INFORM ATION TO USERS The m ost advanced technology h as been used to photo­ graph and reproduce th is m anuscript from th e microfilm m aster. UM I film s th e te x t d irec tly from th e o rig in a l or copy subm itted. T hus, some th esis and dissertation copies are in ty p ew riter face, w hile others m ay be from any type of com puter printer. The q u a lity of th is re p ro d u ctio n is d ep en d en t upon th e quality of th e copy subm itted. B roken or indistinct p rin t, colored or poor q u a lity illu s tra tio n s an d p h o to g rap h s, p rin t bleedthrough, su b stan d ard m argins, and im proper alignm ent can adversely affect reproduction. In th e u n lik ely event th a t th e author did not send UM I a complete m an u scrip t and th ere are m issing pages, these w ill be noted. Also, if u n a u th o riz e d copyright m a te ria l had to be removed, a note will indicate th e deletion. Oversize m aterials (e.g., maps, draw ings, charts) are re ­ produced by sec tio n in g th e o rig in al, b eg in n in g a t th e upper left-hand corner and continuing from left to rig h t in equal sections w ith sm all overlaps. Each original is also photographed in one exposure and is included in reduced form at th e back of th e book. These are also available as one exposure on a stan d ard 35mm slide or as a 17" x 23" b lack a n d w h ite p h o to g ra p h ic p r in t for a n a d d itio n a l charge. P h o to g rap h s in clu d ed in th e o rig in a l m a n u sc rip t have been rep ro d u ced x e ro g rap h ically in th is copy. H ig h e r q u ality 6" x 9" b lack an d w h ite p h otographic p rin ts are available for any photographs or illu stratio n s appearing in th is copy for an ad d itional charge. C ontact UM I directly to order. UMI U n iversity M icro film s Intern ation al A Bell & Howell Inform a tion C o m p a n y 30 0 N o rth Z e e b Road, A n n Arbor, Ml 48106-1346 USA 3 1 3 /7 6 1 -4 7 0 0 8 0 0 /5 2 1 -0 6 0 0 O rder N u m b er 8923834 D is tr ib u tio n and a b u n d a n ce of ich th yop lan k ton in n earsh ore Lake M ich ig a n near L u d in gton an d p o te n tia l tra n sp o rt from a tr ib u ta r y m arsh Brazo, Dan C., Ph.D. Michigan State University, 1989 UMI 300 N. ZeebRd. Ann Arbor, MI 48106 DISTRIBUTION AND ABUNDANCE OF ICHTHYOPLANKTON IN NEARSHORE LAKE MICHIGAN NEAR LUDINGTON AND POTENTIAL TRANSPORT FROM A TRIBUTARY MARSH By Dan C. Brazo A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 1989 ABSTRACT DISTRIBUTION AND ABUNDANCE OF ICHTHYOPLANKTON IN NEARSHORE LAKE MICHIGAN NEAR LUDINGTON AND POTENTIAL TRANSPORT FROM A TRIBUTARY MARSH By Dan C. Brazo Effects of several physical factors (e.g., wind direction, water current, and water temperature) on distribution and abundance of ichthyoplankton in Lake Michigan were studied from 1978 to 1981 on the east-central shore of Lake Michigan. Further, passive transport of fish larvae from a local tributary-marsh was documented. All samples were made with a 1-meter diameter, 3 51 li­ me sh, conical plankton net towed obliquely or mounted on a sled. Four transects were established in Lake Michigan with samples taken at five depth contours, day and night from mid-April through September. Four additional stations were established in Pere Marquette Marsh-Lake in 1981. Alewife (Alosa pseudoharenqus), yellow perch (Perea flavescens), rainbow smelt (Osmerus mordax), and johnny darter (Etheostoma nigrum) larvae were most numerous in Lake Michigan samples, while common carp (Cyprinus carpio), Pomoxis spp., Lepomis spp., yellow perch, and clupeids dominated Pere Marquette Marsh-Lake samples. Mean peak densities of most species in Lake Michigan were considerably higher in 1978 (N = 256) than 1979 (N = 738): alewife, 2618 vs. 623/1000 m 3; rainbow smelt, 1123 vs. 322/1000 m3 ; yellow perch 749 vs. 143/1000 m 3; johnny darter, 669 vs. 295/1000 m3 ; Cottus spp., 59 vs 33/1000 m 3; ninespine stickleback, 126 vs. 262/1000 m 3 ; and spottail shiner, 123 vs. 102/1000 m3 . During 1979, prolonged periods of easterly and northerly winds caused cold hypolimnetic water to upwell and produced a relatively "cold" year on Lake Michigan compared to 1978 which reduced densities of fish larvae. Passive transport of fish larvae from the Pere Marquette Marsh-Lake system was relatively low when contrasted with abundance of fish larvae in nearshore Lake Michigan. Alewife, yellow perch, common carp, gizzard shad (Dorosoma cepedianum), Pomoxis spp., johnny darter, lake whitefish (Coreqonus clupeaformis), and rainbow smelt were all found in the outlet channel, but only the first six taxa apparently hatched in Pere Marquette Marsh-Lake. Further, of these six taxa, only yellow perch were observed in substantial densities with nearly 0.75 million larvae transported to Lake Michigan. Other taxa were either transported in numbers that were inconsequential compared with larval fish densities in Lake Michigan or were collected in such low numbers in Lake Michigan that it suggests: 1.) high mortality of transported larvae or 2.) non-transport of larvae (e.g., Pomoxis spp., gizzard shad, and common carp). ACKNOWLEDGEMENTS The completion of this research would not have been possible without significant input from several sources. Accordingly, I would like to thank the following people. Dr- Niles Kevern, chairman of the Department of Fisheries and Wildlife, and Dr. Charles Liston, chairman of my graduate committee, intervened on my behalf on several occasions to ensure all requirements were met and accepted. Other graduate committee members Drs. Pat Muzzall, Dave Jude, and Cal McNabb are all gratefully acknowledged for their various inputs that brought this research to fruition. Their input has made this dissertation a far better piece of work than the initial product and they have shown me the more scholarly thought processes required of a successful Ph. D. candidate. Consumers Power Company provided funds to accomplish this study and additional support was provided by the Michigan Agricultural Experiment Station. The many graduate students, undergraduate interns, and staff members who aided in data collection and sanity protection are too numerous to mention individually, but the efforts of Joe Bohr, Rich O ’Neal, Leo Yeck, Rick Ligman, Lisa Barnese, Walt Duffy, Greg Peterson, Fred v Koehler, Sara Chubb, Shirley Ehler, and Pat Carlson were all instrumental in ensuring the above two objectives were met. The input of Barb Poppema and Karen Braun cannot be envisioned by solely reading the final manuscript, but their outstanding typing ability, word processing knowledge, endurance, and humor through a myriad of changes made the final product possible. Finally, without the mental and physical support, sacrifice, and endurance given by my family (especially my wife Sue and my mother Lucille) this project would never have been completed. I would also like to acknowledge the memory of my father who first introduced me to the wonders of our natural world. TABLE OF CONTENTS Page LIST OF TABLES......................... LIST OF FIGURES......... ........................ X xvii INTRODUCTION................................. 1 METHODS.................. 7 Lake Michigan...... Pere Marquette Marsh-Lake...................... 9 14 DESCRIPTION OF THE STUDY AREA. ..................... 16 RESULTS..................... 24 Alewife ............. Lake Michigan. .... Temporal distribution....... Spatial distribution.......... Diel distribution....................... Total abundance............ Pere Marquette Marsh-Lake ........ General distribution.................... Inputs to Lake Michigan................. 27 27 27 32 35 37 39 39 41 ............ Rainbow Smelt Lake Michigan. ............... Temporal distribution....... Spatial distribution.................... Diel distribution....... Total abundance...... Pere Marquette Marsh-Lake................... General distribution............. Inputs to Lake Michigan................. 42 42 42 45 47 47 50 50 51 Yellow Perch...... Lake Michigan...... Temporal distribution ............. Spatial distribution. ........... Diel distribution................ Total abundance............... Pere Marquette Marsh-Lake................... General distribution......... Inputs to Lake Michigan................. 52 53 53 53 56 56 59 59 60 vii Johnny Darter.......................... Lake Michigan............................... Temporal distribution................... Spatial distribution.................... Diel distribution....................... Total abundance......................... Pere Marquette Marsh-Lake................ General distribution............. Inputs to Lake Michigan................. 60 61 61 63 65 67 69 69 70 Cottus spp...................................... Lake Michigan............................... Temporal distribution................... Spatial distribution.................... Diel distribution....................... Total abundance....... Pere Marquette Marsh-Lake................... General distribution.................. 71 71 71 72 75 76 78 78 Ninespine Stickleback........................... Lake Michigan............................... Temporal distribution................... Spatial distribution.................... Diel distribution....................... Total abundance...... Pere Marquette Marsh-Lake................... General distribution.................... 78 79 79 81 81 83 83 83 Spottail Shiner..................... Lake Michigan. ............................. Temporal distribution. ............... Spatial distribution.................... Diel distribution............. Total abundance.................. Pere Marquette Marsh-Lake................... General distribution.................... Inputs to Lake Michigan................. 86 86 86 87 90 90 90 90 93 Deepwater Sculpin.............................. . Lake Michigan............................... Temporal distribution................... Spatial distribution.................... Diel distribution....................... Total abundance......................... Pere Marquette Marsh-Lake................... General distribution.............. 93 94 94 95 95 98 98 98 Lake Whitefish.................................. Lake Michigan.......... Temporal distribution................... Spatial distribution.................... Diel distribution....................... Total abundance......................... viii 100 101 101 101 103 103 Pere Marquette Marsh-Lake................... General distribution.................... 103 103 Burbot .. Lake Michigan............................... Temporal distribution................... Spatial distribution.................... Diel distribution....................... Total abundance......................... Pere Marquette Marsh-Lake................... General distribution.................... 106 107 107 107 109 109 112 112 Other Species............ 112 DISCUSSION........................................... Distribution and Abundance...................... Diel Distribution............... Importance of Inputs to Lake Michigan from Pere Marquette Marsh-Lake.......... 117 117 124 SUMMARY............... 132 LIST OF REFERENCES.................................. 145 APPENDIX A. APPENDIX B. APPENDIX C. APPENDIX D. 126 Water temperatures in Lake Michigan 1978 - 1979.......................... 155 Densities of ichthyoplankton in Pere Marquette Marsh-Lake................... 156 Fisheries survey of Pere Marquette Marsh-Lake, 1981....................... 166 Alphabetical list of common and scientific names of fish used in this manuscript. From Robins et al. (1980).......................... 167 ix LIST OF TABLES Table 1. 2. Page Volume of water (thousands of m 3) in an imaginary wedge with a surface rectangle of 2.4 x 9.7 km in Lake Michigan near Ludington, Michigan. Depth and width of contours based on Army Corps of Engineers monthly mean lake levels and base datum from 1975. Description of Lake Michigan sampling transects used during 1978, 1979, and 1981. 13 17 3. Physical and chemical data from Pere Marquette Marsh- Lake taken during larval fish tows, 1981. (WT = water temperature, 0C; DO = dissolved oxygen, mg/1; Cond = conductivity, ymho/cm at 25 C; pH = negative log of hydrogen ion concentration; S = surface; B = bottom) 20 4. Water velocity (V) and discharge (D) in the Pere Marquette Lake (PML) outlet channel to Lake Michigan, 1981. 23 Sampling dates, locations, and number of ichthyoplankton samples taken in Lake Michigan and the Pere Marquette Marsh-Lake 1978, 1979, and 1981. 25 Total number of larval fish collected from 1978 to 1981 in Lake Michigan, Pere Marquette Marsh-Lake, and its connecting channel. 26 5. 6. 7. 8. 9. Average nighttime density and range (in parentheses) of alewife larvae (number/1000 m3 ) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 29 Number of days per month wind direction originated from given compass points near Ludington, Michigan. Data from National Oceanic and Atmospheric Administration (NOAA) Marine Coastal Weather Log - Ludington Station. 33 Mean total length (mm) of alewife larvae at depth contours in Lake Michigan, 1978 and 1979. Number measured in parentheses. 34 x Table 10. 11. Page Comparison of average day - night densities of alewife larvae (number/1000 m3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 36 Estimates based on night densities and 90% confidence interval (+) of alewife abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 38 12. Average nighttime density and range (in parentheses) of rainbow smelt larvae (number/1000 m3 ) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 44 13. Mean total length (mm) of rainbow smeltlarvae at depth contours in Lake Michigan, 1978 and 197 9. Numbers measured in parentheses. 46 14. Comparison of average day - night densities of rainbow smelt larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 15. 16. 17. 48 Estimates based on night densities and 90% confidence interval (+) of rainbow smelt abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 49 Average nighttime density and range (in parentheses) of yellow perch larvae (number/1000 m 3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 54 Comparison of average day - night densities of yellow perch larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 55 xi Page Table 18. 19. 20. 21. 22. 23. 24. 25. 26. Mean total length (mm) of yellow perch larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. 57 Estimates based on night densities and 90% confidence interval (+) of yellow perch abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 58 Average nighttime density and range (in parentheses) of johnny darter larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 62 Mean total length (mm) of johnny darter larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. 64 Comparison of average day - night densities of johnny darter larvae (number/1000 m 3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 66 Estimates based on night densities and 90% confidence interval (+) of johnny darter abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number ofsamples. 68 Average nighttime density and range (in parentheses) of Cottus larvae (number/1000 m 3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means nodata. 73 Mean total length (mm) of Cottus spp. larvae at depth contours in Lake Michigan, 1978 and 1979. Number measured in parentheses. 74 Comparison of average day - night densities of Cottus larvae (number/1000 m3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D - day, N = night. 76 xii Table 27. 28. 29. 30. 31. 32. 33. 34. Page Estimates based on night densities and 90% confidence interval (+) of Cottus spp. abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 77 Average nighttime density and range (in parentheses) of ninespine stickleback larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 80 Mean total length (mm) of ninespine stickleback larvae at depth contours in Lake Michigan, 1978 and 1979. Number measured in parentheses. 82 Comparison of average day - night densities of ninespine stickleback larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 84 Estimates based on night densities and 90% confidence interval (+ ) of ninespine stickleback abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 85 Average nighttime density and range (in parentheses) of spottail shiner larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 88 Mean total length (mm) of spottail shiner larvae at depth contours in Lake Michigan, 1978 and 1979. Number measured in parentheses. 89 Comparison of average day - night densities of spottail shiner larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 91 xiii Page Table 35. 36. 37. 38. 39. 40. 41. 42. Estimates based on night densities and 90% confidence interval (+) of spottail shiner abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 92 Average nighttime density and range (in parentheses) of deepwater sculpin larvae (number/1000 m 3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 96 Comparison of average day - night densities of deepwater sculpin larvae (number/1000 m3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 97 Estimates based on night densities and 90% confidence interval (+) of deepwater sculpin abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 99 Average nighttime density and range (in parentheses) of lake whitefish larvae (number/1000 m 3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 102 Comparison of average day - night densities of lake whitefish larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 104 Estimates based on night densities and 90% confidence interval (+) of lake whitefish abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 105 Average nighttime density and range (in parentheses) of burbot larvae (number/1000 m 3 ) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. 108 xiv Page Table 43. 44. IA. Comparison of average day - night densities of burbot larvae (number/1000 m 3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N - night. 110 Estimates based on night densities and 90% confidence interval (+) of burbot abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 111 Night bottom temperatures (C) at several depth contours on larval fish sampling transects in Lake Michigan 1978 - 1979. S = shallow depth contours, 1 m; I = intermediate depth contours, 3 - 6m; D = deep contours, >6 m. 155 IB. Densities (number/1000 m 3) of alewife larvae in the Pere Marquette Marsh and at the 1-m contour 156 in Lake Michigan, 1981. N = night, D = day. 2B. Densities (number/1000 m 3 ) of rainbow smelt larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. 157 Densities (number/1000 m3 ) of yellow perch larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. 158 Densities (number/1000 rn3 ) of johnny darter larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. 159 Densities (number/1000 m 3 ) of spottail shiner larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. 160 3B. 4B. 5B. 6B. Densities (number/1000 m 3 ) of trout perch larvae in the Pere Marquette Marsh and at the 1-m contour 161 in Lake Michigan, 1981. N = night, D = day. 7B. Densities (number/1000 m3 ) of gizzard shad larvae in the Pere Marquette Marsh and at the 1-m contour 162 in Lake Michigan, 1981. N = night, D = day. xv Table 3B. 9B. ?age Densities (number/1000 m 3) of Pomoxis spp. larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. 163 Densities (number/1000 m 3) of Lepomis spp.larvae in the Pere Marquette Marsh and at the 1-m contour 164 in Lake Michigan, 1981. N = night, D = day. 10B. Densities (number/1000 m 3) of common carp larvae in the Pere Marquette Marsh and at the 1-m contour 165 in Lake Michigan, 1981. N = night, D = day. IC. ID. Total number of juvenile and adult fish collected with several methods in the Pere Marquette LakeMarsh system, 1981. 166 Alphabetical list of common and scientific names of fish used in this manuscript. From Robins et al. 167 (1980). xvi LIST OF FIGURES Page Figure 1. 2. Schematic diagram of larval fish sampling transects (T) in Lake Michigan and stations (S) in Pere Marquette Marsh-Lake. Dotted line depicts reference wedge (2.4 x 9.7 km) used to estimate larval fish abundance. 10 Schematic cross section of the Pere Marquette Marsh-Lake harbor channel depicting sites for measurement of current velocity (®), area of each component cell (m2 ) for calculating discharge of larvae, and other relevant measurements. Not to scale. 21 Mean (-) and range (shaded area) of bottom temperatures on sample nights at 3- to 12-m depth contours at all transects in Lake Michigan, 1978-1979. 30 xvii INTRODUCTION In most cases, fishes rely on success of natural spawning to ensure population maintenance and the most appropriate time to estimate, year-class success may not be at the recruitment stage but rather at the early life (larval) stages (Gulland 1965; Hempel 1965). However, scientists have only recently begun to examine in detail the ecology of early life history stages of fishes. Much of the earlier work was hampered by difficulties in identifying fish larvae, but many of these deterrents have been resolved except for some taxonomic groups that present problems even as adults (e.g., minnows, darters). Though major predators in the current Lake Michigan fishery community do not rely on natural reproduction (but are rather intensively managed to maintain population size) their dependence on a forage base comprised primarily of two exotic species, alewife (common and scientific names of fish used in this manuscript are given in Appendix Table ID) and rainbow smelt, suggests the importance of understanding the biology of forage fish as well. Again a paucity of data exists on early life stages of these fish in the Great Lakes except for work around power plants and general surveys (Jude et al. 1979a, 1980; Cima et al. 1975; 1 2. Wells 1973; Liston et al. 1980, 1981; and others). Important indigenous species of Lake Michigan, including yellow perch, lake whitefish, round whitefish, and bloater, have been or are exploited by commercial and sport fishermen, yet biology of early life history stages is poorly known. One of the primary objectives of my research was to study the temporal and spatial distribution of fish larvae and physical factors affecting their distribution. One of the most important factors affecting larval fish distribution, indeed that of most planktonic organisms, is water current (Arnold 1974). Wind speed and wind direction are major factors affecting water current in lentic habitats (Norcross and Shaw 1984). Wind generates water currents that flow clockwise (to the right of the wind vector) in the northern hemisphere and counter-clockwise (to the left) in the southern hemisphere. The angle between wind and current movement is small at the water surface, but this angle increases with deeper currents and net transport approaches straight down (90 degrees from wind direction). This phenomenon is known as Eckman spiral (Eckman 1905) and is responsible for upwellings of cold, nutrient-rich water from the bottom of large lakes. These currents can transport larvae for long distances and through a wide variety of ecological conditions, both favorable and detrimental (Walford 1938, Bishai 1960, Lasker 1981). Effects of water transport on fish larvae can be conceptualized in two ways: 1.) on movements of fish-food organisms to fish larvae or 2.) on movement of the fish larvae to unfavorable, similar, or more favorable environments. These factors may in turn influence year class strength (Norcross and Shaw 1984). Movement and availability of food organisms would have major impacts on larval fish. The "critical period concept", first developed by Hjort (1914), and reviewed by May (1974) observed that survival may be affected by lack of food at the time of first feeding of larval fish. Indeed, Hjort (1914) considered this aspect most important in larval fish survival, and Hunter (1981) and Lasker (1981) have shown that deprivation of an exogenous food source at time of yolk absorption, and even before (Toetz 1966, Hokanson 1978), can result in high mortalities due directly to starvation and increased susceptibility of starved larvae to predation and other deleterious agents. If currents repeatedly carried away food sources, one could expect substantial effects on year-class strength. Lasker (1981) further suggested favored foods of larvae may be transported away and replaced by less desirable food (e.g., kinds, size, etc.). It is well known that newly hatched larval fish generally eat small, slow-moving zooplankton (Toetz 1966, Siefert 1972, and Nigro and Ney 1982). Thus food densities may remain similar, but final results would again be a severe impact on year-class strength. 4 Passive transport of larvae may be equally critical in determining year-class strength. May (1974) and Hjort (1914) suggest that larvae may be transported to areas of low food abundance or poor quality food. Further, larvae may be transported to physical habitat (water temperature, toxic substances, etc.) that is less than optimal. Extreme changes in water temperature may alone be lethal though Lasker (1981) points out that under natural conditions this is rarely the case. More importantly, rapid temperature changes may cause abnormal growth and development such that mortality follows. If by chance major physical events (e.g., upwelling, longshore currents, seiches, overturn) on large bodies of water, as Lake Michigan, coincide with peak larval fish densities, strong impacts (positive or negative) on year-class strength and population structure may result. Exact mechanisms responsible for survival and mortality remain to be fully identified and indeed are most likely a combination of several factors. A second main objective was to estimate local abundance of ichthyoplankton. Abundance of larval fish may provide an initial indicator of year-class strength, and can be critical when estimating impact of various natural phenomenon. Water intakes (e.g., for power plants, municipal or industrial, or agricultural purposes) have the potential to reduce larval fish abundance, at least on a local level. Introduction of toxic substances may also have severe impacts on ichthyoplankton populations. 5 If ichthyoplankton densities are severely reduced owing to any of the above events, ichthyoplankton recruitment from tributaries to Lake Michigan becomes increasingly important. The other main objective of this study was to document transport of ichthyoplankton to Lake Michigan from a major tributary. Scientists have speculated on the importance of tributary marshes as spawning and nursery sites for many species of fish normally existing in Lake Michigan. The importance of estuarine, salt-grass marshes as nursery and spawning areas for many species of marine fish is well established (Shenker and Dean 1979; Weinstein 1979; Hodson 1979; Birkhead et al. 1977; and others). Copeland (1965) observed that large numbers of organisms entered the Gulf of Mexico through a bridge pass from an estuarine embayment which contributed markedly to the biomass of a commercially important species. Similar types of data from freshwater environments are scanty. Part of the reason may be the lack of extensive tributary marsh systems in the Great Lakes that are analogous to the salt-grass, tidal tributaries of marine systems. The Great Lakes may be viewed as inland freshwater oceans with associated freshwater marsh systems. Recently Mansfield (1984) reported on inputs of larval fish to Lake Michigan from a small tributary stream in the southeastern basin. However, Herdendorf and Hartley (1981) in a Great Lakes wide survey continually refer to the lack of adequate fishery data from these marshes. METHODS Larval fish were sampled approximately once every 2 weeks from mid-April to late September in 1978, 1979, and 1981. All samples were taken with standard gear consisting of a 1-m diameter, conical plankton net, having a 351micron-mesh aperture and a 3:1 net-length-to-mouth-diameter ratio. A 3.6-kg brass depressor was attached to the bridle and a flow meter (General Oceanics, model number 2030) was mounted one-third off center in the net mouth to measure volume of water strained (UNESCO 1968). During the entire study, an average of 126 m 3 (SD = 53) of water per sample was strained. Nets were washed down by pump or buckets of water, and condensed samples were preserved immediately in 10% formaldehyde solution. A few random samples from 1978 plankton nets were washed a second time to check error rate associated with net washing. A mean of 4 larvae per sample as found in the second washing and nearly all were alewife. Based on percentages, 2.3% of the larvae present in the first sample were recovered in the second rinse. Most of these larvae were assumed to be clinging to the net and personal observations suggest percentage of loss was slightly greater when large amounts of debris or algae were present. Samples were sorted in the laboratory over light 7 8 and dark backgrounds with a lighted magnifying lens. Larvae were removed to Davidson’s solution (Lam and Roff 1977) to maintain flexibility. For each species up to 10 larvae per sample were randomly selected and measured with an ocular micrometer mounted in the microscope lens and calibrated to 0.1 mm. Nearly 24% of the 1978 samples were examined a second time to identify sorting accuracy. The mean percentage of missed larvae per sample was 4.9% with a range of percentages from 0 to 42%, In only 8 of 61 samples was the percentage of misses greater than 10%, and in over half of the samples no larvae were found in the second sorting. Alewife and rainbow smelt larvae were most often missed which is not unusual considering their small size. Some burbot and yellow perch were also found on second sorting. The presence of these larger and more robust larvae in second sort samples suggests factors other than physical appearance of larvae may affect sorting efficiency. Individual skill at sorting larvae was an important factor. Of the eight samples in which greater than 10% of the larvae were missed, six samples were initially picked by the same person. Keys and descriptions by Dorr et al. (1976), Hogue et al. (1976) , Lippson and Moran (1974), and Nelson and Cole (1975) were most useful as supplemental aids in identification of larvae, though many of the identifications were based on personal expertise and a 9 knowledge of major species and their spawning times in the study area. Identification verification for some larvae were provided by the Great Lakes Research Division, University of Michigan, when taxonomic problems were unresolved through the literature or in-house. Statistical analyses were conducted on non-transformed density data using non-parametric techniques. Only data for alewife, rainbow smelt, and yellow perch were statistically analyzed because of their greater numbers. Day and night data were compared using the Wilcoxon signed rank test for paired observations (Sokal and Rohlf 1973). An alpha (a) level of <0.1 was used to determine significance. Comparisons of densities among years and stations were made with the Mann-Whitney U-test. Probabilities greater than 0.1 were considered non­ significant . Lake Michigan Collections in Lake Michigan were taken as a series of day and night samples in 1979, but only during the night in 1978 and 1981. No samples were taken in 1980 due to personnel and monetary constraints. In 1979, day samples were usually taken from 0930 - 1500 hours and in all years night samples were taken from 2100 - 0130 hours depending on time of year. In Lake Michigan, simultaneous duplicate samples were taken at the 1-, 3-, 6~, 9-, and 12-m depth contours at each of four transects (Figure 1). Only 10 Figure 1. Schematic diagram of larval fish sampling transects (T) in Lake Michigan and stations (S) in Pere Marquette Marsh-Lake. Dotted line depicts reference wedge (2.4 x 9.7 km) used to estimate larval fish abundance. 11 LUDINGTON PEP B U TTERSVILLE T«p PAR K Km 3m 9m TS 'A i LU D ING TO N ■/(‘ P O W ER P L A N T A IN T A K E S U S -31 fT2] 9m 12m 3m 6m ,LUD1NGT0N S U M M IT |* T w p PA R K 12 transects 1 - 3 and contours .1- 9 m were sampled in 1978, and only transects 3 and 4 at the 1-m contour were sampled in 1981. The 1-m contour was sampled with the standard net attached to an aluminum sled that was towed with a boat, but all other samples were taken with standard nets towed in a stepped oblique fashion. Tow times were 5 min and vessel speed was maintained between 0.6 - 0.9 knots. The total number of larvae within a reference wedge (approximately 2.4 km x 9.7 km) in Lake Michigan was estimated on sample days. The rectangular surface of the wedge bounded an area from the shore out to the 13.7-m contour east and west, and approximately from Buttersville Park to Summit Park 4.5 and 14.2 km, respectively, south of Ludington (Figure 1). The wedge was comprised of several cells with one cell for each contour sampled. Sample depth was considered to be the mid-point between the edges of a cell, i.e., the cell delineated by the 10.7- and 13.7-m depth contour contained the 12.2-m sample contour as mid­ point, and so forth shoreward. By calculating the volume (Table 1) of a cell and multiplying by the larval fish density found within the same cell, the total number of fish larvae for the cell was estimated. Only data from night time samples were used in calculating abundance. Total number of larvae for the wedge was estimated by summing the number of larvae calculated for all cells sampled. Table 1. Contour Sampled (m) 1 Volume of rectangle Depth and mean lake water (thousands of m 3) in an imaginary wedge with a surface of 2.4 x 9.7 km in Lake Michigan near Ludington, Michigan. width of contours based on Army Corps of Engineers monthly levels and base datum from 1975. Range of Depth (m) 1.5 shore - 1 222 2 179 Transect 3 199 4 260 Total 862 3 1.5 - 4.6 1789 1414 1751 1498 6453 6 4.6 - 7.6 4036 3509 3689 3509 14,743 9 7.6 - 10.7 12,410 11,588 13,689 17,861 55,549 12 10.7 - 13.7 30,389 22,982 29,918 27,601 110,890 shore - 13.7 48,847 39,674 49,247 50,730 188,499 Total 14 Day and night data taken in 1979 were used to determine periods of greatest densities and to document some behavioral and ecological factors. Additional laboratory observations were made on larval fish behavior during daylight. Fish were collected with a 2-m diameter plankton net on a vertical tow. bucket and sorted immediately. Samples were washed into a Live fish larvae were transferred to 10-L aquaria with ambient lake water and observed ad libitum. No attempt to feed the fish larvae was made. Pere Marquette Marsh-Lake Pere Marquette Marsh-Lake ichthyoplankton collections were made weekly in May and June and once every 2 weeks in April, July, and August, 1981. Day and night samples were taken at four stations (Figure 1). 1-3 Collections at stations were thought to be representative of fish larvae spawned in the marsh, and station 4 provided data on whether the marsh was a conduit for larval fish movement into Lake Michigan. Duplicate samples at stations 3 and 4 were collected in a 5-minute stepped oblique tow unless current data indicated water was flowing into the marsh from Lake Michigan. On these occasions duplicate 5-minute oblique tows were made to sample cells of water flowing only in a discrete direction. Duplicate samples at stations 1 and 2 were collected by sled tows identical to techniques described earlier in Lake Michigan. Estimates of the number of larvae being transported through the harbor channel (station 4) were made in the following manner. The channel was partitioned in half vertically and currents were estimated by measuring flow with an Endeco Type 110 Recording Meter which recorded current velocity, direction, water temperature, and depth. The three-point method of estimating discharge (Hynes 1970) was used and current readings were taken at 0.8, 0.5, and 0.15 of the depth from the surface at the horizontal midpoint across each channel half. Volume of water discharged for a 24-hour period was then estimated by multiplying by the appropriate time factor. The total number of larvae carried into Lake Michigan or into Pere Marquette Marsh-Lake was estimated by multiplying the density of ichthyoplankton by the total volume of water moved over the 24-hour sampling period. DESCRIPTION OF THE STUDY AREA The Lake Michigan study area was approximately 4.5 to 14.2 km south of Ludington, Michigan and extended from the beach out to the 12-m contour. Bottom substrates were primarily sand with irregular outcroppings of clay and large rocks (Table 2). Some distinct differences occurred along sampling transects. Transects 1 and 4 had similar substrates, primarily sand with a band of gravel near the beach out to approximately the 1.5~in depth at transect 1 and a similar band occurred at the 6-m contour at transect 4. However, the nearshore zone is highly dynamic and shifts in sand continually cover and expose gravel bars. Additionally, transect 4 is located within 3 to 4 km of the mouth of the Pere Marquette Marsh-Lake. Bottom substrates at transects 2 and 3 were considerably more rugose and were characterized by gravel nearshore and a considerable number of large rocks, clay outcroppings, and other debris out to the 12-m contour. Physical and chemical data in the study area were given an extensive treatment by Liston et al. (1976), and were indicative of an oligotrophic lake (Brazo and Liston 1979). Water temperature was typically 2 - 3 C at the onset of sampling in mid-April and rose slowly to summer 16 Table 2. Transect Description of Lake Michigan sampling transects used during 1978,. 1979, and 1981. Location Depth Bottom Substrates 11.5 km south of Ludington harbor Shore to 12-m contour Fine sand, some gravel beds particularly nearshore 7.5 km south of Ludington harbor Shore to 12-m contour Sand to gravel with some large rock beds and boulders 5.5 km south of Ludington harbor Shore to 12-m contour Coarse sand and fine gravel with large boulders 2 km south of Ludington harbor Shore to 12-m contour Fine sand with some bands of gravel nearshore and at the 6-m contour. 18 maxima of 18 - 22 C in late July and August, followed, by autumn cooling when temperatures were 5 - 7 C by the end of November. However, summer water temperatures were strongly affected by upwellings which produce precipitous declines in water temperature. Liston and Tack (1976) recorded drops of 14 C in 17.5 hours near Grand Haven, approximately 90 km south of the present study. They recorded four periods of major upwelling throughout the summer of 1975 at this site. Other chemical parameters measured in the present study area included: dissolved oxygen, 8.9 - 15.0 mg/1; pH, 7.7 - 8.8; total alkalinity, 98 - 136 mg/1; dissolved solids, 161 - 200 mg/1; and turbidity, 0.3 - 17.0 FTU (Formazin Turbidity Units). Pere Marquette Marsh-Lake system is characteristic of a submerged river mouth entering Lake Michigan with the lower portion being primarily open water to depths of 11 m and encompassing 224 ha. Marsh-Lake Station 3 in Pere Marquette was located in this type of environment. The upper portion of Pere Marquette Marsh-Lake was more typically marsh habitat with extensive stands of emergent macrophytes, primarily cattail (Typha spp.) and burreed (Sparqanium spp.), and lush beds of submerged macrophytes, primarily milfoil (Myriophyllum spp.) and pondweed (Potamogeton spp.), bisected in several places by old and existing river channels. Stations 1 and 2 were located in the upper marsh which encompassed about one-half the area of the open water (I'll ha) and had an average depth of 1.5 13 m. Bottom substrates were typically soft and organic in nature in Pere Marquette Marsh-Lake and physical and chemical parameters measured during larval fish collections in 1981 are given in Table 3. In addition to tabled values, total alkalinity was measured on 21 May and 23 July and ranged from 136 - 148 mg/1 and 140 - 153 mg/1, respectively. Pere Marquette River harbor outlet to Lake Michigan (station 4) has been channelized to 9 m to allow large commercial boat traffic access to industry and railroad facilities along the northwestern shore of Pere Marquette Marsh-Lake. Directly south of the Ludington U.S. Coast Guard Station the channel is 116 m wide and has a cross sectional area of slightly greater than 1000 m 2 (Figure 2). Based on measurements made from May through mid-August, the ranges in velocity and discharge from Pere Marquette MarshLake to Lake Michigan were 2.5 (lower detectable limit) 48.5 cm/sec and 164,000 - 15,691,000 m 3 , respectively (Table 4). However, on several occasions (72% of the time) some water was flowing from Lake Michigan into Pere Marquette Marsh-Lake (usually along the bottom) and ranged from 2.5 - 21.5 cm/sec velocity. This phenomenon of reversing inlet currents is primarily due to seiching (Mortimer 1965) which is initiated by storm pressure and wind forces on the lake (Seelig and Sorensen 1977). Table 3. Physical-chemical data from Pere Marquette Marsh-Lake taken during larval fish tows, 1981. (WT = water temperature, °C; DO = dissolved oxygen, mg/1; Cond = conductivity, umho/cm at 25 C; pH - negative log of hydrogen ion concentration; S = surface; B = bottom) Upper Marsh Date WT S DO S Lower Pere Marquette Lake Cond pH S S WT S B DO Cond Outlet S B S B pH S WT S DO B S B Cond S B gH S 9 April 11 10.0 215 7.0 10 10 10.4 10.2 257 - 6.9 9 - 10.0 - 243 - 7.0 23 April 9 10.0 255 7.2 10 - 9.2 8.9 249 248 7.0 10 - 9.8 9.6 241 240 7.2 8.6 8.8 289 281 7.8 12 9 9.0 11.0 269 209 7.3 311 282 7.8 6 May 15 9.4 280 8.2 14 13 20 May 16 8.3 300 7.9 15 13 3 June 16 8.1 345 7.8 16 14 7.7 7.2 328 283 8.5 15 12 9.2 9.1 291 259 17 June 20 7.2 364 7.2 19 18 6.6 5.7 361 357 - 19 18 6.6 6.7 332 290 1 July 22 8.0 378 - - - 7.8 6.4 350 322 - 19 14 8.4 8.3 329 217 13 July 23 7.1 393 7.4 22 19 7.2 5.7 389 338 7.3 21 20 7.2 5.8 330 335 23 July 20 7.8 - 7.8 20 15 7.7 7.8 - - 7.9 17 9 7.8 7.8 - - 10 Aug. 22 5.8 360 7.6 21 16 5.1 4.2 365 325 8.3 20 19 6.4 5.6 280 330 8.0 7.5 15 10 ...................... 8.5 9.3 8.3 o 21 Figure 2. Schematic cross section of the Pere Marquette Marsh-Lake harbor channel depicting sites for measurement of current velocity (®), area of each component cell (m2 ) for calculating discharge of larvae, and other relavent measurements. Not to scale. ■115.8m0.9 m Z.lm 137.0 m 139.0 m (45.6 m 156.3 m ^3.3 m 9.1m L6.lm 149.8 m ,9.1 m 8.2 m 23 Table 4. Date Water velocity (V) and discharge (D) in the Pere Marquette Lake (PML) outlet channel to Lake Michigan, 1981. Diel Period PML to Lake Michigan D V (thousands (cm/sec) of nf3) Lake Michigan to PML V D (thousands (cm/sec) of nf5) 6 May Night Day 2.5 - 5.1 2.5 - 17.9 12 May Night 2.5 20 May Night 2.5 - 12.8 1,087 2.5- 7.7 794 3 June Night Day 2.5 - 25.6 2.5 - 15.3 2,790 2,598 5.1 2.5 - 5.1 482 646 11 June Night 5.1 306 5.1 - 12.8 2,345 17 June Night Day 7.7 - 20.4 7.7 - 25.6 2,625 5,840 2.5 2 .5 24 June Night 7.7 - 17.9 3,972 0 1 July Night Day 7.7 - 16.6 7.7 - 15.3 1,982 2,555 10.2 - 17.9 2.5 - 10.2 13 July Night Day 24.5 - 46.0 5.1 - 23.0 9,586 7,742 0 7 .7 23 July Night Day 5.1- 7.7 7.7 - 48.5 1,106 15,691 12.8 - 21.5 0 10 Aug. Night Day 7.7 - 12.8 5.1 - 26.6 2.936 8,934 5.1 0 478 6,617 164 2.5 0 7.7 1,091 0 5.1 - 12.8 2,332 120 414 0 2,153 2,027 0 211 3,082 0 342 0 RESULTS Ichthyoplankton samples were taken on 13 dates in 1978, 11 dates in 1979, and 13 dates in 1981 yielding 1,206 samples with 41,602 larvae representing 11 families (Tables 5 and 6). The majority of these samples as taken in 1978 and 1979 as ecological baseline data on spatial and temporal distribution of ichthyoplankton in Lake Michigan; they accounted for 93% of the larvae captured. Lake Michigan samples were dominated by larvae of alewife, rainbow smelt, yellow perch, and johnny darter. Ichthyoplankton fauna of Pere Marquette Marsh-Lake consisted primarily of common carp, centrarchids, percids, and clupeids. Three major groups of cohorts were recognized in Lake Michigan: spring, late-spring - early summer, and summer. Species most common in the spring were lake whitefish, burbot, and deepwater sculpin. These species are all recognized as cold water fish spawning in late fall and winter. Late spring-early summer cohorts were mainly yellow perch and rainbow smelt larvae. Alewife larvae were ubiquitous throughout summer and for short periods of time were associated with johnny darter, ninespine stickleback, 24 Table 5. Sampling dates, locations, and number of ichthyoplankton samples taken in Lake Michigan and the Pere Marquette Marsh-Lake 1978, 19 79, and 1981. ________ 19/S_________ 1979___________ 1981_____________ Pore Marquette Lake Michigan Lake Michigan Lake Michigan Marsh-Lake Date_________ Night________ Date________ Day____ Night_____ Date_______ Night______ Day_____ Night 17 April 22 April 6 29 9 April 28 April 6 i May 3 May 6 15 May 15 May 24 31 May 8 8 27 23 April 2 8 S 40 40 6 May 2 8 8 29 May 38 40 12 May 24 12 June 26 24 20 May 14 June 25 28 June 38 40 3 June 27 June 20 10 July 40 39 11 June 1 10 July 25 31 July 40 39 17 June 2 24 July 25 15 August 39 40 24 June 7 August 24 27 August 39 40 1 July 21 August 25 25 September 40 40 13 July 21 September 24 23 July O. c 18 October 22 10 August 1 8 2 6 9 10 10 10 9 9 8 2 9 9 2 8 9 2 8 Table 6. Total number of larval fish collected from 1978 to 1981 in Lake Michigan, Pere Marquette Marsh-Lake, and its connecting channel. Lake Michigan 1978 bake Michigan 1979 Niqht Day Pere Lake Marquette Marsh Michiqan 1981 1981 Outlet Channel 1981 Total Smallnrcnith bass 20,086 2,647 816 1,275 0 266 184 0 116 104 102 81 0 0 12 0 1 0 0 0 0 0 0 1,282 1,161 771 13 0 46 24 0 3 28 1 33 1 3 1 0 8 0 0 0 0 0 0 3,759 2,533 572 978 0 555 187 3 148 117 129 123 2 1 23 4 21 0 0 0 0 0 0 291 186 19 21 0 0 143 0 0 2 4 0 0 0 3 0 1 0 0 0 0 0 0 138 22 384 29 1,325 0 11 344 0 0 1 0 135 109 27 41 0 14 14 14 9 3 2 99 25 13 9 6 0 0 0 0 0 1 0 6 3 0 0 0 0 0 0 0 0 0 25,583 6,574 2,575 2,325 1,331 867 380 347 267 251 239 237 144 116 66 45 31 14 14 14 9 3 2 Total 25,690 3,375 9,155 598 2,622 162 41,602 256 340 398 18 138 56 1,206 Alewife Rainbow smelt Yellow perch Johnny darter Common Carp Ninespine stickleback Spottail shiner Lepomis spp. Cottus spp. Burbot Lake whitefish Deepwater sculpin Gizzard shad Pomoxis spp. Trout-Perch Cyprinidae Bloater Moxostoma spp. White sucker Rock bass Percidae Golden shiner Number of samples yellow perch, and Cottus spp. larvae. Most other species occurred infrequently from June through September. Alewife The spawning migrations of adult alewives into shallow water during early summer is well documented in recent power plant studies (Brazo and Liston 1979; Jude et al. 1980) and in older literature (Norden 1967; Wells 1968). Spawning is generally believed to occur over much of the summer. Eggs are dispersed randomly into the water and hatching generally occurs within 1 week at 15 C (Edsall 1970). Lake Michigan Temporal distribution. Alewife larvae totally dominated Lake Michigan ichthyoplankton in all years, comprising two-thirds of larvae captured in 1978 - 1981. However, seasonal appearance and densities varied markedly among years. Alewife larvae were first collected on 14 June and 28 June in 1978 and 1979, respectively, and were present on all collection dates for the remainder of both years. Initial range of densities was low in both years (1 - 36/1000 m 3), but peak densities varied in magnitude and timing between years. In 1978, alewife larvae were extremely abundant from mid-July to mid-August, reaching peak densities in early August of 9316 larvae/1000 m 3' and 28 well over 1000 larvae/1000 m 3 in many samples on all dates during this period in 1978 (Table 7). After this peak, alewife abundance declined rapidly to averages of 0 - 24 larvae/1000 m 3 in late September and October. In contrast, 1979 densities of alewife larvae remained generally low (<200/1000 m 3) until late summer when peak densities (up to 1298 larvae/1000 m 3) were observed at all transects and contours on 27 August, when water temperatures were 16 - 18 C. Overall, densities of alewife larvae in 1978 were significantly greater (0.02 < p < 0,05) than in 1979. However, substantial numbers of alewife larvae were present on the last sampling date in 1979, 25 September, when the range of average densities was 25 - 227/1000 m 3. The 2 years were vastly different in terms of water temperatures, current, and prevailing winds. Indeed, 1978 could be termed a "warm" year on Lake Michigan. Water temperatures had risen to about 15 C by late May and were above 17 C from mid-August to September reaching peaks of 21 - 22 C in late August (Figure 3 and Appendix Table 1A). By comparison, water temperatures in 1979 remained considerably colder. Bottom temperatures did not exceed 13 C on most sample dates during the summer, but did warm to 17 C by late August. Examination of surface wind data provides some explanation for the annual differences in water temperatures. During 1978, from June to the end of August, winds were from the southerly and westerly quadrants on _ Table 7. Average nighttime density and range (in parentheses) of alewife larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data._______________ __________________________ Depth contour (m)_________________________________________ Date 1 3 6 9 12 1978 14 June 3 (0 - 90) 0 27 June 71 (91 - 186) 586 (10 - 1228) 31 10 July 558 (66 - 930) 1765 (69 - 2910) 701 24 July 1781 (437 - 3760) 742 (429 - 1030) 827 7 Aug. 2.638 (525 - 9316) 2502 (915 - 4403) 1494 (247 - 2741) 2002 (1230 - 2651) 21 Aug. 21 Sept. 0 1 (0 - 4) - 12 (0 - 40) - (287 - 1464) 328 (174 - 614) - (458 - 4073) 794 (348 - 1278) - 2586 (1481 - 4073) 1866 (707 - 2458) - 1004 (0 - 93) (779 - 2930) 808) - 1 (0 - 6) - 8 (4 - 18) - 1 (0 - 3) 0 24 (3 - 73) 21 (3 - 37) 18 28 June 2 (0 - S) 36 (0 - 159) 0 2 (0 - 11) "1 lu July —» (0 - 10) 1 (0 - &) 0 1 (0 - 6) 0 18 Oct. 0 588 (455 - (5 - 50) -979 (0 - 9) 31 July 110 (0 - 200) 162 (34 - 269) 102 (14 - 202) 57 (0 - 144) 24 (0 - 83) 15 Aug. 18 (0 - 38) 37 (0 - 99) 13 (0 - 40) 11 (0 - 30) 14 (0 - 31) 27 Aug. 292 (87 - 702) 623 (322 - 1298) 370 (19S - 575) 381 (194 - 519) 25 Sept. 227 (150 - 725) 163 74 (34 - 148) (69 - 375) 61 (23 - 136) 322 (150 - 682) 25 (4 - 48) 30 Figure 3. Mean(-) and range (shaded area) of bottom temperatures on sample nights at 3- to 12-m depth contours at all transects in Lake Michigan, 1978-1979. Water T em perature (°C) c* o c* — r~ ro o □ a> od r-— i 0 % H SJ CO IE 65.5 days; however, during this same time period in 1979 winds from the same quadrants occurred on only 59 days (Table 8). The remaining time, winds prevailed from the easterly and northerly quadrants. Strong winds from the east and north can induce upwelling of cold water along the eastern shore of Lake Michigan and at the same time drive warm surface waters offshore. This phenomenon was apparently responsible for the depressed and delayed peak abundance of alewife until late August in 1979 compared to larger and earlier peaks in 1978 when water temperatures warmed sooner and rose to markedly higher levels than in 1979. Mean lengths of alewife larvae in August tended to be greater in 1979 than 1978 (range of means 5.2 - 17.5 mm in 1979 versus 7.0 - 10.3 mm in 1978, Table 9). This suggests poor survival of larval alewife hatched later in the summer during 1979 (prolonged upwellings). Though some newly hatched larvae (4.3 - 5.2 mm) were taken in August in both years, the number of this size larvae in 1978 was far greater than in 1979. Spatial distribution. Over both years alewife densities tended to be greater at transects 2 and 3 (those nearest the power plant). This was especially true at the shallower contours (1 and 3 m ) . One explanation is the breakwall and jetties protecting the plant or currents emanating from the plant may offer some habitat the alewife key on for spawning. 33 Table 8. Number of days per month wind direction originated from given compass points near Ludington, Michigan. Data from National Oceanic and Atmospheric Administration (NOAA) Marine Coastal Weather Log - Ludington Station. Month South and West Quadrants North and East Quadrants 1978 May 10.75 6.25 June 23.25 6.75 July 19.25 11.75 August 23.0 8.0 September 15.0 15.0 1979 May 6.25 9.75 June 19.0 11.0 July 21.5 9.5 August 18.5 12.5 September 25.75 4.25 Table 9. Mean total length (inm) of alewife larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. Contour Date 3 1 (rn) 9 6 1978 14 June 4.5 (2) 0.0 27 June 4.7 (13) 4.6 (26) 5.3 10 July 4.9 (52) 5.8 (54) 24 July 6.5 (60) 7.4 7 Aug. 7.9 (50) 21 Aug. 8.6 21 Sept. 18 Oct 0.0 12 4.4 (1) _ (12) 4.8 (8) - 5.9 (60) 5.4 (66) - (60) 6.8 (61) 6,5 (60) - 7.8 (60) 7.7 (60) 7.0 (70) - (57) 8.1 (60) 8.5 (59) 10.3 (67) - 23.0 (1) 0.0 25.2 (1) - 20.6 (21) 20.3 (17) 20.7 23.2 (8) - 1979 28 June 4.1 (3) 4.4 (21) 0.0 4.5 (2) 4.7 10 July 4.5 (3) 5.0 (1) 0.0 4.8 (1) 0.0 31 July 6.3 (54) 5.5 (66) 4.6 (41) 5.2 (34) 5.0 (21) 15 Aug. 11.5 (10) 9.9 (30) 6.2 (11) 5.2 (8) 5.8 (14) 27 Aug 17.5 (58) 14.4 (61) 13.9 (79) 13.8 (76) 14.2 (78) 25 Sept. 19.8 (53) 19.0 (63) 19.1 (56) 17.4 (50) 18.2 (38) 0.0 (14) (2) 35 Horizontal depth, distribution was similar among years. Densities tended to be markedly higher at the 1- and 3~m contours than at deeper contours (Table 7). As the season progressed, density of alewife larvae at deeper contours increased, possibly the result of passive dispersion of larvae by wind and current, or active movement of larger larvae. Examination of length data provide contradictory support for the latter hypothesis. In 1978, data lend some support to the possibility of larger larvae moving to deeper water (Table 7), but in 1979 the exact opposite trend was noticed. Larger larvae tended to be concentrated at shallower contours. These data may be confounded by the effect of the relatively cold year in 1979. Greater densities of alewife larvae at shallower contours have been previously reported for this area (Liston et al. 1980), and for other areas of Lake Michigan (Jude et al. 197 9b). 1979a, Warmer water temperatures at the shallower contours were believed partially responsible for the observed distribution. Water temperatures at the 1-m contour ranged from 11.8 - 20 C on dates of greatest alewife abundance, but were as low as 11.1 C at deeper contours on these dates. Diel distribution. Examination of diel distribution was limited to the 19 79 field season. Differences in larval alewife distribution were apparent. Night tows produced more larvae than day tows (0.10 < p < 0.15), though this trend was not always consistent (Table 10). Table 10. Comparison of average day - night densities of alewife larvae (number/1000 nr3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. Depth contours3 (m) 1 Date 28 June 10 July 6 9 12 D N D N D N D N D N 4 (8) 2 (3) 6 (8) 36 64 2 (5) 0 - 0 - 2 (4) 0 0 2 (3) 2 -1 0 0 (2) 2 (5) 0 - 1 (2) 0 - 0 - (5) (4) - 30 (25) no (70) 155 (102) 162 (95) 42 (61) 102 (69) 34 (38) 57 (50) 102 (47) 18 (10) 29 (32) 37 (35) 8 (11) 13 (12) 7 (5) 11 (11) o 14 (6) (ID 27 Aug. 128 (134) 292 (189) 231 (123) 623 (350) 141 (187) 370 (15) 79 (100) 381 (110) 48 (56) 322 (205) 25 Sept. 58 (111) 227 (99) 12 (10) 163 (90) 8 (16) 74 (41) 6 (8) 61 (37) 7 (10) 25 (15) 31 July 15 Aug. 9 (14) 24 (28) 37 Daytime densities of larvae at shallow contours were occasionally as great or greater than night densities; at deeper contours (9 and 12 m) night tows generally contained many more larvae than day tows. These data suggest that alewife larvae at deeper contours may not be as vulnerable to daytime capture as larvae at shallower contours. This phenomenon may be partially a function of water clarity and distribution of larvae in the water column. Later in the year the number of alewife larvae captured at night was significantly greater than daytime catches (p < 0 .0 1 ). This phenomenon is most likely the result of increased avoidance by larger larvae. Total abundance. In 1978, alewife larvae were extremely abundant in the Lake Michigan reference wedge and ranged from nearly 47 thousand larvae in late September to 160 million larvae in early August (Table 11). From late July to late August point estimates ranged from 40 to 160 million larvae on sample days. Total alewife abundance in 1979 at similar 1978 contours ranged from an estimated 52 thousand larvae on 10 July to only 31 million larvae on 27 August in the reference wedge (Table 11). Densities of larvae at deeper contours were generally less than those at shallow contours, but the greater volume of water contained within the deeper contours tended to equalize the number of alewife larvae among all contours. Early in the alewife hatching season the greatest abundance of larvae was at the shallowest two contours. The great disparity in abundance Table 11. 1978 Date Estimates based on night densities and 90% confidence interval (+ ) of alewife abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. (N = 190) Number 19791 1979“ Date Number Date = (N = 238) Number 14 June 77 (+153) 27 June 4984 (+6020) 28 June 367 (+449) 28 June 553 (+682) 10 July 40,461 (+19,088) 10 July 51 (±95) 10 July 51 (±95) 24 July 62,670 (+29,186) 31 July 5814 (+3057) 31 July 9058 (±5114) 7 Aug. 160,209 (+81,138) 15 Aug. 1073 (+727) 15 Aug. 2650 (+1558) 21 Aug. 61,709 (+21,731) 27 Aug. 30,966 (+7362) 27 Aug. 21 Sept. 7 (±94) 25 Sept. 5 ,723 (+2223) 25 Sept. 18 Oct. 865 (+492) “At 1978 contours only. toAt all contours sampled in 1979. 66,529 (±22,627) 8462 (±3328) 39 of alewife larvae between years was the result of greater densities in 1978 than 1979 due to cliraatological conditions previously discussed. Pere Marquette Marsh-Lake General distribution. In 1981 alewife larvae were first collected in Pere Marquette Marsh-Lake on 3 June. Those first larvae were present only in the upper marsh (station 1) and densities were low (24 - 31 larvae/1000 m 3 ), but this record represents the presence of larvae from 10 to 24 days earlier than sampled in Lake Michigan during 1978 and 1979, and additional samples taken at the 1-m contour in Lake Michigan in 1981 did not contain any alewife until 17 June (Appendix Table IB). Early appearance of larvae in the marsh is not surprising as typical alewife spawning behavior (e.g., splashing at the surface and circular swimming) had been observed in mid-May and peaked in June. Water temperatures at this time were considerably higher in the marsh (14.0 - 20 C, see Table 4) than in Lake Michigan (9.0 - 15 C). Alewife densities remained low (generally <100/1000 m 3) in the marsh, and larvae were present at all stations by mid-June, but they remained for only a short period of time until early July, after which few larvae were collected. In Pere Marquette Marsh-Lake, peak densities of larvae (up to 164/1000 m 3) were in the upper marsh (stations 1 and 2) and the size of larvae 2.9 - 4.3 mm 40 would indicate they were recently hatched, but higher densities were observed most consistently at station 3. Further, overall greatest density occurred in the outlet where up to 308 larvae/1000 m 3 were recorded. Greater densities of alewife larvae in the outlet than in the marsh are somewhat confusing and may be explained several ways. Inflow of larvae from Lake Michigan could have added alewife larvae to the marsh. Indeed, on certain dates there was considerable reverse flow from Lake Michigan into Pere Marquette Marsh-Lake and alewife densities were as high as 308 larvae/1000 m 3, greater than densities measured in Lake Michigan at the same time (Appendix Table IB). Alewife densities at the 1-m contour in Lake Michigan did not peak until mid-July through midAugust in 1981 when few alewife larvae were found in the marsh and nearly all flow was from the marsh to Lake Michigan. A second hypothesis is that the harbor outlet and breakwall area provided alewife spawning sites and many larvae could have originated there. Corroborating evidence to support this premise is provided by the observation of large numbers of spawning alewife in this area. Length data indicate most larvae produced there were newly hatched (e.g., 3.5 - 4.5 mm) though some larger larvae were also present. (6 - 8 mm) A final plausible explanation lies in the dichotomy of observing extensive alewife spawning in the marsh and the presence of many eggs (thought to be alewife based on diameter and seasonality) in marsh 41 samples, but few larvae were actually collected in the marsh. Possibly these eggs drifted towards the harbor outlet and hatched in this area. All the above circumstances may contribute to the high densities of alewife larvae at the harbor outlet. Inputs to Lake Michigan. Alewife larvae were present in the harbor outlet in weekly samples taken from 11 June to 13 July 1981 and once in early August. On those sample days during this period when some reverse flow from Lake Michigan was occurring, alewife larvae were present in both the outflow (to Lake Michigan) and inflow (from Lake Michigan). Densities of larval alewife in water leaving Pere Marquette Marsh-Lake ranged from 50 to 308/1000 m 3 at night and 25 to 84/1000 m 3 during the day, while densities of fish larvae in water flowing in from Lake Michigan were generally lower at 15 to 131/1000 m3 , except in one sample in mid-June when densities were 308/1000 m 3. This sample may have introduced error since it was not collected entirely in that cell of water flowing in from Lake Michigan; however, point estimates on eight sample days when alewife were collected indicate nearly 4 million alewife larvae may have been carried out to Lake Michigan, while only 436,000 larvae were transported into Pere Marquette Marsh-Lake during the sample season. This total compares with 160 million and 66.5 million alewife larvae estimated on a single day during peak abundance in 1978 and 42 1979, respectively, in a reference wedge in nearby Lake Michigan. It must be deduced that some alewife larvae are carried into Pere Marquette Marsh-Lake from Lake Michigan and subsequently returned, thus the impact of these estimates may be further reduced. Because of the ubiquitous distribution of alewife, the true importance of input from Pere Marquette Marsh-Lake to alewife populations in Lake Michigan is probably small and difficult to ascertain. Rainbow Smelt Rainbow smelt are probably the second-most important forage fish in Lake Michigan and tend to fluctuate inversely with alewife populations. Rainbow smelt are generally considered to be anadromous spawners, returning to streams in April and May at water temperatures of about 9 C (Scott and Crossman 1973). However, Rupp (1965) has shown shoreline spawning can be significant, and McCallum and Reiger (1970) reported rainbow smelt spawned in Lake Erie at 5 - 22 m. Lake Michigan Temporal distribution. Rainbow smelt larvae were the second-most abundant larvae collected in both 1978 and 1979 comprising 10.3 and 29.4%, respectively, of the total number of larvae captured annually. Several striking 43 similarities occurred between years. They were first captured at low densities (2 - 48/1000 in3) in mid-May during both years at water temperatures of 7 - 9 C. Larvae at this time ranged from 5.8 to 6.4 mm indicating they were newly hatched (Dorr et al. 197 6 ). Further, rainbow smelt densities increased dramatically to peaks of 697 - 2302 larvae/1000 m 3 by the end of May in each year (Table 12). This peak was followed by a precipitous decline in both years which was most likely due to distribution rather than avoidance because large numbers (densities as high as 980/1000 m 3) of rainbow smelt young-of-the-year (YOY) (>25.4 mm) were captured in late August. YOY rainbow smelt were abundant in collections in mid to late August when water temperatures were from 15 to 2.1 C. Temporal patterns of rainbow smelt distribution were similar among years but there was a noticeable difference in magnitude of densities. Peak densities of larvae were significantly higher (p = 0.05) in 1978 than in 1979, ranging up to over 4000 larvae/1000 m 3 compared with slightly less than 800 individuals/1000 m 3 in 1979. Again these differences may be attributed to climatological and physical conditions on Lake Michigan. During May, 1979 easterly winds dominated (61% of the time) weather patterns and upwelled water kept inshore waters cool (8 C ). Larvae could have been transported offshore to less desirable habitat. By comparison, easterly winds were only present Table 12. Average nighttime density and range (in parentheses) of rainbow smelt larvae (number/1000 nr*) __________ at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data._____ Depth contour (m) Date 1 3 6 12 9 1978 15 May 31 May 29 90) 17 1123 (128 - 4666) 797 (0 (0 64) (73 - 1951) 4 (0 - 9) 254 (214 - 506) (0 - 7) - 300 (97 - 515) - 2 14 June 18 (0 46) 32 (16 60) 25 (0 - 57) 26 (7 - 63) - 27 June 11 (0 23) 10 (0 23) 6 (0 - 13) 22 (0 - 44) - 10 July 0 2 (0 9) 2 (0 - 15) 0 24 July 10 6) 3 (0 - 9) 1 (0 28) 1 (0 (0 - - 6) 1979 15 May 2 (0 - 15) 2. (0 - 10) 0 29 May 211 (0 - 797) 322 (95 - 626) 79 (21 - 193) 70 (0 - 188) (0 50) 39 (11 - 75) 39 (10 - 59) 30 (0 - (0 38) 34 (0 - 127) 25 (0 - 14) 4 (0 - 13) 11 (0 - 14) 0 0 0 0 19 (0 - 45) 47) 45 (12 - 76) (0 - 70) 14 (0 - 38) 4 12 June 16 28 June 4 (0 " 15) 16 10 July 3 (0 - 16) 4 31 July 2 (0 - 37) 0 2 15 Aug. 2 (0 - 12) 0 0 0 0 25 Sept. 0 0 0 1 1 (0 8) (0 -- 30) (0 - 17) (0 - 11) 45 37% of the time in May, 1978 and water temperatures rose to 13 C just prior to peak larval fish abundance. Spatial distribution. Rainbow smelt larvae were rather evenly distributed in both years at all transects along the 10 km of coastline studied. uniform at all transects. Size of fish was These data suggest that shoreline spawning may be important in the study area. Indeed, large numbers of spawning adults were taken annually in the spring in gill nets and seines throughout the study area. Significant differences (p = 0.1, when tested by sample date) in horizontal depth distribution of rainbow smelt larvae existed and were similar in 1978 and 1979. During early hatching and up through peak densities, rainbow smelt larvae exhibited a decided preference for the shallowest contours; densities were generally greater than at deep water contours (Table 12). Mean lengths of rainbow smelt larvae during the first month ranged from 5.4 to mm indicating they were newly hatched. 6 .0 After peak densities, larvae were more uniformly distributed among contours (p > 0.5). Length data at deeper contours showed rainbow smelt larvae averaged 5.7 - 7 . 1 mm total length (Table 13). offshore. This suggests that as larvae grow they move Jude et al. (1979a) observed similar distribution patterns of larval rainbow smelt in southern Lake Michigan and hypothesized that newly hatched rainbow smelt inhabit the beach zone and move offshore as they Table 13. Mean total length (ram) of rainbow smelt larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. Date 1978 15 May 5.8 (15) 6.0 31 May 5.7 (57) 6 14 June 5 .9 (1 0 ) 27 June 20.7 (2 ) 10 July Contour n O 1 0.0 (m) 9 6 (3) — 6.8 (44) - 7.1 (2 1 ) - (17) - (18) 5.7 (58) 6.8 (60) 6.6 (31) 7.1 (2 1 ) 11.1 (5) 6.8 (3) 19.0 (1 ) 13.9 (1 ) 0.0 21.4 1 io ->)\ 22.0 .4 (5) 12 6.3 11.0 ~ - 22.7 (6 ) 22.5 (1 ) 15 May 5.4 (1 ) S.2 (1 ) 29 May 5.7 (57 ) 5.9 (80) 6.2 (59) 6.4 (45) 6.4 (28) 12 June 6.0 (5) 6.1 (15) 6.4 (13) 6.5 (1 2 ) 6.4 (16) 28 June 9.1 (5) 11.0 (15) 9.8 (30) 7.4 (24) 7.5 (14) 10 July 6.5 (3) 31 July 0.0 24 July (1 ) 1979 15 Aug. 25 Sept. 25.2 0.0 9.2 (5) 8.6 16.7 0.0 0.0 25.4 0.0 (1 ) 0.0 0.0 (5) (1 ) 10.1 (1 0 ) 12.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 14.5 (6 ) (1) 47 grow. Later in the year, large larvae and YOY preferred shallow areas. smelt per contours. 1000 Densities of several hundred YOY rainbow m 3 of water were observed at the 1 - and 3-m Scott and Crossman (1973) stated that the shallow water distribution of YOY rainbow smelt is common over most of its distribution. Diel distribution. Differences in day-night distribution in 1979 were generally not detected for rainbow smelt larvae less than 6 .5 mm in length, but highly significant differences (p<0„0039) in diel distribution were apparent for larger larvae (Table 14). During August, densities of YOY fish were substantially greater in night tows which is probably the result of gear avoidance by the larger fish. One exception was at the 1-m contour on 15 August, when densities of rainbow smelt YOY ranged up to 984/1000 m 3 in day tows compared with only 140 larvae/1000 m 3 in night tows. Physical conditions may have affected distribution because day tows were made under high wind conditions and the water was very turbid at shallow contours which most likely reduced gear avoidance. Total abundance. Abundance of larval rainbow smelt in the reference wedge in Lake Michigan was dissimilar between 1978 and 1979. In both years rainbow smelt abundance peaked abruptly in late May at 2.6.5 million and 7.3 million larvae, respectively, and declined gradually (Table 15). Densities of fish were generally greater at shallow contours, but total numbers of fish were greatest at deeper Table 14. Comparison of average day - night densities of rainbow smelt larvae (number/1000 irr1) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 1 Date 15 May D 10 (15) 3 D N 2 (4) 2 (4) 1 (3) jL* (25.4 mm) in lower marsh macrophyte beds in August. Samples taken concurrently in Lake Michigan indicated YOY were at high densities (256/1000 m3) at the 1 -m contour at this time, suggesting these fish may have actively moved into the marsh. Inputs to Lake Michigan. As stated earlier, it is believed that this submerged river mouth and the Pere Marquette River and its tributaries do not support a viable spawning run of rainbow smelt, and thus have virtually no direct contribution to rainbow smelt populations in Lake Michigan. Presence of rainbow smelt in Pere Marquette Marsh-Lake is thought to result from reverse flow of Lake Michigan water carrying larvae into the system. Indeed, on sample days when reverse flow occurred, 267,000 larvae were 52 estimated to be carried into the marsh. Similarly, some 541,000 larvae were estimated transported out the harbor channel. The disparity between inflow and outflow may be explained in two ways since reproduction in the marsh was found to be negligible. Rainbow smelt are known to spawn heavily on the breakwalls and jetties that form the Ludington harbor (personal observation) and larvae may drift with the complex currents in the harbor outlet. Possibly large numbers of rainbow smelt larvae were swept into the marsh during non-sampling periods and were subsequently sampled as they were transported back to Lake Michigan. Regardless of the disparity, numbers of larvae transported out of the marsh only represented 3% of peak abundance of rainbow smelt larvae present in the reference wedge in Lake Michigan. Yellow Perch Yellow perch in Lake Michigan near Ludington spawn from mid-May to mid-June depending on climatic conditions (Brazo et al. 1975). Eggs are deposited in a long gelatinous skein that is usually hung over submerged stumps, limbs, rocks, or other types of underwater objects and hatching usually occurs in 8 to 10 three times as long at temperatures of Crossman 1973). days but may require 8 C (Scott and Dorr (1982) provides an extensive 53 summarization of spawning requirements of yellow perch. Lake Michigan Temporal distribution. Yellow perch larvae were the third-most abundant larvae collected in the present study comprising 6.2% of the catch (Table combined. 6 ) with all years Seasonal distribution of yellow perch larvae in Lake Michigan was very similar for 1978 and 1979. Yellow perch larvae were first collected on 15 May in both years at water temperatures of 7 to 9.8 C. Peak nighttime densities as high as 2120 and 288/1000 m 3 in 1978 and 1979, respectively, were recorded on these dates. Some yellow perch larvae were present in all subsequent samples through late July, after which none were collected. Densities decreased rapidly after May peaks, but a second peak (though slightly smaller) occurred in late June or early July when larval fish densities reached 556 larvae and 194 larvae/1000 m 3 in 1978 and 1979, respectively. Yellow perch spawning seemed to be bimodal but significant differences in spatial distribution of larvae existed. Spatial distribution. Yellow perch larvae collected in May and early June were concentrated at the shallowest contour (1 m) where densities averaged from 4 to 749 larvae/1000 m 3 in night tows and 17 to 409 larvae/1000 m 3 in day tows (Tables 16 and 17). Subsequently, a decline in yellow perch larvae was observed until late June when Table 16, Average nighttime density and range (in parentheses) of yellow perch larvae (number/1000 rrr^) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. ________________________________ Depth contour (m)__________________________________ Date_____________ 1____________________ 3__________________ 6_________________ 9________________ 12 1978 (0 - 2120) 104 (0 - 241) 24 (4 ~ 44) 5 (0 - 12) (0 - 16) 5 (0 - 12) 4 (0 - 19) 3 (0 - 6) - (0 - 0) 17 (0 - 26) 28 (0 - 57) 19 (0 - 56) - 0 15 (0 - 59) 34 (0 - 152) 131 (0 - 556) - 10 July 0 3 (0 - 10) 0 3 (0 - 17) - 24 July 2 0 1 (0 - S) - 15 May 749 31 May 5 14 Jane 4 27 June (0 - 9) 0 - 1979 15 May 143 (15 - 288) 6 (0 - 24) 5 (0 - 18) 1 (0 - 6) 3 (0 - 9) 29 May /b (6 - 168) 12 (0 - 21) 4 (0 - 12) 3 (0 - 12) 2 (0 - 6) 12 June 17 (0 53) 5 (0 - 19) 3 (0 - 11) 3 (0 - 13) 5 (0 - 12) 28 June 15 (0 30) 14 (0 - 38) 27 (0 - 107) 50 (0 - 107) 69 (0 - 190) 10 July 96 (0 - 462) 29 (0 - 136) 41 (7 - 89) 20 (0 - 48) 28 (0 - 56) 31 July 0 2 (0 - 13) 0 0 0 Table 17. Comparison of average day - night densities of yellow perch larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, parentheses. 3 1 )ate 15 May D 1979. One standard deviation in D = day, N = night. N D 409 143 (333) (101) 42 (26) Depth contours (m) 6 N D N 6 (10) 9 12 D N D N 14 (11) 5 (7) 6 (8) 1 (2) 2 (4) 3 (4) 29 May 90 (91) 76 (60) 44 (43) 12 (9) 6 (5) 4 (5) 112 (14) 3 (5) 3 (3) 2 (3) 12 June 17 (25) 17 (19) 8 (3) 5 (9) 10 (12) 3 (6) 2 (3) 3 (6) 0 - 5 (6) 28 June 4 (8) 15 (24) 22 (40) 14 (13) 87 (143) 27 (35) 69 (46) 50 (62) 79 (77) 69 (63) 10 July 14 (16) 96 (178) 14 (19) 28 (49) 38 (32) 41 (33) 17 (27) 20 (22) 26 (21) 28 (21) 31 July 0 - 0 (6) 0 - 0 - 0 - 0 - 0 - 0 - 0 - 2 (5) 56 substantial numbers (up to 556 larvae/1000 m 3 ) were again observed but at deeper contours (>6 m) at most transects. Examination of length data indicates additional disparities in yellow perch larvae. Mean lengths of yellow perch larvae present early in the field season ranged from 6.2 to 9.3 mm (Table 18) suggesting many were not newly hatched. However, when peak densities occurred at deeper contours later in the season, lengths of larvae ranged from 5 . 6 to 6 .3 mm indicating these fish were newly hatched. Evidence from Lake Michigan seems to support the suggested hypothesis of yellow perch cohorts arising from two different environs. Diel distribution. Yellow perch larvae were collected in nearly equal proportions if not slightly higher in day than night tows. Average densities in day tows were as high as 409 larvae/1000 ra3 compared with only 143 larvae/1000 m 3 in peak night tows in 1979. No significant (p > 0.5) diel differences were observed at contours. Total abundance. Abundance of yellow perch in the Lake Michigan reference wedge was slightly lower in 1979 than 1978. Peak densities occurred in May but numbers of larvae were estimated at only 1.9 and 0.28 million during this month in 1978 and 1979, respectively (Table 19). Peak abundance was observed in late June in both years at 3.3 7.9 million fish larvae. The disparity between peak abundance and peak densities is related to yellow perch distribution. During peak densities, larvae were Table 18. Mean total length (mm) of yellow perch larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. Contour Date 3 1 (m) 9 6 12 1978 15 May 6.4 6.4 (42) 6.5 (31) 31 May 7.2 (2 ) 8.3 (5) 9.2 (5) 14 June 5.7 (3) 6.0 (16) 5.7 (17) 27 June 0.0 5.7 (1 0 ) 5.9 (1 2 ) 10 July 0.0 5.6 (2 ) 24 July 7.1 (1 ) 15 May 6.5 (43) 6.2 29 May 6.8 (48) 6.7 12 June 7.9 (8 ) 7.8 28 June 6.0 (2 0 ) 5.8 (13) 10 July 6.-3 (2 2 ) 6 .2 (24) 31 July 0.0 (40) 0.0 0.0 6.2 (9) - 8.4 (4) - 6.0 (15) - 5.6 (29) - 5.5 (2 ) - 5.9 (1 ) 0.0 6.2 (5) 6.7 (1 ) 6.3 (3) 6.9 (5) 6.2 (4) 6.9 (3) 5.6 (1 ) 9.6 (1 ) 9.1 (2 ) 6.0 (2 1 ) 5.8 (33) 5.8 (47) 6.0 (45) 6.4 (2 1 ) 6.1 (37) 1979 0.0 (3) (15) (2 ) 0.0 0.0 5.2 (1 ) Table 19. Estimates based on night densities and 90% confidence interval (+ ) of yellow perch abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. 1978 Date (N = 143) Number N = number of samples. 197 9a Date Number 197913 (N = 2 2 2 ) Date Number 15 May 1924 (+1479) 15 May 282 (+271) 15 May 646 (±617) 31 May 271 (+245) 29 May 400 (+303) 29 May 636 (+523) 14 June 1565 (+1314) 12 June 267 (+604) 12 June 832 (+1388) 27 June 7891 (+10,971) 28 June 3310 (+2723) 28 June 11,000 2015 (+1458) 10 July 5126 (+2992) 31 July 185 (+351) 10 July 24 July 177 (+342) 21 10 July 31 July (+4) aAt 1978 contours only. toAt all contours sampled in 1979. 0 - (+7428) 59 concentrated at shallow contours where there is considerably less water volume. Again abundance of larvae was slightly greater in 1978 than 1979 which may be attributed to the quicker and extended warming of Lake Michigan in 1978. However, high abundance of yellow perch was extended somewhat later in 1979. Pere Marquette Marsh-Lake General distribution. Yellow perch were the secondmost abundant larvae captured in the Pere Marquette MarshLake system (Table 6), but were only present for a short period of time. Larvae were first collected in late April at low densities (4 - 5/1000 m 3 ) at station 3 (Appendix Table 3B). Yellow perch larvae rapidly reached peak densities in early May at levels up to 1188/1000 m 3. At this time, yellow perch larvae were present at all stations, including the harbor outlet., but peak densities were in submersed macrophyte beds in the upper marsh (station 2) while densities were only 8 - 6 6 in outlet samples. larvae/1000 in3 By the end of May no yellow perch larvae were found in the marsh except for a single specimen on 3 June. The time of peak abundance in Pere Marquette Marsh-Lake coincided with low densities of larger (>6.6 mm) yellow perch larvae at the 1-m contour in Lake Michigan. In mid June 1981, another peak of yellow perch larvae occurred at the 1-m contour in Lake Michigan but none had been collected in the Pere Marquette Marsh-Lake for 3 60 weeks. Larvae at this time were 5.2 - 5.9 mm indicating they were newly hatched. Inputs to Lake Michigan. Yellow perch larvae were only present in outlet samples on two dates (May 6 and 12), and then only in water moving from Pere Marquette Lake to Lake Michigan. Densities ranged from 8 to 66 larvae/1000 m 3 and 1244 - 86,259 larvae were estimated to be transported daily from the marsh to Lake Michigan. These figures are not large when compared with estimates of 271 thousand to 1.9 million larvae present in a reference wedge in Lake Michigan during the period of gear vulnerability. These data do suggest nearly one-third of the larvae in an area comparable to the reference wedge in Lake Michigan may have been transported from Pere Marquette Marsh-Lake. Over a 2-week period of gear vulnerability an estimated 0.75 million yellow perch larvae may have been transported to Lake Michigan from the marsh near Ludington in 1981. Johnny Darter Adult johnny darters were the only darters collected in the study area. Though little published material is available on this species in Lake Michigan, some data on spawning times and behavior are available from power plant surveys (Jude et al. 1979a; Brazo and Liston 1979). Johnny darters spawn on the underside of rocks and the male guards the nest (Scott and Crossman 1973). Though these same 61 authors list spawning time as May in Canada, Winn (1958) observed johnny darters to spawn in April inland in southern Michigan. Johnny darters in Lake Michigan near Ludington spawn much later, normally from mid-June through July (Brazo and Liston 1979). Differences in spawning times were undoubtedly related to water temperature. Eggs hatch in 5 - 8 days at 22 - 24 C (Scott and Crossman 1973) and are presumed to take slightly longer in this area of Lake Michigan where water temperatures rarely rise above 20 C and then only for short periods of time. Lake Michigan Temporal distribution. Johnny darters were the fourth-most abundant larvae collected and 22.66 larvae comprised 5.9% of the Lake Michigan catch during 1978 1979. During both years, johnny darter larvae first appeared in late June at the shallow contours of transects 2 and 3 where water temperatures were 12 - 14 C. However, seasonal distribution thereafter was dissimilar for the 2 years. In 1978, peak densities of johnny darter larvae were recorded the first time these larvae were collected (densities as high as 1741/1000 m 3 ) (Table 20). Densities remained high through mid-July (up to 1133 larvae/1000 m 3) after which only a few larvae were captured in August. In 1979, johnny darter larvae were present in low densities (1 - 4 larvae/1000 m3 ) at first occurrence, but increased Table 20. Average nighttime density and range (in parentheses) of johnny darter larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. Date 1 Depth contour (m) 5 3 9 12 1978 27 June 152 (0 - 352) 10 July 36 (0 - 91) 218 (39 - 516) 410 24 July 21 (0 - 43) 106 (32 - 240) 245 (170 - 7 Aug. 6 (0 - 23) 46 (0 - 99) 37 •"7 / (0 - 35) 2 (0 - 8) 28 June 4 (0 - 14) 1 (0 - 10 July ?8 (0 - 157) 57 31 July 22 (0 - 74) 15 Aug. 1 (0 - 9) 27 Aug. 26 (6 - 88) 74 25 Sept. 1 (0 7) 2 21 Aug. 669 (0 - 1741) 50 (0 - (0 - 10) - 253 (0 - 761) - 778) 48 (0 - 82) - 99) 73 (0 - 138) - 0 2 (0 - 6) - 8) 0 0 0 (0 - 246) 4 0 156 (14 - 312) (0 - 13) (8 - 232) (0 - 1133) (0 - ') 1979 - A (0 - (0 - 14) 0 295 (121 - 950) 247 (65 - 619) 99 4 (0 - 20) 24 (0 - 189) 7 155) 57 (0 - 130) 47 (0 - 82) 45 15) 1 6) 0 (0 - 1 (46 - 216) (0 - 48) (0 - 129) (0 - 6) 63 rapidly to 246 larvae/1000 m 3 in early July. Peak larval fish abundance occurred in late July, 1979, when johnny darter larvae were collected at all transects and contours at densities of 14 - 950/1000 m 3. This peak was quickly followed by a precipitous drop and a secondary peak in late August. A few larvae were captured in late September. The bimodal peak in johnny darter distribution in 1979 was apparently the result of a strong upwelling in early August that uniformly reduced water temperatures to 5 - 10 C , after which a second spawning-hatching period occurred. The temporal distribution of johnny darter larvae in 1979 was different from that observed in 1978 when a single peak in abundance was observed from late June through mid-July. Spatial distribution. Spatial distribution of johnny darter was not xiniform and appeared to be related to substrate types. Larvae first appeared near transects 2 and 3 and greatest densities of these larvae were continually recorded at these areas in 1978 and 1979. Examination of length data shows these fish were recently hatched (5.7 - 6.4 mm, Table 21). Bottom substrates at transects 2. and 3 are composed of a much larger percentage of gravel, rocks, clay outcroppings, and other irregularities that provide better spawning sites than large expanses of sand at transects 1 and 4. In fact, densities of larval johnny darters only exceeded 100/1000 m 3 in three samples over 1978 - 1979 and 63% of the samples contained no larval johnny darters at transects 1 and 4. Table 21. Mean total length (mm) of johnny darter larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. (m) Contour Date 3 1 9 6 12 1978 (1 ) - 6.6 (35) - .6 (34) - 6.3 (32) - 11.4 (2 ) - 27 June 6.4 (2 0 ) 6.4 (24) 6.4 (13) 5.9 10 July 6.5 (8 ) 6.5 (51) 6.3 (42) 24 July 6.3 (1 ) 6.3 (51) 6 .5 (43) 6 7 Aug. 5.7 (2 ) 6.4 (37) 6.2 (28) 21 Aug. 7.2 (1 1 ) 6.0 (3) 0.0 28 June 5.7 (4) 6.4 (1 ) 0.0 10 July 6.2 (17) 6.3 (2 0 ) 6 .2 (4) 31 July 6 .2 (15) 6.6 (58) 6.5 (63) 15 Aug. 6 .6 (1 1 ) 6.0 (2 ) 9.0 (2 ) 27 Aug. 6 .6 (28) 6 (52) 6.5 (41) 25 Sept. 6.7 1979 (1 ) .6 6.6 (2 ) 7.3 (1 ) 0.0 0.0 0.0 0.0 6.5 (69) 6.3 (52) 8.6 (1 0 ) 7.4 (8 ) b .2 (42) 6.9 (50) 0.0 10.8 (1 ) 65 At all transects greatest densities of johnny darter larvae generally occurred at the 6 - and 9-m contours, though early in the year densities were greater at shallower contours (Table 20). Appearance of johnny darters at transect 2 was most probably related to substrate differences and early hatching at the shallow contours may be temperature related as these shallow areas warm quicker. A large percentage of larvae at deeper contours later in the season were also newly hatched (5.9 - 6.6 mm, Table 21), but some length data suggest johnny darter larvae moved offshore with increased growth. Diel distribution. Dramatic differences in diel catches of johnny darter larvae were apparent (Table 22). During the period of vulnerability to our gear in 1979, johnny darter larvae only occurred in 3.4% of day tows which accounted for 13 larvae collected in Lake Michigan. The greatest proportion of larvae collected during the day were taken at the 1- and 3-m contours where turbidity values were generally greatest. Two possible factors may be responsible for this disparity in day-night catches: gear avoidance and vertical distribution. Personal observation of johnny darter larvae captured with a 2 -m net and placed in small aquaria for several days indicate that (at least during daylight hours) these larvae are extremely demersal in behavior, and would be relatively invulnerable to gear used in the field. Table 22. Comparison of average day - night densities of johnny darter larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 1 3 Depth contours (m) 6 D N N )ate D N D 28 June 0 - 4 (5) 0 ■- 1 (3) 0 - 10 July 0 - 28 (57) 0 - 57 (106) 31 July 0 - 22 (27) 2 (5) 15 Aug. 27 Aug. 15 (22) 1 (3) 9 12 D N D N 0 - 0 - 0 - 0 - 0 0 -- 4 (5) 0 - 0 - 0 - 0 - 156 (97) 0 - 295 (298) (6) 1 (3) 0 - 4 (6) 0 - 4 (7) 0 26 (28) 0 - 74 (52) 0 - 57 (46) 0 - - - 247 (195) 0 - 99 (68) 27 (72) 0 - 7 (17) 40 (30) 0 - 45 (2) 67 Total abundance. Total johnny darter abundance in the reference Lake Michigan wedge fluctuated considerably over the study period. In 1978, total abundance of larvae was generally near 5 million or greater on all dates except in late August, and peaked at 21.1 million in early July (Table 23). Peak abundance differed slightly from peak density which occurred in late June. As with other species, dates of peak abundance and density differed because peak densities occurred at the 1 -m contour where volume of water is considerably less than at the deeper contours. Abundance estimates in 1979 were substantially lower than in 1978, but johnny darter larvae were present for nearly a month longer in 1979. Peak abundance of 19 million larvae was recorded at the end of July with a secondary peak of 3.6 million larvae at the end of August, 1979. The extended presence of johnny darter larvae in 1979 was undoubtedly due to an interruption of spawning and hatching in early August by a strong upwelling. These conditions may have also slowed growth of johnny darter larvae and thus decreased gear avoidance. By mid to late August 1978 some larvae over 11 mm were collected, but in 1979 by mid to late August the largest johnny darter larvae only averaged 9 mm and at the end of September, the largest larvae collected was only 10.8 m (Table 21). Table 23. 1978 Date Estimates based on night densities and 90% confidence interval (+) of johnny darter abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. (N = 119) Number 1979 a. Date Number 1979to (N = 238) Date Number 27 June 5285 (+7059) 28 June 9 (±15) 28 June 9 (±15) 10 July 21,531 (+21,082) 10 July 450 (+548) 10 July 450 (±54) 24 July 6959 (+6387) 31 July 19,088 (+10,927) 31 July 30,020 (±15,957 7 Aug. 4910 (+5256) 15 Aug. 1586 (+3025) 15 Aug. 2360 (±4278 27 Aug. 3596 (+1841) 27 Aug. 8561 (±5022 21 Aug. 119 (+159) 25 Sept. aAt 1978 contours only. faAt all contours sampled in 1979. 23 (±44) 25 Sept. 102 (±193) 69 Pere Marquette Marsh-Lake General distribution. Johnny darter larvae were not collected in great numbers in the Pere Marquette MarshLake, though adults were one of the most abundant fish there (Appendix Table 1C). No other species of adult darters were collected in the marsh, but a small number of larval logperch were captured. Johnny darter larvae first appeared at the upper marsh stations (1 and 2) in late May (13 larvae/1000 m 3). Some johnny darter larvae were collected on all remaining dates except 23 July, though densities were low (3 - 58/1000 rn3, Appendix Table 4B). Densities of johnny darter larvae were probably underestimated because most larvae were taken at upper marsh stations, where dense growths of macrophytes may have hampered collection of larvae. Greatest densities of johnny darter larvae (77/1000 m 3) occurred in the harbor outlet in water flowing from Lake Michigan to the marsh. Indeed, the only time johnny darter larvae occurred in the harbor outlet was when they were also present in outflow samples. Complex current patterns may have resulted in the catch of johnny darter larvae hatched in Lake Michigan but present in waters flowing out of the marsh. Riprap forming the Ludington harbor breakwalls may have provided spawning sites for johnny darters. Michigan State University scuba divers at the Ludington Pumped Storage Power Plant observed large 70 numbers of adult johnny darters utilizing the base of the protective jetty riprap during spawning season. Inputs to Lake Michigan. As previously mentioned, johnny darter larvae only appeared in the harbor outlet channel on three occasions, once in mid-June, and twice in early July (Appendix Table 4B). The density of larvae was greatest in water flowing from Lake Michigan to the marsh although substantial volumes of water were not entering Pere Marquette Marsh-Lake at this time. Only 9218 larvae were estimated to be transported from Lake Michigan to the marsh. On other dates when johnny darter larvae were present in the harbor channel, they were collected in outflowing water. Densities were generally low (9 - 30/1000 m 3 ) but large volumes of water were moving from the marsh-lake system to Lake Michigan (see Table 3) and some 390 thousand larvae were estimated transported to Lake Michigan on sample days. This estimate compares closely with point estimates of johnny darter larvae in the reference wedge in Lake Michigan during low densities (e.g., start of hatch and end of hatching). However, this estimate is considerably lower than abundance (1 . 6 million larvae) during most of 1978 and 1979. 21.1 71 Cottus spp. Both the mottled and slimy sculpin occur in this area of Lake Michigan but identifying and separating adults with 100% reliability is difficult. Thus all larvae were identified as Cottus spp., though Heufelder and Auer (1980) have recently described techniques for separating the larvae of these two species. Sculpins are protective nesters that lay their eggs on the underside of rocks. Spawning usually occurs from early May to early June in this area of Lake Michigan at water temperatures of 5 - 10 C (Brazo and Liston 1979). Eggs take approximately 4 weeks to hatch at these temperatures (Van Vleit 1964). Lake Michigan Temporal distribution. Sculpin larvae first appeared in ichthyoplankton collections in mid-June in both 1978 and 1979. Densities were low at this time and ranged from 0 - 12 larvae/1000 m 3 in individual tows. Densities of larval Cottus spp. peaked and declined rapidly in both years, lasting about 3 weeks from late June to mid-July. Large unimodal peaks in larval fish densities suggest that spawning occurred within a relatively short time. advanced developmental state of The Cottus spp. larvae at hatching also reduces the length of time of vulnerability to gear. Peak densities recorded in 1978 were near 140 larvae/ 1 0 0 0 m 3, while peaks of 108/1000 m 3 were observed in 72 1979 (Table 24); water temperatures were from 10 - 15 C during peaks of both years. Some larval sculpins were collected in late August each year but densities were low after July and water temperatures were generally >15 C. Annual differences in hatching time and peak densities of larvae may be attributed to physical conditions in Lake Michigan. Waters warmed to 10 C considerably earlier in 1978, whereas, prolonged upwellings in June 1979 may have delayed peak hatching of Cottus spp. until early July. The effect of cold water on delayed hatching and growth is supported by examination of length data. During the warm year of 1978, mean lengths of sculpin larvae ranged from 8.6 to 12.7 mm by mid-July and sizes of 21 - 22 mm were attained by mid-August (Table 25). In contrast, by mid- July of 1979 larval sculpins were on the average 8.4 mm long and had only grown to 13.2 mm by mid-August after which no larvae were collected. Spatial distribution. Sculpin larvae were not uniformly distributed in the study area. Few larvae were captured at the shallowest contour and densities generally increased lakeward (Table 24). One exception was on 27 June 1978 when mean densities of 50 larvae/1000 m 3 and 59 larvae/1000 m 3 were observed at the 1- and 3-m contours. These data are somewhat confusing but local climatological data provide a plausible explanation. Strong south- southwest winds had prevailed for 48 hours prior to collection and continued on the sample day. Quite possibly Table 24. Average nighttime density and range (in parentheses) of Cottus spp. larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. Date Depth Contour (m) 6 3 1 9 12 1978 14 June 0 2 1 (0 - 7) 3 (0 - 8) - 27 June 50 (0 - 140) 59 (0 - 133) 48 (0 - 119) 21 (0 - 44) - 10 July 2 (0 - 14) 3 (0 - 17) 7 (o - 27) 19 (0 - 61) - 24 July 0 3 (0 - 20) 0 3 (0 - 6) - 7 Aug. 0 0 0 0 21 Aug. 0 0 3 12 June 0 0 0 28 June 1(0- 10 July (0 - 9) 3 (0 - -j - 10) LO 1979 0 2 (0 - 9) 8 (0 - 18) 8 (0 - 39) o (0 - 18) 17 (0 - 60) 0 10 (0 - 35) 20 (0 - 71) 21 (0 - 79) 33 (0 - 56) 31 July 0 0 0 3 (0 - 15) 30 (0 - 108) 15 Aug. 0 0 0 0 5) 1 (0 - 8) Table 25. Mean total length (mm) of Cottus spp. larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parentheses. Contour Bate J 1 (m) 9 6 12 1978 14 June 0.0 27 June 9.0 10 July 12.7 7.8 (2 ) (9) 8.7 (27) (1 ) 9.6 13.8 7.2 (1 ) 8.8 (24) (1 ) 8.6 (5) (3) 0.0 24 July 0.0 21 Aug. 0.0 0.0 21.2 12 June 0.0 0.0 0.0 28 June 7.8 10 July (4) 8.2 - (2 ) 8.8 (14) - 9.2 (1 0 ) - 16.9 (2 ) 22.3 (4) - 1979 0.0 7.3 (1 ) 7.8 (8 ) 7.8 (6 ) 8.4 (3) 7.9 (13) 0.0 8.4 (1 2 ) 8.4 (23) 8.5 (18) 8.4 (31) 31 July 0.0 0.0 0.0 9.0 (2 ) 9.1 (2 2 ) 15 Aug. 0.0 0.0 0.0 0.0 13.2 (1 ) (1 ) 75 Cottus spp. larvae could have been transported shoreward via currents driven from these winds. Transects 2 and 3 had the greatest aggregation of sculpin larvae on most dates, and in fact no larvae were collected at transect 1 in 1978. Abundance of larvae at transects 2 and 3 was believed related to substrate factors (see johnny darter discussion). Diel distribution. Nearly all (98%) sculpin larvae were collected at night in 1979 (Table 26). Paucity of larvae in day samples may be attributed to avoidance and dispersal of larvae in the water column. Observations in the laboratory indicated newly hatched sculpin larvae were quite pelagic but become demersal 3 - 4 days after hatching. Total abundance. Estimates of sculpin abundance in the reference wedge of Lake Michigan could be made from June to August in both years. In 1979, total abundance ranged from 187 thousand larvae at the end of this time period to 1.5 million sculpins during peak densities (Table 27). Abrupt declines in abundance after July were most likely due to avoidance of collection gear and the demersal behavior characteristic of larger sculpin larvae and YOY. For the 2-week period in late June - early July, 1978, estimated larval fish abundance in the reference wedge in Lake Michigan was 1.7 - 2.3 million larvae. On all other dates when sculpin larvae were present abundance was Table 26. Comparison of average day - night densities of Cottus spp. larvae (number/1000 m3} at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. Depth contours (m) 12 Date N 12 June 0 0 0 0 0 0 0 0 0 2 (4) 28 June 0 - 1 (7) 0 - 8 (6) 0 - 8 (15) 0 - 8 (7) 0 - 17 (23) 10 July 0 - 0 - 1 (3) 10 (14) 2 (4) 20 (25) 0 - 21 (30) 0 - 33 (32) 01 July 0 - ) 0 - 0 - 0 - 0 - 0 - 0 - 3 (6) 0 - 30 (33) 15 Aug. 0 - 0 - 0 - 0 - 0 - 0 - 0 -- 0 - 0 ~ 1 (3) Table 27. 1978 Date Estimates based on night densities and 90% confidence interval (+ ) of Cottus spp. abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 1979a (N = 144) Number Date 179 (+250) 12 June 27 June 2321 (+1974) 28 June 10 July 1180 (+1366) 24 July 164 (+230) 14 June 7 Aug. 0 Number 12 June 193 (+412) 349 (+421) 28 June 2218 (+2 1 0 2 ) 10 July 1524 (+1434) 10 July 5195 (+3820) 31 July 188 (+235) 31 July 3561 (+2661) 0 - 15 Aug. 0 - 21 Aug. 1979to (N = 182) Date Number 207 (+254) ^At 1978 contours only. toAt all contours sampled in 1979. 15 Aug. 10 (±2 0 ) 78 consistently between 150 and 200 thousand larvae in the wedge. Pere Marquette Marsh-Lake General distribution. No Cottus spp. larvae were found in the Pere Marquette Marsh-Lake system in 1981, and no larvae were collected in the harbor outlet channel. This is not surprising since no adults were collected in the Pere Marquette Marsh-Lake (Appendix Table 1C), and suitable spawning substrate was lacking. Further, no sculpin larvae came in from Lake Michigan which may be related to the offshore distribution. Ninespine Stickleback Adult ninespine sticklebacks are common in this area of Lake Michigan during their spawning period from late spring to early summer (Liston et al. 1980). Wootton (1976) reported ninespine sticklebacks built shallow depression-type nests between rocks and gravel in Lake Huron in the absence of aquatic macrophytes, and similar substrate conditions exist in the study area. Ninespine stickleback larvae comprised 2.3% of the Lake Michigan catch and were the fifth-most abundant larvae collected in the lake in 1978-79. 79 Lake Michigan Temporal distribution. Ninespine stickleback larvae first appeared in mid and late June in 1978 and 1979, respectively. Densities at first occurrence averaged 1 - 24 larvae/1000 m 3 (Table 28). In 1978, peak abundance occurred from late June to early July at densities up to 492 larvae/1000 m 3 . approximately 2 - 3 were In 1979 peak abundances occurred weeks later in late July, when larvae present in 80% of the samples, and densities were as high as 680 larvae/1000 m3 . Water temperatures during the peak density of both years generally ranged from 10 to 15 C. Some ninespine stickleback larvae were collected in August of both years but densities rarely were greater than 15/1000 m 3 and frequency of occurrence was only 32.5%. No larvae were observed in September and October samples. As with most other species of larvae discussed, the dramatic differences in physical conditions in Lake Michigan may explain the variation in annual distribution of ninespine stickleback larvae. They hatched earlier and reached peak densities sooner in 1978 than in 1979. Substantial densities of larvae were present until late August 1979, but only a single larva was collected after .late July in 1978. Inspection of length data revealed little about growth of larvae during the 2 years because only a narrow size range of larvae was collected. Lippson and Moran (1974) report other species of stickleback hatch at 4 - 5 mm. Most fish in the present study were 7 - 9 mm Table 28. Date Average nighttime density and range (in parentheses) of ninespine stickleback larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. ______________________________ Depth contour (m)___________________________________ .1 3 6 9 12 1978 14 June 4 27 June 0 10 July 0 24 July s 7 Aug. (0 - (0 - 14) 14) 0 1 - (0 - 7) 0 0 121 (36 - 259) 92 (0 - 312) 36 (0 - 68) - 6 (0 - 25) 32 (0 - 74) 126 (0 - 492) - 7 (0 - 23) 8 (0 - 24) 4 (0 - 6) - 3 (0 - 11) - 0 0 .979 28 June 2 (0 - 1 ) 20 (0 - 62) 40 (0 - 98) 24 (0 - 59) 23 (0 - 78) 10 July 1 (0 - 8) 24 (0 - 65) 46 (0 - 125) 22 (0 - 57) 20 (0 - 40) 31 July 7 (0 - 38) 24 (0 - 79) 36 (0 - 130) 105 (32 - 273) 262 15 Aug. 2 (0 - 16) 2 (0 - 13) 2 (0 - 8) 11 (0 - 67) 8 (0 - 16) 27 Aug. 0 3 (0 - S) 10 (0 - 40) 0 0 (64 - 680) 81 (Table 29) indicating they were not newly hatched which was further supported by the paucity of yolk in larvae. Possibly these larvae were hatching offshore and passively drifting to nearshore water. However, 7.5 - 8.5-mm larvae were taken routinely into late July and mid-August in 1979 suggesting hatching was more prolonged than in 1978. Spatial distribution. Spatial distribution of ninespine stickleback larvae appeared to be directly correlated with depth. Few larvae were collected at the 1- m contour, and densities tended to increase lakeward with several densities greater than 100 larvae/ 1 0 0 0 m 3 recorded at the 9- and 12-m contours (Table 28). This lakeward distribution was even more pronounced later in the year. One notable exception was on 27 June 1978 when densities of ninespine stickleback larvae decreased lakeward, almost an identical phenomenon as reported above with Cottus spp. larvae, and most probably resulting from similar wind and current conditions. Larvae tended to be equally abundant at transects 2, 3, and 4 though densities at transect 2 were slightly higher. Densities at transect 1 were considerably lower and may be attributed to substrate preferences of spawning adults similar to that theorized for johnny darter and Cottus spp. larvae. Diel distribution. Larvae of ninespine sticklebacks were remarkably more susceptible to our collection techniques at night than during the day (Table 30). Only Table 29. Mean total length (mm) of ninespine stickleback larvae at depth contours in Lake Michigan, 1978 and 1979. Numbers measured in parenthesis. Contour Date .J 1 (m) 9 6 12 1978 14 June 7.6 27 June 0.0 7.8 10 July 0.0 2.4 July 8.3 7 Aug. (4) (1 ) 7.7 (1 ) - 0.0 (32) 7.6 (24) 7.8 (24) - 8.0 (2 ) 7.6 (2 0 ) 7.8 (37) - 17.3 (7) 7.7 (7) 8.1 (3) - 7.7 (3) - 0.0 0.0 0.0 0.0 1979 28 June 7.6 (2 ) 7.9 (19) 7.7 (34) 7.9 (2 2 ) 7.8 (2 0 ) 10 July 6.1 (1 ) 7.8 (29) 7.7 (43) 7.8 (2 2 ) 7.7 (27) 31 July 8.0 (2 ) 7.5 (16) 7.8 (18) 7,8 (54) 7.7 (73) 15 Aug. 7.9 (2 ) 8.1 8,0 (7) 7.6 (8 ) 27 Aug. 0.0 7.8 (3) 7.9 (14) 0.0 (1 ) 8.2 0.0 (2 ) 83 techniques at night than during the day (Table 30). Only 7.6% of the larvae from Lake Michigan in 1979 were observed in day samples, and daytime densities exceeded on but two occasions. 20/1000 m3 The paucity of larvae in day samples appears to be related to an interaction of gear avoidance and distribution in the water column. Personal observations of ninespine stickleback larvae in aquaria indicated that they were demersal during the day. Jude et al. (1978) also observed greater densities of ninespine stickleback larvae in night tows than in day tows at a site in Lake Michigan approximately 93 km south of the present study site. Total abundance. Estimates of ninespine stickleback larvae in the reference wedge in Lake Michigan ranged from slightly less than 150 thousand at the end of August, 1979, to 6.5 million at the end of July in 1979 (Table 31) and from 10 thousand in June to 7.5 million in early July in 1978. On all 1979 sample dates but one, numbers were estimated to be greater than abundance only exceeded 1 1 million individuals, but million larvae on two of five dates in 1978 (Table 31). Pere Marquette Marsh-Lake General distribution. No larvae of ninespine sticklebacks were collected in the Pere Marquette MarshLake system, nor were any found in the harbor-outlet channel. No adults were collected in this habitat in 1981 Table 30. Comparison of average day - night densities of ninespine stickleback larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. -J 1 Date Depth contours (m) 6 N D N D N D 28 June 0 - 2 (3) 0 - 20 (24) 0 - 10 July 1 (3) 1 (3) (6) 23 (24) 0 - 7 (14) 0 - 31 July 15 Aug. 27 Aug. 10 (12) 0 _ o z. (6) 0 ~ 9 12 D N D N 40 (36) 1 (3 24 (22) 2 (4) 23 (32) 0 - 46 (41) 0 - 22 (24) 1 (4) 20 (44) 24 (24) 0 - 4b (48) 0 (9) 22 (76) 1 (6( 20 (204) 35 (39) 2 (4) 3 (9) 2 (4) (3) 11 (25) 1 (2) 8 (16) 0 ~ 0 0 0 — 3 (4) 1 (1) 10 (17) ■” 0 Table 31. 1978 Date 14 June Estimates based on night densities and 90% confidence interval (+ ) of ninespine stickleback abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. (N = 119) Number 1979a Date Number 1979*= (N = 198) Date Number 10 (±19) 27 June 4175 (+3986) 28 June 2071 (+1292) 28 June 4645 (+3684) 10 July 7532 (+8812) 10 July 2075 (+1397) 10 July 4260 (+2612) 24 July 379 31 July (+347) 7 Aug. 19 6505 31 July (+3477) 15 Aug. (+244) 653 15 A u g . (+1079) 27 Aug. 149 (+139) aAt 1978 contours only. toAt all contours sampled in 1979. 35,556 (+18,612) 1528 (+1591) 27 Aug. 1233 (+1389) 86 (Appendix Table 1C), though macrophyte beds may offer suitable spawning sites. Preference for cooler temperatures may partially explain their absence from Pere Marquette Marsh-Lake. This seems to refute Liston et al.'s (1980) notion that some spawning of ninespine stickleback may occur in coastal marsh systems. Spottail Shiner Spottail shiners were the dominant minnow in the Lake Michigan study area. They move into shallow nearshore areas in late spring and early summer where they spawn over sandy shoals or algal mats (Lippson and Moran 1974). Based on spawning times and larval fish presence, spottail shiner eggs may take 2 -3 weeks to hatch at ambient summer temperatures in Lake Michigan (14 - 20 C). This species was the sixth-most abundant larvae collected in Lake Michigan. Lake Michigan Temporal distribution. Larvae of spottail shiners first appeared in late May to early June in 1978 and 1979, respectively. Mean densities at all contours were low at this time (<5 larvae/1000 m 3 , Table 32). Peak densities (up to 394 larvae/1000 m 3) occurred from late June through early July when water temperatures ranged from 7 to 13 C. However, larvae were collected on all dates through late 87 September in both years, indicating an extended spawning period. Though temporal distribution was essentially the same in both years, except for a slightly earlier hatching time in 1978, densities of larvae were substantially lower in 1979 (Table 32). Again the rather cool summer of 1979 may account for this difference. Growth data would seem to support the influence of prolonged cold water. Newly hatched spottail shiner larvae (mean = 5.1 mm) were still occurring in late August 1979 and greatest length of larvae was only 9.2 ram in late September (Table 33). In contrast, no newly hatched spottail shiner larvae were collected after early August 1978, and larvae as large as 12.5 mm were collected in late September. Spatial distribution. Distinct spatial distribution was exhibited by spottail shiner larvae. abundant at the 1 They were most -m contour, though on some occasions in 1978 high numbers were taken at the 3 -m contour (Tcible 32) , which suggests warmer water temperatures may be a key to distribution (see Appendix Table 1A). Jude et a l . (1979a) also observed highest densities of spottail shiner larvae in shallow water in southeastern Lake Michigan. Distribution of larvae among transects was non-uniform and larvae were concentrated at transects 1 and 2. Brazo and Liston (1979) reported spawning adult spottail shiners were markedly more abundant in trawl hauls at transect trawl stations near transects 3 and 4. 1 than at Table 32. Average nighttime density and range (in parentheses) of spottail shiner larvae (number/1000 m3) __________ at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data._______ Depth contour (m) Date 1 3 6 12 9 1978 31 May 2 (0 - 14) 0 0 0 - 14 June 2 (0 - 12) 0 0 0 - 27 June 123 (0 - 196) 10 (0 - 24) 2 (0 - 10) 8 (0 - 36) - 10 July 119 (14 - 394) 49 (0 - 148) 10 (0 - 24) 6 (0 - 25) - (0 - 32) 0 1 (0 - 6) - 24 July 74 (0 - 15 5 ) 6 7 Aug. 45 (0 - 150) 0 15 (0 - 41) 1 1 (0 - 5) .12 June 4 (0 - 28 June 102 10 July 0 - 0 0 - 0 0 0 - 30) 0 0 0 (0 - 381) 9 (0 - 46) 0 'I 1 35 (0 - 106) n iL. (0 - 8) 0 0 31 July 17 (0 - 54) 0 0 1 15 Aug. 12 (0 - 40) 2 (0 - 13) 0 0 0 27 Aug. 13 (0 - 31) 5 (0 - 24) 1 0 0 25 Sept. 3 (0 - 11) 0 21 Aug. 21 Sept. 2 (0 - 12) (0 - 6) 1979 0 (0 - 8) 0 0 (0 - 9) 1 (0 0 (0 - 11) 0 0 10 ) Table 33. Mean total length (mm) of spottail shiner larvae at depth contours in ___________Lake Michigan, 1978 and 1979. Numbers measured in parentheses.______ Contour (m) 1 3 6 Date 9 12 1978 31 May 5.6 (1) 0.0 0.0 0.0 - 14 June 4.9 (1) 0.0 0.0 0.0 - 27 June 5.2 (18) 5.4 (4) 5.7 (1) 5.8 (5) - 10 July 6.0 (20) 5.7 (20) 5.1 (7) 5.6 (4) - 24 July 6.2 (33) 5.1 (6) 0.0 5.6 (1) - 7 Aug. 5.7 (18) 0.0 5.4 21 Aug. 10.5 (7) 0.0 21 Sept. 12.5 (1) 1979 12 June 5.0 28 June 0.0 - 0.0 0.0 - 0.0 0.0 0.0 (1) 0.0 0.0 0.0 4.8 (35) 4.8 (8) 0.0 4.6 10 July 5.0 (29) 5.3 (2) 0.0 0.0 31 July 5.3 (8) 0.0 0.0 8.0 15 Aug. 5.5 (6) 5.7 (1) 0.0 0.0 0.0 27 Aug. 7.3 (13) 5.1 (4) 5.3 0.0 0.0 25 Sept. 9.2 (2) 0.0 0.0 0.0 0.0 (2) (1) 0.0 (1) 5.3 0.0 (1) 0.0 (1) 90 Diel distribution. Substantially more larvae (89%) were eolleeted at night than during the day (Table 34), which is related to behavioral patterns. Spottail shiner larvae are highly demersal and undoubtedly are capable of avoiding the gear during the day. During the night, the larvae become more pelagic and more susceptible to collection gear. Total abundance. Total estimated abundance in the Lake Michigan wedge varied considerably from 1978 to 1979. Peak abundances of 87 6 thousand and 208 thousand larvae, respectively, were attained at nearly the same time (late June - early July) (Table 35). In .1978, larval fish abundance rose quickly from 2000 fish in late May to midJune to peaks of just under 1 million larvae in late June to early July, after which consistent declines to 1200 larvae in late September occurred. In 1979 abundance of larval spottail shiner did not vary as greatly and only rose above 100 thousand larvae on one date (late June). The remainder of the year, larval fish abundance was consistently between 20 thousand and 90 thousand fish except for lower fish numbers early in the season and at the end of the field season (Table 35). Pere Marquette Marsh-Lake General distribution. Spottail shiner larvae were collected from early May to early June in the Pere Marquette Marsh-Lake system in 1989, which was more than a Table 34. Comparison of average day - night densities of spottail shiner larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. late D N D Depth contours (m) 6 D N H 12 June 0 - 4 (11) 0 - 0 - 28 June 3 (5) 102 (56) 0 - 10 July 5 (8) 35 (33) 31 July 0 23 (26) 1 15 Aug. 27 Aug. 25 Sept. -t •J (6) 0 - 9 12 D N D N 0 - 0 - 0 - 0 - 0 - 0 9 (17) 0 - 0 - 0 1 (3) 0 - 1 (3) 0 - 2 (4) 0 -- 0 - 0 - 0 - 0 - 0 - 17 (19) 0 - 0 - 0 - 0 - 0 1 0 0 - 12 (16) 0 - 2 (4) 0 - 0 - 0 - 0 - 0 - 0 - L3 (12) 0 - 5 (9) 0 - 1 (3) 0 - 0 _ 0 - 0 - O (4) 0 - 0 - 0 - 0 - 0 - 0 - 0 - 0 - - - (4) Table 35. 1978 Date 31 May Estimates based on night densities and 90% confidence interval (+) of spottail shiner abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. (N = 192) 1979a Number Date 197913 (N = 262) Number Date Number 2 (±4) 14 June 1 12 June (±3) 3 (±6 ) 12 June 3 (±6 ) 27 June 621 (+873) 28 June 208 (±248) 28 June 347 (±547) 10 July 876 (+1080) 10 July 42 (±35) 10 July 42 (±34) 24 July 172 (+253) 31 July 92 (±157) 31 July 92 (±157) 7 Aug. 64 (±74) 15 Aug. 19 (±27) 27 Aug. 21 Aug. 21 Sept. 1 20 15 Aug. 25 Sept. (±2 ) aAt 1978 contours only. toAt all contours sampled in 1979. 54 (±75) 2 (±2 ) 20 (±29) (±29) 27 Aug. 25 Sept. 54 (±75) 2 (±2 ) 93 month before they appeared in Lake Michigan at the 1-m contour (Appendix Table 5B). marsh were low (6 May. However, densities in the larvae/ 1 0 0 0 m3 ) and peaked in late - 2 2 After early June, no spottail shiner larvae were collected in the marsh. All larvae were taken at upper marsh stations and no spottail shiners were recorded from the harbor-outlet station nor station 3 (immediately upstream). Many adult spottail shiners were collected in small-mesh trap nets at the upper marsh stations in 1981 (Appendix Table 1C). Input to Lake Michigan. There was apparently no transport of spottail shiner larvae between the Pere Marquette Marsh-Lake system and Lake Michigan though they were present in both habitats in 1981 and reached densities as high as 504 larvae/1000 m 3 in Lake Michigan in 1981. These data indicate that spottail shiner larvae may not be overly susceptible to passive transport and indeed are known to be highly demersal in behavior. Deepwater Sculpin Reproductive biology and early life history of the deepwater sculpin is poorly understood (Scott and Crossman 1973) due to their deep water distribution and difficulties in collecting at these depths. Recent evidence of Khan and Faber (1974), Westin (1969), and Jude et al. (1979a) suggests that spawning probably occurs in the winter 94 months. An extended spawning season from October through February is likely based on presence and developmental stage of larvae from collections in several areas of Lake Michigan (Mansfield et al. 1983). Larvae of deepwater sculpins comprised 1.7% (265 individuals) number of larvae collected of the total in 1979, which is significant in light of the total lack ofadults in samples Ludington (Brazo and Liston 1979). collectednear However, exploratory fishing by the U.S. Fish and Wildlife Service and Michigan Department of Natural Resources indicates adult deepwater sculpins are abundant in water >70 m off Ludington (Myrl Keller, personal communication, Michigan Department of Natural Resources). Lake Michigan Temporal distribution. Deepwater sculpin larvae were present in earliest ichthyoplankton collections in midApril of both years. Water temperatures were 3 - 5 C and densities ranged from 3 to 35/1000 m 3 . Mean length (10.8 mm) and length range (8.9 - 13.3 mm) indicated that hatching must have occurred earlier, based on a minimum hatching size of 7 . 6 mm observed by Khan and Faber (1974). However, presence of large and small larvae simultaneously suggests a possible extended spawning season. Jude et al. (1979a) collected deepwater sculpin larvae as early as February in 1978 at a site approximately 93 km south of the present study site. 95 Highest densities of deepwater sculpin larvae were recorded in early May each year and occasional single specimens were collected in every month except September 1979. In 1978, no larvae were collected after early June. No larvae were ever collected at any time when water temperatures exceeded 10 C. Larvae collected in the summer were taken during periods of upwelling and were larger (13 - 22 mm), indicating they were spawned earlier in the year. Spatial distribution. Abundance of deepwater sculpin larvae increased with increasing water depth and greatest densities generally occurred at the 9- and 12-m contours (Table 36) where up to 35 larvae/1000 m 3 were taken. Unfortunately supporting data for 1978 are not available because samples were not taken at deeper contours until mid-May and peak densities may have already occurred judging by the relatively high mean density (19 larvae/1000 m3 ) at the shallowest contour in early May. Deepwater sculpin larvae were relatively uniformly distributed among all transects, suggesting wide and varied transport. Diel distribution. Most deepwater sculpin larvae were collected at night (Table 37), with night tows accounting for 79% of the larvae taken in Lake Michigan. greater than 10 Densities larvae/ 1 0 0 0 m 3 were rare in day samples. However, deepwater sculpin larvae occurred more frequently in day tows (30%) than did other cottid larvae (5%). Low numbers of larvae collected during the day may reflect gear avoidance and demersal distribution, but the greater Table 36. Average nighttime density and range (in parentheses) of deepwater sculpin larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. Date 1978 Depth contour (m) 6 3 1 9 12 22 April 3 (0 - 10 ) - - - - 3 May 19 (0 - 01 ) - - - - 15 May 2 (0 - 11) 9 31 May 0 0 0 14 June 0 0 7 (0 - 27) 3 (0 - 20 ) (0 - 20 ) 10 (0 - 35) 25) 24 (10 - 33) 3 (0 - 9) 6 (0 - 33) (0 - 8) (0 - 9) (0 - 25) 12 (0 - 19) - 18 (11 - 28) - 0 - 1979 c 4 (0 - 13) 8 (0 - 16) 4 (0 - 16) 8 1 May 5 18) 7 (0 - 15) 10 (0 - 34) 10 15 May 0 2 (0 - 9) 1 (0 - 6) 0 29 May 0 1 (0 - '7) 4 (0 - 17) 10 12 June 0 0 0 0 28 June 0 0 0 1 iO July 0 0 0 0 1 31 July 0 0 o (0 - 14) 0 0 15 Aug. 0 0 0 0 -> 17 April (0 - (0 - (0 - 33) 0 (0 - 9) 0 Table 37. Comparison of average day - night densities of deepwater sculpin larvae (number/1000 m 3 ) at several depth contours In nearshore Lake Michigan, parentheses. D = day, N = night. 3 1 Date 17 April 1 May 15 May D N _ 4 (6 ) 12 June 28 June 10 July 31 July ]5 Aug. _ 8 - (7) One standard deviation in 9 12 D N D 4 (7) _ 8 _ ~ (7) - 10 - 10 _ (14) _ “ 5 (7) - 7 (6 ) 0 0 0 2 (4) 1 (2 ) 1 (3) 0 (6 ) 1 (2 ) 2 (3) 4 (6 ) 29 May D Depth contours (m) 6 N D N 1979. - N 10 (15) “ 24 (9) - 5 (5) 3 (3) 1 (2 ) 10 (11 ) 4 (5) 6 (8 ) (9) 0 0 0 - - - 0 0 0 0 0 0 - 2 (5) 0 - - - - - - (3) - 0 0 0 0 0 0 0 0 - - - - - - 1 (3) 0 - - - 0 0 0 0 0 0 1 0 0 1 - - - - _ - (3) - - (3) 0 0 0 0 0 2 0 0 0 0 - - - - - (5) - - - - 0 0 0 0 0 0 2 ~ 1 (2 ) 0 — 0 (3) 3 (4) 98 occurrence of deepwater sculpin in day samples compared to Cottus spp. larvae suggests that the former may be more pelagic than the latter. Total abundance. Abundance of deepwater sculpin larvae at deeper contours cannot be reported for 1978 before mid-May. During May and early June 0.5 - 1.2 million larvae were present in the reference wedge. However, peak abundances may have occurred prior to this date. In 1979, some deepwater sculpin larvae were taken on all dates through mid-August, except in mid-June when water temperatures were 10 - 11 C at all contours. Total abundance in the reference wedge (including the 12 -m contour) ranged from 30.4 thousand individuals on 31 July to 3.5 million larvae on 1 May (Table 38). Abundance in April and May was usually greater than 1 million individuals, but numbers were reduced after May. The paucity of deepwater sculpin larvae later in the year was most likely related to movement into cold deeper waters that are generally considered to be preferred habitat of juveniles and adults. Pere Marquette Marsh-Lake General distribution. No larvae of deepwater sculpin were collected in Pere Marquette Marsh-Lake or its harbor outlet channel in 1981. As with other cottids and ninespine stickleback, no adult deepwater sculpin were collected in Pere Marquette Marsh-Lake (Appendix Table 1C) Table 38. Estimates based on night densities and 90% confidence interval (+) of deepwater sculpin abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples._______________________ 19791978 (N = 91) 1979to (N = 318) Date Number Date Number Date Number 22 April0 3 17 April0 4 17 April 1,621 (+5) (+1204) (±4) 3 May° 15 May 31 May 14 June 16 (+29) 1260 (+594) 0 532 (+488) 1 May° 4 (+5) 1 May 3471 (+1716) 15 May 23 (+45) 15 May 380 (+327) 29 May 650 (+504) 29 May 1276 (+1094) 12 June 28 June 10 July 0 62 (+118) 0 12 June 28 June 10 July 0 62 (+118) 110 (+208) 31 July 15 Aug. aAt 1978 contours only. ^At all contours sampled in 1979. ^Estimates made only for 1-m contour. 30 (±59) 0 31 July 30 (±59) 15 Aug. 32.2 (±301) 100 which is not surprising as adult deepwater sculpins are known to prefer colderr deeper waters in large lakes and are generally restricted to depths of 60 - 70 m and greater (Scott and Crossman 1973). Lake Whitefish Brazo and Liston (1979) reported adult lake whitefish are uncommon in the study area except during the spawning season. During October and November large numbers of fish migrate into the area to spawn over a shelf-like shoal off the Ludington Plant (see Figure 1). This site is historically considered to be a primary lake whitefish spawning site (Organ et al. 1979). Lake whitefish eggs are large, demersal, and require about 120 - 1.40 days to hatch at water temperatures of 0.5 - 2.0 C (Price 1940). Larvae of lake whitefish comprised only a small percentage of all ichthyoplankton collected and were distinguished from round whitefish and bloater larvae by their larger size at developmental stage, melanophore patterns (Hinrichs 1979), and because round whitefish are known to cover their eggs with gravel during spawning and larvae emerge in a more advance state with yolk sac already utilized (Booke 1970). 101 Lake Michigan Temporal distribution. Lake whitefish larvae were spring cohorts of deepwater sculpin and burbot, and were likewise present in the initial samples taken in mid-April of both years. Densities at this time were low (average 1 - 40 larvae/1000 m 3), and mean size and range of larvae (12.2 mm; 11.2 - 13.2 mm) suggest that all larvae were newly hatched with large amounts of yolk. Peak densities of nearly 300 larvae/1000 m 3 were recorded in early May of both years at water temperatures of 5 C (Table 39). Numbers of lake whitefish larvae were reduced dramatically after this, although in 1979 some were collected through early June. May. No larvae were collected in 1978 after early Paucity of lake whitefish after early May was probably related to avoidance due to growth and distributional factors including preferred water temperatures. Spatial distribution. Lake whitefish larvae were taken almost exclusively in water 3 m or less in depth (Table 39). Mean densities were an order of magnitude higher at the 1-m contour than in deeper water in 1979. High mean densities of 272 larvae/1000 m 3 were also recorded for the 1-m contour in 1978, but comparative data at deeper contours were not collected. Faber (1970) also noted larvae of lake whitefish were more abundant in tows made in 1 - 3 m of water than in deeper water in the south bay of Lake Huron. Table 39. Average nighttime density and range (in parentheses) of lake whitefish larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. Date Depth contour (m) 6 3 1 9 12 1978 22 April 3 May 40 272 63) - - - - (0 - 338) - - - - 47) 2 (0 - 9) 0 0 (5 - 291) 4 (0 - 9) 2 (0-7) 1 (0 - (0 - 2 (0 - 10) 1 (0 - 6 ) 0 0 1 (0 - 0 0 0 0 0 0 (9 - 1979 17 April 19 103 15 May 8 29 May 0 12 vlune 2 (0 - 45) 7) 0 7) 4) 1 (0 - 7) T (0 - 7) 102 1 May (10 - 103 Diel distribution. Only a single lake whitefish larva was taken during the day in Lake Michigan in 1979 (Table 40). This may have been a function of distribution in the water column or avoidance of nets, though no day tows were made until 15 May, 1979 well after the period of peak abundance. Total abundance. Abundance estimates ranged from 2300 fish on 12 June to 348 thousand at peak densities in early May in 1979, in the reference wedge out to the 12-m contour (Table 41). Although reliable estimates of abundance in the reference wedge could not be made at deeper contours in 1978 due to lack of samples, some 234 thousand larvae were estimated at the 1 -m contour, suggesting overall abundance may have been greater in 1978 than 1979. After peak abundance, numbers dropped rapidly and consistently in ensuing weeks. Estimated total numbers were substantially lower than for other species because lake whitefish larvae were primarily restricted to the shallowest contour where water volumes were small. Pere Marquette Marsh-Lake General distribution. Only two lake whitefish larvae were present in the Pere Marquette Marsh-Lake system in 1981 and both were collected on the earliest sampling date, 9 April. One larva was present in the harbor outlet channel, and one larva was collected at the nearest station upstream from the harbor. Both larvae were relatively Table 40. Comparison of average day - night densities of lake whitefish larvae (number/1000 m3) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. 1 Date 17 Apri D _ - 1 May - - N D 19 ( 16) - (4) - 4 (5) - 103 (122) _ 2 _ - - 0 “ 9 12 D N _ 0 - (3) - - 1 (2 ) D - N 1 (3) - 2 (3) - 8 (5) 1 (4) 2 (4) 0 0 - 0 0 - 1 (2 ) 0 - - - - 0 0 0 0 0 0 0 0 - - 1 (2 ) 0 - - - - - - - .12 June 0 0 2 0 0 0 0 0 0 0 28 June 0 0 0 0 0 0 0 1 0 0 29 May 0 (3) 104 15 May Depth contours (m) 6 D N N Table 41. 1978 Date (N = 1 2 ) Number 22 April° 3 May^ Estimates based on night densities and 90% confidence interval (+) of lake whietfish abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lake Michigan. N = number of samples. 35 (±43) 234 (±294) 1979a Date Number 1979to (N = 160) Date Number 17 April^ 16 (±1 1 ) 17 April 181 (±362) 1 Mayc 89 (±70) 1 May 348 (±601) 15 May 30 (±56) 15 May 30 (±56) 29 May 5 (±1 0 ) 29 May 5 (±1 0 ) 12 June 2 (±4) aAt 1978 contours only. toAt all contours sampled in 1979. ^Estimates made only for 1-m contour. 12 June 20 (±4 106 large (13.5 - 14.2 rrun) indicating they were not newly hatched. Unfortunately no current direction or velocity measurements were made in the harbor outlet at this time and samples in Lake Michigan were postponed due to inclement weather. However, on the next sample date, 22 April, densities of 17 - 22 larvae/1000 m 3 were recorded in Lake Michigan at the 1-m contour. Most likely these larvae were transported from Lake Michigan to the marsh as no other lake whitefish larvae were collected in the marsh system and no spawning habitat exists. No adult lake whitefish were ever collected in the Pere Marquette MarshLake system in 1981 though sampling included times when peak spawning occurred in Lake Michigan (Appendix Table 1C). Thus presence of lake whitefish larvae in the marsh was only accidental. Burbot Burbot are one of the few freshwater fishes that spawn dxiring the winter (Coberly and Horrall 1980). Brazo and Liston (1979) observed that gonadal development of burbot near Ludington indicated spawning probably occurs from midDecember through January. Jude et al. (1979b) concluded burbot spawned in late December in southeastern Lake Michigan because they collected fish with ripe gonads in December and January. Spawning may occur in tributary streams (Mansfield et al. 1983), on rocky reef areas in 107 shallow water (Bailey 1972), or in deep water (Scott and Crossman 1973). Eggs will hatch in 30 days at water temperatures of 6 - 7 C, but take 70 days to incubate at 0 - 1.5 C (McCrimmon 1959). Thus larvae of burbot are generally only present in late winter and spring. Lake Michigan Temporal distribution. Distribution of burbot larvae was similar between 1978 and 1979. Larvae were present in mid-April samples at densities up to 453/1000 m 3 and water temperatures of 3 - 5 C. This period also was one of peak abundance, though occasional moderate densities (<1 0 / 1 0 0 0 m 3 ) of burbot larvae were observed through mid-June at temperatures up to 12 C (Table 42). Shortly after water temperatures increased substantially, no burbot larvae were collected for the remainder of the year. Burbot were probably present earlier than initial sampling, but the mean size (3.8 mm) and small range (3.1 - 4.4 mm) indicate that most larvae were newly hatched. Spatial distribution. Densities of burbot larvae generally were greatest at the shallowest contour (Table 42.), but low numbers of larvae were widely distributed at all contours. This may reflect a wide range in spawning habitat because nearly all burbot were newly hatched, but more likely, drift of pelagic larvae with water currents was responsible for the observed distribution. burbot were widely distributed among transects. Larval Transects Table 42. Average nighttime density and range (in parentheses) of burbot larvae (number/1000 irf5) __________ at several depth contours in nearshore Lake Michigan, 1978 - 1979. Dash means no data. Depth contour (m) Date 3 1 6 9 12 1978 22 April 61 (8 - 179) - - - - 3 May 68 (0 - 202 ) - - - - 15 May 35 (0 - 113) 6 (0 - 15) i (0 - 4) 1 (0 - 4) - 31 May 2 (0 - 10 ) 1 (0 - 1 (0 - 4) 1 (0 - 4) - 14 June 2 (0 - 12 ) 0 (0 - 453) 9 (0 - 27) 7 12 5) 0 - 0 1979 100 (0 - 27) 8 (0 - 19) b (0 - 18) (0 - 3 (0 - 0 1 May 8 (0 - 28) 4 (0 - 13) 15 May 3 (0 - 16) 2 (0 - 9) 0 0 29 May 7 (0 - 38) 7 (0 - 25) 0 1 12 June 0 2 (0 - 3 28 June 0 0 0 1 10 July 0 0 0 0 1 (0 - 31 July 0 0 2 0 0 15 Aug. 0 0 0 0 3 (0 - 9) 8) (0 - 12 ) (0 - 14) 8) <1 (0 (0 - 6) 0 3) 0 0 (0 - 9) 0 8) 9) 108 17 April 109 1 and 3 generally showed greatest densities during night tows, but transect 2 had greatest densities during the day. Reasons for this distribution are unknown, could not be surmised from the climatological data, and may simply be due to passive transport. Diel distribution. Densities of burbot larvae were substantial in night collections in 1979 (Table 43), but day tows were not initiated until 15 May which was considerably later than peak abundance. When comparable data were available, similar densities of larvae were taken in day and night tows from mid--May through raid-June. Observations of burbot larvae in a small aquarium indicated they were pelagic during the day. This suggests burbot larvae were present in the water column and unable to avoid gear at sizes collected. Total abundance. Total abundance of burbot was always less than one million larvae in the Lake Michigan reference wedge (Table 44), except during peak abundance at first sampling in 1979. Consistent decreases in abundance were recorded from April through mid-May. From mid-May to mid- June , 1979, the number of larvae in the reference wedge was low (generally <100 thousand larvae) but fairly uniform. After mid-June no larvae were observed. An increase in demersal behavior similar to adults and movement offshore to cooler waters is probably responsible for the absence of larvae. Table 43. Date 17 April 1 May 29 May 12 June __________________________ Depth contours (m)_________________________________ 1 3 6 9 .12 D N D N D N D N D N 100 (179) - - - — 9 (11 ) -- 7 (12) - (9) - -• 8 - 6 (8 ) 8 - 4 ■■ 2 - 3 - -- (11 ) - (6 ) ~ (3) - (4) - 4 3 r> 2 3 0 _) T> 0 (9) O) (6 ) (4) (5) - (4) - 2 (6 ) l (3) 7 (13) 3 7 (8 ) 2 0 4 1 0 0 (5) - (8 ) (2 ) - - (6 ) 0 <1 (1 ) 0 0 1 2 0 3 0 0 0 0 - - (2) (5) - (6) - - - - 110 15 May Comparison of average day - night densities of burbot larvae (number/1000 nf5) at several depth contours in nearshore Lake Michigan, 1979. One standard deviation in parentheses. D = day, N = night. Table 44. Estimates based on night densities and 90% confidence interval (+ ) of burbot abundance (thousands of larvae) in a 188.5 million m 3 reference wedge on the east-central shore of Lak e Michigan. 1978 22 April13 15 May 31 May D ate Number 53 58 (+ 52) 147 (+161) 93 17 April13 1 May^ 15 May 2 Nu m b e r 87 (+127) 7 (±6 ) 20 D ate Number 17 April 1294 (+1620) 1 May 238 (+321) 15 May 62 (+109) (±26) 29 May 92 (+119) 29 May 92 (+119) 12 June 58 (+137) 12 June 58 (+137) (+1 2 0 ) 14 June 1979" (N = 1601) 1979a (N = 12) D ate 3 Mayc N = nu m b e r of samples. (+3) aAt 1978 contours only. toAt all contours sampled in 1979. ^Estimates made only for 1-m contour. 112 Pere Marquette Marsh-Lake General distribution. No burbot larvae were collected in Pere Marquette Marsh-Lake during 1981, nor in the harbor outlet channel. Only a single immature fish was collected during routine netting in 1981 (Appendix Table 1C). These data tend to refute spawning in lotic waters by burbot in this area of Lake Michigan. However, yearling burbot were collected far upstream in tributaries to the Pere Marquette River, suggesting some spawning may occur much farther upstream. Most likely mature adults moved upstream after the end of sampling (e.g. winter ), spawned, and returned to Lake Michigan. Other Species An additional six taxa of larvae were collected in Lake Michigan: trout-perch, gizzard shad, crappie (Pomoxis spp.), sunfish (Lepomis spp.), bloater, and longnose dace. Of these, only the former four were also collected in Pere Marquette Marsh-Lake and will be discussed below. Adult trout-perch are an abundant forage fish in Lake Michigan (Brazo and Liston 1979), but few larvae were collected. Trout-perch larvae were collected from late May to late September during 1978 and 1979. They first appeared nearly a month earlier in 1978 than in 1979. Densities were generally <10/1000 m 3 with peak densities near 20 larvae/ 1 0 0 0 m 3 recorded sporadically throughout the 11.3 summer in Lake Michigan. All larvae were collected at the shallowest contours and nearly all (96%) were taken at night. All trout-perch larvae were of similar size (6 - 8 mm) and contained large amounts of yolk, suggesting a prolonged spawning period. This may explain the lower than expected densities of larvae. If a large number of larvae are spread over a long time period, the densities at any one time will be lower. Liston et al. (1980) also hypothesized that trout-perch may spawn at alternate locations (as tributary marshes) thus explaining low densities in Lake Michigan, hut the following data refute this theory. Trout-perch larvae were present in higher densities in the Pere Marquette Marsh in 1981 than observed in Lake Michigan in 1978 - 1979. Larvae occurred in the marsh from early May to mid-June at densities of 3 - 36/1000 m 3 (Appendix Table 6 B ). No larvae were collected in the lower marsh (station 3) nor in the harbor outlet (station 4) indicating there was probably no transport of larvae into or out of the marsh. This suggests discrete populations of trout-perch may exist in Pere Marquette Marsh-Lake and in Lake Michigan. Trawling data in 1981 substantiated this fact, as adult trout-perch were one of the most abundant fish taken in adult fish surveys in the marsh (Appendix Table 1C). Most gizzard shad larvae collected were prolarva and easily separable from alewife larvae. Gizzard shad larvae 114 were collected only on two dates (mid-May and mid-June) in 1979 and then in very low densities (<5/1000 m 3) in Lake Michigan. 3m). All larvae were taken at shallow contours (1 and In 1981, larvae of gizzard shad were collected in the Pere Marquette Marsh-Lake on all mid-July at densities dates from mid-June to of 8 - 565 larvae/1000 m 3 . Most larvae were collected at the lower open water station (3) of Pere Marquette Marsh-Lake and a substantial number of larvae were present in the harbor outlet (station Appendix Table 7B). 4, All gizzard shad larvae present in the outlet channel were taken in water flowing from the inarsh to Lake Michigan and an estimated 450,000 larvae were transported into Lake Michigan on dates when they were present in the outlet in 1981. This compares with 2000 - 54 ,700 larvae present in the reference wedge in Lake Michigan during 1978 - 1979. The data suggest transport of gizzard shad to Lake Michigan from these type of marshes may be sufficient to account for their presence there. Larvae of two centrarchids (Pomoxis spp. and Lepomis spp.) were occasionally found in Lake Michigan at low densities in 1979. Some Pomoxis spp. larvae were collected from late May to late June at mean densities of 1 - 3/1000 m3 . All larvae (except one) were taken at the 1-m contour and estimated total abundance in the reference wedge ranged from 963 to 14,157 larvae. Some Lepomis spp. larvae were collected on a single date in early July at densities of 3/1000 m 3. As with Pomoxis spp. larvae, these larvae were 115 all taken at the 1-m contour and total abundance was estimated at 2,650 in the reference wedge. The low densities of Pomoxis spp. and Lepomis spp. larvae suggest spawning may occur elsewhere, a fact corroborated by lack of suitable habitat for adults in this area of Lake Michigan. Lepomis spp. and Pomoxis spp. larvae were the thirdand sixth-most abundant larvae, respectively, collected from the Pere Marquette Marsh-Lake in 1981. Pomoxis spp. larvae were present as early as 20 May at densities of 4 22/1000 m 3 (Appendix Table 8B). Peak densities were not recorded until late June - early July when densities approached 290/1000 m 3 in samples taken from dense macrophyte beds at upper marsh stations (1 and 2). Densities of Pomoxis spp. larvae were almost restricted to dense submerged macrophyte beds later in the year (Appendix Table 8B). Larvae were present at most stations in the marsh during June, and substantial numbers of Pomoxis spp. larvae (12 - 34/1000 m 3) were recorded from the harbor outlet in mid- to late June. All larvae were present in waters flowing from the marsh and an estimated 210 thousand Pomoxis spp. larvae were transported into Lake Michigan on sample days. This estimate is markedly higher than estimated abundance in Lake Michigan suggesting all Pomoxis spp. larvae result from transport from the marsh to Lake Michigan. 116 Lepomis spp. larvae were present in samples from Pere Marquette Marsh-Lake from late June to early August, approximately a month later than Pomoxis spp. larvae. Initial densities were 52 larvae/1000 m 3^ and peaks of 1080 - 4062 larvae/1000 m 3 were observed in mid to late July (Appendix Table 9B). As with Pomoxis spp. larvae, these peak densities were recorded from dense submerged macrophyte beds. Lepomis spp. larvae were more restricted in distribution than Pomoxis spp. larvae and were found primarily at upper marsh stations, except for a single individual in the open water of Pere Marquette Lake. No Lepomis spp. were collected in the harbor outlet in 1981, suggesting no transport of these larvae out of Pere Marquette Marsh-Lake. However, the few Lepomis spp. larvae collected in Lake Michigan in 1979 probably originated from a tributary marsh. DISCUSSION Distribution and Abundance A seasonal succession of fish larvae is well documented for many bodies of water in both lentic and lotic habitats (Faber 1967; Wells 1973; Amundrud et al. 1974; Jude 1977; Floyd et al. 1984) and obvious seasonal distributions existed in Lake Michigan near Ludington. During this study it became apparent that physical factors such as water temperature, water current, wind direction, and wind speed played a major role in determining spatial and temporal distribution of ichthyoplankton in Lake Michigan. Brattstrom (1968) discusses different mechanisms that induce water currents and concludes that currents have a profound effect on the biology and distribution of aquatic organisms. A variety of physical conditions occurred during this study that may have produced noticeable annual variations in ichthyoplankton density, abundance, and growth. Physical conditions in 1978 produced a "warm" year in Lake Michigan. A relatively rapid spring warming was followed by a slight upwelling and then a very warm period in late summer. In contrast, water warmed slowly in spring 1979 which was accompanied by depressed temperatures until late July to early August due to prolonged upwelling. In addition to physical effects, 117 118 passive transport of fish larvae by wind-induced currents may affect distribution. Distribution of larvae in Lake Michigan was decidedly different between 1978 and 1979. Annual differences were not as great for larvae present in the spring (up to midMay) as for larvae present later in the year because spring temperatures in both years were cold. Deepwater sculpin, lake whitefish, and burbot larvae were abundant in spring of both years. No trends in densities and total abundance were found for these species although analyses of early spring, 1978, were hampered by lack of collections. Faber (1970) and Reckahn (1970) both observed similar spring cohorts in Lake Huron. Mansfield et al. (1983) reported larval burbot and deepwater sculpin were abundant in Lake Michigan from March through May. The only primary difference in the distribution of the three spring cohorts was the presence of these larvae (especially lake whitefish and deepwater sculpin) for longer periods in 1979 than 1978. These coldwater species either preferred the cold water from prolonged upwellings in 1979 or were passively transported with upwelled water, which most likely accounted for the difference. By late May, 1978, water temperature warmed rather suddenly to 13 C and these larvae were absent thereafter. None of the "cold water" larvae showed any input from the Pere Marquette Marsh-Lake to Lake Michigan. Lack of suitable spawning habitat and warm water temperatures in 119 the marsh undoubtedly influenced this. However, the presence of deepwater sculpin larvae in the nearshore zone was surprising as these fish are thought to spawn in very deep water (70 m and greater; Scott and Crossman. 1973 ). Kahn and Faber (1974) report that after hatching, deepwater sculpin larvae become limnetic and widely dispersed. In oceanic surveys off the coast of California, Eldridge (1977) was surprised to find larvae spawned offshore present in a coastal embayment. He attributed this distribution to movement with upwelled water. Also, Smith et al. (1981) and Richardson and Pearcy (1977) noticed species of fish larvae normally found offshore being swept shoreward by current flowing in the opposite direction that the Eckman spiral would predict. Similar phenomenon occur in Lake Michigan, although other parameters such as spring overturn may also produce movement of deep water shoreward. Larvae present in late spring began to show the effects of the cold year in 1979. Densities of yellow perch and rainbow smelt never attained peaks observed in 1978. Water temperatures in late May, 1978, rose rapidly to 13 C which may have promoted a strong hatch of rainbow smelt in 1978; whereas, in May 1979 water temperatures did not exhibit a rapid increase and remained around 7 C. Owens (1982) reported rainbow smelt eggs took 59 days to hatch at water temperatures of 2.9-5.9 C . Correspondingly, 1979 densities of rainbow smelt larvae remained at least 50% lower than densities in 1978 . Tin 120 and Jude (1983) also observed lower densities of rainbow smelt larvae in 1979 than in other years from 1978-1980 in southeastern Lake Michigan. They attributed the lower densities to sample times and not poor reproduction but their data coupled with the present study would indicate climatic conditions in 1979 may have in fact affected rainbow smelt reproduction. Distribution and abundance of yellow perch larvae followed similar patterns of lower densities in 1979 than in 197 8. Even the early peak of larvae, which were transported from local tributary marshes, was considerably lower in 1979 than 1978. This suggests that colder water may have a strong and varied effect. Development and hatching is prolonged, but warm water tends to increase developmental rate and uniformity in size such that very high density peaks with rapid attenuation are observed. Greatest effects of upwelling currents and colder water were observed on larvae present in the summer: Cottus spp., johnny darter, ninespine stickleback, spottail shiner, and alewife. Upwelled water can reduce local water temperatures by as much as 10 - 15 C in less than 24 hours in the summer (Liston and Tack 1974). Besides lowered water temperatures, which would retard growth and development, many larvae may be swept offshore to less desirable habitat. Alewife larvae were present 2 weeks earlier in 1978 and reached densities of 1000 - 5000 larvae/1000 m 3 at nearly all contours from early July 121 through late August. In contrast, in 1979, larvae were not collected until 2 weeks later and peak densities (which averaged only 292 - 623 larvae/1000 m 3 at discrete contours) did not occur until late August. Substantial numbers of alewife larvae were still present in late September 1979. The effect of prolonged upwelling on alewife larvae was quite pronounced by reduced densities and abundance in 1979. This would suggest lower reproductive success possibly due to increased mortality. Heufelder et al. (1982) observed similar effects on alewife distribution in southeastern Lake Michigan from 1977 - 1979. They postulated mortality from thermal shock and passive drift of larvae as mechanisms responsible for reduced alewife density. Hunter (1981) also suggests that one of the main causes of larval fish mortality is the deportation of larvae in large masses of water to areas that are poor in nutrients and food, or the importation of water with temperatures unsuitable for growth. Edsall (1970) noted that jaw apparatus of alewife larvae showed abnormal development at temperatures less than 10 C which suggests negative impacts on survival during periods of prolonged upwelling. The impact of upwelling in early summer on alewife spawning success may be offset somewhat by the prolonged spawning period of alewife which resulted in increased densities of larvae late into September of 1979, a phenomenon not observed in 1978. 122 The rapid increase in water temperature in late May, 1978, triggered hatching of several species of larvae, most notably alewife, Cottus spp. and ninespine stickleback, all of which were not collected until 2 weeks later in 1979. The cold summer of 1979 had other dramatic effects on johnny darter, ninespine stickleback, and spottail shiner larvae. Peak densities of ninespine stickleback larvae occurred earlier in 1978 than 1979, but total abundance was similar between 1978 and 1979. Apparently the relatively cool water in 1979 was more conducive to ninespine stickleback reproductive success than the warmer water in the previous year. This species seems to prefer colder water and rarely was trawled in the study area after June when water temperatures typically rise to 15 - 22 C (Brazo and Liston 1979). Spottail shiner larvae, though exhibiting prolonged occurrence in both years, were markedly more abundant and more widely dispersed to deeper contours in 1978 than in 1979. Again, the negative effect of colder water is implicated. Jude et al. (1980) provide corroborating evidence that spottail shiner tend to select the warmest water available. Finally, cold upwelled water impacted johnny darter larvae similarly. In 1978 johnny darter exhibited an earlier and larger single peak of abundance than in 1979 when the peak was later, much lower, and bimodal. After the initial peak in 1979, a strong upwelling occurred in mid-August which apparently delayed johnny darter spawning and hatching until water 123 temperatures warmed again in late August. Swee and McCrimmon (1966) noted interruption in common carp spawning when water temperatures were lowered by climatic conditions. Some johnny darter larvae were present into late September, substantially prolonging their presence in this area of Lake Michigan in 1979 compared to 1978. The above examples and discussion point to the important effects of density-independent factors, such as water current and temperature on distribution and abundance of ichthyoplankton. Though Lake Michigan water currents were not measured in the present study, Liston and Tack (1976) provide detailed data on the effects of wind direction on water current and temperature in Lake Michigan near Grand Haven, only 93 km south of the present study site. Many factors play key roles in determination of year class strength and distribution [e.g., food availability (Lasker 1981), predation (Hartig et al. 1982), substrate types (Dorr 1982)], but the importance of physical factors should not be underestimated. Effects may be directly on distributional patterns of larvae, or they may indirectly affect food sources and preferred habitat. Further, though effects of upwellings may be deleterious for some species they may also be advantageous for others, as witnessed by the success of ninespine stickleback larvae in 1979. Major physical phenomena must overlap with larval fish presence to have pronounced results. For instance, no demonstrable effect of the cold 1979 year could be shown for larvae 124 present in the spring. At this time of year all water is cold and though strong north or east winds may produce upwellings, the effect is not as dramatic as presence of cold upwelled water in mid-summer. such as: A variety of effects reduced peak densities and abundance (Heufelder et al. 1982), retardation of growth (Taniguchi 1981), interruption of spawning and hatching (Swee and McCrimmon 1966), and prolonged presence of larvae may result. Species of fish with extended spawning seasons are better suited to coping with prolonged upwelling because some larvae may hatch during more favorable conditions as observed with alewife and johnny darter in 1979. Diel distribution. The present study also provided some insight to diel analyses in larval fish distribution though this was not one of the specific objectives. Most species of larval fish were collected in greater numbers at night than during the day. Faber (1970) reported newly hatched lake whitefish larvae exhibited positive phototactic response and personal observations in the laboratory during the current study indicated lake whitefish larvae were active in the water column during the day. Hunter (1981) suggested clupeid larvae are less capable of avoiding predators (collection) at night because of the lack of development of the scotopic visual system, a phenomenon also recorded by O' Connell (1981). Water clarity may also influence catches of larvae. Netsch et al. (1971) observed little difference between diel 125 densities of clupeid larvae taken in a turbid area of a Kansas reservoir. In the present study, unexpected large numbers of large rainbow smelt were taken on days when wind and wave action created highly turbid conditions at very shallow contours. One exception to the decreased catch of larval fish during the day was yellow perch. Densities were consistently similar or slightly greater in day catches compared with night catches. This differs somewhat from data reported by Wong (1972) and Perrone et al. (1983). Wong (1972) reported yellow perch exhibited a diel vertical migration and they were taken at greater densities at dusk than during lowest light intensities of late night. He attributed this to partial night blindness and poor scotopic vision. Perrone et al. (1983) observed greater catches of yellow perch larvae at night than in the day in shallower contours and felt net avoidance was responsible. Quite possibly percid larvae have a well developed scotopic visual system in early life unlike alewife as suggested by Hunter (1981) and are therefore more dispersed in the water column at night making them as susceptible to catch. Further, Privolnev (1956) indicated yellow perch larvae are positive phototactic which explains their presence higher in the water column and makes them more susceptible to capture. 126 Importance of Inputs From Pere Marquette Marsh-Lake Though the importance of tributary marshes in supporting spawning populations of migrating fish species from Lake Michigan has been implied by Great Lakes researchers and managers, the present research suggests that at least for Pere Marquette Marsh and Lake system the inputs for most of the species studied is of small consequence. Only eight taxa of larvae were collected in the outlet channel in 1981 and of these, only five (alewife, yellow perch, common carp, gizzard shad, Pomoxis spp.) actually spawned in Pere Marquette Marsh and provided some input to Lake Michigan (Appendix Tables IB - 10B). Mansfield (1984) indicated substantial inputs of spottail shiner larvae to Lake Michigan occurred at Little Pigeon Creek 104 km south of the present study area, but little drift of other larvae to Lake Michigan seemingly occurred. Alewife were the most abundant larvae collected in the outlet channel, but many of the larvae were collected in water flowing from Lake Michigan into the marsh. The density of alewife larvae in Pere Marquette Marsh-Lake was considerably lower than anticipated. Extensive spawning of adult alewife in the marsh and the presence of what were assumed to be many alewife eggs in collections in mid late June were observed. This suggests that high egg or hatching mortality may be occurring. The larvae of another clupeid, gizzard shad, reached much higher densities than did alewife larvae though few spawning adults were 127 captured. Reasons for this discrepancy are not clear but are most likely related to habitat preferences, larval fish morphology, and type of eggs. Spawning habitat of gizzard shad is typically over rocks and other submerged obstacles where the demersal eggs are readily adhesive and will stick to anything (Miller 1960). On the other hand, alewife tend to spawn pelagically and the eggs are broadcast at random. Though they tend to be demersal they are not adhesive (Graham 1956), and may drift to Lake Michigan before hatching (Mansfield. 1984) . In submerged river mouths and marsh habitats, the advantage of an adhesive egg is clear. Adhesive eggs may stick on submerged objects off thebottom and would be kept clear of debrisby currents, whereas demersal eggs that are non-adhesive could be transported to areas with high siltation rates and probably low dissolved oxygen. An anatomical feature that tends to favor gizzard shad over alewife larvae in submerged river mouths and marshes is the presence of an oil globule in gizzard shad that is lacking in alewife larvae (Lippson and Moran1974). An oil globule would lend more buoyancy to gizzard shad larvae which may be critical in high turbidity environments such as marshes. This would allow larvae to more easily remain higher in the water column where increased light may enhance feeding. Large numbers of larval yellow perch were transported from Pere Marquette Marsh-Lake into Lake Michigan. Yellow perch larvae often exhibited relatively high densities in 128 shallow contours of Lake Michigan prior to spawning of adult yellow perch in Lake Michigan. The following evidence further corroborates this hypothesis. Peak spawning of yellow perch in Lake Michigan occurs around the first of June at 6 - 8 C (Brazo 1973), and hatching time should be 10 - 24 days at those temperatures (Scott and Crossman 1973). Yellow perch larvae from Lake Michigan spawning then should not appear until late June. Also, these "late" larvae were most abundant at sites that Brazo and Liston (1979) suggested as prime spawning areas based on gill net collections. Presence of a bimodal peak and the differential spatial distribution of larval yellow perch have been reported elsewhere in the lake (Wells 1973; Jude et al. 1979a; and others). Results from this study show "early" larvae are the result of yellow perch spawning in inland lakes with tributaries to Lake Michigan, hence the alongshore distribution, and "later" larvae result from Lake Michigan spawning. Perrone et al. (1983) and Dorr (1982) also observed similar distribution of yellow perch larvae in southeastern Lake Michigan. Gizzard shad and Pomoxis spp. larvae were rarely collected in Lake Michigan although they were the outflow from the marsh. present in Jude et al. (1980) reported similar low densities of Pomoxis spp. larvae 97 km south in Lake Michigan and suggested these larvae came out of a local lake. These fish probably do not spawn in Lake 129 Michigan, but are more likely found in connecting tributaries and marshes where more preferred spawning habitat exists (Tin 1982). Thus, survival of transported larvae must be considered. Survivorship of the taxa may differ considerably. Yellow perch larvae become rapidly able to avoid ichthyoplankton gear which may be assumed from the small length range of larvae collected and is well documented by Perrone et al. (1983), Forney (1971), Noble (1970) , Wong (1972), and others. Thus yellow perch larvae transported from tributaries to Lake Michigan may be quite robust and have a higher survival rate. On the other hand, gizzard shad, alewife, and Pomoxis spp. larvae tend to be more vulnerable to ichthyoplankton gear over a longerperiod of time suggesting they do not develop as rapidly and are less robust. These larvae are transported from warm marsh temperatures to considerably colder waters of Lake Michigan which may adversely affect survival. Coutant et al. (1974) and Wolters and Coutant (1976) observed cold shock decreased the ability of bluegill and channel catfish to avoid predation, although temperatures had to differ by more than 6 C to show effects. Also, the number and kinds of zooplankton may differ markedly between the two environments which could also contribute to mortality (Hunter 1981) though specific data are lacking in the present study. In late summer and fall, relatively large numbers of Pomoxis spp. and gizzard shad young-of-the-year (YOY) were 130 trawled periodically in Lake Michigan (Brazo and Liston 1979; Liston et al. 1980). Low numbers of larvae of these two species were transported from the marsh to Lake Michigan so larval fish survival would have to be extraordinarily high to produce the numbers of YOY collected in Lake Michigan. This dichotomy may be resolved by continuing to assume little transport of the larvae of these species occurs, but realizing gizzard shad and Pomoxis spp. may move from the tributary marsh as larger juveniles later in the year. This same phenomenon may be important for other species. If important inputs of fish from tributary marshes to Lake Michigan occur it could be as active transport (swimming) of YOY rather than larvae, though supporting data need to be gathered. Jude et al. (1981) and Tesar et al. (1986) found large numbers of YOY gizzard shad in the fall in nearshore Lake Michigan, near Grand Haven and Bridgeman, Michigan. They suggest these fish were abundant in tributary river mouths and move into Lake Michigan in the fall. For Pere Marquette Marsh-Lake in 1981, inputs of larvae to Lake Michigan were probably of small consequence except in the case of yellow perch. Pere Marquette Marsh- Lake supports a large percentage of "rough" fish with common carp, Catostomus spp., alewife, and gizzard shad comprising a large portion of the ichthyoplankton. Though the use of the marsh as a spawning and nursery site for resident sport fish (e.g., smallmouth bass, northern pike, 131 Pomoxis spp., Lepomis spp.) is potentially very important it was beyond the scope of the present study. Important sport and forage species from Lake Michigan characteristically do not use the Pere Marquette Marsh-Lake to a great extent but this does not preclude the fact that other tributary marshes may have important inputs. Truly important coastal marshes for Lake Michigan spawners may be those marshes immediately on the shoreline and bays (e.g., Bay de Noc) where water transfer is more direct. SUMMARY 1. Main objectives of this study were: 1) to study spatial and temporal distribution of larval fish in Lake Michigan; 2) estimate local abundance of larvae; and 3) assess transport of larvae from a local tributary (Pere Marquette Marsh-Lake) to Lake Michigan. 2. Larval fish were sampled in Lake Michigan in 1978, 1979, and 1981. Samples were taken during the day and night in 1979, but only at night in 1978 and 1981. Samples were taken approximately once every 2 weeks from mid-April through September with conical plankton nets of l.-m diameter mouth and 351 y.-mesh aperture. Four transects were established in Lake Michigan and collections were made at 1-, 3-, 6-, 9-, and 12-m contours along each transect. Total abundance of larvae was estimated for a reference wedge in Lake Michigan with surface rectangular dimensions of 2.4 x 9.7 km and volume of 188.5 million m 3 . 3. Day and night larval fish samples were taken at four stations in Pere Marquette Marsh-Lake in 1981. The number of larvae being transported into and out of the marsh-lake were estimated by measuring velocity and 132 discharge of the Pere Marquette River to Lake Michigan and multiplying by estimated densities of larvae. Overall 1206 samples were taken which contained 41,602 larvae. Dominant larvae in Lake Michigan samples were alewife, yellow perch, rainbow smelt, and johnny darter. Pere Marquette Marsh-Lake samples were dominated by common carp, Pomoxis spp., Lepomis spp., yellow perch, and clupeid larvae. Alewife were the most abundant larvae in Lake Michigan in 1978 and 1979 and comprised two-thirds of all larvae collected. Alewife larvae were present from mid to late June through the end of sampling in September October. Peak densities were as high as 9316/1000 m3 in 1978. Larvae were first collected 2 weeks later in 1979 than in 1978 and reached peak densities in late August 1979 compared to late July - early August in 1978. These differences may be attributed to differences in climatological conditions between 1978 and 1979. Wind direction and speed affected water currents and ultimately water temperature. Initially, alewife larvae were more concentrated at shallow contours (1 and 3 m) but later became more dispersed to deeper contours. Night tows generally produced more alewife larvae than did day tows but these data were confounded somewhat during periods of high turbidity. Local abundance of alewife in the reference wedge in Lake Michigan for different sampling dates was estimated as 47,000 to 160 million larvae in 1978, and 52,000 to 31 million in 1979. In 1981, alewife larvae were not abundant in Pere Marquette Marsh-Lake where densities were generally <100/1000 m 3. However, they did appear nearly 2 weeks earlier than first recorded in Lake Michigan in 1981. There was considerable transport of larvae into and out of Pere Marquette Marsh-Lake but results were confounded somewhat by the complex outlet currents. An estimated 4 million larvae may have been transported to Lake Michigan, compared with slightly less than 0.5 million larvae carried into the marsh-lake from Lake Michigan. Rainbow smelt were the second-most abundant larvae collected in Lake Michigan during 1978 and 1979. Larvae were first collected in mid-May of each year at low densities (2 - 48/1000 m 3 ). Peak densities (697 - 2302 larvae/1000 m 3) were attained by late May, followed by a rapid decline. Though temporal distribution was similar for the 2 years, magnitude of densities as considerably lower in 1979 than 1978. During early hatching, rainbow smelt larvae were markedly more abundant at shallow contours, but as the season progressed they became more uniformly distributed among depth contours. Examination of length data suggests that as larval rainbow smelt grew they moved offshore. Diel differences in larval rainbow smelt distribution were not observed for newly hatched larvae, but were evident for larger larvae. This phenomenon may be attributed to gear avoidance. Peak local abundance of larval rainbow smelt in the reference wedge in Lake Michigan was considerably higher in 1978 (26.5 million fish) than in 1979 (7.3 million fish) which again may be due to colder water temperatures and upwellings in 1979. Larval and YOY rainbow smelt were found in low densities (up to 42 larvae/1000 m 3) in Pere Marquette Marsh-Lake. However, during survey collection of adult and juvenile fish during April - November in Pere Marquette Marsh-Lake, no rainbow smelt were collected. Implications are that rainbow smelt larvae were transported into the marshlake from Lake Michigan. This is corroborated by presence of larvae (up to 60/1000 m 3 ) in reverse flows to the marsh from Lake Michigan. Yellow perch were the third-most abundant larvae collected in Lake Michigan during 1978 - 1979 and were cohorts of rainbow smelt larvae. Yellow perch larvae were first collected in mid-May of 1978 and 1979, with peak densities of 2120 and 288/1000 m3 , respectively. Densities declined rapidly after these early peaks but a second peak occurred in late June - early July at 556 and 194 larvae/1000 ra3 in 1978 and 1979, respectively. Presence of a bimodal peak was complicated by a difference in distribution. Greatest abundance of yellow perch larvae during peaks in mid-May occurred at 136 the 1-m contour, but distribution during the second peak was shifted to deeper contours (6 - 9 ra) . Length data for both peaks indicated most larvae were recently hatched. Evidence suggests that "early" larvae were not hatched in Lake Michigan but came from other sources, and "late" larvae were the result of Lake Michigan spawning. Yellow perch larvae were collected in equal numbers during the day and night. Local abundance in the reference wedge in Lake Michigan peaked in late June in both years at 7.9 and 3.3 million larvae in 1978 and 1979, respectively. Larval yellow perch were the second-most abundant fish larvae captured in the Pere Marquette Marsh-Lake, and were present from late April until early June. Densities of 1188 larvae/1000 m 3 were recorded in the upper marsh in early May. Substantial numbers (8 - 66/1000 m3 ) of larval yellow perch were collected in the outlet and an estimated 0.75 million yellow perch may have been transported into Lake Michigan from the Pere Marquette Marsh-Lake during 1981, suggesting a source for "early" larvae collected at the 1-m contour in Lake Michigan in mid - late May. 8. Johnny darter larvae comprised 5.9% of the Lake Michigan catch during 1978-79. Larval johnny darters first appeared in late June in both years at the shallow contours, but annual distribution was dissimilar thereafter. In 1978, greatest densities 137 (1741 larvae/1000 m 3) were observed in late June, though some densities >1000/1000 m 3 were recorded through mid-July, after which only a few larvae were observed. By contrast, densities of larval johnny darters in 1979 did not peak until late July at 950/1000 m 3 . A second peak was observed in late August and some larvae were present through September. These bimodal peaks in densities was undoubtedly due to strong upwellings in early August 1979 that reduced water temperatures drastically and may have interrupted spawning. Johnny darter larvae were not uniformly distributed in the study area but tended to be more abundant at those transects in Lake Michigan where bottom substrates were more irregular. The majority (98.7%) of larvae were captured at night which may be related to demersal behavior in the daytime. Local abundance in the reference wedge in Lake Michigan was 21.1 million larvae in early July, 1978, and 19 million larvae in late July, 1979. Johnny darter larvae were not collected in high numbers in the Pere Marquette Marsh-Lake, and first appeared at upper marsh stations at densities of 13/1000 m3 . Densities of larvae in Pere Marquette Marsh-Lake did not exceed 60/1000 m3 , though slightly higher densities (77/1O0Q m3 ) were recorded in the outlet. However, these values may represent some input from Lake Michigan. 138 9. Cottus spp. larvae first appeared in low densities (0 12/1000 m 3) in Lake Michigan samples in mid-June 1978 and 1979. Densities peaked and declined rapidly, reaching 140 larvae/1000 m 3 in late June 1978 and 108 larvae/1000 m 3 in late July, 1979. Cottus spp. larvae were not evenly distributed in the area, but like johnny darter larvae were concentrated in similar areas of irregular substrates. Larval sculpins tended to be present in markedly higher numbers at deeper contours (6 - 12 m) than at shallower contours. Nearly all Cottus spp. larvae were collected at night which may be attributed to gear avoidance and demersal behavior. Local abundance in the reference wedge in Lake Michigan peaked at 1.5 - 2.5 million larvae each year. No Cottus spp. larvae nor adults were found in the Pere Marquette Marsh-Lake. 10. Larvae of ninespine stickleback first occurred in ichthyoplankton samples in June of both years, but were delayed 2 weeks in 1979 compared with 1978. Peak abundance was observed in late June - early July 1978 at 492 larvae/1000 m 3, and in late July, 1979 at 680 larvae/1000 m 3 . Hatching began later and was more prolonged in 1979 than 1978, and success seemed to be greater in 1979, which may be the result of colder temperatures. Ninespine stickleback larvae were unequally distributed in Lake Michigan and increased in abundance with increasing depth out to the 12-m 139 contour. Few larvae were collected during the day and observations in aquaria indicated they were demersal during daylight which would decrease vulnerability to gear. Local abundance in the reference wedge in Lake Michigan was estimated as greater than 1 million larvae on most dates in 1979 with a peak of 6.5 million in late July. In 1978, abundance reached 7.5 million larvae in early July, but was markedly reduced thereafter. No ninespine stickleback larvae nor adults were collected in the Pere Marquette Marsh-Lake. 11. Spottail shiner larvae were collected from late May through September in both years which suggests an extended spawning period. Larval fish densities reached nearly 400/1000 m 3 in late June - early July, 1978-1979. Spottail shiner larvae were primarily restricted to the 1- and 3-m contours which may be an effect of preferred temperature selection. Nearly all larvae were collected at night due to demersal behavior during the day. Local abundance in reference wedge in Lake Michigan varied considerably from 876,000 fish at peak densities in 1978 to 208,000 larvae at peak in 1979. Spottail shiner larvae were collected in the Pere Marquette Marsh-Lake more than a month before they occurred in Lake Michigan though densities were never high ( 6 - 2 2 larvae/1000 m 3) in the marsh. No transport of larvae from or into the Pere Marquette Marsh-Lake was recorded. 140 12. The reproductive biology of deepwater sculpin is poorly understood but larvae are known to be limnetic and become widely distributed in Lake Michigan. Deepwater sculpin occurred as early as mid-April in 1978-79 at low densities (0 - 13/1000 m 3). However, length information suggests hatching was prolonged and could have been much earlier. were observed in early May. Peak densities of larvae Larvae were never collected at water temperatures >10 C, but were collected throughout the summer of 1979 during upwellings. Abundance of deepwater sculpin larvae increased with increasing water depth. The percentage of deepwater sculpin larvae taken at night was high (79%) but was considerably lower than the percentage of other cottid larvae taken at night. Local abundance of deepwater sculpin in the reference wedge in Lake Michigan was 0.5 - 1.2 million fish. No larvae or adults of deepwater sculpin were collected in Pere Marquette Marsh-Lake during 1981. 13. Lake whitefish larvae were spring cohorts of deepwater sculpin and burbot, and occurred in the first samples collected in mid-April 1978 - 79 at densities of 1 - 40 larvae/1000 m 3 . However, length data and presence of large amounts of yolk in larvae suggest hatching had only recently occurred. Peak densities of lake whitefish were observed in early May each year (around 300 larvae/1000 m 3 ), and were followed by precipitous 141 declines. Nearly all lake whitefish larvae were taken in water <4 m, and densities were an order of magnitude higher at the 1-m contour than at deeper contours. Only a single lake whitefish larva was taken during day tows, though accurate descriptions of diel distributions were hampered by the late start of day sampling (15 May) in 1979. Local abundance in the Lake Michigan reference wedge ranged from 2300 to 348,000 fish in 1979. No adult lake whitefish were collected in Pere Marquette Marsh-Lake during 1981, though a single larva was recorded from the lower open water station and another larva was collected in the outlet channel during reverse flows from Lake Michigan. 14. Burbot larvae were present in mid-April samples at peak densities of 453/1000 m 3 . Length data suggest hatching had only recently occurred at this time. Burbot larvae were present in low densities (<10/1000 m 3) through late May - early June after which none were collected. Greatest densities of larval burbot existed at the shallowest contours, but some larvae were found at all contours during day and night, though diel comparisons could not be made until well after peak densities because no day samples were collected until 15 May 1979. Local abundance in the reference wedge in Lake Michigan was generally less than a million larvae. No burbot larvae were collected in Pere Marquette Marsh-Lake. 142 15. An additional four taxa of larvae were collected in both Lake Michigan and Pere Marquette Marsh-Lake: trout-perch, gizzard shad, Pomoxis spp., and Lepomis spp. Trout-perch larvae occurred in low densities in both environments (10 - 35/1000 m3 ) and were slightly more abundant in Pere Marquette Marsh-Lake. However, no transport of larvae between the two environments was observed. Gizzard shad larvae were abundant in Pere Marquette Marsh-Lake and reached peak densities of 565/1000 m 3 in mid-July. An estimated 0.5 million larvae were transported to Lake Michigan from the marsh which supports the hypothesis that low densities of gizzard shad observed in Lake Michigan probably originated from tributary marshes and rivers. Pomoxis spp. and Lepomis spp. larvae were also abundant in Pere Marquette Marsh-Lake and reached peak densities of 290 and 4062/1000 m 3, respectively in July. Pomoxis spp. larvae were transported into Lake Michigan, but no Lepomis spp. were observed in the outlet. These data suggest the presence of a few larvae of each of these taxa in Lake Michigan was due to transport from local tributary marshes. 16. This study points out the importance and interaction of water temperature, water current, and wind direction on distribution and abundance of ichthyoplankton in Lake Michigan. Effects of the relatively "cold" year (1979) were most obvious on larvae present in late 143 spring and summer and were manifested in lowered densities, delayed hatching, slowed growth, and prolonged occurrence for most species. Lowered densities may have been the result of increased mortality due to abnormal morphological development, alterations in food sources, and transport of larvae from the area. 17. Several key concepts can be derived from the present study. Though physical factors may have a deleterious effect on some larvae, they may be beneficial to other larvae. For major physical and climatological phenomena to have an effect, they must coincide with peak larval fish densities. Also, species of fish with an extended spawning period may cope better with timelimited, deleterious factors. 18. Overall, inputs of larvae from Pere Marquette MarshLake to Lake Michigan seem to be of little consequence, except in the case of yellow perch. Inputs from the marsh to Lake Michigan of 4 million alewife, 0.75 million yellow perch, 210 thousand Pomoxis spp., and 450 thousand gizzard shad were recorded in 1981. Utilization of the Pere Marquette Marsh-Lake as a spawning-nursery site by migratory fish from Lake Michigan seems to be negligible. Most important inputs of fish spawned in the marsh may occur in late summer and fall as young-of-the-year move from the marsh to Lake Michigan. This may be especially true for 144 alewife, gizzard shad, and Pomoxis spp. Moreover, Pere Marquette Marsh-Lake supports a much larger biomass and spawning site for "rough" fishes, e.g., common carp, catostomids, clupeids, bullheads, and bowfin. LIST OF REFERENCES Amundrud, J. R. , D. J. Faber, and A. Keast. 1974. Seasonal succession of free-swimming perciforin larvae in Lake Opinicon, Ontario. J. Fish. Res. Board Can. 31:1611-1665. Arnold, G. P. 1974. Rheotropism in fishes. Cam. Phil. Soc. 49: 515-576. Biol. Rev. Bailey, M. M. 1972. Age, growth, reproduction, and food of the burbot, Lota lota, (Linneaus), in southwestern Lake Superior. Trans. Amer. Fish. Soc. 101(4): 667674. Birkhead, W. S., C. R. Bennett, E. C. Pendleton, and B. J. Copeland. 1977. Nursery utilization of the Dutchman Creek estuary, N.C., 1971 - 1976. Report to Carolina Power and Light Co., Raleigh, N.C. Report No. 77-2. Bishai, H. M. 1960. The effect of water currents on the survival and distribution of fish larvae. J. Cons., Cons. Perm. Int. Expl. Mer 25: 134-136. Booke, H. E. 1970. Speciation parameters in coregonine fishes: I. Egg-size II. Karyotype. Pages 61-66 in C. C . Lindsey and C . S . Woods (eds.), Biology of coregonid fishes. Univ. Manitoba Press, Winnipeg, Manitoba. Brattstrom, H. 1968. The importance of water movements for biology and distribution of marine organisms. Second European Symposium on Marine Biology. Sarsia 34:9-12. Brazo, D. C. 1973. Fecundity, food habits, and certain allometric features of the yellow perch (Perea flavescens Mitchill) before operation of a pumped storage plant on Lake Michigan. M.S. Thesis, Michigan State University. 75 pp. 145 146 _____ , and C. R. Liston. 1979. The effects of five years of operation of the Ludington Pumped Storage Power Plant on the fishery resources of Lake Michigan (1972 - 1977). Mich. State Univ., Dept. Fish. & Wildl., 1977 Ann. Rep. to Consumers Power Co., Vol. II, No. 1 (Fish. Research). 406 pp. _____ , P.I. Tack, and C. R. Liston. 1975. Age, growth, and fecundity of yellow perch, Perea flavescens (Mitchill) in Lake Michigan near Ludington, Michigan. Trans. Amer. Fish. Soc. 104:726-730. Cima, G. , M. Cochran, and M. Kitchel. 1975., 1975 Annual interpretive report on fish eggs and larvae in Lake Michigan near Zion and Waukegan generating stations. Report to Commonwealth Edison Co., Chicago, 111. Nalco Environmental Sciences. 170 pp. Clady, M. D. .1976. Influence of temperature and wind on the survival of early stages of yellow perch, Perea flavescens. J. Fish. Res. Board Can. 33(9): 18871893. Coberly, C. E., and R. M. Harrall. 1980. Fish spawning grounds in Wisconsin waters of the Great Lakes. Univ. Wise. Sea Grant Inst., WIS-SG-80-235, Madison, Wisconsin. Copeland, B. J . 1965. Fauna of the Arkansas Pass Inlet, Texas. I . Emigration as shown by tide trap collections. Publ. Inst. Mar. Sci. Texas. 10:9-21. Dorr, J. A. III. 1982. Substrate and other environmental factors in reproduction of the yellow perch (Perea flavescens) Ph. D. Dissertation. Univ. Mich., Ann Arbor, Mich. 162 pp. without appendices. _____ , D. J . Jude, F. J . Tesar, and N. J . Thurber. 1976. Identification of larval fishes taken from the inshore waters of southeastern Lake Michigan near the Donald C . Cook Nuclear Plant, 1973 - 1975. Pages 61-82 in J . Boreman (ed.), Great Lakes fish egg and larvae identification. Nat. Power Plant Team. Off. Biol. Ser. U.S. Fish & Wildl. Serv. Ann Arbor, Mich. FWS/OBS-76/23. Eckman, V. W. 1905. On the influence of the earth's rotation on ocean currents. Arch. Math. Astro. Physics 2:1-55. Edsall, T. A. 1970. The effect of temperature on the rate of development and survival of alewife eggs and larvae. Trans. Amer. Fish. Soc. 99 :376-380. 147 Eldridge, M. B. 1977. Factors influencing distribution of fish eggs and larvae over eight 24-hr samplings in Richardson Bay, California. Calif. Fish and Game 63:101-116. Faber, D. 1967. Limnetic larval fish in northern Wisconsin lakes. J. Fish. Res. Board Can. 24:927937. „ 1970. Ecological observations on newly hatched lake whitefish in South Bay, Lake Huron. Pages 481-500 in C. C. Lindsey and C. S. Woods (eds.), Biology of coregonid fishes. Univ. of Manitoba Press, Winnipeg. Floyd, K. B., R. D. Hoyt, and S. Timbrook. 1984. Chronolgy of appearance and habitat partitioning by stream larval fishes. Trans. Amer. Fish. Soc. 113: 217-223. Forney, J. L. 1971. Development of dominant year claases in a perch population. Trans. Amer. Fish, Soc. 100: 734-749. Graham, J. J. 1956. Observations on the alewife, Pomolobus pseudoharengus (Wilson), in fresh water. Publ. 74, Ont. Fish. Res. Lab. 43 pp. Gulland, J. A. 1965. Survival of the youngest stages of fish, and its relation to year class strength. Spec. Publ. ICNAF 6:363-371. Hartig, J. H., D. J. Jude, and M. S. Evans. 1982. Cyclopoid predation on Lake Michigan fish larvae. Can. J. Fish. Aqua. Sci. 39: 1563-1568. Hempel, G. 1965. On the importance of larval survival for the population dynamics of marine food fish. Calif. Coop. Oceanic Fish. Invest., Rep. 10:13-23. Herdendorf, C. E., and S.M. Hartley, eds. 1981. Fish and wildlife resources of the Great Lakes coastal wetlands within the United States. Vol. five: Lake Michigan. U.S. Fish Wildl. Serv., Washington, D.C., FWS/OBS81/02-V5. 1040 pp. Heufelder, G. R , , and N. A. Auer. 1980. A comparison of larval Cottus bardi and Cottus coqnatus from southeastern Lake Michigan. Pages 58-69 in L. A. Fuiman (ed.), Proceedings of the Fourth Annual Larval Fish Conference. U.S. Fish and Wildl. Serv., FWS/OBS 80/43. 148 _____ , D. J. Jude, and F. J. Tesar. 1982. Effects of upwelling on local abundance and distribution of larval alewife (Alosa pseudoharengus) in eastern Lake Michigan. Can. J. Fish. Aqua. Sci. 39: 1531-1537. Hinrichs, M. A. 1979. A description and key of the eggs and larvae of five species of fish in the subfamily Coregoninae. M. S. Thesis, Univ. Wis.-Stevens Point, W I . 73 pp. Hjort, J. 1914. Fluctuations in the great fisheries of northern Europe viewed in light of biological research. Rapp. Proc.-Verh. Rev. Cons. Perm. Intern. Expl. Mer. 20:1-228. Hodson, R. G. 1979. Utilization of marsh habitats as primary nursery areas by young fish and shrimp, Cape Fear Estuary, North Carolina. Report 79-5 to Carolina Power and Light Co. , Raleigh, North Carolina. Hogue, J. J. , Jr., R. Wallus, and L. K. Kay. 1976. Preliminary guide to the identification of larval fishes in the Tennessee River. TVA Div. Forestry, Fish, and Wildl. Devel. 6 6 pp. Hokanson, K. E. F. 1978. Optimum culture requirements of the early life phases of the yellow perch. Pages 2440 in R. W. Soderburg (ed.), Perch fingerling production for aquaculture. Wise. Sea Grant Coll. Prog., Univ. Wise., Madison, W I . Rep. No. 421. Hunter, J . R. 1981. Feeding ecology and predation of marine fish larvae. Pages 33-77 in R. Lasker (ed.), Marine fish larvae: Morphology, ecology, and relation to fisheries. Washington Sea Grant, Univ. Wash., Seattle, WA. Hynes, H. B. N. 1970. The ecology of running waters. Liverpool University Press, Liverpool, England. 555 PP- Jude, D. J . 1977. Entrainment of fish larvae and eggs on the Great Lakes, with special reference to the D. C. Cook Nuclear Plant, southeastern Lake Michigan. Pages 177-199 in L. D. Jensen (ed.) , Third National Workshop on entrainment and impingement. , B. A. Bachen, G . R. Heufelder, H. T. Tin, M. H. Winnell, F. J. Tesar, and J . A. Dorr III. 1978. Adult and juvenile fish, ichthyoplankton and benthos populations in the vicinity of the J . H. Campbell Power Plant, eastern Lake Michigan, 1977. Spec. Rep. No. 65. Great Lakes Res. D i v ., Univ. Mich., Ann Arbor, Mich. 639 pp. 149 _____, G. R. Heufelder, N, A. Auer, H. T. Tin, S. A. Klinger, P. J. Schneeberger, C. P. Mandenjian, T. L. Rutecki, and G. G. Godun. 1980. Adult, juvenile and larval fish in the vicinity of the J. H. Campbell Power Plant, eastern Lake Michigan, 1979. Spec. Rep. No. 79. Great Lakes Res. Div., Univ. Mich., Ann Arbor, Mich. 607 pp. _____, G. R. Heufelder, H. T. Tin, N. A. Auer, S. A. Klinger, P. J. Schneeberger, T. L. Rutecki, C. P. Mandenjian, and P. J. Rago. 1979a. Adult, juvenile and larval fish in the vicinity of the J. H. Campbell Power Plant, eastern Lake Michigan, 1978. Spec. Rep. No. 73. Great Lakes Res. Div., Univ. Mich., Ann Arbor, Mich. 574 pp. _____, F .J. Tesar, J. C. Tomlinson, T. J. Miller, N. J. Thurber, G. G. Godun, and J. A. Dorr III. 1979b. Inshore Lake Michigan fish populations near the D. C. Cook Nuclear Power Plant during preoperational years 1973, 1974 . Spec. Rep., No. 71, Great Lakes Res. Div., Univ. Mich., Ann Arbor, Mich. 529 pp. _____, H. T„ Tin, G. R. Heufelder, P. J. Schneeberger, C . P. Madenijian, T. L. Rutecki, P. J . Mansfield, N. A. Auer, and C. E. Noguchi.1981. Adult, juvenile and larval fish populations in the vicinity of the J . H. Campbell Power Plant, eastern Lake Michigan, 19771980. Spec. Rep. No. 8 6 , Great Lakes Res. Div., Univ. Mich., Ann Arbor, Mich. 364 pp. Khan, N. Y . , and D. J. Faber. 1974. A comparison of larvae of the deepwater and fourhorn sculpin, Myoxocephalus quadricornis L. from North America. 1. Morphological development. Pages 703-712 in J. II. S. Blaxter (ed.), The early life history of fish. Springer-Verlag, New York. 765 pp. .nfiT. .IP. ..I,....11.1— .11^...— ■— II— — Lam, C. N. H., and J. C. Roff. 1977. A method for separating alewife Alosa pseudoharengus from gizzard shad, Dorosoma cepedianum larvae. J. Great Lakes Res. 3:313-316. Lasker, R. 1981. The role of a stable ocean in larval fish survival and subsequent recruitment. Pages 79-87 in R. Lasker (ed.), Marine fish larvae: Morphology, ecology, and relation to fisheries. Wash. Sea Grant, Univ. Wash., Seattle, WA. Lippson, A. J., and identification of the Potomac PPSP-MP-13:282 R. L. Moran. 1974. Manual for of early developmental stages of fishes River estuary. Maryland Dept. Nat. Res. PP. 150 Liston, C . R., D. Brazo, J. Bohr, R. Ligman, R. O ’Neal, and G. Peterson. 1980. Results of 1978 aquatic research at the Ludington Pumped Storage Plant on Lake Michigan including entrainment of fish and invertebrates, turbine mortalities, reservoir residence periods of salmonid fishes, netting and hydroacoustic surveys, and water currents. Mich. State Univ., Dept. Fish. & Wildl., 1978 Ann. Rep. to Consumers Power Co., Vol. I. 2 2 1 pp. , D. C. Brazo, and P. I Tack. 1976. A study of the effects of installing and operating a large pumped storage project on the shores of Lake Michigan near Ludington, Mich. Mich. State Univ., Dept. Fish, and Wildl., 1974 Ann. Rep. to Consumers Power Co., Physical Chemical aspects. Vol. II. 65 pp. , D. Brazo, R. O'Neal, J. Bohr, G. Peterson, and R. Ligman. 1981. Assessment of larval, juvenile and adult fish entrainment at the Ludington Pumped Storage Power Plant on Lake Michigan. Mich. State Univ., Dept. Fish. Wildl., 1980 Ann.Rep. to Consumers Power Co., Vol. I. 275 pp. _____ , and P. I. Tack. 1976. A study of water currents and temperatures in Lake Michigan adjacent to the James H. Campbell Power Plant at Port Sheldon, Michigan. Report submitted to Commonwealth Assoc, by Mich. State Univ., Dept. Fish and Wildl. 6 6 pp. Mansfield, P. J. 1984. Reproduction by Lake Michigan fishes in a tributary stream. Trans. Amer. Fish. Soc. 113:231-237. _____ , D. J. Jude, D. T. Michaud, D. C. Brazo, and J. Gulvas. 1983. Distribution and abundance of larval burbot and deepwater sculpin in Lake Michigan. Trans. Amer. Fish. Soc. 112:162-172. MacCallum, W. R., and M. A. Regier. 1970. Distribution of smelt, Osmerus mordax, and the smelt fishery in Lake Erie in the early 1960's . J . Fish. Res. Board Can. 27:1823-1846. May, R. C. 1974 Larval mortality in marine fishes and the critical period concept. Pages 3-19 in J . H. S. Blaxter (ed.), The early life history of fish. Springer-Verlag, New York. 765 pp. McCrimmon, H. R. 1959. Observations on spawning of burbot in Lake Simcoe, Ontario. J. Wildl. Mgt. 23:447-449. 151 Miller j. R. R. 1960. Systematics and biology of the gizzard shad (Dorosoma cepedianum) and related fishes. U.S. Fish Wildl. Serv., Fish. Bull. 173, Vol. 60:371392. Mortimer, C. H. 1965. Spectra of long surface waves and tides in Lake Michigan and at Green Bay, Wisconsin. 18th Conf. Great Lakes Res.:304-325. Nelson, D. D., and R. A. Cole. 1975. The distribution of larval fishes along the western shore of Lake Erie at Monroe, Michigan. Mich. State Univ., Inst, of Water Res., Tech. Rep. No. 34.4:66 pp. Netsch, H. F., G. M. Kersh, Jr., A. Houser, and R. V. KiIambi. 1971. Distribution of young gizzard shad in Beaver Reservoir. Amer. Fish. Soc. Spec. Publ. 8:95105. Nigro, A. A., and J . J . Ney. 1982. Reproduction and early life accommodations of land-locked alewives to a southern range extension. Trans. Amer. Fish. Soc. 111:559-569. Noble, R. L. 1970. Evaluation of the Miller high-speed sampler for yellow perch and walleye fry. J . Fish. Res. Board Can. 27:1033-1044. Norcross, B. L., and R. F. Shaw. 1984. Oceanic and estuarine transport of fish eggs and larvae: a review. Trans. Amer. Fish. Soc. 113:153-165. Norden, C . R. 1961. The identification of larval yellow perch, Perea flavescens, and walleye, Stizostedion vitreum. Copeia 1961:282-288. _________. 1967. Age, growth and fecundity of the alewife, Alosa pseudoharengus (Wilson), in Lake Michigan. Trans. Amer. Fish. Soc. 96:387-393. O'Connell, C. P. 1981. Development of organ systems in the northern anchovy, Engraulis mordax, and other teleosts. Amer. Zool. 21:429-446. Organ, W. L. , G. L. Towns, M. Walter, R. B. Pelletier, and D. A. Riege. 1979. Past and present known spawning grounds of fishes in the Michigan coastal waters of the Great Lakes. Mich. Dept. Nat. Res., Fish Div., Tech. Rep. 79-1. 502 pp. Owens, R. W. 1982. The ecology of spawning smelt in southeastern Lake Michigan. M. Sc. Thesis. Eastern Michigan University, Ypsilanti, Mich. 78 pp. 152 Perrone, M . , Jr., P. J. Schneeberger, and D. J. Jude. 1983. Distribution of larval yellow perch (Perea flavescens) in nearshore waters of southeastern Lake Michigan. J. Great Lakes Res. 9:517-522. Price, J. W. 1940. Time-temperature relations in the incubation of the whitefish, Coregonus cLupeaformis (Mitchill). J. Gen. Physiology. 23:449-468. Privolnev, T. I. 1956. The reaction of fish to light. Voprosy Ikhtiol. 1956:3-20. Reckahn, J. A. 1970. Ecology of young lake whitefish (Coregonus clupeaformis) in South Bay, Manitoulin Island, Lake Huron. Pages 437-460 in C. C. Lindsey and C„ S. Woods (eds.), Biology of coregonid fishes. Univ. Manitoba Press, Winnipeg. Richardson, S. L . , and W. G. Pearcy. 1977. Coastal and oceanic fish larvae in an area of upwelling off Yaquina Bay, Oregon. Fish. Bull. 75:125-144. Robins, C. R . , R. M. Bailey, C. E. Bond, J. R. Brooker, E. A. Lachner, R. N. Lea, and W. B. Scott. 1980. A list of common and scientific names of fish from the United States and Canada. Amer. Fish. Soc. Spec. Pub. No. 6 . 17 4 pp. Rupp, R. S. 1965. the American 168. Shore-spawning and survival of eggsof smelt. Trans. Amer. Fish. Soc. 94:160- Scott, W. B., and E. J . Crossman. 1973 . Freshwater f ishes of Canada. Bulletin 184. Fish. Res. Board Can. Ottawa, Can. 966 pp. Seelig, W. N ., and R. M. Sorensen. 1977. Hydraulics of Great Lakes inlets. U.S. Coast. Eng. Res. Center, Fort Belvoir, VA. Tech. Paper No. 77-8 . Shenker, J . M . , and J . M. Dean. 1979. The utilization of an intertidal salt marsh creek by larval and juvenile fishes: abundance, diversity and temporal variation. Estuaries 2:154-163. Siefert, R. E. 1972. First food of larval yellow perch, white sucker, bluegill, emerald shiner, and rainbow smelt. Trans. Amer. Fish. Soc. 101:219-225. Smith, W. G., D. McMillan, and A. Wells. 1981. Changes in northeastern ichthyoplankton populations may be weather-related. Coast. Ocean. Climat. News 3:32-33. 153 Sokal, R. R,, and F. J. Rohlf. 197 3. Introduction to biostatistics. W. H. Freeman and Co, San Francisco, CA. 368 pp. Swee, V. B . , and H. R. McCrimmon. 1966. Reproductive biology of the carp, Cyprinus carpio L., in Lake St. Lawrence, Ontario. Trans. Amer. Fish. Soc. 95:372388. Taniguchi, A. R. 1981. Survival and growth of larval spotted seatrout (Cynoscion nebulosus) larvae in relation to temperature, prey abundance and stocking densities. Rapp. Proc. Reun., Cons. Int. Expl. Mer 178:507-508. Tesar, F ., D. J. Jude, H. Tin, P. Mansfield. 1986. Ecology of fish in nearshore southeastern Lake Michigan, 1973-1982. Pages 353-399 in R. Rossman (ed.), Southeastern nearshore Lake Michigan: Impact of the Donald C. Cook Nuclear Plant. Univ. Mich. Great Lakes Res. Div. Pub.22. Tin, H. T. 1982. Family clupeidae, herrings. Pages 64-73 in N. A. Auer (ed.), Identification of larval fishes of the Great Lakes Basin with emphasis on the Lake Michigan drainage. Great Lakes Fish. Comm., Ann Arbor, MI, Spec. Pub. 82-3. 744 pp. _____, and D. J. Jude. 1983. Distribution and growth of larval rainbow smelt in eastern Lake Michigan, 19781981. Trans. Amer. Fish. Soc. 112:517-524. Toetz, D. W. 1966. The change from endogenous to exogenous sources of energy in bluegill sunfish larvae. Invest. Ind. Lakes Streams. 7:115-146. UNESCO. 1968. Zooplankton sampling. Sci. and Cult. Org. 174 pp. United Nations Ed. Van Vliet, W. H. 1964. An ecological study of Cottus cognatus Richardson in northern Saskatchewan. M. A. Thesis. Dept. Biol., Univ. Saskatchewan, Saskatoon, Saskatchewan. 155 pp. Waiford, L. A. 1938. Effects of currents on distribution and survival of the eggs and larvae of the haddock (Melanogrammus aeglefinus) on Georges Bank. Fish. Bull. 49:1-73. Weinstein, M. P. 1979. Shallow marsh habitats as primary nurseries for fish and shellfish in Cape Fear River, North Carolina. Fish. Bull. 77:339-357. 154 Wells, L. 1968. Seasonal depth distribution of fish in southeastern Lake Michigan. Fish. Bull. 67:1-15. _____ . 1973. Distribution of fish fry in nearshore waters of southeastern and east-central Lake Michigan, MayAugust, 1972. Administrative Report. Great Lakes Fish. Lab., U.S. Bur. of Sport Fish, and Wildl. Ann Arbor, 24 pp. Westin, L. 1969. The mode of fertilization, parental behavior and time of egg development in fourhorn sculpin, Myoxocephalus guadricornis (L). Rep. Inst. Freshw. Res. Drottningholm— 49:175-182. Winn, H. E. 1958. Comparative reproductive behavior and ecology of fourteen species of darters (PiscesPercidae). Ecol. Monographs 28:151-191. Wolters, W. R., and C. C . Coutant. 1976. The effect of cold shock on the vulnerability of young bluegill to predation. Pages 162-164 in G. W. Esch and R. W. McFarlane (eds.), Thermal ecology II. Tech. Info. Center, Washington, D. C. Wong, B. 1972. Growth, feeding and distribution of yellow perch fry, Perea flaviatitis flavescens (Mitchill) during their first summer in West Blue Lake. M.Sc. Thesis. Univ. Manitoba, Winnipeg, Manitoba, Canada. 92 pp. Wootton, R. J. 1976. The biology of sticklebacks. Academic Press, NY. 3 87 pp. APPENDICES APPENDIX A Water temperatures in Lake Michigan during 1978 - 1979 APPENDIX A. Table 1A. Night bottom temperatures (C) at several depth contours on larval fish sampling transects in Lake Michigan, 1978 - 1979. (S = shallow depth contours, 1 m; W f ^ Transect 9 _* s Date 1978 Aprii May May May .Tune June July July Aug. Aug. Sept. Oct. D 3 4 I J D s T _L D S D 4 7 7 15 9 6 6 7 14 9 7 15 17 17 19 12 22 21 19 19 14 18 12 8 8 1.4 6 13 9 -\ 7 O 7 _L .J 14 10 11 16 17 17 11 14 IS 22 19 19 12 12 19 8 11 13 NOT 16 SAMPLED 11 18 17 17 18 22 22 21 19 18 19 12 12 12 5 7 7 9 15 18 14 17 IN 1978 155 22. 3 15 31 14 27 10 24 7 21 21 18 1 'i' 1979 17 1 15 29 12 28 10 31 15 27 April May May May June June July July Aug. Aug. 25 Sept. 3 8 6 10 9 6 8 11 10 6 8 16 15 11 10 5 17 10 14 17 5 8 8 7 11 14 17 17 10 11 10 17 18 15 7 .17 14 5 5 6 5 9 10 9 9 7 8 10 8 11 12 12 10 12 7 19 11 8 11 6 10 7 16 7 16 14 15 14 14 14 14 10 11 11 7 17 15 8 8 9 7 16 14 APPENDIX B Densities of ichthyoplankton in Pere Marquette Marsh-Lake 156 APPENDIX B. Tables IB. Densities (number/1000 m 3) of alewife larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night , D = day. i o 3 H N D 2 * 3 June 1 ! ■'3' Date ii* Station Lake Michigan 24 0 0 0 0 0 31 0 0 0 0 - 11 June N 5 0 0 0 15 1.7 June N D 4 0 33 78 56 146 207 25 308 90 24 June N 58 37 57 308 - 1 July N D 164 9 29 18 89 34 104 38 131 23 194 - 13 July N D 0 0 - - 35 - 50 84 — 465 - N D 0 0 N D 0 ... 23 July 10 August — 51 0 105 - 0 0 0 0 — 0 •- - 63 52 .. 15 ” 606 ■ Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. w Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 157 Table 2B. Densities (number/1000 m 3 ) of rainbow smelt larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station 2 3 1 1 * 0 Date 6 4-Itt Lake Michigan N D 0 0 0 42 7 60 67 - 172 - 12 May N 0 39 14 61 - - 20 May N D 0 0 0 0 0 0 7 “• 33 - 438 - N D 0 0 0 0 0 0 0 0 0 0 11 June N 0 0 30 0 15 0 17 June N D 0 0 0 34 0 52 0 0 0 0 0 6 May 3 June 4 - Harbor outlet, water from of Pere Marquette Marsh to Lake Michigan. * Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 158 Table 3B. Densities (number/1000 m 3 ) of yellow perch larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Lake Michiqan 2 3 N D 0 0 0 5 4 0 0 N D 5 7 445 1188 42 29 66 8 12. May N 0 335 0 8 20 May N D 0 11 0 0 0 0 0 0 0 _ - - N D 4 0 0 0 0 0 0 0 0 0 0 - 11 June N 0 0 0 0 0 0 17 June N D 0 0 0 0 0 86 0 0 0 0 0 6 May 3 June 0 H 1 22 April 4-0* 1 Date * Station 0 — - - - 5 - - - —• Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 159 Table 4B. Densities (number/1000 m 3 ) of johnny darter larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station Date 3 4-0" 4-1* Lake Michigan 1 2 N D 13 0 0 0 0 0 0 0 0 — — - N D 24 54 17 0 0 0 0 0 0 0 0 ... 11 June N 14 0 0 0 15 0 17 June N D 4 3 0 18 0 77 0 6 0 0 24 June N 6 56 0 0 - 1 July N D 0 58 0 9 0 22 0 0 0 0 0 - N D 0 8 0 39 N D 0 0 0 0 0 0 - — — 0 — - N D 0 0 0 - 0 20 May 3 June 13 July 23 July 10 August — ” 38 0 - — 28 - 79 - 0 ■ Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. * Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 160 Table 5B. Densities (number/1000 m 3) of spottail shiner larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station 1 . 2 3 * 01 Date 4-1# Lake Michigan N D 0 0 0 0 13 0 0 0 — - 12 May N 6 8 0 0 - - 20 May N D 22 0 0 0 0 0 0 0 0 - - N D 12 0 0 0 0 0 0 0 0 0 0 - 11 June N 0 0 0 0 0 0 17 June N D 0 0 0 0 0 0 0 0 0 0 24 June N 0 0 0 0 - 1 July- N D 0 0 0 0 0 0 0 0 0 0 0 0 0 0 - 0 0 0 0 - 0 - 0 - 6 May 3 June 13 July 23 July 10 August N D — N D 0 N D 0 0 0 0 0 0 _ — 143 82 — 504 — 36 — 90 ■ “* Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 161 Table 6 B. Densities (number/1000 m 3) of trout-perch larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station Lake Michigan 1 2 3 N D 23 0 0 0 0 0 0 0 12 May N 6 8 0 0 - - 2.0 May N D 4 0 0 0 0 0 0 0 0 — - - N D 36 0 0 0 0 0 0 0 0 0 0 - 11 June N 0 0 30 0 15 0 17 June N D 0 0 0 0 0 0 3 0 0 0 0 - 24 June N 0 0 0 0 - ~ 1 July N D 0 0 0 0 0 0 0 0 0 0 0 - N D 0 0 0 0 - Date 6 May 3 June 13 July 4-0* 4-1 — 0 ■- 14 0 Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 162 Table 7B. Densities (number/1000 m 3) of gizzard shad larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station Date 17 June 1 2 3 4-0* 4-1* Lake Michigan N D 0 0 0 0 0 130 565 138 0 0 0 -- 24 June N 29 112 38 24 - - 1 July N D 0 0 0 0 0 56 9 0 0 0 0 - N D 0 8 15 0 - 0 .13 July 0 Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. * Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 163 Table 8 B. Densities (number/1000 m3) of Pomoxis spp. larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. S tation * 0 1 <3* Date Lake Michigan 1 2 3 N D 22 4 0 0 0 14 0 - 0 ~ 0 “ 3 June N D 36 41 0 33 43 11 0 0 0 0 0 - 11 June N 5 8 0 0 0 0 17 June N D 0 3 16 94 18 35 34 12. 0 0 0 24 June N 40 1.50 25 12 - - 1 July N D 0 0 174 0 22 9 0 0 0 0 0 - 13 July N D 0 - 289 — 0 0 0 — - 0 — N D 0 - 26 - 0 - 0 0 0 - 0 - 20 May 23 July — 4-1# _ Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. # Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 164 Table 9B. Densities (number/1000 m 3) of Lepomis spp. larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station Date 3 0 0 0 - 0 N 52 1 July N D 9 N D 23 July 10 August Lake Michigan 2 24 June 13 July 4-jtt 1 4-0* 0 0 0 0 55 0 0 0 73 — 1080 ~ 0 0 - 0 _ - N D 4 -* 4062 - 16 0 0 0 0 - — N D 0 0 - 0 0 76 ”* 0 0 0 - *■" Harbor outlet, water flowing from Pere Marquette Marsh to Lake Michigan. * Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. 165 Table 10B. Densities (number/1000 m 3) of common carp larvae in the Pere Marquette Marsh and at the 1-m contour in Lake Michigan, 1981. N = night, D = day. Station Date Lake Michigan 2 3 N D 256 16 0 0 0 0 0 0 0 0 0 - 11 June N 489 17 0 8 15 0 17 June N D 2325 45 279 47 18 69 77 0 0 0 0 - 24 June N 1209 56 0 0 - - 1 July N D 1107 29 129 0 9 0 0 0 0 0 ~ N D 213 — 76 - 0 0 - 0 “ 0 - — N D 4 0 0 0 0 — **" 0 3 June 13 July 2.3 July 0 4-0* 4-1* 1 0 — Harbor outlet , water flowing from Pere Marquette Marsh to Lake Michigan. * Harbor outlet, water flowing into Pere Marquette Marsh from Lake Michigan. APPENDIX C Fisheries survey of Pere Marquette Marsh-Lake, 1981. 166 APPENDIX C. Table 1C. Total number of juvenile and adult fish collected with several methods in the Pere Marquette Marsh-Lake system, April through November, 1981. Gill Net Trap Net Trawl Species_________ (N = 22)____ (N = 23)_____ (N = 25)____ Total Alewife Trout-perch Johnny darter Yellow perch Mimic shiner Golden shiner Spottail shiner White sucker Gizzard shad Brown bullhead Black crappie Brown trout Pumpkinseed Golden redhorse Northern pike Shorthead redhorse Smallmouth bass Bowfin Silver redhorse Rainbow smelt Rock bass Rainbow trout Chinook salmon Coho salmon Common carp Walleye Black bullhead Lake trout Greater redhorse Bluegill Burbot Largemouth bass Longnose sucker Bluntnose minnow Common shiner Longnose gar Black redhorse 704 4 0 245 0 29 56 66 155 57 74 151 1 41 78 1936 16 0 143 306 35 82 28 3 103 40 2 29 9 272 2474 1684 350 307 545 289 230 13 3 41 1 117 60 2 6 22 13 0 1 49 54 0 7 3 0 1 27 10 23 18 5 3 3 9 2 32 25 7 2912 2494 1684 738 613 609 427 324 171 163 155 154 147 110 86 84 55 36 35 26 26 23 18 5 4 4 0 0 0 0 0 0 0 1 1 0 2 0 0 2 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 1 0 1 0 1 0 1 APPENDIX D Alphabetical list of command scientific names of fish used in this manuscript. From Robins et al. (1980) 167 APPENDIX D Table ID. Alphabetical list of common and scientific names of fish used in this manuscript. Names are from Robins et. al. (1980). Common names Alewife Black bullhead Black crappie Black redhorse Bloater Bluegill Bluntnose minnow Bowfin Brown bullhead Brown trout Burbot Chinook salmon Coho salmon Common carp Common shiner Deepwater sculpin Gizzard shad Golden redhorse Golden shiner Greater redhorse Johnny darter Largemouth bass Lake trout Lake whitefish Logperch Longnose gar Longnose sucker Mimic shiner Mottled sculpin Ninespine stickleback Northern pike Pumpkinseed Rainbow smelt Rainbow trout Rock bass Round whitefish Shorthead redhorse Silver redhorse Slimy sculpin Smallmouth bass Spottail shiner Trout-perch Walleye White sucker Yellow perch Scientific names Alosa pseudoharengus Ictalurus melas Pomoxis nigromaculatus Moxostoma duquesnei Coregonus hoyi Lepomis macrochirus Pimephales notatus Amia calva Ictalurus nebulosus Salmo trutta Lota lota Oncorhynchus tshawytscha Oncorhynchus kisutch Cyprinus carpio Notropis cornutus Myoxocephalus thompsoni Dorosoma cepedianum Moxostoma erythrurum Notemigonus crysoleucas Moxostoma valenciennesi Etheostoma nigrum Micropterus salmoides Salvelinus namaycush Coregonus clupeaformis Percina caprodes Lepisosteus osseus Catostomus catostomus Notropis volucellus Cottus bairdi Punqitius pungitius Esox lucius Lepomis gibbosus Osmerus mordax Salmo gairdineri Ambloplites rupestris Prosopium cylindraceum Moxostoma macrolepidotum Moxostoma anisurum Cottus cognatus Micropterus dolomieu Notropis hudsonius Percopsis omiscomaycus Stizostedion vitreum Catostomus commersoni Perea flavescens