OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIAL§z P‘Iace in book return to team charge from c1 rculation race! ‘7" ".4 - Am . " ,1 ’ ‘ may! ,,'..',-' d . I 1,.- DEFINING ANGLING QUALITY AND ESTIMATING THE DEMAND FOR MICHIGAN'S 1976 GREAT LAKES SALMONID AND NON-SALMONID SPORT FISHERIES BY Charles Sheldon Korson 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 1979 ABSTRACT DEFINING ANGLING QUALITY AND ESTIMATING THE DEMAND FOR MICHIGAN'S 1976 GREAT LAKES SALMONID AND NON-SALMONID SPORT FISHERIES BY Charles Sheldon Korson Angler demand and values are largely dependent upon the "quality" provided. Consumers apparently select any produCt, a particular dinner wine for example, based on some set of attributes distinguishing it from other simi- lar products (wines). Anglers apparently have a similar classification system to distinguish between various fish- ing sites, based on similar sets of characteristics or at- tributes. If anglers believe the important attributes of any two sites are the same, they will usually visit only the more convenient site. This principle was used to de- fine the different kinds of Great Lakes salmonid and non- salmonid angling (respectively) available in Michigan, based upon seasonal surveys of 1976 angling. The results indicated that catch rate and species composition were apparently the most important attributes to Great Lakes open-water anglers. Anglers fishing for anadromous fish consider catch rates, lake throughways, publicity, and regulations of primary importance. This technique of product enumeration rovides the basis for intensive Charles Sheldon Korson demand and supply analysis. Instead of estimating the much more general demand for salmonid fishing as a whole, the demand for each different component is studied separ- ately, producing much more precise estimates of angler behavior. Dedicated to my parents ii AC KNOWLEDGMENTS 1 would like to express thanks to the Michigan Sea Grant Program for providing financial support throughout my graduate program and for making this research project possible. I am especially grateful to my graduate advi- sor and research supervisor, Dr. Daniel R. Talhelm, for his patience and invaluable guidance during the course of this study. I would also like to thank both Dr. Charles R. Lis- ton and Dr. Milton H. Steinmueller for serving as members of my graduate committee. For his assistance with much of the computer program- ming work, Scott Jordan deserves special recognition. If it were not for Scott's unselfish help and cooperation, this study would have taken much longer to complete. Much appreciation for his help with the labeling of certain figures is extended to Joel Soba Fawumi. Most importantly, much of the credit for my present accomplishments goes to my mother and father. I cannot adequately express my thanks and appreciation for their undying moral support, encouragement, confidence in my ability, and financial assistance. I feel very fortunate to have parents as dear as mine, and who always respected iii my decisions and stood behind me throughout my educational career. iv TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . CHAPTER I INTRODUCTION . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . Objectives . . . . . . . . . . . . . . Great Lakes Fishery Resources - Values History . . . . . . . . . . . . . . II LITERATURE REVIEW . . . . . . . . . . . III THEORY . . . . . . . . . . . . . . . . . 1. Demand . . . . . . . . . . . . . . 2. Supply . . . . . . . . . . . . . . 3. Product Classification and Angling Quality . . . . . . . . . . . . I‘7 METHODS O C O O O O O C O O O O O O D O 1. Product Classification Analysis . General Description . . . . . . Data Collection . . . . . . . . Input Data and Organization . . Specific Procedures . . . . . . 2. Price Equations . . . . . . . . . 3. Demand Analysis . . . . . . . . . General Description . . . . . . Specific Procedures . . . . . . V RESULTS AND DISCUSSION . . . . . . . . . 1. Product Classification Analysis . Salmonid Angling in the Great Lakes Angling for Anadromous Fish in Streams . . . . . . . . . . . 00H 0\ 48 Page Non-Salmonid Angling in the Great Lakes 0 O O I O I I O O O O O O O O O 67 2. Demand Analysis . . . . . . . . . . . . . 75 an. General Description . . . . . . . . . r Angling for Great Lakes Salmonids . . . 75 Angling for Anadromous Fish . . . . . . 80 Angling for Great Lakes Non-Salmonids . 84 VI CONCLUDING OBSERVATIONS . . . . . . . . . . . . 89 APPENDICES A DEPARTMENT OF NATURAL RESOURCES SURVEY QUESTIONNAIRE . . . . . . . . . . . . . . . . . 92 B MISCELLANEOUS INFORMATION . . . . . . . . . . . 96 LIST OF REFERENCES . . . . . . . . . . . . . . . . . 129 vi LIST OF TABLES Table Page 1. The Names of the Counties in Michigan . . . . 8 2. The Individual Species or Groups of Fish Whose Catch Rates are Utilized in the Angling Product Classification Procedure . 32 3. Price Equation Coefficients for Salmonid and Non-Salmonid Fishing . . . . . . . . . . . 44 B1. Various Species Combinations Tested as Alter- native Catch Rate Attributes in the Class- ification of Great Lakes Salmonid Angling . 96 BZ. Various Species Combinations Tested as Alter- native Catch Rate Attributes in the Class- iffication of Great Lakes Anadromous Angl- ing I I O O I O I O O I O O O O O O O I O O 99 B3. Various Species Combinations Tested as Alter- native Catch Rate Attributes in the Class- ification of Great Lakes Non-Salmonid Angling . . . . . . . . . . . . . . . . . . 100 B4. Non-Catch Rate Factors Utilized in the Inventory for Great Lakes Salmonid, Non- Salmonid and Anadromous Sport Fishing . . . lOl BS. Demand Equations for the Products Defined for Winter-Spring Great Lakes Salmonid Angling O O O O O O O O O O O I O O O O O O 106 B6. Demand Equations for the Products Defined for Summer Great Lakes Salmonid Angling . . . . 107 B7. Demand Equations for the Products Defined for Fall Great Lakes Salmonid Angling . . . 108 BB. Demand Equations for the Products Defined for Winter-Spring Great Lakes Anadromous Angling D I O O I O O O I O Q 0 O O O O t O 109 B9. Demand Equations for the products defined for Fall Great Lakes Anadromous Angling . . . . 110 Vii Table Page B10. Demand Equations for the Products Defined for Winter-Spring Great Lakes Non-Salmonid Angling . . . . . . . . . . . . . . . . . . 112 B11. Demand Equations for the Products Defined for Summer Great Lakes Non-Salmonid Angling O O I I I O O O O O I O I O O O O 0 ll 3 B12. Demand Equations for the Products Defined for Fall Great Lakes Non-Salmonid Angling . 115 B13. Final Product Identification Numbers for the 41 Coastal Counties Offering Great Lakes Salmonid Angling . . . . . . . . . . . . . 117 B14. Final Product Identification Numbers for the 52 Counties Offering Great Lakes Anadromous Salmon-Steelhead Angling . . . . 119 B15. Final Product Identification Numbers for the 41 Coastal Counties Offering Great Lakes Non-Salmonid Angling . . . . . . . . . . . 131 316. Catch Rate Factors Utilized in the Final Product Classifications for Great Lakes Salmonid Angling. Catch Rates are in Fish~ Per Angler Day (AD) . . . . . . . . . . . . 123 B17. Catch Rate Factors Utilized in the Final Product Classifications for Great Lakes Anadromous Angling. Catch Rates are in Fish Per Angler Day (AD) . . . . . . . . . 125 318. Catch Rate Facotrs Utilized in the Final Product Classifications for Great Lakes Non-Salmonid Angling. Catch Rates are in Fish Per Angler Day (AD) . . . . . . . . . 126 819. Common and Scientific Names of Fish SpSCieS o o o o o o o o o o o o o o o o o o 128 viii LIST OF FIGURES Figure Page 1. The Great Lakes and the 83 counties of Mich— igan (the names of each county are given in Table l) o o o o o o o o o o o o o o o o o o 7 2. Total sport catch for salmonids and non- salmonids in Michigan's Great Lakes and tributary streams from 1970 to 1976 . . . . 13 3. A normal supply curve (A) and a supply curve for angling with a given travel distance requirement (B) . . . . . . . . . . . . . . 20 4. An angling demand curve traced out by supply curves (prices) and quantities of angling for anglers residing at hypothetical loca- tions A and B . . . . . . . . . . . . . . . 23 5. Two attributes defining four specific dinner wine products . . . . . . . . . . . . . . . 25 6. A hypothetical graph showing the catch rates for steelhead at various counties (x). . . . 35 7. An example of six different catch rate attri— butes formed by simply varying salmonid species combinations . . . . . . . . . . . . 36 8. The relative importance to anglers of various angling attributes was judged by examining the excess amounts of money spent in rela- tion to the number of products defined. A curved frontier is formed by the "best" hy- potheses and the result of each hypothesis is represented by an "x" . . . . . . . . . . 41 9. Price curves, winter-spring period, showing the user cost of angling as related to tra- ‘fel distance 0 O O O O O O O O O O I O O O O 45 10. Price curves, summer period, showing the user cost of angling as related to travel dis- tance I I O O O O I I I I O O O O O O 0 O O 4 6 ix Figure Page 11. Price curves, fall period, showing the user cost of angling as related to travel dis- tance O O O O O O O O O I O O O O O O O O 0 4 7 12. The possible general forms for equation 01 = C + bPi + b'/Pi: (A) b < 0, b > 0; (B) b < O, b' < 0; (C) b > 0, b' > O; (D) b > 0, b' < 0 . . . . . . . . . . . . . 51 13. Identification key defining eight different kinds (products) of Great Lakes salmonid fishing during the 1976 winter-spring period . . . . . . . . . . . . . . . . . . 58 14. Identification key defining seven different kinds (products) of Great Lakes salmonid fishing during the 1976 summer period . . . 61 15. Identification key defining eight different kinds (products) of Great Lakes salmonid fishing during the 1976 fall period . . . . 62 16. Identification key defining nine different kinds (products) of Great Lakes anadromous fishing during the 1976 winter-spring period . . . . . . . . . . . . . . . . . . 64 17. Identification key defining thirteen differ- ent kinds (products) of Great Lakes anadro- mous fishing during the 1976 fall period . 66 18. Identification key defining ten different kinds (products)of Great Lakes non-salmonid fishing during the 1976 summer period . . . 68 19. Identification key defining twelve different kinds (products) of Great Lakes non-sal- monid fishing during the 1976 summer period . . . . . . . . . . . . . . . . . . 71 20. Identification key defining twelve different kinds (products) of Great Lakes non-sal- monid fishing during the 1976 fall period . 73 21. Demand curves for the eight salmonid angling products identified by any trout-any sal- mon catch rates in period I: H = high, M = moderate, and O = other . . . . . . . . 77 Figure Page 22. Demand curves for the seven salmonid angling products identified by any trout~aggregate salmon catch rates in period II:H = high, M = moderate, and O = other . . . . . . . . 78 23. Demand curves for the eight salmonid angling products identified by any trout-aggregate salmon catch rates in period III: H = high, M = moderate, and O = other . . . . . . . . 79 24. Demand curves for the nine anadromous angling products identified by steelhead catch rate-publicity- lake throughways- fly fish- ing regulations in period I . . . . . . . . 81 25. Demand curves for the thirteen anadromous angling products identified by aggregate salmon/steelhead catch rate (VH = very high)- lake throughways- snagging regula- tions- fly fishing regulations in period III....................82 26. Demand curves for the ten non-salmonid an- gling products identified by aggregate panfish-aggregate gamefish catch rates— special bass regulations in period I . . . 85 27. Demand curves for the twelve non-salmonid angling products identified by any panfish- aggregate gamefish catch rates- resort areas in period II . . . . . . . . . . . . 86 28. Demand curves for the twelve non-salmonid angling products identified by yellow perch-any other species catch rates- resort areas in period III . . . . . . . . . . . . 87 xi I . INTRODUCTION Overview Over the last thirty years, society has placed in- creasingly heavy demands on the recreational use of our renewable natural resources. All areas of public recrea- tion--fisheries, wildlife, national parks and forests, water-based resources, and many more--have in some way been affected by this preoccupation with outdoor recrea- tion. This is eSpecially true of Michigan's public re- sources, where recreational activity has expanded on all fronts. One such area of public recreation showing con- siderable growth in interest over the last decade has been Great Lakes sport fishing activities. Michigan's sport fishing industry is highly esteemed as one of the most popular forms of outdoor recreation available to the pub- lic. With current emphasis being placed on continually improving and developing opportunities for sport fishing, managers and planners must make more crucial decisions about the proper and intelligent use of the public's sport fishery resources. Questions regarding the direc- tion and efficiency of public programs must be raised to determine if limited public monies are properly invested for the greatest public good. Not only does this require assessing the biological and physical tradeoffs between various management and development plans, but it also necessitates assessing the desires of people for altern- ative recreation uses. Resource economists have commonly addressed the question of social preferences within the context of de- mand models for outdoor recreation. Demand functions express the willingness of users to exchange their per- sonal resources for the type of recreation in question. In addition, they estimate the expected use and recreation benefits associated with the planned develOpment Of vari- ous sites for recreational activities (Dwyer, 23 31,, 1977). These kinds of information help decide which al- ternative public investment decisions secure the greatest return to society, thus maximizing social welfare. The primary management goals of Operational efficiency and a social Optimum are attainable with further development of reliable economic evaluations. One aspect of social choice which has become more important as increasing pressure is put on recreational re- sources is understanding the exact nature of peoples de- mands for recreational Opportunities. What is it about a recreation site or facility that attracts user interest? Should certain features be developed at the expense of others which are somewhat less important to some users? Such questions can be answered more easily by determining how the "quality" of a recreation experience affects an individual's demand for outdoor recreation. It is widely recognized that site quality varies considerably. PeOple presumably select a particular recre- ation site based on certain characteristics or attributes which are most important to them. Important attributes for a salmon fishing experience may include the Species of salmon, success rate, fish size, shoreline access, stream or lake fishing, crowding, special regulations, and other factors. However, the problem is determining just which attributes peOple use to distinguish between alternative recreation sites. Managers, policymakers, and administra- tors would benefit by knowing which of these attributes are most important as a guide to better resource alloca- tion. Quality evaluation frequently is unreliable when differences between sites are based on some arbitrary qual- ity rating system. Personalized judgments may not reflect individuals tastes and preferences. The major goal of this study is to utilize a more satisfactory product classifica- tion approach for determining which recreation sites are alike or different, for the purpose of estimating the de- mands for Michigan's Great Lakes sport fishery resources. The primary objectives are to: (1) Group angling sites into various specific kinds (products) of angling recreation accord- ing to the attributes of angling that anglers apparently consider most important. This is very similar to a biological taxonomy where several levels of aggregation exist and there is some degree of personal interpretation (Talhelm, 1978b); (2) DevelOp separate product classification schemes for Michigan's Great Lakes Open-water salmonid, Open-water non-salmonid, and anadromous salmon- steelhead sport fisheries; (3) Estimate the demand for each of the specific component angling products by integrating the angling classification systems into an inten- sive supply and demand model. These separate but related analyses will generate useful information on: (a) angling supply: the prices or costs of the re- spective component products to anglers; (b) and the willingness of anglers to substitute one kind of angling for another. These demand equations, together with knowledge of present angler costs, serve as the basis for estimating values to anglers of alternative sport fishing management programs in Michigan. Specifically, equations may be used in a simulation model to estimate (1) the net social wel- fare Or benefits of sport fishing accruing to the public, and (2) more importantly, the changes in participation levels and benefits caused by changes in angling quality at certain locations. This will be useful for evaluating the efficiency and the prospects and desirability of many alternatives (Talhelm, 1973b). Future research will util- ize the information provided in this study to estimate the net benefits of salmonid and non-salmonid sport fishing programs. . This research is one phase of an overall project (sponsored in part by the Michigan Sea Grant Program) to document fisheries values for Michigan's Great Lakes. It is anticapted that this study partially fulfills one of the major project goals: to provide information needed for selecting optimal utilization of Great Lakes fisheries by documenting the benefits, costs and other impacts of potential management strategies. The remaining portion of this introduction contains a brief capsule on the valuation and history of Great Lakes fishery resources. The thesis then proceeds with a short review of the literature dealing with recreation demand and its relationship to quality evaluation (Chapter II). Chapter III will discuss the theory behind the definition of angling quality and supply and demand. Chapter IV ex- plains the specific procedures for the research. The re- sults and discussion sections are combined in Chapter V. 1Taken from a report on the 1975-76 Sea Grant Pro- gram in fisheries economics and marketing (Talhelm, 1975). Great Lakes Fishery Resources--Values and History The Great Lakes2 are recognized as supporting one of the outstanding fresh—water sport fisheries in the world. On the other hand, the concomitant regulation of commercial fishing has caused the once thriving commercial fishing industry to become increasingly depressed. No- where is this trend more prominent than in Michigan, where 41% of the total Great Lakes water area lies within state boundaries (Figure 1). As a result, Michigan is provided with 3,200 miles of coastline from which sportfishing and commercial fishing may be pursued. The success of sportfishing in Michigan is attribut- able to a number of factors. Primary among them is recog- nition by the Department of Natural Resources (DNR) that sportfishing is much more valuable to the public than com- mercial fishing, both in terms of social welfare (value) and in terms of the positive economic impact on the state. Various studies have documented the values accruing to the public for Michigan's Great Lakes sport fishery re- sources. Talhelm (1973b) and Ellefson (1973) estimated that the 1970 anadromous salmon-steelhead program produced net "social" benefits of approximately $24 million for licensed Michigan residents in 1970. This is to say that anglers would have been willing to contribute or pay this 2With a water surface area of about 95,000 square miles and over 9,000 miles of shoreline. . 3 game as swim mum wucsoo Sumo mo memc on“: :mmwnoflz mo mmwucooo mm man was wwxmq umwuo one 0;.— gzo .32.. .H whomwm )o. I $124 0 a» S a. o» u. 2 s = v N. .0 In n. On 00 in) S 3 — nu _ a _ n .I .39 on me i V (4.6 1 2 2 en 2 M é I we 8 .c 3 L 2 3 Sr) 3 2. 2. M o 3 3 co 3 co 3 «n m e a... a. S 3 n... 9 . V 3 no 2. S no a N a. _ no 8 3 as e. o p a . % v . 0 v on no a 0a.. A. 1 n. . , ... U .h 0. / /.,::Ws Q Mva a x cm a o o 0U. fie.... 0 re. a .N - ,0, v t a on x t 0' Na 5N s S .n h at / 9%.4) QOsQ / // “4“ / Xv nV eh Table 1. The Names of the Counties in Michigan Number County Number County Number County 1 Alcona 36 Iron 71 Presque Isle 2 Alger 37 Isabella 72 Roscommon 3 Allegan 38 Jackson 73 Saginaw 4 Alpena 39 Kalamazoo 74 St. Clair 5 Antrim 40 Kalkaska 75 St. Joesph 6 Arenac 41 Kent 76 ’Sanilac 7 Baraga 42 Keweenaw 77 Schoolcraft 8 Barry 42 Lake 78 Shiawassee 9 Bay 44 Lapeer 79 Tuscola 10 Benzie 45 Leelanau 80 Van Buren 11 Berrien 46 Lenawee 81 Washtenaw 12 Branch 47 Livingston 82 Wayne 13 Calhoun 48 Luce 83 Wexford l4 Cass 49 Mackinac 15 Charlevoix 50 MaComb l6 Cheboygan 51 Manistee l7 Chippewa 52 Marquette 18 Clare 53 Mason 19 Clinton 54 Mecosta 20 Crawford 55 Menominee 21 Delta 56 Midland 22 Dickinson 57 Missaukee 23 Eaton 58 Monroe 24 Emmet 59 Montcalm 25 Genesee 60 Montmorency 26 Gladwin 61 Muskegon 27 Gogebic 62 Newaygo 28 Grand Traverse 63 Oakland 29 Gratiot 64 Oceana 30 Hillsdale 65 Ogemaw 31 Houghton 66 Ontonagon 32 Huron 67 Osceola 33 Ingham 68 Oscoda 34 Ionia 69 Otsego 35 Iosco 70 Ottawa amount in 1970 to prevent the total loss of salmon-steel- head angling opportunity. (The total all-or-none value for Michigan's entire Great Lakes sport fishery is estimated at around $250 million per year in current dollars (Tal- helm, 1979).) Michigan's sport fishery has also exerted a consid- erable economic impact on the state's economy.< Some $20 million was spent for 1970 Great Lakes salmon-steelhead angling activities by licensed and unlicensed anglers (Talhelm, 1979);) Ellefson (1973) estimated that 60% of the $15.5 million in licensed fisherman expenditures for 1970 salmon-steelhead fishing were made at or near the lo- cation fished.ffThe remaining $6.1 million was spent for goods or services en route to a sites‘ The anadromous fish- ery was responsible for anglers' spending $400 thousand and for providing 21.5 full-time equivalent jobs in the Grand Traverse area of Michigan in 1970 (Kapetsky and Ryckman, 1973). The current total economic impact of Michigan's entire Sport fishery for Great Lakes fish is estimated at $200-300 million annually (Talhelm, 1979). The estimated values of Michigan's commercial fishing industry have all been considerably lower than those for sportfishing- Tyefisesslsconomic impastti§_sSEEWAEsd_at only around $16-20 million_annually (Talhelm, 1979). Fogle (1973) reported the 1971 dockside value for the en- tire commercial catch as $2.7 million. The social surplus or net all-or-none value for Michigan's 1976 commercial IO fishery was estimated at slightly more than $2.6 million (Ghanbari, 1977). ,» The DNR saw the potential benefit to society of expanding sportfishing programs. As a result, commercial fishing Operators were subjected to stricter regulations in an attempt to maintain a more limited but economically viable commercial fishing industry. Over the last ten years, the DNR has been restricting the species, locations and methods of harvest, and reducing the number of licensed commercial fishermen. By 1976, commercial Operators num- bered less than 150 while the number of sport fishermen had risen to some 1.2 million (Talhelm, 1979). Sport fishing is an extremely pOpular recreational activity in Michigan. The total Great Lakes sport harvest is estimated to be about three times (by weight) the pre- sent commercial catch (Talhelm, 1979). Anglers enjoy di- verse angling Opportunities from among a wide variety of gamefish populations, including abundant stocks of salmon, lake trout, steelhead, yellow perch, walleye, bass, pike, and others. However, prior to the mid-1960's and the in- stitution Of a salmonid program, the existence of a sport fishing program was seriously threatened. Stocks of many of these fishes had become severly depleted due to a num- ber of biological and man-induced factors. Primary among them was the invasion of the sea 1am- prey and alewife, apparently through the Welland Canal in the mid-1930's. Lake trout and other important stocks 11 were almost exterminated by extensive sea lamprey depreda- tions. By 1950, the lake trout fishery on Lakes Michigan and Huron was gone, and by 1962 the Lake Superior fishery was closed (Borgson and Tody, 1967). Disruption of preda- tor-prey equilibriums permitted smelt and especially ale- wife to rapidly explode into superabundance on Lakes Mich- igan and Huron beginning in 1955. Alewife further impacted stock levels by devastating small market fish like herring, chubs, perch, and recreational fish such as walleye and smallmouth bass (Tainter and White, 1977). In addition, alewife posed a threat to the spawning success of other species, and became a public nuisance when dead fish accumulated on beaches and in harbor areas. Once other fish populations were seriously depleted, commercial fish- ermen began harvesting smaller, less valuable species. By the mid-1960's, stocks of important commercial and rec- reational species had declined to dangerously low levels. Because fish stocks were so limited, the number of com- mercial fishermen fell from 1,100 in 1950 to approximately 300 in 1969 (Fogle, 1973). By the late 1960's, these ecological disruptions were brought under control through effective lamprey con- trol and intensive restocking programs. These facilitated the recovery of pOpulations of lake trout in Lakes Mich- igan and Huron, and permitted the reestablishment of steel- head, brown, brook, and hybrid (splake) trout. As expected these salmonids preyed on abundant pelagic species to 12 successfully prevent massive die-offs. More importantly, the DNR saw the possibility of utilizing the alewife as forage to produce sport or food fish of maximum interest and value (Tody and Tanner, 1966). This objective was realized when coho and Chinook salmon were first introduced in the middle 1960's. As a result, a highly successful and popular salmonid program has developed in the Great Lakes. Since 1970 the number of angler days spent fishing for Great Lakes salmonids has increased from 2 million to more than 3.3 million. Licensed sportsmen harvested around 23 million pounds of salmonids in 1976 (Jester, 1978 in Talhelm, 1978a). The resulting introductions and rehabilitation ef- forts have enabled most other gamefish species to increase their biomass levels as well. As a result of restoring the ecological balance of the Great Lakes, an outstanding and diversified sport fishery has developed in Michigan since the late 1960's. The total sport harvest for the salmonid and non-salmonid fisheries has sharply increased from 1970 through 1976 (Figure 2).3 3The 1974 harvest data are from DNR survey raw data reports and may not be accurately represented, but these were the only available catch statistics for this year. Therefore, one should interpret these numbers with extreme caution. 13 "Ill MICHIIllIOIS If FISH) 3 - 4 - 1 t l I // //. - 173535352225“ I _.... , r r / ;'//////////7':-~-.fi_ ,A’xx/x/xxzxz/xx/f "fly-f2, /////////////// f n m: ////////////////. z/////////////////// ////////////////////. I/////////////////// // Illélll/liillllll/III. 1970 71 72 73 74 75 75 VEII Figure 2. Total sport catch for salmonids and non-salmonids in Michigan's Great Lakes and tributary streams from 1970 thru 1976 (Jameson and Ellefson, 1970, 1971a, 1971b; Jamesen, 1972, 1973, 1974, 1976, 1977; Michigan DNR, 1978) . II . LITERATURE REVIEW The traditional approaches to estimating demand equa- tions have largely ignored the effects of quality upon the quantities of recreation taken by users. Most demand Stu- dies have focused on travel costs as the primary variable influencing visitation. The Hotelling (1949), and later the Trice and Wood (1958) and Clawson (1959) methods de- fined broad geographic zones around the recreation site, and assumed that the amounts of use by peOple from increas- ing distance zones were caused by the differences in money and time costs of visiting sites. The amounts of partici— pation associated with each level of travel cost are used to derive a demand curve for a single, unique recreation site. One serious drawback in the method is that not all parts of the same zone can be assumed equal. In reality any distance zone is comprised of a number of heterogen- eous areas, each of which may be situated near other rec- reation sites and have people with different tastes and incomes. The amounts of visitation to a recreation site are probably not only a function of distance, but also of prices of alternative forms of recreation, site character- istics, and other socioeconomic variables. The difficulty with this early travel cost approach 14 15 was overcome when various investigators began centering their observations on pOpulation centers rather than homo- geneous distance zones. This permitted one to consider all aspects of a recreation experience when predicting use and values of recreation resources. Brown, Singh, and Castle (1964); Boyet and Tolley (1966); Merewitz (1966); and Johnston and Pankey (1968) are some who used this method, but without the prices of alternatives. Studies by Talhelm (1972, 1973, 1976); Burt and Brewer (1971, 1974 in Dwyer, at 31., 1977); Cesario and Knetsch (1976); Cicchetti, Fisher, and Smith (1976); and Knetsch, Brown, and Henson (1976) extended the revised travel cost pro- cedure by including the influence Of substitute resources on recreation use. An additional dimension of demand which has received more recent attention is that the qualitative characteris- tics of the site are also influential in determining de- mand and usage. This is an important concept when consid- ering the prospective use and development of a number of different sites within a similar recreation system. Be- havioral studies by Hendee and Potter (1971), More (1973), and Hendee (1974) suggested that an important component of resource management was understanding the qualitative factors that motivate the behavior of recreational users. Talhelm (1973a) indicates that optimal management efforts can be achieved by estimating the demand and supply for different varieties or qualities of recreation. However, 16 the major difficulty is in selecting a variable to repre- snet quality in a way which eliminates the bias encountered when making subjective value judgments. Only a few demand studies have dealt explicitly with quality where subjective measurements are avoided. Stevens (1962), using a degree of quantitative measurement of the qualitative characteristics of a sports angling ex- perience, included angling success per unit of angling effort as a variable in demand functions. In this way, he estimated the total angling effort for original levels of angling success, and for some reduced success levels brought about by changes in water quality. Cesario (1975) and Cesario and Knetsch (1976) included an index Of inher- ent appeal or quality when specifying demand functions. Instead of subjectively ranking or measuring quality, an arbitrary scale rating the apparent utility or attrac- tiveness of a site was used to reflect the multitude of site characteristics. Johnston and Pankey (1968) evalua- ted the effects of quality upon total recreation use for seven California reservoirs. Demand functions were esti- mated by including certain reservoir size characteristics as one of a potential group of independent variables. Wennergren and Fullerton (1972) estimated recreational values attributable to qualitative differences in sites for sixteen Utah counties, but failed to identify the factors contributing to recreation quality. A different approach to interrelating demand and 17 recreation quality has been develOped in a number of studies by Talhelm (1972, 1973b, 1976). A consumer behavioral model is used to partition sites or counties into various kinds or varieties of recreation according to specific site attributes. The assumption is that if any sites are per- fect substitutes, users will only go to the least expen- sive (closer) site. Each variety is considered a different quality or "product" of recreation, analogous to one of the many makes and models of automobiles or other products. With the products defined, the demand and supply for each kind of recreation can be estimated. This permits calcula- tions of amounts of use and benefits to users of each pro- duct at each site, and the changes in user benefits and participation levels produced by changes in recreation at- tributes at specific sites over time. III . THEORY4 1. Demand A consumer normally responds to changes in the price of a product by either increasing or decreasing his con- sumption of the good in question. This behavior is tradi- tionally summarized in a demand curve for any market com- modity. Generally, a smaller quantity of a good is demanded by consumers at higher prices, and vica-versa. Demand is more formally defined as a schedule of the max- imum quantities purchased (per unit of time) at every pos- sible price over a specified period of time, if all other influences on demand remain constant. The demand for angling recreation is a similar price- quantity schedule, only here price is not determined by typical market forces. Rather the "price" of angling rep- resents the cost of angling in terms of the money and time resources required of the angler for participation in angling activities. Thus, the demand for angling relates the costs (prices) of participation to the amounts (quanti- ties) of participation per angler day. As the price of angling increases, the quantities of use (days) taken by 4This theory was first proposed by Talhelm (1972). 18 19 anglers will decrease, ceteris paribus. For instance, residents of Detroit commonly fish less for salmon than residents of southwestern Michigan primarily (presumably) because of higher expenses imposed by greater travel dis— tances. In essence, the demand for a particular type of angling is the willingness of anglers to exchange their resources for that kind of angling (Talhelm, 1973b). An angling demand curve illustrates the voluntary rate of exchange between "all other goods" (measured in terms of dollars) and angling-~the total preference for angling relative to other goods. An unbiased price-quantity relationship is estimated by seeing that the values of other factors affecting demand remain unchanged. However, the shapes and positions of demand curves, and therefore, participation rates, are sig- nificantly influenced by a number of important factors. An acceptable means for estimating recreational demand curves should take into account such influences as the availability and quality of alternative forms of angling, and the nonhomogeneity of tastes and income in the pOpula- tion (Dwyer, 33 33., 1977). Because of the countless var- iety of sites within a similar recreational system, it is misleading to consider only one site in isolation from others. To do so could lead to severe overestimates or underestimates of use. 20 2 - $11pr The price producers receive for a good largely de- termines how much is supplied in the marketplace. Gener- ally, more of a good is made available for sale when prices are higher. Such a price-quantity relationship is repre- sented by the supply curve A in Figure 3. It shows the given quantities of goods which will be forthcoming (per unit of time) at various prices during a specified time period, with other influences on production held constant. In a real sense, supply relfects the "abilityt,of society to produce a good (Talhelm, 1973b). A .— - 2 = a u, A G- a V in O ‘ a B- B 00mm ”sumun" Figure 3. A normal supply curve (A) and a supply curve for angling with a given travel distance re- quirement (B). The supply of angling recreation is somewhat differ- ent since consumers are in themselves the producers of angling activity. Anglers must trade their personal time and money resources in exchange for participation in angl- ing. The price of obtaining a unit of angling recreations is based on the monetary and time costs of transportation to the site. Thus, the "supply" of angling is a relation- ship between the costs (price) of going fishing and the amounts of angling available to anglers. This concept is referred to as the "supply of angling effort," since anglers must allocate time and money to travel to a partic- ular angling site (Talhelm, 1973b). Angling cost or "price" equations may be developed to express the price of angling (in dollars per angler day) as a direct function of distance (mileage) to the various angling sites. In other words, the variability in angler costs is primarily a function of angler residence and angling resource location. The major resource costs re- quired of the angler to produce angling are (l) the Oper- ating costs necessary for transportation; (2) the additional monetary costs of food and lodging; (3) the direct expendi- tures on fees, licenses, and equipment necessary for an- gling; and (4) the value of time spent to facilitate the angling experience. This last factor is important because visitation rates are not only influenced by monetary costs, 5Here a unit is the angler day, defined as any part of a day in which an angler fished. 22 but also by the time required to reach a site. A greater travel distance not only increases the time and money costs of travel, but it also reduces the recreation time remaining on a trip (Talhelm, 1972). Time spent angling is at the expense of Opportunities which must be foregone, either in alternative forms of recreation or in non-leisure activities. Thus, the value or Opportunity cost of time is most appropriately measured by the lowest current wage rate or potential wage rate one could have earned by re- allocating time and effort to more "productive" endeavors. For a given travel distance, any number of angling trips may be taken at a relatively constant cost per an— gler day (Talhelm, 1972, 1973b, 1976). Since the varia- tion in trip costs depends on the distance an angler must travel, a perfectly elastic or horizontal angling supply curve is defined for every possible travel distance (Fig- gure 3). Those anglers from more distant locations have higher supply curves (a lesser supply) because of the higher costs of traveling to a particular site. A reduc- tion in travel costs will lower the supply curve because availability is now greater. This explicit treatment of supply, in conjunction with origin-destination patterns of use, forms the basis for statistically estimating demand equations for various kinds of angling recreation. A point on a demand curve is formed by each of the respective price and quantity observations. Hence, with supply curves (prices) and 23 :: ‘ a :3 . g r.” mm (3) ¢ I . II I :: a . ' '34? 3 .' 4.9” «:5 : '4. :5 ' ' L of. sumv (a) : I ‘-.. 0 I ' I I I I '. : rnou a raou A QUANTITY (ANGLE! DAYS PER CAPITA) Figure 4. An angling demand curve traced out by supply curves (prices) and quantities of angling for anglers residing at hypothetical locations A and B. observed use of angling from two hypothetical locations (A and B), a demand function is traced out by points like a and b in Figure 4. 3. Product Classification and Angling Quality —— Analysts often assume that people prefer "high" quality over "low" quality angling, where this notion of quality rating implies making individual value judgments to distinguish between various angling sites. However, a site which is "high" quality for one individual may be con- sidered "low" quality by someone with different personal preferences. Clearly when users' tastes vary considerably, 24 there may be no clear consensus regarding how various attributes define recreation quality along a single scale. Even if there were a consensus, how can we judge the rela- tive importance to society of "good" as Opposed to "out- standing" recreation? A more useful concept is that different products have different attributes of varying importance to users. Consumers apparently select a product based on some set of attributes distinguishing it from other similar pro- ducts. For example, casual wine drinkers may select a dinner wine simply on the basis of wine color: red or white. Others may recognize four specific products by in- cluding taste as an additional attribute (Figure 5). Con- noisseurs may further distinguish between dinner wines by including such attributes as aroma groupings, major brand, and ageing period. The number of products or permutations become greater as attributes are divided into even finer divisions or details. Any hypothesized set of attributes defines a variety of specific goods within a general pro- duct group. Other examples may include the many different cuts of beef, or the many makes and models of automobiles. In an analogous fashion, angling resources can be characterized by enumerating the most important attributes of the recreation experience. A Great Lakes fishing site might be typified by the probability of catching salmonids, the species mix of salmonids, whether or not it has a pier, the extent to which publicity influences anglers' ATTRIBUTES PFORUCt. , Identification Wine Color Taste Number -———SWEET 1 RED h——-—-DRY 2 ——-—SWEET 3 WHITE L—— m 4 Figure 5. Two attributes defining four specific dinner wine products. expectations, and other attributes. Each particular per- mutation of attributes is used to define the different "products" (like the specific dinner wines) of angling recreation. For instance, one specific product may be "high" trout and salmon catch rates, no piers, and "high" publicity; another product may be "high" trout and salmon catch rates, no piers, and "low" publicity. With these three attributes (catch rate, piers, publicity), the dif- ferences in angling sites are classified into a multiple product set, where each site corresponds to a particular product and each product is a separate but related good. In this way, a general kind of angling is segmented into its specific component parts. Certain attributes or combinations of attributes may be hypothesized as being possibly important to anglers in 26 determining where they go to fish. Of course, the problem remains of choosing the best set of angling attributes, and therefore the best classification scheme to define angling quality. Since consumers normally consider the various characteristics and prices of goods before making their choices, anglers should display similar patterns of be- havior when deciding among different products of angling recreation. Rational behavior dictates that few will know- ingly pay more than necessary for a good with a given set of characteristics. Through a similar process, a defini- tion of angler quality is based on the idea that if anglers feel the angling afforded at any two sites is essentially the same, they will usually visit only the more convenient site. Otherwise, if anglers feel one of the sites is more desirable than the other, some anglers will travel farther or longer to visit that site. The degree to which anglers consider sites to be different is dependent on the magni- tude of anglers' willingness to travel farther than nec- essary to reach a supposedly identical site. A classification system showing large numbers of anglers traveling farther than necessary, or high levels of "excess" travel, indicates that some sites classified as identical products are actually considered different by many anglers. In this case, the hypothesized attri- butes do not adequately define the different products of angling available to anglers. An alternative hypothesis would be to reclassify similar sites as different products 27 using a different set of attributes. If a descriptive set of attributes sufficiently defines the various products of angling, then few anglers have reasons 33; to travel to the closest site of a given product (Talhelm, 1973b). Therefore, the most satisfactory or "best" classification system minimizes "excess" travel within each product cate- gory. By analytically trying alternative attribute combin- ations, the set most consistent with this expected pattern of behavior is found. As a result of minimizing "excess" travel, few anglers are incurring higher prices (in time and money costs) than necessary to reach an angling site within a particular product category, assuming perfect knowledge. In other words, the supply (price) of an angling product is the minimum (least expensive) price available for a pro- duct, since similar products are considered perfect sub- stitutes, and anglers have little reason to go to more distant sites. Rather than being an arbitrary classification based on 3 priori value judgments, this multiple-product approach depends on actual Observations of visitors' choices of angling sites. This consumer behavioral model reduces the bias associated with sentimental choices. However, this type of analysis is subject to the limitations imposed by the lack of perfect knowledge among anglers. These may include (1) the absence of information regarding the at- tributes Of various sites, thus causing users to mistakenly 28 go to less advantageous locations; (2) anglers may have multiple reasons for traveling to sites, such as for al- ternative forms of recreation, or to visit friends and relatives who reside near angling sites visited; and (3) attributes of sites hypothesized to be important may not be identified properly, especially when attributes are confounded with one another (i.e. catch rate and fish size). These difficulties are inevitable when conducting a discriminant study of this type. Yet it is not neces- sary that all users' recognize all angling products, just as all wine buyers can't recognize all types of wine; but the greatest proportion must act as if they perceive some difference, in order for analysts to detect the anglers' selection process. This technique of product definition not only pro- vides insight into angler behavior, but also is useful as the basis for intensive supply and demand analysis. Ra- ther than estimating the general demand for a general pro- duct group, such as all Great Lakes salmonid angling, this procedure generates more precise estimates of levels of participation and recreational values by analyzing compon- ent products. IV. METHODS 1. Product Classification Analysis General Description Seasonal product classification schemes are devel- oped for three types of sport fishing: Great Lakes open- water salmonid and non-salmonid fishing, and anadromous salmon-steelhead fishing. Specifically, each species group is examined independently for the winter-spring (Period I, January thru May), summer (Period II, June thru August), and fall (Period III, September thru December) angling seasons.6 Consequently, this study in effect consists of eight separ- ate analyses, one for each type of seasonal fishing. Since the procedural steps for each analysis are virtually ident- ical, only one period and type of fishing will be used as a general example to explain the following methodology. For this purpose, winter-spring angling for Great Lakes salmonids will serve as the model. On occasion, reference to other forms of fishing is necessary when circumstances require reporting certain facts or details. These cases 6Summer anadromous angling is not included in the analysis because of the limited number of anadromous an- gling Opportunities available during the summertime. This was reflected by the paucity of data collected for the summer survey of anadromous fishing activities. 29 30 will also apply to Period I. Data Collection The Department of Natural Resources (DNR) conducted a mail survey of over 40,000 sport fishermen licensed in Michigan about their 1976 seasonal angling activity.7 The seasonal survey samples were taken from only a very small segment of the licensed sport fishing population: 2% in winter-spring, and 1% each for summer and fall. One sample questionnaire showing the type of information requested from fishermen may be found in Appendix A.8 Data were gathered on such questions as origin-destination travel patterns, angling effort at various destinations (counties), and numbers of fish of various species caught by anglers. The data were collected and separated into~five distinct categories of fishing: (1) Great Lakes salmonid; (2) Great Lakes non-salmonid; (3) anadromous salmon and steelhead; (4) inland trout; and (5) inland non-trout. Within each category, anglers' responses were coded and stored on a permanent computer tape file according to season and type of fishing. A total of 19,109 useable records (responses) were collected for the five types of fishing. Data on 7There are 83 counties available for angling activi- ties. 8It should be stated at this point that the nature of the survey asked anglers to report only those species of fish they caught and the effort expended while doing so. Anglers were not asked about the species they actually fished for. This factor is considered when computing catch rate estimates for various species. 31 specific questions were extracted from each record for use in the product classification analysis. Input Data and Organization The relevant data utilized from the period I survey results were as follows: (1) county of residence, (2) site (county) fished; (3) number of days fished in the county; and (4) number of fish of various species caught. Data on catch and effort by species were used to estimate county catch rates for the five types of fishing. Depending on the Species category, catch rates were computed for indi- vidual species or groups of fishes for each of the 83 counties in Michigan (Table 2). The species comprising inland trout and non-trout fishing were used to compute trout, gamefish and panfish catch rate categories. While not included in the actual classification process, the in- land catch rate estimates identify locations for substi- tute kinds of angling. Catch rates were calculated by dividing the total catch for a certain Species or group of Species by the relevant number of angler days fished in a particular county. For each Species (or group of species), county effort included only those angler days in which anglers reported catching fish of that species. Otherwise when an angler failed to catch fish, the effort expended by that angler was excluded form the resultant calculations. For example, if in Berrien county, one respondent reported 32 Table 2. The Individual Species or Groups of Fish Whose Catch Rates are Utilized in the Angling Product Classification Procedure Type Of Fishing Species 1. Lake trout 2. Brown trout Great Lakes salmonid 3. Steelhead (rainbow) trout 4. Coho salmon 5. Chinook salmon 1. Steelhead trout Anadromous 2. Coho salmon 3. Chinook salmon White bass, crappie Yellow perch Bluegill, sunfish Small and largemouth bass Walleye pike, sauger Northern pike, muskellunge Other (smelt, carp) Great Lakes non-salmonid \lmU‘lubUONH Lake trout Rainbow trout Brook trout Brown trout Inland Trout ubUJNl-J Walleye Bass . Pike, muskie Gamefish: WNH Inland non-trout Yellow perch Bluegill, sunfish White bass, crappie 4. Rock bass bowl-4 000 Panfish: 33 catching no fish for one day's effort, and another reported catching two lake trout for two days effort, the catch rate for lake trout is g = 1.0 fish per angler day. This is done because of the manner in which the questionnaire was worded, since it iS difficult to determine what an angler fishes for in cases where no fish or even if fish are har- vested. Aggregate catch rates for various species associa- tions were computed using total catch for the species con- sidered and dividing by aggregate effort. The remaining sources of input data included infor- mation on anglers' residence, angling location, and parti- cipation in Great Lakes salmonid angling activities. From these data, origin-destination patterns of total angler effort for each county were generated for period I. In other words, this is the number of angler days Spent at the various counties offering salmonid angling Opportuni- ties by anglers originating from each location. Anglers origin sites included any one of the following 88 possi- bilities: all 83 Michigan counties, Wisconsin, Illinois, Indiana, Ohio, plus an additional category reserved for any other origin (i.e. other states, Canada). This origin- destination information was provided as an 88 x 83 matrix and stored on a permanent computer file. Specific Procedures The primary purpose of the classification procedure is to determine which attribute or sets of attributes 34 anglers' apparently use to distinguish between various angling Sites. Therefore, preliminary work was devoted to compiling a list of potential attributes which might be relevant in the decision-making process. Every county of- fering salmonid fishing was then inventoried for each of the attributes selected for analysis. Catch rate attributes were finalized after the re- liability and accuracy of the Species catch rate estimates were verified by professional fishery biologists. As a result of both the variability in annual angling success and the small seasonal samples, many catch rate estimates for certain counties were somewhat questionable and required adjustments based upon more reliable information. Follow- ing confirmation, county catch rate estimates for each in- dividual salmonid species were plotted on standard graph paper. Counties were then separated into various levels or sub-divisions, such as high, moderate, and low catch per unit of effort, on the basis of their relative distri- bution (Figure 6A). Many catch rate attributes were generated to test the importance to anglers of specific seasonal hypotheses regarding salmonid species mix. During period I, some anglers may select sites on the basis of the success rates of major species categories, such as the trout and salmon groups. Others may further discriminate within the trout category by considering the catch rates for a certain species or combination of species as being most important 35 3 I I I = l I I I I E 20_ 1 . 2 i 3 . 4 9 l l I h I I I o l I I = I I I S I I I g I l I I I l 10" I I I I I I I I I I I I _ : : : 5 I I I I I z I ‘ ‘ ‘I oII pix xImeIx 31x. x 0.2 0.4 0.6 0.8 1.0 1.2 1.4 CATCH PER UNIT 0F EFFORT(CE) Figure 6. A hypothetical graph showing the catch rates for steelhead at various counties (x). By changing the cut-Off points, two different catch rate attributes are created as follows: (A) catch rate levels initially designated as l = Low, 2 = Moderate, 3+4 = High; (B) catch rate levels alternatively designated as l = Low, 2+3 = Moderate, 4 = High. to them. For example, anglers may regard steelhead or lake trout catch rates independently, steelhead together with either lake trout or brown trout catch rates, or brown trout and aggregate steelhead/lake trout catch rates. A number of different catch rate attributes were created by first varying and recombining species into various asso- ciations (Figure 7). Depending on the species combination and levels of catch rate, the possible permutations for any catch rate attribute range anywhere from a simple 3- way breakdown (for a single species with 3 levels of catch lake trout Trout: steelhead brown trou- coho Salmon. Chinook Figure 7. 36 Any One Either Trout Salmon Any One Aggregate Trout Salmon Aggregate Either Trout Salmon Brown trout Either coho , or Chinook or Steelhead lake trout Any other Ch:200k salmoind lake trout Lake or tr. Brown trout steelhead Either Salmon Example of six different catch rate attributes formed by simply varying salmonid Species com- binations. 37 rate) to a 27-way system (three Species categories each with 3 levels of catch rate). Tables describing the many seasonal catch rate attributes formulated and tested for the three types of fishing may be found in Appendix B. Secondly, additional catch rate attributes are formed by simply redistributing counties into different levels by changing the arbitrary cut-off points on one or more graphs (Figure 6B). Here the counties in group three (high catch per unit of effort), although somewhat differentiated from moderate catch rate counties (group 2), may not truly be classified as "high" by some anglers. An alternative is to lump these counties into the moder- ate category, thus forming a different attribute. Other plausible attributes in addition to catch rate were chosen to more completely describe how anglers dis- criminate between angling sites. These included such factors as the availability of piers, publicity, and the presence of natural bays. A list with complete defini- tions of these other attributes utilized for each type of fishing is found in Table 4 of Appendix B. These selec- tions were made on the basis of personal judgment, sugges- tions by professionals, discussions among fellow students, or remarks by anglers themselves. Many other important factors, such as size of fish, public access, and crowding, were not selected because they are either confounded with other attributes or are difficult to assess on a county- wide basis. 38 A complete inventory is formed by coding each county according to the level of attributes found there. These results, combined with origin-destination use patterns and price data obtained from various price equations (described below), were then simultaneously integrated into a complex Fortran computer program which generated each classifica- tion scheme. Any hypothesis includes a catch rate attri- bute, and may include one or more of the non-catch rate attributes. Each hypothesis was judged by examining user travel patterns in relation to the number of Specific products generated by the classification procedure. Talhelm (1972, 1973b, 1976) proposed that if each product is properly de- fined, the magnitude of the number of users traveling Ifarther than necessary" to reach a more distant but hypothetically identical site will be minimized. Excess travel is minimum in the case where all Sites are unique, and maximum when sites are classified as being alike. Therefore, a reasonable criterion is one which minimizes excess travel within categories while using a "small" number of categories. Instead of measuring excess travel by counting the number of anglers traveling farther than necessary to reach each product, it was decided to measure total ex- cess expenditures incurred by users in traveling to more distant Sites. That is, the extra money and value of time Spent over and above what users could have incurred to 39 reach a hypothetically identical and less expensive (more 3convenient) Site. The percentage of total excess expendi- tures indicates the adequacy of each hypotheses. High percentages suggest that (l) a large proportion of anglers have reasons for incurring additional expenses, or (2) a small number of users are traveling much farther than nec- essary. All hypotheses were examined for their level of ex- cess expenditures. A number of hypotheses were investiga- ted in detail to determine possible explanations for an- gler behavior. Attribute combinations and subdivisions (levels) of attributes were varied depending on which al- ternative hypotheses appeared feasible. For instance, one hypothesis might examine the following attribute for salmonid fishing in period I: any trout and any salmon catch rates. A pattern may emerge whereby a large propor- tion of anglers are consistently paying high prices to reach one or two particular sites classified as "moderate" catch rate products for both species categories. Appar- ently, anglers' find these more distant sites sufficiently different to warrant spending excessive amounts of money. Closer inspection of the inventory may indicate that the farther Sites offer pier fishing while the closest site does not. Therefore, a second attribute (piers) is tested in combination with catch rates. By testing this alter- native hypothesis, the level of excess expenditures is reduced when these counties are separated into a distinct 40 product category ("moderate" catch rates and piers avail- able). Under this new classification scheme, anglers who were originally spending more than necessary for hypothet— ically identical products are traveling to those same sites, but to different products. Since the new products are only available at these distant locations, by defin- ition, anglers must be traveling to the closest counties for this new product. Therefore, anglers' supply prices are necessarily minimized. These procedures were followed until a satisfactory hypothesis was discovered. Various hypotheses were com— pared by plotting them on a graph where one axis has the number of products and the other total excess expenditures (Figure 8). A curved frontier is formed by the most "rea- sonable" hypotheses for various numbers of products de- fined. The final selection was made from among a few pos— sibilities for which excess expenditures were low, con- sidering the number of product categories. 2. Price Equations Price equations were developed to express angler travel costs as a function of travel time and distance (mileage) to the site. Costs attributable to fishing in- clude (l) the estimated value of time, and (2) expenditures on travel, equipment, fees and lodging. The equations used in this study are modifications of those estimated for a study Of Michigan's 1970 salmon-steelhead Sport fishery 41 EXCESS EXPENDITURES NUMBER DE PRODUCTS Figure 8. The relative importance to anglers of various angling attributes was judged by examining the excess amounts of money spent in relation to the number of products defined. A curve fron- tier is formed by the "best" hypotheses and the result of each hypothesis is represented by an "x". Talhelm, 1973b) and a 1972 inland lake study of boating and angling in Michigan (Talhelm, 1976). Expenditure data for the 1970 equations were obtained from mail survey questionnaires asking anglers to report their expenses traveling and in the area fished. It was assumed that they included only those expenses incurred for the pur- poses of angling. However, if some expenses were in- curred for purposes other than fishing, such as travel costs incurred partially for visiting friends or relatives, then it is likely that 1970 angling costs may be slightly 42 exaggerated. The angling price equation estimated for the 1972 study was based upon expenditure data collected by means of personal interviews of users. In this way, expenses for the purposes of fishing are determined more precisely. The 1972 equation was cOnsidered much more reliable for predicting true angling costs. However, it represented costs for angling on inland lakes rather than angling for Great Lakes fish. In the 1970 study, separate price equations were estimated for resident and non-resident salmonid and non- salmonid anglers. Upon closer examination of these equa- tions, it was discovered that the resident non-salmonid price curve (plotted from the equation) began to decline at distances greater than 450 miles, implying that an- gling costs are decreasing with increasing travel dis- tance. Normally one would expect angling costs to continue rising as distance increases. Therefore, to correct this problem and improve the 1970 non-salmonid price equation, the resident costs of non-salmonid angling at distances (in increments of 10 miles) of 0-450 miles were combined with the non-resident costs of non-salmonid angling at distances of 460-1000 miles, and a single 1970 non-salmonid angling cost equation applicable for both residents and non—residents was estimated. This new curve was used as the basis for calculating angling costs in 1976. 43 For this study, the 1970 price equations are re- vised to conform in part with the 1972 equation. This involved (1) graphically comparing each 1970 equation es- timated for resident and non-resident salmonid and non- salmonid anglers with the 1972 equation; (2) determining any differences between curves by examining the heights and slopes of curves at various travel distances; and (3) incorporating these differences by adding or subtract- ing specific constants to the 1970 price equations. Be- cause angling costs were probably somewhat overestimated in 1970, the intercepts for the 1970 curves are greater than the intercept value for 1972. Therefore, various constants were subtracted from the 1970 salmonid and non- salmonid price equations. Different constants were sub- tracted when (1) an origin and distination site happened to be the same; equivalent to zero mileage; or (2) when origin and destination Sites are different. The specific constants were 19 and 16 for salmonid fishing and 13 and 16 for non-salmonid fishing, reSpectively. In addition, to account for inflation over a four year period (1972-1976), an average rate of inflation applicable to angling expenditures is added to the updated 1970 price equations. Consumer price indices for various items pertinent to recreational activities were gathered to estimate this factor. Angling costs were converted to 1976 dollars by assuming a 50% inflationary rate. For each sampling period, the various resident and 44 non-resident price equations are illustrated by type of fishing in Figures 9, 10, and 11, respectively. The func- tional form for these price equations is: P = C + b Distance + b2(Dist)2 + b ln(Dist + l) (l) l 3 The Specific price equation coefficients are shown in Table 3. Table 3. Price Equation Coefficients for Salmonid and Non-Salmonid Fishing Salmonid Non-Salmonid Factor . . . ReSident and ReSident Non ReSident Non-Resident Distance .31078 .29347 .105815 (Distance)2 -.000259 —.000098 .0000044 ln(Distance+l) -.69l95 -l.0942 2.84127 Constants:a . Period 1 25.34822 40.2882 23.9028 Period 2 30.58148 19.4348 24.4153 Period 3 33.97938 38.7328 22.6254 aThe values 19 and 15 for salmonid fishing and 13 and 16 for non-salmonid fishing were subtracted from the specific constants given above to calculate the costs of angling for various travel distances. Hfit ZDDi PRICE (I PEI ANCIER I") g... .9 Q l_ 120‘ ID‘ ADJ Figure 9. 45 l l 100 200 300 I00 500 DISTANCE (IIIES) g 1 Price curves, winter-Spring period, showing the user cost of angling as related to travel dis- tance. 46 2404 200‘ PRICE (I PER ANCIEI III) ISO .. IZD " 80') 7 .“ 40- /’ I I 100 200 300 ICC 500 I DISTANCE (IIIESI Figure 10. Price curves, summer period, showing the user cost of angling as related to travel distance. 47 2404 E Q 32004 5 I ‘ - t , :3. .~ :3 0 3. . ‘160' " 33,- e9 / \ 69‘“) // 120q / / / I :0- . ‘ / , / 40- o 1 A J I I I 100 200 300 ICC 500 DISTANCEINILES Figure 11. Price curves, fall period, showing the user cost of angling as related to travel distance. 48 3. Demand Analysis General Description The final resource classifications define the spe- cific angling products for which demand is estimated. As mentioned previously, the demand analysis gives the advant- age of evaluating consumer preferences for specific com- ponents of a general recreational system. AS before, demand is analyzed independently for each of the eight types of seasonal fishing. Since the analytical procedures are virtually identical for each demand analysis, the fol- lowing description is applicable to all eight cases. Specific Procedures The initial information required for estimating the demand functions are (l) the number of angler days use of each angling product from every origin, and (2) the corres- ponding minimum (supply) prices of each product from every origin. In other words, these pairs of data represent the prices and amounts of use (angler days) at the closest county of a particular product for anglers originating from the same location. Each product provides 88 sets of quant- ities and prices, one set for each origin. Because only a small percentage of the licensed an- gling population were sampled in each season, the seasonal user information was adjusted to reflect quantity in terms of the total populations of each origin. Various expansion factors (provided by the DNR) were utilized to estimate the 49 total number of user days consumed by anglers. Specific- ally, origin user data were multiplied by values Of 115.0, 180.0, and 190.0 for periods I, II, and III, respectively. Quantity was subsequently expressed as numbers of angler days per 1,000 capita by dividing total use by the popula- tion Size of each origin. For this and other studies (Talhelm, 1973b, 1976), a correction was allowed for calculating the use at the closest counties of a given angling product. The amounts of use at those counties within 20 miles or 20% (whichever is greater) of the minimum distance to a particular angling product were combined with the observed amounts of use for that same product. The reasons for this allowance are (1) county to county distances may not necessarily repre- sent actual road distances for various anglers, and (2) anglers wishing to visit two angling products which are close to each other may find it cheaper to visit both Sites together rather than separately, even though one product is Slightly farther. In addition to the preceeding price and quantity variables, observations of the prices of general substi- tute forms of angling and one socioeconomic variable were accumulated for every origin. These included (1) the 333- 3333 price of angling in a county offering at least a "moderate" catch rate for inland trout angling; (2) the minimum price for inland panfish angling (yellow perch, bluegill, crappie) of moderate or better catch rate; 50 (3) the minimum price for inland gamefish angling (bass, walleye, pike, muskellunge)of moderate or better catch rate; (4) the minimum prices for Great Lakes salmonid, non-salmonid, and anadromous stream angling (respectively) of moderate or better catch rate; and (5) personal income per capita. The prices were determined from the price functions given earlier. The non-salmonid price function was used to calculate the respective prices for inland angling substitutes. With information from every origin on (1) the prices and quantities of use of each angling product, (2) the prices of various substitutes, and (3) income per capita, demand equations were estimated for each quality of an- gling. Each demand function is described by the general functional form in equation 2, Qi = bo + bi/Pi + biPi + bij + kak + bss (2) where Qi represents the number of visits per 1,000 capita at the least expensive location for product i; the inde- pendent variables are the minimum available price of pro- duct i (Pi), the prices of the other substitute Specific products (Pj), the prices of the relevant substitute gen- eral angling substitutes (Pk), plus income (S). The four possible forms for this equation are illustrated in Figure 12. The equations were purposely restricted so as to elim- inate any possibility of a positively sloped demand curve (Figure 12D). The demand equations for each specific product were 51 (A) , 0. (B) 0 (C) o. (D) D Figure 12. The possible general forms for equation 01 = C + bPi + b'/Pi: (A) b < 0, b' > 0; (B) b < 0, b' < 0; (C) b > 0, b > 0; (D) b > 0, b' < 0. 52 estimated using ordinary least squares regression analyses. Initially, each regression is run using the complete set of price variables possible for a given product. Through a step-wise elimination process, certain variables are pro- gressively omitted when (1) coefficients are negative, or (2) when specific positive coefficients cannot be retained for statistical reasons. These factors are discussed more completely later in this section. Upon eliminating any questionable variables, the regression was rerun and the process repeated until an acceptable demand relationship was found. In this way, a set of demand functions (one for each angling product) is estimated for each of the seasonal types of fishing analyzed. A typical demand function for period I Great Lakes salmonid angling is represented in equation 3, Q4 = -55.2+l§?7?i + .20P4 + .051?3 + .11P6 - .0011 (3) where, in this case, the estimated quantity of angling con- sumed (per 1,000 anglers) at product four for a given or- igin site is a function of the minimum price of product four, the prices of Specific substitute products three and Six, and county per capita income. Tables giving the de- mand equations for all of the angling products defined for each seasonal type of fishing are found in Appendix B. In any demand function, a positive coefficient or cross elasticity of demand indicates substitution between two goods. Two products are considered substitutes when changes in the price of one have a positive effect on the 53 quantity demanded of the other (i.e. (dxz/dpl)pl/x2 is always positive, where x2 is the quantity of good 2 de- manded). For example, as the price of meat increases (pl), consumers will tend to increase their consumption of poultry or fish (x2) instead (and decrease the quantity of meat demanded). In the case of equation 3, anglers will go to product four less as the price of product six be- comes less expensive (more available to anglers). Higher values for positive coefficients indicate that anglers are more willing to give up one product for another, either be- cause (1) anglers consider the products to be good substi- tutes for each other, so they visit the least expensive of the two; (2) product Six is preferred to four, so anglers switch when the opportunity presents itself; and (3) pro- duct four is generally too expensive relative to product six (Talhelm, 1973b). The rationale for including only those variables with positive coefficients is that specific products in the same general product group are normally substitutes (Talhelm, 1978b). Any negative coefficients, indicating complement- arity between products,9 were eliminated from the regres- sions because they were difficult to justify due to the nature of the classification scheme. Two angling products might be considered complements if they are located 9Two goods are complements when an increase (decrease) in the price of one decreases (increases) the quantity de- manded of the other. 54 directly adjacent to one another, since anglers visiting one product may also find it very convenient to visit a second product which is in close proximity. The only other logical explanation for a negative coefficient would be "learning": the use of site Y whets the appetite for the product at site X (Talhelm, 1978b). However, it is unlikely that quantity of use at Site X would increase as a result of a decrease in the price of site Y, Since a reduction in price effectively means the complement is located closer to a particular origin. As mentioned before, several independent variables were eliminated for specific Statistical reasons. First, and probably the most fundamentally serious, is a source of measurement error in the dependent variable. Because the sample of user origin information is not very inten- sive at any single county, the amount of information available for estimating the equations is both limited and subject to some degree of uncertainty. Frequently as few as ten out of 88 origins provided non-zero data points for any one product. The observations of zero use from some origins may also result from statistical error asso— ciated with sampling, therefore they cannot be arbitrarely excluded from the regressions. With the liklihood of ques- tionable data, it is difficult to justify supporting some twenty independent variables to explain the total variation in the dependent variable. Secondly, prices are correlated with product categories, either because (1) often a 55 particular product is represented by one unique county or by a group of counties in a unique geographical area, or (2) when products near population centers are separated from those farther away. For example, visualize a situa- tion where a product (say number one) is represented by both Grand Traverse and Leelanau counties in northern Lower Michigan. Moreover, another product (say number two) is located in Berrien county in the extreme south- western end of Lower Michigan. For those origins in sout- ernmost lower Michigan, the price of product two is al- ways 1ow relative to the price of product one, and vica versa for origins in the northern part of the State. Con- sequently, when product two is examined as a potential substitute in the equation for product one, a strong sub- stitution effect (high positive coefficient) will be ex- hibited due to the negative correlation between product prices. Other cases may result in positive interrelation- ships between product prices. The major multi-collinearity problems were minimized by excluding those variables sub- ject to these statistical pitfalls. The many demand equations permit generalizations about consumers' revealed preferences for angling activi- ties. By comparing the different product demand curves for a particular seasonal type of fishing, one can assess the relative importance to anglers of various qualities of angling. In general, the nature of the comparative demand for any angling product depends upon (1) the 56 location of the demand curves with respect to one another, and (2) the shapes of the demand curves. Whether a demand curve is to the right or left of other curves indicates anglers' relative preferences for different angling pro- ducts. The shape of a demand curve is an indication of the responsiveness in quantity demanded to changes in the price of a particular angling product, or what is referred to as the price elasticity of demand. The relative positions of each demand curve was approximated by using an average of the 88 minimum supply prices for each of the final substi- tutes, including average county per capita income. The demand curves for each seasonal type of fishing are shown in the following chapter. V. RESULTS AND DISCUSSION 1. Product Classification Analysis Salmonid Angling in the Great Lakes A total of 41 counties offered some form of Great Lakes salmonid angling each period. Anglers fishing for salmonids on the Great Lakes appear to find catch rates and species mix the most important attributes. It was de- cided that a nine-way breakdown describing eight products best identifies winter-spring (Period 1) Great Lakes sal- monid fishing (Figure 13). For this period the level of "excess expenditures" was 15.3%. Apparently, anglers dis- criminate primarily between the trout group (steelhead or rainbow, brown and lake trout) and the salmon group (coho and Chinook). They seem to be interested in the catch rates of any one or more species within these two groups, and do not distinguish much between the Species within the groups. For example, if a particular county has a "high" catch rate for at least one species of trout and "moder- ate" (but not high) for either coho or Chinook salmon, then it falls into the "high" any trout and "moderate" any salmon category (product number 1). If only "moder- ate" catch rates are found for one or more trout, the county is classed as product number 4. Tables 13-15 in 58 ATTRIBUTES Product Number Catch Rates Identification of Any Trout [ Any Salmon Number Counties P——-—- RICH ’ ’ HIGH MODERATE 1 2 I_____ 10W 2 5 ~——- HIGH 3 4 monsnnrs monznnrt 4 7 ,___ l0" 5 ll ._ mu 6 I 1 [CW MODERATE 7 3 ,.___ Low 8 3 Figure 13. Identification key defining eight different kinds (products) of Great Lakes salmonid fish- ing during the 1976 winter-spring period. 59 Appendix B lists the counties corresponding to each of the various products defined for the respective periods and types of fishing. Definitions of all the final catch rate attributes are found in Tables 16-18 of Appen- dix B. Alternative attribute combinations which were found to be almost as satisfactory in period I included (1) the availability of pier fishing in conjunction with any trout and any salmon catch rates, and (2) a nine-way breakdown of aggregate trout and any salmon catch rates. Some addi- tional factors tested in combination with various catch rate attributes included the location of a county on a natural bay, the amount of publicity, and urban/non-urban angling environment. Although none of these combinations proved to be as successful as catch rate alone, these fac- tors could be of secondary importance to anglers. These results are not surprising, in light of the fact that the winter-spring fishery in Michigan provides excellent early Spring runs of steelhead off river mouths, and late Spring lakeshore runs of brown trout, lake trout, and salmon. Although there are anglers who prefer a cer- tain species of fish and/or areas with bays, the results indicate that the majority of anglers feel and behave dif- ferently when fishing for salmonids during the winter- spring season. The summer (period II) and fall (period III) Great Lakes salmonid fisheries differ somewhat from that of 60 spring. Larger coho and more abundant stocks of Chinook are responsible fOr enhancing salmon fishing opportunities. Good creels of lake trout are caught in summer. Steelhead fishing reaches its peak in late fall, usually after salmon have entered the streams for their annual spawning migra- tions. For the 41 counties, seven Specific products with a level of excess expenditures of 12.3% were identified in period II and eight Specific products with 8.7% in period III. During these two periods, anglers were apparently most interested in "any trout" and "aggregate salmon" catch rates (Figures 14 and 15). Other hypotheses which were tested and found somewhat less important indicated that anglers do not differentiate summer lake trout or fall Steelhead fishing from their respective species groups, and aggregate catch rates for all fish (hypothe- sizing that anglers do not differentiate at all between species) were not as satisfactory. In addition, hypotheses testing secondary attributes such as alternative forms of recreation available at sites (complementary recreation), natural bays, pier fishing, and publicity were not as suc- cessful as catch rate alone. Angling_for Anadromous Fish in Streams The winter-spring anadromous fishery in Michigan is based primarily upon the numerous migrating steelhead trout. Upwards of 200,000 Steelhead were caught in the many streams and tributaries during their 1976 Spawing 61 ATTRIBUTES Product Number Catch Rates Identification of Any Trout Aggregate Salmon Number Counties F'_’—' HIGH 1 3 HIGH MODERATE 2 3 l——-— 10?! 3 5 HIGH 4 1 MODERATE MODERATE 5 5 '— LO IV 6 15 '_————- HIGH — - 1.0 W MODERATE — _ [OW 7 9 Figure 14. .Identification key defining seven different kinds (products) of Great Lakes salmonid fish- ing during the 1976 summer period. 62 ATTRIBUTES Product Number Catch Rates Identification of Any Trout Aggregate Salmon Number Counties F—-- HIGH 1 4 HIGH MODERATE 2 1 -——-—- 10W 6 2 HIGH 3 4 MODERATE MODERATE 4 7 I—— IM 5 5 t--- HIGH ' 7 7 [ON MODERATE - - *--—- [OR 8 10 Figure 15. Identification key defining eight different kinds (products) of Great Lakes salmonid an- gling for the 1976 fall period. 63 runs (Michigan DNR, unpublished survey results, January 1-May 31, 1976). Anglers have ample opportunity to pursue steelhead fishing in 52 counties around the state. Fig- ure l6 identifies the nine specific kinds of anadromous angling for period I. The overall level of excess expend- itures is 11.5%. It is apparent that (l) steelhead catch rate (3-way breakdown), (2) publicity, (3) natural lake throughways, and (4) fly-fishing regulations (in this or- der of importance) are the most important attributes used by anglers to differentiate between counties offering winter-spring anadromous fishing.10 Some additional attri- butes examined with catch rate included (1) the presence of dams along streams, (2) anadromous stream mileage in a county, (3) size of streams, and (4) availability of land- locked anadromous fisheries. Stream size appears to be the most important secondary attribute. The fall anadromous fishery is primarily dominated by the Spawning migrations of coho and chinook salmon, al- though in late fall steelhead are also found in streams throughout the State. Thirteen specific products were de- fined for the 52 counties by the following attributes (in their apparent order of importance): (1) aggregate steel- head and salmon catch rate, (2) natural lake throughways, 10The reader will notice that levels for the fly- fishing attribute were left unlabeled for the majority of angling products. This was done to facilitate perusal of the figure. However, one should understand that each per- mutation is characterized by unavailability (NO) for this attribute. Similarly, other attribute levels in Figures 17- 20 are unlabeled, but they also denote unavailability. 64 ATTRIBUTES Product Number ID of Catch Rate . . Lake Fly-fishing - Steelhead PubllClty Throughways Regulations Number Counties YES 1 3 --|HOH-j F—-—-YES 2 1 HION--* ‘—-——-NO 3 2 F—-—-YES - - -— IDW-——‘ *———- NO 4 9 --—-YES 5 l . HI H-- r—— O )—--NO 6 3 MODERATE- _--YES 7 3 L—— lON-——- L----NO 8 l3 r——-—YES - - HICH-——* NO - - [OW r—-—-YES - - )— Iow— “—-'NO 9 17 Figure 16. Identification key defining nine different kinds (products) of Great Lakes anadromous fishing during the 1976 winter-spring period. 65 (3) snagging regulations, and fly—fishing regulations (Figure 17). For period III, the level of total excess expenditures is 5%. Among the numerous alternative attri— butes tested, stream size again appeared to be the most successful secondary attribute in combination with catch rate. These results for anadromous angling may not be sur- prising to those familiar with salmonid angling opportuni— ties around the state. Anglers' interested in salmon and steelhead may be attracted to areas with lake throughways at or near river mouths of major anadromous streams. This is especially true of the many rivers found in counties located along the eastern Shore of Lake Michigan. AS sal- mon and steelhead enter the streams for their annual spawning migrations, they inevitably must pass through these lakes before continuing upstream. Moreover, lakes of this nature provide greater Opportunities for boat fishing and an angling environment which differs Signifi- cantly from that of streams. Since lake throughways are usually protected from strong off—Shore winds, they also offer more favorable weather conditions than open-waters for angling activities. While publicity is to some ex- tent influential in any season, it is apparently more cru— cial in period I. The initial reports of steelhead success are likely to have a greater impact on anglers after anad- romous angling activity has been somewhat slower during the winter. Some anglers may be attracted to salmon 66 ATTRIBUTES Product Number Catch Rate Lake Snagging Fly-fishing ID of Aggregate through regulations regulationsINumber Counties Steelhd/Salmon —- YES I 2 -YES—-' —— NO 2 2 YERY HIGH—- ‘r—— YES 3 2 __.~o___ *~——- ND 4 2 —— YES 5 2 HIGH ———IIO —-4~ ~—-—YES 6 1 .____.'o.____ . —— no 7 3 — YES 8 I e—Ies— L—— RD 9 2 MODERATE-— YES 10 4 u—uo—w L— no 11 12 -—- YES 12 2 [OR --—-NO-—* L— NO 13 17 Figure 17. Identification key defining thirteen different kinds (products) of Great Lakes anadromous fish- ing during the 1976 fall period. 67 snagging areas because little skill is needed while fishing and higher pOpulation densities in the fall increases the probability of catching salmonids. The fact that aggregate salmon-steelhead catch rates are important implies that most fall anglers are interested in all Species rather than one individual species. Non-Salmonid Angling in the Great Lakes A total of 41 coastal counties offer some form of Great Lakes non-salmonid angling activity annually. In period I, ten Specific products are identified by (l) ag- gregate panfish (yellow perch, bluegill, crappie, white bass) and aggregate gamefish (bass, walleye, pike/muskie) catch rates and (2) special bass regulations (Figure 18). The level of excess expenditures is 4%. These particular attributes were selected in lieu of other slightly less important catch rate attributes, namely (1) aggregate panfish and any gamefish, and (2) any panfish and aggre- gate gamefish catch rates. Other attributes were examined to test such hypotheses as (1) whether yellow perch catch rates were considered important because of this species pOpularity during the winter ice fishing season, or (2) that yellow perch, walleye, and any other Species catch rates may be more important to anglers who exploit the pop- ular walleye and yellow perch fisheries during their spring Spawning seasons. However, both of these theories proved less satisfactory for the winter-spring non-salmonid fishery. 68 ATTRIBUTES Catch Rates . Product Number SpeCial Bass ID of Agassi assist .eg........ Numb... IIIOII I I HICII ———MODERATE 2 I L— IN 3 I . YES 4 3 *r- IIIOII——— ~————- NO 5 1 MODERATE -———-MODERATE 6 3 ~— IOIY 7 3 —- IIIOII 8 2 low ——-—NODERATE 9 6 -— low 10 20 Figure 18. Identification key defining ten different kinds (products) of Great Lakes non-salmonid fishing during the 1976 winter-spring period. 69 The wide variety of fish available to non-salmonid anglers in period I may be the reason anglers regard groups or categories of sportfish as more important than individual Species. The Special bass regulation occurs as a result of bass season Opening later (June 17) than normal (May 27) for three Michigan counties: Wayne (Detroit river), MaComb (Lake St. Clair), and St. Clair (St. Clair river) counties. Interestingly, winter-spring survey results in- dicated that large numbers of bass were caught in these areas. Apparently, either (1) some respondents may have mistakenly reported their early summer bass catches (after June 17) instead of their actual catches for the winter- spring period, or (2) anglers are taking advantage of an excellent smallmouth bass catch and release fishery in these special counties and reporting their catches. Since the existence of higher bass catch rates were verified for these counties, it is feasible that anglers are be- having in the latter fashion. At the same time, the class- ification revealed that many anglers also tended to avoid the three special counties in order to reach otherwise identical products at some more distant site. Evidently, most anglers are inclined to fish for bass where they may legally retain their catches. During the summer period, anglers' are apparently primarily concerned with (1) any panfish and aggregate gamefish catch rates and (2) a resort or vacation area fac- tor. Twelve specific products are defined in this 70 classification and the level of total excess expenditures is 3% (Figure 19). The resort factor was selected because a large percentage of anglers were Spending in excess of what was necessary for primarily moderate catch rate pro- ducts in the upper peninsula. In particular, many anglers originating from southern lower Michigan (especially the Detroit area) were avoiding hypothetically identical Sites in the lower peninsula in favor of Chippewa, Mackinac, and other counties. Since the upper peninsula is a popular seasonal recreational area with many aesthetic qualities, it is reasonable to conclude that anglers are also spend-A ing their vacation time in this resort-like setting. The amenities of a site can significantly influence the de— cision-making process. Examples of slightly less satisfactory attributes in period II included (1) aggregate panfish - other (suckers, catfish, Whitefish, etc.) and any gamefish catch rates; and (3) any panfish and any gamefish catch rates. Other attributes rejected as being less important suggested that anglers do not differentiate between individual non-salmonid species during the summer. Also, factors such as bays, piers, and publicity were not as successful in combination with catch rate. A travel pattern similar to period II was found in the fall. In addition to a resort factor, catch rates for yellow perch and "any other species" (bluegill, crappie, bass, pike, walleye) appear to be the most important 71 ATTRIBUTES Product Number Catch Rates Resort or ID of Any Panfish Aggregate Vacation Number Counties Gamefish Areas r--4"GN I 1 HIGH MODERATE 2 .2 '———-— ION 3 2 INCH I l «—— YES 5 I MODERATE —-——— MODERATE—w IL— 00 s 5 '~———-—- LOIY 7 4 *--- HIOH YES 8 2 -— YES 9 S [OH MODERATE— *-—-ND 10 8 .r__ YES 11 s —— Iow _. ~———- NO 12 I Figure 19. Identification key defining twelve different kinds (products) of Great Lakes non-salmonid fishing during the 1976 summer period. 72 attributes. Twelve specific products with total excess expenditures of 2.2% were defined for period III (Figure 20). Various other catch rate attributes of secondary importance included (1) aggregate panfish and any game- fish; (2) aggregate panfish and aggregate gamefish, and (3) yellow perch, walleye, and any other Species catch rates . These results indicate that anglers apparently do discriminate between species in period III (perch in con- trast to the other species). This is probably due to the abundant populations of yellow perch in comparison to other non-salmonid fish Stocks. It is plausible that yel- low perch becomes the dominant species because of an adap- tive capacity for remaining vigorous as water temperatures rapidly cool during the fall. As a result of this toler- ance, ice fishermen frequently harvest large creels of yellow perch in winter. It is evident from each of the preceeding angling classification analyses that catch rate and a particular species mix of fish are the most common and sometimes, as in the case of Great Lakes salmonid angling, the only significant attributes defining angling quality suffic- iently. Instances in which other attributes in combina- tion with catch rate, as for anadromous and non-salmonid angling, are used by anglers to differentiate between an- gling sites indicate the apparent importance of (1) an- gling Sites with the reputation or potential for higher 73 ATTRIBUTES Product Number Catch Rates Resort or ID of ' Numb ' Yellow Perch Aggeggggr Vigggéon er Counties l“OH-—-—‘ '--MDDERATE 2 2 ‘———'YES 3 l “' NIOH--‘ “—_—'YES 5 1 MODERATE———MODERATE—‘ .I___. no 5 2 -———'YES 9 4 LOW —“— MODERATE —* '—--YES 11 7 t—-- lOW -———~ Figure 20. Identification key defining twelve different kinds (products) of Great Lakes non-salmonid fishing during the 1976 fall period. 74 catch rates (publicity, lake throughways, special bass regulations, snagging), and (2) locations which require a greater degree of skill and effort in catching fish (no snagging, no lake throughways, fly-fishing regula- tions). Other product classification studies have shown varied results. Talhelm (1972) found that trout catch rate, stream size, regulations, and in some cases stream- side buildings were the most important attributes defin- ing trout angling quality in the southern Appalachians. In a study of Michigan's 1970 salmon-steelhead fishery (Talhelm, 1973b), angling quality was defined primarily by combinations of catch rates of three species of fish: coho salmon, chinook salmon, and steelhead trout. In add- ition, some other secondary attributes of importance in- cluded (1) urban or non-urban angling environment, (2) pub- licity, (3) early or late salmOn migration, (4) the nature of the streams in which fish migrate, and (5) the availa- bility of complementary types of recreation. Finally, an inland lake study of angling in Michigan classified an- gling sites in terms of (l) lake Size, (2) catch rates and species mix of fish, and (3) whether or not a lake has a public entry point (Talhelm, 1976). 75 3. Demand Analysis General Description Anglers' demands for the different varieties of an- gling are determined by estimating demand functions for each specific angling product. Demand curves are then plotted and compared to assess anglers' relative prefer- ences for angling in Michigan's Great Lakes. The follow- ing figures illustrate the demand curves for all of the angling products identified in the classification of each seasonal type of fishing. A letter code for each permuta- tion of attributes is used to delineate the various quali- ties of angling. In the case of multiple attributes, lev- els of catch rate and those attributes which denote avail- ability or high levels for a product are labeled to facil- itate easy comparison. For example, in Figure 21, H-M is the demand curve for products having high any trout and moderate any salmon catch rates (product number 1 in Fig- ure 13) during the winter-spring season; there are no other attributes describing period I salmonid fishing. In Figure 24, H-PB-LK indicates the demand for angling products with high steelhead catch rates, 3333 publicity, presence of lake throughways, and no fly-fishing regula- tions (product number 1 in Figure 16); H represents the demand for angling products having high steelhead catch rates, low publicity, lake throughways unavailable, and no fly-fishing regulations(product number 4 in Figure 16). The demand equations estimated for each angling product 76 are given in Tables 5-12 of Appendix B. Angling for Great Lakes Salmonids Figures 21, 22, and 23 illustrate the demand curves for the three seasonal periods, respectively, In general, products in greatest demand are those with (1) higher trout catch rates during period I, and (2) higher trout catch rates at relatively low prices and higher salmon catch rates at somewhat high prices in periods II and III. The demand equations also indicated that higher trout and salmon catch rates are generally substitutes for lower catch rate products, particularly in period II. In period II low catch rate products are also substitutes for es- pecially high trout and salmon catch rate products. In addition, anglers substitute inland trout, inland game- fish, and inland panfish angling for low catch rate trout and salmon angling products in period III. These findings reflect two different classes of sal- monid anglers. First, there are those who probably "key" on or exhibit a strong preference for trout throughout the year. This is especially true during winter-spring and fall, when trout abundance and distribution is great- est. At these times, anglers have ample Opportunity to fish for trout around the state, and need not travel far to do so. Second, there are a group of anglers who may be considered more dedicated salmon anglers, as displayed by their willingness to pay higher prices for products 77 III-I 570~ 3 E ‘0v« I = gs»- : n- u- u- n ‘ ‘ ' 20 00 100 100 100 120 440 AIAIEI 001: 0001000 000110 5 O S 3‘ I C C r n 3 = ‘ \ \ ‘I s “T‘\ ‘1... \ ”‘H— 100 220 :40 000100 DAYS PEI 1000 CAPITA Figure 21. Demand curves for the eight salmonid angling products identified by any trout-any salmon catch rates in period I: H = high, M = moder- ate, and O = other. 78 PRICE “PER RRCLER DAY) 2‘ 2 l I ‘81 ID- ID- ID ' ' ——~- 20 SO IOD III ISO 220 “O ANCLER DAYS PER IND CAPITA SD- IId "4 60-1 PRICE IS PER ARClER I") SO- ID-I ID- ID- h 20 SO I” III ISO 220 “D ANCLER DAYS PER I”. CRPITA Figure 22. Demand curves for the seven salmonid angling products identified by any trout-aggregate salmon catch rates in period II: H = high, M = moderate, and O = other. . ID- 70- PRICE (IPER ANCIER DRY) TO- PIICE (S PER ANCLER DAY) / EI-w \ SD- II-( 3D-4 2D‘ 79 III III III ANCLER DAYS PER IOOO CAPITA IID 10 Figure 23. ID SD 100 III III 220 III AICLER DAYS PER IDDD CRPITA Demand curves for the eight salmonid angling products identified by any trout-aggregate salmon catch rates in period III: H = high, M = moderate, and O = other. 80 with high salmon catch rates. Apparently, these anglers are traveling greater distances to reach the rather Spec- ific locations providing higher "quality" salmon fishing, particularly along the western Shore of the lower penin- sula. Angling for Anadromous Fish Demand curves for the various anadromous angling pro- ducts are illustrated in Figures 24 and 25. In period I anglers Show strong preferences for (1) high steelhead catch rate areas receiving little publicity and having no lake throughways, (2) higher steelhead catch rate areas re- ceiving high publicity and containing lake throughways, especially at high prices, and (3) somewhat more moderate steelhead catch rate products with either high publicity or lake throughways present, particularly at lower prices. In addition, anglers were found to substitute (1) higher steelhead catch rates for lower ones, (2) areas with lake throughways for those which have higher catch rates but no lake throughways, and (3) Great Lakes salmonid and non-salmonid angling for low steelhead catch rates when lake throughways are present. During period III, products generally with (1) higher aggregate salmon/steelhead catch rates having either or both lake throughways and snagging, and (2) moderate catch rates and no lake throughways or fly-fishing regula- tions are in greatest demand. Moreover, anglers substitute 81 801 it = 370‘ s g C - t a: l3 S Q. IO- 30~ 20- 10 20 4‘0 00 W 100 140 130 {00 *7‘20 ANCLER DAYS PER IDOO CAPITA 801 .= ‘ - :70- -l 3 C C :50- c' .3. Ear fin 1:0 360 7'50 ANOLER DAYS PER IDOO OAPITA Figure 24. Demand curves for the nine anadromous angling products identified by steelhead catch rate- publicity- lake throughways-fly fishing reg- ulations in period I. 82 on a I ; I g 4 570‘ \I‘ ‘s’ \\ I500- \ '. E \ ~. EM) \ \‘ Bo lu“ no a“ ”ho “01:11 am PER 1000 cmn RD] N a l CR 6 A PRICE (3 PER ANCLER DAY) 90 O A 40- 20y ID 2 40 00 so 100 130 100 :60 7'20 ANCLER DAYS PER IDOO CAPITA Figure 25. Demand curves for the thirteen anadromous angling products identified by aggregate salmon/steelhead catch rate (VH = very high)-lake throughways-snagging regulations- fly fishing regulations in period III. 83 very high catch rates for low ones and in cases where snag- ging is permitted in moderate catch rate areas, moderate catch rates for higher ones. These relationships suggest that (l) in period I steelhead anglers reveal strong preferences for higher catch rate areas which are well publicized and/or because lakes at river mouths provide additional angling opportun- ities, (2) in period I the presence of lake throughways enhances the desirability for angling at lower catch rate areas, (3) in period I anglers find open-water Great Lakes angling roughly equivalent to lower steelhead catch rate areas when lake throughways are present; the opportunities for salmonid and non-salmonid angling are greater because of the convenient access to open-waters and because both can be fished from boats, (4) in period III anglers prefer higher salmon/steelhead catch rates when either or both lake throughways are present and snagging is permitted, (5) in period III anglers prefer moderate salmon/steelhead catch rates, no lake throughways and no special regula- tions in counties situated in the northern lower peninsula and upper peninsula; perhaps anglers expect higher quality angling in the northern part of the state, and (6) in period III anglers are willing to switch to lower catch rate areas when snagging is also permitted. 84 Angling for Great Lakes Non-Salmonids The non-salmonid product demand curves are illus- trated in Figures 26, 27, and 28. Generally, they show demand is greater for (1) higher panfish and gamefish catch rate products in period I, (2) higher panfish catch rate products at lower prices and high gamefish and panfish catch rate products at high prices in period II, (3) high gamefish catch rate products in resort areas in period II, and (4) high yellow perch catch rate products in period III. In addition, the demand equations indicated that in period I higher panfish catch rate angling in the Great Lakes and inland panfish angling are substitutes for lower panfish catch rates in Great Lakes open-water. In period II inland gamefish angling generally is a substitute for more moderate panfish and gamefish angling products in the Great Lakes. In period III both inland panfish and game- fish angling are substitutes for low yellow perch catch rate angling in Great Lakes Open-water. These relationships suggest that in general non- salmonid anglers prefer a yellow perch/panfish fishery which provides higher catch rate possibilities. The great- est demands for higher panfish catch rate products are ob- served at low and high prices during winter—spring and summer. No doubt many anglers feel panfish and perch are easily caught and are an excellent source of food, so they are willing to travel varying distances to obtain high catch rate products. Those anglers interested in gamefish 8 3 8 l l _1 "ICE (S P" "Cl" I") 0‘ O #1 85 1 ml 0 1 "It! (3 P" "GI." I") 30‘.‘ 20‘ ml ' Figure 26. A T00 3&0 030 600 mm ms m 1000 cmn do 060 1'50 mm on: m 1000 cum Demand curves for the ten non-salmonid an- gling products identified by aggregate panfish- aggregate gamefish catch rates-special bass regulations in period I. met (mu mm m) E 33' 7,5 / ,/ l ¥ "ICE (3 "I "Cl“ I") 0'! 9 1_ - o L 30- 20- 86 ..\ J A j 356 440 0 mm ms m 1000 mm ; g. 10 Figure 27. 330 (40 000 0:0 mm 0m m 1000 mm Demand curves for the twelve non-salmonid angling products identified by any panfish- aggregate gamefish catch rates-resort areas in period II. 87 N a e o l __J 1 "ICE (3 PEI mm D") ‘1 :30 4‘00 600 m IIGLER on: m 1000 mm 8 3" 8 l l I "It! (s "I mm D") 0! O 1 403 1” 160 3 400 030 fisho INGLER DAYS PER 1000 “PITA Figure 28. Demand curves for the twelve non-salmonid angling products identified by yellow perch- any other species catch rates-resort areas in period III. 88 . angling products with higher catch rates may believe that this species group is particularly vulnerable during the winter-Spring and early summer seasons; bass, walleye, pike and muskie each spawn at this time and are more sus- ceptible to hook and line fishermen because of their nor- mally more aggressive behavior during spawning season. Not surprisingly, the demand analyses have generally supported the intuitive feeling that demand is greater for higher catch rate products. Other demand studies utiliz- ing this classification procedure have shown somewhat sim- ilar results. Talhelm (1972) found that angling products in greater demand were with higher catch rates and larger stream sizes. A study of salmon-steelhead fishing in Michigan (Talhelm, 1973b) also showed that (I) demand is greatest for high catch rate products, (2) anglers are willing to switch from lower catch rate angling locations to high catch rate locations and (3) a stronger positive relationship exists between personal income per capita in the anglers' origin county and the demand for higher catch rate angling. In addition, Talhelm concluded that salmon- steelhead anglers (1) consider inland trout angling as equivalent to salmon-steelhead angling, (2) they strongly prefer high catch rate salmon—steelhead angling to other gamefish angling, and (3) they strongly prefer high catch rate salmon-steelhead angling to perch-panfish an- gling, particularly during summer. VI . CONCLUDING OBSERVATIONS This research has utilized a discriminant analytical technique to define angling quality for Michigan's 1976 Great Lakes salmonid and non-salmonid sport fisheries. Furthermore, this definition of quality is integrated into a demand and supply model for angling recreation in Mich- igan. These separate but related analyses provide several kinds of useful information to fisheries managers and planners: (l) a description of Great Lakes fisheries in terms of the attributes or characteristics of angling ap- parently most important to anglers; (2) descriptions of the specific varieties (qualities) or products of angling com- prising the general forms of Great Lakes angling; (3) in- dications of anglers' preferences for each of the specific angling products within a general product group; (4) an indication of the willingness of anglers to substitute one kind of angling for another; and (5) the prices or costs of each specific angling product to anglers originating from any location. These kinds of information should prove extremely useful in formulating management programs for Michigan's Great Lakes sport fishery resources. It is recognized that recreation quality is an important demand determinant. 89 90 An explicit consideration of this concept provides greater insight into the preferences of anglers for the different characteristics associated with Great Lakes angling activ- ities. Managers and planners can concentrate on those as- pects of the angling experience in greatest demand and thereby increase the efficiency of management programs by making socially optimal choices. Such important manage- ment questions as what "kinds" of recreation should be developed and where, which characteristics should be pro- vided and expanded, which characteristics should be al- tered or contracted, are more easily answered by under- standing the exact nature of peoples demands for outdoor recreation. To ensure that resources are prOperly util- ized and investment decisions are properly allocated for the greatest public good, the personal perceptions of recreation users must be evaluated in the decision-making process. An additional guide for determining the optimal level of management efforts is to estimate the benefits of vari- ous management programs for Michigan's Sport fisheries. By comparing the trade-offs of several hypothetical choices, managers may select a strategy which maximizes social wel- fare. This kind of information will be provided in future studies using the results of my research in a computerized simulation model. The demand equations will be used to predict the gain or loss in direct benefits to users that would result from planned or unplanned changes in angling 91 attributes over time. This would be equivalent to chang- ing the product ("quality") of fishing at a certain loca- tion. A number of such hypothetical changes in specific attributes will be made to test various alternative manage- ment strategies for each type of Great Lakes fishing. The model will estimate (1) the changes in user benefits and participation levels that would be produced by changes now or in the future at specific counties, (2) the changes in user benefits and participation that would be produced now or in the future by creating new fishing sites with certain attributes, (3) the effects of both types of changes on the participation levels at other counties in the system, and (4) the amounts of participation at present and ex- pected in the future for all counties relevant for a par- ticular type of fishing. APPENDICES APPENDIX A DEPARTMENT OF NATURAL RESOURCES SURVEY QUESTIONNAIRE APPENDIX A DEPARTMENT OF NATURAL RESOURCES SURVEY QUESTIONNARIE STATE OF MICHIGAN m, 15‘”? «we» “sou-cu common ”(Q-6 ' CAIII. T JOHNSON a...» s VI LAITALA WILLIAM G. MILLIKEN. Governor DEAN PRIDGEON mum F st DEPARTMENT OF NATURAL RESOURCES ”W" “- "mm" STEVENS T MAsoN BUllDlNG. sox 30023. LANSING. MlCHlGAN 48909 JOAN l WOT?! T CHARLES G VOUNGLOVE HOWARD A. ANNCR. Dirmor Dear License Holder: You have been selected to participate in aur annual survey of sport fishing in Michigan. Only Iour fishermen out at each one hundred license buyers will be surveyed this year. Therefore, your response is very imponant for an aCCurate «aunt of where people tisned what they fished for, and how many fish they kept in 1976. Please return your completed questionnaire even if you did not go fishing this year. Since we value the response of everyone in the survey, you can expect a reminder in abOut three weeks if we do not hear from you. Thank you for you time in helping us Improve fishing in Michigan. Sincerely. [4.4.2.- Henry J. Vondett Acting Chief. Fisheries Division I976 MICHIGAN SPORT FISHING SURVEY: JANUARY 1 - DECEMBER 31 1 WHERE IS YOUR PERMANENT RESIDENCE" County State ‘2. DID you FISH IN MICHIGAN DURING I976? YES ‘ . Please continue with question # 3 NO l ' Please return this questionnaire by mailing it in the enclosed post paid envelope CONTINUE ON THE OTHER SIDE OF THIs SHEET ' OICI ".[J ' ‘6 93 2 3 4 76 77 73 ’9 30 3. DID YOU FISH ON THE GREAT LAKES OR THEIR CONNECTING WATERS (Lake St. Clair, Detroit. St. Marys and St Clair Rivers) FOR FISH OTHER THAN SALMON OR TROUT IN I976? YES D Please Complete TABLE A and then continue with question # 4 NO D Please continue with question # 4 TAELE A - GREAT lAKES FISHING 55‘ I I “ I . g A I ‘ , mm 9.0 you o~ me one” mes I 0513,,“ I "ow M:;gU;th,gD:gwgrg_§’-A§§~ m ANO CONNECTING WATERS? I LOCATION? . AEED'AT EACH LOCATION? . f ; NUMBER ~N CATC-s I | I ‘ I ’ l ' r e l I I a I .3 E A .3; w I 2 I:- I I c.1229: 2:. : W I If: so , s: 32: o I _ I ,- ... NEAREST . OF DAYS I :2 I g: ‘ : ,-¢ 33...,“ a... g:* ; §:«_- 5 V A "REA' LAKE OR ’OWN CR CITY : FISHED I “ '5 is E: :25, 1’; l -’-< I .7. 3:» = i CONNECTING WATER ‘ l '3" ' £3 i-I "”“c. S; | 5". '3: ‘ I I ‘ l E . i I 1 Q ' : i I 5-7. i iii—To» l iii-i2, I (is-III 'w u (47, isoii so I ism ‘ '50:: o- oei .,. E; - ng'caw as, . Pznconning J; 3 11 Q . 20 I ' T f I T fl T T : O‘I ‘ I I E ‘ . I 442 . . I I I I ' I I 443 ' I 7 i . l , l y ‘ . , T I 4“ I I I ; , . , . : I I i i 3 I ' . I I ' I ‘ . L L A ' L 4. DID YOU FISH ON INLAND LAKES OR STREAMS FOR FISH OTHER THAN TROUT AND SALMON IN I976? YES 3 Please complete TABLE 8 and then continue with question # 5. NO L_:Please continue with question 3 5. TABLE I " INLAND LAKE AND STREAM FISHING 7 DAYS AT A HOW MANY FISH :EXCLUOING taou‘I WHERE DID YOU FISH" I EACH ' DID YOU CATCt-s AND ‘ LOCATION? ‘ KEEP AT EACH LOCATION" I . ‘JLVISER 'N CATCH l I I i ‘ t I - . I 3 z :3 I . __ : 53' 8. a ‘ $992.? 9" ‘ was ”i ‘ cc ; co o , a No STREAh OR LAKE 'O‘I‘v‘: CRSC'TY 3 "9‘50 , g 8- I is . g; g; “1;; g: ‘ S E. E I . ' ‘ . , 32%; g‘ .r ;‘ziI§»I. I , I A I ' *5—7‘ 8-‘0' ; -It-12‘_ I ‘Iz-I‘I MALI Mills?“ (50: '53 56 II5° ‘lb? I s‘ "‘ Ex ' TecI LO"! Negowzee " i 7 1 S ssi , ’ E E 7 . I I l i I I I ' l r a 55’ i ‘ . : i i . s ; , H I r s—“ L J 4 I . i I I I . ' 553 . I . l I I I . , f i I i ' I ! 1! 1L I 1 I i ‘ #4 T i * * z ' I . . 4 ‘i . " I I L‘ 5. HAVE you FISHED FOR TROUT OR SALMON DURING W76" YES [3 Please continue with question g 6 NO 3 Please skip questions :3 6. # 7, and # 8 andoretu'n only "“5 5h”? 'l" 9"“0590 905* paid envelope. 94 D. DID YOU FISH IN l97o ON THE GREAT LAKES OR THEIR CONNECTING WATERS (Lake St. Clair Detroit. ST. Marv s and St. Clair Rivers) FOR COHO SALMON. CHINOOK SALMON. LAKE TROUT. RAINBOW TROUT. OR BROWN TROUT" _ YES DPlease complete TABLE C on the other side of this sheet Please continue with question #7 NO 7. DID YOU FISH IN 1976 FOR STEELHEAD (Rainbow Trout over I5 inches) CHINOOK, COHO. OR ATLANTIC SALMON IN GREAT LAKES TRIBUTARIES DOWNSTREAM FROM THE FIRST BARRIER TO MIGRATING FISH OR IN ANY OF THE FOLLOW- ING LAKES: (Lake Mocatowa. Pentwater Lake. Muskegon Lake. Manistee Lake. Lake Charlevoix. White Lake in Muskegon Caunty. Portage Lake in Manistee Caunty Platte Lake and Loon Lake in Benzie County. Fere Marquette Lake Duck Lake in Muskegan Caurity, and Pigeon Lake in Ottawa Caunty? YES D Please complete TABLE D on The other Side oi this sheer Please cantinue with question # 8. NO — LAKE SUPERIOR STREAMS RIVER Anna Bloc-i Zora Chccoiay Deco ‘aiis Hurgn Laughing W'HTCTI‘H LIrtie Geri-r Ontonagon F'esaue I'sie Silve' Sturgeon 340(0' 'wo Hearted COJNTV Alger Gogebic Marquette Marquette Marquette Baraga Baraga. Marquette Aiger Marquette Ontonagon Hoc'gnton Gogebic Baraga Baraga "Ough'Ofl AIge' LUCC LAKE MICHIGAN STREAMS R‘VER Bear Sear {ree- Be'sie Bloc-i Black Booramon Boyne Big Cedar C'atiiery Cree- Crystal COC‘NTY Emmet Mamet” Benzie .’an Buren Aliegan Mackinac Grand Traverse Charlevoin Menormnee OHM Muueqon Leeienau PRIMARY SALMON AND STEELHEAD STREAMS ' Riv" Eli. Fish Creek F at Gotten Grand JOPCOH Kalamazoo Leiand Little Manistee Big Manistee Man-maue MOI”. Menommee Muskegon Paw Paw Pentweiev pore Marquette Dion. =’rcI'Ie C'eeii Roooit Rogue Soaie Si Joseph Stoney Creek Sioi-ii Creeii Sturgeon River 5m Creek Thornoppi. Whit. Whitetish COUNTY Antrim Ioma Menicolm Kent .omo Berrien Om. Kent onia Antrim Allegan Leeionou M06610. LOIR Moms‘OO Schoolcratt Ciinton ion-a Menomnee Mus-each Newavqc Berrien Oceana Mason Lake Oceana Newavgc Benzie onia Alieqan len' Mow!- Berrien 'onia Clinton Oceana Delta Al'egan iieiu Mus-each Oceana Delta LAKE HURON STREAMS FILER A. Ives AL Scale Blot! 10's Cass ineoavgon Cr poewo C-OFvana Creek Eih Cree: g; r' low-Owun Nooest Cree. Ocoueoc ° :ecn o ;. =;e‘. Pinneoc; :II‘. Scanaw Smowanee Tawas T~-nce' Boy T ”anawassee 'rc.t ‘Nhitne. Drain (CANT Arenac Iosco osco Aiccno Mat-mac C'TIDCOW Sag-nae 2x: : Cheboygan 'sasena Mic‘ona Huron Sen-to: Saginaw Senezsee Bay Presque ‘s-e F'escice iste d~r3i~ :- tono ascc Mac-mo: Lnipoewa Hfllofi A-eroc Ogemaw Bov See new leg-Pow Sn-awas'ee 2x; Alpena 3c; nov- ‘-'-'J-:'c F«no.9 sie Arena: LAKE ST. CLAIR STREAMS RIVER Clinton C" hTV ‘w‘b Macon: Cc- one LAKE ERIE STREAMS R712! Huron Raisin-i COUNT" Monroe Wayne Monro. ' PLEASE ALSO REPORT YOUR FISHING FOR SALMON AND STEELHEAD ON STREAMS NOT LISTED. 8. DID YOU FISH IN 1976 FOR BROOK. BROWN, RAINBOW, LAKE TROUT. OR SPLAKE ON INLAND LAKES OR STREAMS? — YES [3 Please complete TABLE E on the other side of this sheet NOD 95 bl (.0 3‘ . 2 3 4 75 ‘“ Ta ‘= 3 TABLE C - GREAT LAKES SALMON AND TROUT FISHING «HERE 3: ‘I’Cd F 5:4 =0: SA..N.CN Ah: 2»: A' ; How MANV = SH 5:5 vcu Ct-TC— :40 wow ON ME 3REAT LAKEE7 « um .owxom KEE° £3 EACH ;C.C.~'~T.:-\‘ i . NUMBER A 2.91:1 ' I fl 1 v v I I ,cc mm; 0: HARBOR c3u~fv on i NUMSER 1 -°: g g— g g: * g2 * .2 Lo" BAY open 9.x: on NEAREST 0: DAYS 3;; g; , g; i a! j ‘35 CONNECTING WATER 'owN on CW mas: l in 3 g- - n- ‘ - :5. . .3 3, l - '5—7 ‘!--‘0‘ 31-121 3.3-H, ‘ ’23—25 . [26—26 '20-11 25—‘7 35—“ E; .3-«9 Mm“. gar. Humvee 5 ~' In I j I . . 2 . I ‘3 I I i .4 ' f T I TABLE D - STREAM FISHING FOR SALMON AND STEELHEAD 3r; .‘.hICH’ g'STED STREAMS 310 :Av! r HOW MANV W94 9*; VOL; (”skin ANC VO'.’ "3H ‘0‘? SALMON AN: STEELHEAD" 551*: .OU'CN’ [35: AT EACH Lofgflcg‘m‘ ~‘.U.‘.‘.EER m JCS- NAME :3: - ,,... . . I 9 ‘ = V x = .3: wear ’7: “O:"‘Y,?R ”Pies“ iv ' 2: 3? g ‘ '55" ‘30 STPE-‘d; 3:49an O’ U‘YS : 5 f l g: < E I ' .. A P .V F S " 6 = . V ‘ I = I : 4 ,Cr nomod che OWN VR ”I" ' l HCD l 7. j I 1 .4” ' ‘m 5-7 5-10. '1—12 ' 315-14: ' :25-23: ; {as—37 ‘3e—-40\ : ‘o-J: Ex WIN-'0 Rne' Mus-99m 5 A t 22‘ . :2 2.5 324 TABLE E - INLAND LAKE AND STREAM FISHING son TROUT - I - '. , - - - CA“ LT now MANY =-3r- 3:: YOU CATCH Ah: 9 r~ ,‘ _ ' u 9 I «:95 em v _.. ls.- OF $04! sw- .:m «.mma m.nmam h.ommm o.oHHm Mm\b mm. mm. no. oH.H om. ma. om. mm. .mn m.m~I ~.oHHI ~.>I o.omI m.mmI h.noI m.h>I o.oHHI unnumcoo a m n o m e m m H .0 mafiamaa oflcosflmm moxmq umouw mcflHQqumucflz MOM oocflmmo muosooum on» ecu mcoflumovm panama .mm manna 107 moo. mom. Hso. omo. mom. oom. mmo. mm oooo. ooo. HHo. Hoo. mooo. ooo. moo. mzoozH I I I I I I I zomzH I I I I AH. I I zcozH I I I I I I I mezH I I I I I I I mm H.¢Hma H.emma o.m~ma v.omom «.mmam om\b mm.I mo. oo. om. oH.A o.~I we. om. oo.~ Hm. Ann o.mal v.mmI m.mmI m.o~| ~.omHI m.mm| H.mI H.HmI ~.H~mI H.Hol ucmumcoo oH o m A o m e m N H mowamod pocoEHmmIcoz moxmq umouu mcflnmmlumucflz ecu conflmmo muooooum one HON mcowumovm panama .on manna 113 moo. mom. Nam. ooh. oav. omm. m moo. Hoo.I voo.I Hoo.I moo.I moo. WEOWZH I I I I Am. I zamzH I I oo. I oo.N oo. zaozH I I I I I I mezH I I I I I I mmqo I mo. I I I I moo I I I I I I NH I I I I I I HA I I I NH. I I oo I I I I I I o I I I I I I o I I I I I I A I I I I I I o I II I I I I m I I I I I I o I I I a I I m I I I I I I N I I I I I e H nmuououmnom A.omoN o.oNHH H.vo o.ooom m.moooo A.ooo No\n oA. oo. NH. AN.H Hm.a oA.I .mn o.HHHI A.oHI A.oI o.HAHI o.NoNI m.A ncmumcoo o m o m N H No mcflamcfl cocOEHmecoz moxmq ummnu umEESm HOm omoommo muosooum mnu New mGOAumowm panama .Ham moose 114 Aoo. on. vo. ooo. moN. moo. Nm moo. Aooo. I ooo.I Aooo.I Nooo.I msoozH I oN. I I Ao. I zmmzH I I I I I I smozH IN. I I NN. I oo. mszH I I I I I I mNHo I I I I I I . moo e I I I I I NH I I I I I I HH I I « mo. I I oH oo. I I I I I o I I I I e I m I I I I I e h I I I I I I m I I I I I I m om. I I I I I o I I I I I No. m I I I I I I N I I I I I I H mousuwumnom o.omoN o.NAmH H.mmA N.Nmo m.oooN H.oom Hm\n HN. oN. oN.I mA.I oN.I No. Hon N.AoHI m.ooI NN. m.oo m.NoI o.oI ucmumcoo H NH HH oH o o A .o romocHucooo .HHm mHnma 115 mmN. Amm. mom. mmN. mom. mom. Nm ooo. Hoo.I Hoo.I Hoo.I Ho.I Hoo. mzoozH I I I I oo. I zamzH I I I I om. I zeozH I I I I I I mszH I I I I I mo. m