5. «fax wxg... . #1... mafia. www.mufi. . mfiwzr .33."? 5:94 .. S 1. .599 a. 2 Rafi“. .wfihw. .. . if a... (v1; .2: .. nu; . .rnsfi; arm“; @afixw a $339} . 5. Ian... . .ék;n 3. e : . . . .w a; .13 #gxwwga . . d a (A. 9‘... Swan I tin: ”flank fiafluk ‘ I .3: . 512.9; 2:}. . I (r... x... rfimfi. .4 z i. r . 3... via... .\. tn LIBRARIES - MICHIGAN STATE UNIVERSITY EAST LANSING, MICI-I 48824-1048 . j J - Wang This is to certify that the dissertation entitled Foraging Ecology, Bioenergetics and Predatory Impact of Breeding Double-crested Corrnorants (Phalacrocorax aun'tus) in the Beaver Archipelago, Northem Lake Michigan presented by Nancy E. Seefelt has been accepted towards fulfillment of the requirements for the Ph. D degree in ZOLIOQL fVZfflf Date MSU is an Affinnative Action/Equal Opportunity Institution PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE FEB 1 2 2009 2/05 c:7§lm95u-;.lndd-p.15 _ _ —_._._—_-._—-_.__- FORAGING ECOLOGY, BIOENERGETICS AND PREDATORY IMPACT OF BREEDING DOUBLE-CRESTED CORMORANTS (PHAIACROCORAXA URITUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN By Nancy E. Seefelt A DISSERTATION Submitted to Michigan State University in partial fiilfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Zoology 2005 ABSTRACT F ORAGING ECOLOGY, BIOENERGETICS AND PREDATORY IMPACT OF DOUBLE-CRESTED CORMORANTS (PHALA CROCORAX A URI T US) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN By Nancy E. Seefelt As Double-crested Cormorant (Phalacrocorax auritus) populations have increased throughout the Great Lakes, many sport and commercial fish populations have declined. A high density of birds combined with their fish eating habits has led to their implication in these declines. From 1999 through 2002, a Smallmouth Bass (Micropterus dolomieui) population study was rekindled after a twenty plus year hiatus in the Beaver Archipelago. This work documents an intensive study on the population dynamics and foraging ecology of breeding cormorants of the same area between 2000 and 2004. The population size of breeding birds, well as their reproductive output, was estimated and the diet of birds was determined using regurgitates and the stomachs of harvested birds. Pellets were not used in analysis because they did not provide reliable quantitative data regarding bird diet. Breeding population size, as well as reproductive output, appears to vary substantially from year to year, and may be linked to the availability of alewife (Alosa pseudoharengus). Breeding bird diet consists primarily of species of little commercial or sport value; the importance of individual prey species in bird diet varies temporally and spatially. To determine important foraging areas, VHF radio telemetry was used to track the foraging activities of ten cormorants. Using triangulation, birds were monitored from both land and water daily, weather permitting, throughout the breeding season. In addition, rafting locations of cormorants were documented by boat survey throughout the breeding season. Radiotelemetery indicated that cormorants typically foraged 2.5 km for away from the colony, at the northeastern end of Beaver Island. This area overlaps with the area determined by rafting locations, however the latter were centered further south. Neither area overlaps with known bass habitat. Colonial waterbirds are an important component of Great Lakes ecosystems. One important aspect is the role of these birds as top predators in aquatic food webs. In order to investigate this role, bioenergetics models, using allometric equations, were applied to breeding cormorants and their offspring in the study area. The models estimated the total prey biomass consumed as 1444.11 tonnes of prey in 2000, and 1586.17 tonnes of prey in 2001. Each year the majority of the prey biomass was aewife, with these fish comprising a greater percentage of prey biomass in 2001. In addition, two types of simulation models, Rebuild and Forecast, to investigate a possible connection between the declining bass populations and the avian predators. Rebuild models indicate that is was unlikely that cormorants alone caused the observed bass decline; birds may have contributed to the decline in synergism with low recruitment, angling mortality or other factors. Forecast models suggest that direct cormorant predation on bass is not currently the leading factor limiting the bass population size in the region, even if bass are experiencing relatively low recruitment. Combined with the data collected on bird diet, foraging locations and bioenergetics, these models enhance our understanding of the relationships between cormorants and fish populations in the study area and may benefit fisheries managers in other systems. DEDICATION This work is dedicated to my family and fiiends. .. and the natural beauty that abounds in wild places. iv ACKNOWLEDGEMENTS I would like to take this opportunity to thank my committee for their advice, patience and encouragement. Special thanks to my major advisor, Dr. Don Hall, who has not only been a wonderfitl mentor throughout my stay at Michigan State University, but is perhaps the greatest thinker I have ever known. His humor and his guidance are greatly appreciated. I would also like to thank Dr. James Bence, especially for his expertise in fish modeling dynamics and his willingness to help me in my efforts. Also, I‘ would like to thank both Dr. Tom Getty and Dr. Gary Mittelbach for their time, commitment and comments regarding this work. Great thanks and acknowledgment to Dr. James C. Gillingham of Central Michigan University; without his support and effort, this work would not have been possible. Special thanks to the many members of the cormorant field and lab crew at the Central Michigan University Biological Station on Beaver Island, who were always ready to lend a hand. Many thanks go to my lab mates, Mary Martin and Carrie Scheele, who were always helpful and offered me a most precious gift, their fiiendship. In addition, I would like to acknowledge Michigan State . University, Central Michigan University, the Michigan Department of Natural Resources and the US. Fish and Wildlife Service for funding this work. Thanks also go to Jory Jonas of the Michigan Department of Natural Resources — Fisheries Division for her advice and support throughout this study. Additionally, I would like to thank my parents, Robert and Sylviann Seefelt for their encouragement and interest in my work. Finally, I would like to thank my husband, Jeffrey A. Scofield, and my dog, Motega, for assistance in the field, companionship and unyielding support of my endeavors. TABLE OF CONTENTS LIST OF TABLES ................................................................................. viii LIST OF FIGURES ................................................................................. xi CHAPTER 1 THE DOUBLE-CRESTED CORMORANT IN LAKE MICHIGAN: A REVIEW OF POPULATION TRENDS, ECOLOGY AND CURRENT MANAGEMENT ............... 1 Introduction .................................................................................... 2 History of Occurrence and Population Trends ........................................... 3 Ecology .......................................................................................... 5 Current Management ........................................................................ 8 Conclusion .................................................................................. 11 Summary .................................................................................... 11 Acknowledgements ........................................................................ 1 1 References ................................................................................... 12 CHAPTER 2 A COMPARISON OF THREE METHODS TO INVESTIGATE THE DIET OF BREEDING DOUBLE-CRESTED CORMORANTS (PHALA CROC ORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN .................. 15 Abstract ...................................................................................... 16 Introduction .................................................................................. 17 Materials and Methods .................................................................... 20 Study Area ......................................................................... 20 Pellets and Regurgitates ............................................................ 21 Stomach Contents ................................................................. 22 Analysis ............................................................................ 22 Results ....................................................................................... 23 Discussion ................................................................................... 25 Conclusions ................................................................................. 30 Acknowledgements ........................................................................ 3 1 Literature Cited ............................................................................. 32 CHAPTER 3 POPULATION ESTIIVIATES AND DIETARY EVALUATION OF DOUBLE- CRESTED CORMORANTS (PHALACROCORAX A URI TUS) BREEDING IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN ............................ 46 Abstract ...................................................................................... 47 Introduction ................................................................................. 48 Methods ........................................................................................ 50 Study Area .......................................................................... 50 Population Estimates .............................................................. 51 Diet Estimates ..................................................................... 53 Analysis ............................................................................ 55 Results ....................................................................................... 57 Discussion ................................................................................... 62 Literature Cited ............................................................................. 68 CHAPTER 4 USING RADIOTELEMETRY AND RAF TING LOCATIONS TO DETERMINE FORAGING LOCATIONS OF DOUBLE-CRESTED CORMORANT (PHALA CROCORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN ................................................................ 84 Abstract ...................................................................................... 85 Introduction ................................................................................. 86 Methods ...................................................................................... 88 Study Site ........................................................................... 88 Telemetry ........................................................................... 89 Rafting Locations .................................................................. 91 Analysis ...................................................................................... 91 Results ....................................................................................... 92 Discussion ................................................................................... 94 Literature Cited ............................................................................. 97 CHAPTER 5 BIOENERGETICS AND PREY CONSUMPTION OF BREEDING DOUBLE- CRESTED CORMORANTS (PHALA CROC ORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN ....................................... 106 Abstract .................................................................................... 107 Introduction ................................................................................ 108 Methods .................................................................................... 110 Study Area ........................................................................ 110 Model Construction ............................................................. l 1 1 Sensitivity Analysis ............................................................. 114 Results ...................................................................................... 1 15 Discussion ................................................................................. 116 Literature Cited ........................................................................... 121 CHAPTER 6 SIMULATION MODELS TO INVESTIGATE THE POTENTIAL IIVIPACT OF DOUBLE-CRESTED CORMORANTS ON A SMALLMOUTH BASS POPULATION INTI-IE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN... ....134 Abstract .................................................................................... 135 Introduction ................................................................................ 136 General Methods .......................................................................... 140 Study Area ........................................................................ 140 Basic Model Description ....................................................... 140 Rebuild Model ............................................................................ 145 Baseline Model Description .................................................... 145 vii Baseline Model Analysis ....................................................... 146 Baseline Model Results ......................................................... 147 Forecast Model ............................................................................ 148 Baseline Model Description .................................................... 148 Baseline Model Analysis ....................................................... 149 Baseline Model Results ......................................................... 149 Sensitivity Analysis ...................................................................... 150 Tests for Model Robustness ............................................................. 151 Rebuild Model — Simulation Experiments .................................... 151 Forecast Model — Simulation Experiments ................................... 154 Experimental Simulation Analysis ............................................ 156 Experimental Simulation Results .............................................. 157 Discussion ................................................................................... 162 Conclusions... . ............................................................................ 171 Literature Cited ........................................................................... 173 LITERATURE CITED ........................................................................... 208 viii LIST OF TABLES CHAPTER 1 THE DOUBLE-CRESTED CORMORANT IN LAKE MICHIGAN: A REVIEW OF POPULATION TRENDS, ECOLOGY AND CURRENT MANAGEMENT Table 1. Lake Michigan and Great Lakes Double-crested Cormorant population estimates beginning in 1977 and ending in 1997. Both the number of breeding pairs and active colonies have increased substantially over this twenty year period. In 1997, the Lake Michigan colonies comprised 57.55% of the breeding cormorants in the US. waters of the Great Lakes ...................................................................................... 5 Table 2. Numbers of Double-crested Cormorant pairs breeding at colonies in the Beaver Archipelago, northern Lake Michigan beginning in 1984 and ending in 2001. Note that the number of breeding pairs increases until 1997, where after there is an overall population decline. Some colonies, such as Pismire and Gull, have continued to show an increase as the regional population declines, while Timm’ s and Whiskey colonies have dissolved ............................................................................................. 5 CHAPTER 2 A COMPARISON OF THREE METHODS TO INVESTIGATE THE DIET OF BREEDING DOUBLE-CRESTED CORMORANTS (PHALA CROCORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Table 1. 3X2 contingency table showing the actual (and expected) values of the numerical frequency data for Alewife (A losa pseudoharengus) and crayfish (Orconectes) in pellets, regurgitates and stomachs. Chi- Square Tests indicate that the values differ from expected for both Alewife (x2 = 387.06, critical value= 5.999 at d= 0.05, df= 2) and crayfish (,6— -— 119 02, critical value= 5 999 at a= o 05, df= 2) ....................... 37 Table 2. Number of analyzed samples were individual prey items were found in pellets, regurgitates and stomachs (n = 60 for pellets, n = 44 for regurgitates, and n = 25 for stomachs). The percent fiequencies of each prey item are shown in parenthesis. . . . . . 38 Table 3. 3X2 contingency table showing the actual (and expected) values of the number of samples that contained Alewife (A 1030 pseudoharengus) and crayfish (Orconectes) in ‘ pellets, regurgitates and stomachs. Chi-Square Tests for 3x2 comparison indicate that the values difi‘er from expected for both Alewife (x2: 9. 53, cn'tical value= 5.999 at d= 0.05, df= 2) and crayfish (x2= 6. 46, critical value= 5.999 at a= 0.05, df= 2). Comparisons using 2x2 contingency tables show that pellets and stomach contents data for both Alewife (,8- - 3 35, critical value= 3 34 at a= o 05, df= 1) and crayfish (18- -— 1 so, critical value= 3. 84 at a= 0.05, df= 1) were statistically similar. Also, regurgitate and stomach contents data were statistically similar for both Alewife (x2: 0.61, critical value ix = 3.84 at a = 0.05, df= 1) and crayfish (x2 = .074, critical value = 3.84 at a =—- 0.05, df= l) ..................................................................................................... 39 CHAPTER 3 POPULATION ESTIMATES AND DIETARY EVALUATION OF DOUBLE- CRESTED CORMORANTS (PHALACROCORAX A URI TUS) BREEDING IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Table 1. Population estimates of breeding Double-crested cormorants in the Beaver Archipelago, northern Lake Michigan from 1984 to 2004 ................................. 72 CHAPTER 5 BIOENERGETICS AND PREY CONSUMPTION OF BREEDING DOUBLE- CRESTED CORMORANTS (PHALA CROCORAXA URITUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Table 1. Life history characteristics ofDouble—crested Cormorants used to model prey consumption by breeding adult and young birds in the Beaver Archipelago, northern Lake Michigan .................................................................................. 125 Table 2. Proportions of the biomass of individual prey items consumed by breeding adult and young cormorants for the pre-breeding/incubation, nestling, and post-breeding time periods for 2000 and 2001 ..................................................................... 126 Table 3. Average calorie density (kcal/kg) for prey species ............................. 127 Table 4. Total DFC (daily food consumption) in kilograms for adults during each period of the breeding season for 2000 and 2001 .................................................. 128 , Table 5. Total seasonal food consumption (kg) per pre-fledged and post-fledged chicks for 2000 and 2001 .............................................................................. 129 Table 6. Sensitivity analysis .................................................................. 130 CHAPTER 6 SIMULATION MODELS TO INVESTIGATE THE POTENTIAL IMPACT OF DOUBLE-CRESTED CORMORANTS ON A SMALLMOUTH BASS POPULATION IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Table 1. Model Parameters for Baseline Rebuild and Forecast Models ................ 178 Table 2. Sensitivity analysis results for the Baseline Rebuild model showing the influence of each model parameter on the final adult smallmouth bass population size ................................................................................................ 179 ' Table 3. Sensitivity analysis results for the Baseline Forecast model showing the influence of each model parameter on the final adult smallmouth bass population size ................................................................................................ 180 . Table 4. The parameters for the four different scenarios created from the Rebuild Model that allow for varying cormorant predation rates across immature bass age classes. (Cormorant Predation Rate is the proportion of fish in each age class removed per month of the breeding season by birds.) .............................................................. 181 Table 5. The parameters used to construct the two Forecast Constant Predation Pressure Scenario models, where cormorant predation rates remain constant throughout the similulation. (Cormorant Predation Rate is the proportion of fish in each age class removed per month of the breeding season by birds.) ..................................... 182 Table 6. The cormorant predation mortality rates used in the modified Forecast Constant Predation Pressure Scenario models to examine how increase cormorant predation on immature bass due to greater bass availability could influence adult population size. (Cormorant Predation Rate is the proportion of fish in each age class removed per month of the breeding season by birds.) ............................................................. 183 - Table 7. The parameters used for the construction of the Forecast Increased Predation Pressure Scenarios. (Cormorant Predation Rate is the proportion of fish in each age class removed per month of the breeding season by birds.) ..................................... 184 xi LIST OF FIGURES CHAPTER 1 THE DOUBLE-CRESTED CORMORANT IN LAKE MICHIGAN: A REVIEW OF POPULATION TRENDS, ECOLOGY AND CURRENT MANAGEMENT Figure 1. Diet of Double-crested Cormorants in the Beaver Archipelago, 2000 and 2001 combined, shown as percent mass for both regurgitate (n = 1128) and stomachs (n = 150). Alewife (Alosa) provide the most biomass in cormorant diets. For regurgitate samples, Alewife comprise 57,073 g of 79,230 g, while for stomach samples, Alewife comprise 18,603 g of 25,550g. However, crayfish (Orconectes), sculpin (Corals), stickleback (Pungr'n'us) and sucker (Catostomus) are also frequently taken ............................ 8 CHAPTER 2 A COMPARISON OF THREE METHODS TO INVESTIGATE THE DIET OF BREEDING DOUBLE-CRESTED CORMORANTS (PHALA CROC ORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Figure 1. Pellet numerical frequency data showing the diet of Beaver Archipelago cormorants as percentages ....................................................................... 40 Figure 2. Regurgitate numerical frequency data showing the diet of Beaver Archipelago cormorants as percentages ....................................................................... 41 Figure 3. Stomach contents numerical frequency data showing the diet of Beaver Archipelago cormorants as percentages. Percent values for Catostomus and Notropis are small (both 0.23%) were combined for clarity ................................................ 42 Figure 4. Regurgitate biomass data showing the diet of Beaver Archipelago cormorants as percentages ..................................................................................... 43 Figure 5. Stomach contents biomass data showing the diet of Beaver Archipelago cormorants as percentages. Percent values for Pungitius, Etheostoma, and Nolropis are small (1.48%, 1.43% and 1.48%, respectively) were combined for clarity ............... 44 Figure 6. Means of arcsine transformed biomass data. The Mann-Whitney Test , indicated that biomass of Alewife (confidence intervals = 0.01 to 36.21, W = 1652.0, p = 0.1207, adjusted for ties) and crayfish (confidence intervals = - 36.21, -0.01, W = 1428.0, p = 0.1207, adjusted for ties) estimated by each method were not statistically significant fi'om each other at a = 0.05 ..................................................................... 45 xii CHAPTER 3 POPULATION ESTIMATES AND DIETARY EVALUATION OF DOUBLE- CRESTED CORMORANTS (PHALA CROCORAX A URI TUS) BREEDING IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Figure 1. The Beaver Archipelago of northern Lake Michigan ........................... 73 Figure 2. Number of active cormorant nest during the early, mid and late breeding season in 2000 and 2001, for a) Pismire, b) Grape, c) Hat, and d) Gull Island colonies. 74 Figure 3. Number of active cormorant nests counted during the early, mid and late breeding season from 2000 through 2004 at Pismire Island. Note that no nests remained at the end of the 2002 breeding season ......................................................... 75 Figure 4. a) Mean clutch size (with standard error) for all colonies during the early and mid breeding season in 2000 and 2001. Late clutch sizes in mid 2000 and 2001 are significantly different from each other. b) Mean clutch size in the mid breeding season for Pismire Island in 2000, 2001, 2003 and 2004 ............................................. 76 Figure 5. a) Mean number of successful breeding pairs (with standard errors) for all colonies during the 2000 and 2001 breeding seasons. b) Mean number of successful breeding pairs for Pismire Island in 2000, 2001, 2003 and 2004 ........................... 77 Figure 6. a) Estimates for the number of chicks produced for pairs breeding at all colonies during 2000 and 2001 breeding season. b) Estimates for the number of chicks produce for breeding pairs on Pismire Island in 2000, 2001, 2003 and 2004 ............. 78 Figure 7. Numerical frequency data showing the diet of Beaver Archipelago cormorants as percentages for 2000 and 2001 ............................................................... 79 Figure 8. Biomass data showing the diet of Beaver Archipelago cormorants as percentages, 2000, for all colonies, Main Archipelago colonies, and Gull Island colonies ............................................................................................. 80 Figure 9. Biomass data showing the diet of Beaver Archipelago cormorants as percentages, 2001, for all colonies, Main Archipelago colonies, and Gull Island colonies ............................................................................................. 81 Figure 10. Biomass data showing the diet of Main Archipelago cormorants as percentages, 2000, for Pre-nesting and Incubation, Nestling and Fledgling, and Post- nesting periods ..................................................................................... 82 xiii Figure 11. Biomass data showing the diet of Main Archipelago cormorants as percentages, 2001, for Pre-nesting and Incubation, Nestling and Fledgling, and Post- nesting periods ..................................................................................... 83 CHAPTER 4 USING RADIOTELEMETRY AND RAFTING LOCATIONS TO DETERMINE F ORAGING LOCATIONS OF DOUBLE-CRESTED CORMORANT (PHALACROC ORAX A URI T US) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Figure 1. The Beaver Archipelago of northern Lake Michigan ........................ 100 Figure 2. The bays of Garden Island and their proximity to Pismire and SE Garden Colonies ......................................................................................... 101 Figure 3. Bird locations determined by radiotelemetry and triangulation ............. 102 Figure 4. Weighted density contours for foraging areas as determined by radiotelemetry .......... . ........................................................................ 103 Figure 5. Locations of cormorant rafting sites as determined by boat surveys ........ 104 _ Figure 6. Weighted density contours for foraging areas as determined by boat survey ............................................................................................. 105 CHAPTER 5 BIOENERGETICS AND PREY CONSUMPTION OF BREEDING DOUBLE- CRESTED CORMORANTS (PHALACROCORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Figure 1. The Beaver Archipelago of northern Lake Michigan ........................ 131 Figure 2. a) Daily food consumption (g) for pro-fledged birds in 2000 and 2001, and h) daily food consumption (kg) of post-fledged chicks up to the time they attain adult mass in2000 and 2001....... ....................................................................... 132 Figure 3. Biomass of prey consumed (tones) by breeding cormorants and their young in - the Beaver Archipelago in 2000 and 2001 ................................................. 133 CHAPTER 6 SIMULATION MODELS TO INVESTIGATE THE POTENTIAL IMPACT OF DOUBLE-CRESTED CORMORANTS ON A SMALLMOUTH BASS POPULATION IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN Figure 1. The Beaver Archipelago of northern Lake Michigan ........................ 185 xiv Figure 2. Conceptual model depicting the smallmouth bass population ............... 186 Figure 3. The population size of age three and older smallmouth bass in Garden Island Harbor as determined by the Schnabel method (with standard deviations). Estimates for 1970s and 19803 from Lennon (unpublished data) and later estimates fi'om Seider (2003) ............................................................................................ 187 Figure 4. The overall trends in the population size of Double-crested Cormorants nesting in the Main Archipelago and when just Pismire and Grape colonies are combined. Curves are based on actual nest counts in 1984 (Ludwig 1984), 1989 (Scharf and Shugart‘ 1993), 1997 (Cuthbert et a1. 1997) and the current study (2000-2002) ................. 188 Figure 5. Results of the Baseline Rebuild model simulation showing a) the trends in the adult smallmouth bass population size and b) the percent biomass of bird diet composed of smallmouth bass for the entire Main Archipelago cormorant population and the Pismire and Grape population alone ......................................................... 189 Figure 6. Results of the Baseline Forecast model simulation showing a) the trends in the adult smallmouth bass population size and b) the biomass of smallmouth bass consumed by the cormorant population, over the length of the simulation .......................... 190 Figure 7. Results of the modified Rebuild model with no age four predation simulation showing a) the trends in the adult smallmouth bass population size and b) the percent biomass of bird diet composed of smallmouth bass for the entire Main Archipelago cormorant population and the Pismire and Grape population alone ..................... 191 . Figure 8. Results of the modified Baseline Rebuild model with spawning constant set at 0.40 showing the simulated trends in the adult smallmouth bass population size when all immature bass are subject to cormorant predation and when age four bass are invulnerable to cormorant predation ......................................................... 192 Figure 9. Results of the modified Baseline Rebuild model with spawning constant set at 0.40 showing the simulated trends percent biomass consumed by Main Archipelago birds, and Pismire and Grape birds when, a) all immature age classes are vulnerable to predation and b) when age four bass are invulnerable to predation ...................... 193 Figure 10. Simulated adult population size (means and standard deviations) for a) Rebuild Model Scenario 1 and b) Rebuild Model Scenario 2 ............................. 194 Figure 11. Experimental (means and standard deviations) percent biomass of bird diet composed of smallmouth bass as simulated by Rebuild Scenarios 1 and 2 for a) all Main . Archipelago birds, and b) only birds nesting on Pismire and Grape Islands ............ 195 XV Figure 12. Simulated adult population size (means and standard deviations) for a) Rebuild Model Scenario 3 and b) Rebuild Model Scenario 4 .............................. 196 , Figure 13. Experimental (means and standard deviations) percent biomass of bird diet composed of smallmouth bass as simulated by Rebuild Scenarios 3 and 4 for a) all Main Archipelago birds, and b) only birds nesting on Pismire and Grape Islands... ... ... ..‘. 197 Figure 14. a) Simulated adult population sizes when adult populations are subject to varying levels of angling harvest. b) The cormorant predation pressure on immature bass necessary to work in synergism with adult bass angling to drive the adult population to observed levels ................................................................................... 198 Figure 15. a) Simulated number of adult bass removed by anglers at varying harvest levels beginning in 1988 and ending in 1999. b) The biomass (kg) of immature bass needed to be removed by cormorants over the same simulation period when angling pressure varies.......... .......................................................................... 199 Figure 16. Simulated impacts on the adult bass population that increased YOY mortality ‘ (0.99), increased annual natural mortality for age-1 through adult fish (0.535) and increased adult mortality (0.885), in the absence of cormorant predation. Each variable was tested separately while all other variables remained unchanged ..................... 200 Figure 17. The simulated results for the Baseline Forecast Model with spawning constant set at 0.10 showing a) the resulting adult population size when cormorant predation is increased and b) the biomass (kg) of bass consumed by cormorants in each simulation ......................................................................................... 201 Figure 18. The simulated results for the Baseline Forecast Model with spawning constant set at 0.20 showing a) the resulting adult population size when cormorant predation is increased and b) the biomass (kg) of bass consumed by cormorants in each simulation ............... . .......................................................................... 202 Figure 19. Forecast Constant Predation Pressure Scenario 1 output showing a) the resulting adult bass population for experimental (means and standard deviations) simulations and b) the biomass of (kg) immature bass consumed by cormorants in experimental simulations (means and standard deviations) ............................... 203 Figure 20. Forecast Constant Predation Pressure Scenario 2 output showing a) the resulting adult bass population for experimental (means and standard deviations) simulations and b) the biomass (kg) of immature bass consumed by cormorants in experimental simulations (means and standard deviations) ............................... 204 Figure 21. Simulation results for the modified Forecast Constant Predation Pressure Scenario 1 exploring the influence of increasing cormorant pressure as bass numbers recover as compared to constant predation pressure showing a) adult population size and b) the biomass (kg) of immature bass consumed over the entire simulation period... 205 xvi Figure 22. Experimental (means and standard deviations) output for Forecast Increased Predation Pressure Scenario 1 showing a) simulated adult bass population and b) the biomass (kg) of immature bass consumed ................................................... 206 Figure 23. Experimental (means and standard deviations) output for Forecast Increased Predation Pressure Scenario 2 showing a) simulated adult bass population and b) the biomass (kg) of immature bass consumed ................................................... 207 , xvii CHAPTER 1 Seefelt, NE. and 1C. Gillingham. 2004. The Double-crested Cormorant in Lake Michigan: A Review of Population Trends, Ecology, and Current Management. In: T. Edsall and M. Munawar, Ed. The State of Lake Michigan: E00109, Health and Management, Ecovision World Monograph Series. The Double-crested Cormorant in Lake Michigan: A review of population trends, ecology and current management NE. Seefelt, J.C. Gillingham Central Michigan University, Department of Biology Mt. Pleasant, Michigan 48859 Key Words: Phalacrocorax auritus, Beaver Archipelago, colonial waterbirds Introduction The Double-crested Cormorant (Phalacrocorax auritus), or DCCO, is the most widely distributed cormorant of the six North American cormorant species (Hatch and Weseloh, 1999). The breeding range extends from the Pacific Coast (Alaska to Mexico) to the Atlantic Coast (Newfoundland to the Caribbean). The DCCO is the only species of cormorant to breed in large numbers in the interior of the US. and Canada (Hatch and Weseloh, 1999). These breeding interior populations are strongly migratory and mostly winter along the south-eastern Atlantic Coast and Gulf Coast (Root, 1988). In addition, large numbers also winter at inland lakes, rivers and impoundments in the south-eastern US, and have become more numerous at catfish and other aquaculture facilities (Stickley et al., 1992; Mott et aL, 1998). The DCCO is the cormorant species most frequently cited as conflicting with sport and commercial fisheries in North America (Hatch and Weseloh, 1999). There are six allopatric breeding populations of DCCOs, including the West Coast, Alaska, Bahamas, Florida, Atlantic, and Interior Populations (Hatch, 1995). Of these, the Interior Population, which includes the Great Lakes Basin, is the largest, with more breeding pairs than all other populations combined. Although this population is centered in the northern prairies, in the Great Lakes, cormorants The State qf Lake Michigan: Ecology, Health and Management Edited by 71 Edsall & M. Munawar O Ecovision World Monograph Series 2004 Aquatic Ecosystem Health and Management Society N.E. Seefelt and J.C. Gillingham are widely distributed and colonies can be found throughout the lakes and the St. Lawrence River. Large colonies can be found in eastern and western Lake Ontario, western Lake Erie, and across the Upper Lakes (Hatch, 1995; Weseloh et al., 2002). In Lake Michigan, the three major breeding areas are southern and northern Green Bay, and the Beaver Archipelago (Ludwig et al., 1989; Scharf and Shugart, 1998) This work is an attempt to review and consolidate what is currently known about DCCOs in Lake Michigan. This includes an overview of the occurrence of DCCOs in the Great Lakes basin, as well as a discussion of both population trends and ecology of these birds in Lake Michigan. Finally, this work describes the recent changes in management legislation concerning DCCOs in the United States. History of occurrence and population trends DCCO populations have fluctuated greatly over much of their range throughout the past centuries. Early reports state that these birds were abundant across their range in the 18003 (Lewis, 1929), including in the Great Lakes Basin by the end of this century (Wires et al., 2001). However, by 1900, cormorants were not found breeding in the Lake Michigan Basin (Barrows, 1912; Ludwig et al., 1989). Yet, during the early 19003, cormorant populations expanded again in the upper Great Lakes, with the first official breeding record in 1913, in western Lake Superior (Baillie, 1947). By the 19303, several breeding colonies were active in northern Green Bay, Lake Michigan (Baillie, 1947; Ludwig, 1984). Due to persecution, commercial fishing practices and the general use of organochlorine chemicals, cormorant populations declined again from the 1940s through the early 1970s (Baillie, 1947; Ludwig, 1984; Craven and Lev, 1987; Ludwig et al., 1989). According to Ludwig (1984), waterbird surveys from 1959 through 1969, yielded no cormorant nests in Lake Michigan and the US. waters of Lake Huron, and by sometime between 1960 and 1962, DCCOs were completely extirpated as a breeding bird from Michigan. After 1973, DCCOs began their most recent resurgence in the upper Great Lakes (Scharf, 1978; Ludwig et al., 1989; Scharf and Shugart, 1998), with at least a few pairs nesting on two islands and shoreline areas of southern Green Bay (Ludwig, 1984). This recovery has been attributed to the inclusion of DCCOs on the list of protected species under the 1918 Migratory Bird Treaty Act in 1972 (DEIS, 2001). Thus, a decline in human depredation, combined with a decline in both commercial fishing and chemical residue contamination levels (Ludwig, 1984), as well as changes in the fish communities (Hatch, 1995) may have provided new opportunities for cormorant recovery across the Great Lakes. The population expansion of non-native fish commonly found in their diet, including Alewife The Double-crested Cormorant in Lake Michigan (Alosa pseudoharengus), has also been linked to the cormorant population recoveries (Weseloh and Ewins, 1994). Overall, the number of breeding pairs in the Great Lakes has increased from 89 nests in 1970 to 38,000 by 1991 (Weseloh et al., 1995). Formal nest count data for DCCOs in US waters of the Great Lakes are available from 1977 (Scharf, 1978) and additional surveys were conducted in 1984 (Ludwig et al,. 1984), 1989-1990 (Scharf and Shugart, 1998), and 1997 (Cuthbert et al., 1997) (Table l). The number of breeding DCCO colonies in the U.S. Great Lakes region has grown from four in 1977 to sixty-nine in 1997; in Lake Michigan this is paralleled by an increase from three in 1977 to twenty- seven active colonies in 1997 (Cuthbert et al., 1997). DCCOs have shown an overall population increase in the US. Great Lakes region, from 171 pairs in 1977 to 48,931 pairs in 1997. This same trend is true for Lake Michigan alone, where the number of breeding pairs increased from 75 pairs in 1977 to 28,158 pairs in 1997. Interestingly, in 1977 and 1989-1990, Lake Michigan cormorants comprised around 43% of the overall population nesting in US. waters of the Great Lakes. However, by 1997, the Lake Michigan DCCO population comprised over 57% of breeding pairs in the US. Great Lakes (Cuthbert et al., 1997). The revival of cormorant populations has been no less pronounced in the Beaver Archipelago where, as of 1997, they were estimated to comprise almost 39% of the nesting DCCO pairs within Michigan waters of the Great Lakes (Cuthbert et al., 1997; Ludwig and Summer, 1997) and over 41% of the Lake Michigan breeding population (Cuthbert et al., 1997) (Table 2). Historically and more recently, there have been six active breeding DCCO colonies in the Beaver Archipelago, including Grape and Timm’s spits (portions of Hog Island) and Pismire, Hat, Whiskey and Gull Islands. Overall, through 1997, the Beaver Archipelago breeding DCCO population has followed the general increasing trend seen lake wide. However, between 1997 and 2000, the population declined by 13.5%. In addition, between 2000 and 2001, the population size decreased by another 4.1% (Table 2). Colonies on Timm’s Spit and Whiskey Island have disappeared, while both Pismire and Gull Islands experienced growth. The overall decline in the Beaver Archipelago may be indicative of a stabilizing population or changes in suitable nesting habitat due to lower lake levels. Lower lake levels may have allowed greater access to some colonies for mammalian predators, and has also increased the size of some small, more remote islands in the archipelago. Similarly, in Green Bay, DCCO populations also appear to be stabilizing, since there are few remaining unoccupied sites and existing colonies are probably approaching upper size limits (K. Stromborg, US. Fish and Wildlife Service, New Franklin, WI, U.S.A., pers. comm..). ME. Seefelt and J.C. Gillingham Table 1. Lake Michigan and Great Lakes breeding Double-crested Cormorant population estimates beginning in 1977 and ending 1997. Both the number of breeding pairs and active colonies have increased substantially over this twenty year period. In 1997, the Lake Michigan colonies comprised 57.55% of the breeding cormorants in the US. waters of the Great Lakes. Lake Michigan Lake Michigan Great Lakes Great Lakes Erecting Pairs Colonies BreedingPairs Colonies 1977a 75 3 171 4 1984h 684 9 NA NA 1939-9oc 4743 18 1 1099 36 19976' 28158 27 4893 1 69 ‘ Nest count data from Scharf, 1978. " Nest count data rtom Ludwig, 1984. ° Nest count data from Scharf and Shugart, 1998. d Nest count data from Cuthbert etal., 1997. Table 2. Numbers of Double-crested Cormorant pairs breeding at colonies in the Beaver Archipelago, northern Lake Michigan beginning in 1984 and ending in 2001. Note that the number of breeding pairs increases until 1997, where after there is an overall population decline. Some colonies, such as Pismire and Gull, have continued to show an increase as the regional population declines, while Timm’s and Whiskey colonies have dissolved. Pismire Grape Timm’s Hat Whiskey Gull Total 1984. 57 0 0 54 0 139 250 mg" 35 291 0 294 0 260 880 1997° 383 3509 753 4617 560 1887 11709 2000 987 2431 277 4917 0 1532 10125 2001 1035 2146 0 4511 o 2013 9705 ‘ Nest count data from Ludwig, 1984. b Nest count data from Scharf and Shugart, 1998. ° Nest count data from Cuthbert et al., 1997. Ecology DCCOs are seasonal inhabitants of Lake Michigan, typically arriving in April in northern regions. In the Beaver Archipelago, egg laying and incubation begins in May or early June, chicks are batched in mid to late June, and young birds fledge by late July or early August. Most birds leave the area by September; however it is not uncommon to see some DCCOs in October in the archipelago. There are also migrant birds that pass through the region in spring and autumn, and imrirature (non-breeding) birds that summer in the archipelago. Since the resurgence of the The Double-crested Cormorant in Lake Michigan DCCO population in the Beaver Archipelago, there has been a growing concern how this seasonal residency and the migration influx may influence local prey species. DCCOs are opportunistic fish predators that often feed in shallow waters (Lewis, 1929; Birt et al., 1987). Information on prey taken by these birds has been widely gathered, but generally only in context of perceived fishery conflicts (Hatch and Weseloh, 1999). Seasonal variation in cormorant diet is evident in most studies, but simple lists and percentages do not reveal the ecological factors contributing to this predation pattern and the impact on fisheries. Since the local impacts on prey populations and ecosystem dynamics have remained unclear, it has been recommended that research efforts focus more closely on foraging behaviour and predator-prey interactions (Hatch and Weseloh, 1999). Currently, some areas in Lake Michigan, including the Beaver Archipelago, are being investigated to further understand cormorant-fish relationships. Prey species and prey size are factors that may help determine the impact DCCOs have on fisheries (Ainley et al., 1981; Fowle, 1997). Because of the loss of large, native, piscivorous fish in the 19403 and an increase in salmonid stocking programs, prey fish populations have fluctuated across the region (Christie et al., 1987). In addition, introduced Alewife became abundant in all lakes, except Lake Superior, by the 19503 (Hatch and Weseloh, 1999). Other non-native forage fish, such as Rainbow Smelt (Osmerus mordax), have also become established. Breeding cormorants remain relatively close to breeding colonies when foraging (Custer and Bunck, 1992). Although cormorants may have only small and localized effects on fish populations during migration (Kirsch, 1995), it has been demonstrated that these birds may deplete fish prey around breeding colonies in some areas (Birt et al., 1987). In Lake Ontario, for example, cormorants had a significant effect on specific age classes of Smallmouth Bass (Micropterus dolomieui) (Adams et al., 1999; Johnson et al., 1999; Schneider and Adams, 1999). However, most studies reflect that cormorant diets tend to include species that are of little commercial or sport value. Yet these species may be important in community trophic dynamics (Craven and Lev, 1987). Therefore, cormorants may have a secondary effect on sport fisheries by competing with desired species for forage fish and other prey such as crayfish. Although the effects on forage fish numbers may be limited and may only occur in localized areas (Madenjiau and Gabrey, 1995), this combined with direct sport and commercial fish depredation may have some impact on some sport and commercial fish populations. Historically, there have been several studies documenting the diet of cormorants in the upper Great Lakes, including Lakes Huron, Michigan and Superior (Craven and Lev, 1987; Ludwig et al., 1989; Ludwig and Summer, 1997; Maruca, 1997; Neuman et al., 1997). Ludwig et a1. (1989) documented food items (n=8512) in the regurgitates of adults and chicks at several locations in Lakes Huron, Michigan NE. Seefelt and J.C. Gillingham and Superior from 1986 to 1989. By number, Alewife and Nine-spine Sticklebaek (Pungitius pungitius) accounted for 41% of the diet By biomass, the important species included Alewife (5 7%), Yellow Perch (Percaflavescens) (13%), Rainbow Smelt (8%), and White Sucker (Catostomous commersoni) (7%). Diet varied seasonally, and by August, the diet of cormorants in each study area surveyed contained 100% Alewife (Ludwig et al., 1989). In addition, Ludwig and Summer (1997) documented food items (n=6293) in the regurgitates of adults and chicks at nesting colonies in the Les Cheneaux Islands of northern Lake Huron in 1995. By weight, Alewife constituted 72% of the nestling diet. As part of the same study, Maruca (1997), using 373 stomachs, documented that adult cormorant diet contained approximately 48% Yellow Perch during the perch spawning season. In July, however, adults fed primarily on Alewife. Weseloh and Ewins (1994) have suggested that cormorant reproductive success may be intimately linked to Alewife population dynamics. The Beaver Archipelago, and particularly the habitat around Garden and Hog Islands, has long been known for its excellent Smallmouth Bass fishing, and this evaluation has been published in the national media a number of times (Robinson, 1995). Recently, however, there have been numerous reports of a decline in the sport fishery by local anglers (Hooker, 1999a, 1999b, 1999c). In partial response to these reports, Central Michigan University (CMU) and the Michigan Department of Natural Resources (MDNR) initiated an intense population sampling in May of 1999. Compared to similar data gathered using trap nets by researchers at CMU nearly 20 years ago, not only is the Smallmouth Bass population down by an astounding 75-80%, but other fish species, including Brown Bullhead (Ictalurus nebulosus) and Rock Bass (Ambloplites mpestris) have declined by as much as 98% (D. Peterson, University of Georgia, Athens, GA, U.S.A., pers. com.) [I is therefore quite clear that there has been a recent and very rapid decline in the Beaver Archipelago fishery. The factors that have caused these declines have remained unclear. Present research is investigating whether the local DCCO breeding population could have played a role in the decline of local fish populations. Most recent regurgitate and stomach content data suggest that DCCOs in the Beaver Archipelago feed primarily on Alewi fe during the breeding season. During 2000 and 2001, a total of 1128 regurgitate samples (10,600 individual prey items) were collected. When regurgitated food items are compared by mass, Alewife comprised 72.00% of the diet (57,073 g of 79,230 g) (Figure 1). Of the 150 stomachs (3363 individual prey items) analyzed in 2000 and 2001, Alewife mass comprised 72.83% of the diet (18,603 g of 25,550 g) (Figure 1). Other prey commonly found in the diet of Beaver Archipelago DCCOs are crayfish (Orconectes sp.), sculpin (Cottus sp.), Nine-Spine Sticklebaek, and White Sucker. Other miscellaneous prey include Spottail Shiner (Notropis hudsonius), Johnny The Double-crested Cormorant in Lake Michigan 80 70- 604 a” Q40“ OWEN :30. IStomachs 204 104 0 e’d/cfl’éf'f” Figru'e 1.DietofDouble-crestedCormorantsintheBeaverArchipelago, 2000and2001 combined, shownaspercentmassforbothregurgitate(n-1128)audstomeehs(n-150).Alewife(Alosa) povide the most biomass in cormorant diets. For regurgitate samples, Alewife comprise 57,073 g of 79,230, while for stomach samples, Alewife comprise 18,603 g or 25,550 g. However crayfish (Orooneaes), sculpin (Courts), stickleback (Pungittrc) and sucker (Catastarnoru) are also frequaltly taken. Darter (Etheostonra nigrwn), Trout-perch (Percopsis omiscomaycus) and Brook Sticklebaek (Culaea inconstans). Only one Smallmouth Bass was found in this investigation. No sample has yet yielded evidence of Rockbass or Brown Bullhead in the diet of archipelago DCCOs. As in previous studies in the Upper Great Lakes (Ludwig et al., 1989; Ludwig and Summer, 1997), Alewife became increasingly more inrportant in the diet of Beaver Archipelago DCCOs as the breeding season progressed. Current management Cormorant-human conflicts are not a new phenomenon. Because of their perceived negative effect on aquatic communities, several studies have been conducted investigating cormorant influences on fisheries both in Europe (Suter, 1995; Warke and Day, 1995) and in the Great Lakes (Madenjian and Gabrey, 1995; Maruca, 1997; Neuman et al., 1997; Schiavone, 2001). Research suggests that waterbirds actually play central roles in marine food webs (Cairns, 1992), and this probably holds true in Great Lakes community dynamics. In order to get a realistic picture of the impact bird predators have on fish populations in Lake Michigan, it is NE. Seefelt and J.C. Gillingham necessary to incorporate quantitative data on the diets, population size and energy requirements of the cormorant population. In addition, reliable data on the size of the fish populations, with the appropriate spatial and temporal scales, are necessary (Draulans, 198 8). Currently, in the Beaver Archipelago, data of this sort are being collected in order to facilitate the reconstruction of fish communities and to determine the extent of the role the cormorant resurgence may have had in recent fishery declines in the region. Similar approaches have been successfully used in Lake Erie to assess the impact DCCOs have on fish populations (Madenjian and Gabrey, 1995; Hebert and Morrison, 2003). Currently, there is no long-term management plan for DCCO populations in Lake Michigan. In the Wisconsin waters of Green Bay, a landowner has successfully deterred cormorants from nesting on his small island using devices designed to scare the birds. In addition, there has also been an isolated incidence of vandalism at a colony on a northern island in Green Bay (K. Stromborg, US. Fish and Wildlife Service, New Franklin, WI, U.S.A., pers. comm). Yet, overall in Lake Michigan, there have not been actions to manage cormorant populations. Since the resurgence of DCCO populations, research has focused mainly on monitoring breeding colony size, diet and toxicology within this basin. However, there has been an aggressive DCCO management program in the Lake Ontario Basin. In 1992, the New York State Department of Environmental Conservation (NY SDEC) began focused research to determine the actual impacts DCCOs have on fisheries and other aspects of both the Lake Oneida and eastern Lake Ontario ecosystems (Farquhar et al., 2003). This research, although still ongoing, has led to the implementation of a five-year management plan for US. waters of the eastern basin of Lake Ontario beginning in 1999. Management practices have focused on using egg-oiling, nest removal, harassment and habitat modification to control DCCO numbers, without the implementation of lethal control of adults and chicks (Farquhar et al., 2003). Egg-oiling has proved successful at reducing reproductive success of DCCOs in US. waters of Lake Ontario. Mthin five years, the number of breeding pairs should be reduced to target numbers with continued annual oiling activities (Schiavone, 2003). Resolving human-cormorant conflicts has become a focus of the US. Fish and Wildlife Service. In 1998, a Depredation Order (50 CF R 21.47) was enacted which authorized “commercial aquaculture producers in 13 states...to take DCCOs, without a federal permit, when found committing or about to commit depredations to aquaculture stocks” (DEIS, 2001). Since this original action did not allow for Federal management or population control of DCCOs, and did not take into account more recent concerns, it has been considered ineffectual. In the fall of 2001, the US. Department of the Interior Fish and Wildlife Service in conjunction with the US. Department of Agriculture APHIS Wildlife Services released a Draft Environmental Impact Statement (DEIS) for DCCOs. The purpose The Double-crested Cormorant in Lake Michigan of that document was to describe and evaluate alternatives which focused on reducing conflicts between DCCOs and people (commercial, recreational and other issues) and to ensure the long-term health of cormorant populations (DEIS, 2001). Apparently, although most Americans were ambivalent with regard to DCCOs, there were many individuals concerned about the conservation and management of these birds. The DEIS categorized these concerned parties as follows: 1) animal protectionists that support non-lethal management; 2) individuals, including resource professionals, that favour conserving DCCOs and not scapegoating the birds; 3) others, including resource professionals, that emphasize conservative DCCO management; and 4) citizens who are directly affected by DCCOs, including aquaculturists, and favour more aggressive management. These differing viewpoints have added much emotion to the debate and the resolution of cormorant-human conflicts. In order to facilitate both dialog and action, the DEIS proposed six alternatives ranging from no action (allowing current management plans to stand) to a cormorant hunting season. Each alternative outlined in the DEIS was analyzed as to how each would impact cormorant populations, fish, other birds, vegetation, federally listed Threatened and Endangered species, water quality, human health, economic issues and others (DEIS, 2001). The new “Proposed Action” favoured by the Services establishes “a new Depredation Order to address public resource conflicts.” The new action would authorize “State, Tribal and Federal land management agencies to implement a DCCO management program, while maintaining Federal Oversight of DCCO populations via reporting and monitoring requirements” (DEIS, 2001). Participation of State agencies is strictly voluntary and these management plans will allow for local DCCO control with federal oversight. This new Public Resource Depredation Order will allow for some taking of cormorants at breeding and roosting sites, as well as eg oiling and destruction (DEIS, 2001). In March 2003, the Department of the Interior, Fish and Wildlife Service released the proposed rule for DCCO management (50 CFR Part 21, 68 FR 12653). Then, in August 2003, the Final Environmental Impact Statement (FEIS) was released. According to the F E18 (2003), the Public Resource Depredation Order “will cause the estimated take of < 160,000 DCCOs, which is not predicted to have a significant negative impact on...DCCO populations.” In addition, it will minimize the local impact on other birds, reduce both fishery and vegetation impacts on a local scale and reduce depredation on both aquaculture and hatchery facilities. However, this new order “is not likely to significantly benefit recreational fishing economies or commercial fishing” (FIES, 2003). The final rule (50 CFR Part 21.48, 68 FR 25396) released on 08 October 2003, allows for local cormorant control in twenty-four states. It established not only a Public Resource Depredation Order, but also revised the original Aquaculture Depredation Order. In addition, the new rule also requires the monitoring of cormorant numbers and careful record 10 N.E. Seefelt and ./.C. Gillingham keeping to insure that populations remain sustainable. This final rule took effect 07 November 2003. Currently, it is difficult to comment on how this action will influence Lake Michigan populations of DCCOs. Conclusion The DCCO, an indigenous species, has shown a remarkable population recovery over the past three decades in Lake Michigan and other Great Lakes. Often perceived as depredating fish stocks, thus far, in northern Lake Michigan, these birds do not seem to have negatively impacted either sport or commercial fisheries. However, cormorant-fish relationships are still being investigated and soon more information may be available, especially with respect to the northern Lake Michigan ecosystem including the Beaver Archipelago. It will be interesting to see how this new information in combination with the new management rule will influence DCCO populations of Lake Michigan. Summary The Double-crested Cormorant (Phalacrocorax auritus), a colonial waterbird native to North America, has experienced a substantial population increase throughout the Great Lakes, including Lake Michigan, over the past thirty years. This resurgence in combination with a simultaneous decline in some sport and commercial fisheries has led to their implication in fishery depredations. Research in Lake Michigan has largely focused on monitoring breeding population numbers and investigating prey species in the diet. Currently in northern Lake Michigan, specifically the Beaver Archipelago, the most important prey item in the cormorant diet appears to be Alewife (Alosa pseudoharengus). However, research is still being conducted to investigate what impact these birds may have on this species and other fisheries in northern Lake Michigan. In addition, the U.S. Department of Interior’s Fish and Wildlife Service, in conjunction with the U.S. Department of Agriculture’s APHIS Wildlife Services, has released a rule change that will allow for more aggressive management of cormorants when they are in conflict with economic and ecological interests. How this new rule will influence the Lake Michigan cormorant population remains uncertain. Acknowledgements The authors would like to acknowledge the Dr. Francesca J. Cuthbert and her 11 The Double-crested Cormorant in Lake Michigan crew for providing lake wide population estimates and advice. Also, Jory Jonas of the Michigan Department of Natural Resources - Fisheries Division for her valuable contact information and guidance. We would like to thank Dr. Kenneth Stromborg of the U.S. Fish and Wildlife Service for his insight on DCCO populations in Green Bay. In addition, the Michigan Department of Natural Resources — Wildlife Division and Central Michigan University Research Excellence Funds for providing funding. Finally, we are grateful to all members of our field and lab crew. References Adams, C.M._, Schneider, C.P., Johnson, J.H.. 1999. Predicting the size and age of smallmouth bass (Micmpterus dolomieu) consumed by Double-crested Cormorants (Phalacrocorax auritus) in eastern Lake Ontario, 1993-1994. In: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario, Section 6, pp. 1-8. Joint Report of NYSDEC Bureau of Fishes and the US Geological Survey Biology Resource Division, Albany, New York. Ainley, D.G., Anderson, D.W., Kelly, P.R., 1981. Feeding ecology of marine cormorants in southwestern North America. Condor 83, 120-131. Baillie, J.L., 1947. The Double-crested Cormorant nesting in Ontario. The Canadian Field Naturalist 61, 119-126. Barrows, W.B., 1912. Michigan Bird Life. Michigan Agricultural College Special Bulletin, East Lansing, MI. Birt, V.L., Birt, T.P., Goulet, D., Cairns, D.K., Montevecchi, W.A., 1987. Ashmole’s halo: direct evidence for prey depletion by a seabird. Marine Ecology Progress Series 40, 205-208. Cairns, D.K., 1992. Bridging the gap between omithology and fisheries science: use of seabird data in stock assessment models. Condor 94, 811-824. Christie, W.J., Scott, K.A., Sly, P.G., Strus, RH, 1987. Recent changes in the aquatic food web of eastern Lake Ontario. Can. J. Fish. Aquat Sci. 44 (Suppl.2), 37-52. Craven, S.R., Lev, E., 1987. Double-crested Cormorants in the Apostle Islands, Wisconsin, USA: population trends, foods habits and fishery depredations. Colonial Waterbirds 10(1), 64-71. Custer, T.W., Bunck, C., 1992. Feeding flights of breeding Double-crested Cormorants at two 'Wrsconsin colonies. Journal of Field Ornithology 63(2), 203-211. Cuthbert, F.J., McKearnan, J., Wemmer, L., 1997. U.S. Great Lakes Tern and Cormorant Survey: 1997 Progress Report. Department of Fisheries and Mldlife, University of Minnesota, St. PauL DEIS, 2001. Draft Environmental Impact Statement: Double-crested Cormorant Management. U.S. Department of Interior Fish and Wildlife Service and U.S. Department of Agriculture APHIS erdlife Service, Washington, DC. Draulans, D., 1988. Effects of fish-eating birds on fieshwater fish stocks: an evaluation. Biological Conservation 44, 251-263. Farquhar, J.F., McCullough, RD, Mazzocchi, I.M., 2003. Cormorant management activities in Lake Ontario’s eastern basin. NYSDEC Special Report: Section2, pgs 1-5. FEIS, 2003. Final Environmental Impact Statement: Double-crested Cormorant Management. U.S. Department of Interior Fish and Wildlife Service and U.S. Department of Agriculture APHIS Midlife Service, Washington, D.C.. 12 N.E. Seefelt and J.C. Gillingham Fowle, M.R., 1997. Population dynamics, food habits, and bioenergetics of Double-crested Cormorants in Lake Champlain. MS. Thesis, Wildl. and Fish. Biol. Prog., School of Natural Resources, University of Vermont. Hatch, J.J., 1995. Changing populations of Double-crested Cormorants. Colonial Waterbirds 18 (Special Publication 1), 8-24. Hatch, J .J., Weseloh, D.V.C., 1999. Double-crested Cormorant (Phalacrocorax aurirus). In: A. Poole, F. Gill (Eds.), The Birds of North America, Inc., Philadelphia, PA, No. 441. Hebert . C.E., Morrison, H.A., 2003. Consumption of fish and other prey items by Lake Erie waterbirds. Journal of Great Lakes Research 29(2), 213-227. Hooker, C., 1999a. Carrington Challenges Cormorants! ! In: Beaver Beacon (March), pp. 11. Beacon Publishing, Inc. St. James, Michigan. Hooker, C., 1999b. Cormorants Beware! In: Beaver Beacon (March), pp. 1-6. Beacon Publishing, Inc. St. James, Michigan. Hooker, C., 199%. Protected Invaders! Cormorants overrun Great Lakes, Feds Ponder What to Do. In: Beaver Beacon (October), pp. 14-15. Beacon Publishing, Inc. St. James, Michigan. Johnson, J.H., Ross, R.M., Adams, C.M., 1999. Diet composition of Double-crested Cormorants in Eastern Lake Ontario, 1998. In: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario, Section 5, pp. 1-7. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division, Albany, New York. Kirsch, EM, 1995. Double-crested Cormorants along the upper Mississippi River. Colonial Waterbirds 18(Special Issue 1), 131-136. Lewis, HQ, 1929. The Natural History of the Double-crested Cormorant. Ru-Mi-Loo Books, Ottawa, Ontario. Ludwig, J. P., 1984. Decline, resurgence and population dynamics of Michigan and Great Lakes Double-Crested cormorants. Jack Pine Warbler 62, 91-102 Ludwig, J.P., Summer, CL, 1997. Population status and diet of cormorants in Les Cheneaux Islands area. In: 1.8. Diana, G.Y. Belyea, RD. Clark, Jr. (Eds.), History, Status, and Trends in Populations of Yellow Perch and Double-crested Cormorants in Les Cheneaux Islands, Michigan, No. 17, pp. 5-25. State of Michigan Department of Natural Resources, Lansing, MI. Ludwig, J.P., Hull, C.N., Ludwig, M.E., Auman, ILL, 1989. Food habits and feeding ecology of nesting Double-crested Cormorants in the upper Great Lakes, 1986-1989. The Jack-Pine Warbler 67(4), 115-126. Madenjian, C.P., Gabrey, S.W., 1995. Waterbird predation of fish in western Lake Erie: a bioenergenics model application. The Condor 97, 141-153. Maruca, S.L., 1997. The impact of cormorant predation on yellow perch in Les Cheneaux Islands, Lake Huron. In: 1.8. Diana, G.Y. Belyea, RD. Clark, Jr. (Eds.), History, Status, and Trends in Populations of Yellow Perch and Double-crested Cormorants in Les Cheneaux Islands, Michigan, No. 17, pp. 47-70. State of Michigan Department of Natural Resources, Lansing, MI. Mott, D.F., Glahn, J.M., Smith, P.L., Reinhold, D.S., Bruce, K.J., Sloan, C.A., 1998. An evaluation of winter roost harassment for dispersing Double-crested Cormorants away fi'om catfish production areas in Mississippi. Wildlife Soc. Bulletin 26(3), 584-591. Neuman, J., Pearl, D.L., Ewins, PJ., Black, R, Weseloh, D.V.C., Pike, M., Karwowski, K., 1997. Spatial and temporal variation in the diet of Double-crested Cormorants (Phalacrocorax auritus) breeding in the lower Great Lakes in the early 19905. Canadian Journal of Fisheries and Aquatic Sciences 54, 1569-1584 Robinson, J., 1995. Lake Michigan Smallmouths. Field and Stream (April) 16-22. 13 The Double-crested Cormorant in Lake Michigan Root, T.L., 1988. Atlas of Wintering North American Birds: An Analysis of Christmas Bird Count Data. University of Chicago Press, Chicago. Scharf, W.C., 1978. Colonial birds nesting on man-made and natural sites in the U.S. Great Lakes. Technical Report D-78-10, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi. Scharf, W.C., Shugart, G.W., 1998. Distribution and abundance of gull, tern , and cormorant nesting colonies of the Great Lakes, 1989 and 1990. Publication No. l, Gale Gleason Environmental Institute, Lake Superior State University, Sault Ste. Marie, Michigan. Schiavone, A., 2001. Double-crested Cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2000 studies. In: Impact of Double-Crested Cormorant Predation on Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario, 2000 Studies, Section I, pp. 1- 5. New York State Department of Environmental Conservation, Albany, New York. < ' mama? Schiavone, A., 2003. Double-crested cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2002 studies. In: Impact of Double-Crested Cormorant Predation on Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario, 2002 Studies, Section I, pp. 1-4. New York State Department of Environmental Conservation, Albany, New York. < ' WP Schneider, C.P., Adams, C.M., 1999. Estimating the size and age of smallmouth bass (Micmprerus dolomieu) and yellow perch (Perca flavescens) consumed by Double-crested Cormorants (Phalacrocorax auritus) in the Eastern basin of Lake Ontario, 1998. In: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario, Section I, pp. 1-6. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division, Albany, New York. Stickley, A.R., Jr., Warrick, G.L., Glahn, J.P., 1992. Impact of Double-crested Cormorant depredations on Channel Catfish farms. J. World Aquaculture Society. 23(1), 192-204. Suter, W., 1995. The effect of predation by wintering cormorants Phalacrocorax carbon grayling nymallus rhymallus and trout (Salmonidae) populations: two case studies from Swiss rivers. J. of Applied Ecology 32, 29-46. Warke, G.M.A., Day, KR, 1995. Changes in the abundance of cyprinid and percid prey affect the rate of predation by cormorants Phalacrocorax carbo carbo on salmon Salmo salar smolt in northern Ireland. Arden 83,157-166. Weseloh, D.V.C., Ewins, P.J., 1994. Characteristics of a rapidly increasing colony of Double- crested Cormorants (Phalacrocorax aurims) in Lake Ontario: population size reproductive parameters and band recoveries. J. Great Lakes Research 20, 443-456. Weseloh, D.V.C., Ewins, P.J., Struger, J., Mineau, P., Bishop, C.A., 1995. Double-crested Cormorants of the Great Lakes: Changes in population size, breeding distribution and reproductive output between 1913 and 1991. Colonial Waterbirds 18 (Special Publication 1), 48-59. Weseloh, D.V.C., Pekarik, C., Havelka, T., Barrett, 0., Reid, J., 2002. Population trends and colony locations of Double-crested Cormorants in the Canadian Great Lakes and immediately adjacent areas, 1990-2000: a manager’s guide. J. Great Lakes Research 28(2), 125-144. Wires, L.R., Cuthbert, F.J., Trexel, D.R, Joshi, A.R., 2001, Status of the Double-crested Cormorant (Phalacrocorax auritus) in North America. Final Report to U.S. Fish and Wildlife Service, Arlington, V.A. l4 CHAPTER 2 A COMPARISON OF THREE METHODS TO INVESTIGATE THE DIET OF BREEDING DOUBLE-CRESTED CORMORANTS (PHALA CROC ORAX A URI TUS) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN This chapter has been accepted for publication in Hydrobiologia 15 Abstract In order to understand the role of waterbirds in aquatic food webs it is important to first get an accurate depiction of their diet. Three methods of dietary assessment (pellets, regurgitate and stomach contents) are compared here for breeding Double-crested Cormorants (Phalacrocorax auritus) of the Beaver Archipelago, northern Lake Michigan. By numerical frequency (percent number), each method yielded different depictions of the diet. However, in terms of presence and absence (percent frequency) of possible prey types, stomach content data did agree with both pellets and regurgitate data. However, differences were noted between regurgitate and pellets. In terms of biomass measured (percent biomass) in regurgitate and stomachs, data gathered agreed. In essence, pellets underestimate the importance of alewife (Alosa pseudoharengus) and overestimate the importance of crayfish (Orconectes sp.) in the diet when compared to both regurgitate and stomach analysis. The non-lethal method of regurgitate collection and analysis appears most practical in assessing cormorant diet in this system. In combination with information on avian foraging ecology and prey populations, these data may be used to investigate the relationships among cormorants and their prey, and lead to a better understanding of Great Lake food web dynamics. 16 Introduction Research suggests that waterbirds play central roles in marine food webs (Cairns 1992), and this probably holds true in North American Great Lakes community dynamics. Several studies have been conducted investigating the influence of piscivorous birds on fisheries in Europe (Suter 1995, Warke and Day 1995) and the Great Lakes (Maruca 1997, Neuman et al. 1997, Schiavone 2001). Interactions between piscivores and their prey can lead to cascading direct and indirect effects at many trophic levels within lake communities (Kerfoot 1987). To gain insight into the impact avian predators have on fish populations, it is necessary to integrate quantitative data collected on many aspects of the biology and behavioral ecology of the avian populations in question, as well as an accurate account of the prey populations they may influence. In the Beaver Archipelago, data of this sort are being collected in order to facilitate the reconstruction of fish communities and to determine the extent of the role the Double- crested Cormorant (Phalacrocorax auritus), or DCCO, resurgence may have had in recent fishery declines in the region. Similar approaches have been successfully used in Lake Erie to assess the impact DCCOs have on fish populations (Madenjian and Gabrey 1995, Hebert and Morrison 2003). Central to gaining an understanding of the role of piscivorous birds in aquatic systems is the acquisition of accurate dietary data. The DCCO is an opportunistic fish predator that often feeds in shallow waters (Lewis 1929, Birt et al. 1987). Over the past several decades, the population of cormorants inhabiting the interior of North America has increased and expanded (Hatch and Weseloh 1999). High densities of birds combined with their observed fish-eating behaviors have led some natural resource biologists as well as the general public to 17 implicate cormorants in declines of both commercial (Ludwig et al. 1989, Neuman et al. ~ 1997) and recreational fisheries throughout the Great Lakes region (Lantry et al. 1999, Neuman et al. 1997). Although cormorants may have only small and localized effects on fish populations during migration (Kirsh 1995), Birt et al. (1987) documented that this species may deplete fish prey around their breeding colonies in a marine environment. Cormorant diets often include species that are of little commercial value but may be important to community trophic dynamics (Craven and Lev 1987). Therefore, cormorants may have a secondary effect on sport fisheries by competing with desired species for forage fish. Although the effects on forage fish numbers may be limited and only occur in localized areas (Madenjian and Gabrey 1995), combined with direct sport fish depredation, cormorants may impact sport fish distributions and/or numbers. Studies assessing DCCO diet have used several methods including the analysis of pellets, regurgitate, and stomach contents of harvested birds. Pellets, indigestible material such as bones and scales encased in mucous that are typically regurgitated on a daily basis, may easily be collected in large numbers at breeding colonies. In addition, pellet analysis is relatively inexpensive and fairly easy to complete (Carss et a1. 1997). However, pellets have been shown less effective at determining cormorant diet in some studies (Duffy and Laurenson 1983, Johnstone et al. 1990, Blackwell and Sinclair 1995, Trauttmansdorff and Wassermann 1995, Zijlstra and van Eerden 1995, Carss et al. 1997) and these limitations are discussed below. Analysis of stomach contents and regurgitated food items (boli) can be useful tools to investigate cormorant diet because both methods allow for study of relatively fresh material (Carss et al. 1997). Bones and scales of partially digested fish can be used 18 to determine fish age classes, as well as estimate lengths and widths by utilizing fish reference collections (Blackwell et al. 1995, Ross and Johnson 1995). There are drawbacks to stomach analysis, including the necessity of killing birds, potential small samples that may not be representative of breeding population diet, and presence of highly eroded biomass (Wires et al. 2003). However, stomach content analysis is usefirl because such dietary data are accompanied by age, sex and other information for each bird (Carss et al. 1997). Regurgitate samples, like pellets, are easily collected from breeding colonies because both nestling and adult birds will regurgitate stomach contents when disturbed (Lewis 1929). However, these regurgitate samples may not be complete and also show varying levels of digestion (Waneless et al. 1993, Carss et al. 1997). Because good sample sizes are easily collected, regurgitated food items are considered a rigorous method for estimating of nestling diet, but not necessarily adult diet (Wires et al.‘ 2003) This study analyzes the use of each method (pellets, regurgitate and stomach contents) to assess the diet of DCCOs at breeding colonies in the Beaver Archipelago in northern Lake Michigan. The goal of this study is to ascertain which method(s) yields the most accurate portrayal of DCCO diet in northern Lake Michigan. This work is part of a larger study investigating cormorant foraging ecology and fish population dynamics in the Beaver Archipelago. These data have guided efforts in estimating DCCO diet in the study area. 19 Materials and Methods Study Area The Beaver Archipelago is located in Michigan waters of colder, northern basin of Lake Michigan. The islands and surrounding mainland areas are primarily forested, sparsely populated, and considered the Northern Lacustrine-Influenced Ecoregion (Fuller et a1. 1995). Inshore areas consist of sand, cobble, rock and occasional small wetlands (EPA 2000). Open water areas around the islands include areas that exceed 80 m (262 ft) in depth (EPA 2000). Fish communities, although changed and degraded compared to pre-settlement conditions, are still developed within this aquatic ecosystem. Nearshore areas provide habitats for warm water fish, including Centrarchids, and pelagic prey fish, including alewife (Alosa pseudoharengus), dominate open water areas (EPA 2000). Overall, the northern basin of Lake Michigan is characterized as a “typical phosphorus- limited lake ecosystem” (Chen et a1. 2002). The Beaver Archipelago consists of about ten islands. Three of the larger islands (Gull, Hog and Hat Islands) and one small island (Pismire Island) contained nesting colonies of DCCOs that ranged in size from 277 to 4918 nests in 2000. The Hog Island colonies were located on two peninsulas known as Grape Spit and Timms Spit. For this work the diet of cormorants on Pismire Island (987 nests) and Grape Spit (2431 nests), because of their close proximity to each other (approximately 2 km or 1.25 miles), were examined together. 20 Pelletmd Regurgitatgs Pellets and regurgitate samples were collected by hand fi'om the ground adjacent ‘ to individual nests in the Pismire Island and Grape Spit colonies on 24 June 2000. In addition, regurgitates were collected from areas away from nests. Adults were observed regurgitating as they left the colony while young chicks remained in their nests. Therefore, adults likely produced samples collected within the colony but not immediately adjacent to nests. Each sample was placed in a plastic Whirl-pak® bag (510 g) and returned to the lab within 1-3 hours of collection in a cooler. Pellets were subsequently dried at 43°C in an oven for 24 hours and then stored in plastic bags inside a plastic container. Pellets were kept at room temperature. Regurgitate samples were frozen immediately. Sixty pellets (30 from each colony) were rehydrated using warm water. Rehydration allowed for manual removal of the mucous using rinse water and forceps. Pellet contents were further rinsed with cold water and sorted using a No.16 Standard Sieve (1.19 mm opening) and a No. 35 Standard Sieve (0.5 mm opening). All otoliths and some bones, including jaws, pharyngeal bones, operculae, cleithra and vertebrae, were removed and placed in vials containing 70% ethanol to retard any bacterial or fungal growth. Later, using a reference collection (University of Michigan Museum and personal collection), the number and prey species (or genera) were recorded for each pellet. Because most bones and otoliths were eroded, no attempts were made to calculate original length and fresh mass of prey. These methods are similar to those outlined in Carss et al. (1997). 21 A total of 44 regurgitate samples, 31 from Pismire Island and 13 from Grape Spit, were thawed and analyzed. Each prey item was identified to species when possible and recorded. In addition, all identified prey items, including partially digested prey, were individually weighed. Complete fish were measured to the nearest 0.5 mm. Regurgitate samples were then preserved in 70% ethanol. 8mm Contents T wenty-five birds used for the stomach analysis were collected using shotguns on 23 June and O6, 15, 23 July 2000 (U SFWS Permit No. M3022886). These birds were harvested as they returned to their breeding colonies. After birds were collected, they were placed in plastic bags and frozen. Later, the birds were thawed and examined as outlined in Carss et al. (1997). The esophagus, crop and complete stomachs (proventriculus and pylorus) were removed from each bird and total mass of these organs. and their contents were recorded. These organs were then dissected and all prey items were removed and identified to species when possible. All prey items, including partially digested prey, were individually weighed. Complete fish were measured to the nearest 0.5 mm. Stomach contents were then preserved in 70% ethanol. In addition, each bird was sexed by examining reproductive organs. Analysis Numerical frequencies of prey items in the samples were calculated for each method and were converted to percentages (also referred to as percent numbers). Wires et al. (2001) defines percent number as the number of specimens of a taxon as a percent 22 of all specimens in a sample. Raw data from each method for both alewife and crayfish (Orconectes sp.) were analyzed using contingency tables for 3x2 and 2x2 comparisons and Chi-Square Goodness of Fit Tests (Sokal & Rohlf, 1995). Data were also examined by comparing the number of samples that contained a particular prey item for each method. These data, converted to percentages, are referred to as percent frequencies by Wires er a1. (2001). The values for both Alewife and crayfish were compared for each method using 3x2 and 2x2 contingency tables and Chi-Square Goodness of Fit Tests (Sokal and Rohlf 1995). All other prey items were found rather infrequently within the samples and were not further analyzed. Biomasses of prey items for both regurgitate and stomach content data were converted to percents. Percent biomass is defined as the biomass of a taxon as a percent of total biomass (Wires et al. 2002). Because there was a large range of sample masses (2.0 g to 136.7 g for regurgitates and 1.4 g to 413.7 g for stomach contents), these data were converted to proportions; an arcsine transformation was performed to normalize data (Sokal and Rohlf 1995). Transformed data for Alewife and crayfish were then analyzed using a Mann-Whitney Test (Minitab 13 for Windows). Results Analysis of pellets, regurgitate and stomachs shows that in late June-July 2000 the diet of DCCOs in the Beaver Archipelago included alewife (Alosa pseudoharengus), crayfish (Orconectes sp.), sculpin (Cottus sp.), nine-spine stickleback (Pungitius pungr'tius), sucker (Catostomus sp.), johnny darter (Etheostoma nigrum), Trout-perch 23 (Percopsis omiscomaycus), and spottail Shiner (Notropis hudsonius). Birds harvested for stomach contents included nine males and sixteen females. Percent number data indicated that pellets produced by Beaver Archipelago cormorants are comprised of 82.29% crayfish and only 2.54% alewife (Figure 1). Regurgitate samples indicate, by percent number, crayfish constitute 39.54% of the samples, while alewife comprise 28.60% (Figure 2). Stomach content data, by percent number, indicated that crayfish comprised 31.96% of the stomach contents and alewife made up 46.80% (Figure 3). The 3x2 contingency table (Table l) and Chi-Square Goodness of Fit Tests for the raw numerical frequency data indicate that values differ from expected and therefore, each method differed from each other in estimating the diet of DCCOs for both alewife (x2 = 387.06, critical value = 5.99 at a = 0.05, df= 2) and crayfish (x2 = 119.02, critical value = 5.99 at or = 0.05, df = 2). The 2x2 contingency tables (not shown) also indicate that each method differed from the other two in describing the DCCO diet. Table 2 shows the number of samples that contained a particular prey item for each method of dietary assessment. The 3x2 contingency table (Table 3) indicated that these data differed from expected and therefore, each method differed fi'om each other in estimating the diet of DCCOs for both alewife (x2 = 9.53, critical value = 5.99 at CL = 0.05, df = 2) and crayfish (x2 = 6.46, critical value = 5.99 at or = 0.05, df = 2). However, ~ pairwise comparisons using 2x2 contingency tables (not shown) indicate that pellets and stomach content data for both alewife (x2 = 3 .35, critical value = 3.84 at or = 0.05, df = 1) and crayfish (x2 = 1.50, critical value = 3.84 at or = 0.05, df = 1) were statistically similar. The type of method had no effect. In addition, regurgitate and stomach content data, 24 when analyzed using 2x2 contingency table (not shown), were also statistically similar for both alewife (x2 = 0.61, critical value = 3.84 at a = 0.05, df = 1) and crayfish (12 = 0.74, critical value = 3.84 at or = 0.05, df= l). Regurgitate samples indicate that, by percent biomass, crayfish constituted 15.83% of the DCCO diet, while alewife comprise 68.82% of their diet (Figure 4). Stomach content data, by percent biomass, indicate that crayfish comprise 19.74% of the diet and alewife made up 69.24% of the diet (Figure 5). The Mann-Whitney Test (F igure. 6) indicates that biomass of alewife (confidence intervals = 0.01 to 36.21, W = 1652.0, p = 0.1207, adjusted for ties) and crayfish (confidence intervals = -3 6.20 to -—0.01, W = 1428.0, p = 0.1207, adjusted for ties) estimated by each method are not statistically significant from each other at a = 0.05. Both dietary assessment methods appear to be equal predictors of the Alewife and crayfish biomass in the DCCO diet. Discussion Different methods of investigating the diet of DCCOs can lead to different estimations of prey abundance and occurrence in the diet. By numerical frequency, each method yielded different results. However, in terms of presence and absence of possible prey types, each method agreed, with some exceptions. Spottail Shiner appeared in both pellets and stomachs, but not regurgitate samples. Likewise, pellets did not show any evidence of trout-perch in the diet of DCCOs, while the other two methods showed they are captured in small numbers. In addition, stomach content data did agree with both pellets and regurgitate data in terms of number of samples in which alewife and crayfish - occur. Regurgitate and pellets, however, differ from one another. Finally, in terms of 25 biomass measured in regurgitate and stomachs, values for alewife and crayfish were not statistically significant from each other. Historically there have been several studies documenting diet of cormorants in the upper Great Lakes, including Lakes Huron, Michigan and Superior (Craven and Lev 1987, Ludwig et al. 1989, Ludwig and Summer 1997, Maruca 1997, Neuman et al. 1997). Ludwig et al. (1989) documented food items (n=8512) in regurgitates of adults and chicks at several locations in Lakes Huron, Michigan and Superior from 1986 to 1989. By number, alewife and nine-spine stickleback accounted for 41% of the diet. By biomass, the important species included alewife (57%), yellow perch (Percaflavescens) (13%), rainbow smelt (8%), and white sucker (Catostomous commersoni) (7%). Diet varied seasonally, and by August, the diet of cormorants in each study area surveyed contained 100% alewife (Ludwig et al. 1989). In addition, Ludwig and Summer (1997) documented food items (n=6293) in the regurgitates of adults and chicks at nesting colonies in the Les Cheneaux Islands of northern Lake Huron in 1995. By weight, alewife constituted 72% of the diet. As part of the same study, Maruca (1997), examined 373 stomachs and documented that adult cormorant diet contained approximately 48% yellow perch during the perch spawning season. In July, however, adults fed primarily on alewife. With the exception of Lake Superior, throughout the Great Lakes region, open water fish species, including alewife, are important in DCCO diet (Wires et al. 2001). Weseloh and Ewins (1994) have suggested that cormorant reproductive success may be intimately linked to alewife population dynamics. In this study, it appears that in late June and July alewife is an important prey item in Beaver Archipelago DCCOs when 26 analyzing both regurgitate and stomach samples. However, pellet analysis does not support this finding. The limitations of pellet analysis have been demonstrated in other works, including studies with captive birds (J ohnstone et al. 1990, Trauttmansdorff and Wassermann 1995, Zijlstra and van Eerden 1995) and in the field (Duffy and Laurenson ' 1983, Blackwell and Sinclair 1995). However, several studies (Ross and Johnson 1995, 1999, Warke and Day 1995, Johnson etal.1999, 2001a, 2001b, 2003) have relied on pellets as indicators of the diet. In the Beaver Archipelago, evidence of some prey types was not apparent in pellets. This has been documented in other systems, as well (Brown and Ewins 1996). Pellets have been shown less effective at determining cormorant diet in some studies because of species-related differential recovery of prey types (J ohnstone et al. 1990). In essence, small prey and soft-bodied species may be under represented (Brugger 1993). Also, ototliths and bones may be eroded in pellets (da Silva and Neilson 1985, Jobling and Breiby 1986), thus the estimation of prey length and flesh mass are often in error (Carss et al. 1997). Prey found in pellets may also represent secondary consumption by cormorants (Blackwell and Sinclair 1995). The assumption that pellets reflect the remains of prey taken during the previous 24-hour period has been shown to be invalid in some species. Thus, pellets are less useful in estimating daily food intake and energy requirements (Russel er a1. 1995). Additionally, DCCO nestlings digest bones, possibly due to minerals needed for rapid growth (Dunn 1975), and do not produce pellets until about seven weeks of age (Trauttmansdorff and Wassermann 1995, Zijlstra and van Eerden 1995). Therefore, pellet analysis does not reflect nestling diet. 27 However, pellets have proved more useful in describing cormorant diets than feces (Johnson and Ross 1996). In northern Lake Michigan (Ludwig et al. 1989) and in similar systems such as northern Lake Huron (Ludwig and Summer 1997, Maruca 1997), alewife have been shown to be important prey. Because alewife remains are only detected at low levels in the samples, pellet analysis does not appear to accurately depict the importance of these fish in the diet of Beaver Archipelago cormorants. This could indicate different digestion of prey types. However, in eastern Lake Ontario, Johnson et al. (1999, 2001a, - 2001b, 2003) have used pellets to detect the presence and the importance of alewife in the diet of DCCOs. Yet, Derby and Lovvom (1997), when comparing pellets and stomach contents, found that each sampling technique did lead to different estimates of fish and crayfish in the diet of DCCOs in an area with known changes in prey availability. Regurgitate and stomach contents analyzed in this study more accurately depict the importance of alewife in the diet of DCCOs in the Beaver Archipelago, especially in comparison to the work by Ludwig el al. (1989). However, both methods have weaknesses and limitations, including the probability of under- and over- estimating daily food intake (Carss et al. 1997). Therefore, caution should be used when using either method to estimate daily food intake, because some digestion has inevitably occurred prior to sample collection (Wanless et al. 1993). However, with addition of other information (e.g., feeding observations, foraging patch location), use of both regurgitate and stomach content data can be applied to bioenergetics models, and contribute to the understanding of relationships among waterbirds and their prey. 28 Other concerns include the accuracy of both regurgitates and stomach samples in describing the diet of both adults and chicks. However, Lewis (1929) noted by observation at breeding colonies that both male and female birds feed nestlings and adults appear to feed older chicks the same prey types consumed by adults. Therefore, regurgitate samples may provide a more complete assessment of cormorant diet during the breeding season. In addition, collection of regurgitates when nestlings are young may allow examination of seasonal and age-related diet differences, especially because young birds do not produce pellets. Such data are valuable in assessing important prey in the diet, the relative abundance of these prey, how these prey populations may be influenced by cormorants, and if these predator-prey relationships may vary as the breeding season progresses. Choice of dietary assessment method used when investigating the diet of DCCOs may lead to different inferences in prey abundance and importance. According to Derby and Lovvom (1997), daily changes in bird foraging behavior and time of data collection may account for some of these discrepancies. Such discrepancies may be reflected in this study, for birds were harvested for stomach contents over a month long time period, while both pellets and regurgitates were collected in one day. However, regurgitate and stomach content data do suggest that DCCOs in the Beaver Archipelago feed on alewife during the breeding season. During 2000 and 2001, a total of 1128 regurgitate samples (10,600 individual prey items) were collected. Each year, samples were collected on three dates during the breeding season in an attempt to determine seasonal changes in the diet. When regurgitated food items are compared by mass, alewife comprised 72.00% of the samples (57,073 g of 79,230 g) (unpublished data). Of the 150 stomachs (3363 29 individual prey items) collected during the breeding seasons of 2000 and 2001, Alewife mass comprised 72.83% of the samples (18,603 g of 25,550 g) (unpublished data). This supports the findings of previous studies in the Upper Great Lakes (Ludwig et al. 1989, Ludwig and Summer 1997, Maruca 1997), where alewife become increasingly more important in the diet of DCCOs as the breeding season progresses. Under the current Lake Management Plan, Lake Michigan is to be managed by an, ecosystem approach (EPA 2000). Seabirds, such as DCCOs, that occupy high trophic levels are an integral part of aquatic food webs because they are very mobile and can integrate ecosystem processes over wide spatial and temporal scales (Hebert and Sprules 2002). Avian piscivores may be valuable environmental indicators in lake systems (Hebert and Sprules 2002) and, therefore, accurately estimating seabird diet may prove imperative in monitoring ecosystem health and processes. Conclusions Regardless of limitations, pellets can be useful in qualitatively documenting what prey types occur in the diet of DCCOs and other waterbird species. Both regurgitate and stomach analyses appear to be more usefirl in both qualitative descriptions and quantitative analysis of prey importance in the diet of breeding Beaver Archipelago DCCOs. Because it is a non-lethal method, regurgitate collection and analysis is the most practical way to assess cormorant diets in this system. Regurgitates can be collected in good numbers, can be analyzed quickly, and provide information on prey type, length and mass. Information on size and age class of fish taken by birds, as well as total biomass, is important in determining the influence cormorants may have on a fishery (Wires et al. 30 2001). In a lake ecosystem, predation on fish can have complex effects on other trophic levels and help determine community structure (Vanni 1987). The ability to estimate cormorant diet more accurately will strengthen attempts to understand the importance of these birds as predators in this study area. Diet studies alone cannot answer complex questions as to the relationships among DCCOs and their prey. However, it is an important step, in combination with more detail information on avian foraging ecology and prey population dynamics, in investigating community level interactions. Acknowledgements The authors acknowledge Dr. Gerald Smith of the University of Michigan Natural History Museum for the use of a fish reference collection and Susan Maruca for her advice on creating a fish reference collection. Also, we are gratefirl to Dr. John Rowe for his advice on data analysis. Further, we thank Dr. Don Hall for his financial and intellectual assistance and we also acknowledge the U.S. Fish and Wildlife Service for the permits (Permit Number MBOZZ886) to conduct this work. In addition, the Michigan Department of Natural Resources — Wildlife Division, Central Michigan University Research Excellence Funds, and Michigan State University Department of Zoology and - Ecology, Evolutionary Biology and Behavior Program for providing support for this study. Finally, we are indebted to the Central Michigan University Biological Station for logistical support and are grateful to all members of our field and lab crew. 31 Literature Cited Birt, V.L., T.P. Birt, D. Goulet, D.K. Cairns, and WA. Montevecchi. 1987. Ashmole’ 5 halo: direct evidence for prey depletion by a seabird. Marine Ecology Progress Series 40:205-208. Blackwell, B.F., W.B. Krohn, and RB. Allen. 1995. Foods of nestling Double-crested Cormorants in Penobscot Bay, Maine, USA: temporal and spatial comparisons. Colonial Waterbirds 18(Special Publication 1):]99-208. Blackwell, BF. and J .A. Sinclair. 1995. Evidence for secondary consumption of fish by Double-crested Cormorants. Marine Ecology. Progess Series 123: 1-4. Brown, KM. and P.J. Ewins. 1996. Technique-dependent biases in determination of diet composition: an example with Ring-billed Gulls. Condor 98:34-41. Brugger, KB. 1993. Digestibility of three fish species by Double-crested Cormorants. Condor 95:25-32. Cairns, D.K. 1992. Bridging the gap between ornithology and fisheries science: use of seabird data in stock assessment models. Condor 94:811-824. Carss, D.N. and the Diet Assessment and Food Intake Working Group. 1997. Techniques for assessing cormorant diet and food intake: towards a consensus view. Supplemento alle Ricerche di Biologia della Selvaggina 26: 197-230. Chen, C., RJi, D]. Schwab, D. Beletsky, G.L. Fahnenstiel, M.Jiang, T.H. Johengen, H. Vnaderploeg, B. Eadie, J .W. Budd, M.H. Bundy, W. Gardner, J.Cortner, and P.J. Lavrentyev. 2002. A model study of coupled biological and physical dynamics in Lake Michigan. Ecological Modeling 152: 145-168. Craven, SR. and E. Lev. 1987. Double-crested Cormorants in the Apostle Islands, Wisconsin, USA: population trends, foods habits and fishery depredations. Colonial Waterbirds 10:64-71. da Silva, J. and JD. Neilson. 1985. Limitations of using otoliths recovered in scats to estimate prey consumption in seals. Canadian Journal of Fish and Aquatic Sciences 42: 143 9-1442. Derby, CE. and J .R. Lovvom. 1997. Comparison of pellets versus collected birds for sampling diets of Double-crested Cormorants. Condor 99:549-553. Duffy, DC. and L.J.E. Laurenson. 1983. Pellets of Cape Cormorants as indicators of diet. Condor 85:305-307. 32 Dunn, EH. 1975. Caloric intake of nestling Double-crested Cormorants. Auk 92:553- 565. Environmental Protection Agency (EPA). 2000. Lake Michigan Lake Management Plan. Fuller, K., H. Shear, and J. Wittig (eds). 1995. The Great Lakes Atlas. Enviomment Canada and the U.S. Environmental Protection Agency. Hatch, J .J . and D.V.C. Weseloh. 1999. Double-crested Cormorant (Phalacrocorax auritus). In. Poole, A., F.Gill, (eds) The Birds of North America, Inc., Philidelphia, Pennsylvania, No. 441. Hebert , CE. and HA. Morrison. 2003. Consumption of fish and other prey items by Lake Erie waterbirds. Journal of Great Lakes Research 29:213-227. Hebert, CE. and W.G. Sprules. 2002. The relevance of seabird ecology to Great Lakes Management Journal of Great Lakes Research 28291-103. Jobling, M. and A. Breiby. 1986. The use and abuse of fish otoliths in the studies of feeding habits of marine piscivores. Sarsia 71:265-274. Johnson, J .H. and RM. Ross. 1996. Pellets versus feces: their relative importance in describing the food habits of Double-crested Cormorants. Journal of Great Lakes Research 22:795-798. Johnson, J .H., RM. Ross and CM. Adams. 1999. Diet composition of Double-crested Cormorants in Eastern Lake Ontario, 1998. In: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division. Johnson, J.H., RM. Ross, RD. McCullough, and B. Edmonds. 2001a. Diet composition and fish consumption of Double-crested Cormorants from the Pigeon and Snake Island colonies of eastern Lake Ontario in 2000. NYSDEC Special Report: Section 5, pgs. 1-9. Johnson, J .H., RM. Ross, RD. McCullough, and B. Edmonds. 2001b. Diet composition and fish consumption of Double-crested Cormorants from the Little Galloo Island colony of eastern Lake Ontario in 2000. NYSDEC Special Report: Section 3, pgs. 1-7. Johnson, J H RM. Ross, RD. McCullough, and B. Edmonds. 2003. Diet composition and fish consumption of Double-crested Cormorants from the Little Galloo Island colony of eastern Lake Ontario in 2002. NYSDEC Special Report: Section 3, pgs. 1-8. 33 Johnstone, I.G., M.P. Harris, S. Waneless and J .A. Graves. 1990. The usefulness of pellets for assessing the diet of adult Shags Phalacrocorax aristotelis. Bird Study 37:5-11. Kerfoot, WC. 1987. Cascading effects and indirect pathways. In: Kerfoot, WC and A] Sih, (eds) Predation: Direct and Indirect Impacts on Aquatic Communities. University of New England Press, Hanover, New Hampshire, pgs. 57-70. Kirsch, EM. 1995. Double-crested Cormorants along the upper Mississippi River. Colonial Waterbirds 18(Special Publication 1):]31-136. Lantry, B.F., T.H. Eckert, and GP. Schneider. 1999. The relationship between the abundance of Smallmouth Bass and Double-crested Cormorants in the Eastern Basin of Lake Ontario. In: Final Report: To Assess the Impact of Double-crested Cormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 12, pgs. 1-10. Lewis, H. G. 1929. The Natural History of the Double-crested Cormorant. Ottawa, Ontario: Ru-Mi-Loo Books. Ludwig, J .P. and CL. Summer. 1997. Population status and diet of cormorants in Les Cheneaux Islands area. In: J .8. Diana, G.Y. Belyea, RD. Clark, Jr. (eds) History, Status, and Trends in Populations of Yellow Perch and Double-crested Cormorants in Les Cheneaux Islands, Michigan. State of Michigan Department of Natural Resources, No. 17, pp. 5-25. Ludwig, J.P., C.N. Hull. ME. Ludwig and HI. Auman. 1989. Food habits and feeding ecology of nesting Double-crested Cormorants in the upper Great Lakes, 1986- 1989. The Jack-Pine Warbler 67:115-126. Madenjian, GP. and SW. Gabrey. 1995. Waterbird predation of fish in western Lake Erie: a bioenergenics model application. Condor 97: 141-153. Maruca, S.L. 1997. The impact of cormorant predation on yellow perch in Les Cheneaux Islands, Lake Huron. In: J .S. Diana, G.Y. Belyea, RD. Clark, Jr. (eds) History, Status, and Trends in Populations of Yellow Perch and Double-crested Cormorants in Les Cheneaux Islands, Michigan. State of Michigan Department of Natural Resources, No. 17, pgs: 47-70. Neuman, J., D.L. Pearl, P.J. Ewins, R. Black, D.V. Weseloh, M. Pike, and K. Karwowski. 1997. Spatial and temporal variation in the diet of Double-crested Cormorants (Phalacrocorax auritus) breeding in the lower Great Lakes in the early 1990s. Canadian Journal of Fisheries and Aquatic Sciences 54:1569-1584 34 Ross, RM. and J .H. Johnson. 1995. Seasonal and annual changes on the diet of Double-crested Cormorants: implications for Lake Ontario’s fishery. Great Lakes Resource Review 2: 1-9. Ross, RM. and J .H. Johnson. 1999. Effect of altered salmonid stocking methods on cormorant predation in eastern Lake Ontario. . In: Final Report: To Assess the Impact of Double-crested C ormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 11, pp. 1-10. Russel, A.F., S. Wanless, and MP. Harris. 1995. Factors affecting the production of pellets by Shags Phalacrocorax aristotelis. Seabird 17:44-49. Schiavone, A. 2001. Double-crested Cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2000 studies. NYSDEC Special Report: Section 1, pgs. 1-5. Sokal, RR. and RI Rohlf. 1995. Biometry, 3rd Edition. W.H. Freeman and Company, New York, 887 pp. Suter, W. 1995. The effect of predation by wintering cormorants Phalacrocorax carbon grayling Thymallus thymallus and trout (Salmonidae) populations: two case studies from Swiss rivers. Journal of Applied Ecology 32:29-46. Trauttmansdorff, J. and G. Wassermann. 1995. Number of pellets produced by immature cormorants Phalacrocorax carbo sinensis. Ardea 83: 133-134. Vanni, M.J. 1987. Indirect effect of predators on age-structured prey populations: planktivorous fish and zooplankton. In: Kerfoot, WC. and Sih, A. (eds) Predation: Direct and Indirect Impacts on Aquatic Communities. University of New England Press, Hanover, New Hampshire, pgs. 149-160. Wanless, S., M.P. Harris, and AF. Russel. 1993. Factors influencing food-load sizes brought in by Shags Phalacrocorax aristotelis during chick rearing. Ibis 135219- 24. Warke, G.M.A. and KR. Day. 1995. Changes in the abundance of cyprinid and percid prey affect the rate of predation by cormorants Phalacrocorax carbo carbo on salmon Salmo salar smolt in northern Ireland. Ardea 83 :157-166. Weseloh, D.V.C. and P.J. Ewins. 1994. Characteristics of a rapidly increasing colony of Double-crested Cormorants (Phalacrocorax auritus) in Lake Ontario: population‘ size reproductive parameters and band recoveries. Journal of Great Lakes Research 20:443-456. 35 Wires, L.R, D.N. Carss, F.J. Cuthbert and J .J . Hatch. 2003. Transcontinental connections in relation to corrnorant-fishery conflicts: perceptions and realities of a “bete noire” (black beast) on both sides of the Atlantic. Vogelwelt 124:389-400. Wires, L.R, F.J. Cuthbert, D.R Trexel, and AR Joshi. 2001. Status of the Double- crested Cormorant (Phalacrocorax auritus) in North America. Final Report to USFWS. Zilj stra, M. and MR. van Eerden. 1995. Pellet production and the use of otoliths in determining the diet of Cormorants Phalacrocorax carbo sinensis: trials with captive birds. Ardea 83: 123-13 1. 36 Table 1. 3X2 contingency table showing the actual (and expected) values of the numerical frequency data for Alewife (A losa pseudoharengus) and crayfish (Orconectes) in pellets, regurgitates and stomachs. Chi-Square Tests indicate that values differ from expected for both Alewife (x2 = 387.06, critical value = 5.99 at or = 0.05, df = 2) and crayfish (x2 = 119.02, critical value = 5.99 at or = 0.05, df= 2). Sampling Method Alosa Orconectes Totals Pellets 26 (204) 841 (663) 867 Regurgitate 123 (68.9) 170 (224.1) 293 Stomachs 205 (81.1) 140 (263.9) 345 Totals 354 1151 1505 37 9.: 6:4 69. A4: 8% 89. A432 63: £88ch o an 30 a: 3% 8% 5.5 68% cacamsmcx as: 0 ass as» $08 $.va an? 8a“: 322 Emcee? Mameoamk afieamomfim @3588an «SEMEQ .13th $828an 913‘ @0522 mam—952m .mmmofieoeeq E 56% Be So: >05 some do 365295 8083 at. .Amnanom 8m 3 n a new .mocfimwcnmoa com 3» .I. e :30on com on n 5 £an06 28 moaeumwaawoe .333 5 venom £03 £5: >er BeBzeE 20:3 BEES vouch—m5 .«o 3:632 .N 038. 38 Table 3. 3X2 contingency table showing the actual (and expected) values of the number of samples that contained Alewife (A losa pseudoharengus) and crayfish (Orconectes) in pellets, regurgitates and stomachs. Chi-Square Tests for 3x2 comparison indicate that values differ from expected for both Alewife (x2 = 9.53, critical value = 5.99 at or = 0.05, df = 2) and crayfish (x2 = 6.46, critical value = 5.99 at or = 0.05, df = 2). Comparisons using 2x2 contingency tables show that pellets and stomach contents data for both Alewife (x2 = 3.35, critical value = 3.84 at or = 0.05, df= l) and crayfish (x2 = 1.50, critical value = 3.84 at‘a = 0.05, df = 1) were statistically similar. Also, regurgitate and stomach contents data were statistically similar for both Alewife (x2 = 0.61, critical value = 3.84 at or = 0.05, df= 1) and crayfish (x2 = 0.74, critical value = 3.84 at or = 0.05, df= - l). Sampling Method Alosa Orconectes Totals Pellets 18 (29.5) 55 (43.5) 867 Regurgitate 29 (19.4) 19 (28.6) 293 Stomachs 14 (12.1) 16 (17.9) 345 Totals 61 90 1505 39 El Alosa El Orconectes I Cottus E Pungitius I Catostomus I Etheostoma I Notropis Figure 1. Pellet numerical frequency data showing the diet of Beaver Archipelago cormorants as percentages. 40 El Alosa E] Orconectes I Cottus E Pungitius I Catostomus I Etheostoma Percopsis Figure 2. Regurgitate numerical fiequency data showing the diet of Beaver Archipelago cormorants as percentages. 41 El Alosa El Orconectes I Cottus a Pungitius I Etheostoma Percopsis I Catostomus/Notropis Figure 3. Stomach contents numerical frequency data showing the diet of Beaver Archipelago cormorants as percentages. Percent values for Catostomus and Notropis are small (both 0.23%) were combined for clarity. 42 El Alosa E] Orconectes I Cottus E Pungitius I Catostomus I Etheostoma Percopsis Figure 4. Regurgitate biomass data showing the diet of Beaver Archipelago cormorants as percentages. 43 El Alosa El Orconectes I Cottus I Catostomus Percopsis E Pungitius, Etheostoma, Notropis Figure 5. Stomach contents biomass data showing the diet of Beaver Archipelago cormorants as percentages. Percent values for Pungitius, Etheostoma and Notropis are small (1.48%, 1.43% and 1.48%, respectively) were combined for clarity. 44 El Regurgitate I Stomachs Alosa Orconectes Figure 6. Means of arcsine transformed biomass data. The Mann-Whitney Tear tvfitxatsfi tnort Brouaoo 04) Alternate (xovdrtfievxe wrapworho = 0.01 to 36.21, 0 = 1652,0, n = 0.1207, a5¢uore§ (bop tree) avfi xporwdnon (xovdnéevxe wrepmako = —36.20 to -0.01, O = 1428.0, it = 0.1207, a6cpuoteb drop use) scrapers-:5 Bur scram p. 8131105 coepe vor oratiorlxakkw myvubtxavr bpou cam ornep at or = 0.05. 45 CHAPTER 3 POPULATION ESTIMATE AND DIETARY EVALUATION OF DOUBLE-CRESTED CORMORANTS (PHALA CROC ORAX A URI TUS) BREEDING IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN 46 Abstract The Double-crested Cormorant (Phalacrocorax auritus) has shown resurgence in population levels within the Great Lakes Basin, including Lake Michigan, over the past several decades. Since the abatement of the general use of organochlorine pesticides, these birds have made a remarkable comeback, and have been implicated in the decline of both commercial and recreational fisheries. In most regions, the role of cormorants in these declines has remained uncertain. The population size of breeding birds, well as their reproductive output, was estimated over several years in the Beaver Archipelago, northern Lake Michigan. In addition, the diet of birds was determined using regurgitates and the stomachs of harvested birds. The population size of breeding birds has declined since 1997, in the region. However, breeding population size, as well as reproductive output, appears to vary substantially from year to year, and may be linked to the availability of alewife (Alosa pseudoharengus). Breeding bird diet consists primarily of species of little commercial or sport value; the importance of individual prey species in bird diet varies temporally and spatially, in the Beaver Archipelago. Population and diet studies alone cannot answer complex questions as to the relationships among cormorants and their prey, but are important steps, in combination with more detailed information on avian foraging ecology and prey population dynamics, in investigating community level ~ interactions. 47 Introduction The Double-crested Cormorant (Phalacrocorax auritus), or DCCO, is the most widely distributed cormorant of the six North American cormorant species (Hatch and Weseloh 1999). The breeding range extends from the Pacific Coast (Alaska to Mexico) to the Atlantic Coast (Newfoundland to the Caribbean) and is the only species of cormorant to breed in large numbers in the interior of the U.S. and Canada (Hatch and Weseloh 1999). The DCCO has shown resurgence in population levels within the Great Lakes Basin over the several decades (Ludwig 1984, Ludwig and Summer 1997). Although probably inhabiting the Great Lakes Basin since about the turn of the century, and with a slow and steady population increase since its Great Lakes colonization, this species suffered a sharp decline in the 19505 and 603 because of dieldrin-mediated egg- shell thinning (Ludwig et al. 1995). Since the abatement of the general use of organochlorine pesticides, these birds have made a remarkable comeback, to the point of . being implicated in the decline of both commercial (Ludwig et al. 1989, Neuman et al. 1997) as well as recreational fisheries throughout the Great Lakes Region (Lantry et al. 1999, Neuman et al. 1997). Formal nest count data for DCCOs in Lake Michigan are available from 1977 (Scharf 1978) and additional surveys were conducted in 1984 (Ludwig 1984), 1989-1990 (Scharf and Shugart 1998), and 1997 (Cuthbert et al. 1997). The number of DCCO colonies in the U.S. Great Lakes region has grown from four in 1977 to sixty-nine in 1997; in Lake Michigan this is paralleled by an increase from three in 1977 to tvventy- seven active colonies in 1997 (Cuthbert et al. 1997). DCCOs have shown an overall population increase in the U.S. Great Lakes region, increasing from 171 pairs in 1977 to 48 48,931 pairs in 1997. This same trend is true for Lake Michigan, where the number of breeding pairs increased from 75 pairs in 1977 to 28,158 pairs in 1997 (Cuthbert et al. 1997). Interestingly, in 1977 and 1989-1990, Lake Michigan cormorants comprised around 43% (4743 pairs) of the overall population nesting in U.S. waters of the Great Lakes, and by 1997, the Lake Michigan DCCO population comprised over 57% (28,158 pairs) of the breeding population (Cuthbert et al. 1997). The revival of cormorant populations has been no less pronounced in the Beaver Archipelago where, as of 1997, they were estimated to comprise almost 39% of the nesting DCCO pairs within Michigan waters of the Great Lakes (Cuthbert et al. 1997, Ludwig and Summer 1997) and over 41% of the Lake Michigan breeding population (Cuthbert et al. 1997). Several studies have documented the diet in breeding populations of Double- crested Cormorants in the Great Lakes, including Lake Ontario (Johnson et al. 1999, Schiavone 2001, Schiavone 2003), Lake Huron (Ludwig et al. 1989, Ludwig and Summer 1997, Maruca 1997, Neuman et al. 1997) and Lake Superior (Craven and Lev 1987, Ludwig et al. 1989). In these studies, non-game forage fish, such as alewife (Alosa pseudoharengus), typically comprise most of the annual diet. Ludwig et al. (1989) documented food items (n=8512) in regurgitates of adults and chicks at several locations in Lakes Huron, Michigan and Superior from 1986 to 1989. By number, alewife and nine-spine stickleback (Pungitius pungitius) accounted for 41% of the diet. By biomass, the important species included alewife (57%), yellow perch (Perca flavescens) (13%), rainbow smelt (Osmerus mordax) (8%), and white sucker (Catostomous commensoni) (7%). As in other studies, diet varied seasonally, but by 49 biomass, the most important species was alewife. In fact, by August, the diet of cormorants in each study area surveyed contained 100% alewife (Ludwig et al. 1989). Ascertaining the number of potential predators in an ecosystem is necessary to better understand the influence these predators. In addition, in a lake ecosystem, predation on fish can have complex effects on other trophic levels and help determine community structure (Vanni, 1987). This study investigates DCCO breeding colonies, within the Beaver Archipelago, northern Lake Michigan from 2000 to 2004, in order to 1) estimate breeding population numbers, as well as changes in colony size and location, 2) estimate reproductive success and output, and 3) estimate the diet of breeding cormorants and their young. This work will be used to firrther investigate cormorant foraging ecology and bioenergetics, as well as how DCCOs may impact fish population dynamics in the Beaver Archipelago. Methods Study Area The Beaver Archipelago, located in Michigan waters of the colder, northern basin of Lake Michigan, consists of about ten main islands and numerous small islands (Figure 1). The number of smaller islands depends on fluctuating lake levels. Four of the larger islands (Gull, Hog, Hat and Whiskey Islands) and two small islands (Pismire and SE Garden Island) provided sites for nesting DCCOs at some time between 2000 and 2004. ' The Hog Island colonies were located on two peninsulas known as Grape Spit and Timm’s Spit. Throughout this work, in reference to dietary data, Pismire, Grape, Hat, and Timm’s are collectively referred to as the Main Archipelago. (By their location, both 50 Whiskey and SE Garden can be considered in the Main Archipelago, however, no dietary samples were collected at these sites.) Colonies, excluding Whiskey, of the Main Archipelago group are within an area of 78 square kilometers (approximately 21 square miles), to the north and east of Beaver Island. Gull Island consists of two colonies, North and South, on one island and is separated from the main island group by almost 18 km (11 miles) of open water to the west of Beaver Island. From aerial surveys, it was concluded that the cormorants breeding at Gull Island remain close to these colonies and do not regularly travel to the Main Archipelago. Population Estimates In order to estimate the breeding population of DCCOs and to minimize disturbance to the nesting colonies, three boat surveys were conducted during 2000, and again in 2001, within the Beaver Archipelago. Each year, these surveys were made on islands shown to support nesting colonies of cormorants following spring aerial surveys, including Pismire, Grape, Hat and Gull Islands. Timm’s Island was included in 2000, but did not support breeding birds in 2001. In 2000, count dates were 05-12 June, 24 June — 07 July, and 18-20 July. In 2001, count dates were 25-30 May, 20 — 22 June and 18 —25 July. Since islands supporting cormorant colonies were accessed by watercraft, weather conditions were largely responsible for the range of count date duration for each sample period. Year to year variations in the onset of nesting are responsible for different first count dates each year. Dates for each year roughly correspond to time periods when reproductive attempts are at their seasonal high, with egg-laying and incubation being the predominant activities (late May and early June), when most pairs are actively brooding - 51 and rearing young (mid June through mid July) and when chicks are beginning to fledge and nests that remain are those that are successful (late July). Nests were considered successful if they were well maintained and were occupied by large chicks (or large chicks were close proximity to the nests) in July. The methods of these surveys followed that of Cuthbert et al. (1997), where nest, eggs and chicks were hand tallied for each colony. Numbers of nests counted during the first count of each season were used to estimate the number of breeding pairs of cormorants in the Beaver Archipelago. By counting eggs and chicks, as well as the number of active nests as the breeding season progressed, population reproductive success was also estimated (see analysis below). In 2002, breeding birds were again estimated using boat surveys, following aerial surveys. On Pismire Island, three surveys were completed as the season progressed in the same manner described above; count dates were 27 May, 17 June and 13 July in 2002. Other colonies, including Grape, Gull and Hat Islands were counted only once by ground survey during the breeding season on 13 July, 15 July and 24 July, respectively. In addition, a new colony location, SE Garden was established and nests were documented in mid July. In 2003, nest counts were repeated on Pismire Island on 04 June, 09 July and 29 July, in the same manner followed in previous years. In addition, the SE Garden colony ’ was counted on 05 June and again on 29 July. Hat Island colony was counted on 20 June. Gull Island colonies were not counted in 2003. No other colonies were active in 2003. In 2004, the Pismire Island colony was counted on 01 June and 29 July by ground survey. The SE Garden Colony was counted on 15 June and again on 18 June by ground 52 survey. Whiskey Island was visited on 28 June to count a re-established colony and re- visited on 26 July to observe if any DCCOs were successful. Gull and Hat Island colonies breeding bird estimates were obtained from aerial photo taken on 09 June. To verify the validity of the counts from these digital photos, ground counts of Pismire Island were compared to counts obtained from an aerial photo of the Pismire colony. Diet Estimates To assess the diet of cormorants inhabiting the Beaver Archipelago and to maintain consistency with techniques used in previous studies, regurgitate samples and stomach contents of harvested bird were used to estimate diet composition. A third method, pellet analysis, was determined to be less reliable for DCCO diet quantification in the study area (see Chapter 2). Adult and young cormorants readily regurgitate both fresh and partially digested food items when disturbed. Regurgitate samples (boli) were collected by hand from the , ground adjacent to individual nests during colony visits in 2000 and 2001. Regurgitates were also collected from areas away from nests. Adults were observed regurgitating as they left the colony while young chicks remained in their nests. Therefore, adults likely produced samples collected within the colony but not immediately adjacent to nests. Each sample was placed in a plastic Whirl-pak® bag (510 g) and returned to the lab within 1-3 hours of collection in a cooler. Upon return to the lab, regurgitate samples were fiozen immediately. In the lab, regurgitates were thawed and analyzed. Each prey item was identified ‘ to species (or genera) and recorded. All of these prey items, including partially digested 53 (mostly intact prey), were individually weighed. Complete fish were measured to the nearest 0.5 mm. In some cases, otoliths and some bones, including jaws, pharyngeal bones, operculae, cleithra and vertebrae, were used to identify more digested prey. Using a reference collection (University of Michigan Museum and personal collection), these prey were identified to species (or genera) and counted. Because most bones and otoliths were eroded, no attempts were made to calculate original length and fresh mass of prey from these mostly digested samples. These methods are similar to those outlined in Carss et al. (1997). Regurgitate samples were then preserved in 70% ethanol. Birds used for the stomach analysis were collected using shotguns between 23 June and 08 August 2000, and between 31 May and 10 August 2001 (U SFWS Permit No. MB022886). These birds were typically harvested along a flight path as they returned to their breeding colonies. (Most birds were collected near Pismire and Grape Colonies.) A total of 53 birds were harvested in 2000, and 97 birds were harvested in 2001. After birds were collected, they were placed in plastic bags and frozen. Later, the birds were thawed and examined as outlined in Carss et al. (1997). The esophagus, crop and complete stomachs (proventriculus and pylorus) were removed from each bird and total mass of these organs and their contents were recorded. These organs were then dissected. and all prey items were removed and identified to species when possible. All prey items, including partially digested prey, were individually weighed. Complete fish were measured to the nearest 0.5 mm. Stomach contents were then preserved in 70% ethanol. In addition, each bird was sexed by examining reproductive organs. 54 Analysis Breeding DCCO population estimates for 2000 and 2001 were first tallied to describe the changes in the number of breeding DCCOs at each of the colonies as the breeding season progressed. By comparing early nest and egg counts with mid season nest, egg and chick counts at each colony, mean clutch size for the early season birds and. mean clutch size for mid season birds were calculated. These mean clutch sizes were used to calculate mean clutch size for all colonies combined for each season (early and mid) for each of the years (2000 and 2001). Both within and between year differences in mean clutch size were detected. The South Gull Island colony was not used to determine mean clutch size for either season or year. Many nests at this location are in trees and, although an accurate nest count was possible, it was difficult to detemrine the number of eggs and/or chicks in these nests. The mean clutch size data were then analyzed using a Single Factor Anova (Microsoft Excel). Because differences were detected, pair-wise t- tests were then performed to compare early and mid season clutch sizes for each year. In addition, between year comparisons for early season mean clutch sizes and mid season mean clutch sizes were also analyzed using t-tests (Microsoft Excel). A Bonferroni Correction was used to re-set appropriate significance levels for these t-tests (a = 0.008). The proportion of successful nests were determined by comparing the number of active nests at the beginning of the breeding season with the number of active nest remaining at the end of each breeding season for each of the colonies each year. These proportions were then analyzed using a t-test. In addition, the total reproductive output for breeding DCCOs in 2000 and 2001 were estimated by multiplying the mean clutch size with the number of active nests remaining at each of breeding season. Reproductive 55 output varied from colony to colony in each year. These values calculated by overall means were higher then individual colonies combined and may over represent the number of chicks fledged. Dietary data from each year (2000 and 2001) were analyzed separately. The regurgitate and stomach content samples were combined into one data set (see Chapter 2). In 2000, 248 samples were collected, which included 3493 food items (22,0405 g). In 2001, a total of 880 samples were collected, which included 10,512 food items (81,496 g). Initially, numerical frequencies of prey items in the samples were calculated for each. year and were converted to percentages (also referred to as percent numbers). Wires et al. (2001) defines percent number as the number of specimens of a taxon as a percent of all specimens in a sample. All colonies were combined for each year to determine the biomasses of prey items in both regurgitate and stomach content samples and these were converted to percents. Percent biomass is defined as the biomass of a taxon as a percent of total biomass (Wires et al. 2002). Biomass is considered a more appropriate method in determining potential impacts DCCOs may have on their prey (Cairns 1992). The biomass data set was then divided into two categories, the Main Archipelago (Pismire, Hog, Hat Islands) and Gull Island colonies. For the Main Archipelago data set, regurgitates and stomach contents samples were combined because all birds were collected near Pismire and Grape colonies. Main Archipelago and Gull Island data sets for each year were analyzed separately. In addition, the biomass data sets from the Main Archipelago for both 2000 and 2001 were examined to determine whether DCCO diet changes as the breeding season 56 progresses. Each data set was broken in to three phases: Pre—nesting and Incubation (20 April — 14 June), Nestling to Fledging (15 June to 31 July), and Post-Nesting (01 August — 10 September). Results The number of breeding DCCOs in the Beaver Archipelago has fluctuated over the past twenty years (Table 1). From 1984 to 1997, the breeding population increased from 250 to 11709 breeding pairs, over a 46 fold increase. However, between 1997 and 2000, the population declined by 13.5%, from 11709 to 10125 pairs. Between 2000 and 2001, the population size decreased by another 4.1%, from 10125 to 9705 pairs. In addition, between 2001 and 2002, the number of DCCO pair attempting to breed declined by another 31.4%, fi'om 9705 to 6657 pairs. However the actual number of successfiJl ‘ breeders in 2002 was much lower (see below). The number of breeding pairs increased between 2002 and 2003, from 6657 to 9119+ pairs. Because the nests were not counted on Gull Island in 2003, the magnitude of this increase remains uncertain. However, in 2004, the number breeding pairs again declined to 6407 (Table 1). The number of active breeding pairs tends to decline as the season progresses in the Beaver Archipelago (Figure 2). In 2000, the largest number of nests were counted at the onset of the breeding season during the early count season at each of the colonies, including Pismire, Grape, Hat and Gull (North and South, collectively). However, in 2001, this was the case at Gull Island colonies only. At Pismire and Grape, there were modest increases in breeding bird at the mid season count (28 and 34 nests, respectively). At Hat Island, this increase was more pronounced (340 nests) (Figure 2). Regardless, by 57 the late count date each year, the number of active breeders declined at all colonies. In 2000, the final number of active nests was 6652, a decrease of 34.3% from the early nest count. In 2001, the final number of active nests was 7642, a decrease of 21 .3% from the early nest count. The Pismire colony accentuates the year-to-year variability in the proportion on nests that survive to the late count dates (Figure 3). In 2000, the late count date shows a decline of 54.9% (from 967 to 436 nests). Comparably, in 2001, this decline is only 25.0% (from 1035 to 776 nests). During the 2002 breeding season, no nests survived to fledging on Pismire Island. (Similarly, only two nests of 1339 survived to fledging at the Grape Spit colony.) In 2003, the number of nests declined 30.8 % from early to late nest count dates (from 1167 to 807 nests). In 2004, the late count date shows a decline of 50.1% (fiom 725 to 362 nests) (Figure 3). Mean clutch size also varied from early to mid breeding season and from year to year in the Beaver Archipelago in 2000 and 2001 (Figure 4a). Single Factor Anova analysis indicated a significant difference among the mean clutch sizes as compared within each year and between years (F = 11.18755, df= 3, p = 0.000861). Pair-wise t- tests indicated that only the late season clutch size between year comparisons was significant (df= 3, p = 0.00279). Again, Pismire Colony illustrates the variability in mean clutch size from year to year (Figure 4b). In 2000, the mean clutch size of 1.86 was calculated. In 2001, mean clutch sized increased to 2.45 young per pair. Mean clutch size was not calculated in 2002 (few eggs and no chicks were produced). In 2003, mean clutch size was 2.13 and, in 2004, mean clutch size decreased to 0.557 (Figure 4b). 58 The mean number of successful breeders increased from 2000 to 2001 (Figure 5a). This increase, as examined by a t-test, was not significant (df = 3, p = 0.056781). Year to year variability in the number of successfiil breeders is apparent at Pismire colony (Figure 5b). Only an estimated 45% of the initial breeders at Pismire were successfiil in 2000. This estimate increased in 2001 to 75%. No breeders were successfiil in 2002. In 2003, however, an estimated 69% of the pairs were successfiil. Yet, in 2004, this declined to only 50% (Figure 5b). The estimated number of chicks produced in the Beaver Archipelago increase between 2000 and 2001, from 9659 chicks to 16786 chicks, respectively (Figure 6a). This is an increase in chick production of 174%. Year to year variability in chick production is illustrated by Pismire colony (Figure 6b). In 2000, an estimated 811 chicks. were produced. This increased to 1901 chicks in 2001. No chicks were produced in 2002. An estimated 1720 chicks were produced in 2003. This declined to 209 chicks in 2004 (Figure 6b). Figure 7 shows the percent of each prey type, by numerical frequency, when all colonies and all collection days are combined for each year separately (2000 and 2001). In all of the diet figures, the miscellaneous category may include small numbers or masses of species present as categories on other figures. Other species that were found include several rainbow smelt (Osmerus mordax), one burbot (Lota Iota), one smallmouth bass (Micropterus dolomieui) and one salmonid. In 2000, alewife and crayfish comprised 33.2% and 31.8% of the prey in the samples. Other prey included sculpin (Cottus spp.) (12.3%), nine-spine stickleback (Pungitius pungitius) (10.3%), spottail shiner (3.5%), johnny darter (Etheostoma nigrum) (3.1%), and brook stickleback 59 (Culaea inconstans) (3.1%), with some prey, including sucker (Catostomus spp.) and trout perch (Percopsis omiscomaycus), each comprised 1% or less of the samples (Figure 7a). In 2001, alewife comprised 58.1% of the samples, while nine-spine stickleback, sculpin and crayfish comprised 20.2%, 12.6% and 4.1%, respectively. Brook stickleback were present (2.0%) as were sucker, spottail shiner, trout perch and johnny darter (all at 1% or less) (Figure 7b). When all colonies and all collection days are combined in 2000, alewife comprised 55.8 % by biomass (Figure 8a). Crayfish, sucker and sculpin comprised 18.8%, 11.9% and 5.5% of prey biomass, respectively. All other prey items combined ' comprised a total of 10.3% (Figure 8a). Several trends are apparent when this data set was divided into Main Archipelago and Gull Island data sets. The Main Archipelago data set varies only slightly from the combined data set (Figure 8b). However, the Gull Island samples show an increase in the amount of alewife in the diet (77.6%) (Figure So). In addition, Gull Island samples have proportionally more sucker present (15.0%). The other prey in the diet of Gull Island DCCOs included nine-spine stickleback (4.1%) and sculpin (2.3%). No crayfish were present in Gull Island samples (Figure So). When all colonies and all collection days are combined in 2001, alewife became an increasingly important prey item by biomass at 77.1% (Figure 9a). Sucker and sculpin comprised 9.8% and 6.1% of prey biomass, respectively. All other prey items combined, including nine-spine stickleback and crayfish, comprised a total of 7.5% (Figure 9a). Again, when this data set was divided into Main Archipelago and Gull Island data sets, the Main Archipelago data set varies only slightly fi'om the combined data set (Figure 9b). However, the Gull Island samples again show an increase in the amount of alewife 6O in the diet (87.3%)(Figure 9c). In the Gull Island samples, sucker, sculpin and nine-spine stickleback have a combined 12.7% biomass. Again, crayfish were not present in Gull Island samples (Figure 9c). When the Main Archipelago biomass data set for 2000 was examined by collection date, several seasonal changes in the diet were apparent (Figure 10). During Pre-nesting/Incubation, alewife (49.0%) and spottail shiner (29.0%) comprised the majority of the samples. Other prey present included Johnny Darter (11.0%), sucker (7.0%), sculpin (3.0%) and nine-spine stickleback (3.0%)(Figure 10a). As the breeding season progressed to the Nestling/Fledgling stage, Alewife (54.0%) and crayfish (24.0%) dominated the diet. Sucker (1 1.0%) and sculpin (6.0%) comprised most of the remaining prey biomass (Figure 10b). However, during Post-nesting, crayfish biomass became more prevalent (36.0%). Alewife (22.0%), sucker (19.0%) and sculpin (8.0%) comprised much of the remaining biomass. The large miscellaneous category (11.0%) was comprised mostly of unidentifiable flesh. Much was suspected to be either alewife or sucker (Figure 10c). When the Main Archipelago biomass data set for 2001 was examined by collection date, again, several seasonal changes in the diet were apparent (Figure 11). During Pre-nesting/Incubation, nine-spine stickleback (38.0%), sculpin (30.0%) and alewife (22.0%) comprised the majority of the samples. All other prey combined comprised 10% of the biomass (Figure 11a). As the breeding season progressed to the Nestling/Fledgling stage, alewife (81.0%) dominated the diet; sucker comprised 10.0% of prey biomass (Figure 11b). During Post-nesting, alewife biomass declined to 70% and 61 sucker biomass increased to 16.0% (Figure 11c). In 2001, crayfish were not as important in DCCO diet as described by biomass. Discussion DCCO populations have monitored periodically over the last several decades and ‘ it has been recommended in recent management plans that population monitoring on the breeding grounds be continued at regular intervals (FEIS 2003). In the past, waterbird censuses have been conducted in the U.S. waters of the Great Lakes every eight to twelve years (Scharf 1978, Scharf and Shugart 1993, Cuthbert et al. 1997). In the Beaver Archipelago, an additional survey was conducted in 1984 (Ludwig et al. 1984). In this work, where breeding DCCOs were counted each reproductive season beginning in 2000 and ending in 2004, year-to-year variability in breeding population size is quite apparent. Although the earlier, more periodic population counts are invaluable in documenting the - recovery of these birds across the Great Lakes, the current DCCO breeding populations seem to be stabilizing in northern Lake Michigan (Seefelt and Gillingham 2004a). In addition, the within season trends emphasize the importance in timing DCCO censuses when estimating breeding population size, as described by Ewins et a1. (1995). As a reflection of a change in population size, the number of breeding birds at specific colony sites has fluctuated in the Beaver Archipelago from year-to-year. In other areas, such colony dynamics have been attributed to changes in suitable habitat, ofien due to species interactions, predation and/or human disturbance (Kury and Gochfeld 1975, Ellison and Cleary 1978, Verbeek 1982, Gotmark 1992, Cairns et al. 1998, Skagen et a1. 2001, Seefelt and Gillingham, 2004b). In addition, changes in water level have had an 62 impact on the breeding habitats of a wide variety of birds, including Piping Plovers (Charadrius melodus) (North 1986) and pelicans (Pyrovesti 1997). Lowering of lake level has been the trend in the Lake Michigan-Huron basin over the past several years, with the highest rate in lake level reduction occurring between 1998 and 2000, and the lowest water levels documented in 2003 (NCAA 2003). Such changes in lake level affect the shape and size of shoreline habitat. The number of available nest sites may, in part, regulate seabird populations (Croxall 1987). Although DCCOs in the Great Lakes may not be limited by the number available nest sites (Hatch and Weseloh 1999), trends . in colony location and size in the Beaver Archipelago are partially due to changes in availability of desirable nesting locations, which have been influenced by changing water levels (Seefelt and Gillingham 2004b). Not only has the breeding population fluctuated over the past several years in the study area, the reproductive success of these birds has also been variable. In 2002, a decline in reproductive success at the Grape Spit colony can be attributed, in part, to mammalian predators (Seefelt and Gillingham 2004b). However, the major trends across the study area are probably due to year-to-year climatic differences and changes in prey ' availability; such changes in prey availability could be due to colder water (and air) temperatures, as well as unrelated fluctuations prey population sizes. It was noted that during the 2002 field season, both air and water temperatures remained unseasonably low through early July (personal observation). In response, DCCOs at Pismire, SE Garden, and Gull North colonies, as well as the depredated Grape Spit colony, abandoned nesting attempts. Ring-billed Gulls (Larus delawarensis), Herring Gulls (Larus argentatus) and Caspian Terns (Stema caspia) showed similar trends (Seefelt and Gillingham 2004b). In 63 2004, both water and air temperatures were lower than average and a reduced reproductive output was documented at Pismire. Although many factors could be influencing the variability in nesting success and fledged chick numbers, it is apparent that weather patterns and lake water temperatures do impact DCCO colonies in the study . area. Croxall (1987) suggested that another factor that can regulate seabird populations is food shortages. Hatch and Weseloh (1999) remark that the local densities of birds may be affected by the temporal and spatial distribution of prey. A decline in suitable prey could have attributed to reproductive failure at some colonies and a decline in the number of successful breeding birds in 2002 and 2004. Due to lower water temperatures (personal observation), large numbers of adult alewife may not have returned to shallow waters during the prime DCCO breeding season during these years. Although alewife biomass in Lake Michigan has remained relatively stable between the early 19805 and 2003, (Madenjian et al. 2002, Madenjian et al. 2004), local availability of these fishes may have fluctuated. In addition, warm spring temperatures in 1998 led to moderately high levels of age-3 alewives in 2001; in fact, this year class dominated survey catches in 2001 (Madenjian et al. 2004). Age-3 alewife, due to their size, are an attractive prey size to DCCOs. The large number of alewife in the diet of DCCOs in 2001 as compared to 2000, could be in response to this increased availability of alewife of a preferred size. Although there is no evidence that during 2000 or 2001 DCCOs in the Beaver Archipelago were influenced by a food shortage, it does appear that birds shifted their diet based on food availability within the study area. Actual relative availability of prey in the Beaver Archipelago is not known, however, cormorants are opportunistic predators 64 feeding on a variety of prey species (Lewis 1929, Birt et al. 1987). It is likely that these birds shift their diet in response to prey availability, especially when schooling species of an appropriate size are present in large numbers in shallow waters. Ludwig et al. (1989) documented by comparing alewife trawling estimates with DCCO diet data that bird diet does shift based on the availability of alewife when years were compared. Further, the current study supports suggestions that cormorant reproductive success may be linked to alewife availability (Weseloh and Ewins 1994). Although the breeding population declined in 2001 as compared to 2000, mean clutch size and estimates of chicks produced increased Ludwig et al. (1989) found the regional differences diet were apparent when cormorants from several areas of the Upper Great Lakes were compared. For the most part, alewife were important prey in all regions, but other prey, such as yellow perch, lake Whitefish and centrarchids, occurred in the diet in some regions while these prey were uncommon or absent in other regions. These differences were assumed to exist because of differences in prey availability and habitat types. In the current study, such differences are apparent yet again, but on an even finer scale. In the Main Archipelago, open water schooling species such as alewife are prevalent in the diet of DCCOs; however, other species, more prevalent in more inshore habitats, including crayfish, are also quite common. In comparison, crayfish are completely absent in the diet of Gull Island DCCOs and alewife constitute a higher percentage of prey documented in the diet of these birds. Further, the local differences observed in bird diet supports that Main Archipelago and Gull Island birds to not mix during the breeding season. 65 This work supports other work in the Upper Great Lakes region the documented seasonal shifts in the diet of cormorants (Ludwig et al. 1989, Ludwig and Summer 1997, . Hatch and Weseloh 1999). These shifis in the diet may indicate changes in prey availability, but could also reflect prey choice based on other factors including handling time (Stickley et a1. 1992). In the Beaver Archipelago, the overall trend was that alewife became increasingly more important in the diet of DCCOs as the breeding season progressed. This trend is slightly different in 2000, when alewife biomass declined during the Post-nesting phase. Although this could reflect changes in alewife availability, this trend could also be explained by the fact that many birds harvested during this time period in 2000 were young, less experienced birds. These birds were collected near the . colony site, where crayfish are an abundant prey. Seabirds are an important aspect of the Great Lakes ecosystem (Hebert and Sprules 2002). Several studies have been conducted investigating the influence of fish- eating birds on fisheries both in Europe (Suter 1995, Warke and Day 1995) and in the Great Lakes (Madenjian and Gabrey 1995, Maruca 1997, Neuman et al. 1997, Schiavone 2001). Research suggests that waterbirds play central roles in marine food webs (Cairns 1992), and this probably holds true in Great Lakes community dynamics. The ability to estimate cormorant numbers and diet accurately will strengthen attempts to understand the importance of these birds as predators in this study area. These data can be used to develop bioenergetic models that can assess of the importance of avian predators in energy transfer and nutrient cycling in aquatic systems (Wiens and Scott 1975). The primary parameters incorporated into most bioenergetic models include calculated daily energy demands, food type and daily consumption, and population estimates (F owle 66 1997) and such models have been USCfiJl in estimating total fish consumption by avian populations. Although population and diet studies alone cannot answer complex questions as to the relationships among DCCOs and their prey, these are important steps, in combination with more detail information on avian foraging ecology and prey population dynamics, in investigating community level interactions. 67 Literature Cited Birt, V.L., T.P. Birt, D. Goulet, D.K. Cairns and WA. Montevecchi. 1987. Ashmole’ 5 halo: direct evidence for prey depletion by a seabird. Marine Ecology Progress Series 40:205-208. Cairns, D.K. 1992. Bridging the gap between ornithology and fisheries science: use of seabird data in stock assessment models. Condor 94:811-824. Cairns, D.K., R.L. Dibblee and P.-Y.Daoust. 1998. Displacement of a large Double- crested Cormorants, Phalacrocorax auritus, colony following human disturbance. Canadian Field Naturalist l 12:520-522. Carss, D.N , and the Diet Assessment and Food Intake Working Group. 1997. Techniques for assessing cormorant diet and food intake: towards a consensus view. Supplemento alle Ricerche di Biologia della Selvaggina 26:197-230. Croxall, JP. 1987. Conclusions. Pgs. 369-381 in Seabirds: Feeding Ecology and Role in Marine Ecosystems (J .P. Croxall, ed.) Cambridge University Press, Cambridge, Massachusetts, USA. Craven, SR. and E. Lev. 1987. Double-crested Cormorants in the Apostle Islands, Wisconsin, USA: population trends, foods habits and fishery depredations. Colonial Waterbirds 10(1):64-71. Cuthbert, F .J ., J. McKeaman and L. Wemmer. 1997. U.S. Great Lakes Tern and Cormorant Survey: 1997 Progress Report. Department of Fisheries and Wildlife, University of Minnesota, St. Paul. Ellison, L.N. and L. Cleary. 1978. Effect of human disturbance on breeding Double- crested Cormorants. Auk 95:510-517. Ewins, P.J., D.V.C. Weseloh, and H. Blokpoel. 1995. Within-season variation in nest numbers of Double-crested Cormorants (Phalacrocorax auritus) on the Great Lakes: implications for censusing. Colonial Waterbirds 18: 179-182. FEIS. 2003. Final Environmental Impact Statement: Double-crested Cormorant Management. U.S. Department of Interior Fish and Wildlife Service and U.S. Department of Agriculture APHIS Wildlife Service. Fowle, MR. 1997. Population dynamics, food habits, and bioenergetics of Double- crested Cormorants in Lake Champlain. MS. Thesis, Wildl. and Fish. Biol. Prog., School of Nat. Resourc, Univ. of Vermont. Gotmark, F. 1992. The effects of investigator disturbance on nesting birds. Current Ornithology 9:63-104. 68 Hatch, J .J . and D.V.C. Weseloh. 1999. Double-crested Cormorant (Phalacrocorax auritus). In. Poole, A., Gill, F. (Eds) The Birds of North America, Inc., Philidelphia, PA, No. 441, pgs. 1-36. Hebert, CE. and W.G. Sprules. 2001. The relevance of seabird ecology to Great Lakes Management. Journal of Great Lakes Research 28(1):91-103. Johnson, J.H., R.M. Ross and CM. Adams. 1999. Diet composition of Double-crested Cormorants in Eastern Lake Ontario, 1998. In: Final Report: T 0 Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division. Johnstone, I.G., M.P. Harris, S. Waneless and J.A.Graves. 1990. The usefiJlness of pellets for assessing the diet of adult Shags Phalacrocorax aristotelis. Bird Stuay 3725-11. Kury, CR. and M. Gochfeld. 1975. Human interference and gull predation in cormorant colonies. Biological Conservation 8:23-34. Lantry, B.F., TH. Eckert, and GP. Schneider. 1999. The relationship between the abundance of Smallmouth Bass and Double-crested Cormorants in the Eastern Basin of Lake Ontario. In: F inal Report: To Assess the Impact of Double-crested C ormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 12, pgs. 1-10. Lewis, H. G. 1929. The Natural History of the Double-crested Cormorant. Ottawa, Ontario: Ru-Mi-Loo Books. Ludwig, J. P. 1984. Decline, resurgence and population dynamics of Michigan and Great Lakes Double- Crested cormorants. Jack Pine Warbler 62291-102 Ludwig, J .P. and CL. Summer. 1997. Population status and diet of cormorants in Les Cheneaux Islands area. IN: History, Status, and Trends in Populations of Yellow Perch and Double-crested C ormorants in Les C heneaux Islands, Michigan. Ludwig, J.P., H.J. Auman, D.V.C. Weseloh, G.A. Fox, J.P. Giesy, and ME. Ludwig. 1995. Evaluation of the effects of toxic chemicals in Great Lakes cormorants: has causality been established? Colonial Waterbirds 18:60-69. Ludwig, J.P.,C.N. Hull, ME. Ludwig, and HI. Auman. 1989. Food habits and feeding ecology of nesting Double-crested Cormorants in the upper Great Lakes, 1986- 1989. The Jack-Pine Warbler 67(4):]15-126. 69 Madenjian, C.P., T.J. Desorcie, and J .D. Holuszko. 2004. Status and trends of prey populations in Lake Michigan, 2003. Report, Great Lakes Fishery Commission, Lake Michigan Committee Meeting, Ypsilanti, Michigan. Madenjian, C.P., G.L. Fahnenstiel, T.H. Johengen, T.F. Nalepa, H.A. Vanderploeg, G.W. Fleisher, P.J. Schneeberger, D.M. Benjamin, E.B. Smith, J.R. Bence, E.S. Rutherfors, D.S. Lavis, D.M. Robertson, D.J. Jude, and MP. Ebner. 2002. Dynamics of the Lake Michigan food web, 1970-2000. Can. J. Fish. Aquat. Sci. 59:736-753. Maruca, S.L. 1997. The impact of cormorant predation on yellow perch in Les Cheneaux Islands, Lake Huron. IN: History, Status, and Trends in Populations of Yellow Perch and Double-crested C ormorants in Les Cheneaux Islands, Michigan. J. S. Diana, G.Y. Belyea and RD. Clark, Jr., Eds. Michigan Department of Natural Resources, Fisheries Division, Special Report, Number 17 pgs. 47-70. Neuman, J., D.L. Pearl, P.J. Ewins, R. Black, D.V.C. Weseloh, M. Pike and K. Karwowski. 1997. Spatial and temporal variation in the diet of Double-crested Cormorants (Phalacrocorax auritus) breeding in the lower Great Lakes in the early 19905. Canadian Journal of Fisheries and Aquatic Sciences 54:1569-1584 National Oceanic and Atmospheric (NOAA). 2003. Water levels of the Great Lakes. NOAA Great Lakes Environmental Research Laboratory, Ann Arbor, Michigan. North, MR 1986. Piping Plover (Charadrius melodus) nest success on Mallard Island, North Dakota, U.S.A., and implications for water level management. Prairie Naturalist 1 8(2): 1 1 7- 122. Pyrovesti, M. 1997. Integrated management to create new breeding habitat for Dalmatian Pelicans (Pelecanus crispus) in Greece. Environmental Management 21(5):657-667. Scharf, WC. 1978. Colonial birds nesting on man-made and natural sites in the U.S. Great Lakes. Technical Report D-78-10, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi. Scharf, WC. and G.W. Shugart. 1998. Distribution and abundance of gull, tern, and cormorant nesting colonies of the Great Lakes, 1989 and 1990. Publication No. 1, Gale Gleason Environmental Institute, Lake Superior State University, Sault Ste. Marie, Michigan. Schiavone, A. 2001. Double-crested Cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2000 studies. NYSDEC Special Report: Section 1, pgs. 1-5. 70 Schiavone, A. 2003. Double-crested cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2002 studies. NYSDEC Special Report: Section 1, pgs 1-4. ‘ Seefelt, NE. and J .C. Gillingham. 2004a. The Double-crested Cormorant in Lake Michigan: A Review of Population Trends, Ecologyand Current Management. In: T. Edsall and M. Munawar, Ed. State of Lake Michigan: Ecology, Health and Management Ecovision World Monograph Series. Seefelt, NE. and 1C. Gillingham. 2004b. A new colony location for Double-crested Cormorants (Phalacrocorax auritus) and other waterbirds in the Beaver Archipelago, northern Lake Michigan. Michigan Birds and Natural History 11(3)]22-127. Skagen, S.K., C. P. Melchur and E. Muths. 2001. The interplay of habitat change, human disturbance and species interaction in a waterbird colony. American Midland Naturalist 145(1): 18-28. Stickley, A.R. Jr., G.L. Warrick and IF Glahn. 1992. Impact of Double-crested Cormorant depredations on Channel Catfish farms. J. World Aquaculture Society. 23(1): 192-204. Suter, W. 1995. The effect of predation by wintering cormorants Phalacrocorax carbo on grayling (Ihymallus thymallus) and trout (Salmonidae) populations: two case studies from Swiss rivers. Journal of Applied Ecology 32:29-46. Verbeek, N.A.M. 1982. Egg predation by Northwestern Crows: its association with human and Bald Eagle activity. Auk 99:347-3 52. Weseloh, D.V.C. and P.J. Ewins. 1994. Characteristics of a rapidly increasing colony of Double-crested Cormorants (Phalacrocorax auritus) in Lake Ontario: population size reproductive parameters and band recoveries. J. Great Lakes Research 20:443-456. Warke, G.M.A. and KR. Day. 1995. Changes in the abundance of cyprinid and percid prey affect the rate of predation by cormorants Phalacrocorax carbo carbo on salmon Salmo salar smolt in northern Ireland. Ardea 83:157-166. Wiens, J .A. and J.M. Scott. 1975. Model estimation of energy flow in Oregon coastal seabird populations. Condor 77 2439-452. Wires, L.R., F.J. Cuthbert, D.R. Trexel and AR. Joshi. 2001. Status of the Double- crested Cormorant (Phalacrocorax auritus) in North America. Final Report to U.S. Fish and Wildlife Service, Arlington, VA. 71 Table 1. Population estimates of breeding Double-crested Cormorants in the Beaver Archipelago, northern Lake Michigan from 1984 to 2004. SE Pismire Grape Timm’s Hat Whiskey Gull Garden Total 1984a 57 0 0 54 0 139 0 250 1989b 35 291 0 294 0 260 0 880 1997c 383 3509 753 4617 560 1887 0 11709 2000 987 2431 277 4917 0 1532 0 10125 2001 1035 2146 0 4511 0 2013 0 9705 2002 615 1339 0 3659 0 957 87 6657 2003 1164 0 0 7341 0 (7) 614 9119 2004 725 0 0 3 515 95 1274 798 6407 a Nest count data from Ludwig, 1984. b Nest count data from Scharf and Shugart, 1993. ° Nest count data from Cuthbert et al., 1997. 72 Squaw Island Garden Island Hat Island 0 Pismire / f Hog Island SE Garden N A Beaver Island O Whiskey Island 0 Trout Island 0 High Island Gull Island 9 0 18 Kilometers Figure 1: The Beaver Archipelago of northern Lake Michigan. 73 1200 . . . 3000 U) a % g 800 , g 2000 - '06 E12000 “5 1212000 E .2001 '5 I2001 E 400 - '2 1000 _ :I a Z 2 0 , 0 - Early Mid Late Early Mid Late c d. on % a 1800 — a g 4000 - Z '05 [32000 “5 120°_ ‘ £12000 L I- o I2001 3 .2001 ‘E‘ 2000- E 3 3 600 . Z Z O — o - . Early Mid Late Early Mid Late Figure 2. Number of active cormorant nests counted during the early, mid and late breeding season in 2000 and 2001, for a) Pismire, 1)) Grape, c) Hat, and d) Gull Island colonies. 74 1400 1200 — g 1000 - 02000 g 800 - 02001 I... .2002 g 500 ‘ 02003 5 400 A 02004 200 - 0 I I Early Mid Late Figure 3. Number of active cormorant nests counted during the early, mid and late breeding season from 2000 through 2004 at Pismire Island. Note that no nests remained at the end of the 2002 breeding season. 75 3.5 3 l a 25 i 275 .2000 Early f3 2 ' .2000 Mid g i .2001 Early 1.5 , . 5 .2001 Mid 0 s 1 _ 0.5 0 b. 3 2.5 — G N — J ‘2 2 02000 g 1 5 - .2001 E5 ' .2003 C 3 1 _ .2004 s 0.5 - o Figure 4. a) Mean clutch size (with standard error) for all colonies during the early and mid breeding season in 2000 and 2001. Late clutch sizes in mid 2000 and 2001 are significantly different from each other. b) Mean clutch size in the mid breeding season for Pismire Island in 2000, 2001, 2003 and 2004. 76 fi 1 13 0.8 i C E P. g 0.6— .2000 5 02001 E 0.4 - 8 2 0.2 - 0- 0 b. E 0.8 i n_n 0.6 — E 02000 02001 g 0‘4 I 02003 ‘6 02004 ‘5' 0.2 - E 8 8 n. 0 Figure 5. a) Mean number of successful breeding pairs (with standard errors) for all colonies during the 2000 and 2001 breeding seasons. b) Mean number of successful breeding pairs for Pismire Island in 2000, 2001, 2003 and 2004. 77 18000 a 15000- g 5 12000— B U 3 9000- IL ‘6 i 6000- E 3 3000- 0 2000 2001 b. 2000 0 § 1500 ‘ 02000 E U .2001 ‘6 1000 « E .2003 g .2004 :2 500- 0 Figure 6. a) Estimates for the number of chicks produced for pairs breeding at all colonies during the 2000 and 2001 breeding seasons. b) Estimates for the number of chicks produced for breeding pairs on Pismire Island in 2000, 2001, 2003 and 2004. 78 m 3005:0002 300.30 El 2000 I 2001 £0262 £30800 mEQmooEm w: EoumOumU 3:650 3200 mmfiocooao l——l mmo_< O 6 0 200 101 q — _ 0 O 0 5 4 3 1.x... 5:259.“— iota—:32 Figure 7. Numerical frequency data showing the diet of Beaver Archipelago cormorants 79 as percentages for 2000 and 2001. 9005:0022 smmso £0282 £8850 I Main Archipelago ElAll Colonies Gull Island w: anmoumo 35950 3:00 wmfimcooao as. nausea Figure 8. Biomass data showing the diet of Beaver Archipelago cormorants as percentages, 2000, for all colonies, Main Archipelago colonies, and Gull Island colonies. 8O El All Colonies 0 Main Archipelago Gull Island Biomass (%) P i a T Alosa Orconectes E Pungitius Catostomus Miscellaneous P Figure 9. Biomass data showing the diet of Beaver Archipelago cormorants as percentages, 2001, for all colonies, Main Archipelago colonies, and Gull Island colonies. 81 Biomass (%) 8 Alosa \:‘.,“ S3. . 1‘5 < . “f; x i SN >393» $ . a} \‘ \fx §fi> \\‘ 1§\‘ t\§ >\ 1.\\‘ :\\ Orconectes Cottus 5f. punQMUs El Pre-nestinglhcmation I Nestling/Fledgling Post-nesting Catostomus , _ Percopsis Notropis 1—1 Etheostoma .1 Miscellaneous L Figure 10. Biomass data showing the diet of Main Archipelago cormorants as percentages, 2000, for Pre-nesting and Incubation, Nestling and Fledgling, and Post- nesting periods 82 El Pre-nestinglhcrbation I Nestling/Fledgling Post-nesting Biomass (%) —4 —1 AIosa Orconectes cottus pungitius PercopSis NotroP'S Culaeu l Catostomus Miscellaneous :1 .— Figure 11. Biomass data showing the diet of Main Archipelago cormorants as percentages, 2001, for Pre-nesting and Incubation, Nestling and Fledgling, and Post- nesting periods 83 CHAPTER 4 USING RADIOTELEMETRY AND RAFTING LOCATIONS TO DETERMINE FORAGING LOCATIONS OF DOUBLE-CRESTED CORMORANTS (PHALACROC ORAX A URI T US) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN 84 Abstract During summer 2003, VHF radio telemetry was used to track the foraging activities of ten Double-crested Cormorants (Phalocrocorax auritus) nesting on Pismire Island or at the Southeast Garden Colony in the Beaver Archipelago, northern Lake Michigan. Leg . hold traps were used to capture birds and backpack radio-transmitters were harnessed to the birds. Using triangulation, birds were monitored from both land and water on a daily basis, weather permitting, throughout the breeding season. In addition, rafiing locations of cormorants were documented by boat throughout the breeding season. Radiotelemetery indicated that cormorants typically foraged 2.5 km away from the colony, at the northeastern end of Beaver Island. This area overlaps with the area determined by rafting locations, however the latter were centered firrther south. These data allow for better estimation on foraging patch use by cormorants and indicate the birds are not typically concentrating their foraging in areas where the Smallmouth Bass fishery has recently declined. Not only has this led to better understanding of cormorant foraging patterns, but these data also provide information concerning the activity budgets of archipelago birds. Together, these data can be used to estimate cormorant bioenergtics and, in combination with data on prey population dynamics, will help unravel the complex relationships between cormorants and their prey in the study area. 85 Introduction Over the past several decades, the population of Double-crested Cormorants (Phalacrocorax auritus), or DCCOs, has increased substantially in size throughout the Great Lakes Basin (Ludwig 1984, Cuthbert et al. 1997, Ludwig and Summer 1997, Wires et al. 2001), including the Beaver Archipelago of northern Lake Michigan (Seefelt and Gillingham, 2004a). The high density of birds combined with their observed fish eating behaviors have led to their implication in declines of both commercial (Ludwig et al., 1989; Neuman et al. 1997) and recreational fisheries throughout the Great Lakes region (N euman et al. 1997, Lantry et al. 1999). Cormorants are opportunistic fish predators that often feed in shallow waters (Lewis 1929, Birt et al. 1987) and breeding birds remain relatively close to breeding colonies when foraging (Custer and Bunck 1992). Although cormorants may have only small and localized effects on fish populations during migration (Kirsh 1995), it has been demonstrated that these birds deplete some species of prey fish around breeding colonies in marine systems (Birt et al. 1987). The Beaver Archipelago, and particularly the habitat around Garden and Hog Islands, has been considered in the past to have excellent smallmouth bass (Micropterus dolomieu) fishing, and this evaluation has been published in the national media a number of times (Robinson 1995). Cormorants have been documented to feed on smallmouth bass (Blackwell et al. 1997, Neuman et al. 1997, Schiavone 2001, 2003) with records of bass predation in the Beaver Islands (Ludwig et al. 1989). Seider (2003) attempted to estimate the population size of bass in the Beaver Archipelago (Seider 2003), although this species has not been documented as an important prey item in cormorant diet (see Chapter 3). However, compared to similar data gathered by researchers nearly 20 years ~ 86 39 80 Be 181 bu C0 SP1 113 Spi for the .\li 90' the P01 Opl ago (Lennon unpublished data, Seider 2003), the bass population has declined by 7 5- 80%. It is therefore quite clear that there has been a recent and rapid decline in the Beaver Archipelago bass fishery. However, the role of cormorants in this decline has remained uncertain. Cormorant diet often includes primarily species that are of little commercial value but important in community trophic dynamics (Craven and Lev 1987). Therefore, cormorants may have a secondary effect on sport fisheries by competing with desired species for forage fish and other prey such as crayfish. Although the effects on forage fish numbers may be limited and may only occur in localized areas (Madenjian and Gabrey 1995), this combined with direct sport fish depredation may have some impact on sport fish populations. A There are studies in which feeding flights were followed by airplane to learn foraging distances birds travel from the breeding colonies and potential home ranges of these birds (Ainley and Boekelheide 1990, Custer and Bunck 1992). However, in Lake Michigan, there is little information on the localization of foraging sites within such large potential areas. With these data lacking, it is not easy to ascertain the ecological impacts that large numbers of foraging birds may have aquatic communities, including fish populations which may often be localized within particular habitats (ie. littoral zone and ‘ open water). Waneless et al. (1995), Gremillet (1997) and Gremillet et al. (1998, 1999) have successfitlly used radiotelemetry to monitor shags and cormorants. These studies have allowed for better estimation on foraging patch use and seasonal variation, which, in turn, has lead to better understanding of behavior patterns. 87 The objective of this study was to determine the important foraging areas for breeding DCCOs in the Beaver Archipelago of northern Lake Michigan in 2003, by utilizing VHF radiotelemetry and by observing rafting areas throughout the breeding season. By doing so, this work will shed light on whether cormorants are concentrating their foraging in important bass habitats or in other habitat types. This work, in combination with other work on the foraging ecology and bioenergetics of DCCOs in the study area, will enable researchers to ascertain the impact of these birds on the fishery in the present, as well as possible future scenarios. Methods S tudy Site The study was conducted under permit (U .S. Department of the Interior Permit N o - 20852-03-89) in the Beaver Archipelago of northern Lake Michigan. This archipelago consists of about ten main islands and numerous small islands (Figure 1). The number of smaller islands depends on fluctuating lake levels. Two of the larger i 51 ands (Gull and Hat) and two small islands (Pismire and Southeast Garden) provided sites for nesting DCCOs in 2003. Pismire Colony (lat 45° 45.8’N, long 85° 26.6’W) and the Southeast (SE) Garden Colony (lat 45° 45.8’N, long 85° 27 .O’W) are relatively close t()gether (approximately 1.5 km or 0.92 miles apart) and are the focal colonies of this we rk- These colonies were chosen due to their central location within the archipelago and their close proximity to important habitat areas for smallmouth bass. In 2003, 1164 _ breeding pairs were established at Pismire Colony and 615 pairs were documented at SE 88 Gar M Roa 150 radi Dun sean Tell 101i hen: PTOH Garden Colony during June. By late July, the number of nests had declined to 807 and 355, respectively (Seefelt and Gillingham, 2004b). Telemetry VHF transmitters (Model RI-2C Backpack, Holohil Systems Ltd, 112 Cavanagh Road, Carp, Ontario, Canada, KOA 1L0; 6.4 g with whip antennae; fiequency range 150.026-150204 MHz) were used. Harnesses were constructed and attached to radiotransmitters in the lab prior to trapping birds using standard methods outlined in Dunston (1972). Harnesses were composed primarily of Teflon Tubing. At junctions, seams were fastened by stitches (needle and thread), fast-setting marine adhesive and Teflon-coated fiberglass tape. One connection was left free to allow for quick attachment to the birds in the field. The mass of the transmitter and harness was 20g, which is between 1-2% of the body weight of adult birds. On 04 June 2003, eight birds were trapped using soft-catch leg hold traps (model SN 3 5798, Forestry Suppliers, PO. Box 8397, Jackson, Mississippi, USA, 39284) on Pi smire Island. To insure the capture of experienced adult birds, these traps were placed at a well-established, central location in the colony. Once birds were captured, they were pro mptly removed from the traps and placed in wet pillowcases. Birds were then carried to another area on Pismire Island away from cormorant nests and out of site of nesting bi r <18 to be processed. Each bird was weighed and then the harness (with I.a(ii()transmitter) was attached. Seam construction followed the same method used in the [abol‘atory As soon as the harness was attached, birds were released. Handling was kept a - . . . . . . . t a mlnrmum. To mmrmrze colony disturbance at Pismire Island, on 05 June 2003, two 89 birds, for a total of ten, were trapped and harnessed using the same methods as at the SE Garden Colony. Birds were monitored via radiotelemetry from a boat on a daily basis, weather permitting, throughout the breeding season (04 June — 31 July) and during post-breeding ' (01 August — 13 September) using a receiver (model LA l2-Q, AVM Instrument Co. Ltd, 1213 S. Auburn, PO. Box 1898, Colfax. California, U.S.A., 95713; frequency coverage 150.000 — 151.999 MHz) and 8-element yagi aerial. It was difficult to sight birds, so triangulation was used to document bird locations, as follows. Once a bird’s signal was heard, the GPS location of the boat was taken (using a Garmin Handheld GPS Unit) and a compass bearing (Brunton‘D Outback Electronic Compass) based on the strongest signal strength was taken. This process was repeated two more times at two different locations within 10 minutes. In addition, date and time (beginning and ending) were noted. In some cases, signals were detected coming from the colonies. Ifthis was the case, birds were assumed to be at or near the colonies. To avoid disturbing birds, a more accurate location was not attempted. On several occasions, birds were actually sighted foraging in open water. When weather did not permit boat travel, birds were monitored from land on the northeast portion of Beaver Island, referred to as Sucker Point or Gull Harbor (lat 45° 44 ~ 9 ’ N long 85° 29.9’ W). From this station, it was noted whether birds were on colony (or near colony) or whether they were away, presumably foraging. In addition, birds Were also monitored from the southeastern portion of Garden Island (lat 45° 46.5’ N long 8 S 0 2‘7 ’ W) to the same end. In addition, from these data the general daily activity periods at the colony were determined. 90 W Beginning 18 May 2003 and ending 13 September 2003, cormorant rafting locations were marked by boat survey. In general, daily surveys of rafts were conducted on the northern and eastern waters around Beaver Island, weather permitting. Also, in- shore areas of Garden Island that are considered to be primary smallmouth bass habitat, including Garden Island Harbor, Northcut Bay and Sturgeon Bay, were also surveyed (Figure 2). Garden Island Harbor is 2.25 km (1.4 miles) from Pismire Island, Northcut Bay is 1 km (0.62 miles) from Pismire Island, and Sturgeon Bay is 0.5 km (0.3 miles) from Pismire Island. In addition, the Paradise Bay-St. James Harbor area of Beaver Island (3 km or 1.86 miles from Pismire Island) was also surveyed, as were the waters between Garden and Hog Island (north of Pismire Island, south of Grape Spit). These areas are also considered to be good habitat for smallmouth bass. In addition, on several occasions, surveys were conducted around the entire periphery of Beaver Island. Groups of rafting birds ranging from three to approximately 300 birds were located on these surveys. In some cases when rafts were large (exceeding 50 birds) it was necessary to estimate bird numbers as opposed to actually counting individuals. Each rafting site was marked using a Garmin hand-held GPS unit and an approximated wat er depth was noted (as read on a Humminbird Fish-finder Depth sensor). The date arnd ti me were also recorded. W Locations for each bird were determined using triangulation with Locate II (Nams, 2001. These data were then imported into ArcView GIS 3.3 and locations for all I 91 birds were plotted together. Then, Kernel Estimator: Extension Animal Movement (Hooge and Eichenlaub 1997) was used to determine primary foraging areas, based on the density of locations as determined by telemetry. This estimator provided 20% density contours based on the number of actual locations and determined the area where birds were most likely found when away from the colony; these areas were then compared to known smallmouth bass habitat, as documented by Lennon (unpublished data) and Seider (2003). Furthermore, the area within each density contour was then compared to water depth contours (nautical soundings map of Lake Michigan — Waugashance Point to Soul Choix Point, NOAA 1989). In addition, locations were divided into two data sets based on date: 09 June — 31 July (breeding) and 01 August - 07 September (post-breeding). Plots were compared to determine any seasonal differences. Rafting locations, as well as the number of birds per raft and date, were imported into ArcView GIS 3.3. These were plotted and then 20% density contours were determined, which weighted each raft location by the number of birds observed per raft. These areas were then compared to known smb habitats, and the area within each density contour was then compared to water depth contours (NOAA 1989). In addition, rafting locations were divided into three data sets based on date: 18 May —13 June (pre-nesting and incubation), 14 June — 31 July (nestling to fledgling) and 01 August — 10 September ' (po St-—breeding). Plots were compared to determine any seasonal differences. A total of 131 foraging locations were plotted for the ten birds with radio t - r3’11'181rntters from 09 June 2003 through to 07 September 2003 (Figure 2). All birds 92 remained in the Beaver Archipelago until at least 31 August 2003, with eight birds remaining in the area through 13 September 2003. A total of 223 on (or near) colony locations were documented by boat and an additional 185 were documented from Beaver Island. The Garden Island observations indicated the birds typically began leaving their colonies at 05:00 EST and all birds return to roost by 19:00 EST during the breeding season. Figure 3 shows the 20% density contour for all foraging locations marked using radiotelemetry. The harnessed birds tended to concentrate their foraging efforts in an area at the northeastern end of Beaver Island. This area is centered at about 2.5 km (1.55 miles) from the colony and is close to Luney Point and the Paradise Bay — St. James Harbor mouth. It is also an area with dramatic depth contours, with areas of only 1-3 meters dropping rather abruptly to areas greater than 18 meters. In addition, there are several shoals (NOAA 1989) and a channel used by ferry service to Beaver Island. Overall, birds did not show any seasonal changes in overall foraging locations, except for one bird that foraged very close to Pismire Island during June and part of July. These poi nts are not shown in Figure 3, because it was difficult to discern when this bird was on colony or foraging in the water near the colony. However, this bird did forage in the area shown in Figure 4 during the post-breeding time period. A total of 271 observed rafting locations, with an average of 33 birds per raft, are shown in Figure 4. Figure 5 shows the 20% density contours weighted by the number of b i rd S per raft. Interestingly, this area overlaps the foraging areas for birds with radiotransmitters. However, the highest density of weighted rafting locations is fitrther South and is centered near Conn’s Point and the north end of Sand Bay, Beaver Island. 93 Like the area defined by telemetry, the major rafting areas also have rather dramatic depth contours, with areas of 1- 3 meters dropping off to areas of 18 meters or more. The areas highlighted to the southeast and west of Beaver Island are due to a few large rafts (200+ birds) observed at these sites in July. Very few rafts were documented in in-shore areas of Garden Island, St. James Harbor — Paradise Bay, and between Garden and Hog Islands. Discussion Both VHF radiotelemetry and the observation of rafting sites produced a similar pattern in mapped foraging locations. The primary foraging areas for many of the DCCOs in the Beaver Archipelago is centered near the northeastern portion of Beaver Island, especially near dramatic depth contours. Cormorants nesting on Pismire and SE Garden Islands do appear to remain close to the colonies when foraging, but both methods indicate that important habitat areas for smallmouth bass are probably not used extensively by DCCOs in the study area; this includes areas relatively close to the colonies. Cormorants, therefore, are probably not directly competing with bass for local . Prey items. In addition, other studies indicate that birds do tend to forage in areas within a few kilometers of their colonies, as shown in DCCOs breeding in the Wisconsin waters Of Green Bay, Lake Michigan (Custer and Bunck 1992), and in Great Cormorants (Phalacrocorax carbo carbo) breeding at the Chausey Islands, France (Gremillet 1997). Cormorant foraging locations in the Beaver Archipelago seem to correspond with areas that are frequented by alewife (Alosa pseudoharengus). Alewife are presently a maj 0r planktivorous species in the lake community and an important energy link to upper 94 consumer levels. In addition, alewife are an important link between inshore and deep lake communities (Eck et al. 1987, Madenjian et al. 2004). During early summer, alewife begin to move inshore for spawning, only to then return to deepwater in fall. This cycle is repeated each year. Because of this life history and abundance, alewife are considered to be a key species in Lake Michigan and their population trends appear to be a driving force in fish community dynamics (Eck et al. 1987, Madenjian et al. 2002, 2004). Interestingly, in the study area, alewife are documented to be an important prey for cormorants and typically comprise an estimated 72% of the prey biomass consumed by breeding DCCOs and their chicks each year (Seefelt and Gillingham 2004a). Cormorants nesting at Hat and Gull Islands were not studied in this work. These colonies typically support more nesting pairs than either Pismire or SE Garden Colonies, with Hat Island consistently being the largest colony (Seefelt and Gillingham 2004a). Hat Island is relatively near some areas of the smallmouth bass habitat around Hog Island, as described by Seider (2003). However, open-water areas, specifically areas of depth dramatic depth contours are much closer to this colony. Furthermore, there has been little evidence of bass in the diet of these birds (see Chapter 3). The Gull Island Colonies are separated from the main island group by a stretch of open water to the west 0f B eaver Island. From aerial surveys (unpublished data), it was concluded that the CO r1Tlorants breeding at Gull Island remain close to these colonies and do not regularly tr ave] to the Main Archipelago. In addition, DCCO dietary data from Gull Island d<)cllt'nents that these birds are dependent on alewife (see Chapter 3). The primary foraging locations documented using birds harnessed with r ~ . . . . adlotransmrtters and rafting locations of DCCOs are concentrated in areas removed from 95 smallmouth bass habitat, as sampled by Seider (2003). DCCO foraging areas, as determined by telemetry and rafting sites, do support bird dietary data gathered in the study which shows birds primarily feed on alewife and other species unimportant to sport and commerial fisheries. The data presented in this work will be used fitrther to develop, in part, a time-budget analysis bioenergetics model. These data can then be used, in combination with data on prey population dynamics, to investigate the complex relationships between cormorants and their prey in the study area. 96 Literature Cited Ainley, D.G. and RI. Boekelheide, eds. 1990. Seabirds of the Farallon Islands. Stanford Univ. Press, Stanford, California. Birt, V.L., T.P. Birt, D. Goulet, D.K. Cairns, and WA. Montevecchi. 1987 . Ashmole’ 5 halo: direct evidence for prey depletion by a seabird. Marine Ecology Progress Series 40:205-208. Blackwell, B.F., W.B. Krohn, NR Dube, and A]. Godin. 1997. Spring prey use by Double-crested Cormorants on the Penobscot River, Maine, USA. Colonial Waterbirds 20(1):77-86. Craven, SR. and E. Lev. 1987. Double-crested Cormorants in the Apostle Islands, Wisconsin, USA: population trends, foods habits and fishery depredations. Colonial Waterbirds 10(1):64-71. Custer, T.W. and C. Bunck. 1992. Feeding flights of breeding Double-crested Cormorants at two Wisconsin colonies. Journal of Field Ornithology 63(2):203- l 211. Cuthbert, F.J., J. McKeaman and L. Wemmer. 1997. U.S. Great Lakes Tern and Cormorant Survey: 1997 Progress Report. Department of Fisheries and Wildlife, University of Minnesota, St. Paul. Dunston, TC. 1972. A harness technique for radio-tagging raptorial birds. Inland Bird Banding News 44:4-8. Eck, G.W. and EH. Brown. 1987. Recent changes in Lake Michigan’s fish community and their probable causes, with emphasis on the role of the alewife (Alosa pseudoharengus). Can. J. Fish. Aquat. Sci. 44:53-60. Gremillet, D.G. 1997. Catch-per-unit effort, foraging efficiency, and parental investment in breeding Great Cormorants (Phalacrocorax carbo carbo). ICES J. ' of Marine Science 54:63 5-644. GTernillet, D., G. Argentin, B. Schulte and BM. Culik. 1998. Flexible foraging techniques in breeding cormorants Phalacrocorax carbo and shags Phalacrocorax aristotelis: benthic or pelagic feeding? Ibis 140:113-119. C"Tetrlillet, D., R.P. Wilson, S. Storch, and Y. Gary. 1999. Three-dimensional space utilization by a marine predator. Marine Ecol. Prog. Ser. 183 :263 -273. 97 Hooge, P.N and B. Eichenlaub. 1997. Animal movement extention to ArcView. version 1.1. Alaska Biological Science Center, Geological Survey, Anchorage, AK, USA. Kirsch, EM. 1995. Double-crested Cormorants along the upper Mississippi River. Colonial Waterbirds 18(Special Issue l):131-136. Lantry, B.F., T.H. Eckert, and C. P. Schneider. 1999. The relationship between the abundance of Smallmouth Bass and Double-crested Cormorants in the Eastern Basin of Lake Ontario. In: Final Report: To Assess the Impact of Double-crested Cormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 12, pgs. 1-10. Lewis, H. G. 1929. The Natural History of the Double-crested Cormorant. Ottawa, Ontario: Ru-Mi-Loo Books. Ludwig, J. P. 1984. Decline, resurgence and population dynamics of Michigan and Great Lakes Double- Crested cormorants. Jack Pine Warbler 62:91-102 Ludwig, J .P., C.N. Hull, ME. Ludwig, and H]. Auman. 1989. Food habits and feeding ecology of nesting Double-crested Cormorants in the upper Great Lakes, 1986- 1989. The Jack-Pine Warbler 67(4), 115-126. Ludwig, JP. and C. L. Summer. 1997. Population status and diet of cormorants in Les Cheneaux Islands area. IN: History, Status, and Trends in Populations of Yellow Perch and Double-crested C ormorants in Les Cheneaux Islands, Michigan. Madenjian, C.P., T.J. Desorcie, and JD. Holuszko. 2004. Status and trends of prey populations in Lake Michigan, 2003. Report, Great Lakes Fishery Commission, Lake Michigan Committee Meeting, Ypsilanti, Michigan. Madenjian, C.P., G.L. Fahnenstiel, TH. Johengen, T. F. Nalepa, H.A.Vanderploeg, G.W. Fleisher, P.J. Schneeberger, D.M. Benjamin, E.B. Smith, J.R. Bence, E.S. Rutherfor, D.S. Lavis, D.M. Robertson, D.J. Jude, and MP. Ebner. 2002. Dynamics of the Lake Michigan food web, 1970-2000. Can. J. Fish. Aquat. Sci. 59:736-753. Madenjian, CF. and SW. Gabrey. 1995. Waterbird predation of fish in western Lake Erie: a bioenergenics model application. The Condor 97 : 141-153. Nam 3, V0. 2001. Locate II: A program to triangulate radiotelemetry bearings and calculate error ellipses. Department of Environmental Science, NSAC, Truro, Nova Scotia, Canada. 98 National Oceanic and Atmospheric Administration (NOAA). 1989. United States - Great Lakes — Lake Michigan — Michigan: Waugoshance Point to Seul Choix Point Nautical Sounding Chart, 17th ed. US Department of Commerce, Washington DC, USA. Neuman, J., D.L. Pearl, P.J. Ewins, R. Black, D.V.C. Weseloh, M. Pik, and K. Karwowski. 1997. Spatial and temporal variation in the diet of Double-crested Cormorants (Phalacrocorax auritus) breeding in the lower Great Lakes in the early 1990s. Canadian Journal of Fisheries and Aquatic Sciences 54: 1569-1584. Robinson, J. 1995. Lake Michigan Smallmouths. Field and Stream (April) 16-22. Schiavone, A. 2001. Double—crested Cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2000 studies. NYSDEC Special Report: Section 1, pgs. 1-5. Schiavone, A. 2003. Double-crested cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2002 studies. NYSDEC Special Report: Section 1, pgs 1-4. Seefelt, NE. and Gillingham, J .C. 2004a. The Double-crested Cormorant in Lake Michigan: A Review of Population Trends, Ecology and Current Management. In State of Lake Michigan: Ecology, Health and Management (T. Edsall and M. Munawar, ed.) Ecovision World Monograph Series. S eefelt, NE. and Gillingham, J.C. 2004b. A new colony location for Double-crested Cormorants (Phalacrocorax auritus) and other waterbirds in the Beaver Archipelago, northern Lake Michigan. Michigan Birds and Natural History 11(3)122-127. S eider, M.J. 2003. Population dynamics of Smallmouth Bass in the Beaver Archipelago, northern LakeMichigan, 1999 -— 2002. MS. Thesis, University of Georgia. Wanless, S., M.P. Harris, and J .A. Morris. 1995. Factors affectiong daily activity budgets of South Georgian Shags at chick rearing at Bird Island, South Georgia. Condor 97:550-558. Wi res, L.R., F.J. Cuthbert, DR Trexel, and AR. Joshi. 2001. Status of the Double- crested Cormorant (Phalacrocorax auritus) in North America. Final Report to U.S. Fish and Wildlife Service, Arlington, V.A. 99 Squaw Island 0 Hat Island Garden —sland O _o _og Island SE Garden N High Island A Gull Island Beaver Island 9 0 9 18 Kllometers Figure I. The Beaver Archipelago of northern Lake Michigan. 100 SE Garden 1 0 1 2 Kilometers Figure 2. The bays of Garden Island and their proximity to Pismire and SE Garden Colonies. 101 Smallmouth 0 Bass Habitat / . # Telemetry Locations 8 0 8 18 Kilometers e E Figure 3. Bird locations determined by radiotelemetry and triangulation. 102 O o 0 Paradise Bay- 0 St. James Harbor _ , \ Density of Locations - 2° - .0 60 8 5 0 5 10 Kilometers r I Figure 4. Weighted density contours for foraging areas as determined by Iadlotelemetry. 103 Smallmouth ' 0 Bass Habitat /' . 8 0 8 18 Kilometers Figure 5. Locations of cormorant rafting sites as determined by boat surveys. 104 >2 Weighted Raft Density . .., - .0 so so 5 o 5 10 lfllometers r fl Figure 6. Weighted density contours for foraging areas as determined by boat survey. 105 CHAPTER 5 BIOENERGETICS AND PREY CONSUMPTION OF BREEDING DOUBLE-CRESTED CORMORANTS (PHALA CROC ORAX A URI T US) IN THE BEAVER ARCHIPELAGO, NORTHERN LAKE MICHIGAN 106 Abstract Colonial waterbirds are important component of Great Lakes ecosystems. One important aspect is the role of these birds as top predators in aquatic food webs. In order to investigate this role, bioenergetics models, using allometric equations, were applied to breeding Double-crested Cormorants (Phalacrocorax auritus) and their offspring in the Beaver Archipelago, northern Lake Michigan. These models were parameterized using detailed information collected during the breeding seasons between 2000 and 2001, as well as literature values. The breeding season was divided into stages in the models to reflect changes in cormorant diet documented in the study area. The models estimated the total prey biomass consumed as 1444.11 tonnes of prey in 2000, and 1586.17 tonnes ‘ of prey in 2001. Each year the majority of the prey biomass was alewife (Alosa pseudoharengus), with alewife comprising a greater percentage of prey biomass in 2001. An increase in cormorant reproductive success in 2001 may be linked to this increase in al ewife biomass; the breeding bird population size, however, declined in 2001 as compared to 2000. The other prey items, which are not considered to be species of commercial or sport value, were also important contributors to bird diet and did tend to vary from year-to-year. Overall, the application of bioenergetics models allow for greater understanding of the role of cormorants as predators and as energy links in the system. 107 Introduction Over the last several decades, the Double-crested Cormorant (Phalacrocorax auritus), or DCCO, has substantially increased in numbers throughout the Great Lakes Basin (Ludwig 1984, Cuthbert et al. 1997, Ludwig and Summer 1997, Wires et al. 2001), including the Beaver Archipelago of northern Lake Michigan (Seefelt and Gillingham, 2004). The high density of birds combined with their observed fish eating behaviors have led to their implication in declines of both commercial (Ludwig et al. 1989, Neuman et al. 1997) and recreational fisheries throughout the Great Lakes region (Neuman et al. 1 997, Lantry et al. 1999). Cormorants are opportunistic predators (Lewis 1929, Birt et al. 1987) that feed primarily on species of little commercial or sport value, but still i mportant in community trophic dynamics (Craven and Lev 1987). Although the effects on non-game forage fish populations may be limited and may only occur in localized areas (Madenjian and Gabrey 1995), cormorant predation on these fish may have some i m pact on overall aquatic community structure. Waterbird bioenergetics can assess of the importance of avian predators in energy transfer and nutrient cycling in aquatic systems (W ions and Scott 197 5). Models are used to calculate the food requirements necessary for a bird’s daily energy expenditure and thus, can estimate total prey consumption (Cairns et a1. 1991). By this quantification, the demands avian predators place on aquatic communities can be determined (Madenjian and Gabrey 1995). Since cormorants are wide-ranging generalists, it is difficult to measure energy demands directly. However, many indirect modeling approaches have been developed which estimate energetics of metabolism and adjust for other factors that may influence energy requirements (W iens 1984). 108 The primary parameters incorporated into bioenergetic models include calculated daily energy demands, food type and daily consumption, and population estimates (F owle 1 997). Direct measurements of wild bird metabolism have been be used to develop allometric equations that predict avian metabolic rates (Kendeigh et al. 1977). In addition, techniques such as doubly-labeled water and time-budget methods can be used to directly measure metabolism of free ranging birds (Nagy 1989, Birt-Friesen et al. 1 989). Many indirect modeling approaches have been developed to extrapolate individual metabolic estimates to whole populations (Wiens 1984, Cairns et a1. 1991, Glahn and Brugger 1995, Madenjian and Gabrey 1995, Fowle 1997). In all models, energy consumption was based on metabolism. Furthermore, these models have applied specific caloric requirements to daily energy expenditure to estimate food consumption, and have been useful in estimating total fish consumption by avian populations. Several studies have been conducted investigating the influence of fish-eating bi rds on fisheries both in Europe (Suter 1995, Warke and Day 1995) and in the Great Lakes (Madenjian and Gabrey 1995, Maruca 1997, Neuman et al. 1997, Schiavone 2001, 2 003). Research suggests that waterbirds play central roles in marine food webs (Cairns 1 992), and this probably holds true in Great Lakes community dynamics. In order to get a realistic picture of the impact bird predators have on fish populations, it is necessary to incorporate quantitative data on the diets, population size and energy requirements of the co rmorant population. In addition, reliable data on the size of the fish populations, with the appropriate spatial and temporal scales, is necessary (Draulans 1988). Yet even with the latter lacking, bioenergetics approaches can still aid in uncovering the energy demands placed on aquatic systems by avian predators. This can then allow for further 109 investigation to explore possible impacts these predators may have on their prey and to compare the prey consumption by birds to other sources of mortality. In this work, a bioenergetics approach based on allometric equations was used to investigate the energy demands and estimate annual prey consumption of breeding DCCOs and their chicks in the Beaver Archipelago, northern Lake Michigan during the 2000 and 2001 breeding seasons. The biomass of individual prey species consumed by birds were estimated each year and compared. Waterbird predation on fish and other species is an important link in the northern Lake Michigan web and should allow a fithher understanding other ecology of this system. In addition, this work can be directly compared to work in other systems, such as western Lake Erie (Madenjian and Gabrey l 995, Hebert and Morrison 2003), in an effort to understand the role of cormorants in Great Lake ecology. Methods S tudy Area The Beaver Archipelago of northern Lake Michigan consists of about ten islands. Three of the larger islands (Gull, Hog and Hat Islands) and one small island (Pismire Island) contained nesting colonies of DCCOs in 2000 and 2001 (Figure 1). The Hog Island colonies were located on two peninsulas known as Grape Spit, on the west side of ~ the i S] and, and Timm’s Spit, on the east side of the island. Population and dietary data collected in the Main Archipelago in 2000 and 2001 were used in the construction of the bioenergetics models. The Main Archipelago colonies include only Pismire, Grape, Timm’ s and Hat. Gull Island consists of two colonies, North and South, on one island 110 and is separated from the main island group by almost 18 km (11 miles) of open water to the west of Beaver Island. F rom aerial surveys (unpublished data), it was concluded that the cormorants breeding at Gull Island remain close to these colonies and do not regularly travel to the Main Archipelago. The purpose of these models was to determine total prey consumption by breeding DCCOs and their young in the Main Archipelago in 2000 and 2001, and compare these years. The models were limited to the Main Archipelago because this was the area of concern in respect to fish populations. Model Construction The major model parameters include (1) seasonal arrival/departure dates, (2) population estimates of adult and chicks, (3) diet data, and (4) estimates of daily energy expenditure (DEE). Population estimates and diet data were collected in the study area by the author (See Chapter 3) and, whenever possible, model inputs were measured in the H eld each year. Some model inputs, however, were taken from the literature. A complete list of model parameters for each year is shown in Table 1. To provide better accuracy in the model, the breeding season was divided into three categories: pre-nesting/incubation stage (20 April —— 12 June), nestling stage (13 June - 31 July), and post-nesting stage (01 August -10 September). Diet data were categorized by date to reflect the changes in DCCO diet as the breeding season Progresses, and were used to determine the biomass of individual prey species consumed by cormorants over the breeding season (Table 2). Models were applied to each year’s data separately. 111 The bioenergetic models follow a procedure outlined by Madenjian and Gabrey (1995) and then modified to include individual prey species by Hebert and Morrison (2003); however, in this work, models were developed using a spreadsheet (Microsofi Office Excel”). The models are dependent on allometric equations that use body mass in order to determine DEE (daily energy expenditure) for breeding adult DCCOs and their chicks. These equations have been successfully used in the past (Birt-Friesan et al. 1989, Cairns et a1. 1991, Madnejian and Gabrey 1995, Hebert and Morrison 2003). The equation for adults is as follows (Birt-Friesen et al. 1989): DEE = 1737.8W (”27 A body mass (W) for adult cormorants was determined by averaging the 150 birds harvested for the dietary study during 2000 and 2001. The daily energy intake (DEI) is determined by dividing DEE by the assimilation efficiency (given as 0.80 by Fumess 1 978). In addition, the bioenergetics of egg production was included in the model f0 l lowing the work of Kendeigh et al. (1977). The greatest number of nests counted each year was used to best estimate the maximum energy input needed for egg production. Daily food consumption (DFC) is found by dividing DEI by the average cal oric density (ACD) of bird diet. This is calculated using the following equation: n ACD = Z (CDi) * (PROPi) i = 1 Where CD1: calorie density (kcal/kg) of diet category I , PROPi = the proportion of the bird diet (by mass) comprised of prey type i, and n = total number of diet categories. Daily food consumption (DFC) is found by dividing DEI by the average caloric density 112 (ACD) of bird diet for each stage of the breeding season. In addition, the total DEI for all adult birds was summed throughout each segment of the breeding season. The DEI of individual prey types was then determined by multiplying total cormorant DEI by the proportion of occurrence a particular prey species appears in the diet by mass during each stage of the breeding season. Then prey DEI was multiplied by energy content conversion factors for each prey type (Table 3, Cummins and Wuycheck 1971, Bryan et al. 1996) to determine the biomass of each prey type consumed. These values were calculated for each time period of the breeding season (pre-nest/incubation, nestling and post-breeding stages) and then combined to determine the total biomass (tonnes) of prey consumed for each year by adult birds. The time period for cormorant chick models were divided into two segments, pre- fledging and post-fledging, which corresponded to the nestling and post-breeding time segments, respectively, for adult birds. To simplify the model, all eggs hatched si multaneously and all young fledged on the same date. An allometric equation (fiom Kendeigh et al. 1977) is used to determine the daily energy expenditure of a pre-fledged bird (DEEN): DEEN = 1.23ow 07749 Again, an assimilation efiiciency, given as 0.80 by Fumess 1978, was used. Daily food consumption by a pre-fledged bird (DFCN) is determined by the following equation: DFCN = DEEN * PROPi + DG * PROP; 0.80 * ACD 0.80 S i nce pre-fledged birds are growing throughout the season, this equation includes a daily gr OWth increment (DG), which is equal to the food needed to increase the mass of the 113 bird. Growth rate per day is considered to be a linear relationship and modeled as described in Madenjian and Gabrey (1995). Since young birds continue to grow and gain mass after fledging, this same equation was used to determine post-fledged birds’ DFC, however growth was modeled at a slower rate. The total consumption of prey biomass by chicks was determined using the same method used in adult populations. These biomass values were added to adult values to determine total biomass of each prey type consumed for each time period of the breeding season and also yearly totals. In addition, the total biomass of prey consumed by a single chick that survived the entire simulation for each ‘ year was calculated by summing the ACD for one chick during both the pre-fledging and post-fledging time periods. S ensitivity Analysis Sensitivity analysis was performed to determine which model parameters were most influential in calculating total DEI and, therefore prey consumption, by breeding co morants and their chicks. An individual parameter perturbation method was used (B artell et al. 1986, Madenjian and Gabrey 1995). Thirteen model inputs were examined. for a total of 26 simulations. For each sensitivity analysis simulation, only one model input was changed while all other parameters were left at their original values. A Si mulation was conducted for an individual model input by raising it fi'om its original value by 10%; a second simulation was performed for that same model input by lowering its value 10% from its original value. The outputs from these simulations were compared to the original model’s output to determine percent change. 114 Results The models calculated the DFC (daily food consumption) per adult breeding bird in 2000 based on the ACD (average caloric density) of the diet to be 0.55 kg during the pre-nesting/incubation period, 0.55 kg during the nestling period, and 0.65 kg during the post-breeding period (Table 4). The DFC values per adult bird for 2001 were 0.55 kg, 0.48 kg and 0.51 kg, respectively (Table 3). For chicks in 2000, the DFC per chick based on ACD of the diet ranged from 15.0 g to 320.8 g from hatching to the end of the pre- fledging period (Figure 2a) for a total of 9.0 kg of food consumed/chick (Table 5). For the post-fledging period, DFC per chick in 2000 ranged from 0.60 kg at fledging to 0.65 kg, when chicks attained adult size (Figure 2b), for a total of 26.0 kg of food consumed/chick (Table 4). In 2001, the DFC per chick ranged from 13.2 g to 281.4g from hatching to the end of the pre-fledging period (Figure 2a) for a total of 7.9 kg of food consumed/chick (Table 5). For the post-fledging period, DFC per chick in 2001 ranged from 0.47 kg at fledging to 0.51 kg, when chicks attained adult size (Figure 2b), for a total of 20.4 kg of food consumed/chick (Table 5). According to modeling results, breeding DCCOs and their young consumed a total of 1444.11 tonnes of prey in 2000. Of this, 426.73 tonnes were alewife (Alosa pseudoharengus), 354.11 tonnes were crayfish (Orconectes spp.), 59.75 tonnes were sculpin (Cottus spp.), 13.19 tonnes were nine-spine stickleback (Pungitius pungitius), 300.97 tonnes were sucker (Catostomus spp.), 167.50 tonnes were spottail shiner (Notropis hudusonius), and 121.87 tonnes of other prey (Figure 3). In 2001, adult and young birds consumed a combined 1586.17 tonnes of prey. Alewife comprised 695.78 tonnes, crayfish 55.29 tonnes, sculpin 369.37 tonnes, nine-spine stickleback 171.96 115 tonnes, sucker 271.23 tonnes, spottail shiner 5.48 tonnes, and 17.06 tonnes of other prey (Figure 3). For both years, the other category contains primarily johnny darter (Etheostoma nigrum), trout-perch (Percopsis osmiscomaycus), and brook stickleback (Culaea inconstans). Some prey items, such as spottail shiner, were common in the pre- nesting/incubation diet samples but uncommon during the other stages. Because of their rarity, when spottail shiner were found in the nestling or post-breeding diet, they were included in the other category. Sensitivity analysis revealed that adult and young assimilation efficiency, the number of nests, and adult mass were most influential in determining model output (Table 5). The other model inputs had relatively low impacts on model results, with changes two parameters, calories per egg and mass at hatching, showing virtually no influence on model output. Overall, the models were fairly robust in regard to any uncertainty in model inputs. Discussion According to the bioenergetics models, breeding DCCOs and their young consumed an estimated 1444.11 tonnes and 1586.17 tonnes of prey during 2000 and 2001, respectively. During both years, alewife contributed the greatest biomass to these totals, relative to other prey species. Using an average mass of 14 g for alewife, as detemiined by dietary analysis (see Chapter 3), in 2000, it is estimated that cormorants in the Main Archipelago consumed over 30.4 million of these fish. In 2001, it is estimated that over 49.6 million alewives were consumed. In addition, in 2001, alewife comprised over one and a half times the biomass to the birds’ diet as compared to 2000. Alewife 116 biomass in Lake Michigan has remained relatively stable between the early 1980s and 2003 (Madenjian et al. 2002, Madenjian et a1. 2004), however local availability of these fishes may have fluctuated between 2000 and 2001. In addition, according to Madenjian et al. (2004), wann spring temperatures in 1998 led to moderately high numbers of age-3 alewives in 2001 in Lake Michigan; these fish are an attractive prey size for DCCOs. In addition, alewives, as compared to other available prey, have a higher energy density (kcal/kg) (Cummins and Wuycheck 1971, Bryan et al. 1996). This, as well as their size and schooling habits, may make these fish quite a desirable prey to cormorants and explain why birds rely on alewife as a major energy source. In addition to changes in alewife biomass consumed, several other prey species showed year-to-year variability in terms to their importance to cormorant diet. Crayfish and spottail shiner were both prominent in terms of biomass consumed in 2000. However, in 2001, the biomass each species contributed to the diet sharply declined. Interestingly, the other category, which includes only fish, also showed a sharp decline in 2001. This emphasizes the importance of each prominent prey type in 2001, and, likewise, highlights the importance of alternative prey in 2000 in contributing to the overall energy demands of the breeding DCCO population and their offspring. In comparison, both nine-spine stickleback and sculpin biomass increased in importance as contributors to DCCO diet in 2001 as compared to 2000. Only one prey type, sucker, remained relatively constant in terms of biomass when both years are compared. Interestingly, the energy density of fish, with the exception of suckers, is higher then crayfish (Cummins and Wuycheck 1971, Bryan et al. 1996). 117 The number of breeders declined from 2000 to 2001, however the biomass of prey consumed increased according to the model. This increase in prey biomass consumed is due to higher reproductive success (more chicks) in 2001, as compared to 2000. Hatch and Weseloh (1999) remark that the local densities of DCCOs may be affected by the temporal and spatial distribution of prey. Number of young produced by each pair each year (reproductive success) could also be influenced by yearly prey availability patterns; Wesoloh and Ewins (1994) suggest that cormorant reproductive success may be intimately linked to Alewife populations. Interestingly, the DFC for both adult birds and chicks was higher in 2000 than in 2001, and therefore, so was the biomass of prey necessary to raise a chick to fledging. In addition, DFC was shown to change as the breeding season progressed during both years in proportion to ACD of the diet. Yearly variations are due to the overall greater energy ' density of the diet in 2001, as compared to 2000. In essence, less prey were consumed per individual in 2001, but more chicks were successfully raised to fledging. This is reflected in the relatively modest increase of total prey biomass consumed in 2001, even with more chicks produced. From the model, it appears that comparing energy density of prey types consumed, and also prey availability, may be better parameters in forecasting of the number of chicks successfully fledged per year, as opposed to strictly relying on the number of nests. In earlier works (Madenjian and Gabrey 1995, Hebert and Morrison 2003), there 1 was an attempt to model the importance of non-breeding cormorants in the system, as well as the migratory seasons. This is not the case in this work. In the Beaver Archipelago, it was difficult to ascertain the population of non-breeders during the 118 breeding season. When plumage patterns and gonad development (harvested birds only) were used to determine maturity, overall very few non-adults were documented in flocks and harvested birds (unpublished data). An exception to this was the post-breeding season, when some birds collected were fledglings. Better estimation of the number of non-breeding birds during the breeding season and the inclusion of migration would enhance this work. Further work is necessary to fully estimate the biomass of prey consumed by all DCCOs in the system. As with the work of Madenjian and Gabrey (1995), better estimates of prey fish populations is necessary to fiilly understand the impact the breeding DCCOs and their young may have on these prey populations in the study area. Recently, an attempt has been made to estimate the population size of smallmouth bass (Micropterus dolomieui) in the study area (Seider 2003), although this species has not been documented as a prey item in cormorant diet (with the exception of one fish in 2000). However, there is little information regarding the availability of other species, which appear to be far more important in the DCCO diet, on a local level. This work was an attempt to better quantify the role of cormorants as predators on a local scale and how their impact may vary from I year-to year; however, without better estimates of prey availability, understanding whether cormorants are exhibiting prey selectivity, as well as their ecological role, remains unclear. In 2003, alewife was the most abundant prey fish in Lake Michigan with an estimated lake-wide biomass of 42.876 kilotonnes (kt) (kt = 1000 metric tones) (Madenjian et al. 2004). Given that alewife populations have remained relatively stable in recent years (Madenjian et al. 2004), Beaver Archipelago breeding DCCOs consumed 119 approximately 1% of the available lake-wide alewife biomass in 2000, and approximately 1.6% in 2001, based on alewife consumption figures estimated by this work. In the last lake-wide waterbird survey in 1997, Beaver Archipelago cormorants comprised approximately 41% of the breeding DCCOs in Lake Michigan (Cuthbert et al. 1997). If this trend holds true presently, and all Lake Michigan breeding DCCOs have a similar diet, cormorants consumed an estimated 2.4% of the available lake-wide alewife biomass in 2000, and approximately 4% in 2001. None-the-less, it is possible that cormorants are fimctioning as important predators of alewife and other prey fish on a local scale within the Beaver Archipelago. However, it is also likely that predatory fish populations are consuming far more alewife and other prey biomass on an annual basis, both lake-wide and on a local basis (Madenjian and Gabrey 1995). Therefore, predatory fish may be far ' more important in influencing prey populations in Lake Michigan. In fact, long-term trends in alewife biomass suggest that salmonids were effective in reducing lake-wide alewife biomass from historical highs, and have also been effective in maintaining relatively low lake-wide alewife biomass over the last several decades (Madenjian et al. 2002). The Double-crested Cormorant is just one predator that, in synergism with other predators, as well as other ecological factors, contributes to overall Lake Michigan ecosystem fimctioning. 120 Literature Cited Bartell, S.M., J.E. Breck, R.H.Gardener, and A.L.Brenhert. 1986. Individual parameter perturbation and error analysis of fish bioenergetics models. Can. J. of Fish. Sci. 43:160-168. Birt, V.L., T.P. Birt, D. Goulet, D.K. Cairns, and WA. Montevecchi. 1987. Ashmole' 3 halo: direct evidence for prey depletion by a seabird. Marine Ecology Progress Series 40:205-208. Birt-Friesen, V.L, Montevecchi, W.A., Cairns, D.K., and SA. Mackes. 1989. Activity- . specific metabolic rates of free-living Northern Gannets and other seabirds. Ecology 70:357-367. Blomme, C. 1981. Status and breeding success of Double-crested Cormorants in two North Channel (Lake Huron) colonies in 1979. Ont. Field Biol. 35:70-78. Bryan, S.D., C.A. Soupir, W.G. Duffy, and CE. Freiburger. 1996. Caloric densities of three predatory fishes and their prey in Lake Oahe, South Dakota. J. Fresh. Ecol. 11(2):153-161. Cairns, D.K. 1992. Bridging the gap between ornithology and fisheries science: use of seabird data in stock assessment models. Condor 94:811-824. Cairns, D.K., G. Chapdelaine and W.A.Montevecchi. 1991. Prey exploitation by seabirds in the Gulf of St. Lawrence. In J .-C. Therriault (ed.), The Gulf of St. Lawrence: Small Ocean or Big Estuary? Can. Spec. Publ. Fish. Aquat. Sci. 113:277-291. Craven, SR. and E. Lev. 1987. Double-crested Cormorants in the Apostle Islands, Wisconsin, USA: population trends, foods habits and fishery depredations. Colonial Waterbirds 10(1):64-7l. Cummins, K.W. and .1 .C. Wuycheck. 1971. Caloric equivalents for investigations in ecological energetics. Mitt. Int. ereor. Agnew. Limnol. 18:1-51. Cuthbert, F.J., J. McKeaman and L. Wemmer. 1997. U.S. Great Lakes Tern and Cormorant Survey: 1997 Progress Report. Department of Fisheries and Wildlife, University of Minnesota, St. Paul. Draulans, D. 1988. Effects of fish-eating birds on freshwater fish stocks: an evaluation. Biol. Conserv. 44:25 1-263. Fowle, MR. 1997. Population dynamics, food habits, and bioenergetics of Double- crested Cormorants in Lake Champlain. MS. Thesis, Wildl. and Fish. Biol. Prog., School of Nat. Resourc, Univ. of Vermont. 108 pp. 121 Fumess, R.W. 197 8. Energy requirements of seabird communities: a bioenergetics model. J. Anim. Ecol. 47:39-53. Glahn, J .F. and KB. Brugger. 1995. The impact of Double-crested Cormorants on the Mississippi Delta catfish industry: a bioenergetics model. Colon. Waterbirds 18 (Spec. Publ. 1): 168-175. Hatch, J .J . and D.V.C. Weseloh. 1999. Double-crested Cormorant (Phalacrocorax auritus). Pgs 1-36 in The Birds of North America, No. 441 (A. Poole, and F. Gill, eds) The Birds of North America, Inc., Philidelphia, Pennsylvania. Hebert, CE. and HA. Morrison. 2003. Consumption of fish and other prey items by Lake Erie waterbirds. J. Gr. Lakes Res. 29(2):213-227. Kendeigh, S.C., VR Dol’nik and V.M. Gavrilov. 1977. Avian energetics. In J. Pinowski and SC. Kendeigh (eds), Granivorous Birds in Ecosystems. Cambridge University Press, London. pgs. 127-205. Lantry, B.F., TH. Eckert, and GP. Schneider. 1999. The relationship between the abundance of Smallmouth Bass and Double-crested Cormorants in the Eastern Basin of Lake Ontario. In: Final Report: T 0 Assess the Impact of Double-crested C ormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 12, Pp. 1-10. Lewis, H.G. 1929. The Natural History of the Double-crested Cormorant. Ottawa, Ontario: Ru-Mi-Loo Books. Ludwig, JP. 1984. Decline, resurgence and population dynamics of Michigan and Great Lakes Double- Crested cormorants. Jack Pine Warbler 62:91-102 Ludwig, J .P., C.N. Hull, M.E. Ludwig, and H]. Auman. 1989. Food habits and feeding ecology of nesting Double-crested Cormorants in the upper Great Lakes, 1986- 1989. The Jack-Pine Warbler 67(4): 115-126. Ludwig, J .P. and Summer, CL. 1997. Population status and diet of cormorants in Les Cheneaux Islands area. IN: History, Status, and Trends in Populations of Yellow Perch and Double-crested C ormorants in Les Cheneaux Islands, Michigan. Madenjian, C.P., T.J. Desorcie, and J .D. Holuszko. 2004. Status and trends of prey populations in Lake Michigan, 2003. Report, Great Lakes Fishery Commission, Lake Michigan Committee Meeting, Ypsilanti, Michigan. 122 Madenjian, C.P., G.L. Fahnenstiel, T.H. Johengen, T.F. Nalepa, HA. Vanderploeg, G.W. Fleisher, P.J. Schneeberger, D.M. Benjamin, E.B.Smith, J .R. Bence, E.S. Rutherford, D.S.Lavis, D.M. Robertson, D.J. Jude, and MP. Ebner. 2002. Dynamics of the Lake Michigan food web, 1970-2000. Can. J. Fish Aquat. Sci. 59:736-753. Madenjian, GP. and SW. Gabrey. 1995. Waterbird predation of fish in western Lake Erie: a bioenergenics model application. The Condor 97:141-153. Maruca, S.L. 1997. The impact of cormorant predation on yellow perch in Les Cheneaux Islands, Lake Huron. IN: History, Status, and Trends in Populations of Yellow Perch and Double-crested C ormorants in Les Cheneaux Islands, Michigan. J. S. Diana, G.Y. Belyea and RD. Clark, Jr., Eds. Michigan Department of Natural Resources, Fisheries Division, Special Report, Number 17, pgs: 47-70. Nagy, K.A. 1989. Field bioenergetics: accuracy of models and methods. Physio]. Zool. 62:237-252. Neuman, J., D.L.Pearl, P.J. Ewins, R. Black, D.V.C. Weseloh, M. Pike and K. Karwowski. 1997. Spatial and temporal variation in the diet of Double-crested Cormorants (Phalacrocorax auritus) breeding in the lower Great Lakes in the early 19905. Canadian Journal of Fisheries and Aquatic Sciences 54:1569-1584. Schiavone, A. 2001. Double-crested Cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2000 studies. NYSDEC Special Report: Section 1, pgs. 1-5. Schiavone, A. 2003. Double-crested cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2002 studies. NYSDEC Special Report: Section 1, pgs 1-4. Seefelt, NE. and J.C.Gillingham. 2004. The Double-crested Cormorant in Lake Michigan: A Review of Population Trends, Ecologyand Current Management. In State of Lake Michigan (M. Munawar, ed. Ecovision World Monograph Series. Seider, M.J. 2003. Population dynamics of Smallmouth Bass in the Beaver Archipelago, northern Lake Michigan, 1999 —- 2002. MS. Thesis, University of Georgia. Suter, W. 1995. The effect of predation by wintering cormorants Phalacrocorax carbon Grayling Ihymallus thymallus and trout (Salmonidae) populations: two cases studies from Swiss rivers. J. of Applied Ecology 32:29-46. Warke, G.M.A. and KR. Day. 1995. Changes in the abundance of cyprinid and percid prey affect the rate of predation by cormorants Phalacrocorax carbo carbo on salmon Salmo salar smolt in northern Ireland. Ardea 83:157-166. 123 Weseloh, D.V.C. and P.J. Ewins. 1994. Characteristics of a rapidly increasing colony of Double-crested Cormorants (Phalacrocorax auritus) in Lake Ontario: population size reproductive parameters and band recoveries. J. Great Lakes Research 203443-456. Wiens, J .A. 1984. Modeling the energy requirements of seabird populations. IN: Seabird Energetics. G.C. Whittow and H. Rahn, eds. Plenium Press, New York. pgsz255-285. Wiens, J .A. and J .M. Scott. 1975. Model estimation of energy flow in Oregon coastal seabird populations. Condor 77:439-452. Wires, L.R., F.J. Cuthbert, DR Trexel and AR. Joshi. 2001. Status of the Double- crested Cormorant (Phalacrocorax auritus) in North America. Final Report to U.S. Fish and Wildlife Service, Arlington, V.A. 124 Tablel. Life history characteristics of Double-crested Cormorants used to model prey consumption by breeding adult and young birds in the Beaver Archipelago, 2000 and 2001. Characteristic 2000 2001 Number of Nests (seasonal high) 8316 8096 Calories per eg a 75-4 75-4 Clutch Size — Early Season 2.60 2.89 Clutch Size -— Mid Season 1.50 2.38 Number of Incubation Days b 28 28 Hatch Rate (% of eggs laid) c 62 67- Fledge (days) b 50 50 Fledged Rate (% of chicks hatched) 29 49 Hatch-year Mortality (% year) d 42 42 Breeder Mortality (% year) c 20 20 Mass at Hatch (g) b 34 34 Adult Mass (kg) 2.0 2.0 aKendeigh et al. 1977 b Hatch and Weseloh 1999 cBlomme 1981 d Madenjian and Gabrey 1995 e Cairns et al. 1991 All other inputs were measured in the field by the authors 125 .eotom be: 088 .2: macaw :o: 3 Ba: meomtoaoa oz“: 35 860% 58:8 3E bemoan. met. I. No.0 8.0 86 E .o 3.0 So 125 o 0 5o 0 o and assuage assess scam =§oam 2 .o 2 .o 8.0 2 .o :.o :.o 2% ”sequence coo—02m o :3 ”no :3 5.0 Se Anzafiaisfimsfiv sumac—#35 05%..05 Z 8.0 So one 8.0 2:. o 2% s38 £93m 8.0 8.0 0 one v3 0 was. ”masseuse 3.36 OS :3 8.0 NS 3.0 3o esmfieeonsae. e83 0:302 winced“... wezcmoz 5:335 wcfioofi wczaoz 5:335 58— .3...— éom \wccmocbcm -mom \mcumocbi _o¢~ cecN .Som can ooom com weapon 0:5 $685-33 e5 :53: .eocunsoccwfieoofiba 05 Be $58.58 was?» one :33 mam—389 .3 evanmcoo 25: .38 3222?: we 922:03 05 co £59805 .N 033. 126 Table 3. Average caloric density (kcal/kg) for prey species. Prey Species Calorie Density (Real/kg) Alewife (A losa pseudoharengus) a 1.977 Crayfish (Orconectes spp.) a 1.077 Sculpin (Cottus spp.) a 1.493 Nine-spine Stickeback (Pungitius pungitius) a 1-493 Sucker (Catostomus spp.) b 0.884 Spot-tail Shiner (Notropis hudsonius) b 1-193 a From Cummins and Wuycheck (1971) b From Bryan et al. (1996) 127 Table 4. Total DFC (daily food consumption) in kilograms for adults during each period of the breeding season for 2000 and 2001. 2000 2001 Pre-nesting/Incubation Stage 0.55 0.55 Nestling Stage 0.55 0.48 Post-breeding Stage 0.65 0.51 128 Table 5. Total seasonal food consumption (kg) per pre-fledged and post-fledged chicks for 2000 and 2001. 2000 2001 Pre-fledged Chick 9.0 7.9 Post-fledged Chick 26.0 20.4 Season Totals 35.0 28.3 129 Table 6. Sensitivity analysis results. Model Input Input Perturbation Error +10% -10% Number of Nests +10.0 -10.0 Adult Mass +7.2 -7.4 Hatch Rate +2.5 -2.5 Fledge Rate +1.1 -1.1 Hatch-year Mortality -0.8 +0.8 Breeder Mortality -0.6 +0.6 Calories per Egg 0.0 0.0 Adult Assimilation Efficiency -9.1 +11.1 Young Assimilation Efficiency -9.1 +11.1 Energy Density of Prey (kcang) -0.1 +0.1 Mass at Hatching 0.0 0.0 Incubation Length (days) -0.5 +0.5 Fledging Length (days) -0.3 +0.3 130 I Squaw Island Ol [Garden Island I [ Hat Island Whiskey Island 0 3 I Hog Island | Pismire N A Gull Island Beaver Island 9 0 9 18 Kilometers E Figure 1. The Beaver Archipelago of northern Lake Michigan. 131 350 300 - 250 _ 200 — —2000 150 _ —2oo1 100 - 50 - Daily Food Intake (9) O Tlrllllllllllllllllrlllllllllllllllllllll1111111 <22 m e o e e o e @«gwheb~«¢wogflkege ”Exp“ Pre-fledge Chick Mass (9) 0.63 0.61 0.59 0.57 ~ 0.55 0.53 0.51 0.49 0.47 0,45.,.,....T...rir..ul.... fix‘be‘bmebflego‘brfi’wého‘b 99’ \% \‘b \% '\%'\ Kg Kg N9 \g N9 Mass of Post-Fledged Chick (kg) 1 I 1 —-2000 — 2001 l I 1 Daily Food Intake (kg) \ N9 Figure 2. a) Daily food consumption (g) for pre-fledged chicks in 2000 and 2001, and b) daily food consumption (kg) of post-fledged chicks up to the time they attain adult mass in 2000 and 2001. 132 700 ~ —- 600 - 500 - I 2000 400 i T {32001 Tonnes 300 ~ 200 _ 100 ~ O ‘ l I r ' T r 49 d” a: .g§’ <9 69 ‘&§ 99’ c§§ C§§> cormorant Age-1 . natural surv1val ii 14> cormorant Age-2 natural 11 E cormorant Age-3 natural 11 E cormorant Age-4 natural fl L> cormorant Adults natural L> \" angling Figure 2: Conceptual model depicting the smallmouth bass population. 186 2250 a 2000 2 1750 r 1500 ~ 1250 ~ 1000 r j i i 75a 1 500 r Estimated Population (Age 3 and older) 250 — t f T 9 0 1 r 1 r r r 1 1970 1975 1980 1985 1 990 1995 2000 2005 Figure 3. The population size of age three and older smallmouth bass in Garden Island Harbor as determined by the Schnabel method (with standard deviations). Estimates for 19703 and 19803 from Lennon (unpublished data) and later estimates from from Seider (2003) 187 20000 — 3 . c I 2! ° 15000 ~ - ' g —0— Main Archipelago 1; + Pismire & Grape ' c I '3 10000 — e I m g. o . 3 5000 a n I E 3 z . . o o 1 “a" t e a a 1983 1988 1 993 1 998 2003 Figure 4. The overall trends in the population size of Double-crested Cormorants nesting in the Main Archipelago and when just Pismire and Grape colonies are combined. Curves are based on actual nest counts in 1984 (Ludwig 1984), 1989 (Scharf and Shugart. 1993), 1997 (Cuthbert et al. 1997) and the current study (2000-2002). 188 500* 400* 300* 200* Number of Adult Bass 100“ 0 T I I I l 1 1 987 1 989 1 991 1 993 1995 1 997 1 999 1-6 i —o— Main Archipelago 1.4 ~ + Pismire & Grape Percent Biomass 0 f I I 1987 1989 1991 1993 1995 1997 1999 Figure 5. Results of the Baseline Rebuild model simulation showing a) the trends in the adult smallmouth bass population size and b) the percent biomass of bird diet composed of smallmouth bass for the entire Main Archipelago cormorant population and the Pismire and Grape colonies cormorant population alone. 189 800 - 700 - 600 ~ 500 n 400 ~ 300 n 200 ~ 100 a 0 1 fl 1 1 1 1 2001 2003 2005 2007 2009 201 1 2013 Number of Adult Bass Biomass Conssumed (kg) ..s _L N N 0) O OI O 01 O O O O 0 O 01 O l o l l l l 1 2001 2003 2005 2007 2009 201 1 Figure 6. Results of the Baseline Forecast model simulation showing a) the trends in the adult smallmouth bass population size and b) the biomass of smb (kg) consumed by the cormorant population. 190 500~ 400- 300* 200* Number of Adult Bass 100* O I H l I I I 1 987 1 989 1991 1 993 1995 1 997 1999 1-5 ‘ —0— Main Archipelago + Pismire & Grape Percent Blomass 0.5 ~ 0 I l l 1 987 1 989 1 991 1993 1 995 1 997 1 999 l Figure 7. Results of the modified Baseline Rebuild model with no age four predation simulation showing a) the trends in the adult smallmouth bass population size and b) the percent biomass of bird diet composed of smallmouth bass for the entire Main Archipelago cormorant population and the Pismire and Grape colonies cormorant population alone. 191 600 - a 500 I —o—All predation m - g 400 . + No Age Four 2 Predation '6 300 - . as a 200 - E ‘ . . , g . 100 ~ ' . , O 1 1 1 1 1 1 1987 1989 1991 1993 1995 1997 1999 Figure 8. Results of the modified Baseline Rebuild model with spawning constant set at 0.40 showing the simulated trends in the adult smallmouth bass population size when all immature age classes are subject to cormorant predation and when age four bass are invulnerable to cormorant predation. 192 -0— Main Archipelago + Pismire & Grape Percent Blomass 0) I l 1 I 1 987 1 989 1 991 1 993 1 995 1997 1 999 -0— Main Archipelago + Pismire 8. Grape Percent Blomass OJ l 1 1 987 1 989 1 991 1 993 1 995 1 997 1 999 l l Figure 9. Results of the modified Baseline Rebuild model with spawning constant set at 0.40 showing the simulated trends percent biomass consumed by Main Archipelago birds, and Pismire and Grape colonies birds when a) all immature age classes are vulnerable to predation and b) when age four bass are invulnerable to predation. 193 600 ~ 500 - 400 ~ 300 r 200 ~ Number of Adult Bass 100 ~ 0 I J l l l —1 1 987 1 989 1 991 1993 1995 1 997 1 999 300* 200— Nunber of Adult Bass 100* 0 1 1 F l T 1 1987 1 989 1 991 1 993 1 995 1997 1 999 Figure 10. Simulated adult population size (means and standard deviations) for a) Rebuild Model Scenario 1 and b) Rebuild Model Scenario 2. 194 -°- Scenario 1 —0- Scenario 2 Percent Blomass O O) 0 1 1 1 1 v 1 1 1 987 1 989 1991 1 993 1995 1997 1 999 -°- Scenario 1 -0— Scenario 2 Percent Blomass a 0.5 ~ 0 l l l 1 1 987 1 989 1 991 1 993 1995 1 997 1 999 Figure] 1. Experimental (means and standard deviations) percent biomass of bird diet composed of smallmouth bass as simulated in Rebuild Scenarios 1 and 2 for a) all Main Archipelago birds, and b) only birds nesting on Pismire and Grape Islands. 195 600 n 500 - 400 n 300 1 200 a Number of Adult Bass 100 — 0 1 987 I 1 989 l 1991 I 1 993 I 1 995 1M 1997 1 999 7001 600~ 5001 400* 300- 200~ Number of Adult Bass 100‘ 0 1 987 T 1 989 I 1991 I 1993 I 1 995 I 1 997 I 1 999 Figure 12. Simulated adult population size (means and standard deviations) for a) Rebuild Model Scenario 3 and b) Rebuild Model Scenario 4. 196 -°- Scenario 3 -0- Scenario 4 Percent Blomass _l .3 or .0 at I .K_ *— 0 1 1 1 r ' 1 987 1 989 1991 1 993 1995 1997 1 999 6 - -0— Scenario 3 + Scenario 4 5- Percent Blomass A 0 1 1 r rR——— 1 987 1 989 1 991 1 993 1 995 1 997 1 999 Figure 13. Experimental (means and standard deviations) percent biomass of bird diet composed of smallmouth bass as simulated in Rebuild Scenarios 3 and 4 for a) all Main Archipelago birds, and b) only birds nesting on Pismire and Grape Islands. 197 450 — 400 ~ 350 n 300 n 250 n 200 ~ 150 — 100 e 50 ~ —o— No Angling +10% +20% —x— 30% - - at - -40% Number of Adult Bass 0 I I I 1 987 1 989 1 991 1 993 1 995 1 997 1 999 0.16 - 0.14 - 0.12 - 0.1 - 0.08 « 0.06 - 0.04 - 0.02 1 Cormorant Predation Pressure 0 10 20 20 40 Percent Angling Harvest Figure 14. 3) Simulated adult bass population sizes when adult populations are subject to- varying levels of angling harvest. b) The cormorant predation pressure necessary to work in synergism with angling harvest to drive the adult population to observed levels. 198 3000 '1 2500 J 2000 - 1500 1 1000 - Number of Bass Harvested 500 - 1 0 20 30 40 Percent Angling 2800 - 2100 J 1400 - Blomass Consumed (kg) 700 - 10 20 3O 40 Percent Angling Figure 15. a) Simulated number of adult bass removed by anglers at varying harvest levels beginning in 1988 and ending in 1999. b) The biomass (kg) of immature bass needed to be removed by cormorants over the same simulation period when angling pressure varies. 199 1400 — 120° * —o—YOY Recruitment Q . 3 1000 - ---e--NatLralMortallty g —a—Adult Mortality 3 800 . < h 3 600 . I .n S 400 ~ 2 .4 ----- t 1 " “"0 . '0 ---- . ..... . '''' e. 1 200 " “a...” i u. ....... o . i 1 t a 1987 1989 1991 1993 1995 1997 1999 Figure 16. Simulated impacts that increased YOY mortality (0.99), increased annual natural mortality for age 1 through adult fish (0.53 5), and increase adult mortality (0.885), in the absence of cormorant predation. Each variable was tested separately while all other variables remained unchanged. 200 800‘ 700‘ W 3 600 - or g 500— 'u +Odginal 5 40° ‘ —-—750°/ 0 4 . o g 300 - ‘/' \ +10% __ o 5200 1 ‘0-\/><\ x 15” z 2'. X\x\x +20% 100- e ‘ _ \x-x ..-.---25% 0.“... 0 I I I I I I T I I I I QWQ‘thQo’Qbé\Q%QQNQNN (1' '19 '19 '19 '19 '19 '19 '19 '19 '19 '19 '19 b. 700— .0 3600‘ ‘- 5 u 500 a g +15% 3 400 - 3 +7.50% 0 300 . ' +10% ” -‘-—- r 3 ‘W‘ +15% 5 200 7 o ‘. +20% .- ‘. m 100 - e ------ ----25% ..__ 0 N. ---- 9 ..... . 61' 65686" 696‘ 656919 N“ '19 '19 '19 ’19 '19 ’19 '19 '19 '19 ‘19 Figure 17. The simulated results for the Baseline Forecast Model with spawning constant. set a 0.10 showing a) the resulting adult population size when cormorant predation is increased and b) the biomass (kg) of bass consumed by cormorants in each simulation. 20] 16001 1400- a 3 1200 a m 3 1000 ~ 5% :5 800+ —-—7.50% o 0 ‘6 600- +10” a +15% 2 =§.§ + 0 200 — “\. ------ 25% . ........ t ...... 0 I I I l I I I I I I. I e‘ko'boVoPoPé‘oq’o‘va-(‘O as as as as as as as as as as as b. 14001 a 1200 1 5 1g 1000 - 3 800- '--. —-—5% 5 ~. —I—7.50% o 600‘ ~. // | 10% a ‘ \ (I g 400 _ ‘ I. +15%) .5 ”W.” +20‘y0 m 200‘ ‘e -------25% “'0 ..... . _____ . 0 1 1 1 1 1 1 1 1 1 l 61' 6'5 6‘" 6° 6‘9 o" 6‘" o°-’ 139 a“ as as as as as as s as as as Figure 18. The simulated results for the Baseline Forecast Model with spawning constant set a 0.20 showing a) the resulting adult population size when cormorant predation is increased and b) the biomass (kg) of bass consumed by cormorants in each simulation. 202 Number of Adult Bass —t N (a) .5 01 O) \l o 8 8 8 8 8 8 8 I I 1 I l 2001 2003 2005 2007 2009 201 1 700 - 600 ~ 500 r 400 s 300 r 200 ~ Biomass Consumed (kg) 100 - O I 1 I I 2001 2003 2005 2007 2009 201 1 Figure 19. Forecast Constant Predation Pressure Scenario 1 output showing a) the resulting adult bass population for experimental (means and standard deviations) simulations and b) the biomass (kg) of immature bass consumed by cormorants in experimental simulations (means and standard deviations). 203 450 n 400 - 350 — 300 ~ 250 - 200 4 150 — 100 e 50 4 0 I I I I I 2001 2003 2005 2007 2009 201 1 Number of Adult Bass ii 88 Blomass Consumed (kg) e88 I r I I j 2001 2003 2005 2007 2009 201 1 Figure 20. Forecast Constant Predation Pressure Scenario 2 output showing a) the resulting adult bass population for experimental (means and standard deviations) simulations and b) the biomass (kg) of immature bass consumed by cormorants in experimental simulations (means and standard deviations). 204 600 1 500 - g —O—Constant : 400 4 --O"0.05-0.1O % —a—0.10-0.20 5 300 ~ nee-0.15.025 2 +0.20-0.30 g 200 - --e--0.25-0.35 g —i——0.30-O.40 100 ~ 0 I I I I I fl 2001 2003 2005 2007 2009 2011 2013 b. 3500 ~ 3000 n '3" g 2500 ~ g 2000 - C 8 3 1500 1 g on. B 500 —1 o I I I I 75 S 8 «‘4? 8 «‘8 S 3? o 9' o‘ o’ 9' o' 5 £55 .9 3.5 S «‘9 8 0 o o o’ o’ d o’ Figure 21. Simulation results for the modified Forecast Constant Predation Pressure Scenario 1 (Predator Response) exploring the influence of increasing cormorant pressure as bass numbers recover as compared to constant predation pressure showing a) adult population size and b) the biomass (kg) of immature bass consumed over the entire simulation period. 205 600 n 500 n 400 ~ 300 - 200 ~ Number of Adult Bass 100 n 0 I I7 I I I 1 2001 2003 2005 2007 2009 201 1 2013 0 I Hi I I U 2001 2003 2005 2007 2009 201 1 Figure 22. Experimental (means and standard deviations) output for Forecast Predator Response Scenario 1 showing a) simulated adult bass population and b) the biomass (kg) of immature bass consumed 206 600 n 500 ~ 400 4 300 ~ 200 — Number of Adult Bass 100 n 0 fl I I I I fl 2001 2003 2005 2007 2009 201 1 2013 1000 - 900 ~ 800 n 700 - 600 — 500 - 400 ~ 300 1 200 ~ 100 1 O I I I I 7 2001 2003 2005 2007 2009 201 1 Blomass Consumed (kg) Figure 23. Experimental (means and standard deviations) output for Forecast Predator Response Scenario 1 showing a) simulated adult bass population and b) the biomass (kg) of immature bass consumed. 207 LITERATURE CITED Adams, C.M., Schneider, GP. and Johnson, J .H. 1999. Predicting the size and age of smallmouth bass (Micropterus dolomieu) consumed by Double-crested Cormorants (Phalacrocorax auritus) in eastern Lake Ontario, 1993-1994. In: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division. Ainley, D.G. and RI Boekelheide, eds. 1990. Seabirds of the Farallon Islands. Stanford Univ. Press, Stanford, California. Baillie, J.L. 1947. The Double-crested Cormorant nesting in Ontario. The Canadian Field Naturalist 61, 119-126 Barrows, W.B. 1912. Michigan Bird Life. Michigan Agricultural College Special Bulletin. Bartel], S.M., J .E. Breck, RH. Gardener, and AL. Brenhert. 1986. Individual parameter perturbation and error analysis of fish bioenergetics models. Can. J. of Fish. Sci. 43:160-168. Beamesderfer, RC. and J .A. North. 1995. Growth, natural mortality and predicted response to fishing for largemouth bass and smallmouth bass populations in North America. North American Journal of Fisheries Management 15 :688-704 Birt, V.L., T.P. Birt, D. Goulet, D.K. Cairns, and WA. Montevecchi. 1987. Ashmole’ 3 halo: direct evidence for prey depletion by a seabird. Marine Ecologr Progress Series 40:205-208. Birt-Friesen, V.L, Montevecchi, W.A., Cairns, D.K., and SA. Mackes. 1989. Activity- specific metabolic rates of free-living Northern Gannets and other seabirds. Ecology 70:3 57-367. Blackwell, B.F., W.B. Krohn, and RB. Allen. 1995. Foods of nestling Double-crested Cormorants in Penobscot Bay, Maine, USA: temporal and spatial comparisons. Colonial Waterbirds 18(Special Publication 1): 199-208. Blackwell, B.F., W.B. Krohn, NR, Dube, and A.J. Godin. 1997. Spring prey use by Double-crested Cormorants on the Penobscot River, Maine, USA. Colonial Waterbirds 20(1):77-86. Blackwell, BF. and J .A. Sinclair. 1995. Evidence for secondary consumption of fish by Double-crested Cormorants. Marine Ecology. Progess Series 12321-4. 208 Blomme, C. 1981. Status and breeding success of Double-crested Cormorants in two North Channel (Lake Huron) colonies in 1979. Ont. Field Biol. 35:70-78. Brown, KM. and P.J. Ewins. 1996. Technique-dependent biases in determination of diet composition: an example with Ring-billed Gulls. Condor 98:34-41. Brugger, KB. 1993. Digestibility of three fish species by Double-crested Cormorants. Condor 95:25-32. Bryan, S.D., C.A. Soupir, W.G. Duffy, and CE. Freiburger. 1996. Caloric densities of three predatory fishes and their prey in Lake Oahe, South Dakota. J. Fresh. Ecol. 11(2):]53-161. Cairns, D.K. 1992. Bridging the gap between ornithology and fisheries science: use of ‘ seabird data in stock assessment models. Condor 94:811-824. Cairns, D.K., R.L. Dibblee and P.-Y.Daoust. 1998. Displacement of a large Double- crested Cormorants, Phalacrocorax auritus, colony following human disturbance. Canadian Field Naturalist 1 12:520-522. Carss, D.N. and the Diet Assessment and Food Intake Working Group. 1997. Techniques for assessing cormorant diet and food intake: towards a consensus view. Supplemento alle Ricerche di Biologia della Selvaggina 26:197-230. Chen, C., R.Ji, D]. Schwab, D. Beletsky, G.L. Fahnenstiel, M.Jiang, T.H. Johengen, H. Vnaderploeg, B. Eadie, J .W. Budd, M.H. Bundy, W. Gardner, J .Cortner, and P.J. Lavrentyev. 2002. A model study of coupled biological and physical dynamics in Lake Michigan. Ecological Modeling 1522145-168. Christie, W.J., Scott, K.A., Sly, PG, and Strus, RH. 1987. Recent changes in the aquatic food web of eastern Lake Ontario. Canadian Journal of Fisheries and Aquatic Sciences 44 (Suppl.2), 37-52. Craven, SR. and E. Lev. 1987. Double-crested Cormorants in the Apostle Islands, Wisconsin, USA: population trends, foods habits and fishery depredations. Colonial Waterbirds 10(1):64-71. Croxall, J .P. 1987. Conclusions. Pgs. 369-381 in Seabirds: Feeding Ecology and Role in Marine Ecosystems (J .P. Croxall, ed.) Cambridge University Press, Cambridge, Massachusetts, USA. Cummins, K.W. and J .C. Wuycheck. 1971. Caloric equivalents for investigations in ecological energetics. Mitt. Int. Theor. Agnew. Limnol. 18:1-51. 209 Custer, T.W. and C. Bunck. 1992. Feeding flights of breeding Double-crested Cormorants at two Wisconsin colonies. Journal of Field Ornithology 63 (2):203- 21 1. Cuthbert, F.J., J. McKeaman, and L. Wemmer. 1997. U.S. Great Lakes Tern and Cormorant Survey: 1997 Progress Report. Department of Fisheries and Wildlife, University of Minnesota, St. Paul. da Silva, J. and JD. Neilson. 1985. Limitations of using otoliths recovered in scats to estimate prey consumption in seals. Canadian Journal of Fish and Aquatic Sciences 42: 1439-1442. DEIS, 2001. Draft Environmental Impact Statement: Double-crested Cormorant Management. U.S. Department of Interior Fish and Wildlife Service and U.S. Department of Agriculture APHIS Wildlife Service. Derby, CE. and JR. Lowom. 1997. Comparison of pellets versus collected birds for sampling diets of Double-crested Cormorants. Condor 99:549-553. Draulans, D. 1988. Effects of fish-eating birds on freshwater fish stocks: an evaluation. ‘ Biol. Conserv. 44:251-263. Dufi‘y, DC. and L.J.E. Laurenson. 1983. Pellets of Cape Cormorants as indicators of diet. Condor 85:305-307. Dunn, EH. 1975. Caloric intake of nestling Double-crested Cormorants. Auk 92:553- 565. Dunston, TC. 1972. A harness technique for radio-tagging raptorial birds. Inland Bird Banding News 4424-8. Eck, G.W. and EH. Brown. 1987. Recent changes in Lake Michigan’s fish community and their probable causes, with emphasis on the role of the alewife (Alosa pseudoharengus). Can. J. Fish. Aquat. Sci. 44:53-60. Ellison, L.N. and L. Cleary. 1978. Effect of human disturbance on breeding Double- crested Cormorants. Auk 95:510-517. Environmental Protection Agency (EPA). 2000. Lake Michigan Lake Management Plan. Ewins, P.J., D.V.C. Weseloh, and H. Blokpoel. 1995. Within-season variation in nest numbers of Double-crested Cormorants (Phalacrocorax auritus) on the Great Lakes: implications for censusing. Colonial Waterbirds 18:179-182. Farquhar, J .F., R.D. McCullough, and I.M Mazzocchi. 2003. Cormorant management activities in Lake Ontario’s eastern basin. NYSDEC Special Report. Section 2, pgs 1-5. 210 FEIS. 2003. Final Environmental Impact Statement: Double-crested Cormorant Management. U.S. Department of Interior Fish and Wildlife Service and U.S. Department of Agriculture APHIS Wildlife Service. Fowle, MR. 1997. Population dynamics, food habits, and bioenergetics of Double- crested Cormorants in Lake Champlain. MS. Thesis, Wildl. and Fish. Biol. Prog., School of Nat. Resourc., Univ. of Vermont. Fry, F.E.J. and K.E.F. Watt. 1957. Yields of year classes of the smallmouth bass hatched in the decade of 1940 in Manitoulin Island waters. Trans. Am. Fish. Soc. . 85:135-143. Fuller, K., H. Shear, and J. Wittig (eds). 1995. The Great Lakes Atlas. Enviornment Canada and the U.S. Environmental Protection Agency. Fumess, R.W. 1978. Energy requirements of seabird communities: a bioenergetics model. J. Anim. Ecol. 47239-53. Glahn, J .F . and KB. Brugger. 1995. The impact of Double-crested Cormorants on the Mississippi Delta catfish industry: a bioenergetics model. Colon. Waterbirds 18 (Spec. Publ. 1): 168-175. Gotmark, F. 1992. The effects of investigator disturbance on nesting birds. Current Ornithology 9263-104. Gremillet, D.G. 1997. Catch-per-unit effort, foraging efficiency, and parental investment in breeding Great Cormorants (Phalacrocorax carbo carbo). ICES J. of Marine Science 54:63 5-644. Gremillet, D., G. Argentin, B. Schulte and BM. Culik. 1998. Flexible foraging techniques in breeding cormorants Phalacrocorax carbo and shags Phalacrocorax aristotelis: benthic or pelagic feeding? Ibis 140:113-119. Gremillet, D., R.P. Wilson, S. Storch, and Y. Gary. 1999. Three-dimensional space utilization by a marine predator. Marine Ecol. Prog. Ser. 183:263-273. Hatch, J .J . 1995. Changing populations of Double-crested Cormorants. Colonial Waterbirds 18 (Special Publication 1), 8-24. . Hatch, J .J . and D.V.C. Weseloh, D.V.C. 1999. Double-crested Cormorant (Phalacrocorax auritus). Pgs 1-36 in The Birds of North America, No. 441 (A. Poole, and F. Gill, eds) The Birds of North America, Inc., Philidelphia, Pennsylvania. Hebert , CE. and HA. Morrison. 2003. Consumption of fish and other prey items by Lake Erie waterbirds. Journal of Great Lakes Research 29:213-227. 211 Hebert, CE. and W.G. Sprules. 2002. The relevance of seabird ecology to Great Lakes . Management. Journal of Great Lakes Research 28:91-103. Hooge, P.N and B. Eichenlaub. 1997. Animal movement extention to ArcView. version 1.1. Alaska Biological Science Center, Geological Survey, Anchorage, AK, USA. Hooker, C. 1999a. Carrington Challenges Cormorantsll Beaver Beacon (March) p. 11. Beacon Publishing, Inc. St. James, Michigan. Hooker, C. 1999b. Cormorants Beware! Beaver Beacon (March) pp. 1-6. Beacon Publishing, Inc. St. James, Michigan. Hooker, C. 1999c. Protected Invaders! Cormorants overun Great Lakes, Feds Ponder What to Do. Beaver Beacon (October) pp. 14—15. Beacon Publishing, Inc. St. James, Michigan. Jobling, M. and A. Breiby. 1986. The use and abuse of fish otoliths in the studies of feeding habits of marine piscivores. Sarsia 71:265-274. Johnson, J .H. and RM. Ross. 1996. Pellets versus feces: their relative importance in describing the food habits of Double-crested Cormorants. Journal of Great Lakes Research 222795-798. Johnson, J .H., RM. Ross and CM. Adams. 1999. Diet composition of Double-crested Cormorants in Eastern Lake Ontario, 1998. In: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes of the Eastern Basin of Lake Ontario. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division. Johnson, J .H., RM. Ross, RD. McCullough, and B. Edmonds. 2001a. Diet composition and fish consumption of Double-crested Cormorants fi'om the Pigeon and Snake ‘ Island colonies of eastern Lake Ontario in 2000. NYSDEC Special Report: Section 5, pgs. 1-9. Johnson, J .H., RM. Ross, RD. McCullough, and B. Edmonds. 2001b. Diet composition and fish consumption of Double-crested Cormorants from the Little Galloo Island colony of eastern Lake Ontario in 2000. NYSDEC Special Report: Section 3, pgs. 1-7. Johnson, J .H., RM. Ross, RD. McCullough, and B. Edmonds. 2003. Diet composition and fish consumption of Double-crested Cormorants from the Little Galloo Island colony of eastern Lake Ontario in 2002. NYSDEC Special Report: Section 3, pgs. 1-8. 212 Johnstone, I.G., M.P. Harris, S. Waneless and J.A. Graves. 1990. The usefirlness of pellets for assessing the diet of adult Shags Phalacrocorax aristotelis. Bird Study 3725-11. Kendeigh, S.C., VR Dol’nik and V.M. Gavrilov. 1977. Avian energetics. In J. Pinowski and SC. Kendeigh (eds), Granivorous Birds in Ecosystems. Cambridge University Press, London. pgs. 127-205. Kerfoot, W.C.. 1987. Cascading effects and indirect pathways. In: Kerfoot, WC. and A. Sih, (eds) Predation: Direct and Indirect Impacts on Aquatic Communities. University of New England Press, Hanover, New Hampshire, pgs. 57-70. Kieffer, J.D., MR. Kubacki, F.S. Phelan, D.P. Philipp, and BL. Tufis. 1995. Effects of catch-and-release angling on nesting male smallmouth bass. Trans. Am. Fish. Soc. 124270-76. Kirsch, EM. 1995. Double-crested Cormorants along the upper Mississippi River. Colonial Waterbirds 18(Special Issue 1), 131-136. Kroeff, T. 1996. Assessment of spawning populations of smallmouth bass at Little Sturgeon and the Mink River/Rowley Bass area and young-of-the-year surveys along the Door County shoreline in 1996. Department of Natural Resources, Sturgeon Bay, Wisconsin. pgs 1-15. Kroeff, T. 1997. Assessment of the spawning population of smallmouth bass in Detroit Harbor, Washington Island and young-of-the—year surveys along the Door County shoreline in 1997. Department of Natural Resources, Sturgeon Bay, Wisconsin. pgs 1-13. Kury, CR. and M. Gochfeld. 1975. Human interference and gull predation in cormorant colonies. Biological Conservation 8223-34. Lantry, B.F., T.H. Eckert, and GP. Schneider. 1999. The relationship between the abundance of Smallmouth Bass and Double-crested Cormorants in the Eastern Basin of Lake Ontario. In: Final Report: To Assess the Impact of Double-crested C ormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 12, Pp.‘ l - 1 0. Latta, WC. 1963. The life history of smallmouth bass, Micropterus d dolomieui, at Waugoshance Point, Lake Michigan. Bulletin of the Institute for Fisheries Research No. 5. Michigan Department of Conservation and the University of Michigan. Lewis, H. G. 1929. The Natural History of the Double-crested Cormorant. Ottawa, Ontario: Ru-Mi-Loo Books. 213 Ludwig, J. P. 1984. Decline, resurgence and population dynamics of Michigan and Great Lakes double-crested cormorants. Jack Pine Warbler 62:91-102 Ludwig, J.P., H.J. Auman, D.V.C. Weseloh, G.A. Fox, J.P. Giesy, and ME. Ludwig. 1995. Evaluation of the effects of toxic chemicals in Great Lakes cormorants: has causality been established? Colonial Waterbirds 18:60-69. Ludwig, J .P., C.N. Hull. ME. Ludwig and HI Auman. 1989. Food habits and feeding ecology of nesting Double-crested Cormorants in the upper Great Lakes, 1986- 1989. The Jack-Pine Warbler 67(4):]15-126. Ludwig, J .P. and CL. Summer. 1997. Population status and diet of cormorants in Les Cheneaux Islands area. IN: History, Status, and Trends in Populations of Yellow ' Perch and Double-crested C ormorants in Les Cheneaux Islands, Michigan. Madenjian, C.P., T.J. Desorcie, and JD. Holuszko. 2004. Status and trends of prey populations in Lake Michigan, 2003. Report, Great Lakes Fishery Commission, Lake Michigan Committee Meeting, Ypsilanti, Michigan. Madenjian, C.P., G.L. Fahnenstiel, T.H. Johengen, T.F. Nalepa, H.A. Vanderploeg, G.W. Fleisher, P.J. Schneeberger, D.M. Benjamin, EB. Smith, JR. Bence, E.S. Rutherfors, D.S. Lavis, D.M. Robertson, D.J. Jude, and MP. Ebner. 2002. Dynamics of the Lake Michigan food web, 1970-2000. Can. J. Fish. Aquat. Sci. 592736-753. Madenjian, GP. and SW. Gabrey. 1995. Waterbird predation of fish in western Lake Erie: a bioenergenics model application. The Condor 97:141-153. Maruca, S.L. 1997. The impact of cormorant predation on yellow perch in Les Cheneaux Islands, Lake Huron. IN: History, Status, and Trends in Populations of Yellow Perch and Double-crested Cormorants in Les Cheneaux Islands, Michigan. J. S. Diana, G.Y. Belyea and RD. Clark, Jr., Eds. Michigan Department of Natural Resources, Fisheries Division, Special Report, Number 17 pgs: 47-70. Mott, D.F., Glahn, J .M., Smith, P.L., Reinhold, D.S., Bruce, KI and Sloan, CA. 1998. An evaluation of winter roost harassment for dispersing Double-crested Cormorants away from catfish production areas in Mississippi. Wildlife Soc. Bulletin 26(3), 584-591. Nagy, K.A. 1989. Field bioenergetics: accuracy of models and methods. Physiol. Zool. 62:237-252. Nams, V0. 2001. Locate II: A program to triangulate radiotelemetry bearings and calculate error ellipses. Department of Environmental Science, NSAC, Truro, Nova Scotia, Canada. 214 National Oceanic and Atmospheric Administration (NOAA). 2003. Water levels of the Great Lakes. NOAA Great Lakes Environmental Research Laboratory, Ann Arbor, Michigan. National Oceanic and Atmospheric Administration (NOAA). 1989. United States — Great Lakes — Lake Michigan — Michigan: Waugoshance Point to Seul Choix Point Nautical Sounding Chart, 17‘” ed. US Department of Commerce, Washington DC, USA. Neuman, J., D.L. Pearl, P.J. Ewins, R. Black, D.V. Weseloh, M. Pike, and K. Karwowski. 1997. Spatial and temporal variation in the diet of Double-crested Cormorants (Phalacrocorax auritus) breeding in the lower Great Lakes in the early 19903. Canadian Journal of Fisheries and Aquatic Sciences 54: 1569-1584 North, MR 1986. Piping Plover (Charadrius melodus) nest success on Mallard Island, North Dakota, U.S.A., and implications for water level management. Prairie Naturalist 18(2): 1 17-122. Philipp, D.P., C. A. Toline, MF. Kubacki, D.B.F. Philipp, and F.J.S. Phelan. 1997. The ‘ impact of catch-and-release angling on the reproductive success of smallmouth bass and largemouth bass. N. Am. J. Fish. Management 172557-567. Pyrovesti, M. 1997. Integrated management to create new breeding habitat for Dalmatian Pelicans (Pelecanus crispus) in Greece. Environmental Management 21(5):657-667. Raffetto, N.S., J. R. Baylis and S.L. Sems. 1990. Complete estimates of reproductive success in a closed population of smallmouth bass (Micropterus dolomieui). Ecologi 7(4): 1523-1535. Ricker, W.B. 1975. Computation and interpretation of biological statistics of fish populations. Fisheries Research Board of Canada Bulletin 191. Ridgway, MS, and B]. Shuter. 1997. Predicting the effects of angling for nesting male _ smallmouth bass on production of age-0 fish with an individual model. N. Am. J. of Fish. Management 17: 568-5 80. Robinson, J. 1995. Lake Michigan Smallmouths. Field and Stream (April) 16-22. Root, TL. 1988. Atlas of Wintering North American Birds: An Analysis of Christmas Bird Count Data. University of Chicago Press, Chicago. Ross, RM. and J .H. Johnson. 1995. Seasonal and annual changes on the diet of Double-crested Cormorants: implications for Lake Ontario’s fishery. Great Lakes Resource Review 2: 1-9. 215 Ross, RM. and J.H. Johnson. 1999. Effect of altered salmonid stocking methods on cormorant predation in eastern Lake Ontario. . In: Final Report: To Assess the Impact of Double-crested C ormorants Predation on the Smallmouth Bass and Other Fishers of the Eastern Basin of Lake Ontario. NYSDEC Special Report, February, 1999. Section 11, pp. 1-10. Ross, R.M., J .H. Johnson, RD. McCullough and B. Edmonds. 2003. Diet composition and fish consumption of Double-crested Cormorants from the Pigeon and Snake Island colonies of eastern Lake Ontario in 2002. NYSDEC Special Report: Section 5, pgs. 1-10. Rudstam, L.G., A.J. VanDeValk, C.M. Adams, J .T.H. Coleman, J .L. Fomey and M. E. Richmond. 2004. Cormorant predation and the population dynamics of walleye and yellow perch in Oneida Lake. Ecological Applications 14(2): 149-163. Russel, A.F., S. Wanless, and MP. Harris. 1995. Factors affecting the production of pellets by Shags Phalacrocorax aristotelis. Seabird 17:44-49. Scharf, WC. 1978. Colonial birds nesting on man-made and natural sites in the U.S. Great Lakes. Technical Report D-78-10, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi. Scharf, WC. and G.W. Shugart. 1998. Distribution and abundance of gull, tern, and cormorant nesting colonies of the Great Lakes, 1989 and 1990. Publication No. 1, Gale Gleason Environmental Institute, Lake Superior State University, Sault Ste.‘ Marie, Michigan. Schiavone, A. 2001. Double-crested Cormorant predation on Smallmouth Bass and Other fishes of the eastern basin of Lake Ontario: summary of 2000 studies. NYSDEC Special Report: Section 1, pgs. 1-5. Schiavone, A. 2003. Double-crested cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: summary of 2002 studies. NYSDEC Special Report: Section 1, pgs 1-4. Schneider, OR, A. Schiavone, TH. Eckert, RD. McCullough, BF. Lantry, D.W. Einhouse, JR. Chrisman, and CM. Adams. 1999. Double-crested Cormorant predation on Smallmouth Bass and other fishes of the eastern basin of Lake Ontario: Summary and Overview IN: Final Report: To Assess the Impact of Double-crested Cormorant Predation on the Smallmouth Bass and Other Fishes . of the Eastern Basin of Lake Ontario. Joint Report of NYSDEEC Bureau of Fishes and the US Geological Survey Biology Resource Division. Section 1, pgs:1-6. 216 Schneider, J .C., P.W. Laarman, and H. Gowing. 2000. Chapter 17: Length-weight relationships. In Manual of Fisheries Survey Methods 11: With Periodic Updates (J .C. Schneider, ed). Michigan Department of Natural Resources, Fisheries Special Report 25, Ann Arbor. pgs 1-16. Schreer, J.F., S.J. Cooke, and RS. McKinely. 2001. Cardiac response to variable forced exercise at different temperatures: an angling simulation for smallmouth bass. Trans. Amer. Fish. Soc. 130:783-795. Seefelt, NE. and J .C. Gillingham. 2004a. The Double-crested Cormorant in Lake Michigan: A Review of Population Trends, Ecologyand Current Management. In State of LakeMichigan.: Ecology, Health and Management, T. Edsall and M. Munawar, eds. Ecovision World Monograph Series. Seefelt, NE. and J .C. Gillingham. 2004b. A new colony location for Double-crested Cormorants (Phalacrocorax auritus) and other waterbirds in the Beaver Archipelago, northern LakeMichigan. Michigan Birds and Natural History 1(3)]22-127. Seider, M.J. 2003. Population dynamics of smallmouth bass in the Beaver Archipelago, northern Lake Michigan, 1999 — 2002. MS. Thesis, University of Georgia, Athens Georgia, U.S.A. Skagen, S.K., C. P. Melchur and E. Muths. 2001. The interplay of habitat change, human disturbance and species interaction in a waterbird colony. American Midland Naturalist 145(1): 18-28. Sokal, RR and F .J . Rohlf. 1995. Biometry, 3rd Edition. W.H. Freeman and Company, New York, 887 pp. Stickley, AR. Jr., G.L. Warrick and J .F. Glahn. 1992. Impact of Double-crested Cormorant depredations on Channel Catfish farms. J. World Aquaculture Society. 23(1): 192-204. Suter, W. 1995. The effect of predation by wintering cormorants Phalacrocorax carbon Grayling Thymallus thymallus and trout (Salmonidae) populations: two case studies from Swiss rivers. J. of Applied Ecology 32:29—46. Toneys, ML. and D.W. Coble. 1979. Size-related, first winter mortality of freshwater fishes. Trans. Am. Fish. Soc. 108:415-419. Trauttmansdorff, J. and G. Wassermann. 1995. Number of pellets produced by immature cormorants Phalacrocorax carbo sinensis. Ardea 83 2 133-134. 217 Vanni, M.J. 1987. Indirect effect of predators on age-structured prey populations: planktivorous fish and zooplankton. In: Kerfoot, WC. and Sih, A. (eds) Predation: Direct and Indirect Impacts on Aquatic Communities. University of New England Press, Hanover, New Hampshire, pgs. 149-160. Verbeek, N.A.M. 1982. Egg predation by Northwestern Crows: its association with human and Bald Eagle activity. Auk 99:347-352. Warke, G.M.A. and KR. Day. 1995. Changes in the abundance of cyprinid and percid prey affect the rate of predation by cormorants Phalacrocorax carbo carbo on salmon Salmo salar smolt in northern Ireland. Ardea 83:157-166. Wanless, S., M.P. Harris, and AF. Russel. 1993. Factors influencing food-load sizes brought in by Shags Phalacrocorax aristotelis during chick rearing. Ibis 135219- 24. Watt, K.E.F. 1959. Studies on population productivity H. Factors governing productivity in a population of smallmouth bass. Ecological Monographs 29(4):367-392. Weseloh, D.V.C. and J. Casselman. 1992. Calculated fish consumption of Double- crested Cormorants in eastern Lake Ontario. Abstracts of Scientific Papers Presented at the 1992 Annual Meeting of the Colonial Waterbird Society: 76, and Colonial Waterbird Society Bulletin l6(2):63 -64. Weseloh, D.V.C. and P.J. Ewins. 1994. Characteristics of a rapidly increasing colony of Double-crested Cormorants (Phalacrocorax auritus) in Lake Ontario: population size reproductive parameters and band recoveries. Journal of Great Lakes Research 202443-456. Weseloh, D.V.C., Ewins, P.J., Struger, J., Mineau, P., and Bishop, CA. 1995. Double- crested Cormorants of the Great Lakes: Changes in population size, breeding distribution and reproductive output between 1913 and 1991. Colonial Waterbirds 18 (Special Publication 1), 48-59. Weseloh, D.V.C., Pekarik, C., Havelka, T., Barrett, G. and Reid, J. 2002. Population trends and colony locations of Double-crested Cormorants in the Canadian Great ' Lakes and immediately adjacent areas, 1990-2000: a manager’s guide. J. Great Lakes Research 28(2), 125-144. Wiens, J .A. 1984. Modeling the energy requirements of seabird populations. IN: Seabird Energetics. G.C. Whittow and H. Rahn, eds. Plenium Press, New York. pgs:255-285. Wiens, J .A. and J .M. Scott. 1975. Model estimation of energy flow in Oregon coastal seabird populations. Condor 77:439-452. 218 Wires, L.R, F.J. Cuthbert, DR Trexel, and AR. Joshi. 2001. Status of the Double- crested Cormorant (Phalacrocorax auritus) in North America. Final Report to U.S. Fish and Wildlife Service, Arlington, V.A. Zilj stra, M. and MR. van Eerden. 1995. Pellet production and the use of otoliths in determining the diet of Cormorants Phalacrocorax carbo sinensis: trials with captive birds. Ardea 83:123-131. 219 IIIIIIIIIIIIIIIIIIIIIIII 11111111 111111ll11111111111111)“ 3 1293 02736 0 -‘ —-.-_v— .I‘l. _4 fl