Florida State University Libraries

Electronic Theses, Treatises and Dissertations The Graduate School

2014 Distribution and Abundance, Community Structure, and Trophic Ecology of Sharks and in the Florida Big Bend Cheston Thomas Peterson

Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected] FLORIDA STATE UNIVERSITY

COLLEGE OF ARTS AND SCIENCES

DISTRIBUTION AND ABUNDANCE, COMMUNITY STRUCTURE, AND TROPHIC

ECOLOGY OF SHARKS AND TELEOST FISHES IN THE FLORIDA BIG BEND

By

CHESTON THOMAS PETERSON

A Thesis submitted to the Department of Biological Science in partial fulfillment of the requirements for the degree of Master of Science

Degree Awarded: Spring Semester, 2014 Cheston Thomas Peterson defended this thesis on April 2, 2014. The members of the supervisory committee were:

Dean Grubbs Professor Directing Thesis

Joe Travis Committee Member

Walter Tschinkel Committee Member

The Graduate School has verified and approved the above-named committee members, and certifies that the thesis has been approved in accordance with university requirements.

ii

I dedicate this to my grandfather, Chester ‘Pop’ Onks, and my dear friend, Troy Billington.

“Fare thee well, fare thee well I love you more than words can tell Listen to the river sing sweet songs to rock my soul” - Robert Hunter, Brokedown Palace

“Flight of the seabirds, scattered like lost words Wheel to the storm and fly” - John Perry Barlow, Cassidy

iii

ACKNOWLEDGMENTS

Many people have contributed this work, and it is impossible to list them all. First, I would like to thank my advisor, mentor, and friend Dean Grubbs, whose remarkable intelligence, gentle guidance, and strong patience saw me through this entire process. I thank my committee members, Joe Travis and Walter Tschinkel, for their thoughtful comments and suggestions which improved both this research and my ability to think about science and statistics. Many people provided assistance in the field for this work, and without them this work would not have been possible. Specifically, I thank Travis Richards, Ale Mickle, Matthew Kolmann, Lisa Hollensead, Mariah Pfleger, Erica Holdridge, and Johanna Imhoff. Countless undergraduate and outside volunteers gave up their time for this work, and I am profoundly grateful for their help. This work was part of the Gulf Shark Pupping and Nursery (GulfSPAN) survey conducted by the NOAA/NMFS fisheries lab in Panama City, Florida. I thank Dana Bethea and John Carlson for annual funding of the survey, gillnets, and shark tags; all of which facilitated this research. I had a great group of colleagues with whom I could discuss my research with, both formally and informally: Travis Richards, Johanna Imhoff, Matthew Kolmann, Lisa Hollensead, Mollie Taylor, Robert Ellis, as well as the many EERDG participants. This work funded by through GulfSPAN funding provided by NOAA/NMFS, as well a research awards granted through Florida Grant and funded by the Guy Harvey Ocean Foundation, as well as a grant awarded to Dean Grubbs by the Guy Harvey Ocean Foundation. Finally, I thank my family and friends for their continued love and support.

iv

TABLE OF CONTENTS

List of Tables ...... vii List of Figures ...... viii Abstract ...... xv

1. ABUNDANCE, DISTRIBUTION, AND COMMUNTIYCOMPOSTIION OF SHARKS AND TELEOST FISHES IN THE FLORIDA BIG BEND ...... 1 1.1 Introduction ...... 1 1.2 Materials and Methods ...... 3 1.2.1 Study Area ...... 3 1.2.2 Survey Design ...... 3 1.2.3 Sampling ...... 4 1.2.4 Abundance and Distribution of Dominant Species ...... 5 1.2.5 Community Composition and Environmental Correlates ...... 6 1.3 Results ...... 8 1.3.1 General Results ...... 8 1.3.2 Abundance and Distribution of Dominant Species ...... 9 1.3.3 Community Composition and Environmental Correlates ...... 12 1.4 Discussion ...... 16 1.4.1 Ubiquitous Species...... 16 1.4.2 Sexual Segregation in Sharks...... 17 1.4.3 Regional Faunal Zones ...... 18 1.4.4 Community Composition and Environmental Correlates ...... 19 1.4.5 Conclusion ...... 22

2. STABLE ISOTOPE ECOLOGY AND TROPHIC STRUCTURE OF SHARK AND TELEOST FAUNAL ASSEMBLAGES IN THE BIG BEND ...... 70 2.1 Introduction ...... 70 2.2 Materials and Methods ...... 71 2.2.1 Sample Collection and Preparation ...... 71 2.2.2 Stable Isotope Analysis ...... 72 2.2.3 Statistical Analysis ...... 73 2.3 Results ...... 74 2.3.1 General Results for All Taxa ...... 74 2.3.2 Ontogenetic Shift in Dominant Taxa ...... 77 2.3.3 Regional Variation in Stable Isotope Values ...... 77 2.4 Discussion ...... 78 2.4.1 General Results for All Taxa ...... 78 2.4.2 Ontogenetic Shifts in Dominant Taxa ...... 82 2.4.3 Regional Variation in Stable Isotope Values ...... 83 2.4.1 Conclusion ...... 84

v

APPENDICES ...... 105

A. RANK ABUNDANCE CURVES AND OMISSION OF RARE SPECIES ...... 105

B. NMDS DIMENSIONS AND STRESS ...... 106

C. STATION COORDINATES AND CLUSTER DESIGNATIONS ...... 107

D. ACUC LETTER OF APPROVAL ...... 112

REFERENCES ...... 113

BIOGRAPHICAL SKETCH ...... 122

vi

LIST OF TABLES

1 Summary of environmental parameters of sites sampled within each survey region. Means and standard deviations are shown with ranges in parentheses. Mean depth included maximum and minimum depths of each site. For salinity, temperature, and dissolved oxygen, each line represents surface, mid, and bottom values from top to bottom, respectively ...... 23

2 Summary of elasmobranch fishes; longline CPUE ([catch/100 hooks]/hours soaked *100), gillnet CPUE (catch/hours soaked); SE = standard error; m = male, f = female; YOY = young-of-the-year, Juv = juvenile, Mat = mature, FL = fork length, TL = total length, DW = disc width, Individuals for which sex and/or life stage were not recorded were omitted for calculation of sex and maturity ratios ...... 24

3 Summary of bony fishes in taxonomic order by family; LL CPUE ([catch/100 hooks]/hours soaked*100), GN CPUE (catch/hours soaked); SE=standard error; FL=fork length...... 25

4 Diversity indices of each survey region ...... 27

5 Results of indicator species of gillnet cluster analysis. Data shown for the cluster for which each species’ indicator value was highest. Species in bold were significant indicators...... 28

6 Results of indicator species analysis of longline clusters. Species in bold were significant indicators ...... 30

7 Summary of elasmobranch stable isotope values; SE = standard error; YOY = young- of-the-year, Juv = juvenile, Mat = mature ...... 86

8 Summary of bony fishes stable isotope values; SE = standard error ...... 87

9 Results of Tukey’s HSD tests on mean isotope values of elasmobranchs. Cownose rays were not included in the δ15N test due to their depleted δ13C relative to other taxa. Pairwise comparisons were significant at p < 0.05 ...... 90

10 Sample-size corrected standard ellipse area (SEAC) of the four dominant shark species and both ...... 91

11 Pooled regional stable isotope values of dominant species and results of Tukey’s HSD test. Pairwise comparisons were significant at p < 0.05 ...... 91

12 Station number by gear, latitude and longitude, and cluster designation...... 107

vii

LIST OF FIGURES

1 USGS GIS layer of seagrass coverage in the Big Bend ...... 31

2 All sampling stations in the Big Bend. General survey regions are labeled. Dark circles represent longline sets and triangles represent gillnet sets ...... 32

3 Primary (a) and secondary (b) bottom types of sampling sites within survey regions ...... 33

4 Total number of elasmobranch and teleost species in gillnet and on longline ...... 34

5 Shark size selection of each gear . Longline range was 28.5-302.0 cm TL. Gillnet range was 32.0-167 cm TL ...... 34

6 Length frequencies of blacknose sharks captured on longline in each survey region ...... 35

7 CPUE of blacknose sharks (Carcharhinus acronotus) in gillnet [catch/hours soaked] (top) and on longline [(catch/100 hooks)/hours soaked * 100] (bottom). Means for each life stage within each survey region are shown, error bars indicate standard error. No mature blacknose sharks were captured in gillnet ...... 36

8 Map of gillnet and longline CPUE of blacknose sharks (Carcharhinus acronotus) ...... 37

9 Length frequencies of blacktip sharks (Carcharhinus limbatus) captured on longline in each survey region ...... 38

10 CPUE of blacktip sharks (Carcharhinus limbatus) in gillnet [catch/hours soaked] (top) and on longline [(catch/100 hooks)/hours soaked * 100] (bottom). Means for each life stage within each survey region are shown, error bars indicate standard error ...... 39

11 Map of gillnet and longline CPUE of blacktip sharks (Carcharhinus limbatus) ...... 40

12 Length frequencies of Atlantic sharpnose sharks (Rhizoprionodon terraenovae) captured on longline in each survey region ...... 41

13 CPUE of Atlantic sharpnose sharks (Rhizoprionodon terraenovae) in gillnet [catch/hours soaked] (top) and on longline [(catch/100 hooks)/hours soaked * 100] (bottom). Means for each life stage within each survey region are shown, error bars indicate standard error ...... 42

14 Map of gillnet longline CPUE of Atlantic sharpnose sharks (Rhizoprionodon terraenovae) ...... 43

15 Length frequencies bonnethead sharks (Sphyrna tiburo) captured in gillnet in each survey region ...... 44

viii

16 CPUE of Bonnethead sharks (Sphyrna tiburo) in gillnet [catch/hours soaked]. Means for each life stage within each survey region are shown, error bars indicate standard error ..45

17 Map of gillnet CPUE of bonnethead sharks (Sphyrna tiburo) ...... 45

18 Sex ratios of each life stage for blacknose, blacktip, Atlantic sharpnose, and bonnethead sharks ...... 46

19 CPUE of (a) Hardhead (Arius felis) and (b) ( marinus) in gillnet and on longline. Longline CPUE = [(catch/100 hooks)/hours soaked * 100] and gillnet CPUE = [catch/hours soaked]. Means for each gear type within each survey region are shown, error bars indicate standard error ...... 47

20 Maps of gillnet and longline CPUE of (Arius felis, above) and gafftopsail catfish (Bagre marinus, below) ...... 48

21 Per-set species richness, Shannon diversity, Simpson’s diversity, and Pielou’s evenness (top to bottom, respectively) of each survey region for gillnet (left) and longline (right) ...... 49

22 Gillnet dendrogram showing 7 clusters at 92% similarity. Numbers identify individual stations, which are shown with latitude and longitude coordinates in Appendix C...... 50

23 From top to bottom, respectively: max depth, surface salinity, water clarity, and latitude (bottom) of sampling stations in each gillnet cluster ...... 51

24 All gillnet sampling sites represented by symbols according to cluster ...... 52

25 NMDS plot of gillnet data with 75% standard deviation confidence ellipses around the centroid of each cluster and lines connecting each point to its assigned cluster ...... 53

26 NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster. Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern ); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel) ...... 54

ix

27 NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and directions of correlated environmental vectors (p < 0.05). Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel) ...... 55

28 NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of water clarity (cm). Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species. Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel)...... 56

29 NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of surface salinity. Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru

x

(American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel)...... 57

30 NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of latitude. Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel)...... 58

31 Longline dendrogram showing four clusters at 85% similarity. Numbers identify individual stations, which are shown with latitude and longitude coordinates in Appendix C...... 59

32 Minimum depth (top-left), max (top right), water clarity (mid-left), bottom salinity (mid- right), depth difference (bottom-left), and latitude (bottom-right) of sampling stations in each longline cluster ...... 60

33 All longline sampling sites represented by symbols according to cluster ...... 61

34 NMDS plot of longline data with 75% standard deviation confidence ellipses around the centroid of each cluster and lines connecting each point to its assigned cluster ...... 62

35 NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia)...... 63

36 NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and directions of correlated environmental vectors (p < 0.05). Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR,

xi

Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia) .64

37 NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of minimum depth (m). Four letter codes indicate weighted average points for each species. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia) ...... 65

38 NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of water clarity (cm). Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia)...... 66

39 NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of bottom salinity. Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species. Species codes (N=10): BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia) ...... 67

40 NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of depth difference (m). Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse

xii

shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia) ...... 68

41 Correlation of longline set water clarity (cm) and maximum (top) and minimum depth (m) (bottom) ...... 69

42 Mean δ13C and δ15N values of elasmobranchs. Bars denote standard error. Above, all species and life stages of dominant species. Below, RBON and GMIC removed and life stages pooled for dominant species (YOY excluded). Note differences in scales. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); DAME, Dasyatis americana (southern stingray); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); GMIC, Gymnura micrura (smooth ); NBRE, Negaprion brevirostris (lemon shark); RBON, Rhinoptera bonasus (cownose ray); RTER; Rhizoprionodon terraenovae (Atlantic sharpnose shark); SMOK, Sphyrna mokarran ( shark); STIB, Sphyrna tiburo (bonnethead shark) ...... 92

43 Individual δ13C and δ15N values of elasmobranchs. Juvenile and adult life stages were pooled, and YOY excluded. Mean values with error bars representing standard deviation shown for primary producers. Values for phytoplankton, epiphytes, macroalgae, and Halodule were adapted from Moncreiff and Sullivan (2002). Data for Thalassia and Syringodium were provided by Chanton and Harper (unpublished data). Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); DAME, Dasyatis americana (southern stingray); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); GMIC, Gymnura micrura (smooth butterfly ray); NBRE, Negaprion brevirostris (lemon shark); RBON, Rhinoptera bonasus (cownose ray); RTER; Rhizoprionodon terraenovae (Atlantic sharpnose shark); SMOK, Sphyrna mokarran (great hammerhead shark); STIB, Sphyrna tiburo (bonnethead shark) ...... 93

44 Mean δ13C and δ15N values of teleost fishes (N > 10). Error bars denote standard error. Only species with N > 5 are shown. Species codes: AFEL, felis (hardhead catfish); BMAR, Bagre marinus (gafftopsail catfish); BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CARC, Calamus arctifrons (grass porgy); CSTR, Centropristis striata (black sea bass); CNEB, Cynoscion nebulosus (spotted seatrout); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); OCHR, Orthopristis chrysoptera (pigfish); PPAR, Peprilus paru (American harvestfish); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel)...... 94

45 Lipid-normalized (Post, 2002) mean δ13C and δ15N values of teleost fishes (N > 10). Error bars denote standard error for fishes. Only species with N > 5 are shown. Mean values with error bars representing standard deviation shown for primary producers. Values for phytoplankton, epiphytes, macroalgae, and Halodule were adapted from Moncreiff and Sullivan (2002). Data for Thalassia and Syringodium were provided by Chanton and Harper (unpublished data). Species codes: AFEL, Ariopsis felis (hardhead

xiii

catfish); BMAR, Bagre marinus (gafftopsail catfish); BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CARC, Calamus arctifrons (grass porgy); CSTR, Centropristis striata (black sea bass); CNEB, Cynoscion nebulosus (spotted seatrout); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); OCHR, Orthopristis chrysoptera (pigfish); PPAR, Peprilus paru (American harvestfish); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel)...... 95

46 Convex hull and Bayesian standard ellipses (SEAC) of dominant shark species. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark) ...... 96

47 Convex hull and Bayesian standard ellipses (SEAC) of ariid catfishes. Species codes: AFEL, Ariopsis felis (hardhead catfish); BMAR, Bagre marinus (gafftopsail catfish) ....97

48 δ15N values of each life stage of dominant shark species. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark) ...... 98

49 δ13C values of each life stage of dominant shark species. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark) ...... 99

50 Hardhead catfish (Ariopsis felis) δ13C and δ15N values by fork length (FL) ...... 100

51 Gafftopsail catfish (Bagre marinus) δ13C and δ15N values by fork length (FL) ...... 101

52 δ13C and δ15N values of all taxa by region. YOY sharks were removed for this figure ..102

53 Pooled regional means of δ13C and δ15N of dominant taxa (above) and both species of ariid catfishes, individually (below). Bars denote standard error. Gray shapes = gafftopsail catfish (Bagre marinus), white shapes = hardhead catfish (Ariopsis felis). ..103

54 δ13C and δ15N values of all life stages of dominant shark species by region. Species codes: CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark) ...... 104

55 Rank abundance (log abundance) curves for each gear type. Arrow indicates cutoff for species to be considered rare ...... 105

56 Scree plots showing stress at increasing numbers of dimensions in NMDS of data sets for both gear types ...... 106

xiv

ABSTRACT

Community structure and trophic ecology of sharks and large teleost fishes in seagrass beds of the Florida Big Bend were investigated using fishery-independent longline and gillnet surveys and stable isotope analyses. The Big Bend is inhabited by at least 14 species of elasmobranch and 56 species of teleost fishes during the summer. Assemblages of these fauna are spatially variable, and five species dominate. Community structure was analyzed using a combination of cluster analysis, indicator species analysis, and non-metric multidimensional scaling (NMDS). These analyses suggested community composition of fishes in the Big Bend is correlated with water clarity, salinity, and depth. Patterns of inshore and offshore species assemblages were common throughout the Big Bend. There were two distinct faunal zones: one in the central Big Bend characterized by high relative abundance of blacktip sharks (Carcharhinus limbatus) in turbid water with moderate salinity, and a second in the southern Big Bend characterized by high relative abundance of blacknose sharks (Carcharhinus acronotus) in water of high clarity and salinity. High catch rates of young-of-the-year and juvenile blacktip and blacknose sharks in these areas suggest the central and southern Big Bend may act as nursery for these two species, respectively. Carbon and nitrogen stable isotope analyses were used to infer relative trophic structure of these taxa and the potential for regional variation in trophic patterns. Stable isotope analyses suggest this system is trophically diverse, with considerable isotopic overlap across many taxa. These fishes appear to be supported through multiple channels of primary production potentially dominated by epiphytic microalgae and/or macroalgae. Isotopic ontogenetic were not evident in dominant taxa, with exception to weak relationships of stable isotopes and length in two species of marine catfishes. Comparisons of regional stable isotope values of the dominant species suggest the southern Big Bend is isotopically distinct in terms of the balance of primary production and trophic structure. The results of this study suggest an ecological gradient in the Big Bend, culminating in a distinct southern faunal zone; and based on the results of this work I hypothesize patterns of community composition and trophic structure in the Florida Big Bend are related to varying levels of river influence across the habitat.

xv

CHAPTER ONE

ABUNDANCE, DISTRIBUTION, AND COMMUNITY COMPOSITION OF ELASMOBRANCH AND TELEOST FISHES IN THE FLORIDA BIG BEND

1.1 Introduction

Understanding distribution and abundance of species is among the central themes of ecology, and critical to effective resource management. Baseline data of faunal assemblages are essential in evaluating the effects of natural and anthropogenic disturbances. Instances in which pristine coastal remain both unspoiled from anthropogenic influence and understudied in the scientific community are rare in developed countries such as the United States; however the northwest coastline of Florida (henceforth called the Big Bend) is an example of such habitat. The Big Bend, considered to be the coastline extending west from Apalachee Bay and south to the Anclote Keys, is bordered by one of the state’s largest seagrass beds, second only to that of Florida Bay and exceeding that of Tampa Bay (Zieman and Zieman, 1989). Like the Florida Bay seagrass habitat, which is largely protected through its proximity to the Everglades National Park, the Big Bend seagrass habitat is thought to be essentially pristine (Zieman and Zieman, 1989). Although the smaller teleost fish and macroinvertebrate communities of the Big Bend (particularly Apalachee Bay) have been studied in detail by R.J. Livingston and colleagues (e.g. R.J Livingston, 1982; Dugan and Livingston, 1982), the larger teleost and shark community remains to be documented. These taxa are all but absent in many reviews of the fauna of Big Bend seagrass habitats (Zieman and Zieman, 1989; and references therein). Additionally, studies focused on sharks of the Florida Gulf have generally been restricted to the coastline south of the Suwannee River inlet (especially Tampa Bay and south to Florida Bay) and the coastline west of Apalachee Bay (especially Apalachicola Bay and the surrounding area) (e.g. Carlson, 2002; Heupel and Simpfendorfer, 2002; Heupel and Hueter, 2002; Bethea et al., 2004; Heupel, 2007; Hueter and Tyminski, 2007).

1

The relationship between a given species and its habitat use is often complex, especially for large, mobile taxa in marine environments (Froeschke et al., 2010). Multivariate methods are often employed to investigate the relationships of environmental variables with species distributions and catch data (Grubbs and Musick, 2007; Drymon et al., 2010; Froeschke et al., 2010; Drymon et al., 2013); and previous studies have found a suite of environmental factors to be potentially influential in shark distributions (Froeschke et al., 2010; Knip et al., 2010). These include water quality parameters and other abiotic factors such as salinity, temperature, dissolved oxygen, depth, distance to tidal inlets, and day length (e.g. Morrissey and Gruber, 1993; Abel et al., 2007; Grubbs and Musick, 2007; Heithaus et al., 2009; Drymon et al., 2010; Froeschke et al., 2010), as well as biotic factors such as inter-specific competition and predator presence (Heupel and Hueter, 2002; Bethea et al., 2004). Variation also occurs at temporal and spatial scales, and these are often correlated (Grubbs et al., 2007; Heupel, 2007; Froeschke et al., 2010; Knip et al., 2010). It is important to document the elasmobranch and large teleost assemblages in habitats such as the Big Bend to provide a baseline for future management actions and evaluation of effects of natural and anthropogenic disturbances. An understanding of the environmental parameters that drive spatially-variable patterns of community composition is also important in the delineation of essential habitat for these taxa, as the susceptibility to problems related to fishing pressure and habitat loss of elasmobranch fishes is well documented in primary literature (Musick et al., 2000). The objectives of this study are to document the elasmobranch (primarily shark) species, as well as large species, that occur in the Florida Big Bend, and to describe spatial variations in community composition and relate them to environmental correlates. These data will provide a baseline for measures of relative abundance and distribution of these species for future management and analytical work and will act as a first step in delineating essential habitat for some species in the Big Bend. A more detailed analysis of the most abundant species will provide additional insight of ontogenetic differences in habitat use and the potential of these areas as nursery habitat, based on the high occurrence of young-of-the- year and juvenile life stages relative to adults over multiple years (Heupel et al., 2007). I will use a combination of traditional methods to examine relative abundance and distribution and multivariate techniques to explore environmental correlates to community composition and structure.

2

1.2 Materials and Methods 1.2.1 Study Area The Big Bend seagrass bed along the northwest coast of Florida, extending from St. Marks to Tarpon Springs, is the second largest in the state at approximately 3,000 km2 (Iverson and Bittaker, 1986). Zieman and Zieman (1989) reviewed the ecology and physio-chemical parameters of the Big Bend seagrass bed. Three species of seagrass dominate the habitat: turtle grass (Thalassia testudinum), manatee grass (Syringodium filiforme), and to a lesser extent shoal grass (Halodule wrighti). A variety of epiphytic algal species and macroalgal species are also present, and biomass of macroalgae can be quite high (Zieman and Zieman, 1989). Zieman and Zieman (1989) found the area to be characterized by a gentle slope of the wide continental shelf and very little wave action, despite the lack of major barrier islands or other physical obstructions between the greater and the nearshore environment. Temperature during summer is fairly consistent throughout the habitat; and salinity varies locally with river influence and precipitation, with peak precipitation in the summer months. Zieman and Zieman (1989) also reported a gradient of water clarity and river influence in the Big Bend, in which the southern portion is characterized by higher water clarity and therefore a higher maximum depth of the seaward edge of the seagrass bed due to greater light penetration (approximately 7.0 meters in the south, compared to 4.5 meters in the north).

1.2.2 Survey Design The survey was designed employing a spatially-balanced, random sampling design using the function GRTS in the spsurvey software package for the R console (Stevens and Olsen, 2004; Kincaid and Olsen, 2012) in the Comprehensive R Archive Network site (CRAN)-http://cran.r- project.org/ (R Core Development Team, 2013). This program was used to generate 140 stations across the sampling area, which was constrained to a USGS GIS layer of the Big Bend seagrass bed, for each sampling year (Figure 1, Mattson et al., 2006). At least 40 stations were haphazardly chosen for sampling each year based on logistics, habitat quality, depth profile, and proximity to other sampling stations. The survey area was divided into four regions based on geographical features and sampling logistics. Regions, beginning northwest and moving southeast, were St. Marks, from Apalachee Bay to the Steinhatchee River; Steinhatchee, from the Steinhatchee River to Cedar Key; Crystal River, from Waccasassa Bay to Homosassa Bay; and

3

Hernando, from Homosassa Bay to Anclote Key. It was attempted to sample evenly across the entire survey area, and at least 10 stations were sampled within each survey region per year. All sampling stations are depicted in Figure 2.

1.2.3. Sampling Sampling was conducted during the summer months (June-August, primarily July) from 2009 to 2012, using a 26-foot research vessel specifically outfitted for the survey. Two gear types were used: an experimental gillnet and an experimental longline; both of which were fished concurrently for one hour at each station sampled. The experimental gillnet was three meters deep and 183 meters long and consisted of six panels of 0.5 inch incremental mesh sizes ranging from 3.5 inches to 5.5 inches. The experimental longline consisted of an approximately 1,500 meter mainline of 4.0 mm monofilament, anchored and marked with buoys at both ends. Each line held 100 gangions composed of four 25-hook sections separated by buoys with a unique hook size in each section. Four sizes of Mustad circle hooks (10/0, 12/0, 14/0, and 16/0) were used to minimize size-selection bias and allow capture of all possible sharks present from the smallest neonates to the largest adults. Each gangion began with a stainless steel tuna clip attached to two-meters of monofilament (136 kg test for 10/0, 12/0 and 14/0 hooks and 318 kg test for the 16/0 hooks). The monofilament was crimped to an 8/0 stainless steel barrel swivel followed by a one-meter section of 7x7 stainless-steel aircraft cable (1.8mm for 10/0, 12/0, and 14/0 hooks; 2.2mm for 16/0 hooks). Each gangion was terminated by a circle hook crimped to the steel cable. All 10/0, 12/0, and 14/0 hooks were baited with either Atlantic mackerel (Scomber scombrus) or Spanish mackerel (Scomberomorus maculatus); and bait was alternated on 16/0 hooks between mackerel and pieces of small carcharhiniform sharks, which were collected from survey mortalities. All fishes captured were brought aboard, and all fishes caught alive were released. Each specimen was identified to species and measured. Sharks were measured to precaudal (PCL), fork (FL), and total (TL) lengths, batoids were measured to disc width (DW), and teleosts were measured to fork length. Elasmobranchs were sexed, and maturity stage was recorded. Sharks caught in good condition were tagged and released with nylon or stainless steel dart tags with unique identification numbers and contact information for NOAA’s Southeast Fisheries Science

4

Center in Panama City, Florida. White muscle biopsies for carbon and nitrogen stable isotope analysis were collected from up to ten specimens from each species per survey region. Environmental parameters were recorded for each longline and gillnet set. Salinity, temperature (°C), and dissolved oxygen (ppm) were measured at surface, mid-water column, and bottom depths using either YSI 85 or YSI Pro 2030 handheld water quality meters. Water clarity (cm) was measured using a secchi disc. Maximum and minimum depths (m) were recorded for each longline and gillnet set using on-board sonar. Bottom type was also recorded based on qualitative observation of on-board sonar display and direct observation when possible.

1.2.4 Abundance and Distribution of Dominant Species I analyzed catch data for each gear type separately, due to the selectivity differences between gillnets and longlines. Each gillnet or longline set was treated as a single statistical unit. I more closely evaluated the relative abundance and distribution of the four most abundant shark species, as well as two species of ariid catfishes, by comparing per-set catch-per-unit-effort (CPUE) across years and regions using Kruskal-Wallis one-way ANOVA on ranks with a mean- rank adjustment for ties (Kruskal and Wallis, 1952). Longline CPUE was calculated as catch per 100 hook-hour [(catch/100 hooks)/hours soaked * 100] and gillnet CPUE was calculated as catch per net-hour. I also compared CPUE of individual life stages (young-of-the-year, juvenile, and mature) of each shark species across years and regions. In the case of significance (p < 0.05) I conducted a post-hoc multiple comparison test using Dunn’s method. I evaluated departures of sex ratios from an expected 1:1 for all four species using Chi-squared tests. Abundances of the remaining elasmobranch species were generally too low to provide for meaningful statistical analyses. All analyses were conducted in the R statistical console using the software packages coin v.1.0-23 and multcomp v.1.3-1 (Hothorn et al., 2006; Hothorn et al., 2008, R Core Development Team, 2013). Figures were created using the R software packages lattice v.0.20-24 (Deepayan, 2008; R Core Development Team, 2013) and ggplot2 v.0.9.3.1 (Wickham, 2009; R Core Development Team, 2013).

5

1.2.5 Community Composition and Environmental Correlates To assess course-scale regional variation of community composition, I first compared a suite of common diversity indices (species richness, Shannon diversity, Simpson’s diversity, and Pielou’s evenness) calculated for each sampling set across the four survey regions (McCune and Grace, 2002). Temporal variability was also assessed comparing the same suite of diversity indices calculated for each survey year. These comparisons were made for each gear type. Due to departures from normality, I used Kruskal-Wallis analysis of variance (ANOVA) on mean ranks to make comparisons. If significant effects were found (p < 0.05), I conducted a post-hoc multiple comparison test using Dunn’s method. To further investigate community structure I used a combination of hierarchical cluster analysis, non-metric multidimensional scaling (NMDS), indicator species analysis, and environmental fitting, following a general approach similar to those outlined in Field et al. (1982) and Clarke (1993). I chose NMDS as an ordination method as there are not assumptions in the relationships among variables (e.g. the assumption of linearity in principal components analysis) and any measure of distance between samples can be used (Clarke, 1993; McCune and Grace, 2002). These analyses were conducted using the R software packages vegan v.2.0-9 (Oksanen et al., 2013; R Core Development Team, 2013), labdsv v.1.6-1 (Roberts, 2013; R Core Development Team, 2013), cluster v.1.14.4 (Maechler et al., 2013; R Core Development Team, 2013), and permute v.0.8-0 (Simpson, 2012; R Core Development Team, 2013). Both analyses were conducted on Bray-Curtis zero-adjusted dissimilarity matrices (McCune and Grace, 2002; Clarke et. al, 2006) constructed from abundance data standardized as CPUE for both gear types. Bray-Curtis is often recommended as a distance measure in ecological studies based on its semi- metric properties (among three given samples the distance between two samples can exceed the distance between the sum of the other two distances), general robustness, and wide utility (Faith et al., 1987; Clarke, 1993; McCune and Grace, 2002; Clarke et al., 2006). Additionally, I compared NMDS analyses using Jaccard and Kulczynski similarities, two other commonly used ecological distance measures (Faith et al., 1987; McCune and Grace, 2002), and Euclidean distance. Jaccard and Kulczynski distances are similar to Bray-Cutis dissimilarity. All three are proportion coefficients; however the Jaccard distance is metric (among three given samples the distance between two samples cannot exceed the distance between the sum of the other two distances), and the Kulczynski is a semi-metric measure mathematically similar to Bray-Curtis,

6 but relativized by sample size (McCune and Grace, 2002). Results were similar among the analyses using ecological distance measures; however, ordination stress dramatically increased with the use of Euclidean distance. Based on the similarity of results obtained using three different ecological distance measures, I felt confident in the integrity of the analysis using Bray- Curtis dissimilarity. Both ubiquitous and rare species were removed from the data set for the cluster and NMDS analyses to avoid their disproportional influence on results. Species were considered ubiquitous if they were captured at high rates throughout the survey area. Species were considered rare if they occurred less than a total of 5 times in the survey. Rank-abundance curves and further discussion of the removal of rare species are included in Appendix A. Because these analyses cannot operate with missing values, sets with zero catch were omitted, and therefore sets in which only ubiquitous or rare species were captured were also omitted. This left 140 sets and data for 10 species in the longline data set and 112 sets and data for 26 species in the gillnet data set. Average linkage was used in the hierarchical cluster analysis, and similarity profile (SIMPROF) tests (Clarke, 1993; Clarke et al., 2008; Wei et al, 2012) were conducted to investigate statistically significant clusters (p < 0.05) using the ‘simprof’ function in the R software package clustsig v.1.0 (Whitaker and Christman, 2010; R Core Development Team, 2013). Indicator species within each cluster were determined using indicator species analysis (Dufrene and Legendre, 1997) with the ‘indval’ function in the R software package labdsv v1.6-1 (Roberts, 2013; R Core Development Team, 2013). I qualitatively compared the clusters over space by mapping each set and labeling them as their respective cluster and examining box plots of cluster environmental variables. Optimal numbers of dimensions in the NMDS were chosen based on qualitative examination of scree plots (McCune and Grace, 2002), shown and discussed in Appendix B. To assess environmental correlates of community structure, the R software package vegan function ‘envfit’ (Oksanen et al., 2013; R Core Development Team, 2013) was used in conjunction with the two-dimensional NMDS ordination. The ‘envfit’ function wraps the functions ‘vectorfit’, which determines directions in ordination space in which vectors of environmental variables change most drastically and are most strongly correlated with the ordination structure, and ‘factorfit’, which calculates mean ordination scores for each categorical factor level. Significance of environmental vectors and factors are evaluated using squared

7 correlation coefficients as a goodness of fit statistic following 1,000 permutations of the environmental data. Function envfit also calculates empirical p-values for each environmental variable. Included as environmental variables in this analysis were region, bottom type, latitude, longitude, maximum and minimum set depth, depth difference (difference between maximum and minimum depth - a proxy for habitat complexity), water clarity, salinity, temperature, and dissolved oxygen (the latter three at surface, mid, and bottom depths). Bottom type was divided into two categories: primary and secondary. If only one kind of bottom was recorded for a given set it was considered to be both the primary and secondary bottom type. There were five categories of bottom type: seagrass, sand, mud, , and reef. For visual inspection of environmental gradients, environmental vectors determined as significant (p < 0.05) were overlaid as smoothed contours in standard ordination plots using the ‘ordisurf’ function in the R software package vegan v.2.0-9 (Oksanen et al., 2013; R Core Development Team, 2013). It is important to note that sites missing data for any environmental variable are omitted in this analysis, which left 88 out of 140 sites in the longline data and 71 out of 112 in the gillnet data.

1.3 Results

1.3.1. General Results A total of 159 paired gillnet and longline sets were conducted from 2009 to 2012. Environmental parameters of sites sampled in each region are described in Table 1. Primary and secondary bottom types for sites in each region are depicted in Figure 3. The cumulative catch included a total of 4,490 individual fishes (70 species, 34 families), consisting of 14 species of elasmobranchs (Table 2) and 56 species of teleosts (Table 3). Five species were dominant overall in terms of raw abundance (in descending order of overall abundance): Atlantic sharpnose sharks (Rhizoprionodon terraenovae), hardhead catfish (Ariopsis felis), bonnethead sharks (Sphyrna tiburo), blacktip sharks (Carcharhinus limbatus), and gafftopsail catfish (Bagre marinus). These five species combined comprised 80.4% of the total catch. Atlantic sharpnose sharks and hardhead catfish were the dominant species, making up 30.5% and 27.5% of the total catch, respectively. The 14 species of elasmobranchs included three small coastal shark species, eight large coastal species, and three batoids, summarized in Table 2. Blacktip sharks were the only species

8 in the large coastal shark complex (NMFS, 1993) caught in relatively high numbers and across all life stages. Individuals of other large coastal species, such as bull (Carcharhinus leucas), tiger (Galeocerdo cuvier), lemon (Negaprion brevirostris), and great hammerhead (Sphyrna mokarran) sharks were typically large juveniles, although mature lemon sharks were caught occasionally (N = 3, of 19 total). There were differences in catch composition between gillnets and longlines. A greater number of teleost species were captured in gillnets, while a greater number of elasmobranch species were captured on longlines (Figure 4). Bonnethead sharks were the only species of shark captured in much higher numbers in gillnets than on longlines (Table 2). In addition to the differences in catch composition, gillnets sampled a more narrow size range of sharks. Both the smallest and largest sharks caught during the survey were captured on longlines (Figure 5).

1.3.2 Abundance and Distribution of Dominant Species

1.3.2.1 Carcharhinus acronotus, blacknose shark A total of 12 blacknose sharks were captured in gillnets and 45 on longlines. Only young- of-the-year (YOY) and juveniles were captured in gillnets, while YOY, juvenile, and mature sharks were captured on longlines. Length frequencies of blacknose sharks captured on longlines in each survey region are shown in Figure 6. Female to male sex ratio was 1.19, and this did not vary significantly from an expected 1:1 ratio according to a Chi-square test (p = 0.59). Sex ratios of each life stage are depicted in Figure 18. Blacknose sharks were captured at the highest rates in the Hernando region, but sample sizes of blacknose sharks for each gear type were not high enough to make for reliable statistical analysis.

1.3.2.2 Carcharhinus limbatus, blacktip shark A total of 109 blacktip sharks were captured in gillnets and 209 were captured on longlines. All three life stages were captured in both gear types, but only two mature sharks were caught in gillnets, compared to 40 on longlines. Mean fork length of blacktip sharks was 72.2 cm (range 40.0 - 137.0 cm). Length frequencies of blacktip sharks captured on longlines in each survey region are shown in Figure 9. Female to male sex ratio was 1.45, which varied significantly from an expected 1:1 according to a chi-square test (p < 0.01). It is worth noting

9 that this difference primarily occurred between YOY and juvenile sharks, and that female to male sex ratio of mature sharks was nearly 1 (23:19). Sex ratios of each life stage are depicted in Figure 18. Catch rates of all life stages of blacktip sharks did not vary significantly across years for either longlines or gillnets (p = 0.06 in both cases). CPUE of each life stage within each region for both gear types is shown in Figure 10, and a map depicting CPUE at each station is shown in Figure 11. Catch rates of all life stages of blacktip sharks combined varied significantly across regions in both gillnets and on longlines (p < 0.01 in both cases), and pairwise regional differences were consistent across both gear types. Catch rates of blacktip sharks were significantly higher in Crystal River than St. Marks for both gear types (longlines, p = 0.04; gillnets p = 0.01) and Hernando (p = 0.02 in both cases); while differences between Steinhatchee and St. Marks were nearly significant for both gear types (p = 0.06 in both cases). Catch rates of blacktip sharks on longlines were significantly higher in Steinhatchee than Hernando (p = 0.04, gillnet p = 0.07). Sample sizes of each individual life stage of blacktip sharks were too low for reliable statistical analyses; however, YOY and juvenile blacktip sharks were captured at high rates in the Crystal River and Steinhatchee regions.

1.3.2.4 Rhizoprionodon terraenovae, Atlantic sharpnose shark A total of 653 Atlantic sharpnose sharks were captured in gillnets, and 714 were captured on longlines. All three life stages were captured on both gear types, but YOY sharks were much more common on longlines than in gillnets (N = 195 of 205). Length frequencies of Atlantic sharpnose sharks captured on longlines in each survey region are shown in Figure 12 Female to male sex ratio was 0.24, and this was significantly different from an expected 1:1 according to a chi-square test (p < 0.01). In contrast to blacktip sharks, these differences occurred primarily between mature sharks. Very few mature female Atlantic sharpnose sharks were captured (N = 9 out of 460). Sex ratios of each life stage are depicted in Figure 18. . There were no significant patterns among years in catch rates of Atlantic sharpnose sharks in gillnets (p = 0.17). Catch rates of Atlantic sharpnose sharks on longlines varied significantly across years (p < 0.01) with higher catch rates on longlines in 2012 than 2010 (p < 0.01). Catch rates of YOY Atlantic sharpnose sharks on longlines varied across years (N = 195, p < 0.01), with significantly higher rates in 2012 than each other year (2009, p = 0.04; 2010, p <

10

0.01; and 2011, p = 0.02). Catch rates of juvenile and mature Atlantic sharpnose sharks on longlines did not vary across years (N = 298, p = 0.77; and N = 218, p = 0.09, respectively). CPUE of each life stage within each region for both gear types is shown in Figure 13, and a map depicting CPUE at each station is shown in Figure 14. Catch rates of Atlantic sharpnose sharks in gillnets did not vary among regions (p = 0.51). Catch rates of all life stages of Atlantic sharpnose sharks on longlines did not vary significantly across regions, but catch rates of both YOY and juvenile Atlantic sharpnose sharks on longlines did vary significantly across regions (p < 0.01 in both cases). Catch rates of YOY Atlantic sharpnose sharks on longlines were significantly higher in Crystal River than St. Marks (p = 0.01) and Hernando (p < 0.01) and in Steinhatchee than Hernando (p < 0.01). However, catch rates of juvenile Atlantic sharpnose sharks were significantly higher on longlines in St. Marks than Crystal River (p = 0.01) and Steinhatchee (p < 0.01). Catch rates of mature Atlantic sharpnose sharks did not vary across regions (p = 0.42).

1.3.2.5 Sphyrna tiburo, bonnethead shark A total of 413 bonnethead sharks were captured in gillnets, while only 6 were captured on longlines. All three life stages were captured in gillnets, and only juvenile and mature were caught on longlines. Length frequencies of bonnethead sharks captured in gillnets in each survey region are shown in Figure 15. Female to male sex ratio was 0.68, which varied significantly from an expected 1:1 according to a chi-square test (p < 0.01). Sex ratios of each life stage are depicted in Figure 18. Catch rates of all life stages of bonnethead sharks combined in gillnets did not vary across years (p = 0.17). CPUE of each life stage within each region is shown in Figure 16, and a map depicting CPUE at each station for gillnet is shown in Figure 17. Catch rates of all life stages of bonnethead sharks in gillnets did not vary significantly across regions (p = 0.10); however catch rates of mature bonnethead sharks in gillnets did vary across regions (N = 214, p = 0.05). Catch rates of mature bonnethead sharks in gillnets were significantly higher in Steinhatchee than in Hernando (p = 0.05). Catch rates of bonnethead sharks on longlines were not analyzed because they were so infrequently caught on longlines.

11

1.3.2.6 Ariid catfishes: Arius felis, hardhead catfish and Bagre marinus, gafftopsail catfish A total of 262 hardhead catfish were captured in gillnets, and 973 were captured on longlines. Mean fork length of hardhead catfish was 26.9 cm (range 14.0 - 37.0 cm). CPUE of hardhead catfish for each gear type in each region are shown in Figure 19 and a map depicting CPUE at each station is shown in Figure 20. Catch rates of hardhead catfish did not vary significantly across years for either longlines (p = 0.51) or gillnets (p = 0.29). Catch rates of hardhead catfish did not vary significantly across regions on longlines (p = 0.46), but they did vary significantly in gillnets (p < 0.01). The Hernando region had significantly lower catch rates of hardhead catfish in gillnets than St. Marks, Steinhatchee, and Crystal River (p = 0.01, p < 0.01, p = 0.05, respectively). A total of 98 gafftopsail catfish were captured in gillnets, and 171 were captured on longlines. CPUE of gafftopsail catfish for each gear type in each region are shown in Figure 19 and a map depicting CPUE at each station is shown in Figure 20. Sample sizes of gafftopsail catfish on each gear type were too low for reliable statistical analyses.

1.3.3 Community Composition and Environmental Correlates All comparisons of each diversity index were made on per-set mean rank values. There were no significant differences (p < 0.05) in species richness (S), Shannon diversity (H), Simpson’s diversity (D), or Pielou’s evenness (J) between years for gillnet (p = 0.64, 0.40, 0.24, and 0.53, respectively) or longline (p = 0.18, 0.22, 0.41, and 0.96, respectively). When only elasmobranchs were examined, there were no significant differences in the same diversity indices across years for gillnet (p = 0.13, 0.63, 0.87, and 0.29, respectively). Only Pielou’s evenness was significantly different among the same diversity indices across years when including only elasmobranchs for longline (p = 0.15, 0.13, 0.66, 0.01, respectively). A Dunn’s post-hoc multiple comparison test indicated a significance difference in longline elasmobranch evenness between 2010 and 2011 (p = <0.01). When only teleosts were examined, there were no significant differences in the same diversity indices across years for gillnet (p = 0.60, 0.79, 0.29, 0.74, respectively) or for longline (p = 0.09, 0.09, 0.32, 0.32, respectively). Diversity indices for each region and gear type are shown in Table 4. I found significant differences diversity indices across regions for both longline and gillnet. Species richness was significantly different across regions for both gear types (gillnet p < 0.01, longline p < 0.01). A

12 post-hoc multiple comparison test (Dunn’s method) revealed gillnet species richness was lower in the Hernando region than both the Steinhatchee and Crystal River regions (p < 0.01 and p = 0.01, respectively; Figure 21). Longline species richness was significantly higher in the St. Marks region than Steinhatchee, Crystal River, and Hernando (p < 0.01 for all three comparisons; Figure 21). There were significant differences in both Shannon and Simpson’s diversity for gillnet (p < 0.01 and p = 0.03, respectively) and longline (p < 0.01 in both cases). The Hernando region was significantly lower than both the Steinhatchee and Crystal River regions in gillnet Shannon diversity (p < 0.01 and p = 0.01, respectively; Figure 21) and Simpson’s diversity (p = 0.05 and p = 0.05, respectively; Figure 16). The St. Marks region was significantly higher than Steinhatchee, Crystal River, and Hernando regions in both longline Shannon diversity (p < 0.01 for all three comparisons; Figure 21) and Simpson’s diversity (p < 0.01 for all three comparisons; Figure 21). Finally, mean rank per-set Pielou’s evenness was only significantly different for gillnet (p = 0.02, p = 0.44 for longline). In this instance, a post-hoc test indicated evenness in Steinhatchee was lower than Crystal River and Hernando (p = 0.03 in both cases; Figure 21). Hierarchical cluster analysis with SIMPROF tests of the gillnet data suggested 15 significant clusters sharing from 0% to 20% of species. To increase simplicity and integrate small clusters containing only two to four samples into their larger neighboring clusters, I cut the dendrogram to a minimum of 8% similarity, based on visual breaks in the dendrogram. This created 7 clusters in the gillnet data sharing from 0% to 8% of species (Figure 22). Cluster designations, latitude, and longitude coordinates are shown for each station in Appendix C. Cluster 3 was the largest with 41 stations. Clusters 1 and 5 included 23 and 21 stations, respectively. Clusters 2, 4, and 7 included seven stations each. Cluster 6 was the smallest with five stations. there was considerable overlap of the largest, cluster 3, with three others: clusters 1, 5, and 7. Cluster 3 was extremely rich, and included sets with representatives of every species in the analysis, with the exception of striped burrfish. Clusters 2, 4, and 6 were relatively species- poor (S = 3, 4, and 3, respectively) and all three included species in high abundance relative to all other clusters. These three clusters were most distinctly separated in the NMDS plot and were most dissimilar to the other clusters. Clusters 1, 5, and 7 were relatively species rich (S = 12, 12, and 9, respectively), but they included some species that occurred in high relative abundance. Clusters 2 and 5 included stations primarily in the far southern portion of the survey area, as well

13 as the far northern portion. Stations in both clusters 2 and 5 had relatively high water clarity, and stations in cluster 5 had relatively high salinities (Figure 23). A map depicting each gillnet station as a symbol according to its assigned cluster is shown in Figure 24. Indicator species analysis determined 10 of the 26 species in the gillnet data set to be significant indicators of clusters (Table 5). All seven clusters had at least one significant indicator species (p < 0.05). The stress level of the two-dimensional NMDS of the gillnet data was moderate at 0.11. This is well within an acceptable range, according to both Clarke (1993) and McCune and Grace (2002). A scree plot, showing stress at different numbers of NMDS dimensions, for gillnet is shown in Appendix B. Vector and factor fitting of the gillnet NDMS using the envfit function in the vegan software package (Okansen, 2013; R Core Team, 2013) indicated cluster designation to be a significant factor (p < 0.01) in the ordination structure (Figures 25 and 26). Additionally, two vectors of environmental variables were significantly correlated with NMDS structure: water clarity (p = 0.04, r2 = 0.10), surface salinity (p = 0.01, r2 = 0.14), while latitude was nearly significant (p = 0.07, r2 = 0.07). The direction of each vector in NMDS space is shown in Figure 27. Fitted surfaces in NMDS space for water clarity, surface salinity, and latitude are shown in Figures 28, 29, and 30, respectively. Hierarchical cluster analysis of the longline data with SIMPROF tests suggested eight significant clusters, which shared from 0% to 25% of species. Again, to reduce complexity and integrate small clusters containing only two to four samples into their larger neighboring clusters, I cut the dendrogram at 15% similarity. This created four clusters, which shared from 5% to 15% of species (Figure 31). Cluster designations, latitude, and longitude coordinates are shown for each station in Appendix C. Longline clusters one and three were the largest, containing 50 and 43 stations, respectively. Longline cluster four contained 29 stations; and longline cluster two was the smallest, containing 18 stations. Select environmental data for each cluster are shown in Figure 32. There were some spatial patterns between clusters (Figure 33). Stations in longline clusters 1 and 3 were distributed along the entire Big Bend coast. Stations in longline cluster two were primarily restricted to lower latitudes; however three stations in the northern portion of the survey area were classified within longline cluster 2. Stations in longline cluster 4 were generally concentrated in the central portion of the survey area. Stations in longline cluster 1 were typically offshore with maximum depths ranging from 2.4 to 6.6 meters and minimum depths ranging from 1.5 to 5.5 meters. Stations in longline cluster 3 were typically inshore with maximum

14 depths ranging from 1.3 to 5.8 meters and minimum depths ranging from 0.3 to 4.0 meters. Median water clarity generally decreased for each successive longline cluster (cluster 1 to cluster 2, cluster 2 to cluster 3, and so on). There were relatively large ranges of bottom salinity in stations within longline clusters 1, 3, and 4. There was a more narrow range and higher median bottom salinity in stations within longline cluster 2 than all others. Indicator species analysis found seven of the ten species in the longline dataset to be significant indicators of longline clusters (Table 6). Black sea bass (Centropristis striata) and sharksuckers (Echeneis naucrates) were found to be significant indicators of longline cluster one (p < 0.01 for each), blacknose sharks and cobia (Rachycentron canadum) were found to be significant indicators of longline cluster two (p < 0.01 and p = 0.01, respectively), gafftopsail catfish were the single indicator species of longline cluster three (p < 0.01), and blacktip and lemon sharks were found to be significant indicators of longline cluster four (p < 0.01 and p = 0.04, respectively). The stress level of the two-dimensional NMDS of the longline data was marginally high at 0.16. However, this is within an acceptable range, according to both Clarke (1993) and McCune and Grace (2002). Additionally, Clarke (1993) noted that stress typically increases as sample size increases, and McCune and Grace (2002) state that most data sets will have ordination solutions with stress between 0.10 and 0.20. A scree plot, showing stress at different numbers of NMDS dimensions, is shown in Appendix B. Vector and factor fitting of the longline NDMS using the ‘envfit’ function in the vegan software package (Oksanen, 2013; R Core Team, 2010) indicated cluster designation to be a significant factor (p < 0.01) in the ordination structure (Figures 33 and 34). Additionally, five vectors of environmental variables were significantly correlated with NMDS structure: water clarity (p < 0.01, r2 = 0.30), minimum depth (p = 0.02, r2 = 0.09), depth difference (p = 0.02, r2 = 0.09), and surface and bottom salinity (p < 0.01, r2 = 0.13 and p < 0.01, r2 = 0.11, respectively). The direction of each vector in NMDS space is shown in Figure 36. Fitted surfaces in NMDS space for minimum depth, water clarity, bottom salinity, and depth difference are shown in Figures 37, 38, 39, and 40, respectively. Water clarity and maximum and minimum depth were correlated (Figure 41).

15

1.4 Discussion

The faunal assemblages of fishes occurring in the summer months of the Florida Big Bend are diverse and spatially variable, like many coastal ecosystems (Knip et al., 2010). Proportional abundance was dominated by a subset of only a few species. The elasmobranch assemblage is primarily composed of carcharhinid species, along with two sphyrnid species and at least one species of dasyatid, gymnurid, and myliobatid. Batoid elasmobranchs are likely to be under-represented in these data, as the fishing gear in this survey selects for them poorly. The teleost assemblage of that Florida Big Bend is very diverse, with representatives of at least 30 families. The use of two gear types allowed two very different species assemblages to be sampled. Longline was generally much more selective for sharks than gillnet, which was expected. Longline also allowed for sampling of both the smallest and largest sharks. These differences underscore the importance of considering the selective properties of fishing gear when studying marine fish assemblages. The observed differences in catch rates between years could be related to a variety of factors. In 2010 bait availability forced the use of low-quality Atlantic mackerel (Scomber scombrus) for bait, and this may have generally reduced catch rates of sharks. It is also possible there were particularly strong year classes for certain species, such as Atlantic sharpnose sharks in 2012, and the differences between years were simply natural fluctuations in populations. Parsons and Hoffmayer (2007) reported that shark abundance was highly variable from year to year in the north-central Gulf of Mexico. Parts of the Big Bend are potentially nursery habitats for some shark species, based on the high proportion of YOY and juvenile life stages to adults over consecutive years. . Hueter and Tyminski (2007) reported nursery habitat for blacktip sharks in Yankeetown, FL (the Crystal River region of this survey), as well as nursery habitat for blacknose sharks in Tampa Bay, FL (south of the Hernando region of this survey). The results of my study corroborate these findings, considering the high proportion of YOY and juvenile blacktip sharks in the Crystal River region. My results also suggest the blacknose shark nursery documented by Hueter and Tyminski (2007) extends north up to Hernando Beach. The role of nursery habitat in the life histories of some small coastal shark species such as Atlantic sharpnose and bonnethead sharks is

16 debated (Heupel et al., 2007; Drymon et al., 2010; Knip et al., 2010), and I believe my study does not provide the data necessary to evaluate the utility of nursery habitats for these species or to delineate the Big Bend as a nursery for them.

1.4.1 Ubiquitous Species Four of the dominant species were essentially ubiquitous and were captured at relatively high rates throughout the survey area. Due to their ubiquity and high abundance, Atlantic sharpnose sharks, bonnethead sharks, and hardhead catfish were removed from the data set for the multivariate analyses. Course-scale comparisons of regional CPUE indicated there was little evidence of regional differences in distribution of these species. Previous studies have suggested salinity, temperature, and dissolved oxygen are influential in the distribution of Atlantic sharpnose sharks (Parsons and Hoffmayer, 2005 and 2007, Drymon et al., 2013). The narrow temporal window of this study and the limited range of sampled salinities may have not allowed for the detection of distribution patterns of this species. For example, Parsons and Hoffmayer (2007) found no Atlantic sharpnose sharks in salinities lower than 14.0, which is outside of salinity range sampled in this study. Similarly, previous studies found bonnethead sharks showed little site fidelity within their resident (Heupel et al., 2006) and were capable of tolerating a wide range of salinities (Ubeda et al., 2009). Similarly, bonnethead sharks have been reported in a variety of habitats in other studies (Carlson, 2002; McCandless et al., 2007). I believe the narrow temporal window and limited salinity range of sites in this study, as in the case of the Atlantic sharpnose shark, were within the general range for habitats of this species.

1.4.2 Sexual Segregation in Sharks There was dramatic sexual segregation in adult Atlantic sharpnose sharks with a strong bias towards males, which corroborate the findings of previous studies (Parsons and Hoffmayer, 2005; Drymon et al., 2010). Drymon et al. (2010) reported a 3:1 ratio of females to males in depths greater than 30 meters off the coast of Alabama, and mature female Atlantic sharpnose very likely occupy depths in the Big Bend greater than those sampled in this study. Although there was little evidence of explicit sexual segregation in blacktip sharks, a high proportion of female blacktip sharks were captured. Drymon et al. (2010) reported a high proportion of female blacktip sharks to males in waters less than 10 meters off the coast of

17

Alabama, especially mature females. Although significantly more female blacktip sharks were captured than males in this survey, the sex ratio of mature sharks was nearly one. Drymon et al. (2010) noted that a high proportion of females in samples was not anomalous based on other studies. There was little evidence of strong sexual segregation in bonnethead sharks, although a higher proportion of male bonnethead sharks were captured. Similarly, there was no evidence of sexual segregation in blacknose sharks; however, very few mature individuals were captured in this survey. Drymon et al. (2010) found evidence of sexual segregation in blacknose sharks off the coast of Alabama with a bias towards males; however, they primarily captured mature individuals. Drymon et al. (2010 and 2012) also reported a discrete depth preference of blacknose sharks in mid-depths, ranging from 10 to 30 meters. The low catch of adult blacknose sharks in this survey is perhaps related to the restriction of sampling to depths less than ten meters.

1.4.3 Regional Faunal Zones My analysis suggests there is a regional faunal zone in the southern Big Bend characterized by blacknose sharks. Stations in the southern longline cluster had relatively high salinities and relatively clear water; while stations in the blacknose shark gillnet cluster (gillnet cluster 2) had a wider range of salinities, but similarly clear water. This pattern may be related to the gradient of river influence, which is moderate in the north, low in the south, and high in the central Big Bend (Zieman and Zieman, 1989). Carlson (2002) reported blacknose sharks (all life stages) to be most abundant in waters of relatively high salinity (mean 32.1, range 26.0-38.0) in the northwest Gulf of Mexico. Similarly, Hueter and Tyminski (2007) found blacknose sharks (neonate, YOY and older juveniles) in salinities ranging from 26.2 to 34.5. The type of habitat in Tampa Bay is very similar to the Big Bend, especially between the southern Big Bend and northern Tampa Bay where the bottom is dominated by seagrass and river influence is low (Zieman and Zieman, 1989; Hueter and Tyminski, 2007). Carlson (2002) noted a high proportion of YOY and mature female blacknose sharks in similar habitat (high salinity, relatively high clarity, and abundant seagrass), and reported a range of water clarities for YOY blacknose sharks from 100.0-200.0 cm (mean 201.0 cm). However, Hueter and Tyminski (2007) noted the catch of older juvenile blacknose sharks in Yankeetown, FL (the Crystal River region of this survey),

18 and Carlson (2002) reported a wider range of water clarities for juvenile and adult blacknose sharks (8.3-400.0 cm and 50.0-290.0 cm, respectively). These results suggest salinity may be among the most influential environmental variables in terms of distribution of blacknose sharks within a narrow temporal window when temperature is not an issue. Likewise, habitat with high salinity, high water clarity, and seagrass bottom may constitute preferable nursery habitat for blacknose sharks. Similar to blacknose sharks, my analysis suggests a faunal zone characterized by blacktip sharks. This cluster was not as regionally restricted as the southern faunal cluster; however, stations within this cluster occurred most densely in central latitudes of the survey. Stations in the turbid longline cluster were typified by generally lower water clarity and mud as a primary or secondary bottom type; while stations in the gillnet cluster for which blacktip sharks were an indicator species (gillnet cluster 3) had a more narrow range of water clarities. The characteristics of this habitat are likely related to the high river influence in the central Big Bend (Zieman and Zieman, 1989; Hueter and Tyminski, 2007). Previous studies have found blacktip sharks over silt, sand, clay, and mud bottoms (Carlson, 2002; McCandless et al., 2007), and Hueter and Tyminski (2007) noted that older juvenile blacktip sharks move well into estuarine waters, and they captured blacktip sharks in a wide range of salinities (15.3-41.6). Froeschke et al. (2010) reported blacktip sharks were restricted to areas of moderate salinities and proximity to tidal inlets and deeper water. Depth difference was a significant environmental variable in the longline NMDS, and many stations within this cluster had high depth difference values. Frequenting shallow water proximate to deeper channels may be common in blacktip sharks, and habitat with complex bottom is potentially preferable to this species. However, I contend that a tendency to frequent deep channels through shallow habitat is not unique to blacktip sharks and these habitats are likely preferential to other large coastal species, such as lemon and bull sharks.

1.4.4 Community Composition and Environmental Correlates Overall species richness, diversity, and evenness appeared to be relatively constant across years. Both species richness and diversity followed a consistent pattern for each gear type across survey regions, with the highest mean rank per-set richness and diversity on longlines in St. Marks and in gillnets in Steinhatchee and Crystal River. Evenness was relatively constant, and I associate the significant difference found in gillnet evenness between regions to be influenced by

19 two gillnet sets that captured an exceptionally high number of one shark species (one of bonnethead sharks and one of Atlantic sharpnose sharks) in the St. Marks region. Hierarchical cluster analysis and NMDS of the gillnet data were complex. Three clusters, clusters 2, 4, and 6, were separated due to a single species (blacknose sharks, striped burrfish, and Atlantic spadefish, respectively) being caught in high relative abundance with few other species. These clusters were fairly well separated in the NMDS plots, but the sharing of species between the more species rich clusters 1, 3, and 7, was indicated in the heterogeneous dispersion of those stations in the ordination. Likewise, there was overlap between clusters 3 and 5, which shared several species. Salinity, water clarity, and to a lesser extent latitude were correlated with community structure according to environmental fitting analysis of the gillnet ordination. Gillnet clusters 2 and 5 appeared to primarily include stations in the southern Big Bend where water clarity and salinity are relatively high, and this conclusion was supported by the gradients of these variables in the NMDS plots. Indicator species for these two clusters were blacknose sharks (cluster 2) and grass porgies (cluster 5). Grass porgies were not reported in seagrass habitats in the Suwannee Sound (Steinhatchee and Crystal River regions of this survey) by Tuckey and Dehaven (2006), and this species may be more abundant in habitats with reduced river influence. Relationships between environmental variables and the other clusters were difficult to discern, and it may be that the high species richness of teleost catch in gillnets and the high number of potential species combinations may render these community analyses relatively uninformative for many taxa without additional sampling. Additionally, some species may be poorly represented through selection biases related to size, shape, and patterns of movement. I found four fairly distinct clusters in the longline data. Two were distributed along the entire Big Bend and represent inshore and offshore faunal assemblages. Although there was overlap in the occurrence of species in each cluster, there was a transition from shallow habitats dominated by species such as gaftopsail catfish to deeper, less turbid habitats dominated by species such as black sea bass and sharksuckers. There was also a southern cluster characterized by cobia and blacknose sharks, and a deep, turbid cluster which occurred most densely in the central Big Bend and was characterized primarily by blacktip sharks. Relative abundance and frequency of tiger sharks were highest in the offshore cluster, however relatively few tiger sharks were captured overall. Similarly, relative abundance and frequency of bull sharks was highest in

20 the inshore and turbid clusters, but bull sharks were not well represented in the data set, either. It could be that the sample sizes of these species were too low to expose a significant pattern. However, cobia had the lowest CPUE of any species in the longline data set, and they were found to be a significant indicator species for the southern cluster. Therefore, it may be that these large coastal shark species are essentially cosmopolitan, exhibiting some tendencies to occupy certain habitats, but occurring throughout the Big Bend. The NMDS and environmental fitting analysis of the longline data indicated clarity, salinity, minimum depth, and depth difference were most correlated with the NMDS structure and therefore community structure. Based on the cluster analysis, bottom salinity seems to be the most significant factor between the southern cluster and the other three clusters. Water clarity and depth appear to be the most influential environmental variables between the three clusters distributed throughout (or nearly throughout) the survey area and these variables are correlated with each other. Depth difference was generally highest in the inshore and turbid clusters. A high depth difference of inshore sets is sensible since longlines are typically set perpendicular from shore. It is difficult to discern whether the high depth difference values in the turbid cluster are a factor influencing community composition or simply a characteristic of the sites in that cluster in which blacktip and lemon sharks are caught at high rates due to other variables. That salinity and water clarity were significantly correlated with NMDS structure of both the gillnet and longline data suggests, within the temporal window sampled and of the variables assessed, these two parameters may most strongly influence elasmobranch and large teleost faunal assemblages in the Big Bend. There are also correlations between community composition and depth, and water clarity and depth were themselves correlated. Temperature was not a significant variable in my analysis, which was expected given the narrow temporal window of this survey. Previous work has shown temperature, as well as day length, to be related to shark distribution, immigration, and emigration (Grubbs et al., 2007; Heupel, 2007); and it is important to consider temporal range when interpreting the results of this study. Bottom type was not a significant factor, but power to detect an effect of bottom type was likely limited in this analysis, given the survey was designed to sample seagrass habitats. Previous studies have shown relationships between fish distributions and environmental variables to be complex and interactive (Froeschke et al., 2010; Knip et al., 2010), and the results of my study suggest the Florida Big Bend is no exception. The narrow temporal window of this

21 study facilitated the investigation of influential environmental parameters outside of seasonality. However, this is only one category of habitat characteristics that may be important drivers of species distribution patterns. Previous studies have found evidence for habitat partitioning due to interspecific competition (Bethea et al., 2004; Papastamatiou et al., 2006). Heupel and Hueter (2002) concluded prey density was not correlated with the movements of juvenile blacktip sharks and suggested presence of predators may be of greater influence. Carlson (2002) made a similar suggestion, that species with small juvenile life stages may choose habitats in which predator abundance is low. However, Parsons and Hoffmayer (2007) found no significant negative associations between species and size-classes of sharks in the north-central Gulf of Mexico. Based on this and other studies, factors influencing species distributions appear to include both biotic and abiotic processes. It has also been suggested that different populations of a single species may behave differently in separate portions of its range (Drymon et al., 2010; Knip et al., 2010). Therefore, the relative influence of a given factor may change from system to system.

1.4.5 Conclusion The Big Bend is inhabited by at least 14 species of elasmobranch and 56 species of teleost fishes during the summer. Assemblages of these fauna are spatially variable, and five species dominate. These include three species of sharks, two carcharhinids and one sphyrnid, and two species of ariid catfishes. Parts of the region likely act as nursery habitat for at least two shark species: blacktip sharks in the central Big Bend, and blacknose sharks in the south. Environmental variables influencing community composition include salinity, water clarity, and depth. There is a distinct southern faunal zone off Hernando Beach, where salinity and water clarity are relatively high. In the next chapter, I will investigate the trophic structure of these faunal assemblages using carbon and nitrogen stable isotope analysis to determine if trophic structure varies in ways similar to community composition.

22

Table 1. Summary of environmental parameters of sites sampled within each survey region. Means and standard deviations are shown with ranges in parentheses. Mean depth included maximum and minimum depths of each site. For salinity, temperature, and dissolved oxygen, each line represents surface, mid, and bottom values from top to bottom, respectively.

Region N Temperature Salinity Dissolved Water clarity Water depth (°C) oxygen (cm) (m) (mg/l)

30.3 ± 1.1 26.8 ± 2.8 5.96 ± 1.10 (27.9 - 32.6) (15.8 - 30.4) (3.87 - 8.08)

St. Marks 38 29.9 ± 1.1 27.9 ± 2.2 5.88 ± 1.05 309 ± 97 3.4 ± 1.3 (27.6 - 32.1) (23.8 - 30.6) (3.64 - 8.13) (160 - 500) (1.0 - 6.6)

30.0 ± 1.2 28.0 ± 2.3 5.90 ± 1.39 (27.6 - 32.1) (22.4 - 31.4) (3.54 - 8.60)

30.4 ± 1.2 27.4 ± 2.6 5.77 ± 1.21 (28.5 - 32.5) (18.6 - 30.5) (3.78 - 8.88)

Steinhatchee 35 30.9 ± 0.9 27.8 ± 2.9 5.56 ± 1.11 201 ± 88 2.6 ± 1.2 (28.9 - 32.1) (18.9 - 30.5) (3.22 - 7.90) (50 - 450) (0.3 - 6.0)

30.4 ± 1.2 27.9 ± 2.5 5.42 ± 1.60 (28.4 - 32.3) (20.7 - 30.5) (2.04 - 8.55)

29.6 ± 1.6 27.6 ± 3.3 5.72 ± 0.91 (26.6 - 32.7) (20.2 - 36.0) (3.35 - 7.18)

Crystal River 32 30.5 ± 0.8 27.0 ± 2.8 5.84 ± 0.73 219 ± 114 3.4 ± 1.7 (28.7 - 31.7) (20.4 - 30.5) (4.68 - 7.06) (80 - 650) (1.0 - 8.6)

30.4 ± 0.8 27.1 ± 2.6 5.59 ± 1.16 (28.8 - 31.8) (20.4 - 30.5) (3.32 - 8.24)

30.3 ± 1.4 28.5 ± 3.9 5.66 ± 0.60 (27.2 - 32.9) (19.8 - 34.4) (4.60 - 7.21)

Hernando 53 30.7 ± 1.2 28.6 ± 4.0 5.53 ± 0.59 280 ± 80 2.6 ± 0.9 (28.8 - 33.0) (19.8 - 34.5) (4.41 - 7.09) (150 - 500) (1.0 - 5.6)

30.6 ± 1.3 28.5 ± 4.0 5.44 ± 0.78 (28.2 - 33.0) (19.8 - 34.5) (3.64 - 7.31)

23

Table 2. Summary of elasmobranch fishes; longline CPUE ([catch/100 hooks]/hours soaked *100), gillnet CPUE (catch/hours soaked); SE = standard error; m = male, f = female; YOY = young-of-the-year, Juv = juvenile, Mat = mature, FL = fork length, TL = total length, DW = disc width, Individuals for which sex and/or life stage were not recorded were omitted for calculation of sex and maturity ratios.

Genus species N Longline Gillnet FL Range Sex Ratio Maturity (common name) CPUE CPUE (cm) (f:m) Ratio (mean±SE) (mean±SE) (YOY: Juv:Mat) Carcharhinus acronotus 57 0.25 ± 0.06 0.06 ± 0.02 33.0-94.0 31:26 30:26:1 (blacknose shark) (1.19) Carcharhinus brevipinna 4 0.02 ± 0.01 - 51.0-55.0 3:1 1:0:0 (spinner shark) (3.00) Carcharhinus leucas 18 0.09 ± 0.03 - 120.0 - 3:5 0:1:0 (bull shark) 175.0 (0.60)

Carcharhinus limbatus 318 0.99 ± 0.21 0.59 ± 0.26 40.0- 61:42 73:64:21 (blacktip shark) 150.0 (1.45) Dasyatis americana 10 0.02 ± 0.01 0.04 ± 0.02 19.0-90.0 3: 1 1:4:5 (southern stingray) (DW) (3.00) Galeocerdo cuvier 15 0.08 ± 0.02 - 154.0- 13:1 0:1:0 (tiger shark) 196.0 (13.00) Ginglymostoma cirratum 12 0.25 ± 0.06 - 218.0- 1:12 0:1:11 (nurse shark) 252.0 (0.08) (TL) Gymnura micrura 2 - 0.01 ± 0.01 30.0-53.0 - 0:1:0 (smooth butterfly ray) (DW) Negaprion brevirostris 19 0.11 ± 0.03 - 153.0- 9:10 0:16:3 (lemon shark) 211.0 (0.90) Rhizoprionodon 1376 3.94 ± 0.33 3.69 ± 0.37 24.0-83.0 257: 205:667: terraenovae 1081 461 (Atlantic sharpnose shark) (0.24) Rhinoptera bonasus 5 - 0.03 ± 0.02 33.0-68.0 0:1 0:3:2 (cownose ray) (DW) (0.00) Sphyrna mokarran 4 0.02 ± 0.01 - 158.0- 3:1 0:1:0 (great hammerhead shark) 241.0 (3.00) Sphyrna tiburo 419 0.03 ± 0.01 2.28 ± 0.37 45 - 110 167: 5:94:109 (bonnethead shark) 245 (0.68)

24

Table 3: Summary of bony fishes in taxonomic order by family; LL CPUE ([catch/100 hooks]/hours soaked*100), GN CPUE (catch/hours soaked); SE=standard error; FL=fork length

Family species (common name) N LL CPUE GN CPUE FL Range (mean±SE) (mean±SE) (cm)

Lepisosteidae Lepisosteus osseus (longnose gar) 7 - 0.04 ± 0.01 83.0-118.0

Elopidae Elops saurus (ladyfish) 18 - 0.11 ± 0.01 28.0-49.0 Clupeidae Brevoortia patronus (Gulf menhaden) 49 - 0.25 ± 0.12 15.0-37.0

Brevoortia smithi (yellowfin menhaden) 62 - 0.39 ± 0.20 20.5-30

Opisthonema oglinum (threadfin 3 - 0.02 ± 0.01 15.5-16.0 herring) Synodontidae Synodus foetens (inshore lizardfish) 4 0.01 ± 0.01 0.01 ± 0.01 22.0-28.0

Ariidae Arius felis (hardhead catfish) 1235 5.38 ± 0.37 1.54 ± 0.27 14.0-47.0

Bagre marinus (gafftopsail catfish) 269 0.91 ± 0.13 0.52 ± 0.11 21.0-53.0

Batrachoididae Opsanus beta (Gulf toadfish) 2 0.01 ± 0.01 - 22.0-27.0

Ogcocephalidae Ogcocephalus sp. (batfish sp.) 1 - - 17.0

Belonidae crocodilus (houndfish) 3 0.01 ± 0.01 0.01 ± 0.01 71.0-94.0

Serranidae Centropristis striata (black sea bass) 112 0.62 ± 0.11 0.01 ± 0.01 13.0-44.0 Diplectrum formosum (sand perch) 1 0.01 ± 0.01 - 14.0

Epinephelus morio (red grouper) 1 0.01 ± 0.01 - 31.0

Mycteroperca microlepis (gag grouper) 1 - <0.01 ± <0.01 38.0

Pomatomidae Pomatomus saltatrix (bluefish) 8 0.01 ± 0.01 0.04 ± 0.01 35.0-54.0

Rachycentridae Rachycentron canadum (cobia) 21 0.06 ± 0.02 0.04 ± 0.01 44.0-116.0

Echeneidae Echeneis naucrates (sharksucker) 120 0.70 ± 0.14 0.06 ± 0.02 33.0-59.0

Carangidae Caranx crysos (blue runner) 2 - 0.01 ± 0.01 21.0-22.0

Caranx hippos (crevalle jack) 10 - 0.06 ± 0.03 21.5-38.0

Caranx latus (horse-eye jack) 3 - 0.02 ± 0.01 22.0-32.0

Chloroscombrus chrysurus (A. bumper) 1 - <0.01 ± <0.01 19.0

Selene vomer (lookdown) 1 - 0.01 ± 0.01 20.0

Trachinotus carolinus (pompano) 2 - 0.01 ± 0.01 27.0-37.0

25

Table 3 - continued

Family Genus species (common name) N LL CPUE GN CPUE FL Range (mean±SE) (mean±SE) (cm)

Lutjanidae Lutjanus griseus (mangrove snapper) 1 - 0.01 ± 0.01 30.0

Lobotidae Lobotes surinamensis (tripletail) 1 - 0.01 ± 0.01 -

Haemulidae Haemulon parrai (sailors choice) 1 - 0.01 ± 0.01 24.0 Haemulon plumieri (white grunt) 5 0.01 ± 0.01 0.02 ± 0.01 21.0-28.0

Orthopristis chrysoptera (pigfish) 12 0.01 ± 0.01 0.07 ± 0.03 15.0-24.0

Sparidae Archosargus probatocephalus 2 - 0.01 ± 0.01 33.0-36.0 (sheepshead) Calamus arctifrons (grass porgy) 16 - 0.09 ± 0.03 11.5-24.0

Diplodus holbrookii (spottail pinfish) 8 - 0.04 ± 0.02 11.0-20.0

Lagodon rhomboides (pinfish) 22 0.02 ± 0.01 0.10 ± 0.04 11.0-24.0

Sciaenidae Bairdiella chrysoura (American silver 5 0.01 ± 0.01 0.02 ± 0.01 14.0-24.0 perch) Cynoscion arenarius (sand weakfish) 4 0.01 ± 0.01 0.01 ± 0.01 23.0-33.0

Cynoscion nebulosus (spotted sea trout) 22 0.05 ± 0.02 0.08 ± 0.02 34.0-57.0

Leiostomus xanthurus (spot) 3 - 0.02 ± 0.01 22.0-25.0

Menticirrhus sp. (kingfish sp.) 1 - <0.01 ± 33.0 <0.01 Micropogonias undulatus (croaker) 1 - 0.01 ± 0.01 32.0

Pogonias cromis (black drum) 6 - 0.04 ± 0.03 38.0-48.0

Sciaenops ocellatus (red drum) 3 0.01 ± 0.01 0.01 ± 0.01 44.0-60.0

Ephippidae Chaetodipterus faber (Atlantic spadefish) 12 - 0.07 ± 0.03 11.0-23.0

Scaridae Nicholsina usta (emerald parrotfish) 1 - 0.01 ± 0.01 17.0

Mugilidae Mugil cephalus (striped mullet) 5 - 0.03 ± 0.03 30.-5-36.0

Scombridae Scomberomorus maculatus (Spanish 67 0.07 ± 0.03 0.30 ± 0.06 3.0-63.0 mackerel)

Stromateidae Peprilus burti (Gulf butterfish) 1 - 0.01 ± 0.01 18.0

Peprilus paru (American harvestfish) 30 - 0.17 ± 0.07 13.0-25.0

26

Table 3 - continued

Family Genus species (common name) N LL CPUE GN CPUE FL Range (mean±SE) (mean±SE) (cm)

Bothidae Paralichthys albigutta (Gulf flounder) 12 - 0.07 ± 0.02 21.0-45.0

Bothidae Paralichthys lethostigma (southern 1 - 0.01 ± 0.01 38.0 flounder)

Monacanthidae Aluterus monoceros (unicorn filefish) 2 - 0.01 ± 0.01 32.0-34.0 Aluterus schoepfi (orange filefish) 1 - 0.01 ± 0.01 36.0

Aluterus scriptus (scrawled filefish) 3 - 0.02 ± 0.02 35.0-39.0

Monocanthus hispidus (plainhead 1 - 0.01 ± 0.01 8.0 filefish) Ostraciidae Lactophrys quadricornis (scrawled 1 - 0.01 ± 0.01 20.0 cowfish) Lactophrys trigonus (trunkfish) 2 - 0.01 ± 0.01 18.0

Diodontidae Chilomycterus schoepfi (striped 7 - 0.04 ± 0.02 12.0-26.0 burrfish)

Table 4. Diversity indices of each survey region.

Region N Gear Species Shannon Simpson’s Pielou’s richness diversity diversity evenness S H D J

St. Marks 38 Gillnet 32 1.872 0.709 0.540 Longline 21 1.847 0.774 0.607

Steinhatchee 35 Gillnet 42 2.455 0.859 0.657 Longline 16 1.586 0.711 0.572 Crystal River 32 Gillnet 31 2.257 0.842 0.657 Longline 13 1.513 0.732 0.590 Hernando 53 Gillnet 25 1.853 0.716 0.576 Longline 18 1.588 0.694 0.549

27

Table 5. Results of indicator species of gillnet cluster analysis. Data shown for the cluster for which each species’ indicator value was highest. Species in bold were significant indicators.

Species Indicator Relative Relative Indicator P Cluster frequency abundance value

BMAR, gafftopsail catfish, Bagre 3 0.78 0.78 0.61 <0.01 marinus

BPAT, Gulf menhaden, Brevoortia 3 0.17 0.76 0.12 0.20 patronus

BSMI, yellowfin menhaden, Brevoortia 3 0.20 1.00 0.20 0.10 smithi

CACR, blacknose shark, Carcharhinus 2 1.00 0.78 0.78 <0.01 acronotus

CARC, grass porgy, Calamus arctifrons 5 0.62 0.96 0.60 <0.01

CFAB - Atlantic spadefish, 6 1.00 0.96 0.96 <0.01 Chaetodipterus faber

CHIP - crevalle jack, Caranx hippos 7 0.28 0.57 0.16 0.10

CLIM - blacktip shark, Carcharhinus 3 0.49 0.84 0.41 0.03 limbatus

CNEB - spotted seatrout, Cynoscion 5 0.19 0.31 0.06 0.74 nebulosus

CSCH - striped burrfish, Chilomycterus 4 1.00 0.95 0.95 <0.01 schoepfi

DAME - southern stingray, Dasyatis 7 0.14 0.45 0.06 0.54 americana

DHOL - spottail pinfish, Diplodus 7 0.42 0.81 0.35 0.01 holbrookii

ENAU - sharksucker, Echeneis 1 0.30 0.92 0.28 0.01 naucrates

ESAU - ladyfish, Elops saurus 7 0.29 0.31 0.09 0.47

LOSS - longnose gar, Lepisosteus osseus 3 0.17 1.00 0.17 0.10

LRHO - pinfish, Lagodon rhomboides 7 0.86 0.75 0.65 <0.01

MCEP - striped mullet, Mugil cephalus 3 0.02 1.00 0.02 1.00 Table 5 - continued

28

Species Indicator Relative Relative Indicator P Cluster frequency abundance value

OCHR - pigfish, Orthopristis chrysoptera 3 0.17 0.85 0.14 0.15

PALB - Gulf flounder, Paralichthys 7 0.29 0.46 0.13 0.20 albigutta

PCRO - black drum, Pogonias cromis 3 0.05 1.00 0.05 0.63

PPAR - American harvestfish, Peprilus 3 0.17 0.89 0.15 0.13 paru

PSAL - bluefish, Pomatomus saltatrix 5 0.14 0.61 0.09 0.32

RBON - cownose ray, Rhinoptera bonasus 3 0.07 1.00 0.07 0.62

RCAN - cobia, Rachycentron canadum 5 0.19 0.51 0.10 0.29

SMAC - S. mackerel, Scomberomorus 1 0.78 0.49 0.39 <0.01 maculatus

29

Table 6. Results of indicator species analysis of longline clusters. Species in bold were significant indicators.

Species Cluster Relative Relative Indicator Indicator P frequency abundance value cluster

BMAR - gafftopsail catfish 1 0.14 0.07 0.01 3 <0.01 2 0.06 0.16 0.00 Bagre marinus 3 1.00 0.73 0.73 4 0.24 0.19 0.04

CACR - blacknose shark 1 0.08 0.04 0.00 2 <0.01 2 0.89 0.90 0.80 Carcharhinus acronotus 3 0.02 0.02 0.00 4 0.07 0.04 0.00

CLEU - bull shark 1 0.06 0.18 0.01 3 0.42 2 0.06 0.13 0.01 Carcharhinus leucas 3 0.14 0.47 0.07 4 0.14 0.22 0.03

CLIM - blacktip shark 1 0.22 0.05 0.01 4 <0.01 2 0.11 0.17 0.00 Carcharhinus limbatus 3 0.23 0.06 0.01 4 1.00 0.87 0.87

CSTR - black sea bass 1 0.56 0.71 0.40 1 <0.01 2 0.16 0.09 0.01 Centropristis striata 3 0.23 0.13 0.03 4 0.17 0.07 0.01

ENAU - sharksucker 1 0.70 0.73 0.51 1 <0.01 2 0.17 0.10 0.17 Echeneis naucrates 3 0.26 0.17 0.43 4 0.00 0.00 0.00

GCIR - nurse shark 1 0.12 0.53 0.06 1 0.34 2 0.00 0.00 0.00 Ginglymostoma cirratum 3 0.12 0.47 0.05 4 0.00 0.00 0.00

GCUV - tiger shark 1 0.12 0.40 0.05 1 0.62 2 0.06 0.14 0.01 Galeocerdo cuvier 3 0.07 0.27 0.02 4 0.07 0.20 0.01

NBRE - lemon shark 1 0.04 0.08 0.01 4 0.03 2 0.06 0.11 0.01 Negaprion brevirostris 3 0.14 0.31 0.04 4 0.28 0.50 0.14

RCAN - cobia 1 0.06 0.16 0.01 2 0.02 2 0.22 0.64 0.14 Rachycentron canadum 3 0.05 0.14 0.01 4 0.03 0.07 0.00

30

Figure 1. USGS GIS layer of seagrass coverage in the Big Bend.

31

St. Marks

Steinhatchee

Crystal River

Hernando

Figure 2. All sampling stations in the Big Bend. General survey regions are labeled. Dark circles represent longline sets and triangles represent gillnet sets.

32

a) 100% 90% 80% 70% Seagrass 60% Sand 50% Reef 40% Mud Algae 30% 20% 10% 0% St. Marks Steinhatchee Crystal River Hernando

b) 100% 90% 80% 70% Seagrass 60% Sand 50% Reef 40% Mud Algae 30% 20% 10% 0% St. Marks Steinhatchee Crystal River Hernando

Figure 3. Primary (a) and secondary (b) bottom types of sampling sites within each survey region.

33

60

50

40

30 Elasmobranchs Teleosts

Number Number speciesof 20

10

0 Gillnet Longline

Figure 4. Total number of elasmobranch and teleost species in gillnet and on longline.

Longline

Gillnet

0 25 50 75 100 125 150 175 200 225 250 275 300 325 Shark Total Length (cm)

Figure 5. Shark size selection of each gear type. Longline range was 28.5-302.0 cm TL. Gillnet range was 32.0-167 cm TL.

34

Figure 6. Length frequencies of blacknose sharks captured on longline in each survey region.

35

Figure 7. CPUE of blacknose sharks (Carcharhinus acronotus) in gillnet [catch/hours soaked] (top) and on longline [(catch/100 hooks)/hours soaked * 100] (bottom). Means for each life stage within each survey region are shown, error bars indicate standard error. No mature blacknose sharks were captured in gillnet.

36

Figure 8. Map of gillnet and longline CPUE of blacknose sharks (Carcharhinus acronotus).

37

Figure 9. Length frequencies of blacktip sharks (Carcharhinus limbatus) captured on longline in each survey region.

38

Figure 10. CPUE of blacktip sharks (Carcharhinus limbatus) in gillnet [catch/hours soaked] (top) and on longline [(catch/100 hooks)/hours soaked * 100] (bottom). Means for each life stage within each survey region are shown, error bars indicate standard error.

39

Figure 11. Map of gillnet and longline CPUE of blacktip sharks (Carcharhinus limbatus).

40

Figure 12. Length frequencies of Atlantic sharpnose sharks (Rhizoprionodon terraenovae) captured on longline in each survey region.

41

Figure 13. CPUE of Atlantic sharpnose sharks (Rhizoprionodon terraenovae) in gillnet [catch/hours soaked] (top) and on longline [(catch/100 hooks)/hours soaked * 100] (bottom). Means for each life stage within each survey region are shown, error bars indicate standard error.

42

Figure 14. Map of gillnet longline CPUE of Atlantic sharpnose sharks (Rhizoprionodon terraenovae).

43

Figure 15. Length frequencies bonnethead sharks (Sphyrna tiburo) captured in gillnet in each survey region.

44

Figure 16. CPUE of Bonnethead sharks (Sphyrna tiburo) in gillnet [catch/hours soaked]. Means for each life stage within each survey region are shown, error bars indicate standard error

Figure 17. Map of gillnet CPUE of bonnethead sharks (Sphyrna tiburo).

45

Young-of-the-year 100%

80%

60% Male 40% Female

20%

0% Blacknose Blacktip Atl Sharpnose Bonnethead

Juvenile 100%

80%

60% Male 40% Female

20%

0% Blacknose Blacktip Atl Sharpnose Bonnethead

Mature 100%

80%

60% Male 40% Female 20%

0% Blacknose Blacktip Atl Sharpnose Bonnethead

Figure 18. Sex ratios of each life stage for blacknose, blacktip, Atlantic sharpnose, and bonnethead sharks.

46

Figure 19. CPUE of (a) Hardhead catfish (Arius felis) and (b) gafftopsail catfish (Bagre marinus) in gillnet and on longline. Longline CPUE = [(catch/100 hooks)/hours soaked * 100] and gillnet CPUE = [catch/hours soaked]. Means for each gear type within each survey region are shown, error bars indicate standard error.

47

Figure 20. Maps of gillnet and longline CPUE of hardhead catfish (Arius felis, above) and gafftopsail catfish (Bagre marinus, below).

48

Figure 21. Per-set species richness, Shannon diversity, Simpson’s diversity, and Pielou’s evenness (top to bottom, respectively) of each survey region for gillnet (left) and longline (right).

49

Cluster 2

Cluster 6

Cluster 4

Cluster 5

Cluster 7

Cluster 1

Cluster 3

Figure 22. Gillnet dendrogram showing 7 clusters at 92% similarity. Numbers identify individual stations, which are shown with latitude and longitude coordinates in Appendix C.

50

Figure 23. From top to bottom, respectively: max depth, surface salinity, water clarity, and latitude (bottom) of sampling stations in each gillnet cluster.

51

Figure 24. All gillnet sampling sites represented by symbols according to cluster.

52

Figure 25. NMDS plot of gillnet data with 75% standard deviation confidence ellipses around the centroid of each cluster and lines connecting each point to its assigned cluster.

53

Figure 26. NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster. Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

54

Figure 27. NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and directions of correlated environmental vectors (p < 0.05). Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

55

Figure 28. NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of water clarity (cm). Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species. Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

56

Figure 29. NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of surface salinity. Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

57

Figure 30. NMDS plot of gillnet 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of latitude. Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=25). Visible species codes: BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CACR, Carcharhinus acronotus (blacknose shark); CARC, Calamus arctifrons (grass porgy), CFAB, Chaetodipterus faber (Atlantic spadefish); CLIM, Carcharhinus limbatus (blacktip shark); CNEB, Cynoscion nebulosus (spotted seatrout); CSCH, Chilomycterus schoepfi (striped burrfish); DAME, Dasyatis americana (southern stingray); DHOL, Diplodus holbrookii (spottail pinfish); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); MCEP Mugil cephalus (striped mullet), OCHR, Orthopristis chrysoptera (pigfish); PALB, Paralichthys albigutta (Gulf flounder); PCRO Pogonias cromis (black drum); PPAR, Peprilus paru (American harvestfish); PSAL, Pomatomus saltatrix (bluefish); RBON, Rhinoptera bonasus (cownose ray); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

58

Southern (Cluster 2)

Turbid (Cluster 4)

Inshore (Cluster 3)

Offshore (Cluster 1)

Figure 31. Longline dendrogram showing four clusters at 85% similarity. Numbers identify individual stations, which are shown with latitude and longitude coordinates in Appendix C.

59

Figure 32. Minimum depth (top-left), max (top right), water clarity (mid-left), bottom salinity (mid-right), depth difference (bottom-left), and latitude (bottom-right) of sampling stations in each longline cluster.

60

Figure 33. All longline sampling sites represented by symbols according to cluster.

61

Figure 34. NMDS plot of longline data with 75% standard deviation confidence ellipses around the centroid of each cluster and lines connecting each point to its assigned cluster.

62

Figure 35. NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia).

63

Figure 36. NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and directions of correlated environmental vectors (p < 0.05). Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia).

64

Figure 37. NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of minimum depth (m). Four letter codes indicate weighted average points for each species. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia).

65

Figure 38. NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of water clarity (cm). Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia).

66

Figure 39. NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of bottom salinity. Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species. Species codes (N=10): BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia).

67

Figure 40. NMDS plot of longline 75% standard deviation confidence ellipses around the centroid of each cluster and fitted contour of depth difference (m). Four letter codes indicate weighted average points for each species. Colors represent different clusters. Four letter codes indicate weighted average points for each species (N=10). Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); Species codes: BMAR, Bagre marinus (gafftopsail catfish); CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); CSTR, Centropristis striata (black sea bass); ENAU, Echeneis naucrates (sharksucker); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); NBRE, Negaprion brevirostris (lemon shark); RCAN, Rachycentron canadum (cobia).

68

700 R² = 0.1407 600

500

400

300 Water clarity Waterclarity 200

100

0 1 2 3 4 5 6 7 8 9 Maximum depth

700

600 R² = 0.1907

500

400

300 Water clarity Waterclarity 200

100

0 0 1 2 3 4 5 6 7 Minimum depth

Figure 41. Correlation of longline set water clarity (cm) and maximum (top) and minimum depth (m) (bottom).

69

CHAPTER TWO

STABLE ISOTOPE ECOLOGY AND TROPHIC STRUCTURE OF SHARK AND LARGE TELEOST FAUNAL ASSEMBLAGES IN THE BIG BEND

2.1 Introduction

Trophic interactions and competition are fundamental in the dynamics of community structure, and delineation of these processes is important in understanding ecological functioning (Livingston, 1982). Additionally, knowledge of trophic dynamics is essential for effective resource management. A frequently discussed topic in modern fisheries science, ‘ecosystem- based management’ (EBM) or ‘fisheries management in an ecosystem context’ requires information on the managed species, as well as those species with which the managed species interact (Link, 2002). Stable isotope analysis is commonly used in studies of trophic structure, particularly those of large, mobile marine taxa that are logistically difficult to sample, or when investigating the relative trophic structure of several species within a given system. Ratios of heavy and light isotopes of carbon (13C/12C, δ13C) and nitrogen (15N/14N, δ15N) in tissues can provide information on dietary carbon sources and trophic position of an individual relative to its conspecifics or other species in the system. Stable carbon isotope ratios generally fractionate at a consistent and low rate (~ 0.4‰; Post, 2002) in trophic exchanges and can be used to infer sources of primary production (e.g. benthic microalgae and macrophytes, pelagic phytoplankton). Nitrogen isotopes, conversely, fractionate relatively predictably between each and can be used to determine the relative position of a consumer in the food web (Peterson and Fry, 1987; Post, 2002). However, potential variation in diet-tissue discrimination factors (DTDFs) between taxa, trophic levels and ecosystems requires careful interpretation of δ15N values when estimating trophic positions (Vanderklift and Ponsard, 2003; Caut et al., 2009; Olin et al., 2013; Hussey et al., 2014). Stable isotope analysis provides an integrated measure of trophic position over rates of tissue turnover, as opposed to generally narrow temporal windows

70 of traditional diet analyses (Kling et al., 1992; Pinnegar and Polunin, 1999). In addition to describing trophic structure, stable isotope analyses can be used to detect trophic shifts due to ontogeny (Graham et al. 2007), movement (Hobson, 1999), and temporal variation in diet (Ben- David et al., 1997); and these techniques have previously been employed to investigate effects of anthropogenic disturbances on food web structure (e.g. Sanpera et al., 2008). Stable isotope analyses are increasingly being used to compare intraspecific measures of isotopic niche space (Newsome et al., 2007), as well as characteristics of the isotopic structure of entire communities (Layman et al., 2007). These methods have been primarily based on conceptual work by Layman and colleagues (2007) and have been advanced to include the propagation of error through a Bayesian framework using the methods developed by Jackson et al. (2011). Unfortunately, these methods are susceptible to complications with small sample sizes (Syvaranta et al., 2013), and conclusions from these analyses must be made with caution, especially with highly mobile species that have mobile and isotopically-variable prey (Cummings et al., 2012). The objectives of this study were to describe the isotopic trophic structure of fishes in a large seagrass ecosystem in the Florida Big Bend to provide insight in community dynamics, and to evaluate potential ontogenetic shifts in dominant taxa and regional patterns in trophic structure. While Livingston (1982) documented the trophic ecology of small teleost fishes in the Big Bend in detail, the larger fishes of this system are relatively poorly studied.

2.2 Materials and Methods 2.2.1 Sample Collection and Preparation White muscle biopsies were collected from specimens captured during the Big Bend fishery-independent gillnet and longline surveys described in the previous chapter. Biopsies were collected from the dorsal musculature with a 3mm biopsy punch and stored on ice until they could be frozen and later processed. Samples were collected from up to ten specimens per species per survey region each year, and an attempt was made to sample across size ranges of any species possible. Biopsies were thawed, rinsed in DI water, and dried for approximately 48 hours at 60°C to reach a constant mass. Samples were ground to a homogenous powder using a mixer mill (SPEX Sample Prep 5100 Mixer Mill). Lipids were not extracted from samples from bony fish,

71 following Post’s (2002) recommendations to only lipid extract when the ratio of carbon concentration to nitrogen concentration is over 3.4. However, lipids and urea were extracted from elasmobranch samples (Hussey et al., 2012a and 2012b). Lipid extraction followed Folch (1957). Each elasmobranch sample was homogenized in a 2:1 chloroform-methanol solution using a MODEL T91 SHAKER, and centrifuged for 2 minutes. The supernatant was then decanted, and a 1:1 solution of methanol and DI water was added to the sample as a rinse. Samples were shaken and centrifuged again, the rinse was decanted, and the tissue was dried at 60°C for 48 hours and again homogenized in a mixer mill. The rinse step in this protocol was found to effectively remove water-soluble, δ15N-depleted urea retained for osmotic balance in elasmobranch tissue (Imhoff et al., in prep), which has been shown to potentially confound trophic positions inferred from stable isotope analysis results of elasmobranchs (Kim and Koch, 2011; Hussey et al., 2012a) through the physiological suppression of δ15N values and therefore inferred relative trophic position. Given the two protocols used for sample processing, untreated samples were standardized using Post’s (2002) lipid extraction normalization equation, which adjusts stable isotope values of untreated tissue according to the ratio of bulk carbon and nitrogen concentrations.

2.2.2 Stable Isotope Analysis Stable isotope analyses were conducted at the National High Magnetic Field Laboratory in Tallahassee, Florida. For each sample, approximately 0.5 to 0.8 mg of homogenized tissue was weighed on a Mettler-Toledo A T21 mass comparator balance (Mettler-Toledo International, Inc., Columbus, OH, USA) and wrapped into 4x6 mm tin capsules. Carbon and nitrogen stable isotope analyses were conducted on a continuous flow ThermoFinnigan Delta Plus mass spectrometer coupled to a CHNO elemental analyzer (Thermo Scientific, Waltham, MA, USA). Calibration curves were constructed by analyzing integrated subsamples of five standard materials (sucrose, phenylalanine, PeeDee Belemnite, and urea in two concentrations) during stable isotope analysis. Stable isotope values are relative to international standards of PeeDee Belemnite (PDB) and Air for carbon and nitrogen, respectively, and are reported in standard δ notation in which units are parts per thousand (‰) differences from the standard:

72

Where X is the stable isotope of C or N (and R is the ratio ) of the heavy to light isotope of the element (i.e. 13C:12C or 15N:14N). Isotope values are considered “enriched” when a sample has a higher ratio of the heavier isotope than the standard and “depleted” when a sample has a lower ratio than the standard. Analytical error of the mass spectrometer was assessed by duplicating every 10th sample in a run of the machine, allowing variation to be assessed across taxa and runs. Mean difference between all duplicate samples was 0.21‰ (SD = 0.5) for δ13C and 0.23‰ (SD = 0.5) for δ15N. Mean difference between duplicate samples did not vary across runs on the mass spectrometer and elemental analyzer (Kruskal-Wallis ANOVA on ranks; δ13C, p = 0.57; δ15N, p = 0.96). Mean differences of δ13C between duplicate samples were slightly higher in bony fishes (0.09±0.7‰; elasmobranchs, 0.07±0.4‰), while mean differences of δ15N between duplicates were higher in elasmobranchs (0.08±0.7‰; bony fishes, 0.04±0.2‰).

2.2.3. Statistical Analysis Stable isotope values were initially plotted and compared qualitatively. Because primary producers were not sampled, inferences based on literature values for primary producers in sub- tropical seagrass beds were compared with δ13C data for all taxa. Size and amount of overlap of Bayesian standard ellipses, a proxy for isotopic niches (Jackson et al., 2011) were calculated for the dominant species. These analyses were conducted using the R software package siar v.4.2 (Jackson et al., 2011; R Core Development Team, 2013). Correlations between δ13C and δ15N, as well as stable isotope values and fish length, were investigated using Pearson’s correlation coefficient. Potential ontogenetic shifts for the three most abundant shark species were also investigated using one-way analysis of variance (ANOVA) on δ13C and δ15N values for each life stage. Differences in δ15N values between species of elasmobranchs were assessed using one-way ANOVA and Tukey’s honestly significant difference (HSD) post-hoc test. Bony fish stable isotope values were compared using bivariate analysis of variance (i.e. MANOVA) for all taxa with at least 10 samples. Regional differences in δ13C and δ15N values were assessed using a two-way MANOVA on taxon and region on pooled values for taxa with at least 6 samples per region, which included only the five dominant species. The dominant species were Atlantic sharpnose sharks (Rhizoprionodon

73 terraenovae), bonnethead sharks (Sphyrna tiburo), blacktip sharks (Carcharhinus limbatus), hardhead catfish (Arius felis), and gafftopsail catfish (Bagre marinus). Regional stable isotope values were also compared using one-way ANOVA for each of the dominant species individually. This set of analyses was performed in the R statistical console using the software packages Hmisc v.3.14-3 (Harrel, Jr., et al., 2014) and MASS v.7.3-30 (Venables and Ripley, 2002). Figures were produced in the R console and using Microsoft Excel© (2010).

2.3 Results 2.3.1 General Results for All Taxa A total of 1287 muscle biopsies were collected from 13 species of elasmobranchs and 53 species of bony fishes from 2009 to 2012. Stable isotope values are summarized for each species in Tables 7 and 8. Mean values with SE are shown in Figures 42 and 43 for elasmobranchs and Figures 44 and 45 for bony fishes. Figures 42 and 43 include mean and standard deviations of primary producers adapted from Moncreiff and Sullivan (2002) and Chanton and Harper (unpublished data). There was a fairly wide range of δ13C values across species (-25.8 to -12.5), as well as δ15N values (5.7 to 17.7). There was also a significant negative relationship between δ13C and δ15N in each of the dominant species individually (for all five species r = -0.40 to -0.57, p < 0.01). This correlation has previously been reported (e.g. Moncreiff and Sullivan, 2002) and is due to differences in fractionation during nitrogen assimilation between primary producers (Peterson and Fry, 1987). Mean values of both δ13C and δ15N varied significantly among years in hardhead catfish (δ13C: F = 5.82, p < 0.01; δ15N: F = 2.48, p = 0.05), with mean δ13C enriched in 2012 relative to 2010 (p = 0.04) and 2011 (p < 0.01) and mean δ15N enriched in 2009 relative to 2010 (p = 0.05). However, the effects sizes of year were relatively low (δ13C: η2 = 0.12; δ15N: η2 = 0.06). Similarly, mean values of δ15N varied significantly among years in bonnethead sharks (F = 2.96, p = 0.08) with mean δ15N enriched in 2009 relative to 2011 (p = 0.04), but the effect size of year was low (η2 = 0.08). Mean stable isotope values did not vary among years in the remaining dominant taxa. Because these patterns were not consistent across taxa and the magnitudes of differences were low, I pooled samples from all years for all remaining analyses. I did not compare lumped values for all species across years as any differences could be due to actual variation between years or a function of the species and individuals captured in a particular year.

74

Young-of-the-year (YOY) Atlantic sharpnose sharks, a placental viviparous species, were highly 15N-enriched relative to adults and juveniles (Table 7), a phenomenon observed in studies of other matrotrophic sharks (Olin et al., 2011). Due to the effects of this maternal enrichment, all YOY sharks were removed for statistical analyses. This removed the three spinner sharks from all analyses, as only YOY individuals of this species were captured. Only one smooth butterfly ray was sampled. Sample sizes were low for great hammerhead sharks (N = 3) and cownose rays (N = 4), however these species were included in most analyses. Sample size was highly variable across bony fish species. There was significant variation (F = 6.20, p < 0.01) in mean δ13C values among some elasmobranch taxa, but the magnitude of effect was small (η2 = 0.12). Tukey’s HSD indicated mean δ13C values of bull and bonnethead sharks were depleted relative to those of blacknose sharks (p = 0.01 for each comparison), implying a difference in the proportions of primary productivity sources between these species. Cownose rays, which had the most 13C-depleted (i.e. lowest) δ13C values (Figure 42), were significantly different from all other species (Table 9); again suggesting this species is supported through a different source of primary productivity. There were also significant differences in mean δ15N values among elasmobranchs (p < 0.01). Cownose rays were removed from this analysis because of their depleted mean δ13C value relative to other elasmobranch taxa and the associated enrichment of baseline δ15N as δ13C decreases (Figure 43). The differences in δ15N observed between primary producers could confound conclusions of relative trophic position inferred from δ15N values among taxa with very different δ13C values, such as cownose rays. Tukey’s HSD pairwise comparisons of mean δ15N values for elasmobranchs are summarized in Table 9. To compare the ‘isotopic niches’ of the dominant species, I calculated sample size- corrected standard ellipses (SEAC) using the methods developed by Jackson et al. (2011). Of the four dominant shark species, bonnethead sharks had the largest SEAC, indicating this taxon is more isotopically diverse, followed by Atlantic sharpnose sharks, blacknose sharks, and blacktip sharks (Table 10, Figure 46). The distinctions in isotopic niches were primarily due to differences in δ15N and suggest differences in relative trophic position, but similarities in sources of primary productivity. Blacktip sharks had the highest δ15N values, bonnetheads the lowest, and those of the other two species were intermediate. These results were expected based on the diets of these taxa (Castro, 1996; Cortes et al., 1996; Cortes, 1999; Bethea et al., 2004; Bethea et

75 al., 2006; Bethea et al., 2007). Overlap in SEAC, which provides a measure of isotopic niche overlap between taxa, was low between blacktip sharks and each other dominant species, and between bonnethead and blacknose sharks (< 3% in each case). There was marginal SEAC overlap between bonnethead and Atlantic sharpnose sharks (4% and 6%, respectively), but there was considerable SEAC overlap between Atlantic sharpnose and blacknose sharks (56% and 63%, respectively).

Values of SEAC were relatively high in both species of ariid catfishes, indicating high isotopic diversity and potentially high trophic diversity. There was very little overlap between both catfish species' SEAC (Figure 47). This was unsurprising, considering gafftopsail catfish feed on higher proportions of portunid crabs and teleost fishes while hardhead catfish feed more heavily on panopeid crabs (Holdridge, 2013). There was some overlap of hardhead catfish and bonnethead shark SEAC (8% and 11%, respectively), both of which feed heavily on crabs (Cortes et al., 1996; Bethea et al., 2007; Holdridge, 2013), and gafftopsail catfish and Atlantic sharpnose sharks (12% and 19%, respectively), whose diets include higher proportions of shrimp and teleost fishes (Bethea et al., 2004; Bethea et al., 2006; Holdridge, 2013). There was relatively high SEAC overlap between gafftopsail catfish and bonnethead sharks (37% and 39%, respectively), again likely due to both species feeding on high proportions of portunid crabs. Due to the greater range of mean δ13C values across bony fish species, δ13C and δ15N were compared for these taxa using one-way MANOVA. I included only species with at least 10 samples in this analysis. There was significant variation in mean δ13C and δ15N values among teleost species (Wilk’s λ = 0.32, F13, 581 = 27.48, p < 0.01). Univariate ANOVAs indicated both δ13C and δ15N varied significantly among teleosts, but the effect of δ15N was slightly greater (δ13C: F = 17.41, p < 0.01, η2 = 0.38; δ15N: F = 24.79, p < 0.01, η2 = 0.44). Values of δ13C for three species were depleted relative to all remaining species: American harvestfish (Peprilus paru), Gulf menhaden (Brevoortia patronus), and yellowfin menhaden (Brevoortia smithi). These three species are presumed to be planktivorous based on the diets of conspecifics in the northwest Atlantic (Bowman et al., 2000) and their depleted δ13C suggest they are indeed supported through phytoplankton production (Moncreiff and Sullivan, 2002). A Tukey HSD test indicated, among the species included in this analysis, there were only significant pairwise differences in δ13C between these three species and all other taxa (p < 0.01 in all cases), and between Spanish mackerel (Scomberomorus maculatus) and grass porgies (Calamus arctifrons),

76 which were enriched relative to Spanish mackerel, (p = 0.03). These results suggest similar sources of primary productivity for the remaining teleosts in this analysis.

2.3.2 Ontogenetic Shifts in Dominant Taxa There was little evidence for ontogenetic shifts in the δ13C and δ15N values of the four dominant shark species for which multiple life stages were captured There were no relationships between fork length (FL) and δ15N or δ13C (Figure 48 and 49, respectively)., nor were there differences in mean δ15N or δ13C of individual life stages for the same species. There were, however, significant relationships between stable isotope values and FL of hardhead and gafftopsail catfishes (Figures 50 and 51, respectively). There was a weak, but significant, positive relationship between hardhead catfish FL and δ13C values (r = 0.40, p < 0.01), suggesting a weak shift in either diet or habitat; however there was no relationship between hardhead catfish FL and δ15N values (r < 0.01, p = 0.98). There was a very weak, but significant, positive relationship between gafftopsail catfish and δ13C values (r = 0.22, p =0.02), as well as a significant positive relationship between FL and δ15N values (r = 0.40, p < 0.01), which may indicate gafftopsail catfish feed at slightly higher relative trophic levels over ontogeny.

2.3.3 Regional variation in Stable Isotope Values Values of δ13C and δ15N for all taxa by region are shown in Figure 52. I compared regional mean δ13C and δ15N values for all species with at least 6 samples per region, which left only the 5 dominant species, shown in Figure 53. Results of a two-way MANOVA indicated significant multivariate effects of both region (Wilk’s λ = 0.69, F = 70.273,615, p < 0.01) and taxon (Wilk’s λ = 0.28, F = 130.824,614, p < 0.01). There was a significant interaction of region and taxon, but it explained a small proportion of the variance (Wilk’s λ = 0.89, F = 2.8412,607, p < 13 0.01). Univariate two-way ANOVAs suggested a greater effect of region on δ C (F3,615 = 73.92, 2 2 p < 0.01, η = 0.28) than taxon (F4,614 = 44.20, p < 0.01, η = 0.16), but a greater effect of taxon 15 2 2 on δ N (F4,614 = 214.96, p < 0.01, η = 0.50) than region (F3,615 = 103.30, p < 0.01, η = 0.13). The effect of the interaction of region and taxon explained a very low proportion of the variance 13 2 15 2 in both cases (δ C: F = 1.9912,607, p = 0.02, η = 0.02; δ N: F = 3.0812,607, p < 0.01, η = 0.02). These results imply there is greater regional variation in δ13C, suggesting shifts in the balance of baseline production across the Big Bend, than among taxa. There is greater variation in δ15N

77 among taxa than across regions, which is expected since these taxa occupy a range of relative trophic positions. Tukey HSD tests showed Hernando mean δ13C was enriched and mean δ15N depleted relative to each other region (p < 0.01 for each comparison). Additionally, mean δ15N was enriched in Crystal River relative to each other region (p < 0.01 for each comparison), but there was not a significant pairwise difference in mean δ15N between St. Marks and Steinhatchee (p = 0.15). There were similar patterns in regional means of δ13C and δ15N values in both species of catfishes (Figure 53), with significant differences in both stable isotope values (hardhead catfish: δ13C, F = 10.83, p < 0.01, η2 = 0.20; δ15N, F = 8.07, p < 0.01, η2 = 0.16; gafftopsail catfish: δ13C, F = 26.55, p < 0.01, η2 = 0.43; δ15N, F = 9.37, p < 0.01, η2 = 0.21). Mean δ13C values of both catfishes were significantly higher in Hernando than each other region (p < 0.01 in all cases). Mean hardhead catfish δ15N values were significantly higher in Crystal River than St. Marks and Steinhatchee (p < 0.01 and p = 0.04, respectively). Similarly, mean gafftopsail catfish δ15N values were significantly lower in Hernando than each other region (p < 0.01 in each case). Patterns of regional mean δ13C and δ15N values in the three dominant shark species were less distinct (Figure 54). Although there were statistically significant effects of region on mean values of both δ13C and δ15N (p < 0.01 for both isotopes in all three species), the pairwise differences among region were small in all three species.

2.4 Discussion 2.4.1 General Results for All Taxa There was considerable variation in δ13C and δ15N within and across species of marine fishes in the Big Bend. These fauna are likely supported by multiple primary production pathways, ranging from seagrass to epiphytic microalgal and pelagic (i.e. phytoplankton) production, based on the reported δ13C and δ15N values for primary producers in subtropical seagrass beds (Chanton and Lewis, 2002; Moncreiff and Sullivan, 2002). While some species, such as the three planktivorous teleosts previously mentioned, likely feed on a food web based solely on pelagic phytoplankton production, other taxa appear to be primarily supported through benthic epiphytic or macroalgal production. The importance of epiphytes in seagrass ecosystems is fairly well-documented (Fry et al., 1987; Moncreiff and Sullivan, 2002). However, δ13C values of individuals of several species suggested a mixture of epiphytic/macroalgal production and

78 phytoplankton or seagrass production, suggesting the relative importance of different primary production pathways may vary among individuals. More thorough stable isotope data for primary producers and analysis using stable isotope mixing models (e.g. Parnell et al., 2010) are required to further investigate the balance of primary production in the Big Bend. There was relatively little variation in mean δ13C values of elasmobranchs. Cownose rays were the most 13C depleted, suggesting this species is primarily supported through pelagic production, however the sample size was small (N = 4). Cownose rays feed on benthic invertebrates, including filter-feeding bivalves, benthic decapods, and worms (Collins et al., 2007), so a δ13C signal reflective of pelagic phytoplankton production is sensible. Pairwise differences between relatively 13C-depleted bull and bonnethead sharks and relatively 13C- enriched blacknose sharks suggest juvenile bull (no mature bull sharks were captured) sharks and bonnethead sharks feed in higher proportions from river-influenced or pelagic-based food webs in the Big Bend than do blacknose sharks (Chanton and Lewis, 2002). The remaining elasmobranch species appeared to be supported primarily through epiphytic and/or macroalgal production. Patterns of elasmobranch δ15N values appeared as expected based on published information on these species’ diets. Southern stingrays and bonnethead sharks, which feed primarily on benthic invertebrates (Gilliam and Sullivan, 1993; Cortes et al., 1996; Bethea et al, 2007), had the lowest δ15N values; whereas Atlantic sharpnose, blacknose, nurse, and lemon sharks, which eat higher proportions of bony fishes (Cortes 1999; Castro, 2000; Bethea et al, 2006)., had slightly higher mean δ15N values. The remaining species (blacktip, tiger, bull, and great hammerhead sharks) had the highest δ15N values and are those that have been reported to have some proportion of elasmobranchs in their diet, in addition to a variety of other vertebrate taxa (e.g. marine reptiles, birds, and mammals) in the case of tiger sharks (Castro, 1993; Cortes, 1999). Bull and great hammerhead sharks had the highest mean δ15N values of the elasmobranchs, and were reported to have the highest proportion of elasmobranchs in their diets (Cortes, 1999). There was considerable overlap in the stable isotope values of bony fishes. This was expected, as high degrees of omnivory and trophic overlap were reported in small teleosts during peaks of high productivity, such as during summer, in Apalachee Bay, and many undergo ontogenetic shifts in food selection (Livingston, 1982). Whereas both δ13C and δ15N were significant factors in a MANOVA on bony fish species with sufficient sample size, taxon had a

79 slightly greater effect on δ15N than δ13C. Isotopic differences between teleost species were greater in δ15N than δ13C, therefore differences in relative trophic positions were stronger drivers than differences in primary production sources, with exception to the strictly planktivorous species. Mean δ13C values were generally lower in bony fish species than the elasmobranchs, with the exception of cownose rays. There are multiple possible explanations for this pattern. Portions of the shark’s diet may simply be absent in this data set, as it contains only fishes. There are also potential differences in stable carbon fractionation between taxa. Although consistent and limited fractionation is assumed when interpreting δ13C values, reported levels of enrichment 13 13 in trophic transfers (i.e. δ Cconsumer - δ Cdiet) range from -3.0‰ to 4.0‰ (Peterson and Fry, 1987; Post, 2002). Furthermore, stable isotope values of long-lived, highly mobile taxa like these coastal sharks likely reflect feeding over a long-term integrated scale determined by taxon- specific tissue turnover rates, which would include time outside of the study area (Hussey et al., 2012a). While comparisons of mean stable isotope values across species can be informative, it is important to consider the ranges and intraspecific variation of δ13C and δ15N. The variations in δ13C observed in many of these fishes, especially in omnivorous and predatory species, indicate these taxa are supported through multiple channels of primary productivity, and likely couple primary production pathways (Chasar et al., 2005; Rooney et al., 2006) and habitats (e.g. inshore and offshore systems) through migration and production subsidies (McCann et al., 2005; Nelson et al., 2012). As noted by several authors, trophic structure in diverse, productive systems inhabited by opportunistic, omnivorous, and mobile species is a spectrum (Darnell, 1961), especially considering ontogenetic progressions of dietary selection (Darnell, 1961; Livingston, 1982). Furthermore, the correlation between δ13C and δ15N in primary producers, and therefore consumers, makes baseline stable isotope values vital in calculations of numerical, even if continuous, trophic positions (Vander Zanden and Rasmussen, 1999; Post, 2002). These analyses are further complicated by multiple potential production pathways, physiological differences in excretion of nitrogenous wastes, and variable diet-tissue discrimination factors (DTDFs) across taxa and prey items (Darnell, 1961; Caut et al., 2009; Hussey et al., 2011, 2012a, and 2014; Olin et al., 2013). Of the dominant sharks, bonnethead sharks appear to be the most isotopically diverse. Bonnethead sharks feed primarily on crabs, especially blue crabs, Calinectes sapidus (Cortes et

80 al., 1996; Bethea et al., 2007). Blue crabs are opportunistic predators and scavengers and are known to go through ontogenetic shifts in diet and stable isotope values (Dittel et al., 2006). These data suggest this variation carries into bonnethead sharks, which may individually specialize on crabs in particular habitats or of certain sizes and do not appear to be convergent with low variation in isotopic signatures. Conversely, blacktip sharks show the lowest degree of isotopic diversity of the four dominant shark species. Juvenile blacktip sharks were reported to eat high proportions of clupeid fishes (primarily menhaden) transitioning from sciaenids and clupeids as YOY in Apalachicola Bay, Florida (Bethea et al., 2004), and are thought to transition to mostly clupeids and small elasmobranchs as adults (Castro 1993). However, blacktip shark mean 13C was considerably enriched relative to those of the planktivorous teleosts, suggesting clupeids may constitute a small proportion, if any, of blacktip shark diets in the Big Bend. Heupel and Hueter (2002) reported a high proportion of sparids in diets of YOY blacktip sharks in Tampa Bay, although a large proportion of teleost material was unidentified in their study. Blacktip sharks may feed on a higher proportion of sparids and sciaenids in the Big Bend, which have relatively enriched 13C values relative to menhaden, and are abundant families in the region.

There was a relatively high degree of overlap in SEAC between Atlantic sharpnose and blacknose sharks, indicating isotopic overlap. These results could be indicative of trophic overlap between these species, although these analyses cannot rule out the possibility of different prey items for each species that are not isotopically distinct. However, Bethea et al. (2004) observed high dietary overlap between early life stages of four sympatric shark species in Apalachicola Bay, and it seems likely that these taxa feed from similar prey bases with some dietary overlap Both species of catfishes had relatively high degrees of isotopic diversity in terms of

SEAC size, but there was little overlap between the two species. The vertical displacements of

SEAC areas of the catfishes suggest gafftopsail catfish feed at a higher trophic level than hardhead catfish. This conclusion is supported by reported diet data of both species, as gafftopsail catfish have been found to feed on higher proportion of portunid crabs, and teleost fishes than hardhead catfish in the Big Bend, which feed more heavily on panopeid crabs (Holdridge, 2013). These analyses and reported diet information suggest trophic overlap between the ariid catfishes and small coastal shark species such as Atlantic sharpnose and bonnethead sharks (Cortes et al., 1996; Bethea et al., 2004; Bethea et al., 2006; Bethea et al., 2007; Holdridge, 2013). . .

81

Diverse resource pools provide opportunities for individual specialization, especially in large predators (Araujo et al., 2011; Matich et al., 2011), and the data in this study suggest potential for individual specialization in some species (e.g. bonnethead sharks); however levels of specialization or generalization in consumers must be evaluated using stable isotopes from multiple tissue types with variable turnover rates or in concert with additional data from gut content analysis (Matthews and Mazumder, 2004; Layman and Post, 2008; Araujo et al., 2011; Matich et al., 2011).

2.4.2 Ontogenetic shifts in Dominant Taxa There were not clear ontogenetic shifts in δ13C and δ15N in the four dominant shark species, despite ontogenetic diet shifts being reported in both Atlantic sharpnose and blacktip sharks in Apalachicola Bay (Bethea et al., 2004). Cortes et al. (2006) found no evidence for ontogenetic shifts in dietary items of bonnethead sharks, but they did report a relationship between shark size and crab prey size. Isotopic ontogenetic shifts would not necessarily be expected in bonnethead sharks, unless ontogenetic shifts were distinct enough between crab sizes and this predator-prey-size relationship was very strong. The lack of distinct ontogenetic shifts in stable isotope values could be related to a variety of factors. Prey species must be isotopically distinct to indicate a shift. The degree of the diet shift must also be sufficiently distinct to alter stable isotope values. Livingston (1982) described ontogenetic shifts as ‘ontogenetic progressions’ during which individuals gradually showed increased selection of different or larger prey with increasing size. A gradual transition in diet over ontogeny may not be evident in stable isotope values. Furthermore, Scharf et al. (2000) reported an asymmetric relationship between predator and prey size in which small prey items were found continually over the ontogeny of predator species, and that those findings corroborated those of other studies. Therefore the continued inclusion of small prey items or prey species consumed when the predator was of small size may isotopically cloud ontogenetic shifts in diet. Ontogenetic shifts in habitat use may also expand the prey breadth of a species, however isotopic detection of such shifts may be confounded by spatial differences in the isotopic baseline (Graham et al., 2007). Finally, an inverse relationship between δ15N DTDFs and dietary δ15N has been demonstrated in recent literature (Caut et al., 2009; Dennis et al., 2010; Olin et al., 2013). Therefore, the relative

82 difference in δ15N between a predator and its prey decreases as prey δ15N increases, a phenomenon which may make dietary ontogenetic shifts difficult to discern isotopically. Although there was little evidence of ontogenetic shifts in the dominant shark species, there was some evidence for ontogenetic shifts in both species of catfishes. There were weak, but statistically significant, positive relationships between hardhead catfish δ13C and FL and gafftopsail catfish δ15N and FL (Figures 7 and 8, respectively). These relationships suggest hardhead catfish may transition from a pelagic-based or river-influenced food web (relatively depleted δ13C) to a benthic-based food web (relatively enriched δ13C) over ontogeny through a shift in either diet, such as from epipelagic shrimp to benthic crabs, or habitat use, such as from an estuarine river mouth to a marine seagrass bed (Chanton and Lewis, 2002; Moncreiff and Sullivan, 2002; Graham et al., 2007). Additionally, gafftopsail catfish may feed at a gradually higher relative trophic position over ontogeny. A relationship between trophic position and FL of hardhead catfish may have been masked by the relationship with δ13C, as δ15N values tended to decrease with increasing δ13C. There was only a very weak relationship between FL and δ13C in gafftopsail catfish, suggesting the pattern of increasing δ15N was not a function of shifting δ13C.

2.4.3 Regional Variation in Stable Isotope Values Mean δ13C was enriched and δ15N depleted in Hernando, the southernmost region, relative to each other region. This pattern was observed among the dominant taxa when pooled by region, and in each species of catfishes alone. There was, however, little evidence of distinct regional effects on stable isotope values in the dominant shark species individually. These highly mobile and migratory species may traverse survey regions, and therefore have stable isotope values that synthesize regional variation. It is also possible that turnover rates, which are functions of metabolic rates and growth and therefore vary among taxa and life stages (Nelson et al., 2011), in these species' muscle tissue are so low that their stable isotope values reflect foraging across regions or outside of the Big Bend (Hussey et al., 2012a). That this pattern is consistent in both species of catfishes, assuming consistent diet of each species among Big Bend regions, suggests a shift in the isotopic baseline or balance of primary productivity in Hernando relative to the other regions. The lower mean δ15N is likely related to generally enriched δ13C, which are correlated across primary producers and in the dominant taxa of this study (Moncreiff and Sullivan, 2002; and see above).

83

The shift in δ13C to a more enriched signal in the southern Big Bend could be related to the balance of primary productivity in the Hernando region, where river influence is relatively low (Zieman and Zieman, 1989). River influence has been found to strongly alter trophic dynamics and production balance in a river-dominated when river flow fell below a threshold (Livingston et al., 1997), shifting the system from one dominated by pelagic primary production with high water column turbidity to a weakly productive, clear water system. Hernando therefore may receive fewer dissolved nutrient subsidies commonly demonstrated in heavily river-influenced systems (Chanton and Lewis, 2002) relative to other parts of the Big Bend, altering the balance of primary production to include a higher proportion of seagrass derived production (13C enriched) relative to phytoplankton or epiphyte derived production (13C depleted). However, there is a large biomass of macroalgae in the southern Big Bend, and a shift in the balance of primary production may not necessarily be due to the water being nutrient poor in that region. It is also possible that there is a shift in the trophic web in which fishes and prey fauna feed more heavily on seagrasses and seagrass fauna, which have relatively enriched δ13C values (Chanton and Lewis, 2002; Moncreiff and Sullivan, 2002), and the proportions of primary production aren’t necessarily different. The remaining regions appear to be supported primarily through an epiphytic and/or macroalgal production base, and these sources appear to dominate the Big Bend seagrass system.

2.4.4 Conclusion Fishes in the Florida Big Bend are isotopically diverse. Values of δ13C suggest these fauna are supported through multiple primary productivity channels, and there is variation in intra- and interspecific δ13C. Values of δ15N corroborate published diet information for elasmobranch fishes. There were not isotopic ontogenetic shifts in dominant elasmobranch taxa; however the timing, progression, and degree of ontogenetic diet shifts, along with inversely proportional DTDFs, may inhibit the ability of these shifts to be demonstrated isotopically. Regional differences in stable isotope values were not evident in elasmobranch species, but there were differences in mean δ13C and δ15N values across regions in both species of ariid catfishes and in the pooled values of dominant taxa. Although the mechanism is unclear, this analysis suggests a shift in the relative balance of primary production sources in the southern region of the Big Bend, which I hypothesize is related to reduced river influence. This work highlights the

84 important notion that stable isotope data from a single tissue type provide interesting information, but conclusions that can be confidently drawn from their patterns are limited. However, the results of this study indicate the Big Bend system is ripe for further study of trophic structure in heterogeneous environments supported by multiple channels of primary production and the potential for individual specialization and habitat-coupling in mobile taxa.

85

Table 7. Summary of elasmobranch fish stable isotope values; SE = standard error; YOY = young-of-the-year, Juv = juvenile, Mat = mature.

Genus species Life Stage N δ13C range δ15N range (common name) (mean±SE) (mean±SE) Carcharhinus acronotus YOY 24 -16.6 - -13.2 10.4 -14.4 (blacknose shark) (-15.1 ± 0.2) (12.2 ± 0.2) Juv 21 -17.3 - -13.7 10.7 - 13.0 (-15.3 ± 0.2) (11.6 ± 0.1) Carcharhinus brevipinna YOY 3 -15.7 - -16.3 11.0 - 15.4 (spinner shark) (-15.1 ± 0.2) 13.8 ± 1.4 Carcharhinus leucas Juv 15 -19.8 - -13.6 12.9 - 17.7 (bull shark) (-16.8 ± 0.4) (13.9 ± 0.3)

Carcharhinus limbatus YOY 29 -17.3 - -13.3 11.7 - 14.3 (blacktip shark) (-15.7 ± 0.1) (13.4 ± 0.1) Juv/Mat 70 -17.6 - -12.8 11.7 - 16.0 (-16.0 ± 0.1) (13.2 ± 0.1) Dasyatis americana YOY 1 -17.8 11.9 (southern stingray) Juv/Mat 7 -17.5 - -14.9 9.8 - 12.4 (-16.4 ± 0.4) (10.7 ± 0.3) Galeocerdo cuvier Juv 11 -17.2 - -14.8 10.4 - 14.4 (tiger shark) (-16.0 ± 0.2) (13.1 ± 0.3) Ginglymostoma cirratum Juv/Mat 8 -16.5 - -12.5 10.5 - 12.7 (nurse shark) (-15.7 ± 0.5) (11.9 ± 0.2) Gymnura micrura - 1 -18.6 13.1 (smooth butterfly ray) Negaprion brevirostris Juv/Mat 19 -17.3 - -13.8 10.7 - 14.1 (lemon shark) (-15.7 ± 0.2) (12.2 ± 0.2) Rhinoptera bonasus Juv/Mat 4 -20.6 - -18.0 8.1 - 10.1 (cownose ray) (-19.5 ± 0.6) (9.7 ± 0.5) Rhizoprionodon terraenovae YOY 30 -17.8 - -13.0 11.9 - 16.2 (Atlantic sharpnose shark) (-16.0 ± 0.2) (14.3 ± 0.2) Juv/Mat 181 -17.9 - -12.6 9.4 - 15.8 (-16.0 ± 0.1) (12.1 ± 0.1) Sphyrna mokarran Juv 3 -16.6 - -14.0 13.1 - 14.1 (great hammerhead shark) (-15.7 ± 0.5) (13.6 ± 0.2) Sphyrna tiburo YOY 5 -18.1 - -13.2 9.2 - 11.3 (bonnethead shark) (-16.9 ± 0.9) (10.7 ± 0.5) Juv/Mat 117 -19.3 - -12.7 9.1 - 13.0 (-16.4 ± 0.1) (10.9 ± 0.1)

86

Table 8: Summary of bony fishes stable isotope values; SE = standard error.

Family Genus species (common name) N δ13C range δ15N range (mean±SE) (mean±SE) Lepisosteidae Lepisosteus osseus (longnose gar) 4 -20.0 - -17.7 12.5 - 15.2 (-18.9±0.5) (13.0±0.5) Elopidae Elops saurus (ladyfish) 16 -19.7 - -15.4 8.8 - 15.6 (-17.8±0.5) (11.3±0.5) Clupeidae Brevoortia patronus (Gulf menhaden) 30 -23.7 - -15.7 9.4 - 12.4 (-20.9±0.3) (10.6±0.2) Brevoortia smithi (yellowfin menhaden) 25 -25.5 - -18.8 8.7 - 12.3 (-20.4±0.3) (10.3±0.1) Opisthonema oglinum (threadfin herring) 2 -19.6 - -19.9 10.9 - 11.0 (-19.8±0.2) (10.9±0.1) Synodontidae Synodus foetens (inshore lizardfish) 3 -15.1 - -17.7 9.1 - 11.1 (-16.5±0.8) (10.1±0.6) Arius felis (hardhead catfish) 132 -20.7 - -12.9 8.1 - 12.5 (-17.3±0.1) (9.8 ± 0.1) Bagre marinus (gafftopsail catfish) 112 -20.0 - -14.8 9.9 - 15.0 (-17.4±0.1) (11.9±0.1) Batrachoididae Opsanus beta (Gulf toadfish) 2 -19.5 - -15.7 8.0 - 11.6 (-17.6±1.9) (9.8±1.8) Ogcocephalidae Ogcocephalus sp. (batfish sp.) 1 -18.5 12.0

Belonidae Tylosurus crocodilus (houndfish) 2 -14.5 - -14.0 7.9 - 10.9 (-14.2±0.2) (9.4±1.5) Serranidae Centropristis striata (black sea bass) 58 -17.1 - -19.3 8.1 - 11.0 (-17.4±0.2) (9.4±0.1) Diplectrum formosum (sand perch) 1 -15.4 6.9

Mycteroperca microlepis (gag grouper) 1 -18.9 13.4

Pomatomidae Pomatomus saltatrix (bluefish) 7 -18.7 - -16.0 9.6 - 15.2 (-17.2±0.3) (13.3±0.7) Rachycentridae Rachycentron canadum (cobia) 17 -18.5 - -14.6 9.8 - 11.8 (-16.5±0.2) (10.7±0.2) Echeneidae Echeneis naucrates (sharksucker) 95 -25.8- -11.5 5.7 - 14.0 (-15.8±0.4) (7.6±0.4)

87

Table 8 - continued

Family Genus species (common name) N δ13C range δ15N range (mean±SE) (mean±SE) Carangidae Caranx crysos (blue runner) 2 -19.2- -18.8 11.5 - 12.2 (-19.0±0.2) (11.8±0.3) Carangidae Caranx hippos (crevalle jack) 7 -16.7- -19.6 10.7 - 13.9 (-18.6±0.4) (12.0±0.5) Caranx latus (horse-eye jack) 3 -19.3- -19.9 12.7 - 13.8 (-19.5±0.2) (13.1±0.4) Selene vomer (lookdown) 1 -17.5 11.3

Trachinotus carolinus (pompano) 2 -21.4- -19.3 8.4 - 12.1 (-20.4±1.0) (10.2±1.8) Haemulidae Haemulon parrai (sailors choice) 1 -19.5 10.1 Haemulon plumieri (white grunt) 3 -17.8- -16.9 9.9 - 10.0 (-17.2±0.3) (10.0±0.0) Orthopristis chrysoptera (pigfish) 10 -21.6- -15.8 7.5 - 10.1 (-17.5±0.5) (9.1±0.3) Sparidae Archosargus probatocephalus 1 -17.5 8.8 (sheepshead) Calamus arctifrons (grass porgy) 11 -18.0- -13.6 7.5 - 10.0 (-16.1±0.4) (8.7±0.2) Diplodus holbrookii (spottail pinfish) 8 -21.0- -15.9 8.6 - 10.1 (-18.9±0.7) (9.4±0.2) Lagodon rhomboides (pinfish) 17 -20.1- -14.0 6.8 - 10.6 (-17.5±0.4) (8.8±0.2) Sciaenidae Bairdiella chrysoura (American silver 4 -16.5- -18.3 8.3 - 12.9 perch) (-17.3±0.4) (10.9±1.0) Cynoscion arenarius (sand weakfish) 1 -19.1 13.6 Cynoscion nebulosus (spotted sea trout) 20 -19.4- -14.5 9.1 - 13.6 (-17.2±0.3) (11.0±0.2) Leiostomus xanthurus (spot) 2 -19.1- -18.5 8.7 - 10.6 (-18.8±0.3) (9.7±1.0) Menticirrhus sp. (kingfish sp.) 1 -17.8 11.1 Micropogonias undulatus (croaker) 1 -22.5 9.4 Pogonias cromis (black drum) 5 -19.4- -16.8 8.5 - 10.3 (-18.1±0.4) (9.4±0.3)

88

Table 8 - continued

Family Genus species (common name) N δ13C range δ15N range (mean±SE) (mean±SE) Sciaenidae Sciaenops ocellatus (red drum) 2 -21.1- -18.6 9.0 - 10.1 (-19.8±1.2) (9.6±0.6) Ephippidae Chaetodipterus faber (Atlantic 6 -22.1- -19.8 8.0 - 11.9 spadefish) (-18.5±0.9) (10.3±0.6) Scaridae Nicholsina usta (emerald parrotfish) 1 -13.6 6.1

Mugilidae Mugil cephalus (striped mullet) 2 -20.2- -19.5 7.9 - 10.0 (-19.9±0.4) (9.0±1.0) Scombridae Scomberomorus maculatus (Spanish 55 -22.1- -19.8 9.5 - 14.0 mackerel) (-18.5±0.9) (10.3±0.6) Stromateidae Peprilus burti (Gulf butterfish) 1 -20.1 11.1

Peprilus paru (American harvestfish) 16 -22.1- -18.4 10.4 - 13.3 (-20.4±0.2) (12.0±0.2) Bothidae Paralichthys albigutta (Gulf flounder) 7 -19.4- -15.7 8.7 - 11.1 (-17.5±0.6) (10.1±0.3) Bothidae Paralichthys lethostigma (southern 1 -15.6 9.1 flounder) Monacanthidae Aluterus scriptus (scrawled filefish) 3 -19.9- -18.5 9.6 - 10.2 (-19.0±0.5) (9.9±0.2) Ostraciidae Lactophrys quadricornis (scrawled 1 -16.2 7.7 cowfish) Diodontidae Chilomycterus schoepfi (striped 1 -15.2 8.4 burrfish)

89

Table 9. Results of Tukey’s HSD tests on mean isotope values of elasmobranchs. Cownose rays were not included in the δ15N test due to their depleted δ13C relative to other taxa. Pairwise comparisons were significant at p < 0.05.

Genus species(common name) δ13C δ15N

Carcharhinus acronotus (blacknose shark) A C

Carcharhinus leucas (bull shark) B A

Carcharhinus limbatus (blacktip shark) AB A

Dasyatis americana (southern stingray) AB D

Galeocerdo cuvier (tiger shark) AB AB

Ginglymostoma cirratum (nurse shark) AB BCD

Negaprion brevirostris (lemon shark) AB BC

Rhinoptera bonasus (cownose ray) C -

Rhizoprionodon terraenovae (Atlantic sharpnose shark) AB C

Sphyrna mokarran (great hammerhead shark) AB A

Sphyrna tiburo (bonnethead shark) B D

90

Table 10. Sample-size corrected standard ellipse area (SEAC) of the four dominant shark species and both catfishes.

Genus species SEAC (common name) (units2) Carcharhinus acronotus 2.16 (blacknose shark)

Carcharhinus limbatus 1.57 (blacktip shark) Rhizoprionodon terraenovae 2.43 (Atlantic sharpnose shark) Sphyrna tiburo 3.70 (bonnethead shark) Ariopsis felis 4.81 (hardhead catfish) Bagre marinus 3.91 (gafftopsail catfish)

Table 11. Pooled regional stable isotope values of dominant species and results of Tukey’s HSD test. Pairwise comparisons were significant at p < 0.05.

Region N δ13C range δ15N range Tukey’s HSD Tukey’s HSD (mean±SE) (mean±SE) δ13C δ15N

St. Marks 178 -20.7 - -15.1 8.1 - 15.8 A B (-17.1 ± 0.1) (11.4 ± 0.1) Steinhatchee 154 -20.2 - -14.5 8.4 - 16.0 A B (-16.9 ± 0.1) (11.6 ± 0.1) Crystal River 153 -20.3 - -12.8 8.6 - 15.8 A A (-17.0 ± 0.1) (12.2 ± 0.1) Hernando 133 -20.5 - -12.6 8.5 - 13.2 B C (-15.2 ± 0.1) (10.7 ± 0.1)

91

15.0

14.0

13.0

12.0 N 15 δ 11.0

10.0

9.0

8.0 -22.0 -21.0 -20.0 -19.0 -18.0 -17.0 -16.0 -15.0 δ13C CACR CLEU CLIM DAME GCUV GCIR

GMIC NBRE RTER SMOK STIB RBON

Figure 42. Mean δ13C and δ15N values of elasmobranchs. Bars denote standard error. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); DAME, Dasyatis americana (southern stingray); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); GMIC, Gymnura micrura (smooth butterfly ray); NBRE, Negaprion brevirostris (lemon shark); RBON, Rhinoptera bonasus (cownose ray); RTER; Rhizoprionodon terraenovae (Atlantic sharpnose shark); SMOK, Sphyrna mokarran (great hammerhead shark); STIB, Sphyrna tiburo (bonnethead shark).

92

18.0

16.0

14.0

12.0

10.0

8.0 Phytoplankton Macroalgae Halodule wrightii 6.0

Epiphytes 4.0

Syringodium filiforme 2.0 Thalassia testudinum

0.0 -23.0 -21.0 -19.0 -17.0 -15.0 -13.0 -11.0 -9.0

CACR CLEU CLIM DAME GCUV GCIR

GMIC NBRE RBON RTER SMOK STIB

Phytoplankton Epiphytes Macroalgae Halodule Thalassia Syringodium

Figure 43. Individual δ13C and δ15N values of elasmobranchs. Juvenile and adult life stages were pooled, and YOY excluded. Mean values with error bars representing standard deviation shown for primary producers. Values for phytoplankton, epiphytes, macroalgae, and Halodule were adapted from Moncreiff and Sullivan (2002). Data for Thalassia and Syringodium were provided by Chanton and Harper (unpublished data). Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLEU, Carcharhinus leucas (bull shark); CLIM, Carcharhinus limbatus (blacktip shark); DAME, Dasyatis americana (southern stingray); GCIR, Ginglymostoma cirratum (nurse shark); GCUV, Galeocerdo cuvier (tiger shark); GMIC, Gymnura micrura (smooth butterfly ray); NBRE, Negaprion brevirostris (lemon shark); RBON, Rhinoptera bonasus (cownose ray); RTER; Rhizoprionodon terraenovae (Atlantic sharpnose shark); SMOK, Sphyrna mokarran (great hammerhead shark); STIB, Sphyrna tiburo (bonnethead shark).

93

14.0

13.0

12.0 N 15 δ 11.0

10.0

9.0

8.0

7.0 -22.0 -21.0 -20.0 -19.0 -18.0 -17.0 -16.0 -15.0 δ13C

AFEL BMAR BPAT BSMI CARC CSTR CNEB ENAU ESAU LRHO OCHR PPAR RCAN SMAC

Figure 44. Mean δ13C and δ15N values of teleost fishes (N > 10). Error bars denote standard error. Only species with N > 5 are shown. Species codes: AFEL, Ariopsis felis (hardhead catfish); BMAR, Bagre marinus (gafftopsail catfish); BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CARC, Calamus arctifrons (grass porgy); CSTR, Centropristis striata (black sea bass); CNEB, Cynoscion nebulosus (spotted seatrout); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); OCHR, Orthopristis chrysoptera (pigfish); PPAR, Peprilus paru (American harvestfish); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

94

14

12

10 N

15 Phytoplankton δ 8

Macroalgae 6 Epiphytes Halodule wrightii

4

Syringodium filiforme 2

Thalassia testudinum 0 -23 -21 -19 -17 -15 -13 -11 -9 δ13C_normalized

AFEL BMAR BPAT BSMI CARC CSTR CNEB ENAU ESAU LRHO OCHR PPAR RCAN SMAC Phytoplankton Epiphytes Macroalgae Halodule Thalassia Syringodium

Figure 45. Lipid-normalized (Post, 2002) mean δ13C and δ15N values of teleost fishes (N > 10). Error bars denote standard error for fishes. Only species with N > 5 are shown. Mean values with error bars representing standard deviation shown for primary producers. Values for phytoplankton, epiphytes, macroalgae, and Halodule were adapted from Moncreiff and Sullivan (2002). Data for Thalassia and Syringodium were provided by Chanton and Harper (unpublished data). Species codes: AFEL, Ariopsis felis (hardhead catfish); BMAR, Bagre marinus (gafftopsail catfish); BPAT, Brevoortia patronus (Gulf menhaden); BSMI, Brevoortia smithi (yellowfin menhaden); CARC, Calamus arctifrons (grass porgy); CSTR, Centropristis striata (black sea bass); CNEB, Cynoscion nebulosus (spotted seatrout); ENAU, Echeneis naucrates (sharksucker); ESAU Elops saurus (ladyfish); LRHO, Lagodon rhomboides (pinfish); OCHR, Orthopristis chrysoptera (pigfish); PPAR, Peprilus paru (American harvestfish); RCAN, Rachycentron canadum (cobia); SMAC, Scomberomorus maculatus (Spanish mackerel).

95

Figure 46. Convex hull and Bayesian standard ellipses (SEAC) of dominant shark species. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark).

96

Figure 47. Convex hull and Bayesian standard ellipses (SEAC) of ariid catfishes. Species codes: AFEL, Ariopsis felis (hardhead catfish); BMAR, Bagre marinus (gafftopsail catfish).

97

18.0 CACR YOY 16.0 JUV 14.0 N 15

δ 12.0 10.0 8.0 20 40 60 80 100 120 140 160 FL (cm)

18.0 CLIM YOY 16.0 JUV 14.0 N 15

δ 12.0 10.0 8.0 20 40 60 80 100 120 140 160 FL (cm)

18.0 RTER YOY 16.0 JUV MAT 14.0 N 15

δ 12.0 10.0 8.0 20 40 60 80 100 120 140 160 FL (cm)

18.0 STIB YOY 16.0 JUV MAT 14.0 N 15

δ 12.0 10.0 8.0 20 40 60 80 100 120 140 160 FL (cm)

Figure 48. δ15N values of each life stage of dominant shark species. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark).

98

-10.0 CACR YOY -12.0 JUV -14.0 C 13

δ -16.0 -18.0 -20.0 20 40 60 80 100 120 140 160 FL (cm)

-10.0 CLIM YOY -12.0 JUV -14.0 C 13

δ -16.0 -18.0 -20.0 20 40 60 80 100 120 140 160 FL (cm)

-10.0 RTER YOY -12.0 JUV MAT -14.0 C 13

δ -16.0 -18.0 -20.0 20 40 60 80 100 120 140 160 FL (cm)

-10.0 STIB YOY -12.0 JUV MAT -14.0 C 13

δ -16.0 -18.0 -20.0 20 40 60 80 100 120 140 160 FL (cm)

Figure 49. δ13C values of each life stage of dominant shark species. Species codes: CACR, Carcharhinus acronotus (blacknose shark); CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark).

99

-10.0

-12.0

-14.0 R² = 0.1623 C

13 -16.0 δ

-18.0

-20.0

-22.0 10 15 20 25 30 35 40 45 50 FL (cm)

13.0

12.0

11.0 N 15 δ 10.0

9.0

8.0 10 15 20 25 30 35 40 45 50 FL (cm)

Figure 50. Hardhead catfish (Ariopsis felis) δ13C and δ15N values by fork length (FL).

100

-10.0

-12.0

-14.0

C R² = 0.0587

13 -16.0 δ

-18.0

-20.0

-22.0 10 20 30 40 50 60 FL (cm)

16.0 15.0 14.0 R² = 0.1519 13.0 N

15 12.0 δ 11.0 10.0 9.0 8.0 10 20 30 40 50 60 FL (cm)

Figure 51. Gafftopsail catfish (Bagre marinus) δ13C and δ15N values by fork length (FL).

101

17.0

15.0

13.0 N 15 δ 11.0

9.0 St. Marks

Steinhatchee 7.0 Crystal River

Hernando 5.0 -26.0 -24.0 -22.0 -20.0 -18.0 -16.0 -14.0 -12.0 δ13C

Figure 52. δ13C and δ15N values of all taxa by region. YOY sharks were removed for this figure.

102

14.0

13.0

12.0 St. Marks N

15 11.0

δ Steinhatchee

10.0 Crystal River Hernando 9.0

8.0 -20.0 -19.0 -18.0 -17.0 -16.0 -15.0 -14.0 δ13C

14.0

13.0

12.0

St. Marks N

15 11.0

δ Steinhatchee Crystal River 10.0 Hernando 9.0

8.0 -20.0 -19.0 -18.0 -17.0 -16.0 -15.0 -14.0 δ13C

Figure 53. Pooled regional means of δ13C and δ15N of dominant taxa (above) and both species of ariid catfishes, individually (below). Bars denote standard error. Gray shapes = gafftopsail catfish (Bagre marinus), white shapes = hardhead catfish (Ariopsis felis).

103

14.0 14.0 CLIM RTER 13.0 13.0 St. Marks St. Marks 12.0 12.0 Steinhatchee Steinhatchee N N

15 11.0 15 11.0 Crystal River Crystal River δ δ 10.0 Hernando 10.0 Hernando 9.0 9.0 8.0 8.0 -20.0 -18.0 -16.0 -14.0 -20.0 -18.0 -16.0 -14.0 δ13C δ13C

14.0 STIB 13.0 St. Marks 12.0 Steinhatchee N

15 11.0 Crystal River δ 10.0 Hernando 9.0 8.0 -20.0 -18.0 -16.0 -14.0 δ13C

Figure 54. δ13C and δ15N values of all life stages of dominant shark species by region. Species codes: CLIM, Carcharhinus limbatus (blacktip shark); RTER, Rhizoprionodon terraenovae (Atlantic sharpnose shark); STIB, Sphyrna tiburo (bonnethead shark).

104

APPENDIX A

RANK ABUNDANCE CURVES AND OMISSION OF RARE SPECIES

The multivariate analyses detailed above are often susceptible to disproportionate influence of rare species, which are consequently removed in many studies (McCune and Grace, 2002). In this analysis, species occurring less than 5 times in each data set (gillnet and longline) were considered rare and omitted for analysis, based on rank abundance curves of each gear type (Figure 42).

Gillnet Rank Abundance 3 2.5 2 1.5 1

Log Log abundance 0.5 0 0 10 20 30 40 50 60 70 Rank

Longline Rank Abundance 3.5 3 2.5 2 1.5 1

Log Log abundance 0.5 0 0 5 10 15 20 25 30 Rank

Figure 55. Rank abundance (log abundance) curves for each gear type. Arrow indicates cutoff for species to be considered rare.

105

APPENDIX B

NMDS DIMENSIONS AND STRESS

Dimensionality in NMDS analyses are evaluated through a balance of ordination stress and number of dimensions, with a goal of reducing stress while maintaining interpretability in the NMDS plot (McCune and Grace, 2002). Scree plots showing stress for NMDS ordinations of increasing dimensions of the data set for each gear type are shown in Figure 42. Numbers of dimensions are often chosen at an inflection point in the scree plot, at which additional dimensions only marginally reduce stress (McCune and Grace, 2002). I chose ordinations of two dimensions, as the stress levels were within interpretable ranges (Clarke, 1993; McCune and Grace, 2002). I evaluated the longline NMDS at three dimensions based on the longline scree plot, but found the results to be approximately the same.

0.15 Gillnet 0.1

Stress Stress 0.05

0 0 1 2 3 4 5 6 7 Dimensions

0.25 Longline 0.2 0.15

Stress Stress 0.1 0.05 0 0 1 2 3 4 5 6 7 Dimensions

Figure 56. Scree plots showing stress at increasing numbers of dimensions in NMDS of data sets for both gear types.

106

APPENDIX C

STATION COORDINATES AND CLUSTER DESIGNATIONS

Table 12. Station number by gear, latitude and longitude, and cluster designation.

Longline Latitude Longitude Longline Gillnet Station Latitude Longitude Gillnet Station Number Cluster Number Cluster 1 30.04650 -84.15587 1 1 30.05010 -84.16432 1 2 29.99860 -84.09680 1 2 29.91940 -83.83702 1 3 29.94610 -83.94620 1 3 29.89357 -83.81048 2 4 29.92320 -83.82452 1 4 29.83692 -83.73972 1 5 29.89670 -83.80478 2 5 29.86288 -83.67600 1 6 29.83410 -83.73573 1 6 29.69147 -83.54053 3 7 29.85680 -83.67327 3 7 29.64633 -83.43127 4 8 29.68890 -83.54253 4 8 29.49908 -83.44585 3 9 29.64180 -83.43747 4 9 29.36038 -83.21610 3 10 29.55320 -83.44207 4 10 29.43292 -83.37257 4 11 29.50330 -83.44475 4 11 29.31675 -83.25655 3 12 29.35790 -83.21595 4 12 29.74772 -83.62542 5 13 29.42780 -83.37102 1 13 29.07337 -82.83753 3 14 29.30780 -83.25507 4 14 28.84833 -82.83017 3 15 29.20640 -83.21375 1 15 28.88877 -82.74070 1 16 29.74410 -83.63195 1 16 28.62735 -82.74447 4 17 29.05090 -82.83872 4 17 28.78105 -82.81570 3 18 28.86150 -82.82465 3 18 29.01793 -82.92245 3 19 28.83650 -82.76888 2 19 28.95040 -82.92667 3 20 28.89620 -82.73370 4 20 29.00187 -82.83355 3 21 28.63680 -82.75072 4 21 28.56025 -82.81862 2 22 28.78820 -82.80820 1 22 28.51103 -82.71593 5 23 29.00990 -82.91130 3 23 28.40757 -82.74970 5 24 28.96880 -83.16963 4 24 28.33152 -82.77935 2 25 29.01260 -82.82002 1 25 28.28908 -82.78972 2

107

Table 12 - continued

Longline Latitude Longitude Longline Gillnet Station Latitude Longitude Gillnet Station Number Cluster Number Cluster 26 28.62690 -82.68893 3 26 28.57050 -82.67213 5 27 28.56000 -82.82463 2 27 28.58090 -82.75743 2 28 28.45940 -82.83220 1 28 28.21392 -82.81230 1 29 28.40200 -82.75300 2 29 28.26842 -82.75517 5 30 28.37070 -82.86067 1 30 28.51554 -82.70904 3 31 28.33970 -82.77223 4 31 28.44326 -82.85034 6 32 28.29990 -82.78730 3 32 28.60677 -82.77242 4 33 28.56500 -82.67645 3 33 28.72207 -82.80463 3 34 28.26100 -82.83635 2 34 29.01094 -82.93638 5 35 28.42850 -82.93877 1 35 29.05939 -82.82788 7 36 28.58880 -82.74855 2 36 28.99930 -82.83361 3 37 28.22296 -82.81850 1 37 29.59441 -83.42195 3 38 28.26596 -82.82700 1 38 29.67869 -83.55734 1 39 28.32771 -82.80995 2 39 29.73719 -83.64440 1 40 28.34450 -82.74989 2 40 29.65369 -83.41032 1 41 28.51028 -82.70554 3 41 29.52296 -83.46761 6 42 28.44771 -82.75853 1 42 29.43312 -83.33330 3 43 28.43599 -82.84534 2 43 29.49140 -83.35978 6 44 28.63508 -82.71877 3 44 29.83179 -83.65167 3 45 28.60831 -82.78222 4 45 29.91143 -83.69579 3 46 28.55361 -82.83365 4 46 29.37968 -83.25715 3 47 28.81393 -82.81526 3 47 29.31550 -83.25535 3 48 29.01318 -82.92664 4 48 30.04670 -84.06584 7 49 29.05140 -82.83003 4 49 30.03011 -83.95721 4 50 28.99072 -82.83543 3 50 29.83574 -83.65783 3 51 29.60311 -83.42222 2 51 29.83525 -83.74155 1 52 29.68277 -83.54929 1 52 29.50791 -83.40658 1 53 29.74490 -83.64461 4 53 29.39207 -83.28867 3 54 29.64603 -83.41274 4 54 29.31667 -83.25673 1

108

Table 12 - continued

Longline Latitude Longitude Longline Gillnet Station Latitude Longitude Gillnet Station Number Cluster Number Cluster 55 29.52723 -83.46352 1 55 29.47258 -83.35716 7 56 29.43913 -83.33990 4 56 29.68944 -83.58556 3 57 29.49784 -83.36159 3 57 29.74415 -83.61945 3 58 29.84541 -83.66553 1 58 30.06077 -83.99717 3 59 29.90200 -83.69230 3 59 30.05171 -84.04427 5 60 29.37427 -83.25951 3 60 29.89926 -83.73180 4 61 29.30872 -83.25249 3 61 29.91839 -83.84063 1 62 29.98980 -84.13905 2 62 29.99077 -84.03342 5 63 30.05223 -84.06039 3 63 30.02390 -83.95532 3 64 30.02386 -83.94980 3 64 29.92862 -83.92632 2 65 29.99468 -84.03924 1 65 29.98495 -83.92622 7 66 29.90201 -83.84696 1 66 30.03918 -84.12274 3 67 29.83045 -83.74593 1 67 29.11099 -82.90301 6 68 29.84607 -83.65427 3 68 29.06261 -81.01411 5 69 29.83187 -83.73972 1 69 29.05235 -82.83674 3 70 29.50948 -83.41197 1 70 28.99788 -82.83598 3 71 29.54074 -83.46547 1 71 28.81577 -82.80367 1 72 29.63619 -83.44038 4 72 28.67027 -82.71658 1 73 29.38138 -83.28238 3 73 28.63148 -82.73660 1 74 29.68830 -83.55863 3 74 28.60264 -82.78058 3 75 29.74027 -83.62550 3 75 28.27004 -82.80838 5 76 30.04914 -84.01876 3 76 28.22638 -82.85212 5 77 30.05171 -84.04427 3 77 28.22421 -82.77249 5 78 29.89548 -83.73067 1 78 28.32890 -82.81979 5 79 29.90518 -83.78772 1 79 28.32577 -82.74821 3 80 29.92309 -83.84921 1 80 30.02428 -84.09556 5 81 29.99205 -84.01649 1 81 30.03403 -84.00230 3 82 29.92532 -83.91557 1 82 29.99579 -84.01115 1 83 29.98514 -83.91113 3 83 29.81833 -83.63589 1

109

Table 12 - continued

Longline Latitude Longitude Longline Gillnet Station Latitude Longitude Gillnet Station Number Cluster Number Cluster 84 30.03090 -84.12756 3 84 29.88259 -83.71976 1 85 29.11756 -82.89325 4 85 29.95110 -83.82032 3 86 29.07078 -83.01973 1 86 29.96947 -83.91325 5 87 29.04804 -82.84704 4 87 29.82134 -83.72691 7 88 29.00748 -82.82668 4 88 29.45147 -83.32220 3 89 28.81534 -82.81129 3 89 29.31708 -83.25697 7 90 28.71194 -82.81053 3 90 29.43378 -83.37579 3 91 28.67537 -82.71942 1 91 29.71850 -83.53075 7 92 28.88840 -82.70660 3 92 29.72331 -83.59678 5 93 28.57720 -82.68992 1 93 29.68105 -83.49062 3 94 28.59646 -82.77572 2 94 29.59504 -83.42696 3 95 28.48331 -82.75859 1 95 29.51109 -83.44376 6 96 28.42864 -82.83588 1 96 29.09994 -83.04228 1 97 28.40166 -82.74123 3 97 29.11011 -82.90474 3 98 28.26052 -82.81080 3 98 29.04650 -82.80070 3 99 28.22071 -82.85554 1 99 28.89606 -82.71001 3 100 28.21765 -82.77667 2 100 28.77072 -82.78479 5 101 28.52487 -82.68543 3 101 28.69367 -82.79548 1 102 30.01591 -84.10544 3 102 28.67961 -82.71040 1 103 30.03350 -84.01619 3 103 28.87507 -82.78715 3 104 29.98818 -84.01914 1 104 28.52652 -82.79031 5 105 29.99879 -83.94802 1 105 28.31932 -82.80813 5 106 29.81110 -83.64227 3 106 28.45113 -82.74966 5 107 29.88577 -83.72838 1 107 28.29760 -82.77951 5 108 29.93842 -83.80867 3 108 28.29381 -82.77136 4 109 29.91760 -83.85459 1 109 28.27239 -82.85172 3 110 29.97747 -83.92078 1 110 28.35355 -82.74306 2 111 29.89626 -83.77396 1 111 28.38673 -82.85252 1 112 29.81994 -83.71429 1

110

Table 12 - continued

Longline Latitude Longitude Longline Gillnet Station Latitude Longitude Gillnet Station Number Cluster Number Cluster 113 29.76786 -83.60730 3 114 29.44210 -83.31937 3 115 29.32204 -83.26537 4 116 29.43925 -83.36066 4 117 29.71113 -83.53168 3 118 29.71782 -83.60851 3 119 29.67306 -83.49669 3 120 29.61948 -83.43332 3 121 29.51936 -83.45775 3 122 29.02171 -82.91922 4 123 29.10724 -83.03343 4 124 29.10901 -82.89178 4 125 29.00856 -82.83696 1 126 29.04437 -82.81114 4 127 28.88831 -82.70716 3 128 28.77361 -82.79729 2 129 28.70111 -82.80544 3 130 28.67632 -82.71581 1 131 28.87224 -82.80103 1 132 28.58539 -82.75458 1 133 28.53888 -82.79193 1 134 28.32844 -82.81882 1 135 28.44343 -82.74340 2 136 28.25920 -82.77812 1 137 28.30379 -82.77530 2 138 28.26656 -82.84126 4 139 28.35666 -82.75288 2 140 28.35666 -82.75288 2

111

APPENDIX D

ACUC APPROVAL

112

REFERENCES

Abel, D.C., R.F. Young, J.A. Garwood, M.J. Travaline, and B.K. Yednock. 2007. Survey of the shark fauna in two South Carolina estuaries and the impact of salinity structure. American Fisheries Society Symposium 50: 109-123.

Araujo, M.S., D.I. Bolnick, C.A. Layman. 2011. The ecological causes for individual specialization. Ecology Letters 14: 948-958.

Ben-David, M., R.W. Flynn, D.M. Schell. 1997. Annual and seasonal changes in the diets of martens: evidence from stable isotope analysis. Oecologia 111: 280-291.

Bethea, D.M., J.A. Buckel, and J.K. Carlson. 2004. Foraging ecology of the early life stages of four sympatric shark species. Marine Ecology Progress Series 268: 245-264.

Bethea, D.M., J.K. Carlson, J.A. Buckel, M. Satterwhite. 2006. Ontogenetic and site-related trends in the diet of the Atlantic sharpnose shark, Rhizoprionodon terraenovae from the northeast Gulf of Mexico. Bulletin of Marine Science 78: 287-307.

Bethea, D.M., L. Hale, J.K Carlson, E. Cortes, C.A. Manire, J. Gelsleichter. 2007. Geographic and ontogenetic variation in the diet and daily ration of the bonnethead shark, Sphyrna tiburo, from the eastern Gulf of Mexico. Marine Biology 152: 1009-1020.

Bowman, R.E., C.E. Stillwell, W.L. Michaels, M.D. Grosslein. 2000. Food of northwest Atlantic fishes and two common species of squid. NOAA Technical Memorandum NMFS-NE- 155.

Carlson, J.K. 2002. Shark nurseries in the northeastern Gulf of Mexico. In: McCandless CT, Pratt HL Jr, Kohler NE (eds) Shark nursery grounds of the Gulf of Mexico and east coast waters of the United States: an overview. An internal report to NOAA’s Highly Migratory Species Office. NOAA Fisheries, Narragansett, RI, 165–182.

Castro, J.I. 1996. Biology of the blacktip shark, Carcharhinus limbatus, off the southeastern United States. Bulletin of Marine Science 59: 508-522.

Castro, J.I. 2000. The biology of the nurse shark, Ginglymostoma cirratum, off the Florida east coast and Bahama Islands. Environmental Biology of Fishes 58: 1-22.

Caut, S., E. Angulo, and F. Courchamp. 2009. Variation in discrimination factors (Δ15N and Δ13C): the effect of diet isotopic values and applications for diet reconstruction. Journal of Applied Ecology 46: 443-453.

Chasar, L.C., J.P. Chanton, C.C. Koenig, and F.C. Coleman. 2005. Evaluating the effect of environmental disturbance on the trophic structure of Florida Bay, U.S.A.: Multiple 113

stable isotope analyses of contemporary and historical specimens. Limnology and Oceanography 50: 1059-1072.

Chanton, J.P., and F.G. Lewis. 2002. Examination of coupling between primary and secondary production in a river-dominated estuary: Apalachicola Bay, Florida, U.S.A. Limnology and Oceanography 47: 683-697.

Clarke, K.R. 1993. Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18: 117-143.

Clarke, R. K., P. J. Somerfield, and M. G. Chapman. 2006. On resemblance measures for ecological studies, including taxonomic dissimilarities and a zero-adjusted Bray–Curtis coefficient for denuded assemblages. Journal of Experimental Marine Biology and Ecology 330:55–80.

Clarke, K.R., Somerfield, P.J., and Gorley R.N. 2008. Testing of null hypothesis in exploratory community analyses: similarity profiles and biota-environment linkage. Journal of Experimental Marine Biology and Ecology 366: 56-69.

Collins, A.B., M.R., Heupel, R.E. Hueter, P.J. Motta. 2007. Hard prey specialist or opportunistic generalists? An examination of the diet of the cownose ray, Rhinoptera bonasus. Marine and Freshwater Research 58: 135-144.

Cortes, E., C.A. Manire, R.E. Hueter. 1996. Diet, feeding habits, and diel feeding chronology of the bonnethead shark, Sphyrna tiburo, in southwest Florida. Bulletin of Marine Science 58: 353-367.

Cortes, E. 1999. Standardized diet compositions and trophic levels of sharks. ICES Journal of Marine Science 56: 707-717.

Cummings, D.O., J. Buhl, R.W. Lee, S.J. Simpson, S.P. Holmes. 2012. Estimating niche width using stable isotope s in the face of habitat variability: a modelling case study in the marine environment. PLOS One 7: 1-14.

Darnell, R.M. 1961. Trophic spectrum of an estuarine community, based on studies of Lake Pontchartrain, Louisiana. Ecology 42: 553-568.

Dennis, C.A., M.A. MacNeil, J.Y. Rosati, T.E. Pitcher, and A.T. Fisk. 2010. Diet discrimination factors are inversely related to δ15N and δ13C of food for fish under controlled conditions. Rapid Communications in mass Spectrometry 24: 3515-3520.

Deepayan, S. 2008. Lattice: Multivariate Data Visualization with R. Springer, New York, NY. 265 pp. Dittel, A.I., C.E. Epifanio, M.L. Fogel. 2006. Trophic relationships of juvenile blue crabs (Calinectes sapidus) in estuarine habitats. Hydrobiologia 568: 379-390.

114

Drymon, J.M., S.P. Powers, J. Dindo, B. Dzwonkowski, and T.A. Henwood. 2010. Distributions of sharks across a continental shelf in the northern Gulf of Mexico. Marine and Coastal Fisheries: Dynamics, Management, and Ecosystem Science 2: 440-450.

Drymon, J.M., L. Carassou, S.P. Powers, M. Grace, J. Dindo, and B. Dzwonkowski. 2013. Multiscale analysis of factors that affect the distribution of sharks throughout the northern Gulf of Mexico. Fishery Bulletin 111: 370-380.

Dufrene, M. and P. Legendre. 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67: 345-366.

Dugan, P.J. and R.J. Livingston. 1982. Long-term variation of macroinvertebrate assemblages in Apalachee Bay, Florida. Estuarine, Coastal, and Shelf Science 14: 391-403.

Faith, D.P., P.R. Minchin, L. Belbin. 1987. Compositional dissimilarity as a robust measure of ecological distance. Vegetatio 69: 57-68.

Field, J.G., K.R. Clarke, and R.M. Warwick. 1982. A practical strategy for analyzing multispecies distribution patterns. Marine Ecology Progress Series 8: 37-52.

Folch, J., M. Lees, and G.H. Sloane Stanley. 1957. A simple method for the isolation and purification of total lipides from animal tissues. The Journal of Biological Chemistry 497-509.

Froeschke, J., G.W. Stunz, and M.L. Wildhaber. 2010. Environmental influences on the occurrence of coastal sharks in estuarine waters. Marine Ecology Progress Series 407: 279-292.

Fry B., S.A. Macko, J.C. Zieman. 1987. Review of stable isotopic investigations of food webs in seagrass meadows. In: Durako MJ, Phillips RC, Lewis RR (eds) Proceedings of a Symposium on Subtropical-Tropical Seagrasses, Southeast United States. Florida Department of Natural Resources, Bureau of Marine Research, St. Petersburg, 117–138 (Florida Mar Resources Publ No. 42)

Gilliam, D. and K.M. Sullivan. 1993. Diet and feeding habits of the southern stingray Dasyatis americana in the central Bahamas. Bulletin of Marine Science 52: 1007-1013.

Graham, B.S., D. Grubbs, K. Holland, B.N. Popp. 2007. A rapid ontogenetic shift in the diet of yellowfin tuna from Hawaii. Marine Biology 150: 647-658.

Grubbs, R.D. and J.A. Musick. 2007. Spatial delineation of summer nursery areas for juvenile sandbar sharks in Chesapeake Bay, Virginia. American Fisheries Society Symposium 50: 63-85. Grubbs, R.D., J.A. Musick, C.L. Conrath, and J.G. Romine. 2007. Long-term movements, migration, and temporal delineation of a summer nursery for juvenile sandbar sharks in the Chesapeake Bay region. American Fisheries Society Symposium 50: 87-107.

115

Harrell, F.E., Jr. 2014. Hmisc: Harrell Miscellaneous. R package version 3.14-3. http://CRAN.R-project.org/package=Hmisc

Heupel, M.R. 2007. Exiting Terra Ceia Bay: an examination of cues stimulating migration from a summer nursery area. American Fisheries Society Symposium 50: 265-280.

Heupel, M.R., J.K. Carlson, and C.A. Simpfendorfer. 2007. Shark nursery areas: concepts, definition, characterization and assumptions. Marine Ecology Progress Series 337: 287- 297.

Heupel, M.R. and R.E. Hueter. 2002. Importance of prey density in relation to the movement patterns of juvenile blacktip sharks (Carcharhinus limbatus) within a coastal nursery area. Marine and Freshwater Research 53: 543-550.

Heupel, M.R. and C.A. Simpfendorfer. 2002. Estimation of mortality of juvenile blacktip sharks, Carcharhinus limbatus, within a nursery area using telemetry data. Canadian Journal of Fisheries and Aquatic Sciences 59: 624-632.

Heupel, M.R., C.A. Simpfendorfer, A.B. Collins, and J.P. Tyminski. Residency and movement patterns of bonnethead sharks, Sphyrna tiburo, in a large Florida estuary. Environmental Biology of Fishes 76: 47-67.

Heithaus, M.R., B.K. Delius, A.J. Wirsing, and M.M. Dunphy-Daly. 2009. Physical factors influencing the distribution of a top predator. Limnology and Oceanography 54: 472-482.

Holdridge, E.M. 2013. Trophic ecology of ariid catfishes in the Gulf of Mexico. Honors thesis. Florida State University: U.S.A.

Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous inference in general parametric models. Biometrical Journal 50: 346--363.

Hothorn, T., K. Hornik, M.A. van de Wiel, and A. Zeileis. 2006. A Lego system for conditional inference. The American Statistician 60: 257-263.

Hueter, R.E. and J.P. Tyminski. 2007. Species-specific distribution and habitat characteristics of shark nurseries in Gulf of Mexico waters off peninsular Florida and Texas. American Fisheries Society Symposium 50: 193-223.

Hussey, N.E., S.F.J. Dudley, I.D. McCarthy, G. Cliff, A.T. Fisk. 2011. Stable isotope profiles of large marine predators: viable indicators of trophic position, diet, and movement in sharks? Canadian Journal of Fisheries and Aquatic Sciences 68: 2029-2045.

Hussey, N.E., M.A. MacNeil, J.A. Olin, B.C. McMeans, M.J. Kinney, D.D. Chapman, and A.T. Fisk. 2012a. Stable isotopes and elasmobranchs: tissue types, methods, applications, and assumptions. Journal of Fish Biology 80: 1449-1484.

116

Hussey, N.E., J.A. Olin, M.J. Kinney, B.C. McMeans, and A.T. Fisk 2012b. Lipid extraction effects on stable isotope values (δ13C and δ15N) of elasmobranch muscle tissue. Journal of Experimental Marine Biology and Ecology 434: 7-15.

Hussey, N.E., M.A. MacNeil; B.C. McMeans, J.A. Olin, S.F.J. Dudley, G. Cliff, S.P Wintner, S.T. Fennessy, and A.T. Fisk. 2014. Rescaling the trophic structure of marine food webs. Ecology Letters 17: 239-250.

Iverson, R.L. and H.F. Bittaker. 1986. Seagrass distribution in the eastern Gulf of Mexico. Estuarine Coast Shelf Science 22: 577-602.

Jackson, A.L., R. Inger, A.C Parnell, S. Bearhop. 2011. Comparing isotopic niche widths among and within communities: SIBER - stable isotope Bayesian ellipses in R. Journal of Animal Ecology 80: 595-602.

Kim, S.L. and P.L. Koch. 2012. Methods to collect, preserve, and prepare elasmobranch muscle tissue for stable isotope analysis. Environmental Biology of Fishes 95: 53-63.

Kincaid, T.M., and A.R. Olsen. 2012. spsurvey: Spatial survey design and analysis. R package version 2.5. http://CRAN.R-project.org/package=spsurvey

Knip, D.M., M.R. Heupel, and C.A. Simpfendorfer. 2010. Sharks in nearshore environments: models, importance, and consequences. Marine Ecology Progress Series 402: 1-11.

Kling, G.W., B. Fry, and W.J. O’Brien. 1992. Stable isotopes and planktonic trophic structure in Arctic lakes. Ecology 73: 561-566.

Kruskal, W.H. and W.A. Wallis. 1952. Use of ranks in one-criterion variance analysis. Journal of the American Statistical Association 47: 583-681.

Layman, C.A., D.A. Arrington, C.G. Montana, and D.M. Post. 2007. Can stable isotope ratios provide for community-wide measures of trophic structure? Ecology 88: 42-48.

Layman, C.A. and D.M. Post. 2008. Can stable isotope ratios provide for community-wide measures of trophic structure? Reply. Ecology 89: 2358-2359.

Link, J.S. 2002. What does ecosystem-based fisheries management mean? Fisheries 27: 18-21.

Livingston, R.J., 1982, Trophic organization of fishes in a coastal seagrass system Marine Ecology Progress Series 7: 1-12.

Maechler, M., P. Rousseeuw, A. Struyf, M. Hubert, and K. Hornik. 2013. cluster: Cluster Analysis Basics and Extensions. R package version 1.14.4. http://CRAN.R- project.org/package=cluster

117

Matich, P., M.R. Heithaus, C.A. Layman. 2011. Contrasting patterns of individual specialization and trophic coupling in two marine apex predators. Journal of Animal Ecology 80: 294- 305.

Matthews, B. and A. Mazumder. 2004. A critical evaluation of intrapopulation variation in δ13C and isotopic evidence of individual specialization. Oecologia 140: 361-371.

Mattson, R. A., T. K. Frazer, J. A. Hale, S. Blitch, and L. Ahijevych. 2006. Florida Big Bend. USGS.

McCandless, C.T., N.E. Kohler, and H.P. Pratt, Jr. 2007. Species-specific shark nursery habitat summary tables. American Fisheries Society Symposium 50: 365-390.

McCann, K.S., J.B. Rasmussen, and J. Umbanhowar. 2005. The dynamics of spatially coupled food webs. Ecology Letters 8: 513-523.

McCune, B. and J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, Oregon. 300 pp.

Musick, J.A., G. Burgess, G. Cailliet, M. Camhi, and S. Fordham. 2000. Management of sharks and their relatives (). Fisheries 25: 9-13.

Moncreiff, C.A. and M.J. Sullivan. 2002. Tropic importance of epiphytic algae in subtropical seagrass beds: evidence from multiple stable isotope analyses. Marine Ecology Progress Series 21: 93-106.

Morrissey, J.F. and S.H. Gruber. 1993. Habitat selection by juvenile lemon sharks, Negaprion brevirostris. Environmental Biology of Fishes 38: 311-319.

National Marine Fisheries Service (NMFS). 1993. Fishery management plan for sharks of the Atlantic Ocean. National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Office of Sustainable Fisheries, Highly Migratory Species Management Division, Silver Spring, MD. Public Document.

Nelson, J., J. Chanton, F. Coleman, C. Koenig. 2011. Patterns of stable carbon turnover in gag, Mycteroperca microlepis, an economically important marine piscivore determined with a non-lethal surgical biopsy procedure. Environmental Biology of Fishes 90: 243-252.

Nelson, J., R. Wilson, F. Coleman, C. Koenig, D. DeVries, C. Gardner, J. Chanton. 2012. Flux by fin: fish mediated carbon and nutrient flux in the northeastern Gulf of Mexico. Marine Biology 159: 365-372.

Newsome, S.D., C. Martinez del Rio, S. Bearhop, D.L. Phillips. 2007. A niche for isotopic ecology. Frontiers in Ecology and the Environment 5: 429-436.

118

Oksanen, J., F.G. Blanchet, R. Kindt, P. Legendre, P.R. Minchin, R.B. O'Hara, G.L. Simpson, P. Solymos, M., Henry, H. Stevens, and H. Wagner. 2013. vegan: Community Ecology Package. R package version 2.0-9. http://CRAN.R-project.org/package=vegan

Olin, J.A., N.E. Hussey, M.W. Fritts, M.R. Heupel, C.A. Simpfendorfer, G.A. Poulakis, and A.T. Fisk. 2011. Maternal meddling in neonatal sharks: implications for interpreting stable isotopes for young animals. Rapid Communications in Mass Spectrometry 25: 1008- 1016.

Olin, J.A., N.E. Hussey, A. Grgicak-Mannion, M.W. Fritts, S.P. Wintner, and A.T. Fisk. 2013. Variable δ15N diet-tissue discrimination factors among sharks: implications for trophic position, diet and food web models. PLOS One 8: 1-11.

Papastamatiou, Y.P., B.M. Wetherbee, C.G. Lowe, and G.L. Crow. 2006. Distribution and diet of four species of carcharhinid shark in the Hawaiian Islands: evidence for resource partitioning and competitive exclusion. Marine Ecology Progress Series 320: 239-251.

Parnell, A.C., R. Inger, S. Bearhop, A.L. Jackson. 2000. Source partitioning using stable isotopes: coping with too much variation. PLOS One 5: 1-5.

Parson, G.R. and E.R. Hoffmayer. 2005. Seasonal changes in the distribution and abundance of the Atlantic sharpnose shark, Rhizoprionodon terraenovae, in the north-central Gulf of Mexico. Copeia 2005: 914-920.

Parsons, G.R. and E.R. Hoffmayer. 2007. Identification and characterization of shark nursery grounds along the Mississippi and Alabama Gulf . American Fisheries Society Symposium 50: 301-315.

Peterson, B.J. and B. Fry. 1987. Stable isotopes in ecosystem studies. Annual Review of Ecology and Systematics 18: 293-320.

Pinnegar, J.K. and N.V.C. Polunin. 1999. Differential fractionation of δ13C and δ15N among fish tissues: implications for the study of trophic interactions. Functional Ecology 13: 225- 231.

Post, D.M. 2000. Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83: 703-718.

R Development Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.org/.

Roberts, D.W. 2013. labdsv: Ordination and Multivariate Analysis for Ecology. R package version 1.6-1. http://CRAN.R-project.org/package=labdsv

119

Rooney, N., K. McCann, G. Gellner, and J.C. Moore. 2006. Structural asymmetry and the stability of diverse food webs. Nature 442: 265-269.

Sanpera, C., S. Valladares, R. Moreno, X. Ruiz, and L. Jover. 2008. Assessing the effects of the Prestige oil spill on the European shag (Phalacrocorax aristotelis): trace elements and stable isotopes. Science of the Total Environment 407: 242-249.

Scharf, F.S., F. Juanes, R.A. Rountree. 2000. Predator size - prey size relationships of marine fish predators: interspecific variation and effects of ontogeny and body size on trophic niche-breadth. Marine Ecology Progress Series 208: 229-248.

Simpson, G.L. 2012. permute: Functions for generating restricted permutations of data. R package version 0.7-0. http://CRAN.R-project.org/package=permute

Stevens, D.L., and A.R. Olsen. 2004. Spatially balanced sampling of natural resources. Journal of American Statistical Association 99: 262-278.

Syvaranta, J., A. Lensu, T.J. Marjomaki, S. Oksanen, and R. I. Jones. 2013. An empirical evaluation of the utility of convex hull and standard ellipse areas for assessing population niche widths from stable isotope data. PLOS One 8: 1-8.

Tuckey, T.D., and M. Dehaven. 2006. Fish assemblages found in tidal-creek and seagrass habitats in the Suwannee River estuary. Fisheries Bulletin 104: 102-117.

Ubeda, A.J., C.A. Simpfendorfer, and M.R. Heupel. 2009. Movements of bonnetheads, Sphyrna tiburo, as a response to salinity change in a Florida estuary. Environmental Biology of Fishes 84: 293-303.

Vanderklift, M.A. and S. Ponsard. 2003. Sources of variation in consumer-diet and δ15N enrichment: a meta-analysis. Oecologia 136: 169-182.

Vander Zanden, M.J. and J.B. Rasmussen. 1999. Consumer δ13C and δ15N and the trophic position of aquatic consumers. Ecology 80: 1395-1404.

Venables, W. N. and B.D. Ripley. 2002. Modern Applied Statistics with S. Fourth Edition. Springer, New York. ISBN 0-387-95457-0

Wei, C., G.T. Rowe, R.L Haedrich, G.S. Boland. 2012. Long-term observations of epibenthic fish zonation in the deep northern Gulf of Mexico. PLOS One 7: 1-14.

Wickham, H. 2009. ggplot2: elegant graphics for data analysis. Springer New York, New York. 213 pp. Whitaker, D. and M. Christman. 2010. clustsig: Significant Cluster Analysis. R package version 1.0. http://CRAN.R-project.org/package=clustsig

120

Zieman, J.C. and R.T. Zieman. 1989. The ecology of the seagrass meadows of the west coast of Florida: a community profile. U.S. Fish and Wildlife Service Biological Report 85(7.25). 155 pp.

121

BIOGRAPHICAL SKETCH

Curriculum Vitae

CHESTON T. PETERSON

Education  M.Sc. - Florida State University - Biological Science, anticipated 2013 Thesis title: Trophic ecology of sharks and teleost fishes in the Florida Big Bend as determined by stable isotope analysis; Advisor: Dr. R. Dean Grubbs

 B.S. - University of North Carolina Wilmington - Marine Biology, cum laude, 2009 Honor’s thesis title: Dental sexual dimorphism in bluntnose stingrays, Dasyatis say: implications for male feeding ability, Advisor: Dr. Thomas Lankford

Research Experience  Florida State University Coastal and Marine Laboratory, St. Teresa, FL - Elasmobranch Ecology Laboratory: M.S. Research Assistant, August 2010-present Supervisor: Dr. R. Dean Grubbs

 North Carolina State University, Center for Marine Sciences and Technology, Morehead City, NC - Zoology Laboratory: Research technician, March-July 2010 Supervisors: Dr. Jeff Buckel and Sarah Friedl (M.Sc. student)

 University of North Carolina Wilmington, Ichthyology Laboratory, Wilmington, NC: Honors student, Jan. 2008-May 2009 Advisor: Dr. Thomas Lankford

 Mote Marine Laboratory, Center for Shark Research, Sarasota, FL: Intern, Jan-Aug. 2008. Supervisor: Jack Morris

 University of North Carolina Wilmington, Fisheries Ecology Laboratory, Wilmington, NC: Volunteer, Aug. 2007-May 2008 Supervisor: Dr. Fred Scharf

 University of North Carolina Wilmington, Marine Biofluiddynamics and Ecology Lab, Wilmington, NC: Laboratory assistant, May 2007-Aug. 2007. Supervisor: Dr. Chris Finelli

Teaching Experience 122

 Florida State University, Department of Biological Science, Tallahassee, FL Teaching assistant, Aug. 2010-July 2011 o Tropical Marine Ecology, BSC 4933-003, Summer 2011 o Elasmobranch Fish Biology and Ecology, BSC 4933-004, Summer 2011 o Animal Diversity Laboratory, BSC 2011L, Fall 2010 and Spring 2011 Presentations

2013 Peterson, Cheston T., Sharks and large teleost fishes of Florida’s Big Bend: a preliminary analysis of distribution, abundance, and community structure. American Elasmobranch Society meeting contributed paper (oral presentation), Albuquerque, NM

2013 Peterson, Cheston T., Sharks and large teleost fishes of Florida’s Big Bend: a preliminary analysis of distribution, abundance, and community structure. FSU Biology Dept. Ecology and Evolution Seminar, Tallahassee, FL

2012 Peterson, Cheston T., R. Dean Grubbs, Investigating the trophic ecology of sharks and teleost fishes in the Florida Big Bend using stable isotope analysis. American Society of Ichthyologists and Herpetologists (poster), Vancouver, British Columbia, Canada.

2011 Peterson, Cheston T., R. Dean Grubbs, Investigating the relationship between two diskfishes, the sharksucker (Echeneis naucrates) and the remora (Remora remora) and their shark hosts using stable isotope analysis. William and Lenore Mote Symposium (poster), Sarasota, FL

2011 Peterson, Cheston T., R. Dean Grubbs, Investigating the trophic ecology of sharks and teleost fishes in the Florida Big Bend using stable isotope analysis. American Fisheries Society, Larval Fish Conference (poster), Wilmington, NC

2011 Peterson, Cheston T., R. Dean Grubbs, Investigating the trophic ecology of sharks and teleost fishes in the Florida Big Bend using stable isotope analysis. Southeastern Ecology and Evolution Conference (poster), Auburn, AB

2009 Peterson, Cheston T. Implications of Seasonally Sexually Dimorphic Dentition in Regard to Feeding Ability and Effects of Tooth Damage in the Bluntnose Stingray, Dasyatis say. UNC Wilmington CMS (Honors thesis defense), Wilmington, NC

Awards and Research Funding  2011 FSU Coastal and Marine Laboratory Graduate Research Scholarship - $800  2011 FSU Coastal and Marine Laboratory Graduate Research Scholarship - $1000  2010 Guy Harvey Excellence Award, Graduate Research Fellowship - $5000

Society Memberships and Service

123

 FSU Ecology and Evolution Discussion and Reading Group o Social chair (elected), 2011-2012, 2012-2013  American Elasmobranch Society 2009-present o Corporate Donations Manager on Student Affairs Committee, 2013  American Society for Ichthyologists and Herpetologists 2009-2010  UNCW Subunit of the American Fisheries Society 2006-2009 o Secretary (elected), 2007-2009

Outreach  Co-manager of ‘Shark Biology’ table at the 2012 Florida Seafood Festival  Member of ‘Experienced TA Panel’ for 2012 FSU Biology Teaching Workshop  Judge for 2011 Capital Regional Science and Engineering Fair - Tallahassee, FL

124