W&M ScholarWorks

Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects

2003

Historical Diet Analysis of Loggerhead (Caretta caretta) and Kemp's Ridley (Lepidochelys kempi) Sea Turtles in Virginia

Erin E. Seney College of William and Mary - Virginia Institute of Marine Science

Follow this and additional works at: https://scholarworks.wm.edu/etd

Part of the Fresh Water Studies Commons, Oceanography Commons, and the Zoology Commons

Recommended Citation Seney, Erin E., "Historical Diet Analysis of Loggerhead (Caretta caretta) and Kemp's Ridley (Lepidochelys kempi) Sea Turtles in Virginia" (2003). Dissertations, Theses, and Masters Projects. Paper 1539617799. https://dx.doi.org/doi:10.25773/v5-hc11-p771

This Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Dissertations, Theses, and Masters Projects by an authorized administrator of W&M ScholarWorks. For more information, please contact [email protected]. H istorical diet Analysis of loggerhead {Ca r e t t a c a r e t t a) a n d

KEMP’S RIDLEY (LEPIDOCHELYSKEMPT) SEA TURTLES IN VIRGINIA

A Thesis Presented to

The Faculty of the School of Marine Science The College of William and Mary in Virginia

In Partial Fulfillment Of the Requirements for the Degree of Master of Science

by Erin E. Seney 2003 APPROVAL SHEET

This thesis is submitted in partial fulfillment of

the requirements for the degree of

Master of Science

Erin E. Seney

Approved, April 2003

. Musick, Ph.D. Chairman/Advisor

John M. Brubaker, Ph.D.

Robert H. George, D.V.M. Head Veterinarian, NMFS and Virginia Marine Mammal and Sea Turtle Stranding Networks Gloucester, Virginia

Roger~L. Mann, Ph.D. For my parents, Becky and Frank, for their love and encouragement over the years, and for my cat, Cuervo, who tolerated me when I smelled like turtle guts.

ii TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS...... v LIST OF TABLES...... vi LIST OF FIGURES...... viii ABSTRACT...... xii INTRODUCTION...... 2 METHODS...... 14 Sample Collection...... 14 Data Collection...... 17 Data Analysis...... 18 Diet descriptions...... 18 Diet comparisons...... 20 RESULTS...... 22 Loggerheads...... 22 Interannual diet...... 37 Size-specific diet...... 60 Sex-specific diet...... 68 Interseasonal diet...... 68 Kemp’s Ridleys...... 73 Interannual diet...... 87 Interspecific Competition...... 93 DISCUSSION...... 103 Loggerheads...... 103 Interannual diet...... 103 Size-specific diet...... 106 Sex-specific diet...... 106 Interseasonal diet...... 107 Kemp’s Ridleys...... 107 Interspecific Competition...... 108 Page

Concluding Remarks ...... 108 APPENDICES...... 110 LITERATURE CITED...... 115 PREY IDENTIFICATION REFERENCES...... 122 VITA...... 123

iv ACKNOWLEDGEMENTS

I would like to thank my major advisor, Dr. John (Jack) Musick, for his support and guidance throughout this project. I am also grateful to my committee members, Drs. John Brubaker, Robert George, and Roger Mann, for their time and advice. Although he was not on my committee, Dr. David Evans provided much needed advice on data analysis and reviewed a draft of this thesis. Dr. John E. Graves served as moderator for both my qualifying exam and my defense.

This study would not have been possible without the efforts made by many individuals throughout the VIMS Sea Turtle Program’s existence. As such, I would like to recognize and thank every VIMS student, employee, and volunteer, as well as the many state cooperators, who contributed to the collection of sea turtle gut samples and stranding data from 1980 to 2002. Roy Pemberton showed me the VIMS sampling protocol, and Meredith Fagan, Katy Frisch, Amber Knowles, and Anne Morrison helped me considerably with sample collection, sieving, and preservation during the 2001 and 2002 stranding seasons. I am also thankful to Trish Bargo, Lisa McFarland, Wendy Walton, and the other employees and volunteers from the Virginia Marine Science Museum Stranding Team who collected samples during 2001 and 2002. My ofificemate, Kate Mansfield, has taught me much about sea turtles, fieldwork, and the inner-workings of a turtle program, as well as being a willing proofreader and great friend.

Jim Gartland provided valuable insight and assistance during all steps of my project, from the literature search to prospectus edits to data analysis. Julia Ellis assisted with many prey identifications and questions, as well as editing a thesis draft. Melanie Harbin of the VIMS Fish Collection helped with fish identification, answered many questions, and assisted me whenever I needed more preservation supplies. Ken Goldman, Chris Hager, Juli Harding, Beth Hinchey, Mike Vecchione, and John Walter also aided in various prey identifications. Bobby Harris helped me develop a Microsoft Access database to link my data to the existing turtle database.

Joey Brown, Marilyn Lewis, Charles McFadden, and Diane Walker of the VIMS Library provided assistance finding reference materials at VIMS, with interlibrary loans, and in my search for horseshoe crab data. I am also grateful for the valuable reprints and correspondence provided by Mike Frick (Caretta Research Project, Georgia), Rom Lipcius (VIMS), Pamela Plotkin (East Tennessee State University), Carl Shuster (VIMS emeritus), and Dale Youngkin (formerly Florida Atlantic University). Stephanie Iverson, Rob O’Reilly, and Todd Watkins of the Virginia Marine Resources Commission answered countless questions and provided landings data.

Lastly, I would like to thank my parents, Frank Seney and Becky Mazzarella; my sisters, Lauren and Megan; my brother, Kevin; and my grandparents for their support of my academic and scientific endeavors over the years.

v LIST OF TABLES

Table Page

1. National Marine Fisheries Service condition codes for stranded sea turtles...... 16

2. Virginia Institute of Marine Science condition codes for stranded sea turtles...... 16

3. Composition of loggerhead diet data and samples collected during 1980 to 2002...... 27

4. Percent occurrence of all prey items found in loggerhead samples from Virginia during 1983 to 2002 (n = 166) ...... 45

5. Percent occurrence, percent dry weight, percent number, and percent index of relative importance of all prey items found in whole loggerhead gut samples from Virginia during 1983 to 1987 (n = 13)...... 47

6. Percent occurrence, percent dry weight, percent number, and percent index of relative importance of all prey items found in whole loggerhead gut samples from Virginia during 1988 to 1992 (n = 49)...... 48

7. Percent occurrence, percent dry weight, percent number, and percent index of relative importance of all prey items found in whole loggerhead gut samples from Virginia during 1993 to 1997 (n = 20)...... 50

8. Percent occurrence, percent dry weight, percent number, and percent index of relative importance of all prey items found in whole loggerhead gut samples from Virginia during 1998 to 2002 (n = 46)...... 52

9. Prey species with five highest percent index of relative importance values for loggerheads stranding in Virginia from 1983 to 2002 (n = 297)...... 56

10. Percent occurrence of all fish species found in loggerhead samples from Virginia during 1983 to 2002 (n = 166) ...... 61

11. Composition of Kemp’s ridley diet data and samples collected during 1987 to 2002...... 81

12. Frequency of occurrence of all prey items found in Kemp's ridley samples from Virginia during 1987 to 2002 (n = 33)...... 95 Table Page

13. Percent occurrence, percent dry weight, percent number, and percent index of relative importance of all prey items found in whole Kemp’s ridley gut samples from Virginia during 1991 to 1994 (n = 5)...... 97

14. Percent occurrence, percent dry weight, percent number, and percent index of relative importance of all prey items found in whole Kemp’s ridley gut samples from Virginia during 2000 to 2002 (n = 18)...... 98

vii LIST OF FIGURES

Figure Page

1. Annual reported horseshoe crab (Limulus polyphemus) landings for (a) all of Virginia and (b) Chesapeake Bay, Virginia from 1980 to 2002...... 8

2. Annual reported whelk landings for Virginia from 1992 to 2001...... 9

3. Annual reported blue crab (Callinectes sapidus) landings for Virginia from 1980 to 2002...... 11

4. Approximate locations of all loggerheads (n = 179) and Kemp’s ridleys (n = 33) from Virginia and Maryland from which samples were collected during 1980 to 2002...... 23

5. Approximate locations of all loggerheads from Virginia and Maryland from which samples were collected during 1980 to 1989 (n = 60)...... 24

6. Approximate locations of all loggerheads from Virginia and Maryland from which samples were collected during 1990 to 1997 (n = 70)...... 25

7. Approximate locations of all loggerheads from Virginia and Maryland from which samples were collected during 2000 to 2002 (n = 49)...... 26

8. Approximate locations of all loggerheads and Kemp’s ridleys from North Carolina from which samples were collected during 1990 to 1992 (n = 8)...... 28

9. Virginia sea turtle stranding regions...... 29

10. Annual distribution of (a) loggerhead samples (n = 185) and (b) loggerhead diet data by state (n = 327, excluding 24 empties) ...... 31

11. Annual distribution of loggerhead samples and data by data type (n = 351, 185 samples, 142 data only, 24 empty)...... 32

12. Distribution of loggerhead samples by month and state (n = 185) ...... 33

13. Size distribution of (a) Virginia loggerhead whole samples (n = 128), (b) all Virginia loggerhead samples (n = 169), and (c) all Virginia loggerheads with diet data (n = 303)...... 34

14. Cumulative prey curve for whole loggerhead samples collected in Virginia during 1983 to 2002 (n = 128)...... 36 Figure Page

15. Linear regressions of loggerhead straight carapace length versus (a) total dry weight of Virginia whole samples and (b) diy weights of large prey items in Virginia whole samples collected during 1983 to 2002 (n = 128)...... 38

16. Linear regressions of loggerhead straight carapace length versus (a) total number of prey items in Virginia whole samples and (b) total number of large prey items in Virginia whole samples (n = 128)...... 39

17. Cumulative prey curves for whole loggerhead samples collected in Virginia during (a) 1983 to 1987 (n = 13), (b) 1988 to 1992 (n = 49), (c) 1993 to 1997 (n = 20), and (d) 1998 to 2002 (n = 46)...... 40

18. Frequency of occurrence of (a) major prey groups in all loggerhead diet data (n = 297) and (b) major prey items in all loggerhead samples (n = 166) collected in Virginia during 1983 to 2002, grouped by five-year intervals...... 42

19. Biplots of year division and prey type principal components for PCI and PC2 of a correspondence analysis using (a) %F values of all four general diet categories and (b) %F values of three general diet categories from all Virginia loggerhead diet data collected during 1983 to 2002 (n = 297)...... 43

20. (a) Frequency of occurrence, (b) percent dry weight, (c) percent number, and (d) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia from 1983 to 2002 (n = 128), grouped by five-year intervals...... 54

21. Biplot of year division and prey type principal components for PCI and PC2 of a correspondence analysis using %F values for major prey items from all loggerhead samples collected in Virginia during 1983 to 2002 (n = 166)...... 57

22. Biplots of year division and prey type principal components for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 128)...... 58

23. Cumulative prey curves for whole samples from Virginia loggerheads with straight carapace lengths of (a) 40.0 to 49.9 cm (n = 13), (b) 50.0 to 59.9 cm (n = 44), (c) 60.0 to 69.9 cm (n = 48), (d) 70.0 to 79.9 cm (n = 9), (e) 80.0 to 89.9 cm (n = 8), and (f) 90.0 to 99.9 cm (n = 6)...... 62

24. (a) Frequency of occurrence and (b) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia during 1983 to 2002 (n = 128), grouped by ten-centimeter size classes (SCL)...... 65

ix Figure Page

25. Biplots of size class and prey type principal components for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 128)...... 66

26. Cumulative prey curves for whole loggerhead samples from Virginia taken from (a) females (n = 73), (b) males (n = 26), and (c) unknowns (n = 29) ...... 69

27. (a) Frequency of occurrence and (b) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia during 1983 to 2002 (n = 128), grouped by sex of turtle...... 70

28. Biplots of turtle sex and prey type principal components for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 128)...... 71

29. Cumulative prey curves for whole loggerhead samples from Virginia during (a) Spring (March to May, n = 47), (b) Summer (June to August, n = 64), and (c) Fall (September to November, n = 14) ...... 74

30. (a) Frequency of occurrence and (b) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia during 1983 to 2002 (n = 128), grouped by season...... 75

31. Biplots of season and prey type principal components for PC 1 and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 125, spring, summer, and fall only)...... 76

32. Approximate locations of all Kemp’s ridleys from Virginia from which samples were collected during 1987 to 1989 (n = 6)...... 78

33. Approximate locations of all Kemp’s ridleys from Virginia from which samples were collected during 1990 to 1994 (n = 9)...... 79

34. Approximate locations of all Kemp’s ridleys from Virginia from which samples were collected during 2000 to 2002 (n = 18)...... 80

35. Annual distribution of (a) Kemp’s ridley samples (n = 35) and (b) Kemp’s ridley diet data by state (n = 62, excluding 4 empties) ...... 82

x Figure Page

36. Annual distribution of Kemp’s ridley samples and data by data type (n = 66, 35 samples, 27 data only, 4 empty)...... 84

37. Distribution of Kemp’s ridley samples by month and state (n = 35) ...... 85

38. Size distribution of (a) Virginia Kemp’s ridley whole samples (n = 23), (b) all Virginia Kemp’s ridley samples (n = 33), and (c) all Virginia Kemp’s ridleys with diet data (n = 59)...... 86

39. Cumulative prey curve for whole Kemp’s ridley samples collected in Virginia during 1991 to 1994 and 2000 to 2002 (n = 23)...... 88

40. Linear regressions of Kemp’s ridley straight carapace length versus (a) total dry weight of Virginia whole samples and (b) dry weights of large prey items in Virginia whole samples (n = 23)...... 89

41. Linear regressions of Kemp’s ridley straight carapace length versus (a) total number of prey items in Virginia whole samples and (b) total number of large prey items in Virginia whole samples (n = 23)...... 90

42. Cumulative prey curves for whole Kemp’s ridley samples collected in Virginia during (a) 1991 to 1994 2002 (n = 5) and (b) 2000 to 2002 (n = 18)...... 91

43. Frequency of occurrence of (a) major prey groups in all Kemp’s ridley diet data (n = 59) and (b) major prey items in all Kemp’s ridley samples (n = 33) collected in Virginia during 1983 to 2002...... 92

44. Biplots of year division and prey type principal components for PCI and PC2 of a correspondence analysis using (a) %F values of general diet categories for all Kemp’s ridley data (n = 59) and (b) %F values of major prey items from all Kemp’s ridley samples (n = 33) collected in Virginia during 1983 to 2002...... 94

45. (a) Frequency of occurrence, (b) percent dry weight, (c) percent number, and (d) percent index of relative importance (IRI) in whole Kemp’s ridley gut samples collected in Virginia from 1991 to 1994 and 2000 to 2002 (n = 23)...... 100

xi ABSTRACT

The Chesapeake Bay and coastal waters of Virginia, U.S.A. serve as foraging grounds for loggerhead (Caretta caretta) and Kemp’s ridley (.Lepidochelys kempi) sea turtles from approximately May to October each year. Both loggerheads and Kemp’s ridleys are known to feed primarily on benthic invertebrates as juveniles and adults, but specific prey preferences vary between geographic regions. The Virginia Institute of Marine Science Sea Turtle Program has collected diet data and gut samples from stranded and incidentally caught sea turtles in Virginia since 1979. Examination of turtles that stranded in Virginia during the late 1970s and early 1980s indicated that loggerheads fed primarily on Atlantic horseshoe crab (Limulus polyphemus) and Kemp’s ridleys primarily on blue crab ( Callinectes sapidus).

During 1980 to 1994, 1997, and 2000 to 2002, 128 whole digestive tract samples and 41 partial gut samples were collected from loggerheads in Virginia. Diet information was noted on stranding datasheets for an additional 134 loggerheads from 1980 to 2002. Twenty-three whole samples and 10 partial samples were collected in Virginia from Kemp’s ridleys during 1987 to 1994 and 2000 to 2002, and data were available on an additional 26 ridleys from 1983 to 2002. Prey items in the samples were identified to the lowest possible taxonomic level, and dry weights and prey item counts were recorded. Results indicate a shift in loggerhead diet from predominantly horseshoe crab during the early to mid- 1980s to predominantly blue crab during the late 1980s and early 1990s. Loggerhead diet in the mid- 1990s and 2000 to 2002 was dominated by fmfish, particularly menhaden (Brevoortia tyrannus) and croaker ( Micropogonias undulatus). These diet shifts suggest that fishery-related declines in horseshoe crab and blue crab populations have caused loggerheads to instead forage on fish caught in nets or on discarded bycatch. A slight seasonal effect on diet was also detected, and the diet of juvenile loggerheads differed somewhat from that of the adults. The small Kemp’s ridley dataset suggests that blue crabs and spider crabs (Libinia spp.) were important components of ridley diet in Virginia during 1987 to 2002. H istorical Diet analysis of loggerhead (Ca r e t t a c a r e t t a) a n d Kemp’S R idley (L epidochelyskempi) Sea T urtles in Virginia INTRODUCTION

Six species of sea turtles are found in the waters of the United States, and five of these are found along the Atlantic and Gulf coasts. The loggerhead, Caretta caretta, and the Kemp’s ridley, Lepidochelys kempi, are the most abundant species within the United

States and in Virginia (Keinath et al. 1987; USFWS 1998). The loggerhead sea turtle is classified as “Threatened” under the U.S. Endangered Species Act of 1973, while the

Kemp’s ridley is listed as “Endangered” and is the least abundant sea turtle species in the world (USFWS 1998).

Loggerheads are found circumglobally in tropical, subtropical, and temperate waters, and they nest in temperate zones and in the subtropics. At least four genetically distinct nesting populations are recognized and managed separately in the Northwestern

Atlantic (TEWG 2000). Two of the world’s largest loggerhead rookeries are found on the Atlantic beaches of central and southern Florida, and other nesting grounds occur along the U.S. coast from Virginia to Louisiana (NRC 1990). A petition was filed in

2002 to upgrade the Northern and Florida Panhandle subpopulations to “Endangered”

(NOAA 2002), and it is currently under evaluation by the National Marine Fisheries

Service. Mitochondrial DNA analysis indicates that about 36% of Virginia’s juvenile loggerheads originate from nesting beaches in North Carolina, South Carolina, and

Georgia, and approximately 69% of Virginia’s loggerheads originate from the Northern subpopulation as a whole (Virginia to Northeast Florida) (Norrgard and Graves 1996).

2 3

Kemp’s ridleys are restricted primarily to the Gulf of Mexico and Northwestern

Atlantic Ocean. Although they forage in large numbers in U.S. waters, the only important nesting site for the Kemp’s ridley is found at and around the beaches of

Rancho Nuevo in Tamaulipas, Mexico (TEWG 2000). Juvenile ridleys encountered in

Virginia and along the rest of the U.S. east coast are able to return to these nesting sites as adults (TEWG, 2000; VIMS mark-recapture data).

The Chesapeake Bay, Virginia serves as an important seasonal foraging ground for an estimated 5,000 to 10,000 sea turtles each year (Bellmund et al. 1987; Musick

1988; Keinath et al. 1987). These turtles are predominantly benthic-stage juvenile loggerheads and Kemp’s ridleys, and the numbers of loggerheads that enter the Bay are estimated at roughly ten times those of the Kemp’s ridley (Musick 1988; Keinath et al.

1994). The Atlantic green turtle (Chelonia my das) and Atlantic leatherback

(Dermochelys coriacea) also enter Virginia waters, but in substantially smaller numbers, and there have only been two records of hawksbill sea turtles (Eretmochelys imbricata) in

Virginia since 1979 (Virginia stranding data). Sea turtles begin to enter Virginia’s coastal waters and the Chesapeake Bay when water temperatures reach approximately 18 to 20° Celsius (Byles 1988; Coles 1999; Musick 1988), generally in mid- to late-May, and they begin to migrate south in the fall once temperatures decrease (Coles 1999;

Mansfield et al. 2001).

The Virginia Institute of Marine Science (VIMS) Sea Turtle Program has served as the Virginia center for sea turtle standings, monitoring, and research since its inception in 1979. Standings are responded to by VIMS, the Virginia Marine Science

Museum (VMSM), and various state cooperators. Virginia’s standings typically number 4 between 200 and 400 annually (Mansfield et al. 2002a), and most strandings occur between mid-May and mid-October. A large stranding peak is usually seen in the late spring or early summer as the turtles complete migrations to the Chesapeake Bay

(Mansfield et al. 2001). Data recorded on each stranded turtle include stranding location, turtle measurements, decomposition state, noticeable abnormalities and wounds, and gut contents. Tissue, bone, epibiota, and gut content samples are also collected, depending on decomposition state of the turtle and research interests and requests.

Gut content analysis can aid in conservation and management decisions by providing insight to the role of an organism in its environment. Diet composition studies may also be employed to determine competition and habitat segregation between and within sea turtle species (Burke et al. 1993; Keinath et al. 1987; Shaver 1991). The presence of predominantly pelagic or benthic prey items can be used to determine the life stages of stranded turtles (Van Nierop and Den Hartog 1984; Plotkin 1989; Plotkin 1996).

Immediately after hatching, both loggerhead and Kemp’s ridley sea turtles enter a multi-year pelagic stage, during which the young turtles associate with Sargassum spp. and feed on pelagic molluscs, crustaceans, and coelenterates, as well terrestrial insects and anthropogenic debris (Bjomdal 1997; Bolten and Balazs 1995; Carr 1986; Musick and Limpus 1997; Richardson 1991). More is known about the diet habits of these two species after the pelagic stage, once the turtles have returned to coastal waters. Both loggerhead and Kemp’s ridley sea turtles are known to be benthic carnivores as juveniles and adults (Bjomdal 1985; Hendrickson 1980; Mortimer 1979), but their diets vary by region. 5

The diets of the small juvenile loggerheads and Kemp’s ridleys that seasonally forage off Long Island, New York have been evaluated using whole digestive tracts from stranded animals and fecal samples from turtles caught incidentally in poundnets (Burke et al. 1993; Burke et al. 1994; Morreale and Standora 1991). Crustaceans accounted for the vast majority of prey items, and two slow-moving crabs, the nine-spined spider crab

(Libinia emarginata) and Atlantic rock crab ( Cancer irroratus), were encountered most frequently in samples from both species.

Reports from Georgia (Creech and Allman 1997; Frick 1997; Frick and Mason

1998) indicate that stranded Kemp’s ridleys from northern Georgia had consumed

“shallow water inhabitants” including blue crabs ( Callinectes sapidus), mud snails

(Nassarius spp.), stone crabs (Menippe mercenaria), spider crabs {Libinia spp.), benthic finfish, and Carolina diamondback terrapin {Malaclemys terrapin centrata) during the mid-1990s. Stranded loggerheads collected during the mid- to late- 1990s on Wassaw

Island and Cumberland Island, Georgia have contained predominantly spider and stone crabs, as well as whelks {Busycon spp.), moon snails{Neverita spp.), and horseshoe crabs

{Limulus polyphemus) (Frick et al. 2001). In a long-term (1979 to 1999) diet study using whole digestive tracts from loggerheads that stranded on Cumberland Island (n = 369),

Youngkin and Wyneken (in press) noted four diet shifts between crab and mollusc dominance. Fish were found in approximately 50% of all samples, while horseshoe crabs were only found in about 3%. Amounts of crab and fish varied significantly with curved carapace length (CCL), with larger turtles eating proportionately more crabs and smaller turtles eating proportionately more fish. 6

In Southern Texas, the diets of stranded loggerheads and Kemp’s ridleys have been well documented. Plotkin et al. (1993) noted seventeen different prey groups in samples from 82 stranded benthic-stage juvenile loggerheads. Sea pens, including

Virgualaria presbytes, were found in 56.1% of the turtles and accounted for 58.7% of the total dry weight. Crabs occurred in 87.6% of the gut samples, but accounted for only

28.8% of the total dry weight. Walking crabs such as spider crabs, calico crabs ( Hepatus epheliticus), and purse crabs (Persephona mediterranea) seemed to be more important than portunid crabs in the diet. Plotkin (1989, 1996) also identified some of the smaller loggerheads stranding in Texas as pelagic-stage animals based on gut contents such as

Sargassum spp. and pelagic crustaceans and molluscs.

Shaver (1991) examined the gut contents of 50 wild and 51 head-started Kemp’s ridleys, predominantly subadults (benthic-stage juveniles) that stranded in Southern

Texas. (Head-starting is a technique in which hatchlings are raised in captivity for at least several months to reduce hatchling mortality (NRC 1990).) Crabs, including blue, spider, and purse crabs, were most frequently observed and accounted for the greatest proportion of the dry weight in both wild and head-started turtles. However, head-started subadults consumed greater amounts of molluscs and fewer species of crabs than wild subadults. The head-started turtles also consumed larger quantities of fish and shrimp, probably in the form of bycatch.

In the first known publication referring to sea turtle diet in Virginia, Hardy (1962) reported that crabs from the genus Callinectes accounted for approximately 95% of the gut contents of a Kemp’s ridley found in Northumberland County. Later reports

(Bellmund et al. 1987; Lutcavage and Musick 1985; Musick et al. 1984) concluded that 7 loggerheads collected from 1979 to 1984 had consumed primarily horseshoe crabs, and blue crabs, spider crabs, rock crabs, clam bodies, and fish were found in some digestive tracts. During this same time period, Kemp’s ridleys examined had consumed mainly blue crabs (Bellmund et al. 1987; Lutcavage and Musick 1985; Musick et al. 1984).

These prey preferences appear to correspond to the partitioning of habitat seen in biotelemetric studies, with loggerheads feeding on horseshoe crabs in river mouths and channels and Kemp’s ridleys foraging for blue crabs in shallower waters, including seagrass beds (Byles 1988; Keinath et al. 1987; Lutcavage 1981; Musick et al. 1984).

Horseshoe crabs were once used in the U.S. for fertilizer and livestock feed, but the commercial fishery has primarily caught horseshoe crabs for bait and biomedical uses since the 1970s (Berkson and Shuster 1999; Botton and Ropes 1987). During the 1980s,

Virginia’s reported horseshoe crab landings averaged about 154,000 pounds per year, and landings within the Virginia portion of the Chesapeake Bay peaked at about 80,000 pounds in 1986 (VMRC data 2003; Figure 1). A ban on all trawling in Virginia waters was instated in 1989 (VMRC 1995), and Virginia horseshoe crab landings averaged only about 21,000 pounds per year during 1990 to 1995 (VMRC data 2003; Figure la).

However, the Virginia whelk pot fishery took off from its inception in the early 1990s

(VMRC data 2003; Figure 2), and horseshoe crab became the bait of choice (Fisher 2000;

Mills 2000). Virginia’s commercial horseshoe crab landings, particularly from the ocean, increased considerably during 1998 to 2000, peaking at nearly 1.9 million pounds in 1999

(Figure la). FIGURE 1. Annual reported horseshoe crab (Limulus polyphemus) landings for (a) all of Virginia and (b) Chesapeake Bay, Virginia from 1980 to 2002. Data courtesy of Virginia Marine Resources Commission. Note: 2002 data is preliminary.

(a) 2.0

€ 1-5 c o3 OL o 1.0 (A C o

0.5 n

0.0 l.lllayi... ______..1 i U9 | i a | o CM ^ CO 0 0 O CM Mf CO 00 O CM 00 0 0 CO CO 00 05 05 05 0 5 0 5 O O 0 5 0 5 0 5 0 5 05 05 05 05 0 5 0 5 O O CM CM

I Ocean (VA) □ Chesapeake Bay (VA) □ Bay Tributaries (VA)

(b)

100,000

80,000

« 60,000 c 3 0- 40,000

20,000 ______n n 0 I I I I I I I I I IJ1 I I nJuuiI I 111 OCMM-COOOOCM^fCOOOOCM 00 00 00 00 00 05 0 5 0 5 0 5 0 5 0 0 050505050505 0505050500 T - T - T - T - T - T - T- T- 1- r- CM CM

□ Chesapeake Bay (VA) 19 Bay Tributaries (VA) 9

FIGURE 2. Annual reported whelk landings for Virginia from 1992 to 2001. Data courtesy of Virginia Marine Resources Commission. Note: mandatory reporting of commercial whelk landings was enstated in 1993.

500,000

400,000

300,000 ■o

200,000 I

100,000

CM CO •'J- in CO t'- 05 O 05 05 05 05 05 05 05 O O 05 05 0 5 05 05 05 05 o O t- t- t- t- t- t- t-CMCM

□ Channel Whelk £S Knobbed Whelk 10

Fishing pressure is believed to have reduced the observable horseshoe crab

population in the Northwestern Atlantic by 50% from 1990 to 1998, and drastic declines

have been seen in Delaware Bay (ASMFC 1998; Fisher 2000; Tanacredi 2001). In light

of the implementation of the Atlantic States Marine Fisheries Commission (ASMFC)

Horseshoe Crab Fisheiy Management Plan (1998) and apparent declines in Virginia

waters (Mills 2000), the Virginia Marine Resources Commission (VMRC) enacted a

number of regulations on the fishery starting in 2001. These included setting the state

horseshoe crab landing quota at 152,495 animals (approximately 400,000 pounds using

an average weight of 2.6 pounds) (ASMFC 1998; VMRC 2000). Just over 145,000

pounds were landed in 2001 (Figure la).

The level of fishing for the blue crab in the Chesapeake Bay has been considered

“intensive” since the early 1900s (Van Engel 1958). Virginia’s commercial landings

seldom exceeded 20 million pounds per year in the 1930s and 1940s, but they exceeded

32 million pounds per year in the late 1940s, and annual landings regularly exceeded 40

million pounds during 1965 to 1990 (Kirkley 1997). Average annual blue crab landings

dropped from about 41 million pounds during 1990 to 1995 to about 35 million pounds

during 1996 to 2001 (VMRC data 2003; Figure 3). Recent findings (Lipcius and

Stockhausen 2002) indicate that the Lower Chesapeake Bay blue crab spawning stock,

larval abundance, and postlarval recruitment were significantly lower in 1992 to 1999 than in 1985 to 1991. The decrease in all of these variables was rapid, occurring over one to two years, indicating a phase shift rather than a progressive decrease. Poor recruitment

in 1991 in concert with high fishing and natural mortality are the proposed cause of the diminished spawning stock in 1992. The blue crab stock, larval abundance, and 11

FIGURE 3. Annual reported blue crab (Callinectes sapidus) landings for Virginia from 1980 to 2002. Data Courtesy of Virginia Marine Resources Commission. Note: 2002 data is preliminary.

60

E 40 E?EQ(?(Q(WE

o 30

= 20

■ Ocean (VA) □ Chesapeake Bay (VA) 11 Bay Tributaries (VA) 12 recruitment are not likely to rebound without significant reductions in fishing and natural mortality, along with conditions conducive to successful recruitment (Lipcius and

Stockhausen 2002).

Further characterizing and quantifying the diet of loggerheads and Kemp’s ridleys should provide insight to indirect and direct effects of Virginia’s fisheries on sea turtles.

The apparent declines in horseshoe crab and blue crab populations may have implications for the diets of these turtle species, and the turtles could be forced to shift their diets to other prey. In addition, it is believed that turtles are typically “not agile enough to capture fish under natural conditions” (Bellmund, et al. 1987), and thus would only consume large quantities of finfish by interacting with fishing gear (Bellmund et al. 1987) or bycatch (Shoop and Ruckdeschel 1982). The presence of scavenging mud snails, such as Nassarius spp. andIlyanassa spp., along with fish in the gut may suggest that a turtle has been feeding on dead, discarded bycatch (Plotkin et al. 1993; Shaver 1991).

Loggerhead and Kemp’s ridley gut content samples were collected by VIMS and subsequently preserved and archived during 1980 to 1997. Six loggerhead samples from

1980 were quantified by Lutcavage (1981), but most of the samples collected afterwards were examined in a cursory manner and saved for later analysis. The quantification of the archived samples, with the addition of samples collected from 2000 through 2002, represented one of the first long-term gut content analyses for both juvenile loggerheads and Kemp’s ridleys. In addition to expanding on the limited knowledge of the diet of these two species, this undertaking allowed for the examination of loggerhead and

Kemp’s ridley diet in Virginia over time as compared to temporal changes in prey abundances. The objectives of this study were threefold: 13

1. To describe the diets of loggerhead and Kemp’s ridley sea turtles in Virginia

and help assess the role of sea turtles in the Chesapeake Bay food web.

2. To compare Virginia turtle diet over the time period of 1980 to 2002 and

between size classes, the sexes, seasons, and the two species.

3. To assess potential impacts of horseshoe crab and blue crab population

declines on loggerhead and Kemp’s ridley diets. METHODS

Sample Collection

Opportunities for collecting sea turtle gut samples are somewhat limited because these turtles are protected species. This contrasts to many studies conducted on finfish, which can employ a variety of fishing gear to catch large numbers of fish per site

(Creaser and Perkins 1994; Hacunda 1981; Hynes 1950; Moore and Moore 1974). Shark diet studies often have smaller sample sizes than finfish studies, and specimens are often collected by hook and line (Stillwell and Kohler 1982) and longline sets (Gelsleichter et al. 1999). Sample sizes tend to be even smaller in sea turtle diet studies, and gut contents are often collected over a span of several years from strandings (Bellmund et al. 1987;

Burke et al. 1994; Plotkin et al. 1993; Shaver 1991) and as fecal samples (Burke et al.

1993). Gastric (esophageal) lavage can be performed on live turtles, but it is more feasible for green turtles (Forbes 1999) and hawksbills (Mayor et al. 1998), which eat primarily plant matter and sponges respectively.

The VIMS Sea Turtle Stranding Program and various stranding cooperatives collected partial and whole digestive tract samples and fecal samples during 1980, 1983 to 1994, and 1997. These samples were stored in jars at VIMS in either 10% formalin or

70% ethanol. Those in 10% formalin were switched to 70% ethanol prior to data collection. The portion of the digestive tract sampled was somewhat inconsistent prior to

1988, but most samples from later years were obtained from whole digestive tracts.

14 15

Additional samples, predominantly from whole digestive tracts, were collected from June

2000 to October 2002. The collection methods, sieve size, and preservation methods used for the 2000 through 2002 samples were based on the techniques used to collect and preserve the earlier samples (D. E. Barnard, pers. comm.; J. A. Keinath, pers. comm.; R.

A. Pemberton, pers. comm.), so that the samples would be comparable.

To obtain a whole gut content sample, the entire digestive tract, from the beginning of the esophagus to the end of the large intestine, was dissected from a dead stranded or incidentally caught loggerhead or Kemp’s ridley. These turtles were primarily fresh dead (NMFS condition “1”, VIMS condition “3”; Tables 1 and 2) and moderately decomposed (NMFS “2”, VIMS “3” and “4”), as well as some at the low end of severely decomposed (NMFS “3”, VIMS “4”). After its removal, each digestive tract was frozen, placed in 10% formalin, or immediately sieved. Each digestive tract was cut open (after thawing or rinsing where appropriate), and the contents were emptied into a

500 jam brass sieve and rinsed thoroughly with water. The gut contents were placed in a jar filled with 10% freshwater formalin and kept in formalin for at least 24 to 48 hours.

The formalin was then poured off, and each sample was allowed to soak in freshwater for several hours before filling the jar with 70% ethanol. Stomach samples and other partial gut samples were collected from some stranded animals, and fecal samples were collected from some incidentally caught and live stranded turtles. These samples were collected predominantly in the 1980s, and preservation procedures matched those used for whole samples. Lavage was not attempted on live animals, due to low success of the procedure on loggerheads (R. H. George, pers. comm.), the large and sharp nature of prey items, the temperament of both species, and the size of most loggerheads in Virginia. TABLE 1. National Marine Fisheries Service condition codes for stranded sea turtles.

Condition code Description 0 Alive 1 Fresh dead 2 Moderately decomposed 3 Severely decomposed 4 Dried carcass 5 Skeleton, bones only

TABLE 2. Virginia Institute of Marine Science condition codes for stranded sea turtles.

Condition code Description 1 Live, healthy 2 Live, sick 3 Dead, slight bloat 4 Dead, bloated, gray/green skin 5 Dead, bones showing, decomposed 6 Dead, bones only,scattered 17

The preservation procedure of 10% formalin followed by 75% ethanol is often

used to prepare scientific specimens for museums (Hangay and Dingley 1985), and 10%

formalin followed by 70% ethanol is currently standard for the VIMS Fish Collection (M.

Harbin, pers. comm.). Formalin is effective in hardening stomach contents (Creaser and

Perkins 1994) and preventing tissue decay (DiStefano et al. 1994); however, freshwater

formalin is also known to increase fish wet weights and decrease fish lengths (DiStefano

et al. 1994; Parker 1963). DiStefano et al. (1994) concluded that 10% formalin had no

significant effects on weights and measurements of a freshwater crustacean, the virile

crayfish (Orconectes virilis). Despite its potential effects on wet weights, the use of

formalin is considered acceptable if the methods are consistent throughout a diet study

(Hyslop 1980). Ethanol is also effective at preventing tissue decay in fishes and

crustaceans, and it lacks the potential serious risks to human health associated with

formalin (DiStefano et al. 1994).

Data Collection

Each gut sample was sorted into major groups such as crustaceans, fish, horseshoe

crabs, molluscs, and plants, and prey items were identified to the lowest possible

taxonomic level. All samples were kept after sorting to provide a reference collection of

prey items, and whole fish, scales, and skeletal preparations from the VIMS Fish

Collection were also used for reference. Numbers of prey items were determined when possible, and wet weights were obtained to the nearest tenth of a gram. The samples

were dried at approximately 60° Celsius in a drying oven for 24 to 48 hours (Burke et al.

1994; Forbes 1999; Plotkin et al. 1993; Shaver 1991) and dry weights were taken to the 18 nearest tenth of a gram after cooling. Numerical counts were collected because they provide information about feeding behavior, and weights because they can reflect dietary nutritional values (Macdonald and Green 1983); however, the large amounts of indigestible crab, fish, and mollusc parts present in whole sea turtle digestive tracts and fecal samples may result in some bias (Burke et al. 1993; Plotkin et al. 1993).

All gut content data were linked to the original stranding data in a Microsoft

Access database. Additional diet information on turtles that were necropsied but not sampled was compiled from the VIMS stranding database. All data entry was verified prior to analysis.

Data Analysis

Diet descriptions

Cumulative prey curves were used to evaluate whether a sufficient number of whole samples had been examined to adequately describe the diet of each species and various temporal, size class, and sex subsets (Cortes 1997; Gelsleichter et al. 1999; Ferry and Calliet 1996). These curves were constructed by randomizing the order in which samples were analyzed ten times using Resampling Stats, Inc. Excel Add-in Version 2.0

(www.resample.com) and then plotting the mean cumulative number of prey types versus the number of samples examined (Gartland 2002; Gelsleichter et al. 1999). The minimum number of gut samples that are needed to obtain “precise” and “more reliable” results occurs once the curve has reached its asymptote (Cortes 1997).

Percent occurrences (frequency, %F) of general prey groups (horseshoe crab, crustaceans, molluscs, and fish) were determined for all diet data from Virginia, and 19

percent occurrences of specific prey items were determined for all Virginia samples and

whole samples. Percent dry weight (%W) and percent number (%N) were also calculated

for prey items in the whole samples, since multiple measures are often necessary when

prey items differ in size (Ferry and Calliet 1996). Wet weights were not analyzed

because of the large discrepancy between total sorting time for individual samples

(ranging from several minutes to several hours).

Compound indices that incorporate occurrence, bulk, and numbers appear to

provide a more accurate description of dietary importance, and they can facilitate

comparative studies (Cortes 1997). The index of relative importance (IRI) (Pinkas et al.

1971), a compound diet index used frequently in the fish literature, was calculated for

general prey groups and individual prey items in the whole loggerhead and Kemp’s ridley

gut samples from Virginia. The original index combined numeric, occurrence, and

volumetric data into one value (Pinkas et al. 1971), but the modification introduced by

Hacunda (1981) that incorporated weight instead of volume was used in this study:

IRI/ = (%N/ + %W/) * %F/

where IRI/ is the index of relative importance value for prey type /. To promote

consistency and facilitate comparisons, IRI was also calculated for the whole samples on

a percentage basis (Cortes 1997):

%IRI/ = 100 * IRI/ / X IRI/ where X IRI/ is the sum of the IRI values all prey categories.

Linear regressions were performed using Minitab Version 12.1 to determine if there was any relationship between turtle size (straight carapace length) and dry weights and numbers of prey items in whole samples from all Virginia loggerheads and Kemp’s 20 ridleys. Minitab performs an analysis of variance (ANOVA) in conjunction with each linear regression to determine if the null hypothesis (Ho) that the slope generated by the regression is zero can be rejected. If this null hypothesis is rejected, the regression is considered significant (Moore and McCabe 1998).

Diet comparisons

To increase sample sizes, the Virginia loggerhead data were grouped into five- year increments from 1983 to 2002. The data were also examined in ten-centimeter size classes from 40 to 100 cm, by sex of turtle, and by season. Due to the limited amount of

Kemp’s ridley data, the year divisions of 1983-1984, 1987-1989, 1991-1994, 1999-2000, and 2001-2002 were examined for the general diet information. Whole Kemp’s ridley samples were only collected from 1991 to 1994 and 2000 to 2002, so these divisions were used in whole sample analysis.

Correspondence analysis (CA) is a multivariate ordination technique similar to principal component analysis (PCA), and it is considered appropriate for count data presented in contingency tables (Davis 1989; Gartland 2002). The CA technique was used to identify variations in Virginia loggerhead diet between the five-year groupings, size classes, sexes, and seasons using %F, %W, %N, and %IRI values. Correspondence analysis was also used to examine %F for all Kemp’s ridley data and samples during the various year divisions. Diet variations between stranding regions were not examined because dead turtles may float with winds, currents, and tides for days to weeks prior to stranding on land (Mansfield et al. 2002a, 2002b). 21

The %F, %W, %N, and %IRI values for large, frequently occurring prey items

(horseshoe crab, all crab species, bony fish, Busycon whelks {Busycon spp.), and Atlantic moon snail (Neverita duplicata)) were entered into worksheets in Minitab Version 12.1 to run correspondence analyses. Each turtle grouping (years, sizes, sexes, seasons) was examined separately for each data type (%F, %W, %N, %IRI) due to the small number of whole samples in each of the divisions. All values were rounded to the nearest whole integer, as required by Minitab to run a CA. The CA technique operates on a matrix derived from a contingency table, and row (years, sizes, sexes, or seasons) and column

(prey type) principal components (PCs) are calculated using eigenanalysis (Davis 1986).

Biplots of the first and second PCs for both rows and columns were constructed for each turtle grouping by data type to identify potential patterns in diet.

Diet overlap between the species and various subsets of loggerheads was computed using the Schoener (1970) index:

a = 1 - 0.5(E |pxi- p yi\) where pXi is the proportion of food categoryi in the diet of species x, andpyi is the proportion of food categoryi in the diet of speciesy (Schoener 1970; Wallace 1981).

This index ranges from zero to one (no diet overlap to complete overlap) and can be multiplied by 100 to yield percent overlap. A comparison of four diet overlap indices by

Wallace (1981) concluded the Schoener index is one of the “least objectionable” indices available when resource availability data are absent. Additionally, Wallace concluded that the average of individual volume percentages was a less objectionable measure than percent occurrence in all guts, percent of total number of prey items in all guts, and percent of total volume in all guts for evaluating diet overlap. RESULTS

Loggerheads

Eighty-two whole digestive tract samples, 38 partial gut samples, and two fecal samples were collected from loggerheads in Virginia during 1980, 1983 to 1994, and

1997 and archived at VIMS (Figures 4-6). An additional 46 whole samples and one partial sample were collected during 2000 to 2002 (Figure 7). The VIMS Sea Turtle

Stranding Database yielded general diet information on another 134 loggerheads from

Virginia (Table 3). A few samples were also collected from Maryland (Figures 4-7) and the Outer Banks of North Carolina (Figure 8). Six fecal samples and two whole samples were archived from Maryland during 1984 to 1994, and two fecal samples were collected in 2002. Six whole samples were collected from North Carolina from 1990 to 1992.

Additional diet information was available on seven loggerheads from Maryland and one from North Carolina (Table 3). Twenty-three empty digestive tracts were noted in

Virginia from 1980 to 2002 and one was noted for North Carolina in 1994.

The loggerhead samples (n = 185) were distributed geographically as follows

(number of whole samples in parentheses): Western Chesapeake Bay - 111 (91),

Southern Chesapeake Bay - 16 (12), Eastern Shore Bayside - 17 (10), Eastern Shore

Oceanside - 2 (1), Virginia Beach Oceanside - 23 (14), Maryland - 10 (2), and North

Carolina - 6 (6) (Figures 4-9, Table 3). Available loggerhead data (n = 327) were

22 23

FIGURE 4. Approximate locations of all loggerheads (n = 179) and Kemp’s ridleys (n = 33) from Virginia and Maryland from which samples were collected during 1980 to 2002.

Loggerhead * Kemp's ridley 24

FIGURE 5. Approximate locations of all loggerheads from Virginia and Maryland from which samples were collected during 1980 to 1989 (n = 60). 25

FIGURE 6. Approximate locations of all loggerheads from Virginia and Maryland from which samples were collected during 1990 to 1997 (n = 70). 26

FIGURE 7. Approximate locations of all loggerheads from Virginia and Maryland from which samples were collected during 2000 to 2002 (n = 49). TABLE 3. Composition of loggerhead diet data and samples collected during 1980 to 2002 (excludes 23 empty tracts from Virginia and one from North Carolina).

Turtles with Samples Diet Data (Whole, Partial, Whole (Includes and Fecal Digestive Tract Samples) Samples) Samples Western Bay, Virginia 179 111 91 Southern Bay, Virginia 40 16 12 Eastern Shore Bayside, Virginia 38 17 10 Eastern Shore Oceanside, Virginia 9 2 1 Virginia Beach Oceanside, Virginia 37 23 14 Maryland 17 10 2 North Carolina 7 6 6 Total 327 185 136 28

FIGURE 8. Approximate locations of all loggerheads (n = 6: 2 strandings, 4 trawler takes) and Kemp’s ridleys (n = 2: 1 stranding, 1 trawler take) from North Carolina from which samples were collected during 1990 to 1992.

Loggerhead ALBEMARLE SOUND Kemp's ridley

ATLANTIC OCEAN

PAMLICO SOUND 29

FIGURE 9. Virginia sea turtle stranding regions. From Mansfield et al. 2001.

Virginia Sea Turtle Stranding Regions

Maryland

estern Shore Eastern Shore Bay Bay Side

Eastern Shore - Ocean Side

Southern Bay

Virginia Beach Ocean

5 0 5 10 Miles Virginia

North Carolina 30

distributed similarly: Western Chesapeake Bay - 179, Southern Chesapeake Bay - 40,

Eastern Shore Bayside - 38, Eastern Shore Oceanside - 9, Virginia Beach Oceanside -

37, Maryland - 17, and North Carolina - 7 (Table 3).

The annual distribution of sample and data collection is shown in Figure 10. The

majority of samples, including 90 of the 169 Virginia samples, were collected in 1990,

1991, 2001, and 2002. More general diet information was collected on stranding data

sheets in other years (Figures 10-11). The monthly distribution of samples (Figure 12) is

consistent with Virginia stranding patterns (Mansfield et al. 2001), and the majority of

samples were collected in May and June. The Maryland samples were collected from

May to October, and all of the North Carolina samples were collected in January and

December.

For Virginia, loggerheads with whole samples (n = 128) ranged in size from 41.6

to 98.5 cm straight carapace length (SCL) (mean = 63.2 cm, standard deviation (SD) =

11.9 cm), and all those with samples (n = 169) ranged in size from 39.0 to 98.5 cm SCL

(mean = 62.6 cm, SD = 11.8 cm) (Figure 13a-b). Virginia loggerheads with data (n =

303) ranged from 33.0 to 98.7 cm SCL (mean = 63.6 cm, SD = 12.3 cm) (Figure 13c).

The majority of these turtles are “benthic immatures” (TEWG 2000), and the size range

is representative of the overall loggerhead strandings in Virginia (Musick and Limpus

1997). Maryland loggerheads with samples (n = 10) measured 57.2 to 98.0 cm SCL

(mean = 73.9 cm, SD = 14.7 cm), and those with data (n= 17) ranged from 47.3 to 98.0 cm SCL (mean = 72.6 cm, SD = 16.6 cm). North Carolina loggerheads sampled (n = 6) ranged from 42.9 to 64.9 cm SCL, and those with data (n = 7) ranged from 42.9 to 64.9 cm SCL. All measurements presented are straight carapace length measured from notch 31

FIGURE 10. Annual distribution of (a) loggerhead samples (n = 185) and (b) loggerhead diet data by state (n = 327, excluding 24 empties).

(a)

45 40 o(fl 35 Q. 30 reE (/) 25 20 15 10

ja. O r M CO '^•inCDh'-COO^OT-CNCO’^LOCOf^OOul.t Ol O r M CO CO CO CO COOOOOCOOOOO O O O Oa O O O O O 0 ) 0 0 0 00005 OOOO0O000O00OOO Oooo

■ Virginia □ Outer Banks, NC 3 Potomac R. or Solomon’s Is., MD

(b)

45 ' 40 - 35 - a>CA t 30 - rs 3 1- H- 25 - o L- a> 20 - n n n 1 E 15 - n n H z 10 - I 1 il Q 5 - n 1 I 1 1 I n 1 ! ! R 1 I o -*1, , ,1,11,1,1,1,1111111m b I ox-csjco^iocor^coo OT-cNcorj-mcor^coo oooococooooococococo oooooooooo oooooooooo oooooooooo CN CN Csl

I Virginia □ Outer Banks, NC ti Potomac R. or Solomon's Is., MD 32

FIGURE 11. Annual distribution of loggerhead samples and data by data type (n = 351, 185 samples, 142 data only, 24 empty).

45 40 «tf> 35 a 30 Ero CO 25 20 15 10

:l 1 O r (N CO 000)0>0)0>0)0>d)0>d}0> 00 00 00 ’III S 8? Year

I Whole Gut ■ Partial Gut □ Fecal Sample a Data only □ Empty 33

FIGURE 12. Distribution of loggerhead samples by month and state (n = 185).

90 80

| 60 w 50 o £ 40 -Q | 30 z 20

10 - m - 0 F=* _ -- 1-- n i I n n Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

■ Virginia □ Outer Banks, NC ■ Potomac R. or Solomon's Is., MD 34

FIGURE 13. Size distribution of (a) Virginia loggerhead whole samples (n = 128), (b) all Virginia loggerhead samples (n = 169), and (c) all Virginia loggerheads with diet data (n = 303).

(a)

70 60 50 40 30 20 10 0 - I I 0 ) 0 ) 0 5 0) O) O) O) O) O) O) O) oi rf o> CO CO ^ ^ £ 8

Straight Carapace Length (cm)

(b)

70 8 60 i 50 t0 40 30 1 20 z 10 0 1------1— “ T t 00) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) O) ^ O) ^ O) ^ O) O) O ) ^ O) o> -OOOOOOOOOOOOOOO^(OCO^^lOU>Cf(ipSN0C)^)O)O)O) ■?QWjOU>QlQQIQOU>QIQQU> > -Jcoco-^-T^iOLncocors-r^coooo)0)0 Straight Carapace Length (cm)

(C) ------• 5 6070 • 3 50 • I ■ 1 o 4 0 ' ■ aJ 30 •fi 9fl ■ E 20 3 1(1 . 1 1 z 10 UA 1. ~ □ 1 in i a a aj i Under 30 30.0-34.9 35.0-39.9 50.0-54.9 55.0-59.9 60.0-64.9 80.0-84,9 85.0-89.9 90.0-94.9 95.0-99.9 40.0-44.9 45.0-49.9 65.0-69.9 " 70.0-74.9 75.0-79.9 Over Over 99.9 Straight Carapace Length (cm) 35

to notch. This particular measurement was chosen because it was recorded the most

frequently for the turtles included in this analysis. The length conversions presented in

Coles (1999) were used for loggerheads for which notch-to-notch SCL was not recorded.

Due to the small number of samples from the other states, only Virginia data are

discussed further. The gut contents for each loggerhead sampled from Maryland and

North Carolina are listed in Appendix I.

Of those loggerheads sampled in Virginia (n = 169), 74.0% (125) had no obvious

wounds or abnormalities, 11.8% (20) had become entangled in or ingested fishing gear

and/or had constriction marks, and 6.5% (11) had propeller-like or crushing injuries.

Seven appeared to have illnesses, four were probable cold stuns, one turtle had a gaff- or

bullet-like wound, and one appeared to have been struck with a blunt object. This

distribution was similar for all loggerheads from Virginia with diet data (n = 303):

73.3% (222) had no visible abnormalities, 13.2% (40) showed signs of fisheries

interactions, and 7.3% (22) had propeller-like or crushing wounds. In some cases,

however, it is difficult to tell whether injuries have occurred pre- or post-mortem

(personal observation).

The cumulative prey curve for all loggerhead whole samples collected in Virginia

during 1983 to 2002 (n = 128) appears to reach an asymptote (Figure 14). This suggests that most of the major prey items for the data set as a whole were encountered. Overall,

whole loggerhead samples had an average dry weight of 77.7 g (SD = 115.2 g, median =

42.0 g), and there was an average of 12.9 prey items per whole sample (SD = 17.9, median = 8.0) and 9.2 large prey items (horseshoe crab, crabs, fish, whelks, moon snails)

(SD = 15.7, median = 5.0) per whole sample. Linear regression indicates that there was a 36

FIGURE 14. Cumulative prey curve for whole loggerhead samples collected in Virginia during 1983 to 2002 (n = 128). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation.

v)o> 60 Q.>< >> 50

0 20 40 60 80 100 120 140 Cumulative Number of Samples 37

slight positive relationship between loggerhead SCL and total and large prey dry weights

(Figure 15) and numbers (Figure 16).

Interannual diet

The cumulative prey curves for the five-year divisions for whole loggerhead

samples from Virginia are shown in Figure 17. The curves for the two divisions with

more samples, 1988-1992 (n = 49) and 1998-2002 (n = 46), appear to have reached

asymptotes, indicating that all common prey items were probably encountered. The other

two divisions, 1983-1987 (n = 13) and 1993-1997 (n = 20), probably have not reached

asymptotes, but the slopes of the both curves decrease enough by the last sample to

suggest that most common prey items were encountered. As such, all four of these year

groups were included in analyses.

General prey data were compiled from 297 turtles examined during 1983 to 2002

and divided into five-year increments (Figure 18a). These data suggest that horseshoe

crabs occurred less frequently in the Virginia loggerhead diet from 1983 to 1997, while

fish occurred more frequently during this same time period. Crustaceans were present in

at least 50% of samples in each five-year period, and frequency peaked during 1993 to

1997. Correspondence analysis supports the trend from horseshoe crab dominance in

1983 to 1987 to crustacean and fish dominance in later years, and removing the mollusc

data (and thus small, incidental prey items) does not affect this trend (Figure 19). Small unfamiliar, and rare prey items were probably missed for the 131 turtles examined by necropsy only (Ruckdeschel and Shoop 1988), and precise prey species information was not available for all turtles. 38

FIGURE 15. Linear regressions of loggerhead straight carapace length (SCL) versus (a) total dry weight of Virginia whole samples and (b) dry weights of large prey items (horseshoe crab, crustaceans, fish, whelks, Atlantic moon snail) in Virginia whole samples collected during 1983 to 2002 (n = 128).

Y = -173.201 + 4.18693X

(a) R-Sq = 16.9 % ANOVA: p = 0.000 1000 3 900 800 I: g> 700 CD 600 § 500 400

O 300

03 200

O 100 0

40 50 60 70 80 90 100 SC L (cm )

Y = -115.465 + 3.06103X

R-Sq = 10.0 % O) ANQVA; p = Q.QQQ 900

800 O) 700

600

500

400

300

200 O) 100

40 6070 ( 9050 100 SC L (cm ) 39

FIGURE 16. Linear regressions of loggerhead straight carapace length (SCL) versus (a) total number of prey items in Virginia whole samples and (b) total number of large prey items (horseshoe crab, crustaceans, fish, whelks, Atlantic moon snail) in Virginia whole samples (n = 128).

Y = -13.1479 + 0.413245X

(a) R-Sq = 7.6 % ANOVA: p = 0.002 150

CL 100 O i_ _QCD E 50

0

40 50 60 70 80 90 100 SC L (cm )

= -17.1564 + 0.417960X (b) R-Sq =10.1 % ANOVA: p = 0.000 140 CD 120 0 CD 100

CD 80

60 0 -Q 40 E 3 20 0

40 50 60 90 100 SC L (cm ) 40

FIGURE 17. Cumulative prey curves for whole loggerhead samples collected in Virginia during (a) 1983 to 1987 (n = 13), (b) 1988 to 1992 (n = 49), (c) 1993 to 1997 (n = 20), and (d) 1998 to 2002 (n = 46). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation.

(a) 60 0)W > 50 a> a.1— H- 40 o a> -Q 30 E 3 z T < J. xsTO 3 —i -i E 10 i •i o3 ----- 1------1------1------1------1------10 20 30 40 50 60 Cumulative Number of Samples

>iQ.

40

20

20 30 40 Cumulative Number of Samples 41

FIGURE 17 (Continued).

r

Cumulative Number of Samples

(d ) 6 0

5 0 a> a i->» o>>. 4 0 u. T... .■ ■' .1 CL ' J-*L 3 0 tIIi1 tL""I _ 4 _ 20

10 p t

10 20 30 40 50 60 Cumulative Number of Samples 42

FIGURE 18. Frequency of occurrence of (a) major prey groups in all loggerhead diet data (n = 297) and (b) major prey items in all loggerhead samples (n = 166) collected in Virginia during 1983 to 2002, grouped by five-year intervals.

(a)

100

90

80

70 © o2 60 3 50 Oo 55 40

30 20 ffl 10 0 Horseshoe crab Crustacean Mollusc Fish

[■ 1983-1987 (n = 95) ■1988-1992 (n = 82) □ 1993-1997 (n = 27) B 1998-2002 (n = 93) |

1983-1987(n = 38) ■ 1988 - 1992 (n = 61) Q 1993-1997(n = 20) H 1998-2002 (n = 47) 43

FIGURE 19. Biplots of year division and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using (a) %F values of all four general diet categories and (b) %F values of three general diet categories from all Virginia loggerhead diet data collected during 1983 to 2002 (n = 297). PCI and PC2 account for 98.86% and 100.0% of the data in (a) and (b), respectively. Number of samples is in parentheses.

1 (a) 1 0 . 8 - 1 1 1 1 1 1 1 1 1 1

Mollusc | C ru st^ ean I 1 0 .0 -- Mg97 (27) , • Horseshoe

1998-2^)2 (gb)

Fish I 1 1

0.0 0.4 0.8 Component 1

0 .4 - (b)

1998-2002 (93) F'St 0.2 —| Hons ashoe crab 1983-1987 (95) 0.0-- >- 27)

C rustacean I -0 . 2 - 1988-^992 (82)

-0.4 -

- 0 . 6 -

—i------1 i 1------1------r -0.6 -0.4 -0.2 0.0 0.2 0.4 Component 1 44

More specific data were acquired for those turtles that were sampled (Tables 4-8).

Small molluscs, such as blue mussels (Mytilus edulis) and mud snails (Nassarius spp.,

Ilyanassa spp.), as well as various mollusc fragments, plant matter, and debris were

assumed to be consumed incidentally from the benthos (Youngkin and Wyneken, in

press), as gut contents of other prey (Mansour 1992), or as scavengers on other prey

items (Shaver 1991). Larger items considered to be target prey species were horseshoe

crabs, decapod crustaceans, fish, Busycon whelks, and moon snails. Atlantic moon snails

appear to be consumed alive, as indicated by the occasional presence of opercula, but

their empty shells may also be inhabited by hermit crabs {Pagurus spp.) (Frick et al.

2001). Similar trends arise over time for large, frequently occurring prey items in all of

the samples (n = 166) (Figure 18b) and all of the whole samples (n = 128) (Figure 20a).

One notable difference occurs because few whole samples were collected in earlier years,

and the whole samples collected during 1983 to 1987 appear to have a bias towards the

turtles that had eaten fish.

Index of relative importance is probably the most useful measure to compare diets

since “large” prey items varied in size (e.g. horseshoe crab versus hermit crab) (Ferry and

Calliet 1996) (Table 9). Regardless of the measure examined, a shift in loggerhead diet

from horseshoe crab dominance to Callinectes spp. (mostly blue crab, C. sapidus) to fish

dominance (supplemented by hermit crabs) is apparent (Figure 20, Table 9).

Correspondence analyses of all sample data (%F only) and all whole sample data

(especially %IRI) show a general trend along the first principal component (PCI) from horseshoe crabs to blue, rock, and spider crabs to fish, hermit crabs, and purse crabs

(Figures 21-22). TABLE 4. Percent occurrence of all prey items found in loggerhead samples (whole and partial) from Virginia during 1983 to 2002 (n = 166).

1983- 1988- 1993- 1998- 1987 1992 1997 2002 Overall (n=38) (n=61) (n=20) (n=47) (n = 166) Chelicerates 39.5 19.7 5.0 19.1 22.3 Horseshoe crab Limulus polyphemus 39.5 19.7 5.0 19.1 22.3 Crustaceans 63.2 93.4 80.0 76.6 80.1 Decapods Blue crab Callinectes sapidus 50.0 67.2 55.0 25.5 50.0 Unidentified portunid Callinectes sp. 0.0 8.2 10.0 2.1 4.8 Rock crab Cancer irroratus 10.5 49.2 20.0 12.8 26.5 Spider crab spp. Libinia spp. 26.3 39.3 25.0 23.4 30.1 Lady crab Ovalipes ocellatus 0.0 0.0 0.0 8.5 2.4 Hermit crab spp. Pagurus spp. 5.3 11.5 25.0 36.2 18.7 Purse crab Persephona mediterranea 0.0 8.2 5.0 12.8 7.2 Mantis shrimp Squdla empusa 5.3 3.3 0.0 4.3 3.6 Other Crustaceans Acorn barnacle Balanus sp. 0.0 4.9 0.0 0.0 1.8 Crab barnacle Chelonibia patula 0.0 6.6 0.0 4.3 3.6 Unidentified barnacle - 2.6 0.0 0.0 2.1 1.2 Unidentified crustacean -- 0.0 0.0 5.0 17.0 5.4 Unidentified crustacean tissue - 10.5 42.6 10.0 12.8 22.9 Fish 21.1 26.2 55.0 61.7 38.6 Bony fish -- 21.1 23.0 55.0 63.8 38.0 Elasmobranchs - 0.0 3.3 0.0 4.3 2.4 Unidentified fish tissue - 0.0 0.0 0.0 2.1 0.6 Molluscs 28.9 72.1 60.0 63.8 58.4 Bivalves Ark spp. Anadara spp. 5.3 1.6 5.0 6.4 4.2 Common jingle shell Anomia simplex 2.6 0.0 10.0 0.0 1.8 Common razor clam Ensis directus 2.6 1.6 0.0 6.4 3.0 Hard clam/clam bodies Mercenaria mercenaria/unknown 2.6 1.6 5.0 0.0 1.8 Little surf clam Mulinia lateralis 0.0 6.6 0.0 2.1 3.0 Blue mussel Mytilus edulis 2.6 47.5 15.0 29.8 28.3 Tellin sp. Tellina sp. 2.6 0.0 0.0 0.0 0.6 Unidentified bivalve - 7.9 18.0 20.0 19.1 16.3 Gastropods Lunar dove shell Astyris lunata 0.0 0.0 0.0 2.1 0.6 Knobbed and channel whelk Busy con spp. 7.9 6.6 20.0 10.6 9.6 Slipper shell sp. Crepidula sp. 0.0 1.6 15.0 4.3 3.6 Wentletrap Epitonium sp. 0.0 1.6 0.0 2.1 1.2 Thick-lipped oyster drill Europleura caudata 0.0 0.0 0.0 2.1 0.6 Eastern mud snail Ilyanassa obsoleta 0.0 0.0 5.0 2.1 1.2 Three-line mud snail Ilyanassa trivittatus 5.3 4.9 15.0 14.9 9.0 Unidentified mud snail Ilyanassa or Nassarius sp. 2.6 1.6 0.0 4.3 2.4 Spotted northern moon snail Lunatia triseriata 0.0 0.0 0.0 2.1 0.6 Mottled dog whelk Nassarius vibex 5.3 6.6 5.0 8.5 6.6 Atlantic moon snail Neverita duplicata 5.3 9.8 5.0 31.9 14.5 Pyramid shell Turbonilla sp. 0.0 0.0 0.0 2.1 0.6 Atlantic oyster drill Urosalpinx cinerea 0.0 0.0 0.0 2.1 0.6 Unidentified gastropod - 0.0 4.9 15.0 2.1 4.2 Other Molluscs Squid (beaks only) Class Cephalopoda 0.0 0.0 5.0 2.1 1.2 Unidentified mollusc - 2.6 0.0 0.0 12.8 4.2 Plants 52.6 68.9 80.0 83.0 70.5 Rockweed Fucus sp . 0.0 1.6 0.0 4.3 1.8 Widgeon grass Ruppia maritima 2.6 0.0 0.0 6.4 2.4 Gulfweed sp. Sargassum sp. 0.0 0.0 5.0 4.3 1.8 Cordgrass sp. Spartina sp . 0.0 1.6 0.0 2.1 1.2 Sea lettuce Ulva lactuca 2.6 3.3 0.0 0.0 1.8 Eelgrass Zostera marina 28.9 42.6 50.0 59.6 45.2 Unidentified marine plant - 42.1 47.5 50.0 57.4 49.4 Unidentified terrestrial plant - 0.0 1.6 10.0 12.8 5.4 Miscellaneous 52.6 59.0 60.0 76.6 62.7 Bird feather Class Aves 0.0 0.0 5.0 0.0 0.6 Grasshopper Class Insecta 0.0 0.0 0.0 4.3 1.2 Hydroid Sertularia argentea or unknown 2.6 3.3 0.0 0.0 1.8 Rocks - 10.5 34.4 25.0 40.4 29.5 Sand dollar Echinarachnius parma 0.0 0.0 0.0 2.1 0.6 Sponge Phylum Porifera 0.0 0.0 10.0 4.3 2.4 Trumpet worm/tubes Pectinaria gouldii 2.6 0.0 0.0 0.0 0.6 Tube worm/tubes Asabellides oculata 5.3 8.2 5.0 4.3 6.0 Tunicate Molgula manhattensis 0.0 1.6 15.0 6.4 4.2 Unidentified bryozoan Class Bryozoa 0.0 0.0 5.0 0.0 0.6 Unidentified eggs - 2.6 0.0 0.0 0.0 0.6 Unidentified gelatinous animal - 0.0 1.6 0.0 0.0 0.6 Unidentified polychaete Class Polychaeta 0.0 1.6 0.0 0.0 0.6 Unidentified tissue - 42.1 31.1 25.0 51.1 38.6 Anthropogenic Items 2.6 3.3 5.0 8.5 4.8 Gillnet - 0.0 0.0 0.0 2.1 0.6 Glass - 2.6 1.6 0.0 0.0 1.2 Hook and line gear - 0.0 1.6 5.0 4.3 2.4 Latex - 0.0 0.0 0.0 2.1 0.6 Plastic -- 0.0 0.0 0.0 2.1 0.6 TABLE 5. Percent occurrence (%F), percent dry weight (%W), percent number (%N), and percent index of relative importance (%IRI) of all prey items found in whole loggerhead gut samples from Virginia during 1983 to 1987 (n = 13). Note: colonial animals such as hydroids and tube worms were given a count (number) of one.

%F %W %N %IRI Chelicerates 69.2 27.6 13.2 27.5 Horseshoe crab Limulus polyphemus 69.2 27.6 13.2 38.7 Crustaceans 53.8 41.8 31.6 38.5 Decapods Blue crab Callinectes sapidus 46.2 28.7 21.1 31.4 Rock crab Cancer irroratus 7.7 0.4 3.9 0.5 Spider crab spp. Libinia spp. 23.1 7.0 3.9 3.5 Mantis shrimp Squilla empusa 7.7 0.0 1.3 0.1 Other Crustaceans Unidentified barnacle - 7.7 0.0 1.3 0.1 Unidentified crustacean tissue - 7.7 5.6 —— Fish 38.5 22.5 17.1 14.9 Bony fish - 38.5 22.5 17.1 20.9 Molluscs 38.5 4.4 32.9 14.0 Bivalves Ark spp. Anadara spp. 7.7 0.0 2.6 0.3 Hard clam/clam bodies Mercenaria mercenaria!unknown 7.7 0.2 3.9 0.4 Blue mussel Mytilus edulis 7.7 0.0 3.9 0.4 Unidentified bivalve - 7.7 1.2 5.3 0.7 Gastropods Knobbed and channel whelk Busy con spp. 7.7 1.6 1.3 0.3 Mottled dog whelk Nassarius vibex 7.7 0.1 2.6 0.3 Other Molluscs Unidentified mollusc - 7.7 1.1 13.2 1.5 Plants 46.2 0.2 —— Widgeon grass Ruppia maritima 7.7 0.0 -- Sea lettuce Ulva lactuca 7.7 0.0 -- Eelgrass Zostera marina 30.8 0.1 -- Unidentified marine plant ~ 46.2 0.1 —- Miscellaneous 61.5 3.3 5.3 5.1 Hydro id Sertularia argentea or unknown 7.7 0.0 1.3 0.1 Rocks - 23.1 0.5 -- Trumpet worm/tubes Pectinaria gouldii 7.7 0.0 1.3 0.1 Tube worm/tubes Asabellides oculata 15.4 0.4 2.6 0.6 Unidentified tissue .. 38.5 2.4 —— Anthropogenic Items 7.7 0.3 -- Glass - 7.7 0.3 — — TABLE 6. Percent occurrence (%F), percent dry weight (%W), percent number (%N), and percent index of relative importance (%IRI) of all prey items found in whole loggerhead gut samples from Virginia during 1988 to 1992 (n = 49). Note: colonial animals such as hydroids and tube worms were given a count (number) of one.

%F %W %N %IRI Chelicerates 22.4 1.7 1.7 0.5 Horseshoe crab Limulus polyphemus 22.4 1.7 1.7 1.1 Crustaceans 93.9 70.4 52.7 71.5 Decapods Blue crab Callinectes sapidus 67.3 34.2 22.2 52.3 Unidentified portunid Callinectes sp. 8.2 0.1 0.6 0.1 Rock crab Cancer irroratus 51.0 4.2 12.3 11.6 Spider crab spp. Libinia spp. 42.9 13.7 7.4 12.5 Hermit crab spp. Pagurus spp. 12.2 1.0 5.9 1.2 Purse crab Persephona mediterranea 4.1 0.1 0.3 0.0 Mantis shrimp Squilla empusa 4.1 0.0 0.3 0.0 Other Crustaceans Acorn barnacle Balanus sp. 6.1 0.0 2.6 0.2 Crab barnacle Chelonibia patula 6.1 0.0 1.1 0.1

Unidentified crustacean tissue - 42.9 17.1 -— Fish 20.4 6.2 6.3 1.6 Bony fish - 18.4 6.1 6.0 3.1 Elasmobranchs - 4.1 0.0 0.3 0.0 Molluscs 75.5 5.5 37.9 20.3 Bivalves Ark spp. Anadara spp. 2.0 0.0 0.2 0.0 Common razor clam Ensis directus 2.0 0.0 0.2 0.0 Hard clam/clam bodies Mercenaria mercenaria!unknown 2.0 0.1 3.7 0.1 Little surf clam Mulinia lateralis 6.1 0.0 1.1 0.1 Blue mussel Mytilus edulis 51.0 0.5 21.0 15.1 Unidentified bivalve - 20.4 0.5 4.3 1.4 Gastropods Knobbed and channel whelk Busy con spp. 6.1 4.2 3.4 0.6 Slipper shell sp. Crepidula sp. 2.0 0.0 0.2 0.0 Wentletrap Epitonium sp. 2.0 0.0 0.3 0.0 Three-line mud snail Ilyanassa trivittatus 6.1 0.0 0.9 0.1 Unidentified mud snail Ilyanassa or Nassarius sp. 2.0 0.0 0.2 0.0 Mottled dog whelk Nassarius vibex 6.1 0.0 0.8 0.1 Atlantic moon snail Neverita duplicata 10.2 0.2 1.2 0.2 Unidentified gastropod - 6.1 0.0 0.6 0.1

Plants 69.4 0.2 — —

Rockweed Fucus sp. 2.0 0.1 - — Cordgrass sp. Spartina sp. 2.0 0.0 -- Sea lettuce Ulva lactuca 2.0 0.0 -- Eelgrass Zostera marina 44.9 0.1 -- Unidentified marine plant - 49.0 0.1 —— Unidentified terrestrial plant - 2.0 0.0 ~ — Miscellaneous 57.1 16.0 1.4 6.1 Hydro id Sertularia argentea or unknown 4.1 0.0 0.3 0.0 Rocks -- 30.6 0.6 -- Tube worm/tubes Asabellides oculata 10.2 0.1 0.8 0.1 Unidentified gelatinous animal - 2.0 0.0 0.2 0.0 Unidentified polychaete Class Polychaeta 2.0 0.0 0.2 0.0 Unidentified tissue - 30.6 15.3 —— Anthropogenic Items 4.1 0.0 —- Glass - 2.0 0.0 -- Hook and line gear - 2.0 0.0 — - TABLE 7. Percent occurrence (%F), percent dry weight (%W), percent number (%N), and percent index of relative importance (%IRI) of all prey items found in whole loggerhead gut samples from Virginia during 1993 to 1997 (n = 20). Note: colonial animals such as hydroids and tube worms were given a count (number) of one.

%F %W %N %IRI Chelicerates 5.0 0.4 0.6 0.0 Horseshoe crab Limulus polyphemus 5.0 0.4 0.6 0.1 Crustaceans 80.0 46.2 36.0 49.2 Decapods Blue crab Callinectes sapidus 55.0 13.7 14.6 28.5 Unidentified portunid Callinectes sp . 10.0 0.2 1.2 0.3 Rock crab Cancer irroratus 20.0 13.3 6.7 7.3 Spider crab spp. Libinia spp. 25.0 13.3 7.3 9.4 Hermit crab spp. Pagurus spp. 25.0 0.2 4.9 2.3 Purse crab Persephona mediterranea 5.0 0.5 0.6 0.1 Other Crustaceans Unidentified crustacean ~ 5.0 0.2 0.6 0.1 Unidentified crustacean tissue ~ 10.0 4.8 —- Fish 55.0 17.9 18.9 15.1 Bony fish - 55.0 17.9 18.9 37.0 Molluscs 60.0 27.7 39.0 29.9 Bivalves Ark spp. Anadara spp. 5.0 0.9 3.0 0.4 Common jingle shell Anomia simplex 10.0 0.3 1.2 0.3 Hard clam/clam bodies Mercenaria mercenaria!unknown 5.0 19.0 3.0 2.0 Blue mussel Mytilus edulis 15.0 0.0 4.9 1.3 Unidentified bivalve - 20.0 1.8 6.7 3.1 Gastropods Knobbed and channel whelk Busy con spp. 20.0 2.8 4.3 2.6 Slipper shell sp. Crepidula sp. 15.0 1.3 5.5 1.9 Eastern mud snail Ilyanassa obsoleta 5.0 0.1 0.6 0.1 Three-line mud snail Ilyanassa trivittatus 15.0 0.1 2.4 0.7 Mottled dog whelk Nassarius vibex 5.0 0.1 0.6 0.1 Atlantic moon snail Neverita duplicata 5.0 0.1 1.2 0.1 Unidentified gastropod -- 15.0 1.0 2.4 1.0 Other Molluscs Squid (beaks only) Class Cephalopoda 5.0 0.0 3.0 0.3 Plants 80.0 0.4 —- Gulfweed sp. Sargassum sp. 5.0 0.1 -- Eelgrass Zostera marina 50.0 0.1 -- Unidentified marine plant ~ 50.0 0.2 -- Unidentified terrestrial plant - 10.0 0.0 -- Miscellaneous 60.0 7.2 5.5 5.7 Bird feather Class Aves 5.0 0.0 0.6 0.1 Rocks - 25.0 0.1 —- Sponge Phylum Porifera 10.0 0.8 1.2 0.4 Tube worm/tubes Asabellides oculata 5.0 0.0 0.6 0.1 Tunicate Molgula manhattensis 15.0 0.1 2.4 0.7 Unidentified bryozoan Class Bryozoa 5.0 0.0 0.6 0.1 Unidentified tissue - 25.0 6.2 -- Anthropogenic Items 5.0 0.1 — ~ Hook and line gear - 5.0 0.1 -- TABLE 8. Percent occurrence (%F), percent dry weight (%W), percent number (%N), and percent index of relative importance (%IRI) of all prey items found in whole loggerhead gut samples from Virginia during 1998 to 2002 (n = 46). Note: colonial animals such as hydroids and tube worms were given a count (number) of one.

%F %W %N %IRI Chelicerates 19.6 5.8 1.6 1.0 Horseshoe crab Limulus polyphemus 19.6 5.8 1.6 2.4 Crustaceans 78.3 36.9 47.4 47.9 Decapods Blue crab Callinectes sapidus 26.1 2.2 4.4 2.8 Unidentified portunid Callinectes sp. 2.2 0.0 0.1 0.0 Rock crab Cancer irroratus 13.0 0.3 1.6 0.4 Spider crab spp. Libinia spp. 23.9 17.3 7.3 9.7 Lady crab Ovalipes ocellatus 8.7 1.8 7.7 1.4 Hermit crab spp. Pagurus spp. 37.0 1.4 20.2 13.1 Purse crab Persephona mediterranea 13.0 2.2 3.7 1.3 Mantis shrimp Squilla empusa 4.3 0.0 0.3 0.0 Other Crustaceans Crab barnacle Chelonibia patula 4.3 0.0 0.7 0.0 Unidentified barnacle - 2.2 0.0 0.4 0.0 Unidentified crustacean - 17.4 0.0 1.0 0.3 Unidentified crustacean tissue - 13.0 11.5 —- Fish 63.0 43.7 15.0 26.9 Bony fish - 63.0 40.8 14.6 57.2 Elasmobranchs - 4.3 0.4 0.4 0.1 Unidentified fish tissue - 2.2 2.5 —— Molluscs 65.2 4.7 34.9 18.8 Bivalves Ark spp. Anadara spp. 6.5 0.0 0.4 0.0 Common razor clam Ensis directus 6.5 0.0 0.4 0.0 Little surf clam Mulinia lateralis 2.2 0.0 1.4 0.1 Blue mussel Mytilus edulis 28.3 0.1 3.9 1.9 Unidentified bivalve - 17.4 0.1 2.6 0.8 Gastropods Lunar dove shell Astyris lunata 2.2 0.0 0.1 0.0 Knobbed and channel whelk Busycon spp. 10.9 2.6 8.0 1.9 Slipper shell sp. Crepidula sp. 4.3 0.1 1.3 0.1 Wentletrap Epitonium sp. 2.2 0.0 0.4 0.0 Thick-lipped oyster drill Europleura caudata 2.2 0.1 0.3 0.0 Eastern mud snail Ilyanassa obsoleta 2.2 0.0 0.1 0.0 Three-line mud snail Ilyanassa trivittatus 15.2 0.2 5.1 1.3 Unidentified mud snail Ilyanassa or Nassarius sp. 4.3 0.0 0.7 0.0 Spotted northern moon snail Lunatia triseriata 4.3 0.0 0.1 0.0 Mottled dog whelk Nassarius vibex 8.7 0.0 0.5 0.1 Atlantic moon snail Neverita duplicata 32.6 1.3 7.4 4.7 Pyramid shell Turbonilla sp. 2.2 0.0 0.1 0.0 Atlantic oyster drill Urosalpinx cinerea 2.2 0.0 0.1 0.0 Unidentified gastropod - 2.2 0.0 0.1 0.0 Other Molluscs Squid (beaks only) Class Cephalopoda 2.2 0.0 0.1 0.0 Unidentified mollusc - 13.0 0.1 1.6 0.4

Plants 87.0 0.2 — —

Rockweed Fucus sp. 4.3 0.0 — -

Widgeon grass Ruppia maritima 6.5 0.0 --

Gulfweed sp. Sargassum sp . 4.3 0.0 - - Cordgrass sp. Spartina sp . 2.2 0.0 --

Eelgrass Zostera marina 60.9 0.1 - -

Unidentified marine plant - 56.5 0.1 - -

Unidentified terrestrial plant - 13.0 0.0 —— Miscellaneous 76.1 8.5 1.2 5.3 Grasshopper Class Insecta 4.3 0.0 0.3 0.0

Rocks « 39.1 0.6 -- Sand dollar Echinarachnius parma 2.2 0.0 0.1 0.0 Sponge Phylum Porifera 4.3 0.0 0.3 0.0 Tube worm/tubes Asabellides oculata 2.2 0.0 0.1 0.0 Tunicate Molgula manhattensis 6.5 0.0 0.4 0.0

Unidentified tissue - 52.2 7.7 —-

Anthropogenic Items 6.5 0.3 — —

Hook and line gear - 4.3 0.2 - -

Latex - 2.2 0.0 -- Plastic — 2.2 0.0 —— 54

FIGURE 20. (a) Frequency of occurrence, (b) percent dry weight, (c) percent number, and (d) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia from 1983 to 2002 (n = 128), grouped by five-year intervals.

■ 1983-1987 (n = 13) ■ 1988 - 1992(n = 49) a 1993-1997 (n = 20) ■ 1998-2002 (n = 46) 55

FIGURE 20 (Continued).

(c)

50

40

.a<5 30 E 3

■ 1983-1987 (n = 13) ■ 1988-1992 (n = (49) □ 1993-1997 (n = 20) a 1998-2002 (n = 46)

100

■ 1983-1987 (n = 38) ■ 1988 -1992 (n = 61) □ 1993-1997 (n = 20) Q 1998-2002 (n = 47) TABLE 9. Prey species with five highest percent index of relative importance values for whole loggerheads samples from Virginia during 1983 to 2002 (n = 128).

1983-1987 (n=13)______1988-1992 (n=49) Horseshoe crab (38.7%) Blue crab (52.3%) Blue crab (31.4%) Blue mussel (15.1%) Bony fish (20.9%) Spider crab spp. (12.5%) Spider crab spp. (3.5%) Rock crab (11.6%) Unidentified mollusc (1.5%) Bony fish (3.1%)

1993-1997 (n=20) 1998-2002 (n=46) Bony fish (37.0%) Bony fish (57.2%) Blue crab (28.5%) Hermit crab spp. (13.1%) Spider crab spp. (9.4%) Spider crab spp. (9.7%) Rock crab (7.3%) Atlantic moon snail (4.7%) Unidentified bivalve (3.1%) Blue crab (2.8%) 57

FIGURE 21. Biplot of year division and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using %F values for major prey items from all loggerhead samples collected in Virginia during 1983 to 2002 (n = 166). PCI and PC2 account for 84.28% of the data. Number of samples is in parentheses.

1.5-r

1.0 -

CN He "seshoe crab

Lady era § 0 .5 - §_ 198 1-1287 (38) E Oo Atlantic moon snail 1998-2^02 (47) Spider c^ab spp 0 .0 -- -B rin y rrferr fi spp. MW• Busycqn ^helk spp. tocl^crab ! 993.^7 (20)

T T 0.0 0.5 1.0 1.5 Component 1 58

FIGURE 22. Biplots of year division and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 128). PCI and PC2 account for 86.77%, 88.22%, 93.14%, and 96.03% of the data in (a), (b), (c), and (d), respectively. Number of samples is in parentheses. (a)

1 . 0 -

Lady era 3

Hors sshoe crab 0 .5 - 383-1987 (13)

B on^fish EL Purse■ crab Busycon whelk spp. Hermit crab spp. 8 °-°-+...... Spider crabl^p r1997(20) Callinecte^sppi 1^2 (49j -0.5-

r -0.5 0.0 0.5 1.0 Component 1

1 - Horseshoe crab (b) 198p-1987 (13)

Callinectes spp. Atlgnli|i moflrLSoaii 4 Purse crab • 1988-1992 (49) elk spp.

19^3-1997 (20)

- 1- Roc l^ crab

0 Component 1 59

FIGURE 22 (Continued).

Horseshoe crab

0.5 1983-1987 (13) idy crab

Purse crab Atlantic rh?§fl 0.0 -- Hermit whelk spp. linectes spp. " Spidel

l-1g92 (49) E -0 .5 - Rock crab

-1.0-

i i i i r -1.5 -1.0 -0.5 0.0 0.5 Component 1

Horseshoe crab

1983-1987 (13)

Atlantic mo in snail, lady crab & Purse crab iny fish Hermit crab ^ p .

(20) C allinectes spp. > spp. ■

1988-1992 (49)

1------1------T -1 0 1 Component 1 60

Interestingly, Atlantic moon snails tended to group with hermit crabs in the biplots,

suggesting that the crabs are using the shells.

The fish species encountered in all samples (partial and whole) are listed in Table

10. Atlantic menhaden (Brevoortia tyrannus), striped bass (Morone saxatilis) and

bluefish (Pomatomus saltatrix) were encountered in all four of the five-year periods.

Menhaden and croaker (Micropogonias undulatus) were the two most frequently

encountered fish species during 1993-1997 and 1998-2002. Batoids (skates and rays)

were encountered in several samples, and a sharpnose shark (Rhizoprionodon

terraenovae) was identified from teeth found in a sample from 2001.

Size-specific diet

As with the annual Virginia data, only the cumulative prey curves for size classes

with larger numbers of whole samples (50.0-59.9 cm and 60.0-69.9 cm SCL) reached

asymptotes (Figure 23); however, data for the smaller size class and the three larger ones

were still examined (Figure 24). Although sample size discrepancies should be kept in

mind, horseshoe crab, hermit crab, and whelk consumption (and %IRI) increased with

size (and therefore age) of loggerheads, while blue crab consumption appeared to

decrease with size. Turtles in the middle size classes consumed bony fish most

frequently, and none of the largest turtles (90.0-99.9 cm SCL) had consumed fish.

Correspondence analysis of %F values separated loggerheads into three groups along

PCI: 40.0-69.9 cm SCL, 70.0-89.9 cm SCL, and 90.0-99.9 cm SCL (Figure 25a).

Correspondence analyses using the %W, %N, and %IRI values separated the 90.0-99.9 cm SCL turtles from all other size classes (Figure 25b-d). By calculating the Schoener TABLE 10. Percent occurrence of all fish species found in loggerhead samples (whole and partial) from Virginia during 1983 to 2002 (n = 166).

1983- 1988- 1993- 1998- 1987 1992 1997 2002 (n=38) (n=61) (n=20) (n=47) Bony Fish Striped anchovy Anchoa hepsetus 0.0 0.0 5.0 0.0 Menhaden Brevoortia tyrannus 2.6 16.4 35.0 23.4 Seatrout sp. Cyno scion sp. 0.0 1.6 5.0 10.6 Spot Leiostomus xanthurus 0.0 0.0 0.0 2.1 Croaker Micropogonias undulatus 0.0 0.0 20.0 40.4 Striped bass Morone saxatilis 5.3 1.6 5.0 4.3 Oyster toadfish Opsanus tau 0.0 1.6 0.0 0.0 Bluefish Pomatomus saltatrix 7.9 4.9 10.0 4.3 Unidentified clupeid Family Clupeidae 0.0 0.0 0.0 6.4 Unidentified flatfish Order Pleuronectiformes 2.6 1.6 0.0 0.0 Unidentified bony fish - 10.5 9.8 20.0 17.0 Elasmobranchs Cleamose skate Raja eglanteria 0.0 0.0 0.0 2.1 Sharpnose shark Rhizoprionodon terraenovae 0.0 0.0 0.0 2.1 Stingray sp. Family Dasyatidae 0.0 1.6 0.0 0.0 Skate egg case Family Rajidae 0.0 1.6 0.0 0.0 Unidentified batoid — 0.0 0.0 0.0 2.1 62

FIGURE 23. Cumulative prey curves for whole samples from Virginia loggerheads with straight carapace lengths of (a) 40.0 to 49.9 cm (n = 13), (b) 50.0 to 59.9 cm (n = 44), (c) 60.0 to 69.9 cm (n = 48), (d) 70.0 to 79.9 cm (n = 9), (e) 80.0 to 89.9 cm (n = 8), and (f) 90.0 to 99.9 cm (n = 6). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation.

>; 40 h >%

30

20

0 10 20 30 40 50 Cumulative Number of Samples

50

30 -Q 20

4020 30 4020 Cumulative Number of Samples 63

" LIBRARY of the VIRGINIA INSTi i U of . MARINE SCiENC

FIGURE 23 (Continued).

(C ) 50

(A h->* 40 >* 0)k. Q. **-0 30 1 E3 z 20 ©> 3 E 3 o

0 10 20 30 40 50 Cumulative Number of Samples

>; 40

I

Cumulative Number of Samples 64

FIGURE 23 (Continued).

(e) 50

a>: 40 a)>* a.V. H— o 30 na> E3 20

3E 10 o

10 20 30 40 50 Cumulative Number of Samples

(f) 50

(A £ 40 >*

—i------1------1------1— 10 20 30 40 50 Cumulative Number of Samples 65

FIGURE 24. (a) Frequency of occurrence and (b) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia during 1983 to 2002 (n = 128), grouped by ten-centimeter size classes (SCL).

(a)

100 90 80 70 60 50 40 30 20 10 0

■40.0-49.9 cm (n=13) 150-59.9 cm (n=44) 0 60-69.9 cm (n=48) □ 70-79.9 cm (n=9) 180.0-89.0 cm (n=8) □ 90.0-99.9 cm (n=6)

(b)

100 90 80 70 60

40 30 20 10 0

■ 40.0^9.9 cm (n=13) ■ 50-59.9 cm (n=44) □ 60-69.9 cm (n=48) O 70-79.9 cm (n=9) ■ 80.0-89.0 cm (n=8) P 90.0-99.9 cm (n=6) 66

FIGURE 25. Biplots of size class and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 128). Label represents lower end of size class (e.g. 40 includes samples from 40.0 to 49.9 cm). PCI and PC2 account for 95.23%, 86.60%, 89.38%, and 82.46% of the data in (a), (b), (c), and (d), respectively. Number of samples is in parentheses. (a)

e■v, c in whelk spp.

Spider rabspp. 40£13) Calllnectesa 50 £45)

60 £18) CN 0 .0 -- C C

.0 - - i------r r~ - 1.0 -0.5 0.0 0.5 Component 1

70(9) (b) Spldqrcraffspp. Atl. moon snail

50^45) Bony fish 0 ------90^6) ------.Uocseehae-cnab- - a Busyc in whelk spp Rock crab Hermit crab spp. 40°iT) r« 3)60^48)3 ) Calllnectes spp.

8. E Oo -1 -

- 2-

-1 Component 1 67

FIGURE 25 (Continued).

Hermit crab spp.

A * . m o o t Horseshoe crat)

CM iipidebuaUspp. Busyc , Rock crab c I 4(f f 13) CD aC E Oo

- 1 -

■1 1 Component 1

Hermjt crab spp. I » 80(7)

Atl. moon snail Horseshoe crab

C Busyc in whelPspp hsh

-2 •1 1 Component 1 68

(1970) index using average percent dry weight values from whole samples, a diet overlap of 80.5% was found between loggerheads 40.0-69.9 cm SCL (n = 106) and those 70.0-

99.9 cm SCL (n = 22) for the more prominent prey items (horseshoe crabs,Callinectes spp., spider, hermit, and rock crabs, bony fish, whelks, and Atlantic moon snails).

Sex-specific diet

The cumulative prey curves for whole samples collected in Virginia from female, male, and unknown sex loggerheads all reached reasonable asymptotes (Figure 26). The ratio of females to males examined in this study was almost three to one, whereas reported values for Virginia have been about two to one (Musick et al. 1984; Bellmund et al. 1987). Discounting the unknown turtles, there did not appear to be any major diet differences between female (n = 73) and male (n = 26) loggerheads (Figure 27).

Examination of CA biplots along PCI yielded the same conclusion (Figure 28). The perceived difference in whelk %IRI was due to an adult male sample containing 57

Busycon whelk opercula. An 89.4% overlap was found between the sexes using the

Schoener (1970) index and average percent dry weight values for horseshoe crabs,

Callinectes spp., spider, hermit, and rock crabs, bony fish, whelks, and Atlantic moon snails. If whelks were excluded, the overlap was 92.1%.

Interseasonal diet

Samples were also grouped according to season. Only three Virginia loggerhead samples were collected in Virginia during winter months (December to February). 69

FIGURE 26. Cumulative prey curves for whole loggerhead samples from Virginia taken from (a) females (n = 73), (b) males (n = 26), and (c) unknowns (n = 29). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation. (a) 70 60 K>> !> 50 CL a2 40 .o E = 30 > 20 3 3E o 10 0 0 10 20 3040 50 60 70 80 Cumulative Number of Samples

(b) 70 8 60 H £^ 50 o 2 40 £ |3 30 ®> | 20 3 E o 10 0 0 10 20 30 40 50 60 70 80 Cumulative Number of Samples

(c) 70 I* CLI' 50 O 2 40 2E Z3 30 o .2 3 20 3 E O 10 0 0 10 20 30 4050 60 70 80 Cumulative Number of Samples 70

FIGURE 27. (a) Frequency of occurrence and (b) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia during 1983 to 2002 (n = 128), grouped by sex of turtle.

B Fem ale (n = 73) □ Male (n = 26) 13 Unknown (n = 29)

M Female (n = 73) D Male (n = 26) a Unknown (n = 29) 71

FIGURE 28. Biplots of turtle sex and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 128). PCI and PC2 account for 100% of the data in all four analyses. Number of samples is in parentheses.

------r — — - - I ( ® ) 0 .6 - I I i 0 .5 - I I 1 0 .4 - i I 1 0 .3 - I 1 Busycon wt CM 1 » c 0 .2 - i a> c i 8. o . i - Spider crabs [ M al^(26) E ftermit crabs nown(29RLq5^ teh * 1

Bus) con whelks (b) 0 .5 - '

Male (26) Horsesflbe crab Spider-crabs

Rock crab

0.0------OnttnO^Ti (2ST v c Callinectes spp. 8. E Oo F e m ^ (73)

-0.5-

: rjjo o n s

-0.5 0.0 0.5 Component 1 72

FIGURE 28 (Continued).

(C) spider crabs

Rock>c)rS#ncjjrn ( !9)

Bus; con whelks Male (26) Atlantic moon■ sr ail 0.0 -.------n itcrabs

Callinc jtes spp. Horseshoe crab Femate (73) 8. OI -0.5-|

- 1.0 -

“I------1.0 -0.5 0.0 Component 1

(d) Atlantic modn• I snail Femate (73) Bon^fish ,

Unknown (29)

Rock crab o.o- ...... -■------1Hem|it-cmbs ------

Spider crabs

I Mal^26)|_|ors0shoe cr. -0.5-

- 1. 0- B u sy co j wh< Iks ~I - 1.0 -0.5 0.0 Component 1 73

Therefore, diet was examined among only spring (March to May), summer (June to

August), and fall (September to November). The spring and summer cumulative prey curves reached asymptotes, while the fall curve appeared close to an asymptote (Figure

29). Although no major prey differences were apparent, the spring and summer samples were dominated slightly by Callinectes spp. and fish, while horseshoe crabs and hermit crabs slightly dominated fall diet (Figure 30). Spring and summer grouped away from fall in CA biplots (Figure 31).

Kemp’s Ridleys

Nine partial digestive tract samples and one fecal sample were collected from

Kemp’s ridleys in Virginia and archived at VIMS during 1987 to 1994. Five whole ridley gut samples were archived from 1991 to 1994, and 18 were collected from 2000 to

2002 (Figures 4, 32-35, Table 11). Two whole samples were collected from North

Carolina during 1990 to 1991 (Figure 8). Additional data were available on 26 Kemp’s ridleys from Virginia and one from North Carolina (Table 11). Three ridleys from

Virginia and one from North Carolina had empty digestive tracts.

The Kemp’s ridley samples (n = 35) were distributed geographically as follows

(number of whole samples in parentheses): Western Chesapeake Bay - 13 (9), Southern

Chesapeake Bay - 4 (3), Eastern Shore Bayside - 6 (6), Virginia Beach Oceanside - 10

(5), and North Carolina - 2 (2) (Figures 32-34, Table 11). Available Kemp’s ridley data

(n = 62) were distributed similarly: Western Chesapeake Bay - 19, Southern Chesapeake

Bay - 15, Eastern Shore Bayside -1 1 , Virginia Beach Oceanside - 14, and North

Carolina-3 (Table 11). 74

FIGURE 29. Cumulative prey curves for whole loggerhead samples from Virginia during (a) Spring (March to May, n = 47), (b) Summer (June to August, n = 64), and (c) Fall (September to November, n = 14). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation.

70

2 60

a. o £40 E ^3 30 a .2 n 20 3 E O 10 0 0 10 20 30 40 50 60 70 Cumulative Number of Samples

70

60

£g- 50 £° 40 E = 30 o .2 m 20 3 E 5 10 o 0 10 20 30 40 50 60 70 Cumulative Number of Samples

(C)

20 30 40 50 60 70 Cumulative Number of Samples 75

FIGURE 30. (a) Frequency of occurrence and (b) percent index of relative importance (IRI) in whole loggerhead gut samples collected in Virginia during 1983 to 2002 (n = 128), grouped by season.

■ Spring (Mar - May, n =47) □ Summer (June - Aug, n = 64) a Fall (Sept - Nov, n = 14)

(b )

100

90 80 70 60

■ Spring (Mar - May, n = 47) a Summer (June-A ug, n = 64) a Fall (Sept - Nov, n = 14) 76

FIGURE 31. Biplots of season and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using (a) %F, (b) %W, (c) %N, and (d) %IRI values for major prey items from all whole loggerhead samples from Virginia during 1983 to 2002 (n = 125, spring, summer, and fall only). PCI and PC2 account for 100% of the data in all four analyses. Number of samples is in parentheses.

(a) 0.5-

Bus) con whelks

Spring (47)

CM tylantic moon snail c0 0 .0 --...... c H orseshoe cf ^ llineci es*s PP a g o n ^ fish E Oo l Sum rrer (64)

-0.5-

-0.5 0.0 0.5 Component 1

------1------1 (b) 0.5- 1 1 1 1 Sum ftlP^ess' • 1 1 1 1 Horseshoe crab | Spider crabs " | 0.0- a im 4) " i I Bon^fis 1 l 1 a I E I o Spring (47) O i

Rock crab -0.5 - a 1

Atlantic moon snail . ■ i i 1 I I Busycon whelks i

-0.5 0.0 0.5 Component 1 77

FIGURE 31 (Continued).

Bus} con whelks (C)

Rock crab 0.5- Spring (47)

C

Callinectes spp. Bon^fish -0.5- Summer (64) "I------1 -0.5 0.0 0.5 Component 1

(d) 0 .5 - Summgr (i 4)

Horseshoe crab Callinectes sj p.

Hefmit crabs • Fallg14) o.o-. Atlantic moon snail Spider dabs

8 8. Spring (47)

O -0.5-

Busycon whelks

- 1. 0- I ~~r - 1.0 -0.5 0.0 0.5 Component 1 78

FIGURE 32. Approximate locations of all Kemp’s ridleys from Virginia from which samples were collected during 1987 to 1989 (n = 6). 79

FIGURE 33. Approximate locations of all Kemp’s ridleys from Virginia from which samples were collected during 1991 to 1994 (n = 9).

f

• 1992 80

FIGURE 34. Approximate locations of all Kemp’s ridleys from Virginia from which samples were collected during 2000 to 2002 (n = 18).

Y

\ TABLE 11. Composition of Kemp’s ridley diet data and samples collected during 1987 to 2002 (excludes three empty tracts from Virginia and one from North Carolina).

Turtles with Samples Diet Data (Whole, Partial, Whole (Includes and Fecal Digestive Tract Samples) Samples) Samples Western Bay, Virginia 19 13 9 Southern Bay, Virginia 15 4 3 Eastern Shore Bayside, Virginia 11 6 6 Eastern Shore Oceanside, Virginia 0 0 0 Virginia Beach Oceanside, Virginia 14 10 5 Maryland 0 0 0 North Carolina 3 2 2 Total 62 35 25 82

FIGURE 35. Annual distribution of (a) Kemp’s ridley samples (n = 35) and (b) Kemp’s ridley diet data by state (n = 62, excluding 4 empties).

(a) 10 9 8 otfl a 7 reE 6 (0 5 na> 4 E 3 3 II 2 1 I II 0 ______! r m u i i i L . CM 8 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1994 1995 1997 1998 1999 1993 1996 o n rn 2000 CM

■ Virginia □ Outer Banks, NC

■ Virginia □ Outer Banks, NC 83

The majority of the Kemp’s ridley samples, including 18 of the 33 Virginia samples, were collected during 2000 to 2002 (Figure 35). Some general diet information was collected on stranding data sheets in other years (Figures 35-36). The monthly distribution of samples (Figure 37) is consistent with Virginia spring and fall stranding peaks (Mansfield et al. 2001), and the vast majority of samples were collected in May,

June, and October. The two North Carolina samples were collected in December.

Kemp’s ridleys from Virginia with whole samples taken (n = 23) ranged in size from 23.1 to 49.9 cm SCL (mean = 36.7 cm, SD = 7.3 cm) (Figure 38a), and all those with samples (n - 33) ranged in size from 23.1 to 53.4 cm SCL (mean = 37.9 cm, SD =

8.5 cm) (Figure 38b). Virginia Kemp’s ridleys with data (n = 59) ranged from 23.0 to

53.4 cm SCL (mean = 36.0 cm, SD = 8.6 cm) (Figure 38c). All of the Kemp’s ridleys examined were “benthic immatures” (TEWG 2000), and the size range included all but the largest ridleys encountered in Virginia (Musick and Limpus 1997). The North

Carolina ridleys sampled were 30.0 cm and 38.5 cm SCL. As with the loggerhead data, all measurements presented are notch-to-notch SCL, and conversions presented in Coles

(1999) were used for ridleys for which this measurement was not available.

Only Virginia data are discussed further. The contents of the two North Carolina

Kemp’s ridley samples are listed in Appendix II.

Of those Kemp’s ridleys from Virginia with samples (n = 33), 78.8% (26) had no obvious abnormalities, and 15.2% (5) had propeller-like or crushing injuries. One ridley sampled had ingested hook and line fishing gear, and one was an incidental dredge take.

This distribution was similar for all Virginia Kemp’s ridleys with diet information 84

FIGURE 36. Annual distribution of Kemp’s ridley samples and data by data type (n = 66, 35 samples, 27 data only, 4 empty).

Year

■ Whole Gut ■ Partial Gut □ Fecal Sample I I Data Only (3 Empty 85

FIGURE 37. Distribution of Kemp’s ridley samples by month and state (n = 35).

11 10 9 8 ©(A 7 a. raE 6 V) 5 <4o- % 4 3 2 1 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month

■ Virginia □ Outer Banks, NC 86

FIGURE 38. Size distribution of (a) Virginia Kemp’s ridley whole samples (n = 23), (b) all Virginia Kemp’s ridley samples (n = 33), and (c) all Virginia Kemp’s ridleys with diet data (n = 59).

(a) 10 8 6

4 2 0 o o> o> o> 05 o> 05 o> CN a> o> oi CM 3 CO s s

Straight Carapace Length (cm)

(b) 10 8 6

4 2

0 0 5 05 0 5 05 05 05 05 ''t 05 05 05 TT CN CN CO CO in

Straight Carapace Length (cm)

Straight Carapace Length (cm) 87

(n = 59): 79.7% (47) had no visible abnormalities, 11.9% (7) had propeller-like wounds,

and 5.1% (3) showed evidence of fisheries interactions.

The cumulative prey curve for the Kemp’s ridley whole samples collected in

Virginia from 1991 to 2002 (n = 23) appears to reach an asymptote, suggesting that most

major prey items were encountered (Figure 39). Overall, whole Kemp’s ridley samples

had an average dry weight of 51.1 g (SD = 51.4 g, median = 33.2 g), and there were an

average of 12.3 prey items per whole sample (SD = 16.9, median = 7.0) and 6.3 large

prey items (horseshoe crab, crabs, fish, whelks, moon snails) (SD = 4.1, median = 6.0)

per whole sample. Linear regression indicated a positive relationship between Kemp’s

ridley SCL and total and large prey dry weights (Figure 40), but there was no significant

linear relationship between SCL and prey numbers (Figure 41).

Interannual diet

The cumulative prey curve for whole Kemp’s ridley samples collected in Virginia

during 2000-2002 (n = 18) appeared to be approaching an asymptote, but the 1991-1994

curve (n = 5) did because of the small number of samples (Figure 42). Due to these

sampling limitations, only general diet data (n = 59,1983-2002) and %F data for all

samples (n = 33, 1987-2002) were examined with correspondence analysis.

General prey data was compiled from 59 Kemp’s ridleys examined during 1983 to

2002 and was divided into five different year groups (1983-1984, 1987-1989, 1991-1994,

1999-2000, and 2001-2002) (Figure 43a). Crustaceans were dominant throughout the ridley data set. The rare occurrence of horseshoe crabs and fish may be an artifact of 88

FIGURE 39. Cumulative prey curve for whole Kemp’s ridley samples collected in Virginia during 1991 to 1994 and 2000 to 2002 (n = 23). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation.

Cumulative Number of Samples 89

FIGURE 40. Linear regressions of Kemp’s ridley straight carapace length (SCL) versus (a) total dry weight of Virginia whole samples and (b) dry weights of large prey items (horseshoe crab, crustaceans, fish, Atlantic moon snail) in Virginia whole samples (n = 23).

-109.951 + 4.44451X (a) R-Sq = 40.9 % ANOVA: p = 0.001 200 U) JZ D) CD 5 100

Q 03 O \— 0

20 30 40 50 SCL (cm)

Y = -106.254 + 4.28814X (b) R-Sq = 37.3 % ANOVA: p = 0.002 U) 2 00 £ O) CD §

100 — Q >> CD i_ CL CD CD L_ 03

20 30 40 50 SCL (cm) 90

FIGURE 41. Linear regressions of Kemp’s ridley straight carapace length (SCL) versus (a) total number of prey items in Virginia whole samples and (b) total number of large prey items (horseshoe crab, crustaceans, fish, Atlantic moon snail) in Virginia whole samples (n = 23). Note: neither of these regressions is significant.

V = 1.48853 + 0.293551 X (a) R-Sq = 1.6 % ANOVA: p = 0.564 90

80

70 CD CL 60 4— o 50 0 40 -Q E 30 3 20 • •

10 ** - 0

2 0 30 40 50 SCL (cm)

Y = 2.02812 + 0.117714X (b) R-Sq = 4.5 % ANOVA: p = 0.333 2 0 — 0 Q_ 0 0) CD _l 10 — 4o— 0 E 13 z o — 2 0 30 40 50 SCL (cm) FIGURE 42. Cumulative prey curves for whole Kemp’s ridley samples collected Virginia during (a) 1991 to 1994 (n = 5) and (b) 2000 to 2002 (n= 18). Samples were randomized ten times and numbers of prey types averaged. Error bars represent ± one standard deviation.

Cumulative Number of Samples

Cumulative Number of Samples 92

FIGURE 43. Frequency of occurrence of (a) major prey groups in all Kemp’s ridley diet data (n = 59) and (b) major prey items in all Kemp’s ridley samples (n = 33) collected in Virginia during 1983 to 2002. Note: Due to inconsistent sampling, these time series are not continuous.

(a)

Horseshoe crab Crustacean Mollusc Fish

■ 1983-1984 (n = 2) ■ 1987-1989 (n = 8) □ 1991-1994 (n = 10) □ 1999-2000 (n = 13)______B 2001-2002 (n = 26) ______93 small sample size, but neither has been reported previously in the literature as part of the diet in Virginia (Bellmund et al. 1987; Keinath et al. 1987; Lutcavage and Musick 1985).

Additionally, the recent addition of fish to ridley diet would not be surprising given trends observed in the loggerhead data. Correspondence analysis also suggests that crustaceans remained a key component of Kemp’s ridley diet during the years examined

(Figure 44a).

More specific data were acquired for those Kemp’s ridleys that were sampled

(Tables 12-14). Target prey species appeared to include horseshoe crabs, decapod crustaceans, fish, and moon snails. As with horseshoe crabs and fish, the appearance of hermit crabs, purse crabs, and moon snails in the 2000-2002 diet data may be due to the increased sampling effort (Figures 43b and 45). The disproportionate sampling may also be the cause for the separation of the recent samples from the 1987-1989 and 1991-1994 samples in the CA biplot seen in Figure 44b. At best, it can be concluded that Callinectes spp. (predominantly blue crabs) and Libinia spp. (spider crabs) were important components of ridley diet in Virginia from 1987 to 2002.

Only three ridleys were examined that had consumed fish. One ridley from 2000 had consumed croaker, and two from 2002 had consumed both croaker and menhaden.

Due to small sample sizes, Kemp’s ridley data were not examined by size class, sex, or season.

Interspecific Competition

The Schoener (1970) index was used to estimate diet overlap between loggerhead and Kemp’s ridleys in Virginia during 2001 and 2002. Earlier years were not examined 94

FIGURE 44. Biplots of year division and prey type principal components (PCs) for PCI and PC2 of a correspondence analysis using (a) %F values of general diet categories for all Kemp’s ridley data (n = 59) and (b) %F values of major prey items from all Kemp’s ridley samples (n = 33) collected in Virginia during 1983 to 2002. PCI and PC2 account for 81.54% and 100.0% of the data in (a) and (b), respectively. Number of samples is in parentheses.

(a)

1.0 -

Fish CM 0 .S E Mollusc c 3c 8. 99-00 (13) § o.o-

-0.5 0.0 0.5 1.0 Component 1

(b) 1 - Horseshoe crab a

1987-1989(6)

Fish, Spider p 8 e t o m b Hermit crab| ______200p-2gg2_(18)_ P u rse crab 8. o. CalTinecTes spp. a E Oo 1991-1^94 (9)

Lady^crab -1 - I I I -1 0 1 Component 1 TABLE 12: Frequency of occurrence of all prey items found in Kemp's ridley samples (whole and partial) from Virginia during 1987 to 2002 (n = 33).

1987- 1991- 2000- 1989 1994 2002 Overall (n=6) (n=9) (n=18) (n = 33) Chelicerates 16.7 0.0 5.6 6.1 Horseshoe crab Limulus polyphemus 16.7 0.0 5.6 6.1 Crustaceans 100.0 100.0 100.0 100.0 Decapods Blue crab Callinectes sapidus 83.3 100.0 72.2 81.8 Unidentified portunid Callinectes sp. 16.7 0.0 5.6 6.1 Rock crab Cancer irroratus 16.7 11.1 27.8 21.2 Spider crab spp. Libinia sp p . 50.0 33.3 66.7 54.5 Lady crab Ovalipes ocellatus 0.0 11.1 5.6 6.1 Hermit crab spp. Pagurus spp. 0.0 0.0 33.3 18.2 Purse crab Persephona mediterranea 0.0 0.0 44.4 24.2 Mantis shrimp Squilla empusa 0.0 0.0 5.6 3.0 Other Crustaceans Crab barnacle Chelonibia patula 16.7 0.0 0.0 3.0 Unidentified crustacean tissue - 33.3 22.2 61.1 45.5 Fish 0.0 0.0 16.7 9.1 Bony fish - 0.0 0.0 16.7 9.1 Molluscs 33.3 44.4 50.0 45.5 Bivalves Eastern American oyster Crassostrea virginica 0.0 0.0 5.6 3.0 Blue mussel Mytilus edulis 16.7 22.2 22.2 21.2 Unidentified bivalve - 16.7 11.1 11.1 12.1 Gastropods Cerith sp. Bittium sp. 0.0 0.0 11.1 6.1 Wentletrap Epitonium sp. 0.0 0.0 5.6 3.0 Eastern mud snail Hyanassa obsoleta 0.0 0.0 5.6 3.0 Three-line mud snail Ilyanassa trivittatus 0.0 11.1 22.2 15.2 Unidentified mud snail Hyanassa or Nassarius sp. 0.0 0.0 11.1 6.1 Mottled dog whelk Nassarius vibex 0.0 11.1 5.6 6.1 Atlantic moon snail Neverita duplicata 0.0 0.0 5.6 3.0 Unidentified gastropod - 0.0 0.0 5.6 3.0 Plants 16.7 77.8 55.6 54.5 Widgeon grass Ruppia maritima 16.7 0.0 16.7 12.1 Gulfweed sp. Sargassum sp. 0.0 11.1 5.6 6.1 Eelgrass Zostera marina 0.0 44.4 38.9 33.3 Unidentified marine plant - 0.0 44.4 22.2 24.2 Unidentified terrestrial plant -- 0.0 0.0 5.6 3.0 Miscellaneous 27.3 22.2 Rocks — 16.7 11.1 5.6 9.1 Unidentified tissue -- 33.3 11.1 16.7 18.2 Anthropogenic Items 0.0 0.0 5.6 3.0 Hook and line gear 0.0 0.0 5.6 3.0 TABLE 13. Percent occurrence (%F), percent dry weight (%W), percent number (%N), and percent index of relative importance (%IRI) of all prey items found in whole Kemp’s ridley gut samples from Virginia during 1991 to 1994 (n = 5). Note: colonial animals were given a count (number) of one.

%F %W %N %IRI Crustaceans 100.0 85.7 32.4 68.3 Decapods Blue crab Callinectes sapidus 100.0 29.5 20.6 65.6 Rock crab Cancer irroratus 20.0 7.0 2.9 2.6 Spider crab spp. Libinia spp. 40.0 14.5 5.9 10.7 Lady crab Ovalipes ocellatus 20.0 0.3 2.9 0.9 Other Crustaceans

Unidentified crustacean tissue - 40.0 34.3 -- Molluscs 80.0 0.9 67.6 31.7 Bivalves Blue mussel Mytilus edulis 40.0 0.1 8.8 4.7

Unidentified bivalve - 20.0 0.3 2.9 0.9 Gastropods Three-line mud snail Hyanassa trivittatus 20.0 0.4 55.9 14.7

Plants 100.0 0.4 -- Eelgrass Zostera marina 60.0 0.2 --

Unidentified marine plant - 80.0 0.3 - -

M iscellaneous 20.0 13.0 - -

Rocks - 20.0 0.0 --

Unidentified tissue — 20.0 12.9 — — TABLE 14. Percent occurrence (%F), percent dry weight (%W), percent number (%N), and percent index of relative importance (%IRI) of all prey items found in whole Kemp’s ridley gut samples from Virginia during 2000 to 2002 (n = 18). Note: colonial animals were given a count (number) of one.

%F %w %N %IRI Chelicerates 5.6 0.3 0.4 0.0 Horseshoe crab Limulus polyphemus 5.6 0.3 0.4 0.0 Crustaceans 100.0 94.0 51.2 85.6 Decapods Blue crab Callinectes sapidus 72.2 21.7 16.1 34.6 Unidentified portunid Callinectes sp. 5.6 0.1 4.4 0.3 Rock crab Cancer irroratus 27.8 0.8 3.2 1.4 Spider crab spp. Libinia spp. 66.7 40.1 12.9 44.7 Lady crab Ovalipes ocellatus 5.6 0.1 0.8 0.1 Hermit crab spp. Pagurus spp. 33.3 0.2 3.6 1.6 Purse crab Persephona mediterranea 44.4 7.1 9.7 9.5 Mantis shrimp Squilla empusa 5.6 0.0 0.4 0.0 Other Crustaceans Unidentified crustacean tissue - 61.1 23.7 - ~ Fish 16.7 3.0 2.0 0.5 Bony fish - 16.7 3.0 2.0 1.1 Molluscs 50.0 0.7 46.4 13.9 Bivalves Eastern American oyster Crassostrea virginica 5.6 0.0 0.4 0.0 Blue mussel Mytilus edulis 22.2 0.0 6.9 1.9 Unidentified bivalve - 11.1 0.0 0.8 0.1 Gastropods Cerith sp. Bittium sp. 11.1 0.0 5.6 0.8 Wentletrap Epitonium sp . 5.6 0.0 0.4 0.0 Eastern mud snail Hyanassa obsoleta 5.6 0.1 0.8 0.1 Three-line mud snail Hyanassa trivittatus 22.2 0.1 6.5 1.8 Unidentified mud snail Hyanassa or Nassarius sp. 11.1 0.2 2.4 0.4 Mottled dog whelk Nassarius vibex 5.6 0.2 20.6 1.5 Atlantic moon snail Neverita duplicata 5.6 0.0 0.8 0.1 Unidentified gastropod - 5.6 0.0 1.2 0.1 Plants 61.1 0.1 -- Widgeon grass Ruppia maritima 11.1 0.0 -- Gulfweed sp. Sargassum sp. 11.1 0.0 -- Eelgrass Zostera marina 33.3 0.0 -- Unidentified marine plant - 27.8 0.0 - - Unidentified terrestrial plant - 5.6 0.0 - -

Miscellaneous 16.7 2.0 —— Rocks — 5.6 0.0 Unidentified tissue — 16.7 2.0 Anthropogenic Items 5.6 Hook and line gear 5.6 gear missing 100

FIGURE 45. (a) Frequency of occurrence, (b) percent dry weight, (c) percent number, and (d) percent index of relative importance (IRI) in whole Kemp’s ridley gut samples collected in Virginia from 1991 to 1994 and 2000 to 2002 (n = 23).

1991-1994 n = 5) 02000 - 2002 n = 18)

1991 -1994 n = 5) □ 2000 - 2002 (n = 18) FIGURE 45 (Continued).

■ 1991 -1994 (n = 5) 0 2000-2002 (n = 18) 102 due to small numbers of ridley samples, and only whole samples were used in this calculation. Average dry weight values were used in the calculations, and die prey items included were horseshoe crabs, Callinectes spp., spider, hermit, rock, and purse crabs, bony fish, whelks, and Atlantic moon snails, and. The overlap calculated for 2001 (23 loggerheads, 7 ridleys) was 33.4%, and the 2002 (22 loggerheads, 9 ridleys) overlap was

58.0%. Combining the two years gave an overlap value of 46.1% between the loggerheads and Kemp’s ridleys examined. DISCUSSION

Loggerheads

Interannual diet

The diet of loggerhead sea turtles in Virginia appears to have shifted from horseshoe crab dominance during the early to mid- 1980s to blue crab dominance in the late 1980s and early 1990s, and then from a blue crab-dominated diet to a fish-dominated diet by the mid- 1990s and in 2001 to 2002. Additionally, spider crabs and rock crabs became more prevalent in the diet during the 1990s, and hermit crabs (and their moon snail shells) were observed more frequently in samples from 2001 and 2002. These diet shifts are not surprising given the declines in horseshoe crab and blue crab abundance.

No estimates are available for the Chesapeake Bay horseshoe crab population in any year (C. N. Shuster, pers. comm.) and records of horseshoe crab distribution and abundance within the Bay are fragmentary and largely unpublished (Shuster 1985); however, information can be gained from Virginia landings data (Figure 1). Reported

Virginia horseshoe crab landings within the Bay and its tributaries peaked in 1986 at about 80,000 pounds (approximately 20,000 to 31,000 animals depending on average weight), and they decreases to less than half of that in the next two years. Given that horseshoe crabs require about ten years to mature (Berkson and Shuster 1999), these diminished landings probably reflect a decrease in Chesapeake Bay, Virginia horseshoe

103 104

crab abundance. Continued fishing pressure in conjunction with this long time to

maturity likely diminished horseshoe crab numbers in the late 1980s and early 1990s,

possibly so much that loggerheads shifted their diets. Additionally, horseshoe crabs from

offshore areas may enter the Bay to spawn during May and June (Williams 1985; C. N.

Shuster, pers. comm.). As such, the explosion of the horseshoe crab bait fishery in the

late 1990s and the observed 50% population decline in the northwestern Atlantic

(Tanacredi 2001) undoubtedly affected spring and summer horseshoe crab numbers in the

Bay during more recent years.

Although the Virginia blue crab fishery remained intensive over the entire time

span of this study, the phase shift in blue crab abundance identified by Lipcius and

Stockhausen (2002) is a likely cause for the shift away from blue crab dominance in

loggerhead diet. The blue crab phase shift occurred around 1992, and loggerhead diet

had noticeably shifted by the mid- 1990s. Blue crabs were infrequent in the diet by 2001-

2002 and had a %IRI value of less than 3% during these years. The decreased

importance of blue crabs in the diet is likely because spawning stock, recruitment, female

size, and size at maturity values during 1992-2000 were all lower than those for 1985-

1991 (Lipcius and Stockhausen 2002). The blue crab population was certainly

compromised further by the Virginia blue crab landings during 1993-2001, which were at

or near 1985-1991 levels (Figure 3).

The apparent severe declines in horseshoe crab and blue crab populations in the

Chesapeake Bay seem to have caused loggerheads to shift their diet to one dominated by finfish. Sea turtles are not considered to be fast or agile enough to catch large quantities of fish (Bellmund et al. 1987; Shoop and Ruckdeschel 1982). Aside from the occasional 105 benthic inhabitant, fish found in sea turtle gut contents are assumed to be acquired either as discarded bycatch (Shaver 1991; Shoop and Ruckdeschel 1982; Tomas et al. 2001;

Youngkin and Wyneken, in press) or in nets (Bellmund et al. 1987). This dependence on fisheries to provide food may put the turtles at more risk for boat strikes or entanglement in nets.

Menhaden, croaker, seatrout (Cynoscion sp.), striped bass, and bluefish were the fish species most frequently encountered in Virginia loggerhead samples (Table 10), and all are commercially important in Virginia’s gillnet and poundnet fisheries (Mansfield et al. 2001, 2002a). The fish species composition and the fact that few turtles had consumed both fish and scavenging mud snails (four in 2001 and two in 2002), suggests that the turtles examined were feeding primarily on live and fresh dead fish from nets.

Sea turtles are known to enter the pound or “heart” of poundnets successfully and unharmed, and VIMS has relied on poundnetters to provide live turtles since the turtle program’s inception (Lutcavage 1981). Turtles are also at risk of becoming entangled and drowning in poundnet leaders, particularly those with large mesh (greater than 12- inch stretch) or stringers (Bellmund et al., 1987). One turtle that stranded in 2002 was found entangled in a section of gillnet (Virginia stranding data), and its stomach contained several squares of mesh in addition to bluefish, seatrout, and spot (Leiostomus xanthurus). Although some turtles entangle in nets and drown (Bellmund et al. 1987;

Mansfield et al. 2002b), the presence of fish in the gut alone does not definitively implicate a fishery-related death, and many of the fish encountered in this study were comprised of disarticulated bones and tissue in later stages of digestion. 106

Size-specific diet

Examination of the whole loggerhead samples suggests that a partial diet shift

occurs after benthic immature stage animals reach maturity. The 40.0-69.9 cm SCL

loggerhead diet differs somewhat from the 70.0-99.9 cm SCL diet. The turtles in the

latter size class (n = 106) were probably all immature, whereas the majority of the turtles

in the second size class (n = 22) were likely to be mature adults (TEWG 1998; Van

Buskirk and Crowder 1994). The data suggest that as Virginia loggerheads mature, fish

become less prevalent in the diet, whereas horseshoe crabs and whelks become more

important; however, this conclusion is made hesitantly due to the small sample size for

larger loggerheads. This sampling limitation occurred because only about 5% of

loggerheads that enter Virginia waters are adults (Musick and Limpus 1997; Figure 13).

Loggerheads from Cumberland Island, Georgia sampled during 1979 to 1999 also

appeared to eat less fish as they increased in size, but the amount of crustaceans

consumed increased with size (Youngkin and Wyneken, in press).

Sex-specific diet

There appears to be little or no partitioning of food resources between the male and female loggerheads examined. The same conclusion was reached in a comparison of

122 males and 217 females from Cumberland Island, Georgia (Youngkin and Wyneken, in press). 107

Interseasonal diet

Spring and summer loggerhead diet differed somewhat from fall diet with regards

to fish and horseshoe crab consumption and the species of crustaceans consumed, but the

sample size for the fall was only 14, compared to 47 for spring and 64 for summer. The

smaller sample size for fall months may have skewed the results. Youngkin and

Wyneken (in press) did not find any seasonal effects on the consumption of crustaceans

or fish by loggerheads stranding on Cumberland Island, but consumption of moon snails

and whelks increased significantly from spring (n = 105) to summer (n = 214) to fall (n =

48).

Kemp’s Ridleys

All of the Kemp’s ridleys examined were in the benthic immature size class, and

the majority of samples were collected during 2000 to 2002; however, it is clear that blue

crabs and spider crabs were key components of Virginia ridley diet from 1987 to 2002.

In comparison, the smaller Kemp’s ridleys found in New York appear to concentrate their

foraging efforts on slower-moving types of crabs, including spider crabs and rock crabs

(Burke et al. 1993; Burke et al. 1994; Morreale and Standora 1991). Ridleys from

Georgia (Frick and Mason 1998) and Texas (Shaver 1991) appear to consume a large

amount of blue crabs and other portunid (swimming) crabs, in addition to the slower,

walking crabs. Both the turtle size range and prey composition of the Virginia samples

are more similar to the ridleys examined in Georgia and Texas than to those from New

York. The appearance of hermit crabs, purse crabs, and fish in the Virginia 2000-2002

samples could be due to the small sample sizes in earlier years, or it may suggest that 108

Chesapeake Bay blue crab declines (Lipcius and Stockhausen 2002) are beginning to

affect ridley diet

Interspecific Competition

Sample sizes were only sufficient for 2001 and 2002 to compare the diet of

Virginia loggerheads and Kemp’s ridleys. The overlap for major prey items in 2001 was

33.4%, while the 2002 value was 58.0%. The discrepancy between the years is due in

part to items that were found in 2002 samples but not in 2001 samples: purse and rock

crabs for the loggerheads and horseshoe crabs and fish for the ridleys. The overlap for

the two years combined (46.1%) may be a better estimation of the overlap between

loggerhead and ridley diet in recent years. The actual foraging range overlap may be

minimal, as telemetry data suggest that loggerheads forage in deeper waters than ridleys

(Byles 1988; Keinath et al. 1987).

Concluding Remarks

In addition to establishing a baseline for the current diet of both loggerheads and

Kemp’s ridleys in Virginia, this long-term diet study suggests two shifts in loggerhead

diet that were likely due to fishery-induced prey declines. In general, Virginia’s

loggerheads and Kemp’s ridleys are benthic carnivores, but loggerheads appear to have become more opportunistic in response to prey declines. In order to better understand the effects of the horseshoe crab and blue crab fisheries on turtle diet, as well as to investigate potential interactions between turtles and frnfish fisheries, turtle gut contents should continue to be monitored. At the very least, necropsies should be conducted on 109 strandings whenever possible, and the contents of the entire digestive tract examined and recorded. Such information may be useful in the management of various fisheries in

Virginia, as well as in conservation schemes for these two protected sea turtle species. APPENDIX I. Contents of loggerhead samples from Maryland (n - 10) and North Carolina (n = 6).

Marine turtle (MT) numbers are expressed as "MT-year-month-day-turtle number by day". Letters after horseshoe crabs and blue crabs indicate sex of animals: F (female), M (male), U (unknown).

Turtle: MT-84-10-17-02 (VIMS Tags: G1055, G1056, K4004, K4005) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) CALLINECTES SAPIDUS BLUE CRAB (U) 3 37.7 14.5 LIMULUS POLYPHEMUS HORSESHOE CRAB (U) 1 26.7 8.7 - ROCKS — 0.3 0.1

Turtle: MT-87-06-11-01 (VIMS Tags: G1064, G1065) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces (2 samples)

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) LIMULUS POLYPHEMUS HORSESHOE CRAB (2 M) 2 453.9 142.3 CANCER IRRORATUS ROCK CRAB 1 114.5 39.4

- ROCKS - 14.7 14.2 GEUKENSIA DEMISSUS RIBBED MUSSEL 2 10.9 7.8 - UNIDENTIFIABLE TISSUE - 26.3 4.4 CALLINECTES SAPIDUS BLUE CRAB (U) 2 15.2 3.5

- UNIDENTIFIED BIVALVE 1 0.8 0.4 UNIDENTIFIED HERMIT CRAB 1 1.4 0.2 MYTILUS EDULIS BLUE MUSSEL 1 0.1 0.1

Turtle: MT-87-06-29-02 (VIMS Tags: K6431.K6432) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces (2 samples)

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) CALLINECTES SAPIDUS BLUE CRAB (1M, 2 MATURE F) 3 270.5 100.6 BREVOORTIA TYRANNUS MENHADEN 1 7.4 3.9 LIMULUS POLYPHEMUS HORSESHOE CRAB (U) 1 0.5 0.2

Turtle: MT-87-06-29-06 (VIMS Tags: K4568, K4569, PPN149) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces (2 samples)

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) CALLINECTES SAPIDUS BLUE CRAB (U) 5 296.8 76.0 LIMULUS POLYPHEMUS HORSESHOE CRAB (F) 1 180.5 47.0 MYTILUS EDULIS BLUE MUSSEL 1 0.2 0.1 ~~ UNIDENTIFIED MARINE PLANT — 0.1 0.1 Turtle: MT-87-07-17-03 (VIMS Tags: PPN143, PPN144) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g)

- ROCKS - 27.8 27.1 CALLINECTES SAPIDUS BLUE CRAB (U) 1 41.6 11.9 - UNIDENTIFIED BONY FISH 1 3.9 1.4 - UNIDENTIFIED BIVALVE 1 0.8 0.4 TELLINA SP. TELLIN SP. 1 0.1 0.1 ~ UNIDENTIFIED MARINE PLANT - 0.2 0.1

Turtle: MT-90-05-25-08 (VIMS Tags: QQB422, QQB424, QQB499, QQB500) Location: Mouth of Potomac River, MD (Western Bay) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) BREVOORTIA TYRANNUS, BONY FISH 6 192.5 5S.S CALLINECTES SAPIDUS BLUE CRAB (U) 1 16.2 4.0 NASSARIUS VIBEX MOTTLED DOG WHELK 4 2.7 1.8

UNIDENTIFIED MARINE PLANT - 1.6 0.1 - UNIDENTIFIED LEAVES - 0.2 0.1 - UNIDENTIFIED BIVALVES 8 0.2 0.1

- ROCK - 0.2 0.1 ZOSTERA MARINA EELGRASS - 0.1 0.1

Turtle: MT-90-12-09-01 Location: Hatteras, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) -- UNIDENTIFIABLE CRUSTACEAN TISSUE - 208.7 78.5 PARAUCHTHYS DENTATUS SUMMER FLOUNDER 1 147.5 43.1 LIBINIA EMARGINATA NINE-SPINED SPIDER CRAB 1 41.7 18.5 PERSEPHONA MEDITERRANEA PURSE CRAB 2 29.1 15.4 OVALIPES OCELLATUS LADY CRAB 9 28.3 10.5 CALLINECTES SAPIDUS BLUE CRAB (U) 3 9.9 5.4 PAGURUS POLLICARIS BROAD-CLAWED HERMIT CRAB 3 5.5 2.1 NEVERITA DUPLICATA ATLANTIC MOON SNAIL 1 1.5 1.3

- UNIDENTIFIED GASTROPOD 1 0.5 0.4 CHELONIBIA PATULA CRAB BARNACLE 4 0.5 0.4 BALANUS SP. ACORN BARNACLE 3 0.3 0.3 ZOSTERA MARINA EELGRASS ~ 0.2 0.1 - ROCKS - 0.2 0.1 — UNIDENTIFIED BIVALVE 1 0.3 0.1

Turtle: MT-91-05-30-01 (VIMS Tags: QQB478, QQB479, QQB374, QQB375) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) LIMULUS POLYPHEMUS HORSESHOE CRAB (U) 1 30.5 14.3 Turtle: MT-91-12-02-01 Location: Flounder Trawler, 30 miles offshore of Oregon Inlet, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g)

- UNIDENTIFIED BIVALVES 4 3.1 3.0 PAGURUS POLLICARIS BROAD-CLAWED HERMIT CRAB 5 7.3 2.4 NEVERITA DUPUCATA ATLANTIC MOON SNAIL 1 2.5 2.4 CALLINECTES SP. UNIDENTIFIABLE PORTUNID CRAB 1 3.4 1.7 ASABELLIDES OCULATA SOFT WORM TUBE 1 0.7 0.5 ILYANASSA TRIVITTATUS THREE-LINE MUD SNAIL 1 0.4 0.4 - UNIDENTIFIED TERRESTRIAL LEAF - 0.1 0.1 - ROCK — 0.1 0.1

Turtle: MT-91-12-09-02 Location: Flounder Trawler, off Frisco, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g)

- UNIDENTIFIED BIVALVES 9 17.5 15.0

- UNIDENTIFIABLE TISSUE - 32.0 13.9 CRASSOTREA VIRGINICA OYSTER 1 7.5 6.3 ARGOPECTEN SP. SCALLOPS 4 6.6 5.7 OVALIPES OCELLATUS LADY CRAB 1 9.5 4.1 ECHINARACHNIUS PARMA SAND DOLLAR 1 5.1 3.0 MERCENARIA MERCENARIA HARD CLAM 1 3.2 2.9

- ROCKS - 2.4 2.3 CREPIDULA SP. SLIPPER SHELL SP. 1 0.9 0.7 OLIVELLA SP. OLIVE SHELL 2 0.7 0.6

- ALUMINUM - 0.8 0.4 PAGURUS POLLICARIS BROAD-CLAWED HERMIT CRAB 1 2.0 0.3 SARGASSUM SP. GULFWEED -- 2.1 0.2 ASABELLIDES OCULATA SOFT WORM TUBE 1 0.5 0.2

- PLASTIC ~ 0.4 0.1

ZOSTERA MARINA EELGRASS - 1.4 0.1

- UNIDENTIFIED MARINE PLANT ~ 0.2 0.1 ANADARA SP. ARK SP. 1 0.1 0.1 — UNIDENTIFIED GASTROPOD 1 0.1 0.1

Turtle: MT-91-12-10-01 Location: Flounder Trawler, off Hatteras Inlet, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) ASABELLIDES OCULATA SOFT WORM TUBE 1 3.4 0.8

- UNIDENTIFIABLE TISSUE - 4.1 0.5 NEVERITA DUPUCATA ATLANTIC MOON SNAIL 1 0.3 0.2

~ UNIDENTIFIED BIVALVES 2 0.1 0.1 ZOSTERA MARINA EELGRASS — 0.1 0.1 Turtle: MT-91-12-18-01 Location: Hatteras, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME______COMMON NAME______# WET WT (g) DRY WT (g)

- UNIDENTIFIABLE TISSUE — 13.1 3.5 BUSYCON SP. 17 EGG CASES (CONNECTED) 1 21.8 2.8

- UNIDENTIFIED BIVALVES 2 0.8 0.5 ARGOPECTEN SP. SCALLOP 1 0.3 0.3 SARGASSUM SP. GULFWEED - 1.1 0.1 ZOSTERA MARINA EELGRASS - 0.1 0.1 — UNIDENTIFIED MARINE PLANT - 0.1 0.1

Turtle: MT-92-01-20-01 Location: Flounder Trawler, offshore of NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g)

- UNIDENTIFIABLE TISSUE 3.9 0.8 - ROCK 0.1 0.1

Turtle: MT-94-05-31-01 (VIMS Tags: QQM791, PPX807, PPX808, PPX816) Location: Solomon's Island, MD (Western Bay) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) MORONS SAXATILIS, STRIPED BASS, SEATROUT SP. 2 '' 165.3 66.8 UNIDENTIFIED MARINE PLANT 0.5 0.1

Turtle: MT-02-06-11-07 (VIMS Tags: SSB919, SSV642) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) LIMULUS POLYPHEMUS HORSESHOE CRAB (M) 1 88.8 25.5

CALLINECTES SAPIDUS BLUE CRAB (U) 1 6.3 1.4 ZOSTERA MARINA EELGRASS — 0.2 0.1

Turtle: MT-02-06-18-01 (VIMS Tags: XXF771, XXF772) Location: Poundnet, Potomac River Mouth, MD (Western Bay) Sample Type: Feces

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) BALANUS SP. ACORN BARNACLE 15 48.1 38.2 BREVOORTIA TYRANNUS, 2 MENHADEN, 1 UNIDENTIFIED UNKNOWN BONY FISH 3 29.2 10.9

- WOOD - 1.5 0.4

- UNIDENTIFIED MARINE PLANT - 0.1 0.1 — PLASTIC -- 0.1 0.1 APPENDIX II. Contents of Kemp's ridleys samples from North Carolina (ii = 2).

Marine turtle (MT) numbers are expressed as "MT-year-month-day-turtle number by day". Letters after horseshoe crabs and blue crabs indicate sex of animals: F (female), M (male), U (unknown).

Turtle: MT-90-12-09-02 Location: Frisco, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) - UNIDENTIFIABLE CRUSTACEAN TISSUE - 19.4 5.8 LIBINIA DUBIA SIX-SPINED SPIDER CRAB 1 9.5 3.0 CALLINECTES SP. SWIMMING CRAB 1 2.8 1.2 - UNIDENTIFIED BIVALVES 4 0.5 0.4 - UNIDENTIFIED GASTROPOD 1 0.1 0.1 - ROCKS - 0.2 0.1 ZOSTERA MARINA EELGRASS — 0.3 0.1

Turtle: MT-91-12-09-01 Location: Flounder Trawler, off Frisco, NC (Outer Banks) Sample Type: Whole Tract

SCIENTIFIC NAME COMMON NAME # WET WT (g) DRY WT (g) HEPATUS EPHELITICUS DOLLY VARDEN/CALICO CRAB 1 13.3 8.3 OVALIPES OCELLATUS LADY CRAB 2 8.4 5.1

- UNIDENTIFIABLE CRUSTACEAN TISSUE ~ 8.7 3.0 ZOSTERA MARINA EELGRASS - 0.8 0.1

- ROCKS - 0.2 0.1 — UNIDENTIFIED MARINE PLANT — 0.1 0.1 LITERATURE CITED

Atlantic States Marine Fisheries Commission. 1998. Fishery Management Report No. 32 of the Atlantic States Marine Fisheries Commission: Interstate Fishery Management Plan for the Horseshoe Crab. 52 pp.

Bellmund, S. A., J. A. Musick, R. C. Klinger, R. A. Byles, J. A. Keinath, and D. E. Barnard. 1987. Final Contract Report on the Ecology of Sea Turtles in Virginia. Virginia Institute of Marine Science Special Scientific Report No. 119. Virginia Institute of Marine Science, Gloucester Point, Virginia, 48 pp.

Berkson, J. and C. N. Shuster, Jr. 1999. The horseshoe crab: The battle for a true muMple-use resource. Fisheries 24(11): 6-11.

Bjomdal, K. A. 1985. Nutritional ecology of sea turtles. Copeia 1985(3): 736-751.

Bjomdal, K. A. 1997. Foraging ecology and nutrition of sea turtles. In :Lutz, P. L. and J. A. Musick (eds.). The Biology o f Sea Turtles. CRC Press, Boca Raton, FL, 199-231.

Bolten, A. B. and G. H. Balazs. 1995. Biology of the early pelagic stage - The “Lost Year.” In : Bjomdal, K. (ed.). Biology and Conservation o f Sea Turtles, Revised Edition. Smithsonian Institution Press, Washington, 579-581.

Botton, M. L. and J. W. Ropes. 1987. The horseshoe crab, Limulus polyphemus, fishery and resource in the United States. Marine Fisheries Review 49(3): 57-61.

Burke, V. J., E. A. Standora, and S. J. Morreale. 1993. Diet of juvenile Kemp’s ridley and loggerhead sea turtles from Long Island, New York. Copeia 1993(4): 1176- 1180.

Burke, V. J., S. J. Morreale, and E. A. Standora. 1994. Diet of the Kemp’s ridley sea turtle, Lepidochelys kempii, in New York waters. Fishery Bulletin 92(1): 26-32.

Byles, R. A. 1988. Behavior and Ecology of Sea Turtles from Chesapeake Bay, Virginia: A dissertation presented to the faculty of the School of Marine Science, The College ofWilliam and Mary in Virginia. Gloucester Point, VA, 112 pp.

115 116

Carr, A. 1986. Rips, FADS, and little loggerheads. Bioscience 36(2): 92-100.

Creaser, E. P. and H. C. Perkins. 1994. The distribution, food, and age of juvenile bluefish, Pomatomus saltatrix, in Maine. Fishery Bulletin 92(3): 494-508.

Creech, L. and P. E. Allman. 1997. Stomach and gastrointestinal contents of stranded Kemp’s ridley (Lepidochelys kempii) sea turtles in Georgia. In: Epperly, S. P., and J. Braun (eds.). Proceedings o f the Seventeenth Annual Symposium on Sea Turtle Conservation and Biology, NOAA Technical Memorandum NMFS- SEFSC-415, 176-169.

Coles, W. C. 1999. Aspects of the Biology of Sea Turtles in the Mid-Atlantic Bight: A dissertation presented to the faculty of the School of Marine Science, The College of William and Maiy in Virginia. Gloucester Point, VA, 153 pp.

Cortes, E. 1997. A critical review of methods of studying fish feeding based on analysis of stomach contents: Application to elasmobranch fishes. Can. J. Fish. Aquat. Sci. 54:726-738.

DiStefano, R. J., M. J. Roell, B. A. Wagner, and J. J. Decoske. 1994. Relative performances of four preservatives on fish and crayfish. Transactions o f the American Fisheries Society 123: 817-823.

Ferry, L. A. and G. M. Calliet. 1996. Sample size and data analysis: Are we characterizing and comparing diet properly?In :MacKinlay, D. and K Shearer (eds.). Feeding Ecology and Nutrition in Fish: Proceedings of the Symposium on the Feeding Ecology and Nutrition in Fish, International Congress on the Biology of Fishes, San Francisco, CA, 14-18 July 1996. American Fisheries Society, San Francisco, CA, 70-80.

Fisher, B. 2000. Update: Horseshoe crab research. Virginia Marine Resource Bulletin 32(2): 16-17.

Forbes, G. A. 1999. Diet sampling and diet component analysis. In: Eckert, K. L., K. A. Bjomdal, F. A. Abreu-Grobois, and M. Donnelly (eds.). Research and Management Techniques fo r the Conservation o f Sea Turtles. IUCN/SSC Marine Turtle Specialist Group, Publication No. 4, 144-148.

Frick, M. G. 1997. Lepidochelys kempi (Kemp’s ridley), Caretta caretta (loggerhead), and Malaclemys terrapin centrata (Carolina diamondback terrapin) diet and predation. Herpetological Review 28(3): 149.

Frick, M. G. and P. A. Mason. 1998. Lepidochelys kempi (Kemp’s ridley sea turtle) diet. Herpetological Review 29(3): 166-168. 117

Frick, M. G., K. L. Williams, and L. Pierrard. 2001. Summertime foraging and feeding by immature loggerhead sea turtles ( Caretta caretta) from Georgia. Chelonian Conservation and Biology 4(1): 178-181.

Gelsleichter, J., J. A. Musick, and S. Nichols. 1999. Food habits of the smooth dogfish, Mustelus canis, dusky shark, Carcharhinus obscurus, Atlantic sharpnose shark, Rhizoprionodon terraenovae, and the sand tiger, Carcharias taurus, from the northwest Atlantic Ocean. Environmental Biology o f Fishes 54: 205-217.

Godley, B. J., S.M. Smith, P. F. Clark, and J. D. Taylor. 1997. Molluscan and crustacean items in the diet of the loggerhead turtle, Caretta caretta (Linnaeus 1758) [Testudines: Chelonidae] in the eastern Mediterranean. Journal o f Molluscan. Studies 63: 474-476.

Hangay, G. and M. Dingley. 1985. Biological Museum Methods: Volume 1, Vertebrates. Academic Press, Sydney, Australia, pp. 17-19.

Hardy, J. D. 1962. Comments on the Atlantic ridley turtle, Lepidochelys olivacea kempi, in the Chesapeake Bay. Chesapeake Science 3: 217-220.

Hendrickson, J. R. 1980. The ecological strategies of sea turtles. American Zoologist 20: 597-608.

Hynes, H. B. N. 1950. The food of fresh-water sticklebacks ( Gasterosteus aculeatus and Pygosteus pungitius), with a review of methods used in studies of the food of fishes. Journal o f Animal Ecology 19: 36-58.

Hyslop, E. J. 1980. Stomach contents analysis - A review of methods and their application. Journal o f Fish Biology 17:411-429.

Keinath, J. A., D. E. Barnard, J. A. Musick, and B. A. Bell. 1994. Kemp’s ridley sea turtles from Virginia waters. In: Bjomdal, K. A., A. B. Bolten, D. A. Johnson, and P. J. Eliazar (Compilers). Proceedings o f the Fourteenth Annual Symposium on Sea Turtle Conservation and Biology, NOAA Technical Memorandum NMFS- SEFSC-351,70-73.

Keinath, J. A., J. A. Musick, and R. A. Byles. 1987. Aspects of the biology of Virginia’s sea turtles: 1979-1986. Virginia Journal o f Science 38(4): 329-336.

Kirkley, J. 1997. Virginia’s Commercial Fishing Industry: Its Economic Performance and Contributions. Special Report in Applied Marine Science and Ocean Engineering, No. 337. Virginia Institute of Marine Science, Gloucester Point, VA, 77 pp. 118

Lipcius, R. N. and W. T. Stockhausen. 2002. Concurrent decline of the spawning stock, recruitment, larval abundance, and size of blue crab in Chesapeake Bay. Marine Ecology Progress Series 226: 45-61.

Lutcavage, M. 1981. Status of Marine Turtles in Chesapeake Bay and Virginia Coastal Waters: A thesis presented to the faculty of the School of Marine Science, The College of William and Mary in Virginia. Gloucester Point, VA, 137 pp.

Lutcavage, M. and J. A. Musick. 1985. Aspects of the biology of sea turtles in Virginia. Copeia 38(4): 329-336.

MacDonald, J. S. and R. H. Green. 1983. Redundancy of variables used to describe the importance of prey species in fish diets.Canadian Journal o f Fisheries and Aquatic Sciences 40: 635-637.

Mansfield, K. L., J. A. Musick, and R. A. Pemberton. 2001. Characterization of the Chesapeake Bay Pound Net and Whelk Pot Fisheries and Their Potential Interactions with Marine Turtle Species. Final Report Submitted to: Northeast National Marine Fisheries Service. Virginia Institute of Marine Science, Gloucester Point, VA, 66 pp.

Mansfield, K. L., E. E. Seney, and J. A. Musick. 2002a. Evaluation of Sea Turtle Abundances, Mortality and Fisheries Interactions in Chesapeake Bay, Virginia. Final Draft Report Submitted to: Northeast National Marine Fisheries Service. Virginia Institute of Marine Science, Gloucester Point, VA, 102 pp.

Mansfield, K. L., E. E. Seney, M. A. Fagan, J. A. Musick, K. L. Frisch, and A. E. Knowles. 2002b. An Evaluation of Interactions between Sea Turtles and Poundnet Leaders in the Chesapeake Bay, Virginia. Final Draft Report Submitted to: Northeast National Marine Fisheries Service. Virginia Institute of Marine Science, Gloucester Point, VA, 118 pp.

Mayor, P. A., B. Phillips, and Z. M. Hillis-Starr. Results of the stomach content analysis on the juvenile hawksbill turtles of Buck Island National Monument, U.S.V.I. In: Epperly, S. P. and J. Braun (compilers).Proceedings of the Seventeenth Annual Sea Turtle Symposium. U.S. Department of Commerce. NOAA Technical Memorandum NMFS-SEFSC-415, pp. 244-246.

Mills, S. 2000. Virginia’s pot fisheries. Virginia Marine Resource Bulletin 32(2): 2-9.

Moore, D. S. and G. P. McCabe. 1998. Introduction to the Practice o f Statistics, Third Edition. W. H. Freeman and Company, New York, pp. 684-686.

Moore, I. A. and J. W. Moore. 1974. Food of shorthorn sculpin, Myoxocephalus scorpius, in the Cumberland Sound area of Baffin Island. Journal o f the Fisheries Research Board o f Canada 31: 355-359. 119

Morreale, S. J. and E. A. Standora. 1991. Habitat use and feeding activity of juvenile Kemp’s ridleys in inshore waters of the northeastern U.S. In: Salmon, M. and J. Wyneken (compilers). Proceedings o f the Eleventh Annual Workshop on Sea Turtle Biology and Conservation. National Oceanic and Atmospheric Administration, Washington D.C., 75-77.

Mortimer, J. A. 1995. Feeding ecology of sea turtles. In: Bjomdal, K. (ed.). Biology and Conservation o f Sea Turtles, Revised Edition. Smithsonian Institution Press, Washington, 103-109.

Musick, J. A. 1988. The Sea Turtles o f Virginia, Second Revised Edition. Virginia Sea Grant Program, Virginia Institute of Marine Science, Gloucester Point, VA, 22 pp.

Musick, J. A., R. Byles, R. Klinger, and S. Bellmund. 1984. Mortality and Behavior of Sea Turtles in the Chesapeake Bay: Summary Report for 1979 through 1983. Virginia Institute of Marine Science, Gloucester Point, VA, 100 pp.

Musick, J. A. and C. J. Limpus. 1997. Habitat utilization and migration in juvenile sea turtles. In: Lutz, P. L. and J. A. Musick (eds.). The Biology o f Sea Turtles. CRC Press, Boca Raton, FL, 137-163.

National Oceanic and Atmospheric Administration. 2002. Listing endangered and threatened wildlife and designating critical habitat; 90-day finding for a petition to reclassify the Northern and Florida Panhandle subpopulations of the loggerhead as distinct population segments with endangered species status to designate critical habitat. Federal Register 61(107): 38549-38461.

National Research Council, Committee on Sea Turtle Conservation. 1990. Decline o f the Sea Turtles: Causes and Prevention. National Academy Press, Washington, D.C., 247 pp.

Norrgard, J. W. and J. E. Graves. 1996. Determination of the natal origin of a juvenile loggerhead turtle {Caretta caretta ) population in Chesapeake Bay using mitochondrial DNA analysis. In: Bowen, B. W. and W. N. Witzell (eds.). Proceedings o f the International Symposium on Sea Turtle Conservation Genetics. NOAA Technical Memorandum NMFS-SEFSC-396, pp. 129-136.

Parker, R. R. 1963. Effects of formalin on the length and weight of fishes. Journal o f the Fisheries Research Board o f Canada 20(6): 1441-1455.

Pinkas, L. M., S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in Californian waters. California Fish and Game Fish Bulletin 152: 1-105. 120

Plotkin, P. 1989. Feeding ecology of the loggerhead sea turtle in the northwestern Gulf of Mexico. In: Eckert, S. A., K. L. Eckert, and T. H. Richardson (eds.). Proceedings o f the Ninth Annual Workshop on Sea Turtle Conservation and Biology. National Oceanic and Atmospheric Administration, Washington, D. C., 139-141.

Plotkin, P. L. 1996. Occurrence and diet of juvenile loggerhead sea turtles, Caretta caretta, in the northwestern Gulf of Mexico. Chelonian Conservation and Biology 2(1): 78-80.

Plotkin, P. T., M. K. Wicksten, and A. F. Amos. 1993. Feeding ecology of the loggerhead sea turtle Caretta caretta in the Northwestern Gulf of Mexico. Marine Biology 115: 1-15.

Richardson, J. I. 1991. Post-hatchling loggerhead turtles eat insects in Sargassum community. Marine Turtle Newsletter No. 55: 2-5.

Ruckdeschel, C. and C. R. Shoop. 1988. Gut contents of loggerheads: Findings, problems, and new questions. In: Schroeder, B. A. (Compiler). Proceedings of the Eighth Annual Workshop on Sea Turtle Conservation and NOAA Biology. Technical Memorandum NMFS-SEFSC-214, 97-98.

Schoener, T. W. 1970. Nonsynchronous spatial overlap of lizards in patchy environments. Ecology 51(3): 408-418.

Shaver, D. J. 1991. Feeding ecology of wild and head-started Kemp’s ridley sea turtles in South Texas waters. Journal o f Herpetology 25(3): 327-334.

Shoop, C. R. and C. Ruckdeschel. 1982. Increasing turtle standings in the southeast United States: A complicating factor. Biological Conservation 23: 213-215.

Shuster, C. N. 1985. Introductory remarks on the distribution and abundance of the American horseshoe crab, Limulus polyphemus spawning in the Chesapeake Bay area. In: Chase, V. (ed.). The Chesapeake: Prologue to the Future, Proceedings from the Chesapeake Bay Symposium. 34-38.

Stillwell, C. E. and N. E. Kohler. 1982. Food, feeding habits, and estimates of daily ration of the shortfin mako ( Isurus oxyrinchus) in the northwest Atlantic. Canadian Journal o f Fisheries and Aquatic Sciences 39:407-414.

Tanacredi, J. T. 2001. Horseshoe crabs imperiled? In: Tanacredi, J. T. (ed.). Limulus in the Limelight: A Species 350 Million Years in the Making and in Kluwer Peril? Academic/Plenum Publishers, New York, 7-14.

Tomas, J., F. J. Aznar, and J. A. Raga. 2001. Feeding ecology of the loggerhead turtle Caretta caretta in the Western Mediterranean. Journal o f Zoology 255: 525-532. 121

Turtle Expert Working Group. 1998. An Assessment of the Kemp’s Ridley {Lepidochelys kempii) and Loggerhead {Caretta caretta ) Sea Turtle Populations in the Western North Atlantic. U.S. Department of Commerce. NOAA Technical Memorandum NMFS-SEFSC-409, 96 pp.

Turtle Expert Working Group. 2000. Assessment Update for the Kemp’s Ridley and Loggerhead Sea Turtle Populations in the Western North Atlantic. U.S. Department of Commerce. NOAA Technical Memorandum NMFS-SEFSC-444, 115 pp.

U.S. Fish and Wildlife Service. 1998. Sea Turtles. Washington, D.C.

Van Buskirk, J. and L. B. Crowder. 1994. Life-history variation in marine turtles. Copeia 1994(1): 66-81.

Van Engel, W. A. 1958. The blue crab and its fishery in Chesapeake Bay: Part 1 - Reproduction, early development, growth, and migration. Commercial Fisheries Review 20(6): 6-17.

Van Nierop, M. M. and J. C. den Hartog. 1984. A study on the gut contents of five juvenile loggerhead turtles, Caretta caretta (Linnaeus) (Reptilia, Cheloniidae), from the south-eastern part of the North Atlantic Ocean, with emphasis on coelenterate identification. Zoologische Mededlingen 59(2): 35-54.

Virginia Marine Resources Commission. 1995. Regulation 4 VAC 20-240-10 ET. SEQ. From http://www.mrc.state.va.us/fr420.htm (accessed February 2003).

Virginia Marine Resources Commission. 2000. Regulation 4 VAC 20-900-12 ET. SEQ. From http://www.state.va.us/mrc/fr900.htm (accessed December 2001).

Virginia Marine Resources Commission. 2003. Blue crab, horseshoe crab, and whelk landings data. Newport News, VA.

Wallace, Jr., R. K. 1981. An assessment of diet-overlap indexes. Transactions o f the American Fisheries Society 110: 72-76.

Williams, J. P. 1985. Horseshoe crabs - A test of endurance. Chesapeake Bay Magazine 15(1): 27.

Youngin, D. and J. Wyneken. In press. A long-term dietary analysis of loggerhead sea turtles {Caretta caretta ) from Cumberland Island, GA, USA. In: X, X (Compilers). Proceedings o f the Twenty-First Annual Symposium on Sea Turtle Conservation and Biology, NOAA Technical Memorandum NMFS-SEFSC-XXX. PREY IDENTIFICATION REFERENCES

Abbott, T. R. 1990. Seashells. Gallery Books, New York, 176 pp.

Gosner, K. L. 1978. A Field Guide to the Atlantic Seashore from the Bay o f Fundy to Cape Hatteras. Houghton Mifflin Company, Boston, 329 pp.

Hurley, L. M. 1989. Field Guide to the Submerged Aquatic Vegetation o f Chesapeake Bay. U.S. Fish and Wildlife Service, Chesapeake Bay Estuary Program, Annapolis, MD, 51 pp.

Lippson, A. J. and R. L. Lippson. 1997.Life in the Chesapeake Bay, Second Edition. Johns Hopkins University Press, Baltimore, 294 pp.

Murdy, E. O., R. S. Birdsong, and J. A. Musick. 1997. Fishes o f Chesapeake Bay. Smithsonian Institution Press, Washington, D.C., 324 pp.

Williams, A. B. 1984. Shrimps, Lobsters, and Crabs o f the Atlantic Coast o f the Eastern United States, Maine to Florida. Smithsonian Institution Press, Washington, D.C., 550 pp.

122 VITA

Erin E. Seney

Bom in Lexington, Kentucky on October 1,1978. Graduated from Homer L. Ferguson High School in Newport News, Virginia in June 1996. Earned Bachelor of Arts in Biology, with a minor in Environmental Sciences, from the University of Virginia in May 2000. Entered Master of Science program at the College of William and Mary, School of Marine Science in August 2000.

123