rates for lake trout >533 mm were substantially underestimated using the wounding rates from central Lake Huron. Sea lamprey-induced mortality rates for lake trout in northern Lake Huron are in Table 44 of the Appendix. Model output Southern Lake Huron WH-3/4/5), 1984-1993 Based on the results of statistical catch-at-age analysis of the southern Lake Huron population model, the estimated mean annual abundance of lake trout from 1984- 92 1993 was 1.1 million (Table 28, Appendix). Mean annual abundance for mature lake trout (ages 8+) was estimated at about 70,000 fish. Total annual abundance was estimated to be lower during 1990-1993 than 1984-1989. This was due to lower stocking rates in 1987, 1988, and 1990 (Table 18, Appendix). Estimated mortality rates were relatively constant during this time period. On average, sea lamprey-induced mortality was estimated to be higher than all other sources of mortality (Figure 27, A3= 1; Table 11; also see Tables 29-30 in Appendix). For lake trout ages most selected by sea lampreys and recreational fishing (ages 3-10), it was estimated that 43% of lake trout deaths were caused by sea lamprey parasitism, recreational fishing accounted for 21% of the deaths, while natural mortality killed 36% (Figure 28). Estimates of annual deaths due to each mortality source for each age are listed in Tables 31-33 of the Appendix. Central Lake Huron (1101-2), 1984-1993 During 1984-1993, estimated mean annual abundance of lake trout in region MH-2 was about 385,000 (Table 34, Appendix). Mean abundance estimated for ages 8+ was approximately 26,000 fish. An increasing trend in total abundance can be attributed to higher stocking rates over time (Table 18, Appendix). Overall recruitment in central Lake Huron, as indexed by age-1 abundance, was lower than southern Lake Huron. This was due to lower stocking rates and the high emigration rate (60%) from central to northern Lake Huron assumed in the population models. This was reflected in the lower mean annual abundance estimated in the central region as compared to the south. Sea lamprey-induced mortality was overwhelmingly the dominant source of lake trout death 93 I Sea Lamprey El Commercial Fishing I Recreational Fishing 55— 50* 45- 40* 35- 301 25- . 20- Percent of deaths MH-1 (84-93) MH-1 (91-93) MH-z MH-3/4/5 Region Figure 28. Allocation of estimated lake trout deaths (ages 3-10) in the main basin of Lake Huron from 1984-1993. MH-1= north, MH-2= central, and MI-I-3/4/5= south. 94 in central Lake Huron (Figure 29; Table 10; also see Tables 29, 35-36 in Appendix). In contrast, commercial and recreational fishing mortality were minor. In relation to numbers of ages 3-10 lake trout killed in the central area from 1984-1993, sea lamprey parasitism was estimated to account for more than half of all deaths (Figure 28). Recreational fishing accounted for 2%, commercial fishing accounted for 7%, and natural mortality 39% of ages 3-10 lake trout deaths. Estimates of total deaths by year and age are in Tables 37-40 of the Appendix. Northern Lake Huron (MH-I), 1984-1993 Estimated abundance of lake trout in northern Lake Huron averaged 1.4 million fish per year from 1984-1993 (Table 41, Appendix). However, estimated mean abundance of mature lake trout (ages 8+) was about 3,000 fish per year. Total lake trout abundance was estimated to be highest in the north compared with the rest of the main basin of Lake Huron, and was dominated by immature fish. This was due to the higher stocking rates in the north and the high immigration from central Lake Huron (Table 18, Appendix). An increasing trend in estimated total annual abundance was observed over the time series. This trend reflects recruitment as indicated by age-1 abundance (Tables 41, 18, Appendix). Mortality rates changed dramatically fiom 1984-1994. Commercial fishing mortality for ages 3-10 lake trout was the highest source of death during 1987-1989, whereas sea lamprey-induced mortality was the dominant source during 1984-1985 and 1991-1993 (Figure 30; Tables 29, 42-44, Appendix). For ages 4-7 lake trout during 95 Total + Sea Lamprey 0.55 — +Comm Fishing - - o - - Rec Fishing 0.60 — 0.50 1 0.45 5 0.40 -1 0.35 4 0.30 - 0.25 - Instantaneous Mortality Rate 0.20 — 0.15 4 0.10 4 0.05 . -.- .. ..- C. . .C. ... 0.00 'f“4.£ll. r r I r F 1 I 1 12 34 56 7 8 9101112+ Age Figure 29. Age-specific estimates of instantaneous mortality rates (year'l) for lake trout in central Lake Huron. Mortality rates averaged from 1984-1993. 96 . ‘ —0—Sea Lamprey 4.0 - : --I~Comm. Fishing 3.8 ~ -' '1. +1266. Fishing Instantaneous mortality rate N 0| 0.01—e 4 =0 4—4—4—r 6 I 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Year Figure 30. Temporal patterns in estimated instantaneous mortality rates (year'l) averaged for ages 3-10 lake trout in northern Lake Huron. 97 1987-1988, estimated instantaneous mortality rates due to commercial fishing ranged from 3.81 to 9.15 year". Lake trout are not a target species in the commercial fishery, and are harvested as bycatch in the lake Whitefish (Coregonus clupeaformis) large-mesh gill net fishery (M. Ebener, Chippewa-Ottawa Treaty Fishery Management Authority, pers. com). The high commercial fishing mortality rates estimated correspond to the highest levels of tribal gill net effort for lake Whitefish during 1984-1993 (Table 26, Appendix). From 1991-1993, when mortality rates were relatively constant, the dominant source of mortality for lake trout in northern Lake Huron was due to sea lampreys (Figure 31; Tables 29, 42-44, Appendix). Commercial fishing was also a significant source of lake trout mortality starting at age-4. In contrast, recreational fishing was an insignificant source of mortality for lake trout in the north. Although total mortality was estimated to be extremely high for the older lake trout, there were very few fish older than age-8 in the population, because most fish were killed at earlier ages (Table 41, Appendix). In terms of the average number of ages 3-10 lake trout killed in the northern region, fi'om 1984-1993 commercial fishing caused 54%, sea lamprey parasitism 30%, recreational fishing less than 1%, and natural mortality 16% of deaths (Figure 28; Tables 4548, Appendix). However, during the most recent period (1991-1993), sea lampreys caused 44% of deaths for ages 3-10 lake trout, while commercial fishing accounted for 33% (Figure 28). 98 1.6 1.5 4 1.4 - —Total 13 - +Sea Lamprey +Comm. Fishing 12 1 +Rec. Fishing 1.11 1.0 ~ .0 co 1 0.8 5 0.7 ~ 0.6 - Instantaneous Mortality Rate 0.5 - 0.4 ~ 0.3 - 0.2 J 0.1 . 0.0 . 4 - . 4. 4- :. . . . . 1 2 3 4 5 6 7 8 9 10 11 12+ e 1» Figure 31. Estimates of age-specific instantaneous mortality rates (year") for lake trout in northern Lake Huron. Mortality rates averaged from 1991-1993. 99 Total mortality rates in Lake Huron In southern Lake Huron, estimated instantaneous rates of total mortality (Z) for lake trout ages 5 and older were above the GLFC lake trout rehabilitation target maximum of 0.59 year" during 1987, 1990, and 1993(Table 49, Appendix). Overall, total mortality rates in southern Lake Huron were below the lake trout rehabilitation target. For central Lake Huron, estimates of Z were below the GLFC target maximum in all years from 1984-1993 (Table 50, Appendix). The total mortality rates estimated for lake trout ages 5+ in northern Lake Huron exceeded the rehabilitation target maximum in all years from 1984-1993 (Table 51, Appendix). Model prgjegtigns Southern Lake Huron (MH-3/4/5) Scenario 1: Total Allowable Catch (TAC) with maximum Z=0.59 year’l Under the TAC scenario, abundance of lake trout ages 8 and older in southern Lake Huron is projected to decrease 56% by the year 2010 if sea lamprey-induced mortality was equal to current estimated rates (Figure 323). If sea lamprey-induced mortality was eliminated, total abundance of lake trout ages 8+ is projected to still decrease 54% by the year 2010. TAC is projected to increase 194% by the year 2010 under current conditions and is projected to increase 783% if sea lamprey-induced mortality was reduced to 0 (Figure 32b). These results were based under the assumption that fishing mortality could be increased to make total mortality equal to the target rate of 0.59 year'l. 100 100 90 _ + Current 70 4 +0.50 60 3 +0.25 50 ~ +0 40 * 30 ~ 20 1 10 * 0fr[IrrrrrrrrrrrrrrTrjrrrr 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Number x1000 100 90 — b 80 - 70 4 60 ~ 50 - 40 - 30 . Number x1000 20 1 + current + 0.75 10 - +0.50 +0.25 +0 0 T Tfifrj I I I I I I I I I I T I I I I I T Iijfi I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 32. Model estimates of lake trout (a) abundance for ages 8+, and (b) total harvest under a total allowable catch (TAC) management scenario in southern Lake Huron from 1984-2010. Maximum total instantaneous mortality for projections was 0.59 year". Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (ZL): Current= average ZL for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. 1 0 1 Scenario 2: Current fishing mortality rate Iffishing mortality remained constant during the projection period (equal to average of 1991-1993), total abundance of ages 8+ lake trout is projected to decrease 29% under current sea lamprey-induced mortality rates (Figure 338). However, if 2,, was reduced to 0, abundance would increase 318% by 2010 (Figure 33a). Under this management regime, harvest would increase 66% by the year 2010 with current sea lamprey-induced mortality rates and would increase 353% if ZL=0 (Figure 33b). Scenario 3: No fishing Under this scenario, total abundance of lake trout older than age-7 are projected to increase 7% under current sea lamprey-induced mortality rates and to increase 678% by the year 2010 if ZL was 0 (Figure 34). This management option provides the highest projected spawner population increase under current stocking, natural mortality, and sea lamprey-induced mortality rates. Central Lake Huron (NIH-2) Scenario 1: Total Allowable Catch (TAC) with maximum Z=0.59 year'l Total abundance of ages 8+ lake trout in the central region is projected to decrease 15% by 2010 under this management plan with current sea lamprey-induced mortality rates (Figure 35). If sea lamprey-induced mortality was reduced to zero, ages 8+ abundance is projected to increase 51% by 2010. This is because of the differential age-selectivity of sea lamprey-induced and fishing mortality rates. Thus, it is more 102 350 a +Current 8300 +0.75 § 250 7 +0.50 5200 _ +0.25 I- 3 150 - +0 51001 . M 2 50_ g=,M 0 I I I I I I I I r I I I j I I I I I l I I l I I I l 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year 60 b +current 50- +0.75 3 +0.50 +0.25 9- 40‘ +0 x 8 3‘" M /W :E, 20~ £5...” 2 10~ OIrijriiTithIIrrrirrTiIIII 1984 1987 1990 1993 19Y96 1999 2002 2005 2008 ear Figure 33. Model estimates of lake trout (a) abundance for ages 8+ and, (b) total harvest under a constant fishing mortality management scenario in southern Lake Huron from 1984-2010. Fishing mortality rates for projections were based on the average of 1991- 1993 rates. Projections (1994-2010) were according to varying levels of sea lamprey- induced mortality (21,): Current= average 2., for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. 103 700 600 ~ +Current 500 8 +0.75 8 +0.50 5:: 400 ‘ +0.25 E's +0 8 a 300 "‘ Z 200 — I 100 - / > //\’—\ 4 0 l I I I T I l I l I r‘rfi I I l I l I I I I I I I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 34. Model estimates of ages 8+ lake trout abundance in southern Lake Huron from 1984-2010. Projections were based on a no fishing management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average 2., for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 21, is 0. 104 40 351 30 - o 25 T O O X EZO- , E 3 z 15 ~ +current 10* +0.75 +0.5 +0.25 +0 5- 0 I I 1 r I r r I I rfi r I r 1 T I i I 1 1 1 r T T 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 35. Model estimates of ages 8+ lake trout abundance under a total allowable catch (TAC) management scenario in central Lake Huron from 1984-2010. Maximum total instantaneous mortality for projections was 0.59 year". Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (21,): Current= average Z], for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2., is 0. 105 beneficial to allocate the maximum mortality rate to fishing than to sea lampreys because fishing tends to target a smaller range of ages than sea lampreys. If 2,, was equal to current rates, commercial harvest is projected to increase 157% under the TAC plan (Figure 36a). IfZL was zero, TAC is projected to increase 500% by 2010. Similar increases in projected recreational harvest were observed (Figure 36b). Scenario 2: Current fishing mortality rate Under current fishing and sea lamprey-induced mortality levels, total abundance of ages 8+ lake trout is projected to increase 50% by the year 2010 (Figure 37). If ZL was reduced to zero, abundance of ages 8+ in central Lake Huron is projected to increase 924% by 2010. Commercial harvest of lake trout is projected to increase 27% with current sea lamprey conditions, and to increase 134% when ZL was zero (Figure 38a). Recreational harvest had a higher level of projected increase than commercial harvest. Under current sea lamprey-induced mortality rates, recreational harvest is projected to increase 49% by 2010 (Figure 38b). If ZL was reduced to zero, projected harvest increases 357% by 2010. Scenario 3: No fishing Under this management plan, ages 8+ lake trout abundance is projected to increase 124% by the year 2010 given current sea lamprey-induced mortality rates (Figure 39). Total abundance is projected to increase 1,578% if Z1, was zero. Compared to the other two plans, zero fishing would allow for maximum spawner population indu 0.50. mom 106 60 55 * a o 50— +current g 45- ,; £4 +0.75 3 30_ +0.5 :5: 5(5)“ +0.25 3 154 +0 2 1o- 5.. O I I I I I I I fi r r V I I T a I I I I I I I I I I I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year 1% ii“ '° ° 7 +c rrent 8 101 +0u75 .- 94 - x 8* +0.55 8 7- n 6_ +0.25 E 51 +0 :1 4* 2 3* 2 .1 0IlirirrfirtrrlIfrtrrrrITrfiI 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 36. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in central Lake Huron from 1984-2010. Projections were based on a total allowable catch (TAC) management scenario according to varying levels of sea lamprey- induced mortality (ZL): Current= average Z], for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. Maximum total instantaneous mortality for projections was 0.59 year". 107 275 250* 225* 200) +current +0.75 +0.5 631757 +025 o 2 +0 x150“ t- 3 5125* 5 100* 75— 50* 25‘ .‘=‘ 0IrrrfiiIrrrrrtirtrritirrrl 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 37. Model estimates of ages 8+ lake trout abundance under a constant fishing mortality management scenario in central Lake Huron from 1984-2010. Fishing mortality rates for projections were based on the average of 1991-1993 rates. Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (ZL): Current= average ZL for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. + current + 0.75 Number x1000 8 r r r r 1 r r r r r I r r r r r I r r r r r 1 r 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year + current +0.75 +0.5 +025 M 3 ‘ /‘ M +0 5 0 1111irrrrrrrrrrrrrrrrrrrrn 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Number x1000 A Figure 38. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in central Lake Huron from 1984-2010. Projections were based on a constant fishing mortality management scenario according to varying levels of sea lamprey- induced mortality (ZL): Current= average 2., for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. Fishing mortality rates for projections were based on the average of 1991-1993 rates. 109 450 425 * 400 1 375 * 350 * 325 * 300 _ +current +0.75 8 275 ‘ +0.5 55 250 - +0.25 '- 225 1 +0 E 200 - 5 175 - 150 - 125 4 100 _. 75 — 50 - 25 J n; \\\ Il\\ 0 fiFIIIIIIIIrrlrrrrrrfIIIIfi 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 39. Model estimates of ages 8+ lake trout abundance in central Lake Huron from 1984-2010. Projections were based on a zero fishing management scenario according to varying levels of sea lamprey-induced mortality (ZL): Current= average ZL for 1991- 1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. l 10 regeneration in central Lake Huron. Northern Lake Huron (MH—I) Scenario 1: Total Allowable Catch (TAC) with maximum Z=0.59 year" Following the TAC management plan, total abundance of ages 8+ lake trout in northern Lake Huron is projected to increase 10,784% by the year 2010 (Figure 40). However, no harvest would be allowed since sea lamprey-induced and natural mortality rates exceeded the target maximum rate (Figure 41). This enormous increase in ages 8+ abundance in the projections was due to low fishing mortality rates in comparison with the extremely high rates during 1987-1989. This high fishing mortality period essentially eliminated fish that would be ages 8+ (see Tables 41-44, Appendix). Moreover, under the TAC plan, no harvest was allowed until sea lamprey-induced mortality was reduced to 25% of current rates. The highest increase in ages 8+ lake trout abundance (52,976%) is projected to occur if sea lamprey-induced mortality was reduced to zero (Figure 40). Furthermore, when sea lamprey-induced mortality was reduced to 25% of current rates, the projected increase in ages 8+ abundance was less than when ZL was reduced only by 50%. This lower increase in abundance was due to the increase in fishing mortality to scale total mortality to the target of 0.59 year". Scenario 2: Current fishing mortality rate Under current (1991-1993) fishing mortality rates, ages 8+ abundance in northern Lake Huron is projected to increase under all levels of sea lamprey-induced mortality 111 150 140 4 130 _ 120 * +CUIT6I'II 110 3 +0.75 100 « +95 +0.25 90 _ +0 80* 70* 60* Number x1000 50— 40- 30- y 10 /\ / \ 0 ..\ I I I I I I I I I I I I I I I I I I I I T I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 40. Model estimates of ages 8+ lake trout abundance under a total allowable catch (TAC) management scenario in northern Lake Huron from 1984-2010. Maximum total instantaneous mortality for projections was 0.59 year". Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (ZL): Current= average Z; for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. 112 2004 a §175~ :150‘ 5125‘ +current 3100a +0.75 E 75 ~ +0.5 5 50_ +0.25 254 +0 0 I I l I l l l j I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year 1.3 1.2 - b 1.1 - a 1— 2 09‘ X 0'8 “ +current g. 0.7 '* o 0.6+ +0.75 "5’ 0.5- +0.5 2 . 1 0.2— +0 0.1 - 0 , I I , . , . . 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 41. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in northern Lake Huron from 1984-2010. Projections were based on a total allowable catch (TAC) management scenario according to varying levels of sea lamprey- induced mortality (21,): Current= average ZL for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= 2;, is 0. Maximum total instantaneous mortality for projections was 0.59 year". 113 (Figure 42). Under current ZL, projected abundance of ages 8+ lake trout would increase 2,885% by 2010, and increase 86,305% if 21, was zero. This high increase was due to the current fishing mortality rates being significantly lower than the mortality rates during 1987-1989, which in turn allowed for the resurgence of older fish in the projection period even with similar sea lamprey-induced mortality rates. During 1989-1993, there were very few fish older than age-8 in the population (Tables 22, 41, Appendix). Natural mortality was estimated to be highest for ages 1-3, commercial fishing mortality impacted the population at age-3 and was most selective for ages 4-6, while sea lampreys started to impact lake trout at age-5 and increased with age. When fishing mortality rates fiom 1987-1989 were used instead of 1991-1993 rates in this scenario, ages 8+ abundance is projected to decrease by 99.9% or more by the year 2010 under all levels of sea lamprey-induced mortality. Increases in projected commercial harvest by the year 2010 ranged from 26% under current sea lamprey conditions to 135% increase when ZL was zero (Figure 43a). Similarly, recreational harvest is also projected to increase, although in higher proportions (Figure 43b). Recreational harvest is projected to increase 67% by the year 2010 under current sea lamprey-induced mortality rates and by 418% if ZL was zero. Scenario 3: No fishing The maximum increase in ages 8+ abundance in northern Lake Huron is projected to occur under this strict management plan (Figure 44). With current sea lamprey- induced mortality rates, abundance of ages 8+ lake trout is projected to exceed 30,900 114 250 225 - 200 - 175 . +current +0.75 +0.5 8150 1 +0.25 2 +0 x 3125 - .0 E 3100 . 75 — 50 4 , -/W o/XfifTrr‘fvarifitTfifiivll 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 42. Model estimates of ages 8+ lake trout abundance under a constant fishing mortality management scenario in northern Lake Huron from 1984-2010. Fishing mortality rates for projections were based on the average of 1991-1993 rates. Projections (1994-2010) were according to varying levels of sea lamprey-induced mortality (ZL): Current= average ZL for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= ZL is 0. 115 + current = +0.75 2 50 4 +0.5 _ + 0.25 25 + 0 0 IIIIjrrlrrlrIIFfrfirjrrrr 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year 1.6 _ +current ° ' +0.75 0,0 I r 1 I l l l I l r r l 1 l 1 r 1 l I r I r r l I I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 43. Model estimates of lake trout (a) commercial harvest, and (b) recreational harvest in northern Lake Huron from 1984-2010. Projections were based on a constant fishing mortality management scenario according to varying levels of sea lamprey- induced mortality (ZL): Current= average 2;, for 1991-1993; 0.75= 7 5% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z1, is 0. Fishing mortality rates for projections were based on the average of 1991-1993 rates. 116 1300 1200 - 1100 ~ 1000 I + current +0.75 +0.5 + 0.25 800 _‘ +0 900- 700— 600 1 Number x1000 500 - 400 + 300 d 200 -* 10°“ W {/W /. N .— 0 I if I7 I I T I I I - I I I 1 I I I I I I j I I I 1984 1987 1990 1993 1996 1999 2002 2005 2008 Year Figure 44. Model estimates of ages 8+ lake trout abundance in northern Lake Huron from 1984-2010. Projections were based on a zero fishing management scenario according to varying levels of sea lamprey-induced mortality (21,): Current= average ZL for 1991-1993; 0.75= 75% of current; 0.50= 50% of current; 0.25= 25% of current; 0.0= Z}, is 0. 117 fish by 2010, an increase of greater than 10,700%. If sea lamprey-induced mortality was reduced to zero, ages 8+ lake trout would increase more than 470,000% or reach an abundance of 1.3 million fish. Mortflig trade-ofi‘: sea lamprey-induced vsr. fishing mortw Under 1991-1993 sea lamprey and fishing conditions in the main basin of Lake Huron, decreases in sea lamprey-induced mortality yield a larger increase in projected ages 8+ abundance than equivalent percentage decreases in fishing mortality. However, this was not true over the entire period for northern Lake Huron. Assuming current sea lamprey-induced mortality rates and the much higher fishing mortality rates from 1987- 1989, decreases in fishing mortality are projected to yield greater gains in ages 8+ lake trout abundance than equivalent decreases in sea lamprey-induced mortality (Figure 45). 118 28 25 * 23 * —o—Fishing Mortality 20 _ - a - Sea Lamprey- lnduced Mortality 18* 15* Ln (% change in ages 8+ abundance) l 0 25 50 75 100 % decline in mortality rate Figure 45. Change in projected abundance of ages 8+ lake trout in the year 2010 due to decreases in fishing and sea lamprey-induced mortality rates for northern Lake Huron. Fishing mortality was based on the average of 1987-1989 rates, sea lamprey-induced mortality was based on the average of 1991-1993 rates. DISCUSSION The primary goal of lake trout rehabilitation in Lake Huron is to re-establish self- sustaining populations that are capable of supporting harvest (DesJardine et al. 1995). In addition, the rehabilitation plan states that total annual mortality rates should not exceed 45% to facilitate the achievement of the primary goal. Although the rehabilitation efforts in Lake Huron have been ongoing since the late 19605, progress has been limited by the fact that lake trout populations in the main basin are still totally dependent on hatchery stockings. The lack of significant natural recruitment may be due to spawning habitat deficiencies, poor spawning site homing ability, poor genetic fitness of hatchery lake trout, insufficient spawning stock biomass, or a combination of these factors. The failure of lake trout to re—establish self-sustaining populations is likely due to several of these factors, however, high mortality rates have played an important role in limiting population growth, especially for populations that were starting from near extinction levels, as in the case with lake trout in the main basin of Lake Huron. This study examined the effects of fishing and sea lamprey parasitism on lake trout abundance and showed that temporal variations and age-selectivity of these mortality sources have greatly affected population growth in Lake Huron, particularly in northern waters. 119 120 Rola af saa lampreys in la_ke trout rehabilitation Based on the evaluations of the patterns in sea lamprey wounding rates on lake trout, my results indicated that sea lampreys target larger lake trout in Lake Huron, and thus inflict higher mortality rates on older fish. This finding has been previously documented, but has not been reported for the main basin of Lake Huron. Although I did not detect any overall temporal trends in wounding rates fi'om 1984-1993, there did appear to be a cyclic pattern. This cyclic phenomenon may be related to variations in sea lamprey year-class strength associated with treatment of streams and rivers with chemical toxicants by sea lamprey control programs. I also detected a geographic gradient in sea lamprey-induced mortality rates with the highest rates in northern Lake Huron. This was based on the results from ANOVA models that compared the rates between central and southern Lake Huron, and the calibration of the northern population model which estimated sea lamprey-induced mortality rates much higher than the other regions. These sea lamprey-induced mortality rates were based on the assumption that the laboratory values for the probability of survival from a sea lamprey attack reported by Swink (1990) were realistic values. An attempt was made to evaluate these probabilities using statistical catch-at-age analysis, but due to the lack of sufficient contrast in wounding rates in the time series, no conclusions could be made as to the accuracy of these values. Future research should focus on validating these survival probabilities in natural systems. Overall, the analyses of the patterns in wounding rates showed that sea lamprey- induced mortality rates were not constant across age, time, or geographic area. The implications for lake trout rehabilitation are that lake trout p0pulation growth is highly 12 1 dependent on sea lamprey dynamics. The high mortality rates caused by sea lamprey parasitism in northern Lake Huron was one of the most influential factors in inhibiting lake trout population increase. Survival and abundance of lake trout rin 1984-1993 Lake trout total mortality rates were lower in the central region of the main basin of Lake Huron than in the other regions during 1984-1993. Total instantaneous mortality in the central region was below the rehabilitation target of 0.59 year'l (A=0.45). In southern Lake Huron, total mortality was higher than the central region mostly due to higher fishing mortality rates. Similar to central Lake Huron, total mortality rates in the south were usually below the target maximum mortality rate during 1984-1993. In northern Lake Huron, total mortality has exceeded the target rate in every year with instantaneous rates reaching values up to 9.5 year'l. During the late 19808, high commercial fishing mortality, combined with high levels of sea lamprey parasitism caused the age structure of the population to be truncated with virtually no fish older than age-8 from 1988 to the present. These mortality rates do not provide promise for lake trout re-establishment, particularly for a population that is recovering fiom virtual extinction. Abundance of mature lake trout, an index of potential natural recruitment, was highest in southern Lake Huron and lowest in northern Lake Huron. There was approximately a twenty-fold difference in mean abundance of ages 8+ lake trout between the two regions during 1984-1993. This was not due to differential stocking rates, but 122 can be attributed to the lower sea lamprey-induced and fishing mortality rates in the south. The lack of commercial exploitation has contributed in allowing the high abundance of mature lake trout in southern Lake Huron. There were eight times as many mature lake trout in central Lake Huron than in the north. Even with 60% immigration from central Lake Huron, abundance of ages 8+ lake trout in the north only averaged about 3,200 fish during 1984-1993. Such low spawning stock biomass probably explains why there has been no natural recruitment in northern Lake Huron. Similarly, low spawning stock biomass in central Lake Huron, which is likely due to the high emigration (60%) to northern Lake Huron, is also precluding natural recruitment, while lack of sufficient suitable spawning substrate is also an important factor. Although there are reports of some natural recruitment in central Lake Huron in Thunder Bay (Johnson and VanAmberg 1995) and on the mid-lake Six Fathom-Yankee Reef complex (C. Bowen, II, National Biological Service, pers. comm), these observations were localized and are probably not contributing significantly to the regional population at this time. Despite the high numbers of mature lake trout in southern Lake Huron (annual mean of approximately 70,000 fish), lack of suitable spawning habitat has probably reduced the likelihood for natural reproduction (Hansen 1994; Eshenroder et al. 1995). Ironically, spawning habitat has been reported to be abundant in northern Lake Huron (Eshenroder et al. 1995), but the low abundance of mature fish there due to high mortality rates has diminished the potential for natural recruitment. This is despite immigration of lake trout from central Lake Huron. Unless mortality rates are reduced in 123 northern Lake Huron, rehabilitation will not be achieved under current conditions. Maaagement trade-off: fishing vs. sea lampray—induced mortality Northern Lake Huron Progress towards lake trout rehabilitation, as indicated by changes in spawner abundance (ages 8 and older), was evaluated through a series of trade-off analyses between the management of fishing and sea lamprey-induced mortality. In northern Lake Huron (MH-l), there has been concern about the high influx of parasitic phase sea lampreys from the St. Mary’s River, and mortality caused by the tribal gill net fishery. Under 1991-1993 fishing and sea lamprey-induced mortality rates, there is the potential for increase in mature lake trout abundance. However, the amount of increase may not produce sufficient spawning stock biomass to allow natural recruitment. Currently, there is no quantitative reference to what spawning stock biomass must be for natural recruitment, which is the first step towards self-sustainability. Commercial fishing mortality has fluctuated temporally and drastically affected the age structure of the population in concert with sea lamprey-induced mortality. For example, fishing intensities during 1987-1989, which were the highest in the time series, resulted in a highly truncated age structure with very few fish in the population older than age-8. If fishing mortality were allowed to reach those high rates again, spawning stock biomass will decrease. Under current sea lamprey-induced mortality rates, model simulations indicated that the maximum abundance of ages 8+ lake trout would be 9,500 fish under current fishing mortality rates, and 31,000 fish under a zero fishing scenario. 124 Given the large spatial area of northern Lake Huron and the ongoing high mortality rates due to sea lampreys, these results suggest that re-establishment of self-sustaining lake trout populations is unlikely until sea lamprey abundance is reduced. Although optimal levels of fishing and sea lamprey control depend upon economic costs of reducing mortality due to each source, the trade-off analysis suggests that a percentage drop in sea lamprey-induced mortality produces more mature lake trout than a similar decrease in fishing mortality. However, it is imperative that both sea lamprey-induced and commercial fishing mortality be managed closely so that total mortality rates do not reach the levels comparable to 1987-1989. Ifmortality rates are to remain high in northern Lake Huron, the only way to increase the abundance of mature lake trout would be to significantly increase hatchery stockings. This is not a wise option since it is financially costly and does not account for possible depensatory responses from sea lampreys and fishing. Central Lake Huron Results from trade-ofi‘ analyses indicated that reductions in sea lamprey-induced mortality would produce a higher increase in mature lake trout abundance than equivalent reductions in fishing mortality. Overall, fishery exploitation has been low on this population when compared to sea lamprey-induced mortality and to the situation in northern Lake Huron. Under current conditions, there is promise for population grth in central Lake Huron. Simulation results indicated that total abundance of mature lake trout would increase 50% by the year 2010 with current fishing and sea lamprey-induced 125 mortality rates. Iffishing mortality was to be regulated, the TAC management plan with a target of A=0.45 would not be a logical choice. Under current conditions, simulations indicate that adoption of the TAC plan would result in a 15.4% decrease in ages 8+ abundance by the year 2010. No increase in ages 8+ abundance would be observed unless sea lamprey-induced mortality was reduced to 50% of current rates. The TAC management strategy does not seem appropriate for populations that are recovering from extinction levels and are not self-sustaining. Maintaining current mortality levels in central Lake Huron will lead to an increase in abundance of mature lake trout. Higher stocking rates would accelerate this increase, however the issue of successfirl spawning still needs to be investigated. The ongoing research at the mid-lake Six Fathom Bank-Yankee Reef complex (C. Bowen, II, National Biological Service, pers. comm.) may provide a quantitative measure for the potential for natural recruitment in central Lake Huron. There has been low levels of natural recruitment detected on this reef complex. In addition, there are indications that mortality rates are lower in this region than in other parts of central Lake Huron and that certain genetic strains of lake trout suffer lower sea lamprey wounding rates (C. Bowen, II, National Biological Service, pers. comm). Southern Lake Huron Current fishing and sea lamprey-induced mortality rates are at levels that do not allow increases in mature lake trout abundance. Under current conditions, abundance of ages 8+ lake trout are projected to decrease 29%. Sea lamprey-induced mortality 126 accounts for most of the lake trout deaths in southern Lake Huron. Therefore, similar to central Lake Huron, adoption of a TAC management strategy would inhibit the increase in the numbers of mature lake trout in southern Lake Huron. In fact, under current sea lamprey wounding rates, the TAC plan would decrease ages 8+ abundance approximately 50% by the year 2010. Based on model simulation results, more emphasis should be placed on reducing sea lamprey-induced mortality than reducing fishing mortality. However, this assumes that recreational fishing mortality remains constant at current rates and does not take into account the relative economic costs to control each source of mortality. Status and potentials of lake trout reha:bilitation Ifsuflicient suitable spawning sites are available and sufficient numbers of hatchery lake trout are being stocked, significant progress towards lake trout rehabilitation can occur as exhibited by lake trout populations in Lake Superior (Hansen 1994). The results of this study partly answers why the goals of lake trout rehabilitation have not been attained in the main basin of Lake Huron. In northern Lake Huron, commercial fishing and sea lamprey-induced mortality rates were too high to allow sufficient accumulation of mature fish, despite sufficient spawning habitat. Although mortality rates were not excessive in central Lake Huron, low population size due to high emigration to the north, and moderate levels of sea lamprey parasitism, as well as lack of sufficient spawning habitat are factors that have precluded the existence of a self- sustaining population in this region. In southern Lake Huron, sea lamprey-induced 127 mortality has reduced the rate of population growth. Although abundance of mature lake trout is highest in this region of the main basin, lack of natural recnritment is likely due to insufficient spawning habitat. However, the failure may be also be partly due to insufficient spawning stock biomass. In order to rehabilitate lake trout in the main basin of Lake Huron, mortality rates must be effectively reduced and managed. This means that sea lamprey control must be increased and the commercial gill net fishery must be managed. Current research on the St. Mary’s River, a major source of sea lampreys in the main basin, indicates that localized application of larnpricides in areas where ammocoetes are highly concentrated may be highly efiicacious (Lake Huron Technical Committee, Great Lakes Fishery Commission, pers. com). This strategy is currently being pursued by the Great Lakes Fishery Commission. Commercial fishing mortality on lake trout must be reduced for rehabilitation to proceed. The high lake trout harvest is a result of incidental harvest in the lake whitefish gill net fishery. A promising management strategy is to convert the lake Whitefish fishery gear from gill nets to trap nets. In comparison to gill nets, trap nets have been reported to dramatically reduce capture and mortality of non-target species such as lake trout (Schorflraar and Peck 1993). Further research on gear conversion fi'om gill to trap nets should be pursued with emphasis on the social, economic, and biological impacts. Stocking of hatchery-raised lake trout should continue as a management tool to increase population size. Stocked lake trout have contributed significantly to the successful re-establishment of populations in Lake Superior (Hansen 1994). However, 128 this tool can only be effective if total mortality rates are reduced and effectively managed in northern Lake Huron, where there is high potential for natural recruitment. Furthermore, criteria, based on quantitative analyses, must be established as to when stocking should cease. Results fi'om the ongoing genetic research on the differential fitness of various lake trout strains should also be applied to the stocking program. Lake trout mortality rates in Lake Huron appear to vary over time and depend upon age. The statistical catch-at-age method used here allowed me to estimate these rates without the acceptance of unrealistic assumptions. In contrast, catch curve techniques, which have been used in the past to estimate mortality rates of lake trout, are based on the assumption of age-independent mortality rates, equal vulnerability to the sampling gear for the ages used in the analysis, and equal recruitment for all cohorts (Ricker 1975). The results of this study exemplify the utility of approaches such as statistical catch-at-age analysis in describing the dynamics of Great Lakes fish populations such as lake trout. My results showed that mortality rates were age- and year dependent, which had important implications to population grth and age structure of lake trout in the main basin of Lake Huron. However, these results also caused some difliculty in applying the 45% target rate, since in any given year, there was no single mortality rate. A goal of this study was to gauge progress towards rehabilitation by reference to the GLFC target maximum mortality rate (A=0.45, Z=0.59 year"). The fact that mortality rates vary with lake trout age brings forward a question of interpretation: to what ages should the target of 45% annual losses apply? In the model projections, I 129 assumed that total annual mortality should not exceed this level for lake trout ages 5 and older. However, this was in some sense an arbitrary choice, and if the age-specific patterns were different, very different dynamics could occur for populations experiencing the same peak mortality rate. This could even be the case when mortality rates had the same average over a broad range of ages. Furthermore, gauging rehabilitation progress using a target mortality rate seems more pertinent to self-sustaining populations, which is not the situation in the main basin of Lake Huron. Since a preliminary step towards rehabilitation is the establishment of self-sustaining lake trout populations, it would be more logical to set goals in terms of spawning stock biomass produced per fish rather than a mortality rate. Moreover, as populations become self-reproducing, stock- recruitment relationships, harvest allocations, and hatchery stocking should be evaluated in terms of population stability and rehabilitation objectives and goals. In closing, lake trout p0pulations in the main basin of Lake Huron face a difficult path towards self-sustainability due to sea lamprey parasitism and commercial fishing. If successful rehabilitation is to be achieved, total mortality in northern Lake Huron will have to be limited through coordinated multi-agency management of fishery harvest and sea lamprey control. APPENDIX- ADDITIONAL TABLES 130 Table 15. Joint age-length distribution for lake trout in northern Lake Huron (MI-I-l). Data from Michigan Department of Natural Resources annual spring gill net surveys from 1 984-1 994. Length Class (mm) Age 432-533 534-635 63 6-737 >737 Total 1 0 0 0 0 0 2 0 0 0 0 0 3 18 1 0 0 l9 4 222 5 0 0 227 5 128 34 0 0 162 6 15 22 1 0 38 7 6 3 3 0 12 8 O 1 2 0 3 9 0 l 2 1 4 10 0 l 2 1 4 1 1 0 0 1 1 2 12 0 0 0 1 1 13 0 0 0 1 1 14 0 0 0 l 1 15 0 0 O 1 1 16 0 0 0 l 1 17 0 O O l 1 18 0 0 0 1 1 19 0 0 0 1 l 20 0 O 0 l 1 >20 0 0 0 l 1 Total 389 68 1 1 13 481 131 Table 16. Joint age-length distribution for lake trout in central Lake Huron (MH—Z). Data fiom Michigan Department of Natural Resources annual spring gill net surveys from 1984-1994. Length Class (mm) Age 43 2-533 S34-635 63 6-737 >737 Total 1 O O O 0 0 2 O 0 0 O O 3 67 3 0 O 70 4 682 l 15 3 0 800 5 271 339 42 O 652 6 21 158 13 5 4 318 7 8 42 73 6 129 8 2 6 20 9 37 9 O O l 1 8 19 10 0 0 3 5 8 1 1 0 0 2 1 3 12 0 0 1 3 4 l3 0 0 2 2 4 14 0 0 0 l l 15 0 0 O l 1 l6 0 0 O 1 l 17 0 0 0 l 1 18 O 0 O l 1 l9 0 O O 1 1 20 O O 0 l 1 >20 0 O 0 l 1 Total 1,05 1 663 292 47 2,053 132 Table 17. Joint age-length distribution for lake trout in southern Lake Huron (MH- 3/4/5). Data from Michigan Department of Natural Resources annual spring gill net surveys from 1984-1994. Length Class (mm) Age 432-533 534-635 636-737 >737 Total 1 0 0 0 0 0 2 O 0 0 0 0 3 89 9 O 0 98 4 808 299 8 1 1,1 16 5 335 1,025 292 2 1,654 6 25 454 903 18 1,400 7 9 73 614 73 769 8 1 18 362 192 573 9 1 3 135 180 319 10 1 2 68 166 237 11 0 1 18 80 99 12 O 0 6 74 80 13 0 0 4 36 40 14 0 0 6 32 38 15 0 0 1 13 14 16 0 0 1 4 5 l7 0 0 0 6 6 18 0 0 0 9 9 19 0 0 0 2 2 20 O O 0 2 2 >20 0 0 0 2 2 Total 1,269 1,884 2,418 892 6,463 133 Table 18. Assumed age-1 abundance (x 1000) of lake trout in the main basin of Lake Huron. Data, adjusted for migration, were based on number of yearlings and fall fingerlings (age-O) stocked. Fall fingerlings were converted to yearling-equivalents based on the assumption that 40% of fingerlings survived to the yearling stage. Sixty percent of lake trout stocked in MH-2 were assumed to migrate to MH-l (J. Johnson, Alpena Fisheries Research Station, Michigan Department of Natural Resources, pers. comm). Region Year North (MH-l) Central (MH-2) South (MH-3/4/5) Basin total 1972 0 0 0 0 1973 384.6 0 100.0 484.6 1974 850.9 71.6 187.0 1,109.5 1975 707.4 72.8 331.0 1,111.2 1976 659.5 82.8 395.5 1,137.8 1977 713.0 81.2 361.0 1,155.2 1978 654.4 88.0 550.0 1,292.4 1979 555.0 75.2 777.8 1,408.0 1980 751.8 95.2 605.0 1,452.0 1981 245.3 15.2 555.0 815.5 1982 634.3 115.4 612.8 1,362.5 1983 529.1 84.0 650.4 1,263.5 1984 136.8 45.2 360.0 542.0 1985 489.8 87.6 482.1 1,059.5 1986 943.2 205.5 638.9 1,787.5 1987 480.1 105.2 169.6 754.9 1988 645.7 114.8 157.0 917.5 1989 658.6 120.4 390.8 1,169.8 1990 565.6 110.8 240.0 916.4 1991 967.1 185.9 339.0 1,492.0 1992 859.7 362.7 416.8 1,639.2 1993 657.3 293.0 389.5 1,339.8 134 Table 19. Sport harvest and effort of lake trout in Michigan waters of Lake Huron. Harvest reported in numbers of fish and effort expressed as angler hours. Data from Michigan Department of Natural Resources. Region North (MH-l) Central (MH-Z) South (MH-3/4/5) Year Harvest Effort Harvest Effort Harvest 1-3f‘fort'r 1984 1,861 * 99,413“ 381" 86,337" 27,827“ 723,572.7* 1985 1,861 99,413 454 102,860 27,827*** 723,572.7*** 1986 3,410 160,634 283 55,590 50,993 1,169,127 1987 974 82,698 380 72,306 40,255 1,059,693 1988 1,631 153,954 1,188 143,814 34,162 1,248,123 1989 869 130,019 67 4,627 38,615 685,205 1990 444 1 19,390 167 6,467 30,698 1,176,035 1991 1,968 108,959 1,689 129,022 14,351 581,542.5 1992 1,216 70,318 1,443 153,210 10,581 535,071 1993 264 69,408 424 142,517 5,450 410,962.5 I = Does not include data fi'om Harbor Beach, MI. * No data available, assumed to equal 1985 values. ** Estimated value based on ratio of 1984 to 1985 Canadian harvest in MH-2, 1984 sport harvest and effort = 0.8394 of 1985 harvest and effort. *** Estimated value based on ratio of 1985 to 1986 in MH-l, 1985 harvest = 0.5457 of 1986 harvest, 1985 effort = 0.6189 of 1986 effort. 135 Table 20. Age composition of sport fishery harvest of lake trout in Michigan waters of Lake Huron. Data, expressed as proportions at age, were from Michigan Department of Natural Resources sport harvest monitoring program. n= sample size. Year Region North (MK-1) Central (MH-Z) South (MH-3/4/5) 0 ooxrmmrswug O 3+ WQQUt-bw 9+ 3 OO\IO\MADJ 11 1985 0.09412 0.24706 0.35294 0.15294 0.11765 0.02353 0 0.01176 85 0.13699 0.30822 0.23288 0.19178 0.06849 0.02055 0.04110 146 0.02443 0.10860 0.26489 0.25795 0.17480 0.05737 0.11 196 375 1986 0 0.04310 0.64655 0.25862 0.03448 0.00862 0 0.00862 1 16 0.03004 0.40343 0.32618 0.12446 0.03433 0.05150 0.03004 233 0.01081 0.21364 0.16902 0.20798 0.15676 0.14963 0.09216 458 1987 0.02344 0.34375 0.25781 0.28125 0.08594 0.00781 0 0 128 0.02362 0.12598 0.29921 0.22047 0.24409 0.03150 0.05512 127 0.02779 0.09458 0.40052 0.16079 0.10236 0.10381 0.11015 323 1988 0 0.29710 0.52899 0.06522 0.07246 0.02899 0.00725 0 138 0.05000 0.65000 0.10000 0.15000 0.05000 0 0 20 0.00946 0.13061 0.21991 0.33218 0.11108 0.08627 0.11050 220 1991 0.04545 0.13636 0.31818 0.22727 0.18182 0.09091 0 0 22 0 0.13514 0.37838 0.43243 0.05405 0 0 37 0.03483 0.11946 0.23611 0.30978 0.14613 0.03142 0.12227 189 1992 0 0.10989 0.53846 0.30769 0.03297 0.01099 0 0 91 0 0.44318 0.15909 0.19318 0.20455 0 0 88 0.01233 0.20834 0.17703 0.12040 0.19017 0.06506 0.22667 202 136 Table 21. Canadian harvest of lake trout in southern Lake Huron (OH-3, OH-4 and OH- 5). Annual yield data from Ontario Ministry of Natural Resources. Harvest in numbers estimated by dividing yield by average mass per fish of Michigan sport harvest for each year. OH-3 OH-4/5 OH-3 + OH-4/5 Year Yield (kg) Numbers Yield (kg) Numbers Yield (kg) Numbers 1984 1,309 445 27,117 9,226 28,426 9,672 1985 368 125 20,235 6,885 20,603 7,010 1986 109 36 29,724 9,768 29,833 9,804 1987 107 36 29,829 10,154 29,936 10,191 1988 191 61 17,956 5741 18,147 5,802 1989 901 346 15,134 5,820 16,035 6,166 1990 1,625 572 11,985 4,221 13,610 4,793 1991 2,006 748 14,736 5,495 16,742 6,244 1992 1,564 510 21,355 6,959 22,919 7,469 1993 3,980 1,370 10,354 3,565 14,334 4,935 1994 7,769 r 2,675 10,393 3,578 18,162 6,253 137 Table 22. Catch and effort of lake trout from Michigan Department of Natural Resources annual spring gill net surveys in northern Lake Huron (MH-l). Efl‘ort expressed as meters of gill net per day. No data available for 1990. Year Age 1984 1985 1986 1987 1988 1989 1991 1992 1993 1994 1 0 0 0 0 2 0 0 0 0 0 2 12 4 22 5 l 31 2 6 5 1 3 124 82 17 4O 81 28 33 42 68 73 4 187 76 91 8 29 34 17 65 34 33 5 87 21 24 11 2 5 3 5 8 22 6 16 3 10 5 1 2 1 0 2 0 7 9 1 1 1 0 0 0 0 0 0 8 3 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 ll 1 1 0 0 0 0 0 0 0 0 12 0 0 0 0 0 0 0 0 O 0 13 0 0 0 0 0 0 O 0 0 0 l4 0 0 0 0 O 0 O 0 0 0 15 0 0 0 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0 0 0 0 l7 0 0 0 0 0 0 0 0 0 0 18 0 0 0 0 0 0 0 0 0 0 19 0 0 0 0 0 O 0 0 0 0 20 0 0 0 0 0 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 Total 439 188 165 70 116 100 56 118 117 129 Effort 3,018 3,018 3,018 3,018 3,018 3,018 3,018 3,018 3,018 3,018 138 Table 23. Catch and effort of lake trout fi'om Michigan Department of Natural Resources annual spring gill net surveys in central Lake Huron (MB-2). Effort expressed as meters of gill net per day. Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1 0 0 0 0 0 0 0 0 0 0 0 2 35 12 76 23 15 24 68 0 0 2 4 3 150 157 57 173 187 119 98 11 3 9 33 4 156 195 185 56 65 203 53 25 91 22 51 5 90 51 84 99 11 71 33 33 59 91 45 6 29 37 5 47 7 11 5 89 29 9 50 7 21 17 12 4 4 12 2 4 43 5 5 8 3 7 7 1 0 4 2 7 0 5 l 9 3 0 2 2 1 0 1 6 2 0 2 10 2 2 1 0 1 0 0 0 2 0 0 11 0 0 0 0 0 1 0 l 1 0 0 12 1 2 0 0 0 0 0 0 1 0 0 13 0 1 2 0 0 0 0 0 0 1 0 14 0 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 O 0 0 0 16 0 0 0 0 0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 1 0 0 18 0 0 0 0 0 0 0 0 1 0 0 19 0 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0 0 0 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 0 Tina] 490 481 431 405 291 445 262 176 233 144 191 Effort 3,018 3,018 3,018 3,018 3,018 2,012 3,018 1,372 1,372 1,554 1,852 139 Table 24. Catch and effort of lake trout fi'om Michigan Department of Natural Resources annual spring gill net surveys in southern Lake Huron (MH-3/4/5). Effort expressed as meters of gill net per day. Year .Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1 0 0 0 0 13 11 18 0 0 O 1 2 22 10 35 20 6 14 2 0 0 3 6 3 93 113 81 58 146 39 15 15 1 30 35 4 198 146 318 90 139 152 18 21 57 29 96 5 229 270 145 267 146 143 168 13 57 185 39 6 224 243 163 79 227 76 101 58 34 30 166 7 105 165 78 71 42 115 60 37 61 14 21 8 68 110 66 38 39 16 143 24 41 22 6 9 51 54 18 37 27 12 13 44 32 18 10 10 26 31 10 29 24 8 23 6 55 14 9 11 0 16 5 8 9 14 4 4 10 24 5 12 0 16 6 8 3 8 7 7 l6 5 4 13 0 0 8 1 4 4 1 12 3 7 0 14 0 0 l 4 7 2 2 O 12 6 4 15 0 0 0 0 2 1 0 1 6 3 1 16 0 0 0 0 0 0 0 0 0 4 1 17 0 0 0 0 0 0 0 0 4 1 1 18 0 0 0 0 0 0 0 0 8 0 l 19 0 0 0 0 0 0 0 0 1 0 0 20 0 0 0 0 0 0 0 0 3 0 0 >20 0 0 0 0 0 0 0 0 0 0 0 Total 1,0161,174 934 710 834 615 575 242 401 395 406 Effort 3,018 3,018 2,012 1,555 1,303 1,303 1,463 1,143 1,097 1,573 1,481 140 Table 25. Canadian harvest of lake trout in OH-l and OH-2 in northern and central Lake Huron. Forty percent of the harvest from zone 4-1 in district OH-l were assumed to be from the northern area. Sixty percent of lake trout harvested in zone 4-1 of OH-l, and all harvest in OH-2 were assumed to be fiom the central area. Annual yield data from Ontario Ministry of Natural Resources. Harvest in numbers for Canadian removals fiom the MH-l stock estimated by dividing reported yield by average mass per fish of tribal gill net harvest in MH-l for each year. Harvest in numbers for Canadian removals from the MH-Z stock estimated by dividing reported yield by average mass per fish of Michigan sport harvest for each year. Northern Central OH-l (MH-l) OH-l + 0H-2 (MB-2) Year Yield (kg) Numbers Yield (kg) Numbers 1984 249.2 207 737.8 381 1985 116.0 93 879.0 453 1986 112.8 115 2,484.2 1,361 1987 435.6 376 1,903.4 810 1988 506.0 771 2,104.0 1,222 1989 588.8 1,039 3,884.2 2,160 1990 613.6 697 5,409.4 3,008 1991 886.8 831 5,633.2 3,004 1992 1,386.8 1,211 7,041.2 4,029 1993 2,150.4 5,532 14,817.6 7,710 14 1 Table 26. Reported tribal commercial harvest and effort of lake trout in northern Lake Huron (MH-l). Data provided by Chippewa-Ottawa Treaty Fishery Management Authority. Effort expressed as meters of large-mesh gill net targeted at lake Whitefish and lake trout. Year Yield (kg) Effort (m) 1984 89,151.45 2,239,579 1985 102,468.24 2,782,824 1986 105,370.37 3,822,680 1987 78,583.02 3,310,555 1988 75,575.20 3,702,863 1989 76,512.34 4,122,511 1990 35,945.53 3,296,442 1991 35,557.25 3,386,999 1992 43,579.62 2,334,097 1993 56,659.63 2,362,779 1 42 Table 27 . Parameters estimated to calibrate the northern and central lake trout population models. f c, y =commercial fishing intensity (year") in year y, u’ = proportionality coeficient for sea lamprey-induced mortality, and qR= catchability coemcient for the recreational fishery (angler hours'l) , p. = survival proportion for age a for cohorts before 1984 to estimate abundance in 1984 for ages>1. Modeled Region Parameter MH-l MH-2 fc. 1934 0.485177 0.007804 fc, 1985 0.897677 0.010657 fc, 1986 1.760810 0.032434 fc,1987 9.148661 0.019773 fc. 1988 6.931023 0.026980 fc_ 1939 3.344381 0.034468 fc. 1990 0.953025 0.042784 fc, 1991 0.398272 0.043089 fc, 1992 0.331953 0.062046 fc, 1993 0.392221 0.130979 qR 4.29318 mo"8 1.05413 no"7 p.’ 4.059982 not estimated pl 0.513611 0.513611 p2 0.725195 0.727298 p3 0.781318 0.841338 p4 0.683716 0.878265 p5 0.614042 0.874634 p6 0.575742 0.840974 p7 0.125603 0.827164 pg 0.099222 0.807527 p9 0.096822 0.784633 p10 0.102303 0.778539 pll 0.092245 0.787351 p12 0.087107 0.774820 p13 0.087107 0.782319 914+ 0.087107 0.767396 143 Table 28. Model estimates of lake trout abundance in southern main basin of Lake Huron (MH-3/4/5). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 360,000 482,100 638,900 169,600 157,000 390,800 240,000 339,000 416,800 389,500 2 434,858 184,900 247,612 328,146 87,108 80,637 200,719 123,267 174,114 214,073 3 127,660 316,270 134,477 180,086 238,658 63,353 58,647 145,981 89,651 126,632 4 134,857 96,427 252,322 107,631 146,469 190,872 47,465 43,794 114,019 59,010 5 117,584 99,824 77,165 195,076 86,604 113,639 140,783 34,193 33,817 76,955 6 98,613 74,625 72,609 49,641 138,358 57,260 72,435 86,672 23,136 22,006 7 40,918 59,169 50,874 42,874 32,831 89,836 34,204 44,227 55,905 14,507 8 26,166 24,389 39,478 28,995 27,657 21,522 52,961 20,978 28,560 34,077 9 18,064 15,462 16,131 21,176 18,238 17,821 12,542 31,219 13,725 16,636 10 985 10,646 10,223 8,245 13,112 11,585 10,340 7,135 20,785 7,695 1 1 2,758 580 7,048 5,085 5,065 8,255 6,710 5,757 4,803 1 1,402 12 912 1,620 384 3,427 3,102 3,164 4,770 3,671 3,905 2,590 13 0 535 1,073 182 2,074 1,923 1,824 2,563 2,512 2,065 14 0 0 354 512 110 1,288 1,109 984 1,750 1,334 15 0 0 0 171 311 69 743 604 669 938 16 0 0 0 0 103 193 40 399 413 353 17 0 0 0 0 0 65 111 22 271 223 18 0 0 0 0 0 0 37 59 15 141 19 O 0 0 0 0 0 0 20 41 8 20 0 0 0 0 0 0 0 0 14 21 >20 0 0 0 0 0 0 0 0 0 7 Total 1,363,377 1,366,547 1,548,648 1,140,848 956,8021,052,283 885,440 890,545 984,904 980,174 144 Table 29. Estimates of instantaneous rates of natural mortality (M) for lake trout in main basin of Lake Huron based on statistical catch-at-age analysis of the southern Lake Huron population model. Rates were assumed constant from 1984-1993. Age M (year'l) 1 0.666 2 0.318 3 0.172 4 0.123 5 0.108 6 0.103 7 0.101 8 0.100 9 0.100 10 0.100 11 0.100 12 0.100 13 0.100 14 0.100 15 0.100 16 0.100 17 0.100 18 0.100 19 0.100 20 0.100 >20 0.100 Table 30. Model estimates of instantaneous rates of recreational fishing mortality 145 (year'l) for lake trout in southern Lake Huron (MH-3/4/S). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 4.77x10" 5.33x10“ 7.96x10“ 6.48x10“ 6.371110" 4.57x10‘ 9.33::10‘s 4.21x10“ 3.99x10" 2.64x1045 3 0.004 0.004 0.006 0.005 0.005 0.003 0.007 0.003 0.003 0.002 4 0.036 0.041 0.061 0.050 0.049 0.035 0.071 0.032 0.030 0.020 5 0.100 0.112 0.167 0.136 0.134 0.096 0.196 0.089 0.084 0.056 6 0.109 0.122 0.182 0.148 0.146 0.105 0.213 0.096 0.091 0.060 7 0.107 0.120 0.179 0.146 0.143 0.103 0.210 0.095 0.090 0.059 8 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 9 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 10 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 11 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 12 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 13 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 14 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 15 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 16 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 17 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 18 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 19 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 20 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 >20 0.110 0.123 0.183 0.149 0.147 0.105 0.215 0.097 0.092 0.061 146 Table 31. Model estimates of number of lake trout deaths (x1000) due to natural mortality in region southern Lake Huron (MI-I-3/4/5). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 175.100 234.488 310.754 82.492 76.363 190.081 116.733 164.886 202.727 189.449 2 118.586 50.422 67.524 89.486 23.755 21.990 54.736 33.615 47.481 58.378 3 19.102 48.579 20.688 27.919 36.701 9.443 8.728 22.195 12.572 19.443 4 14.380 10.673 27.469 11.944 15.972 20.318 4.997 4.765 11.640 6.510 5 10.175 9.207 6.720 17.780 7.638 9.854 12.013 3.048 2.961 6.968 6 7.917 6.356 5.787 4.168 11.519 4.588 5.862 7.194 1.895 1.825 7 3.221 4.906 3.924 3.500 2.700 7.036 2.728 3.612 4.446 1.169 8 2.040 2.003 2.946 2.327 2.244 1.667 4.125 1.714 2.212 2.712 9 1.404 1.267 1.176 1.684 1.467 1.376 0.959 2.567 1.043 1.317 10 0.077 0.873 0.736 0.653 1.050 0.893 0.783 0.589 1.563 0.608 1 1 0.214 0.048 0.502 0.401 0.404 0.635 0.504 0.477 0.358 0.899 12 0.071 0.133 0.027 0.269 0.247 0.243 0.355 0.306 0.289 0.204 13 0 0.044 0.076 0.014 0.165 0.148 0.136 0.213 0.186 0.163 14 0 0 0.025 0.040 0.009 0.099 0.083 0.082 0.130 0.105 15 0 0 0 0.013 0.025 0.005 0.055 0.050 0.049 0.074 16 0 0 0 0 0.008 0.015 0.003 0.033 0.031 0.028 17 0 0 0 0 0 0.005 0.008 0.002 0.020 0.018 18 0 0 0 0 0 0 0.003 0.005 0.001 0.011 19 0 0 0 0 0 0 0 0.002 0.003 0.001 20 0 0 0 0 0 0 0 0 0.001 0.002 >20 0 0 0 0 0 0 0 0 0 0.001 Total 352.287 368.999 448.353 242.691 180.268 268.396 212.811 245.355 289.610 289.883 147 Table 32. Model estimates of number of lake trout deaths (x1000) due to recreational fishing mortality in southern Lake Huron (MH-3/4/5). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0.002 0.001 0.002 0.002 0 0 0.002 0 0.001 0 3 0.406 1.154 0.734 0.807 1.042 0.193 0.363 0.417 0.223 0.229 4 4.239 3.518 13.527 4.790 6.294 5.749 2.882 1.242 2.871 1.064 5 9.458 9.570 10.434 22.483 9.490 8.791 21.852 2.505 2.302 3.592 6 8.421 7.559 10.282 6.032 16.378 4.683 12.201 6.766 1.687 1.077 7 3.423 5.831 6.967 5.061 3.836 7.177 5.674 3.395 3.954 0.689 8 2.231 2.449 5.381 3.461 3.280 1.750 8.826 1.657 2.024 1.645 9 1.541 1.555 2.156 2.514 2.152 1.449 2.060 2.491 0.958 0.802 10 0.084 1.071 1.350 0.975 1.541 0.941 1.682 0.572 1.436 0.370 1 1 0.235 0.058 0.921 0.599 0.593 0.670 1.083 0.463 0.329 0.548 12 0.078 0.163 0.050 0.403 0.362 0.256 0.764 0.297 0.265 0.124 13 0 0.054 0.139 0.021 0.242 0.156 0.292 0.207 0.171 0.099 14 0 0 0.046 0.060 0.013 0.104 0.179 0.079 0.120 0.064 15 0 0 0 0.020 0.036 0.006 0.119 0.049 0.045 0.045 16 0 0 0 0 0.012 0.016 0.006 0.032 0.028 0.017 17 0 0 0 0 0 0.005 0.018 0.002 0.018 0.011 18 0 0 0 0 0 0 0.006 0.005 0.001 0.007 19 0 0 0 0 0 0 0 0.002 0.003 0 20 0 0 0 0 0 0 0 0 0.001 0.001 >20 0 0 0 0 0 0 0 0 0 0 Total 30.117 32.983 51.989 47.227 45.272 31.945 58.008 20.180 16.438 10.385 148 Table 33. Model estimates of number of lake trout deaths (x1000) due to sea lamprey- induced mortality in southern Lake Huron (MH-3/4/5). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 l 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 3 11.726 14.215 5.424 4.891 10.042 6.253 5.762 9.350 17.845 6.015 4 16.414 5.072 16.249 4.293 10.563 24.023 5.394 3.970 22.552 4.530 5 23.325 8.438 10.371 16.455 12.216 22.560 20.246 5.504 6.548 12.577 6 23.107 9.837 13.666 6.610 20.625 13.785 10.145 16.808 5.046 4.940 7 9.885 8.954 10.988 6.657 4.773 22.662 4.824 8.661 13.429 3.542 8 6.434 3.805 9.974 4.970 4.311 5.562 8.791 3.881 7.688 8.668 9 4.473 2.416 4.554 3.866 3.033 4.657 2.387 5.375 4.030 4.314 10 0.245 1.655 3.052 1.552 2.266 3.041 2.118 1.171 6.384 2.011 11 0.688 0.090 2.197 0.982 0.904 2.180 1.452 0.912 1.525 3.006 12 0.229 0.252 0.125 0.681 0.570 0.840 1.088 0.557 1.286 0.689 13 0 0.083 0.346 0.036 0.379 0.510 0.411 0.393 0.821 0.549 14 0 0 0.112 0.100 0.020 0.341 0.243 0.154 0.563 0.353 15 0 0 0 0.034 0.057 0.018 0.170 0.091 0.220 0.250 16 0 0 0 0 0.018 0.051 0.009 0.064 0.131 0.093 17 0 0 0 0 0 0.017 0.026 0.003 0.091 0.060 18 0 0 0 0 0 0 0.009 0.009 0.005 0.038 19 0 0 0 0 0 0 0 0.003 0.014 0.002 20 0 0 0 0 0 0 0 0 0.005 0.006 >20 0 0 0 0 0 0 0 0 0 0.002 Total 96.526 54.817 77.059 51.127 69.779 106.502 63.075 56.905 88.183 51.645 149 Table 34. Model estimates of lake trout abundance in central main basin of Lake Huron (MI-L2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 45,200 87,600 205,472 105,200 114,800 120,400 1 10,800 185,920 362,720 293,040 2 43,143 23,215 44,992 105,533 54,032 58,962 61,839 56,908 95,490 186,297 3 43,122 31,376 16,883 32,712 76,738 39,287 42,868 44,956 41,371 69,407 4 4,777 29,613 21,684 13,259 27,319 57,466 31,934 35,540 35,806 29,600 5 26,277 3,304 21,183 17,262 1 1,249 20,561 46,251 26,190 28,167 25,279 6 18,155 16,084 2,346 16,118 13,730 7,778 14,483 33,973 18,913 18,398 7 17,866 10,621 11,384 1,670 11,917 9,116 4,920 10,191 22,665 11,799 8 13,636 10,788 7,563 8,028 1,226 8,058 5,797 3,557 6,841 14,397 9 11,229 8,425 7,671 5,221 5,796 827 5,089 4,136 2,386 4,330 10 7,746 7,051 5,980 5,143 3,674 3,886 512 3,573 2,743 1,498 l 1 5,931 4,852 4,996 4,024 3,621 2,442 2,415 350 2,407 1,734 12 0 3,737 3,456 3,374 2,846 2,455 1,521 1,735 233 1,532 13 0 0 2,642 2,322 2,371 1,876 1,524 1,020 1,174 147 14 0 0 0 1,780 1,639 1,589 1,167 1,065 684 745 15 0 0 0 0 1,246 1,062 985 751 728 429 16 0 0 0 0 0 807 658 634 513 456 17 0 0 0 0 0 0 501 423 433 322 18 0 0 0 0 0 0 0 322 289 271 19 0 0 0 0 0 0 0 0 220 181 20 0 0 0 0 0 0 0 0 0 138 >20 0 0 0 0 0 0 0 0 0 0 Total 237,084 236,664 356,251 321,646 332,205 336,571 333,264 411,243 623,784 659,999 150 Table 35. Model estimates of instantaneous rates of recreational fishing mortality (year'l) for lake trout in central Lake Huron (MH-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 3.95x10" 4.71x10" 5.85x10‘7 6.28x10’7 6.58x10’7 6.3rx10" 6.39x10" 5.90x10" 7.01x10" 103le 3 3.02x10" 3.60x10" 4.471110" 4.80x10" 5.03x10" 4.82x10" 4.89x10" 4.51x10" 5.36x10" 7.91x10" 4 0.003 0.004 0.004 0.005 0.005 0.005 0.005 0.005 0.005 0.008 5 0.008 0.010 0.012 0.013 0.014 0.013 0.013 0.012 0.015 0.022 6 0.009 0.011 0.013 0.014 0.015 0.014 0.015 0.014 0.016 0.024 7 0.009 0.011 0.013 0.014 0.015 0.014 0.014 0.013 0.016 0.023 8 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 9 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 10 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 11 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 12 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 13 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 14 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 15 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 16 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 17 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 18 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 19 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 20 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 >20 0.009 0.011 0.013 0.014 0.015 0.015 0.015 0.014 0.016 0.024 151 Table 36. Model estimates of instantaneous rates of commercial fishing mortality (year") for lake trout in central Lake Huron (MB-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 l O 0 0 0 0 0 0 O 0 0 2 7.801110" 1.071110" 3.241110" 1.981110" 2.701110" 3.451110" 4.281110" 4.311110" 0.001 0.001 3 7.801110" 1.071110" 3.241110" 1.981110" 2.701110'3 3.451110'3 4.281110" 4.3111103 6.201110" 0010 4 0.006 0.008 0.024 0.015 0.020 0.026 0.032 0.032 0.047 0.098 5 0.008 0.011 0.032 0.020 0.027 0.034 0.043 0.043 0.062 0.131 6 0.007 0.009 0.028 0.017 0.023 0.030 0.037 0.037 0.053 0.113 7 0.004 0.006 0.018 0.011 0.015 0.019 0.024 0.024 0.034 0.072 8 0.004 0.005 0.016 0.010 0.013 0.017 0.021 0.021 0.030 0.064 9 0.003 0.004 0.013 0.008 0.011 0.013 0.017 0.017 0.024 0.051 10 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 11 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 12 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 13 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 14 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 15 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 16 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 17 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 18 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 19 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 20 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 >20 0.002 0.002 0.006 0.004 0.005 0.007 0.009 0.009 0.012 0.026 Table 37 . Model estimates of number of lake trout deaths (x1000) due to natural 152 mortality in central Lake Huron (NIH-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 21.985 42.608 99.939 51.168 55.838 58.561 53.892 90.430 176.423 142.532 2 11.765 6.331 12.268 28.776 14.733 16.077 16.860 15.516 26.033 50.772 3 6.168 4.501 2.574 5.137 11.435 6.089 6.708 6.900 6.033 9.941 4 0.493 3.107 2.394 1.510 2.936 6.378 3.577 3.915 3.733 2.971 5 2.236 0.301 1.996 1.662 1.013 1.868 4.286 2.407 2.470 2.130 6 1.441 1.394 0.204 1.426 1.155 0.640 1.252 2.864 1.546 1.457 7 1.415 0.908 0.969 0.145 0.994 0.739 0.424 0.847 1.837 0.935 8 1.085 0.917 0.634 0.687 0.102 0.648 0.493 0.294 0.550 1.136 9 0.899 0.714 0.633 0.441 0.478 0.066 0.429 0.339 0.191 0.342 10 0.619 0.597 0.494 0.434 0.302 0.309 0.043 0.295 0.220 0.119 1 1 0.475 0.412 0.413 0.340 0.300 0.195 0.206 0.029 0.194 0.138 12 0 0.316 0.285 0.284 0.233 0.195 0.125 0.144 0.019 0.122 13 0 0.000 0.218 0.196 0.195 0.149 0.128 0.084 0.094 0.012 14 0 0.000 0.000 0.150 0.133 0.126 0.094 0.089 0.055 0.059 15 0 0 0 0 0.101 0.084 0.080 0.062 0.058 0.034 16 0 0 0 0 0 0.064 0.053 0.053 0.041 0.036 17 0 0 0 0 0 0 0.040 0.035 0.035 0.026 18 0 0 0 0 0 0 0 0.027 0.023 0.022 19 0 0 0 0 0 0 0 0 0.018 0.014 20 0 0 0 0 0 0 0 0 0 0.011 >20 0 0 0 0 0 0 0 0 0 0 Total 48.581 62.104 123.021 92.355 89.948 92.189 88.690 124.331 219.573 212.811 T2 fis 153 Table 38. Model estimates of number of lake trout deaths (x1000) due to recreational fishing mortality in central Lake Huron (MB-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 3 0.011 0.009 0.007 0.014 0.034 0.017 0.019 0.018 0.019 0.046 4 0.012 0.090 0.087 0.059 0.119 0.249 0.141 0.143 0.162 0.190 5 O. 172 0.028 0.228 0.203 0.130 0.230 0.534 0.277 0.338 0.430 6 0.127 0.146 0.027 0.200 0.170 0.090 0.179 0.377 0.242 0.336 7 0.125 0.095 0.126 0.020 0.146 0.104 0.060 0.112 0.287 0.216 8 0.098 0.099 0.085 0.099 0.015 0.094 0.072 0.040 0.089 0.270 9 0.082 0.077 0.085 0.064 0.072 0.010 0.063 0.046 0.031 0.081 10 0.056 0.065 0.067 0.063 0.046 0.045 0.006 0.040 0.036 0.028 11 0.043 0.045 0.056 0.049 0.045 0.028 0.030 0.004 0.031 0.033 12 0 0.034 0.038 0.041 0.035 0.028 0.018 0.020 0.003 0.029 13 0 0 0.029 0.028 0.030 0.022 0.019 0.011 0.015 0.003 14 0 0 0 0.022 0.020 0.018 0.014 0.012 0.009 0.014 15 0 0 0 0 0.015 0.012 0.012 0.008 0.009 0.008 16 0 0 0 0 0 0.009 0.008 0.007 0.007 0.009 17 0 0 O 0 0 0 0.006 0.005 0.006 0.006 18 0 0 0 0 0 0 0 0.004 0.004 0.005 19 0 0 0 0 O 0 0 0 0.003 0.003 20 0 0 0 0 0 0 0 0 0 0.003 >20 0 0 0 0 0 0 0 0 0 0 Total 0.726 0.689 0.835 0.862 0.878 0.956 1.182 1.124 1.289 1.710 154 Table 39. Model estimates of number of lake trout deaths (x1000) due to commercial fishing mortality in central Lake Huron (MH-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0.003 0.002 0.012 0.018 0.012 0.017 0.023 0.021 0.051 0.209 3 0.028 0.028 0.049 0.059 0.180 0.122 0.167 0.173 0.218 0.759 4 0.023 0.201 0.472 0.181 0.481 1.335 0.929 1.025 1.407 2.364 5 0.162 0.030 0.601 0.305 0.254 0.598 1.703 0.963 1.423 2.591 6 0.094 0.125 0.055 0.237 0.261 0.185 0.449 1.035 0.805 1.601 7 0.060 0.053 0.171 0.016 0.146 0.139 0.099 0.199 0.622 0.668 8 0.041 0.048 0.100 0.066 0.013 0.109 0.103 0.062 0.167 0.727 9 0.027 0.030 0.080 0.034 0.050 0.009 0.072 0.057 0.046 0.174 10 0.010 0.013 0.032 0.017 0.016 0.021 0.004 0.025 0.027 0.031 11 0.007 0.009 0.027 0.013 0.016 0.013 0.018 0.002 0.024 0.036 12 0 0.007 0.019 0.011 0.013 0.013 0.011 0.012 0.002 0.032 13 0 0 0.014 0.008 0.011 0.010 0.011 0.007 0.012 0.003 14 0 0 0 0.006 0.007 0.009 0.008 0.008 0.007 0.016 15 0 0 0 0 0.005 0.006 0.007 0.005 0.007 0.009 16 0 0 0 0 0 0.004 0.005 0.005 0.005 0.010 17 0 0 0 0 0 0 0.003 0.003 0.004 0.007 18 0 0 0 0 0 0 0 0.002 0.003 0.006 19 0 0 0 O 0 0 0 0 0.002 0.004 20 0 0 0 0 0 0 0 0 0.0 0.003 >20 0 0 0 0 0 0 0 0 0.0 0 Total 0.457 0.544 1.633 0.972 1.467 2.592 3.611 3.606 4.832 9.249 155 Table 40. Model estimates of number of lake trout deaths (x1000) due to sea lamprey- induced mortality in central Lake Huron (MB-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 l 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 3 7.302 5.152 0.994 0.183 7.624 1.124 0.434 2.059 5.501 10.868 4 0.944 5.032 1.470 0.260 3.222 3.252 1.096 2.290 5.225 4.808 5 7.623 0.600 2.239 1.362 2.075 3.382 5.756 3.629 5.538 4.980 6 5.872 3.034 0.389 2.339 3.028 1.943 2.413 7.031 4.522 4.286 7 5.479 2.002 2.089 0.263 2.572 2.336 0.780 2.192 5.522 2.841 8 3.986 2.053 1.523 1.379 0.268 2.119 0.993 0.775 1.705 3.524 9 3.170 1.625 1.730 1.009 1.310 0.232 0.952 0.950 0.620 1.099 10 2.210 1.381 1.363 1.008 0.869 1.095 0.109 0.805 0.726 0.384 11 1.669 0.931 1.126 0.775 0.805 0.684 0.427 0.082 0.626 0.445 12 0 0.738 0.792 0.666 0.690 0.693 0.347 0.385 0.062 0.392 13 0 0 0.599 0.452 0.546 0.527 0.301 0.233 0.308 0.038 14 0 0 0 0.357 0.417 0.451 0.300 0.229 0.185 0.190 15 0 0 0 0 0.317 0.301 0.253 0.161 0.197 0.109 16 0 0 0 0 0 0.229 0.169 0.136 0.139 0.116 17 0 0 0 0 0 0 0.129 0.091 0.117 0.082 18 0 0 0 0 0 0 0 0.069 0.078 0.069 19 0 0 0 0 0 0 0 0 0.059 0.046 20 0 0 0 0 0 0 0 0 0 0.035 >20 0 0 0 0 0 0 0 0 0 0.000 Total 38.256 22.547 14.317 10.053 23.741 18.369 14.458 21.118 31.132 34.314 156 Table 41. Model estimates of lake trout abundance in northern main basin of Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 136,800 489,840 943,176 480,060 645,720 658,580 565,620 967,100 859,680 657,260 2 271,751 70,262 251,587 484,425 246,564 331,649 338,254 290,508 496,713 441,541 3 236,241 196,688 50,645 179,785 321,519 167,317 233,274 243,678 210,446 360,061 4 71,386 155,194 124,798 24,929 52,655 119,788 87,679 172,419 155,096 133,847 5 149,587 37,455 59,333 21,354 20 233 7,840 37,301 91,720 84,383 6 67,808 53,868 9,642 5,118 2 0 5 2,348 14,883 43,163 7 46,032 14,958 10,754 629 1 0 0 l 655 6,182 8 6,300 13,978 4,511 1,511 2 0 0 0 0 275 9 578 996 2,877 411 6 O 0 O 0 0 10 60 94 220 312 4 0 0 0 0 0 11 7 11 25 33 17 0 0 0 0 0 12 0 1 3 4 2 1 0 0 0 0 l3 0 0 0 0 0 0 0 O 0 0 14 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 l6 0 0 0 0 O 0 0 0 O 0 l7 0 0 0 0 0 0 0 0 0 0 18 0 0 0 0 0 0 0 0 0 0 l9 0 0 0 0 0 0 0 0 0 0 20 O 0 0 0 O 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 Totafl 986,552 1,033,344 1,457,570 1,198,572 1,266,512 1,277,569 1,232,672 1,713,355 1,829,193 1,726,712 157 Table 42. Model estimates of instantaneous rates of recreational fishing mortality (year") for lake trout in northern Lake Huron (MH-l). Agp \Rmu 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 \OOOQQLII-BUJNr—n r—u—t—h—t—nu—Iv—Iv—nu—u— \OOOQO\M-5UNh—o 20 >20 0 1.851110" 1.421110" 1.411110" 0 0 1.851110" 2.991110" 1.421110" 2.291110" 1.411110" 2.291110" 3.891110" 3.891110" 4.241110" 4.241110" 4.161110" 4.161110" 4.261110" 4.261110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 4.271110" 6.291110" 6.851110" 6.731110" 6.881110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 6.901110" 0 0 0 0 0 1.541110" 2.871110" 2.421110" 2.221110" 2.031110" 1.181110" 2.191110" 1.181110" 2.191110" 3.241110" 6.021110" 3.531110" 6.561110" 3.461110" 6.451110" 3.541110" 6.601110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.551110" 6.611110" 3.55x10" 6.611110" 1.851110" 1.851110" 1.701110" 1.551110" 1.701110" 1.551110" 5.091110" 4.671110" 4.261110" 5.541110" 5.451110" 5.57x10" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.581110" 5.091110" 4.641110" 5.001110" 4.561110" 5.121110" 4.671110" 5.131110" 4.681110" 5.13x10" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.681110" 5.131110" 4.68x10" 0 0 1.311110" 1.291110" 1.001110" 9.891110" 1.001110" 9.871110" 2.751110" 2.721110" 3.001110" 2.961110" 2.951110" 2.911110" 3.011110" 2.971110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.021110" 2.981110" 3.02x10" 2.981110" 3.021110" 2.981110" 1 5 8 Table 43. Model estimates of instantaneous rates of commercial fishing mortality (year'l) for lake trout in northern Lake Huron (MH- 1 ). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0.005 0.009 0.018 0.091 0.069 0.033 0.010 0.004 0.003 0.004 3 0.049 0.090 0.176 0.915 0.693 0.334 0.095 0.040 0.033 0.039 4 0.364 0.673 1.321 6.861 5.198 2.508 0.715 0.299 0.249 0.294 5 0.485 0.898 1.761 9.149 6.931 3.344 0.953 0.398 0.332 0.392 6 0.417 0.772 1.514 7.868 5.961 2.876 0.820 0.343 0.285 0.337 7 0.267 0.494 0.968 5.032 3.812 1.839 0.524 0.219 0.183 0.216 8 0.238 0.440 0.863 4.483 3.396 1.639 0.467 0.195 0.163 0.192 9 0.189 0.350 0.687 3.568 2.703 1.304 0.372 0.155 0.129 0.153 10 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 11 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 12 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 13 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 14 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 15 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 16 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 17 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 18 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 19 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 20 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 >20 0.097 0.180 0.352 1.830 1.386 0.669 0.191 0.080 0.066 0.078 159 Table 44. Model estimates of instantaneous rates of sea lamprey-induced mortality (year") for lake trout in northern Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 3 0.200 0.193 0.361 0.141 0.122 0.140 0.035 0.240 0.248 0.093 4 0.156 0.163 0.319 0.120 0.097 0.093 0.015 0.207 0.235 0.053 5 0.425 0.348 0.576 0.253 0.254 0.382 0.140 0.409 0.311 0.298 6 0.987 0.733 1.106 0.540 0.588 0.988 0.415 0.827 0.488 0.814 7 0.820 0.600 0.886 0.507 0.532 0.808 0.426 0.674 0.581 0.681 8 1.503 1.036 1.427 0.941 1.009 1.543 0.914 1.133 1.036 1.342 9 1.525 1.054 1.427 0.985 1.049 1.565 0.974 1.147 1.115 1.372 10 1.525 1.054 1.427 0.985 1.049 1.565 0.974 1.147 1.115 1.372 11 1.534 1.050 1.370 1.060 1.114 1.573 1.094 1.130 1.295 1.405 12 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 13 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 14 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 15 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 16 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 17 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 18 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 19 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 20 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 >20 1.592 1.108 1.428 1.118 1.171 1.631 1.151 1.187 1.353 1.463 160 Table 45. Model estimates of number of lake trout deaths (x1000) due to natural mortality in northern Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 66.538 238.253 458.751 233.496 314.071 320.326 275.112 470.387 418.139 319.684 2 73.937 19.079 68.040 126.547 65.081 89.024 91.828 79.073 135.242 120.186 3 33.102 27.119 6.226 17.766 35.065 21.150 34.549 33.647 29.048 53.296 4 0.004 0.010 0.008 0.002 0.003 0.007 0.005 0.011 0.010 0.008 5 10.109 2.211 2.388 0.242 0 0 006 0.491 2.632 6.946 6.262 6 3.591 2.748 0.339 0.062 0 0 0 0.136 1.017 2.522 7 2.716 0.880 0.476 0.011 0 0 0 0 0.044 0.394 8 0.289 0.705 0.172 0.027 0 0 0 0 0 0.014 9 0.027 0.051 0.116 0.009 0 0 0 0 0 0 10 0.003 0.005 0.010 0.010 0 0 0 0 0 0 1 1 0 0.001 0.001 0.001 0.001 O 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 l3 0 0 0 0 0 0 0 0 0 0 14 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0 0 0 0 l7 0 0 0 0 0 0 0 0 0 0 l8 0 0 0 0 0 0 0 0 0 0 19 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0 0 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 Total 190.316 291.062 536.526 378.173 414.221 430.514 401.985 585.887 590.446 502.365 161 Table 46. Model estimates of number of lake trout deaths (x1000) due to recreational fishing mortality in northern Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 3 0.027 0.022 0.008 0.012 0.045 0.023 0.034 0.030 0.017 0.031 4 0.074 0.141 0.134 0.004 0.021 0.076 0.100 0.198 0.116 0.105 5 0.365 0.080 0.139 0.007 0 0 0.021 0.104 0.177 0.158 6 0.148 0.113 0.023 0.002 0 0 0 0.006 0.030 0.073 7 0.112 0.036 0.032 0 0 0 0 0 0.001 0.011 8 0.012 0.030 0.012 0.001 0 0 0 0 0 0 9 0.001 0.002 0.008 0 0 0 0 0 0 0 10 0 0 0.001 0 0 0 0 0 0 0 11 0 0 0 0 0 0 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 13 0 0 O 0 0 0 0 0 0 0 14 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 0 0 18 0 0 0 0 0 0 0 0 0 0 19 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0 0 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 Total 0.740 0.425 0.356 0.028 0.066 0.099 0.156 0.339 0.341 0.378 162 Table 47 . Model estimates of number of lake trout deaths (x1000) due to commercial fishing mortality in northern Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 1.127 0.538 3.762 36.359 14.166 9.350 2.748 0.989 1.410 1.480 3 9.358 14.186 6.388 94.713 141.614 41.216 19.186 7.809 5.619 12.181 4 19.144 67.123 77.378 24.052 50.271 102.989 42.132 38.190 28.923 31.397 5 45.470 18.398 38.959 20.534 0.019 0.199 4.341 9.717 21.384 22.778 6 14.590 20.657 4.999 4.729 0.001 0 0.002 0.454 2.827 8.287 7 7.177 4.303 4.561 0.559 0.001 0 0 0 0.080 0.843 8 0.683 3.089 1.476 1.221 0.002 0 0 0 0 0.026 9 0.050 0.180 0.793 0.312 0.004 0 0 0 0 0 10 0.003 0.009 0.035 0.185 0.002 0 0 0 0 0 1 1 0 0.001 0.004 0.019 0.008 0 0 0 0 0 12 0 0 0 0.002 0.001 0 0 0 0 0 13 0 0 0 0 0 0 0 0 0 0 14 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 0 0 13 o 0 0 0 0 0 0 0 0 0 19 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0 0 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 Total 97.603 128.484 138.357 182.685 206.090 153.755 68.409 57.160 60.244 76.992 163 Table 48. Model estimates of number of lake trout deaths (x1000) due to sea lamprey- induced mortality in northern Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 3 38.562 30.564 13.093 14.639 25.010 17.251 7.088 47.098 41.917 28.730 4 8.217 16.287 18.698 0.419 0.937 3.813 0.869 26.524 27.329 5.610 5 39.791 7.128 12.734 0.569 0.001 0.023 0.640 9.969 20.058 17.292 6 34.528 19.601 3.652 0.325 0 0 0.001 1.097 4.828 20.007 7 22.053 5.229 4.174 0.056 0 0 0 0.001 0.254 2.661 8 4.320 7.278 2.441 0.256 0 0 0 0 0 0.182 9 0.406 0.542 1.648 0.086 0.002 0 0 0 0 0 10 0.044 0.055 0.141 0.100 0.001 0 0 0 0 0 1 1 0.005 0.006 0.015 0.011 0.007 0 0 0 0 0 12 0 0.001 0.002 0.001 0.001 0.001 0 0 0 0 13 0 0 0 0 0 0 0 0 0 0 14 0 0 0 0 0 0 0 0 0 0 15 0 0 0 0 0 0 0 0 0 0 16 0 0 0 0 0 0 0 0 0 0 17 0 0 0 0 0 0 0 0 0 0 13 0 0 0 0 0 0 0 0 0 0 19 0 0 0 0 0 0 0 0 0 0 20 0 0 0 0 0 0 0 0 0 0 >20 0 0 0 0 0 0 0 0 0 0 Total 147.927 86.691 56.599 16.463 25.959 21.088 8.598 84.688 94.387 74.482 164 Table 49. Model estimates of instantaneous rates of total mortality (year'l) for lake trout in southern main basin Lake Huron (MH-3/4/5). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 2 0.322 0.424 0.369 0.363 0.348 0.365 0.432 0.432 0.391 0.562 3 0.186 0.317 0.236 0.250 0.221 0.257 0.325 0.308 0.278 0.413 4 0.192 0.411 0.283 0.339 0.272 0.331 0.441 0.337 0.348 0.382 5 0.251 0.519 0.434 0.486 0.404 0.387 0.612 0.374 0.431 0.436 6 0.259 0.534 0.469 0.533 0.440 0.385 0.641 0.377 0.436 0.468 7 0.265 0.537 0.470 0.586 0.458 0.396 0.645 0.409 0.418 0.509 8 0.275 0.542 0.474 0.637 0.477 0.412 0.654 0.446 0.402 0.548 9 0.279 0.543 0.473 0.664 0.485 0.421 0.656 0.468 0.391 0.569 10 0.283 0.545 0.473 0.687 0.492 0.429 0.658 0.485 0.383 0.587 11 0.287 0.546 0.473 0.712 0.500 0.437 0.660 0.503 0.374 0.606 12 0.286 0.546 0.473 0.706 0.498 0.435 0.660 0.499 0.376 0.602 13 0.284 0.545 0.474 0.695 0.495 0.431 0.659 0.490 0.381 0.593 14 0.287 0.546 0.473 0.712 0.500 0.437 0.660 0.504 0.374 0.607 15 0.283 0.545 0.474 0.686 0.492 0.428 0.659 0.483 0.384 0.586 16 0.290 0.547 0.472 0.727 0.504 0.442 0.661 0.515 0.368 0.618 17 0.290 0.547 0.472 0.727 0.504 0.442 0.661 0.515 0.368 0.618 18 0.290 0.547 0.472 0.727 0.504 0.442 0.661 0.515 0.368 0.618 19 0.290 0.547 0.472 0.727 0.504 0.442 0.661 0.515 0.368 0.618 20 0.290 0.547 0.472 0.727 0.504 0.442 0.661 0.515 0.368 0.618 >20 0.210 0.502 0.436 0.522 0.415 0.381 0.606 0.398 0.394 0.475 165 Table 50. Model estimates of instantaneous rates of total mortality (year'l) for lake trout in central main basin Lake Huron (MB-2). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 2 0.318 0.319 0.319 0.319 0.319 0.319 0.319 0.319 0.319 0.320 3 0.376 0.369 0.242 0.180 0.289 0.207 0.187 0.228 0.335 0.373 4 0.369 0.335 0.228 0.164 0.284 0.217 0.198 0.233 0.348 0.429 5 0.491 0.343 0.273 0.229 0.369 0.350 0.309 0.326 0.426 0.512 6 0.536 0.346 0.340 0.302 0.410 0.458 0.352 0.405 0.472 0.540 7 0.505 0.340 0.349 0.309 0.391 0.453 0.324 0.399 0.454 0.502 8 0.482 0.341 0.371 0.326 0.393 0.460 0.338 0.399 0.457 0.499 9 0.465 0.343 0.400 0.351 0.400 0.481 0.354 0.411 0.466 0.497 10 0.468 0.345 0.396 0.351 0.409 0.475 0.379 0.395 0.458 0.471 11 0.462 0.339 0.393 0.346 0.389 0.473 0.331 0.406 0.452 0.472 12 0.470 0.347 0.398 0.353 0.417 0.476 0.400 0.390 0.461 0.471 13 0.465 0.342 0.395 0.349 0.400 0.475 0.359 0.400 0.456 0.472 14 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 15 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 16 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 17 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 18 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 19 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 20 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 >20 0.475 0.351 0.401 0.357 0.434 0.478 0.441 0.381 0.467 0.470 166 Table 51. Model estimates of instantaneous rates of total mortality (year") for lake trout in northern main basin Lake Huron (MH-l). Year Age 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 0.666 2 0.323 0.327 0.336 0.410 0.388 0.352 0.328 0.322 0.322 0.322 3 0.420 0.455 0.709 1.228 0.987 0.646 0.302 0.452 0.453 0.303 4 0.645 0.962 1.765 7.106 5.421 2.726 0.855 0.631 0.609 0.471 5 1.021 1.357 2.450 9.513 7.298 3.839 1.206 0.919 0.754 0.800 6 1.511 1.611 2.730 8.514 6.657 3.973 1.342 1.277 0.879 1.257 7 1.192 1.199 1.962 5.643 4.451 2.754 1.056 0.998 0.867 1.001 8 1.845 1.581 2.397 5.527 4.512 3.288 1.487 1.433 1.302 1.637 9 1.818 1.509 2.221 4.657 3.859 2.975 1.450 1.407 1.348 1.628 10 1.726 1.338 1.886 2.918 2.542 2.340 1.269 1.331 1.285 1.553 11 1.736 1.334 1.829 2.993 2.607 2.348 1.390 1.314 1.465 1.587 12 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 13 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 14 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 15 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 16 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 17 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 18 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 19 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 20 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 >20 1.793 1.391 1.887 3.051 2.664 2.405 1.447 1.371 1.522 1.644 LIST OF REFERENCES LIST OF REFERENCES Baldwin, N.S., R.W. Saalfeld, M.A. Ross, and HI. Buettner. 1979. Commercial fish production in the Great Lakes 1867-1977. Great Lakes Fishery Commission Technical Report No. 3. Bence, IR, A. Gordoa, and J .E. Hightower. 1993. Influence of age-selective surveys on the reliability of stock synthesis assessments. Canadian Journal of Fisheries and Aquatic Sciences 50: 827-840. Bergstedt, RA. and C .P. Schneider. 1988. Assessment of sea lamprey (Petromyzon marinas) predation by recovery of dead lake trout (Salvelinus namaycush) fi'om Lake Ontario, 1982-85. Canadian Journal of Fisheries and Aquatic Sciences 45: 1406-1410. Berst, AH. and GR. Spangler. 1973. Lake Huron: the ecology of the fish community and man’s effects on it. Great Lakes Fishery Commission Technical Report No. 21. Budd, J .C., and FBI. Fry. 1960. Further observations on the survival of yearling lake trout planted in South Bay, Lake Huron. Canadian Fish Cultun'st 26: 7-13. Budd, J .C., F.E.J. Fry, and P.S.M. Pearlstone. 1969. Final observations on the survival of planted lake trout in South Bay, Lake Huron. Journal of the Fisheries Research Board of Canada 26: 2413-2424. Christie, W.J. 1974. Changes in the fish species composition of the Great Lakes. Journal of the Fisheries Research Board of Canada 31: 827-854. Coble, D.W., R.E. Bruesewitz, T.W. Fratt, and J.W. Scheirer. 1990. Lake trout, sea lampreys, and overfishing in the upper Great Lakes: a review and reanalysis. Transactions of the American Fisheries Society 119: 985-995. Cochran, RA. 1985. Size-selective attack by parasitic lampreys: consideration of alternate null hypotheses. Oecologia 67: 137-141. 167 168 Deriso, R.B., T.J. Quinn 11, and RR. Neal. 1985. Catch-age analysis with auxiliary information. Canadian Journal of Fisheries and Aquatic Sciences 42: 815-824. DesJardine, R.L., T.K. Gorenflo,'N.R. Payne, and JD. Schrouder. 1995. Fish- community objectives for Lake Huron. Great Lakes Fishery Commission Special Publication 95-1. Doubleday, W.G. 1976. A least squares approach to analysing catch at age data. International Commission for the Northwest Atlantic Fisheries Research Bulletin 12: 69-81. Ebener, M.P., J. Selgeby, M. Gallinat, and D. Donofiio. 1989. Methods for determining total allowable catch of lake trout in the 1842 treaty-ceded area within Michigan waters of Lake Superior, 1990-1994. Great Lakes Indian Fish and Wildlife Commission Administrative Report 89-11. Eschmeyer, PH. 1957. The near extinction of the lake trout in Lake Michigan. Transactions of the American Fisheries Society 85: 102-119. Eshenroder, R.L., R.A. Bergstedt, D.W. Cuddy, G.W. Fleischer, C.K. Minns, T.J. Morse, N.R. Payne, and R.G. Schorfhaar. 1987. Report of the St. Mary’s sea lamprey task force. Great Lakes Fishery Commission. Ann Arbor, Michigan. Eshenroder, R.L., D.W. Coble, R.E. Bruesewitz, T.W. Fratt, and J .W. Scheirer 1992. Decline of lake trout in Lake Huron. Transactions of the American Fisheries Society 121: 548-554. Eshenroder, R.L. and J .F . Koonce. 1984. Recommendations for the standardizing the reporting of sea lamprey marking data. Great Lakes Fishery Commission Special Publication 84-1. Eshenroder, R.L., N.R. Payne, J.E. Johnson, C.A. Bowen II, and MP. Ebener. 1995. Lake trout rehabilitation in Lake Huron. Journal of Great Lakes Research 21 (supplement 1): 108-127. Farmer, GI. and F.W.H. Beamish. 1973. Sea lamprey (Petromyzon marinas) predation on freshwater teleosts. Journal of the Fisheries Research Board of Canada 30: 601-605. F ournier, D. and GP. Archibald. 1982. A general theory for analyzing catch at age data. Canadian Joumal of Fisheries and Aquatic Sciences 39: 1195-1207. Francis, G.R., J.J. Magnuson, H.A. Regier, and DR. Talhelm. 1979. Rehabilitating Great Lakes ecosystems. Great Lakes Fishery Commission Technical Report No. 169 37. Fry, F.E.J. 1953. The 1944 year class of lake trout in South Bay, Lake Huron. Transactions of the American Fisheries Society 82: 178-192. Greig, L., D. Meisner, and G. Christie. 1992. Manual for the management protocol for the implementation of integrated management of sea lamprey in the Great Lakes Basin. Version 1.1. Great Lakes Fishery Commission Report. Hansen, M.J. 1994. Dynamics of the recovery of lake trout (Salvelinus namaycush) in US. waters of Lake Superior. Doctoral dissertation. Michigan State University, East Lansing, MI. Hatch, R.W. 1983. Population dynamics and species interactions, p. 4-9. In R.L. Eshenroder, T.P. Poe, and CH. Olver, editors. Strategies for rehabilitation of lake trout in the Great Lakes: proceedings of a conference on lake trout research, August 1983. Great Lakes Fishery Commission Technical Report No. 40. Healey, MC. 1978. The dynamics of exploited lake trout populations and implications for management. Journal of Wildlife Management 42: 307-328. Hilbom, R. and CI. Walters. 1992. Quantitative fisheries stock assessment: choice, dynamics, and uncertainty. Chapman and Hall, New York, NY. Hile, R. 1949. Trends in the lake trout fishery of Lake Huron through 1946. Transactions of the American Fisheries Society 76: 121-147. Jacobson, L.D. 1989. New approach to measuring the frequency of sea lamprey wounds in fish stocks in the Great Lakes. North American Journal of Fisheries Management 9: 23-33. Johnson, J .E. and J .P. VanAmberg. 1995. Evidence of natural reproduction of lake trout in western Lake Huron. Journal of Great Lakes Research 21 (supplement 1): 253-259. Johnson, J.E., G.M. Wright, D.M. Reid, C.A. Bowen, II, and NR. Payne. 1995. Status of the cold-water fish community in 1992, p. 21 -71 In M. P. Ebener, editor. The state of Lake Huron 1n 1992. Great Lakes Fishery Commission Special Publication 95 2. King, E.L. Jr. 1980. Classification of sea lamprey (Petromyzon marinas) attack marks on Great Lakes lake trout (Salvelinus namaycush). Canadian Journal of Fisheries and Aquatic Sciences 37: 1989-2006. 170 Koonce, J .F ., R.L. Eshenroder, and G.C. Christie. 1993. An economic injury level approach to establishing the intensity of sea lamprey control in the Great Lakes. North American Journal of Fisheries Management 13: 1-14. Law, AM. and W.D. Kelton. 1982. Simulation modeling and analysis. McGraw-Hill, New York, NY. Lawrie, AH. 1970. The sea lamprey in the Great Lakes. Transactions of the American Fisheries Society 99: 766-775. Martin, NV. and CH. Olver. 1980. The lake charr, Salvelinus namaycush, p. 205-277. In E.K. Balon, editor. Charrs, salmonid fishes of the genus Salvelinus. Dr. W. Junk bv Publishers, The Hague, The Netherlands. Megrey, BA. 1989. Review and comparison of age-structured stock assessment models fi'om theoretical and applied points of view. American Fisheries Society Symposium 6: 8-48. Mema, J .W., J .C. Schneider, GR. Alexander, W.D. Alward, and R.L. Eshenroder. 1981. Manual of Fisheries Survey Methods. Michigan Department of Natural Resources Fisheries Management Report No. 9. Methot, RD. 1990. Synthesis model: an adaptable framework for analysis of diverse stock assessment data, p. 259-277. In L. Low, editor. Proceedings of the symposium on applications of stock assessment techniques to Gadids. International North Pacific Fisheries Commission Bulletin 50. Miller, D.M. 1984. Reducing transformation bias in curve fitting. The American Statistician 38: 124-126. Morman, RH. 1979. Distribution and ecology of lampreys in the lower peninsula of Michigan, 1957-75. Great Lakes Fishery Commission Technical Report No. 33. Morse, T.J., R.J. Young, and J .G. Weise. 1995. Status of sea lamprey populations in 1992, p. 101-107. In MP. Ebener, editor. The state of Lake Huron in 1992. Great Lakes Fishery Commission Special Publication 95-2. Pycha, R.L. and GR. King. 1975. Changes in the lake trout population of southern Lake Superior in relation to the fishery, the sea lamprey, and stocking, 1950-70. Great Lakes Fishery Commission Technical Report No. 28. Rakoczy, GP. 1992. Charter boat catch and effort from the Michigan waters of the Great Lakes, 1991. Michigan Department of Natural Resources Fisheries Technical Report No. 92-9. 171 Rakoczy, GP. and RD. Rogers. 1991a. Charter boat catch and effort from the Michigan waters of the Great Lakes, 1989. Michigan Department of Natural Resources Fisheries Technical Report No. 91-11. Rakoczy, GP. and RD. Rogers. 1991b. Charter boat catch and efl‘ort from the Michigan waters of the Great Lakes, 1990. Michigan Department of Natural Resources Fisheries Technical Report No. 91-12. Rakoczy, GP. and RF. Svoboda. 1993. Charter boat catch and efl‘ort from the Michigan waters of the Great Lakes, 1992. Michigan Department of Natural Resources Fisheries Technical Report No. 93-2. Rakoczy, GP. and RF. Svoboda. 1994a. Sportfishing catch and effort fi'om the Michigan waters of Lakes Michigan, Huron, Erie, and Superior, April 1, 1992- March 31, 1993. Michigan Department of Natural Resources Fisheries Technical Report No. 94-6. Rakoczy, GP. and RF. Svoboda. 1994b. Charter boat catch and effort from the Michigan waters of the Great Lakes, 1993. Michigan Department of Natural Resources Fisheries Technical Report No. 94-7. Ricker, W.E. 1975. Computation and interpretation of biological statistics of fish populations. Fisheries Research Board of Canada Bulletin 191. Rybicki, R.W. 1990. Survival rates of 1- and 2-year-old hatchery-reared lake trout in the west arm of Grand Traverse Bay, Lake Michigan. Michigan Department of Natural Resources Research Report No. 1978. SAS Institute. 1985. SAS° user’s guide: statistics, version 5 edition. SAS Institute Inc., Cary, NC. Schneider, C.P., R.W. Owens, R.A. Bergstedt, and R. O’Gorman. In press. Predation by sea lampreys (Petromyzon marinus) on lake trout (Salvelinus namaycush) in southern Lake Ontario, 1982-92. Canadian Journal of Fisheries and Aquatic Sciences. Schorflraar, R.G. and J .W. Peck. 1993. Catch and mortality of non-target species in lake Whitefish trap nets in Michigan waters of Lake Superior. Michigan Department of Natural Resources Fisheries Research Report No. 1974. Seber, G.A.F. and C. J. Wild. 1989. Nonlinear Regression. John Wiley and Sons, New York, NY. 172 Shetter, D8. 1949. A brief history of the sea lamprey problem in Michigan waters. Transactions of the American Fisheries Society 76: 160-176. Smith, B.R., H.J. Buettner, and R. Hile. 1961. Fishery statistical districts of the Great Lakes. Great Lakes Fishery Commission Technical Report No. 2. Smith, BR. and J .J . Tibbles. 1980. Sea lampreys (Petromyzon marinas) in Lakes Huron, Michigan, and Superior: history of invasion and control, 1936-78. Canadian Journal of Fisheries and Aquatic Sciences 37: 1780-1801. Smith, SH. 1972. Factors of ecologic succession in oligotrophic fish communities of the Laurentian Great Lakes. Journal of the Fisheries Research Board of Canada 29: 717-730. Swink, W.D. 1990. Effect of lake trout size on survival after a single sea lamprey attack. Transactions of the American Fisheries Society 119: 996-1002. Swink, W.D. 1991. Host-size selection by parasitic sea lampreys. Transactions of the American Fisheries Society 120: 637-643. Swink, W.D. and L.H. Hanson. 1989. Survival of rainbow trout and lake trout aficr sea lamprey attack. North American Journal of Fisheries Management 9: 35-40. Technical Fisheries Review Committee. 1992. Status of the fishery resource-1991. A report by the Technical Fisheries Review Committee on the assessment of lake trout and lake whitefish in treaty-ceded waters of the upper Great Lakes: state of Michigan. Mimeo. Rep. 87p. Wisconsin State/Tribal Technical Committee. 1984. Recommended lake trout harvest for the Apostle Islands region of Lake Superior for the 1985, 1986, and 1987 fishing years. Internal report. "lllllllllllllllllllll