ABSTRACT

INVASIVE SPECIES SUSTAIN DOUBLE-CRESTED CORMORANTS IN SOUTHERN LAKE MICHIGAN

Patrick Madura, MS Department of Biological Sciences Northern Illinois University, 2015 Holly Jones, Director

Double-crested cormorants (Phalacrocorax auritus) have long and often been implicated in having detrimental effects on fisheries. Resulting persecution, as well as DDT water contamination, led to a major decline of the species throughout its range. Research shows that the main components of cormorant diets vary significantly among forage, invasive, or economically important fish species. A recent, rapid increase in cormorant abundance in the

Great Lakes has led, in some instances, to calls for the management of cormorant populations.

Thus the objective of this study was to determine the prey composition of the double-crested cormorant colony in East Chicago, Indiana. This study builds on previous work in the north basin of Lake Michigan by focusing on cormorant diet composition at their only significant nesting colony in the southern basin where cormorant diet has been unstudied. Regurgitated pellets were collected from the colony and diagnostic bones were used to elucidate diet composition. ( pseudoharengus), round goby (Neogobius melanostomus), Lepomis spp., white perch (Morone americana) and yellow perch (Perca flavescens) were the most frequently found prey items depending on cormorant breeding stage. Invasive species (i.e., alewife, round goby, white perch) contributed over ninety percent of the individuals and eighty percent of the biomass to cormorant diet over the study period. No salmonine species were detected suggesting that negative effects on this important fishery would likely occur only via direct competition for prey (e.g., alewife). Predation on yellow perch, which occurred mainly prior to and during the perch spawning season (i.e., of age-1 and older individuals), may warrant further study to quantify the effects on the local recreational yellow perch fishery. However, because yellow perch abundance is thought to be currently limited by poor recruitment at age-0, these results do not support active management of the cormorants at this colony to protect local fisheries.

NORTHERN ILLINOIS UNIVERSITY DEKALB, ILLINOIS

MAY 2015

INVASIVE SPECIES SUSTAIN DOUBLE-CRESTED CORMORANTS IN SOUTHERN LAKE MICHIGAN

BY

PATRICK T. MADURA

A THESIS SUBMITTED TO THE GRADUATE SCHOOL

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE

MASTER OF SCIENCE

DEPARTMENT OF BIOLOGICAL SCIENCES

Thesis Director: Holly P. Jones

ACKNOWLEDGEMENTS

My thesis advisor, Professor Holly Jones, and committee members Professor Nick

Barber, Dr. Owen Gorman, and Professor Emeritus Carl von Ende, provided consistent encouragement and guidance throughout the process of developing this project, carrying out its implementation, analyzing the resultant data, and editing drafts of manuscripts. Robert Page

(AECOM) assisted with site access. Randy Brindza, John Castrale, Jake Fitzhugh (IN-DNR),

Heather Herakovich, Stephanie Kong, Taylor McDonald, and Alyse Olson (NIU) donated their time collecting samples. Samples were collected under Indiana Scientific Purposes License #14-

001, which was provided by the Indiana Department of Natural Resources – Division of Fish and

Wildlife. Suzanne ven Housen and Taylor McDonald (NIU) gave much needed help sorting otoliths. Christopher Dalton (Cornell) shared his knowledge of sample processing. Owen

Gorman (USGS), Joshua Dub, Sara Creque, Rebecca Redman (INHS), Brian Breidert (IN-DNR),

Steve Krueger (IL-DNR), and Michael Brien (NIU) donated reference specimens. Jim Johnson

(USGS) and Joshua Dub (INHS) helped with identification of otoliths. Steve Lewis (US-FWS) and Brian Breidert (IN-DNR) provided crucial support of this project at a critical moment.

DEDICATION

In memory of my cousin Christopher, whose generous heart nourished me through graduate school and blessed countless children who struggled with life-threatening illnesses.

TABLE OF CONTENTS

Page

LIST OF TABLES ...... v

LIST OF FIGURES ...... vi

Chapter

1. INVASIVE SPECIES SUSTAIN DOUBLE-CRESTED CORMORANTS IN SOUTHERN LAKE MICHIGAN ...... 1

Introduction ...... 2

Materials and Methods ...... 6

Data Analysis ...... 10

Results ...... 15

Discussion ...... 20

REFERENCES...... 27

LIST OF TABLES

Table Page

1. Mean sagittal otolith lengths and the sources of regression equations used for calculating biomass consumed by cormorants ...... 12

2. Composition of cormorant diets at East Chicago breeding colony during the years 2013-2014 ...... 16

3. Temporal overlap in the numerical composition of cormorant diets ...... 17

4. Results of multinomial logistic regression of numerical composition of major prey taxa in cormorant diets by breeding stage and year...... 17

LIST OF FIGURES

Figure Page

1. East Chicago, IN. study site with 30 km foraging radius, bathymetry in meters, and fish stocking locations ...... 8

2. Percent frequency of prey taxa in cormorant diets by cormorant breeding stage ...... 19

3. Comparison of numerical and estimated biomass composition of cormorant diets at the East Chicago colony with alewife, yellow perch, and round goby consumption further broken down into cormorant breeding stages ...... 20

4. Contribution by biomass of invasive species to cormorant diets ...... 21

CHAPTER 1

INVASIVE SPECIES SUSTAIN DOUBLE-CRESTED CORMORANTS

IN SOUTHERN LAKE MICHIGAN

2 Introduction

Humans have often perceived wildlife to be at odds with their economic needs. For centuries we have been in conflict with the double-crested cormorant (Phalacrocorax auritus, hereafter cormorant), a piscivorous, colonial-nesting waterbird with numerous breeding colonies throughout North America (Wires and Cuthbert, 2006). Cormorants can be factors in the decline of sport and commercial fisheries (Burnett et al., 2002; Lantry et al., 2002; Rudstam et al., 2004), alter soil characteristics and vegetation (Boutin et al., 2011; Rush et al., 2011), and potentially displace other bird species near their nesting colonies (Cuthbert et al., 2002; Somers et al., 2011).

Fishery impacts are a main cause of contention regarding cormorant colonies and are often the main impetus for cormorant control.

Cormorants are native to the Great Lakes region (Wires and Cuthbert 2006), but were subject to persecution and exposure to dichlorodiphenyltrichloroethane (DDT) that resulted in a marked decline of cormorant populations through the 1960s and extirpation of breeding birds from Lake Michigan (Wires et al., 2001). After legal protection and the banning of DDT, the cormorant is now widespread (Wires and Cuthbert, 2006). Still, the cormorant metapopulation has not reached biological carrying capacity or even its historical abundance, but — perhaps in part because cormorants had recently been so relatively rare — they often exceed their social carrying capacity (Wires and Cuthbert, 2006), the maximum population size considered to be acceptable to stakeholders and the public. In a few cases, cormorants are actively controlled using techniques such as oiling eggs to limit reproduction and culling adults to control their numbers (Sullivan et al., 2006). Such intervention occurs at select breeding colonies in the Great

3 Lakes and elsewhere, as well as their wintering range in southern US states. In some cases, control measures have shown some benefit to fish populations (e.g., Fielder, 2010), and estimated breeding cormorant abundance has recently declined in the Great Lakes region

(Cuthbert and Wires, 2013). However, no studies to date have demonstrated that cormorants have a consistently negative effect on fisheries over broad geographical regions.

For instance, cormorants were found to cause declines in smallmouth bass (Micropterus dolomieu) populations in Lake Ontario’s east basin (Lantry et al., 2002). At Oneida Lake, just east of Lake Ontario, cormorants were the likely cause of a major decline in the yellow perch

(Perca flavescens) and walleye (Sander vitreus) sport fisheries (Rudstam et al., 2004). However, they have not negatively affected the smallmouth bass fishery at the Beaver Archipelago in northern Lake Michigan (Kaemingk et al., 2012) where they primarily consume invasive alewife

(Alosa pseudoharengus) (Seefelt and Gillingham, 2008). Additionally, abundant invasive species may help buffer against predation on economically important fish (Coleman et al., 2012; Johnson et al., 2010). After round goby (Neogobius melanostomus) invaded Lake Ontario’s east basin, cormorants consumed less yellow perch and smallmouth bass in favor of the round goby

(Johnson et al., 2010). Round goby also makes up a significant portion of cormorant diet in the western basin of Lake Ontario (Somers et al., 2003) and in the Niagara River (Coleman et al.,

2012).

Usually described as opportunistic piscivores (e.g., Hatch and Weseloh, 1999), some studies suggest cormorants could potentially be selective. For example, cormorant consumption of gizzard shad (Dorosoma cepadianum) at Oneida Lake increased, even while yellow perch and walleye density remained high (DeBruyne et al., 2013). Stable isotope analysis using samples

4 from three Saskatchewan lakes found cormorants might have a dietary niche of foraging for pelagic, schooling fish (Doucette et al., 2011). And, attempts to feed salmonines to captive birds were met with reluctance and followed by illness (Lewis, 1929). However, perhaps the most consistent result of cormorant dietary studies is that their prey composition is highly variable over space and time. Variable cormorant diets that sometimes rely on commercially important fish and other times do not suggest that cormorant impacts on fisheries should be evaluated and management decisions made on a case-by-case basis.

In Lake Michigan, the potential predation on salmonines by cormorants is a special concern as they are the taxa most targeted by Indiana fishers (Palla, 2011), make up most of the biomass taken by recreational fishers (Breidert et al., 2014), and salmonine (i.e., salmon and trout) populations are reliant on hatchery-augmented reproduction (Crawford, 2001). Salmonines are members of the family Salmonidae, which is comprised of the salmonines, coregonines (i.e., whitefishes) and the thymallines (i.e., graylings). Introduced, non-native salmonine populations were established in the Great Lakes mainly to control invasive alewife die-offs and rejuvenate ailing fisheries that had been decimated by the invasive sea lamprey (Petromyzon marinus)

(Crawford, 2001). Introduced salmonines of particular importance to sportfishers and managers are chinook (Oncorhynchus tshawytscha) and coho (Oncorhynchus kisutch) salmon, and brown

(Salmo trutta) and rainbow (Oncorhynchus mykiss) trout. Because salmonines are stocked with hatchery-produced juveniles, they may be at a higher risk of predation than native populations due to increased stress, disease, or a naiveté of predators (Collis et al., 2002).

No studies to date have suggested salmonines are a major component of cormorant diet in the Great Lakes, but they sometimes comprise a small portion (e.g., Ross and Johnson, 1997;

5 Somers et al., 2003). Cormorant-salmonine interactions are most likely to occur when colonies are located next to stocking locations, such as harbors or streams, and when salmonine migratory patterns take them within foraging and diving depth range of cormorants. In spring, salmonines are generally located nearshore, but they move offshore to deeper, cooler water in the summer and later return to shallow water in the fall (Adlerstein et al., 2008; Becker, 1983; Nettles et al.,

1987; Stewart and Bowlby, 2009). Salmonine movements are also dependent on those of their prey like the alewife (Adlerstein et al., 2008). Alewife move inshore to spawn in early summer

(Becker, 1983) and were found to be an important prey of cormorants in northern Lake Michigan

(Seefelt and Gillingham, 2008).

Yellow perch constitutes another important Lake Michigan fishery. It was the second most sought-after taxa by Indiana fishers (Palla, 2011) and represented nearly 3% of the total fish biomass taken from the lake (Breidert et al., 2014). Economically important yellow perch and their prey fish, the invasive round goby, are present nearshore throughout the cormorant breeding season (Becker, 1983; Kornis and Mercado-Silva, 2011). Each of these species have been found to make significant contributions to cormorant diets elsewhere (Coleman et al., 2012; Fielder,

2008; Johnson et al., 2010; Rudstam et al., 2004). Although cormorant diet is relatively well- studied in northern Lake Michigan, no studies have been done in southern Lake Michigan. The objective of this study was to determine the dietary composition of the cormorants at their only significant nesting colony in southern Lake Michigan (Wires and Cuthbert, 2010) with a focus on potential predation on economically important fish.

6 Materials and methods

Study Species

The double-crested cormorant (Phalacrocorax auritus) is a piscivorous (i.e., fish-eating) waterfowl. It gains its name from the specialized nuptial (i.e., pre-breeding) feathers located above and behind each eye. The color of these feathers varies among subpopulations from white to black (Hatch and Weseloh, 1999). Cormorants in the Great Lakes region typically have black crests (Hatch and Weseloh, 1999). They also have feathers that are able to take up water and reduce body buoyancy (Lewis, 1929). Cormorants forage by diving from the water surface and propelling themselves underwater using their large totipalmate (i.e., all four toes are webbed) feet (Lewis, 1929). Upon returning to land, they will often rest with wings spread out, likely to facilitate the drying of their feathers (Hatch and Weseloh, 1999).

The cormorant subpopulation in the interior of North America is migratory with birds arriving in the Great Lakes region from the Gulf of Mexico and neighboring states in late March or April (Hatch and Weseloh, 1999). Cormorants are colonial-nesting waterbirds and may form breeding colonies ranging from dozens to thousands of pairs (Wires and Cuthbert, 2010); though there was a colony recorded in 1913 at San Martin Island in Baja California, Mexico that was

200,000-350,000 pairs (Wires and Cuthbert, 2006). However, by 1970 breeding cormorants had been extirpated from Lake Michigan with only an estimated 89 breeding pairs remaining in the

Great Lakes, but by 1997, the number had increased to 88,000 pairs (Wires et al., 2001). More

7 recently however, the estimated number of cormorants breeding here had declined 28% from

2007-2012 (Cuthbert and Wires, 2013).

Study site

This research analyzed the diet of the cormorant colony present at East Chicago, Indiana

(approx. 25km. SSE of Chicago, IL.) on the property of a large steel mill (Figure 1). The steel mill is situated on a manmade, backfilled peninsula that juts prominently into the southwest corner of Lake Michigan. This colony is the only significant cormorant colony occurring in the southern basin of Lake Michigan (Wires and Cuthbert, 2010). Cormorants were first recorded nesting at this colony in 2004 (Wires and Cuthbert, 2010). By 2013, the colony was comprised of

2,764 cormorant nests; in 2014, the nest count shrank to 2,166 (John Castrale, Indiana Dept. of

Natural Resources, personal communication). In 2013, cormorants were co-nesting adjacent to greater numbers of ring-billed (Larus delawarensis) and gulls (Larus argentatus) and a few dozen great egrets (Ardea alba), but in 2014 there was one order of magnitude fewer gulls observed (P.T. Madura, personal observation). The cormorants at this colony had not been and were not being managed during the course of the study.

While cormorants typically forage close to their breeding colonies (Hatch and Weseloh,

1999), they can travel up to 30 km away when feeding (Bugajski et al., 2013). One salmonine stocking location is located only 4 km away (FWS/GLFC, 2010). At the seven stocking locations that are within the 30km cormorant foraging range, approximately 3.5 million juvenile salmonines have been stocked over the five-year period 2008-2012. Stocked salmonines are

8

Figure 1 East Chicago, IN. study site (O) with 30 km foraging radius, bathymetry in meters (NOAA/NGDC, 1996), and fish stocking locations (Δ) (FWS/GLFC, 2010).

60-183 mm in length (FWS/GLFC, 2010); this range is within of cormorants' preferred prey size

(<150mm, Hatch and Weseloh 1999). Cormorants are capable of diving to depths of 25 m

(Coleman, 2009). Local water depth is less than 10 m (NOAA/NGDC, 1996), and therefore cormorants at this location are able to forage throughout the pelagic and demersal zones where salmonines and yellow perch are located, respectively.

9 Sample collection and processing

Pellets egested by cormorants were collected twice in 2013 (April 30 and September 11) and twice per month throughout the 2014 cormorant breeding season (i.e., April through

August). Fewer pellets were collected in 2013 because of restricted access to the breeding colony. Pellets were collected randomly within the colony with no more than one pellet collected per nest and preference was given to the collection of fresh pellets. Approximately 40 pellet samples were obtained during each collection. Each pellet was bagged and labeled, placed on ice, and transported back to the lab in a cooler and then stored in a freezer until further processing.

Pellets were immersed in 1M NaOH and allowed to soak for 24 hours (Dalton et al.,

2009) to dissolve mucus and soft tissue and to prevent further acidic decay of the calcium carbonate composing bony tissue. Pellet contents were then rinsed in a 0.5mm sieve (No. 35

Std.) using cold water. Remaining hard tissue was then allowed to dry overnight at room temperature. Using a stereomicroscope, diagnostic bones and other calcified tissue (i.e., otoliths, cleithra, pharyngeal jaw fragments, spinous rays, and scales) were separated manually and stored in vials. Manually comparing sagittal otoliths with a reference collection was the primary means of identifying prey items. Otoliths are paired bones in the skulls of fishes. Only 2% of diagnostic otoliths (i.e., those retaining identifying characteristics), found in 10% of sampled pellets (35 of

350), were not identifiable to species but were instead identified to the lowest possible taxon.

Furthermore, cyprinids and other Ostariophysi have reduced sagittal otoliths, resulting on reliance on other bony tissues for identification. For example, spines and pharyngeal teeth (when

10 present) were keyed according to Schofield et al. (2005). Cleithra were compared with Traynor et al. (2010); this method detected an additional prey taxon: Ictaluridae.

Dietary analysis using otoliths from pellets is subject to bias due to differential digestibility of otoliths (e.g., smaller otoliths disintegrate more quickly in the stomach of cormorants) and items within pellets may belong to prey secondarily consumed by piscivorous fish (Carss, 1997). However, this method allowed for a simple analysis of diet composition throughout the cormorant breeding season without the need to euthanize birds. An examination of various methods of cormorant dietary analysis found the importance of prey items depended on the method used, but pellets were considered useful in qualitatively describing cormorant diet

(Seefelt and Gillingham, 2006).

Data Analysis

Similar to Johnson (2010), pellet samples were categorized according to the cormorant breeding stage during which they were collected: pre-chick (before egg hatch, April), chick

(nestlings present, May 7 through July 11), post-chick (juveniles fledged and observed in flight,

July 23 and after). Otolith counts were then used for estimating proportional diet composition.

Numerical composition (percent number, %N) was calculated for each prey taxon and for each breeding stage as follows:

11 푁 푥푖 ∑푖=1[ ] 푛푖 %푁 = ∗ 100% 푁

where xi is the number of otoliths belonging to the prey taxon in pellet i, ni is the total number of otoliths in pellet i, and N is the total number of pellets collected during that breeding season.

Frequency of occurrence (percent number, %F) was calculated for each prey taxon and for each breeding stage as follows:

푝 %퐹 = ∗ 100% 푁

where p is the number of pellets containing otoliths belonging to the prey taxon and N is the total number of pellets collected during that breeding stage.

As numerical composition may not reflect the proportional biomass consumed by cormorants, prey mass was calculated from otolith lengths using regressions obtained from the literature (Table 1). Regressions relating otolith length directly to fish mass were available only for round goby (Sokolowska and Fey, 2011); but, otolith length to fish length regressions

12 Table 1 Mean sagittal otolith lengths and the sources of regression equations used for calculating biomass consumed by cormorants.

Mean otolith length: mm. (+/- 1 S.D.) Otolith length to Fish total length to 2013 2013 2014 2014 2014 fish total length fish fresh mass Pre Post Pre Chick Post Alewife N/D 2.07 N/D 2.25 2.42 (Ross et al., 2005) (Schneider et al., 2000) (0.25) (0.27) (0.15) Common a N/A N/A N/A N/A N/A N/A e (Schneider et al., 2000) Channel catfish b N/A N/A N/A N/A N/A N/A e (Schneider et al., 2000) Gizzard shad N/A N/A N/A N/A N/A N/A e (Schneider et al., 2000) Goldfish a N/A N/A N/A N/A N/A N/A e (Schneider et al., 2000) Pumpkinseed c 5.22 N/D 4.39 5.44 N/D (Ross et al., 2005) (Schneider et al., 2000) (1.11) (0.68) (1.36) Round goby 3.89 3.09 3.73 2.78 3.25 (Sokolowska and Fey, 2011) f (0.38) (0.52) (0.89) (0.43) (0.42) Spottail shiner d N/A N/A N/A N/A N/A N/A e (Schneider et al., 2000) White perch 7.88 N/D 7.39 7.98 N/D (Dalton et al., 2009) (Schneider et al., 2000) (0.41) (0.99) (1.73) Yellow perch 5.73 5.00 5.54 5.07 5.04 (Ross et al., 2005) (Schneider et al., 2000) (1.13) (0) (0.97) (1.58) (1.29) a Common carp regression used for common carp, goldfish, and Cyprininae estimates. b Channel catfish used for Ictaluridae estimates. c Pumpkinseed regression used for Lepomis spp. estimates. d Spottail shiner regression used for spottail shiner and Leuciscinae estimates. e Assumed an average total length of 103.5mm. f Fresh mass was calculated directly from otolith length. N/A is not applicable because it was a minor species. N/D is not detected.

were available for alewife, pumpkinseed (Lepomis gibbosus), white perch (Morone americana), and yellow perch (Ross et al., 2005; Dalton et al., 2009). For these prey species during each cormorant breeding stage, a mean sagittal otolith length was obtained from a subsample of otoliths and used to calculate the length (mass for round goby) of the average fish consumed; only the least degraded otoliths were used for this estimation. For other species, the average total length of a fish was assigned as 103.5 mm. The assignment of 103.5 mm total length to these less common species was derived from the arithmetic mean of field-collected regurgitated voucher specimens, which reflects the average size of fish consumed by cormorants. Fish mass was then estimated using fish length to weight regressions (Schneider et al., 2000). Since length-weight

13 regressions were not available for all species that occurred in cormorant diets, the body mass was approximated by using length-weight regressions of taxonomically close species: channel catfish

(Ictaluris punctatus) was used for unidentified ictalurids; common carp (Cyprinus carpio) was used for goldfish (Carassius auritus) and unidentified (Cyprininae); pumpkinseed

(Lepomis gibbosus) was used for Lepomis spp.; spottail shiner (Notropis hudsonius) was used for unidentified (Leuciscinae).

Percent biomass (%B) was then computed for each prey taxon and for each breeding stage as follows:

푁 푥 ∗ 푚 %퐵 = ∑ [ ] ∗ 100% ∑푠 [푥 ∗ 푚 ] 푖=1 푗=1 푗 푗

where x is the number of otoliths belonging to the prey taxon of interest, m is the estimated mass of an average individual belonging to the prey taxon of interest during the breeding stage, xj is the number of otoliths belonging to prey taxon j in pellet i, mj is the estimated mass of an average individual of prey taxon j, s is the number of taxa detected in the pellet, and N is the total number of pellets collected during that breeding stage. Since proportional results were used for analysis, total number or absolute biomass of prey consumed was not determined.

14 To account for temporal variation in the numerical composition of prey in cormorant diets, multinomial logistic regression models with a cumulative logit link function were created using the Generalized Linear Models function in IBM® SPSS® version 22. For each model, percent numerical composition by a major prey taxon was used as the response variable, and year collected and cormorant breeding stage were used as predictors. A likelihood ratio chi-square

(omnibus) test of the fitted verses a thresholds-only model was used to validate the model. A

Wald chi-square test of model effects was then performed to determine which of these predictors were significant for a given prey taxon. Finally, overlap in mean numerical composition of whole cormorant diets between years and between breeding stages was quantified by calculating Cλ

(Johnson et al., 2010):

푠 2 ∑푖=1[푋푖푌푖] 퐶휆 = 푠 2 푠 2 ∑푖=1[푋푖 ] + ∑푖=1[푌푖 ]

where Cλ is the overlap value, s is the total number of prey taxa detected, Xi is the numerical proportion of diet during period X contributed by prey taxon i, and Yi is the numerical proportion of diet during period Y contributed by prey taxon i. Values of Cλ range from zero to one representing no overlap to identical composition between periods, respectively.

15 Results

Detected prey taxa consisted of alewife; gizzard shad; Lepomis spp.; round goby; white perch; yellow perch; common carp (Cyprinus carpio), goldfish (Carassius auratus), and unidentified Cyprininae; spottail shiner (Notropis hudsonius) and unidentified Leuciscinae, and

Ictaluridae (Table 2). No salmonine species were detected in cormorant diets during the years

2013-2014.

Temporal overlap (Cλ) in the numerical composition of whole cormorant diet was strong between years (Cλ = 0.8627) indicating that diets were very similar between 2013 and 2014

(Table 3). This was not the case between breeding stages, except between the chick and post- chick stages (Cλ = 0.8497), suggesting that cormorant diet varies between the pre-chick and chick breeding stages (Cλ = 0.1874).

Significant omnibus tests of factorial (i.e., predictors and their interaction), multinomial logistic regression models of percent number round goby and yellow perch indicated that the models outperformed null thresholds-only models (Table 4). Main effects (i.e., factors only) models of percent number alewife, Lepomis spp. and white perch also outperformed null models

(Table 4); a factorial model was not created for these species due to quasi-complete separation in the data (i.e., Lepomis spp. were not detected during the 2013 post-chick stage, alewife was not detected during the 2013 or 2014 pre-chick stage, and white perch was not detected during the

2013 or

16 Table 2 Composition of cormorant diets at East Chicago breeding colony during the years 2013-2014. Samples were collected during the pre-chick hatch, chick present in nest, and post-chick fledging breeding stages. No samples were collected during the chick breeding stage of 2013.

Percent Frequency Percent Number Percent Biomass 2013 2014 2013 2014 2013 2014

Pre Post Pre Chick Post Pre Post Pre Chick Post Pre Post Pre Chick Post n=17 n=16 n=47 n=164 n=106 n=17 n=16 n=47 n=164 n=106 n=17 n=16 n=47 n=164 n=106 Yellow perch 65% 25% 68% 30% 19% 42% 6% 49% 3% 2% 48% 12% 55% 7% 4% Round goby 41% 100% 11% 76% 91% 19% 88% 5% 48% 76% 14% 80% 5% 36% 70% Alewife 0% 25% 0% 79% 49% 0% 6% 0% 43% 20% 0% 7% 0% 51% 23% White perch 18% 0% 21% 3% 0% 8% 0% 13% 2% 0% 11% 0% 14% 3% 0% Lepomis spp. 12% 0% 26% 6% 0% 5% 0% 14% 3% 0% 5% 0% 13% 3% 0% Gizzard shad 29% 0% 17% 1% 0% 13% 0% 8% 0% 0% 8% 0% 5% 0% 0% Cyprininae 6% 6% 4% 0% 0% 2% 1% 2% 0% 0% 1% 1% 2% 0% 0% Common carp 6% 0% 2% 0% 0% 6% 0% 2% 0% 0% 6% 0% 2% 0% 0% Goldfish 6% 0% 4% 1% 1% 6% 0% 4% 0% 1% 6% 0% 3% 0% 1% Leuciscinae 0% 0% 4% 3% 0% 0% 0% 2% 0% 0% 0% 0% 1% 0% 0% Spottail 0% 0% 2% 0% 0% 0% 0% 1% 0% 0% 0% 0% 0% 0% 0% shiner Ictaluridae 0% 0% 0% 0% 1% 0% 0% 0% 0% 1% 0% 0% 0% 0% 1%

17 Table 3 Temporal overlap in the numerical composition of cormorant diets. A Cλ value of zero indicates no overlap, whereas a value of one signifies identical composition.

Between Years Cλ 2013-2014 0.8627 Between Stages

Pre-Chick  Chick 0.1874 Chick  Post-Chick 0.8497 Pre-Chick  Post 0.1785

Table 4 Results of multinomial logistic regression of numerical composition of major prey taxa in cormorant diets by breeding stage and year. A cumulative logit link function was used for all models. Zero consumption was used as the reference category.

Omnibus Testb Breeding Stage Year Collected Breeding Stage * Year Collected c Alewife a X2=162.374, df=3, p<0.001 X2=35.112, df=2, p<0.001 X2=2.850, df=1, p=0.091 N/A

Lepomis spp.a X2=34.229, df=3, p<0.001 X2=13.050, df=2, p=0.001 X2=1.408, df=1, p=0.235 N/A

Round goby X2=167.807, df=4, p<0.001 X2=90.806, df=2, p<0.001 X2=6.745, df=1, p=0.009 X2=2.739, df=1, p=0.098

White perch a X2=31.970, df=3, p<0.001 X2=13.526, df=2, p=0.001 X2=0.173, df=1, p=0.677 N/A

Yellow perch X2=90.029, df=4, p<0.001 X2=52.225, df=2, p<0.001 X2=0.073, df=1,p=0.787 X2=1.083, df=1, p=0.298

a Main effects only model was performed due to quasi-complete separation in these data. b Significant omnibus (likelihood ratio chi-square) tests indicate that the fitted model outperforms a null (thresholds-only) model. c Test of an interaction between these two factors.

2014 post-chick stages). There was a strong, but insignificant interaction between breeding stage and year on alewife and round goby consumption by cormorants, but this was not the case with yellow perch (Table 4). Breeding stage was a significant factor in the consumption of all species, but year was only a significant predictor and marginally significant predictor for consumption of round goby and alewife, respectively (Table 4). Given the result of strong overlap of whole diet

18 between years and between chick and post-chick breeding stages, the high contribution by alewife and round goby during the 2014 chick and post-chick stage, and the absence of pellet collections from the 2013 chick stage, it is likely that the significance of year collected for estimating round goby consumption in the model is spurious. Therefore, results from each year were subsequently combined.

The most frequent taxa recovered from cormorant pellets were alewife, round goby, white perch, yellow perch, and Lepomis spp. (Figure 2). Generally, gizzard shad, yellow perch, white perch, and Lepomis spp. were prevalent during the pre-chick breeding stage (i.e., April samples) but subsequently gave way to a diet that was chiefly composed of alewife and round goby. These taxa also composed the highest proportion of prey by number and estimated biomass consumed (Figure 3) during the study period.

19

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Percent Frequency Percent Pre 0% Chick Post

Figure 2 Percent frequency of prey taxa in cormorant diets by cormorant breeding stage (combining years).

20 Numerical Contribution Biomass Contribution

White Yellow Yellow Yellow Yellow Perch Perch Perch Perch Perch 1% (Pre) (Pre) (Chick) (Post) 4% 8% 4% 1% White Alewife Perch (Chick) 4% Alewife 20% Alewife (Post) (Chick) 8% 30% Round Goby Round (Post) Goby 38% (Post) 21% Alewife (Post) Round Lepomis Round 10% Goby spp. Goby (Chick) 2% (Chick) 23% Round 17% Goby Lepomis (Pre) spp. 4% 5%

Figure 3 Comparison of numerical (left) and estimated biomass (right) composition (combining years) of cormorant diets at the East Chicago colony with alewife, yellow perch, and round goby consumption further broken down into cormorant breeding stages. Contributions of less than 1% were omitted for readability.

Discussion

Given the economic importance of the Lake Michigan fishery, the potential negative effects of cormorant predation on economically important fishes, and the monetary, ecological, and ethical concerns of managing cormorants, this study focused on determining the diet composition of cormorants at a previously unstudied nesting colony with a focus on the consumption of salmonine and other game fish. However, the diet analysis in this study does not

21 support the hypothesis that cormorants have a negative direct impact on Lake Michigan fisheries.

Invasive species formed the major component of cormorant diets at this colony (Figure 4). Over ninety percent of the numerical contribution and eighty percent of the biomass contribution to cormorant diet over the study period were made by invasive alewife and round goby, as well as white perch (Figure 3). Cormorants appear to have benefited from ecosystem changes that have

100%

90%

80% 70% 60% 50% 40% 30% 20% Percent Biomass Percent Non-Native 10% Invasive 0% Native

Figure 4 Biomass contribution of invasive species to cormorant diets at East Chicago breeding colony (combining years). Pre-chick breeding stage occurred during April, Chick stage May through July 11, and Post-Chick stage after July 11.

occurred in Lake Michigan since the mid-20th century: proliferation of non-native sea lamprey, alewife, dreissenid muscles, and round goby. The results of this study suggest that cormorants

22 are not drivers of these ecosystem changes but rather, they have capitalized on the increased abundance of non-native alewife and round goby in nearshore waters.

Ecosystem changes have resulted in the loss or decline of native species (e.g., yellow perch in nearshore waters frequented by cormorants). Because invasive species can be so devastating to native species and ecosystems, these negative effects are often spotlighted by the media. However, potential benefits provided by invasive species may be overshadowed

(Schlaepher et al., 2011). For instance, Lake Erie Water Snake (Nerodia sipedon insularum) growth rates increased after a change in diet in which round goby became the principal component (King et al., 2006). The snake had been listed for protection under the Endangered

Species Act, but was subsequently delisted after populations recovered with this new food source

(FWS, 2011). Similarly, the widespread availability and high abundance of alewife and round goby in Lake Michigan likely acted to promote the establishment and maintenance of the cormorants at this colony and thus aided the recovery of this native species. Along with such changes in the cormorant’s breeding range, Wires et al. (2001) list several other reasons for the recent, rapid population growth of cormorants: the banning of DDT, Migratory Bird Treaty Act protection, the creation of new habitat in the form of reservoirs and dredge spoil islands, and the proliferation of new food sources from aquaculture operations in their wintering range.

Invasive fishes may also benefit native fish species by buffering them from predation by cormorants. At the study locale, buffering by alewives would occur mostly during the chick breeding stage when both alewife and yellow perch are present nearshore. Conversely, buffering by alewives is limited during the pre-chick stage (i.e., April) when relative cormorant consumption of yellow perch was highest and alewife weren’t detected in cormorant diet. These

23 differences may be due to the relative timing of spawning between these species. In Lake

Michigan, yellow perch spawn earlier (i.e., middle to late spring) in the year than alewives do

(i.e., early to mid-summer) (Becker, 1983). Likewise, abundant round gobies may buffer predation on yellow perch throughout the whole of the cormorant breeding season. Gobies prefer nearshore habitat and may remain there to spawn multiple times per season (Kornis et al., 2012).

Although I detected nonnative common carp and goldfish, I did not find invasive Asian carps in cormorant diets. This is consistent with Jerde (2011) who had collected positive bighead

(Hypophthalmichthys nobilis) and silver carp (H. molitrix) eDNA samples in the Calumet

Harbor/Calumet Sag Channel system (and where a bighead carp individual had been recovered in 2010), but not at the sites closest to the study colony (e.g., Indiana Harbor, Gary boat slip;

Jerde et al., 2013). Nevertheless, given the proximity of these eDNA detections and predation on other carp species by the cormorants at this location, researchers monitoring invasive Asian carp dispersal should consider factoring cormorants into their sampling design.

Since salmonines were not detected in cormorant diet at this location over the study period, it can be reasonably inferred that this colony of cormorants has little direct negative effect of predation on the salmonine fisheries. Zero predation could be the result of low encounter rates with salmonines due to minimal overlap of cormorant and salmonine habitat or to sufficient overlap but low relative abundance of salmonines, poor attack success rates of cormorants on salmonines, or selection by cormorants for other species of fish. Salmonines have relatively cold water requirements compared to other fishes causing them to escape predation risk by moving to deeper, cooler parts of the lake. The Great Lakes region had one of its coldest springs in 2014 (NOAA, 2014). Even when factoring in annual fluctuations in temperature,

24 nearshore areas in southern Lake Michigan may be too warm for salmonines to remain in foraging range of cormorants. In Lake Michigan, 50% of chinook salmon are recruited by natural reproduction (Robillard et al., 2012) and thus spawn in tributaries. The mouths of these streams may not be located near enough to this cormorant colony for juvenile salmon (i.e., smolts) to be at risk of predation. Adult salmon and trout are top-predators in the lake, and although supplemented by stocking juveniles, it can be reasonably assumed that their relative abundance is less than prey fishes. This is supported by an Indiana creel survey (Palla, 2011) that showed salmons including smolts were caught a tenth as often (i.e., catch per unit effort) as yellow perch.

Juvenile salmonines may also escape predation by growing rapidly in cooler, productive areas of

Lake Michigan, which can contribute to low encounter rates and poor attack success rates by cormorants. Finally, cormorants may be selecting other species of fish, as was suggested by

Lewis (1929) when cormorants initially rejected the fresh trout he offered them and became ill after consuming them. However, Collis et al. (2002) showed that salmonids were a major food source for cormorants nesting in the Columbia River estuary in the Pacific Northwest, but salmonines have been detected in low numbers in Great Lakes studies of cormorant diets (e.g.,

Ross and Johnson, 1997; Somers et al., 2003).

While cormorants may not have had a direct predatory effect on salmonines at this study location, they may affect economically important fisheries via competition over shared prey. The occurrence of alewife in cormorant diets is notable because this species is the primary prey of stocked Chinook salmon (Jacobs et al., 2013). Although cormorant predation on alewife may reduce the forage base available to salmonines, consumption of alewives by cormorants may be beneficial to the larvae of native lake trout (Salvelinus namaycush), burbot (Lota lota), and

25 yellow perch, which are subject to predation by alewives (Bunnell et al., 2006). Additionally, the shift to a diet primarily composed of invasive round goby in the post-chick cormorant breeding stage may act in competition with yellow perch as they have been shown to heavily consume gobies (Truemper et al., 2006). However, critical yellow perch population bottlenecks occur during the larval, planktivorous portion of their life cycle. Further research is needed to quantify the importance of competition between cormorants and game fishes at this location.

In contrast to the absence of salmonines, the prevalence of yellow perch in the pre-chick diets of the cormorants at this colony suggests that cormorants could have a direct, negative effect on the perch population. Yellow perch is a popular target species of anglers and represents an important fishery on its own. Unfortunately, it is one that has experienced decline. The main basin of Lake Michigan has been closed to commercial yellow perch harvesting since 1997, and all states have implemented creel limits for the recreational fishery (Santucci, Jr. et al., 2014).

Since yellow perch occurred most frequently in cormorant diets early in the year (before and/or during the yellow perch spawning season), a significant number of the individuals consumed are likely to be reproductive adults or age-1 and older juveniles. However, given that recent yellow perch declines are mainly being driven by poor survival of planktivorous age-0 larvae due to reduced lake-wide primary productivity, introduced dreissenid mussels, and resultant declines in zooplankton communities (Santucci, Jr. et al., 2014), the effects of predation on yellow perch by this colony of cormorants is likely to be limited. Furthermore, because cormorants function at a higher trophic level than yellow perch larvae, they would not act in direct competition for forage with one another. Nonetheless, since adult yellow perch disperse west along the southern shore

(Santucci, Jr. et al., 2014) and would likely congregate as they negotiate the obstacle produced

26 by Indiana Harbor and since managers deem the preservation of yellow perch spawning stock important to the fishery (Santucci, Jr. et al., 2014), quantifying the effect of cormorant predation on yellow perch in this area may warrant further study.

Cormorants have been implicated in impacting prized fisheries. Contributing factors in this judgment are their high visibility in the public eye because of novelty, aggregating in large conspicuous colonies, and their consumption of fish that we want to catch for food, sport, or income. Because double-crested cormorants have been only occasionally shown to cause declines in fishery yields, it is important that stakeholders consider the economic and ecological effects these birds may have on a case-by-case basis. At the study colony, cormorants primarily consumed invasive species that are known to either compete or prey upon important native fish species. Economically important salmonines were not found in egested cormorant pellets.

Although cormorants preyed on yellow perch, especially before and/or during the spawning season, the reduction in yellow perch abundance occurred before cormorants colonized this site and is currently being driven by poor recruitment of age-0 individuals (Santucci, Jr. et al., 2014).

So, it is unlikely that cormorants have played a significant role in the decline and current state of the southern Lake Michigan yellow perch fishery. In consideration of these results, active management of cormorants would not be supported in order to protect the local fisheries.

References

Adlerstein, S.A., Rutherford, E.S., Claramunt, R.M., Clapp D.F., Clevenger J.A., 2008. Seasonal movements of Chinook salmon in Lake Michigan based on tag recoveries from recreational fisheries and catch rates in gill-net assessments. Trans. Am. Fish. Soc. 137, 736-750. Becker, G.C., 1983. Fishes of Wisconsin. London: University of Wisconsin Press. Boutin, C., Dobbie, T., Carpenter, D., Hebert, C.E., 2011. Effects of double-crested cormorants (Phalacrocorax auritus Less.) on island vegetation, seedbank, and soil chemistry: evaluating island restoration potential. Restor. Eco. 19, 720-727. Breidert, B., Wright, K., Dickinson, B., Legler, N., Robillard S., Kolb, T., Hanson, D., DeSorcie, T., 2014. Harvest of Fishes from Lake Michigan in 2013. Presented at: 2014 Upper Lakes Meeting (25 March 2014, Windsor, ON.) Great Lakes Fisheries Commission: Lake Michigan Committee. Bugajski, A., Reudink, M.W., Doucette, J.L., Franks, S.E., Wissel, B., Somers, C.M., 2013. The complexity of cormorants: stable isotopes reveal multiple prey sources and feeding site switching. Can. J. Fish. Aquat. Sci. 70, 271-279. Bunnell, D.B., Madenjian, C.P., Claramunt, R.M., 2006. Long-term changes of the Lake Michigan fish community following the reduction of exotic alewife (Alosa psuedoharengus). Can. J. Fish. Aquat. Sci. 63, 2434-2446. Burnett, J.A., Ringler, N.H., Lantry, B.F., Johnson, J.H., 2002. Double-crested cormorant predation on yellow perch in the eastern basin of Lake Ontario. J. Great Lakes Res. 28, 202- 211. Carss, D.N., and 27 co-authors., 1997. Techniques for assessing cormorant diet and food intake: towards a consensus view, in: Bacetti, N., Cherubini, G. (Eds.), European Coference on Cormorants. Supplemento alle Ricerchi di Biologia della Selvaggini, Vol. XXVI, pp. 197- 230. Coleman, J.T.H., 2009. Diving behavior, predator-prey dynamics, and management efficacy of double-crested cormorants in New York State. Ph.D. Dissertation, Cornell University, Ithaca, NY. Coleman, J.T.H., Adams, C.M., Kandel, M., Richmond, M.E., 2012. Eating the invaders: the prevalence of round goby (Apollonia melanostomus) in the diet of double-crested cormorants on the Niagara River. Waterbirds 35, 103-113, Special Publication 1. Collis, K., Roby, D.D., Craig, D.P., Adamany, S., Adkins, J.Y., Lyons, D.E., 2002. Colony size and diet composition of piscivorous waterbirds on the lower Columbia River: implications for losses of juvenile salmonids to avian predation. Trans. Am. Fish. Soc. 131, 537-550. 28 Crawford, S.S., 2001. Salmonine introductions to the Laurentian Great Lakes: an historical review and evaluation of ecological effects. NRC Res. Press, Ottawa. Cuthbert, F.J., Wires, L.R. 2013. Monitoring colonial waterbirds at Great Lakes priority sites. Presented at: Great Lakes Double-crested Cormorant Management Working Group Meeting (5 March 2014, Lansing, MI.). Cuthbert, F.J., Wires, L.R., McKearnan, J.E., 2002. Potential impacts of nesting double-crested cormorants on great blue herons and black-crowned night-herons in the U.S. Great Lakes region. J. Great Lakes Res. 28, 145-154. Dalton, C.M., Ellis, D., Post, D.M., 2009. The impact of double-crested cormorant (Phalacrocorax auritus) predation on anadromous alewife (Alosa pseudoharengus) in south- central Connecticut, USA. Can. J. Fish. Aquat. Sci. 66, 177-186. DeBruyne, R.L., Coleman, J.T.H., Jackson, J.R., Rudstam, L.G., VanDeValk, A.J., 2013. Analysis of prey selection by double-crested cormorants: a 15-year diet study in Oneida Lake, New York. Trans. Am. Fish. Soc. 142, 430-446. Doucette, J. L., Wissel, B. and Somers, C. M., 2011. Cormorant-fisheries conflicts: stable isotopes reveal a consistent niche for avian piscivores in diverse food webs. Ecol. Appl., 21, 2987-3001. Fielder, D.G., 2008. Examination of the factors contributing to the decline of the yellow perch population and fishery in Les Cheneaux Islands, Lake Huron, with an emphasis on the role of double-crested cormorants. J. Great Lakes Res. 34, 506-523. Fielder, D.G., 2010. Response of yellow perch in Les Cheneaux Islands, Lake Huron to declining numbers of double-crested cormorants stemming from control activities. J. Great Lakes Res. 36, 207-214. FWS, 2011. Endangered and threatened wildlife and plants: Removal of the Lake Erie Water snake (Nerodia sipedon insularum) from the Federal List of Endangered and Threatened Wildlife. [Online]. Docket No: FWS–R3–ES–2010–0039. Available from: http://www.fws.gov/policy/frsystem/default.cfm. [Accessed: March 24, 2014]. FWS/GLFC, 2010. Great Lakes Fish Stocking Database. [Online]. Available from http://www.glfc.org/fishstocking/. [Accessed: June 22, 2013]. Hatch, J.J. , Weseloh, D.V., 1999. Double-crested cormorant (Phalacrocorax auritus), in: Poole, A., Gill, F., (Eds.), The Birds of North America. No. 441. The Birds of North America, Inc., Philadelphia, PA. Jacobs, G.R., Madenjian, C.P., Bunnell, D.B., Warner, D.M., Caramunt, R.M., 2013. Chinook salmon foraging patterns in a changing Lake Michigan. Trans. Am. Fish. Soc. 142, 362-372.

29 Jerde, C.L., Chadderton, W.L., Mahon, A.R., Renshaw, M.A., Corush, J., Budny, M.L., Mysorekar, S. Lodge, D.M., 2013. Detection of Asian carp DNA as part of a Great Lakes basin-wide surveillance program. Can. J. Fish. Aquat. Sci. 70, 1-5. Jerde, C.L., Mahon, A.R., Chadderton, W.L., Lodge, D.M., 2011. "Sight unseen" detection of rare aquatic species using environmental DNA. Conserv. Lett. 00, 1-8. Johnson, J.H., Ross, R.M., McCullough, R.D., Mathers, A., 2010. Diet shift of double-crested cormorants in eastern Lake Ontario associated with the expansion of the invasive round goby. J. Great Lakes Res. 36, 242-247. Kaemingk, M.A., Galarowicz, T.L., Clevenger, J.A., Clapp, D.F., Lenon, H.L., 2012. Fish assemblage shifts and population dynamics of smallmouth bass in the Beaver Archipelago, northern Lake Michigan: a comparison between historical and recent time periods amidst ecosystem changes. Trans. Am. Fish. Soc. 141, 550-559. King, R.B., Ray, J.M., Stanford, K.M., 2006. Gorging on gobies: beneficial effects of alien prey on a threatened vertebrate. Can. J. Zool. 84, 108-115. Kornis, M.S., Mercado-Silva, N., 2011. Round goby, Neogobius melanostomus. In: Lyons J, editor. Fishes of Wisconsin E-book [Internet]. Madison (WI): Wisconsin Department of Natural Resources, and Middleton (WI): U.S. Geological Survey; [Accessed: March 16, 2014]. Available from http://www.fow-ebook.us/. Kornis, M.S., Mercado-Silva, N., Vander Zanden, M.J., 2012. Twenty years of invasion: a review of round goby Neogobius melanostomus biology, spread and ecological implications. J. Fish Biol. 80, 235-285. Lantry, B.F., Eckert, T.H., Schneider, C.P., Chrisman, J.R., 2002. The relationship between the abundance of smallmouth bass and double-crested cormorants in the eastern basin of Lake Ontario. J. Great Lakes Res. 28, 193-201. Lewis, H.F., 1929. The natural history of the double-crested cormorant. Ph.D. Thesis. Cornell University, Ithaca, NY. NOAA/NGDC, 1996. Bathymetry of Lake Michigan. [Online]. Available from http://www.ngdc.noaa.gov/mgg/greatlakes/michigan.html. [Accessed: June 22, 2013]. NOAA, 2014. Quarterly climate impacts and outlook - Great Lakes region, June 2014. [Online]. Available from http://mrcc.isws.illinois.edu/pubs/pubsGreatLakes.jsp. [Accessed: Dec. 9, 2014]. Nettles, D.C., Haynes, J.M., Olson, R.A. and Winter, J.D., 1987. Seasonal movements and habitats of brown trout (Salmo trutta) in southcentral Lake Ontario. J. Great Lakes Res. 13, 168-177. Palla, J.S., 2011. Lake Michigan, 2010 creel survey report. Indiana Dept. of Natural Resources. Indianapolis., IN.

30 Robillard, S.R., Breidert, B., Eggold, B.T., Gorenflo, T.K., Wesley, J., 2012. Conclusions, in: Bunnell, D.B,(Ed.), The State of Lake Michigan in 2011. Great Lakes Fisheries Commission Spec. Pub. 12-01. Ross, R.M., Johnson, J.H., 1997. Fish losses to double-crested cormorant predation in eastern Lake Ontario, 1992–97. In: Tobin, M.E. (Ed.), Symposium on Double-crested Cormorants: Population Status and Management Issues in the Midwest (9 December 1997, Milwaukee, WI). U.S. Department of Agriculture Technical Bulletin No. 1879, pp. 61–70. Ross, R.M., Johnson, J.H., Adams, C.M., 2005. Use of fish-otolith regressions to infer size of double-crested cormorant prey fish from recovered otoliths in Lake Ontario. Northeast. Nat. 12, 133-140. Rudstam, L.G., VanDeValk, A.J., Adams, C.M., Coleman, J.T.H., Forney, J.L., Richmond, M.E., 2004. Cormorant predation and the population dynamics of walleye and yellow perch in Oneida Lake. Ecol. Appl. 14, 149–163. Rush, S.A., Verkoeyen, S., Dobbie, T., Dobbyn, S., Hebert, C.E., Gagnon, J., Fisk, A.T., 2011. Influence of increasing populations of double-crested cormorants on soil nutrient characteristics of nesting islands in western Lake Erie. J. Great Lakes Res. 37, 305-309. Santucci Jr., V.J., Eggold, B.T., Kalish, T.G., Price, J., Gorenflo, T.K., 2014. Summary report. Lake Michigan Yellow Perch Summit (22 March 2014, Chicago, IL). Great Lakes Fisheries Commission: Lake Michigan Committee. Schlaepher, M.A., Sax, D.F., Olden, J.D., 2011. The potential conservation value of non-native species. Conserv. Biol. 25, 428-437. Schneider, J.C., Laarman, P.W., Gowing, H., 2000. Length-weight relationships, in: Schneider, J.C. (Ed.), Manual of Fishery Survery Methods II: with periodic updates. Michigan Department of Natural Resources, Fisheries Special Report 25. Schofield, P.J., Williams, J.D., Nico, L.G., Fuller, P., Thomas, M.R., 2005. Foreign nonindigenous carps and minnows () in the United States: a guide to their identification, distribution, and biology. United States Geological Survey Special Investigations Report 2005-5041. Scott, W.B., Crossman, E.J., 1973. Freshwater Fishes of Canada. Ottawa: Freshwater Research Board of Canada, Vol. 184. Seefelt, N.E., Gillingham, J.C., 2006. A comparison of three methods to investigate the diet of breeding double-crested cormorants (Phalacrocorax auritus) in the Beaver Archipelago, northern Lake Michigan. Hydrobiologia 567, 57-67. Seefelt, N.E., Gillingham, J.C., 2008. Bioenergetics and prey consumption of breeding double- crested cormorants in the Beaver Archipelago, northern Lake Michigan. J. Great Lakes Res. 34, 122-133.

31 Sokolowska, E., Fey D.P., 2011. Age and growth of the round goby Neogobius melanostomus in the Gulf of Gdansk several years after invasion. Is the Baltic Sea a new promised land? J. Fish Biol. 78, 1993-2009. Somers, CM., Doucette, J.L., Weseloh, D.V., Kjoss, V.A., Brigham, R.M., 2011. Interactions between double-crested cormorants and other ground-nesting species. Waterbirds 34, 168- 176. Somers, C.M., Lozer, M.N., Kjoss, V.A., Quinn, J.S., 2003. The invasive round goby (Neogobius melanostomus) in the diet of nestling double-crested cormorants (Phalacrocorax auritus) in Hamilton Harbour, Lake Ontario. J. Great Lakes Res. 29, 392-399. Stewart, T.J. and Bowlby, J.N., 2009. Chinook salmon and rainbow trout catch and temperature distributions in Lake Ontario. J. Great Lakes Res. 35, 232-238. Sullivan, K.L., Curtis, P.D., Chipman, R.B., McCullough, R.D., 2006. The double-crested cormorant: issues and management. Cornell University, Ithaca, N.Y. Traynor, D., Moerke, A., Greil, R., 2010. Identification of Michigan fishes using cleithra. Great Lakes Fish. Comm. Misc. Publ. 2010-02. Truemper, H.A., Lauer, T.E., McComish, T.S., Edgell, R.A., 2006. Response of yellow perch diet to a changing forage base in southern Lake Michigan, 1984-2002. J. Great Lakes Res. 32, 806-816. Wires, L.R., Cuthbert, F.J., Trexel, D.R., Joshi, A.R., 2001. Status of the double-crested cormorant in North America (Phalacrocorax auritus). Final Report to USFWS. Wires, L.R., Cuthbert, F.J., 2006. Historic populations of the double-crested cormorant (Phalacrocorax auritus): implications for conservation and management in the 21st century. Waterbirds 29, 9-37. Wires, L.R., Cuthbert, F.J., 2010. Characteristics of double-crested cormorant colonies in the U.S. Great Lakes island landscape. J. Great Lakes Res. 36, 232-241.