...m.-. ‘ . ' qw‘lhl" :r/ .... .u mm; 1%: “a". u, I ‘. Wm“ . 11346" ’ 3 $4,. ‘tv. 1‘. ‘ I4 1 ; ‘ "‘15" K 1 V i "‘3: am% . ".3 “we. 6 fififig -3; m ‘6' fifi fiwé an, .l. ' J "(Es-g4}: is. I“. .myu‘fiié? 2&3fi$§fi$fi%’ fifififififififi$g ar3¢¢£g%%$€®?aku *wwmg 1‘r 4 I 1??“ «19%gaw géwwwfi .‘_; n... kw‘éfi’r'rv _ ; filing», I‘d" 1v $13.33.! N fife,“ ~ \ . % I". Q ”15;" r ,I. 'l :‘U 4,9 . n L; (W C? ixa‘vigyr' 54 ”451-4?“ 41 ’1- m... 3 ‘x ., THEarS This is to certify that the thesis entitled The Diet of Yellow Perch (Perca flavescens) in Lake Michigan, Near Ludington presented by Douglas Lee Peterson, Jr. has been accepted towards fulfillment of the requirements for Master of Science degree in Fisheries and Wildlife 72/46 fl x/fzw/V/L/ Major professor Date July 1, 1993 0-7539 MS U is an Affirmative Action/Equal Opportunity Institution Iliiii'iiil iiiiiliiiiiliiiIi’liliiifiil 1 3 1293 01022 0204 r LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative Action/Equal Opportunity institution ammo-n1 The Diet of Yellow Perch (Perca flavescens) in Lake Michigan, Near Ludington BY Douglas Lee Peterson Jr. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Fisheries and Wildlife 1993 ABSTRACT THE DIET OF YELLOW PERCH (PERCA FLAVESCENS) IN LAKE MICHIGAN, NEAR LUDINGTON by Douglas L. Peterson Jr. Yellow perch (Perca flavescens) were sampled with variable mesh gill nets from May to August in 1991 and 1992. Diet analysis of 2,544 yellow perch collected in 1992 showed that alewife (Alosa pseudoharenqus) was the most important prey species, comprising 64% of total prey biomass consumed during the growing season. Other fish important in the diet included rainbow smelt (Osmerus mordax) and bloater chub (Coregonus hgyi). In addition to preying on other fish, yellow perch also preyed heavily on Bythotrephes cederstroemi, an exotic zooplankton introduced into Lake Michigan in 1987. Although a. cederstroemi was present in the stomachs of nearly all yellow perch collected after June 24, it was especially important in the diet of yellow perch 100-200 mm TL in total length. These findings suggest that the food habits of yellow perch may have important implications regarding the production of other Lake Michigan sport fish. Copyright by Douglas Lee Peterson Jr. 1993 To Dad and Kathy iv ACKNOWLEDGMENTS I would like to thank my major professor, Dr. Niles Kevern, for his continued support throughout this project. I would also like to express my gratitude to the other members of my graduate committee, Dr. Tom Coon and Dr. Patrick Muzzall, for their helpful advice and review of this publication. I would like to extend special thanks to my lab and field assistants John Neuman, Kate Whitmore, Tom Schild, and Janine Scott. I am also grateful to Rob Elliott, a fellow graduate student who took time from his own research to assist me with study design and field logistics. His expertise regarding fisheries data bases also was instrumental in the analysis of my data. Finally, I would like to thank Asa Wright and Tom Rozich of the Michigan Department of Natural Resources for their support in dealing with private agencies. Tom’s encouragement and assistance at Ludington was greatly appreciated. TABLE OF CONTENTS Page List of Figures . . . . . . . . . . . . . . . . . . . viii Introduction . . . . . . . . . . . . . . . . . . . . . 1 Area of Study . . . . . . . . . . . . . . . . . . . . . 6 Materials and Methods . . . . . . . . . . . . . . . . . 10 Field Collections . . . . . . . . . . . . . . . . . . 10 1991 Field Season . . . . . . . . . . . . . . . . 10 1992 Field Season . . . . . . . . . . . . . . . . 12 Laboratory Analysis . . . . . . . . . . . . . . . . . 14 Descriptive Categories . . . . . . . . . . . . . . . 15 Results . . . . . . . . . . . . . . . . . . . . . . . . 17 Diets of Large Yellow Perch (>200 mm TL) . . . . . . 26 Diets of Small Yellow Perch (5200 mm TL) . . . . . . 31 Diurnal Effects on Diet . . . . . . . . . . . . . . . 31 Effects of Substrate on Diet . . . . . . . . . . . . 36 Sex Ratio . . . . . . . . . . . . . . . . . . . . . . 36 Discussion . . . . . . . . . . . . . . . . . . . . . . 40 Diets of Large Yellow Perch (>200 mm TL) . . . . . . 41 Diets of Small Yellow Perch (5200 mm TL) . . . . . . 45 Diurnal Effects on Diet . . . . . . . . . . . . . . . 49 Effects of Substrate on Diet . . . . . . . . . . . . 50 Sex Ratio . . . . . . . . . . . . . . . . . . . . . . 52 vi Limitations of the Data Impacts of g. cederstroemi Impacts of Alewife Summary List of References vii 53 56 57 65 67 Figure Figure Figure Figure Figure Figure Figure Figure Figure LIST OF FIGURES Area of study. Map and location of sampling stations in the vicinity of the Consumer’s Power Pumped Storage Plant in Ludington, Michigan Seasonal catch by station, 1992 Seasonal catch by station, 1991 Annual diet composition of yellow perch based on percent biomass of each prey type consumed, 1992. *The "Other" category contained unidentified fish remains, aquatic insects, and fish eggs. Total lengths of yellow perch vs the biomass of prey type consumed by each, 1992 . . . . . . Frequency of occurrence of the three most common prey types consumed by large yellow perch (>200 mm TL), 1992 . . . . . . . . . . Weekly diet composition of large yellow perch (>200 mm TL) based on percent biomass, 1992 . . . Weekly catch and number of non-empty stomachs present for small (5200 mm TL) and large (>200 mm TL) yellow perch, 1992 . . . . . . . . Diurnal effects on catch, foraging intensity, and diet composition of large yellow perch (>200 mm TL) in June, July, and August, 1992 viii Page 19 21 23 25 28 30 33 35 Figure Figure Figure Figure Figure 10. ll. 12. 13. 14. Diet composition of yellow perch collected from two different substrate types near Ludington, Michigan during the week of August 14, 1992. *The "Other" category contained smelt, bloater, and unidentified fish remains Diet composition of small (5200 mm TL) and large (>200 mm TL) yellow perch based on percent biomass. *Insufficient sample size of fish with non-empty stomachs . . . Length frequency distribution of yellow perch caught near Ludington, Michigan, 1991 and 1992 Length frequency distribution of alewife consumed by yellow perch, 1992 Total lengths of yellow perch vs total lengths of alewife consumed by each, 1992 ix 38 48 55 59 62 INTRODUCTION In 1967 the Michigan DNR implemented a management plan, the Great Lakes Rehabilitation Program, that radically altered the future of Great Lakes fisheries. This plan, which initiated the introduction of pacific salmon into Lake Michigan, met with unprecedented success. While reducing nuisance populations of alewife, it created a multi-million— dollar sport fishery which has continued to date. However, the new fishery also brought with it an entirely new set of management problems, many of which remain today. The accidental introduction of unwanted exotics, the spread of toxic contaminants, and changes in forage composition, are a few examples of problems which seriously threaten the future of Lake Michigan’s sport fishery. In recent years, several significant lake-wide changes have occurred within the aquatic community of Lake Michigan. Notably, the dramatic decline in alewife (Algga pseudoharengus) populations in the mid 1980's and the subsequent decline of the salmon fishery are of major economic concern. The invasion and accidental introduction of exotic species such as the zebra mussel (Dreisena 2 polymorpha), the ruffe (Gymnocephalus cernuus), and the spiny water flea (Bythotrephes cederstroemi) have also had significant impacts on community structure (Mills et al. 1993). While Lake Michigan’s salmon fishery has undergone serious decline over the last 10 years, yellow perch (Pgrgg flavescens) populations in the lake have greatly increased (Wells and Hatch 1984, Jude and Tesar 1985, Eck and Wells 1987). Most biologists agree that the recent resurgence in yellow perch populations can be attributed to declines in alewife populations, because adult alewife compete with and prey on larval yellow perch (Smith 1970, Wells 1977). Several studies have documented sharp increases in larval yellow perch abundance in years immediately following alewife declines (Eck and Wells 1987, Jude and Tesar 1985, Christie 1974). Several hypotheses have been put forth to explain recent declines in Lake Michigan alewife including salmonid predation (Stewart et al. 1981), competition with native planktivores (Crowder and Binkowski 1983), and weather related effects (Eck and Brown 1985). Although all of these explanations may be at least partially correct, alewife predation by yellow perch has not been well studied. Many biologists believe that reduced alewife stocks may be an important stressor contributing to bacterial kidney disease mortality in Lake Michigan salmon. Consequently, alewife predation by a burgeoning yellow perch population 3 may be having a significant detrimental effect on the Lake Michigan salmonid fishery. Some fisheries biologists are skeptical of the potential of yellow perch to significantly impact alewife populations. They contend that Lake Michigan alewife populations exploded in the mid 1960’s at a time when yellow perch were extremely abundant (Christie 1974, Eck and Wells 1987, Hatch et al. 1981). However, since that time ecological change within Lake Michigan may have affected change in yellow perch diet. Previous studies at Ludington, (Brazo 1973, Hauer 1974), indicated that sculpins (Cottus cognatus and Cottus bairdii), not alewife, were once the primary forage species for yellow perch. However, recent studies on forage abundance in Lake Michigan (Wells and Hatch 1984, Eck and Wells 1987) indicate that sculpin populations have greatly declined in recent years. In 1991 this study was initiated to investigate the relative importance of different species in the diet of Lake Michigan yellow perch. Because it is one of the lake’s most abundant fish species, a major focus of this project was to determine the extent of yellow perch predation on alewife as a means of identifying a potential for competition with salmonid species. In addition to interactions with alewife and salmonid species, yellow perch may also be effected by the recent invasion of Bythotrephes cederstroemi. Accidentally introduced into Lake Michigan from ballast water in 1987, g. 4 cederstroemi has rapidly spread throughout the entire lake (Evans 1988, Lehman 1987). Laboratory feeding trials using B. cederstroemi and juvenile yellow perch (Barnheisel 1991) have shown that the long caudal spine of B. cederstroemi make them difficult for small yellow perch to ingest. Barnheisel has also shown that young yellow perch eventually learn to avoid B. cederstroemi. Field studies on Great Lakes yellow perch (Schneeberger 1991, Baker 1992) corroborate these findings. Because B. cederstroemi is a voracious zooplankton predator, it may compete with many fish species that depend on the native zooplankton, either as juveniles or adults. Although its effects on trophic structure are currently unknown, most biologists agree that B. cederstroemi has the potential to significantly alter planktonic food webs in the Great Lakes (Lehman 1987, Scavia et a1. 1988). However, Keitly (1990) suggests that predation by planktivorous fish may limit its abundance, thereby reducing its impacts on native zooplankton. Although B. cederstroemi is not readily eaten by small fish, field studies suggest that it is, at times, a principle prey species for adults of many Great Lakes species, including yellow perch and alewife (Schneeberger 1991, Keitly 1990). Although the impacts of yellow perch predation on B. cederstroemi densities were not measured, yellow perch predation on B. ggdgggprggmi was another important focus of this study. 5 In addition to its obvious impacts on growth and reproduction, the diet of a species largely defines its niche. As a generalist, yellow perch may learn to exploit new prey species when changes in forage availability occur. Consequently, dramatic changes in forage availability could potentially affect a change in diet. In the past century, Lake Michigan has undergone dramatic ecological changes as a direct result of overexploitation by commercial fisheries and invasion by exotic species. Intentional introductions of exotic sport fish have compounded these changes. Although Lake Michigan currently supports a sport fishery of immense economic value, ecological instability has created uncertainty about its future. As one of Lake Michigan's most abundant and popular sport fishes, yellow perch may play a key role in future management strategies. Consequently, the purpose of this study was to gain a better understanding of the role yellow perch play in the lake Michigan food web. The specific objective of this study then, was to describe the diet of yellow perch in Lake Michigan by determining the relative importance of different prey species. Of particular concern was the question of whether alewife and B. cederstroemi are important prey species of yellow perch. Area of Study The study site consisted of six sampling stations located in the nearshore waters surrounding the Ludington Pumped Storage Plant (Figure 1). The plant is located on the shore of Lake Michigan, approximately three miles SSE of the Ludington harbor. During summer months these waters support an excellent sport fishery for both yellow perch and salmonids. The bottom substrate of the area is mostly sand, however large stretches of rock and gravel are common in nearshore areas. Historical data on Ludington fish populations (Liston and Tack 1972, Brazo 1973) indicated that in late May and early June vast numbers of yellow perch move into this area to spawn and feed. The concentration of yellow perch in this area provided an excellent opportunity to conduct diet studies of this species over an entire growing season. In 1991 yellow perch were collected within a 0.5 mile radius of the Ludington Pumped Storage Hydroelectric Plant (stations A1-B2). In 1992, fish were collected approximately 1.5 miles north of the plant, just south of Ludington harbor (stations C and D). Surface water temperature ranged from a low of 2W: in early April to a high of 18W: in August. Temperature fluctuations at the shallow water stations were frequently dramatic depending on wind speed and direction. In addition to yellow perch, several other fish species inhabited this area. Other species commonly collected Figure 1. Area of study. Map and location of sampling stations in the vicinity of the Consumer’s Power Pumped Storage Plant in Ludington, Michigan. /////////// ///// US LUDINGTON é / // MICHIGAN p... // Marquette // Lake Buttersville Park 82 A2 Pumped I] Storage Reservoir A - 1991 2 m Stations B - 1991 12 at Stations C . 1992 2 111 Station D - 1992 12 m Station Bl Al 1/2 1 i i SCALE OF MILES LUDINGTON PUMPED STORAGE PROJECT SAMPLING STATIONS 1991 & 1992 9 included alewife, steelhead (Oncorhynchus mykiss), brown trout (Salmo trutta), lake trout (Salvelinus namaycush), Chinook salmon (Oncorhygchus tshawytscha), longnose sucker (Catostomus catostomus) spottail shiner (Notropis hudsonius), and round Whitefish (Prosopium cylindracuem). The most common zooplankton were copepods. Few cladocera were present until July, when B. cederstroemi emerged and became the most abundant zooplankton species (Barner 1993). In 1991 all fish were collected within a 1 mile radius of the plant. However, at no time were yellow perch captured inside the Consumers Power Company (CPCo) barrier net used in this study. In 1992 sampling stations were moved approximately 1 mile north of the plant (stations C and D) so as not to disturb CPCo operations as well as to minimize effects from plant activities. Although fish were collected at the same depths each week, the exact placement of the gill nets at stations C and D was varied to minimize contact with local anglers. METHODS AND MATERIALS Field Collections In 1991 and 1992 this study used gill netting efforts of the CPCo Barrier Net Monitoring Program to gather data on Ludington’s yellow perch population. In 1991 CPCo permitted limited samples of fish (40 per week) to be taken from their gill nets for diet analysis. Information obtained from these samples was used to design a more thorough study of yellow perch diets in 1992. Because extensive examination of CPCo yellow perch catches was not possible, only general data regarding total catch and forage abundance from the CPCo data will be discussed. In 1992 yellow perch collected from the CPCo study were used only to supplement independent field collections taken from stations C and D in Figure 1. All diet data discussed in this study were taken exclusively from 1992 samples. 1991 Field Season In 1991, CPCo conducted gill netting twice weekly from mid April to mid October. The nets were set at sunset and removed shortly after dawn, typically from 7 pm to 8 am. Two types of variable mesh, bottom gill nets were used to collect fish from the four stations (labeled A1, A2, B1 and B2) illustrated in Figure 1. The first gill net type, designed to sample the 10 ll littoral zone, was 100.6 m long and 1.8 m deep. The second type also measured 100.6 m in length, but was 7.3 m deep. The larger gill nets were used to sample fish from a depth contour of 12 m, approximately 1/2 km from shore. Both types of nets were composed of 11 distinct 9.2 m panels, each consisting of a different mesh size. The panels were connected in order of ascending size, with stretch mesh widths of 2.5 cm, 3.75 cm, 5 cm, 6.25 cm, 6.875 cm, 7.5 cm, 10 cm, 11.25 cm, 12.5 cm, 15 cm, and 17.5 cm. Samples for the CPCo study were sometimes taken in other locations around the plant not indicated in Figure 1. However, only stomachs of yellow perch caught outside the barrier net enclosure were used in this study, because diets of yellow perch inside the net might not be representative of fish in the rest of the area. Typically, yellow perch were subsampled from one shallow (2 m) and one deep (12 m) station. The use of the shallow and deep stations was not intended to determine differences in diet based on depth, but rather to ensure an adequate catch. The nets were retrieved from the lake shortly after dawn and transported to the CPCo processing facility. There, all species of fish were removed, sorted, counted, and total catch for each species was recorded. Once removed from the nets, yellow perch from all four stations were packed in ice and kept separate until a subsample could be taken and analyzed further. Subsampled yellow perch were 12 measured, weighed, and sexed before stomachs and scale samples were removed. 1992 Field Season Many of the prey items in yellow perch stomachs collected in 1991 were too digested for positive identification. In an attempt to acquire fresher samples several other sampling methods were tested. In 1992 bottom trawling was the initial method of choice because size selectivity and digestion during collection could be minimized. An otter trawl, with a mouth opening of 4.9 m, was towed through the study area for periods of 5 to 15 minutes per haul. However, rock substrate, common within the study area, made trawling ineffective. On every attempt made the cod end of the trawl was torn and no yellow perch were captured. Trap nets were also tested but their use was discontinued since they produced extremely variable catches without eliminating the gill net biases previously encountered. On the first day trap nets were set, one trap yielded more than 600 yellow perch (all but 7 were released alive), while the other trap remained empty. Over the next three days no fish were caught in either trap. In addition to their inconsistent performance, the numerous buoys on the traps made them highly conspicuous and tampering by local anglers was a second major drawback of trap net sampling. 13 After other methods had proved ineffective, modifications of 1991 sampling methods were used to attain yellow perch stomachs that had undergone minimal digestion. This was done by reducing the duration of net sets and by expediting net processing following retrieval. In 1992 nylon gill nets with four 15 m panels were used. The stretch mesh sizes of these panels were 5 cm, 6.25 cm, 7.5 cm, and 10 cm. Data from CPCo 1991 collections indicated that these panels were the most effective for sampling yellow perch. By eliminating mesh sizes that caught few yellow perch, non target catch was greatly reduced, thereby reducing processing time. Later setting and earlier retrieval of gill nets in 1992 reduced the length of time yellow perch remained in the nets. These modifications also eliminated exposure of fish to sunlight, helping to minimize digestion after capture. Modifications to gill nets and sampling protocol did not appear to reduce catches of yellow perch and ultimately led to the acquisition of higher quality samples in 1992. In addition to the overnight gill netting conducted twice weekly, three 24-hour gill net sets were also conducted. These sets were made at the end of June, July, and August. During each 24-hour sampling period, nets were pulled, cleared of fish, and reset every 6 hours. Samples collected from 24-hour gill netting were used to identify diurnal patterns in yellow perch diets. One other collection of yellow perch was made on August 14 13, 1992, by use of a 305 m x 30 m lampara net (similar to a purse seine) in three meters of water. Located approximately six miles north of the weekly gill net stations, this second site contained only sand substrate. No patches of rock were detected by sonar or during earlier lampara net sampling‘. Stomachs of yellow perch collected at this site were used to identify differences in diet that may have been attributable to habitat differences. Laboratory Analysis In both years stomachs from representative samples were preserved in 10% formalin and stored individually in coded Whirl-Pak bags. Preserved stomachs were later transported to the Michigan State University laboratory where they were dissected and their contents recorded. Prey items in stomachs were separated, counted, and weighed (wet weight) to the nearest 0.01 grams. Total lengths of forage fish consumed were determined whenever possible. Fish that were too decomposed to accurately measure, were weighed and identified by vertebrae structure using a 60X dissecting scope. Fish that were too digested to identify were categorized as "unidentified fish remains". 1In his study of chinook salmon, fellow graduate student Rob Elliott conducted lampara netting at this and numerous other Lake Michigan locations several times throughout the summeru Elliott reports that although small numbers of yellow perch were often collected, only one large catch of yellow perch was made outside of the Ludington study area. Thirty- two of the 219 yellow perch from that catch were randomly subsampled and kept for the diet comparisons discussed here. 15 Descriptive Categories An obstacle to a thorough diet analysis of any fish species is how to determine the relative importance of different prey types in the diet on a seasonal basis. In species such as yellow perch, this problem is compounded by the wide variety of fish and invertebrates consumed as well as temporal and spatial changes in diet. The categories that best described data here were frequency of occurrence and percent total biomass. Other descriptive categories frequently used in diet studies include percent total volume and percent of total organisms consumed. These categories were not used in this study because small prey items, such as zooplankton, might be consumed in very large numbers and volume, but not contribute significantly to the total biomass or calories consumed. Percent of total prey biomass for each prey type consumed is probably the best single indicator of the relative importance of each prey type. However, diet analysis of any species based on biomass alone may be biased by the consumption of large quantities of atypical prey items by a few individuals. Frequency of occurrence is useful in diet studies because it indicates the extent to which fish in the population function as a single feeding unit (Nielsen and Johnson, 1983). However, frequency of occurrence only indicates the percentage of the fish sampled where a l6 particular prey type is present. It says nothing about the relative importance of different prey types. When a high percentage of a population consumes a relatively small biomass of a particular prey, a high frequency of occurrence is obtained for that prey species. However, since small prey usually contain few calories, many must be consumed to equal the calories acquired from a single large prey. Yellow perch collected in this study, often had consumed prey species of differing sizes. Consequently, percent of total prey biomass was used with frequency of occurrence to describe diet. RESULTS During 13 weeks of sampling from mid May to mid August of 1992, stomachs from 2,544 yellow perch were collected and preserved. Although weekly catch varied, the target catch of 200 fish per week was reached or exceeded in nearly every week (Figures 2 and 3). Dissection of preserved yellow perch stomachs showed alewife, crayfish, and B. cederstroemi to account for more than 84% of the prey biomass consumed over the entire season (Figure 4). For this reason, most discussion of diet will focus on these three prey species. Although alewife and crayfish were heavily preyed upon by yellow perch greater than 200 mm TL, they were much less important to smaller yellow perch (Figure 5). Because diet appeared to change once yellow perch reached 200 mm TL, diet analysis was performed separately for large (>200 mm TL) and small fish (5200 mm TL). 17 18 Figure 2. Seasonal catch by station, 1992. l9 700- 600 _ 2 m Station 400— Catch 300— 200- 100 ° * f f I . ll'l'l'l'i'l'l'l'l’ 13 20 27 3 10 17 24 1' 8 15 22 29 5 12 19 800 7.. _ 60° F 12 m Statlon 500 t- 400— Catch 300- 200—- 100— 13202731017241815222951219 May June July August Date 20 Figure 3. Seasonal catch by station, 1991. Catch Catch §§§§§ 700 § § § § 200 100 § 700 O 21 2 m Stations WI PM 1320273 10172418 1522295 1219 12 m Stations 13202731017241815222951219 May June ' July August Date Figure 4. 22 Annual diet composition of yellow perch based on percent biomass of each prey type consumed, 1992. *The "Other" category contained unidentified fish remains, aquatic insects, and fish eggs. 23 63.8% B. cederstroemi {373?- 5? Bloater E Alewife Crayfish Sculpin I Other“ Figure 5. 24 Total lengths of yellow perch vs the biomass of prey type consumed by each, 1992. Biomass (g) 88883 O 50 88 10 S 30 10 25 Yellow Perch Total Length (mm) -— IALIWVHWE L— “+- l l l 100 150 400 CJLKYFTSEI I I D I 100 150 400 ._ IllCEZHERSTTRQEmlI I 0"“ I “”fix’ 4 ‘ .,--. I I 100 150 200 250 300 350 400 26 Diets of Large Yellow Perch (>200 mm TL) The percentage of yellow perch with non—empty stomachs and the frequency of occurrence of the three most important prey species was determined for each sampling week (Figure 6). Most yellow perch captured before June 26 had empty stomachs, but those found with non-empty stomachs fed almost exclusively on alewife. Throughout May and June the percentage of yellow perch with non-empty stomachs continued to increase. During the same period alewife became increasingly common in the diets of large yellow perch, while crayfish and other species were rarely eaten. By June 26 frequency of occurrence had increased to 87% for alewife and 7% for crayfish. As the frequency of alewife in the diet declined after June, both crayfish and B. cederstroemi became increasingly common. In the last two weeks of the study, frequency of occurrence of B. cederstroemi increased to 65% and 63%, respectively. Diet analysis based on biomass of prey types consumed (Figure 7) also showed that alewife were the most important prey of large yellow perch. From May to mid June alewife comprised at least 85% of the total prey biomass consumed in each week. In August frequency of occurrence of B. cederstroemi was much higher than for any other prey type, however, the biomass of each prey type consumed was about equal. 27 Figure 6. Frequency of occurrence of the three most common prey type consumed by large yellow perch (>200 mm TL), 1992. Frequency of Occurrence :8888 28 ‘ PREY TYPE _[:l NON-EMPTY STOMACHS - B. CEDERSTROEMI Eflflmfl HH |l|||| 22295121926 May June July Week Ending 29 Figure 7. Weekly diet composition of large yellow perch (>200 mm TL) based on percent biomass, 1992. O B10mass(g) Hid H I SEE: 30 - :1 AVG RATION SIZE CRAYFISH *EEEQQ B. CEDERSIROEMI :TWWW.Iflmflmflflfl— 2229'5121926310172431714 May June July | August Week Ending 31 Diets of Small Yellow Perch (<200 mm TL) A complete seasonal diet analysis of small yellow perch was not possible due to an insufficient sample size of small fish with non—empty stomachs. Diet analysis of small yellow perch was limited to sampling weeks from July 31 to August 14. During these three weeks catches of non—empty small yellow perch were 25, 28, and 15, respectively. Once B. cederstroemi became abundant in late July, it was an important prey species for most small yellow perch. In the last three sampling weeks B. cederstroemi was found in 73% (July 31), 78% (August 7), and 100% (August 14) of all small yellow perch collected. During the same three weeks, it composed 49%, 84%, and 83%, respectively, of total prey biomass consumed. Alewife and crayfish were rarely found in stomachs of small yellow perch, even during periods when large yellow perch preyed heavily on them. Diurnal Effects on Diet Diet analysis of yellow perch collected during 24-hour gill netting efforts showed that diurnal changes in mean ration size, diet composition, and catch were apparent but largely unpredictable (Figure 9). Mean ration size was highest during the 0800 to 1400 period in every 24-hour collection. However, diet composition did not appear to change during any of the 24-hour sampling periods, except in August when the percentage of alewife in the diet increased sharply from 0800 to 1400. 32 Figure 8. Weekly catch and number of non—empty stomachs present for small (5200 mm TL) and large (>200 mm TL) yellow perch, 1992. 400 350 250 200 150 100 so} 400 350 300 — 250 150 33 SMALL YELLOW PERCH _ (5200 mm) WWDD [—1 LARGE YELLOW PERCH (> 200 mm) 22 29 5 12 19 26 3 10 17 24 31 May June July Week Ending 7 14 August 34 Figure 9. Diurnal effects on catch, foraging intensity, and diet composition of large yellow perch (>200 mm TL) in June, July, and August 1992. Prey Biomass (g) 35 EMean Ration Size _ .3 Alewife June 30 ._,_, 0200 - 0800 0800 - 1400 Sampling Period August 18 (saq 9 / list!) “3190 36 Catch per effort varied markedly with time of day. In all three 24-hour collections the 0200 to 0800 period yielded the highest catch and was followed by a decline in catch from 0800 to 1400. Effects of Substrate on Diet Thirty-two of the 219 yellow perch collected over sand substrate were subsampled for diet comparisons with yellow perch collected on the same day over the rocky substrate at the gill net stations. Crayfish and alewife constituted 43.1% and 16.5%, respectively, of diet biomass of yellow perch over the rock substrate (Figure 10). The diet of yellow perch collected over the sand substrate contained 55.4% alewife but no crayfish. The percentage of B. cederstroemi in the diets of yellow perch was similar at both sites; 18.2% over sand and 27.3% over rock. Other prey types, primarily bloater and smelt, comprised 25.1% of the diet biomass in yellow perch collected over sand but only 14.7% of the diet biomass of yellow perch collected over rock. we Catches in May of both years were comprised of an overwhelming percentage of ripe males. In 1992, males constituted 93.3% of the 1,918 yellow perch caught in May. Most males captured at this time had readily running milt, although some were already partly spent by our first 37 Figure 10. Diet composition of yellow perch collected from two different substrate types near Ludington, Michigan, during the week of August 14, 1992. *The "Other" category contained smelt, bloater, and unidentified fish remains. 38 B. cederstroemi 18.2% Alewife 55.4% Sand Substrate n = 104 B. cederstroemi 27.3% Crayfish 43.1% Other“ 13.1% Alewife 16.5% Rock and Sand Substrate n=3l 39 sampling day on May 13, 1992. Males continued to comprise the majority of the catch until mid June when females became more abundant than males. By late July in both years the percentage of males in the catch had dropped to less than 20%. In 1992 the first spent female was caught on May 26 and the last ripe female was caught on June 15. The spring of 1992, however, was unusually cold and probably prolonged spawning in that year. In 1991 weather conditions were more typical and no ripe females were captured after June 4. In both years, very few females were caught until the last week in May when at least a few had already spawned. DISCUSSION In order to determine which prey species were most important, the total biomass of each prey type was summed over the entire season and then averaged over total number of fish examined (approximately 200 per week). By converting these values to percentages, the relative importance of each prey species was determined for the entire growing season (Figure 4). Although this is a crude representation of the diet, it shows that 84.5% of the biomass consumed during the growing season was composed of alewife, crayfish, and Bythotrephes. Plotting yellow perch length vs biomass for each of these three prey species (Figure 5), showed a change in diet once yellow perch attained a "critical length" of about 200 mm. Because yellow perch size appeared to have a strong effect on diet, a separate diet analysis was conducted for small (5200 mm TL) and large (>200 mm TL) fish. Although diet analysis is often conducted for different age classes, that practice was not used here for two reasons. First, the number and type of prey species available to yellow perch depend on predator size, not age. Since the timing at which yellow perch attain the critical length may not correspond to a discrete age classification, separation by age class is not the best criteria for diet analysis. Secondly, females grow more rapidly than males, 40 41 allowing them to attain the critical length at a younger age. Therefore, analysis by age class would also necessitate a further separation by sex as well. A Chi- square test comparing the percentage of large empty and non- empty fish, captured within one week at the same location, showed no significant differences between sexes (X?=0.0134). For these reasons, yellow perch were separated for diet analysis by length at which an obvious change in diet occurred and not by age or sex. Diets of Large Yellow Bgrch (>200 mm TL) To illustrate seasonal trends, diet data frequently have been plotted against sampling week. Frequency of occurrence (Figure 6) indicates the percentage of fish sampled that had consumed each prey type, whereas percent biomass (Figure 7) indicates the proportion of the diet by weight attributed to each. In Figure 6 the percentages indicated by the shaded bars were calculated by dividing the number of yellow perch containing at least 0.1 grams of each prey type by the total catch for that day. This calculation is sometimes made using only the total of non—empty fish, however this practice was not used here because it tends to inflate values, especially when foraging is low. In comparing the frequency of occurrence for the three major prey species, alewife are consumed by a higher percentage of the fish sampled than any other species until 42 late July. However, as alewife predation dropped, after peaking on June 26, crayfish and B. cederstroemi became increasingly important. In 1992 B. cederstroemi was first detected in the study area in late June, however it did not become abundant until the last week of July (Barner 1993). This may explain why B. cederstroemi did not appear in the diet until July 31. Once it became abundant, frequency of occurrence of B. cederstroemi was higher than for any other prey species. However, to accurately assess the importance of B. cederstroemi in yellow perch diets, biomass of each prey species must be compared. Weekly diet composition based on percent biomass (Figure 7) shows that large yellow perch rely almost exclusively on alewife throughout most of the growing season; including periods of peak food consumption. Despite its high frequency of occurrence in August, B. cederstroemi comprises only about one-third of the total prey biomass consumed by large yellow perch. After June 10, alewife consumption stabilized at about 1.0 grams per fish, despite a slight increase in mean ration size. Since alewife abundance declined after mid June (CPCo 1993), fluctuations in mean ration size were probably associated with availability of alternate prey species. These data suggest that alewife are the staple of large yellow perch diets, but that crayfish and B. cederstroemi are also eaten in late summer when alewife are scarce, or "unavailable". The distinction between abundance and 43 availability of prey may be important in the case of alewife. Die-offs of alewife in Lake Michigan are still common, although greatly diminished in magnitude since their decline in the 1980’s. On several occasions, stressed alewife were observed swimming erratically at the lake surface. Alewife in this condition were undoubtedly easy prey, making them highly available to predators. Periods of alewife stress were not always observed when alewife were highly abundant in the area, but rather, seemed linked to rapid increases in nearshore water temperatures. Studies on Lake Michigan alewife (Graham 1956 and Otto et al. 1976) have shown that alewife die-offs are frequently caused by thermal shock induced by sharp temperature gradients. Although relatively unimportant to the diet as a whole, several other prey species were identified. If the biomass of alewife, crayfish, and B. cederstroemi are subtracted from the total prey biomass, the category of "unidentified fish remains" contributed about 12% of the remaining total. As explained earlier, this category contained fish too digested for positive identification, probably in roughly the same proportion as the identified species. Differences in average length of the various forage fish species undoubtedly affected rates of digestion, making quantitative diet reconstruction impossible. Because alewife constituted the vast majority of forage fish consumed, it is likely that most of the unidentified fish 44 remains found was actually alewife. Consequently, the disparity between alewife biomass and all other prey types is probably slightly greater than the data indicate. The relative unimportance of other forage fish in yellow perch diets was unexpected. However, CPCo gill netting seining data from both years showed that alewife constituted the vast majority of the forage fish collected within the study site. Although some of the forage species listed were not susceptible to the gill nets (ie. johnny darter and sculpins), data from CPCo beach seining (CPCo 1992, CPCo 1993) also showed that these species were much less abundant than alewife. When compared with Brazo’s 1973 diet data, the findings of this study suggest that a change in Lake Michigan yellow perch diets has occurred. Although it is unclear why yellow perch diets have changed over the last 20 years, changes in forage availability are the most likely explanation. Although these data may have important implications regarding sport fish production, perhaps of equal interest is what was not found. As one of the most abundant piscavores in Lake Michigan, yellow perch probably play an influential role in determining the production dynamics of other fish species. Explanations for recent declines in salmon populations have focussed on the interaction of BKD and alewife declines. However, Horn (pers. com. 1982) suggests that yellow perch predation on young salmon smolts might be an important contributing factor. 45 In light of this possibility, yellow perch stomachs were carefully examined for salmon smolts. Using freshly thawed samples, close study of smolt skeletal structure was made prior to stomach dissections so that even highly digested smolts could have been identified if present. Despite these preparations, no salmon smolts were found in any of the 2,544 yellow perch stomachs examined. Since yellow perch populations undoubtedly exceed yearly smolt numbers by several orders of magnitude, significant smolt predation by yellow perch might not be detected, even if many thousand yellow perch stomachs were examined. Although yellow perch in Lake Michigan have been found to prey on young salmon in some instances (Horn pers. comm), available data suggest that these cases most likely result from high smolt densities during stocking. Consequently instances of yellow perch predation on young salmon can probably be characterized as localized stocking mortality. Diets of Small Yellow Perch (<200 gm TL) Diet analysis of small yellow perch was less revealing than for large yellow perch. Of the 252 small yellow perch collected 38% had non—empty stomachs, as compared with 49% for large yellow perch. A Chi-squared test showed that small yellow perch had significantly more empty stomachs than large yellow perch (X%=9.29). This comparison provides a second rationale for analyzing diets of small and large fish separately. Unfortunately the relative scarcity of 46 small yellow perch with non-empty stomachs made a complete seasonal diet analysis impossible, although some findings were meaningful. Once B. cederstroemi became abundant in late July, it was an important prey species for small yellow perch. In the last three sampling weeks B. cederstroemi occurred in 73% (July 31), 78% (August 7), and 100% (August 14) of all small yellow perch collected. Frequency of occurrence of B. cederstroemi was also high for large yellow perch at this time, however it composed a much higher percentage of the prey biomass in small yellow perch (Figure 11). Although the scarcity of data for small yellow perch make conclusions uncertain, they do suggest that B. cederstroemi may be an important food source for yellow perch 5200 mm TL. These results may seem contrary to other studies indicating that small yellow perch avoid B. cederstroemi because of its protective spines (Baker 1992 and Barnheisel 1991). However, no yellow perch under 100 mm were collected in this study. Baker (1992), who did collect yellow perch less than 100 mm from Lake Michigan, found that these fish rarely consumed B. cederstroemi. Baker also found that frequency of occurrence of B. cederstroemi increased rapidly as length of yellow perch increased. These findings are compatible with the data presented here and may suggest that B. cederstroemi is an important food source for yellow perch 100-200 mm TL. 47 Figure 11. Diet composition of small (5200 mm TL) and large (>200 mm TL) yellow perch based on percent biomass. *Insufficient sample size of fish with non-empty stomachs. Percent of Total Biomass 60 50 30 20 10 48 Alewife .. . Crayfish - B. cederstroemi Small Yellow Perch ( 5 200 mm TL) Large Yellow Perch (>200 mm TL) 19 17 24 July 31 Week Ending 49 Diurnal Effects on Diet Diurnal changes in catch, foraging activity, and diet composition were apparent but showed no consistent pattern. In June, ration size was highest from 0200 to 0800, however, in July and August it peaked during the 0800 to 1400 period. Catch also varied at different times of day. In all three 24-hour samples the 0200-0800 interval yielded the highest catch followed by a precipitous decline from 0800-1400. These results were probably indicative of an increase in yellow perch activity at dawn making them more vulnerable to gill net capture during this period. Despite the variability associated with diurnal effects, the diet data collected during 24—hour sampling are similar to seasonal diet data. Data from both sampling methods show throughout May, June, and July, yellow perch consume far more alewife than any other prey species. In August the proportion of the diet changed with time of day and the decline of alewife in the diet was apparently offset by an increase in predation on crayfish and B. cederstroemi. Because diurnal changes in catch and foraging activity were unpredictable, a clear determination of the best time of day to collect yellow perch was not possible. Consequently, collection of yellow perch for diet analysis should probably be made at hourly intervals over each 24- hour sampling period. Such a sampling method would minimize quantitative biases incurred from differential rates of prey digestion. Twenty-four hour sampling would be an essential 50 element of a quantitative diet analysis that might also include bomb calorimetry of prey types, laboratory feeding studies to determine evacuation rates for different prey types, and bioenergetics modeling. Effects of Substrate on Diet Although no studies on crayfish abundance were conducted, the sandy substrate at the lampara net sampling site was presumably unsuitable for crayfish. However, crayfish were frequently observed in shallow waters of the gill net study site. The absence of crayfish in the diets of yellow perch collected at the sand only site was apparently countered by a substantial increase in alewife predation in those fish. While alewife contributed only 16.5% to the prey biomass at the gill netting sites, alewife constituted 55.4% of prey biomass at the lampara net site. At both locations, crayfish and alewife together made up 55% to 60% of the diet biomass, but in different proportions. Studies by Werner and Hall (1977) point out that the value of a prey depends not only on the energy acquired from it, but also the energy expended in capturing it. As alewife abundance dropped at the gill netting sites in late summer, it is likely that yellow perch had to expend more energy to capture them than in early summer when they were abundant. Thus, yellow perch over rocky substrates could begin preying on crayfish when alewife availability 51 declined. Yellow perch over sand only substrate would not have this option. Consequently, yellow perch occupying sandy areas would be at a competitive disadvantage unless alewife abundance remained high, or other alternative prey types became available. Data from the lampara net samples suggest that both of these circumstances may occur. In addition to their greater alewife consumption, yellow perch collected over the sand substrate contained more than twice the biomass of "other" prey types than those collected over the rock substrate. Analysis of the "other" prey category showed that bloaters and smelt constituted 22.4% and 2.4%, respectively, of the prey biomass in the lampara net samples. In contrast, yellow perch diets at the rocky site, contained only 10% bloater and 1.7% smelt. Furthermore, the percentage of bloater found in the diets of yellow perch collected over the sand substrate was much higher than was found at any time in yellow perch collected over rock substrate. The data from these samples have two important implications. First, they show that although alewife is the most important prey for yellow perch in both habitats, it is especially important in sandy habitats where crayfish are not available. Secondly, these data suggest that Lake Michigan yellow perch may be even more dependent on alewife than this study has shown. Because rocky substrate is rare in Lake Michigan, competition may limit yellow perch abundance in 52 these habitats. As density dependent factors intensify in rocky habitats, yellow perch may switch to foraging in sandy areas where alewife may comprise an even higher percentage of the diet. Sex Ratio High ratios of males:females in spring catches of yellow perch have been observed in several studies. In 1973, Brazo found spring sex ratios of yellow perch to be 8 males:1 female. He offers the explanation that differential behavior of males and females might be responsible. This explanation seems plausible since Harrington (1947) found that large numbers of ripe males often follow individual females. El-Zarka (1959) also noted that male yellow perch arrived at spawning areas in Saginaw Bay several days before females. Although these findings may explain skewed sex ratios in early spring catches, they do not explain why sex ratios might remain skewed throughout the spawning season, as was the case in this study. These uncertainties are compounded by Tsai and Gibson (1971) who reported a sex ratio of 1:1 for spawning yellow perch. A second behavioral explanation may be that actual spawning took place in areas away from our gill nets, such as rock pilings, breakwaters, and CPCo barrier 1’1th structuresz. 2Because of its immense size (over two miles in length) the CPCo barrier net attracts large schools of spawning yellow perch. Each spring CPCo divers have observed numerous yellow 53 Studies by LeCren (1958) and Muncy (1962) may offer yet another explanation. Both studies report that male yellow perch mature by age 2, but that females do not mature until age 3. Consequently, the proportion of mature males in the population might be greater than that of mature females. Since only mature fish move inshore to spawn, it may be reasonable to expect a larger proportion of males in spring catches. Limitations of the Data The length frequency distribution of yellow perch collected in 1991 (Figure 12) shows that size classes larger than 150 mm TL are well represented. However, despite the use of mesh sizes designed to capture small fish (1.0" and 1.5" stretch), yellow perch less than 150 mm TL were rarely caught in 1991. Although ineffective in capturing yellow perch less than 150 mm TL, the small mesh sizes were extremely effective in capturing small non target species, such as alewife, bloaters, and spottail shiners. Because of the time required to remove these species from the gill nets, small mesh sizes were not used in 1992. Consequently, size classes smaller than 180 mm TL are virtually absent from the 1992 data. The scarcity of small yellow perch in the 1991 catch perch egg masses deposited on the net. Impacts on local yellow perch populations from egg losses incurred during routine net cleaning and maintenance have not been studied. 54 Figure 12. Length frequency distribution of yellow perch caught near Ludington, Michigan, 1991 and 1992. Number Sampled Number Sampled 150 140 130 120— 110 100 90 80 70 50 40 30 10 55 1991 (CPCo Gill Nets) 100 120 140 160 180 200 220 240 260 280 300 320 340 360 n=2292 1992 (MSU Gill Nets) 100 120140160180200220240260280300320340360 Total Length (mm) 56 can probably be explained in two ways. First, it is likely that small yellow perch were not as susceptible to the gear as larger yellow perch, even though small mesh sizes were used. Size selectivity is a known bias of gill netting, although variable mesh gill nets help minimize, but do not eliminate, this problem (Nielsen and Johnson 1983). However, if this was the only reason for the Scarcity of small yellow perch in the 1991 catch, then small fish of other species would not have been caught either. In light of the numerous small non-target fishes gill netted in 1991, the scarcity of small yellow perch in the 1991 catch can not be sufficiently explained by gear selectivity alone. As discussed earlier, 1991 catch data showed that most yellow perch captured in the study area were mature. Although a week year class in 1990 could account for these findings, a more probable explanation is that most juvenile yellow perch inhabit other areas of the lake. If this is the case, then diet data on small yellow perch collected from the study area may not be representative of most juveniles. Impacts of B. cederstroemi Since its accidental introduction into Lake Michigan in 1987 (Evans 1988), B. cederstroemi has become a serious concern to the future of Great Lakes fisheries. A number of recent studies (Barnheisel 1991, Baker 1992, and Schneeberger 1991) have shown that yellow perch below 100 mm 57 TL may avoid B. cederstroemi due to its protective spines. Diet data presented here show that B. cederstroemi composes a significant portion of the food consumed by yellow perch between 100-200 mm TL. While this may suggest that B. cederstroemi is beneficial to this particular size class, some biologists believe that the caudal spines of B. cederstroemi may cause intestinal damage in some species. While many yellow perch collected in late summer, had consumed large quantities of B. cederstroemi, it appeared that the tailspines were able to pass through the guts as easily as other prey items. As omnivores, yellow perch may have a less specialized gastrointestinal anatomy than other fish species. Consequently, yellow perch may be better able to digest B. cederstroemi. However, when B. cederstroemi is consumed in large quantities, its long chitinous tailspines occupy a large proportion of the available volume in fish stomachs. Because chitin has no caloric value, B. cederstroemi may be a less valuable food source by volume than other prey species. Ippacts on Alewife The frequency distribution of alewife consumed (Figure 13) shows that most of the alewife eaten were between 80 and 130 mm TL. Alewife age data (Smith 1956) show that this size class is made up primarily of yearlings although some two-year-olds may also be included since the distribution 58 Figure 13. Length frequency distribution of alewife consumed by yellow perch, 1992. Number Consumed HHHHHHI—IHHHSN cwuuauaqme H uuhuaqae 59 60 70 80 90 100 110 120 130 140 150 160 170 Alewife Total Length (mm) 180 60 was constructed using alewife consumed over the entire summer. Consequently, large yearlings eaten in August may have grown enough to overlap with the distribution of two- year-olds eaten in May. Figure 14 shows total lengths of yellow perch plotted against the total lengths of alewife each had consumed. The absence of alewife smaller than 60 mm TL indicates an important characteristic of the alewife forage base that may be important to yellow perch. Forage studies on Lake Michigan alewife have shown that most young— of—the-year (YOY) alewife normally reach a total length of 51—75 mm TL by fall (Smith 1956). Much of the size variability in YOY alewife is often lost in the following year because winter survival drops sharply for YOY alewife less than 60 mm TL (Brown 1972). High winter mortality of YOY alewife less than 60 mm TL removes them from the forage base until fall, when the next generation of YOY alewife have grown large enough to become suitable prey. Since yellow perch below 200 mm TL are probably too small to eat alewife larger than 60 mm TL, alewife are not available to them throughout most of the summer. This may explain why yellow perch must reach a critical length before they begin preying on alewife. Nonetheless, some YOY alewife should have grown large enough to become suitable prey for yellow perch by late fall. Although a few unidentified larval fish were found in yellow perch stomachs, they constituted a very small percentage of the diet (less than 0.1%). 61 Figure 14. Total lengths of yellow perch vs total lengths of alewife consumed by each, 1992. fe Total Length (mm) Alewn 62 190 + 180— _ + 170 + + 160— + ++ i + + 150— + + + + + + 140— + + + 130— + + + + + + + + ++ + ++ + + + + 120— ++ + + ++ + ++ ++ 110— ++ ti + + + + + ++ ++ i++ + 1+ + + + 100— #1 ++++++ ++ +++ + +1; +1 ++ +I+ + + 90— + is + ++ 71+: 1 +I + leafs:r f + + +++ + + +++ ++ ++ ++ $ 80— 1 ++ ++if + + ++ + me + + 60 l l l l l l l l l l | l l l l l l 200 210 220 230 240 250 260 270 280 290 300 310 320 330 340 350 360 Yellow Perch Total Length (mm) 63 The lack of YOY alewife found in yellow perch diets probably results from vertical separation of yellow perch and YOY alewife. Life history studies of Lake Michigan alewife (Norden 1967) show that YOY alewife are extremely size selective when feeding. According to Norden (1967), YOY alewife prey primarily on small copepods and cladocerans that congregate near the surface. Depth distribution of the CPCo yellow perch catch showed that nearly all yellow perch collected at the 12 m stations (90% in 1991 and 93% in 1992) were caught within 2 m of the lake bottom (CPCo 1992, CPCo 1993). Although yellow perch are known to move off the bottom when foraging (Scott and Crossman 1973), it is doubtful that they ever venture close enough to the surface to encounter YOY alewife. Although older alewife also feed on zooplankton, they spend much more time at mid depths and near the bottom where they feed on benthic invertebrates as well (Morsell and Norden 1968, Scott and Crossman 1973). By the end of their second year alewife are probably too large to be preyed upon by most yellow perch. Consequently, age-l (60-130 mm TL) alewife are preyed upon most heavily by yellow perch. Although the effects of yellow perch predation on alewife are probably affecting other alewife predators as well, it is difficult to prove that competition is actually occurring. Brazo (1973) and Hauer (1974) both showed that sculpin, not alewife, were the primary forage fish of yellow 64 perch near Ludington in the early 1970s. In 1993 sculpin made up only 2.1% of yellow perch diets as compared with 63.8% for alewife. This apparent change in diet may be an important factor inhibiting the recovery of Lake Michigan alewife populations since their decline in the 1980’s. Because alewife are also the primary forage of many Lake Michigan salmonids, recent increases in yellow perch populations may be adversely affecting salmonid production. Given the extent of alewife predation observed in this study, fisheries management aimed at restoring and maintaining Lake Michigan's sport fishery may need to consider the effects of yellow perch diet on future production of other alewife dependent gamefish. SUMMARY Alewife populations in Lake Michigan have not recovered as expected since their decline in the mid 1980’s. Reduced alewife abundance is believed to be an important stress factor contributing to recent increases in mortality of chinook salmon infected with bacterial kidney disease. Alewife declines may also have allowed yellow perch populations to increase over the last ten years, because adult alewife compete with and prey on larval yellow perch. Diet data from yellow perch collected near Ludington in the early 1970’s indicated that alewife were not a major prey species of yellow perch. However, diet analysis of 2,544 yellow perch gill netted from Lake Michigan in 1992 shows that alewife currently comprise 65.8% of the total prey biomass consumed during the growing season. Throughout most of the summer, including the weeks of highest foraging intensity, yellow perch greater than 200 mm TL preyed almost exclusively on alewife. Based on percent biomass and frequency of occurrence, alewife was the most important prey species for yellow perch greater than 200 mm TL. After it became abundant in late July, B. cederstroemi was preyed upon by nearly all yellow perch sampled. However, no yellow perch under 100 mm TL were captured and other studies have shown that B. ggggpgpppgmi are not eaten by yellow perch smaller than 100 mm TL. Although most 65 66 yellow perch larger than 200 mm TL consumed some B. cederstroemi, it contributed a relatively small percentage to prey biomass in those fish. However, B. cederstroemi comprised 49 to 84 percent of the prey biomass consumed by yellow perch 150-200 mm TL. indicating that it may be an important food source for yellow perch in that size range. Because of their yellow perch are among the most abundant alewife predators in Lake Michigan, it is likely that they are having negative impacts on alewife populations. If shortages of alewife are inhibiting the recovery of chinook salmon in Lake Michigan, as many biologists believe, it may be necessary to consider the ecological impacts of yellow perch in future fisheries management plans. REFERENCES 67 LIST OF REFERENCES Baker E. L. 1992. Evidence for yellow perch predation on Bythotrephes cederstroemi in southern Lake Michigan. Journal of Great Lakes Research. 18(1): 190-193. Barner, G. H. 1993. Data on zooplankton populations near Ludington, Michigan. Unpublished. Barnheisel, D. R. 1991. Zooplankton spine induces aversion in small fish predators. Oecologia, 88: 444-450. Brazo, D. C. 1973. Fecundity, food habits, and certain allometric features of the yellow perch, Perca flavescens (Mitchell), before operation of a pumped storage Plant on Lake Michigan. Masters thesis. Mich. State Univ. 75 pp. Brown, E. H. Jr. 1972. Population biology of alewives, Alosa pseudoharengus, in Lake Michigan, 1949-1970. Journal of the Fisheries Research Board of Canada 29: 477-500. Christie, W. J. 1974. Changes in the fish species composition of the Great Lakes. Journal of the Fisheries Research Board of Canada 31(5): 827- 854. Consumers Power Company and The Detroit Edison Company, 1992. Ludington Pumped Storage Project 1991 annual report of barrier net operation. Consumers Power Company. Consumers Power Company and The Detroit Edison Company, 1993. Ludington Pumped Storage Project 1992 annual report of barrier net operation. Consumers Power Company. Eck, G. W. and L. Wells. 1987. Recent changes in Lake Michigan’s fish community and their probable causes, with emphasis on the role of the alewife (Alosa pseudoharengps). Canadian Journal of Fisheries and Aquatic Science 44(2): 53-60. El-Zarka, S.E.D. 1959. Fluctuations in the 68 population of yellow perch, Perca flavescens (Mitchell), in Saginaw Bay, Lake Huron. U.S. Fish and Wildlife Service Bulletin. 59: 365-415. Evans, M. S. 1988. Bythotrephes cederstroemi: its new appearance in Lake Michigan. Journal of Great Lakes Research 14: 234-240. Graham, J.J. 1956. Observations on the alewife, Pomolbus pseudoharengus (Wilson), in fresh water. Ontario Fisheries Research Laboratory. Harrington, R. W. 1947. Observations on the breeding habits of the yellow perch, Perca flavescens (Mitchell). Copeia 1937: 199-200. Hatch, R.W., P.M. Haack, and E.H. Brown, Jr. 1981. Estimation of alewife biomass in Lake Michigan, 1967-1978. Transactions of the American Fisheries Society. 110: 575-584. Hauer, R. F. 1974. Comparison of preoperational and operational abundance, age, length-weight relationships, and food habits of yellow perch, Perca flavescens (Mitchell), in Lake Michigan at the Ludington Pumped Storage Reservoir. Horn, B. 1992. Personal communication at the 1992 Great Lakes Littoral Fisheries Research Committee Meeting. Michigan City, Indiana. Jude, K.J., and F.J. Tesar. 1985. Recent changes in the inshore forage fish of Lake Michigan. Canadian Journal of Fisheries and Aquatic Sciences. 42: 1154-1157. Keitly, T.J. 1990. Evidence for alewife (Alosa pseudoharengus) predation on the european cladoceran Bythotrephes cederstroemi in Northern Lake Michigan. Journal of Great Lakes Research 16: 330-333 Lecren, E.D. 1958. Observations on the growth of perch (Perca fluviatilis L.) over twenty-two years with special reference to the effects of temperature and changes in population density. The Journal of Animal Ecology. 27(2): 287-334. Lehman, J. T. 1987. Palearctic predator invades North American Great Lakes. Oecologia 74: 478-480. 69 Liston, C. R. and P. I. Tack. 1972. A study of the effects of installing and operating a large pumped storage project on the shores of Lake Michigan near Ludington, Michigan. Mich. State Univ. Final Report to Consumer's Power. 113 pp. Mills, E. L., J. Leach, J. Carlton, and C. Secor. 1993. Exotic species in the Great Lakes: a history of biotic crises and anthropogenic introductions. Journal of Great Lakes Research 19(1): 1-54. Morsell, J. W., and C. R. Norden. 1968. Food habits of the alewife, Alosa pseudoharengus (Wilson), in Lake Michigan. Journal of Great Lakes Research. 1968: 96-102. Muncy, R.J. 1962. Life History of the yellow perch (Perca flavescens) in esturine waters of Severn River, a tributary of Chesapeake Bay, Maryland. Chesapeake Science. 3(3): 143-159. Nielsen, L. A. and D. L. Johnson. 1983. Fisheries Techniques. American Fisheries Society. 468 pp. Norden, C. R. 1967. Morphology and food habits of the larval alewife, Alosa pseudoharengus (Wilson), in Lake Michigan. Proceedings of the 11th Conference of Great Lakes Research, 1967. p. 70-78. International Association of Great Lakes Research. Otto, R. G., M. A. Kitchell and J. O. Rice. 1976. Lethal and preferred temperatures of the alewife (Alosa pseudoharengps) in Lake Michigan. Transactions of the American Fisheries Society. 105(1): 96-106. Scavia, D., Lang, G.A., and Kitchell, J.F. 1988. Dynamics of Lake Michigan plankton: a model evaluation of nutrient loading, competition, and predation. Canadian Journal of Fisheries and Aquatic Sciences. 45: 165-177. Schneeberger, P. J. 1991. Seasonal incidence of Bythotrephes cederstroemi in the diet of yellow perch (ages 0-4) in Little Bay De Noc, Lake Michigan, 1988. Journal of Great Lakes Research 17(2): 281-285. Scott, W. B. and E. J. Crossman. 1973. Freshwater fishes of Canada. Bulletin 184,Fisheries Research Board of Canada, Ottawa, Canada. 70 Smith, S. H. 1956. Life history of Lake herring of Green Bay, Lake Michigan. U.S. Fish & Wildlife Service. Fisheries Bulletin. 57: 87-138. Smith, S. H. 1970. Species interaction of the alewife in the Great Lakes. Transactions of the American Fisheries Society. 99(4): 754-765. Stewart, D.J., J.F. Kitchell, and L.B. Crowder. 1981. Forage fishes and their salmonid predators in Lake Michigan. Transactions of the American Fisheries Society. 110: 751-763. Tsai, C. and G.R. Gibson. 1971. Fecundity of the yellow perch (Perca flavescens), in the Patuxent River, Maryland. Science 12(4): 270-274. Wells, L. 1977. Changes in Yellow Perch (Perca flavescens) populations of Lake Michigan, 1954- 1975. Journal of the Fisheries Research Board of Canada. 34: 1821-1829. Wells, L. and R.W. Hatch. 1984. Status of bloater chubs, alewives, smelt, slimy sculpins, deepwater sculpins, and yellow perch in Lake Michigan, 1983, p. 29-36. In Appendices to Lake Michigan Committee 1984 Annual Meeting Minutes. Great Lakes Fishery Commission, Ann Arbor, MI. Werner, E. E. and D. J. Hall. 1977. Species packing and niche complementarity in three sunfishes. American Naturalist 111: 553-578. "Illilllllllllllllllllill“