USF Libraries
USF Digital Collections

Assessing the health of coral reef ecosystems in the Florida Keys at community, individual and cellular scales

MISSING IMAGE

Material Information

Title:
Assessing the health of coral reef ecosystems in the Florida Keys at community, individual and cellular scales
Physical Description:
Book
Language:
English
Creator:
Fisher, Elizabeth M
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Bioindicators
Biomarkers
Foraminifera
Montastraea
Regeneration
Dissertations, Academic -- Marine Science -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Coral reefs are threatened in Florida and worldwide. Successful resource management requires rapid identification of anthropogenic sources of stress before they affect the reef community. I tested a multi-scale approach for assessing reef condition at seven reefs within the Florida Keys National Marine Sanctuary and Biscayne National Park between 2001 and 2003. I examined multiple environmental parameters to identify potential sources of stress. I utilized the Atlantic and Gulf Rapid Reef Assessment Biotic Reef Index to assess benthic community structure and an indicator species of Foraminifera (Amphistegina gibbosa) to determine if environmental conditions were suitable for calcareous organisms that host algal endosymbionts. Small tissue samples were extracted from colonies of Montastraea annularis species complex to assay a suite of cellular biomarkers to elucidate possible mechanisms of the coral stress response. I monitored regeneration rates of the resultant lesions to determine if the coral colonies were capable of recovering from damage. Multivariate data analyses indicated that corals at all study sites were experiencing stress with different degrees of response and decline. On reefs with coarse grain sediments that are adjacent to an intact mangrove shoreline, the Cellular Diagnostic System indicated that corals were responding to a xenobiotic stress but appeared to be compensating as evidenced by consistently high lesion regeneration rates, a high percentage of healed lesions, low coral mortality and high abundances of A. gibbosa. On reefs with silt-sized sediments adjacent to developed coastlines, corals also were responding to xenobiotic stresses, but were negatively affected as evidenced by low regeneration rates, a low percentage of healed lesions, high coral mortality, and low abundances of A. gibbosa. Corals at an 18 m offshore site exhibited abnormally low biomarker levels and some died during the study, indicating that sampled colonies were incapable of upregulating necessary protective proteins. Further research will be required to determine stressor sources. This study demonstrates that a multiple-indicator approach, spanning scales from cellular to community, can provide marine resource managers with data linking decline of coral populations to specific environmental conditions and events, thereby providing potential for early detection of stressors allowing for preventive management.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Elizabeth M. Fisher.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 267 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001914305
oclc - 175300220
usfldc doi - E14-SFE0001972
usfldc handle - e14.1972
System ID:
SFS0026290:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001914305
003 fts
005 20071022152355.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071022s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001972
035
(OCoLC)175300220
040
FHM
c FHM
049
FHMM
090
GC11.2 (ONLINE)
1 100
Fisher, Elizabeth M.
0 245
Assessing the health of coral reef ecosystems in the Florida Keys at community, individual and cellular scales
h [electronic resource] /
by Elizabeth M. Fisher.
260
[Tampa, Fla.] :
b University of South Florida,
2007.
3 520
ABSTRACT: Coral reefs are threatened in Florida and worldwide. Successful resource management requires rapid identification of anthropogenic sources of stress before they affect the reef community. I tested a multi-scale approach for assessing reef condition at seven reefs within the Florida Keys National Marine Sanctuary and Biscayne National Park between 2001 and 2003. I examined multiple environmental parameters to identify potential sources of stress. I utilized the Atlantic and Gulf Rapid Reef Assessment Biotic Reef Index to assess benthic community structure and an indicator species of Foraminifera (Amphistegina gibbosa) to determine if environmental conditions were suitable for calcareous organisms that host algal endosymbionts. Small tissue samples were extracted from colonies of Montastraea annularis species complex to assay a suite of cellular biomarkers to elucidate possible mechanisms of the coral stress response. I monitored regeneration rates of the resultant lesions to determine if the coral colonies were capable of recovering from damage. Multivariate data analyses indicated that corals at all study sites were experiencing stress with different degrees of response and decline. On reefs with coarse grain sediments that are adjacent to an intact mangrove shoreline, the Cellular Diagnostic System indicated that corals were responding to a xenobiotic stress but appeared to be compensating as evidenced by consistently high lesion regeneration rates, a high percentage of healed lesions, low coral mortality and high abundances of A. gibbosa. On reefs with silt-sized sediments adjacent to developed coastlines, corals also were responding to xenobiotic stresses, but were negatively affected as evidenced by low regeneration rates, a low percentage of healed lesions, high coral mortality, and low abundances of A. gibbosa. Corals at an 18 m offshore site exhibited abnormally low biomarker levels and some died during the study, indicating that sampled colonies were incapable of upregulating necessary protective proteins. Further research will be required to determine stressor sources. This study demonstrates that a multiple-indicator approach, spanning scales from cellular to community, can provide marine resource managers with data linking decline of coral populations to specific environmental conditions and events, thereby providing potential for early detection of stressors allowing for preventive management.
502
Dissertation (Ph.D.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 267 pages.
Includes vita.
590
Advisor: Pamela Hallock Muller, Ph.D.
653
Bioindicators.
Biomarkers.
Foraminifera.
Montastraea.
Regeneration.
690
Dissertations, Academic
z USF
x Marine Science
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1972



PAGE 1

Assessing the Health of Co ral Reef Ecosystems in the Florida Keys at Community, Individual, and Cellular Scales by Elizabeth M. Fisher A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Marine Science University of South Florida Major Professor: Pamela Hallock Muller, Ph.D. John E. Fauth, Ph. D. Walter Jaap, B. S. Joseph Torres, Ph. D. Cheryl M. Woodley, Ph. D. Date of Approval: March 23, 2007 Keywords: bioindicators, biomarkers, foraminifera, Montastraea regeneration Copyright 2007 Elizabeth M. Fisher

PAGE 2

Dedication This is dedicated to my fianc Chris, my parents and fa mily for their constant patience, love and support.

PAGE 3

Acknowledgments I thank my advisor, Pamela Hallock Muller who provided endless input, support and encouragement. I thank my committee members, John Fauth, Walter Jaap, Jose Torres and Cheryl Woodley for providing valuab le input along the way. Craig Downs of Haereticus Environmental Labo ratory (formally EnVirtue Biotechnologies) conducted all of the cellular biomarker assays, assisted wi th interpretation of cellular diagnostic data, and participated in the fieldwor k. I also thank K. R. Clarke and R. Muller for statistical assistance and B. Keller for his comments on the manuscript. Underwater camera equipment was provided by Florida Fish and Wildlife Research Institute. This work would not have been possible without field as sistance from E. Carnahan, R. Curry, J. C and J. F. Halas, G. Klungness, C. Moses, S. Ryan, and S. Viehman, as well as members of the Reef Indicators Lab in the College of Marine Science, USF who also provided laboratory assistance. Boating assistance was provided by NOAA Florida Keys National Marine Sanctuary, Biscayne National Park, U. S. Geological Survey and Quiescence. J. Bellucci, M. Callahan, C. Dreher, D. Hickey and M. Lybolt ably assisted with the community assessments. James Ivey of the Ocean Optics Laboratory at University of South Florida provided Figure 2.16A. Housing was provided by NOAAs National Undersea Research Center and Everglades Natio nal Park. This work was funded, in part, by National Sea Grant, Environmental Ma rine Biotechnology Awa rd No. NA86RG0052, Am.7.1 (CM Woodley, PI), by NOAA-NURC-UNC W subcontract 2004-19B (P Hallock Muller, PI), and by NSF Grant DEB9727039 (JE Fauth, PI). Support also was provided by the University of South Florida, College of Marine Science Sanibel-Captiva Shell Club Endowed Fellowship and the NOAA Hurri cane Alliance Grant (Thomas Mason and P Hallock Muller, PIs). Th is research was conducted under permit numbers BISC-2001SCI-0022, BISC-2002-SCI-0012 and BISC-2003SCI-0019 from the National Park Service and FKNMS-2001-008 fr om the Florida Keys National Marine Sanctuary.

PAGE 4

i Table of Contents List of Tables v List of Figures xii Abstract xxii 1. Introduction 1 1.1. Reef Degradation 1 1.2. Florida Reef Tract 2 1.3. Need for New Methodologies 4 1.4. Overview of Dissertation Research 4 1.4.1. Approach 4 1.4.2. Description of Sampling Sites 6 1.4.3. Biology and Ecology of Montastraea annularis complex 6 1.4.4. Specific Objective of Dissertation 8 2. Environmental Assessments 14 2.1. Introduction 14 2.1.1. Coastal Wetlands 14 2.1.2. Sedimentation and Turbidity 15 2.1.3. Temperature and Light 17 2.1.4. Nutrients 18 2.2. Methods 20 2.2.1. Site Descriptions 20 2.2.2. Sedimentation/Turbidity 20 2.2.3. Temperature 20 2.2.4. Nutrients 21 2.2.5. Additional Environmental Data 21 2.2.6. Data Analysis 22 2.3. Results 22 2.3.1. Coastal Wetlands 22 2.3.2. Sedimentation 22 2.3.3. Turbidity 23 2.3.4. Temperature 24 2.3.5. Nutrients 24 2.3.6. Additional Environmental Data 25 2.4. Discussion 25 2.5. Conclusions 28

PAGE 5

ii 3. Community Assessments 48 3.1. Introduction 48 3.1.1. Common Coral Species 49 3.1.2. Coral Colony Condition and Mortality 50 3.1.3. Fish Assemblage Structure 51 3.1.4. Algal Biomass and Herbivory 51 3.1.5. Comparisons with regional Caribbean values for the AGRRA Biotic Reef Index (Kramer 2003) 52 3.2. Methods 52 3.2.1. Benthic Assessment 52 3.2.2. Fish Assessment 53 3.2.3. Data Analysis 53 3.3. Results 54 3.3.1. Community Structure 54 3.3.2. Coral Mortality and Condition 55 3.3.3. Fish Community Structure 55 3.3.4. Algal Abundance and Herbivory 57 3.3.5. AGRRA biotic index 57 3.4. Discussion 58 3.4.1. Community Structure 58 3.4.2. Mortality and Coral Colony Condition 59 3.4.3. Recruitment 60 3.4.4. Fish Assemblage Structure 60 3.4.5. Algal Biomass and Herbivory 60 3.5. Conclusions 61 4. Symbiont-bearing Foraminifera as Indicators of Reef Health 85 4.1. Abstract 85 4.2. Introduction 85 4.2.1. Larger Benthic Foraminifera as Indicators of Reef Condition 85 4.2.2. Rationale for Assessing Populations of Amphistegina 86 4.2.2.1. Bleaching in Amphistegina : Evidence for Photoinhibitory Stress 87 4.2.3. Other Larger Benthic Foraminifers 89 4.2.4. Study Goals 90 4.3. Methods 90 4.3.1. Study Sites 90 4.3.2. Sampling and Assessment of Symb iont-bearing Foraminifers 91 4.3.3. Data Analysis 92 4.4. Results 93 4.4.1. 6 m Sites 93 4.4.1.1. Responses of Amphistegina 93 4.4.1.2. Other Symbiont-bearing Foraminifera 94 4.4.2. Depth Gradient 96

PAGE 6

iii 4.4.2.1. Responses of Amphistegina 96 4.4.2.2. Other Symbiont-bearing Foraminifera 97 4.5. Discussion 98 4.6. Conclusions 102 5. Lesion Regeneration Rates in Reef Building Corals ( Montastraea spp.) as Indicators of Colony Condition 121 5.1. Abstract 121 5.2. Introduction 122 5.3. Methods 123 5.3.1. Benthic Community Assessment 124 5.3.2. Lesion Regeneration 124 5.3.3. Data Analysis 126 5.3.3.1. Benthic Community Assessment 126 5.3.3.2. Lesion Regeneration 126 5.4. Results 128 5.4.1. Community Data 128 5.4.2. Regeneration Model 128 5.4.3. Short-Term (45 154 d) Regeneration Rates 128 5.4.4. Quasi-Annual Regene ration Rate (319 376 d) 129 5.4.5. Healed and Type II lesions 130 5.5. Discussion 131 5.5.1. 6 m Sites 132 5.5.2. Depth Gradient 133 5.5.3. Comparisons among All Study Sites 134 5.5.4. Effect of Colony and Lesion Characteristics 134 5.6. Conclusions 135 6. Environmental Links to Coral Stress Response 144 6.1. Introduction 144 6.1.1. Linking Potential Stressors to Organism Responses 144 6.1.2. Potential Candidates of Stress 145 6.1.3. Metabolic Costs of Stress on Corals 145 6.1.4. Cellular Diagnostic System 146 6.1.5. Linking Cellular Biomarkers to Higher Order Processes 148 6.2. Objectives 149 6.3. Methods 150 6.3.1. Study Sites 150 6.3.2. Cellular Diagnostic Sampling 150 6.3.3. Data Analysis 151 6.3.3.1. Cellular Diagnostic System 151 6.3.3.2. Relating Coral Cellular Biomarkers to Coral Regeneration Rates 152 6.3.3.3. Linking Environmental Data to Coral Regeneration Rates 153

PAGE 7

iv 6.3.3.4. Analysis Routines 153 6.4. Results and Discussion 153 6.4.1. Evidence for Temperature or Light Stress? 154 6.4.2. Evidence for Local Xenobiotic Stress 156 6.4.3. Enhanced Local Effects with Heavy Rainfall 158 6.4.4. Are Coral Regeneration Ra tes and Protein Production Energylimited? 162 6.4.5. Diagnosing Reef Condition 163 6.4.5.1. 6 m Sites 164 6.4.5.2. Depth Gradient 167 6.5. Conclusions 170 7. Conclusions and Future Research 225 7.1. Multivariate Approach to Assessing Reef Condition 225 7.2 Strengths and Caveats of Individual Indicators 225 7.3 Recommendations for Management and Future research 229 References 233 Appendices 259 About the Author End Page

PAGE 8

v List of Tables Table 1.1. Sampling dates for 2001 2003 at Biscayne National Park (BNP), Algae Reef (AR), White Banks (WB) and Key Largo (KL) depth gradient (3, 6, 9 and 18 m). 10 Table 2.1. Deployment and collection of sediment traps in 2001 and 2002; NA indicates no traps added; NC indicates no traps collected and swapped; A indicates number of traps added; C indicates number of traps collected and swapped. 29 Table 2.2. SERC Water Quality Monito ring Network sampli ng list including summary statistics for all surface water quality variables for all FKNMS outer reef stations between 1995 a nd 2005 (Boyer & Jones 2002, Boyer & Briceo 2005). 30 Table 2.3. Study sites and associated Southeast Re search Center (SERC) water quality stations. SERC sampling dates for KL depth gradient and WB are the same. SERC sampling dates at KL 6 m, KL 9 m, KL 18 m and WB are the same as KL 3 m. 31 Table 3.1. Benthic parameters measured and calculated at each site 63 Table 3.2. Coral species assessed by Atlantic and Gulf Rapid Reef Assessment 64 Table 3.3. Fish species included in AGRRA assessments and established length and mass relationships fo r Caribbean fishes (Marks & Klomp 2003). Fish biomass was calculated using the power function: W = a Lb, where W is the mass (grams), L is the length (cm), and a and b are parameters estimated by linear regression of logarithmically transformed length-mass data. 65 Table 3.4. Identification of key coral sp ecies that discriminated among sites; represents species that did not contribute to 90% of dissimilarity between specific sites. Bold values represent species primarily responsible for differen ces between sites. A bbreviations of coral species are as shown in Table 3.2. 67

PAGE 9

vi Table 3.5. Comparison of benthic parame ters; values represent mean ( SE) at four 6m patch reefs from 10 m tr ansects using Atlantic and Gulf Rapid Reef Assessment protocol Means not connected by the same letter differed significantly (p < 0.05). 68 Table 3.6. Pairwise comparison of 6 m sites based on ANOSIM of fish composition; significant R values are in bold (significance level = 0.1%) 69 Table 3.7. Kruskal-Wallis ( 2) and Wilcoxon pairwise comparison of 6 m sites based on biomass of fish families; n.s., not significant (p > 0.05). No significant differences were found for the groups Acanthuridae, Balistidae, Pomacanthidae, Serranidae, Stromateidae and other. 70 Table 3.8. Kruskal-Wallis ( 2) and Wilcoxon pairwise comparison of 6 m sites based on densities of fish fa milies; n.s., not significant (p > 0.05). No significant differences (p > 0.05) were found for the groups Acanthuridae, Balistidae, Pomacanthidae, Scaridae, Serranidae, Stromateidae and other. 70 Table 3.9. Characterization of functiona l algal groups and density of coral recruits; values represent mean ( SE). *Algal index = % relative abundance of macroalgae x canopy height. Means not connected by the same letter differed significantly (p < 0.05). 71 Table 3.10. Comparisons of study site s with AGRRA re gional baselines (modified from Kramer et al. 2003 ) based on indicators selected for biotic reef health index for corals >25 cm maximum diameter. 72 Table 3.11. SIMPER results of Biotic Reef Index; represents parameters that did not contribute to 90% of dissi milarity among sites. Large and small coral density, diseased corals, Diadema density and carnivorous fish density did not contribute to differences among sites. 73 Table 3.12. SIMPER results of Biotic Reef Index showing dissimilarities among study sites and Caribbean means; represents parameters that did not contribute to 90% of dissimilarity among sites. Small coral density, recent mortality, diseased corals, Diadema density and carnivorous fish density did not contribute to differences among sites. 74

PAGE 10

vii Table 3.13. SIMPER results of Biotic Reef Index showing dissimilarities among study sites and Caribbean best values; represents parameters that did not contribut e to 90% of dissimilarity among sites. Small coral density, recent mortality and diseased corals did not contribute to differences among sites. 75 Table 3.14. SIMPER results of Biotic Reef Index showing dissimilarities among study sites and Caribbean worst values; represents parameters that did not contribut e to 90% of dissimilarity among sites. Large and small coral density, % crustose coralline, Diadema density and carnivorous fish density did not contribute to differences among sites. 76 Table 3.15. SIMPER results of Biotic Reef Index showing dissimilarities amongCaribbean mean, best and worst values; represents parameters that did not contribut e to 90% of dissimilarity among regional values. Large coral densit y and diseased corals did not contribute to differences among regional values. 77 Table 4.1. Population parameters of Amphistegina (late spring early autumn unless otherwise noted) and th eir interpretive value. 104 Table 4.2. Habitats of common larger benthic foraminifers found on Florida reefs (adapted from Levy 1991, Hallock & Peebles 1993, Hallock 1999, and Hallock pers. comm.) 105 Table 4.3. ANOSIM2 results for diffe rences among 6 m sites (averaged across all sampling periods); Gl obal R = 0.28, significance level = 5% 106 Table 4.4. ANOSIM2 results for di fferences among sampling periods (averaged across all 6 m sites); Global R = 0.23, si gnificance level = 5% 106 Table 4.5. Identification of Key Discri minating Larger Benthic Foraminifers among the 6 m sites between A ugust 2001 and February 2003 107 Table 4.6. ANOSIM2 results for differe nces among depths (averaged across all sampling periods); Global R = 0.53, significance level = 5% 108 Table 4.7. ANOSIM2 results for di fferences among sampling periods (averaged across all depths); Global R = 0.18, significance level = 5% 108

PAGE 11

viii Table 4.8. Identification of Key Discri minating Larger Benthic Foraminifers along the depth gradient between October 2001 and February 2003 109 Table 5.1. Comparison of benthic parameters (mean SE) along 10 m transects at four 6 m-patch reef s. Methods followed the Atlantic Gulf and Rapid Reef Assessment pr otocol. Data not connected by the same superscript letter diffe red significantly (p < 0.05). 136 Table 5.2. Mean ( SE) r2 values of the five colonies for the regression decay model, y = yo + ae-b*time. A zero r2 value was assumed for all lesions that did not fit this model. The last column includes overall mean ( SE) slope (b) (cm2 d-1 x 102). Site abbreviations as in Fig. 1. 137 Table 5.3. Percentage of healed and T ype II lesions (no l onger enclosed by living tissue) at each site. Total number of lesions was <35 at KL 9 m, KL 18 m, AR and BNP due to breakage during sampling (as discussed in methods). 138 Table 6.1. List of biomarkers assayed, representing four cellular subsystems including the sampling period and organism (cnidarian host or algal symbiont) tested for each biomarker. See Appendix 6.1 for a description of each biomarker. 172 Table 6.2. Descriptive statistics for each cellular biomarker from 2000 (Downs et al. 2005a). Mean valu es from March 2000 at Biscayne National Park represent a "stre ssed" condition and overall mean values from pooled Key Largo site s represent "basal" condition; indicates reference conditions for this biomarker are unavailable. 173 Table 6.3. ANOSIM2 results for differe nces between sampling periods based on all cellular biomarkers (average d across all 6 m sites); Global R = 0.38, 0.1% significance level; Multivariate Dispersion Indices (MVDISP) shown in shaded area for each sampling period. 174 Table 6.4. ANOSIM2 results for diffe rences among 6 m sites based on all cellular biomarkers (averaged acr oss all sampling periods); Global R = 0.13, 0.1% significance level; Multivariate Dispersion Indices (MVDISP) shown in shaded area for each site. 174

PAGE 12

ix Table 6.5. Global R values of ANOSIM significance tests for differences among 6 m sites based on all cell ular biomarkers during each sampling period. No significant differences among sites were observed in March 2001, June 2001, August 2001 or March 2002. The R statistic for pairwise comparison of 6 m sites based on ANOSIM of all biomarkers duri ng each sampling period also is shown; n.s. represents not significant (>5%). 175 Table 6.6. ANOSIM2 results for differe nces between sampling periods based on all cellular biomarkers (average d across all depths); Global R = 0.32, 0.1% significance level; Mu ltivariate Dispersion Indices (MVDISP) shown in shaded area for each sampling period. 176 Table 6.7. ANOSIM2 results for differen ces along depth gradient based on all cellular biomarkers (averaged acr oss all sampling periods); Global R = 0.27, 0.1% significance level; Multivariate Dispersion Indices (MVDISP) shown in shaded area for each depth. 176 Table 6.8. Global R values of ANOSIM significance tests for differences among depths based on all cellular biomarkers during each sampling period. No significant differences among depths were observed in March 2002. The R statistic for pairwise comparison of depths based on ANOSIM of all biomarkers during each sampling period also is shown; n.s represents not significant (> 5%). 177 Table 6.9. Repeated measures MANO VA results for individual cellular biomarker levels among 6 m sites; bold values significant p<0.05 (Note: degrees of freedom fo r site effect: numerator = 3, denominator = 16; degrees of fr eedom for time and site x time effect change with sampling fr equency of each biomarker; see Table 6.2) 178 Table 6.10. Repeated measures MANOVA for individual cellular biomarker levels along depth gradient. Bold values significan t p<0.05; (Note: degrees of freedom for site eff ect: numerator = 3, denominator = 16; degrees of freedom for time and site x time effect change with sampling frequency of each biomarker; see Table 6.2) 179

PAGE 13

x Table 6.11. Overall mean ( SE) biomarke r concentrations (averaged across all depths) for each sampling period. Shaded values represent biomarkers that did not vary significantly with time or time x site interactions; represent biomarkers that varied si gnificantly with time but not time x site interactio ns; means for each biomarker that are not connected by the same letter differed significantly based on Tukey HSD test (p < 0.05); represents biomarkers that were not sampled during that time period. 180 Table 6.12. Overall mean ( SE) cellula r biomarker concentrations (averaged across all 6 m sites) for each sampling period. Shaded values represent biomarkers that did not vary significantly with time or time x site interactions. represent biomarkers that varied significantly with time (ANOVA sta tistics shown) but not time x site interactions; means for each biomarker that are not connected by the same letter differed significantly based on Tukey HSD (p < 0.05). represents biomarkers that were not sampled during that time period; units as shown in Table 6.2. 182 Table 6.13. BEST results for coral rege neration rates and SERC environmental parameters. Abbreviations include alkaline phosphatase activity (APA), chlorophyll-a (CHLA), dissolved inorganic nitrogen (DIN), dissolved oxygen (DO) nitrate (NO3), nitrite (NO2), soluble reactive phosphate (SRP), total nitrogen (TN), total organic carbon (TOC), total organic nitrogen (TON) and total phosphorus (TP) 184 Table 6.14. BIO-ENV and RELATE result s for coral regeneration rates and biomarker concentrations during a given time period at the 6 m Sites; n.s. represents not significantly (> 5%) 185 Table 6.15. Descriptive statistics for each cellular biomarker at the 6 m sites over the entire study period betw een March 2001 and February 2003. represents biomarkers th at differed signi ficantly among sites but not with time x site inte ractions. Means of biomarkers not connected by the same letter differed significantly based on Tukey HSD test (p < 0.05) 186 Table 6.16. Coefficient of variance for each cellular biomarker at the 6 m sites over the entire study period betw een March 2001 and February 2003. 188

PAGE 14

xi Table 6.17. Descriptive statistics for each cellular biomarker along the depth gradient over the entire study period between March 2001 and February 2003. represents biomar kers that differed significantly among sites but not with time x site interactions. Means of biomarkers not connected by the same letter differed significantly based on Tukey HSD test (p < 0.05) 189 Table 6.18. Coefficient of variance for each cellular biomarker along the depth gradient over the entire study period between March 2001 and February 2003. 191 Table 6.19. BIO-ENV and RELATE result s for coral regeneration rates and biomarker concentrations during a given time period along the depth gradient; n.s. represents not significantly (> 5%) 192 Table 7.1 Indicators of reef condition 232

PAGE 15

xii List of Figures Figure 1.1. Multi-scale approach to st udy the effects of stress on the marine environment. 11 Figure 1.2. Chart of seven sampling sites in the Florida Keys National Marine Sanctuary and Biscayne National Park. squares designate the 6 m sites included four patch reef s Key Largo (KL) 6 m (25 01.092' N, 80 23.844' W), White Banks (W B) (25 02.232' N, 80 22.496' W), Algae Reef (AR) (25 08.799' N, 80 17.579' W), and Alinas Reef (BNP) (25 23.185 N, 80 09.775' W). circles designate sites along the depth gradient, including two patch reefs, Key Largo 3 m (25 02.447' N, 80 25. 442' W), Key Largo 6 m, and two fore reefs, Key Largo 9 m (25 00.146' N, 80 23.626' W), Key Largo 18 m (25 00.206' N, 80 23.023' W). 12 Figure 1.3. Representative colonies of (A) Montastraea faveolata (B) M. annularis and (C) M. franksi. Pictures taken by Roy Price. 13 Figure 2.1. Map showing location of study sites (l arge yellow circles), Southeast Research Center Water Quality monitoring stations (small blue circles and blue stati on numbers) and area coverage of mangrove and wetlands in the study region. 32 Figure 2.2. Mean sedimentation rate (mg cm d-1 SE) between March 2001 and February 2003 at (A) the 6 m sites and (B) along the depth gradient. The KL 6 m site is incl uded in both panels. Note smaller scale along depth gradient. 33 Figure 2.3. Mean ( SE) turbidity (n ephelometric turb idity units: NTU) between October 2001 and February 2003 at (A) the 6 m sites and (B) along the depth gradient. 34 Figure 2.4. Mean temperature (C) between January 2001 and February 2003 at (A) the 6 m sites and (B) al ong the depth gradient. Molasses Reef (MR) data from the SEAKEY S C-MAN buoy also is plotted. The dotted line represents the temperature where corals typically begin to bleach (31C; Andrews et al. 2005). 35

PAGE 16

xiii Figure 2.5. (A) Nitrate (NO3), (B) nitrite (NO2), and (C) ammonium (NH4) for SERC stations between 2001 and 2 003; Station 224 is closest to KL 6 m and WB, Station 218 is clos est to AR, and 206 is closest to BNP; dashed line represents median values for FKNMS reef stations between 1995 and 2005. 36 Figure 2.6. (A) Total nitrogen (TN), (B) dissolved inorganic nitrogen (DIN), and (C) total organic nitrogen (T N) for SERC stations between 2001 and 2003; Station 224 is clos est to KL 6 m and WB, Station 218 is closest to AR, and 206 is closest to BNP; dashed line represents median values for F KNMS reef stations between 1995 and 2005. 37 Figure 2.7. (A) Total phosphorus (TP) (B) alkaline phos phatase activity (APA), (C) soluble reactiv e phosphorus (SRP) and (D) chlorophyll-a (CHLA) for SERC stations between 2001 and 2003; Station 224 is closest to KL 6 m and WB, Station 218 is closest to AR, and 206 is closest to BNP; dashed line represents median values for FKNMS reef stations between 1995 and 2005. 38 Figure 2.8. Total precipitation (cm) in Miami between January 2001 and February 2003. 39 Figure 2.9. (A) Salinity, (B) turbidity, (C) temperature, and (D) dissolved oxygen for SERC stations between 2001 and 2003; Station 224 is closest to KL 6 m and WB, Statio n 218 is closest to AR, and 206 is closest to BNP; dashed line represents median values for FKNMS reef stations between 1995 and 2005. 40 Figure 2.10. (A) Salinity, (B) turbidity, (C) temperature, and (D) dissolved oxygen for SERC stations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, resp ectively; dashed line represents median values for FKNMS reef stations between 1995 and 2005. 41 Figure 2.11. (A) Total nitrogen (TN), (B) dissolved inorganic nitrogen (DIN), and (C) total organic nitrogen (T ON) for SERC stations along a Key Largo depth gradient be tween 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed line represents median values for FKNMS reef stations between 1995 and 2005. 42 Figure 2.12. Mean (bars) and maximum (line) monthly wind speeds (knots) recorded by SEAKEYS C-MAN buoy at Molasses Reef between January 2001 and February 2003. 43

PAGE 17

xiv Figure 2.13. (A) Total nitrog en to total phosphorus (TN:TP), (B) dissolved inorganic nitrogen to total phosph orus (DIN:TP), and (C) total organic carbon (TOC) for SERC st ations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed line represents median values fo r FKNMS reef stations between 1995 and 2005. 44 Figure 2.14. (A) Total phosphorus (TP) (B ) soluble reactive phosphorus (SRP), (C) alkaline phosphatase activit y (APA) and (D) chlorophyll-a (CHLA) for SERC stations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, resp ectively; dashed line represents median values for FKNMS reef stations between 1995 and 2005. 45 Figure 2.15. Figure 2.15 (A) Nitrate (NO3), (B) nitrite (NO2), and (C) ammonium (NH4) for SERC stations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed line represents median values fo r FKNMS reef stations between 1995 and 2005. 46 Figure 2.16. (A) Ozone (dobsons) and daily dose UV (J m-2 d-1) between 1998 and 2003. Figure taken from Ivey, J. (B) Mean irradiance (W m-2) of UVA between 1997 and 2003. (C) Mean irradiance of UVB (W m-2) between 1997 and 2003. 47 Figure 3.1. Relative abundance of coral species >10 cm maximum diameter. 78 Figure 3.2. Mean ( SE) recent and old mortality for dominant coral species at the four study sites. Means not connected by the same letter (capital and lowercase for old a nd recent mortality, respectively) were not significantly different by Wilcoxons test (p < 0.05). 79 Figure 3.3. Frequency of coral colonies by size class (maximum diameter) and the relationship of colony size with recent and old mortality. Solid Line represents old mortality ( ) and the dashed line represents recent mortality (x). 80 Figure 3.4. Fish biomass (g x 103/100 m2) by family. 81 Figure 3.5. Density (individuals/100m2) of herbivorous and carnivorous fish by size class (cm). 82 Figure 3.6. Relative abundance of functional algal groups 83

PAGE 18

xv Figure 3.7. Cluster analysis based on Atlantic Gulf and Rapid Reef Assessment (AGRRA) biotic health indicators comparing study sites to regional AGRRA values for Caribbean reefs >5 m. 84 Figure 4.1. Pictures of bleaching and damage in A. gibbosa Top left: normal color with no damage; Bottom left: partly bleached with no damage; Top right: pale and broke n; Bottom right: partly bleached and chipped. 110 Figure 4.2. Mean ( SE) densities of Amphistegina gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. Densities are pl otted on a log scale. 111 Figure 4.3. Mean ( SE) percentage of juvenile A. gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. 112 Figure 4.4. Mean ( SE) diameters of A. gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. 113 Figure 4.5. Mean ( SE) percentages of adult A. gibbosa exhibiting any degree of bleaching from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. 114 Figure 4.6. Mean ( SE) percentages of damaged tests in populations of A. gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. 115 Figure 4.7. Mean ( SE) densities of all symbiont-bearing (larger) foraminifera from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth grad ient. Densities are plotted on a log scale. 116 Figure 4.8. Multi-dimensional scaling plots (MDS) illustrate the ordination of samples collected between August 2001 and February 2003 based on (A) the entire assemblage of LBF, (B) the assemblage with A. gibbosa removed, and (C) A. gibbosa alone. 117 Figure 4.9. Mean ( SE) densities of other dominant symbiont-bearing foraminifera at the 6 m sites (A) Archaias angulatus (B) Laevipeneroplis proteus (C) Heterostegina depressa and (D) Broekina orbitolitoides 118

PAGE 19

xvi Figure 4.10. Multi-dimensional scaling plots (MDS) illustrate the ordination of samples collected along the dept h gradient between October 2001 and February 2003 based on the entire assemblage of LBF. 119 Figure 4.11. Mean densities (individuals m-2 SE) of other dominant symbiontbearing foraminifera along the depth gradient (A) Cyclobiculina compressus (B) Asterigerina carinata (C) Heterostegina depressa and (D) Broekina orbitolitoides Densities are plotted on a log scale. 120 Figure 5.1 Examples of lesions at 6 m sites between October 2001 and November 2002 showing two extremes. (A) Algae Reef (AR) lesion completely healed by June 2002. (B) Alinas Reef (BNP) lesion joined with other sampli ng lesions in June 2002 and became covered with turf algae. Black arrow points to the lesion of interest. 139 Figure 5.2 Mean lesion size ( SE) through time for each season between June 2001 and March 2002 at (A) the 6 m sites and (B) along the depth gradient. Axes staggered to align sampling dates. Note expanded y-axis in panels show ing lesion regeneration along depth gradient in August 2001. Merging of two sampling-induced lesions occurred at KL 9 m (in March 2002), at KL 18 m (in February 2003) and at BNP (in June 2002 and August 2002). An additional lesion joined with the previously merged lesions at BNP in October 2002. Lesions that progressed into Type II lesions or data removed for other reasons (as di scussed in methods: breakage or initial size >3.4 cm2) were not included in means. 140 Figure 5.3 Regeneration rates standardized to initial lesion perimeter (mean T/P SE) for each season from one sampling event until the next. Comparisons (A) among 6 m sites and (B) along depth gradient. Regeneration rates were calculated between June and August 2001 (54 13 d), August and October 2001 (56 d), October 2001 and March 2002 (153 2 d), March and June 2002 (91 1 d), June and August 2002 (48 13 d), August and November 2002 (74 d) and November 2002 and February 2003 (99 1 d). 141

PAGE 20

xvii Figure 5.4 Regeneration rates standardized to initial lesion perimeter (mean T/P SE) for each season between June 2001 and March 2002 from the time of sampling until the following year. Compared (A) among the 6 m sites and (B) along the 3 18 m depth gradient. Regeneration rates were calcu lated between June 2001 and 2002 (357 10 d), August 2001 and 2002 (355 1 d), October 2001 and November 2002 (374 2 d) and March 2002 and February 2003 (321 2 d). 142 Figure 5.5 Cumulative percentage (%) of lesions completely healed with time (days) at (A) the 6 m sites and (B) along the depth gradient. Numbers adjacent to lines in the shaded area are total numbers of completely healed lesions at each site. 143 Figure 6.1. Protein Metabolic Conditi on at the 6 m sites including (A) cnidarian heat shock protein (H sp) 60, (B) dinoflagellate heat shock protein 60, (C) cnidarian heat shock protein 70, (D) dinoflagellate heat shock prot ein 70 and (E) ubiquitin. Data presented as means ( SE) in pmol/ng TSP for cnidarian Hsp 60 and 70, dinoflagellate Hsp 60 an d 70 and in fmol/ng TSP for ubiquitin. The red and blue dashed line represents stressed and basal levels, respectively as defined by Downs et al. 2005a. Means for dinoflagellate Hsp 70 were all below stressed levels (1.68 pmol/ng TSP). 193 Figure 6.2. Oxidative damage and res ponse at the 6 m si tes including (A) cnidarian copper/zinc superoxide dismutase (Cu/Zn SOD), (B) dinoflagellate Cu/Zn SOD, (C) cn idarian manganese superoxide dismutase (MnSOD), (D) dinoflage llate Mn SOD, (E) cnidarian glutathione peroxidase (GPx), (F) dinoflagellate GPx and (G) catalase. Data presented as means ( SE) in pmol/ng TSP. The red and blue dashed line represents stressed and basal levels, respectively as defined by Downs et al. 2005. Means for cnidarian GPx were below both basal and str essed levels at all sites (70 and 171 pmol/ng TSP, respectively). Stressed or basal levels are not available for cnidarian and dinoflagellate Cu/Zn SOD or catalase. 195 Figure 6.3. Metabolic Condition at the 6 m sites including (A) heme oxygenase, (B) ferrochelatase, (C) metallothionein, (D) cnidarian small heat shock protein (sHsp) and (E) chloroplast small heat shock protein (ChlpsHsp). Data presented as means ( SE) in Eunits/ng TSP. Baseline condition is not available for these biomarkers. 197

PAGE 21

xviii Figure 6.4. Xenobiotic Detoxification a nd Response at the 6 m sites including (A) cytochrome P450 2-class (C YP-2), (B) cytochrome P450 3class (CYP-3), (C) cytochrome P450 6-class (CYP-6), (D) cnidarian glutathione-S-transfera se (Cn GST), (E ) dinoflagellate GST and (F) multixenobiotic resistance protein (MXR). Data presented as means ( SE) in Eu nits/ng TSP for CYP-2 and MXR; relative units/ng TSP for CYP-3 a nd CYP-6; and pmol/ng TSP for Cn and Dn GST. Baseline informa tion is not available for any of the cytochrome P450 classes or fo r MXR. The red and blue dashed line represents stressed and basal levels, respectively as defined by Downs et al. 2005. Baseline information is not available for CYP-2, CYP-3, CYP-6 or MXR. 199 Figure 6.5. Protein Metabolic Condition along the depth gradient including (A) cnidarian heat shock protein (Hsp) 60, (B) dinoflagellate heat shock protein 60, (C) cnidarian heat shock protein 70, (D) dinoflagellate heat shock prot ein 70 and (E) ubiquitin. Data presented as means ( SE) in pmol/ng TSP for cnidarian Hsp 60 and 70, dinoflagellate Hsp 60 an d 70 and in fmol/ng TSP for ubiquitin. The red and blue dashed line represents stressed and basal levels, respectively as defined by Downs et al. 2005a. Means for dinoflagellate Hsp 70 were all below stressed levels (1.68 pmol/ng TSP). 201 Figure 6.6. Oxidative damage and respons e along the depth gradient including (A) cnidarian copper/zinc superoxi de dismutase (Cu/Zn SOD), (B) dinoflagellate Cu/Zn SOD, (C) cn idarian manganese superoxide dismutase (MnSOD), (D) dinoflage llate Mn SOD, (E) cnidarian glutathione peroxidase (GPx), (F) dinoflagellate GPx and (G) catalase. Data presented as means ( SE) in pmol/ng TSP. The red and blue dashed line represents stressed and basal levels, respectively as defined by Downs et al. 2005. Means for cnidarian GPx were below both basal and str essed levels at all sites (70 and 171 pmol/ng TSP, respectively). Stressed or basal levels are not available for cnidarian and dinoflagellate Cu/Zn SOD or catalase. 203 Figure 6.7. Metabolic Condition along th e depth gradient including (A) heme oxygenase, (B) ferrochelatase, (C) metallothionein, (D) cnidarian small heat shock protein (sHsp) and (E) chloroplast small heat shock protein (ChlpsHsp). Data presented as means ( SE) in Eunits/ng TSP. Baseline condition is not available for these biomarkers. 205

PAGE 22

xix Figure 6.8. Xenobiotic Detoxification a nd Response along the depth gradient including (A) cytochrome P450 2class (CYP-2), (B) cytochrome P450 3-class (CYP-3), (C) cytoch rome P450 6-class (CYP-6), (D) cnidarian glutathione-S-transfera se (Cn GST), (E ) dinoflagellate GST and (F) multixenobiotic resistance protein (MXR). Data presented as means ( SE) in Eu nits/ng TSP for CYP-2 and MXR; relative units/ng TSP for CYP-3 a nd CYP-6; and pmol/ng TSP for Cn and Dn GST. Baseline informa tion is not available for any of the cytochrome P450 classes or fo r MXR. The red and blue dashed line represents stressed and basal levels, respectively as defined by Downs et al. 2005. Baseline information is not available for CYP-2, CYP-3, CYP-6 or MXR. 207 Figure 6.9. Plots of Principle Compone nt (PC1) scores for each sampling period at (A) the 6 m sites and (B) along the depth gradient. Vertical bars show the range of values for each sampling period, squares indicate the sample mean for each period. Shaded areas represent sampling periods when there were significant differences among sites based on ANOSIM (6 m sites: Table 6.5; depth gradient: Table 6.8). 209 Figure 6.10. Plots of Principle Component (PC1) scores at each site for (A) March 2001, (B) June 2001, (C) August 2001 and (D) October 2001. Vertical bars show the range of values for each site, squares indicate the sample mean fo r each site. Eigenvalues and eigenvectors as shown in Appendix B. 210 Figure 6.11. Plots of Principle Component (P C1) scores at each site in February 2003. Vertical bars show the range of values for each site; squares indicate the sample mean fo r each site. Eigenvalues and eigenvectors as shown in Appendix C. 211 Figure 6.12. Plots of Principle Component (PC1) scores at each site for (A) October 2001 and (B) June 2002. Vert ical bars show the range of values for each site, squares indica te the sample mean for each site. Eigenvectors and eigenvalues as shown in Appendix C. 212

PAGE 23

xx Figure 6.13. (A) Multi-dimensional scali ng (MDS) plot for regeneration rates (T/P) between October 2001 and March 2002 at the 6 m sites; larger circles represent higher rege neration rates. (B) MDS plot of October 2001 cellular biomarkers (CDS) selected by BEST routine (Table 6.14), which included me tallothionein and cnidarian copper/zinc superoxide disputas e (Cn Cu/Zn SOD). Regeneration rate MDS superimposed by individual CDS biomarkers (C) metallothionein and (D) Cn Cu/Zn SOD; circle size increases with increasing concentration. 213 Figure 6.14. (A) Multi-dimensional scali ng (MDS) plot for regeneration rates (T/P) between August and October 2001 at the 6 m sites; larger circles represent higher regenerati on rates. (B) MDS plot of August 2001 cellular biomarkers (CDS) sel ected by BEST routine (Table 6.14), which included dinoflagellate heat shock protein 60 (Dn Hsp 60) and dinoflagellate copper/zi nc superoxide dismutase (Dn Cu/Zn SOD). Regeneration rate MDS superimposed by individual CDS biomarkers including (C) Dn Hsp 60 and (D) Dn Cu/Zn SOD; circle size increases with increasing concentration. 214 Figure 6.15. (A) Multi-dimensional scali ng plots for regeneration rates (T/P) between June and August 2002 at th e 6 m sites; larger circles represent higher regeneration rate s. (B) MDS plot of June 2002 cellular biomarkers (CDS) select ed by BEST routine (Table 6.14), which included cnidarian heat shock protein (Cn Hsp 70), cnidarian small heat shock prot ein (Cn sHsp), metallothionein, cnidarian glutathione-S-transferase and cytochrome P450 6-class. Regeneration rate MDS supe rimposed by individual CDS biomarkers (C) Cn Hsp 70, (D) Cn sHsp, (E) metallothionein, (F) Cn GST and (G) CYP-6; circle size increases with increasing concentration. 215 Figure 6.16. (A) Multi-dimensional s caling (MDS) and bubble plot of Key Largo 6 m regeneration rates (T/P); larger circles represent higher regeneration rates. (B) Regenera tion rate MDS superimposed by total nitrogen, which was selected by the BEST routine (Table 6.13); circles increase in size wi th increasing co ncentration. 217 Figure 6.17. (A) Multi-dimensional s caling (MDS) and bubble plot of Key Largo 9 m regeneration rates (T/P); larger circles represent higher regeneration rates. Regenera tion rate MDS superimposed by environmental variables selected by BEST routine (Table 6.13) including (B) ratio of total nitr ogen to total phosphorus (TN:TP), (C) nitrite, (D) chlorophyll-a and (E ) turbidity; circles increase in size with increasing concentration. 218

PAGE 24

xxi Figure 6.18. (A) Multi-dimensional s caling (MDS) and bubble plot of Key Largo 18 m regeneration rates (T/P); larger circles represent higher regeneration rates. Regenera tion rate MDS superimposed by environmental variables selected by BEST routine (Table 6.13) including (B) soluble reactive phosphate, (C) ration of total nitrogen to total phosphorus (TN: TP), (D) chlorophyll-a (E) nitrite and (F) total organic carbon; circ les increase in size with increasing concentration. 219 Figure 6.19 (A) Multi-dimensional scal ing (MDS) and bubble plot of Algae Reef regeneration rates (T/P); larger circles represent higher regeneration rates. Regenera tion rate MDS superimposed by environmental variables selected by BEST routine (Table 6.13) including (B) total organic nitrogen, (C) ratio of total nitrogen to total phosphorus (TN:TP), (D) ratio of dissolved inorganic nitrogen to total phosphorus (DIN:TP) a nd (E) total organic carbon; circles increase in size wi th increasing concentration. 221 Figure 6.20. Plots of Principle Component (PC1) scores at each site for (A) June 2002, (B) August 2002, (C) November 2002 and (D) February 2003. Vertical bars show the range of values for each site; squares indicate the sample mean for each site. Eigenvalues and vectors as shown in Appendix B. 222 Figure 6.21. (A) Multi-dimensional scal ing (MDS) and bubble plot of White Banks regeneration rates (T/P); larger circles represent higher regeneration rates. (B) Regenera tion rate MDS superimposed by total nitrogen, which was selected by BEST routine (Table 6.13); circles increase in size with increasing concentration. 223 Figure 6.22. Physiological status of co rals at each study site based on the relationship between regeneration ra tes, a surrogate indicator and cellular diagnostic markers (modi fied from Allen & Moore 2004). Regeneration rates and densities of symbiontbearing foraminifera (LBF) are represented by the dashed blue line and cellular biomarker levels are represente d by the dashed red line. The position of each site is represente d by where the circle intersects these two lines. Note: regenerati on rates and densities of LBF follow similar trends with the exception of KL 3 m, where densities are low but regeneration rates are high, and with KL 9 m and KL 18 m where densities are high but regeneration rates are low. In these cases, density of LBF were not considered due to the caveats of this indicator with depth. Site abbreviations are the same as those used in Fig. 1.1. 224

PAGE 25

xxii Assessing the Health of Cora l Reef Ecosystems in the Florida Keys at Community, Individual, and Cellular Scales Elizabeth M. Fisher ABSTRACT Coral reefs are threatened in Florid a and worldwide. Successful resource management requires rapid identification of anthropogenic sources of stress before they affect the reef community. I tested a multi-s cale approach for assessing reef condition at seven reefs within the Florida Keys Nationa l Marine Sanctuary a nd Biscayne National Park between 2001 and 2003. I examined multipl e environmental parameters to identify potential sources of stress. I utilized the A tlantic and Gulf Rapid Reef Assessment Biotic Reef Index to assess be nthic community structure and an indicator species of Foraminifera ( Amphistegina gibbosa ) to determine if environmental conditions were suitable for calcareous organisms that host algal endosymbionts. Small tissue samples were extracted from colonies of Montastraea annularis species complex to assay a suite of cellular biomarkers to elucidate possible mechanisms of the coral stress response. I monitored regeneration rates of the resultant lesions to determine if the coral colonies were capable of recovering from damage. Multivariate data analyses indicated that corals at all study sites were experien cing stress with different degr ees of response and decline. On reefs with coarse grain sediments that ar e adjacent to an intact mangrove shoreline, the Cellular Diagnostic System indicated that corals were respondi ng to a xenobiotic stress but appeared to be compensating as evidenced by consistently high lesion regeneration rates, a high percentage of healed lesions, low coral mortality and high abundances of A. gibbosa On reefs with silt-sized sediments adjacent to developed coastlines, corals also we re responding to xenobiotic st resses, but were negatively affected as evidenced by low regeneration rates, a low percentage of healed lesions, high coral mortality, and low abundances of A. gibbosa Corals at an 18 m offshore site

PAGE 26

xxiii exhibited abnormally low biomarker levels and some died duri ng the study, indicating that sampled colonies were incapable of upregulating necessary protective proteins. Further research will be required to determin e stressor sources. This study demonstrates that a multiple-indicator approach, spanni ng scales from cellular to community, can provide marine resource managers with da ta linking decline of coral populations to specific environmental conditions and even ts, thereby providing potential for early detection of stressors allowing for preventive management.

PAGE 27

1. Introduction 1.1. Reef Degradation Coral reefs are threatened resources in Florida and many coastal regions worldwide (Bryant et al. 1998, Hoegh-Gul dberg 1999, Hughes et al. 2003, Bellwood et al. 2004, Waddell 2005, many others). Yet these resources are rapidly being degraded or lost to a combination of global and local st ressors. The economic value of reefs is $7.7 billion per year in goods and se rvices for South Florida (Johns et al. 2001, Andrews et al. 2005) and $375 billion worldwid e (Costanza et al. 1997). On global scales, rising sea-surface temper atures (especially during El Nio) and increasing ultraviolet radiation (due to ozone thinning) (HoeghGuldberg 1999, Dustan 2000, Hallock 2001, Buddemeier et al. 2004, ma ny others) were implicated as in bleaching and disease in corals (Porter et al. 2001, Sutherland et al. 2004, Marshall & Schuttenberg 2006, many others), as well as in other reef organisms. Coral bleaching has become common since 1983, affecting every region worldwide and in many cases resulting in significant coral mortality (Marshall & Schuttenberg 2006). The 1997-1998 bleaching event resulted in 90% mortality to 16% of reefs worldwide. Moreover, 97% of Caribbean reefs were impacted by disease (G reen & Bruckner 2000). In the Florida Keys National Marine Sanctuary (FKNMS), the number of locations exhibiting disease increased by a factor of four between 1996 and 1998 (Porter et al. 2001), and disease prevalence only recently leveled off or bega n to decrease (Beaver et al. 2005). Changes in water chemistry and increasingly rapid sea-level rise also threaten the reef-building capacity of corals (Kleypas et al. 2001, Gu inotte et al. 2003, Buddemeier et al. 2004, Hallock 2005, Pelejero et al. 2005). Intensif ied African dust storms and the microbes they transport also may contribute to the de cline of Caribbean re efs (Shinn et al. 2000, Hayes et al. 2001). Local impacts on reefs increase with increasing urbanization and growing human populations (Bryant et al 1998, Causey et al. 2000, Dustan 2000).

PAGE 28

Although substantial effort is being expended to monitor coral communities and water quality associated with reefs (e.g., SEAKEYS/C-MAN Project; Jaap et al. 2000, Boyer and Jones 2002), why corals are dying at unprecedented rates (Wilkinson 2000, Porter et al. 2001) stil l is not well known. To better understand how to conserve coral reef ecosystems, scientists must determine their status on community scales and understand the underlying mechanisms on popul ation, individual, and cellular scales (Jameson et al. 2001, 2002, Downs 2005, Downs et al. 2005b). 1.2. Florida Reef Tract Reef degradation has been attributed to coastal development and associated stressors, with > 70% of coral reefs worldw ide directly threatened by human-associated activities (Bryant et al. 1998, Waddell 2005). Florida ha s 12 of the top 100 fastest growing counties in the United States (US Census 2000). Miami-Dade Countys population grew from 298 in 1889 to 495,047 in 1950 to over 2 m illion in 2000 (US Census). The upper and lower Florida Keys, particularly Key Lar go and Key West, have experienced substantial increases in human population growth a nd urban development over the past four decades without adequate increases in supporti ng infrastructure to control runoff, groundwater po llution and sewage (Causey et al. 2000, Dustan 2000). Within the last 40 years, Monroe County s human population has increased by 40% to approximately 79,000 people (US Census 2000). This does not include the substantial tourist population (> 25,000 people), which re sults in more than 100,000 people in the Keys any given time during winter mont hs (Kruczynski & McManus 2002). Patterns of coral decline generally correspond with human population centers, with the largest declines doc umented in the upper and lowe r Keys (Jaap et al. 2000). Paradoxically, despite being closer to potentia l coastal impacts, in shore patch reefs on average appear to be in better condition a nd have higher coral cove r relative to offshore reefs (Beaver et al. 2005). Reasons for thes e differences are not well understood but may be related to inshore corals being adapted to more variable environments, which allows them to tolerate anthropogenic stressors better than colonies in historically more stable offshore environments (e.g., Soto 2006).

PAGE 29

Water quality in the Florida Keys declined over the last thirty years due to changes in water flow patterns from Florid a Bay, sedimentation (from boat traffic and development) and increased near-shore nutrien t concentrations (fro m local wastewaters, freshwater upwelling, fertilizers and industria l pollutants; Szmant & Forrester 1996, Lang et al. 1998, Causey et al. 2000, Porter et al. 2001, Andrews et al. 2005 ). Over the same time period, live coral cover decreased througho ut the Florida Keys with 50 70 % loss in live coral cover since the 1970s (Dustan & Halas 1987, Lang et al. 1998, Causey et al. 2000, Porter et al. 2002, Palandro et al 2003, Beaver et al. 2005, Palandro 2006). Significant loss occurred in major reef-building corals such as the Montastraea annularis complex, Acropora palmata and A. cervicornis (Beaver et al. 2005). Mats of sedimenttrapping turf algae are graduall y overgrowing corals and re stricting settlement of new recruits (Lang et al 1998, Petersen et al 2005, Nugues & Szmant 2006). Large storms produce runoff laden with heavy metals (G lynn et al. 1989, Canti llo et al. 1997), pesticides and herbicides (Gardinali et al. 2002, Owen et al. 2002, 2003, Downs et al. 2006), and microbial pathogens as sociated with local sewage (Paul et al. 1995a, b, Paul et al. 1997, Griffin et al. 1999, Lipp et al. 2002). Pest icides and herbicides such as dibrom, which is heavily used in South Florida to c ontrol mosquitoes, induced stress responses in corals in the Upper Keys (Morgan & Snell 2002, Owen et al. 2003). Such toxins can interfere with chemical signals or larval behavior, thereby inhi biting coral reproduction and recruitment (e.g., Richmond 1993, 1997, Peters 1997, McKenna et al. 1999, ReicheltBrushett & Harrison 2005). Prevalence of coral diseases and bleachi ng also dramatically increased over the past three decades (Lang et al. 1998, Caus ey et al. 2000, Jaap et al. 2000, Green & Bruckner 2000, Harvell et al. 2004, Sutherland et al. 2004, Santavy et al. 2005). Some diseases appear most prevalent in popula tions stressed by anthropogenic pollution (Goreau et al. 1998, Richardson 1998, Kaczmarsky et al. 2005). For example, the causal agent of white pox disease was identif ied as the human fecal bacterium Serratia marcescens (Patterson et al. 2002). Nutrient enrichment increased severity of aspergillosis in a sea fan ( Gorgonia ventalina ) and yellow band disease in the M.

PAGE 30

annularis complex (Bruno et al. 2003). White pl ague and black band disease were more prevalent in proximity to sewage effluent (Kaczmarsky et al. 2005). 1.3. Need for New Methodologies Monitoring changes in species composition, abundance and coverage are important for 1) determining change in co mmunity dynamics and de tecting patterns and trends over long time periods, 2) providing base line data to compare present conditions, and 3) predicting how human activities might affect ecosystems (Rogers et al. 1994, Hughes & Connell 1999, Jaap et al. 2000). Howe ver, these traditional methods can only detect a disturbance after the community is altered, and often do not provide sufficient evidence of cause for managers to take spec ific actions. Theref ore, more sensitive techniques are needed to detect stress re sponses before impacts begin to degrade a community (Brown 1988, Depledge et al 1993, Risk 1999, Jameson et al. 2001, Downs 2005, Downs et al. 2005). Early detection of stressors enables implementation of preventive management rather than dependi ng on post-damage restoration. Communitybased bioindicators are needed for effective reef assessm ent, which underlies policy, legislation and management (Risk 1999, McCa rty et al. 2002). There is a need for integrating monitoring with research de signed to identify stre ssors and determine causality (Brown 1988, Risk 1999). Existing reef-monitoring programs commonly do not fulfill this need. 1.4. Overview of Dissertation Research 1.4.1. Approach This dissertation directly addresses the deficiency de scribed by Brown (1988) and Risk (1999) by integrating monitoring of envi ronmental conditions and responses of reef populations with cellular data, which has the poten tial to identify specific stressors before they cause community degradation. The goal of my project is to determine if a suite of cellular, physiological, and comm unity parameters can (1) dis tinguish between levels of physiological condition (e.g., nominal vs. diseased state), (2) identify types of stressors, and (3) elucidate mechanisms of stress response My project also evaluates the strengths

PAGE 31

and caveats of using individual bioindicators to detect differences among sites, times and types of stressors. To accomplish these goals, this project took a multi-scale approach (Fig. 1.1), which was carried out over two years. Dr awing on multiple assessment endpoints across hierarchical levels provides a mechanistic understanding of th e causes of reef degradation (Downs 2005), which allows researchers to determine whether (1) an organism is responding to a stress and (2) that stress ha s resulted in reduced physiological function (Downs 2005). I assessed the condition of a reef ecosystem, including traditional community assessment (i.e., the Atlantic and Gulf Rapid Reef Assessment protocol of Lang 2003) and monitoring of selected envi ronmental parameters, populations of a surrogate indicator (i.e., Hallock et al 2004), coral-colony condition (Williams 1994, Meesters et al. 1997a), and cellu lar physiological responses of coral colonies as indicated by a diagnostic prof ile of cellular parameters (D owns 2005, Downs et al. 2000, 2002, 2005, 2006). With this approach, I characterized both reef and envir onmental conditions, while quantifying temporal changes in popula tions and individual coral colonies. I compared coral responses (e.g., regeneration rate s) with results from Cellular Diagnostic System (CDS) assessments of the same col onies. The CDS has the potential of detecting deviations in cellular function before th ey alter physiological functions (e.g., regeneration, growth or reproduction) and degrade the community (Downs 2005). I applied a diagnostic approach (Jameson et al. 2001, Downs et al. 2005) to assess reef condition at my study site s including (1) examination a nd preliminary classification of each reef (e.g., community and environmen tal assessments), (2) characterization of reef condition based on compar ison with reference values of key organisms, (3) if an altered state is apparent, de veloping hypotheses e xplaining deviations from the nominal state, including investigating appropriate environmental parameters, (4) implementing relevant methodology to test hypotheses and build evidence for th e greatest likelihood explaining the phenomenon, and (5) diagnostic interpretation based on weight of evidence or reevaluation of hypotheses and methods if necessary. A diagnostic method requires knowledge of baseline or reference va lues. However, these values should not be considered fixed and can continually be m odified as new information is gained.

PAGE 32

1.4.2. Description of Sampling Sites I sampled four patch reefs and two depths on one forereef within FKNMS, near Key Largo (KL), and one patch reef in Biscay ne National Park (BNP), which is along the Northern Florida Reef Tract (F ig. 1.2). These sites were chosen in consultation with John Halas, Resource Manage r (now Upper Keys Regiona l Manager), Upper Keys Region of FKNMS, and Richard Curry, Scienc e Coordinator, BNP. Biomarker sampling had been conducted along the depth gradient within FKNM S near Molasses Reef since March 1999 to determine if levels of oxida tive-damage products, an tioxidant enzymes, and specific components of cellular struct ural integrity in the star coral ( Montastraea annularis species complex) varied with coral bleaching, seas onal and increased SST, and water depth (Downs et al. 2002). Downs et al. (2002) chose these sites (Fig. 1.2, KL 3 m KL 18 m) because they were near long-te rm monitoring locations, including those of the EPA-FKNMS Coral Reef Evaluation and Monitoring Project (e.g., Wheaton et al. 1998) and the Molasses Reef SEAKEYS Pr ogram C-MAN buoy, which records hourly weather and water quality parameters (http://coral.aoml.noaa.gov/cman/). Sampling at these sites continued during my project. Algae Reef (AR) was added because of potential groundwater contamination, as indicated by a cyanobacterial outbreak in the early 1990s (e.g., Kuta & Richardson 1997) and White Ba nks (WB) was added because it was considered to be a relatively pristine site Alinas Reef (BNP) was chosen for its proximity to Miami and to coincide with ot her long-term monitoring projects. Previous research indicated that a quarterly sampling design was sufficient to detect changes in coral physiology resulting from seasonal and stressor variation, therefore sampling was conducted in March/April, June, August, and October/November of 2001 and 2002, plus February 2003 (Table 1.1). 1.4.3. Biology and Ecology of Montastraea annularis complex My study focused on corals of the Montastraea annularis species complex (Fig. 1.3), which is an important reef-building coral found throughout the Caribbean over a range of depths (app rox. 1 50 m). The M. annularis complex is made up of three morphotypes/species including M. annularis M. faveolata and M. franksi. Taxonomic

PAGE 33

differences among these morphologies rema in uncertain (Weil & Knowlton 1994, Szmant et al. 1997, Severance et al. 2004a,b, Fukami & Knowlton 2005, Severance & Karl 2006). The corallite structure is simila r, suggesting that they may be one species that exhibits different morphologies with ch anges in environmental conditions. However, all three morphotypes can be found in th e same habitat and some physiological differences (e.g., aggression, growth) are known (Weil & Knowlton, 1994). M. annularis grows in columns with living polyps re stricted to the top of each column. Montastraea faveolata grows in large mounds with bumps aligned in regular rows that extend down the mound, and its cora llites are evenly extended. Montastraea franksi is usually found in deeper water and tends to grow in smaller mounds or flattened plates with irregular bumps and it has unevenly dist ributed and extended corallites. Recently, molecular techniques attempted to differentiate among the three morphotypes. Lopez et al. (1999) did not detect comparable differences between M. franksi and M. annularis with either AFLPs or a micr osatellite locus, while Fuka mi and Knowlton (2005) found low genetic variability among the three members of the M. annularis complex using complete mitochondrial DNA sequences. Members of the M. annularis complex are protogynous hermaphroditic, broadcast spawners (Fadlallah 1983). Spawning occurs annually in late summer (mid-August to mid-September; usually immediately after the full moon) releasing approximately 7202016 eggs/cm2 of coral tissue (Mergner 1971, Szmant 1986, 1991, Mendes & Woodley 2002). Colonies <100 cm2 are rarely reproductive (Szmant 1986). Oogenesis begins in mid-May and spermatogenesis in mid-July (Szmant 1986). Juvenile recruitment of M. annularis from sexual reproduction is apparently in frequent, as there are typically fewer juveniles than adults in a population. Asexual reproduction accounts for most recruitment of new colonies of M. annularis through fission or sepa ration resulting from mortality. Growth rates of M. annularis are approximately 0.50 1.2 cm/yr in the Florida Keys (Hudson et al. 1976). Growth rates vary with depth, water clarity and temperature, and historically have been greatest in mids hore reef areas of the FKNMS, where water is shallow and consistently clear with low temperature variability (Hudson 1981).

PAGE 34

1.4.4. Specific Objective of Dissertation The specific objectives of my study were to 1) assess reef condition using a hierarchical approach includi ng selected environmental, community, popul ation, colony and cellular parameters; 2) evaluate the abil ity of individu al indicators to distinguish differences among sites, times and stressors; and 3) diagnose the physiological state of selected reefs based on weight of evid ence through the integration of multiple indicators. The overall scope of this pr oject is outlined below (also see Fig. 1.1). I. Environmental assessments of study s ites (Chapter 2), including water temperature, turbidity, nutrient concen trations, and sedimentation, addressed these three questions: A. Were potential environmental stressors detected during the study period? B. What time periods were most stressful? C. Were there any environmental differences among sites? II. Characterization of reef condi tion at hierarchical scales: A. I assessed community-scale condition of selected patch reefs using the Atlantic and Gulf Rapid Reef Asse ssment (AGRRA) protocol (Lang 2003) in March 2002. This assessment determined the condition of reefs by evaluating major benthic taxa: cora ls, fish and algae (Chapter 3), addressing these four questions: A. Does reef structure indicate whet her conditions in the recent past were suitable for reef growth and development? B. Are significant differences ev ident among the reefs examined? C. Is there evidence of recent change (e.g., recent mortality)? D. How do data from these reefs compare with Caribbean-wide AGGRA data sets? B. I monitored population densities and vi sually assessed bleaching and shell breakage in a surrogate indicato r group, symbiont-bearing (larger) benthic Foraminifera, which lived near the corals (Chapter 4) to address two questions:

PAGE 35

A. Are water quality and other enviro nmental conditions suitable for calcifying organisms that host algal endosymbionts? B. Do foraminifers indicate expos ure to chronic or acute photic stress? C. I monitored sampling-induced lesions and assessed overall condition (e.g., bleaching, disease, overgrowth, etc.) of individual colonies of Montastraea annularis species complex (Chapter 5) to address these two colony-scale questions: A. Can corals at the study sites re cover from mechanically-induced lesions? B. Is there evidence for compromised physiological function of corals (e.g, reduced regeneration rates or increased mortality) at any site? D. I used indicators of cellular phys iology (Chapter 6) acquired by collaborators at MUSC/NOAA a nd Envirtue Biotechnology using a Cellular Diagnostic System (CDS) to address four questions: A. Did corals deviate from a nominal cellular state? B. Did cellular profiles indicate that corals were stressed and if so, where and when? C. To what types of stress were the corals responding? D. What were likely mechanisms of stress? III. Diagnosis based on weight of evidence (Chapter 6) addressed these two questions: A. Where did sampled corals fall on a p hysiological scale of nominal to diseased state? B. Can potential stressors be linke d to physiological function? IV. Conclusions (Chapter 7) examines the following: A. evaluation of strengths and cav eats of individual indicators, B. summary of diagnosis, and C. recommendations for future research.

PAGE 36

Table 1.1. Sampling dates for 2001 2003 at Bi scayne National Park (BNP), Algae Reef (AR), White Banks (WB) and Key Largo (K L) depth gradient (3, 6, 9 and 18 m). Sampling period Sampling Date Site Sampled March/April 2001 3/30/01 BNP 3/31/01 KL 6 m, WB, AR 4/1/01 KL 3 m, KL 9 m, KL 18 m June 2001 6/26/01 BNP 6/27/01 KL 6 m, WB, AR 6/28/01 KL 3 m, KL 9 m, KL 18 m August 2001 8/29/01 KL 6 m, WB, AR 8/30/01 KL 3 m, KL 9 m, KL 18 m 8/31/01 BNP October 2001 10/23/01 BNP 10/24/01 KL 6 m, WB, AR 10/25/01 KL 3 m, KL 9 m, KL 18 m March 2002 3/22/02 BNP 3/24/02 KL 6 m, WB, AR 3/25/02 KL 3 m, KL 9 m, KL 18 m June 2002 6/23/02 KL 6 m, WB, AR, KL 3 m, KL 9 m, KL 18 m 6/25/02 BNP August 2002 8/19/02 KL 6 m, WB, AR, KL 3 m, KL 9 m, KL 18 m 8/21/02 BNP November 2002 11/1/02 KL 6 m, WB, AR 11/2/02 BNP 11/3/02 KL 3 m, KL 9 m, KL 18 m February 2003 2/8/03 KL 6 m, WB, AR 2/9/03 KL 3 m, KL 9 m, KL 18 m 2/10/03 BNP

PAGE 37

Stressor Input Biochemical Responses Physiological Responses Altered Physiological Performance Reproduction Growth Population Impact Community Structure and Dynamics Ecosystem Structure and Function Minutes to Days Hours to Weeks Days to Months Months to Years Years to Decades Cellular Diagnostic System (Ch. 6) Coral Regeneration Rates (Ch. 5) Symbiont-bearing foraminifera densities (Ch. 4) Coral Recent Mortality (Ch. 3) Bleaching in A. gibbosa (Ch. 4) Community Assessments (AGRRA) (Ch. 3) Environmental Assessments (Ch. 2) Figure 1.1. Multi-scale approach to study the effects of stress on the marine environme nt. Time Scale

PAGE 38

Figure 1.2. Chart of seven sampling sites in the Florida Keys Nati onal Marine Sanctuary and Biscayne National Park. squares designate the 6 m sites, including the four patch reefs Key Largo (KL) 6 m (25 0192' N, 80 23.844' W), White Banks (WB) (25 02.232' N, 80 22.496' W), Algae Reef (AR) (25 08.799' N, 80 17.579' W) and Alinas Reef (BNP) (25 23.185' N, 80 09.775' W). circles designate sites along the depth gradient, including two patch reefs, KL 3 m (25 02.447' N, 80 25.442' W) and KL 6 m, and two depths on one forereef, KL 9 m (25 00.146' N, 80 23.626' W) and KL 18 m (25.206' N, 80 23.023' W).

PAGE 39

Figure 1.3. Representative colonies of (A) Montastraea faveolata (B) M. annularis and (C) M. franksi. Pictures taken by Roy Price. A B C

PAGE 40

2. Environmental Assessments 2.1. Introduction Florida reefs have undergone severe degr adation over the past several decades (Andrews et al. 2005), though reasons for cora l loss are not fully understood. Diagnosing reef condition requires knowing reef history and potential stressors to whic h the reef was exposed (Depledge et al. 1999, Jameson et al 2001, Downs et al. 2005b). Managers of the Florida Keys National Marine Sanctu ary have requested information on the relationship between water quality and the incidence of coral disease and mortality (Bruckner 2002). Effective management also requires identifying indirect and direct stressors on reefs (Bruckner 2002). Coral reef degradation has been linked to changes in the natural coastline and increases in sedimentation, tu rbidity, temperature, light and nutrients (Waddell 2005). The objective of this chapter is to characterize these major environmental parameters, using data collected at my study sites and datasets available from other sources collected in the vicinity of my sites. 2.1.1. Coastal Wetlands Coastal wetlands filter runoff, stabilize sediments and absorb nutrients, thereby helping maintain the clear, relatively nutrient-poor coastal waters required for coral reefs (Yentsch et al. 2002). Coastal waters off de veloped areas of the Florida Keys have an estimated 42% higher nitrogen and 79% highe r phosphorus load due to stormwater than undeveloped areas (Kruczynski & McManus 2002). Coastal development increases turbidity through increased erosion and runoff, reduces filtering by wetlands and destabilizes sediments (Kruczynski & McMa nus 2002). Higher turb idity follows high rainfall during the wet season (May through November), reducing solar energy for photosynthesis by coral zooxanthe llae, while lower turbidity occurs during the dry season (December through April), leaving zooxanthellae more susceptible to photo-oxidative stress from solar radiation.

PAGE 41

Since 1960, development and constructi on of canals has resulted in > 15 % (approx. 1278 ha) loss of coastal mangroves in the Upper Keys (Strong & Bancroft 1994). Mangroves and seagrasses are important producers of colored dissolved organic matter (CDOM), which has a variety of benefits to coastal organisms (Coble et al. 2004). Alteration and destruction of watershed and coastal wetlands reduce natural and consistent sources of CDOM that are tidally flushed into reef waters. For example, waters overlying Algae Reef, which lies off a mangrove coastline, have a consistent source of CDOM, whereas waters at sites al ong highly developed porti ons of Key Largo, such as KL 6 m, have more variable con centrations of CDOM (A youb et al. 2006). CDOM acts as a scavenger of a variety of tr ace metals and organic pollutants including polyaromatic hydrocarbons (PAHs), removing them from solution and decreasing their toxicity to marine organisms (Coble et al 2004). CDOM also ac ts as a sunscreen by rapidly absorbing shorter wavelengths such as ultraviolet (UV) light. Shorter, higher energy wavelengths (including blue, viol et, UV-A and UV-B) can result in photooxidative stress in zooxanthe llae (e.g, Lesser et al. 1990, Lesser 1996), which has been linked to coral bleaching (Downs et al. 2002). Therefore, decreases in natural suncreens, such as CDOM, can result in an increase in bleaching of corals a nd other reef-dwelling, symbiont-bearing organisms (Williams 2002, Hallock et al. 2006a,b). 2.1.2. Sedimentation and Turbidity As human populations have increased in the Florida Keys, the combination of dredging and coastal runoff has increased sedi ment loads and turbidity on reefs. The average underwater visibility dropped from 175 ft. to approximately 35 ft. following the construction of finger-fill canals in the ear ly 1970s (Krucynski & McManus 2002), and since then has increased slightly to betw een 50 and 80 ft. dependi ng on location (Yentsch et al. 2002). Sedimentation rates between 1 to 10 mg cm-2 d-1 and suspended sediment concentrations <10 mg l-1 are considered averag e for Caribbean reefs; values above these are considered potentia lly stressful to corals (Rogers 1983, 1990). Sedimentation limits coral reef development because sediments ma y block sunlight needed for photosynthesis,

PAGE 42

abrade coral tissues, change nutrient supply, and increase the energy corals must expend to rid themselves of sediment (Woolfe & Larcombe 1999). Nutrie nt-rich sediments can smother corals by increasing microbial activit y and creating anoxic conditions (Weber et al. 2006). Experimentally increased sedimentation and turbidity significantly increased coral respiration rates (Telesnicki & Goldberg 1995, Abdel-Salem & Port er 1998), negatively affected growth and calcif ication rates (Dodge et al. 1974, Bak 1978, Kendall et al. 1983, Tomascik & Sander 1985, Hubbard 1986), re production (Kojis & Quinn 1984), and recruitment (Wittenberg & Hunte 1992, Richmond 1997). Meesters et al. (1992) found that the regeneration rate of boulder coral, Montastraea annularis, was lower in areas of higher sedimentation. Labor atory experiments with M. annularis have shown variable results, with some suggesti ng it is an inefficient sedime nt rejecter (7.5 15 mg h-1) (Bak and Elgershuizen 1976) and others showing that these corals have relatively high clearing rates (28 66 mg cm-2 h-1; Abdel-Salem and Porter 1988). Sediment properties, including grain-size a nd organic and nutrient content, play key roles in determining sedimentation stre ss in corals and their ability to remove sediments (Weber et al. 2006). Silt-sized (< 63 m), organic-rich sediments can stress corals after a short-term exposure, wher eas sandy, organic-poor sediments have little effect (Weber et al. 2006). Although prior research classified sediment ation and turbidity as major stressors to corals, in the Florida Keys many ins hore patch reefs experience relatively high sedimentation and turbidity (B oyer & Briceo 2005) and still have higher coral cover and diversity than offshore reefs (Beaver et al 2005). Organic matter associated with sediment may provide a food source for co rals (Rosenfeld et al. 1999, Anthony & Fabricius 2000, Anthony 2006) but assimilation abilities and effects on physiology vary with coral species (Anthony 1999, Mills & Se bens 2004, Anthony 2006). High turbidity generally limits reef growth to <10 m de pth (Yentsch et al. 2002, Fabricius 2005).

PAGE 43

2.1.3. Temperature and Light Coral bleaching is considered one of the main causes of degradation of coral reefs worldwide, resulting in destruction of an estimated 16% of the worlds reefs (Hughes et al. 2003, Wilkinson 2004, Marshall & Schuttenbe rg 2006). Bleaching is a process whereby corals and other invertebrates that host symbiotic algae (e.g., some foraminifers and bivalves) lose or experien ce degradation of their algal symbionts due stressors such as disease, sedimentation, pollu tants, and changes in salinity temperature or light (Brown 1997a). In situ observations in the Caribbe an, Indian Ocean and South Pacific coral reefs suggest a correlation between coral bleaching and high seasurface temperatures (SST) (Goreau et al. 1992, Goreau & Hayes 1994, Brown et al. 1996, Wilkinson 1998, HoeghGuldberg 1999, Fitt et al. 2001). Large-scale bleaching events appear to primarily be caused by heat stress (Brown 1997b, HoeghGuldberg 1999, Marshall & Schuttenberg 2006), which results in photo-oxidative stress in the organism (Lesser et al. 1990, Downs et al. 2002). Bleaching stress can leave corals vulnerable to other stressors, including disease. For example, in the US Virgin Islands, corals that were severely bleached and later regained pigmentation died the following spring from infection by White Plague disease, resulting in a 26 48% loss of cora l cover (Miller et al. 2006). Increased seasurface temperatures, combined with ot her means of coral degradation (e.g., sedimentation, pollution, photic stress, disease, predation, et c.) threaten the health and vitality of coral reef ecosystems worldw ide (Dustan 1999, Marshall & Schuttenberg 2006). Bleaching of corals (Lesser et al 1990, Gleason & Wellington 1993, Glynn 1993, Jones et al. 1998, Lesser & Fa rrell 2004) and other symbiont -bearing marine organisms (Jokiel 1980, Williams & Hallock 2004, Hallock et al. 2006a, b) also has been attributed to high levels of solar radiation, particularly shorter, higher energy wavelengths (blue to UV, 290 490 nm). Bleaching events usually coincide with periods of calm winds, resulting in increased pene tration of solar radiation (Gleason & Wellington 1993, Glynn 1996, Wilkinson 1998). Stratospheric ozone de pletion of 10 15% in mid-latitudes, caused by anthropogenic inputs of chlorofluor ocarbons (Shick et al. 1996) increased

PAGE 44

harmful UVB (290-320 nm) reaching the Florida Reef by roughly 20 30% in spring and early summer months (Shick et al. 1996, Moran & Sheldon 2000, Hallock et al. 2006b). Negative impacts of UVB on reef organisms ra nge from death (Jokiel 1980), to depressed reproduction (Gleason & Wellington 1993), decreas ed calcification and growth (Roth et al. 1982), and increased susceptibility to ot her stressors (Drohan et al. 2005, Hallock et al. 2006b). The first two widespread coral bleaching events, in 1982-83 and 1987-88, coincided with ENSO events but also follo wed accelerated ozone depletion associated with major volcanic eruptions (Hallock et al. 1993). Global ozone depletion of approximately 4%, which occurred followi ng the Mt. Pinatubo eruption in May-June 1991, did not trigger mass coral bleaching (Shi ck et al. 1996), probably because the massive eruption also caused temporary glob al cooling. However, widespread bleaching in reef-dwelling foraminifers began shortly after th e eruption; these foraminifers are more sensitive to photic stress than temperatur e stress (Talge & Hallock 2003, Williams & Hallock 2004). 2.1.4. Nutrients Nutrification can be defined as an increa se in nutrient input to an environment that results in a detectable change in community structure (Hallock 2001, Fabricius 2005). A coral-dominated reef community ca n shift to a mixed coral-algal dominated communities following limited increase in nutrient input and to domination by nonsymbiotic filter-feeders and bioeroders followi ng a substantial nutrient increase (Hallock 1988). How much nutrient input can induce community change is difficult to quantify because so many environmental variables are involved, including how nutrient depleted offshore waters are, rates of exchange w ith nutrient-depleted offshore waters, and respiration rates which are a function of temperature (Hallock 1988, Hallock et al. 1991, 1993. Mangrove and seagrass detritus provide na tural sources of nutrients to coastal reefs. Presently, numerous point sources of nutrients exist thr oughout the Florida Keys (e.g., septic discharges, watertreatment plant discharges), as well as non-point sources (e.g., Everglades, storm-water and agricultura l run-off, groundwater contamination).

PAGE 45

Existing wastewater facili ties in Monroe County incl ude about 23,000 private onsite systems (66% permitted septic tanks, 31% unknown systems and 3% aerobic treatment units) and 246 small wastewater treatmen t plants (WWTPs) (Monroe County 2000). About 2800 of the 7200 unknown systems are suspected illegal cesspools (Monroe County 2000). These systems provide mini mal treatment and nutrient removal and, together with other non-point sources, impact coastal waters in the Florida Keys. Due to the extremely porous limestone st ructure of the Keys, wastewater from onsite systems can be detected in adjacent can als within hours and along coastal surface waters within days (Paul et al. 2000). Twelve hot spots in Key Largo were designated by the USEPA (1993) as having known or suspected degraded water qualit y as a result of poorly designed canals, use of septic tanks or cesspits, and untreated runoff (Krucyznski & McManus 2002). Loading from the canal drainage system and inshore groundwater led to elevated concentrations of nitrate and other nutrients (e.g., 1 g/l chl a; 1 M NO3) in Biscayne Bay and the Upper Florida Ke ys (Szmant & Forrester 1996, Boyer & Jones 2002). Nevertheless, nutrient concentrations meas ured in offshore waters in the Upper Keys tend to be relatively low (e.g., chl a 0.25 g/l; NO3 0.25 M; NH4 0.10 M; Szmant & Forrester 1996, Boyer & Jones 2002). Nutrients and contaminants also can be transported offshore (approximately 8 km) by surface-water and groundwater movement (Shinn et al. 1994, Reich et al. 2002). Howeve r, measuring dissolved inorganic nutrient concentrations in the water to determine po ssible nutrification can be misleading because nutrients are rapidly incorporated into reef and plankton biomass (Laws & Redalje 1979). Therefore, more sensitive a nd quantifiable biological indica tors are needed to quantify the affects of nutrients on reefs. Koop et al (2001) found th at increased nutrients caused increased mortality and reduced reproduction in corals, but that nu trients alone did not result in a shift from a coral-dominated to algal-dominated community. The Florida Keys reefs are unusual for their very low coral c over and high algal abund ance despite having abundant herbivorous fish (Szmant 2001). It is possible that nutrients and other stresses increase coral mortality and open substrat e for algae to colonize (Koop et al. 2001, Szmant 2001).

PAGE 46

2.2. Methods 2.2.1. Site Descriptions An ARC-GIS map was produced to show th e location of my study sites, Southeast Research Centers Water Quality Monitori ng Network (SERC-WQMN) sites, and area coverage of mangrove forest and wetlands (F ig. 2.1). Wetland data were obtained from the National Wetlands Inventory ( http://wetlands.fws.er.usgs.gov/NWI/Index.html ). 2.2.2. Sedimentation/Turbidity I used sediment traps to quantify sediment ation at my study sites. Each trap was 5.1 cm in diameter and 0.61 m long, constructe d from PVC pipe capped at the bottom, and secured to rebar with stainless steel hose clamps. The rebar was secured into the bottom substrate. In early March 2001, three traps were placed at Algae Reef (AR) and White Banks (WB) sites (Fig. 2.1) next to th e corals to be sampled. During sampling in late March, three traps were emplaced at th e KL 6 m and BNP sites. In August 2001, three traps were deployed at the KL 3 m site In March 2002, two mo re traps were added at each 6 m site, and one trap was placed at each of the 9 m and 18 m sites (Table 2.1). During each sampling, the pipe was capped at the top, swapped with a new trap, and brought to the surface (Table 2.1). Then at th e onshore lab, sediment was filtered with a 2.00 mm sieve to remove seagrass and coarse se diment, washed with distilled water to remove salts, dried at 70 C, and weighed. Data were converted to sedimentation rates (mg cm-2 d-1) by dividing dry sample weight by trap area and the number of days the trap was deployed. Beginning in October 2001, during each sampling five water samples were taken by SCUBA divers in the proxim ity of the corals for turbidity analysis using a portable turbidimeter (Orbeco-Hellige model 966, Farm ingdale, NY). Data were recorded as nephelometric turbidity units (NTU). 2.2.3. Temperature Temperature was measured at hourly inte rvals throughout the 2year study using temperature sensors (HOBO Data Logger H 08-001-02 series; accuracy 0.7 C) placed

PAGE 47

at each coral colony sampled at each site. Du e to failures of some sensors, sea-surface temperature (SST) data also were obtaine d from external databases such as NOAA SEAKEYS C-MAN ( http://www.coral.noaa.gov/cman/ ) at Molasses Reef (for 2001 2003. 2.2.4. Nutrients Water quality data (Table 2.2) for the Fl orida Reef tract have been collected by the Southeast Research Centers Water Quality Monitoring Network (SERC-WQMN) ( http://serc.fiu.edu/wqmnetwork/FKNMS-CD/index.htm ) quarterly since 1995. I matched my study sites to WQMN sites (Fig. 2.1, Table 2.3) using methods similar to those described by Callahan (2005), using an ARC view query tool and based on the following criteria: 1) proximity to study sites, 2) depth similarity, 3) distance to shore, and 4) benthic cover similarity under the WQNP station. Due to proximity, KL 6 m and WB were assigned the same water quality st ation. I compared surface samples for all water quality parameters (Table 2.1) during the time frame of my study. 2.2.5. Additional Environmental Data Miami rainfall data were obtained from Florida State Universitys Florida Climate Center ( http://www.coaps.fsu.edu/climate_center/ ). For comparison with quarterly data, I summed total precipitation (cm) for the tw o months prior to each sampling date. Maximum wind speeds were obtained from the NOAA SEAKEYS C-MAN at Molasses Reef for 2001 2003. Data on in tensity of ultra-violet (UV) radiation were obtained from UV sensors (Brewster UV radiometers) ( www.epa.gov/uvnet/ ) at Everglades National Park. Daily dose is defined as total energy from sunlight at 287 to 363 nm reaching one square meter of ground surface over an entir e day. Ozone data were obtained from NASAs Total Ozone Mapping Spectrometer ( http://toms.gsfc.nasa.gov/ozone/ozone. html ).

PAGE 48

2.2.6. Data Analysis I performed repeated-measures MANOVA to determine if sites differed significantly in sedimentation or turbidity. To interpret effects detected by MANOVA, I used one-way ANOVA followed by Tukey-Kr amer HSD method. Statistical analyses were performed using JMP statistical software (SAS Institute, Inc., Cary, NC, USA). 2.3. Results 2.3.1. Coastal Wetlands Algae Reef, which was offshore from John Pennekamp Coral Reef State Park, was along the coastline with the highest densit y of intact mangrove forest and wetlands (Fig. 2.1). Sites along the KL depth gradient (KL 3, 6, 9 and 18 m) were off a developed portion of Key Largo with low densities of mangrove forest and we tlands (Fig. 2.1). Some mangrove forest occurs along the barrier islands near BNP but these forests and associated wetlands are not as extensive as those found in John Pennekamp Coral Reef State Park (Fig. 2.1). 2.3.2. Sedimentation Sedimentation rates ranged from 2 to 187 mg cm-2d-1 with a mean ( SE) of 35 ( 3) mg cm-2 d-1 (n = 128) at the 6 m sites (Fig. 2.2). Se dimentation varied significantly among 6 m sites (repeated meas ures MANOVA: site effect F3,8 = 37.9, p < 0.001), with time ( F7,56 = 125.6, p < 0.001), and time x site interactions ( F21,56 = 4.0, p < 0.001; Fig. 2.2). Mean sedimentation rate was highest at AR relative to the other 6 m sites throughout the year (67 9 vs. 24 2 mg cm-2 d-1; Fig. 2.2). Throughout the entire study period, AR had significantly higher mean sedimentation rate than BNP, KL 6 m (except between June August 2001) and WB (except between June 2001 and October 2001) (Tukey HSD). White Ba nks also had signi ficantly higher mean sedimentation rates throughout the study period than BN P (except between March and June 2001, August and October 2001 and June and Augus t 2002) and KL 6 m (except between June and October 2001) (Tukey HS D). BNP had significantly higher mean sedimentation

PAGE 49

rates than KL 6 m between March and June 2001 and June and August 2002 (Tukey HSD). 2.3.3. Turbidity Turbidity ranged from 0.13 to 1.4 NTU with a mean ( SE) of 0.65 ( 0.03) NTU (n = 71) for the 6 m sites and from 0.26 to 1.36 with a mean of 0.65 ( 0.03) NTU (n = 71) along the depth gradient. No corre lation was found between turbidity and sedimentation rates. Turbidity differed significantly among the 6 m sites (repeated measures MANOVA: site effect F3,7 = 10.0, p < 0.007), with time ( F5,35 = 27.7, p < 0.0001), and the time x site interactions ( F15,35 = 11.5, p < 0.0001; Fi g. 2.3A). Turbidity was highly variable at BNP, where turbidity was signifi cantly lower than all other 6 m sites in October 2001 (BNP: 0.17 0.02 vs. KL 6 m: 0.58 0.02, WB: 0.55 0.02, AR: 0.62 0.06 NTU; ANOVA: F3,8 = 32.0, p < 0.0001) and significantly higher than all other 6 m sites in August 2002 (BNP: 1.17 0.19 vs KL 6 m: 0.49 0.01, WB: 0.55 0.02, AR: 0.50 0.01 NTU; ANOVA: F3,8 = 12.3, p < 0.003). Turbidity was significantly lower at both BNP and KL 6 m than AR and WB in March 2002 (BNP: 0.62 0.01, KL 6 m: 0.69 0.04 NTU vs. WB: 1.08 0.03, AR: 1.04 0.05 NTU; ANOVA: F3,8 = 48.6, p < 0.0001). In June 2002, turbidity at KL 6 m was significantly lower than all other 6 m sites and significantly higher at BNP th an AR (KL 6 m: 0.57 0.03 NTU vs. WB: 0.98 0.04, AR: 0.83 0.07, BNP: 1.03 0.02 NTU; ANOVA: F3,8 = 22.9, p = 0.0003). Turbidity at BNP also was sign ificantly higher than turbidit y at KL 6 m in February 2003 (BNP: 0.57 0.03 vs. KL 6 m: 0.37 0.03 NTU; ANOVA: F3,8 < 4.7, p = 0.04). Turbidity also significantly differed among sites along the depth gradient (repeated measures MANOVA: site effect F3,7 = 91.8, p < 0.001), with time ( F5,3 = 37.0, p < 0.007), and the time x site interactions ( F15,8.7 = 11.2, p < 0.007; Fig. 2.3B). Turbidity was significantly high er at KL 3 m and KL 6 m than at KL 9 m and KL 18 m in October 2001 (KL 3 m: 0.60 0.02, KL 6 m: 0.58 0.02 NTU vs. KL 9 m: 0.46 0.03 and KL 18 m: 0.46 0.03 NTU; ANOVA: F3,8 = 9.3, p < 0.006). Turbidity was significantly higher at KL 3 m than at all othe r sites along the depth gradient in June 2002

PAGE 50

(KL 3 m: 1.08 0.01 vs. KL 6 m: 0.57 0.03, KL 9 m: 0.59 0.06, KL 18 m: 0.56 0.01 NTU; ANOVA: F3,8 = 57.7, p < 0.0001), August 2002 (K L 3 m: 0.88 0.11 vs. KL 6 m: 0.49 0.01, KL 9 m: 0.44 0.03, KL 18 m: 0.48 0.03 NT U; ANOVA: F3,8 = 12.7, p < 0.003), November 2002 (KL 3 m: 1.33 0.02 vs. KL 6 m: 0.51 0.08, KL 9 m: 0.38 0.04, KL 18 m: 0.95 0.08 NTU; ANOVA: F3,8 = 62.8, p < 0.001) and February 2003 (KL 3 m: 1.07 0.03 vs. KL 6 m: 0.37 0.03, KL 9 m: 0.30 0.01, KL 18 m: 0.29 0.02 NTU; ANOVA: F3,8 = 357.2, p < 0.001). Turbidity also was significantly higher at KL 18 m than at KL 6 m and 9 m in November 2002. 2.3.4. Temperature Temperature ranged from 15 32 C with a mean of 27 C at the 6 m sites. Temperature differences among the 6 m sites during my study period were within the precision of the instrument; 0.7 C (Fig. 2.4A). Mean temperat ure also was 27 C along the depth gradient. The highest (33 C) and lowest (15 C) temperatures were observed at KL 3 m in August 2001 and Janua ry 2003, respectively (Fig. 2.4B). 2.3.5. Nutrients While nutrient concentrations were genera lly low, elevated concentrations of dissolved inorganic nitrogen (Fig. 2.5A-C, 2.6B) and organic phosphorus (Fig. 2.7A) at SERC stations in proximity to AR and BN P followed heavy rainfall in October 2001 (Fig. 2.8). These increases were associated with decreases in salinity (Fig. 2.9A), and increases in turbidity (Fig. 2.9B) and chlorophyll a (Fig. 2.7D). Elevated turbidity (Fig. 2.9B and Fig. 2.10B) and organic nitrogen (Fig. 2.6C and Fig. 2.11C) concentrations at SERC stations in proximity to the KL depth gradient and WB were observed in February 2002 and April 2002, respectively, which corres ponds with high winds (Fig. 2.12) that may have caused sediment resuspension (Fig. 2.2A, B). Organic nitrogen concentrations and dissolved oxygen were highest at the offs hore site in proximity to KL 9 and 18 m (Fig. 2.11C). High TN:TP (Fig. 2.13A) a nd DIN:TP (Fig. 2.13B) suggest that phosphorus limited algal growth in that area duri ng that time. Nitrite concentrations were elevated near BNP site in August 2002, follo wing heavy summer rainfall in Miami (Fig.

PAGE 51

2.8) but other nutrient concen trations remained low. Elevated inorganic phosphorus concentrations were observed at AR (Fig. 2.7A C) and along the KL depth gradient (Fig. 2.14A, C) in late October 2002 but did not corr espond with stormy weat her. Nitrate also was elevated along the KL depth grad ient at that time (Fig. 2.15A). 2.3.6. Additional Environmental Data Rainfall was higher in 2001 than in 2002, with the highest rainfall in September 2001 (46 cm) and lowest in February 2001 (0 .13 cm) (Fig. 2.8). Rainfall in 2002 was highest in July 2002 (39 cm) a nd lowest in January 2002 (0.5 6 cm) (Fig. 2.8). Mean and maximum wind speeds were highest in Nove mber 2001 (15 and 38 knots, respectively; Fig. 2.13) over the two year study period. The ratio of ozone to daily dose UV wa s higher between 2001 and 2003 compared to 1998 (Fig. 2.16A), when coral bleaching wa s high. Additionally, daily dose of UV (J m-2 d-1) between 2001 and 2003 was low in co mparison to 1998 (Fig. 2.16A). UV-A irradiance was highest in 1998 and relatively low in 2001 (Fig. 2.16B). High UV-B irradiance was observed in 1999 and 2002 (F ig. 2.16C) but overall was temporally variable. 2.4. Discussion Sedimentation was consisten tly highest at AR throughout the entire study period and was consistently above thres hold stress levels (>10 mg cm-2 d-1; Rogers 1983) at AR and WB. However, sediment collected in my traps was a combination of settled suspended sediment and resuspended sediment s, and sedimentation stress varies with sediment characteristics (Weber et al. 2006). Sediments at AR a nd WB were dominated by coarse carbonate sands (particularly Halimeda ), whereas KL 6 m and BNP were dominated by smaller grain sizes (i.e., muds). Sedimentation rates at KL 6 m and BNP also frequently exceeded threshold levels in 2001. Along the depth gradient, sedimentation rates were generally low and dominated by silty sediments. S ilt-sized and organic-rich sediments are more stressful to corals than sandy sedi ments or organic-poor

PAGE 52

silts (Weber et al. 2006), therefore sedime nts at KL 6 m and BNP are likely more detrimental to coral physiology th an those at AR and WB. Higher sedimentation rates followed peaks in rainfall and maximum wind speeds in 2001. These peaked in early fall as a resu lt of Hurricane Gabrielle that made landfall on September 14, 2001 as a tropical storm n ear Venice, FL, and then passed over the peninsula exiting near Titusville, FL. The hi ghest sedimentation rates were observed in winter months between October 2001 and March 2002, when precipitation was low, and were likely the consequence of resuspension of sediments during winter storms. There was no correlation between sedi mentation rates and turbidity, which may be a result of different sampling frequencies or differences in sediment type among sites. In 2002, sedimentation rates were again lowest in summer and early fall (June through November), but in general sedimentation rate s were considerably lower in 2002 than in 2001 and showed no relationship to mean preci pitation. No hurricane or tropical storm activity affected South Florida in 2002; the combination of less rain and lower winds in 2002 may explain lower sediment ation rates that year. Turbidity was highly variable at BNP, with the highest turbidity in the summer and lowest in winter. Turbidity was consis tently highest at KL 3 m, where sediments were muddy. High turbidity can block sunlig ht essential for photosynthesis in symbiontbearing organisms such as corals (Yentsch et al. 2002). On the other hand, corals and symbiont-bearing organisms on reefs with lo w turbidity (e.g., KL 6 m) may be more susceptible to photic stress (Williams 2002, Le sser & Farell 2004). Turbidity at KL 6 m was the most consistent and rela tively low throughout the year. Temperature did not vary among the 6 m sites but small differences were seen along the depth gradient, with the shallow inshore patch reef (KL 3 m) showing the largest range in temperature. Due to the high ratio of ozone to daily UV between 2001 and 2003 and infrequent high temp eratures (>30 C), the probab ility of coral bleaching as a consequence of high levels of oxidative resp onse (due to temperature and light) were low during my study. No bleaching wa s observed in 2001-2002, unlike 1998 when bleaching was observed Caribbean-wide includ ing Florida (Beaver et al. 2005), and in 1999 when bleaching was observed at KL 9 m and 18 m (Fauth et al. 2003).

PAGE 53

In the Upper Florida Keys, elevated n itrogen and chlorophy ll concentrations occur near marinas and canals (e.g., 1 g/l chl a; 1 M NO3) but generally decrease to oligotrophic concentrations (e.g., chl a 0.25 g/l; NO3 0.25 M; NH4 0.10 M) within 0.5 km of the coast (Szmant & Fo rrester 1996). However, following heavy rainfall and high wind events, elevated levels of total di ssolved phosphorus (0.30 M; within days) and elevated concentrations of ammonia and chlorophyll a (4.0 M and 0.45 g/l, respectively; within 1 3 wks) can r each offshore reef sites (Lapointe & Matzie 1996). High rainfall in September 2001 was fo llowed by decreased salinity and increased turbidity, inorga nic nitrogen (NO3, NO2, NH4), organic phosphorus (TP, APA) and cholorophyll a in October 2001 near my study s ites. Elevated chlorophyll a indicates increased phytoplankton abundance and biomass. Phytoplankton growth in the FL Keys is typically limited by phosphorus (Boyer & Briceo 2005). Alkaline phosphatase enables phytoplankton to use organic phosphate for growth when dissolved inorganic phosphate concentrations are low (Dyhrman & Ruttenberg 2006) and inorganic nitrogen is available. In 2002, rainfall was highest in the early summer (June) but it is difficult to determine if it had an affect on nutrient concentrations due to the low sampling frequency. Higher nutrient c oncentrations did not always correspond with increased rainfall. The predominant form of nutrients in the Florida Keys water column is organic (Szmant & Forrester 1994, Boyer & Briceo 2005). Elevated orga nic nitrogen in proximity to the Key Largo depth transect in June 2002 may result from high wind speeds during that time and sediment resuspen sion, as previously observed during winter storms in the Florida Keys (Szmant & Forre ster 1996) and along the Great Barrier Reef (Ullman & Sandstorm 1987). Periodic upwellin g along the shelf edge is another source of elevated phosphorus ( 0.2 M PO4) to offshore reefs (Szmant & Forrester 1996) and may account for elevated concentrations at AR and along the KL depth gradient in October 2002. In this case, inorganic nitr ogen may actually be limiting to phytoplankton growth as shown by low ratios of nitrogen to phosphorus. In June 2002, the State of Florida created the Key Largo Wastewater Treatment District with the goal to elim inate septic tanks and illegal cesspits and decrease nutrient

PAGE 54

loads to coastal waters. By 2010, designated Priority Hot Spots including coastal areas between KL 6 m and AR in South Key La rgo will receive community wastewater collection with advanced wastewater tr eatment (Monroe County 2000), which will potentially reduce the nutrient load and asso ciated contaminants to these reefs. I recommend continued monitoring at these study sites to determ ine if advances in water quality result in improved reef condition. 2.5. Conclusions Overall, large differences were observed among 6 m sites in sedimentation rates and turbidity but not in temper ature. Sites along highly deve loped coastlines (KL 3 m, 6 m and BNP) were dominated by silt-sized sedi ments, which are poten tially more stressful to corals. Algae Reef, offshore a mangrove lin ed coastline, had the highest sedimentation rates, which were sand-dominated and likely resulted from resuspension. Precipitation and wind speeds were generally higher in 2001 than 2002, resulting in decreased salinity and higher sedimentation rates, turbidity and nutrie nt loads near most study sites in the fall and winter 2001. Environmental conditions potentially were stressful during that time. No extreme temperatures or irradian ce levels and no coral bl eaching were observed during my study period. Relationships among the environmental data discussed above and other indicators are examined in Chapter 6.

PAGE 55

Table 2.1. Deployment and collection of sediment traps in 2001 and 2002; NA indicates no traps added; NC indicates no traps collected and swapped; A indicates nu mber of traps added; C indicates nu mber of traps collected and swapped. March 2001 June 2001 August 2001 October 2001 March 2002 June 2002 August 2002 November 2002 KL 3m NA/NC NA/NC 8/30/01 (3A) 10/25/01 (3C) 3/25/02 (2A & 3C) 6/24/02 (5C) 8/19/02 (5C) 11/1/02 (5C) KL 6m 3/31/02 (3A) 6/27/01 (3C) 8/29/01 (3C) 10/24/01 (3C) 3/24/02 (2A & 3C) 6/23/02 (5C) 8/20/02 (5C) 11/2/02 (5C) KL 9m NA/NC NA/NC NA/NC NA/NC 3/25/02 (1A) 6/23/02 (1C) 8/19/02 (1C) 11/1/02 (1C) KL 18m NA/NC NA/NC NA/NC NA/NC 3/25/02 (1A) NA/NC 8/19/02 (1C) 11/1/02 (1C) WB 3/14/01 (3 A) 3/31/01 (3C) 6/27/01 (3C) 8/29/01 (3C) 10/24/01 (3C) 3/24/02 (2A & 3C) 6/24/02 (5C) 8/20/02 (5C) 11/2/02 (5C) AR 3/14/01 (3 A) 3/31/01 (3C) 6/27/01 (3C) 8/29/01 (3C) 10/24/01 (3C) 3/24/02 (2A & 3C) 6/24/02 (5C) 8/20/02 (5C) 11/2/02 (5C) BNP 3/30/01 (3A) 6/26/01 (3C) 8/31/01 (3C) 10/23/01 (3C) 3/22/02 (2A & 3C) 6/25/02 (5C) 8/21/02 (5C) 11/3/02 (5C)

PAGE 56

Table 2.2. SERC Water Quality Monitoring Network sampling list including summary statistics for all surface wate r quality variables for all FKNM S outer reef stations between 1995 and 2005 (Boyer & Jones 2002, Boyer & Briceo 2005). Water Quality Parameters Median Min Max Salinity (practical scal e salinity) 36.2 26.7 37.8 Temperature (C) 26.9 16.3 32.2 Dissolved oxygen (DO, mg/l) 5.9 0.1 13.5 Turbidity (NTU) 0.33 0.00 10.14 Nitrate (NO3-, M) 0.06 0.00 2.30 Nitrite (NO2 -, M) 0.03 0.00 0.71 Ammonium (NH4+, M) 0.24 0.00 2.73 Total organic nitrogen (TON, M) 8.95 0.00 67.72 Total nitrogen (TN, M) 9.42 1.00 67.85 Soluble reactive phosphate (SRP, M) 0.02 0.00 0.23 Total organic carbon (TOC, M) 144.2 18.4 1054.8 Total phosphorus (TP, M) 0.17 0.00 1.22 Chlorophyll a (CHL a, g/L) 0.21 0.00 3.12 Alkaline phosphatase activity (APA, M/h) 0.04 0.01 0.79

PAGE 57

Table 2.3. Study sites and associated Southeas t Research Center (SERC) water quality st ations. SERC sampling dates for KL dep th gradient and WB are the same. SERC sa mpling dates at KL 6 m, KL 9 m, KL 18 m and WB are the same as KL 3 m. Study Site SERC Station SERC Station # Distance from Study Site (m) Depth (m) Distance from Shore (m) SERC Sampling Dates Study site SERC Study site SERC KL 3 m Mosquito Bay 223 2100 3 3.5 4200 2800 1/26/2001, 6/19/2001, 7/27/2001, 11/16/2001, 2/6/2002, 4/25/2002, 8/29/2002, 10/23/2002, 2/5/2003 KL 6 m Molasses Reef Channel 224 1400 6 7.5 8000 7100 KL 9 m Molasses Reef 225 2400 9 36 9700 10700 KL 18 m Molasses Reef 225 1400 18 36 10100 10700 WB Molasses Reef Channel 224 2200 6 7.5 8000 7100 AR White Banks 218 1400 6 3.5 6200 5800 1/25/2001, 6/14/2001, 7/20/2001, 11/7/2001, 2/6/2002, 4/24/2002, 8/28/2002, 10/23/2002, 2/3/2003 BNP Ajax Reef 206 6000 6 8 6000 3800 1/25/2001, 6/13/2001, 7/20/2001, 11/7/2001, 2/5/2002, 4/23/2002, 8/27/2002, 10/22/2002, 1/31/2003

PAGE 58

WB KL 6 m AR KL 3 m KL 9 m KL 18 m BNP Unconsolidated Bottom Forest Scrub -Shrub Aquatic Bed SERC stations Study Sites Main Land 223 224 225 218 206South Florida Figure 2.1. Map showing location of study sites (large yellow circ les), Southeast Research Center Water Quality monitoring stations ( small blue circles and blue station numbers) and area coverage of mangrove and wetlands in the study region.

PAGE 59

0 20 40 60 20 40 60 80 100 120 140 160 180 0 KL 6 m WB AR BNP AR -JUN01 JUN -AUG01 AUG 01-OCT01 OCT01-MAR02 MAR -JUN02 JUN-AUG02 AUG-NOV02 NOV02-FE B03 M KL 3 m KL 6 m KL 9 m KL 18 mA BSedimentation rate (mg cm -2 d -1 )Figure 2.2. Mean ( SE) sedimentation rate (mg cm -2 d -1 ) between March 2001 and February 2003 at (A) the 6 m sites and (B) along the depth gradient. T he KL 6 m site is included in both panels. Note smaller scale along depth gradient.

PAGE 60

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6A Turbidity (NTU) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003B KL 3 m KL 6 m KL 9 m KL 18 m KL 6 m WB AR BNPFigure 2.3. Mean ( SE) turbidity (nephelometric turbidity units: NTU ) between October 2001 and February 2003 at (A) the 6 m sites and (B) along the depth gradient.

PAGE 61

Figure 2.4. Mean temperature (C) between January 2001 and February 2003 at (A) the 6 m sites and (B) along the depth gradient. Molass es Reef (MR) data from the SEAKEYS C-MAN buoy also is plotted. The dotted line represents the temperature where corals typically begin to bleach (31C; Andrews et al. 2005). A B MR KL 6 m WB AR BNP 18 20 22 24 26 28 30 32Temperature ( o C) MR KL 3 m KL 6 m KL 9 m KL 18 m 18 20 22 24 26 28 30 32Jan 2001 Apr 2001 Jul 2001 Oct 2001 Feb 2002 May 2002 Aug 2002 Dec 2002 Feb 2003

PAGE 62

0 0 2 0 4 0 6 0 8 1 1 2 1 4 1 6 1 8 0 0 0 5 0 1 0 1 5 0 2 0 0 2 0 4 0 6 0 8 1 1 / 1 / 0 1 4 / 1 1 / 0 1 7 / 2 0 / 0 1 1 0 / 2 8 / 01 2 / 5 / 0 2 5 / 1 6 / 0 2 8 / 2 4 / 0 2 1 2 / 2 / 0 2 224 218 206 NO 3 ( m M) NO 2 ( m M) NH 4 ( m M)A B C Figure 2.5. (A) Nitrate (NO 3 ), (B) nitrite (NO 2 ), and (C) ammonium (NH 4 ) for SERC stations between 2001 and 2003; Station 224 is closest to KL 6 m and WB, Sta tion 218 is closest to AR, and 206 is closest to BNP; dashed li ne represents median values for FKNMS reef stations between 1995 and 2005.

PAGE 63

0 0.5 1 1.5 2 2.5 3 0 5 10 15 20 25 0 5 10 15 20 25 1/1/014/11/017/20/0110/28/012/5/025/16/028/24/0212/ 2/02 224 218 206 A B CTON ( m M) DIN ( m M) TN ( m M)Figure 2.6. (A) Total nitrogen (TN), (B) dissolved inorganic nitrogen ( DIN), and (C) total organic nitrogen (TN) for SERC stations between 2001 and 2003; Station 224 i s closest to KL 6 m and WB, Station 218 is closest to AR, and 206 is closest to B NP; dashed line represents median values for FKNMS reef stations between 1995 and 2005.

PAGE 64

0 0.1 0.2 0.3 0.4 0.5 0 0.02 0.04 0.06 0.08 0.1 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 1/1/014/11/017/20/0110/28/012/5/025/16/028/24/0212/ 2/02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 A B C DTP ( m M) APA ( m M/h) SRP ( m M) CHLA ( m M/L) 224 218 206Figure 2.7. (A) Total phosphorus (TP), (B) alkaline phosphatase activity (APA), (C) soluble reactive phosphorus (SRP) and (D) chlorophy ll-a (CHLA) for SERC stations between 2001 and 2003; Station 224 is closest to KL 6 m and WB, Station 218 is closest to AR, and 206 is closest to BNP; dashed line represents median values for FKNMS reef stations between 1995 and 2005

PAGE 65

0 5 10 15 20 25 30 35 40 45 50Jan-01 Mar-01 May-01 Ju l-01 Sep-01 No v-01 Jan 02 Mar-02 May-02 Jul-02 Se p-02 No v-02 Jan-03Figure 2.8. Total precipitation (cm) in Miami between January 2001 and February 2003.Precipitation (cm)

PAGE 66

33.5 34 3 4.5 35 3 5.5 36 36.5 37 0 5 10 15 20 25 30 35 1/1/014/11/017/20/0110/28/012/5/025/16/028/24/0212/ 2/02 0 0.5 1 1.5 2 2.5 3 224 218 206 0 1 2 3 4 5 6 7 8 9 10 Salinity (psu) Turbidity (NTU) Temperature (C) Dissolved Oxygen (mg/L)A B C D Figure 2.9. (A) Salinity, (B) turbidity, (C) tempe rature, and (D) dissolved oxygen for SERC stations between 2001 and 2003; Station 224 is closest to KL 6 m and W B, Station 218 is closest to AR, and 206 is closest to BNP; dashed line represent s median values for FKNMS reef stations between 1995 and 2005.

PAGE 67

32.5 33 33.5 34 34.5 35 35.5 36 36.5 37 37.5 0 5 10 15 20 25 30 35C Temperature (C)1/1/2001 4/11/2001 7/20/2001 10/28/2001 2/5/2002 5/16/2002 8/24/2002 12/2/2002 0 2 4 6 8 10 12D Dissolved Oxygen (mg/L) 0 0.5 1 1.5 2 2.5 3B Turbidity (NTU)A Salinity (psu)Figure 2.10. (A) Salinity, (B) turbidity, (C) temper ature, and (D) dissolved oxygen for SERC stations along a Key Largo depth gradient between 2001 and 2003; Sta tion 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; das hed line represents median values for FKNMS reef stations between 1995 and 2005. 223 224 225

PAGE 68

0 5 10 15 20 25 30 35 40A 0 0.2 0.4 0.6 0.8 1B 0 10 20 30 40C 1/1/2001 4/11/2001 7/20/2001 10/28/2001 2/5/2002 5/16/2002 8/24/2002 12/2/2002Figure 2.11. (A) Total nitrogen (TN), (B) dissolved inorganic nitrogen (DIN), and (C) total organic nitrogen (TON) for SERC stations alon g a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed line represents median values for FKNMS reef stations between 1995 and 2005.TON ( m M) DIN ( m M) TN ( m M) 223 224 225

PAGE 69

Ja n -0 1 Ma r01 Ma y-01 J ul 0 1 Se p 01 Nov-0 1 J a n-02 Ma y -02 Jul-0 2 Se p 02 N o v-0 2 J an-03 M ar-0 2 16 14 12 10 8 6 4 2 0 40 35 30 25 20 15 10 5 0Mean Wind Speed Maximum Wind Speed Figure 2.12. Mean (bars) and maximum (line) monthly wind speeds (knots) r ecorded by SEAKEYS C-MAN buoy at Molasses Reef between January 2001 and February 2003

PAGE 70

0 500 1000 1500 2000 2500 3000ATN:TP 0 2 4 6 8 10 12 14 16DIN:TPB 0 50 100 150 200 250TOC ( m M) C1/1/2001 4/11/2001 7/20/2001 10/28/2001 2/5/2002 5/16/2002 8/24/2002 12/2/2002Figure 2.13. (A) Total nitrogen to total phosphorus (TN:TP) (B) dissolved inorganic nitrogen to total phosphorus (DIN:TP) and (C) total organic carbon ( TOC) for SERC stations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed l ine represents median values for FKNMS reef stations between 1995 and 2005. 223 224 225

PAGE 71

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8A TP ( m M) 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09B SRP ( m M) 0 0.02 0.04 0.06 0.08 0.1 0.12C APA ( m M/h) 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.51/1/2001 4/11/2001 7/20/2001 10/28/2001 2/5/2002 5/16/2002 8/24/2002 12/2/2002D CHLA ( m M/L) 223 224 225Figure 2.14. (A) Total phosphorus (TP) (B) soluble reactive phosphorus (SRP), (C) alkaline phosphatase activity (APA) and (D) chlorop hyll-a (CHLA) for SERC stations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed line r epresents median values for FKNMS reef stations between 1995 and 2005.

PAGE 72

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 223 224 225 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 A B C NO 3 ( m M) NO 2 ( m M) NH 4 ( m M)1/1/2001 4/11/2001 7/20/2001 10/28/2001 2/5/2002 5/16/2002 8/24/2002 12/2/2002Figure 2.15. (A) Nitrate (NO 3 ), (B) nitrite (NO 2 ), and (C) ammonium (NH 4 ) for SERC stations along a Key Largo depth gradient between 2001 and 2003; Station 223, 224, and 225 is closest to KL 3 m, KL 6 m and KL 9/18 m, respectively; dashed l ine represents median values for FKNMS reef stations between 1995 and 2005.0 0.2 0.4 0.5 0.1 0.3 0.6

PAGE 73

B A C Figure 2.16. (A) Ozone (dobsons) and daily dose UV (J m -2 d -1 ) between 1998 and 2003. Figure taken from Ivey, J. (B) Mean irradiance (W m -2 ) of UVA between 1997 and 2003. (C) Mean irradiance of UVB (W m -2 ) between 1997 and 2003.Irradiance (W m -2 )UVA UVBIrradiance (W m -2 ) 6 8 10 12 14 16Jan-97 Jul-97 Jan-98 Jul-98 Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9Jan-97 Jul-97 Jan-98 Jul-98 Jan-99 Jul-99 Jan-00 Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03

PAGE 74

3. Community Assessments 3.1. Introduction Coral reefs have declined ra pidly over the past severa l decades, with dramatic changes in the structure and composition of these dynamic ecosystems (Byrant et al. 1998, Hughes et al. 2003, Bellwood et al. 2004, Waddell 2005). Co ral cover in the Florida Keys and Caribbean-wide was reduced 50 90% since the 1970s (Porter et al. 2002, Gardner et al. 2003, Palandro et al 2003, Palandro 2006). Between 1996 and 2004, local, regional an d global stressors caused declin es in abundance of stony coral species at 79% of the Coral Reef Evaluati on and Monitoring Proj ect (CREMP) stations and decreased mean live stony coral cover from 12% to 7% on Florida Keys reefs (Beaver et al. 2005). Patch reefs have higher remaining mean stony coral cover than offshore reefs in Biscayne National Park (1020% versus 2%; Miller et al. 2000) and in the Florida Keys (15% compared to <5%; Beav er et al. 2005). Therefore, it is important to understand what shapes community dynami cs of these reefs. The structural complexity of coral reefs protects shor elines (e.g., Kunkel et al. 2006) and provides habitat for associated organisms, which bene fit fisheries, tourism, and pharmaceuticals. The annual economic value of South Florida re efs is $7.7 billion dollars (Andrews et al. 2005) and depends on the diversity and abundance of reef organisms. Ecosystem assessments and monitoring provide baseline information on reef condition, which can be used to improve reso urce management. Monitoring detects and quantifies change in the re ef community, which can help explain underlying dynamic processes and characterize how they are disrupted by anth ropogenic and natural stresses (Williams 1994, Jameson et al. 2001, Porter et al. 2001, Hallock et al. 2004). To better understand the diverse factors affecting reefs, scientists need to understand responses of coral communities to environmental vari ation and how population dynamics differ between degraded and undegraded reefs (Done 1992, Bak & Meesters 1999, Jameson et al. 2001, 2003).

PAGE 75

Reef vitality depends on complex relatio nships among corals, fish, algae, and other organisms, whereby changes in one component can influenc e and even disrupt dynamics of another. For example, if abunda nce of herbivorous fi sh decreases, algal abundance often increases and subsequently fu rther contributes to decreased coral cover (Hughes 1994). The Atlantic and Gulf Rapid Reef Asse ssment (AGRRA) protocol focuses on assessing the condition of principa l scleractinian and hydrozoan corals that contribute most to the three-dimensional stru cture and complexity of reefs (Lang 2003). This protocol has been used throughout the Caribbean to assess over 400 reefs (see www.agrra.org) and provides a snapshot charac terization based on sele cted structurally or functionally important benthic and fish indi cators (Lang 2003 and references therein). Structural indicators include abundance of key species, benthic cover and rugosity (community structure). Functional indicator s include coral and fish size-frequency distributions and coral recrui tment (community dynamics, recruitment); recent and old partial mortality, prevalence of disease, bl eaching, and predation (coral condition); and density of herbivorous fish and urchins, and abundance of f unctional algal groups (herbivory). This chapter characterizes the community-s cale condition of f our 6 m deep patch reefs (Fig. 1.1) in March 2002 using data co llected with the AGRRA protocol (Lang 2003). I address four questions. (1) Does reef structure indicate whether conditions in the recent past were suitable for reef growth? (2) Are significant differences evident among the reefs examined? (3) How do these reefs compare with norms established by Caribbean-wide AGRRA datasets? (4) Is ther e evidence of recent change? To answer these questions, I evaluated selected para meters for coral, fish and algae. 3.1.1. Common Coral Species Understanding variation in coral abundan ce is important for understanding why some coral species may be better adapted to certain conditions than others (Bak & Meesters 1999). Corals with high rates of recruitment (e.g., Porites spp ) are favored in shallow, disturbed environments over robust, massive species with low recruitment rates (e.g., Montastraea annularis complex; Hughes & Jackson 19 85, Bythell et al. 1993). The density and size of Montastraea an important framework-building coral, provides

PAGE 76

information on environmental stability. High densities of large (>1 m) Montastraea are indicative of a stable environment (e.g., Flower Gardens Bank) whereas low densities are indicative of a marginal reef environm ent (e.g., Costa Rica; Kramer 2003). 3.1.2. Coral Colony Condition and Mortality Percent live coral cover is used as an indicator of reef health by many traditional reef-monitoring protocols (e.g., Coral Reef Evaluation and Monitoring Project (CREMP) and Global Coral Reef Monitoring Network (GCRM N)). The long-term integrity of reefs depends on recruitment, survival and growth of structure-produci ng scleractinian and hydrozoan corals (Dustan & Halas 1987, Done 1997, Kramer 2003). By quantifying the amount of recent mortality, and therefore the extent of damage, predictions can be made about whether corals are likely to recover. Co rals with large amounts of damage are unlikely to have sufficient energy to fully reco ver (Lang 2003). In general, a healthy reef is expected to have relatively low co ral mortality, so high levels of recent coral mortality indicate a major disturbance occurre d in the previous days to months (Lang 2003). Colonies are expected to have some old mortality, which represents an integration of polyp loss over time (Hughes & Connell 1999, Kramer 2003). Large, long-lived, broadcast-spawning corals such as the Montastraea annularis complex tend to exhibit higher amounts of mortality, while sma ller, short-lived brooding species (e.g. Porites ) tend to exhibit either complete or no mort ality (Bythell et al. 1993, Kramer 2003). Partial mortality also sometimes varies with coral-colony size a nd morphology (Hughes & Jackson 1980) and with predator distribution (Babcock 1985). Size frequencies of corals depend on the processes of settlement, grow th, survival, reproduction and mortality, and therefore also can be affected by envir onmental variation (Hughes & Jackson 1980, 1985, Bak & Meesters 1999). Smaller colonies typi cally have lower par tial mortality but are more susceptible to complete mortality (Hughes & Jackson 1985, Byth ell et al. 1993). Estimates of colony size also can provide information on rugosity and architectural complexity of the reef because large colonies of branchi ng and boulder corals typically provide more 3-dimensional structure than do small colonies (Kramer 2003).

PAGE 77

3.1.3. Fish Assemblage Structure My study was concerned with fish speci es that influence benthic-community dynamics and have key roles in reef ecol ogy (e.g., herbivorous fish, predatory fish, corallivorous fish) or are commercially importa nt. Presence of larg e coral heads, amount of relief on reefs, or condition of coral colonies (% live coral cover) sometimes are correlated with aspects of fish-population dynamics, such as density or diversity (Carpenter et al. 1981, Bell & Galzin 1984, Kuffner et al. 2007 but also see Roberts & Ormond 1987, Bellwood et al. 2004). Chabanet et al. (1997) found fish density only was correlated with coral cover or coral diversity on disturbed reef sites. All Florida reef sites are disturbed to some degree (Andrews et al. 2005), so such measur es are particularly appropriate. 3.1.4. Algal Biomass and Herbivory Herbivorous fishes can affect the distribut ion and abundance of al gae (Brock et al. 1979, Morrison 1988, Miller & Hay 1998, Chabanet et al. 1997), so disrupted reef fish assemblages can contribute to increased alga l abundances and decrea sed coral cover. Over the past two decades, many Caribbean r eefs have dramatically changed from coraldominated communities to algal-dominate d communities (Hallock et al. 1992, Hughes 1994, Lapointe 1997, 1999, Ostrander et al. 2000, Szmant 2001, Littler et al. 2006). A biologically intact reef is exp ected to have a low macroalgae to crustose coralline ratio, whereas declining reefs have high abunda nces of fleshy macroalgae, sometimes associated with a high abundance of Halimeda (Steneck & Detheir 1994, Kramer 2003). The cause of dramatic changes in comm unity composition and population density continues to be debated (e.g., Szmant 2001). The top-down hypothe sis contends that significant declines in herbivores, such as Diadema antillarum and herbivorous fish, are primarily responsible for increased algal abundance (e.g., Hughes 1994, Williams & Polunin 2001). The alternative, bottom-up hy pothesis is that increased nutrient flux to reef waters drives algal blooms and shifts in community compos ition (e.g., Lapointe 1997, 1999). Moreover, these processes are not mutually exclusive and can act together to change community structure.

PAGE 78

3.1.5. Comparisons with regional Caribbean values for the AGRRA Biotic Reef Index (Kramer 2003) Natural variation in community dynamics over spatial and temporal scales makes it difficult to establish baselines for a health y coral reef. However, comparisons with Caribbean-wide AGRRA values for means, and for best and worst cases, can be made using data collected using si milar protocols as an indica tor of deviations from the norm. Kramer (2003) chose selected AGRRA parameters for a biotic health index to evaluate overall reef condition. They charact erized a functional cora l reef as having at least some of the following key attributes: hi gh coral cover; high dens ities of coral > 25 cm; mid to high coral recruitm ent; low percentages of recent mortality; low abundance of macroalgae and high relative abundance of crus tose coralline algae; low occurrence of bleaching and disease; and complex trophic webs including high densities of key herbivores (fish and Diadema urchins) and carnivores. I evaluated all these reef components at my 6 m deep sites. 3.2. Methods 3.2.1. Benthic Assessment Dives were made at each 6 m site (Fi g. 1.1) in early March 2002 to assess the benthic community, focusing on corals and al gae, using the rapi d assessment methods (Table 3.1) as described by Lang (2003). At each site, I haphazardly placed a 10-m transect line just above the reef surface, then estimated live coral cover using a 1 m measuring device to estimate the proportion of the line underlain by living coral. Then swimming back along the transect line, for each coral >10 cm in diameter (Table 3.2), I recorded the following information: speci es, maximum diameter, maximum height, percent recent mortality, percent old mortality, and any apparent conditions including bleaching, disease, or overgrowths. I estimated size (maximum diameter including live and dead areas) in planar view perpendicular to the axis of growth to the nearest cm. Partial mortality was visually quantified by es timating the percentage of dead area from above in planar view (see Lang 2003). "Recently dead" was defined as any non-living parts of the coral in which corallite structures were white and either still intact or covered by a thin layer of algae or fine mud (Lang 2003). "Long dead" was defined as any non-

PAGE 79

living parts of the coral in which corallite structures were either gone or covered by organisms that were not easily removed (e.g., certa in algae and invertebrates; Lang 2003). While each colony was examined, I recorded the number of territorial gardening damselfish ( Stegastes diencaeus S. fuscus S. planifrons and S. variabilis ), total number of damselfish gardens on each head, and th e number of planktivorous damselfish, Stegastes planifrons (bicolor damselfish). I also recorded the number of Diadema along each transect. I estimated % abundance and he ight of macroalgae (f leshy and calcareous) and % abundance of crustose coralline algae within 25 X 25 cm quadrats located 1, 3, 5, 7, and 9 m along the transect. I also recorded the number a nd species of co ral recruits (<2 cm maximum diameter) within the quadrat. I measured the maximum relief (rugosity) as the highest and lowest point within a meter radius of each quadrat. 3.2.2. Fish Assessment I used belt transects at each reef to assess densities and sizes of selected key reeffish families including Acanthuridae, Balistidae, Chaetodontidae, Haemulidae, Lutjanidae, Pomacanthidae, Scaridae, and Serrani dae (Table 3.3). Divers swam a total of ten, 30 m transects, record ing fish found within a 2 m wide, visually estimated belt transect. Size of each fish was estimated and assigned to a category (<5, 5-10, 10-20, 2030, 30-40, >40 cm) using a one meter T-bar marked with 5 cm increments. I estimated fish biomass using the power function: W = a Lb, where W is the mass (grams), L is the length (cm), and a and b are parameters estimated by a linear regression of logarithmically transformed length-mass data (Marks & Klomp 2003; Table 3.3). 3.2.3. Data Analysis I tested specific hypotheses using non-parametric statis tics, including an analysis of similarities (ANOSIM) test, to determine if sites differed significantly based on a group of benthic or fish parameters (Cla rke & Warwick 2001). I used group average cluster analysis to determ ine how sites grouped based on coral or fish abundances, followed by SIMPER (similarity percentages) analyses to determine which species were primarily responsible for group ing of sites (Clarke & Warw ick 2001). Herbivorous fish included all Acanthuridae, all Scaridae, Microspathodon chrysurus and Melichthys niger

PAGE 80

Carnivorous fish included all Haemulidae, all Lutjanidae, all Serridae and Sphyraena barracuda I used a one-way ANOVA followed by the Tukey-Kramer Honestly Significant Difference (HSD) method to determine if site s differed significantly in live coral cover and coral colony density. When data did not meet the normality assumptions of ANOVA, I used Kruskal-Wallis followed by Wilcoxon rank-sum tests to test for differences among sites (Sokal & Rohlf 1995). These tests and correlation analyses were run using JMP statistical software (SAS Institute, Inc., Cary, NC). I used cluster analysis and SIMPER to assess how my study sites compared with regional means, and the best and worst regional values for a set of biotic health indices, based on data collected on Caribbean r eefs >5 m depth (Lang 2003). For these comparisons, colonies <25 cm diameter along th e transects were excl uded from analyses to allow for comparisons with regional means. The biotic health indices used in this cluster analysis included live coral cover, dens ity of corals with >25 cm diameter, density of coral recruits (<2 cm diam eter), maximum diameter of Montastraea spp., mean recent and old mortality, % diseased corals, macr oalgal index (both fleshy and calcareous algae), relative abundance of crustose coralline algae, Diadema density, and densities of herbivorous, carnivorous and total fish. For the biotic index, herbivorous and carnivorous fish were defined as above, with the exception that Haemulidae were not included in the carnivorous fish data to a llow for comparisons wi th regional means. ANOSIM, cluster analysis and SIMPER were performed using PRIMER-e v. 5.2.8 statistical software. 3.3. Results 3.3.1 Community Structure Porites astreoides P. porites and Siderastrea siderea comprised >50% of the coral species at all si tes combined (Fig. 3.1). Sites differed significantly in coral assemblages (ANOSIM Global R = 0.21, p = 0.002, Table 3.4). Porites astreoides was the most common species at Algae Reef and WB (n = 39 and 11, respectivily), Porites porites at KL 6 m (n = 25), and Montastraea annularis at BNP (n = 15) (Fig. 3.1). Key Largo 6 m had a significantly different co ral assemblage than AR (ANOSIM R = 0.60, p

PAGE 81

= 0.001) and WB (ANOSIM R = 0.21, p < 0.05) due to differences in abundance of Porites spp. (Table 3.4). Algae Reef also had a significantly differe nt coral assemblage than BNP (ANOSIM R = 0.19, p < 0.05) due to differences in abundance of Porites spp. and Montastraea annularis (Table 3.4). Percent live coral cover was low at all sites (Table 3.5) ranging from 3 18% with a mean (hereinafter SE ) of 10 % (n = 17). Mean coral cover at AR (16%) was statistically higher than at WB, BNP and KL 6 m (ANOVA F3,13 = 11.4, p < 0.0007; Table 3.5). Density of coloni es (> 10 cm) ranged from 0.4 1.4 colonies/m with a mean of 0.72 .04 (n = 31). Rugosity (maximum relief) did not differ significantly among sites, averaging 10 1 cm2. 3.3.2. Coral Mortality and Condition Old mortality ranged from 0 to 67% of a coral colony with a me an of 12 2% (n = 31 transects). Mortality di ffered by coral genus, with the Montastraea annularis complex having the highest percentage of mortality and Porites spp. tending to have a low percentage of mortality (old mortality: 2 = 61.8, df = 4, p < 0.0001; recent mortality: 2 = 13.6, df = 4, p < 0.01; Fig. 3.2). All si tes were dominated by small colonies but mean colony size (maximum diameter) was sign ificantly larger at BNP and AR, as a few large colonies were recorded at these sites ( 2 = 36.5, df = 3, p < 0.001; Table 3.5, Fig 3.3). Old mortality was positively correlate d with colony size (maximum diameter: Spearmans Rho = 0.46, p < 0.0001; maximum height: Spearmans Rho = 0.47, p< 0.0001) but no relationship was found between colony size and recent mortality (Fig. 3.3). Very little disease (<4 %) and no bleaching were observed at any of the study sites. Evidence of fish predation on corals was observed at all sites except for AR. 3.3.3. Fish Community Structure Haemulon sciurus Sparisoma viride and H. plumieri comprised > 50% of fish biomass surveyed at my sites (Fig. 3.4). Ho wever, fish assemblages differed significantly among sites (ANOSIM Global R = 0.33, p = 0.001), with assemblages at WB differing significantly from those at KL 6 m, AR and BNP (ANOSIM; Table 3.6). Fish assemblages at AR also differed significantl y from those at KL 6 m and BNP (ANOSIM;

PAGE 82

Table 3.6). White Banks and AR were domin ated by grunts (Haemulidae), KL 6 m by parrotfish (Scaridae), and BN P by a mix of parrotfish and grunts (Fig. 3.4). Biomass of Haemulidae, Lutjanidae and Scaridae differed significantly among sites (Kruskal-Wallis ANOVA: Haemulidae: 2 = 23.4, df = 3, p < 0.0001; Lutjanidae: 2 = 16.4, df = 3, p < 0.002; Scaridae: 2 = 11.2, df = 3, p < 0.02; Fig. 3.4); biomass of all three families was significantly lower at KL 6 m than at all other sites (Table 3.7). Haemulid biomass also was significantly lo wer at BNP than AR and WB (Table 3.7), while scarid biomass was significantly lower at WB than AR and BNP (Table 3.7). Densities of herbivorous and carnivorous fish differed significantly among sites (Kruskal-Wallis ANOVA: herbivores: 2 = 10.0, df = 3, p < 0.02; carnivores: 2 = 23.0, df = 3, p < 0.0001; Fig. 3.5). Algae Reef and KL 6 m had significantly higher densities of herbivorous fish (predominantly Scarid ae and Acanthuridae) than WB, which had a significantly higher density of carnivorous fish than the other three reefs (Table 3.8). Algae Reef also had a significantly higher dens ity of carnivorous fish than KL 6 m and BNP (Table 3.8). Carnivorous fish at WB and BNP were predominantly grunts with a mean size of about 15 cm (Fig. 3.6); mean si ze of carnivorous fish at KL 6 m was very small (approx. 7 cm) compared to the other site s, resulting in a low biomass (Fig. 3.4). The parrotfish assemblage at all four reefs consisted mainly of three species: Striped ( Scarus croicensis ), Stoplight ( Sparisoma viride ) and Redband Parrotfish ( S. aurofrenatum ). The grunt assemblage consisted pr imarily of four sp ecies, including Bluestriped ( Haemulon sciurus ), White ( H. plumieri ) and French Grunt ( H. flavolineatum ) plus Tomtate ( H. aurolineatum ). Although there were few snappers at any reef, those at WB consiste d mainly of Schoolmaster ( Lutjanus apodus ), at AR of Gray Snappers ( L. griseus ), and at BNP of Yellowtail Snappers ( Ocyurus chrysurus ). Acanthurids were relatively equally mixed between Ocean Surgeonfish ( Acanthurus bahianus ), Doctorfish ( A. chirurgus ), and Blue Tang ( A. coeruleus ) except at AR, which was dominated by Blue Tang. No relationship was found be tween site relief and fish densities or biomasses.

PAGE 83

3.3.4. Algal Abundance and Herbivory Macroalgae were abundant at all sites, w ith the highest relative abundance at KL 6 m (56%) and lowest at AR (43%) (Table 3.9 ). Relative abundance of coralline algae ( 2 = 8.2, df = 3, p < 0.05) and fleshy macroalgae ( 2 = 17.7, df = 3, p < 0.0005) differed significantly among sites. Algae Reef had sign ificantly more corallin e algae than KL 6 m and WB (Table 3.9). Key Largo 6 m had signi ficantly more fleshy macroalgae than WB and AR, and BNP also had significantly more fleshy algae than AR (Table 3.9). The fleshy macroalgal index (estimate of algal biomass) differed signi ficantly among sites ( 2 = 38.0, df = 3, p < 0.001; Fig. 3.6). Algae Reef had a significantly lower, and KL 6m a significantly higher, fleshy macr oalgal index than all the other sites (Table 3.9). The ratio of macroalgae to crustose coralline algae also differed significantly among sites ( 2 = 33.0, df = 3, p < 0.001). KL 6 m had a signif icantly higher index than all other sites and AR had a significantly lowe r index than all ot her sites (Table 3.10). The number of herbivorous or bicolor damselfish did not di ffer significantly among site s (Table 3.9). No Diadema were found along transects at any of the st udy sites. No rela tionship was found between herbivore abundance and algal a bundance. However, % coral cover was negatively correlated with fleshy macroalg ae biomass (Spearman Rho = -0.53, p < 0.04). 3.3.5. AGRRA Biotic Index Kramer (2003) recommended using select ed AGRRA parameters to create a biotic health index to evaluate overall reef condition (Table 3.10). A group-averaged cluster analysis based on a Bray-Curtis simila rity matrix (Fig. 3.7) showed the greatest similarity between KL 6 m and regional wors t values (78%), which clustered with WB (66%). High similarity also existed betw een AR and Caribbean mean values (77%), which together clustered with BNP (73%). Low similarity (44 %) was observed between the study sites and best values for the Caribbean region (Fig. 3.7). Major contributors to differences among my sites included maximum size of the Montastraea complex, macroalgal index, and fish de nsities (Table 3.11). Algae Reef and BNP were most similar to Caribbean mean values (Table 3.12). Several dissimilarities separated the 6 m sites from regional best va lues, including high macr oalgal index values, low densities of fishes and Diadema and low live coral cover (Table 3.13). Members of

PAGE 84

the Montastraea complex also were smaller at all site s (except AR) than the regional best values. No colonies of Montastraea were >25 cm at either WB or KL 6 m. Key Largo 6 m showed a high similarity with the regional worst values except for the lack of Montastraea >25 cm diameter, lack of coral diseas e, and higher densities of herbivorous fish (Table 3.14). Compared to regional wors t values, macroalgal index values were low at WB, AR and BNP. Densities of fishes were high at WB and AR relative to the regional worst values. 3.4. Discussion 3.4.1. Community Structure Reefs in FKNMS are not stony coral-dominated communities (<7% of benthic cover) but rather are algal-dominated (~76% of benthic cover, of which 14% is macroalgae; all CREMP sites, Beaver et al. 2005). Explanations for algal dominance of reefs include loss of key herbivorous fish a nd urchins (Hughes et al. 1999), increases in nutrient flux to reef waters (Lapointe 1997), and increased ava ilable substrate due to coral loss to other ailments (e.g., disease, bleachi ng; Szmant 2001). Coral cover at my study sites (3-18%) was comparable to other reefs throughout the Florida Keys but low compared to other Caribbean reefs (e.g., Kramer 2003). Mean living stony coral cover at reefs deeper than 5 m in the Caribbean was 26 13%, with a rang e of 3 to 58% (Lang 2003). Overall mean density of corals > 25 cm at my study sites was approximately half (0.46 colonies/m) that of Caribbean sites ( 0.93 colonies/m), indicating recruitment was limited. The most commonly encountered corals were small brooding species (e.g., Porites spp ) and stress-tolerant species (e.g., Siderastrea siderea ) at my FKNMS study sites and M. annularis complex at BNP. These observa tions are consistent with 2004 observations at CREMP sites, where Montastraea annularis complex, M. cavernosa Siderastrea siderea Porites astreoides Colpophyllia natans and Millepora complanata were most common (Beaver et al. 2005). Significant declines in percent coral cover in the Florida Keys has occurred throu gh losses in reef-building corals, M. annularis and Acropora palmata (Beaver et al. 2005). Density of M. annularis was very low at my FKNMS sites. The Montastraea annularis complex continues to be most common

PAGE 85

species in deeper reef (> 5 m) assemblages throughout the Caribbean, with M. annularis as the most common, followed by M. faveolata M. cavernosa and M. franksi (Lang 2003). However, the Montastraea annularis complex comprised only 32% of coral density at BNP and <15% at the other sites. Most coral colonies at my study sites were in the 10 29 cm size class (maximum diameter) compared to Caribbean-wide, where the most frequent coral size class was 3040 cm. This could indicate that juveniles ar e not capable of surviving or accreting past a certain size (Miller et al. 2000) or that larger colonies have died or been partly bioeroded. While values are not directly comparable because my study only counted colonies >10 cm, similar trends were reported by Beaver et al. (2005) throughout the Florida Keys where 70% of coral colonies within CREMP value-added sites (VAS) sites were <11 cm and only 5% were >50 cm. Algae Reef a nd BNP had the highest frequency of coral colonies 30 cm, with 0.41 and 0.42, respectively. Nevertheless, <10% of coral colonies were larger than one meter, indicating that conditions were only marginal for reef growth (Lang 2003). 3.4.2. Mortality and Coral Colony Condition Old mortality at BNP was high relative to other sites, with th e highest percentage of mortality in intermediate-si zed colonies (30 130 cm). The Montastraea annularis complex, which was predominantly found at BNP, had the highest percentage of mortality. Mortality was low compared to Caribbean-wide values (~22%), but small colonies (<30 cm) are less like ly to have partial mortality and more likely to experience complete mortality because they are susceptible to colony edge (i.e., bottom-associated) mortality (Meesters et al. 1996, Kramer 2003). Decreases in live coral cover may depend more on rates of coral regeneration and recr uitment than on mortality rates (Hughes & Connell 1999, Kramer 2003). Recent mortality wa s low at all of my sites. However, while monitoring recent mortality (Chapter 5) I observed overgrowth of lesions by algae and other bioeroders occurred relatively quickly (<1 year), possibly making recent mortality less evident. Mort ality from predation and tissue n ecrosis also is higher in the M. annularis complex than other coral species (Bythe ll et al. 1993). Thus, if regeneration

PAGE 86

were reduced at my study site s, it could lead to permanent patches of old mortality on corals. 3.4.3. Recruitment Abundance of coral recruits was proportio nal to adult abundance at the study sites, i.e., dominated by Porites and Siderastrea The most abundant recruit species were Porites porites (brooder), Siderastrea siderea (broadcast spawner) and Stephanocoenia intersepta (broadcast spawner). White Banks had the highest density of recruits (6.8 recruits/m2) and other sites (4.3 recruits/m2) were comparable to the regional Caribbean mean for reefs > 5 m depth (4.4 recruits/m2). High juvenile mortality rather than low recruitment may be the dominant process a ffecting community structure on offshore bank reefs in BNP (Miller et al. 2000). 3.4.4. Fish Assemblage Structure The most abundant fish families at my study sites were haemulids and scarids; other surveyed fish families were relativ ely scarce. Carnivore density (mainly invertivores ) was higher or equivalent to the density of herbivores at all sites except KL 6 m, which had the highest density of herbivores The highest densities and biomass of fish were found at WB, primarily due to large sc hools of haemulids. No relationship was found between the density or biomass of fish and the structural complexity of the sites (rugosity or coral-colony size ). Other possible factors in fluencing fish assemblage structure at my sites could include fishi ng, environmental conditions and contaminants (Downs et al. 2006), predator-prey intera ctions (Hixon 1991), larval supply and recruitment (Cowen et al. 2000), history of disturbance (Syms & Jones 2000), quality and intactness of adjacent habitats (e.g., Munday 2002), and natural spatial variation within species (Kramer 2003). 3.4.5. Algal Biomass and Herbivory No clear relationship was observed between macroalgae and herbivory. Macroalgal biomass was lowest at AR, whic h had the highest biomass of herbivorous fish. Macroalgal biomass was highest at KL 6 m, which had the highest density and

PAGE 87

second highest biomass of herbivorous fishes. However, herbivorous fishes at KL 6 m were relatively small, suggesting that smalle r fish may have less effect on macroalgae. Fleshy macroalgae found at thes e sites may not be palatable to herbivorous fish, which tend to prefer turf algae (Choat 1991). Als o, AGRRA methods do not include algal cover estimates, but rather relative abundances of fu nctional algal groups. Sites such as KL 6 m and BNP, which had higher macroalgal biom ass also had a lower percentage of live coral cover. The macroalgal index at thes e two sites was considerably higher than the regional mean of 82, particularly at KL 6 m. High abundance of m acroalgae at my study sites may be explained by low abundance of the long-spined sea urchin, Diadema antillarum which is an indiscriminate herbi vore, or by nutrient inputs (Lapointe 1997, 1999, Shinn et al. 2002, others). 3.5. Conclusions All study sites showed signs of dec line and stress as evidenced by high dissimilarity with regional best values. Low coral cover (<20%), relatively small colonies and the low abundance of framework corals such as the Montastraea annularis complex at all study sites indicates marginal reef development. Th ese sites likely have experienced declines in coral cover over the past three decades, as have many reefs in the Upper Florida Keys (Dustan 1999, Porter et al. 2001, Beaver et al. 2005, Andrews et al. 2005). The three FKNMS sites (AR, WB and KL 6 m) represent a clear gradient from best to worst based on the biotic reef health index. Ba sed on AGRRA survey methods, WB and KL 6 m were similar in coral community stru cture, with a high similarity between KL 6 m and worst regional conditions. Fish assemblages at these two sites were very different, with WB domina ted by larger invertivores and KL 6 m by smaller herbivores. Low biomass of macr oalgae and abundant coral recruits at WB indicates a better chance for juvenile coral survival at WB than KL 6 m. Low fish biomass, particularly of top predators, in dicates overall poor condition of reef fish, possible overfishing, or unsuitable habitat. Th e number of top predators was low at all sites, particularly at KL 6 m. Topographic relief did not correlate with fish biomass; what is controlling these differences rema ins unknown. Algae Reef and BNP appeared to

PAGE 88

be similar to each other and to the Ca ribbean-wide regional mean based on coral assemblage structure (colony si ze). Index values for Algae Reef were most similar to regional best values and therefore this reef appears to be in the best condition among study sites. High macroalgal abundance, particularly at KL 6 m and BNP, is an indication of poor reef condition. Coralline algae, which are important contributors to reef structure and facilitate coral recruitment, were most abundant at AR. However, relatively low coral cover and fish densitie s indicate suboptimal conditions at AR for continued reef accretion. Prevalence of large colonies of the Montastraea annularis complex combined with high recent mortality and high macroa lgal abundance indicates that BNP is experiencing a recent decline, with low poten tial for juvenile survival. However, the cause of stress is not apparent from co mmunity assessments and requires further investigation. Recent exposure to stress in corals at BNP in 2000 and throughout this study were evident by increased protein tur nover and oxidative and metabolic stress in response to a xenobiotic (Downs et al. 2005a, also see Chapter 6) Colonies at this site were generally large in size, with substantial contiguous area s of living tissue. Therefore if stressors can be identified and alleviate d, these colonies may survive. White Grunt ( Haemulon plumieri ) also experienced an endocrine-disrupting stress presumably in response to a pesticide at BNP during this study (Downs et al. 2006).

PAGE 89

Table 3.1. Benthic parameters meas ured and calculated at each site Site Information Depth (m) Community Structure Coral colony/meter (> 10 cm) Live coral cover (%) Mean colony diameter (all >10 cm) Mean colony height (all > 10 cm) Coral Colony Condition % old mortality % recent mortality Prevalence of disease (%) Prevalence of bleached or pale corals (%) Community dynamics/ Recruitment Coral recruits (<2 cm) (#/m2) Colony size distributions (max diameter) Algae/Herbivory % crus tose coralline algae % calcareous macroalgae Macroalgae/Crustose coralline algae Fleshy macro height (cm) Calcareous macro height (cm) Fleshy macroalgae index Calcareous macroalgae index Macroalgae index Diadema (#/10m2) Herbivorous damselfish density Planktivorous damselfish density

PAGE 90

Table 3.2. Coral species assessed by A tlantic and Gulf Rapid Reef Assessment Acropora cervicornis AC Meandrina meandrites MEAN Acropora palmata AP Millepora complanata MILC Agaricia agaricites AGA Montastraea annularis MILA Agaricia lamarcki AGL Montastraea cavernosa MC Agaricia tenuifolia AGT Montastraea faveolata MAF Colpophyllia natans CN Montastraea franksi MFR Dendrogyra cylindrus DEN Porites astreoides PA Dichocoenia stokesii DIC Porites furcata PF Diploria clivosa DC Porites porites PP Diploria labyrinthiformis DL Siderastrea sidereal SS Diploria strigosa DS Solenastrea bournoni SB Madracis decacitis MAD Solenastrea Hyades SH Madracis mirabilis MM Stephanocoenia intersepta SI

PAGE 91

Table 3.3. Fish species included in AGRRA assessments and established length and mass relationships for Caribbean fishes (Marks & Klomp 2003). Fish biomass was calculated using the power function: W = a Lb, where W is the mass (grams), L is the length (cm), and a and b are parameters estim ated by linear regression of logarithmically transformed length-mass data. Scientific Name Common Name A b Pomacanthidae (Angelfishes) Holacanthus ciliaris Queen Angelfish 0.0337 2.9004 Holacanthus tricolor Rock Beauty 0.0428 2.8577 Pomacanthus arcuatus Gray Angelfish 0.0344 2.9680 P. paru French Angelfish 0.0203 3.1264 Centropyge argi Cherubfish 0.0601 2.6920 Stromateidae (Butterflyfishes) Chaetodon aculeatus Longsnout Butterflyfish 0.0220 3.1897 Chaetodon capistratus Foureye Butterflyfish 0.0220 3.1897 Chaetodon ocellatus Spotfin Butterflyfish 0.0318 2.9838 Chaetodon sedentarius Reef Butterflyfish 0.0252 3.0760 Chaetodon striatus Banded Butterflyfish 0.0222 3.1395 Haemulidae (Grunts ) Anisotremus surinamensis Black Margate 0.0059 3.3916 Anisotremus virginicus Porkfish 0.0148 3.1674 Haemulon album White Margate 0.0167 3.0423 Haemulon aurolineatum Tomtate 0.0100 3.2077 Haemulon carbonarium Caesar Grunt 0.0147 3.0559 Haemulon chrysargyreum Smallmouth Grunt 0.3971 2.1567 Haemulon flavolineatum French Grunt 0.0127 3.1581 Haemulon macrostomum Spanish Grunt 0.0244 3.0295 Haemulon parra Sailors choice 0.0199 2.9932 Haemulon plumieri White Grunt 0.0121 3.1612 Haemulon sciurus Bluestriped Grunt 0.0194 2.9996 Scaridae (Parrotfishes) Scarus coelestinus Midnight Parrotfish 0.0153 3.0618 Scarus coeruleus Blue Parrotfish 0.0124 3.1109 Scarus croicensis Striped Parrotfish 0.0147 3.0548 Scarus guacamaia Rainbow Parrotfish 0.0155 3.0626 Scarus taeniopterus Princess Parrotfish 0.0335 2.7086 Scarus vetula Queen Parrotfish 0.0250 2.9214 Sparisoma atomarium Greenblotch Parrotfish 0.0121 3.0275 Sparisoma aurofrenatum Redband Parrotfish 0.0046 3.4291 Sparisoma chrysopterum Redtail Parrotfish 0.0099 3.1708 Sparisoma rubripinne Redfin Parrotfish 0.0156 3.0641 Sparisoma viride Stoplight Parrotfish 0.0250 2.9214

PAGE 92

Table 3.3 (cont.). Fish species included in AGRRA assessments and established length and mass relationships for Caribbean fishes (Marks & Klomp 2003). Fish biomass was calculated using the power function: W = a Lb, where W is the mass (grams), L is the length (cm), and a and b are parameters estim ated by linear regression of logarithmically transformed length-mass data. Serranidae (Groupers) Epinephelus adscensionis Rock Hind 0.0111 3.1124 Epinephelus cruentatus Graysby 0.0135 3.0439 Epinephelus fulvus Red Grouper 0.0175 3.0000 Epinephelus guttatus Red Hind 0.0111 3.1124 Epinephelus striatus Nassau Grouper 0.0065 3.2292 Mycteroperca bonaci Black Grouper 0.0068 3.2051 Mycteroperca interstitialis Yellowmouth Grouper 0.0068 3.2051 Mycteroperca tigris Tiger Grouper 0.0094 3.1200 Mycteroperca venenosa Yellowfin Grouper 0.0069 3.1400 Lutjanidae (Snappers) Lutjanus analis Mutton Snapper 0.0162 3.0112 Lutjanus apodus Schoolmaster 0.0194 2.9779 Lutjanus cyanopterus Cubera Snapper 0.0151 3.0601 Lutjanus griseus Gray Snapper 0.0232 2.8809 Lutjanus jocu Dog Snapper 0.0308 2.8574 Lutjanus mahogoni Mahogany Snapper 0.0429 2.7190 Lutjanus synagris Lane Snapper 0.0295 2.8146 Ocyurus chrysurus Yellowtail Snapper 0.0405 2.7180 Acanthuridae (Surgeonfishes) Acanthurus bahianus Ocean Surgeonfis h 0.0237 2.9752 Acanthurus chirurgus Doctorfish 0.0040 3.5328 Acanthurus coeruleus Blue Tang 0.0415 2.8346 Balistidae/Monocanthida e (Leatherjackets) Aluterus scriptus Scrawled Filefish 0.8230 1.8136 Balistes vetula Queen Triggerfish 0.0267 2.9903 Cantherhines macroceros Whitespotted Filefish 0.0562 2.6534 Cantherhines pullus Orangespotted Filefish 0.0684 2.5632 Canthidermis sufflamen Ocean Triggerfish 0.0176 3.0554 Melichthys niger Black Durgon 0.0562 2.6534 Xanthichthys ringens Sargassum Triggerfish 0.0562 2.6534 Other fishes Bodianus rufus Spanish Hogfish 0.0144 3.0532 Caranx ruber Bar Jack 0.0074 3.2370 Lachnolaimus maximus Hogfish 0.0203 2.9880 Microspathodon chrysurus Yellowtail Damselfish 0.0239 3.0825 Sphyraena barracuda Great Barracuda 0.0050 3.0825 Kyphosus secatator Bermuda Chub 0.0174 3.0800

PAGE 93

Table 3.4. Identification of key cora l species that discriminated among sites; -represents species that did not contribute to 90% of dissimilarity between specific sites. Bold values represent species primarily re sponsible for differences between sites. Abbreviations of coral species are as shown in Table 3.2. AR/ KL 6 m WB/ BNP AR/ BNP WB/ KL 6 m KL 6m/ BNP AR/ WB Mean Dissimilarity 78 72727169 68 PA 22 10 18 97 15 PP 15 99 1316 7 SS 6 9899 8 SI 6 8779 5 DIC 4 6265 5 MA 4 11 11413 3 MC 3 525-5 AC 3 4432 4 DC 2 --------MILA 2 4--4-4 MAF 2 --2---2 MEAN --------2 CN -------2 -AGA -----3---

PAGE 94

Table 3.5. Comparison of benthic parameters ; values represent mean ( SE) at four 6 m patch reefs from 10 m transects using Atlantic and Gulf Rapid Reef Assessment protocol. Means not connected by the same letter di ffered significantly (p < 0.05). Site n Colonies Live Coral Coral Coral Recent Old Herbivorous Bicolor (#/m) Cover Height Diameter Mort ality Mortalit y Damselfish Damselfish (%) (cm) (cm) (%) (%) (#/m) (#/m) KL 6 m 7 0.64 7 A 12 A 21 A 3 A 11 A 0.06 A 0.04 A (0.07) (1) (3) (5) (2) (4) (0.04) (0.04) WB 8 0.71 9 A 13 A 19 AB 2 A 8 A 0.05 A 0.01 A (0.07) (1) (1) (2) (1) (2) (0.04) (0.01) AR 8 0.88 16 B 23 B 40 C 2 A 12 A 0.07 A 0.00 A (0.10) (1) (4) (4) (1) (3) (0.04) (0.00) BNP 8 0.64 8 A 25 AB 37 BC 4 A 19 A 0.03 A 0.03 A (0.05) (1) (6) (6) (2) (8) (0.02) (0.03)

PAGE 95

Table 3.6. Pairwise comparison of 6 m s ites based on ANOSIM of fish composition; significant R values are in bold (significance level = 0.1%) WB AR BNP KL 6 m 0.47 0.37 0.07 WB 0.21 0.48 AR 0.45

PAGE 96

Table 3.7. Kruskal-Wallis ( 2) and Wilcoxon pairwise comparison of 6 m sites base d on biomass of fish families; n.s., not significant (p > 0.05). No significant diffe rences were found for the groups Acanthuridae, Balistidae, Pomacanthidae, Serranidae, Stromate idae and other. Kruskal-Wallis Wilcoxon pairwise comparison 6 m Sites KL 6 m/WB KL 6 m/AR KL 6 m/BNP WB/AR WB/BNP AR/BNP Haemulidae 23.4 12.4 13.4 5.8 n.s. 9.2 4.2 Lutjanidae 16.4 13.3 10.4 5.1 n.s. n.s. n.s. Scaridae 11.2 n.s. n.s n.s. 7.9 10.0 Herbivores 11.5 n.s. n.s. n.s. 8.3 9.6 n.s. Carnivores 24.1 11.4 12.4 9.8 n.s. 8.7 6.6 Table 3.8. Kruskal-Wallis ( 2) and Wilcoxon pairwise comparison of 6 m sites based on densities of fi sh families; n.s., not significant (p > 0.05). No significant differences (p > 0.05) were found for the groups Acanthuridae, Balistid ae, Pomacanthidae, Scaridae, Serranidae, Stromateidae and other. Kruskal-Wallis Wilcoxon pairwise comparison 6 m Sites KL 6 m/WB KL 6 m/AR KL 6 m/BNP WB/AR WB/BNP AR/BNP Haemulidae 20.7 13.8 5.9 n.s. 7.9 9.3 n.s. Lutjanidae 16.0 13.3 10.4 4.6 n.s. 4.2 n.s. Herbivores 10.0 6.3 n.s. n.s. 5.3 n.s. n.s. Carnivores 23.0 15.3 9.4 n.s. 5.4 9.8 4.0

PAGE 97

Table 3.9. Characterization of func tional algal groups and density of coral recruits; valu es represent mean ( SE). *Algal ind ex = % relative abundance of macroalgae x canopy height. Means not connected by the same letter differed significantly (p < 0.05). Site Name # Quadrats % Coralline algae % Fleshy algae % Calcareous algae Fleshy height (cm) Calcareous Height (cm) Fleshy algal index* Calcareous algal Index* Macro: Crustose # Coral Recruits/ m2 KL 6 m 35 9 A (3) 42 A (5) 12 A (3) 4.9 A (0.4) 2.3 A (0.2) 210 A (30) 33 A (10) 28 C (5) 4 A (1) WB 40 12 BC (3) 24 A (3) 14 A (3) 2.4 BC (0.3) 2.6 A (0.3) 79 B (12) 44 A (9) 12 B (3) 7 A (2) AR 40 16 C (3) 20 B (4) 13 A (3) 1.0 C (0.2) 2.2 A (0.2) 40 C (9) 29 A (5) 6 A (2) 4 A (1) BNP 40 10 B (3) 31 A (3) 12 A (2) 2.9 B (0.3) 3.2 B (0.3) 91 B (12) 46 A (9) 16 B (4) 4 A (1)

PAGE 98

Table 3.10. Comparisons of study sites with AGRRA regional baselines (modified from Kramer et al. 2003) based on indicators selected for biotic reef health index for corals >25 cm maximum diameter AGRRA Parameter Regional mean Regional (best value): Regional (worst value): KL 6 m WB AR BNP optimal suboptimal (this study) Live Coral Cover (%) 26 56 3 7 9 16 8 Large (>25 cm) Coral Density 9 18 4 1 1 4 3 Small Coral (<2 m) Density 4 15 2 4 7 5 5 Max Diameter of MA complex 71 115 49 0 0 128 89 Mean Recent Mortality (%) 4 1 18 8 2 1 5 Mean Old Mortality (%) 22 8 31 17 9 15 29 Diseased corals (%) 5 0 18 0 0 0 0 Macroalgal index 82 12 215 194 116 45 129 % Crustose coralline algae 29 42 11 10 12 16 10 Diadema Density 2 23 0 0 0 0 0 Total Fish Density 49 123 21 39 60 48 28 Herbivorous Fish Density 31 54 15 29 9 22 17 Carnivorous Fish Density 6 26 0.4 0.3 4.1 5.8 1.2

PAGE 99

Table 3.11. SIMPER results of Biot ic Reef Index; represents pa rameters that did not contribute to 90% of dissimilarity among sites. Large and small coral density, diseased corals, Diadema density and carnivorous fish density did not contribute to differences among sites. KL 6 m/WB KL 6 m/AR KL 6 m/BNP WB/AR WB/BNP AR/BNP Average Dissimilarity 27 53 31 46 31 29 Maximum Diameter Montastraea annularis complex 21 14 24 16 6 Macroalgal Index 14 24 10 13 2 13 Total Fish Density 4 1 2 6 3 Old Mortality 2 2 4 2 Herbivorous Fish Density 4 2 2 1 Live Coral Cover 1 1 Recent Mortality 1 % Crustose Coralline 1

PAGE 100

Table 3.12. SIMPER results of Biotic Reef Index showing dissimilarities among study sites and Caribbean means; represents para meters that did not contribute to 90% of dissimilarity among sites. Small coral de nsity, recent mortality, diseased corals, Diadema density and carnivorous fish density did not contribute to differences among sites. KL 6 m/Mean WB/Mean AR/Mean BNP/Mean Average Dissimilarity 40 36 23 25 Macroalgal Index 17 6 6 7 Max Diameter Montastraea annularis complex 11 12 9 3 Live Coral Cover 3 3 2 3 % Crustose Coralline 3 3 2 3 Total Fish Density 2 2 3 Herbivorous Fish Density 4 1 2 Large Coral Density 1 1 1 1 Old Mortality 2 1 1

PAGE 101

Table 3.13. SIMPER results of Biotic Reef Index showing dissimilarities among study sites and Caribbean best values ; represents parameters that did not contri bute to 90% of dissimilarity among sites. Small coral densit y, recent mortality and diseased corals did not contribute to differences among sites. KL 6 m/Best WB/Best AR/Best BNP/Best Average Dissimilarity 71 67 37 52 Macroalgal Index 22 14 4 14 Total Fish Density 10 8 9 11 Max Diameter Montastraea annularis complex 14 15 3 Live Coral Cover 6 6 5 6 Herbivorous Fish Density 3 6 4 4 % Crustose Coralline 4 4 3 4 Diadema Density 3 3 3 3 Carnivorous Fish Density 3 2 3 Old Mortality 3 2 Large Coral Density 2

PAGE 102

Table 3.14. SIMPER results of Biotic Reef Index showing dissimilarities among study sites and Caribbean worst values; represents parameters that did not contribute to 90% of dissimilarity among sites. Large and small coral density, % crustose coralline, Diadema density and carnivorous fish density did not contribute to differences among sites. KL 6 m/Worst WB/Worst AR/Worst BNP/Worst Average Dissimilarity 22 42 52 29 Macroalgal Index 3 17 25 12 Max Diameter Montastraea annularis complex 7 8 12 6 Total Fish Density 3 7 4 1 Diseased Corals 3 3 3 3 Recent Mortality 1 3 3 2 Old Mortality 1 3 Live Coral Cover 1 2 Herbivorous Fish Density 2

PAGE 103

Table 3.15. SIMPER results of Biotic Reef Index showing dissimilarities among Caribbean mean, best and worst values; repres ents parameters that did not contribute to 90% of dissimilarity among regional values. Large coral density and diseased corals did not contribute to differences among regional values. Mean/Worst Mean/Best Worst/Best Average Dissimilarity 42 39 71 Macroalgal Index 19 8 23 Total Fish Density 4 9 12 Max Diameter Montastraea annularis complex 3 5 8 Live Coral Cover 3 4 6 Herbivorous Fish Density 2 3 4 % Crustose Coralline 3 2 4 Old Mortality 2 3 Diadema Density 2 3 Carnivorous Fish Density 2 3 Recent Mortality 2 Small Coral Density 1

PAGE 104

SI 9% 9%PP 57%SS MA CN DC DL MCMILAPA AC 5% 4% 7% PP 14% SI 9% SS 18%PA 20%PA 20% 9% AGA 5% MEAN 4% MILA 5% PA 12% PP 20% SI 16% SS 14% MA 30% DIC AC MAF PA 56% PP 10% SI 4% SS 7% AC DC DIC MAF MC MFR MILA MILC MA DS PP 23% PA 26% SS 12% SI 9% MA 10%AGA 1% AC 4%MC 3% MAF 1% MILA 2% MILC 1%DIC 5% CN DC DL DS MEAN 1% <1% 1% <1% <1% MFR <1%All 6 m Sites KL 6 m AR WB BNP7% DICFigure 3.1. Relative abundance of coral species >10 cm maximum diameter. DIC MA MC

PAGE 105

ab 0 10 20 30 40 50 60 70% Mortality Old Recentn=108n=32n=26n=10 n=20Stephanocoenia spp Dichocoenia sp Siderastrea spp Montastraea spp Porites spp .A B C C AB ab a b b Figure 3.2. Mean ( SE) recent and old mortality for dominant coral s pecies at the four study sites. Means not connected by the same letter (capital and lowercase for old and recent mortality, respectively) were not significan tly different by Wilcoxons test (p < 0.05).

PAGE 106

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9FrequencyPartial Mortality (%) Maximum Colony Diameter (cm) 10-2930-4950-6970-8990-109110-129>130 KL 6 m WB AR BNP 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9Figure 3.3. Frequency of coral colonies by size cla ss (maximum diameter) and the relationship of colony size with recent and old mortality. Solid line represents old mortality (p) and the dashed line represents recent mortality (x).

PAGE 107

292 1680 0 00 BNP AR WB KL 6 mAcanthuridae Scaridae Haemulidae Lutjanidae Serranidae Balistidae Other0 1 2 3 5 7 4 6Biomass (g x 10 3 /100 m 2 )0 1 2 3 5 7 4 6 0 1 2 3 5 7 4 6 0 1 2 3 5 7 4 6Pomacanthidae StromateidaeFigure 3.4. Fish biomass (g x 10 3 /100 m 2 ) by family. 74 17 12 136 84 603 4673 588 0 51 000 000 113 1839 2458 862 16 137 111 1817 986 144 30 102 0 0 607

PAGE 108

WB45 40 35 30 25 20 15 10 5 0 AR45 40 35 30 25 20 15 10 5 0 Herbivore CarnivoreKL 6 m 45 40 35 30 25 20 15 10 5 0 BNP0-56-1011-2021-3031-40>40 Size (cm) 45 40 35 30 25 20 15 10 5 0 Density (individuals/100 m 2 )Figure 3.5. Density (individuals/100m 2 ) of herbivorous and carnivorous fish by size class (cm)

PAGE 109

Figure 3.6. Relative abundance of functional algal groups Coralline Calcareous Fleshy KL 6 mWBARBNP 100% 80% 60% 40% 20% 0%

PAGE 110

Figure 3.7. Cluster analysis based on Atlantic Gulf and Rapid Reef Assessment (AGRRA) biotic health indicators comparing study sites to regional AGRRA values for Caribbean reefs >5 m. KL 6 m Regional Worst WB BNP AR Regional Best Regional Mean100 80 60 40

PAGE 111

4. Symbiont-bearing Foraminifera as Indicators of Reef Health 4.1. Abstract Symbiont-bearing (larger) benthic Foraminifera (LBF) assemblages were examined at four 6 m deep patch reefs w ithin Biscayne National Park (BNP) and the upper Florida Keys National Marine Sanc tuary (FKNMS), and along a 3 18 m depth transect in FKNMS between August 2001 and February 2003. Populations of Amphistegina gibbosa the dominant LBF, were assess ed based on densities, size distributions, prevalence and severity of bleaching, and shell damage. These criteria were used in conjunction with physical and e nvironmental data to assess the suitability of these reef sites to support growth and reproduc tion of calcifying organisms that host algal endosymbionts (i.e., reef health). Densities of A. gibbosa and other LBF were typically higher at Algae Reef (AR), a reef adjacent to an intact mangrove shoreline, than at Key Largo 6 m, which is closer to developed shoreline. Biscayne National Pa rk had the lowest de nsities of all LBF, suggesting that water quality there was genera lly unsuitable for survival to maturation. Concurrent studies of lesion reco very on colonies of the coral Montastraea annularis species complex showed the same ranking of 6 m sites as LBF abundances, with both lesion recovery rates and LBF abundances high est at AR and lowest at BNP. Bleaching and breakage of A. gibbosa indicated chronic stress at all sites, with no evidence for acute photic stress during the study period. Simila rly, no coral bleaching was observed during the study. Evaluation of LBF populations can provide manage rs with a relatively quick, low-cost method for determining presence and re lative intensity of stressors influencing calcifying organisms that host algal symbionts.

PAGE 112

4.2. Introduction 4.2.1. Larger Benthic Foraminifera as Indicators of Reef Condition The biogenic constituents in reef sedi ments reflect the makeup of the benthic community and thereby characteri ze coral reef health. For example, reef sediments that favor calcifying organisms dependent on algal endosymbionts (i.e., mixotrophic organisms) over autotrophs (e.g., macroalgae) and heterotrophs (e.g., mollusks, sponges) are typically dominated by she lls of larger benthic forami nifers (LBF) and physically degraded coral debris (e.g., Hallock 1988, 2005). Benthic foraminifers that host algal endosymbionts are useful bioindicators fo r reef studies because they 1) have physiological analogies with zooxanthellae corals and ther efore similar water-quality requirements, 2) are abundant in healthy reef ecosystems and collected with minimal effort and effect on reef resources, and 3) ha ve relatively short life spans and therefore comparably rapid responses to environmen tal stressors (Cockey et al. 1996, Hallock 2000, Hallock et al. 2003). Responses of populations of Amphistegina can indicate the presence and intensity of photo-oxidative stress, as well as general water quality suitability on times scales of weeks to m onths (Hallock et al. 2006a,b). My project provided the opportunity to assess LBF assemb lages, including detailed assessments of populations of Amphistegina at the same time coral assemblages and health of individual coral colonies were being assessed. This provided another line of evidence about coral reef health and also the opportunity to de termine which LBF parameters had responses similar to coral parameters. 4.2.2. Rationale for Assessing Populations of Amphistegina Amphistigina is the dominant, algal symbiont-bearing foraminiferal genus found on reefs worldwide (Langer & Hottinger 1997, Hallock 1999), commonly living on coralline and filamentous alga e on reef substrate, as well as on some macrophytes. The two most similar species, A. gibbosa in the Caribbean and A. lessonii in the Indo Pacific, can be found from the shallowest subtidal zones to >100 m depth, and tend to be most abundant between 15 m and 40 m depth (Hallock 1999). Th e distribution of Amphistegina appears to be controlled re gionally by temperature (~12-33o C) and locally by hydrodynamics, water quality, light and substrate (Hallock 1999). Stress symptoms

PAGE 113

similar to those reported for corals and cora l-reef communities (e.g., bleaching, predation, and algal infestation) have been observed in populations of Amphistegina (Hallock et al. 2006b). Amphistegina are sensitive to water quality and bleach in response to excess radiant energy (Hallock et al. 2006b). These pr otists respond to environmental conditions within days to weeks and provide a low-cost method to quickly distinguish between local environmental conditions (e.g., water quality ) and photo-oxidative stress at my study sites. Amphistegina host diatom endosymbionts in a relationship analogous to coralzooxanthellae symbioses (e.g., L ee & Anderson, 1991). Therefore, light is necessary for growth and calcification. However, high intensit ies of solar radiation, particularly shorter wavelengths (blue, violet and ultraviolet: 290-490 nm; 0.10 W m-2), can induce photoinhibition (Muller 1978, Lee et al. 1980, Williams & Hallock 2004), loss of symbionts (Hallock et al. 1986b, 1995, Talge & Hallock 2003, Williams & Hallock 2004) and suppressed growth rates (Williams & Hallo ck 2004). Bleaching has been attributed to high levels of solar radiation, particular ly shorter, higher energy wavelengths (UV-B, 290 320 nm) in corals (Lesser et al 1990, Gleason & Welli ngton 1993, Glynn 1993, Jones et al. 1998, Lesser & Fa rrell 2004) and other symbiont -bearing marine organisms (Jokiel 1980, Williams & Hallock 2004, Hallock et al. 2006b). Bleaching was first documented in Amphistegina in laboratory experiments (Hallock et al. 1986b) and in field populations on the Florid a reef tract beginning in su mmer 1991 (see Hallock et al. 2006a, b and references therein). The abrupt onset of bleaching in A. gibbosa in late June 1991 was postulated to be associated with stratospheric ozone deple tion following eruptions of Mt. Pinatubo in May-June 1991 (Hallock et al. 1992, 1995). Symbiont color loss is caused by degradation and digestion of the diatom endosymbionts, and partial bleaching induced by photoinhibition in laboratory experiments was id entical to that seen in field-collected specimens (Talge & Hallock 1995, 2003). Populations of Amphistegina in the Florida Keys were monitored between 1991 and 1999 (Hallock et al. 1995, Williams et al. 1997, Hallock et al. 2006b) for changes in size di stributions and condition (i.e., visible [with stereo microscope] color changes and shell damage) providing information (Table 4.1) on these protists for comparison with this study.

PAGE 114

While thermal stress has been linked to coral bleaching (Marshall & Schuttenberg 2006), bleaching in Amphistegina is clearly not influenced by temperatures currently experienced along the Florida r eef tract (Hallock et al. 200 6b). Salinity also was ruled out as a potential cause of bleaching in A. gibbosa (Williams 2002). Bleaching incidence and intensity in these foraminifers typically increases in March, when water temperatures are near their lowest, peaking in July following the summer solstice, and declining in late summer when water temperatures are warmes t (Hallock et al. 1995, Williams et al. 1997, Williams 2002). In culture, exposure to 32o C at either 6-8 or 13-15 mol photon m-2 s-1 of photosynthetically active radiation (PAR) 12 hr d-1 for 4 weeks induced significant symbiont loss (Talge & Hallock 2003). Howeve r, in the same experiments, 4 weeks at 25o C and 13-15 mol photon m-2 s-1 PAR induced twice as much symbiont loss as exposure to 32o C at 6-8 mol photon m-2 s-1 PAR for the same duration. The trials at 25o C and 6-8 mol photon m-2 s-1 PAR induced no symbiont loss. Sea surface PAR intensities in summer are appr oximately 1200-1500 umol photon m-2 s-1, so Amphistegina normally avoid photic stress by phototaxic beha vior (Lee et al. 1980). Hallock (2001) and Williams (2002) concluded that bleaching in A. gibbosa was linked to solar radiation, based on laboratory experiments (Hallock et al. 1986b, Talge & Hallock 2003, Williams & Hallock 2004), timing of onset, seasonal and latitudinal tre nds in bleaching in field populations (Hallock et al. 1995, Williams et al. 1997), and significantly higher bleaching prevalence at field sites with significantly higher water tr ansparency. Coral bleaching events usually coincide with periods of unus ually warm sea surface temperatures (SST) and calm winds, resulting in increased pene tration of solar radiation (Gleason & Wellington 1993, Glynn 1996, Wilkin son 1998). Bleaching in A. gibbosa indicates that corals are likely being exposed to photoinhib itory stress and, along with SST anomalies, can predict susceptibility of corals to bleaching and disease (Hallock et al. 2006b). Bleaching in A. gibbosa is typically progressive a nd degenerative; severity of bleaching typically increases with increasing size of individual foraminifers (Talge & Hallock 1995, Williams et al. 1997). Prior to the onset of stress in 1991, all adult individuals observed were normally colore d. Since then in affected populations, individuals smaller than 0.5 mm in diameter seldom exhibited symbiont loss whereas specimens larger than 1.0 mm were seldom nor mally colored. Therefore, for this study

PAGE 115

we only evaluated bl eaching in adult A. gibbosa (> 0.6 mm) as recommended by Williams (2002) and Hallock et al. (2006a, b). Bleaching also affects reproductive success and recruitment of A. gibbosa (Hallock et al. 1995, Williams 2002), and th erefore population densities and their seasonality. Detailed studies, especially Williams (2002) and others summarized by Hallock et al. (2006b), demonstrated normal seasonality in populatio n densities, size distributions, and other parameters, and cha nges in those parameters with increased prevalence and intensity of bl eaching (Table 4.1). Medium to high population densities indicate environments suitable to support grow th and reproduction of these protists over recent weeks to months. Size-frequency distributions of individuals making up the populations require context-dependent inte rpretations, based on season and population densities. For example, high population densiti es with high proportions of juveniles in summer months indicate favorable environm ental conditions, while very low population densities dominated by juveniles indicate envi ronmental conditions suited temporarily for survival of juveniles carried into the envi ronment by currents, but unsuitable for their growth and survival to reproduction. Sim ilarly, high population dens ities and infrequent bleaching indicates that water quality is suita ble and photo-oxidative stress is limited. High population densities and hi gh incidences of bleach ing indicate acute photooxidative stress that occurred after reproducti on peaked. Chronic stress from bleaching also is associated with increased shell breakage (Toler & Hallock 1998, Toler 2002), susceptibility to predation (Hallock & Talg e 1994) and endolithic infestation (Hallock 2000). Therefore, damage assessment indicates whether chronic stress is affecting the population (Table 4.1). 4.2.3. Other Larger Benthic Foraminifers Along a Molasses Reef-Rodriguez Key tran sect (upper Florida reef tract), foraminiferal tests shifted from those domin ated by LBF in 1959-61 (Lidz & Rose 1989), to those dominated by smaller herbivorous an d detrivorous taxa in 1991 (Cockey et al. 1996). This assemblage shift is consistent with ecological/sedimentological models of community response to incr eased nutrient flux (e.g., Hallock 1988, Hallock 2001). I

PAGE 116

assessed abundances of live LBF to determin e environmental suitability for supporting these mixotrophic calcifying organisms. 4.2.4. Study Goals The purpose of my study was to use a suite of parameters (Table 4.1) to determine if (1) water quality and other environmenta l conditions were suitable for calcifying, symbiont-bearing organisms based on densities of live symbiont-bearing foraminifers and (2) foraminifers indicated exposure to chro nic or acute photic st ress, expressed as bleaching or shell damage in A. gibbosa My study provided a unique opportunity to monitor LBF populations at sites where mo rphological and physiol ogical conditions of scleractinian corals and other reef organisms also were be ing monitored (Downs et al. 2005a, 2006, Fisher et al. in pres s, Ch. 6, Fisher et al. in pr ep), an essential step for refining protocols for using foraminifers as bioindicators in reef monitoring and risk assessment. 4.3 Methods 4.3.1. Study Sites I collected foraminifers at one patch reef in Biscayne National Park (BNP) and four patch reefs plus two depths on one fo rereef in the upper Florida Keys National Marine Sanctuary (FKNMS). These seven si tes comprised both a latitudinal transect with four sites at 6 m depth and a depth tran sect from Key Largo (KL) 3 m to KL 18 m (Fig. 1.1). The 6 m sites represented a sp ectrum of possible anth ropogenic influence based on distance from urbanized coastal development. Locations of all sites 6 m were suitable for LBF with respect to salinity and hydrodynamics. The Key Largo (KL) sites were located offshore from the most urbanized coastline of Key Largo, where natural ve getation was removed (see Fig. 2.1), natural topography altered to maximize waterfront prop erties and the coastline was armored with seawalls. These sites also were positioned along the route taken by recreational boaters and commercial dive operators to reach Molass es Reef and other heav ily used outer reefs, so pollutants such as hydr ocarbon combustion products may be more prevalent. In

PAGE 117

addition, larger boats regularly stirred up se diments, potentially remobilizing nutrients and chemical pollutants (K ruczynski & McManus 2002). Algae Reef (AR) also was offshore from Key Largo but situated mid-way between the Key Largo and the BNP sites, ad jacent to John Pennekamp Coral Reef State Park. Most of the extensive natural coastlin e was native, intact and relatively vegetated with coastal hammock mangroves and seagrass beds (see Fig. 2.1). White Banks (WB) was close to the KL sites, but still adjacent to the state park (Fig. 1.1). Alinas Reef (BNP) in Biscayne National Park was closest to the urban Miami area, including a large landfill, a nuclear power plant, and watershed ca nals that drain into Biscayne Bay. This site also may be influenced by extensive agricultural area south and west of Miami and associated nutrients and chemicals that enter watershed canals. However, the patch reefs of Biscayne National Park are somewhat protected from anthropogenic influence by distance, including Biscayne Bay and by uninhabited barrier islands. For example, Carnahan et al. (in press) reported that heavy metal concentrations in Biscayne Bay sediments decline with distance from urban Miami, the landfill and agricultural areas to the south. 4.3.2. Sampling and Assessment of Sy mbiont-bearing Foraminifers Between August 2001 and February 2003, three (August and October 2001) or five (March 2002 February 2003) reef rubbl e samples were collected quarterly except sampling along the depth gradient began in October 2001. Met hods of sampling LBF and populations of Amphistegina assessment are described in detail elsewhere (Hallock et al. 1995, Williams et al. 1997, Hallock et al. 200 6a). Reef rubble, although not the only habitat available to LBF, is easily compared among study site s. Suitable rubble for LBF was readily available at all si tes except KL 3 m, where rubb le was both hard to find and often partially buried in fine sediments. Rubble was placed in plastic bags, which were brought to the surface and placed in a shaded bucket for transport to shore. Onshore, rubble was scrubbed using a small brush to detach associated sediment, algae and meiofauna, including foraminifers. I traced the outline of the rubble to determine its approximate area of bottom cover. The surface area of the rubble outline was determined using image analysis software to permit

PAGE 118

density estimates. I distributed resultant re sidue among 150 mm diamet er petri dishes and placed them in a culture chamber maintain ed between 23 and 28 C (depending on the time of year) on a 12-hour light/dark schedule at ~10 mol photon m-2 s-1 PAR. I removed all LBF whose behavior or color in dicated a high probability of being alive (Hallock et al. 1986a) and placed them in a 100 mm diameter petri dish of seawater. I identified to species and counted each sp ecimen verified to be alive based on pseudopodial activity. Maximum diameter of A. gibbosa was measured to the nearest 0.05 mm, and each specimen was characterized a ccording to symbiont color and presence of damage, either from breakage, predation, endolithic infestation or deformation (Fig. 4.1; methods similar to Hallock et al. 1995). Individual Amphistegina were visually characterized as unbleached (uniform br own color), partly bleached (specimens possessing white patches with at least half retaining some color), or bleached (specimens with <50% brown color remaining). If a sample contained larger numbers of individuals (> 150), all were counted, but a subsample of 150 200 individuals was haphazardly selected, measured and characte rized. Other LBF (see list in Table 4.2) encountered live in each sample were counted but not measured or otherwise assessed. 4.3.3. Data Analysis Data were analyzed in two groups: (1) by sites of 6 m depth along the northeast southwest traverse (BNP, AR, WB and KL 6 m), and (2) by sites along the depth gradient: KL 3, 6, 9 and 18 m. The 6 m site was common to both gr oups (Fig. 1.1). I calculated mean density of all LBF, and, for A. gibbosa only, percent juveniles (individuals < 0.6 mm in diameter), mean (maxim um) diameter and percent of adults that exhibited bleaching (e.g., were partly bl eached or bleached; Williams, 2002). I also calculated the percentage of specimens that exhibited shell damage (i.e., chipped, broken, or deformed; Toler & Hallock 1997, Tole r 2002) in samples between March 2002 and February 2003. I used repeated-measures MANOVA to determine if significant differences existed in % juveniles and % shell damage among sites. I checked model assumptions (e.g., sphericity, homogeneity of variances, normality, and independence) using residual plots. Density and diameter data were log10 transformed to meet these assumptions. In

PAGE 119

cases where the sphericity assumption was not met, I applied a univariate (unadjusted epsilon) approach. To interpret differences detected by MANOVA, I used one-way ANOVA followed by the Tukey-Kramer Hones tly Significant Difference (HSD) method. I used ANOSIM2 (two-way analysis of similarities) to determine if LBF assemblages differed significantly among site s (averaged over the entire study period) and times (averaged across all sites). To do this, I first calculated Bray-Curtis similarity matrices for all log-transformed LBF densities. For each sampling period, I used ANOSIM (one-way analysis of similarities) to determine if sites differed significantly based on assemblages of all LBF. I used MDS (multi-dimensional scaling) to determine how sites clustered based on densities of all LBF followed by SIMPER (similarity percentages) analyses to determine whic h species were primar ily responsible for grouping of sites (Clarke and Warwick, 2001) I performed ANOSIM2, ANOSIM, MDS and SIMPER using PRIMER v.5 (Plymouth R outines in Multivariate Ecological Research PRIMER-E Ltd., Plymouth). All other statistical analyses used JMP v.3.2. (SAS Institute Inc., Cary, NC, USA), with = 0.05 for all hypothesis tests. 4.4. Results 4.4.1. 6 m Sites 4.4.1.1. Responses of Amphistegina Live densities of Amphistegina gibbosa ranged from 4.8 x 102 to 4.37 x 104 individuals m-2 with a mean ( SE hereinafter) of 1.2 x 104 ( 8.5 x 102) m-2 (n = 124). Densities differed significantly among the 6 m sites (repeated measures MANOVA: site effect F3,8 = 20.6, p < 0.0004) and sampling dates (time effect F6,3 = 38.1, p < 0.007; Fig. 4.2A), but there were no site x time interac tions. Algae Reef had significantly higher densities than KL 6 m and BNP, and WB al so had significantly hi gher densities than BNP (ANOVA: F3,120 = 16.5, p < 0.0001). Population densities were significantly higher in June and August 2002 than in August 2001 and February 2003 (ANOVA: F6 ,117 = 4.8, p = 0.0002). Percentages of juvenile A. gibbosa ranged from 25 to 92%, with a mean of 62 1 % (n = 122). The percentage of juveniles changed signi ficantly with time (repeated measures MANOVA: time effect F6,48 = 7.6, p < 0.0001; Fig. 4.3A) but not among sites

PAGE 120

or with site x time interactions. Percentage of juveniles was signifi cantly higher in June and August 2002 than in February 2003 (ANOVA: F6,115 = 11.4, p < 0.0001). Mean diameter of A. gibbosa ranged from 0.40 to 0.77 mm with an overall mean for all samples of 0.53 0.01 mm (n = 124). Mean diameter changed significantly with time (repeated measures MANOVA: time effect F6,48 = 11.2, p < 0.0001; Fig. 4.4A), but not among sites or with site x time interactio ns. Mean diameter was significantly lower in summer 2002 (June and August) than all ot her months and significantly highest in February 2003 (ANOVA: F6,117 = 14.7, p < 0.0001). Percentage of adults exhibiting any degr ee of bleaching ranged from 0 to 100 %, with an overall mean of 24 1 % (n = 124). Incidence of bleaching differed significantly among the 6 m sites (repeated measures MANOVA: F3,8 = 7.4, p < 0.02; Fig. 4.5A) but not with time or time x site interactions. Incidence of partial bleaching at AR was significantly higher than at KL 6 m (ANOVA: F3,80 = 3.3, p < 0.03). The percentage of the population with da maged tests ranged from 0 to 43 % with an overall mean of 17 1 % (n = 10 0). The percentage of damaged A. gibbosa differed significantly among sites and with time (re peated-measures MANOVA: site effect F3,16 = 7.0, p < 0.004; time effect F4,13 = 4.3, p < 0.02; Fig. 4.6A) but th eir interactions were not significant. Alinas Reef had a signi ficantly higher percentage of damaged A. gibbosa tests than WB (ANOVA: F3,96 = 4.2, p < 0.01). Percentage of test damage was significantly lower in Augus t 2002 than in November 2002 and February 2003 (ANOVA: F4,95 = 3.9, p < 0.006). 4.4.1.2. Other Symbiont-bearing Foraminifera Densities of LBF ranged from 6.7 x 102 to 9.97 x 104 individuals m-2 with a mean of 2.2 x 104 1.6 x 103 m-2 (n = 130; Fig. 4.7). Geog raphic location was the primary factor affecting LBF densities (averaged across all time periods; ANOSIM2: Global R = 0.28, p = 0.10%) with significant differences am ong all sites (Table 4.3). The highest dissimilarity was between WB and BNP and the lowest between WB and AR (Table 4.3). Densities of LBF were highest at AR and lo west at BNP (Fig. 4.7A, 4.8A-C). Densities of LBF differed significantly with time (a veraged across all 6 m sites; ANOSIM2: Global

PAGE 121

R = 0.23, p = 0.10%). The highest dissimilari ty was between October 2001 and June 2002 (Table 4.4). Similarity percentages analyses (SIMPER) revealed that Amphistigena gibbosa was the dominant LBF at all sites and therefor e the best indicator among 6 m sites (Table 4.5; Fig. 4.8). Archaias angulatus was the second most abundant and second most diagnostic, followed by Laevipeneroplis proteus, Heterostegina depressa, and Broekina orbitolitoides. Together with A. gibbosa these taxa were the dominant contributors to differences among 6 m sites (T able 4.5). Based on LBF, Algae Reef and BNP were the most dissimilar, while White Banks represented a midpoint between AR (dissimilarity 36.8%) and KL 6 m (dissimilarity 36.4%). A linas Reef was more dissimilar to all FKNMS sites than they were to each other. Densities of Archaias angulatus ranged from 0 to 3.10 x 104 individuals m-2 with an overall mean of 3.98 x 103 514 individuals m-2 (n = 124). Densities of Archaias angulatus differed significantly among site s (repeated measures MANOVA: F3,8 = 35.0, p < 0.0001); BNP had significantly lower densitie s than the other th ree sites (474 122 vs. 5150 640 individuals m-2, respectively; ANOVA: F3,120 = 14.7, p < 0.0001; Fig. 4.9A). Densities of Laevipeneropolis proteus ranged from 0 to 1.68 x 104 individuals m-2 at the 6 m sites with an overall mean of 2.53 x 103 264 individuals m-2 (n = 124). Densities of this LBF differed significantly among sites, seasons a nd their interactions (repeated measures MANOVA: site effect F3,8 = 4.1, p < 0.05; time effect F6,48 = 2.7, p < 0.03; site x time interaction F18,48 = 2.5, p < 0.006; Fig. 4.9B). Densities of L. proteus were significantly lower at BNP than the other sites in August 2001 (207 53 vs. 2139 437 individuals m-2, respectively; ANOVA: F3,8 = 17.6, p = 0.0007) and October 2001 (54 28 vs. 2826 882 individuals m-2, respectively; ANOVA: F3,8 = 10.9, p < 0.004; Fig. 4.9B). Densities of Heterostegina depressa ranged from 0 to 6.94 x 103 individuals m-2 with an overall mean of 645 88 individuals m-2 (n = 124). Densities of this shallowwater species differed significantly among site s and site x time interactions (repeated measures MANOVA: site effect F3,8 = 4.7, p < 0.04; site x time effect F18,9 = 3.6, p < 0.03; Fig. 4.9C). Densities of H. depressa were significantly higher at AR than BNP in

PAGE 122

August 2001 (1717 700 vs. 90 90 individuals m-2, respectively; ANOVA: F3,8 = 4.8, p < 0.04). Densities of H. depressa were significantly higher at AR and WB than KL 6 m and BNP in October 2001 (2189 498 and 164 30 vs. 26 26 and 0 0 individuals m-2, respectively; ANOVA: F3,8 = 21.4, p = 0.0004). Algae Reef had significantly higher densities of H. depressa than BNP in November 2002 (1755 507 vs. 105 53 individuals m-2, respectively; ANOVA: F3,16 = 4.1, p < 0.03; Fig. 4.9C). Densities of Broekina orbitolitoides ranged from 0 to 5.31 x 103 individuals m-2 with an overall mean of 452 86 individuals m-2 (n = 124). Densities of this LBF differed significantly different among sites, seasons and their interactions (repeated measures MANOVA: site effect F3,8 = 5.1, p < 0.03; time effect F6,3 = 220.0, p = 0.0005; site x time interaction F18,9 = 4.2, p < 0.02; Fig. 4.9D). In August 2001, densities of B. orbitolitoides were significantly higher at AR than at KL 6 m and BNP, and significantly higher at WB than at BNP (1465 410 vs. 83 42 and 0 0 individuals m-2, respectively; ANOVA: F3,8 = 14.6, p < 0.002). In Oc tober 2001, densities of B. orbitolitoides were significantly higher at AR and WB than at KL 6 m and BNP (3287 738 and 336 32 vs. 32 32 and 0 0 individuals m-2, respectively; ANOVA: F3,8 = 21.9, p = 0.0003; Fig. 4.9D). Densities of B. orbitolitoides dropped in June 2002 at all sites except BNP, where densities were consistently low (Fig. 4.9D). 4.4.2. Depth Gradient 4.4.2.1. Responses of Amphistegina Along the KL depth gradient, densities of live Amphistegina gibbosa ranged from 5.5 x 102 to 1.1 x 105 individuals m-2 with an overall mean of 1.6 x 104 1.7 x 103 individuals m-2 (n = 111). Densities of Amphistegina gibbosa and differed significantly among depths (repeated measures MANOVA: F3,8 = 12.9, p < 0.002; Fig. 4.2 B) but not with time or their interactions. Key La rgo 9 m and KL 18 m ha d significantly higher densities than KL 3 m and KL 6 m (ANOVA: F3,107 = 14.9, p < 0.0001). The percentage of juvenile A. gibbosa ranged from 9 to 100 % with an overall mean of 52 2 % (n = 108). Percentage juveniles di ffered significantly with depth and time (repeated measures MANOVA: site effect F3,7 = 29.2, p = 0.0002; time effect F5,3 = 10.6, p < 0.04; Fig. 4.3B). Key Largo 3 m and KL 6 m had higher percentages of

PAGE 123

juveniles than KL 9 m and KL 18 m (ANOVA: F3,104 = 11.1, p < 0.0001). Juveniles were more prevalent in June and August 2002 than in February 2003 (ANOVA: F5,102 = 4.2, p < 0.0002). Mean diameter of A. gibbosa ranged from 0.36 to 0.96 mm with an overall mean of 0.61 0.01 mm (n = 112). Mean diameter differed significantly with depth and time (repeated measures MANOVA: site effect: F3,8 = 61.5, p < 0.0001; time effect F5,4 = 13.4, p < 0.02; Fig. 4.4B) but not their interactions. Mean diameter of A. gibbosa was significantly larger at Key Largo 9 m a nd KL 18 m than at KL 3 m and KL 6 m (ANOVA: F3,108 = 24.8, p < 0.0001). Si gnificantly larger A. gibbosa were found in February 2003 than in June, August, and October 2002 (ANOVA: F5,106 = 5.1 p < 0.0004). The percentage of adult A. gibbosa that were partly bleached ranged from 0 to 73% with a mean of 28 2 % (n = 112). Percentage bleach ed differed significantly with depth (repeated measures MANOVA: F3,8 = 21.6, p < 0.0003; Fig. 4.5B) but not with time or their interactions. Key Largo 9 m a nd 18 m had significantly higher percentages of partly bleached adult A. gibbosa than at KL 3 m and KL 6 m (ANOVA: F3,108 = 13.2, p < 0.0001). The percentage of A. gibbosa with damaged tests ranged from 0 to 53 % with a mean of 16 1 % (n = 100). Percentage of A. gibbosa with damaged tests differed significantly with depth (repeated -measures MANOVA: site effect F3,16 = 10.9, p < 0.0005; Fig. 4.6B) but not with time or th eir interactions. Key Largo 3 m had a significantly lower percentage of damaged A. gibbosa tests than other depths (ANOVA: F3,96 = 9.3, p < 0.001). 4.4.2.2. Other Symbiont-bearing Foraminifera Densities of LBF ranged from 1.8 x 103 to 4.39 x 105 individuals m-2 along the depth gradient with a mean of 4.34 x 104 6.1 x 103 individuals m-2 (n = 125). Depth was the primary factor affecting LBF de nsities (averaged across all time periods; ANOSIM2: Global R = 0.53, p = 0.10%) with si gnificant differences among all depths (Table 4.6). The highest dissimilarity was between KL 3 m and KL 9 m and the lowest between KL 9 m and KL 18 m (Table 4.6). Dens ities of LBF were highest at KL 9 m and

PAGE 124

lowest at KL 3 m (Fig. 4.7B, 4.10). LBF de nsities differed signif icantly with time (averaged across depths; ANOSIM2: Global R = 0.18, p = 0.10%) with the highest dissimilary between October 2001 and June 2002 (Table 4.7). While Amphistegina gibbosa generally was the dominant LBF at all sites, SIMPER analyses revealed that A. gibbosa did not contribute highly to dissimilarities along the depth gradient (Table 4.8). Cyclobiculina compressus was the primary discriminating species along th e depth gradient, followed by Asterigerina carinata Heterostegina depressa and Broekina orbitolitoides (Table 4.8, Fig. 4.11 A-D). Based on LBF, the largest differences were between KL 3 m and the deeper sites: KL 9 m (51.2%) and KL 18 m (46.9%). The lo west dissimilarity (19.8%) was between the two deep sites, KL 9 m and KL 18 m. 4.5. Discussion Over the last thirty years, water quality has declined in the Florida Keys due to changes in water flow patterns from Florid a Bay, sedimentation (from boat traffic and development), and increased near-shore nutrien t concentrations (from local wastewaters, freshwater upwelling, fertilizer s, and industrial pollutants; Szmant and Forrester 1996, Lang et al. 1998, Causey et al. 2000, Porter et al. 2001, Andrews et al. 2005). Changes in water quality corresponded with significant decreases in live co ral cover throughout the Florida Keys (Dustan and Halas 1987, Lang et al. 1998, Causey et al. 2000, Beaver et al. 2005) and a shift from symbiont-bearing fora minifers to heterotrophic foraminifers (Cockey et al. 1996). In this study, I evaluated LBF densities and populations of Amphistigina gibbosa at seven coral reefs and related th is information to concurrent studies on environmental assessments, co mmunity condition and coral physiological condition. The pattern of LBF densities among my study sites is consistent with susceptibility to reduced water quality (e.g., Co ckey et al. 1996, Hallock et al. 2003). If conditions were optimum at all sites, LBF densities should exceed 104 individuals/m2, with highest densities (approaching 105) in summer months and lo west in winter or early spring (i.e., December through March; Halloc k et al. 1986a). Mean densities of both A. gibbosa and total LBF at my study sites were approximately 104 individuals/m2,

PAGE 125

indicating generally marginal to suitable c onditions. Densities of LBF were highest at AR, a site adjacent to John Pennekamp Coral R eef State Park, which is characterized by extensive mangrove and seagrass flats that sepa rate this site from urban development and provide a consistent source of colored di ssolved organic matter (CDOM; Ayoub et al. 2006), which preferentially absorbs higher ener gy wavelengths of solar radiation (e.g., Kirk 1996). In contrast, KL 6 m ha d significantly lower densities of A. gibbosa and the other LBF. This site is o ff a highly developed coastline and has more variable CDOM concentrations (Ayoub et al. 2006). White Banks, which is physically closer to the KL 6 m site than AR but still adjacent to state park waters, had intermediate densities of A. gibbosa and other LBF. Densities of A. gibbosa and all other symbiont-bearing foraminifers were lowest in BNP, offshore of Miami, FL, indicating that some aspect of water quality there was unsuitable. Seasonality would be expect ed along the depth transect with increasing densities of A. gibbosa with depth over the depth range studie d. Relative proportions of other taxa also would be expected to change with depth, since some species, e.g., Archaias angulatus and Laevipeneroplis proteus tend to be shallower dwelling than others, e.g., Cyclorbiculina compressa and Broeckina orbitolitoides (see Table 4.2). Along the depth gradient, densities of A. gibbosa and other LBF were higher at KL 9 m and KL 18 m than at the shallower sites (KL 3 m and KL 6 m), which was anti cipated because these protists prefer depths of 15 40 m (Hallock 1999). De nsities of LBF were higher at KL 9 m than at KL 18 m, suggesting better conditions at KL 9 m relative to KL 18 m. Key Largo 9 m and KL 18 m were more co mparable to sites monitored by Williams (2002): Conch (CR 10 m, CR 18 m) and Tennessee (TN 8 m, TN 20 m) Reef. Densities at KL 9 m and KL 18 m were lower than densities observed at CR and TN during periods of low stress (> 104 individuals m-2) but higher than densities at CR and TN following acute bleaching events (103 to 104 individuals m-2) in 1991 and 1998 (Williams 2002). No acute stress events occurred during my st udy (Ch. 2) but chronic stress presumably limited population densities as indicated by intermed iate densities and bleaching. Healthy Amphistigina populations typically reproduce by alteration of semelparous asexual and sexual generations (Harney et al. 1998). Asexual reproduction commonly occurs in the spring and each la rge individual can produce broods of

PAGE 126

approximately 100 to 300 offspring (Williams 2002). Thus, population densities can increase dramatically in summer. Se xual reproduction by gamete broadcasting commonly occurs in the fall. Under id eal conditions, percen tages of juvenile A. gibbosa should be highest (on the order of 40-60%) in early summer (May-July) and lowest, generally <40% in winter and early spring (December-March; Williams 2002). Similarly, mean individual diameters should be highest (approaching 1 mm) in late winter-early spring (February-March) a nd lowest (~0.6-0.7 mm) in mid summer. Increases in both population densities and pe rcentages of juvenile s at my study sites indicate successful reproduction. The percentage of juveniles at all sites was relatively high (> 40% all year). Williams (2002) typically observed 20 40% percentage juveniles during spring and winter months and between 40 60% during summer months in low-stress years at nearby Conch Reef, whic h is located immediatel y south of the area shown in Fig. 1.1. The high percentage of juveniles year-round at my study sites may indicate that zygotes or juveni les were consistently carried to my sites by currents, yet relatively few grew to adult sizes (> 0.6 mm in diameter). For example, Williams (2002) typically observed a mean diameter > 0.7 mm, whereas most A. gibbosa in this study were approximately 0.5 mm in diameter. In addition, summer abundances at BNP were comparable to the lowest densities at AR and WB, so conditions at BNP maybe unsuitable for survival of A. gibbosa throughout its life cycle. In the absence of photo-oxidative stress, no bleaching and minimal breakage (<10% of the A. gibbosa individuals) would be anticipat ed. With strong seasonal photooxidative stress, one expects frequent bleachi ng in early summer and lowest percentages in winter to early spring. Because breakag e is cumulative though the seasonal cycle of a bleaching-stressed population, highest percentages of shell damage are expected in the fall or winter (Toler 2002). Alinas Reef (BNP), which had the lowest LBF densities, also had the highest percentage of A. gibbosa with shell damage. Low-level bleaching stress increases susceptibility of these protists to predation and infestation (Toler 2002). Along the depth gradient, shell damage was highest at sites that also had higher percentages of bleached adults (KL 9 m an d KL 18 m). In the Florida Keys, shell breakage, breakage and repair, and in cidences of shape anomalies of A. gibbosa increased approximately 3-fold even in unbleached i ndividuals after 1992, following the onset of

PAGE 127

bleaching (Hallock et al. 1995, Toler 2002). During this time, profoundly damaged asexual broods defined as producing fewer than 50 offspring, at least 10% of which were malformed were frequently observed (Hallock et al. 1995). Breakage incidences in samples collected before 1990 from Hawaii, Ca yman Islands, Puerto Rico and Florida Keys were consistently low (between a bout 5 and 6 %; Hallock 1995), relative to breakage observed following bleaching events (10 40 %; Williams 2002) and during my study. Breakage tends to be more preval ent under chronic rather than acute stress, perhaps because chronically-stressed indivi duals survive longer and have a higher probability of encountering predators or ex periencing physical damage (Toler 2002). Shell damage during my study was consis tent with chronic stress affecting A. gibbosa with the largest affect on populations at KL 9 m, KL 18 m and BNP. Amphistigina bleach in response to excess solar energy, particularly higher-energy (blue, violet, and ultraviolet) wavelengths Bleaching was unknown in field populations of Amphistigina before 1988; since 1992 it has been observed annually on the Florida reef tract and in all oceans (Hallock 2000, Hallock et al. 200 6b). Prevalence and severity of bleaching declined in Florida reef tract populations between 1992 and 1997, with a sharp increase in summer 1998 (Williams 2002, Hallock et al. 2006b). In summers of 1997 and 1999, bleaching prevalence and severity did not vary significantly with season but observations of bleaching in about 20% of adults indicated that photo-inhibitory stress was still chronic (Williams 2002, Hallock et al. 2006b). I frequently observed partly bleached individuals, but the percentage of the popula tion exhibiting bleaching was generally low (~ 25%) and show ed no clear seasonal trends. Thus, bleaching during my study likely was caused by a chronic rather than an acute stress, similar to observations made in 1997 and 1999 (Williams 2002). Percentage of bleached A. gibbosa was highest at AR and at the deeper sites (KL 9 m and KL 18 m). Damage accumulates with test size (Williams et al. 1997, Williams 2002) because bleaching in A. gibbosa progresses as a degenerative disease, with permanent symbiont loss in affected chambe rs. Larger individuals at AR, KL 9 m and KL 18 m therefore had more time to accumulate damage. A paradox of chronic bleaching in these foraminifers is that incidences tend to be highest at sites otherwise suitable for growth and reproductio n. At sites where juvenile s recruit but other stressors

PAGE 128

limit survival, percentage of juveniles will be higher, but both abundance and incidences of bleaching will be lower. Successful reproduction, as shown by in creases in juven iles during summer months, indicates that bleaching stress wa s not affecting reproduction. No coral bleaching was observed during my study, supporting this conclusion. However, the prevalence of smaller sized indi viduals indicated that some unidentified stressors limited growth and survival. Chronic bleaching stress may result from oxidative damage passed onto offspring or increased photosensitivity from exposure to photoinhibitors (e.g., herbicides, PAHs) acting alone or synergistically affecting A. gibbosa at these sites. Further investigation of UV a nd PAR levels, CDOM concentr ations and concentrations of known photoinhibitors, along with cellular biomarker da ta, would help elucidate mechanisms and causes of bleaching in A. gibbosa at these sites. 4.6. Conclusions Densities and conditions of symbiont-b earing foraminifera support conclusions based on other indicators (e.g., co ral lesion regeneration rates; Fi sher et al., in press; Ch. 5) that the relatively favorable AR site cont rasted most sharply w ith the unfavorable BNP site. All LBF parameters except bleaching prevalence indicated that A. gibbosa were most stressed at the BNP site, likely due to unfavorable water quality (Cockey et al. 1996, Hallock et al. 2006a, b). Downs et al. ( 2005a, 2006) found evidence for a toxic response in both corals ( Montastraea annularis ) and white grunts ( Haemulon plumieri ) at BNP. Reef community assessments (Ch. 2) indicate that coral decline at BNP likely began relatively recently (within the last 10 years) based on high recent coral mortality and high macroalgal abundance. Densities of LBF we re higher at KL 9 m than at KL 18 m suggesting that KL 9 m provides slightly better habitat than KL 18 m, which corroborates conclusions based on coral lesi on regeneration rates (Ch. 5; Fi sher et al. in press) and biomarker profiles (Ch. 6; Fi sher et al. in prep). Abundances of Amphistegina were critical to distinguishing among the 6 m sites, supporting use of LBF as indicators of reef condition. Incidences of bleaching and breakage in A. gibbosa indicated chronic stress at al l sites during the study, with no evidence for acute photic stress during the st udy period. Similarl y, no coral bleaching

PAGE 129

was observed during my study. Using LBF, I determined that water quality was least favorable for calcifying organisms at BNP and less than optimal at all study sites. However the stressor(s) remain unknown. Fu rther assays (e.g., cellular diagnostic biomarkers, histological studie s, contaminant analysis) may be able to provide further insight into potential stressors that are co mpromising water quality at BNP and thereby direct management actions. Along the depth gradient, KL 3 m appeared to be in good condition based on high coral cover and lesion regeneration rates. However, densities of A. gibbosa were low indicating that these foraminifers are not good indicators at such nearshore shallow environments, due to their preference for sandy sediments. Amphistegina are most suitable for assessing reef environments 10 20 m deep but can be used to compare among shallow reefs that are neither too high-energy nor too low-energy, e.g, very shallow, nearshore reef environments such as KL 3 m that have very silty sediments. Densities of A. gibbosa also were low at KL 6 m and BNP, both of which are dominated by silty sediments (see Ch. 2). Archaias angulatus, Sorites marginalis and Laevipeneroplis proteus may be very abundant in low energy, nearshore environments where water quality is suitable (Fujita & Hall ock 1999) and therefore may be more useful bioindicators species there. Low densities of all LBF (especially Archaias angulatus ) at KL 3 m, KL 6 m, and BNP indicates so mething is limiting LBF populations, because LBF occur in high densities at some low-energy sites in the Florida Keys (e.g., Hallock et al. 1986a, Fujita & Hallock 1999).

PAGE 130

Table 4.1. Population parameters of Amphistegina (late spring early autumn unless otherwise noted) and th eir interpretive value Parameter(s) Range Interpretation Density (#/m-2) High (>5x104) Environmental conditions very good Medium (1-5x104) Environmental conditions suitable Low (5-10x103) Environment marginal Very low (<5x103) Environment stressed Absent Environment unsuitable Bleaching High (>50%) Acute photo-oxidative (photic) stress prevalence Medium (10-50%) Chronic photic stress Low (<10%) Minimal chronic photic stress Absent No photic stress Density/bleaching High/high Ac ute photic stress post-reproduction High/Medium Chronic photic stress post-reproduction Medium/Medium Chronic photic stress impacting reproduction; may include other stress Low/low Environmental stress probably not photic Low/high Ongoing, acute photic stress Shell damage High (>30%) Highly su sceptible to predation/infestation Medium (10-30%) Chronically susceptible to pred/infestation Low (<10%) Minimally susceptible to pred/infestation Juveniles High (>50%) With high density, conditions good With low density, unsuitable at time scales of weeks to months Medium (25-50%) Interpret in c ontext of density and bleaching Low (<25%) Reproduction impacted or suppressed Mean diameter ~0.6-0.7 mm With me dium to high densities, indicates early summer reproductive success >0.8 mm With bleaching and low densities, indicates suppressed reproduction late winter >0.9 mm Large in dividuals available for reproduction

PAGE 131

Table 4.2. Habitats of common larger bent hic foraminifers found on Florida reefs (adapted from Levy 1991, Hallock & P eebles 1993, Hallock 1999, and Hallock pers. comm.) Taxon Reef Habitat Depth (optimum) Amphistegina gibbosa Backreef, open shelf, reef margin <100 m (10-40 m) Asterigerina carinata Open shelf, reef margin <40 m (not known) Gypsina spp. Highly variable not known Heterostegina depressa Deep shelf or reef margin <100 m (30-50 m) Archaias angulatus Backreef, reef and open shelf <40 m (<10 m) Borelis pulchra Backreef, reef and open shelf <40 m (not known) Broekina orbitolitoides Backreef, reef, open shel f <40 m (10-30 m) Cyclorbiculina compressus Backreef, reef and open shelf < 40 m (5-30 m) Laevipeneroplis bradyi Backreef, reef and open shelf <40 m (10-30 m) Laevipeneroplis proteus Backreef, reef and open shelf <40 m (0-20 m) Peneroplis pertusus Backreef, reef and open shelf <40 m (0-20 m) Sorites marginalis Backreef, reef and open shelf <40 m (0-20 m)

PAGE 132

Table 4.3. ANOSIM2 results for differences among 6 m sites (averaged across all sampling periods); Global R = 0.28, significance level = 5% Table 4.4. ANOSIM2 results for differences among sampling periods (a veraged across all 6 m sites); Global R = 0.23, significance level = 5% KL 6 mWBARBNP KL 6 m0.240.300.37 WB0.140.40 AR 0.36 Aug-01Oct-01Mar-02Jun-02Aug-02Nov-02Feb-03 Aug-01n.s.n.s.0.460.36n.s.0.23 Oct-01n.s.0.510.33n.s.0.4 Mar-020.220.330.24n.s. Jun-02 0.250.380.34 Aug-02 0.170.26 Nov-02 0.24

PAGE 133

Table 4.5. Identification of Key Discriminating Larger Benthi c Foraminifers among the 6 m sites between August 2001 and Februa ry 2003 AR/BNP WB/BNP KL 6 m/BNP KL 6 m/AR WB/AR KL 6 m/WB Mean dissimilarity 52.4 49.3 46.5 41.5 36.8 36.4 Amphistegina gibbosa 13.7 12.0 10.6 9.4 7.9 7.6 Archias angulatus 9.5 10.4 9.2 7.3 6.8 6.7 Asterigerina carinata --2.5 --1.7 Borelis pulchra --2.2 1.7 1.4 1.4 Broekina orbitolitoides 4.5 3.2 -3.7 3.4 2.4 Cyclobiculina compressus 2.3 2.0 1.9 1.9 1.8 1.7 Heterostegina antillarium 4.7 3.7 3.7 3.7 2.8 2.6 Gypsina sp. 2.9 -1.8 2.4 2.3 -Laevipeneroplis proteus 7.2 6.8 6.5 5.1 4.9 4.9 Peneroplis pertusus 2.3 2.5 2.9 2.1 2.0 2.3 Sorites marginalis 1.7 1.9 2.3 1.6 -1.8

PAGE 134

Table 4.6. ANOSIM2 results for differences among depths (averaged a cross all sampling periods); Global R = 0.53, significance level = 5% Table 4.7. ANOSIM2 results for differences among sampling periods ( averaged across all depths); Global R = 0.18, significance level = 5% KL 3 mKL 6 mKL 9 mKL 18 m KL 3 m 0.530.820.71 KL 6 m 0.660.45 KL 9 m 0.32 Oct-01Mar-02Jun-02Aug-02Nov-02Feb-03 Oct-01n.s.0.360.22n.s.0.34 Mar-02 n.s.0.230.14n.s. Jun-020.250.250.19 Aug-02 0.20.32 Nov-02 n.s.

PAGE 135

Table 4.8. Identification of Key Discriminating Larger Benthic Foraminifers along the depth grad ient between October 2001 and February 2003 KL 3m/6m KL 3m/9m KL 3 m/18 m KL 6 m/9 m KL 6 m/18m KL 9 m/18 m Mean dissimilarity 40.7 51.2 46.9 33.5 32.8 36.4 Amphistegina gibbosa ----7.9 7.6 Archias angulatus 4.6 4.7 3.5 1.8 6.8 6.7 Asterigerina carinata 3.4 6.8 6.7 3.8 -1.7 Borelis pulchra 3.7 4.1 -1.7 .4 1.4 Broekina orbitolitoides 4.0 3.7 3.7 4.3 3.4 2.4 Cyclobiculina compressus 2.5 7.6 6.2 5.5 .8 1.7 Heterostegina antillarium 4.7 5.0 5.4 3.7 2.8 2.6 Gypsina sp. 2.9 -3.3 2.2 2.3 -Laevipeneroplis bradyi -4.6 5.5 3.4 4.9 4.9 Laevipeneroplis proteus 3.4 2.7 3.1 -4.9 4.9 Peneroplis pertusus 4.0 3.2 3.3 2.3 2.0 2.3 Sorites marginalis 3.9 4.9 3.2 2.8 -0.8

PAGE 136

Figure 4.1. Pictures of bleaching and damage in A. gibbosa Top left: normal color with no damage; Bottom left: partly bleached with no dam age; Top right: pale and broken; Bottom right: partly bleached and chipped. 0.5 mm

PAGE 137

Figure 4.2. Mean ( SE) densities of Amphistegina gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. D ensities are plotted on a log scale. Densities (individuals/m 2 ) 100000 10000 1000 100 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003 KL 3 m KL 6 m KL 9 m KL 18 m 100000 10000 1000 100 KL 6 m WB AR BNP6 m Sites Depth Gradient A B

PAGE 138

Figure 4.3. Mean ( SE) percentage of juvenile A. gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003 KL 3 m KL 6 m KL 9 m KL 18 m KL 6 m WB AR BNP80 100 20 40 60 6 m Sites Depth Gradient 80 100 20 40 60% JuvenilesA B

PAGE 139

AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003 0.9 0.8 0.7 0.5 0.6 0.4 0.9 0.8 0.7 0.5 0.6 0.4 6 m Sites Depth GradientDiameter (mm) KL 6 m WB AR BNP KL 3 m KL 6 m KL 9 m KL 18 mA B Figure 4.4. Mean ( SE) diameters of A. gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient.

PAGE 140

Figure 4.5. Mean ( SE) percentages of adult A. gibbosa exhibiting any degree of bleaching from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. 80 20 40 60 0 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003 KL 3 m KL 6 m KL 9 m KL 18 m 80 20 40 60 0 KL 6 m WB AR BNP6 m Sites Depth Gradient% Adults Partly BleachedA B

PAGE 141

Figure 4.6. Mean ( SE) percentages of damaged tests in populations of A. gibbosa from August 2001 to February 2003 at (A) the 6 m sites and (B) along the depth gradient. 0 10 20 30 40 KL 6 m WB AR BNP MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003 KL 3 m KL 6 m KL 9 m KL 18 m% Damaged Individuals0 10 20 30 406 m Sites Depth Gradient A B

PAGE 142

KL 6 m WB A R BNP AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003 KL 3m KL 6m KL 9m KL 18m100,000 10,000 1,000 100 1,000,000 6 m Sites Depth GradientDensities (individuals/m 2 )A B 100,000 10,000 1,000 100 1,000,000 Figure 4.7. Mean ( SE) densities of all symbiont-b earing (larger) foraminifera from August 2001 to February 2003 at (A) the 6 m sites an d (B) along the depth gradient. Densities are plotted on a log scale.

PAGE 143

Stress: 0.14 Stress: 0.18 KL 6M WB AR BNP Stress: 0.01 Stress: 0.01 KL 6 m WB AR BNP A B C AR WB KL 6 M BNP Figure 4.8. Multi-dimensional scaling plots (MDS) il lustrate the ordination of samples collected between August 2001 and February 2003 based on (A) the entire ass emblage of LBF, (B) the assemblage with A. gibbosa removed, and (C) A. gibbosa alone.

PAGE 144

Laevipeneroplis proteus 10 100 1000 10000 100000 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Density (#/m 2 ) KL 6m WB AR BNP Archaias angulatus 10 100 1000 10000 100000 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Density (#/m 2 ) KL 6m WB AR BNP A B C D Heterostegina depressa 1 10 100 1000 10000 100000 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Density (#/m 2 ) KL 6m WB AR BNP Broekina orbitolitoides 1 10 100 1000 10000 100000 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Density (#/m 2 ) KL 6m WB AR BNP Figure 4.9. Mean ( SE) densities of other dominant symbiont-bearing f oraminifera at the 6 m sites (A) Archaias angulatus (B) Laevipeneroplis proteus (C) Heterostegina depressa and (D) Broekina orbitolitoides

PAGE 145

OTHERFORAMS2001-2004 KL 3M KL 6M KL 9M KL 18M Stress: 0.17 Stress: 0.17 KL 3 m KL 6 m KL 9 m KL 18 m Figure 4.10. Multi-dimensional scaling plots (MDS) illustrate the ordination of samples collected along the depth gradient between October 2001 and February 2003 base d on the entire assemblage of LBF

PAGE 146

A 0 1 2 3 4 5 6B C OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003D KL 3m KL 6m KL 9m KL 18m OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003 0 1 2 3 4 5 6Density (individuals m -2 )Cyclobiculina compressusAsterigerina carinata Heterostegina depressa Broekina orbitolitoides Figure 4.11.Mean densities (individuals m -2 SE) of other dominant symbiont-bearing foraminifera along the depth g radient (A) Cyclobiculina compressus (B) Asterigerina carinata (C) Heterostegina depressa and (D) Broekina orbitolitoides Densities are plotted on a log scale.

PAGE 147

5. Lesion Regeneration Rates in Reef-building Corals ( Montastraea spp.) as Indicators of Colony Condition* 5.1. Abstract Regeneration rates of coral le sions reflect the ability of colonies to repair damage and therefore can be useful indicators of co ral health and environmental conditions. I quantified regeneration ra tes of boulder coral ( Montastraea spp.) at four, 6 m deep patch reefs within Biscayne Nationa l Park (BNP) and the upper Fl orida Keys National Marine Sanctuary (FKNMS), and along a 3 18 m depth transect in FKNMS. Coral lesions (approx. 2 cm2) created during sampling for cellular-d iagnostic analysis were monitored quarterly in 2001 and 2002, and in February 2003. Regeneration was a dynamic process, continuing longer than previous ly reported (>300 d after lesi on formation). Geographic location was the strongest factor affecting re generation rate at my study sites. Lesion regeneration differed signifi cantly among 6 m deep sites; sites offshore from John Pennekamp Coral Reef State Park (Algae R eef and White Banks) consistently had the highest regeneration rates, w ith colonies exhibiting exponen tial declines in lesion size and a high percentage of completely-healed le sions. Along the depth gradient, corals at the 3 m site regenerated significantly faster than corals at 6, 9 and 18 m. Colonies at the latter sites had highly variable and overall low regeneration rates, a low percentage of healed lesions, and a high occurrence of br eakage or Type II le sions lesions that increased in size by merging with areas of de nuded tissue on the colony. These results suggest that corals sampled at FKNMS 6, 9 and 18 m sites and BNP were in poor physiological condition or were exposed to sub-optimal environmental conditions This chapter is in press in Marine Ecol ogy Progress Series as Fisher EM, Fauth JE, Hallock P, Woodley CM (in pr ess) Lesion regeneration rate s in reef-building corals ( Montastraea spp.) as indicators of col ony condition. Mar Ecol Prog Ser

PAGE 148

5.2. Introduction Most corals are colonial organisms; a co lony can experience partial mortality in which part remains alive while another porti on dies. When a disturbance produces a lesion (partial mortality), the exposed coral skeleton becomes vulnerable to invasion by sessile organisms such as algae, resulting in th e lesion increasing in size. Alternatively, healing can occur if tissue regeneration is not impeded (Kawaguti 1937). Lesions that fail to heal completely within about two m onths are likely to become permanent patches of mortality (Meesters et al. 1994). Percent mortality of coral colonies is a useful gauge of reef condition (Ginsburg et al. 2001) because it can reveal a recent or chronic disturbance (Lang 2003) and influence colony growth and reproduction (Meester s et al. 1994, Van Veghe l & Bak 1994, Lirman 2000a). Williams (1994) proposed using coral lesions as indicators of environmental stress because they are a gene ralized response to a range of disturba nces, are independent of reef type, and can be monitored by managers easily and inexpensively. Williams noted that the frequency of coral lesions va ries among sites, with polluted sites having more lesions than relatively unpolluted sites. Quantifying colony damage and recovery rates also are essential for predicting dem ographic changes in coral populations (Bak & Meesters 1999). In corals, lesion regeneration begins with growth of an undifferentiated tissue layer created by the coenenchyme and polyps surrounding the lesion (Bak et al. 1977). After about two weeks, polyps begin to develo p in the new tissue (Meesters et al. 1994) and secrete thecal walls and a basal plate. These give rise to numerous radially arranged calcareous partitions (septa), which project inward and support the polyp mesenteries. Pigmentation and zooxanthellae return at the en d of the regeneration process (Bak et al. 1977, Kramarsky-Winter & Loya 2000). Coral re generation rates can vary with species (Kawaguti 1937, Bak et al. 1977, Nagelkerken & Bak 1998) and are in fluenced by lesion characteristics including the type of injury and its initial size, perimeter and shape (Meesters et al. 1994, Meesters et al. 1997b, Oren et al. 1997, Lirman 2000b, Hall 2001), and colony characteristics such as size (K ramarsky-Winter & Loya 2000, Oren et al. 2001). Under normal conditions, lesion size decreases expone ntially; deviations from

PAGE 149

this response suggest resource limitation (Meesters et al. 1997b), stress due to environmental conditions (Lester & Bak 1985, Meesters et al. 1992, Meesters & Bak 1993, Mascarelli & Bunkley-Willia ms 1999, Croquer et al. 2002, Fine et al. 2002) or competition (Hall 2001). The present study was part of a long-term project in the Florid a Keys testing the use of an integrated molecular biomarker system in corals (Downs et al. 2000, 2005a, Fauth et al. 2003). Here I compare the ability of star boulder corals ( Montastraea species complex) within two ma rine protected areas to regene rate biopsy-induced lesions. Lesion regeneration rates were assessed to characterize coral condition at these sites. Three specific questions were addressed: (1) Do regenera tion rates differ among sites, seasons or years? (2) Do regeneration rates va ry with depth? (3) Do regeneration rates vary with lesion parameters (e.g., initial lesion size, peri meter, shape) or colony characteristics (e.g., morphotype/species, size, % mortality)? In chapters 6 and 7, I will relate coral regeneration rates to ecological and cellular indicators to further identify potential sources of stress at my study sites. 5.3. Methods I assessed reef condition at community and colony scales at one patch reef in Biscayne National Park (BNP), and four patch reefs and two fore reef sites in the upper Florida Keys National Marine Sanctuary (FKN MS), as part of an ongoing study of coral ecophysiology (e.g., Downs et al. 2000, 2005a, Fauth et al. 2003). These seven sites (Fig. 1.1) comprised both a latitudinal transect with four sites at 6 m depth and a depth transect (Key Largo (KL) 3 m KL18 m) and were chosen in consultation with resource managers to reflect gradients in environmen tal conditions. Algae Reef (AR) and White Banks (WB) were adjacent to the extensive John Pennekamp Coral Re ef State Park, with intact coastal hammock, mangroves and seagra ss beds. Key Largo (KL) 6 m was located offshore from the most urbanized coastline of Key Largo, from which natural vegetation has been removed, natural topography has been altered to maximize waterfront properties and coastlines are lined with seawalls. This site lies alon g the route that recreational boaters and commercial dive operators take to reach popular Molasses Reef and other

PAGE 150

outer reefs in the Upper Keys reef tract. Mola sses Reef is the most heavily visited reef in the Upper Keys for diving (FKNMS website ). Alinas Reef, which is in Biscayne National Park (BNP) is offshore from urban Miami, FL. This site is potentially influenced by the extensive agricultural area south and west of Miami that drains into Biscayne Bay. Biscayne National Park also is near a nuclear power plant and major landfill. 5.3.1. Benthic Community Assessments In March 2002, a dive was made at each 6 m site to assess the benthic organisms using the rapid assessment methods descri bed by Lang (2003). At each site, a 10-m transect line was placed just above a haphaza rdly selected area of reef surface and live coral cover was determined by estimating the amount of living coral directly beneath the line. For each coral >10 cm in diameter lyi ng beneath the transect, I recorded species, maximum diameter and height, and percent recent and old mortality. "Recently dead" was defined as any non-living part s of the coral in which the corallite structures were still intact or covered by a thin layer of al gae or fine mud (Lang 2003). "Long dead" was defined as any non-living parts of the coral in whic h the corallite struct ures either were gone or covered by organisms that were not easily removed (L ang 2003). Due to differences in reef types, I could not us e this method for comparisons along the depth gradient. 5.3.2. Lesion Regeneration Between June 2001 and Febr uary 2003, I collected tissue samples that created standard-sized lesion) approximately qua rterly (February/March, June, August, October/November) from the same five colonies at each site. Prev ious studies showed that quarterly sampling was adequate to detect changes in coral physi ology as a result of seasonal and stressor variation (Downs et al. 2000, Fauth et al. 2003). I pr eferentially chose Montastraea faveolata for this study but sampled the morphotypes M. annularis and M. franksi when M. faveolata was not available. A si ngle morphotype was not found at all study sites: I sampled M. faveolata at all sites except KL 18 m, M. annularis at KL

PAGE 151

6 m and WB, and M. franksi at KL 9 m, KL 18 m, WB, a nd AR. I estimated size (maximum diameter) and percent partial mortal ity of each colony at the start of the study. I measured diameter (live and dead areas) to the nearest cm in planar view perpendicular to the axis of growth using a meter stick. Partial mortality was visually quantified by estimating the percentage of dead area from above in pl anar view as recommended by Lang (2003). I removed coral tissue and skeleton using a leather punch, which created a circular lesion 1-2 cm2 in area and 3 mm deep. Experimental lesions always were completely surrounded by live tissue. I immediately fill ed the hole with clay (Roma plastilina, medium grey; Blick Art Materials, Gales burg, IL) to fill the void produced by removing the underlying skeleton and limit intrusion of fouling and bioeroding organisms (Fig. 5.1). Use of clay filler was a decision made by park managers when permitting biopsy of these corals for molecular biomarker analysis. Clay provided corals with a flat surface over which to regenerate tissue but, as seen in this study, did not prevent fouling or bioerosion. However, regeneration rates repor ted here may represent maximal rates due to a possible reduction in bi ofouling. I then photographed each lesion using a Nikonos V 35mm camera with a close-up adapter and frame, calibrating measurements with a 4.5 mm long bar. I re-photographed each lesion du ring subsequent quarterly samplings to observe changes in size over time (Fig. 5.1). I scanned photographs to digital images and used image-analysis software (Image Pro) to calculate area (A) and perimeter (P) of all lesions that remained comple tely surrounded by live tissue (Type I lesions: Meesters et al. 1997a). If a lesion enlarged, thereby merging with an area of the colony that lacked tissue (Type II lesions: Meesters et al. 1997a), I c onservatively assumed no change in lesion size for that sampling date and removed it from further analyses because subsequent changes in area were unconstrained. When lesions merged with other sampling lesions (Fig. 5.1B), I calculate d their area as AL= AT/n; where AL is the area of the lesion used for further analyses, AT is the total area of all lesions joined together and n is the total number of lesions joined together. This calculation provided a conservative estimate of lesion area increase. In the few cases where initial lesion size was unavailable due to

PAGE 152

camera malfunctions, I substituted mean initi al lesion size for that sampling period. When photographs of final lesion size were unavailable, I used in situ measurements to calculate lesion area and perimeter using the equation for an ellipse, AL = ab, and PL = 2 *sqrt[(a2 +b2)/2], where AL is the area of the lesion, PL is the perimeter of the lesion, and a and b are one-half of lesion length and width, respectively. Larger lesions resulted from breakage of the coral skeleton, which ofte n was highly bioeroded. Due to the effect of initial lesion size on regenera tion, I removed lesions >3.4 cm2 from further analyses. 5.3.3. Data Analysis 5.3.3.1. Benthic Community Assessments I used one-way ANOVA followed by the Tukey-Kramer Honestly Significant Difference (HSD) method to determine if site s differed significantly in live coral cover and coral colony density. Data on coral diamet er and height, and r ecent and old mortality did not meet the normality assumptions of ANOVA. For these data, I tested for differences among sites using Kruskal-Wallis followed by Wilcoxon rank-sum tests. 5.3.3.2. Lesion Regeneration Data were analyzed in two groups: (1) by sites at 6 m depth along the northeast southwest traverse (BNP, AR, WB and KL 6 m), and (2) by sites along the depth gradient (KL 3, 6, 9 and 18 m). The KL 6 m site wa s common to both groups (Fig. 1.1). I examined lesion changes in three differ ent ways to answer specific questions. (1) Did lesion size decrease exponentially w ith time (cm2 d-1) and did this differ among sites and seasons? Can deviations from this model be used as an indicator of stress? I used least-squares regression to f it an exponential model of regeneration with an asymptote as recommended by Meesters et al. (1994, 1997b): y = yo + a eb *time where yo is the asymptote, a is the amount of tissue regenerated, and b is the slope of the curve. I only applied the exponential model to lesions with a minimum of one year of observations.

PAGE 153

(2) Did regeneration rates differ among s ites and seasons? I calculated rates for two different periods: short-term (45-154 d) and quasi-annual (319 376 d). The shortest period monitored reflects a time fram e similar to previous regeneration studies (e.g., Meesters et al. 1994, 1997b, VanVeghel & Bak, 1994). I calculated the quasiannual rates to determine how lesion size ch anged over multiple seas ons. For each lesion, I calculated the amount of tissue regenerated or lost ( T) as T = % change in lesion size*initial lesion size/time. I standardized regeneration to initial lesion perimeter (P) because this influenced regeneration rate s. I used repeated-measures MANOVA to determine whether standardized regeneration rates ( T/P) differed among sites, seasons and their interactions. I checked model a ssumptions (e.g., sphericity, homogeneity of variances, normality and independence) using residual plots. In cases where the sphericity assumption was not me t, I applied a univariate (unadj usted epsilon) approach. To interpret effects detected by MANOVA, I used one-way ANOVA followed by the Tukey-Kramer HSD method. I re gressed residuals of the regeneration-rate model against lesion (A, P and shape (P/A)) and colony (species, size, % mortality) parameters to determine if they affected regeneration rates. All colony characteri stics that explained significant variation in residuals were used as covariates in the MANOVA model. I also regressed quasi-annual regeneration rates ag ainst short-term rates to determine if monitoring for short time periods could be used to predict qua si-annual trends. (3) Were lesions among all sites capable of completely healing and did the number of Type II Lesions differ among sites and seasons? I used G-tests of independence with Williams correction (Sokal & Rohlf 1995) to determine if the number of lesions that closed complete ly or progressed into Type II lesions differed among sites. I performed non-linear regression using SigmaPlot 2000 (Systat Software, Inc.) and all other statistical analyses using JMP v.3.2. (SAS Institute Inc., Cary, NC, USA), with = 0.05 for all hypothesis tests.

PAGE 154

5.4. Results 5.4.1. Community Data Mean percent live coral cover and coral co lony density were low at all 6 m sites; both were highest at AR and lowest at KL 6 m (coral cover: ANOVA F3,13 = 11.4, p < 0.0007; coral colony density: ANOVA F3,26 = 3.7, df = 3, p < 0.03; Ta ble 5.1). Colonies at AR were significantly larger (maximum diameter) than colonies at WB and KL 6 m; colonies at BNP also were significantly larger than colonies at KL 6 m ( 2 = 9.2, df = 3, p < 0.03; Table 5.1). 5.4.2. Regeneration Model After removing lesions >3.4 cm2 from further analyses, initial lesion area ranged from 0.75 to 3.02 cm2 with a mean ( SE he reinafter) of 1.75 0.04 cm2 (n = 136) for the 6 m sites and from 0.68 to 3.32 cm2 with a mean of 1.80 0.05 cm2 (n = 128) along the depth gradient. Lesion size decreased exponen tially over time at AR, WB and KL 3 m (Fig. 5.2) as indicated by large r2 values and slopes (Table 5.2). W ith few exceptions, lesion size at BNP, KL 6 m, KL 9 m and KL 18 m either ch anged little or in some cases increased over time (Fig. 5.2). Lesions on corals at these s ites deviated from the expected decay model and fit either an expone ntial growth (increase in lesion si ze) model or a reduced model as indicated by low r2 values and slopes (Table 5.2). 5.4.3. Short-term (45 154 d) Regeneration Rates Short-term regeneration rates ( T/P) ranged from to 65 x 10-4 cm d-1 with a mean of 13 1 x 10-4 cm d-1 (n = 136) at 6 m sites and from to 91 x 10-4 cm d-1 with a mean of 13 2 x 10-4 cm d-1 (n = 127) along the depth gradient. Mean short-term regeneration rates di ffered significantly among the 6 m sites (repeated measures MANOVA: site effect F3,10 = 10.6, p < 0.002; Fig. 5.3A), but not among species, seasons or their interactions. Mean short-term regene ration rates at AR were significantly higher than at the other 6 m sites, and sh ort-term regeneration rates at WB were significantly higher than KL 6 m and BNP (AR: 23 2 cm d-1 x 104, WB: 15

PAGE 155

2 cm d-1 x 104, KL 6 m: 7 2 cm d-1 x 104, BNP: 6 2 cm d-1 x 104, Tukeys HSD Test). Short-term regeneration rates also diffe red significantly along the depth gradient (repeated measures MANOVA: site effect F3,8 = 4.4, p < 0.05; Fig. 5.3B) but not with season or depth x season interactions. Shor t-term regeneration rates at KL 3 m were significantly faster than at KL 6 m (19 3 cm d-1 x 104 vs. 7 2 cm d-1 x 104, respectively, Tukeys HSD Test) but not at KL 9 m and KL 18 m (Fig. 5.3B). Species differences partially explained va riation (<7 %) in the residuals of the regeneration-rate model for the 6 m sites (period: 45 154 days), with M. annularis having regeneration rates higher than M. franksi (1.3 0.3 cm d-1 x 104 vs. 0.8 0.4 cm d-1 x 104, respectively). However, this largely resulted from uneven distributions of morphotypes among sites, and the species eff ect was not significant when added as a covariate to the regeneration-rate model. Sp ecies differences did not significantly explain variation in the residua ls along the depth gradient. Initia l lesion size, perimeter and shape explained less than 10% of the residual error in the regeneration (T/P) model at the 6 m sites ([A] r2 = 0.09, p = 0.002; [P] r2 = 0.06, p < 0.01; [P/A] r2 = 0.09, p = 0.002). Residuals were positively correlated with bo th initial lesion size and perimeter, and negatively correlated with P/A. Along the depth gradient, lesion size, perimeter and shape were independent of model residuals 5.4.4. Quasi-Annual Regeneration Rate (319 376 days) At the 6 m sites, mean quasi-annual regeneration rates differed among sites (repeated measures MANOVA: site effect F 3,12 = 14.8, p = 0.0002), season ( F3,36 = 11.2, p < 0.0001), and with the site x season interactions ( F9,36 = 4.2, p < 0.0009). Corals at AR and WB regenerated significantly faster than corals at KL 6 m and BNP between June 2001 and 2002 (ANOVA: F3,15 = 7.3, p < 0.003; 9 1 cm d-1 x 104 and 9 0 cm d-1 x 104 vs. 3 1 cm d-1 x 104 and 2 3 cm d-1 x 104, respectively, Tukey HSD, Fig. 5.4A). Corals at BNP regenerated significantly sl ower than corals at WB between August 2001 and 2002 (ANOVA: F3,14 = 3.4, p < 0.05; -4 5 cm d-1 x 104 vs. 8 1 cm d-1 x 104 respectively, Tukey HSD) and corals at AR between October 2001 and 2002 (ANOVA: F3,14 = 4.8, p < 0.01; 0 3 cm d-1 x 104 vs. 8 1 cm d-1 x 104, respectively, Tukey HSD).

PAGE 156

Corals at AR also regenerated significantl y faster than colonies at KL 6 m and BNP between March 2002 and February 2003 (ANOVA: F3,15 = 5.3, p < 0.01; Fig. 5.4A; 21 5 cm d-1 x 104 vs. 7 1 cm d-1 x 104 and 2 5 cm d-1 x 104, respectively Tukey HSD). Short-term regeneration rates were significan tly correlated with quasi-annual trends (r2 = 0.37, p = 0.0001; regression equation: T/P (annual) = 0.29 T/P (short) + 2.3). Mean regeneration rates varied signifi cantly along the depth gradient (repeated measures MANOVA: site effect F 3,10 = 4.1, p = 0.04), with season ( F3,8 = 38.3, p < 0.0001), and the season x site interactions ( F9,19.6 = 5.6, p < 0.0007; Fig. 5.4B). Regeneration rates at KL 3 m exceeded those at all other sites between March 2002 and February 2003 (ANOVA: F3,15 = 5.3, p < 0.02; 18 4 cm d-1 x 104 vs. 7 1 cm d-1 x 104, 6 2 cm d-1 x 104 and 4 2 cm d-1 x 104, respectively Tukey HSD ). Along the depth gradient, short-term regeneration rates explai ned little variation in quasi-annual trends (r2 = 0.10, p = 0.007; regression equation: T/P (annual) = 0.12 T/P (short) + 3.3) due to high variability, especially among coloni es at the KL 9 m and KL 18 m site. 5.4.5. Healed and Type II lesions Coral colonies at AR completely healed significantly more lesions (30%) than colonies at the other 6 m sites (Gadj = 15.8, df = 3, p < 0.005). Along the depth gradient, significantly more lesions healed completely at 3 m depth (31%) than at other depths (Gadj = 12.8, df = 3, p < 0.01; Fig. 5.5). These results indicate signi ficant heterogeneity among sites in healing (Table 5.3). Of a total of 170 lesions created at the KL 3 m and all 6 m sites combined, only two merged with other lesions to become Type II lesions (Table 5.3). In contrast, at the deepest sites (KL 9 m and KL 18 m combined), 26% of lesions merged to become Type II lesions (Gadj = 12.2, df = 3, p < 0.01; Table 5.3). In two cases, lesions joined with another sampled lesion before merging with pa rtial mortality on other parts of the colony, becoming Type II lesions. In all other cases, lesions joined with partial mortality that occurred naturally on the colony, often asso ciated with increases in algae.

PAGE 157

5.5. Discussion Following recommendations of Williams (1994) and utilizing the extensive work of Bak, Meesters and coworkers, I evaluated lesion regeneration as an indicator of coralcolony condition at seven reefs in BNP and FKNM S. Using this bioindicator, I detected significant differences among sites in mean lesion regeneration rates. While most previous studies of regeneration monito red lesions for 60 -150 days; I followed regeneration for up to 595 days which allowed me to observe changes in recovery trends that might be missed by a study of shorter du ration. Short-term re generation rates were useful predictors of longer-term regenerati on rates among 6 m sites, but explained little variation along the depth gradient. Monito ring long-term regeneration appears necessary when comparing coral colonies living in different reef types/depths. Long-term regeneration rates were time dependent whereas short-term regeneration rates were not. Coral lesions regenerate at a rate de termined by the number of polyps surrounding each lesion (Meesters et al. 1997b, Oren et al. 1997, Lirman 2000b) and normally follow an exponential-decay model with an asymptot e at full healing (Meesters et al. 1994, 1997a, Lirman 2000b). I found that change s in lesion size were dynamic and sitedependent and often deviated from the expect ed exponential-decay model. Some lesions that initially increased in size later regenerate d, and other lesions th at initially began to regenerate later increased in size, especially at sites with high algal growth (e.g., BNP, KL 9 m, KL 18 m). If lesions with a P/A ratio >2 cm-1 should be able to fully regenerate (Meesters et al. 1997a), then most lesions in my study should have healed completely. However, only 14% (n = 228) fully regene rated. The largest lesion (2.0 cm2) that fully regenerated did so after 243 days; after 151 da ys, this lesion had regenerated 78% of its area to a size of 0.45 cm2. Most lesions that healed completely regenerated most (>70%) of their area within 151 days, but complete healing often requ ired a year or longer. One lesion that completely healed after 270 days increased 66% in size in the first 56 days before beginning to regenerate My study confirms that re generation can continue for a year or more and that lesions that do not init ially regenerate (or even increase in size) can regenerate later if conditi ons become favorable.

PAGE 158

5.5.1. 6 m Sites Mean lesion regeneration rates varied si gnificantly among sites at the same depth, suggesting that lesion regeneration may be a useful indicato r of variation in environmental conditions. Of the 6 m site s, corals consistently had the highest regeneration rates at AR, which is adjacent to the extensive John Pennekamp Coral Reef State Park. Colonies at AR had significantly more completely healed lesions than the other 6 m sites. AR also had the highest live coral cover with relatively large colonies. Lesions at the other site adjacent to the st ate park (WB) also re generated exponentially but many failed to heal complete ly, leaving those corals suscep tible to fouling organisms. Partial coral mortality of the community was lo west at this site. In contrast, KL 6 m, located offshore from the most urbanized co astline of Key Largo, had low regeneration rates and low overall live coral cover. Cora ls from the site in Biscayne National Park, offshore from urban Miami, FL, had the lowe st regeneration rates; lesions there often increased in size. Large increases in lesi on size at BNP often were associated with seasonal increases in algae (e.g., June 2002), which sometimes resulted in lesions merging together. During my study, BNP co rals had poor lesion recovery and also exhibited mortality elsewhere on the colonies Mean partial mortality estimated along transects also was highest at this site. In 2000, Montastraea colonies at BNP experienced a severe oxidative and protein denaturing st ress, likely due to chemical contaminant exposure (Downs et al. 2005a). The colonies I sampled were generally large in size, with substantial contiguous areas of living tissue, suggesting that the stressor(s) causing poor lesion recovery and partial mortality likely we re recent, within the last 10 15 years or less. Therefore, if stresses can be identified and alleviated at this site, these large coral colonies may survive. Responses of other reef organisms (e.g., white grunts and foraminifers) at these sites are consistent with observations of lesion regeneration. Downs et al. (2006) compared biomarker levels in white grunts ( Haemulon plumieri ) at BNP, WB and KL 6 m, finding evidence for a toxic response to a xenobiotic at BNP. Concentrations of pesticides in grunt livers were highest at KL 6 m and lowest at WB (Downs et al. 2006). Hallock (2000) proposed using abundances of reef-dwelling foraminifers that host algal

PAGE 159

symbionts to indicate whether environmen tal conditions support calcifying organisms dependent upon algal symbioses; Fisher and others (in prep.) found densities of such foraminifers lowest at BNP and KL 6 m a nd highest at WB and AR (see Ch. 4). 5.5.2. Depth Gradient Mean regeneration rate also varied am ong depths, but qualitative results depended on how long lesions were monitored. When monitored for <1 year (45 154 d), shallowwater (3 m) corals regenerated significantly fa ster than corals at 6 m but not those at 9 and 18 m. Key Largo 3 m also had significant ly more healed lesions than other sites along the depth gradient. For lesions sampled in March 2002 and monitored for approximately 1 year, shallow-water corals (3 m) showed higher re generation rates than all deeper water corals (6 18 m) along the Key Largo transect. Deeper-water corals typically receive less radiant energy and theref ore may have lower carbon reserves than corals in shallow water (Nagelkerken et al 1999). However, this does not explain why KL 9 m and KL 18 m had mean short-term re generation rates similar to the shallowest site. Also, differences in mean regenera tion rates were not seen between the KL 6 m and the deeper sites. Regeneration rates of corals from KL 9 m and KL 18 m were highly variable. Lesions that initially decreased in size often later increased in size and overall live coral cover at these sites was low (<7%). These Montastraea colonies were bioeroded by clionid sponges, making them susceptible to br eakage and resulting in greater patchiness of live tissue, possibly reduci ng the corals ability to reco ver from damage. Type II lesions developed more frequently in corals from these deeper sites. Many lesions joined with dead regions that were unrelated to my sampling. In two cases, the entire colony died; one each at KL 9 m and KL 18 m. Lesion growth often was associated with increased algal turf, particularly thick turfs mixed with fine sediments. Hall (2001 ) reported that regeneration was negatively correlated with algal settlement and cover (p articularly macroalgae), which requires large energy expenditure by corals to overgrow. In my study, algal turfs and macroalgae fluctuated in abundance, possibly associated with seasonal cha nges (as in Lirman &

PAGE 160

Biber 2000). Some lesions at my sites regene rated when algal biomass declined but later increased in size as algae grew, shadi ng and possibly killing polyps surrounding the lesions. Particularly in spring and su mmer, I observed dark reddish cyanobacterial blooms that formed thick mats on the bottom and overgrew portions of these corals. 5.5.3. Comparisons among All Study Sites All sites I sampled had relatively low coral cover (<20%) and appeared to be experiencing stress (Fisher et al. unpublis hed data). Connell (1997) observed that chronically stressed reefs were less likely to recover from acute or physical disturbances than reefs that were not chronically stre ssed. I observed that coral colonies along developed portions of the coastline (i.e., BNP, KL 6 m, KL 9 m, a nd KL 18 m) were least capable of recovering from damage and mortalit y. Colonies at site s offshore from John Pennekamp State Park (AR and WB) recove red from damage despite exposure to potential stressors (e.g., photic stress, contaminants). Although KL 3 m is along the same portion of coastline as KL 6, 9, and 18 m, lesion recovery and coral condition (36 13 % live coral cover) at this site was good. Other studies also found that Floridas inshore patch reefs appear to be in better condition and have higher coral cover relative to offshore reefs (Beaver et al. 2005). For ex ample, corals at KL 9 m and KL 18 m bleached in 1999, while those at KL 3 m and KL 6 m did not (Fauth et al. 2003). 5.5.4. Effect of Colony and Lesion Characteristics Colony size and previous partial tissue mort ality did not affect regeneration rates, probably because colony size was not sma ll enough to limit resources allocated to regeneration (Oren et al. 2001). Once regenera tion rate was standardized to perimeter, which is a measure of coral tissue availa ble for regrowth in the surrounding margin (Meesters et al. 1994), lesion ar ea explained only a small perc entage of the variation in regeneration rate. Colony morphotype did not affect regeneration rate but the three types were not evenly sampled among sites, which c ould influence the results. However, low variation among colonies with in sites containing differe nt species suggests that morphotype was not a major factor affec ting regeneration. Taxonomic differences

PAGE 161

between these morphologies remain uncerta in (Lopez et al. 1999, Fukami & Knowlton 2005). 5.6. Conclusions Based on observations of lesion regeneration rates, coral coloni es under relatively favorable environmental conditi ons (e.g., AR, WB, and KL 3 m) consistently have: high regeneration rates, where lesion sizes decrea se exponentially over tim e; a high percentage of healed lesions; and infrequent Type II lesions. Likewise, under less favorable conditions (e.g., KL 6 m, KL 9 m, KL 18 m, BNP) lesions e xhibit little regeneration, or high variability including increases in lesi on size (overall low regeneration rates); low percentage of healed lesions; frequent Type II lesions; and a high pe rcentage of breakage (indicative of bioerosion). Ca uses of differences in coral regeneration at small spatial scales deserve furthe r investigation. To standardize comparisons of lesion regeneration rates, I recommend (1) monitoring lesions of a similar size and perime ter, (2) comparing sites similar in depth and habitat type (e.g., patch reef, fore reef), and (3) monitoring lesions for more than one year because many lesions may require >200 da ys to heal. I also recommend recording the percentage of healed le sions and the occurrence of T ype II lesions. Regeneration rates of coral lesions reflect the ability of co lonies to repair damage and therefore can be useful, inexpensive indicators of reef cora l condition or of environmental conditions. A caveat of this bioindicator is that it is not capable of separating effects of coral health versus external environmental factors on lesion regeneration rate. More expensive assays can then be applied to distinguish between st ressor types at sites where coral regeneration is compromised.

PAGE 162

Table 5.1. Comparison of benthic parameters (mean SE) along 10 m transects at four 6 m-patch reefs. Methods followed the Atlantic Gulf and Rapid Reef Assessment protocol. Data not connected by the same superscrip t letter differed sign ificantly (p < 0.05). Site n Colonies Live Coral Coral Coral Recent Old Density Cover Height Diameter Mortality Mortality (#/m) (%) (cm) (cm) (%) (%) KL 6 m 7 0.64 A 7 A 11 A 21 A 3 A 11 A (0.07) (1) (3) (5) (2) (4) WB 8 0.71 AB 9 A 13 A 19 AB 2 A 8 A (0.07) (1) (1) (2) (1) (2) AR 8 0.94 B 16 B 23 A 40 C 2 A 12 A (0.10) (1) (4) (9) (1) (3) BNP 8 0.64 A 8 A 25 A 37 BC 4 A 19 A (0.05) (1) (6) (8) (2) (8)

PAGE 163

Table 5.2. Mean ( SE) r2 values of the five colonies fo r the regression decay model, y = yo + a e-b*time. A zero r2 value was assumed for all lesions th at did not fit this model. The last column includes overall mean ( SE) slope (b) (cm2 d-1 x 102). Site abbreviations as in Fig. 1. Location JUN 2001AUG 2001OCT 2001MAR 2002OVERALL SLOPE JUN 2002 AUG 2002 OCT 2002 FEB 2003 MEAN KL 3 m 0.95 (0.03) 0.52 (0.22) 0.94 (0.05) 0.98 (0.01) 0.85 (0.07) 2.1 (0.7) KL 6 m 0.55 (0.18) 0.61 (0.15) 0.33 (0.18) 0.88 (0.07) 0.59 (0.09) 0.3 (0.1) KL 9 m 0.58 (0.19) 0.39 (0.24) 0.55 (0.23) 0.72 (0.24) 0.55 (0.11) 0.1 (0.1) KL 18 m 0.56 (0.22) 0.37 (0.22) 0.72 (0.24) 0.71 (0.24) 0.59 (0.11) 0.4 (0.2) WB 0.97 (0.01) 0.98 (0.01) 0.98 (0.02) 0.90 (0.07) 0.96 (0.02) 1.2 (0.2) AR 0.96 (0.01) 0.80 (0.20) 1.00 (0.00) 1.00 (0.00) 0.93 (0.05) 1.7 (0.3) BNP 0.23 (0.23) 0.36 (0.22) 0.37 (0.22) 0.21 (0.20) 0.30 (0.10) 0.2 (0.2)

PAGE 164

Table 5.3. Percentage of healed and Type II lesions (no longer encl osed by living tissue) at each site. Total number of lesions was < 35 at KL 9 m, KL 18 m, AR and BNP due to breakage during sampling (as discussed in methods). Site Total # # Healed % Healed # Type II % Type II KL 3 m 35 11 31 0 0 KL 6 m 35 2 6 1 3 KL 9 m 29 4 14 6 21 KL 18 m 29 1 3 9 31 WB 35 1 3 0 0 AR 33 10 30 0 0

PAGE 165

October 2001 March 2002 June 2002 August 2002 November 2002 B HEALED A Figure 5.1. Examples of lesions at 6 m sites between October 2001 and November 2002 showing two extremes. (A) Algae Reef (AR) lesion completely healed by June 2002. (B) Alinas Reef (BNP) lesion joi ned with other sampling lesions in June 2002 and became covered with turf algae. Black arrow points to the lesion of interest.

PAGE 166

0100200300400500600 Time (days)4 3 5 0 1 2Lesion size (cm 2 ) 0100200300400500600 Time (days)4 3 5 0 1 2Lesion size (cm 2 )0100200300400500600 Time (days) 4 3 5 0 1 2Lesion size (cm 2 )March 2002 Lesion 4 3 5 0 1 2Lesion size (cm 2 )0 2 4 6 810 12 0100200300400500600 Time (days)0 1 2 3Lesion size (cm 2 ) 0100200300400500600 Time (days)0 1 2 3Lesion size (cm 2 ) 0100200300400500600 Time (days) 0100200300400500600 Time (days)0 1 2 3Lesion size (cm 2 )March 2002 Lesion KL 6 m WB AR BNPA B KL 3 m KL 6 m KL 9 m KL 18 m June 2001 Lesion June 2001 LesionAugust 2001 Lesion October 2001 Lesion August 2001 Lesion October 2001 LesionLesion size (cm 2 )0100200300400500600 Time (days)Figure 5.2. Mean lesion size ( SE) through time for each season between June 2001 and March 2002 at (A) the 6 m sites and (B) along the depth gradient. Ax es staggered to align sampling dates. Note expanded y-axis in panels showing lesion regeneration along depth gradient in August 2001. Merging of two sampling-induced lesions occurr ed at KL 9 m (in March 2002), at KL 18 m (in February 2003) and at BNP (in June 2002 and August 2002). An additional lesion joined with the previously merged lesions at BNP in October 2002. Lesions that progressed into Type II lesions or data removed for other reasons (as discussed in methods: breakage or initial size >3.4 cm 2 ) were not included in means.

PAGE 167

JUN AUG 2001 AUG NOV 2001 NOV 2001 MAR 2002 MAR JUN 2002 JUN AUG 2002 AUG NOV 2002 NOV 2002 FEB 2003Depth Gradient B 6 m Sites A KL 6 m WB RAR BNP KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m 0 10 20 30 40 50 -20 -10 0 10 20 30 40 50 -20 -10D T/P (cm d -1 x 10 4 )Figure 5.3. Regeneration rates standardized to initial lesion perim eter (mean D T/P SE) for each season from one sampling event until the next. Comparisons (A ) among 6 m sites and (B) along depth gradient. Regeneration rates were cal culated between June and August 2001 (54 13 d), August and October 2001 (5 6 d), October 2001 and March 2002 (153 2 d), March and June 2002 (91 1 d), June and August 2002 (48 13 d), August and November 2002 (74 d) and November 2002 and February 2003 (99 1 d).

PAGE 168

6 m Sites A KL 6 m WB RAR BNP 0 -10 10 20 30 -20June 2001-2002 August 2001-2002 October 2001 2002 March 2002 February 2003 0 -10 10 20 30 -20 Depth Gradient BD T/P (cm d -1 x 10 4 ) KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m Figure 5.4. Regeneration rates standardized to initial lesion perime ter (mean D T/P SE) for each season between June 2001 and March 2002 from the time of sampling until the following year. Compared (A) among the 6 m sites and (B) along the 3 18 m depth gradient. Regeneration rates were calculated between June 2001 and 2002 (357 10 d), August 2001 and 2002 (355 1 d), October 2001 and November 2002 (374 2 d) and March 2002 and February 2003 (321 2 d).

PAGE 169

0 10 20 30 40 50 1 2 10 KL 6 m WB AR BNP Total number of completely healed lesions 6 m Sites 3M 0100200300400 0 10 20 30 40 50Cumulative % Lesions HealedDepth Gradient 11 2 4 1 Time (days) KL 3 m KL 6 m KL 9 m KL 18 m Total number of completely healed lesions Figure 5.5. Cumulative percentage (%) of lesions completely healed with time (days) at (A) the 6 m sites and (B) along the depth gradient. Numbers adjacent to lines in the shaded area are total numbers of completely healed lesions at each site.

PAGE 170

6. Environmental Links to Coral Stress Response 6.1. Introduction 6.1.1. Linking Potential Stressors to Organism Responses Florida reefs have undergone severe degr adation over the past several decades (Andrews et al. 2005), with a documented Keys-wide loss of 45% live coral cover between 1996 and 2004 (Beaver et al. 2005) and 50 90% since the 1970s (Porter et al. 2002, Gardner et al. 2003, Palandr o et al. 2003 Palandro 2006). Reef decline has been attributed to a number of global and regional stressors including climate change and bleaching, disease, tropical storms, coastal development and runoff, over-harvesting and pollution (Porter et al. 1999, Bellwood et al. 2004, Waddell 2005). Current monitoring tools have limited ability to differentiate among these stressors and are incapable of determining mechanisms of decline (Downs et al. 2005b). My study examined effects of selected stressors on corals ability to heal by comparing lesion regeneration rates (Ch. 5) with environmental data sets (Ch. 2) and cellular diagnostic data (this chapter). Regeneration rates reflect a corals ability to heal from a disturbance and have been used as indicators of coral physiological condition (F auth et al. 2005, Fisher et al. in press). Regeneration can affect coral fitness by competing for energy with other critical processes such as growth (Bak 1983, Guzm an et al. 1994, Meeste rs et al. 1994) and reproduction (Guzman et al. 1994, VanVeghel & Bak 1994, Kramarksy & Loya 2000, Lirman 2000a, Oren et al. 2001, Kramarsky-Winter 2004). Causal inference can be used to link eff ects of stressors to responses in corals (Suter et al. 2002). Causal in ference is defined as analyzi ng available information, which may include spatial or temporal associations of potential cause and effect, field or lab results, and diagnostic evidence from affected organisms to generate evidence against a particular stressor (Bro-Rasmussen & Lkke 1984, Fox 1991, Suter et al. 2002). If the cause cannot be identified with sufficient c onfidence, effects are re-evaluated and the process starts over. A bette r understanding of the causes of reef decline provides managers with greater confidence when targ eting remediation efforts against stressors

PAGE 171

and determining when and what type of action is necessary (Jameson et al. 2002, Suter et al. 2002, Adams 2005). 6.1.2. Potential Candidates of Stress The first step of causal inference is lis ting potential stress candidates based on available information (Suter et al. 2002). Rege neration rates are reduced with increases in nutrients (Koop et al. 2001), turbidity (Croquer et al. 2002), sedimentation and resuspension rates (Meesters et al. 1992, Croquer et al. 2002), and enhanced with increases in water temperature (Lester & Bak 1985, Kramarsky-Winter & Loya 2000, Paz-Garca & Reyes-Bonilla 2006). Other f actors known to stress corals that may negatively affect regeneration rates in clude algal competition (Lirman 2001), high irradiance (Lesser & Farrell 2004, Lesser 2006), and marine po llutants (e.g., heavy metals, pesticides; Guzman & Jimenez 1992, Morgan & Snell 2002, Owen et al. 2002, Downs et al. 2005a). Reduced availability of food or autotrophically derived energy also can reduce regeneration rates, as evidenced by decreased regenerati on rates with reduced light levels (with increasing depth; Nagelk erken et al. 1999) and with symbiont loss during bleaching (Meester s & Bak 1993, Fine et al. 2002, Mascarelli & BunkleyWilliams 1999). In some cases, energy may be gained heterotrophically (Rinkevich 1996, Henry & Hart 2005) or by reallocating en ergy from other life history processes (Guzman et al. 1994, Kramarsky-Wi nter 2004, Henry & Hart 2005). 6.1.3. Metabolic Costs of Stress on Corals An animal usually functions in a homeostatic state and has a limit of compensation for changes in any environmen tal factor (Sindermann 1996). Stress is defined as any environmental al tercation that extends homeost atic or protective processes into a compensatory state be yond the normal limits of an or ganism (Seyle 1955, Bayne et al. 1985, Moore 2002). If compensatory limits are exceeded, the organism experiences increased energy expenditure and disabilitie s begin to appear (Depledge et al. 1993, Sindermann 1996). Corals in stressed enviro nments are likely to increase energy expenditure as cellular defenses neutralize or dissipate the effects of stress and restore cellular or tissue damage, thereby reduc ing resources (Koehn & Bayne 1989, Williams

PAGE 172

1994, Morgan & Snell 2002) available for rege neration, growth and reproduction. Small changes in protein/maintenance metabolism can have major effects on energy status (Hawkins 1991, Coustau et al. 2000). Consequently, stress renders the i ndividual, and ultimately the population, at a disadvantage through reduced growth, impaired reproduction, and increased susceptibility to disease and mortality (Bayne et al. 1985, Adams 2005, Downs 2005, Downs et al. 2005b) The extent of population response (collective individual responses ) and the ability of the popul ation to recover depend on the intensity and dura tion of environmental change (Si ndermann 1996). There are three phases of physiological respons es to stress: alarm, resist ance and exhausti on (Sindermann 1996). The alarm phase includes immediate or short-term behavioral, biochemical, or physiological responses to non-optimum changes in the environment. The resistance or adaptation phase includes longer-term bioche mical/ physiological resp onses that improve the likelihood of survival in the non-optimal environment; and the exhaustion phase includes failure of critical biochemical functions, leading to physiological and morphological disorders and death (Sinde rmann 1996). This study provided an opportunity to determine where coral colonies at my study sites fit w ithin these phases of physiological stress response by examining bot h coral cellular parameters and colony responses (e.g., regeneration rates). 6.1.4. Cellular Diagnostic System Environmental stressors affect organi sms by overwhelming defenses at lower levels of the biological hierar chy: molecular, cellular, and organismal-level homeostatic processes (Moore 2002, Downs 2005). The Ce llular Diagnostic System (CDS) is a systematic approach to defining and inte grating cellular biomarkers based on their functionality within the cell and how deviatio ns in their behavior may reflect overall cellular operation or performance (Downs 2005) Cellular Diagnostic parameters can be split into functional groups including (1) pr otein metabolic condition, (2) oxidative damage and response, (3) metabolic cond ition and integrity and (4) xenobiotic detoxification (Appendix A; Downs 2005). Pr otein metabolic condition involves the process of protein synthesi s, maturation and degradati on. Oxidative damage and response involve antioxidant pathways that a llow the cell to func tion in an oxygen-rich

PAGE 173

environment. Metabolic condition and integr ity involve a number of sub-processes or metabolic pathways that maintain the cell in a differentiated state from its environment. Xenobiotic detoxification invol ves the process of preventi ng or reducing the adverse effects of exposure to foreign chemicals (e.g., pesticides). The CDS diagnoses organismal hea lth by (1) quantifying cellular and physiological condition (e.g., prot ein metabolism, genomic inte grity), (2) characterizing types of cellular physiological stress (e.g., oxi dative stress, xenobiotic stress), and (3) determining if defenses have been built up ag ainst a particular stress (e.g., pesticide, heavy metal, or polycyclic aromatic hydr ocarbons; Downs 2005, Downs et al. 2000, 2001a,b, 2005a, 2006). The CDS was used in this study to determine (1) if coral biomarker profiles differed among study sites or sampling periods, (2) if coral biomarker profiles reflected stressed conditions in the coral, and (3) what stressor(s) the corals likely experienced (e.g., pollutants vs. increased oc ean temperatures) and possible mechanisms of stress. Shifts in the stea dy-state biomarker levels indicate a shift in the equilibrium of the sub-systems that they represent. Devia tion of the behavior of a specific parameter from the reference is an altered state, wh ich is defined as a pa thology or diseased condition if that phenotype is associated with conditi ons that adversely affect performance (e.g., reduced regeneration rate s or growth; Peters 1997, Downs 2005). Unlike traditional ecological monitoring methods, a cellular di agnostic approach can distinguish among different stressors beca use they elicit a sp ecific biological response in exposed organisms (Depledge et al. 1993) and the cellular function of each biomarker is well understood (Downs 2005). Through examination of multiple parameters, CDS also can elucid ate cellular mechanisms of stress. For example, in the Florida Keys, concentrations of specific diagnostic markers (e .g., lipid peroxide and chloroplast small heat shock protein) were linked to elevated water temperatures, disruption of homeostatic mechanisms and bleaching of the coral, Montastraea annularis (Downs et al. 2002). Certain diagnostic ma rkers (e.g., chloroplast small heat shock protein [ChlpsHsp]) taken in the context of other cellular parameters could predict which coral colonies would bleach six months in adva nce (Fauth et al. 2003). Corals with high levels of antioxidant enzymes (e.g., Cu/Zn superoxide dismutase, Mn superoxide dismutase) were less likely to bleach than those with low levels (Fauth et al. 2003).

PAGE 174

Downs et al. (2002) used this evidence to fo rm the oxidative theory of coral bleaching based on the following evidence: (1) pigment loss and bleaching followed oxidative damage; (2) increased oxidative damage pr oducts (protein carbonyl and lipid peroxide) and protein turnover activity we re highly correlated with ble aching; (3) cellu lar integrity was compromised by oxidative stress; (4) cellu lar defenses (e.g., antioxidant defenses and stress proteins) were capable of providing protection from bleaching to corals; and (4) negative relationships between ChlpsHsp and oxidative damage indicated that breakdown of photosystem II was the primary generator of reactive oxygen species and therefore the underlying source of oxidative stress and temper ature-associated cora l bleaching. Downs et al. (2002) proposed that ble aching is the corals final de fense against oxidative stress. Some caveats of using a biomarker appr oach include: (1) na tural populations often are exposed to multiple stressors, many of which can act synergistically, making interpretations difficult, (2) different organism s can have variable biomarker responses to similar stresses (e.g., Downs et al. 2001a,b), and (3) biomarker responses fluctuate seasonally and with changes in nutritional st ate, and in developmental or reproductive stages (Depledge et al. 1993). Use of an integrated biom arker system in addition to laboratory studies can elucidate the synergistic effects of multiple stressors. For example, CDS showed that normal photosynthetically act ive radiation (PAR) increased effects of heat stress on Montastraea faveolata by increasing levels of oxidative stress, overwhelming antioxidant defenses, and resulti ng in high levels of protein denaturation (Downs et al. 2000). 6.1.5. Linking Cellular Biomarkers to Higher Order Processes Understanding how multiple stressors effect reefs requires a hierarchical, mechanistic approach based on multiple lines of evidence (Adams 2005, Downs 2005, Yeom & Adams in press). This allows resear chers to determine whether (1) an organism is responding to a stressed c ondition and (2) that stress re sulted in reduced physiological function (Downs 2005, Moore et al. 2006). Lowe r levels of biological organization (e.g., cellular biomarkers) can provide informati on on the mechanism of decline, whereas higher levels of biological organization (e.g., coral re generation rates, community

PAGE 175

condition) provide information on the effect of stress on overall organismal fitness and function (Fig. 1.1). A major challenge in biomarker work is connecting cellular processes with higher order processes (Moore et al. 2006, Yeom & Ad ams in press). Molecular and cellular biomarkers have the potential to provide early distress signals of reduced performance, impending pathology and diminished h ealth (Moore 1990, Moore & Simpson 1992, Depledge et al. 1993, Downs 2005, Downs et al. 2005b). For example, lysosomal stability was a predictive tool of cellular in jury and pathology in marine mussels (Moore et al. 2006). Accumulations of biomarkers of xenobiotic response and decreased protein turnover were associated with decreased regeneration rates in the mustard hill coral, Porites astreoides (Fauth et al. 2005). Ubiquitin, cy tochrome P450 2-class (CYP-2) and cytochrome P450 6-class (CYP-6) explained 24 % of variation in re generation rates of P. astreoides Corals with high levels of ubiquitin and low levels of CYP-2 and CYP-6 had the highest regeneration rates, indicating that exposure to a xenobiotic resulted in reduced physiological condition. My study further exam ines the relationship between changes in coral cellular condition and changes in physiological cond ition (e.g., regeneration rates and mortality). 6.2. Objectives My study is a continuation of biomarker st udies in the Upper Florida Keys that began in 1999 (e.g., Downs et al. 2000, 2005a, Fa uth et al. 2003). Objectives of this chapter include (1) to differe ntiate between global (e.g., te mperature) versus local (e.g., pollutants) stressors, (2) to de tect subtle and chronic effect s of environmental stress on corals and (3) to diagnose coral condition at my study sites based on parameters of the cellular diagnostic system and coral colony responses (e.g., regeneration rates). My hypothesis was that a coral colony, for whic h the CDS indicated stress (e.g., xenobiotic, oxidative, etc.), was less likely to regenerate th an a coral with a lower stress signal, due to allocation of resources to cellular maintenan ce. I worked in collaboration with C. M. Woodley, C. A. Downs, and J. E. Fauth to interpret these data.

PAGE 176

6.3. Methods 6.3.1 Study Sites Coral condition was assessed using the cellular diagnostic system at one patch reef in Biscayne National Park (BNP), and four patch reefs plus two depths on one forereef in the upper Florida Keys Nationa l Marine Sanctuary (FKNMS). These seven sites (Fig. 2.1) comprised both a latitudinal transect with fo ur sites at 6 m depth and a depth transect [Key Largo (KL) 3 m KL18 m] and were chosen in consultation with resource managers to reflect a spectrum of possible anthro pogenic influence, based on distance from urbanized coastal development. The study was carried out approximately quarterly between March 2001 and February 2003. The Key Largo (KL) depth transect is located offshore from the most urbanized co astline of Key Largo, from which natural vegetation has been removed and natural topography altere d to maximize waterfront properties; the coastline is lin ed with seawalls. These s ites lay along the route used by recreational boaters and commercial dive opera tors to reach Molass es Reef and other heavily used outer reef s in the Upper Keys re ef tract, so pollutan ts such as hydrocarbon combustion products may be more prevalent. In addition, larger boats regularly stir up sediments, potentially remobilizing nutrien ts and chemical pollutants (Kruczynski & McManus 2002). Algae Reef (A R) also is offshore from Ke y Largo but is adjacent to John Pennekamp Coral Reef State Park and s ituated mid-way between the Key Largo and the BNP sites. The natural coastline is native, intact and relatively vegetated with coastal hammock mangroves and seagrass beds (see Fig. 2.1). White Banks (WB) is close to the KL 6m site, but still adjacent to the state park (Fig. 2.1). Biscayne National Park (BNP) is closest to urban Miami, FL. This site is potentially influenced by the extensive agricultural area to th e south and west that drains into Biscayne Bay through a series of watershed canals. BNP also is located rela tively near Turkey Point nuclear power plant and a major landfill in the Black Point area. 6.3.2 Cellular Diagnostic Sampling For cellular diagnostic sampling, a 1-2 cm2 plug of coral tissue was obtained using a leather punch and hammer from each of five separate colonies of the Montastraea annularis complex (Fig. 5.1). Samples were kept in the dark by plac ing them in opaque

PAGE 177

film canisters underwater. On deck, water was removed from the canisters and samples were immediately transferred to a liquid nitroge n dry shipper. Samples were stored at 80 C until analyses were conducted by EnVi rtue Biotechnologies (Winchester, VA, USA). Frozen coral samples were ground w ith a liquid nitrogen-chilled mortar and pestle. A suite of 20 biomarkers (Table 6.1) was assayed during each sampling period. Concentrations of all biomarkers were de termined using enzyme-linked immunosorbent assays (ELISA). Samples were assayed in tr iplicate. Detailed description of the ELISA assays are in Downs (2005) and Downs et al. (2005a). 6.3.3. Data Analysis Data used in the analysis for this chapte r also include those presented in Chapter 2 (environmental assessments) and Chapter 5 (les ion regeneration). Sampling sites, dates and methods, as well as sample processing me thods, were detailed previously in those chapters. 6.3.3.1. Cellular Diagnostic System Data were analyzed in two groups: by sites at 6 m depth along the northeast southwest traverse, and by sites along the dept h gradient. The KL 6 m site was common to both groups. For each cellular biomarker, I calculated descriptive statistics: mean, standard error, range and coefficient of vari ation. For descriptive statistics, biomarkers were organized based on their grouping in one of the four subsystems (Table 6.1). I compared mean levels of selected biomarke rs with stressed and basal levels as defined by Downs et al. (2005a) (Table 6.2). I used these values as a reference for comparison but they should be adjusted as additional information is obtained. I used repeated-measures MANOVA on each biomarker for the entire study period (March 2001 to February 2003) to determine if individual biomarker concentrations varied with site, time and th eir interactions. I ch ecked model assumptions (e.g., sphericity, homogeneity of variances, normality, and independence) using residual plots. Biomarker data were log transforme d to meet these assumptions. To interpret effects detected by MANOVA, I used one-w ay ANOVA followed by the Tukey-Kramer HSD method.

PAGE 178

Similarity matrices were constructed using Euclidean distance for cellular diagnostic data to allow for comparison of samples. Cellular diagnostic data were normalized and log-transformed to minimize the effect of using parameters measured at different scales. To determine if the entire suite of biomarkers differed among sites or sampling periods, I used ANOSIM2 and ANOSIM analyses. I used ANOSIM2 (twoway analysis of similarities) to determine if all biomarkers differed significantly among sites (averaged over the entire study period) and time (average d across all sites). I used ANOSIM (one-way analysis of similarities) to determine if sites differed significantly for each sampling period (i.e., there were site x time interactions) based on all biomarkers. I used Principle Components Analysis (PCA) to interpret the differences found in ANOSIM2 and ANOSIM. 6.3.3.2. Relating Coral Cellular Biomarkers to Coral Regeneration Rates I compared regeneration rates with biom arker concentrations among sites for a given time period using the BEST routine to determine which biomarkers reflected observed trends in coral regeneration. This routine selects the biomarkers that best explain regeneration rates, by maximizing a rank correlati on between their respective resemblance matrices (Clarke & Gorley 2006). The BEST routine is si milar to a stepwise regression but does not assume independence or normality, is non-additive (R2 does not increase with addition of parameters to the f unction), and can handle da ta sets with a high number of parameters and low sample size (Clarke & Gorley 2006). This procedure is based on the weighted Spearman rank coefficient ( w) between the ranked regeneration and CDS similarity matrices. I then used the RELATE routine to determine the significance of the relationship between the tw o similarity matrices based on the set of markers chosen using BEST. Visualizations of this relationshi p were presented by superimposing the cellular biomarker concentr ations on to the regeneration rate multidimensional scaling (MDS) plot and compar ing this with the bubble plot of the regeneration rates.

PAGE 179

6.3.3.3. Linking Environmental Data to Coral Regeneration Rates Regression analyses were used to determ ine if sedimentation rates influenced regeneration rates among the 6 m sites. Se dimentation rates were log transformed to meet assumptions of normality and homoscedasti city. I used data obtained from the Southeast Research Centers Water Quality Monitoring Network (SERC-WQMN) to examine the effects of changes in environmen tal parameters on rege neration rates. Due to the overlap in environmental datasets fo r some study sites (see Ta ble 2.3), comparisons between regeneration rates and SERC water qu ality parameters (Table 2.2) were made within sites to determine which parameters affected changes in re generation rates during the study period. I first tested for collinearity by calculating Pearson correlation coefficient between all pairs of environmental parameters. I then selected parameters that accounted for the majority of the variation based on PCA and were not duplicative. I used the BEST routine to determine which environmental parameters best reflected trends seen in regeneration at each site. Sim ilarity matrices used in this analysis were constructed using Euclidean di stance for regeneration and en vironmental data. The sets of parameters with the largest w were considered to provi de the best match with regeneration rates. Visualiz ation of this relationship wa s presented through bubble plots of selected environmental parameters on the regeneration rate multidimensional scaling (MDS) plot. 6.3.3.4. Analysis Routines ANOSIM2, ANOSIM, PCA, BEST, RELATE, MDS and Pearson correlations were performed using PRIMER v. 6 (Plymout h Routines in Multivariate Ecological Research, PRIMER-E Ltd., Plymouth). All ot her analyses were performed using JMP v.3.2. (SAS Institute Inc., Ca ry, NC, USA) except repeated-measures MANOVA, which was performed using SAS 9.1 (SAS Institute Inc., Cary, NC, USA) with = 0.05 for all hypothesis tests. 6.4. Results and Discussion I utilized data acquired from the cellula r diagnostic system (Fig. 6.1 6.8) to address these three objectives (1) differentiate between global (e.g., temperature) versus

PAGE 180

local (e.g., pollutants) stressors (2) detect subtle and chro nic effects of environmental stress on corals and (3) diagnose coral conditi on at my study sites based on parameters of the cellular diagnostic system and coral col ony responses (e.g., regeneration rates). 6.4.1. Evidence for Temperature or Light Stress? The CDS was applied to distinguish be tween global stressors (e.g., sea-surface temperatures; Downs et al. 2000) and local stressors (e.g., pesticides; Downs et al. 2005a). Characterization of candidate stre ssors includes elimination, diagnosis and strength of evidence (Suter et al. 2002). One of the criteria in causal inference is temporal association (e.g., time order), therefor e it is important to determine when corals were stressed and how that relates to envi ronmental conditions. Time of sampling was the dominant factor affecting the entire suit e of cellular biomarke r levels among the 6 m sites (time: averaged across all 6 m site s; ANOSIM2: Global R = 0.38, p = 0.10%; Table 6.3; site: averaged across all time periods; ANOSIM2: Global R = 0.13, p = 0.10%; Table 6.4; time x site; ANOSIM; Table 6.5) and along the depth gradient (time: averaged across depth gradient; ANOSIM2: Global R = 0.32, p = 0.10%; Table 6.6; site: averaged across all time periods; ANOSIM2: Global R = 0.27, p = 0.10%; Table 6.7; time x site; ANOSIM; Table 6.8). At the 6 m sites, al l individual biomarker levels between 2001 and 2003 were significant with time except for Cn Hsp 60 and Catalase (Repeated Measures MANOVA; Table 6.9). Most in dividual biomarker levels along the depth gradient also were significantly different with time, except for CYP-3, ChlpsHsp, Heme and Catalase (Repeated-measures MANOVA; Table 6.10). A moderate increase in protein, oxidative and metabolic markers is expected in the summer months because coral metabolic rates increase with warmer water temperatures. For example, Hsp 70 and ubiqu itin levels in bivalves increased in summer months, with temperatures normally experienced by these organisms (Hofmann & Somero 1995). Several coral biomarkers (e.g., heat shock proteins, antioxidant enzymes) upregulate further in response to temperatur e stress (Black et al 1995, Lesser 1996, 1997, Fang et al. 1997, Downs et al. 2000, 2002, Brown et al. 2002a,b). For example, in 1999 along the KL depth gradient, high levels of antioxidants and oxidative damage products were observed in Montastraea annularis between May and September. Increases in

PAGE 181

biomarkers corresponded with increases in water temperatures and preceded bleaching that September (Fauth et al. 2003). During my study, high biomarker levels, par ticularly those indicative of protein denaturation and oxidative stress, occurred along the depth gradient at KL 3 m and KL 9 m in August 2001 (Fig. 6.5 6.6) but did not a ppear to be responses to temperatures, which were not abnormally high at that time (30 and 29 C, respectively; Fig. 2.4) and no coral bleaching was observed at any of my study sites throughout my study. Biomarkers levels also were generally not as high as in March and October 2001 (Table 6.11, Fig. 6.9B) when temperatures are cooler. Biomarker leve ls were highest at KL 9 m (Fig. 6.10C), with high levels of all biomarkers. Cnidarian GS T reached stressed levels at KL 3 m (Fig. 6.8D) and Dn GST reached str essed levels at KL 3m and KL 9m (Fig. 6.9E), suggesting that corals were responding to a xenobiotic stress. Xenobiotic markers do not increase in response to a temperature stress. Potential sources of xenobiotic stressors in summer months include increased use of insecticides to control mosquito populations (e.g., dibrom in August; Morgan & Snell 2002) and increased boat use (May through August). High wind speeds were obs erved at the Molasse s Reef buoy in August 2001 (Fig. 2.12), which could have resuspen ded sediments and exposed corals to associated contaminants. In 2000, no sea surface temperature anomalies occurred and biomarker levels no longer correlated with temper ature (Downs et al. 2005a). Instead, corals at BNP responded to a xenobiotic stress in March that resulted in a severe oxidative, metabolic, and protein-denaturing stress (D owns et al. 2005a). Similar biomarker profiles were observed at the 6 m sites during my study; low levels of all biomarkers were observed in summer months June 2001, August 2001 and August 2002, whereas all biomarker levels were high in the spring and fall months of March 2001, October 2001 and February 2003 (Table 6.12, Fig. 6.9B). Temperature can be eliminated as a potential cause of stress because (1) no significant differences in temperature were observed among sites a nd therefore it could not account for intersite differences in cellula r diagnostic markers or regeneration rates, (2) no relationship was observed between chan ges in temperature during the study and changes in regeneration rate s within sites (Tab le 6.13), and (3) temperatures did not

PAGE 182

exceed known stress thresholds during the period of the study. Limited information on light levels (PAR and UV) did not allow for comparisons among sites. However, stressors were generally highest in winter months and no extreme light levels were observed during my study (Fig. 2.16), indicatin g that excess light was most likely not a major cause of stress. Consistent methodol ogy in cellular diagnostic sampling will allow for future comparisons among studies and long time-series comparison to further elucidate seasonal differences in biomarkers a nd establish baseline values for biomarkers. Experimental setups with corals will allo w for comparison of biomarker levels during routine metabolism versus defenses against sp ecific stressors (particularly contaminants such as pesticides and herbicides). 6.4.2. Evidence for Local Xenobiotic Stress Higher biomarker levels during cooler months (March 2001, February 2003) suggest that local conditions rather than te mperature were the source of stress. The highest stress levels at the 6 m sites were observed in March 2001 (Table 6.12) with no significant differences among sites (Table 6.8) High protein turnover and denaturation affecting both the cnidarian host and the sy mbiont (dinoflagellate ) occurred in March 2001 at all 6 m sites as indi cated by significant increases in Cn Hsp70, Dn Hsp70, Dn Hsp60 and ubiquitin (Fig. 6.1; Table 6.12). Cn idarian Hsp70 reached stressed levels and Dn Hsp70 exceeded basal levels at all 6 m sites (Fig. 6.1). Mean Dn Hsp60 reached stressed levels at BNP and basal levels at all 6 m sites (Fig. 6.1). Oxidative stress, particularly in the di noflagellate symbionts, was ev ident in March 2001 at all 6 m sites as indicated by significant elevations in Dn Cu/Zn SOD, Cn Mn SOD, Dn Mn SOD and Dn GPx (Fig. 6.2; Table 6.12). Cnidarian Mn SOD and Dn MnSOD reached stressed levels at BNP in March 2001 (Fig. 6.2). Metabolic condition was compromised in March 2001 as indicated by el evated concentrations of HO and Cn sHsp (Fig 6.3; Table 6.12). Elevations of CYP2, Cn GST, Dn GST and MXR in March 2001 indicated that corals were responding to a xe nobiotic, particularly at KL 6 m and BNP. Cnidarian and dinoflagellate GST reached stressed levels at KL 6 m (Fig. 6.4). In February 2003, the highest biomarker le vels were found at BNP and the lowest at AR (Fig. 6.11). High prot ein turnover and denaturation o ccurred at BNP as indicated

PAGE 183

by stressed levels of Cn Hsp70 and high er ubiquitin levels (Fig. 6.1). Weaker oxidative stress was evident in both the cn idarian host and dinoflagellate symbiont as indicated by significant increases in Dn Cu/Z n SOD, Cn Mn SOD, Cn GPx and Dn GPx; basal levels of Cn Mn SOD, Dn Mn SOD and Dn GPx were exceeded at all 6 m sites (Fig. 6.2; Table 6.12). Metabolic condition wa s compromised as indicated by elevated ferrochelotase, ChlpsHsp and metallothion ein in February 2003 (F ig. 6.3; Table 6.12), most likely in response to oxidative stre ss in both the cnidaria n and dinoflagellate symbiont. A xenobiotic response also was apparent as indicated by elevated concentrations of MXR, Cn GST, Dn GST, CYP-2 and CYP-6 (Fig. 6.4; Table 6.12). Lower biomarker levels at AR in February 2003 suggest that stress levels were lowest at this site. Regeneration rates were highest at AR relative to other sites between November 2002 and February 2003 (Fig. 5.3), supporting this interpretation. Potential stressors during South Florida winters include increased resuspension of bottom sediments (Fig. 2.2) and associated c ontaminants due to higher wind speeds (Fig. 2.11) during winter storms (Kruczyinski & McManus 2002); increased pesticide use by agriculture (Miles & Pfeu ffer 1997, Pfeuffer & Rand 2004) ; and increased tourism (November through April; Krucyznski & McManus 2002). Sedimentation was consistently highest at AR (Fig. 2.2), but was unrelated to regeneration rates when compared among sites; therefore sediment load s alone are an unlikely stressor. Sediment properties, including grain-size and organic an d nutrient content, pl ay a key factor in determining sedimentation stress in corals (W eber et al. 2006). Silt-sized (< 63 m) and organic-rich sediments can stress corals afte r a short-term exposure and are more likely to bind with pollutants including heavy metals an d pesticides (Wu et al. 2004, Weber et al. 2006), whereas sandy, organic-poor sediments have little effect (Weber et al. 2006). Sediments at Key Largo 6 m and BNP were dominated by finer sands and muds, whereas WB and AR are characterized by coarse sediments. Therefore sediments at KL 6 m and BNP are likely more detr imental to coral physiol ogy than at AR and WB. Microbial communities, particularly pathogens, found in the suspended sediment can have detrimental effects on corals ability to repair damage (Hodgson 1990, Kramarsky-Winter 2004). Organic pollutants can bind with fine marine sediments that are ingested by benthic organisms such as corals a nd snails (Anthony et al. 2006).

PAGE 184

Maximum pesticide detections (dominat ed by atrazine, ametryin, bromacil, simazine and norflurazon) in S outh Florida typically occur in winter to late spring with peaks in February and March (Miles & Pfeuffer 1997). Atrazine con centrations in South Florida canals also are typical ly high in November but highly variable (Harman-Fetcho et al. 2005). The largest number of endosulfan de tections (a chemical of concern in Miami Dade County) corresponded with the agri cultural growing season (October/November and March/April; Pfeuffer & Rand 2004, Harman-F etcho et al. 2005). Concentrations of Irgarol 1051, a herbicide comm only found in anti-f ouling paints, were higher in March 2001 than July 2001 in the Miami River (Gardinali et al. 2004). In summary, cellular stress was highest in winter months and was observed at all study sites. Corals were responding to a xeno biotic stress that resulted in increased protein denaturation and turnover, as well as oxidative and metabolic stress. The largest effect was observed at sites with silty se diments (e.g., BNP and KL 6 m), most likely due to the resuspension of sediments and associat ed contaminants during winter storms. Concurrent contaminant analysis of sediments with future st udies should be conducted at these sites to narrow down potential xenobiot ics (e.g., endosulfan, Igrarol 101) that may be affecting corals. Often concentrations of any one contaminant are lower than what is typically thought to have a biological eff ect but a better understanding of chronic exposure to low dose contaminants (e.g., Owen et al. 2003) and the synergistic effect of multiple stressors (e.g., Porter et al. 2001) is needed. 6.4.3. Enhanced Local Effects with Heavy Rainfall High rainfall also can increase input of stressors (e.g., nutrients, viruses, chemicals) to offshore reefs (Lapointe & Ma tzie 1996, Lapointe 1997, Paul et al. 2000). Large increases in nutrients and sedimentation followed large rainfall events in September 2001 and June 2002 (see Ch. 2). Higher biomarker levels in October 2001 and June 2002 may correspond with high rainfall prior to sampling (see Ch. 2, Fig. 2.8). Rainfall was highest in September 2001 as a result of Hurricane Gabrielle that passed over central Florida and was associated with decreased salinity and increased turbidity, inorganic nitrogen, or ganic phosphorous and chlorophylla, particularly in areas near BNP and AR (Fig. 2.5 2.7 ; 2.9). Increases in nutrien ts during this study indicate

PAGE 185

that terrigenous stressors reached my study site s after large rainfall events and therefore nutrients or other land-based stressors may have influenced these corals. High levels of all biomarkers separated BNP and AR from WB and KL 6 m, with the highest stress found at the most northern site (BNP) and the le ast stress at the most southern site (KL 6 m) (Fig. 6.12A). High protein turnover and de naturation affecting both the cnidarian and symbiont occurred in October 2001, partic ularly at BNP and AR, as indicated by significant increases in Cn Hsp70, Dn Hsp70 a nd ubiquitin (Table 6.5). Mean Cn Hsp70 reached stressed levels at all 6 m sites and mean Dn Hsp70 exceeded basal levels at WB, AR and BNP (Fig. 6.1). Oxidative stre ss also was evident in October 2001 as indicated by significant increas es in Dn Cu/Zn SOD, Dn Mn SOD and Dn GPx (Fig. 6.2; Table 6.12). Mean Dn Mn SOD reached stres sed levels at BNP and AR (Fig. 6.2). Cnidarian Cu/Zn SOD was significantly high er in October 2001 at AR relative to previous months (3.98 0.69 vs. 1.45 0.14 Eunits/ng TSP; ANOVA: F3,16 = 15.0, p <0.0001; Fig. 6.3). Elevated concentrations of Heme, metallothionein and Cn sHsp in October 2001 (Fig. 6.3; Table 6.12 ), indicated that an oxidative or xe nobiotic stress was compromising host mitochondria. Corals were responding to a xenobiotic stress as indicated by elevations in CYP-2, Dn GST and MXR (Fig. 6.4; Table 6.12). Regeneration rates between October 2001 and March 2002 were best explained by October 2001 levels of Cn Cu/Zn SOD and me tallothionein (Table 6.14). High levels of Cn Cu/Zn SOD in October 2001 positively co rrelated with regeneration rates between October 2001 and March 2002 (Fig. 6.13C; separa ting coral colonies at AR from all colonies at KL 6 m and most colonies at BNP (Fig. 6.13B-C). This indicates that while corals at AR were stressed, increased levels of antioxidants protecte d these corals from further damage. This also was observed in August 2001 when high levels of Dn Hsp 60 and Dn Cu/Zn SOD were found in the corals with the highest rege neration rates (Fig. 6.14). While most colonies at KL 6 m did not appear to be res ponding to stress in October 2001, high metallothionein in the colony with the lowest regeneration rates (Fig. 6.13D) suggest that a heavy metal exposur e reduced colony performance. Rainfall also was high in June 2002 but wa s not associated with an increase in inorganic nutrients. High concentrations of Cn Hsp 60, Ubiquiti n, Dn Cu/Zn SOD, Dn MnSOD, Cn GPx, Cn MnSOD, Cn GST, CYP-6 and low con centrations of Dn Hsp60,

PAGE 186

ChlpsHsp, CYP-2, CYP-3 separated KL 6 m from all other sites in June 2002 (Fig. 6.12B). Corals were responding to a xenobiotic, particularly at KL 6 m, where elevated concentrations of Cn GST indicated a stre ssed condition (Fig. 6.4). Cytochrome P450 2-class and 3-class were nega tively correlated with concentr ations of CYP-6 in June 2002 at KL 6 m (Fig. 6.12B). Cytochrome P450 6-cl ass is specific to invertebrates and is recognized as a major contributor to insect icide resistance (Feyereisen 1999, irok & Drastichov 2004); it oxidizes (and thus can be upregulated by exposure to) pesticides such as aldrin, dieldrin, diazinon, chlorpyr ifos, deltamethnin and a wide range of pyrethrin-like compounds. Cytochrome P450 2-class and 3-cla ss play a role in steroidogenesis (both sex steroids and choleste rol) and prostaglandin synthesis, but also are important monooxygenases for de toxifying different xenobiotics. Cytochrome P450 3-class was significantly depressed in corals at KL 6 m compared to corals from AR and BNP in June 2002 (0.26 0.07 vs. 0.54 0.0 7 and 0.51 0.07 relative units/ng TSP, respectively; ANOVA: F 3,16 = 3.70, p < 0.04; Fig. 6.4) indica ting that steroid metabolism may have been disrupted. Hi gh protein turnover and denatura tion occurred at KL 6 m in June as indicated by stressed levels of Cn Hsp70 and higher ubiquitin levels (Fig. 6.2). Metabolic condition was compromised as indica ted by elevated concentrations of FC and ChlpsHsp (Fig. 6.3; Table 6.12). The strongest relationship (Rho = 0.43, s.l. = 0.4%) between biomarkers and regeneration rates was observ ed in June 2002, when mean re generation rates at KL 6 m were at their highest for the entire study pe riod (Table 6.14). The combination of five biomarkers (cnidarian Hsp70, sHsp, metalloth ionein, cnidarian GS T, CYP-6) clearly separated colonies at KL 6 m and BNP from those at WB and AR (Fig. 6.15). However, interpretation becomes difficult because th e highest biomarker concentrations were observed both at the site with the highest (KL 6 m) and lo west (BNP) regeneration rates for that period (Fig. 6.15), with the exception of Cn GST, which was high and variable at KL 6 m only. While corals at KL 6 m and BNP generally both had low regeneration rates, in June 2002 regeneration rates at KL 6 m were highest among the 6 m sites (Fig. 5.3). Chemicals can interfere with the basic cellular mechanisms of regeneration, which are likely controlled by steroi d hormones. Prolonged exposure to low

PAGE 187

concentrations of polychlor inated biphenyls (PCBs) affected the physiology of the crinoid ( Antedon mediterranea) by interacting with cellula r mechanisms that regulate growth and cell proliferation /turnover, which resulted in accelerated but abnormal arm regeneration (Candia Carnevali et al. 2001, Barbaglio et al. 2006). High contaminant levels, on the other hand, suppressed growth a nd increased mortality (Barbaglio et al. 2006). Scleractinian corals pro duce a variety of steroids (Tarrant et al. 2001, 2003, 2004, Tarrant 2004, 2005). Abnormal regeneration ra tes at KL 6 m in June 2002 may be a result of endrocrine disrupting (EDs) contaminants such as PCBs, as evidenced by (1) increased levels of Cn GST, CYP-6, and MXR; (2) inhibition of CYP-2 and CYP-3, which are known for steroid metabolism; (3 ) increased levels of protein turnover indicated by increased ubiquitin and Cn Hs p60; and (4) oxidative st ress as indicated by Cn GPx and Dn Cu/Zn SOD. The corallivorous snail ( Coralliophora abbreviata ), which feeds on these corals, also showed evidence of endocrine disruption at KL 6 m in June 2002 as evidenced by severe metabolic and oxi dative distress, high levels of protein turnover and evidence of DNA damage (Woodley et al. unpublished da ta). Xenobiotic stress is implicated by high levels of GST, CYP-3, CYP-6, and MXR (Bard 2000). While White Grunts ( Haemulon plumieri ) did not show cellular evidence of endocrine disruption at KL 6 m, pesticide concentrations in livers of these fish were highest at this site relative to WB and BN P (Downs et al. 2006). High rainfall in June 2002 could have le d to an increase in pollutant exposure, particularly at KL 6 m, which lies off a developed coastline. Insecticide application is highest in late spring to summer months (P ierce 1998) during the warm, wet season when mosquitoes and other insects are most prevalent. Colo red dissolved organic matter (CDOM), which scavenges various organic pollutants and there by decreases th eir toxicity to marine organisms (Coble et al. 2004), tends to be lower at KL 6 m (Ayoub et al. 2006) relative to the AR and WB sites. Ultraviolet (UV-B) radiation levels also peaked in April and May 2002 (Ch 2), which can leave corals and other organisms more sensitive to chemical pollutants (Shick et al 1996, Owen et al. 2003, Jones 2005).

PAGE 188

6.4.4. Are Coral Regeneration Rates and Protein Production Energy-limited? Another explanation for increased regenera tion rates and biomarker levels in June 2002 at KL 6 m could be related to increased food availability in the form of organic nitrogen (Table 6.13; Fig. 6.16). Unusually hi gh regeneration rates also were observed at KL 9 m and 18 m, where organi c nitrogen also was high (Table 6.13; Fig. 6.17 and 6.18). Organic nitrogen increased in th e area of the WB and the Ke y Largo depth gradient prior to the peak in rainfall in April 2002 (Fig. 2.6A and 2.10A). Individual corals can grow heterotrophically by feeding on dissolved and particulate organic matter (Edinger et al. 2000). Alternatively, these corals may ha ve higher zooxanthell ae and chloroplast densities and grow autotrophically usi ng enhanced zooxanthellae photosynthesis fertilized by increased dissolved organic nutri ents (Edinger et al. 2000) Stable isotopes have shown that corals can consume terri genous organic matter or sewage-derived inorganic nitrogen (Mendes et al. 1997). In creases in inorganic nitrogen leads to increased protein synthesis in zooxanthe llae (Muscantine et al. 1989). Nitrogen availability also influences heat shock prot ein production in higher plants, suggesting that Hsp production might be resource-l imited (Heckathorn et al. 1996). Unusually high biomarker concentrations at KL 18 m in November 2002 (Fig. 6.20) followed increases in to tal phosphorus and nitrate (NO3) in October 2002 (Fig. 2.14A, 2.16A). Periodic upwel ling along the shelf edge is a periodic source of phosphorus to offshore reefs (Szmant & Forres ter 1996). Increased nutrients did not correspond to an increase in regeneration ra tes at KL 18 m in November 2002 (Fig. 6.18), which were lowest between August and N ovember 2002 (Fig. 5.3). High levels of protein denaturation and turnove r were evident by increased levels of Cn Hsp60 (Fig. 6.5; Table 6.11). While Cn Hsp70 was generally low in corals at KL 18 m, stressed levels were indicated at KL 18 m in November 2002 (F ig. 6.5). At that time, samples from both KL 3 m and KL 18 m had significantly higher ubiquitin levels than from KL 6 m (521 21 and 613 14 vs. 371 40 fmols/ng TSP, respectively; ANOVA: F3,16 = 8.8, p < 0.002). Oxidative stress was also appare nt at KL 18 m by significantly higher concentration of Cn GPx in corals from KL 18 m than in corals from KL 6 m and KL 9 m (55.5 1.5 vs. 32.8 3.7 and 38.9 2.7 pm ol/ng TSP, respectively; ANOVA: F3,16 = 9.3, p < 0.001). Ferrochelotase at KL 18 m was five times higher in November 2002 than

PAGE 189

June 2001 (0.50 0.08 vs. 0.10 0.05 Eunits/ng TSP; ANOVA: F8,45 = 2.6, p <0.03), likely in response to a need for increased heme oxygen age production for either antioxidant or xenobiotic de toxification pathways. Xenobiotic stress was evident in corals from KL 18 m, with significantly el evated CYP-6 in November 2002 compared to June 2002 (0.53 0.12 vs. 0.10 0.04 re lative units/ng TSP; ANOVA: F4,19 = 3.0, p < 0.05). Stressed levels of Dn GST were exceeded at KL 18 m in November 2002 (Fig. 6.8). Key Largo 3 m and KL 18 m had significantly higher le vels of MXR than KL 6 m (0.74 0.04 and 0.78 0.04 vs. 0.53 0.06 Eunits/ng TSP, respectively; ANOVA: F3,16 = 6.7, p < 0.004). In November 2002, MXR was elevated at KL 3 m relative to June 2001 (0.74 0.04 vs. 0.26 0.10 Eunits/ng TSP, respectively; ANOVA F8,35 = 2.9, p < 0.02) at KL 9 m relative to June 2001 (0.67 0.03 vs. 0.35 0.05 Enits/ng TSP, respectively; ANOVA F8,36 = 5.2, p < 0.0003) and at KL 18 m relative to March 2001, June 2001 and August 2002 (0.78 0.04 vs. 0.23 0.12, 0.16 0.02 and 0.24 0.07 Eunits/ng TSP, respectively; ANOVA F8,35 = 2.9, p < 0.02). Cytochrome P450 2-class was significantly depressed in corals from KL 18 m relative to corals from KL 3 m and KL 9 m in November 2002 (0.12 0.07 vs. 0.57 0.03 and 0.50 0.04 Eunits/ng TSP, respectively; ANOVA: F3,16 = 9.8, p < 0.001). Cytochrome P450 2-class also was significantly higher at KL 3 m in Novemb er 2002 than June 2001 (0.57 0.03 vs. 0.27 0.08 Eunits/ng TSP, respectively; ANOVA: F8,36 = 8..2, p <0.0001) but was not as high as concentrations in Marc h and October 2001 (Fig. 6.8). In summary, periodic increases in nutrien ts preceded increases in regeneration rates and protein production along the depth gr adient in June 2002 and protein production at KL 18 m in November 2002. These results suggest that rege neration and protein production may be energy-limited along the Key Largo depth gradient. While increased nutrient supply can explain increases in prot ein production, it does not explain why corals at these sites were responding to a xenobiotic stress. 6.4.5. Diagnosing Reef Condition Integrating biomarkers and organism respons es can be valuable tools by providing insight into where an organism lies on the heal th status curve (Fig. 6.22); that is, whether an organism is healthy and exhibiting ho meostasis or has initiated compensatory

PAGE 190

responses associated with a decline in h ealth (Depledge et al. 1993). Organisms are generally more sensitive to environmental chan ges during certain stages of development. Candia Carnevali et al. (2001) recommended monitoring adult developmental stages (e.g., regeneration) of crinoids as an indicator of pollution-induced stress. In this study, I compared lesion regeneration rates of the coral, Montastraea annularis with selected environmental parameters and cellular diagnos tic tools of stress. Cellular biomarker levels indicate whether or not an organism is responding to a stress. Monitoring of regeneration rates and densities of symbi ont-bearing (larger) benthic foraminifera (LBF; Chapter 4) indicate if organisms are responding to a st ress at a colony level and a population level, respectively. 6.4.5.1. 6 m Sites Biomarker levels differed significan tly among 6 m sites with significant differences among all sites (Table 6.4). Over all, the highest dissimilarity was between WB and BNP, while biomarker profiles for WB and AR were the most similar (Table 6.5). Among the 6 m sites, samples from BNP had the highest overall means for all biomarkers except Dn Hsp70, Cn GST and CY P-6 (Table 6.15). Samples from WB and KL 6 m tended to have the lowest overall means for biomarkers, however, variability as indicated by the coefficient of variance wa s highest in samples from KL 6 m (Table 6.16). At the 6 m sites, the relationship between regeneration rates and biomarkers was generally low (Table 6.14). This incons istency is a result of high biomarker concentrations at a site that had the lowest re generation rates (BNP) and also at a site that had the highest regeneration rates (AR) (see Ch 5 and 6). On th e other hand, the sites with the lowest biomarker concentrations had low (KL 6 m) to intermediate (WB) regeneration rates. A need for increased protein synthesis or denaturation of damaged proteins was evident at AR and BNP due to significantl y higher Dn Hsp60 than at KL 6 m (Table 6.15). Dn Hsp60 exceeded basal levels at all 6 m sites throughout the study (Fig. 6.1). Ubiquitin is a 76-residue polypeptide that is c onjugated to proteins slated for degradation by the 26S proteosome. Mean ubiquitin wa s above basal levels at all 6 m sites

PAGE 191

throughout the study period but ne ver reached stressed levels (Fig. 6.1). Significantly higher ubiquitin levels at BNP than KL 6 m and WB (Table 6.15) indicates that more proteins were being damaged and targeted fo r rapid degradation. Oxidative stress in the dinoflagellate symbiont at BNP was high as evidenced by significantly higher Dn GPx at BNP than KL 6 m throughout the study (Table 6.15). Corals at KL 6 m, which is located offs hore from the most urbanized coastline in Key Largo, tended to have low biomarker concentrations thr oughout the study period, except for March 2001 and June 2002 when se veral biomarkers (including Cn GST, Ubiquitin, Cn Hsp 60 and sHsp) increased, indi cating a xenobiotic response resulting in protein denaturation and turnover. Biomarker concentrations also were ge nerally low in white grunts ( Haemulon plumieri ) at KL 6 m and WB, while liver concentrations of numerous pesticides (including DDT a nd its metabolites DDD and DDE, oxadiazon, heptachlor, endosulfan sulfate and others) we re highest at KL 6 m relative to WB and BNP (Downs et al. 2006). While biomarker levels of corals were not consistently high at KL 6 m, periodic exposures to xenobiotic(s ) may have created re latively unfavorable conditions for reef organisms (Fig. 6.22) as evidenced by low coral cover (Ch. 3), low densities of symbiont-bearing foraminifera (Ch. 4) and low regeneration rates (Ch. 5). Increased biomarker levels at this site follo wed winter storms (Mar ch 2001 and February 2003), when sediments and associated contamin ants were likely resuspended, and high rainfall in June 2002. Corals at WB generally had low to intermediate levels of all biomarkers throughout the study, with the highest stress ev ident in March 2001. Biomarkers at WB were generally above basal levels but belo w stressed levels as defined by Downs et al. (2005); exceptions include Cn Hsp70 (2001) and ChlpsHsp (2002). Corals at WB appeared to be compensating for stress (Fig. 6.22) as indicated by low coral mortality (Ch. 3), intermediate to high densities of symbiont-bearing forami nifera (Ch. 4) and intermediate to high coral re generation rates (Ch. 5). Corals at AR, which is adjacent to John Pennekamp Coral Reef State Park, had intermediate to high levels of biomarkers, particularly in Marc h and October 2001, with biomarker levels indicating reduced stre ss throughout 2002 and 2003. Reef organisms appeared to tolerate occasional high stress conditions, as occurred in 2001, as evidenced

PAGE 192

by relatively high coral cover (Ch. 3), high de nsities of symbiont-bearing foraminifera (Ch. 4) and consistently high coral regeneratio n rates and percentage of healed lesions (Ch. 5). Unlike KL 6 m, where biomarker and regeneration rates were variable, both regeneration rates and biomarkers were consis tently high at AR. Th erefore, colonies at AR appear to be stressed in response to a xenobiotic stress but protective cellular pathways are allowing the colonies to f unction normally at this site (Fig. 6.22). Contaminant exposures may be too low to ca use overt effects, but over longer time periods may manifest into adverse conditions and mortality (Depledge et al. 1993). Therefore, management should identify and allevi ate potential stressors at this site before they result in further degradat ion of the reef community. This provides an example of where CDS can be used to detect stress prio r to a perceived affect on the organism at higher levels. This site lies along the most protected portion of coas tline, with intact mangroves and wetlands, that provide a consis tent source of colored dissolved organic matter (CDOM) (Ayoub et al. 2006). Colored DOM acts as a scavenger to a variety of trace metals and organic pollutants (Coble et al. 2004) and acts as a sunscreen to ultraviolet radiation. Furthe r investigation is needed to determine the influence of CDOM or other factors that ma y have contributed to higher re generation rates at this site relative to the others. Corals from BNP, offshore from urban Miami, FL, tended to have the highest biomarker levels, particularly those indicat ive of high protein denaturation and turnover (ubiquitin and Dn Hsp60) and oxidative stre ss (Dn Cu/Zn SOD and Dn GPx) in the symbiotic dinoflagellates. Small changes in protein metabolism may result in a negative energy balance and greater variability in physiological performance (Koehne & Bayne 1989, Hawkins 1991). Refolding a pr otein with the help of stress proteins may cost in excess of 100 ATP molecules; therefore, increa sed synthesis of stress proteins can greatly increase energetic costs (Werner & Hinton 1999). Colonies at BNP had considerable recent coral mortality (Ch. 3), low densititie s of symbiont-bearing foraminifera (Ch. 4) and low coral regeneration rates (Ch. 5) indi cating that conditions were unsuitable for reef organisms during the study period. Biom arker levels at BNP indicate that these colonies are likely being stressed by a chemi cal pollutant, which is starting to overwhelm cellular defenses, resulting in physiologica l decline of the organisms (Fig. 6.22).

PAGE 193

Although stress levels were generally lowe r in summer months, some biomarkers (e.g., Cn Hsp 70, Dn MnSOD) remained at str essed levels at BNP in June and August 2001. In August 2001, BNP show ed evidence of oxidative st ress with significantly higher Cn Cu/Zn SOD than WB and KL 6 m. Corals at BNP and AR both responded to a xenobiotic in October 2001. At the same time BNP also experienced oxidative stress. High rainfall in South Florida (associated with Tropical Storm Gabrielle in September 2001) decreased salinity and increased turbidit y and nutrients at water quality stations near AR and BNP (Ch. 2). Watershed canal s flowing into Bisca yne and Florida Bay potentially carry large loads of nutrients a nd pesticides (Harman-Fetcho et al. 2005). In March 2000, corals at BNP experien ced a protein-dena turing stress (as indicated by increased ubiquitin and Hsp70), likely in response to a xenobiotic (e.g., a fungicide, an organometalloid, endosulfan) affecting both the cnidarian host and the symbiotic zooxanthellae (Downs et al. 2005a). Corals likely responded to the xenobiotic by conjugating glutathione to it by glutat hione-s-transferase (GST), and cellular exclusion of the GSH-conjugated xenobiotic by a P-glucoprotein 140/160 pump action (a.k.a. MXR; Downs et al. 2005a). The CDS also indicated that White Grunt ( Haemulon plumieri ) from BNP were exhibiting a xenobiotic response in August 2002 that adversely affected heme metabolism, resulting in e ndocrine disruption a nd elevated protein turnover (Downs et al. 2006). Contaminant levels in fish livers inte rpreted in the context of biomarker response profiles led to four probable suspects includ ing the pesticides hexachlorobenzene, endrin, PCB 105 and Mirex (Downs et al. 2006). 6.4.5.2. Depth Gradient Biomarker levels differed significantly al ong the depth gradient with significant differences among all depths (Table 6.7). La rger differences in biomarker levels were observed along the depth gradient during my study than in 2000, when no difference was seen along the KL depth gradient and overall biomarker levels were low compared to 2001-2003 (Downs et al. 2005a). However, tren ds in biomarker levels did not follow a straightforward depth trend. Instead, KL 3 m and KL 9 m tended to have higher biomarker levels, whereas KL 6 m and KL 18 m tended to have lower but more variable levels (Table 6.17, Table 6.18). Along the dept h gradient, the highest dissimilarity was

PAGE 194

between KL 9 m and KL 18 m and the lowest between KL 9 m and KL 3 m (Table 6.7). Key Largo 18 m had the lowest overall mean s for all biomarkers except for Dn Hsp60, which was low at both KL 6 m and KL 18 m. Pairwise comparisons of the biomarker suite (Table 6.8) revealed that KL 3 m a nd KL 9 m were the most similar, differing significantly only in August 2001. These two sites differed significantly from KL 18 m on at least two-thirds of the sample dates, with no significant diffe rences in June 2001, October 2001, and March 2002. A similar pattern was observed in 1999 in response to a temperature stress. Corals at KL 3 m had high levels of anti oxidant enzymes between May and September, which protected corals at this site from oxi dative damage, as evidenced by low levels of oxidative damage products (lip id peroxide and protein carbon yl) and no bleaching (Fauth et al. 2003). Lipid peroxide (LPO) is a product of oxidati ve damage to cell membranes and protein carbonyl is a product of oxidative dama ge to proteins. Corals at KL 6 m, KL 9 m and KL 18 m had low levels of antioxidant enzymes despite exposure to temperature stress as evidenced by high levels of oxi dative damage products (LPO and protein carbonyl) and bleaching (Fauth et al. 2003). In my study, corals did not appear to be experiencing a temperature stress and no bleaching was observed, but low regeneration rates and increased mortality at KL 6 m, KL 9 m and KL 18 m indicate that these corals were stressed (Fisher et al. in press; Ch. 5). The relations hip between cellular biomarkers and regeneration rates along the depth gradie nt were generally low (Table 6.19). During my study, KL 3 m and KL 9 m bot h had high biomarker concentrations indicative of oxidative and metabolic stress a nd high protein turnover likely as a result of a xenobiotic stress, particularly in March, August and October 2001. However, physiological status at these two sites was ve ry different. Despite exposure to high stress conditions, corals at KL 3 m had consiste ntly high regeneration rates and a large percentage of completely healed lesions (Fig. 6.22). In cont rast, corals from KL 9 m had highly variable regeneration rates, few comple tely healed lesions, extensive mortality and one colony died during the study period. Live coral co ver at KL 3 m (36 %) was approximately five times that of KL 9 m (< 7 %). Corals in deeper water typically receive less radiant energy and therefore may have lo wer carbon reserves than those in shallower water (Nagelberken et al. 1999). Therefor e, increased protein production at KL 9 m

PAGE 195

might have drained energy reserves for othe r physiological proce sses (e.g., regeneration; Fig. 6.22) Corals at the deepest site, KL 18 m, cons istently had very low biomarker levels (except for November 2002) and showed signs of physiological stress, indicating that these corals were in a diseased stat e (Fig. 6.22; Moore et al. 2006); stressors overwhelmed allostatic defenses and reduced th e ability of corals to recover from damage (Downs 2005). Catalase, Dn GPx and CYP-3 were depressed in corals at KL 18 m compared to corals at other depths (Fig. 6.20; Table 6.17), despite high coral mortality and low regeneration rates, indicating that some thing is limiting production of these proteins. The coral, Madracis mirabilis had depressed levels of catalase and cnidarian GPx following exposure to elevated concen trations of Igrarol 2051, an herbicide commonly used in antifouling pa int (Downs & Downs 2007). Concentrations of Igrarol 1051 are a function of boat density and activity with concentrations as high as 182 ng/L in the Key Largo Harbor canals in Se ptember 2001 (Gardinali et al. 2004). Concentrations as low as 63 ng/L can affect carbon uptake in coral symbionts (Owen et al. 2002), however, concentrations <2 km offs hore were generally below detection limits (1 ng/L; Gardinali et al. 2004). Other biomarkers concentra tions remained low at KL 18 m, particularly CYP-2, ChlpsHsp, Dn Hsp 60 and Cn MnSOD. Down -regulation (as well as up-regulation) of stress proteins may indicat e stress (Werner & Hinton 1999) Mechanisms leading to reduced protein levels may involve disruption of protein synthesis at the transcription, translational or post-translat ional (e.g., phosphorylation/deph osphorylation ev ents) level, or may result from pathological effects or reduced energy (ATP) availability (Werner & Hinton 1999). When energy reserves are depleted, organisms may use proteins for additional energy and gl ucose supplies. Increased synthe sis of stress proteins would represent an additional energy burden. This may explain why increas es in total nitrogen at this site were positively correlated with increased regeneration rates. Exposure to hypoxia in clams resulted in e ither no change or reduced levels of antioxidant enzymes and heat s hock proteins even though this st ress ultimately resulted in mortality (Joyner-Matos et al. 2006). Indirect exposure to turf algae can lead to coral mortality through enhanced microbial activity leading to hypoxic conditions (Smith et al.

PAGE 196

2006). Corals along the Key Largo depth gradie nt, particularly KL 6 m, KL 9 m and KL 18 m, experienced increased mortality as a re sult of overgrowth by turf algae. In many cases, corals paled in color prior to death and overgrowth by these al gae. Low levels of antioxidant enzymes and stress pr oteins indicates that either oxidative stress did not play a role in mortality at KL 18 m or severe oxidative stress may have left corals incapable of responding to stress conditions (Werner & Hinton 1999, Joyner-Matos et al. 2006). Probes for hypoxic conditions in the mucus laye r of these corals w ill help test this hypothesis. Additional biomarker assays of damage products (e.g., proteins, membranes, DNA adducts), in conjunction with studies of physiological status (e.g., growth, reproduction, histological examin ation), can help distinguish where corals sit along the stress gradient (healthy vs. stressed vs. di seased, Moore et al. 2006). Based on the evidence above, it would be expected that leve ls of damage products would be high at KL 18 m as observed in 1999 (Downs et al. 2002, Fauth et al. 2005). 6.5. Conclusions Local stressors, specifically xenobiotics, im pacted corals at my study sites, with the highest stress levels in the winter months and following heavy rainfall. No evidence for temperature or light stress wa s observed during this study. This study emphasized the importance of using a hierarch ical, mechanistic approach to assessing reef condition. Reef s with high regeneration, high densities of LBF and intermediate cellular biomarker levels can be considered hea lthy. None of my study sites were considered to be in a wholly healthy state. If biomarker profiles indicate stressed c onditions but no changes are seen at the organismal or population level, this indicates corals are responding to stress but it is not affecting their performance; colonies effectiv ely are compensating to stress. All of the biomarkers measured in this study are involve d in pathways that protect the cell from further damage. Corals at bo th AR and KL 3 m were compen sating to a xenobiotic stress but this did not appear to affect re generation rates duri ng the study period.

PAGE 197

Low coral regeneration rates, high biomarke r levels, and high mortality, with low LBF densities at the sites, i ndicate stress is negatively aff ecting organisms, overwhelming protective cellular pathways and resulting in reduced performance. Corals at both BNP and KL 9 m were responding to a stress that reduced regeneration rates and increased mortality. Stressors were likely recent at BNP as indicated by community assessments, specifically high recent morality and macroalgal biomass. Where both cellular biomarker levels and regeneration rates were low, corals have been severely damaged physiologically and have reached an incurable state (Allen & Moore 2004). Abnormally low cellular biom arkers, low regeneration rates and high mortality at KL 18 m indicated that corals at this site were responding to severe stress, which has left these colonies incap able of recovering from damage.

PAGE 198

Table 6.1. List of biomarkers assayed, repr esenting four cellular subsystems including the sampling period and organism (cnidarian host or algal symbiont) tested for each biomarker. See Appendix 6.1 for a description of each biomarker. Biomarker Sampling Period Cnidarian/Algae Protein Metabolic Condition Heat-shock protein (Hsp70) Cnidarian (Cn) Hsp70 2001 2003 Cnidarian Dinoflagellate (Dn) Hsp70 2001 Algae Heat-shock protein (Hsp60) Cnidarian (Cn) Hsp60 2001 2003 Cnidarian Dinoflagellate (Dn) Hsp60 2001 2003 Algae Ubiquitin 2001-2003 Cnidarian/Algae Oxidative damage and response Copper/Zinc Superoxide Dismutase (Cu/Zn SOD) Cnidarian (Cn) Cu/Zn SOD 2001 Cnidarian Dinoflagellate (Dn) Cu/Zn SOD 2001 2003 Algae Manganese Superoxide Dismutase (Mn SOD) Cnidarian (Cn) MnSOD 2001 2003 Cnidarian Dinoflagellate (Dn) MnSOD 2001 2003 Algae Glutathione Peroxidase (GPx) Cnidarian (Cn) GPx 2001 2003 Cnidarian Dinoflagellate (Dn) GPx 2001 2003 Algae Catalase 2002 2003 Cnidarian Metabolic Condition Heme Oxygenase (HO) 2001 Cnidarian Ferrochelatase (FC) 2001 2003 Cnidarian Metallothionein (Met) 2001 2003 Cnidarian Chloroplast small heat-shock protein (ChlpsHsp) 2001 2003 Algae Invertebrate small heat-shock protein is oforms (sHsp) 2001 2003 Cnidarian Xenobiotic detoxification Cytochrome P450 2E homologue (CYP P450) CYP 450 2 class (CYP-2) 20012003 Cnidarian CYP 450 3 class (CYP-3) 2002 2003 Cnidarian CYP 450 6 class (CYP-4) 2002 2003 Cnidarian Glutathione-S-transferase (GST) Cnidarian (Cn) GST 2001 2003 Cnidarian Dinoflagellate (Dn) GST 2001 2003 Algae Multiple Xenobiotic Resistance Protein (MXR) 2001 2003 Cnidarian/Algae

PAGE 199

Table 6.2. Descriptive statistics for each cel lular biomarker from 2000 (Downs et al. 2005a). Mean values from March 2000 at Bisc ayne National Park represent a "stressed" condition and overall mean values from pooled Key Largo sites represent "basal" condition; indicates reference conditions for this biomarker are unavailable. March 2000 BNP 2000 KL sites Biomarker units "Stressed" "Basal" Protein Metabolic Condition Ubiquitin fmol/ng TSP 968 67 Cnidarian Hsp70 pmol/ng TSP 0.50 0.14 Dino Hsp 70 pmol/ng TSP 1.59 0.18 Cnidarian Hsp60 pmol/ng TSP 16.2 8.8 Dino Hsp60 pmol/ng TSP 0.126 0.014 Oxidative Response Cnidarian Cu/Zn SOD Eunits/ng TSP Dino Cu/Zn SOD Eunits/ng TSP Cnidarian MnSOD pmol/ng TSP 0.086 0.024 Dino MnSOD fmol/ng TSP 2283 553 Cnidarian GPx pmol/ng TSP 171 70 Dino GPx pmol/ng TSP 3.30 0.33 Catalase pmol/ng TSP Metabolic Condition Heme oxygenase Eunits/ng TSP Ferrochelatase Eunits/ng TSP Metallothionein Eunits/ng TSP Cnidarian sHsp Eunits/ng TSP ChlpsHsp Eunits/ng TSP Xenobiotic Response MXR Eunits/ng TSP Cnidarian GST pmol/ng TSP 2.36 0.82 Dino GST pmol/ng TSP 9.72 6.39 CYP450 2-class Eunits/ng TSP CYP450 3-class relative units/ng TSP CYP450 6-class relative units/ng TSP

PAGE 200

Table 6.3. ANOSIM2 results for differences between sampling peri ods based on all cellular biomar kers (averaged across all 6 m sites); Global R = 0.38, 0.1% significance level; Multivariate Dispersion Indices (MVDISP) shown in shaded area for each sampli ng period. Mar-01 Jun-01 Aug-01Oct-01Mar-02Jun-02Aug-02 Nov-02Feb-03 MVDISP 1.24 1.18 1.31 1.53 0.91 0.82 0.96 0.49 0.67 Mar-01 0.36 0.39 n.s. 0.48 0.61 0.63 0.68 0.65 Jun-01 n.s. 0.23 0.39 0.55 0.44 0.61 0.61 Aug-01 0.21 0.43 0.53 0.44 0.58 0.60 Oct-01 0.34 0.52 0.51 0.54 0.51 Mar-02 0.23 0.33 0.39 0.11 Jun-02 0.42 0.16 0.41 Aug-02 0.31 0.36 Nov-02 0.51 Table. 6.4. ANOSIM2 results for differences among 6 m sites based on all cellular biomarkers (a veraged across all sampling peri ods); Global R = 0.13, 0.1% significance level; Mult ivariate Dispersion Indices (MVDISP) s hown in shaded area for each site. KL 6 m WB AR BNP MVDISP 1.13 0.98 0.95 0.97 KL 6 m 0.13 0.17 0.17 WB 0.06 0.18 AR 0.12

PAGE 201

Table 6.5. Global R values of ANOSIM si gnificance tests for differences among 6 m sites based on all cellular biomarkers duri ng each sampling period. No significant differences among sites were observed in March 2001, June 2001, A ugust 2001 or March 2002. The R statistic for pairwise compar ison of 6 m sites based on ANOSIM of all biomarkers during each sampling period also is shown; n.s. represents not significant (> 5%). OCT 2001 JUN 2002 AUG 2002 NOV 2002 FEB 2003 Global R 0.15 0.27 0.26 0.19 0.11 KL 6 m vs. WB n.s. 0.53 0.38 0.20 n.s. KL 6 m vs. AR 0.39 0.44 n.s. n.s. n.s. KL 6 m vs. BNP 0.37 0.31 n.s. n.s. n.s. WB vs. AR n.s. n.s. n.s. n.s. 0.23 WB vs. BNP n.s. 0.17 0.73 0.50 n.s. AR vs. BNP n.s. 0.26 0.34 n.s. 0.20

PAGE 202

Table 6.6. ANOSIM2 results for differences between sampling periods based on all cellular biomarkers (averaged across all depth s); Global R = 0.32, 0.1% significance leve l; Multivariate Dispersion Indi ces (MVDISP) shown in shaded area for each sampling perio d. Mar-01 Jun-01 Aug-01Oct-01Mar-02Jun-02Aug-02 Nov-02Feb-03 MVDISP 1.45 1.08 1.31 1.47 0.98 0.96 0.67 0.64 0.49 Mar-01 0.24 0.15 n.s. 0.32 0.49 0.49 0.57 0.49 Jun-01 n.s. 0.25 0.34 0.49 0.57 0.69 0.58 Aug-01 n.s. 0.33 0.47 0.48 0.61 0.50 Oct-01 0.28 0.49 0.46 0.50 0.47 Mar-02 0.22 0.21 0.41 0.13 Jun-02 0.16 0.35 0.32 Aug-02 0.28 0.11 Nov-02 0.50 Table 6.7. ANOSIM2 results for differences alon g depth gradient based on all cellular biomarkers (averaged across all sampling periods); Global R = 0.27, 0.1% sign ificance level; Multivaria te Dispersion Indices (MVDISP) show n in shaded area for each dept h. KL 3 m KL 6 m KL 9 m KL 18 m MVDISP 0.97 1.05 0.92 1.07 KL 3 m 0.21 0.14 0.46 KL 6 m 0.20 0.18 KL 9 m 0.47

PAGE 203

Table 6.8. Global R values of ANOSIM signifi cance tests for differences among depths ba sed on all cellular biomarkers during e ach sampling period. No significant differences among depths were observed in March 2002. The R statistic for pairwise comparison of depths based on ANOSIM of all biomarkers during each sampling peri od also is shown; n.s. represen ts not significant (> 5%). MAR 2001 JUN 2001 AUG 2001 OCT 2001 JUN 2002 AUG 2002 NOV 2002 FEB 2003 Global R 0.32 0.21 0.37 0.24 0.46 0.30 0.31 0.16 KL 3 m/6 m 0.43 n.s. 0.28 n.s. 0.52 n.s. 0.34 n.s. KL 3 m/9 m n.s. n.s. 0.31 n.s. n.s. n.s. n.s. n.s. KL 3 m/18 m 0.62 0.33 n.s. n.s. 0.83 0.68 0.61 0.52 KL 6 m/9 m n.s. n.s. 0.77 0.45 0.25 n.s. n.s. n.s. KL 6 m/18 m n.s. n.s. n.s. n.s. 0.83 0.30 0.32 0.46 KL 9 m/18 m 0.46 n.s. 0.67 n.s. 0.52 0.45 0.58 0.36

PAGE 204

Table 6.9. Repeated measures MANOVA results for individual cellu lar biomarker levels among 6 m sites; bold values significant p<0.05 (Note: degrees of freedom for site effect: numerator = 3, denomin ator = 16; degrees of freedom for time and site x time effect change with sampling frequency of each biomarker; see Table 6.2) Site Time Site x Time Protein Metabolic Condition F p Num DF Den DF F p Num DF Den DF F p Ubiquitin 3.99 0.0269 8 124 2.8 0.007 24 124 1.28 n.s. Cnidarian Hsp70 2.7 n.s. 8 124 2.72 0.0086 24 124 0.68 n.s. Dino Hsp 70 2.46 n.s. 3 44 18.72 <0.0001 9 44 1.95 n.s. Cnidarian Hsp60 1.32 n.s. 8 124 1.72 n.s. 24 124 0.85 n.s. Dino Hsp60 3.57 0.0377 8 123 6.93 <0.0001 24 123 1.21 n.s. Xenobiotic Response and Damage MDR 1.55 n.s. 8 122 15.89 <0.0001 24 122 1.4 n.s. CYP450 2-class 2.01 n.s. 8 122 10.65 <0.0001 24 122 1.35 n.s. Cnidarian GST 1.75 n.s. 8 124 6.08 <0.0001 24 124 1.00 n.s. Dino GST 0.86 n.s. 8 124 2.36 0.0212 24 124 0.32 n.s. CYP450 3-class 2.7 n.s. 4 64 3.06 0.0227 12 64 2.16 0.0243 CYP450 6-class 0.55 n.s. 4 64 11.29 <0.0001 12 64 0.62 n.s. Metabolic Condition Ferrochelatase 2.66 n.s. 8 124 5.96 <0.0001 24 124 0.72 n.s. Metallothionein 1.50 n.s. 8 124 3.28 0.002 24 124 0.46 n.s. sHsp 6.08 0.0058 8 124 17.2 <0.0001 24 124 3.13 <0.0001 ChlpsHsp 2.02 n.s. 8 123 5.51 <0.0001 24 123 1.09 n.s. Oxidative Damage and Response Heme oxygenase 1.87 n.s. 3 44 6.62 0.0009 9 44 1.89 n.s. Cnidarian Cu/Zn SOD 2.64 n.s. 3 44 9.44 <0.0001 9 44 3.99 0.0009 Dino Cu/Zn SOD 3.43 0.0424 8 124 4.54 <0.0001 24 124 0.86 n.s. Cnidarian MnSOD 1.41 n.s. 7 109 4.17 0.0004 21 109 0.52 n.s. Dino MnSOD 3.14 n.s. 8 124 5.63 <0.0001 24 124 1.08 n.s. Cnidarian GPx 2.55 n.s. 8 124 3.83 0.0005 24 124 0.65 n.s. Dino GPx 4.00 0.0266 8 124 6.53 <0.0001 24 124 1.02 n.s. Catalase 0.41 n.s. 4 64 0.62 n.s. 12 64 0.49 n.s.

PAGE 205

Table 6.10. Repeated measures M ANOVA for individual cellular biomar ker levels along depth gradient Bold values significant p<0.05; (Note: degrees of freedom for site effect: numerator = 3, denominator = 16; degrees of freedom for time and site x time effect change with sampling frequency of each biomarker; see Table 6.2) Site Time Site x Time Protein Metabolic Condition F p Num DF Den DF F p Num DF Den DF F P Ubiquitin 6.92 0.0034 8 122 2.85 0.0061 24 122 1.61 0.0487 Cnidarian Hsp70 10.00 0.0006 8 122 2.30 0.0246 24 122 1.79 0.0211 Dino Hsp 70 5.19 0.0108 3 44 9.02 <0.0001 9 44 2.20 0.04 Cnidarian Hsp60 11.38 0.0003 8 122 2.53 0.0139 24 122 1.58 n.s. Dino Hsp60 8.70 0.0012 8 122 7.16 <0.0001 24 122 1.29 n.s. Xenobiotic Response and Damage MDR 10.83 0.0004 8 121 11.63 <0.0001 24 121 2.10 0.0047 CYP450 2-class 15.26 <0.0001 8 121 8.40 <0.0001 24 121 1.90 0.0125 Cnidarian GST 11.54 0.0003 8 122 5.72 <0.0001 24 122 2.22 0.0025 Dino GST 4.76 0.0148 8 124 2.76 0.0077 24 124 1.11 n.s. CYP450 3-class 21.75 <0.0001 4 62 1.20 n.s. 12 62 1.60 n.s. CYP450 6-class 10.54 0.0005 4 62 4.01 0.0059 12 62 2.13 0.0272 Metabolic Condition Ferrochelatase 9.93 0.0006 8 122 4.39 0.0001 24 122 1.96 0.0094 Metallothionein 12.03 0.0002 8 122 2.34 0.0224 24 122 1.59 n.s. sHsp 5.19 0.0108 8 122 18.86 <0.0001 24 124 2.16 0.0034 ChlpsHsp 7.94 0.0018 8 122 1.79 n.s. 24 123 1.56 n.s. Oxidative Damage and Response Heme 5.17 0.0109 3 44 2.05 n.s. 9 44 2.15 0.0447 Cnidarian Cu/Zn SOD 6.77 0.0037 3 44 3.21 0.0321 9 44 2.37 0.0279 Dino Cu/Zn SOD 5.27 0.0102 8 122 5.18 <0.0001 24 122 1.63 0.0443 Cnidarian MnSOD 13.97 <0.0001 8 115 2.85 0.0064 22 115 2.15 0.0049 Dino MnSOD 12.22 0.0002 8 122 4.38 0.0001 24 122 2.38 0.0011 Cnidarian GPx 10.03 0.0006 8 122 5.25 <0.0001 24 122 1.8 0.0207 Dino GPx 8.67 0.0012 8 122 6.41 <0.0001 24 122 2.11 0.0045 Catalase 14.53 <0.0001 4 62 0.94 n.s. 12 62 1.22 n.s.

PAGE 206

Table 6.11. Overall mean ( SE) biomarker co ncentrations (averaged acros s all depths) for each sampling period. Shaded values represent biomarkers that did not vary significantly with time or time x site interactions; represent biomarkers that varied significantly with time but not time x site interactions; means for each biomarker that are not connected by the same letter di ffered significantly based on Tukey HSD test (p < 0.05); represents biomarkers that were not sampled during that time period. F Num Den p-value Mar-01 Jun-01 Aug-01 Oct-01 Mar-02 Jun-02 Aug-02 Nov-02 Feb-03 Protein Metabolic Condition Ubiquitin 616 317 322 452 344 416 388 491 410 (125) (80) (51) (135) (48) (33) (25) (24) (20) Cnidarian Hsp70 0.85 0.44 0.69 0.74 0.38 0.36 0.35 0.46 0.39 (0.13) (0.09) (0.11) (0.13) (0.04) (0.04) (0.03) (0.02) (0.03) Dino Hsp 70 0.26 0.13 0.18 0.25 (0.03) (0.03) (0.03) (0.04) Cnidarian Hsp60* 2.3 8 163 0.0224 40AB 22A 29AB 36AB 27AB 35AB 30AB 39B 31AB (4) (2) (5) (3) (4) (3) (2) (2) (2) Dino Hsp60* 9.8 8 166 <0.0001 0.137C 0.071AB 0.096BC 0.096BC 0.027A 0.024A 0.029A 0.026A 0.029A (0.027) (0.016) (0.019) (0.020) (0.003) (0.004) (0.002) (0.003) (0.002) Oxidative Damage and Response Cnidarian Cu/Zn SOD 2.8 1.5 2.1 2.2 (0.5) (0.2) (0.4) (0.3) Dino Cu/Zn SOD 0.44 0.22 0.29 0.33 0.39 0.50 0.42 0.50 0.43 (0.07) (0.04) (0.05) (0.06) (0.05) (0.05) (0.02) (0.02) (0.02) Cnidarian MnSOD 0.076 0. 046 0.082 0.055 0.056 0.045 0.051 0.041 0.061 (0.010) (0.011) (0.016) (0.011) (0.007) (0.006) (0.003) (0.005) (0.004) Dino MnSOD 2310 786 2105 1632 642 785 872 820 756 (381) (189) (561) (274) (83) (59) (142) (50) (42) Cnidarian GPx 32 19 25 26 30 36 34 43 36 (4) (3) (3) (3) (4) (4) (2) (2) (2) Dino GPx 2.00 0.73 1. 14 1.33 0.71 0.33 0.57 0.46 0.81 (0.37) (0.30) (0.30) (0.25) (0.09) (0.05) (0.04) (0.02) (0.04) Catalase 16.7 15.9 15.2 16.1 18.0 (2.2) (2.6) (1.7) (2.3) (1.1)

PAGE 207

Table 6.11. (cont.) Overall mean ( SE) biomarker concentrations (averaged across all depths) for each sampling period. Shaded values represent biomarkers that did not va ry significantly with time or time x site interactions; represent biomarkers that varied significantly with time but not time x site interactions; means for each biomarker that are not connected by the same letter di ffered significantly based on Tukey HSD test (p < 0.05); represents biomarkers that were not sampled during that time period. F Num Den p-value Mar-01 Jun-01 Aug-01 Oct-01 Mar-02 Jun-02 Aug-02 Nov-02 Feb-03 Metabolic Condition Heme oxygenase 0.48 0.26 0.37 0.40 (0.07) (0.05) (0.08) (0.06) Ferrochelatase 0.41 0.22 0.50 0.41 0. 34 0.42 0.38 0.48 0.42 (0.05) (0.05) (0.12) (0.07) (0.05) (0.04) (0.03) (0.03) (0.02) Metallothionein* 2.2 8 166 0.0310 0.54AB 0.26A 0.40AB 0.86B 0.36AB 0.29AB 0.35AB 0.33AB 0.41AB (0.09) (0.06) (0.07) (0.36) (0.05) (0.03) (0.02) (0.01) (0.02) sHsp 0.055 0.026 0.045 0.045 0.010 0.006 0.005 0.014 0.011 (0.008) (0.006) (0.006) (0.005) (0.001) (0.001) (0.001) (0.001) (0.001) ChlpsHsp 0.45 0.23 0.30 0.60 0.80 0.68 0.72 0.72 0.85 (0.06) (0.05) (0.05) (0.17) (0.10) (0.12) (0.10) (0.11) (0.07) Xenobiotic Response and Damage MXR 0.72 0.24 0.37 0.49 0.50 0.59 0.51 0.68 0.59 (0.11) (0.04) (0.05) (0.08) (0.07) (0.04) (0.05) (0.03) (0.04) Cnidarian GST 2.71 0. 66 1.42 1.58 1.20 1.89 1.26 1.59 1.41 (0.59) (0.13) (0.37) (0.27) (0.16) (0.49) (0.10) (0.07) (0.08) Dino GST* 2.7 8 165 0.0086 12.0B 9.6AB 10.9AB 12.4B 8.0A 9.7AB 9.8AB 11.7B 9.8AB (0.8) (0.7) (1.0) (1.0) (1.2) (0.8) (0.5) (0.7) (0.6) CYP450 2-class 0.92 0.19 0.28 0.74 0.37 0.33 0.39 0.38 0.41 (0.29) (0.04) (0.05) (0.14) (0.05) (0.05) (0.03) (0.05) (0.03) CYP450 3-class 0.50 0.47 0.58 0.54 0.58 (0.07) (0.08) (0.05) (0.08) (0.05) CYP450 6-class 0.54 0.42 0.48 0.49 0.66 (0.07) (0.05) (0.03) (0.04) (0.03)

PAGE 208

Table 6.12. Overall mean ( SE) cellular biomarker concentrati ons (averaged across all 6 m site s) for each sampling period. Sh aded values represent biomarkers that did not vary significantly with time or time x site interactions. represent biomarkers that varied significantly with time (ANOVA statistics shown) but not time x site interactions; means for each biomarker that are not connec ted by the same letter differed significantly based on Tukey HSD (p < 0.05). re presents biomarkers that were not sampled during tha t time period; units as shown in Table 6.2. F Num Den p-value Mar-01 Jun-01 Aug-01 Oct-01 Mar-02 Jun-02 Aug-02 Nov-02 Feb-03 Protein Metabolic Condition Ubiquitin* 3.9 8 164 0.0003 620C 320AB 309A 494BC 396BC 434BC 295ABC 381ABC 436BC (96) (38) (56) (76) (46) (25) (29) (12) (27) Cnidarian Hsp70* 6.2 8 167 <0.0001 0.73BC 0.48AB 0.54ABC 0.83BC 0.41A 0.41A 0.36A 0.32A 0.44AB (0.08) (0.08) (0.07) (0.14) (0.04) (0.02) (0.02) (0.02) (0.03) Dino Hsp 70* 9.9 3 72 <0.0001 0.28B 0.12A 0.14A 0.28B (0.03) (0.02) (0.02) (0.04) Cnidarian Hsp60 38 27 27 36 32 38 28 31 34 (3) (3) (3) (4) (4) (1) (2) (2) (2) Dino Hsp60* 15.2 8 167 <0.0001 0.115C 0.066B 0.054AB 0.082BC 0.029A 0.030A 0.029A 0.029A 0.032AB (0.012) (0.016) (0.008) (0.011) (0.003) (0.002) (0.001) (0.002) (0.001) Oxidative Damage and Response 2.1 1.4 1.2 2.8 Cnidarian Cu/Zn SOD (0.3) (0.2) (0.2) (0.4) Dino Cu/Zn SOD* 6.3 8 167 <0.0001 0.48B 0.22A 0.25A 0.40B 0.47B 0.53B 0.37AB 0.43B 0.49B (0.05) (0.03) (0.06) (0.08) (0.05) (0.02) (0.03) (0.01) (0.02) Cnidarian MnSOD* 5.8 7 149 <0.0001 0.070C 0.034A 0.042AB 0.065BC 0.052ABC 0.052ABC 0.044ABC 0.074C (0.009) (0.010) (0.007) (0 .007) (0.002) (0.003) (0.003) (0.003) Dino MnSOD* 5.3 8 164 <0.0001 2050C 1602ABC 706A 1924BC 748ABC 771AB 673AB 714AB 820AB (221) (540) (110) (225) (82) (34) (38) (24) (38) Cnidarian GPx* 8.6 8 167 <0.0001 31AB 23AB 20A 31AB 35CD 41CD 32BC 35CD 43D (2) (3) (2) (3) (4) (2) (2) (1) (2) Dino GPx* 11.8 8 167 <0.0001 2.19D 0.70AB 0.79AB 2.27CD 0.81ABC 0.40A 0.65AB 0.40AB 0.93BC (0.28) (0.12) (0.19) (0.91) (0.09) (0.03) (0.04) (0.01) (0.04) Catalase 18.8 20.0 15.8 17.7 20.4 (2.2) (1.4) (1.1) (1.1) (1.3)

PAGE 209

Table 6.12 (cont.). Overall mean ( SE) cellular biomarker concen trations (averaged across all 6 m sites) for each sampling per iod. Shaded values represent biomarkers that did not vary significantly with time or time x site interactions. represent biomarke rs that varied significantly with time (ANOVA statisti cs shown) but not time x site interacti ons; means for each biomarker that are not connected by the same letter differed significantly based on Tukey HSD (p < 0.05). represents biomarkers that were not sampl ed during that time period; un its as shown in Table 6.2. F p-value Mar-01 Jun-01 Aug-01 Oct-01 Mar-02 Jun-02 Aug-02 Nov-02 Feb-03 Metabolic Condition Heme oxygenase* 5.5 3 72 0.0019 0.45B 0.27AB 0.22A 0.42B (0.05) (0.04) (0.04) (0.07) Ferrochelatase* 5.9 8 167 <0.0001 0.40BC 0.26AB 0.19A 0.38A 0.40BC 0.60C 0.36ABC 0.41BC 0.48C (0.04) (0.05) (0.03) (0.06) (0.05) (0.14) (0.02) (0.02) (0.02) Metallothionein* 5.0 8 167 <0.0001 0.43ABC 0.23A 0.24A 0.60C 0.42ABC 0.31ABC 0.36ABC 0.30AB 0.49BC (0.06) (0.03) (0.04) (0.15) (0.05) (0.01) (0.02) (0.01) (0.02) sHsp 0.039 0.017 0.026 0.047 0.010 0.006 0.006 0.007 0.014 (0.006) (0.003) (0.003) (0.006) (0.001) (0.002) (0.001) (0.001) (0.001) ChlpsHsp* 15.9 8 167 <0.0001 0.40AB 0.22A 0.15A 0.42AB 0.92C 0.84C 0.58BC 0.84C 0.90C (0.04) (0.05) (0.02) (0.07) (0.10) (0.08) (0.09) (0.06) (0.08) Xenobiotic Response and Damage MXR* 11.7 8 167 <0.0001 0.61B 0.19A 0.25A 0.53B 0.59B 0.58B 0.47B 0.55B 0.69B (0.08) (0.03) (0.04) (0.07) (0.07) (0.03) (0.04) (0.02) (0.03) Cnidarian GST* 6.6 8 167 <0.0001 1.92D 0.70A 0.81AB 0.95ABC 1.40BCD 1.96D 1.17ABCD 1.39CD 1.51CD (0.49) (0.12) (0.15) (0.14) (0.15) (0.48) (0.09) (0.05) (0.09) Dino GST* 3.5 8 167 0.0008 12.3B 11.0B 8.6AB 11.5B 9.3A 10.1AB 9.3AB 9.9AB 11.3B (0.6) (0.5) (0.6) (0.9) (1.1) (0.5) (0.5) (0.3) (0.5) CYP450 2-class* 7.6 8 167 <0.0001 0.53D 0.23AB 0.20A 0.64D 0.43BCD 0.40BCD 0.29ABC 0.40BCD 0.50CD (0.06) (0.06) (0.04) (0.12) (0.05) (0.03) 0.04 (0.02) (0.02) CYP450 3-class 0.60 0.60 0.47 0.61 0.73 (0.07) (0.05) (0.06) (0.04) (0.03) CYP450 6-class* 12.1 4 95 <0.0001 0.65B 0.47AB 0.51AB 0.42A 0.82C (0.07) (0.02) (0.03) (0.02) (0.04)

PAGE 210

Table 6.13. BEST results for coral regeneration rates and SERC environmental parameters. Abbreviations include alkaline phosphatase activity (APA), chlorophyll-a (CHLA), dissolved inorganic nitrogen (D IN), dissolved oxygen (DO) nitrate (NO3), nitrite (NO2), soluble reactive phosphate (SRP), to tal nitrogen (TN), total organic carbon (TOC), total organic nitrogen (TON) and total phosphorus (TP) Site Rho # variables Selected environmental parameters KL 3 m 0.60 5 NO3, DIN, TOC, salinity, DIN:TP KL 6 m 0.81 1 TN KL 9 m 0.75 4 NO2, CHLA, turbidity, TN:TP KL 18 m 0.80 5 NO2, SRP, CHLA, TOC, TN:TP WB 0.68 1 TP AR 0.38 4 TON, TOC, TN:TP DIN:TP BNP 0.19 3 NO2, APA, DO

PAGE 211

Table 6.14. BEST and RELATE results for coral regeneration rates and biomarker concentrations dur ing a given time period at the 6 m Sites; n.s. represents not significantly (> 5%) Regeneration Rate Biomarker Rho Significance level # variables Selected Biomarkers JUN AUG 2001 JUN 2001 0.16 n.s. AUG OCT 2001 AUG 2001 0.30 1% 2 Dn Hsp60, Dn Cu/Zn SOD OCT 2001 MAR 2002 OCT 2001 0.24 3.1% 2 Met, Cn Cu/Zn SOD MAR JUN 2002 MAR 2002 0.09 n.s. JUN AUG 2002 JUN 2002 0.43 0.4% 5 Cn Hsp70, sHsp, Met, Cn GST, CYP-6 AUG NOV 2002 AUG 2002 0.17 3.6% 2 Ubiquitin, Cn MnSOD NOV 2002 FEB 2003 NOV 2002 0.11 n.s.

PAGE 212

Table 6.15. Descriptive statistics for each cellular biomarker at the 6 m sites over the entire study period between March 200 1 and February 2003. represents biomarkers that differed significantly among sites but not wi th time x site interactions. Means of biomarkers not connected by the same letter differed significantly based on Tukey HSD test (p < 0.05) Overall (all 6 m sites) Overall mean ( SE) Biomarker units n Max Min Mean SE Median KL 6 m WB AR BNP Protein Metabolic Condition Ubiquitin* fmol/ng TSP 176 2121 -59 409 18 399 382A 359A 374AB 517B (52) (25) (28) (32) Cnidarian Hsp70 pmol/ng TSP 176 2.40 -0.01 0.50 0.02 0.42 0.47 0.44 0.48 0.60 (0.06) (0.04) (0.04) (0.06) Dino Hsp 70 pmol/ng TSP 76 0.62 0.01 0.21 0.02 0.19 0.13 0.18 0.26 0.25 (0.03) (0.03) (0.03) (0.03) Cnidarian Hsp60 pmol/ng TSP 176 86.4 -3.7 32.5 0.9 33.9 31.1 30.4 32.3 35.9 (2.6) (1.8) (1.7) (1.4) Dino Hsp60* pmol/ng TSP 176 0.313 -0.001 0.051 0.003 0.034 0.033A 0.040AB 0.061B 0.065B (0.004) (0.005) (0.009) (0.007) Oxidative Damage and Response Cnidarian Cu/Zn SOD Eunits/ng TSP 76 6.62 0.08 1.84 0.16 1.49 1.24 1.44 2.08 2.49 (0.37) (0.25) (0.31) (0.26) Dino Cu/Zn SOD Eunits/ng TSP 176 1.34 -0.07 0.41 0.02 0.42 0.37 0.36 0.43 0.46 (0.04) (0.03) (0.04) (0.03) Cnidarian MnSOD pmol/ng TSP 157 0.186 -0.007 0.054 0.002 0.054 0.048 0.052 0.056 0.060 (0.005) (0.005) (0.005) (0.004) Dino MnSOD fmol/ng TSP 176 10104 -61 1096 78 815 910 920 1128 1412 (126) (96) (124) (230) Cnidarian GPx pmol/ng TSP 176 59 -3 32 1 35 30 30 34 35 (3) (2) (2) (1) Dino GPx* pmol/ng TSP 176 17.98 -0.08 1.01 0.12 0.73 0.71A 0.78AB 1.04AB 1.48B (0.14) (0.09) (0.12) (0.40) Catalase pmol/ng TSP 100 30.6 -2.5 18.5 0.7 19.4 17.9 18.7 17.8 19.8 (1.8) (1.2) (1.2) (1.1)

PAGE 213

Table 6.15 (cont.). Descriptive statistics for each cellular biom arker at the 6 m sites over the entire study period between M arch 2001 and February 2003. represents biomarkers that differed significantly among sites but not with time x site interactions. Mean s of biomarkers not connected by the same letter differed significantly based on Tukey HSD test (p < 0.05) Overall (all 6 m sites) Overall mean ( SE) Biomarker units n Max Min Mean SE Median KL 6 m WB AR BNP Metabolic Condition Heme oxygenase Eunits/ng TSP 76 1.01 0.01 0.34 0.03 0.30 0.24 0.29 0.37 0.45 (0.06) (0 .04) (0.05) (0.05) Ferrochelatase Eunits/ng TSP 176 3.27 -0.04 0.39 0.02 0.39 0.35 0.32 0.38 0.50 (0.03) (0.02) (0.02) (0.07) Metallothionein Eunits/ng TSP 176 2.93 -0.04 0.38 0.02 0.35 0.38 0.35 0.38 0.39 (0.07) (0 .03) (0.02) (0.03) Cnidarian sHsp Eunits/ng TSP 176 0.134 -0.001 0.019 0.002 0.013 0.018 0.015 0.020 0.023 (0.004) (0.002) (0.003) (0.003) ChlpsHsp Eunits/ng TSP 176 1.35 -0.11 0.59 0.03 0.56 0.51 0.56 0.60 0.68 (0.07) (0.06) (0.06) (0.06) Xenobiotic Response and Damage MXR Eunits/ng TSP 176 1.65 -0.06 0.50 0.02 0.54 0.52 0.44 0.50 0.54 (0.05) (0.04) (0.03) (0.04) Cnidarian GST pmol/ng TSP 176 11.04 -0.08 1.32 0.09 1.37 1.56 1.15 1.13 1.45 (0.34) (0.09) (0.06) (0.10) Dino GST pmol/ng TSP 176 17.00 -0.92 10.34 0.23 10.53 9.74 10.09 10.31 11.18 (0.56) (0.46) (0.38) (0.42) CYP450 2-class Eunits/ng TSP 176 2.21 -0.03 0.40 0.02 0.42 0.33 0.39 0.39 0.49 (0.04) (0.04) (0.03) (0.05) CYP450 3-class relative units/ng TSP 100 1.01 -0.06 0.60 0.02 0.66 0.55 0.55 0.62 0.68 (0.06) (0.05) (0.04) (0.04) CYP450 6-class relative units/ng TSP 100 1.14 -0.04 0.57 0.02 0.56 0.58 0.53 0.60 0.58 (0.04) (0.04) (0.05) (0.05)

PAGE 214

Table 6.16. Coefficient of variance for each cel lular biomarker at the 6 m sites over the entire study period between March 2001 and February 2003. Biomarker units n KL 6 m WB AR BNP Protein Metabolic Condition Ubiquitin* fmol/ng TSP 176 87 47 50 42 Cnidarian Hsp70 pmol/ng TSP 176 79 61 56 63 Dino Hsp 70 pmol/ng TSP 76 100 61 58 48 Cnidarian Hsp60 pmol/ng TSP 176 53 38 35 26 Dino Hsp60* pmol/ng TSP 176 79 88 93 72 Oxidative Damage and Response Cnidarian Cu/Zn SOD Eunits/ng TSP 76 119 77 68 46 Dino Cu/Zn SOD Eunits/ng TSP 176 68 50 56 43 Cnidarian MnSOD pmol/ng TSP 157 58 54 55 45 Dino MnSOD fmol/ng TSP 176 88 70 74 109 Cnidarian GPx pmol/ng TSP 176 53 43 32 29 Dino GPx* pmol/ng TSP 176 128 77 80 180 Catalase pmol/ng TSP 100 50 32 33 27 Metabolic Condition Heme oxygenase Eunits/ng TSP 76 100 62 65 56 Ferrochelatase Eunits/ng TSP 176 57 50 42 92 Metallothionein Eunits/ng TSP 176 121 57 45 46 Cnidarian sHsp Eunits/ng TSP 176 133 107 105 83 ChlpsHsp Eunits/ng TSP 176 90 73 65 57 Xenobiotic Response and Damage MXR Eunits/ng TSP 176 63 55 42 50 Cnidarian GST pmol/ng TSP 176 139 51 37 44 Dino GST pmol/ng TSP 176 37 30 25 25 CYP450 2-class Eunits/ng TSP 176 82 67 49 65 CYP450 3-class relative units/ng TSP 100 53 49 29 26 CYP450 6-class relative units/ng TSP 100 38 42 43 41

PAGE 215

Table 6.17. Descriptive statistics for each cellular biomarker al ong the depth gradient over the entire study period between Ma rch 2001 and February 2003. represents biomarkers that differed si gnificantly among sites but not w ith time x site interactions. Means of biomarkers not connected by the same letter diffe red significantly based on T ukey HSD test (p < 0.05) Overall (along depth gradent) Overall Mean Biomarker units n Max Min Mean SE Median KL 3 m KL 6 m KL 9 m KL 18 m Protein Metabolic Condition Ubiquitin fmol/ng TSP 175 2690 -62.5 417 24 402 482 382 516 282 (44) (52) (55) (33) Cnidarian Hsp70 pmol/ng TSP 175 2.06 -0.02 0.51 0.03 0.44 0.68 0.47 0.60 0.29 (0.06) (0.06) (0.06) (0.04) Dino Hsp 70 pmol/ng TSP 76 0.66 0.01 0.20 0.02 0.19 0.29 0.13 0.26 0.127 (0.04) (0.03) (0.02) (0.03) Cnidarian Hsp60* pmol/ng TSP 175 108.1 -5.0 32.1 1.1 34.1 36.6C 31.1B 37.3BC 23.1A (1.6) (2.6) (2.1) (2.2) Dino Hsp60* pmol/ng TSP 175 0.424 -0.001 0.059 0.005 0.033 0.072B 0.033A 0.093B 0.035A (0.010) (0.004) (0.015) (0.008) Oxidative Damage and Response Cnidarian Cu/Zn SOD Eunits/ng TSP 76 6.18 0.08 2.13 0.19 1.86 3.07 1.24 3.08 0.96 (0.36) (0.37) (0.25) (0.21) Dino Cu/Zn SOD Eunits/ng TSP 175 1.03 -0.07 0.39 0.02 0.42 0.43 0.37 0.49 0.28 (0.03) (0.04) (0.02) (0.03) Cnidarian MnSOD pmol/ng TSP 166 0.284 -0.009 0.057 0.003 0.056 0.074 0.048 0.071 0.035 (0.007) (0.005) (0.004) (0.006) Dino MnSOD fmol/ng TSP 175 9162 -82 1188 97 837 1583 910 1592 630 (247) (126) (215) (89) Cnidarian GPx pmol/ng TSP 175 60.0 -4.1 31.4 1.1 34.0 36.0 30.0 36.5 22.3 (1.6) (2.6) (1.2) (2.4) Dino GPx* pmol/ng TSP 175 5.83 -0.09 0.89 0.08 0.62 1.28 0.71 1.05 0.51 (0.19) (0.14) (0.14) (0.08) Catalase* pmol/ng TSP 99 29.0 -3.1 16.4 0.9 19.6 20.2 17.9 20.6 6.37 (1.2) (1.8) (1.0) (1.6)

PAGE 216

Table 6.17 (cont.). Descriptive statistics for each cellular biom arker along the depth gradient over the entire study period b etween March 2001 and February 2003. represents bi omarkers that differed significantly among sites but not with time x site interact ions. Means of biomarkers not connected by the same letter differed significantly based on Tukey HSD test (p < 0.05) Overall (along depth gradent) Overall Mean Biomarker units n Max Min Mean SE Median KL 3 m KL 6 m KL 9 m KL 18 m Metabolic Condition Heme oxygenase Eunits/ng TSP 76 1.55 0.01 0.38 0.03 0.35 0.48 0.24 0.57 0.20 (0.07) (0.06) (0.04) (0.05) Ferrochelatase Eunits/ng TSP 175 2.29 -0.06 0.40 0.02 0.41 0.53 0.35 0.46 0.24 (0.05) (0.03) (0.03) (0.03) Metallothionein* Eunits/ng TSP 175 6.75 -0.06 0.42 0.04 0.36 0.59B 0.38AB 0.49B 0.22A (0.14) (0.07) (0.03) (0.03) Cnidarian sHsp Eunits/ng TSP 175 0.134 -0.001 0.024 0.002 0.013 0.026 0.018 0.033 0.017 (0.004) (0.004) (0.005) (0.003) ChlpsHsp* Eunits/ng TSP 175 3.03 -0.11 0.60 0.04 0.60 0.83 0.51 0.77 0.27 (0.08) (0.07) (0.05) (0.05) Xenobiotic Response and Damage MXR Eunits/ng TSP 175 1.65 -0.07 0.52 0.02 0.57 0.59 0.52 0.65 0.33 (0.04) (0.05) (0.03) (0.04) Cnidarian GST pmol/ng TSP 175 11.04 -0.14 1.53 0.11 1.43 1.93 1.56 1.73 0.87 (0.21) (0.34) (0.10) (0.13) Dino GST* pmol/ng TSP 175 19.64 -1.07 10.42 0.29 10.80 11.6AB 9.74AB 11.87B 8.38A (0.50) (0.56) (0.41) (0.65) CYP450 2-class Eunits/ng TSP 175 4.56 -0.06 0.44 0.04 0.51 0.50 0.33 0.73 0.19 (0.04) (0.04) (0.13) (0.03) CYP450 3-class* relative units/ng TSP 99 0.96 -0.09 0.53 0.03 0.63 0.68B 0.55B 0.68B 0.21A (0.04) (0.06) (0.03) (0.05) CYP450 6-class relative units/ng TSP 99 0.98 -0.05 0.51 0.02 0.56 0.54 0.58 0.58 0.36 (0.03) (0.04) (0.03) (0.05)

PAGE 217

Table 6.18. Coefficient of variance for each ce llular biomarker along the depth gradient over the entire study period betw een March 2001 and February 2003. Biomarker units n KL 3 m KL 6 m KL 9 m KL 18 m Protein Metabolic Condition Ubiquitin fmol/ng TSP 175 61 87 71 79 Cnidarian Hsp70 pmol/ng TSP 175 57 79 63 103 Dino Hsp 70 pmol/ng TSP 76 59 100 38 98 Cnidarian Hsp60 pmol/ng TSP 175 28 53 37 65 Dino Hsp60 pmol/ng TSP 175 94 79 104 146 Oxidative Damage and Response Cnidarian Cu/Zn SOD Eunits/ng TSP 76 52 119 36 98 Dino Cu/Zn SOD Eunits/ng TSP 175 47 68 33 68 Cnidarian MnSOD pmol/ng TSP 166 62 58 42 114 Dino MnSOD fmol/ng TSP 175 104 88 91 94 Cnidarian GPx pmol/ng TSP 175 30 53 22 73 Dino GPx pmol/ng TSP 175 103 128 89 104 Catalase pmol/ng TSP 99 29 50 23 123 Metabolic Condition Heme oxygenase Eunits/ng TSP 76 67 100 33 105 Ferrochelatase Eunits/ng TSP 175 68 57 37 75 Metallothionein Eunits/ng TSP 175 163 121 45 86 Cnidarian sHsp Eunits/ng TSP 175 100 133 94 106 ChlpsHsp Eunits/ng TSP 175 64 90 40 119 Xenobiotic Response and Damage MXR Eunits/ng TSP 175 47 63 35 79 Cnidarian GST pmol/ng TSP 175 74 139 38 102 Dino GST pmol/ng TSP 175 29 37 23 51 CYP450 2-class Eunits/ng TSP 175 48 82 121 105 CYP450 3-class relative units/ng TSP 99 29 53 19 110 CYP450 6-class relative units/ng TSP 99 30 38 26 69

PAGE 218

Table 6.19. BEST and RELATE results for coral regeneration rates and biomarker concentrations dur ing a given time period along the depth gradient, n.s. repres ents not significantly (> 5%) Regeneration Rate Biomarker Rho Significance level # variables Selected Biomarkers JUN AUG 2001 JUN 2001 0.08 n.s. AUG OCT 2001 AUG 2001 0.25 n.s. OCT 2001 MAR 2002 OCT 2001 0.24 1.1% 1 Cn Hsp70 MAR JUN 2002 MAR 2002 -0.05 n.s. JUN AUG 2002 JUN 2002 0.06 n.s. AUG NOV 2002 AUG 2002 0.27 2.4% 1 Cn GPx NOV 2002 FEB 2003 NOV 2002 0.07 n.s.

PAGE 219

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol /ng TS P 0 0.5 1 1.5 2 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TS P 0 0.1 0.2 0.3 0.4 0.5 MAR 2001JUN 2001AUG 2001OCT 2001pmol/ng TS P KL 6 m WB RAR BNP 0 10 20 30 40 50 60 70 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TS P A B C D Cnidarian Hsp 60 Cnidarian Hsp 70 Dinoflagellate Hsp 60 Dinoflagellate Hsp 70 Figure 6.1. Protein Metabolic Condition at the 6 m sites including (A) cnidarian heat shock protein (Hsp) 60, (B) dinoflagellate heat shock protein 60, (C) cnidarian heat shock protein 70, (D) dinoflagellate hea t shock protein 70 and (E) ubiquitin. Data presented as means ( SE) in pmol/ng TSP for cnidarian Hsp 60 and 70, dinoflagellat e Hsp 60 and 70 and in fmol/ng TSP for ubiquitin. The red and blue dashed line represents stressed and basal levels, res pectively as defined by Downs et al. 2005a. Means for dinoflagellate Hsp 70 were all below stressed levels (1.68 pmol/ng TSP).

PAGE 220

0 200 400 600 800 1000 1200 1400 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003fmol/ng TSP KL 6 m WB RAR BNP E Ubiquitin Figure 6.1 (cont.). Protein Metabolic Condition at the 6 m sites including (A) cnidarian heat shock protein (Hsp) 60, (B) dinoflagellate heat shock protein 60, (C) cnidarian heat shock protein 70, (D) dinoflagella te heat shock protein 70 and (E) ubiquitin. Data presented as means ( SE) in pmol/ng TSP for cnidarian Hsp 60 and 70, dinoflagel late Hsp 60 and 70 and in fmol/ng TSP for ubiquitin. The red and blue dashed line represents stressed and basal levels, res pectively as defined by Downs et al. 2005a. Means for dinoflagellate Hsp 70 were all below stressed levels (1.68 pmol/ng TSP).

PAGE 221

0 0.05 0.1 0.15 0.2 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSPNA 0 1000 2000 3000 4000 5000 6000 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003fmol/ng TS P Dinoflagellate Mn SOD Cnidarian Mn SOD C D B A Cnidarian Cu/Zn SODDinoflagellate Cu/Zn SOD 0 1 2 3 4 5 6 MAR 2001JUN 2001AUG 2001OCT 2001Eunits/ng TSP 0 0.3 0.6 0.9 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Euni ts/ng TSP KL 6 m WB RAR BNP Figure 6.2. Oxidative damage and response at the 6 m sites including ( A) cnidarian copper/zinc superoxide dismutase (Cu/Zn SOD), (B) dinoflagellate Cu/Zn SOD, (C) cnidarian manganese superoxide dismutase (MnSOD), (D) dinoflagellate Mn SOD, (E) cnidarian glutathione peroxidase (GPx), (F) dinoflagellate GPx and (G) cat alase. Data presented as means ( SE) in pmol/ng TSP. The re d and blue dashed line represents stressed and basal levels, respec tively as defined by Downs et al. 2005. Means for cnidarian GPx were below both basal and stressed levels at all sites ( 70 and 171 pmol/ng TSP, respectively). Stressed or basal levels are not available for cnidarian and dinoflagellate Cu/Zn SOD or catalase.

PAGE 222

-5 0 5 10 15 20 25 30 MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003p mol/ng TSPCatalase KL 6 m WB RAR BNP 0 10 20 30 40 50 60 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSPCnidarian GPx 0 1 2 3 4 5 6 7 8 9 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pm ol/ng TS P Dinoflagellate GPx E F G Figure 6.2 (cont.). Oxidative damage and response at the 6 m sites inc luding (A) cnidarian copper/zinc superoxide dismutase (Cu/Zn SOD), (B) dinoflagellate Cu/Zn SOD, (C) cnidarian manganese superoxide dismutase (MnSOD), (D) dinoflagellate Mn SOD, (E) cnidarian glutathione peroxidase (GPx), (F) dinoflagellate GPx a nd (G) catalase. Data presented as means ( SE) in pmol/ng TS P. The red and blue dashed line represents stressed and basal levels, respectively as defined by Down s et al. 2005. Means for cnidarian GPx were below both basal and stressed levels at all sites (70 and 171 pmol/ng TSP, resp ectively). Stressed or basal levels are not available for cnidarian and dinoflagellate Cu/Zn SOD or catalase.

PAGE 223

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSPFerrochelatase 0 0.5 1 1.5 2 2.5 3 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSP KL 6 m WB RAR BNP 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 MAR 2001JUN 2001AUG 2001OCT 2001 Heme OxygenaseE units/ng TSPA B C Metallothionein Figure 6.3. Metabolic Condition at the 6 m sites i ncluding (A) heme oxygenase, (B) ferrochelatase, (C) metallothionein, (D) cnidarian small heat shock protein (sHsp) and (E) chloroplast small heat shock protein (ChlpsHsp). Data presented as means ( SE) in Eunits/ng TSP. Baseline condition is not available for these biomarkers.

PAGE 224

0 0.5 1 1.5 2 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSPAChloroplast sHsp 0 0.02 0.04 0.06 0.08 0.1 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSP KL 6 m WB RAR BNP A Cnidarian sHsp Figure 6.3 (cont.). Metabolic Condition at the 6 m sites including (A) heme oxygenase, (B) ferrochelatase, (C) metallothionein, (D) cnidar ian small heat shock protein (sHsp) and (E) chloroplast small heat shock protein (ChlpsHsp). Data present ed as means ( SE) in Eunits/ng TSP. Baseline condition is not available for these biomarkers.

PAGE 225

0 0.2 0.4 0.6 0.8 1 MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003CYP-3 relative units/ng TSP KL 6 m WB RAR BNP 0 0.2 0.4 0.6 0.8 1 1.2 MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003CYP-6 relative units/ng TSP 0 0.5 1 1.5 2 2.5 3 3.5 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TS PCYP-2 Figure 6.4. Xenobiotic Detoxification and Response at the 6 m sites including (A) cytochrome P450 2-class (CYP-2), (B) cytochrome P45 0 3-class (CYP-3), (C) cytochrome P450 6-class (CYP-6), (D) cnidarian glutathione-S-transf erase (Cn GST), (E) dinoflagellate GST and (F) multixenobiotic resistance protein (MX R). Data presented as means ( SE) in Eunits/ng TSP for CYP-2 and MXR; relative units /ng TSP for CYP-3 and CYP-6; and pmol/ng TSP for Cn and Dn GST. Baseline information is not available for any of the cytochrome P450 classes or for MXR. The red and blue dashed line represents stressed and basal levels, respectively as def ined by Downs et al. 2005. Baseline information is not available for CYP-2, CYP-3, CYP-6 or MXR. A B C

PAGE 226

0 1 2 3 4 5 6 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TS P Cnidarian GST 0 2 4 6 8 10 12 14 16 18 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TS P Dinoflagellate GST MXR 0 0.2 0.4 0.6 0.8 1 1.2 1.4 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSP KL 6 m WB RAR BNP Figure 6.4 (cont.). Xenobiotic Detoxification and Response at the 6 m sit es including (A) cytochrome P450 2-class (CYP-2), (B) cytochrome P45 0 3-class (CYP-3), (C) cytochrome P450 6-class (CYP-6), (D) cnidarian glutathione-S-transf erase (Cn GST), (E) dinoflagellate GST and (F) multixenobiotic resistance protein (MX R). Data presented as means ( SE) in Eunits/ng TSP for CYP-2 and MXR; relative units /ng TSP for CYP-3 and CYP-6; and pmol/ng TSP for Cn and Dn GST. Baseline information is not available for any of the cytochrome P450 classes or for MXR. The red and blue dashed line represents stressed and basal levels, respectively as def ined by Downs et al. 2005. Baseline information is not available for CYP-2, CYP-3, CYP-6 or MXR. D E F

PAGE 227

Cnidarian Hsp60 0 10 20 30 40 50 60 70 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSP 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pm ol/ng TSP Dinoflagellate Hsp60 0 0.5 1 1.5 2 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSP Cnidarian Hsp70 0 0.1 0.2 0.3 0.4 0.5 MAR 2001JUN 2001AUG 2001OCT 2001pmol/ng TS P Dinoflagellate Hsp70 KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m A B C D Figure 6.5. Protein Metabolic Condition along the d epth gradient including (A) cnidarian heat shock pr otein (Hsp) 60, (B) dinoflagellate heat shock protein 60, (C) cnidarian heat shock protein 70, (D ) dinoflagellate heat shock protein 70 and (E) ubiquitin. Data presented as means ( SE) in pmol/ng TSP for cnidarian Hs p 60 and 70, dinoflagellate Hsp 60 and 70 and in fmol/ng TSP for ubiquitin. The red and blue dashed line represents stressed and basal levels, respectively as def ined by Downs et al. 2005a. Means for dinoflagellate Hsp 70 were all below stressed levels (1.68 pmol/ ng TSP).

PAGE 228

KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 mUbiquitinF 0 200 400 600 800 1000 1200 1400 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003fmol/ng TSP Figure 6.5 (cont.). Protein Metabolic Condition along the depth gradient incl uding (A) cnidarian heat shock protein (Hsp) 60, (B) dinoflagellate heat shock protein 60, (C) cnidarian heat shock protein 70, ( D) dinoflagellate heat shock protein 70 and (E) ubiquitin. Data presented as means ( SE) in pmol/ng TSP for cnidarian Hs p 60 and 70, dinoflagellate Hsp 60 and 70 and in fmol/ng TSP for ubiquitin. The red and blue dashed line represents s tressed and basal levels, respectively as defined b y Downs et al. 2005a. Means for dinoflagellate Hsp 70 were all below stressed levels (1.68 pmol/ ng TSP).

PAGE 229

0 1 2 3 4 5 6 MAR 2001JUN 2001AUG 2001OCT 2001Eunits/ng TSP 0 0.3 0.6 0.9 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003E units/ng TSPA B Cnidarian Cu/Zn SOD Dinoflagellate Cu/Zn SOD KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m 0 0.05 0.1 0.15 0.2 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSP 0 1000 2000 3000 4000 5000 6000 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003fm ol/ng TSPNA C D Dinoflagellate Mn SOD Cnidarian Mn SOD Figure 6.6. Oxidative damage and response along the depth gradient including (A) cnidaria n copper/zinc superoxide dismutase (Cu/Zn SOD), (B) dinoflagellate Cu/Zn SOD, (C) cnidarian m anganese superoxide dismutase (MnSOD), (D) dinoflag ellate Mn SOD, (E) cnidarian glutathione peroxidase (GPx), (F) dinoflagellate GPx a nd (G) catalase. Data presented as means ( SE) in pmol/ng TS P. The red and blue dashed line represents stressed and basal levels, respectively as defined by Down s et al. 2005. Means for cnidarian GPx were below both basal and stressed levels at all sites (70 and 171 pmol/ng TSP, resp ectively). Stressed or basal levels are not available for cnidarian and dinoflagellate Cu/Zn SOD or catalase.

PAGE 230

0 10 20 30 40 50 60 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSPCnidarian GPx 0 0.5 1 1.5 2 2.5 3 3.5 4 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSPDinoflagellate GPx -5 0 5 10 15 20 25 30 MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003pmol/ng TSPCatalaseG KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m Figure 6.6 (cont.). Oxidative damage and response along the depth gradient including (A) cnidarian copper/zinc superoxide dismutase (Cu/Zn SOD), (B) dinoflagellate Cu/Zn SOD, (C) cnidarian manganes e superoxide dismutase (MnSOD), (D) dinoflagellate Mn SOD, (E) cnidarian glutathione peroxidase (GPx), (F) dinoflagellate GP x and (G) catalase. Data presented as means ( SE) in pmol/ng TSP. The red and blue dashed line represents stressed and basal levels, respectively as defined by Down s et al. 2005. Means for cnidarian GPx were below both basal and stressed levels at all sites (70 and 171 pmol/ng TSP, resp ectively). Stressed or basal levels are not available for cnidarian and dinoflagellate Cu/Zn SOD or catalase. E F

PAGE 231

0 0.5 1 1.5 2 2.5 3 3.5 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSPMetallothionein 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSPFerrochelatase KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 MAR 2001JUN 2001AUG 2001OCT 2001Eunits/ng TSPHeme OxygenaseFigure 6.7. Metabolic Condition along the depth gradient including (A) heme oxygenase, (B) ferrochelatase, (C) metallothionein, (D) cnidarian small he at shock protein (sHsp) and (E) chloroplast small heat shock protein (ChlpsHsp). Data presented as means ( SE) in Eunits/ng TSP. Baseline condition is not available for these biomarkers. A B C

PAGE 232

Figure 6.7 (cont.). Metabolic Condition along the depth gradient including (A) heme oxygenase, (B) ferrochelatase, (C) metallothionein, (D) cnidarian small heat shock protein (sHsp) and (E) chloroplast small heat shock protein (ChlpsHsp). Data presented as means ( SE) in Eunits/ng TSP. Baseline condition is not avail able for these biomarkers. 0 0.5 1 1.5 2 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSPChloroplast sHsp 0 0.02 0.04 0.06 0.08 0.1 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSPsHsp KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m

PAGE 233

0 0.2 0.4 0.6 0.8 1 MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003relative units/ng TSPCYP-3 0 0.2 0.4 0.6 0.8 1 1.2 MAR 2002JUN 2002AUG 2002NOV 2002FEB 2003relative units/ng TSPCYP-6 KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m 0 0.5 1 1.5 2 2.5 3 3.5 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003E units /ng TS PCYP-2 Figure 6.8. Xenobiotic Detoxification and Response along the depth gradient including (A) cytochrome P450 2-class (CYP-2), (B) cytochrome P450 3-class (CYP-3), (C) cytochrome P450 6-class (CYP-6), (D) cnidarian glutathione-S-transf erase (Cn GST), (E) dinoflagellate GST and (F) multixenobiotic resistance protein (MX R). Data presented as means ( SE) in Eunits/ng TSP for CYP-2 and MXR; relative units /ng TSP for CYP-3 and CYP-6; and pmol/ng TSP for Cn and Dn GST. Baseline information is not available for any of the cytochrome P450 classes or for MXR. The red and blue dashed line represents stressed and basal levels, respectively as def ined by Downs et al. 2005. Baseline information is not available for CYP-2, CYP-3, CYP-6 or MXR.

PAGE 234

MXR 0 0.2 0.4 0.6 0.8 1 1.2 1.4 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Eunits/ng TSP 0 2 4 6 8 10 12 14 16 18 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pmol/ng TSPDinoflagellate GST 0 1 2 3 4 5 6 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003pm ol/ng TSP Cnidarian GST KL 3M KL 6M KL 9M KL 18MKL 3 m KL 6 m KL 9 m KL 18 m Figure 6.8 (cont.). Xenobiotic Detoxification and R esponse along the depth gradient including (A) cytochrome P450 2-class (CYP-2), (B) cytochrome P450 3-cl ass (CYP-3), (C) cytochrome P450 6-class (CYP-6), (D) cnidarian glutathione-S-transfer ase (Cn GST), (E) dinoflagellate GST and (F) multixenobiotic resistance protei n (MXR). Data presented as means ( SE) in Eunits/ng TSP for CYP-2 and M XR; relative units/ng TSP for CYP-3 and CYP-6; and pmol/ng TSP for Cn and Dn GST. Baseline informa tion is not available for any of the cytochrome P450 classes or for MXR. The re d and blue dashed line represents stressed and basal levels, resp ectively as defined by Downs et al. 2005. Baseline information is not available for CYP-2, CYP-3, CYP-6 or MXR.

PAGE 235

-15 -10 -5 0 5 10 15 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 20036 m Sites PC1 -1 0 5 0 5 1 0 1 5 MAR 2001 JUN 2001 AUG 2001 OCT 2001 MAR 2002 JUN 2002 AUG 2002 NOV 2002 FEB 2003Depth GradientPC1 Figure 6.9. Plots of Principle Component (PC1) scores for each sampli ng period at (A) the 6 m sites and (B) along the depth gradient. Vertical bars show the range of values for each sampling period, squares indicate the sample mean for each peri od. Shaded areas represent sampling periods when there were significant difference s among sites based on ANOSIM (6 m sites: Table 6.5; depth gradient: Table 6.8). Low HighAll Biomarkers Low HighAll Biomarkers A B

PAGE 236

8 -6 -4 2 0 2 4 6 8 10March 2001 June 2001 August 2001 October 2001 -10 8 6 4 2 0 2 4 6 8 6 4 2 0 2 4 6 8 -6 -4 -2 0 2 4 6 8KL 3 mKL 6 mKL 9 mKL 18 mPC1 PC1PC1 PC1Figure 6.10. Plots of Principle Component (PC1) sc ores at each site for (A) March 2001, (B) June 2001, (C) August 2001 and (D) October 2001. Vertical bars show the range of values for each site, squares indicate the s ample mean for each site. Eigenvalues and eigenvectors as shown in Appendix B. Low HighAll Biomarkers Low HighAll Biomarkers Low HighAll Biomarkers Low HighAll Biomarkers A B C D

PAGE 237

-10 -8 -6 -4 -2 0 2 4 6 8 10 KL 6 mWBAR BNPPC1February 2003 Low HighAll BiomarkerFigure 6.11. Plots of Principle Component (PC1) scores at each site in February 2003. Vertical bars show the range of values for each site; squares i ndicate the sample mean for each site. Eigenvalues and eigenvectors as shown in Appendix C.

PAGE 238

Low HighAll Biomarkers -8 -6 -4 -2 0 2 4 6 8 10 KL 6 mWBAR BNPPC1 -10 -8 -6 -4 -2 0 2 4 KL 6 mWBAR BNPPC1October 2001 June 2002 Figure 6.12. Plots of Principle Component (PC1) sc ores at each site for (A) October 2001 and (B) June 2002. Vertical bars show the range of values for each s ite, squares indicate the sample mean for each site. Eigenvecto rs and eigenvalues as shown in Appendix C. CYP-2 CYP-3 Dn Hsp 60 Cn GST Cn Hsp 60 CYP-6 A B

PAGE 239

A B October 2001 Metallothionein C October 2001 Cn Cu/Zn SOD D Figure 6.13. (A) Multi-dimensiona l scaling (MDS) plot for regeneration rates (T/P) between October 2001 and March 2002 at the 6 m sites; larger circle s represent higher regeneration rates. (B) MDS plot of October 2001 cellular biomarkers (CDS) selected by BEST routine (Table 6.14), which included me tallothionein and cnidarian copper/zinc super oxide disputase (Cn Cu/Zn SOD). Regeneration rate MDS superimposed by indivi dual CDS biomarkers (C) metallothionein and (D ) Cn Cu/Zn SOD; circle size increases with increasing concentration. Cellular biomarker MDS

PAGE 240

A B August 2001 Dn Hsp 60 C D August 2001 Dn Cu/Zn SOD Figure 6.14. (A) Multi-dimensional scaling (MDS) plot for regene ration rates (T/P) between August and October 2001 at the 6 m sites; larger circles represent higher rege neration rates. (B) MDS plot of August 2001 cellular biomarkers (CDS) selected by B EST routine (Table 6.14), which includ ed dinoflagellate heat shock protein 60 (Dn Hs p 60) and dinoflagellate copper/zinc superoxide dismutase (Dn Cu/Zn SOD). Regeneration ra te MDS superimposed by individual CDS biom arkers including (C ) Dn Hsp 60 and (D) Dn Cu/Zn SOD; circle size increases w ith increasing concentration. Cellular biomarker MDS

PAGE 241

A B June 2002 Cn Hsp 70 Cellular biomarker MDS D June 2002 Cn sHsp Figure 6.15. (A) Multi-dimensional scaling plots for regeneration rates (T/P) between June and August 2002 at the 6 m sites; l arger circles represent higher regeneration rates. (B) MDS plot of June 2002 cellular biomar kers (CDS) selected by BEST routine (Tab le 6.14), which included cnidarian heat shock pr otein (Cn Hsp 70), cnidarian small heat s hock protein (Cn sHsp), metallothionein, cnidarian glutathione-S-transferase and cytoch rome P450 6-class. Regeneration rate MDS s uperimposed by individual CDS biomarkers (C) Cn Hsp 70, (D) Cn sHsp, (E) me tallothionein, (F) Cn GST and (G) CYP-6; ci rcle size increases with increasing concentration. C

PAGE 242

June 2002 Metallothionein E June 2002 Cn GST June 2002 CYP-6 G Figure 6.15 (cont.). (A) Multi-dimensional scaling plots for regeneratio n rates (T/P) between June and August 2002 at the 6 m sites; larger circles represent higher regeneration rates. (B) MDS plot of June 2002 cellular biomarkers (CDS) selected by BEST routi ne (Table 6.14), which included cnidaria n heat shock protein (Cn Hsp 70), cnidarian small heat shock protein (Cn sHsp), metallothionein, cnidarian glutathione-S-transf erase and cytochrome P450 6-class. Regenera tion rate MDS superimposed by individual CDS biomarkers (C) Cn Hsp 70, (D ) Cn sHsp, (E) metallothionein, (F) Cn GST and (G) CYP-6; circle size increases with increasing concentration. F

PAGE 243

Figure 6.16. (A) Multi-dimensional scaling (MDS) and bubble plot of Key Largo 6 m regeneration rates (T/P); larger circles represent higher regeneration rates. (B) Re generation rate MDS superimposed by total ni trogen, which was selected by the BEST routine (Table 6.13); ci rcles increase in size with increasing concentration. A B

PAGE 244

Figure 6.17. (A) Multi-dimensional scaling (M DS) and bubble plot of Key Largo 9 m regene ration rates (T/P); larger circles represent higher regeneration rates. Rege neration rate MDS superimposed by environmental variables selected by BEST routine (Table 6.13) including (B) ratio of total nitrogen to total phosphorus (TN:TP), (C) nitrite, (D) chlor ophyll-a and (E) turbidit y; circles increase in size with incr easing concentration. A B C D E

PAGE 245

Figure 6.18. (A) Multi-dimensional scaling (M DS) and bubble plot of Key Largo 18 m regene ration rates (T/P); larger circles represent higher regeneration rate s. Regeneration rate MDS superimposed by envi ronmental variables selected by BEST routine (Table 6.13) including (B) soluble reactive phosphate, (C) ration of total nitrogen to total phosphorus (TN:TP), (D) chlorophyl l-a (E) nitrite and (F) total orga nic carbon; circles increase in size with increasing concentration. A B C D

PAGE 246

Figure 6.18 (cont.). (A) Multi-dimensional sc aling (MDS) and bubble plot of Key Largo 18 m regeneration rates (T/P); larger ci rcles represent higher regeneration rate s. Regeneration rate MDS superimposed by envi ronmental variables selected by BEST routine (Table 6.13) including (B) soluble reactive phosphate, (C) ration of total nitrogen to total phosphorus (TN:TP), (D) chlorophyl l-a (E) nitrite and (F) total orga nic carbon; circles increase in size with increasing concentration. E F

PAGE 247

A B C D E Figure 6.19 (A) Multi-dimensional scaling (M DS) and bubble plot of Algae Reef regenera tion rates (T/P); larger circles represe nt higher regeneration rates. Regeneration rate MDS superimposed by environmental variab les selected by BEST routine (Table 6.13) including (B) total organic nitrogen, (C) rati o of total nitrogen to total phosphorus (T N:TP), (D) ratio of dissolved inorganic nitrogen to total phosphorus (DIN:TP) and (E) total organic carbon; circles increase in size with increasing concentration.

PAGE 248

-6 -4 -2 0 2 4 6 8 10PC1 -6 -4 -2 0 2 4 6 8 10PC1 -8 -6 -4 -2 0 2 4 6 8 10PC1 -8 -6 -4 -2 0 2 4 6 8 10 12PC1KL 3 mKL 6 mKL 9 mKL 18 m February 2003 June 2002 August 2002 November 2002 Figure 6.20. Plots of Principle Component (PC1) scores at each site for (A) June 2002, (B) August 2002, (C) November 2002 and (D) February 2003. Vertical bars show the range of values for each site; squares indicate the sample mean for each site. Eigenvalues and vectors as shown in Appendix B. Catalase CYP-2 Dn Hsp 60 Cn GPx Cn Hsp 70 Dn GPx Low HighAll Biomarkers Low HighAll Biomarkers Low HighAll Biomarkers A B C D

PAGE 249

Figure 6.21. (A) Multi-dimensional scaling (MDS) and bubble plot of White Banks regene ration rates (T/P); la rger circles repre sent higher regeneration rates. (B) Regeneration rate MDS superimposed by tota l nitrogen, which was selected by BEST routine (Table 6.13); circles increase in size with increasi ng concentration. A B

PAGE 250

Organism ResponseHigh Low Healthy Minimal Stress High Stress Compensating High Stress Reduced Physiological Condition Incurable KL 18 m AR & KL 3 m BNP & KL 9 m KL 6 m WB Regeneration Rates Densities of LBFCellular BiomarkersPhysiological Status Periodic Exposure to Stress?? Figure 6.22. Physiological status of corals at each study site bas ed on the relationship between regeneration rates, a surrogate indicator a nd cellular diagnostic markers (modified from Allen & Moore 2004). Regeneration r ates and densities of symbiontbearing foraminifera (LBF) are represented by the dashed blue line and cellular biomarker levels are represented by the dashed red line. The position of each site is represented by where the circle intersects these two lines. N ote: regeneration rates and densities of LBF follow similar trends with the exception of KL 3 m, where densities are low but regeneration rates are high, and with KL 9 m and KL 18 m wher e densities are high but regeneration rates are low. In these cases, density of L BF were not considered due to the caveats of this indicator with depth. Site abbreviations ar e the same as those used in Fig. 1.1.

PAGE 251

225 7. Conclusions and Future Research 7.1. Multivariate Approach to Assessing Reef Condition No single metric is adequate to study th e complex and inherently variable effects of environmental change on marine ecosystems (Adams 2005). A mechanistic understanding of the effects of multiple stressors on reefs requi res a hierarchical approach based on multiple lines of evidence (Downs 2005), which allows researchers to determine whether an organism is responding to a stress ed condition and, if so whether that stress has resulted in reduced physiological f unction (Downs 2005, Moore et al. 2006). Lower levels of biological organization can provide information on the mechanism of decline, whereas higher levels of biological organi zation provide information on the effect of stress on the overall fitness and function of the organism, populat ion, community or ecosystem. The specific objectives of my study were to 1) eval uate the ability of individual indicators to distinguish differences among sites, times and stressors; 2) assess reef condition using a hierarchal, multi-scale appr oach including selected environmental, community, population, colony and cellular parameters ; and 3) diagnose the physiological state of selected reefs based on ‘weight of evidence’ th rough the integration of multiple indicators. In this chapter, I address the strengths and caveats of individual bioindicators used in this study, summarize the major conclusions of this study, and make recommendations for management of these ecosystems and future research. 7.2. Strengths and Caveats of Individual Indicators All indicators used in this study were cap able of distinguishing among study sites. Cellular biomarkers were the most sensitive to changes in environmental conditions with time. Each indicator provided a different pe rspective into reef condition. However, each indicator also had its lim itations (Table 8.1).

PAGE 252

226 Community assessments provided insight in to past conditions. For example, low coral cover, small coral-colony size and lo w fish biomass indicat ed that past reef conditions were sub-optimal for reef growth and development. An advantage of using the Atlantic Gulf and Rapid Reef Assessment (AGRRA) for community assessment is that it has been widely used throughout th e Caribbean, which provi ded valuable baseline data for comparison. Similar methods have be en used Keys-wide, providing information on long-term changes in the Florida Keys. Algae Reef (AR) showed the highest similarity to Caribbean regional “best” va lues and Key Largo (KL) 6 m showed the highest similarity to Cari bbean regional “worst ” values. Alina’s Reef (BNP) had community characteristics similar to both AR and regional means, but high recent mortality and high abundances of macroalgae i ndicate that this site has experienced decline that began relatively recently. Changes at the community scale can occur over years to decades and often the stressor remains unknown. For example, the cause for increased macroalgae on reefs has been widely debated, with reduced herbivory (Hughes 1994, Williams & Polunin 2001), increased nutrients (Lapointe 1997, 1999) and in creased available substrate due to coral mortality (Szmant 2001) among the postulated factors. Methods such as AGRRA are unsuitable for detecting interannual changes in reef communities. Long term and chronic exposure to environmental stress, including ch emical pollutants or other anthropogenic factors, rarely result in rapid and catastrophic change. In stead, the effects are most likely gradual, subtle and difficult to separate from the effects of natural environmental change (Moore et al. 2004). Foraminiferal assemblages and condition prov ided insight into the suitability of the environment for symbiont-bearing calci fying organisms (e.g., nutrient-depleted, minimal pollution) and the presence of photic stress (e.g., Hallock et al. 2003, 2006). An advantage of monitoring symbiont-bearing (‘la rger’) benthic foraminifera (LBF) is that these foraminifera have short life spans (a pprox. 1 yr) and therefore population changes respond quickly to environmental changes in comparison to long-lived species such as corals and fish. By monitoring LBF, I wa s able to detect a reduction in stress, particularly at BNP, between 2001 and 2002. Monitoring of LBF also has been used

PAGE 253

227 throughout the Florida Keys and in other regi ons providing baseline data for comparison (e.g., Hallock et al. 1986, Williams 2002; Hallock et al. 2006). Population densities of LBF were low in the vicinity of sampled corals at KL 6 m and BNP, indicating suboptimal water quality or othe r environmental conditions at those sites. Intermediate population densities and bleaching in Amphistegina gibbosa the dominant LBF species, at AR, KL 9 m and KL 18 m indicated that chronic photic stress and possibly other chronic stresses were effecting these sites. This indicator was unable to identify the environmental parameter(s) or aspect(s) of water quality that we re impacting population densities. Previous research has shown that bleaching in A. gibbosa is a response to photic stress but not to temperature (W illiams 2002, Hallock et al. 2006). Further investigation is needed to determine if othe r stressors (e.g., chemical pollutants) make these organisms more susceptibl e to photic stress. Another caveat of using LBF as an indicator is that they are not suitable in low-energy, near shore shallow environments such as KL 3 m. These foraminifera also prefer deeper depths, limiting comparisons among different depths. While a strong relatio nship was seen between densities of LBF and regeneration rates among 6 m sites, a poor relationship was found along the depth gradient. Thus, my research demonstrated th at comparisons should be restricted to sites of similar depths. Lesion regeneration provided insight into the physiological condition of the framework-producing corals, Montastraea annularis complex. Colony-scale studies indicated significant differences among sites in the ability of M. annularis complex (Ch 5) to recover from damage. Low regenerati on rates, increases in mortality and high breakage indicated that physiological condition of corals along deve loped portions of the coastline at KL 6 m, BNP, KL 9 m a nd KL 18 m was compro mised. Reefs along undeveloped portions of the coastline (e.g., AR and WB) and the nearshore patch reef (KL 3 m) had consistently high regeneration ra tes. Regeneration rates of coral lesions reflect the ability of colonies to repair from damage, providing a useful, inexpensive indicator of coral condition or e nvironmental conditions. A caveat of this indicator is that it is not capable of separating effects of coral health versus external environmental factors (e.g., sedimentation, temperature, pollutio n) on lesion regeneration rates.

PAGE 254

228 The cellular diagnostic sy stem (CDS) indicated that co rals from all sites deviated from a nominal cellular physiological state, with the highest stress observed in winter months and following high rainfall. Certain cellular parameters assayed indicated that corals were responding to a xenobiotic stress, connecting coral condition to local stressors (e.g., pesticides or herbicides). Most cellular parameters also respond to changes in the environment within days to weeks, although elevated levels of some proteins may persist for longer periods of time, complicating interpreta tions of the data. Among the 6 m sites, corals at BNP and AR te nded to have higher levels of biomarkers, whereas those at WB and KL 6 m tended to ha ve lower levels of biomarkers. Along the depth gradient, corals at KL 3 m and KL 9 m tended to have higher levels of biomarkers, whereas, those at KL 6 m and KL 18 m tended to have lower levels of biomarkers. When selecting a set of cellular parameters, an investigator must consider the questions that they are interested in addressing. For example, additional cellular parameters indicative of a xenobiotic response or cellular damage can be used to help narrow down the list of potential stressors and provide a more t horough understanding of the mechanisms of stress. A couple caveats of this indicator is that it can be cost-prohibitive, requires consumptive sampling and the relative novelty of this indicator means that information on baselines are limited. Data collected in th is and previous studies (e.g., Downs et al. 2005) provide a basis for comparison where actual concentrations are known. Further development of biomarker and bioindicator baselines is needed to gain a better understanding of what is “nor mal” versus “stressed.” Th ese baselines can be further defined through controlled field and laboratory experiments. Use of additional cellular biomarkers indicative of specific damage, al ong with targeted func tional studies (e.g., histology), might provide defini tive evidence that corals are in a diseased or incurable state. Environmental assessments provided the oppor tunity to connect responses in reef organisms to changes in the reef environment. Environmental assessments indicated that sedimentation was highest in the winter mont hs, associated with high winds. Increased nutrients at my study sites followed heavy rain falls, indicating that land-based stressors were reaching these reefs. Algae Reef a nd WB overall had higher sedimentation rates

PAGE 255

229 than KL 6 m and BNP, though sediments at the latter sites tended to be finer and therefore potentially more damaging. No significant differences were observed in temperature and no bleaching was observed at my study sites during this project. Both AGRRA and monitoring of LBF are relatively low cost and low impact indicators providing information on the suitability of the environment for reef inhabitants. Monitoring of coral-regenerati on rates is another low cost indicator, which requires minimal training. Because this method require s creation of small le sions on corals, it is readily paired with cellular diagnostic sampling, which requires sampling of small amounts of coral tissue. The cellular diagnostic methods are costly and require technical training and biochemical background to pro cess samples and interpret results. By drawing from multiple lines of evidence at multiple scales, I was able to diagnose the physiological condition of coral colo nies at these reefs (Fig. 6.22). None of our study sites was considered to be in a “health y” state. Corals at both AR and KL 3 m were compensating to a xenobiotic stress but this did not appear to affect regeneration rates. Corals at both BNP and KL 9 m were responding to a stress that reduced regeneration rates and increased mortality. Stressors were likely recent at BNP as indicated by community assessments, specifi cally high recent morality and macroalgal biomass. Abnormally low cellular biomarkers concentrations, low regeneration rates and high mortality at KL 18 m indicated that cora ls at this site were responding to severe stress, which has left these colonies incapable of recovering from damage. 7.3. Recommendations for Management and Future Research Reefs are both ecologically and economically important resources. Yet the inability to identify stressors has left manage ment incapable of preventing or alleviating stressors that have resulted in drastic coral loss since the 1970s. Inclusion of these reef sites in a marine protected area (Florida Keys National Ma rine Sanctuary and Biscayne National Park) has not adequate ly prevented their deteriora tion (Jameson et al. 2002). One of the major objectives of my study was to test the use of an integrated Cellular Diagnostic System in the characterization of coral condition. The CDS was capable of detecting subtle stress conditi ons due to periodic events (e.g., following heavy rainfall)

PAGE 256

230 and chronic stress conditions (e.g., sustained high biomarker levels at BNP and KL 9 m). Using CDS, I was able to determine that lo cal stressors, specifica lly xenobiotics, were affecting corals at my sites. Possible mechanisms of stress included an endocrine disrupting stress in June 2002 at KL 6 m and de pressed protein levels at KL 18 m either due to hypoxic conditions or exposure to cont aminants. Therefore, this study was an important contribution in the process of diagnosing reef condition by providing avenues for future research to help narrow down the id entification of stressors to these reefs. Further investigations in contaminant exposure and organi sm responses at these sites are needed to help managers identify and alleviate these stressors, which included temperature (e.g., Downs et al. 2002, Fauth et al. 2005) and pest icides (e.g., Downs et al. 2005, Downs et al. 2006). Exposure to xenobioti c stressors may make these corals more susceptible to predicted climate changes (e .g., increased temperat ures, reduced pH). While all sites experienced st ress, my approach distinguish ed between reefs that were compensating for stress (e.g., Algae Reef and Key Largo 3 m) and those that appeared beyond repair (e.g., Key Largo 18 m), as defined by Moore et al. (2006). This information also can help managers target their efforts to reefs that are capable of recovery, as recommended by Jameson et al. (2002). Further investigation is needed to dete rmine what factors contribute to better conditions (i.e., higher regeneration rates, highe r densities of LBF) at some sites relative to others. For example, sites along a less developed portion of co ast (e.g., AR and WB) were in better condition than those al ong developed coastlines (KL sites and BNP), indicating the importance of intact coastlines and wetlands. Currently, the Florida Department of Transportation is removing up to 106 acres of coastal wetlands to widen the 18 mile stretch of highw ay between the mainland and the Upper Florida Keys, potentially increasing sediment loads and re ducing inputs of colored dissolved organic matter to these KL reefs in future years. I recommend continued monitoring of this area to determine if removal of mangroves and othe r natural vegetation affects reef condition. Transplant experiments can help determin e if organisms can acclimate to stress conditions through increased production of pr oteins involved in protective pathways (e.g., heat shock proteins and an tioxidants) allowing them to compensate for the stress.

PAGE 257

231 For example, KL 3 m had consistently high regeneration rates and biomarkers levels. Shallow nearshore patch reefs are generally exposed to a wider range of natural and anthropogenic stressors and have likely beco me adapted to these conditions through the consistent upregulation of protective enzymes and chaper one proteins (Moore et al. 2006). For example, echinoderms from a variab le (intertidal) environmental showed a distinctly sustained expressi on pattern of Hsp72 compared with animals from a stable (benthic) environment, suggesting a functi onally adaptive and dynamic stress response (Patruno et al. 2001). My study also provides opportunities for preventive management by pinpointing areas of high stress where cora ls still appear to be physio logically healthy. High stress levels at Algae Reef and Key Largo 3 m c ould be an indication of early stages of physiological change so that sli ght increases in stress loads at these sites may result in reduced colony function (e.g., reduced regeneration rates, growth, reproduction). Contaminant exposures may be too low to ca use overt effects, but over longer time periods may manifest into adverse conditions and mortality (Depledge et al. 1993). Thresholds could be tested experimentally in the lab and in the field by gradually applying an additional stressor (e.g., pesticid e) to colonies and monitoring changes in cellular diagnostics and colony function (e .g., regeneration rate s). Therefore, management should identify and alleviate poten tial stressors at thes e sites before they result in further degradati on of the reef community.

PAGE 258

Table 7.1. Indicators of reef condition Indicator Description Spatial Scale Time Scale Training required for analyses Cause of stress readily identified (Y/N) Seasonally dependent (Y/N) Recommended frequency of sampling Sampling Cost (High/Low) Atlantic and Gulf Rapid Reef Assessment (AGRRA) Assessment of coral, fish and algal community Community Years Decades Species identification and calibration among observers N Macroalgal abundance can change seasonally Once every five years Low Symbiontbearing foraminifera Density of symbiontbearing organisms and assessment of bleaching and condition Population Weeks Months Species identification Can identifiy if photic stress is present Best to sample in late spring or late summer Biannually Cost of stereo microscope Regeneration rates Ability of coral to heal from damage over time Colony Weeks Months Basic photography and computer analysis N Shortterm (54 -154 d): N Quasi-annual (319 376 d ): Y Biannually/ Quarterly Cost of underwater camera and image analysis software Cellular Diagnostic System Monitors changes in concentrations of stress proteins Cellular Days Weeks Lab analyses and knowledge of biochemical pathways Can distinguish between types of stress and mechanisms of stress Y Monthly or Quarterly High

PAGE 259

References Abdel-Salem HA, Porter JW (1988) Physiol ogical effects of sediment rejection on photosynthesis and respiration in three Ca ribbean reef corals. Proceedings, 6th International Coral Reef Sy mposium, Australia, p 285-292 Adams SM (2005) Assessing cause and effect of multiple stressors on marine systems. Marine Pollution Bulletin 51:649-657 Allen JI, Moore MN (2004) Environmental prognostics: Is current use of biomarkers appropriate for environmental risk eval uation? Marine Environmental Research 58: 227-232 Andrews K, Nall L, Jeffrey C, Pittman S (2005) The State of Coral Reef Ecosystems of Florida. In: Waddell JE (ed) The State of Coral Reefs Ecosystems of the United States and Pacific Freely Associ ated States: 2005. NOAA Technical Memorandum NOS NCCOS 11. NOAA/NCCOS Center for Coastal Monitoring and Assessment's Biogeography Te am, Silver Springs, MD, p 150-200 Anthony KRN (1999) Coral suspension feeding on fine particulate matter. Journal of Experimental Marine Bi ology and Ecology 232:85-106 Anthony KRN (2006) Enhanced ener gy status of corals on coasta l, high-turbidity reefs. Marine Ecology Progress Series 319:111-116 Anthony KRN, Fabricius KE (2000) Shifti ng roles of heterotr ophy and autotrophy in coral energetics under varyi ng turbidity. Journal of E xperimental Marine Biology and Ecology 252:221-253 Ayoub L, Hallock P, Fisher E, Cannizzaro J (200 6) Can spatial and temporal variation in solar radiation provide clues to managi ng bleaching stress in coral reefs? Proceedings, 2nd Annual USF Interdiscip linary Graduate Research Symposium Babcock R (1985) Growth and mort ality in juvenile corals ( Goniastrea Platygyra and Acropora ): the first year. Proceedings, 5th International Coral Reef Congress 4:355 Bak RPM (1978) Lethal and s ublethal effects of dredgi ng on reef corals. Marine Pollution Bulletin 9:14-16 Bak RPM (1983) Neoplasia, regeneration a nd growth in the reef-building coral Acropora palmata Marine Biology 77:221-227

PAGE 260

Bak RPM, Elgershuizen JHBW (1976) Patter ns of oil-sediment rejection in corals. Marine Biology 37:105-113 Bak RPM, Meesters EH (1999) Population st ructure as a response of coral communities to global change. American Zoologist 39:56-65 Bak RPM, Brouns JJWM, Heys FML (1977) Regeneration and aspects of spatial competition in the scleractinian corals Agaricia agaricites and Montastrea annularis. Proceedings, 3rd International Coral Reef Symposium, Miami, Florida, p 143-148 Barbaglio A, Mozzi D, Sugni M, Tremolada P, Bonasoro F, Lavado R, Porte C, Carnevali MDC (2006) Effects of exposure to ED contaminants (TPT-Cl and Fenarimol) on crinoid echinoderms: comp arative analysis of regenerative development and correlated steroi d levels. Marine Biology 149:65-77 Bard SM (2000) Multixenobiotic resistance as a cellular defense mechanism in aquatic organisms. Aquatic Toxicology 48:357-389 Bayne BL, Brown DW, Burns K, Dixon DR, Ivanovici A, Livingstone DR, Lowe DM, Moore MN, Stebbing ARD, Widdows J (1985) The effects of stress and pollution on marine animals. New York, Praeger, 384 p. Beaver CR, Jaap WC, Porter JW, Wheaton J, Callahan M, Kidney J, Kupfner S, Torres C, Wade S, Johnson D (2005) Coral r eef evaluation and monitoring program (CREMP), 2004 Executive Summary, Florid a Fish and Wildlife Conservation Commission and University of Georgia Bell JD, Galzin R (1984) Influence of live co ral cover on coral-reef fish communities. Marine Ecology Progress Series 15:265-274 Bellwood DR, Hughes TP, Folke C, Nystrom M ( 2004) Confronting the coral reef crisis. Nature 429:827-833 Black NA, Voellmy R, Szmant AM ( 1995) Heat shock protein induction in Montastraea faveolata and Aiptasia pallida exposed at elevated temperatures. Biological Bulletin 188:234-240 Borst P, Elferink RO (2002) Mammalian ABC tran sporters in health and disease. Annual Review of Biochemistry 71:537-592 Boyer JN, Briceno HO (2005) FY2005 Annual Re port of the Water Quality Monitoring Project for the Water Quality Protection Program of the Florida Keys National Marine Sanctuary, Southeast Enviro nmental Research Center, Florida International University, Miami, FL

PAGE 261

Boyer JN, Jones RD (2002) A view from the bri dge: external and internal forces affecting the ambient water quality of the Flor ida Keys National Marine Sanctuary (FKNMS). In: Porter JW, Porter KG (eds) The Everglades, Florida Bay, and coral reefs of the Florida Keys: an ecosy stem sourcebook. CRC Press, Boca Raton, p 609-628 Brock RE, Lewis C, Wass RC (1979) Stability and structure of a fish community on a coral patch reef in Hawaii. Marine Biology 54:281-292 Bro-Rasmussen F, Lkke H ( 1984) EcoepidemiologyA casuist ic discipline describing ecological disturbances and damages in relation to their specific causes: Exemplified by chlorinated phenols and chlorophenoxy acids. Regulatory Toxicology and Pharmacology 4:391-399 Brown BE (1988) Assessing environmental impacts on coral reefs. Proceedings, 6th International Coral Reef Sy mposium, Australia p. 71-80 Brown BE (1997a) Adaptations of reef corals to physical environmental stress. Advances in Marine Biology 31:221-299 Brown BE (1997b) Coral bleaching: cause s and consequences. Coral Reefs 16:S129S138 Brown BE, Downs CA, Dunne RP, Gibb SW (2002a) Exploring the basis of thermotolerance in the reef coral Goniastrea aspera Marine Ecology Progress Series 242:119-129 Brown BE, Dunne RP, Chansang H (1996) Coral bl eaching relative to elevated seawater temperature in the Andamen Sea (Indian Ocean) over the last 50 years. Coral Reefs 15:151-152 Brown BE, Dunne RP, Goodson MS, Douglas AE (2002b) Experience shapes the susceptibility of a reef coral to bleaching. Coral Reefs 21:119-126 Bruckner AW (2002) Priorities for effec tive management of coral diseases, NOAA Technical Memorandum, NMFS-OPR-22 Bruno JF, Petes LE, Harvell CD, Hettinger A (2003) Nutrient enrichment can increase the severity of coral diseas es. Ecology Letters 6:1056-1061 Bryant D, Burke L, McManus J, Spalding M (1998) Reefs at risk: a map-based indicator of threats to the world's coral reefs. Washington, DC, World Resources Institute (web site: www.wri.org/indictrs/reefrisk.htm)

PAGE 262

Buddemeier RW, Kleypas JA, Aronson RB (2004) Coral reefs and globa l climate change: potential contributions of climate change to stresses on coral reef ecosystems, Pew Center on Global Climate Ch ange, Arlington, Virginia, USA Bythell JC, Gladfelter EH, Bythell M (1993) Chronic and catastrophic natural mortality of three common Caribbean reef corals. Coral Reefs 12:143-152 Callahan M (2005) Distribution of Clionid Sponges in the Florida Keys National Marine Sanctuary (FKNMS), 2001-2003. Master of Science Thesis University of South Florida, Tampa (website: http://etd .fcla.edu/SF/SFE0001017/Callahan-MichaelCallahanthesis2005.pdf) Candia Carnevali MD, Galassi S, Bonaso ro F, Patruno M, Thorndyke MC (2001) Regenerative response and endocrine di srupters in crinoid echinoderms: arm regeneration in Antedon mediterranea after experimental exposure to polychlorinated biphenyls. Journal of Experimental Biology 204:835-842 Cantillo AY, Lauenstein GG, O'Connor TP (1997) Mollusc and sediment contaminant levels and trends in South Florida coas tal waters. Marine Pollution Bulletin 34:511-521 Carnahan EA, Hoare AM, Hallock P, Lidz BH Reich CD (2007) Dist ributions of heavy metals and foraminiferal assemblages in sediments in Biscayne Bay, Florida, USA. Journal of Coastal Research (in press) Carpenter KE, Miclat RI, Albaladejo VD, Co rpuz VT (1981) The influence of substrate structure on the local abundance and diversity of Philippine reef fishes. Proceedings, 4th International Coral Reef Symposium 2:497-502 Causey B, Delaney J, Diaz E, Dodge D, Garcia JR, Higgins J, Jaap W, Matos CA, Schmahi GP, Rogers C, Miller MW, Turg eon DD (2000) Status of the coral reefs in the U.S. Caribbean and Gulf of Mexi co: Florida, Texas, Puerto Rico, U.S. Virgin Islands and Navassa. In: Wilkinson C (ed) Status of Coral Reefs of the World 2000. Global Coral Reef Monitoring Network. Australian Institute of Marine Science, Townsville, p 239-259 Chabanet P, Ralambondrainy H, Amanieu M, Faure G, Galzin R (1997) Relationships between coral reef substrata and fish. Coral Reefs 16:93-102 Choat JH (1991) The biology of herbivorous fi shes on coral reefs. In: Sale PF (ed) The ecology of fishes on coral reefs. Academic Press Inc., San Diego, pp 120-155 Clarke KR, Gorley RN (2006) PRIMER v6: User Manual/Tutorial. PRIMER-E Ltd, Plymouth

PAGE 263

Clarke KR, Warwick RM (2001) Change in marine communities: an approach to statistical analysis and interpretation. PRIMER-E Ltd, Plymouth Coble P, Hu C, Gould RW, Jr., Chang G, Wood AM (2004) Colored dissolved organic matter in the coastal ocean: an optical tool for coastal zone environmental assessment and management. Oceanography 17(2):50-59 Cockey E, Hallock P, Lidz BH (1996) Decadal-scale changes in benthic foraminiferal assemblages off Key Largo, Fl orida. Coral Reefs 15:237-248 Connell JH (1997) Disturbance and recovery of coral assemblages. Coral Reefs 16:S101S113 Costanza R, dArge R, deGroot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem S, Oneill RV, Paruelo J, Raskin RG, Sutton P, van denBelt M (1997) The value of the world's ecosystem services a nd natural capital. Nature 387:253-260 Coustau C, Chevillon C, ffrench-Constant R (2000) Resistance to xenobiotics and parasites: can we count the cost? TREE 15:378-383 Cowen RK, Lwiza KMM, Sponaugle S, Pari s CB, Olson DB (2000) Connectivity of marine populations: open or closed? Science 287:857-859 Croquer A, Villamizar E, Noriega N (2002) Environmental factors affecting tissue regeneration of the reef-building coral Montastraea annularis (Faviidae) at Los Roques National Park, Venezuela. Re vista de Biologa Tropical 50:1055-1065 Depledge MH, Amaral-Mendes JJ, Daniel B, Halbrook RS, Kloepper-Sams P, Moore MN, Peakall DB (1993) The conceptual basis of the biomarker approach. In: Peakall DB, Shugart LR (eds) Biomarke rs: Research and A pplication in the Assessment of Environmental Health, Vol 68. Springer-Verlag, Berlin, Heidelberg, p 15-29 Dodge RE, Aller RC, Thomson J (1974) Coral growth related to resuspension of bottom sediments. Nature 247:574-577 Done TJ (1992) Phase-Shifts in Cora l-Reef Communities and their ecological significance. Hydrobiologia 247:121-132 Done TJ (1997) Four performance indicators fo r integrated reef resources management. Workshop on Integrated Reef Resources Management in the Maldives, p 237-351 Downs CA (2005) Cellular diagnostics and its application to aquatic and marine toxicology. In: Ostrander G (ed) Tech niques in Aquatic Toxicology, Vol 2. CRC Press, Inc., Boca Raton, p 181-207

PAGE 264

Downs C, Downs A (2007) Preliminary examina tion of short-term cellular toxicological responses of the coral Madracis mirabilis to acute Irgarol 1051 exposure. Archives of environmental contam ination and toxicology 52:47-57 Downs CA, Dillon RT, Fauth JE, Woodley CM (2001a) A molecular biomarker system for assessing the health of gastropods ( Ilyanassa obsoleta ) exposed to natural and anthropogenic stressors Journal of E xperimental Marine Biology and Ecology 259:189-214 Downs CA, Fauth JE, John C H, Dustan P, Bemiss J, Woodley CM (2002) Oxidative stress and seasonal coral bleaching. Fr ee Radical Biology and Medicine 33:533543 Downs CA, Fauth JE, Robinson CE, Curry R, Lanzendorf B, Halas JC, Halas J, Woodley CM (2005a) Cellular diagnostics and coral h ealth: declining cora l health in the Florida Keys. Marine Pollution Bulletin 51:558-569 Downs CA, Fauth JE, Wetzel D, Hallock P, Halas JF, Halas JC, Curry R, Woodley CM (2006) Investigating co ral reef degradation at Alina's Reef in the Florida Keys: cellular physiology of white grunt ( Haemulon plumieri ) as a biological indicator. Environmental Forensics 7:15-32 Downs CA, Fauth JE, Woodley CM (2001b) Assessing the health of grass shrimp ( Palaeomonetes pugio ) exposed to natural and anthropogenic stressors: a molecular biomarker system. Marine Biotechnology 3:380-397 Downs CA, Mueller E, Phillips S, Fauth JE, Woodley CM (2000) A molecular biomarker system for assessing the health of coral ( Montastraea faveolata ) during heat stress. Marine Biotechnology 2:533-544 Downs CA, Woodley CM, Richmond RH, Lanning LL, Owen RR (2005b) Shifting the paradigm of coral-reef health assessment. Marine Pollution Bulletin 51:486-494 Drohan AF, Thoney DA, Baker AC (2005) Syne rgistic effect of high temperature and ultraviolet-B radiation on the gorgonian Eunicea tourneforti (Octocorallia : Alcyonacea : Plexauridae). Bulletin of Marine Science 77:257-266 Dustan P (1999) Coral reef under stress: sources of mortality in the Florida Keys. Natural Resources Forum 23:147-155 Dustan P (2000) Florida Keys. In: Sheppard C (ed) Seas at the Millennium: An Environmental Evaluation. El sevier Science, p 405-414 Dustan P, Halas JC (1987) Changes in th e Reef-Coral Community of Carysfort Reef, Key-Largo, Florida 1974 to 1982. Coral Reefs 6:91-106

PAGE 265

Dyhrman ST, Ruttenberg KC (2006) Presence and regulation of al kaline phosphatase activity in eukaryotic phytoplankton from the coastal ocean: Implications for dissolved organic phosphorus remine ralization. Limnology and Oceanography 51:1381-1390 Edinger EN, Limmon GV, Jompa J, Widjat moko W, Heikoop JM, Risk MJ (2000) Normal coral growth rate s on dying reefs: are coral growth rates good indicators of reef health? Marine Pollution Bulletin 40:404-425 Fabricius KE (2005) Effects of terrestrial runo ff on the ecology of corals and coral reefs: review and synthesis. Marine Pollution Bulletin 50:125-146 Fadlallah Y (1983) Sexual reproduction, develo pment and larval biology in scleractinian corals. A review. Coral Reefs 2:129-150 Fang L, Huang S, Lin K (1997) High temperat ure induces the synthesis of heat-shock proteins and the elevation of in tracellar calcium in the coral Acropora grandis Coral Reefs 16:127-131 Fauth JE, Downs CA, Halas JC, Dustan P, W oodley CM (2003) Midrange prediction of coral bleaching: a molecular diagnostic sy stem approach. In: Valette-Silver N, Scavia D (eds) Ecological forecasting: new tools for coastal and ecosystem management. NOAA Technical Me morandum NOC NCCOS, p 5-12 Fauth JE, Dustan P, Ponte E, Banks K, Va rgas-Angel B, Downs CA (2005) Final report: Southeast Florida coral biomarker lo cal action study. U.S. Environmental Protection Agency, 69 p. (http://www.dep.s tate.fl.us/coastal/programs/coral/ reports/Biomarker_F inal_Report_v4.pdf) Feyereisen R (1999) Insect P450 enzymes. Annual Review of Entomology 44:507-533 Fine M, Oren U, Loya Y (2002) Bleach ing effect on regeneration and resource translocation in the coral Oculina patagonica Marine Ecology Progress Series 234:119-125 Fisher EM, Fauth JE, Hallock P, Woodley CM (in press) Lesion re generation rates in reef-building corals ( Montastraea spp.) as indicators of colony condition: strengths and caveats. Marine Ecology Progress Series Fitt WK, Brown BE, Warner ME, Dunne RP (2001) Coral bleaching: interpretation of thermal tolerance limits and thermal thresholds in tropical corals. Coral Reefs 20:51-65 Fox G (1991) Practical causal inference for ecoepidemiologists. Journal of Toxicology and Environmental Health 33:359-373

PAGE 266

Fujita K, Hallock P (1999) A comparis on of phytal substrate preferences of Archais angulatus and Sorites orbiculus in mixed macroalgal-seagrass beds in Florida Bay. Journal of Foraminiferal Research 29:143-151 Fukami H, Knowlton N (2005) Analysis of complete mitochondrial DNA sequences of three members of the Montastraea annularis coral species complex (Cnidaria, Anthozoa, Scleractinia ). Coral Reefs 24:410-417 Gardinali PR, Plasencia M, Mack S, P oppell C (2002) Occurrence of Irgarol 1051 in coastal waters from Biscayne Bay, Flor ida, USA. Marine Pollution Bulletin 44:781-788 Gardinali PR, Plasencia MD, Maxey C (2004) Occurrence and transp ort of Irgarol 1051 and its major metabolite in coastal waters from South Florida. Marine Pollution Bulletin 49:1072-1083 Gardner TA, Cote IM, Gill JA, Grant A, Wa tkinson AR (2003) Long-term region-wide declines in Caribbean corals. Science 301:958-960 Ginsburg RN, Gischler E, Kien e WE (2001) Partial mortalit y of massive reef-building corals: an index of patch reef condition, Florida reef tract. Bulletin of Marine Science 69:1149-1173 Gleason DF, Wellington GM (1993) Ultraviolet-radiation an d coral bleaching. Nature 365:836-838 Glynn PW (1993) Coral-reef bleaching ecolo gical perspectives. Coral Reefs 12:1-17 Glynn PW (1996) Coral reef bleaching: Fact s, hypotheses and implications. Global Change Biology 2:495-509 Glynn PW, Szmant AM, Corcoran EF, Cofer-S habica SV (1989) Condition of coral reef cnidarians from the Northern Florida reef tract: pesticides, heavy metals, and histopathological examination. Marine Pollution Bulletin 20:568-576 Goreau TJ, Hayes RL (1994) Coral bleachi ng and ocean "hot spots". AMBIO 23:176-180 Goreau TJ, Cervino J, Goreau M, Smith G, Hayes R, Richardson L, Williams E, Nagelkerken I, Porter J, Porter K, Garz on-Ferreira J, Quirol o C, McField M, Gil D, Bruckner A, Santavy D, Peters E, Little r M, Littler D, Patterson K, J. C (1998) Rapid spread of diseases in Caribbean coral reefs. Costa Rica AMLC, Revista de Biologica Tropical 46:157-171

PAGE 267

Goreau TJ, Hayes RL, Clark JW, Basta DJ, Robertson CN (1992) Elevated sea surface temperatures correlate with Caribbean coral reef bleaching. In: Global Warming Forum: Scientific, Economic, and Le gal Overview. CRC Press, Boca Raton, Florida, p Section 1, Chapter 4 Green EP, Bruckner AW (2000) The significance of coral disease epizootiology for coral reef conservation. Biologi cal Conservation 96:347-361 Griffin DW, Gibson CJ, III, Lipp EK, Riley K, Paul JH, III, Rose JB (1999) Detection of viral pathogens by reverse transcript ase PCR and of microbial indicators by standard methods in the canals of the Florida Keys. Applied and Environmental Microbiology 65:4118-4125 Guinotte JM, Buddemeier RW, Kleypas JA (200 3) Future coral reef habitat marginality: temporal and spatial effects of climate change in the Pacifi c basin. Coral Reefs 22:551-558 Guzman HM, Burns KA, Jackson JBC (1994) Injury, regeneration and growth of Caribbean reef corals afte r a major oil spill in Pana ma. Marine Ecology Progress Series 105:231-241 Guzman HM, Jimenez CE (1992) Contamination of coral reefs by hea vy metals along the Caribbean coast of Central America (Cos ta Rica and Panama). Marine Pollution Bulletin 24:554-561 Hall VR (2001) The response of Acropora hyacinthus and Montipora tuberculosa to three different types of colony damage : scraping injury, tissue mortality and breakage. Journal of Experimental Marine Biology and Ecology 264:209-223 Hallock P (1988) The role of nutrient availa bility in bioerosion: consequences to carbonate buildups. Palaeogeography, Pa laeoclimatology, Palaeoecology 63:275291 Hallock P (1999) Symbiont-bearing foramini fera. In: Sen Gupta BK (ed) Modern Foraminifera. Kluwer Academic Publishers, Boston, p 123-139 Hallock P (2000) Larger foraminifera as indicat ors of coral-reef vita lity. In: Martin RE (ed) Environmental Micropaleontolo gy, Vol 15. Kluwer Academic/ Plenum Publishers, New York, p 121-150 Hallock P (2001) Coral Reefs, carbonate sedi ments, nutrients, and global change. In: Jr. GDS (ed) The History and Sedimentology of Ancient Reef Systems. Kluwer Academic/Plenum Publishers, New York, p 387-427 Hallock P (2005) Global change and modern coral reefs: New opportunities to understand shallow-water carbonate depositional processes. Sedimentary Geology 175:19-33

PAGE 268

Hallock P, Talge HK (1994) A predatory foraminifera, Floresina amphiphaga n. sp., from the Florida Keys. Journal of Fo raminiferal Research 24(4):210-213 Hallock P, Barnes K, Fisher EM (2004) Cora l-reef risk assessment from satellites to molecules: a multi-scale approach to environmental monitoring and risk assessment of coral reefs. Journal of Environmental Micropaleontology Microbiology and Meiobenthology 1:11-39 Hallock P, Cottey TL, Forward LB, Halas J (1986a) Population dynamics and sediment production by Archaias angulatus (Foraminiferida) in Largo Sound, Florida. Journal of Foraminiferal Research 16:1-18 Hallock P, Forward LB, Hansen HJ (1986b) Influence of environment on the test shape of Amphistegina Journal of Foraminiferal Research 16:224-231 Hallock P, Lidz BH, Cockey-Burkhard EM, Donnelly KB (2003) Foraminifera as bioindicators in coral reef assessmen t and monitoring: The FORAM Index. Environmental Monitoring and Assessment 81:221-238 Hallock P, Muller-Karger FE, Halas JC (1993) Coral reef decline. National Geographic Research and Exploration 9:358-378 Hallock P, Peebles MW (1993) Foraminifera with chlorophyte endosymbionts: habitats of six species in the Florida Ke ys. Marine Microp aleontology 20:277-292 Hallock P, Premoli-Silva I, Boersma A (1991) Similarities between planktonic and larger foraminiferal evolutionary trends thro ugh Paleogene paleoceanographic changes. Palaeogeography, Palaeoclimatol ogy, Palaeoecology 83:49-64 Hallock P, Talge HK, Cockey EM, Muller RG (1995) A new disease in reef-dwelling foraminifera: implications for coastal sedimentation. Journal of Foraminiferal Research 25:280-286 Hallock P, Talge HK, Smith K, Cockey EM (1992) Bleaching in a Reef-Dwelling Foraminifer, Amphistegina gibbosa. Proceedings, 7th International Coral Reef Symposium, Guam, p 44-49 Hallock P, Williams DE, Fisher EM, Toler SK (2006a) Bleaching in foraminifera with algal symbionts: Implications for reef m onitoring and risk assessment. Anurio do Instituto de Geoc incias 29:108-128 Hallock P, Williams DE, Toler SK, Fisher EM, Talge HK (2006b) Bleaching in reefdwelling foraminifers: a 20 year retrospective. Proceedings, 10th International Coral Reef Symposium, Okinawa, p 729-737

PAGE 269

Harman-Fetcho JA, Hapeman CJ, McConnell LL, Potter TL, Rice CP, Sadeghi AM, Smith RD, Bialek K, Sefton KA, Sc haffer BA, Curry R (2005) Pesticide occurrence in selected South Florida canals and biscayne bay during high agricultural activity. Jour nal of Agricultural and Food Chemistry 53:6040-6048 Harney JN, Hallock P, Talge HK (1998) Ob servations on a trimorphic life cycle in Amphistegina gibbosa populations from the Florida Keys. Journal of Foraminiferal Research 28:141-147 Harvell D, Aronson R, Baron N, Connell J, Dobs on A, Ellner S, Gerber L, Kim K, Kuris A, McCallum H, Lafferty K, McKay B, Port er J, Pascual M, Smith G, Sutherland K, Ward J (2004) The rising tide of ocean diseases: unsolved problems and research priorities. Frontiers in Ecology and the Environment 2:375-382 Hawkins AJS (1991) Protein turnover: a func tional appraisal. F unctional Ecology 5:222233 Hayes ML, Bonaventura J, Mitc hell TP, Prospero JM, Shi nn EA, Van Dolah F, Barber RT (2001) How are climate and marine biological outbreaks f unctionally linked? Hydrobiologia 460:213-220 Heckathorn SA, Poehler GJ, Coleman JS, Hallber g RL (1996) Nitrogen availability alters patterns of accumulation of heat-stress-induced proteins in plants. Oecologia 105: 413-418 Henry L-A, Hart M (2005) Regeneration from injury and resource allocation in sponges and corals a review. Internati onal Review of Hydrobiology 90:125-158 Hixon MA (1991) Predation as a process structuring coral-reef fish communities. In: Sale PF (ed) The Ecology of Fishes on Coral Reefs. Academic Press, San Diego, p 475-508 Hodgson G (1990) Sediment and the settle ment of larvae of the reef coral Pocillopora damicornis Coral Reefs 9:41-43 Hoegh-Guldberg O (1999) Climate change, co ral bleaching and the future of the world's coral reefs. Marine and Fr eshwater Research 50:839-866 Hoffmeister JE, Multer HG (1964) Growth-rate es timates of a Pleistoc ene coral reef of Florida. Geological Society of America Bulletin 75:353-358 Hofmann GE, Somero GN (1995) Evidence fo r protein damage at environmental temperatures: seasonal changes in levels of ubiquitin conjugates and hsp70 in the intertidal mussel Mytilus trossulus Journal of Experi mental Biology 198:1509 1518

PAGE 270

Hubbard DK (1986) Sedimentation as a contro l of reef development St. Croix, USVI. Coral Reefs 5:117-125 Hudson JH (1981) Response of Montastrea annularis to environmental change in the Florida Keys. Proceedings, 4th Internati onal Coral Reef Symposium, Manila, p 233-240 Hudson JH, Shinn EA, Halley RB, Lidz B (1976) Sclerochr onology a tool for interpreting past environments. Geology 4:361-364 Hughes TP (1994) Catastrophes, phase shifts, and large-sclae degradation of a Caribbean coral reef. Science 265:1547-1551 Hughes TP, Connell JH (1999) Multiple stressors on coral reefs: a long-term perspective. Limnology and Oceanography 44:932-940 Hughes TP, Jackson JBC (1980) Do corals lie about their age some demographic consequences of partial mortalit y, fission, and fusion. Science 209:713-715 Hughes TP, Jackson JBC (1985) Population dyn amics and life histories of foliaceous corals. Ecological Monographs 55:141-166 Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C, Grosberg R, Hoegh-Guldberg O, Jackson JBC, Kle ypas JA, Lough LM, Marshall P, Nystrom M, Palumbi SR, Pandolfi JM, Rosen B, Roughgarden J (2003) Climate change, human impacts, and the resilience of coral reefs. Science 301:929-932 Jaap WC, Porter JW, Wheaton J, Hackett K, Lybolt M, Callahan M, Tsokos C, Yanev G, Dustan P (2000) Coral reef monitoring project executive summary, EPA Science Advisory Panel, Key Colony Beach Jameson SC, Erdmann MV, Karr JR, Potts KW (2001) Charting a course toward diagnostic monitoring: A continuing review of coral reef attributes and a research strategy for creating coral re ef indexes of biotic inte grity. Bulletin of Marine Science 69:701-744 Jameson SC, Tupper MH, Ridley JM (2002) The three screen doors: Can marine "protected" areas be effective. Marine Pollution Bulletin 44:1177-1183 Johns GM, Leeworthy VR, Bell FW, Bonn MA (2001) Socioeconomic study of reefs in Southeast Florida, Final Report. Br oward County Environmental Protection Department, Technical Report 01-10 Jokiel PL (1980) Solar Ultr aviolet-Radiation and Cora l-Reef Epifauna. Science 207:1069-1071

PAGE 271

Jones RJ, Hoegh-Guldberg O, Larkum AWD, Schreiber U (1998) Temperature-induced bleaching of corals begins with impairment of the CO2 fixation mechanism in zooxanthellae. Plant Cell and Environment 21:1219-1230 Jones R (2005) The ecotoxicological effect s of Photosystem II herbicides on corals. Marine Pollution Bulletin 51:495-506 Jones RJ, Hoegh-Guldberg O, Larkum AWD, Schreiber U (1998) Temperature-induced bleaching of corals begins with impairment of the CO2 fixation mechanism in zooxanthellae. Plant Cell and Environment 21:1219-1230 Joyner-Matos J, Downs CA, Julian D (2006) In creased expression of stress proteins in the surf clam Donax variabilis following hydrogen sulfide exposure. Comparative Biochemistry and Physiology a-Molecu lar & Integrative Physiology 145:245-257 Kaczmarsky LT, Draud M, Williams EH (2005) Is there a relationship between proximity to sewage effluent and the prevalence of coral disease. Ca ribbean Journal of Science 41:124-137 Kawaguti S (1937) On the physiology of reef corals. III. Regeneration and phototropism in reef corals. Palao Tropical Bi ological Station Studies 1:209-216 Kendall JJ, Powell EN, Conner SJ, Bright TJ (1 983) The effect of drilling fluids (mud) and turbidity on the growth and metabolic state of the coral Acropora cervicornis with comments on methods of normalizati on for coral data. Bulletin of Marine Science 33:336-352 Kirk JTO (1996) Light and Phot osynthesis in Aquatic Ecosystems. Cambridge University Press, Bristol, 509 p. Kleypas JA, Buddemeier RW, Gattuso JP (2001) The future of coral reefs in an age of global change. Internati onal Journal of Earth Sciences 90 (2): 426-437 Koehn RK, Bayne BL (1989) Towards a physiol ogical and energetic understanding of the energetics of the stress response. Biol ogical Journal of the Linnean Society 37:157-171 Kojis BL, Quinn NJ (1984) Seasonal a nd depth variation in fecundity of Acropora palifera two reefs in Papua New Guinea. Coral Reefs 3:165-172 Koop K, Booth D, Broadbent A, Brodie J, Bucher D, Capone D, Coll J, Dennison W, Erdmann M, Harrison PL, Hoegh-Guldberg O, Hutchings P, Jones GB, Larkum AWD, O'Neil J, Steven A, Tentori E, Wa rd S, Williamson J, Yellowlees D (2001) ENCORE: The effect of nutrient enrichment on coral reefs. Synthesis of results and conclusions. Marine Pollution Bulletin 42:91-120

PAGE 272

Kramarsky-Winter E (2004) What can regenerati on processes tell us about coral disease? In: Rosenberg E, Loya Y (eds) Coral Health and Disease. Springer, New York, p 217-230 Kramarsky-Winter E, Loya Y (2000) Tissue regeneration in the coral Fungia granulosa : the effect of extrinsi c and intrinsic factors. Marine Biology 137:867-873 Kramer PA (2003) Synthesis of coral reef health indicato rs for the Western Atlantic: results of the AGRRA program (1997-2000) Atoll Research Bulletin 496:1-58 Kruczynski WL, McManus F (2002) Water quality concerns in the Florida Keys: sources, effects, and solutions. In: Porter JW, Porter KG (eds) The Everglades, Florida Bay, and Coral Reefs of the Florida Keys: An Ecosystem Sourcebook. CRC Press, Boca Raton, p 827-881 Kuffner IB, Brock JC, Grober-Dunsmore R, Bo nito VE, Hickey TD, Wright CW (2007) Relationships between reef fish communities and remotely sensed rugosity measurements in Biscayne National Par k, Florida, USA. Environmental Biology of Fishes 78:71-82 Kunkel CM, Hallberg RW, Oppenheimer M (2006) Coral reefs reduce tsunami impact in model simulations. Geophysic al Research Letters 33:612 Kuta KG, Richardson LL (1997) Black band di sease and the fate of diseased coral colonies in the Florida Keys. Proceedings, 8th International Coral Reef Symposium 1: 575-578 Lang J, Alcolado P, Pablo Carricart-Ganivet J, Chiappone M, Curran A, Dustan P, Gaudian G, Geraldes F, Gittings S, Smith R, Tunnel W, Wiener J (1998) Status of coral reefs in the northern areas of the wider Caribbean. In: Wilkinson C (ed) Status of Coral Reefs of the Worl d: 1998. Global Coral Reef Monitoring Network. Australian Institute of Ma rine Science, Townsville, p 123-134 Lang JC (ed) (2003) Status of coral reefs in the Western Atlantic: Results of initial surveys, Atlantic and Gulf Rapid R eef Assessment (AGRRA) program. Atoll Research Bulletin, Vol 496. National Muse um of Natural History Smithsonian Institution, Washington, D.C. Langer MR, Hottinger L (2000) Biogeography of selected "larger" foraminifera. Micropaleontology 46:105-126 Lapointe BE (1997) Nutrient thresholds fo r eutrophication and macroalgal blooms on coral reefs in Jamaica and southeas t Florida. Limnology and Oceanography 42:1119-1131

PAGE 273

Lapointe BE (1999) Simultaneous top-down a nd bottom-up forces control macroalgal blooms on coral reefs (Reply to the comment by Hughes et al.). Limnology and Oceanography 44:1586-1592 Lapointe BE, Matzie WR (1996) Effects of stormwater nutrient discharges on eutrophication processes in nearshore wa ters of the Florida Keys. Estuaries 19:422-435 Laws EA, Redalje DJ (1979) Effect of sewage enrichment on the phytoplankton population of a tropical estuar y. Pacific Science 33:129-144 Lee JJ, Anderson OR (1991) Symbiosis in fo raminifera. In: Lee JJ, Anderson OR (eds) Biology of Foraminifera. Academic Press, San Diego, p 157-220 Lee JJ, McEnery ME, Garrison JR (1980) Experime ntal studies of larg er foraminifera and their symbionts from the Gulf of Elat on the Red Sea. Journal of Foraminiferal Research 10:31-47 Lesser MP (1996) Elevated temperatures and ultraviolet radiation cause oxidative stress and inhibit photosynthesis in symb iotic dinoflagellates. Limnology and Oceanograpy 41:271-283 Lesser MP (1997) Oxidative stress causes co ral bleaching during exposure to elevated temperatures. Coral Reefs 16:187-192 Lesser MP (2006) Oxidative stress in marine environm ents: biochemistry and physiological ecology. Annual Review of Physiology 68:253-278 Lesser MP, Farrell JH (2004) E xposure to solar radiation in creases damage to both host tissues and algal symbionts of corals during thermal stress. Coral Reefs 23:367377. Lesser MP, Stochaj WR, Tapley DW, Sh ick JM (1990) Bleaching in coral reef anthozoans: effects of irradiance, ultrav iolet radiation, and temperature on the activities of protective enzymes against active ox ygen. Coral Reefs 8:225-232 Lester RT, Bak RPM (1985) Effects of envir onment on regeneration rate of tissue lesions in the reef coral Monstastrea annularis (Scleractinia). Ma rine Ecology Progress Series 24:183-185 Levy A (1991) Present Distribution of Soritid s and Peneroplids and Thanatocenoses of the Florida-Keys (USA). Oceanologica Acta 14:515-524 Lidz BH, Rose PR (1989) Diagnostic forami niferal assemblages of Florida Bay and adjacent shallow waters a comparison. Bulletin of Marine Science 44:399-418

PAGE 274

Lipp EK, Jarell JL, Griffin DW, Lukasik J, Jacukiewicz J, Rose JB (2002) Preliminary evidence for human fecal contamination in corals of the Florida Keys, USA. Marine Pollution Bulletin 44:666-670 Lirman D (2000a) Fragmentati on in the branching coral Acropora palmata (Lamarck): growth, survivorship, and reproduction of colonies a nd fragments. Journal of Experimental Marine Bi ology and Ecology 251:41-57 Lirman D (2000b) Lesion regene ration in the branching coral Acropora palmata : effects of colonization, colony size, lesion size and lesion shape. Marine Ecology Progress Series 197:209-215 Lirman D (2001) Competition between macr oalgae and corals: effects of herbivore exclusion and increased algal biomass on coral survivorship and growth. Coral Reefs 19:392-399 Lirman D, Biber P (2000) Seasonal dynamics of macroalgal commun ities of the Northern Florida reef tract. Bo tanica Marina 45:303-314 Littler MM, Littler DS, Brooks BL (2006) Ha rmful algae on tropical coral reefs: Bottomup eutrophication and top-down herbivory. Harmful Algae 5:565-585 Lopez JV, Kersanach R, Rehner SA, Knowlt on N (1999) Molecular determination of species boundaries in corals : Genetic analysis of the Montastraea annularis complex using amplified fragment length polymorphisms and a microsatellite marker. Biological Bulletin 196:80-93 Marks KW, Klomp KD (2003) Fish biomass conversion equations. Atoll Research Bulletin 496:625-628 Marshall P, Schuttenberg H (2006) A Reef Manager's Guide to Coral Bleaching, Great Barrier Reef Marine Park Authority, Townsville Mascarelli PE, Bunkley-Williams L (1999) An e xperimental field evaluation of healing in damaged, unbleached and artifici ally bleached star coral, Montastrea annularis Bulletin of Marine Science 65:577-586 McCarty IS, Power M, Munkittrick KR (2002) Bioindicators versus biomarkers in ecological risk assessment. Human a nd Ecological Risk Assessment 8:159-164 McKenna SA, Richmond RH, Roos G (1999) Qu antifying sub-lethal e ffects of stress in corals: Can growth and fec undity of corals be used to detect stress before mortality? American Zoologist 39:724 Meesters EH, Bak RPM (1993) Effects of cora l bleaching on tissue re generation potential and colony survival. Marine Ec ology Progress Series 96:189-198

PAGE 275

Meesters EH, Bos A, Gast GJ (1992) Effects of sedimentation and lesion position on coral regeneration. Proceedings, 7th Intern ational Coral Reef Symposium, Guam, p 671-678 Meesters EH, Noordeloos M, Bak RPM (1994) Damage and regeneration: links to growth in the reef-building coral Montastrea annularis Marine Ecology Progress Series 112:119-128 Meesters EM, Pauchli W, Bak RPM (1997a) Pr edicting regeneration of physical damage on a reef-building coral by regenerati on capacity and lesion shape. Marine Ecology Progress Series 146:91-99 Meesters EH, Weeseling I, Bak RPM (1997b) Co ral colony tissue damage in six species of reef-building corals: pa rtial mortality in relation with depth and surface area. Journal of Sea Research 37:131-144 Meesters EH, Wesseling I, Bak RPM (1996) Part ial mortality in three species of reefbuilding corals and the relation with colony morphology. Bulletin of Marine Science 58:838-852 Mendes JM, Risk MJ, Schwarcz HP, Woodley JD (1997) Stable isotopes of nitrogen as measures of marine pollution: a preliminar y assay of coral tissue from Jamaica. Proceedings, 8th International Coral Reef Symposium 2:1869-1872 Mendes JM, Woodley JD (2002) Timing of reproduction in Montastraea annularis : relationship to environmental variables. Marine Ecology Progress Series 227:241251 Mergner H (1971) Cnidaria. In: Reverberi G (ed) Experimental Embryology of Marine and Fresh-water Invertebrates. Nort h-Holland Publishing Company, Amsterdam Miles CJ, Pfeuffer RJ (1997) Pesticides in canals of south Florida. Archives of Environmental Contamination and Toxicology 32:337-345 Miller J, Waara R, Muller E, Rogers C (2006) Coral bleaching and disease combine to cause extensive mortality on reefs in US Virgin Is lands. Coral Reefs 25:418-418 Miller MW, Hay ME (1998) Effects of fish predation and seaweed competition on the survival and growth of corals. Oecologia 113:231-238 Miller MW, Weil E, Szmant AM (2000) Coral recruitment and juvenile mortality as structuring factors for reef benthic comm unities in Biscayne National Park, USA. Coral Reefs 19:115-123 Mills MM, Sebens KP (2004) Ingestion and assimilation of nitrogen from benthic sediments by three species of coral. Marine Biology 145:1097-1106

PAGE 276

Monroe County (2000) Monroe County Sa nitary Wasterwate r Master Plan. http://monroecofl.virtualtownhall.net/pa ges/MonroeCoFL_MarineRes/wastewater Moore MN (1990) Lysosomal cytochemistr y in marine environmental monitoring. Histochemical Journal 22:187-191 Moore MN (2002) Biocomplexity: the post-ge nome challenge in ecotoxicology. Aquatic Toxicology 59:1-15 Moore MN, Allen JI, McVeigh A (2006) Enviro nmental prognostics: an integrated model supporting lysosomal stress responses as pr edictive biomarkers of animal health status. Marine Environmental Research 61:278-304 Moore MN, Depledge MH, Readman J, W ., Leonard DRP (2004) An integrated biomarker-based strategy for ecotoxicologi cal evaluation of risk in environmental management. Mutation Research 552:247-268 Moore MN, Simpson MG (1992) Molecular and cellular pathology in environmental impact assessment. Aquatic Toxicology 22:313-322 Moran MA, Sheldon MJ (2000) Carbon loss and optical property changes during longterm photochemical and biological degrad ation of estuarine dissolved organic matter. Limnology and Oceanography 45:1254-1264 Morgan MB, Snell TW (2002) Characterizing stress gene expression in reef-building corals exposed to the mosquitoside dibrom. Marine Pollution Bulletin 44:12061218 Morrison D (1988) Comparing fish and urchin grazing in shallow and deeper coral-reef algal communities. Ecology 69:1367-1382 Muller PH (1978) 14Carbon fixation and symbiont loss in a foraminiferal-algal symbiont system. Journal of Foraminiferal Research 8:35-41 Mundy PL (2002) Does habitat availability determine geographical-scale abundances of coral-reef dwelling fish es? Coral Reefs 21:105-116 Muscatine L, Falkowski PG, Dubinsky Z, C ook PA, McCloskey LR (1989) The effect of external nutrient resources on the population dynamics of zooxanthellae in a reef coral. Proceedings, Royal Society of London. Series B, Biological Sciences 236 (1284):311-324. Nagelkerken I, Bak RMP (1998) Differentia l regeneration of ar tificial lesions among sympatric morphs of the Caribbean corals Porites astreoides and Stephanocoenia michelinii Marine Ecology Progress Series 163:279-283

PAGE 277

Nagelkerken I, Meesters EH, Bak RPM (1999) Depth-related va riation in regeneration of artificial lesions in the Caribbean corals Porties astreoides and Stephanocoenia michelinii Journal of Experimental Ma rine Biology and Ecology 234:29-39 Nugues MM, Szmant AM (2006) Coral settlement onto Halimeda opuntia : a fatal attraction to an ephemeral s ubstrate? Coral Reefs 25:585-591 Oren U, Benayahu Y, Loya Y (1997) Effect of lesion size and shape on regeneration of the Red Sea coral Favia favus. Mari ne Ecology Progress Series 146:101-107 Oren U, Benayahu Y, Lubinevsky H, L oya Y (2001) Colony integration during regeneration in the stony coral Favia favus Ecology 82:802-813 Ostrander GK, Armstrong KM, Knobbe ET, Ger ace D, Scully EP (2000) Rapid transition in the structure of a coral reef commun ity: The effects of coral bleaching and physical disturbance. Proceedings of th e National Academy of Sciences of the United States of America 97:5297-5302 Owen R, Knap A, Ostrander N, Carbery K (2003) Comparative acute toxicity of herbicides to photosynthesis of coral z ooxanthellae. Bulletin of Environmental Contamination and Toxicology 70:541-548 Owen R, Knap A, Toaspern M, Carbery K (2002) Inhibition of coral photosynthesis by the antifouling herbicide Irgarol 105 1. Marine Pollution Bulletin 44:623-632 Palandro DA (2006) Coral reef habitat change and water clarity a ssessment (1984-2002) for the Florida Keys National Marine Sa nctuary using Landsat satellite data. Ph.D. Dissertation, University of South Florida, Tampa, FL Palandro D, Andrefouet S, Dustan P, Muller-Kar ger FE (2003) Change detection in coral reef communities using Ikonos satellite sensor imagery and historic aerial photographs. International Journa l of Remote Sensing 24:873-878 Pandolfi JM, Jackson JBC, Baron N, Br adbury RH, Guzman HM, Hughes TP, Kappel CV, Micheli F, Ogden JC, Possingham HP, Sala E (2005) Ecology Are US coral reefs on the slippery slope to slime? Science 307:1725-1726 Patruno M, Thorndyke MC, Carnevali MDC, B onasoro F, Beesley P (2001) Changes in ubiquitin conjugates and Hsp72 levels duri ng arm regeneration in echinoderms. Marine Biotechnology 3:4-15 Patterson KL, Porter JW, Ritchie KB, Polson SW, Mueller E, Peters EC, Santavy DL, Smith GW (2002) The etiology of white pox, a lethal disease of the Caribbean elkhorn coral, Acropora palmata Proceedings, National Acadademy of Sciences, USA 99:8725-8730

PAGE 278

Paul JH, McLaughlin MR, Griffin DW, Li pp EK, Stokes R, Rose JB (2000) Rapid movement of wastewater from on-site dis posal systems into surface waters in the lower Florida Keys. Estuaries 23:662-668 Paul JH, Rose JB, Brown J, Shinn EA, Mille r S, Farrah SR (1995a) Viral tracer studies indicate contamination of marine waters by sewage disposal practices in Key Largo, Florida. Applied and E nvironmental Microbiology 61:2230-2234 Paul JH, Rose JB, Jiang S, Kellogg C, Shi nn EA (1995b) Occurrence of fecal indicator bacteria in surface waters and the subs urface aquifer in Key Largo, Florida. Applied Environmental Microbiology 61:2235-2241 Paul JH, Rose JB, Jiang SC, Zhou X, Co chran P, Kellog CA, Kang JB, Griffin DW, Farrah SR, Lukasik J (1997) Evidence for groundwater and surface marine water contamination by waste disposal wells in the Florida Keys. Water Research 31:1448-1454 Paz-Garcia DA, Reyes-Bonilla H (2006) Tempor al variation in the regeneration rate of artificial lesions in two morphotypes of Porites panamensis Ciencias Marinas 32:187-194 Pelejero C, Calvo E, McCulloch MT, Marshall JF, Gagan MK, Lough JM, Opdyke BN (2005) Preindustrial to modern interdecadal variability in coral reef pH. Science 309:2204-2207 Peters EC (1997) Diseases of coral-reef organisms. In: Birk eland C (ed) Life and Death of Coral Reefs. Kluwer Academ ic Publishers, Boston, p 114-139 Petersen D, Laterveer M, Schuhmacher H (2005 ) Spatial and temporal variation in larval settlement of reefbuilding corals in mariculture. Aquaculture 249:317-327 Pfeuffer RJ, Rand GM (2004) South Florid a ambient pesticide monitoring program. Ecotoxicology 13:195-205 Pierce RH (1998) Effects of Mosquito Contro l Measures on Non-Targeted Organisms in the Florida Keys National Marine Sanctu ary. Technical Report 609. Mote Marine Laboratory, Sarasota, Florida USA, 40 pp Porter JW, Dustan P, Jaap WC, Patters on KL, Kosmynin V, Meier OW, Patterson ME, Parsons M (2001) Patterns of spread of coral disease in the Florida Keys. Hydrobiologia 460:1-24

PAGE 279

Porter JW, Kosmynin V, Patterson KL, Port er KG, Jaap WC, Wheaton JL, Hackett K, Lybolt M, Tsokos CP, Yanev G, Marcinek DM, Dotten J, Eaken D, Patterson M, Meier OW, Brill M, Dustan P (2002) Detection of coral reef change by the Florida Keys coral reef monitoring progr am. In: Porter JW, Porter KG (eds) The Everglades, Florida Bay, and Coral Reef s of the Florida Keys: an ecosystem sourcebook. CRC Press, Boca Raton, p 749-769 Porter JW, Lewis SK, Porter KG (1999) The e ffect of multiple stressors on the Florida Keys coral reef ecosystem: a landscap e hypothesis and a physiological test. Limnol. Oceanogr. 44:941-949 Reich CD, Shinn EA, Hickey TD, Tihans ky AB (2002) Tidal and meteorological influences on shallow marine groundwate r flow in the upper Florida Keys. In: Porter JW, Porter KG (eds) The Everglades, Florida Bay, and the Coral Reefs of the Florida Keys: an ecosystem s ourcebook. CRC Press, Boca Raton, p 659-675 Reichelt-Brushett AJ, Harrison PL (2005) Th e effect of selected trace metals on the fertilization success of se veral scleractinian coral species. Coral Reefs 24:524-534 Richardson LL (1998) Coral diseases: what is really known? Trends in Ecology and Evolution 13 (11): 438-443 Richmond RH (1993) Coral reefs: present prob lems and future concerns resulting from anthropogenic disturbance. American Zoologist 33:524-536 Richmond RH (1997) Reproduction and recruitm ent in corals: cri tical links in the persistence of reefs. In: Birkeland C (ed) Life and Death of Coral Reefs. Chapman and Hall, New York, p 175-197 Rinkevich B (1996) Do reproduction and rege neration in damaged corals compete for energy allocation? Marine Ec ology Progress Series 143:297-302 Risk MJ (1999) Paradise lost: how marine sc ience failed the world's coral reefs. Marine Freshwater Research 50:831-837 Roberts CM, Ormond RFG (1987) Habitat comple xity and coral-reef fish diversity and abundance on red-sea fringing reefs. Ma rine Ecology Progress Series 41:1-8 Rogers CS (1983) Sublethal and lethal effect s of sediments applied to common Caribbean reef corals in the field. Mari ne Pollution Bulletin 14:378-382 Rogers CS (1990) Responses of coral reefs a nd reef organisms to sedimentation. Marine Ecology Progress Series 62:185-202 Rogers CS, Garrison G, Grober R, Hillis Z, Franke MA (1994) Coral reef monitoring manual for the Caribbean and Western A tlantic. Virgin Islands National Park

PAGE 280

Rosenfeld M, Bresler V, Abelson A (1999) Sediment as a possible source of food for corals. Ecology Letters 2:345-348 Roth AA, Clausen CD, Yahiku PY, Clausen VE, Cox WW (1982) Some effects of light on coral growth. Pacific Science 36:65-81 Santavy DL, Summers JK, Engle VD, Harwell LC (2005) The condition of coral reefs in South Florida (2000) using coral dise ase and bleaching as indicators. Environmental Monitoring and Assessment 100:129-152 Severance EG, Karl SA (2006) Contrasting popu lation genetic structur es of sympatric, mass-spawning Caribbean corals Marine Biology 150:57-68 Severance EG, Szmant AM, Karl SA (2004a) Microsatellite loci isolated from the Caribbean coral, Montastraea annularis Molecular Ecology Notes 4:74-76 Severance EG, Szmant AM, Karl SA (2004b) Si ngle-copy gene markers isolated from the Caribbean coral, Montastraea annularis Molecular Ecology Notes 4:167-169 Seyle H (1955) Stress and disease. Science 122:625-631 Shick JM, Lesser MP, Jokiel PL (1996) Eff ects of ultraviolet radiation on corals and other coral reef organisms. Global Change Biology 2:527-545 Shinn EA, Reese RS, Reich CD (1994) Fate a nd pathways of injection-well effluent in the Florida Keys, U.S. Geologica l Survey Open-File Report 94-276 Shinn EA, Smith GW, Prospero JM, Betzer P, Hayes ML, Garrison V, Barber RT (2000) African dust and the demise of Caribb ean coral reefs. Geophysical Research Letters 27:3029-3032 Shinn EA, Reich CD, Hickey TD (2002) S eepage meters and Bernoulli's revenge. Estuaries 25:126-132 Simpkins CO, Zhao HL, Torrence CA ( 1994) Effect of metallothionein I on mitochondrial oxygen consumption. Life Sciences 55:221-226 Sindermann CJ (1996) Ocean pollution: eff ects on living resources and humans. CRC Press, Boca Raton, 275 p Siroka Z, Drastichova J (2004) Bioche mical markers of aquatic environment contamination Cytochrome P450 in fi sh. A review. Acta Veterinaria Brno 73:123-132

PAGE 281

Smith JE, Shaw M, Edwards RA, Obura D, Pantos O, Sala E, Sandin SA, Smriga S, Hatay M, Rohwer FL (2006) Indirect e ffects of algae on coral: algae-mediated, microbe-induced coral mortal ity. Ecology Letters 9:835-845 Sokal RR, Rohlf FJ (1995) Biometry, 3rd e d., Vol. W. H. Freeman and Company, New York Soto I (2006) Environmental variability in th e Florida Keys: Impacts on coral reef health. Master of Science Thesis, Univer sity of South Florida, Tampa Steneck RS, Dethier M (1994) A functional-gr oup approach to the structure of algaldominated communities. Oikos 69:476-498 Strong AM, Bancroft GT (1994) Patterns of de forestation and fragmentation of mangrove and deciduous seasonal fore sts in the Upper Florida Keys. Bulletin of Marine Science 54:795-804 Suter GW, Norton SB, Cormier SM (2002) A methodology for inferring the causes of observed impairments in aquatic ecosy stems. Environmental Toxicology and Chemistry 21:1101-1111 Sutherland KP, Porter JW, Torres C (2004) Disease and immunity in Caribbean and Indo-Pacific zooxanthellae corals. Marine Ecology Progress Series 266:273-302 Syms C, Jones GP (2000) Disturbance, habitat structure, and the dynamics of a coral reef fish community. Ecology 81:2714-2729 Szmant AM (1986) Reproductive ecology of Ca ribbean reef corals Coral Reefs 5:43-54 Szmant AM (1991) Sexual reproducti on by the Caribbean reef corals Montastrea annularis and M. cavernosa Marine Ecology Prog ress Series 74:13-25 Szmant AM (2001) Why are coral reefs wo rld-wide becoming overgrown by algae? 'Algae, algae everywhere, and nowhere a bite to eat!' Coral Reefs 19:299-302 Szmant AM, Forrester A (1996) Water colu mn and sediment nitrogen and phosphorous distribution patterns in the Florid a Keys, USA. Coral Reefs 15:21-41 Szmant AM, Weil E, Miller MW, Colon DE (1997) Hybridization within the species complex of the scleractinan coral Montastraea annularis Marine Biology 129:561-572 Talge HK, Hallock P (1995) Cytological exam ination of symbiont loss in a benthic foraminifera, Amphistegina gibbosa Marine Micropale ontology 26:107-113

PAGE 282

Talge HK, Hallock P (2003) Ultrastruc tural responses to bleaching in Amphistegina gibbosa ( Foraminifera). Journal of Eucaryotic Microbiology 50:324-333 Tang W, Kido T, Gross WA, Nogawa K, Sabbi oni E, Shaikh ZA (1999) Measurement of cadmium-induced metallothionein in urine by ELISA and prevention of overestimation due to polymerization. J ournal of Analytical Toxicology 23:153158 Tarrant AM (2004) Love on the rocks? Horm onal signaling and signal disruption in coral reef ecosystems. Chemical Re search in Toxicology 17:1763-1763 Tarrant AM (2005) Endocrine-like signaling in cnidarians: Current understanding and implications for ecophysiology. Integr ative and Comparative Biology 45:201-214 Tarrant AM, Atkinson MJ, Atki nson S (2001) Uptake of estrone from the water column by a coral community. Marine Biology 139:321-325 Tarrant AM, Atkinson MJ, Atki nson S (2004) Effects of st eroidal estrogens on coral growth and reproduction. Marine Ecology Progress Series 269:121-129 Tarrant AM, Blomquist CH, Lima PH, Atki nson MJ, Atkinson S (2003) Metabolism of estrogens and androgens by scleractinian corals. Comparative Biochemistry and Physiology B-Biochemistry & Molecular Biology 136:473-485 Telesnicki GJ, Goldberg WM (1995) Effect s of turbidity on the photosynthesis and respiration of two South Florida reef co ral species. Bulletin of Marine Science 57:527-539 Toler SK (2002) Shell damage in bleaching stressed Amphistegina gibbosa (Class Foraminifera) from the Florida Keys. P h.D. Dissertation, University of South Florida, Tampa, FL Toler SK, Hallock P (1998) Shell malformation in stressed Amphistegina populations: relation to biomineralization and pa leoenvironmental potential. Marine Micropaleontology 34:107-115 Tomascik T, Sanders F (1985) Effects of eutrophication on reef-building corals. I. Growth rate of the reef-building coral Montastrea annularis Barbados, West Indies. Marine Biology 94:53-75 Ullman WJ, Sandstrom MW (1987) Dissolved nutrient fluxes from the nearshore sediments of Bowling Green Bay, centra l Great-Barrier-Reef Lagoon (Australia). Estuarine Coastal and Shelf Science 24:289-303 U.S. Census Bureau (2000) www.census.gov

PAGE 283

U.S. Environmental Protection Agency (1993) Water quality Portec tion Program for teh Florida Keys National Marine Sanctuary. Phase II Report, Contract No. 68-c20134 Van Veghel MLJ, Bak RPM (1994) Reproductiv e characteristics of the polymorphic Caribbean reef building coral Montastrea annularis III. Reproduction in damaged and regenerating colonies. Ma rine Ecology Progress Series 109:229-233 Waddell JE (ed) (2005) The State of Coral R eef Ecosystems of the United States and Pacific Freely Associated States: 2005. NOAA Technica l Memorandum NOS NCCOS 11. NOAA/NCCOS Center for Co astal Monitoring and Assessment's Biogeography Team, Silver Springs Weber M, Lott C, Fabricius KE (2006) Sedi mentation stress in a scleractinian coral exposed to terrestrial and marine sediments with cont rasting physical, organic and geochemical properties. Journal of E xperimental Marine Biology and Ecology 336:18-32 Weil E, Knowlton N (1994) A multi-characte r analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M. Faveolata (Ellis and Solander, 1786) and M. Franksi (Gregory, 1895). Bulletin of Marine Science 55:151-175 Werner I, Hinton DE (1999) Field validation of hsp70 stress proteins as biomarkers in Asian clam ( Potamocorbula amurensis ): is downregulation an indicator of stress? Biomarkers 4:473-484 Wheaton J, Jaap WC, Dustan P, Porter J, Meier OW (1998) EPA Coral Reef and Hardbottom Monitoring Project, 1997 Annua l Report, Florida Marine Research Institute, St. Petersburg, FL Wilkinson C (ed) (1998) Status of Coral Reefs of the World: 1998. Global Coral Reef Monitoring Network. Australian Institute of Marine Science, Townsville Wilkinson C (ed) (2000) Status of Coral Reefs of the World: 2000. Global Coral Reef Monitoring Network. Australian Institute of Marine Science, Townsville Wilkinson C (ed) (2004) Status of the cora l reefs of the world: 2004. Global Coral Reef Monitoring Network. Australian Institute of Marine Science, Townsville Williams DE (2002) Population ecology of bleaching-stressed Amphistegina gibbosa in the Florida Keys (1991-1999): influence of solar radiation on reef-dwelling foraminifera. Dissertation, University of South Florida, St. Petersburg

PAGE 284

Williams DE, Hallock P (2004) Bleaching in Amphistegina gibbosa d'Orbigny (Class Foraminifera): observations from labor atory experiments using visible and ultraviolet light. Marine Biology 145:641-649 Williams DE, Hallock P, Talge HK, Harney JN, McRae G (1997) Responses of Amphistegina gibbosa populations in the Florida Keys (U.S.A.) to a multi-year stress event (1991-1996). Journal of Foraminiferal Research 27:264-269 Williams I (1994) An investigation into th e use of coral lesions as indicators of environmental pollution. MS thesis, Un iversity of Newcastle, Newcastle upon Tyne, UK Williams ID, Polunin NVC (2001) Large-scale associations between macroalgal cover and grazer biomass on mid-depth reefs in the Caribbean. Coral Reefs 19:358-366 Wittenberg GM, Hunte W (1992) Effects of eu trophication and sedimentation of juvenile corals. 1. Abundance, mortality, and co mmunity structure. Marine Biology 112:131-138 Woolfe KJ, Larcombe P (1999) Terrigenous sedimentation and coral reef growth: a conceptual framework. Marine Geology 155:331-345 Wu Q, Riise G, Lundekvam H, Mulder J, Haugen LE (2004) Influences of suspended particles on the runoff of pesticides fro m an agricultural field at Askim, SENorway. Environmental Geoche mistry and Health 26:295-302 Ye B, Maret W, Vallee BL (2001) Zinc metallo thionein imported into liver mitochondria modulates respiration. Proceedings, Na tional Academy of Sciences USA 98:23172322 Yeom DH, Adams SM (in press) Assessing eff ects of stress across le vels of biological organization using an aquatic ecosyst em health index. Ecotoxicology and Environmental Safety Yentsch CS, Yentsch CM, Cullen JJ, Lapoi nte B, Phinney DA, Yentsch SW (2002) Sunlight and water transparency: corner stones in coral research. Journal of Experimental Marine Bi ology and Ecology 268:171-183

PAGE 285

Appendices

PAGE 286

Appendix A. Parameters measured by Cellular Dia gnostic System and their biological significance Parameter Description Protein Metabolic Condition Heat-shock protein (Hsp60) and Heat-shock protein (Hsp70) Heat shock proteins 60 and 70 are mol ecular chaperones. Chaperones regulat e protein structure and function under normal physiological conditions as well as during and followi ng stress by renaturing denatured proteins into active states in an ATP-dependent manner. Both Hsp60 and Hsp70 are found in all phyla of life and are essential components for correct conformation of protein structure. Heat shock proteins 60 and Hs p70 levels increase in response to stress, specifically in re sponse to increased protein synthesis and denaturation. These two chaperones are indicators that the house-k eeping proteins in the cell are experiencing denaturing conditions. Ubiquitin Ubiquitin is a 76-residue protein fo und in most phyla of life that is conjugated to proteins slated for degradation b y the 26S proteosome. Proteins, during stress, are targeted for degradation usually b ecause these proteins have undergone an irreversible denaturation. Increases in ubiquitin levels are an indication of increased levels of protein degradation, and hence, increased prot ein turnover. Consequently, to compen sate for decreased functional protein levels due to stress, the cell will incr ease production of these same proteins Thus measurement of levels of ubiquitin is an index of the structural in tegrity of the protein component of the superstructure of the cell. Increased ubiquitin levels indicates: (1) a prot ein denaturing stress is occurring; (2 ) increased expenditure of energy is required to compensate for this stressinduced protein turnover; and (3) in comparison to baselin e data of this parameter for a particular species, may act as an indicator of individual fitness.

PAGE 287

Appendix A (Continued). Parameters measured by Cellula r Diagnostic System and thei r biological significance Oxidative Damage and Response Superoxide dismutases (SODs) Superoxide dismutases (SOD) play a large role in cellular antioxidant defenses by catalyz ing a reaction of superoxide ions and two protons to form hydrogen peroxide and O2, thereby reducing the harmful effects of oxidants. Copper/zinc SOD is an enzy me involved with antioxidant defenses localized in the cytosol of animal cells and in the cytosol and chloroplast in plants and al gae. Manganese SOD is localized in the mitochondria of eukaryotic cells and is therefore a sp ecific index that the mitochondria are experiencing an oxidative stress. Glutathione Peroxidase (GPx) Glutathione peroxidase is another impor tant antioxidant enzyme with the majo rity of activity in the cytoplasm but also is involved in mitochondrial func tion. This selenoprotein catalyzes the reaction that detoxifies hydroperoxides and organic peroxides to their corres ponding alcohol by oxidizing glut athione to glutathione disulphide and water. Catalase A heme-containing enzyme that catalyzes the br eakdown of hydrogen peroxide into water and oxygen. Metabolic Condition Heme Oxygenase (HO) Heme oxygenase, also known as Hsp32, is an enzyme that catalyzes decomposition of heme to biliverdin, ferrous iron, and carbon monoxide. Biliverdin is further catalyzed to bilirubin, which is a powe rful lipophic antioxidant. Heme production can increase in response to an increased demand for 1) membra ne associated antioxidants and 2) the breakdown of hemin as a result of CYP P450 suicide r eactions and the production of N-alkyl porphyrins.

PAGE 288

Appendix A (Continued). Parameters measured by Cellula r Diagnostic System and thei r biological significance Metabolic Condition (cont.) Ferrochelatase (FC) Ferrochetalase is an enzyme that catalyzes the final step in heme synthesis by inser ting inserts ferrous iron into protoporphyrin IX to form heme. Both cellular detoxifi cation pathways and essentia l cellular metabolism require heme or porphoryn-based substrates. For example, cytochrome c uses a form of heme in order to become an active electron carrier. Further, the cla ss of monooxygenases, cytochrome P450, requires heme to function. As an organism up-regulates metabolic or xenobiotic detoxification pathways, it increases heme production; thus, ferrochetalase is up-regulated too. Metallothionein Metallothioneins are cysteine-rich, low-molecular-weight pr oteins that will bind a vari ety of metals depending on the class of metallothionein. Metallot hionein often is used as a biomarker of heavy metal expos ure because it accumulates in response to exposure to different heavy metals, such as cadmium (Tang 1999, Downs et al. 2001a, b). However, metallothionein also can hyper-accumulate in response to bacterial infection, exposure to some types of mitochondrial inhibitors (e.g., pestic ides), oxidative stress, developmenta l changes, and growth factors. For example, metallothionein type 1 localizes to the m itochondrial inter-membrane sp ace and can help mitigate superoxide production by controlling as pects of oxidative phosphorylation (Sim pkins et al. 1994, Ye et al. 2001, Downs et al. 2006).

PAGE 289

Appendix A (Continued). Parameters measured by Cellular Di agnostic System and their bi ological significance. Metabolic Condition (cont.) Cnidarian small heat-shock protein (Cn sHsp) Total small heat-shock proteins includes -crystallin, Hsp22, Hsp23, Hsp26, and Hs p28. In most cases, the small heat-shock proteins are not present during optimal gr owing conditions and are on ly elicited by stress. -crystallin is a small heat-shock protein found only in the cytosol of animals, where it protects the cytoskeletal elements during stress. The presence and concentration of diff erent small heat-shock proteins helps determine the physiological status of several metabolic processes. Chloroplast small heatshock protein (ChlpsHsp) The ChlpsHsp is a small heat-shock prot ein found only in the choloroplast in re sponse to a stressed condition. The ChlpsHsp specifically associates with the oxygen evolving complex of photosystem II thereby protecting photosystem II activity during heat stre ss, ultraviolet radiation exposure, dehydration, and oxidative stress, most likely via a recycling anti-oxid ant mechanism. These proteins also will upregulate in response to some herbicides (e.g., atrazine).

PAGE 290

Appendix A (Continued). Parameters measured by Cellular Di agnostic System and their bi ological significance. Xenobiotic Response and Detoxification Xenobiotic detoxification involves a three-ph ase process including multixenobiotic resistan ce proteins (MXR), cytochrome P450s (CYP) and glutathione-s-transferase (GST) th at either prevents or reduces the advers e effects of xenobiotic exposure. Phase I of this process involves the enzymatic adduction of polar groups (e.g., hydroxyl) to the xe nobiotic via cytochrome P450s. In Phase II, these new polar metabolites are conjugated with endogenous substrates by enzymes that include glutathione-s-transferase. Phase III involves the export of these water-soluble pr oducts either to th e lysosomes for further metabolism, lysosome-like structures fo r containment or out of the cell through active diffusion transporters, such as ATP-binding cassette transporters (e.g., MXR; Bor st & Elfrink 2002). Multiple Xenobiotic Resistance Protein (MXR) Multixenobiotic resistance proteins (MXR), also known as Pglycoproteins, play a role in xenobiotic detoxification by actively transporting certain xenobiotic s out of the cell (Bard 2000). The level of P-glycoproteins increases with a sustained exposure to certain xenobiotics (Downs et al. 2005a 2006). If this process becomes overwhelmed or if the xenobiotic is not recognized by MX R, it can be metabolized into a hydr ophilic compound that can be easily removed from the cell in Phase I or Phase II.

PAGE 291

Appendix A (Continued). Parameters measured by Cellular Di agnostic System and their bi ological significance. Xenobiotic Response and Detoxification (cont.) Cytochrome (CYP) P450 Cytochrome P450 oxidizes ethanol to acetaldehyde via a m onooxygenase mechanism, as well as other xenobiotics. Cytochrome P450 2E has both physiologically relevant oxidative and reductive reactions and associates and catalyzes as many as 60 xenobiotic-based substrates. One of the primary reasons for using the 2E class of cytochrome P450s is that it is not induced by heat stress but can respond to hypoxia/reper fusion events in mammals. CYP-2 and CYP-3 are involved with dr ug and steroid metabolism and detoxifi cation of electro philic carcinogens, drugs and environmental pollutants. CYP-6 has been impli cated in the evolution of pe sticide resistance, including DDT. GlutathioneS-transferase (GST) Glutathione transferases are usually associated with detoxificaiton by conjunction of genotoxic and cytotoxic xenobiotic electrophiles derived from drugs, carcinogens, and environmenta l pollutants. During a xenobiotic challenge, glutathione may be conjugat ed to a xenobiotic by glutathione S-transferase and represent a major detoxification pathway. Add itionally, glutathione-S-transferase (GST) may detoxify DNA hydroperoxides, and thus may play an important role in DNA repair.

PAGE 292

Appendix B. Eigenvalues and eigenvectors fo r principle components analys is on cellular biomarkers al ong the depth gradient for selected sampling periods. March 2001 June 2001 August 2001 October 2001 June 2002 August 2002 November 2002 February 2003 PC PC1 PC1 PC1 PC 1 PC1 PC1 PC1 PC1 Eigenvalue 14.6 15.9 15.3 13.8 12.5 13.6 9.9 15.5 % Variation 77.1 79.5 76.3 72.4 62.7 67.8 49.2 77.3 Variables Eigenvectors Cnidarian Hsp70 -0.242 -0.226 -0.229 -0.224 -0.213 -0.248 -0.287 -0.208 Dino Hsp70 -0.149 -0.233 -0.196 -0.217 n/s n/s n/s n/s Cnidarian Hsp 60 -0.227 -0.214 -0 .233 -0.266 -0.26 -0.161 -0.233 Dino Hsp60 -0.2 -0.232 -0.222 -0 .203 -0.201 -0.253 0.162 -0.244 Ubiquitin -0.247 -0.242 -0.235 -0. 245 -0.199 -0.232 -0.274 -0.186 Cnidarian Cu/ZnSOD -0.243 -0.236 -0.241 -0.252 n/s n/s n/s n/s Dino Cu/ZnSOD -0.243 -0.238 -0.224 -0.231 -0.25 -0.203 -0.193 -0.197 Cnidarian MnSOD -0.221 -0.202 -0. 223 -0.256 -0.243 0.149 -0.242 Dino MnSOD -0.243 -0.215 -0.24 -0 .254 -0.206 0.02 -0.247 -0.245 Cnidarian GPx -0.244 -0.234 -0.24 -0 .252 -0.267 -0.188 -0.285 -0.227 Dino GPx -0.232 -0.195 -0.234 -0 .242 -0.229 -0.103 -0.292 -0.24 Catalase n/s n/s n/s n/s -0.21 -0.252 0.184 -0.227 Heme oxygenase -0.251 -0.218 -0.239 -0.242 n/s n/s n/s n/s Ferrochelatase -0.229 -0.233 -0.222 -0.216 -0.259 -0.254 -0.245 -0.168 Metallothionein -0.243 -0.223 -0.249 -0.168 -0.237 -0.244 -0.17 -0.249 Cnidarian sHsp -0.201 -0.195 -0.206 -0.241 0.032 -0.119 -0.281 -0.238 Chloroplast sHsp -0.245 -0.24 -0.219 -0.151 -0.208 -0.224 0.153 -0.212 MXR -0.23 -0.189 -0.104 -0.238 -0.251 -0.256 -0.26 -0.229 Cnidarian GST -0.234 -0.238 -0.233 -0 .247 -0.147 -0.257 -0.265 -0.246 Dino GST -0.239 -0.224 -0.247 -0 .246 -0.246 -0.209 -0.27 -0.227 CYP 450-2 -0.195 -0.221 -0.215 -0 .227 -0.203 -0.247 0.159 -0.209 CYP-3 n/s n/s n/s n/s -0.204 -0.253 0.166 -0.227 CYP-6 n/s n/s n/s n/s -0.264 -0.235 -0.124 -0.199

PAGE 293

Appendix C. Eigenvalues and eigenvectors fo r principle components analys is on cellular biomarkers at the 6 m sites for selected sampling periods. Sampling Period October 2001 June 2002 August 2002 November 2002 February 2003 PC PC1 PC1 PC1 PC1 PC1 Eigenvalue 15.1 9.3 13 9.6 10.9 % Variation 79.2 46.4 65.2 47.9 54.6 Variable Eigenvectors Cnidarian Hsp70 -0.23 -0.191 -0.162 -0.172 -0.251 Dino Hsp70 -0.248 n/s n/s n/s n/s Cnidarian Hsp60 -0.242 -0.272 -0.249 -0.216 -0.149 Dn Hsp60 -0.234 0.284 -0.262 -0.134 -0.238 Ubiquitin -0.235 -0.233 -0.198 -0.306 -0.19 Cnidarian Cu/ZnSOD -0.245 n/s n/s n/s n/s Dn Cu/ZnSOD -0.209 -0.246 -0.251 -0.228 -0.236 Cnidarian MnSOD n/s -0.251 -0.245 -0.164 -0.275 Dn MnSOD -0.244 -0.241 -0.258 -0.302 -0.283 Cnidarian GPx -0.231 -0.215 -0.265 -0.308 -0.208 Dino GPx -0.217 0.077 -0.177 -0.235 -0.283 Catalase n/s 0.178 -0.247 -0.069 -0.117 Heme oxygenase -0.243 n/s n/s n/s n/s Ferrochelatase -0.235 -0.003 -0.248 -0.232 -0.254 Metallothionein -0.173 0.031 -0.241 -0.31 -0.202 Cnidarian sHsp -0.249 -0.188 -0.109 -0.093 -0.065 Chloroplast sHsp -0.227 0.283 -0.183 -0.153 -0.169 MXR -0.21 -0.19 -0.213 -0.276 -0.258 Cnidarian GST -0.238 -0.257 -0.243 -0.299 -0.216 Dino GST -0.236 -0.158 -0.247 -0.287 -0.269 CYP450-2 -0.198 0.292 -0.163 -0.121 -0.269 CYP450-3 n/s 0.284 -0.196 -0.116 -0.277 CYP450-6 n/s -0.281 -0.236 -0.172 -0.039

PAGE 294

About the Author Elizabeth M. Fisher received her Bach elor’s degree in Marine Science and Biology with a minor in Chemistry from the Un iversity of Miami in 1999. While at the University of Miami, she worked with Dr Robert Ginsburg and Dr. Philip Kramer conducting coral reef assessments in South Fl orida, Mexico, Jamaica and Bahamas using the Atlantic Gulf and Rapi d Reef Assessment protocol. She started her graduate work at the Univ ersity of South Florida with Dr. Pamela Hallock Muller in the College of Marine Sc ience in 2000 focusing on indicators of reef condition. While at USF she received th e Sanibel-Captiva Shell Club Endowed Fellowship and a teaching assistantship in Envi ronmental Science. She has been actively involved in education outreach and recei ved the National Science Foundation GK-12 OCEANS Fellowship and was the National Ocean Sciences Bowl Regional Coordinator. She also is an active diver a nd served on the USF Diving Safety and Control Board.