Introduction

Benthic Microalgae and Nutrient Flux in Florida Bay, USA

Merrie Beth Neely

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Marine Science University of South Florida

Co-Major Professor: Gabriel A. Vargo, Ph.D. Co-Major Professor: Kent A. Fanning, Ph.D. Paul R. Carlson, Ph.D. Peter Howd, Ph.D. Karen Steidinger, Ph.D. Laura Yarbro, Ph.D.

Date of Approval: November 20, 2008

Keywords: phosphorus, nitrogen, chlorophyll, mesocosm, microphytobenthos

Acknowledgements

Portions of this project were funded by: USDOC/NOAA Award # NA06OP0519 to Dr. Gabriel A. Vargo and Dr. Gary L. Hitchcock (PI‟s) and a Florida Institute of Oceanography Student Grant in Aid to Merrie Beth Neely and Jennifer Jurado. Special thanks go to the Keys Marine Laboratory Staff; Jennifer Jurado, Gary Hitchcock, and Chris Kelble from the University of Miami; Dr. Rob Masserini, USF College of Marine Science; and Bill Sargeant and Dr. Cynthia Heil, Florida Fish and Wildlife Conservation Commission-Fish and Wildlife Research Institute; the Florida Institute of Oceanography SEAKEYS buoy system; Dr. Carmelo Tomas and Dr. Larry Cahoon, University of North Carolina,Wilmington.

Table of Contents

List of Tables ...... vi

List of Figures ...... viii

Chapter 1. Benthic microalgae and nutrient flux in Florida Bay, USA ...... 1 Introduction ...... 1 Florida Bay Ecosystem ...... 2 Benthic Microalgae ...... 4 Benthic Microalgae and Sediment Resuspension ...... 5 Dissertation Objectives ...... 6

Chapter 2. Sediment phosphate flux in Florida Bay, USA...... 7 Introduction ...... 7 Materials and Methods ...... 8 Results and Discussion ...... 12 Nitrogen ...... 44 Summary ...... 45

Chapter 3. Benthic Microalgal Chlorophyll a Standing Stock in Florida Bay ...... 48 Introduction ...... 48 Benthic Microalgae and Sediment Resuspension ...... 51 Benthic Microalgal primary productivity and models ...... 52 Benthic Microalgae in Florida Bay ...... 54 Goals and objectives ...... 57 Materials and methods ...... 57

ii Results ...... 59 Statistical analysis ...... 63 Discussion ...... 64 Summary ...... 66

Chapter 4. Mesocosm bioassays as a tool to investigate microalgal benthic- pelagic coupling in Florida Bay ...... 67 Introduction ...... 67 Florida Bay Nutrient Limitation ...... 67 Bioassays...... 70 Materials and methods ...... 71 Results ...... 75 Nutrients: Initial Field Conditions ...... 75 Mesocosm results ...... 76 Chlorophyll concentration ...... 98 Discussion ...... 107 Nutrient concentrations ...... 107 Chlorophyll concentration ...... 108 Summary ...... 112

Chapter 5. Short term response of benthic microalgae to 33P additions to the sediment...... 114 Introduction ...... 114 Materials and Methods ...... 116 Results ...... 121 Preliminary experiments ...... 121 Laboratory experiments ...... 122 Sediment-water column P pools ...... 126 Discussion ...... 129 Summary ...... 130

iii

Chapter 6. The use of dissolved oxygen probes and Pulse Amplitude Modulated (PAM) fluorometry in the evaluation of benthic microalgal primary production in Florida Bay ...... 131 Introduction ...... 131 Oxygen probes ...... 132 PAM fluorometry ...... 132 Materials and Methods ...... 134 Dissolved Oxygen ...... 134 PAM Fluorometry ...... 136 Results ...... 137 DO Electrodes ...... 137 PAM Fluorometry ...... 149 Discussion ...... 158 Summary ...... 161

Chapter 7. Conclusion ...... 163 Sediment Phosphate Flux ...... 163 Benthic microalgal chlorophyll a standing stock ...... 164 Mesocosms as a tool to measure limiting nutrients in benthic microalgae ...... 165 Radiolabelled tracers of P flux through benthic microalgae ...... 166 Measuring primary production in benthic microalgae ...... 166 Everglades restoration efforts ...... 167 Future work ...... 168

Literature Cited ...... 169

Bibliography ...... 179

Appendices ...... 183

iv Appendix A. 184

Appendix B 185

Appendix C. Primary Production Equations 186

About the Author End Page

List of Tables

Table 2.1. Student's T-test results of TDP, SRP and DOP flux from the sediment by treatment type between locations and by time of day between locations. Arsnicker represents Arsnicker Key in the southeast bay, Carl Ross represents Carl Ross Key in the western bay and End represents End Key in the central bay...... 13

Table 2.2. Student‟s T-test results showing statistical significance between replicate light and dark treatments by date and parameter...... 42

Table 2.3. Average DOP flux in Florida Bay...... 43

Table 2.4. Average TDN values and range of flux in Florida Bay. The last column provides the range of the percent increase from initial values during incubation...... 45

Table 3.1 (reproduced from Cahoon, 2006, with permission). Spatial distribution of intertidal and subtidal studies measuring microphytobenthic production by all methods as of 2003. Data are numbers of published studies from Cahoon (1999) and more recent references...... 49

vi Table 3.2. (reproduced from Cahoon, 2005, with permission). Values of microphytobenthic photosynthetic parameters reported in

published literature: Ek (saturating light intensity, mol -2 -1 photons m s ), Pmax (maximum, biomass-normalized photosynthetic rate, mg C mg chl a-1 h-1), alpha (slope of P-E relationship, mg C mg chl a-1 h-1 (mol photons m-2 s-1) -1), P (biomass-normalized photosynthetic rate mg C mg chl a-1 h-1...... 50

Table 3.3 Ranges and average pigment values for water column and benthic microalgae in Florida Bay...... 61

Table 3.4 ANOVA levels of significance for comparison of chlorophyll a and phaeopigment among locations ...... 64

Table 4.1. Initial nutrient field concentration in M L-1 ...... 75

Table 6.1 Net benthic primary productivity (NBPP), sediment respiration (SR) and gross benthic primary production (GBP) – see Appendix C for calculations...... 138

vii

List of Figures

Figure 1.1. Map of bottom type in Florida Bay (Prager and Halley, 1997). Dark brown tan and orange areas represent bottom unvegetated by seagrass, but available for benthic microalgal production...... 2

Figure 2.1. Typical light and dark incubation chambers (<7L) used for the measurement of nutrient flux to and from the sediment. The photo shows the sampling ports and the ambient water reservoir (IV bags)...... 9

Figure 2.2. A map of Florida Bay indicating the major physical and biological regions and the red dots indicate the locations sampled during this study...... 11

Figure 2.3. Summary of morning TDP flux by chamber and by date for Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 14

Figure 2.4. Summary of afternoon TDP flux by chamber and by date for Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 15

viii Figure 2.5. Summary of Total TDP flux for Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 16

Figure 2.6. Summary of morning SRP flux at Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 17

Figure 2.7. Summary of afternoon SRP flux rates at Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 18

Figure 2.8. Summary of Total SRP flux rates at Carl Ross key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 19

ix Figure 2.9 Summary of morning DOP flux rates (calculated) at Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 20

Figure 2.10. Summary of afternoon SRP flux rates (calculated) at Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 21

Figure 2.11. Summary of Total DOP flux rates (calculated) at Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 22

Figure 2.12. Summary of morning TDP flux rates at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 23

x Figure 2.13. Summary of afternoon TDP flux rates at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 24

Figure 2.14. Summary of Total TDP flux at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 25

Figure 2.15. Summary of morning SRP flux rates at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 26

Figure 2.16. Summary of afternoon SRP flux at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 27

xi Figure 2.17. Summary of total SRP flux rates at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 28

Figure 2.18. Summary of morning DOP flux rates (Calculated) at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 29

Figure 2.19, Summary of afternoon DOP flux rates (calculated) at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 30

Figure 2.20. Summary of total DOP flux rates (calculated) at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 31

xii Figure 2.21. Summary of the morning TDP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 32

Figure 2.22. Summary of the afternoon TDP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 33

Figure 2.23. Summary of the total TDP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 34

Figure 2.24. Summary of morning SRP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment...... 35

xiii Figure 2.25. Summary of afternoon SRP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 36

Figure 2.26. Summary of total SRP flux at End basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 37

Figure 2.27. Summary of morning DOP flux (Calculated) at End basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. A positive flux is into the sediment, while negative flux is out of the sediment...... 38

Figure 2.28. Summary of afternoon DOP flux rates (calculated) at End basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment...... 39

xiv Figure 2.29. Summary of total DOP flux rates (calculated) at End basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment...... 40

Figure 3.1. October 2005, Map of colored water in Florida Bay. Reproduced with permission, FWC-FWRI...... 55

Figure 3.2. November 1995, Map of colored water in Florida Bay. Reproduced with permission FWC-FWRI...... 55

Figure 3.3. December 1995, map of colored water in Florida Bay. Reproduced with permission, FWC-FWRI...... 56

Figure 3.4. Mean chlorophyll a and phaeopigment in western Florida Bay from replicate sediment cores (first Y axis) and a single water column sample when applicable (second Y axis). Error bars represent the standard error of replicate samples. Note that the 1st and 2nd Y axis scales differ by an order of magnitude. This station represents the intertidal microphytobenthos. The water column data was provided by Jennifer Jurado, University of Miami...... 60

Figure 3.5. Mean chlorophyll a and phaeopigment for replicate sediment cores in central Florida Bay. Error bars represent the standard error of replicate samples. This site was representative of subtidal microphytobenthos...... 62

Figure 3.6. Mean chlorophyll a and phaeopigment for replicate sediment cores in southeastern Florida Bay. Error bars represent the standard error of replicate samples. This site was representative of subtidal microphytobenthos...... 63 xv Figure 4.1 (reproduced from Tomas et al. 1999, with permission) Depiction of nutrient limitation zones within Florida Bay based upon bioassays of natural pelagic phytoplankton populations, from Tomas et al. 1999. Zone A represents areas where nitrogen and possibly silica are limiting during most of the year. Zone B represents areas where primarily N and sometimes P were limiting. Zone C represents areas primarily P limiting with rare N limitation...... 69

Figure 4.2. The site locations where sediment and water were collected for the mesocosm experiments, Carl Ross Key and End Key. The sites are considered representative of western and central Florida Bay, respectively...... 72

Figure 4.3. The mesocosms arranged in the randomized block design within the lagoonal flow-through seawater system at the field laboratory in Layton, FL...... 73

Figure 4.4. Average TDP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) during winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 77

Figure 4.5. Average TDP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) during summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 78

xvi Figure 4.6. Average TDP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 79

Figure 4.7. Average SRP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 80

Figure 4.8. Average SRP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 81

Figure 4.9. Average SRP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 82

Figure 4.10. Average TDN results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 84

xvii Figure 4.11. Average TDN results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 85

Figure 4.12. Average TDN results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 86

Figure 4.13. Average silica results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 87

Figure 4.14. Average silica results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 88

Figure 4.15. Average silica results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 89

xviii Figure 4.16. Average nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 90

Figure 4.17. Average nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 91

Figure 4.18. Average nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 92

Figure 4.19. Average nitrate plus nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 93

Figure 4.20. Average nitrate plus nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 94

xix Figure 4.21. Average nitrate plus nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 95

Figure 4.22. Average ammonium results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 96

Figure 4.23. Average ammonium results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 97

Figure 4.24. Average ammonium results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The error bars represent the standard error, and are only shown in one direction to facilitate viewing...... 98

Figure 4.25. Average water column chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during summer 2002. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily...... 100

xx Figure 4.26. Average sediment chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during summer 2002. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily...... 101

Figure 4.27. Average water column chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily...... 102

Figure 4.28. Average sediment chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily...... 103

Figure 4.29. Average sediment chlorophyll a (g L-1) in western Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily...... 104

Figure 4.30. Average water column chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily...... 106

xxi Figure 4.31. Temperature (left y-axis) and salinity (right y-axis) in August 2002 for the central Florida Bay mesocosm experiment...... 109

Figure 4.32. Temperature (left y-axis) and salinity (right y-axis) in January 2003 for the western Florida Bay mesocosm experiment...... 110

Figure 4.33. Temperature (left y-axis) and salinity (right y-axis) in January 2003 for the central Florida Bay mesocosm experiment...... 111

Figure 5.1. The locations where field cores and water were collected in Florida Bay for all the experiments...... 117

Figure 5.2. Cartoon depiction of experimental setup. From left to right: initial cores are received with BMA layer intact, the overlying water column is removed and sodium azide is added to the killed controls, filtered seawater collected from the field location is added to restore the water column and samples sit undisturbed overnight, the experiment commences upon the addition of carrier-free 33P or

PO4 (blank) to the sediment in killed and live treatments, over time the amount of 33P in the water column is measured...... 119

Figure 5.3. Various depth of injection below the sediment-water interface were evaluated to determine the minimum depth necessary to prevent water column contamination by the radiolabelled 33P using killed BMA communities from End Key and Carl Ross Key in western and central Florida Bay, respectively...... 122

Figure 5.4. Average benthic flux of 33P (femtomoles L-1 h-1) from the sediment in cores taken in August 2003 from western Florida Bay. Error bars represent standard error of 5 replicates. Carrierfree 33 PO4 in low nutrient seawater was injected into the sediment at 2cm below the sediment-water interface. An enhancement dose of

xxii 37 M of P was added to the water column in the live and killed treatments after 8 hours, but not controls...... 123

Figure 5.5. The average 33P flux (femtomoles L-1 h-1) from the sediment in cores taken from western Florida Bay, May 2004. Error bars represent standard error of 5 replicates per sampling interval...... 125

Figure 5.6. The average 33P flux (femtomoles L-1 h-1) from the sediment in cores taken from central Florida Bay, May 2004. Error bars represent standard error of 5 replicates per sampling interval. Note the Y axis is an order of magnitude less than the western Florida Bay figure...... 126

Figure 5.7. Daily flux of TDP to the overlying water column from the western Florida Bay sediment during the August 2003 experiment. All results indicate flux from the sediment to the water column...... 127

Figure 5.8. Daily TDP flux to the overlying water column from the sediment pool in western Florida Bay during the course of the May 2004 experiment. The measurement from the light 5 sample was lost. All results indicate flux from the sediment to the water column...... 128

Figure 5.9. Daily TDP flux to the overlying water column from the central Florida Bay sediment during the course of the May 2004 experiment. The measurement from the dark 2 sample was lost. All results indicate flux from the sediment to the water column...... 129

Figure 6.1. Station map indicating the locations were DO in benthic chambers (Carl Ross Key and End Key only) and PAM fluorometry of benthic microalgae were measured...... 135

xxiii Figure 6.2. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on July 17, 2002 in central Florida Bay. PAR was not measured at the final sampling period...... 139

Figure 6.3. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on August 2, 2002 in central Florida Bay. PAR was not measured at the final sampling period...... 140

Figure 6.4. DO concentration (left y-axis) within a light benthic chamber and ambient PAR values (right y-axis) over time on September 11, 2000 in western Florida Bay. PAR was not measured at the final sampling period...... 141

Figure 6.5. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on November 13, 2000 in western Florida Bay...... 142

Figure 6.6. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on November 14, 2000 in western Florida Bay...... 143

Figure 6.7. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on March13, 2001 in western Florida Bay...... 144

Figure 6.8. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on March 15, 2001 in western Florida Bay...... 145

xxiv Figure 6.9. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on March 16, 2001 in western Florida Bay...... 146

Figure 6.10. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on May 16, 2001 in western Florida Bay...... 147

Figure 6.11. DO concentration (left y-axis) within a light benthic chamber and ambient PAR values (right y-axis) over time on May 7, 2002 in western Florida Bay...... 148

Figure 6.12. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on August 8, 2002 in western Florida Bay...... 149

Figure 6.13. Time of day effect on Yield versus Electron Transport Rate (ETR) derived from instantaneous measurements at the three sites in western, central and southeastern Florida Bay from field sampling events between 2000 and 2002...... 150

Figure 6.14. Seasonal effect on Yield versus Electron Transport Rate (ETR) derived from instantaneous measurements at the three sites in western, central and southeastern Florida Bay from field sampling events between 2000 and 2002...... 151

Figure 6.15. Regional differences in Yield versus Electron Transport Rate (ETR) derived from instantaneous measurements at the three sites in western, central and southeastern Florida Bay from field sampling events between 2000 and 2002...... 152

xxv Figure 6.16. Electron transport rate (ETR) from triplicate rapid light curves in southeastern Florida Bay on May 2, 2002 at the 10:00 am and 12:00pm sampling interval...... 153

Figure 6.17. Electron transport rate (ETR) from duplicate rapid light curves in central Florida Bay on April 4, 2002 at the 10:00 am, 12:00pm and 1:30pm sampling interval...... 154

Figure 6.18. Electron transport rate (ETR) from duplicate rapid light curves in central Florida Bay on July 17, 2002 at the 9:00 and 11:00 am and 1:00 and 3:00 pm sampling interval...... 155

Figure 6.19. Electron transport rate (ETR) from triplicate rapid light curves in western Florida Bay on March 15, 2001 at the 9:00 and 11:30 am and 1:15pm sampling interval...... 156

Figure 6.20. Electron transport rate (ETR) from duplicate rapid light curves in western Florida Bay on March 16, 2001 at the 10:00 and 11:45 am sampling interval...... 157

Figure 6.21. Electron transport rate (ETR) from triplicate rapid light curves in western Florida Bay on May 4, 2002 at the 3:00 pm sampling interval...... 158

xxvi

Benthic Microalgae and Nutrient Flux in Florida Bay, USA

Merrie Beth Neely

ABSTRACT

The objective of this study was to address the relationship between benthic microalgal communities and the phosphate nutrient dynamics of Florida Bay sediments and how they relate to benthic and water column primary production. In situ phosphate (P) flux between the sediment and the water column was measured in three regions of Florida Bay. Differences in the ratio of inorganic to organic phosphate flux were found between the three regions in relation to the amount of phosphate measured in the water column. Based upon the average sediment flux in my study, more than 1600 metric tons of P would be supplied by the sediment per year in Florida Bay. Based upon my measurements, dissolved nutrient flux from the sediment can be an important contribution to pelagic phytoplankton blooms in Florida Bay, accounting for 6.5 – 41% of demand and TDN accounts for 100% of the N demand. My findings were similar to others for both benthic nutrient flux and benthic microalgal chlorophyll a concentration. Benthic microalgae in Florida Bay contribute 700 kg Chl a per day to the system. Mesocosm experiments demonstrated that benthic microalgae and water column phytoplankton can respond differently to changes in nutrient availability. The dissolved nutrient in least supply in the water column does not necessarily correspond to the limiting nutrient for benthic microalgae. 33P acted as a tracer between sediment and water column dissolved P pools. The presence of benthic microalgae enhanced the transport of 33P to the water column as compared to simple Fickian diffusion. This was supported by the positive flux of dissolved P from the sediment to the water column pools in control treatments with a living benthic microalgal layer. Primary production by benthic microalgae were measured using dissolved O2 evolution and PAM fluorometry. Primary production for BMA habitat in Florida Bay was between 400 and 800 tons of C per day, based upon O2 production and PAM fluorometry, respectively.

xxvii

Chapter 1. Benthic microalgae and nutrient flux in Florida Bay, USA Introduction The benthic microalgae that live in the upper centimeters of the sediment where the bottom lies within the euphotic zone have been termed the “secret garden” (MacIntyre et al. 1996). These benthic microalgae can form dense mats at the sediment surface (Grant and Gust, 1987) and are comprised mostly of cyanobacteria and diatoms (MacIntyre and Cullen, 1995; Pinckney et al. 1995). The pennate diatoms are both attached to, and free living in and on, the sediments (MacIntyre and Cullen, 1995). Primary production by the microphytobenthos is poorly understood, and ignored in most monitoring programs. Those measurements of benthic microalgal biomass and primary production that have been published are generally from European estuaries or subtropical Pacific waters (Cahoon, 2006; Colijn and De Jonge, 1984). Numerous studies have shown that benthic microalgal biomass and primary production can equal or exceed that of the water column (Cahoon and Cooke, 1992, Nelson et al. 1999, Charpy and Charpy-Roubaud, 1990), but spatially, benthic microalgal biomass can be very patchy (Rizzo and Wetzel, 1985). Even within seagrass beds the microphytobenthos is an important contribution to overall benthic primary production (Moncreiff, 1992). Benthic primary production is effectively limited to the upper millimeter or so of the sediment; basically that depth to where light penetrates permitting photosynthesis by the cells (MacIntyre and Cullen, 1995). However, viable cells have been found to several centimeters in depth, provided the sediments didn‟t become anoxic (MacIntyre and Cullen, 1995; Nelson et al. 1999). Vertical migration by motile benthic diatoms in the intertidal zone may be used as a means to escape desiccation (Pinckney and Zingmark, 1991) and UV exposure at low tide (Cooke, 1991). Other reasons cells are found at depth in sediments is due to burial (physical mixing of the sediments) from tides, currents and waves (MacIntyre and Cullen, 1990, MacIntyre et al. 1996), and bioturbation.

1 Photoinhibition has generally not been shown in benthic microalgae (Colijn and van Buurt, 1975; MacIntyre and Cullen, 1995; Mills and Wilkinson, 1986; Charpy and Charpy-Roubaud, 1990), probably as a consequence of the potential light limited conditions due both to turbidity at the sediment/water interface and lack of light penetration beyond the upper few millimeters of sediment. In contrast, MacIntyre and Cullen (1995) found photoinhibition to be prevalent in both pelagic phytoplankton and benthic microalgae resuspended into the water column in San Antonio Bay, although measurements of nonresuspended benthic microalgae were not part of the study. Florida Bay Ecosystem Florida Bay (Figure 1.1) is a triangular shaped, shallow (average depth <2m) embayment dominated by mud banks forming large basins within the interior of the bay (Schomer and Drew, 1982; Fourqurean et al. 1992; Fourqurean et al. 1993). It is bounded to the north by the Everglades, to the east and south by the Florida Keys, and to the west by the Gulf of Mexico. Parts of Florida Bay are seasonally hypersaline, due to the restricted tidal influence caused by the mud banks and the high degree of evaporation relative to direct precipitation (Schomer and Drew, 1982; Phlips and Badylak, 1996).

Figure 1.1. Map of bottom type in Florida Bay (Prager and Halley, 1997). Dark brown tan and orange areas represent bottom unvegetated by seagrass, but available for benthic microalgal production.

2 Considerable attention has been focused on the area since massive die-off of seagrasses were reported during the late 1980‟s (Carlson et al. 1990). As a result, the macrophyte dominated benthic primary production has been transitioning to a phytoplankton dominated system in many basins (Phlips and Badylak, 1996). Die-off events continue today in some parts of the bay, although it is not as widespread or catastrophic as a few years ago. Subsequent to the seagrass die-off, mixed diatom-cyanobacterial blooms were noted in the central portions of the bay, but these too have varied in intensity and area (Steidinger, et al. 1998, 2001). Recently, predominantly cyanobacterial blooms have been co-occurring in the central bay and also the eastern bay and Card Sound area, where they had never previously been reported (personal observation, Heil unpublished data). Diminished light transmittance to benthic macrophytes became a problem due to sediment resuspension in the die-off areas and from the water column phytoplankton blooms, which may have inhibited recolonization of the seagrasses (Kelble et al. 2005).The cyanobacterial blooms have also been implicated in major benthic die-off in the Rankin Bight region, including sponges and megafauna (Butler et al. 1995). There are three primary sources of water input to Florida Bay: The Gulf of Mexico, the Atlantic Ocean across the reef tract, and the Everglades (Schomer and Drew, 1982). The latter has undergone extensive water redistribution projects during the last century in an effort to reduce flooding and provide drinking water to urbanized areas of south Florida, and for irrigation of the expanding agricultural interests of the area (Schomer and Drew, 1982; Smith et al. 1989). As a result of this diversion of freshwater input from the north, the naturally hypersaline conditions in Florida Bay have been exacerbated in the mid-late 20th century (Phlips and Badylak, 1996; Fourqurean et al. 1992; Rudnick et al. 1997). Florida Bay is a nutrient limited system (Fourqurean et al. 1992). Nutrient limitation of phytoplankton and seagrasses in Florida Bay depends upon geographical location within the bay (Fourqurean et al. 1992, Fourqurean et al. 1993, Phlips et al. 1995; Phlips and Badylak, 1996). Tomas et al. (1995,1999), using bioassays, described P limitation increasing across Florida Bay from west to east and indicated that N and Si can

3 be co-limiting in western Florida Bay during the annual winter diatom bloom. central Florida Bay varies between nitrogen and phosphate limitation (Fourqurean et al. 1992; Fourqurean et al. 1993; Phlips and Badylak, 1996, Tomas et al. 1995 &1999). Many researchers have documented the patterns in water column nutrient availability in the bay, beginning with Fourqurean et al. (1993). This general trend has been supported by several other researchers in large and small scale field and laboratory studies (Vargo et al. 2001 a and b, Jurado and Hitchcock 2001, Brand et al. 2001). Coarser coral-algal sediments are found in the eastern region of the bay (Fourqurean et al. 1993). Fine grained mud or sands are prevalent elsewhere, especially in the more restricted basins. These fine-grained sediments and the attached or free- living microalgae are susceptible to resuspension into the water column. The bays carbonate sediments chemically bind most available phosphorus (P), leading to nutrient limitation of benthic and water column productivity. Unlike the weak bonds to clays and other organic particles, the bonds between the carbonate and P molecules are very tight and rarely break except during redox-reactions (Millero, 2000). Silica and nitrogen via surface and groundwater flow into the bay, combine with the tight chemical adsorption of P on to the carbonate sediments and lead to temporal and spatial nutrient limitation of the water column phytoplankton and benthic macroalgae and seagrasses (Tomas et al. 1996, 1999). Benthic Microalgae It is unknown just how important benthic-pelagic coupling is to the nutrient cycling of the Florida Bay ecosystem. Relatively few rates of benthic primary productivity or nutrient flux are available. The role of benthic microalgae in contributing to bay-wide primary productivity and nutrient cycling has also not been quantified. Further complicating matters is the unknown, but unarguably widespread distribution of patches of benthic microalgae. Where seagrass die off has occurred the only source of benthic primary production would now be from macroalgae (drift and attached mats) or benthic microalgae. Benthic algal communities within the bay vary from a thin film of diatoms over relatively unconsolidated sediments to inch thick Laurencia sp. mats carpeting the bottom (personal observation). Such communities might

4 vary widely in primary productivity, and would certainly vary in sediment stabilization capability as well as biomass. Benthic microalgae cannot be considered in the same context as water column phytoplankton or seagrasses because they exist as an inter-related community of autotrophic and heterotrophic bacteria. These communities involve several different types of microalgae and their extended polysaccharide matrix (MacIntyre et al. 1996). Some microalgae are associated with sediment substrate (epipelic) and others are free- living (epipsammic) (MacIntyre et al. 1996). Community composition can vary by depth in the sediment and the associated photic and redox environment (Underwood and Barnett, 2006). Free-living benthic microalgae migrate within and over the substrate, even colonizing seagrass (MacIntyre and Cullen, 1995). One poorly understood aspect of benthic microalgal communities is their effect on the carbon and nutrient flux between the water column and the sediments. Microalgal layers have been shown to modify the flux of nutrients from the sediment into the water column in laboratory experiments sand agar petri dish experiments (Sundback and Graneli, 1988) and in the field measuring along a tidal bore (Keizer et al. 1989). Bioturbation from megafauna and different rates of porewater advection due to fine scale bottom topography and grain size sorting can influence nutrient flux rates (Huettal and Gust, 1992, Sundback and McGlathery, 2005). A glaring hole in the ongoing research in Florida Bay is the role of benthic microalgal communities in the coupling of benthic - pelagic primary production and mediation of nutrient flux from the sediments. Benthic Microalgae and Sediment Resuspension Some microalgae can secrete a mucilaginous sheath or polysaccharide matrix binding themselves both to other organisms and the sediment grains (Yallop et al. 1994), thus providing an effective mechanism to stabilize the sediment on which they reside and minimize turbidity at the sediment/water interface. In addition to their contribution to primary production in coastal ecosystems, benthic microalgal mats have been shown to dramatically increase the turbulence and shear velocity needed to erode sediments (Yallop et al. 1994). The binding capabilities of their extruded exopolymer polysaccharides are extraordinary and can even increase with desiccation (Yallop et al.

5 1994). When erosion does occur, benthic microalgal mats can quickly re-establish themselves, often in as little as a few hours. The sediment types associated with microalgal mats can be quite diverse and the stabilizing properties of these mats is dependent to some degree on the type of sediments on which they are found and the species of microalgae growing there (Yallop et al. 1994; Amspoker and McIntire, 1978). Dissertation Objectives The specific objectives of my dissertation were to measure the benthic microalgal chlorophyll a standing stock and benthic nutrient flux, primarily phosphate, across Florida bay. Another objective was to find evidence in support of benthic microalgal mediation of phosphate flux to, or from, the sediment to the water column. The final objective was to evaluate nutrient limitation on benthic microalgal growth across Florida Bay for comparison with known zones of nutrient limitation in pelagic phytoplankton.

Chapter 2. Sediment phosphate flux in Florida Bay, USA. Introduction Florida Bay is a subtropical lagoonal estuary south of the Everglades system in Florida, USA. It is considered a primarily phosphate limited system in the interior and eastern portions of the bay, while the northern and western bay temporally differ in limiting nutrients (Tomas et al. 1999). Adverse ecological effects to the bay have occurred since the 1980‟s (Carlson et al. 1990; Butler et al. 1995). Water management practices in the Everglades system, which supplies freshwater to the bay, have been most- implicated as the cause of these perturbations. Recent and proposed changes to the Everglades were brought about to alleviate and restore Florida Bay. In order to better understand this ecosystem and the consequences of Everglade‟s restoration efforts to Florida Bay, scientists have been gathering information to construct a nutrient budget and 3D models for the region (Madden and MacDonald, 2007). The sediment/water column nutrient flux is an important component of this nutrient budget. Seagrasses and attached and drift macroalgae dominate the 2200 km2 submerged vegetative landscape of Florida Bay. These highly productive macrophytes also dominate the primary production of the bay (Nielsen et al. 2005; Yarbro and Carlson, 2008). However, nearly 1/3 of the bay bottom is devoid of seagrass and in these areas benthic microalgae (BMA) would be a source of primary production (Prager and Halley, 1997). Field studies have indicated that nutrification of the water column can be ameliorated as benthic microalgae near nutrient sources take up excess nutrients (Welsh, 1980; Admiraal, 1977). This can occur very quickly, on the order of a few days (Darrow, et al. 2003). Ulthicke and Klumpp (1997) found that ammonium addition led to increased chlorophyll and phaeopigment content of the benthic microalgal layer, which also led to higher maximum gross and daily net production. Heil et al. (2004) suggests that this buffering capacity of the microphytobenthos has reduced the effects of coastal nutrification to the Great Barrier Reef, Australia. Sundback and McGlathery (2005) 7 reference the buffering capacity of microphytobenthos in conceptual interaction models of coastal eutrophication. Fickian diffusive processes provide the baseline for the flux of nutrients to and from the sediments. However, other physical processes, such as tidal pumping and the passage of waves, increase porewater nutrient advection rates (Zeitzschel, 1980, Harrison, et al. 1983) from the bottom to the overlying water. Keizer et al. (1989) found increased water column P and Si concentrations in the “swash zone” of incoming tides also due to changes in pressure and water velocity. Bioroughness, or microtopography of the bottom, causes increased porewater flushing immediately upstream and downstream of ripples, and within small depressions due to small-scale changes in pressure (Thibodeaux, 1987; Huettel and Gust, 1992). This localized flushing delivers nutrients to the benthic microalgal resources on the surface of the sediment and may explain patchiness. Finally, the sediment/water column nutrient flux is also an important missing link, since Florida Bay is a nutrient limited system (Fourqurean et al. 1992; Fourqurean et al. 1993; Lapointe, 1989; Phlips et al. 1995; Phlips and Badylak, 1996). Whether the BMA play a role in sediment/water column flux of Florida Bay is completely unknown. Such information would better characterize any historical adverse impact man has had on these environments and how meaningful any attempt at remediation might be. The objective of this study was to address the relationship between benthic microalgal communities and the phosphate nutrient dynamics of Florida Bay sediments and how they relate to benthic and water column primary production. I present my findings on whether the nutrient fluxes that may limit benthic microalgal productivity differ between regions in Florida Bay. Materials and Methods In situ measurement of soluble reactive phosphate (SRP), total dissolved phosphate (TDP) and total dissolved nitrogen (TDN) were made from benthic chambers (<7L) inserted into submerged sediment devoid of seagrass in Florida Bay. Clear and opaque chambers (Fig. 2.1) were deployed for 3 to 9 hours per day. Between 1 and 4 replicates of each chamber type were used in each deployment.

To ensure a good seal, the acrylic chambers are inserted into the sediment and held down with dive weights if necessary. These chambers contain a gentle water circulation device (a fan blade that was manually turned), but effectively measure only diffusive flux. These chambers are similar to those used by Cahoon and Cooke (1992). The chambers had 1 liter of ambient water reservoirs (collapsible IV bag filled with site seawater collected at the time of deployment) attached to them to replace sample volumes. This replacement reservoir was used to prevent extraction of sediment porewater because of vacuum pressure caused by reducing the volume in the chamber. Redox reactions within the sediment can mobilize P release into the porewaters and overlying water column. Therefore, dissolved oxygen (DO) levels within chambers were frequently monitored. DO levels, measured by microelectrodes inserted into the chambers, were typically between 6 and 8 mg l-1 and never decreased below 1.7 mg l-1 during the monitored incubations. Placement and recovery of the chamber and sample collection was accomplished by wading. Concurrent with the field incubations, ambient irradiance was measured

9 periodically using a Licor 2Quantum sensor and light meter. Temperature and salinity measurements were made periodically during the incubation period. Studies were conducted bimonthly between May 2000 and August 2002 and were not synoptic. A major disadvantage of using benthic chambers is their disruption of the normal mixing of the water column and surface sediments by eliminating wave and current effects (Malan and McLachlan, 1991). Enclosure of the bottom may also alter the vertical distribution of microorganisms within the benthos. Additionally, decreased diffusion rates of gases into and out of the sediments can also occur, limiting photosynthesis (MacIntyre and Cullen, 1995). However, these chambers are the best method available for measuring oxygen production, since a semi-closed system is required. Sampling sites were Whipray Basin, Arsnicker Keys, and Carl Ross/Sandy Key (Figure 2.2); providing representation of north-central, eastern and western Florida Bay, respectively. Carl Ross/Sandy Key was the site most exposed to wind and tidal currents and a bird rookery was located on a nearby mangrove island. This site was open to the Gulf of Mexico to the south and west. There was less tidal action at Arsnicker Key; however, a tidal cut between a gap in the mangroves was located adjacent to the sampling site. Whipray Basin had no noticeable tide range, but was much more influenced by brackish water moving southward out of the Everglades and had a greater salinity range than the other two sites. All three sites were adjacent to mangrove islands and seagrass beds. Seagrasses (or rhizomes) and macroalgae were rarely located within any of the chambers. Approximately 60mls of sample water was extracted from each chamber at each sample period throughout the day. Depending upon the chamber design, water was removed using a syringe with an 18 gauge needle that pierced a septum or through a three way valve located on the side of the chamber.

10 Arsnicker Keys

Figure 2.2. A map of Florida Bay indicating the major physical and biological regions and the red dots indicate the locations sampled during this study.

Phosphate (P) and nitrogen (N) samples were immediately filtered through a fired GF/F syringe filter assembly. Filtrate was placed into a fired scintillation vial, and only in the case of TDP‟s 0.2mls of 0.17M MgSO4 was added. Samples were held in the field on ice until transported back to the laboratory at USF St. Petersburg where they were held at <0ºC until further analysis. Phosphate concentration was determined from discrete replicate samples using the colorimetric method of Solórzano and Sharpe (1980) on a Beckman Model DU520 spectrophotometer with a 5 cm cell path. Field samples were usually analyzed within two to four weeks following collection. The sample absorbance at 885λ was compared with a standard and matrix (de-ionized water) blanks and reagent blanks were subtracted from the samples to remove any error associated with the chemical additions. Dissolved

11 organic phosphate (DOP) was calculated by subtraction of SRP from TDP in the same chamber and sampling interval. Samples that were possibly contaminated, based upon comparisons among replicates, were culled from the dataset prior to statistical analysis. Dissolved organic nitrogen (DON) measurements were performed by Dr. Rob Masserini, USF on a Technicon Autoanalyzer II following Valderama (1981). DON was - determined after the oxidation of all N in the sample to NO3 . The organic value is then derived by subtracting the inorganic value, in this case NO3, NO2, and NH4, from the total dissolved nutrient value. Time, tide, cloud cover, wind direction and speed estimates were also made and compared with measurements obtained from nearby monitoring stations in the SEAKEYS data buoy system (data courtesy of the Florida Institute of Oceanography (FIO) and the National Oceanic and Atmospheric Administration (NOAA)). Parametric statistics were used to analyze the data, since values were normally distributed, and homogeneous in variances. Results and Discussion TDP, SRP and DOP are presented graphically in Figures 2.3 to 2.29. Not all chambers were able to be sampled at all times of the day due to failures so, when available, a portion of an incubation is presented for some chambers. These failures were due to high tides that prevented sampling or extreme weather events. Phosphate flux to and from the sediment in Florida Bay had high standard error and was variable by date and by location. It was frequently near the limits of detection by the colorimetric method of Solórzano and Sharp (1980). The absolute range of TDP flux was 8.1 M m-2 h-1 into the sediment (positive values) and 3.3 M m-2 h-1 out of the sediment (negative values). The greatest TDP flux out of the sediment occurred at Arsnicker Key in one dark chamber on July 11, 2001 (Figs. 2.12-2.14). This same chamber showed increased flux out of the sediment as the day progressed while most other chambers at that location shifted to flux into the sediment during afternoon hours. There were several instances of TDP flux into the sediment at both Carl Ross Key and Whipray Basin that were near the maxima, however the greatest TDP flux into the sediment occurred at Carl Ross Key also in July 2001 (Figs. 2.3-2.5).

12 SRP was variable by date and location and had a high standard error, but was always much lower than TDP. The absolute range of SRP flux was 6.6 M m-2 h-1 into the sediment (positive values) and 1.8 M m-2 h-1 out of the sediment (negative values). The greatest SRP flux into the sediment occurred at Carl Ross key in May 2002 (Figs. 2.6-2.8) and the greatest SRP flux out of the sediment occurred at Whipray in May 2001 (Figs. 2.24-2.26). It should be noted that SRP flux out of the sediment was very rare throughout the study in all locations and was usually near the limits of detection (Figs. 2.7, 2.16, and 2.25). Yet significant differences in SRP flux rates in the morning vs. the afternoon were found at all stations (Table 2.1). SRP flux out of the sediment only occurred in the afternoons and more frequently in dark chambers than in light chambers, but this treatment difference was not statistically significant between any of the locations when samples were grouped (Table 2.1). There were significant differences (p<0.01) in SRP flux between all locations when comparing time (morning vs. afternoon measurements) when samples were grouped, and in DOP flux between Arsnicker and Carl Ross Keys (p<0.01, Table 2.1).

Parameter Location Arsnicker vs Carl Ross Carl Ross vs End Arsnicker vs End Light vs Dark TDP 0.447 0.388 0.265 SRP 0.190 0.232 0.112 DOP 0.413 0.229 0.408 morning vs afternoon TDP 0.066 0.246 0.477 SRP 2E-07** 8E-12** 4E-19** DOP 0.0178** 0.460 0.456 **p=0.01

13 Sandy Key Morning TDP Summary 8.0 dark 1 dark 2

6.0 dark 3 1

- dark 4

h light 1

2 4.0 - light 2 light 3

2.0 M TDP m TDPM

 0.0

-2.0

-4.0

-6.0

1 4.0 -

h light 2 2 - light 3 2.0

TDP m TDP 0.0

M  -2.0

-4.0

-6.0

4.0 light 1 1

- light 2

h 2 - 2.0 light 3

0.0

TDP m TDP

M M  -2.0

-4.0

-6.0

Figure 2.5. Summary of Total TDP flux for Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment. 16 Sandy Key morning SRP summary 8.0 dark 1 dark 2 6.0 dark 3 dark 4

4.0 light 1

1 -

h light 2 2 - 2.0 light 3

0.0 SRP m SRP

M -2.0

-4.0

-6.0

4.0 light 1 1

- light 2

h light 3 2 - 2.0

0.0 SRP m SRP

M -2.0

-4.0

-6.0

18 Sandy Key Total SRP Summary 8.0 dark 1 6.0 dark 2 dark 3 dark 4 4.0 light 1

1 light 2 -

h 2.0 light 3

2 -

0.0 SRP m SRP

-2.0 M

-4.0

-6.0

light 2 1

- 2.0 light 3

2 - 0.0

DOP m DOP -2.0 M -4.0

-6.0

Figure 2.9 Summary of morning DOP flux rates (calculated) at Carl Ross Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment. 20 Sandy Key Afternoon DOP Summary 8.0 dark 1 dark 2 6.0 dark 3 dark 4 light 1 4.0 light 2

light 3 1

- 2.0

2 - 0.0

DOP m DOP -2.0 M

-4.0

-6.0

1 4.0

- light 1

h 2

- light 2 2.0 light 3

DOP m DOP 0.0 M -2.0

-4.0

-6.0

light 2 1

- 2.0

2 - 0.0

07/11/01 05/02/02 TDP m TDP

-2.0 M

-4.0

-6.0

6.0 dark 2

light 1 4.0

light 2 1

- 2.0

2 - 0.0

07/11/01 05/02/02 TDP m TDP

-2.0 M

-4.0

-6.0

Figure 2.13. Summary of afternoon TDP flux rates at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment. 24 Arsnicker Total TDP Summary 8.0 dark 1 6.0 dark 2

4.0 light 1

light 2 1

- 2.0

2 - 0.0

07/11/01 05/02/02 TDP m TDP

-2.0 M

-4.0

-6.0

light 2 1

- 2.0

2 - 0.0

07/11/01 05/02/02 SRP m SRP

-2.0 M

-4.0

-6.0

dark 1 6.0 dark 2

4.0 light 1

1 -

h light 2

2 - 2.0

SRP m SRP 0.0

07/11/01 05/02/02 M -2.0

-4.0

-6.0

6.0 dark 2

4.0 light 1 1

- light 2 h

2 2.0 -

0.0

07/11/01 05/02/02 SRP m SRP

M -2.0

-4.0

-6.0

Figure 2.17. Summary of total SRP flux rates at Arsnicker Key. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment. 28 Arsnicker Morning DOP summary 8.0 dark 1 6.0 dark 2

4.0

light 1

1 -

h 2.0 light 2

2 -

0.0

07/11/01 05/02/02 DOP m DOP

-2.0 M

-4.0

-6.0

6.0 dark 2

light 1 1

- 4.0 h

2 light 2 - 2.0

DOP m DOP 0.0

07/11/01 05/02/02 M -2.0

-4.0

-6.0

30 Arsnicker Total DOP Summary 8.0 dark 1 6.0 dark 2 4.0 light 1

2.0 light 2

h 2

- 0.0 07/11/01 05/02/02

-2.0 DOP m DOP

-4.0 M

-6.0

-8.0

light 2 1 - 2.0

h light 3

2 - 0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 TDP m TDP

-2.0 M

-4.0

-6.0

Figure 2.21. Summary of the morning TDP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Morning flux represents the net P flux in that chamber through the mid-day sampling period – usually between 12:00 and 13:00 EST. Positive flux is into the sediment, while negative flux is out of the sediment. 32 Whipray Afternoon TDP Summary 8.0 dark 1 dark 2 6.0 dark 3 light 1

4.0 1

- light 2

h 2 - light 3 2.0

TDP m TDP 0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 M -2.0

-4.0

-6.0

4.0 light 1

1 -

h light 2 2 - 2.0 light 3

TDP m TDP 0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 M -2.0

-4.0

-6.0

4.0 1

- light 2 h

2 light 3 - 2.0

SRP m SRP 0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 M -2.0

-4.0

-6.0

light 3 1

- 2.0

2 - 0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 SRP m SRP

-2.0 M

-4.0

-6.0

Figure 2.25. Summary of afternoon SRP flux at End Basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Afternoon flux represents the net P flux in that chamber since the mid-day sampling period. Positive flux is into the sediment, while negative flux is out of the sediment. 36 Whipray Total SRP Summary 8.0 dark 1 6.0 dark 2 dark 3 light 1 4.0 light 2

light 3 1

- 2.0

2 - 0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 SRP m SRP

-2.0 M

-4.0

-6.0

light 3 1

- 2.0

2 - 0.0 05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02

DOP m DOP -2.0 M

-4.0

-6.0

4.0 light 1 1

- light 2

h 2 - 2.0 light 3

0.0 DOP m DOP 05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02

M -2.0

-4.0

-6.0

Whipray Total DOP Summary 8.0 dark 1 dark 2 6.0 dark 3 light 1 4.0

light 2

1 -

h light 3 2

- 2.0

0.0

05/15/01 06/13/01 07/16/01 11/09/01 01/17/02 01/18/02 04/04/02 05/03/02 05/04/02 07/17/02 08/02/02 M DOP m DOP M

 -2.0

-4.0

-6.0

Figure 2.29. Summary of total DOP flux rates (calculated) at End basin. Individual dark chambers represented in blue colored bars and individual light chamber results represented in yellow colored bars. Total flux represents the net P flux in that chamber at the end of the incubation period. Positive flux is into the sediment, while negative flux is out of the sediment. 40

In general, trends in P flux were consistent among chambers at the same location and sampling date, and only differed in the magnitude of the flux. There were some instances when one or more treatment chambers deviated substantially from the others in the direction of P flux (example, Figure 2.13). Due to the relatively small chamber size, these inconsistencies between chambers were uncommon and may represent bioturbation or seeps and sinks in the underlying sediment/groundwater system. Whipray Basin and Arsnicker Keys exhibited similar phosphate flux values as Carl Ross Key during the same sampling periods (Figs. 2.3-2.29) despite differences in sediment grain size (observed) and reduced tidal pumping. Hourly P flux rates reported here are similar to those of Yarbro and Carlson (1999; 2008). There is no clear pattern between phosphate flux in light and dark chambers at different locations, when enough measurements were made to allow statistical tests. Nor was there any statistically significant differences between these treatments (Table 2.1). No clear seasonal pattern of phosphate flux was found. For the most part, time comparisons (morning vs. afternoon) between light and dark treatments at each station for each date were not statistically significant, although there were several exceptions no overall pattern was discernable (Table 2.2). DOP is the primary component of the available P pool in the water column, although the amount and TDP:DOP ratio varies across Florida Bay from West to East, with greatest values in the central region (Table 2.3). Flux of organic P dominates the system, whether into or out of the sediment, since organic P concentrations exceeded inorganic P concentrations in all locations at most times of the year and in most sampling chambers. Table 2.3 shows the average hourly DOP flux rates by bay region for the entire study period. The ratio of organic to inorganic P flux is high except at the southeastern station. The central station had the highest ratio (71% DOP). This finding is consistent with what others have found in Florida Bay, but contrasts with most oceanic systems where inorganic forms of P are predominant (Riley and Chester, 1972).

Table 2.2. Student‟s T-test results showing statistical significance between replicate light and dark treatments by date and parameter.

Date Location TotalTDP TotalSRP TotalDOP PMTDP PMSRP PMDOP AMTDP AMSRP 07/11/01 Arsnicker 0.065+ 05/02/02 Arsnicker 11/13/00 Sandy Key 11/14/00 Sandy Key 0.046* NV NV NV 01/24/01 Sandy Key NV NV NV NV NV NV 03/13/01 Sandy Key 03/15/01 Sandy Key NV NV NV 03/16/01 Sandy Key 0.026* 05/16/01 Sandy Key 0.004** 0.056+ 0.051* 05/17/01 Sandy Key 07/12/01 Sandy Key NV NV NV NV NV NV NV NV 07/13/01 Sandy Key NV NV NV 07/14/01 Sandy Key 09/19/01 Sandy Key 09/20/01 Sandy Key 11/29/01 Sandy Key 11/30/01 Sandy Key 05/07/02 Sandy Key 0.089+ 0.019* 0.081+ 05/15/01 Whipray 06/13/01 Whipray 07/16/01 Whipray 11/09/01 Whipray 01/17/02 Whipray 0.035* 01/18/02 Whipray 0.072+ 0.070+ 04/04/02 Whipray 0.062+ 0.092+ 0.061+ 0.03* 05/03/02 Whipray 05/04/02 Whipray 07/17/02 Whipray 0.097+ 42

Table 2.3. Average DOP flux in Florida Bay.

Treatment Location Light Dark (Mm-2h-1) (Mm-2h-1) Western 0.227 0.189 %DOP 57% 56%

Central 0.262 0.265 %DOP 71% 71%

Southeastern 0.055 0.076 %DOP 18% 22%

As mentioned before, dissolved organic P flux at all sites tended to be into the sediment in the morning hours with diminished magnitude of flux into the sediment, or flux out of the sediment, during afternoon. The reason for this is unknown, but I speculate that it is due to uptake by BMA that are increasing primary production in response to light availability. The central bay (End Key) station more frequently exhibited total dissolved organic P flux into the sediment from the water column and this trend was of greater magnitude in light chambers (Figs. 2.27-2.29). In eleven of sixty-six instances there were statistically significant differences (p<0.1) between all forms of P flux in the light and dark treatments, three at the p=0.05 level (Table 2.2). The differences in the light chambers is presumed to be indicative of enhanced algal uptake or release of nutrients in the water – whether BMA or pelagic phytoplankton. In contrast, the Carl Ross Key station more frequently exhibited total dissolved organic P flux out of the sediment into the water column (Figs. 2.9-2.11), while the magnitude and direction of flux was more variable at Carl Ross Key. Total organic P flux was more balanced between light and dark chambers at the Carl Ross Key station than at the End station. However, light chambers exhibited release of DOP into the water column in the morning hours while dark chambers showed diminished DOP release into the water column or flux into the sediment. As with End, the magnitude of P flux 43 diminished in the afternoon at Carl Ross Key. In ten of ninety-six instances, there were statistically significant differences between all forms of P flux in the light and dark treatments, five at the p=0.05 level and one at the p=0.01 level. (Table 2.2). Arsnicker Key was only sampled twice with few consistencies between the treatments among sampling dates. This location had a trend similar to Carl Ross Key where DOP flux was out of the sediment in the morning and then into the sediment in the afternoon, leading to a net flux of DOP out of the sediment for all treatments on both sampling events (Fig.2.21-2.23). Student‟s T-test indicated total SRP values between light and dark treatments were significantly different (p=0.1) in one of the two sampling events (Table 2.2). Based upon growth rate requirements estimated from in situ growth rates and biomass in western Florida Bay (Vargo et al. 1999), DOP flux from the sediment could supply 6.5 - 41% of the P required by water column phytoplankton during bloom and nonbloom conditions, respectively. Nitrogen Total dissolved nitrogen (TDN) was comprised mostly of dissolved organic forms throughout Florida Bay, in excess of 50%. TDN flux ranged from -408.69 (out of the sediment) to 393.59 M m-2h-1 (into the sediment), but averages 38.38 M l-1 throughout the bay. Table 2.4 shows the average TDN values and fluxes by location in the bay. Absolute values were more variable in the central bay; however, average values were quite similar among the three locations. Southeastern Florida Bay showed the least variable flux rate with no flux into the sediment, but this was probably because it was only sampled twice. I did not evaluate statistical differences in morning vs. afternoon or light vs dark chambers for TDN. Still, TDN flux from the sediment could supply 100% of the nitrogen demand of the water column phytoplankton, providing the form of available nitrogen is usable by the cells.

44 Table 2.4. Average TDN values and range of flux in Florida Bay. The last column provides the range of the percent increase from initial values during incubation.

TDN % of initial Avg H2O Column Flux TDN Location (M l-1) (Mm-2h-1)

-208.08 to Western 37.58 153.99 47-275%

-408.64 to Central 40.86 393.59 33-359%

Southeastern 47.94 14.32 to 47.51 111-136%

Summary Despite temporal and spatial variability, phosphorus flux from the sediment would be an important source of P to both the benthic microalgae, macrophytes and phytoplankton in the overlying water column. I estimate that 6-41% of the water column phytoplankton needs could be met by benthic flux. Based upon the area of bay at 2200km2, 1/3 of which is available for microphytobenthos colonization, and a loss of 4000 hectares of seagrass due to dieoff, my average TDP flux values would contribute ~1650 metric tons of P to the entire system per year, 9.1 metric tons per year from areas newly devoid of seagrass due to dieoff. My calculation could be an overestimate since the dark incubation rates were simulated and were not done during the normal dark cycle when diurnal processes could have resulted in net P uptake by the sediment. Repeated anoxia events, or shoaling and deepening of the anoxic layer, would release loosely bound P buried in the sediment. SRP values were frequently near the limits of detection, yet significant differences in SRP flux rates in the morning vs. the afternoon at all stations indicate that benthic microalgal community could be mediating P fluxes during periods of high photosynthesis. Since organic forms of phosphorus dominate Florida Bay, P is being recycled through organisms like benthic microalgae. 45 Although more typical of a shallow coastal lagoon system, autochthonous organic matter accumulation is high in the bay. This is further enhanced by the long residence time of some basins and the lack of tidal flushing for more ~½ of the area of the bay. This suggests the bay would be a sink for nutrients from the upstream sources. Rizzo (1990) and Sundback et al. (1991) both documented nutrient uptake by the microphytobenthos both from the sediment and the water column. Nutrients taken up from the water column allow the microphytobenthos to act as a temporary nutrient sink in Florida Bay. Nutrient flux from the sediment would be a source and my findings indicate it could support water column production. The duration and magnitude of what this nutrient flux would support depends upon pelagic phytoplankton biomass and need, which would vary. Sundback and McGlathery (2005) suggest the trophic status of the sediment controls whether benthic-pelagic interactions are coupled or decoupled. Autotrophic (oxic) sediments, and tropical carbonate sediments, lead to a decoupled benthic-pelagic system as the microphytobenthos traps and recycles nutrients, preventing release to the water column. Heterotrophic sediments are coupled with the water column system as anoxic conditions promote the flux of nutrients from the sediment through the microphytobenthos community and into the overlying water column. In their scenario, the coupling/decoupling persists for weeks to months in the environment before switching in response to changing physical and chemical conditions. Due to the amount of organic matter in the bay (Halley et al.1999) and proximity of the anoxic layer to the sediment surface, conditions in Florida Bay where microphytobenthos dominate, might favor this alternating scenario, but occurring on the order of hours to days – perhaps even on a diurnal cycle. The presence of Laurencia mats during some times of the year, although not investigated in this study, would enhance sediment-water column interactions. This is because the mats cause shoaling of the sediment anoxic layer beneath the mat and in some cases, depending upon thickness of the mat, would include anoxic conditions within this microhabitat. Yarbro and Carlson (2008) do not report consistent diel patterns in P flux in benthic chambers deployed within seagrass beds, despite diurnal changes from net

46 autotrophy to net heterotrophy in the chambers. However, the seagrass and epiphyte community is known to dominate nutrient cycling both by changing oxic conditions in the rhizosphere and by direct uptake and release of nutrients from both the sediment and water column (Koch, et al. 2005; Nielsen, et al. 2005; Yarbro and Carlson, 2008). The diel effect of a complete shift between autotrophy and heterotrophy on nutrient flux in the microphytobenthos community would be dampened relative to that found within seagrass beds. Although unresolved in this study, tidal pumping of P from the sediments may be important in western Florida Bay due to the variability in the sediment grain size, although peak P concentrations did not seem to correspond to any consistent tidal level. Tidal influences are diminished at Arsnicker Keys and there is nearly no tidal range evident in End Basin. This research will enhance the current understanding of benthic- pelagic interactions in western Florida Bay and near shore environments, and provide baseline estimates for nutrient exchange between the benthos and water column which include biotic factors.

Chapter 3. Benthic Microalgal Chlorophyll a Standing Stock in Florida Bay Introduction Benthic Microalgae (Microphytobenthos) Because their production is underappreciated, benthic microalgae that live in the upper centimeters of the sediment where the bottom lies within the euphotic zone have been termed the “secret garden” (MacIntyre et al. 1996). These benthic microalgae (BMA) can form dense mats at the sediment surface (Grant and Gust, 1987) and are comprised mostly of cyanobacteria and diatoms (MacIntyre and Cullen, 1995; Pinckney et al. 1995). Benthic microalgae can also be termed microphytobenthos. Numerous studies have shown that benthic microalgal biomass and primary production can equal or exceed that of the water column (Cahoon and Cooke, 1992; Nelson et al. 1999; Charpy and Charpy-Roubaud, 1990), but spatially, benthic microalgal biomass can be very patchy (Rizzo and Wetzel, 1985). Even within seagrass beds the microphytobenthos is an important contribution to overall benthic primary production (Moncreiff, 1992). Primary production by the microphytobenthos is poorly quantified on a broad geographic scale, and ignored in most monitoring programs (Cahoon, 2006). Those measurements of benthic microalgal biomass and primary production that have been published are generally from European estuaries or US waters (Tables 3.1 and 3.2, from Cahoon, 2006; Cahoon, 1999; Heil et al. 2004; Colijn and De Jonge, 1984). Benthic primary production is effectively limited to the upper millimeter or so of the sediment; basically that depth to where light penetrates permitting photosynthesis by the cells (MacIntyre and Cullen, 1995). However, viable cells have been found to several centimeters in depth, provided the sediments didn‟t become anoxic (MacIntyre and Cullen, 1995; Nelson et al. 1999; Aller et al. 2001).

48 Table 3.1 (reproduced from Cahoon, 2006, with permission). Spatial distribution of intertidal and subtidal studies measuring microphytobenthic production by all methods as of 2003. Data are numbers of published studies from Cahoon (1999) and more recent references. Basin Subasin Subtidal Intertidal Atlantic Ocean 14 31 Gulf of Mexico/Caribbean 8 2 Mediterranean 4 - Black Sea - - Baltic Sea 3 -

Antarctic Ocean 2 -

Arctic Ocean 4 - Russian Arctic - - Canadian Arctic - -

Indian Ocean 2 1 Persian Gulf - -

Pacific Ocean 7 7 Bering Sea - - East Asia - 2 Gulf of California 3 - SE Asia-Australia 1 1 New Zealand 1

49 Table 3.2. (reproduced from Cahoon, 2006, with permission). Values of microphytobenthic photosynthetic parameters reported in published literature: Ek -2 -1 (saturating light intensity, mol photons m s ), Pmax (maximum, biomass-normalized photosynthetic rate, mg C mg chl a-1 h-1), alpha (slope of P-E relationship, mg C mg chl a-1 h-1 (mol photons m-2 s-1) -1), P (biomass-normalized photosynthetic rate mg C mg chl a-1 h-1. b Reference Ek P max a P Steele & Baird 1968 - - - 0.14-1.78 Admiraal & Peletier 1980 - 0.1-13 - - Hargrave et al. 1983 - 0.1-7 - - Rasmussen et al. 1983 160-360 - 0.013-0.028 0.16-0.57 Colijn & de Jonge 1984 - 0.43-0.49 - - Palmisano et al. 1985 11 0.21 0.022 - Mills & Wilkinson 1986 300->750 3.74 - - Rivkin & Putt 1987 6 - 0.08-0.1 0.53-0.60 Blanchard & Montagna 1992 108-215 2.98-20.0 0.01-0.16 - Brotas & Catarino 1995 - - 0.0021 - MacIntyre & Cullen 1995 - - 0.035-0.08 ~1-12 Barranguet et al. 1998 150-450 3-12.7 0.008-0.042 - Kromkamp et al. 1998 450-1200 - - - Wolfstein & Hartig 1998 56-297 0.66-1.5914C 0.011-0.057 - 1.7-4.10DOE Ulthicke & Klumpp 1998 482-975 - - - Meyercoordt & Meyer-Reil 1999 27-367 4.04-54.2 0.017-0.339 - Meyercoordt et al. 1999 26-240 4.8-17.0 0.026-0.188 0.5-8.80 Barranguet & Kromkamp 2000 200-400 18-Feb 0.015-0.035 - Goto et al. 2000 110 0.95 0.008540.4-0.8 Guarini et al. 2000b 460 11.2 0.024 - Miles & Sundback 2000 - - - 0.26-0.52 Kuhl et al. 2001 4.6-6.9 - - - Perkins et al. 2001 750 0.2 - - Glud et al. 2002 33 - - - Guarini et al. 2002 16-30 2.3-4.7 0.033-0.165 - Perkins et al. 2002 160-420 - 0.021-0.028 -

Pennate diatoms can be found free-living in and on the sediments (MacIntyre and Cullen, 1995) or living within polysaccharide tubes interstitially (Yallop et al. 1994). Vertical migration by motile benthic diatoms in the intertidal zone may be used as a means to escape desiccation (Pinckney and Zingmark, 1991) and UV exposure at low tide (Cooke, 1991). Other reasons cells are found at depth in sediments is due to burial (physical mixing of the sediments) from tides, currents and waves (MacIntyre and Cullen, 1995; MacIntyre et al. 1996), and bioturbation.

50 Benthic Microalgae and Sediment Resuspension Some microalgae can secrete a mucilaginous sheath or polysaccharide matrix binding themselves both to other organisms and the sediment grains (Yallop et al. 1994), thus providing an effective mechanism to stabilize the sediment on which they reside and minimize the effects of turbidity at the sediment/water interface. In addition to primary production in coastal ecosystems, benthic microalgal mats have been shown to dramatically increase the turbulence and shear velocity needed to erode sediments (Yallop et al. 1994). The binding capabilities of their extruded exopolymer polysaccharides are extraordinary and can even increase with desiccation (Yallop et al. 1994). When erosion does occur, benthic microalgal mats can quickly re-establish themselves, often in as little as a few hours. The sediment types associated with microalgal mats can be quite diverse and the stabilizing properties of these mats is dependent to some degree on the type of sediments on which they are found and the species of microalgae growing there (Yallop et al. 1994; Amspoker and McIntire, 1978). Benthic microalgae physically alter the benthic environment for infaunal and epifaunal organisms by stabilizing the sediments. The are an important food source for benthic filter feeders like gastropods, bivalves, sponges and infaunal worms (Cahoon and Cooke, 1992), which in turn support higher trophic levels. Benthic cells re-suspended into the water column due to tides or wind events are potential food sources for both benthic and pelagic organisms (MacIntyre and Cullen, 1996). Baillie and Welsh (1980) showed that tidally re-suspended benthic microalgae were transported directly to the oyster reefs found at the fringes of mudflats adjacent to a saltmarsh. The locations of greatest re-suspended cell transport coincided with these oyster reefs indicating the oyster reefs flourished where the maximal food availability was found (i.e. re-suspended benthic microalgae). Baillie and Welsh, (1980) also calculated that if the top 1 mm of sediment were resuspended in 10-15% of the mudflat in Branford Harbor, Connecticut then the benthic microalgal cell numbers in the water column would equal that of the pelagic phytoplankton numbers. De Jonge, (1985) found more benthic microalgae in the water column due to resuspension by physical processes than was found in the upper 0.5 cm of

51 the sediment. Resuspension of sediment along with benthic microalgal cells may offset any potential increase in water column primary productivity by decreasing light availability at depth and to the benthos. MacIntyre et al. (1996) suggest that the water column and sediment should not be treated as mutually exclusive environments, but that this interface is a dynamic area with influence from both systems. C3 plants are more photosynthetically efficient at low to intermediate light availability than high light, which can have implications for competition (Lawlor, 2001). Photoinhibition has generally not been shown in benthic microalgae (Colijn and van Buurt, 1975; MacIntyre and Cullen, 1995; Mills and Wilkinson, 1986; Charpy and Charpy-Roubaud; 1990); probably as a consequence of the potential light limited conditions from turbidity at the sediment/water interface and lack of light penetration beyond the upper few millimeters of sediment. In contrast, MacIntyre and Cullen (1995) found photoinhibition to be prevalent in both pelagic phytoplankton and benthic microalgae resuspended into the water column in San Antonio Bay, although measurements of nonresuspended benthic microalgae were not part of the study. Benthic Microalgal primary productivity and models Photosynthesis is the process by which light energy, water and carbon dioxide are converted into complex, energy-rich organic molecules by chlorophyll containing plants (Lawlor, 2001). More specifically it is an oxidation-reduction reaction, where electron donor and acceptor molecules in a cell bearing photosynthetic pigments and exposed to light split water molecules resulting the release of heat energy and O2 (Lawlor 2001). Respiratory losses balance photosynthetic gains of oxygen. The photosynthetic rate of plants involves the orientation of the cell to the light, the amount of light, temperature and available nutrients (Forster and Kromkamp, 2006). Beyond the limiting nutrients of nitrogen, phosphorus and silica, dissolved inorganic carbon may become limiting to the microphytobenthos during periods of high production and high porewater O2 (Forster and Kromkamp, 2006). As a result of only some of the microphytobenthos cells being functionally able to photosynthesize, by their proximity to the sediment surface and as a result of migratory behavior, they are termed the Photosynthetically-Active Biomass (PAB) (Serodio et al. 1997).

52 Different methods of modeling primary production by the PAB have been developed, including the SML (surface monolayer), UNI (uniform distribution of cells with depth) and the KY models (exponentially decreasing gradient of chlorophyll with depth in the sediment). A comparison of these models, by Forster and Kromkamp (2006) indicated that all three models were capable of simulating actual measured field conditions, although all overestimated actual primary production. Some models were more strongly affected by the application of various rates of respiration between 1 and 10% of gross primary production. However, the rates used in this model were estimated because respiration is seldom measured and varies due to light and temperature regimes. However, all the models would be enhanced by the application of more site specific measurements and more refined estimates of all parameters, most particularly respiration. Darrow et al. (2003) developed a model of benthic primary production on the West Florida Shelf that coupled regenerated nutrients from a pelagic phytoplankton bloom to subsequent production by benthic diatoms. Their one dimensional model, with 16 variables, was used to model the benthic diatom community response to a pulse of riverine water over the shelf that resulted in a bloom. The remnants of the pelagic bloom was transported to the sediment and provided nutrients sufficient to stimulate benthic production. The benthic production effectively served as a temporary sink of regenerated nutrients on the shelf. Global microphytobenthic primary production has been estimated at 0.34 – 0.5 Gt C yr-1 based upon average productivity to a depth of 50m (Charpy- Roubaud and Sournia, 1990) and depth weighted production estimates (Cahoon, 1999), respectively. Tropical areas tend to have higher productivity rates, as expected with warmer, clearer water and greater light levels of longer duration, averaging 527 g C m-2 yr-1 and biomass values of 90-350 mg Chl a m-2 (Cahoon, 2006). However, these are likely overestimates since they are scaled up from smaller studies, utilize different measuring techniques and do not adequately account for localized differences in light and temperature regimes that control productivity. Significant differences in productivity exist between inter- and subtidal zones. It is impractical to make broad spatial and temporal synoptic measurements and,

53 in general, there are too few studies worldwide to have an adequate understanding of the range of global productivity of the microphytobenthos. Benthic Microalgae in Florida Bay Florida Bay began experiencing massive die-off of seagrasses in the late 1980‟s (Carlson et al. 1990), and, more recently, persistent algal blooms (Figs.3.1-3.3). Diminished light transmittance to the benthos has been an ongoing problem in some basins due to sediment resuspension and algal blooms. This problem also has recently increased in magnitude. As a result of the die-off, the macrophyte dominated benthic primary production has been shifting to a phytoplankton dominated system in many basins (Phlips and Badylak, 1996). Chlorophyll a and primary production of pelagic phytoplankton are commonly measured water quality parameters. Extensive monitoring data exists for Florida Bay, although it has only recently received increased attention from the scientific community (Fourqurean et al. 1992; Rudnick et al. 1999; Tomas et al. 1999). Tomas et al. (1999) found that the lowest water column biomass and growth rates were in the eastern regions of Florida Bay, with intermediate values in the western bay while central Florida bay exhibited both the highest biomass and growth rates. This finding is partially consistent with the distribution of dissolved nutrients although phosphate availability appears to be limiting throughout most of the bay (Fourqurean et al. 1992. As information about Florida Bay accumulates, it is becoming obvious that the biomass of the benthic microalgal community has been under sampled and little is known about the coupling of benthic - pelagic primary production. The contribution of benthic microalgal productivity in seemingly unvegetated bottom areas to the baywide primary productivity has also not been quantified. Approximately one third of the 2200km2 bay is unvegetated, inter- or sub-tidal bottom (Prager and Halley, 1997) and about 4000 hectares of seagrass have been permanently lost due to the die-off (Carlson, et al. 1990; Durako and Zieman, 2007) making that bottom now available for enhanced benthic algal production, varying from a thin film of diatoms over relatively unconsolidated sediments to inch thick Laurencia sp. mats carpeting the bottom.

Figure 3.1. October 2005, Map of colored water in Florida Bay. Reproduced with permission, FWC-FWRI.

Figure 3.2. November 1995, Map of colored water in Florida Bay. Reproduced with permission FWC-FWRI.

Figure 3.3. December 1995, map of colored water in Florida Bay. Reproduced with permission, FWC-FWRI.

The contribution of benthic microalgal productivity in seemingly unvegetated bottom areas to the baywide primary productivity has also not been quantified. Approximately one third of the 2200km2 bay is unvegetated, inter- or sub-tidal bottom (Prager and Halley, 1997) and about 4000 hectares of seagrass have been permanently lost due to the die-off (Carlson, et al. 1990; Durako and Zieman, 2007) making that bottom now available for enhanced benthic algal production, varying from a thin film of diatoms over relatively unconsolidated sediments to inch thick Laurencia sp. mats carpeting the bottom. Coarser coral-algal sediments are found in the eastern region of the bay (Fourqurean et al. 1993). Fine grained mud or sands are prevalent elsewhere, especially in the more restricted basins. These fine-grained sediments and the attached or free- living microalgae are susceptible to resuspension into the water column. Many different causes of benthic microalgal resuspension have been suggested, ranging from continental shelf sediments stirred up by the passage of hurricanes and strong storms off North Carolina (Cahoon and Cooke, 1992) to resuspension of both fine and coarse grained sediments by tides and waves (MacIntyre and Cullen, 1996; Baillie 56 and Welsh, 1980). The latter forces would probably be most implicated in Florida Bay given its shallow depth. In fact, I have personally observed that a sustained 10-20 knot wind has reduced secchi depth from >1m to <0.4m in a matter of minutes in western Florida Bay. Resuspension of sediments during sustained wind events has been less evident to me in central and southeastern Florida Bay, but, algal blooms do decrease water clarity at these locations. Goals and objectives The purpose of this study was to determine the chlorophyll a concentration of microphytobenthos in Florida Bay and calculate the potential areal benthic microalgal production. These are key measurements to include in any model of Florida Bay. The goal was to measure the temporal and spatial distribution of benthic microalgal chlorophyll a within Florida Bay. Specifically, whether benthic microalgal biomass contributes significantly to the overall ecosystem primary productivity where seagrasses are sparse or absent, and if benthic microalgal chlorophyll values vary seasonally, or spatially. Materials and methods Samples were collected from three locations within Florida Bay, USA between May 2000 and August 2002. Sampling sites were Whipray Basin, Arsnicker Keys, and Carl Ross/Sandy Key; providing representation of north-central, southeastern and western Florida Bay, respectively. The Carl Ross/Sandy Key location was representative of intertidal environments in the bay, while the other two locations were both representative of subtidal microphytobenthos. Chlorophyll is a lipophilic molecule, a good electron donor and receptor (when oxidized) and probably the most abundant biological pigment on earth (Lawlor, 2001). Benthic chlorophyll samples were collected at the end of the sampling day from sediment where benthic nutrient flux chambers (light and dark, ~7 liter volume) had been deployed for coincident nutrient flux measurements throughout the day. Five replicates sediment mini-cores selected by chance were taken from within the 0.06m2 area of sediment the chamber enclosed. Usually 4 or 5 chambers were deployed resulting in up to 25 benthic chlorophyll samples per sampling day.

57 Mini-corers were 10cc syringes with the ends cut off. The sleeves of these „mini- corers‟ were inserted into the sediment while simultaneously retracting the plunger to preserve the upper layer containing the majority of the benthic microalgae. Due to the loose consistency of the sediment, I used my fingertip to „cap‟ the mini-corer prior to removal from the sediment. Seagrasses, macroalgae and large debris were removed from the mini-cores. Mini-cores were standardized to 2cm depth, with any excess sediment below 2cm depth expelled from the plunger prior to capping with Parafilm® strips. Replicate samples were placed into bags labeled with the chamber treatment, date and location and placed on ice for immediate transport back to the laboratory at USF. There was undoubtedly some degradation in chlorophyll a due to storage on ice for ~10 hours until freezing at the laboratory, but this was the most practical sampling and transport method. Every sample was treated this way so the associated sampling error would be consistent among samples. Samples were held for up to 2 months at -18ºC prior to extraction in 90% acetone with hexane fractionation, according to the method outlined by Whitney and Darly (1979). The only deviation from this procedure was the elimination of the addition of MgCO3, which was determined to be an unnecessary process. The Whitney and Darly (1979) method minimizes overestimation of chlorophyll a caused by chlorophyllide a contamination (a degradation product) which absorbs light in the same wavelengths as chlorophyll a. Chlorophyll a most strongly absorbs light at 430 and 660nm wavelengths. Briefly, this procedure requires extraction of the sediment in 90% acetone for 12 -24 hours at 4ºC in the dark. Then samples are centrifuged and decanted into separatory funnels to which hexane is then added. The acetone/hexane mixture is then shaken vigorously on a reciprocating shaker for 5 minutes. The acetone fraction, containing the chlorophyllide a and other degradation products, is then drained off the bottom and the overlying hexane fraction, containing the viable chlorophyll, is analyzed spectrophotometrically. The hexane fractions of the samples were measured, using a Beckman DU520 spectrophotometer, at four wavelengths both before and 2 minutes after acidification with 50% hydrochloric acid. This yielded measurements for chlorophyll a, b, and

58 phaeopigments other than chlorophyllide a. Replicate benthic chlorophyll samples were pooled for each treatment chamber for each sampling day and location. Benthic chlorophyll a and phaeopigment concentrations were determined on an areal basis (mg Chl a m-2) for the upper 2cm of sediment. Jennifer Jurado, University of Miami, collected water column chlorophyll a samples at the start of the sampling day only at the western Florida Bay location. A known volume of ambient station water was filtered through a GF/F filter in the field and stored in the dark in aluminum foil on ice until she returned to the laboratory. Samples were then frozen at 0ºC for up to 2 weeks until analysis. Samples were extracted in 100% methanol with DMSO according to the method of Shoaf and Lium, (1976) by Jennifer Jurado at the University of Miami. Samples were analyzed fluorometrically using a Turner Designs Model 10AU fluorometer, both before and after acidification with 1N hydrochloric acid. Results Benthic chlorophyll a values show seasonal patterns and interannual variability in Florida Bay. Benthic chlorophyll a values ranged from 0.56 - 112 mg Chl a m-2 throughout the bay. Benthic phaeopigment values ranged from 7.07 to ~350 mg phaeopigments m-2. The average baywide chlorophyll a value during my study was 14.98 mg Chl a m-2. The average baywide phaeopigment value during my study was 72.79 mg Chl a m-2. Depth integrated water column values from western Florida Bay ranged from 0.58 – 2.63 mg Chl a m-2, but averaged 1.48 mg Chl a m-2(Table 3.3). Phaeopigments dominated the pigment ratio spatially and temporally in the bay. The average chlorophyll a for all Florida Bay samples was 23% of the measured pigment per sample (ranged from 1-86%), and very rarely exceeded 50% (<1% of the total number of samples). Benthic microalgal biomass contained proportionally more phaeopigment than water column measurements, about 55 – 93% phaeopigment (Figs. 3.4- 3.6) when averaged among replicates compared to about 40% phaeopigment in the water column. This was due to the accumulation of dead plant matter at the sediment water interface (pelagic phytoplankton, macroalgal and seagrass detritus).

59 My sediment pigment values were also similar to those measured subtidally on the WFS near the middle grounds (Darrow et al. 2003 and Vargo, unpublished data). Highest chlorophyll concentrations in the bay were found during the summer, lower concentrations were found in the wintertime (Figs.3.4-3.6) and coincident with fall phytoplankton blooms in the western bay (Fig. 3.4).

Western Florida Bay Chlorophyll a and Phaeopigment

140 3.0

120 2.5

100

2.0

-2 -2 80

1.5

g Chl a m

 mg Chl a m 1.0 40

0.5 20

0 0.0 May 9 00May 10May 00 31 Sep00 11 Sep00 12 Nov00 13 Nov00 14 Jan00 24 01Mar 15 Mar01 16 May01 16 July01 12 July01 13Sep 01 20 Nov01 29 Nov01 30 01

chl a phaeo Water column chl

Benthic chlorophyll a values near Carl Ross Key in western Florida Bay were 5- 75 times greater than water column concentrations during non-bloom conditions (Fig. 3.4). Depth integrated water column concentrations were equal to benthic concentrations during blooms (8.3-27.4 mg Chl a m-2) in western Florida Bay based upon earlier measurements by Vargo, et al. (1999, 2001). Average benthic chlorophyll a in the intertidal western bay was 17.4 mg m-2 and phaeopigment was 91.6 mg m-2. A seasonal trend in benthic chlorophyll a concentration is evident in the data with highest values in the spring and summer and lowest values in the winter (Fig. 3.4). This is in contrast to water column chlorophyll a with peaks in the fall and winter coincident with the seasonal diatom bloom (Jurado and Hitchcock, 2001). Peaks in benthic phaeopigment followed the seasonal peaks in benthic chlorophyll in the western bay, being highest in the early summer and late fall - winter. Phaeopigment at the western Florida Bay also had the highest phaeopigments compared (p<0.001) to the central and southeastern sites.

Table 3.3 Ranges and average pigment values for water column and benthic microalgae in Florida Bay.

Location Water chl a Benthic Chl a Benthic Phaeo (mg m-2) (mg m-2) (mg m-2) Western Range 0.58-2.63 0.56-111.77 7.07-350 Average 1.48 17.39 91.58 Central Range 2.26-42.9 10.2-77.71 Average 9.17 28.32 Southeastern Range 3.39-35 27.12-153 Average 17.43 61.07

61 Central Florida Bay Chlorophyll and Phaeopigment

20 Chl a mg m-2 10

0 May 15 01 Jun 13 01 July 16 01 Jan 17 02 April 4 02 May 3 02 May 4 02 Aug 2 02 Nov 17 02

phaeo chl a

Central Florida Bay chlorophyll a and phaeopigment values appeared more stable seasonally (Fig. 3.5), however, there were fewer sampling dates over a shorter time period at this location than western Florida Bay. There was much less variability between sampling dates, however, the lowest values were still found in the fall and winter and the highest in the summer. Average benthic chlorophyll a in the central bay was 9.17 mg m-2 and phaeopigment was 28.3 mg m-2. Highest phaeopigments in the central bay were found during and after summer 2001 bloom in the water column, otherwise benthic phaeopigment and chlorophyll a at this location remained fairly constant. A bloom did not occur at this location the following summer. Average phaeopiment was the lowest of the three sites and significantly different from the other two locations (p<0.001).

62 Southeastern Florida Bay Chlorophyll and Phaeopigment

100 90 80 70 60 50 40

Chl a mg m-2 30 20 10 0 July 11 01 May 2 02 May 10 02

phaeo chl a

Southeastern Florida Bay was only sampled on three occasions for benthic chlorophyll a and always in the spring/summer (Fig. 3.6). Average benthic chlorophyll a in the southeastern bay was 17.4 mg m-2 and phaeopigment was 61.1 mg m-2. Samples taken in the central bay during the same time period were comparable in chlorophyll a magnitude to those from the southeast bay. Phaeopigment in the southeast was significantly different from the other two locations (p<0.001), but limited sampling prevented an analysis of seasonal differences in phaeopigment at this location. Statistical analysis All chlorophyll a measurements were normally distributed and expressed homogeneous variances, therefore parametric statistics were used to analyze differences between the locations and treatments. Using a one way ANOVA, there was no statistical

63 difference in chlorophyll a or phaeopigment between light and dark chambers at all locations. However, all the study sites were significantly different from each other (Table 3.4). The synoptic sampling events in July 2001 and May 2002 support this finding. Similar results were found when only the July 2001 and May 2002 samples from the southeastern and central stations were compared using a one way ANOVA.

Table 3.4 ANOVA levels of significance for comparison of chlorophyll a and phaeopigment among locations.

Location Benthic Chl a Benthic Phaeo

West v Central <0.001 <0.001

West v Southeast <0.01 <0.001

Central v Southeast <0.01 <0.001

Discussion The average measurements of chlorophyll a concentration from my study in Florida Bay, while comparable to other measurements from Florida, were less than those reported by others working in both temperate and tropical environments (Cahoon, 1999; Ulthicke and Klumpp, 1997). Possible explanations for this difference between Florida Bay and temperate waters include nutrient limitation of biomass and grazing pressure on the microphytobenthos. However, the microphytobenthic chlorophyll a concentration substantially exceeded water column chlorophyll a concentration underscoring the importance of quantifiying benthic microalgal production in areas of the bay not vegetated by seagrasses and sediment nutrient dynamics. Based upon my average areal chlorophyll a measurement and the estimation of the unvegetated portion of the bay at 773 km2, areal benthic microalgal biomass in Florida Bay would be 708.84 kg Chl a. The similarity in biomass values between light and dark chambers likely stems from the 2 cm depth of my sample cores and that they were collected from an enclosed chamber which was inserted into the sediment. This procedure likely captured the cells within a confined area despite any treatment effect on vertical or horizontal migration. 64 Although initially similar, Aller et al. (2001) found that chlorophyll a in diffusively open sediment degraded at different rates after about 10 days under oxic and anoxic conditions. This was the result of both reactivity rate and comparative pool size under anoxic conditions but decomposition was favored under oxic conditions. Re-exposure and re- oxidation of sediments by bioturbation or resuspension translates into permanent decreases in carbon preservation when compared to unidirectional burial or continuously anoxic conditions. The redox layer is very shallow in Florida Bay sediments. Sediments are easily resuspended by winds and bioturbation as well. Both these factors favor preservation of chlorophyll a and carbon buried in Florida Bay sediments. The similarity in benthic microalgal biomass among sample locations is surprising given the differences in dissolved nutrient availability and sediment type across Florida Bay. However, results were variable among chambers, which would reduce the probability for statistical significance between locations. There were no significant differences in benthic chlorophyll concentration between light and dark chambers (p>0.05). There was variability among some chambers on a sampling date and this was caused by patchiness. This patchiness has been reported by many researchers, and is not restricted to benthic microalgal communities. Variability in available nutrients within sediment can arise through biological, chemical and physical properties, including bioturbation. Redox changes not only affect nutrient availability, but also the micro- community (Aller, et al. 2001). Other physical forces, including fluid dynamics of porous sediments as a result of microtopography can affect the nutrient microhabitat (Huettel and Gust, 1992). Temporal differences in benthic microalgal biomass, particularly in western Florida Bay were evident. The pattern of higher benthic microalgal biomass in the summer, coincident with lower water column biomass is consistent with greater light transmission to the bottom as well as warmer water temperatures. Lower microalgal chlorophyll a concentration in the winter was coincident with increased water column chlorophyll a concentration, possibly as a result of diatom blooms or sediment resuspension due to the passage of fronts. This temporal pattern did not hold true at the other locations. This may have been due fewer sampling events or the fact that the other

65 two locations were subtidal compared to intertidal western Florida Bay. However, diminished benthic microalgal chlorophyll a concentration was noted in the central region during the summer of 2001 when a major water column bloom was ongoing. This suggests microphytobenthic chlorophyll a concentration may be more regulated by light availability and secondarily by temperature in Florida Bay. Summary Changes in potential areal microphytobenthic production resulting from perturbations like seagrass die off, may be offset in some areas of Florida Bay by other perturbations - pelagic phytoplankton blooms and increased turbidity diminishing light to the bottom. Temporal but not spatial differences were evident in microphytobenthic chlorophyll a concentration across the bay, despite anticipated nutrient limitation. Small scale patchiness was evident in chlorophyll a concentration measurements, as a result of microhabitats and bioturbation. Chlorophyll a concentration values in my study were similar to other measurements from Florida, but less than temperate and other tropical systems. As in other studies, the benthic microalgal chlorophyll a concentration equaled or far exceeded that of the water column, yet it remains relatively uninvestigated and not included in modeling efforts for Florida Bay. In summary, my results have shown that microphytobenthic production, as evidenced by BMA chlorophyll a concentration, would be an important component to overall primary production in Florida Bay and should be included in existing and future monitoring programs and modeling efforts.

Chapter 4. Mesocosm bioassays as a tool to investigate microalgal benthic-pelagic coupling in Florida Bay Introduction Benthic microalgae are important contributors to primary productivity in areas where the euphotic zone intercepts the bottom, with chlorophyll a standing stock often exceeding that of the water column (MacIntyre et al. 1996; Cahoon, 1992). This is particularly true in Florida Bay, a shallow subtropical lagoon located at the southern tip of peninsular Florida. Florida Bay is a large estuary, about 2200 km2, with an average depth of <2m. It is bounded to the north by the Everglades, to the east by the Florida Keys and to the west by the Gulf of Mexico (Schomer and Drew, 1982). Florida Bay has been the subject of intense scientific investigation as a result of perturbations in the late 1980‟s (Durako and Zieman, 2007; Carlson et al. 1999; Butler et al. 1995; Phlips et al. 1995) and recent efforts at Everglades restoration (Rudnick et al. 1999). The Gulf of Mexico, the Atlantic Ocean, and the Everglades are the three primary sources of water input to Florida Bay (Schomer and Drew, 1982). The Everglades have undergone extensive water redistribution projects during the last century in an effort to reduce flooding and provide drinking water to urbanized areas of south Florida, and for irrigation of the expanding agricultural interests of the area (Schomer and Drew, 1982; Smith et al. 1989). These water sources and their associated nutrient load drive primary productivity. Although poorly quantified, other important sources of nutrients to the Bay include atmospheric deposition and benthic flux from both groundwater in the Everglades region and sediment porewaters in the Bay itself (Price, 2005). Florida Bay Nutrient Limitation Among all the resources necessary for growth or maintenance of an organism, that in shortest supply is considered limiting. Nutrients limit phytoplankton and seagrass biomass and primary productivity in Florida Bay (Fourqurean et al. 1992; Fourqurean et al. 1993; Phlips et al. 1995; Phlips and Badylak, 1996; Kelble et al. 2005). Light may 67 also limit benthic primary productivity in some regions of the bay and at certain times of year when blooms or turbidity reduce transparency. However, due to the shallowness of Florida bay and typically clear waters, light is not generally limiting to primary production (Kelble et al. 2005). Many researchers have documented patterns in water column nutrient availability in the bay beginning with Fourqurean et al. (1993). Phosphorus is in very low concentrations, near the limits of detection, in a majority of the Bay. The greatest concentrations of P are found in the western Bay where Gulf of Mexico water is P enriched in comparison to the remainder of the bay. Dissolved nitrogen is available in such high concentrations in the bay that it is not considered limiting except during phytoplankton blooms. This general trend has been supported by several other researchers in large and small scale field and laboratory studies (Vargo et al. 2001 a and b; Jurado and Hitchcock, 2003; Brand et al. 2001). The relationship between dissolved nutrient supply and phytoplankton demand is neither simple, nor static. Diversity among phytoplankton includes nutrient preference, reducing competition for a limited resource. Temporal and spatial variation in nutrient limitation occurs in coastal areas where nutrified freshwater mixes with oligotrophic marine water, rather than the steady state condition of either end member. Bioassays have been used to identify conditions that limit phytoplankton growth and biomass. (D'Elia et al. 1986; Pennock and Sharp, 1994; Smith and Hitchcock, 1996; Jurado and Hitchcock, 2003).

Figure 4.1 (reproduced from Tomas et al. 1999, with permission) Depiction of nutrient limitation zones within Florida Bay based upon bioassays of natural pelagic phytoplankton populations, from Tomas et al. 1999. Zone A represents areas where nitrogen and possibly silica are limiting during most of the year. Zone B represents areas where primarily N and sometimes P were limiting. Zone C represents areas primarily P limiting with rare N limitation.

Tomas et al. (1995 & 1999) used nutrient bioassays to determine growth rate and chlorophyll biomass for natural pelagic phytoplankton populations. They described phosphate (P) limitation increasing across Florida Bay from West to East in the pelagic population and indicated that nitrogen (N) and silica (Si) can be co-limiting in western Florida Bay during the annual winter diatom bloom (Figure 4.1). Central Florida Bay varies between N and P limitation (Fourqurean et al. 1992; Fourqurean et al. 1993; Phlips

69 and Badylak, 1996; Tomas et al. 1999). However, N or P limitation was dependent upon the pelagic phytoplankton community structure and standing stock. Resources can be limiting to phytoplankton in several ways. At the most basic level, N and P are necessary components to cellular processes that regulate primary production – like nucleic acids and ATP. Species can differ in their uptake efficiency, allowing those species to more rapidly accumulate nutrients available in short supply or species may differ in their growth rate allowing faster growing species to dominate the community and utilize more resources – or both. Competition for nutrient resources is at the core of both scenarios. From a management perspective, factors affecting phytoplankton growth and biomass are most relevant, because the input of N and P can be controlled while community structure, competition and primary production cannot. Nutrient limitation of pelagic phytoplankton growth and biomass (Tomas et al. 1999) and dissolved nutrient availability (Fourqurean et al. 1993) are well documented in Florida Bay, but do not always coincide spatially. Nutrient availability does not account for remineralization and the importance of the microbial loop. Bioassays Hydrodynamic and biological models of the Bay are being fine tuned to assist managers in making distribution and water quality decisions as part of restoration efforts. Since limiting resources, e.g. nutrients in Florida Bay, are important controls on primary production, and thus higher trophic level production, it is important to quantify the spatial and temporal differences in nutrient limitation of production and biomass. Although often ignored, and certainly undersampled in comparison to pelagic phytoplankton and seagrass primary production, benthic microalgae are significant sources of primary production. Their proximity to the sediment and buried nutrients is often presumed to relieve them of the nutrient limitation that pelagic phytoplankton might suffer (Sundback and McGlathery, 2005). However, this is not necessarily true, especially in Florida Bay where the carbonate sediments tightly bind phosphate through chemical mechanisms. This is coupled with a relatively low concentration of iron bound phosphorus which would be likely to release P to the sediment porewaters during redox reactions under anoxic conditions (Nixon et al. 1980; Krom and Berner, 1981; Caraco et

70 al. 1992). But, balancing this is a shallow redox zone minimizing the thickness of the diffusive layer through which any loosely bound sedimentary P would have to travel. As a result of this capacity for release of sediment phosphorus, nitrogen limitation is most often implicated for marine waters (Ryther and Dunstan, 1971; Howarth, 1988). Lesser nitrogen fixation rates (Howarth, 1988) and greater denitrification rates (Seitzinger, 1988) compared to freshwater systems are also contributing factors to N limitation in marine waters. Silica limitation of diatoms has also been reported in Florida Bay (Tomas et al. 1999; Jurado and Hitchcock, 2001) and coastal regions (Malone et al. 1980; Officer and Ryther, 1980). Mesocosm bioassays are one tool to assess nutrient limitation of algal biomass (Elser et al. 1988; Tomas et al. 1999). Which nutrient is limiting is determined by comparing the changes in algal biomass (g chl a L -1) or growth rates after additions of the limiting nutrient alone and in combination with other nutrients over several days to control assays. These types of assays were performed at two sites in Florida Bay to examine if the nutrient(s) limiting benthic microalgal biomass differ(s) between western and central Florida Bay and if the nutrients limiting benthic microalgal biomass differ spatially and temporally from those that limit pelagic microalgal production. Materials and methods Sediment was collected from two field sampling sites; Carl Ross Key in western Florida Bay and Whipray Basin in central Florida Bay (Fig. 4.2). Sediments ~ 4cm deep were placed into the mesocosms constructed of shallow, clear plastic boxes (17.5cm wide x 15 cm tall x 32.5 cm long) containing a coiled tubing nutrient delivery system in the bottom. A single experiment was conducted in August 2002 at the central Florida Bay location and in January 2003 at both locations. After filling at the field site, the mesocosms (Fig. 4.3) were transported to the Keys Marine Laboratory in Layton, FL where they were placed into a water table that circulated ambient seawater to maintain temperature for the duration of the experiment. The mesocosms were exposed to ambient light, but were protected from evaporation and rainfall by a thin layer of plastic food wrap over the top of the mesocosm which was secured by a thin plastic rim. Water overlying the sediment in each mesocosm was

71 removed and replaced with 2 liters of 0.45  filtered water collected from the field site. A gentle circulation device (miniature water pump) was placed in each mesocosm to continuously recirculate the water column above the sediments. Once these preparations were complete, the benthic microalgal community within the mesocosm was allowed to equilibrate for ~48 hours prior to commencing the experiment.

72 = Recirculating pump

Seawater In

+N +P +NP Control Control +NP

+P +NP +N +P +N Control

Seawater Out

Continuous Flow Ambient Seawater Bath

Figure 4.3. The mesocosms arranged in the randomized block design within the lagoonal flow-through seawater system at the field laboratory in Layton, FL.

Dissolved nutrient availability in the water column was measured as the change in concentration over time. Nutrients were only added to the sediment so increases in dissolved nutrients in the water column had to come from benthic flux. Three nutrient treatments (N alone, P alone, and NP) and controls (no nutrient additions) were tested. Triplicate mesocosms were sampled for each treatment and control utilizing a randomized block experimental design to assign treatments to the mesocosms in the ambient seawater bath enclosure. Inorganic forms of N (NH4Cl) and P (KH2PO4) were added to the sediment daily through tubes buried in the sediment, followed by a flush of ambient water from the mesocosm to ensure complete delivery of the nutrient pulse to the -1 -1 sediment. The daily concentrations added were 0.667 M L PO4, 5.33 M L NH4. Control mesocosms had the nutrient delivery tubing in the bottom but received only

73 flushes of ambient water from the mesocosms. The nutrient concentrations in the water column were sampled immediately before and after the nutrient/ambient water additions. Samples were collected at approximately the same time each morning. The methods of Solórzano and Sharp (1980) for total dissolved phosphate (TDP) and soluble reactive phosphate (SRP), Valderama (1981) for total dissolved nitrogen (TDN) and Gordon et al. (1993) for inorganic nutrients were used. Dissolved organic nitrogen (DON) and dissolved organic phosphate (DOP) were determined by subtraction of the inorganic forms I measured from total values. Initial salinities were maintained by the addition of MilliQ® water. Temperature and salinity were recorded daily. Three benthic chlorophyll cores (top ~1 cm of sediment) and one or two water column chlorophyll samples were taken daily immediately prior to the nutrient addition. Experiments were maintained for at least six days plus the initial equilibration period. Changes in chlorophyll a, photochemical Yield (PAM fluorometer) and dissolved nutrients as a function of nutrient enrichment in the treatment mesocosms were measured against the control mesocosms as indications of limiting nutrients. The summer 2002 experiment was limited to the central Florida Bay only, due to logistics. A rain event caused contamination in the mesocosms on the evening of Day 6, dropping salinity by ~50% and potentially providing an unquantified input of atmospheric N and P in the form of rainwater. This experiment was nine days in duration. No nutrient additions were made after the rainfall contamination, although measurements continued. The winter 2003 experiments were conducted at locations in western and central Florida Bay. The experiments were conducted during the passage of a cold front, which persisted from Day 2 through Day 6. These experiments were seven days in duration. Nutrient additions and dissolved nutrient measurements were made only for the first seven days in the summer experiment and the first two days and the final day in the winter experiment. The benthic microalgal biomass was measured for nine days in the summer experiments and seven days in the winter experiment, to ensure sufficient time elapsed for a change in biomass to occur.

74 Results Nutrients: Initial Field Conditions Conditions at the time when the field collection of sediment for the mesocosm experiments occurred will be discussed first and are summarized in Table 1. Central Florida Bay showed some overlap in the initial TDP concentration of the water column at the collection site for both mesocosm experiments. TDP was greater in the western than central portion of the bay, in keeping with the in situ chamber findings outlined elsewhere in this dissertation. Water column SRP at the collection site was initially near the limits of detection in central Florida Bay mesocosms during the winter experiments, but was an order of magnitude greater in the summer at the same location. Water column SRP at the collection site was initially greater in western Florida Bay compared to central Florida Bay in the winter experiment, mirroring the TDP findings.

Table 4.1. Initial nutrient field concentration in M L-1 with the predicted limiting nutrient determined from the Redfield Ratio of the listed nutrient species of available dissolved nutrients. Summer Winter Winter Central Central Western TDP(M) 0.8 0.47 0.76 TDN(M) 69.77 58.29 22.87 SRP(M) 0.12 0.009 0.027

NH4(M) 5.51 6.82 2.22

NO3+NO2(M) 0.03 0.19 0

NO2(M) 0.00 0.09 0.00 SIL(M) 57.67 9.44 14.59

Predicted Limiting Nutrient TDN:TDP P limited P limited P limited

NH4:SRP P limited P limited P limited NO3+NO2:SRP N limited P limited N limited

At both locations, dissolved silica was much greater (~5 times) initially in summer, but both had similar ranges in winter. During the winter, nitrite was initially at

75 the limit of detection in western Florida Bay. Average initial nitrite was higher in central Florida Bay in January than in August. Nitrate+nitrite was also at the limits of detection in western Florida Bay in January and central Florida Bay in August. Nitrate+Nitrite was at its greatest concentration in January in central Florida Bay and about ½ of this was due to nitrite. Mesocosm results The results of the multi-day mesocosm experiments were mixed. In central Florida Bay TDP concentrations were variable throughout the experiment, with few significant differences between treatments and control (Figs. 4.4 and 4.5). Despite the P additions, TDP did not accumulate in the water column during the two experiments (n=6). The increase in TDP concentration on August 7th could be the result of higher concentration nutrient additions that were accidently applied to some unidentified mesocosms. However in western Florida Bay mesocosms, TDP did accumulate in the water column in the P and the NP treatment additions (Fig. 4.6). This difference was significant when compared to the N and control treatments by Dday 2 of the experiment. TDP in the P only treatment differed significantly from the NP addition and this also occurred by Day 2 of the experiment. Water column TDP concentration increased throughout the course of the experiment in the P addition, but after an initial increase instead remained level after Day 2 in the NP treatment. SRP was variable during both mesocosm experiments using sediment from central Florida Bay (Figs. 4.7 and 4.8). The winter experiment did not reveal any statistically significant differences between treatments and controls except for the first two days of the experiment. In the summer, there were significant differences between P addition treatments for the first, third and fourth days of the experiment. SRP always decreased in the water column throughout the duration of the experiment in central Florida Bay, regardless of treatment, however, this trend was not statistically significant. Presumabley this loss was to the BMA community. The increase in SRP concentration on August 7th could be the result of higher concentration nutrient additions that were accidently applied to some unidentified mesocosms.

76 In western Florida Bay, SRP did leave the sediment and accumulate in the water column in the P and the NP treatment additions (Fig. 4.9). This difference was significant in these treatments versus the N only and the control by Day 2 of the experiment. The P and NP additions differed significantly from each other and N and control by Day 2 of the experiment. As seen with TDP, accumulation of SRP in the overlying water column did not increase throughout the experiment in the NP treatment, but instead remained level or slightly decreased after Day 2. Central Florida Bay mesocosms were significantly different by treatment from each other on Days 1 and 2 of the experiment only in January and Days 1, 3 and 4 in the August experiment.

January 2003 Central Florida Bay TDP 1.4 C N 1.2 NP P

1.0 1 1

- 0.8 M L M  0.6

0.4

0.2

0.0 1/25/2003 1/26/2003 1/27/2003 1/28/2003 1/29/2003 Figure 4.4. Average TDP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) during winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

77 August 2002 Central Florida Bay TDP 2.0

1.8

1.6

1.4

1 C -

1.2 N M L M

 1.0 NP P 0.8

0.6

0.4

0.2

0.0 8/2/2002 8/3/2002 8/4/2002 8/5/2002 8/6/2002 8/7/2002 8/8/2002

Figure 4.5. Average TDP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) during summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

78 January 2003 Western Florida Bay TDP 5.0

4.5

4.0

3.5

C 1 - 3.0 N NP M L M 2.5

 P 2.0

1.5

1.0

0.5

0.0 1/25/2003 1/26/2003 1/27/2003 1/28/2003 1/29/2003

Figure 4.6. Average TDP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

79 January 2003 Central Florida Bay Mesocosms SRP 0.016 C 0.014 N NP 0.012 P

0.010

1 -

0.008

M L M  0.006

0.004

0.002

0.000 1/25/2003 1/26/2003 1/27/2003 1/28/2003 1/29/2003 1/30/2003

Figure 4.7. Average SRP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

80 August 2002 Central Florida Bay Mesocosms SRP 0.40

0.35 C 0.30

0.25 N 1 - 0.20 NP

M L M 0.15 P  0.10

0.05 0.00 8/4/2002 8/5/2002 8/6/2002 8/7/2002 8/8/2002 8/9/2002 8/10/2002

Figure 4.8. Average SRP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

81 January 2003 Western Florida Bay Mesocosms SRP 0.40

0.35

0.30 C

1 0.25 - N

M L M 0.20 NP  P 0.15

0.10

0.05

0.00 1/25/2003 1/26/2003 1/27/2003 1/28/2003 1/29/2003 1/30/2003

Figure 4.9. Average SRP results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

TDN was variable in the three mesocosm experiments (Figs. 4.10-4.12). TDN concentration increased over time in all three experiments for the N addition treatments. The NP treatment showed similar results for both locations during the winter experiment (Figs. 4.11-12). During the summer experiment TDN concentration in the water column was balanced by uptake in the phytoplankton, either benthic or pelagic, by Day 4, and actually resulted in loss of TDN from the water column by Day 7 (Fig. 4.10). The control and P only treatments did not show an increase in TDN except in January 2003 in the central Florida Bay when all mesocosms had greater concentrations by the end of the experiment (Fig. 4.12). The increase in TDN concentration on August 7th could be the

82 result of higher concentration nutrient additions that were accidently applied to some unidentified mesocosms. During the summer experiment, statistically significant differences in water column TDN concentrations were found by Day 3 in the N, and NP, addition treatments versus P only, and controls. Significant differences were noted between N and NP addition treatments versus P only and controls at the conclusion of the winter central Florida Bay experiment, but were not evident during the first two days of the experiment. Due to high variability between samples within treatments, western Florida Bay mesocosm treatments did not exhibit statistically significant differences, but followed the same trend with higher TDN concentrations in the overlying water column in the N and NP addition treatments, respectively. At the conclusion of the experiment, TDN was mostly organic forms of nitrogen and possibly ammonium. Because ammonium values were offscale (at the upper limit of detection), I was unable to determine the proportion of ammonium present in the TDN.

August 2002 Central Florida Bay Mesocosms

300 Total Dissolved Nitrogen

250

200 C

N 1 - 150 NP

P M L M  100

0 080202 080402 080502 080602 080702 080802

Figure 4.10. Average TDN results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

84 January 2003 Western Florida Bay Mesocosms Total Dissolved Nitrogen 900

800

700

600 C N

500 1

- NP

400 P M L M  300

200

100

0 012403 012503 013003

Figure 4.11. Average TDN results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

85 January 2003 Central Florida Bay Mesocosms

600 Total Dissolved Nitrogen

500

400 C

N 1 - 300 NP

P M L M

 200

100

0 012403 012503 013003

Figure 4.12. Average TDN results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

Central Florida Bay mesocosms in the summer all showed a dramatic drop in dissolved silica concentration in the water column between Days 1 and 4 in all treatments (Fig. 4.13). By Day 5 the dissolved silica concentration decreases in all but the NP treatments, followed by another drop on day 6 and another dramatic increase on Day 7. There was high variability among replicates pointing to a potential error in the laboratory analysis. The high values for NP on Days 5 and 7 were replicated among samples, but they were all analyzed on the same day, which still doesn‟t preclude laboratory error. In the winter experiments, dissolved silica increased by the second day of the experiment and on the final day remained the same or decreased again to levels similar to Day 1 (Figs. 4.14-15). Again, P treatments had the lowest average dissolved silica value of

86 either treatment or controls. High variability in the second sample again indicates a potential error in the laboratory analysis, among replicates analyzed on different days.

August 2002 Central Florida Bay Mesocosms Silica 120

100

80 1 - 60 C

M L M N  40 NP P

0 080202 080402 080502 080602 080702 080802

Figure 4.13. Average silica results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

87 January 2003 Western Florida Bay Mesocosms Silica 250

200 C N NP 150

1 -

M L M 100 

0 012403 012503 013003

Figure 4.14. Average silica results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

88 January 2003 Central Florida Bay Mesocosms Silica 90

70 C 60 N

50 NP

1 - P

M L M  30

0 012403 012503 013003

Figure 4.15. Average silica results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

Dissolved nitrite in central Florida bay mesocosms during summer showed significant differences among treatments by Day 2 of the experiment (Fig. 4.16). In contrast to the silica concentration pattern for Days 4-6, the nitrite peaked on Day 5, dropping back to levels similar to Days 2-5 on the final day for the N treatments, and to below the limits of detection in the NP treatments. As expected N, followed by NP treatments were always greater in dissolved nitrite than controls and P treatments, except the NP treatments on Day 6, as mentioned previously. The increase in nitrite concentration on August 7th could be the result of higher concentration nutrient additions that were accidently applied to some unidentified mesocosms.

89 During the winter, nitrite was initially below the limit of detection in western Florida Bay (Fig. 4.17). By day 2 of the mesocosm experiment it was measureable, but still very low and there were not significant differences between treatments. This same result was found on the last day of the experiment in western Florida Bay mesocosms. Central Florida Bay mesocosms showed a drop in N, NP and controls on day two, while P remained near the values for day 1 (Fig. 4.18). None of these differences was significant however. By the last day of the experiment, nitrite levels in the N, NP and controls were similar in concentration to the start of the experiment, but again not significantly different from each other.

August 2002 Central Florida Bay Mesocosms Nitrite 2.0

1.8 C 1.6 N 1.4 NP P

1.2

1 -

1.0 M L M

 0.8

0.6

0.4

0.2

0.0 080202 080402 080502 080602 080702 080802

Figure 4.16. Average nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

January 2003 Western Florida Bay Mesocosms 0.25 Nitrite

0.20

0.15 N 1

- NP P

M L M 0.10 

0.05

0.00 012403 012503 013003

Figure 4.17. Average nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

91 January 2003 Central Florida Bay Mesocosms

0.40 Nitrite

0.35

0.30

C 1 - 0.25 N

M L M NP  0.20 P 0.15

0.10

0.05

0.00 012403 012503 013003

Figure 4.18. Average nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

Nitrate+nitrite generally mirrored the nitrite patterns (Figs. 4.19-4.21), with one exception. In winter in western Florida Bay, significant differences were found between treatments and controls by the end of the experiment (Fig. 4.20). Nitrate+nitrite values were an order of magnitude greater than nitrite alone in western Florida, but only 2-3 times as great in central Florida Bay, indicating there was little available nitrate at this location. The increase in Nitrate+nitrite concentration on August 7th could be the result of higher concentration nutrient additions that were accidently applied to some unidentified mesocosms.

92 August 2002 Central Florida Bay Mesocosms Nitrate+Nitrite 14

12 C 10 N

8 NP

1 - P

M L M  4

0 080202 080402 080502 080602 080702 080802

Figure 4.19. Average nitrate plus nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

93 January 2003 Western Florida Bay Mesocosms Nitrate+Nitrite 3.0

2.5

C 2.0 N

1 NP - 1.5

M L M  1.0

0.5

0.0 012403 012503 013003

Figure 4.20. Average nitrate plus nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

94 January 2003 Central Florida Bay Mesocosms Nitrate+Nitrite 2.0

1.8

1.6 C 1.4 N

1.2 NP 1 - 1.0 P

M L M 0.8  0.6

0.4

0.2

0.0 012403 012503 013003

Figure 4.21. Average nitrate plus nitrite results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

Central Florida Bay mesocosms showed clear differences in ammonium concentration between both N and NP treatments and P only treatments as well as controls in the summer (Fig. 4.22) but not the winter experiments (Fig. 4.23). Ammonium was near the upper limit of detection in the summer by the third day for the N and NP treatments in the summer experiment. In the winter at both locations (Fig. 4.23 and 4.24), ammonium was at the upper limit of detection by the second day of the experiment for all treatments and controls. Because of this, differences between treatments were not able to be determined during the winter.

95 August 2002 Central Florida Bay Mesocosms Ammonium 35

25 Control

20 N 1 - NP P

M L M 15 

0 080202 080402 080502 080602 080702 080802

Figure 4.22. Average ammonium results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in summer. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

96 January 2003 Central Florida Bay Mesocosms Ammonium 35

25 C

20 N 1 - NP

15 P

M L M 

0 012403 012503 013003

Figure 4.23. Average ammonium results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from central Florida Bay (End Key, Whipray Basin) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

97 January 2003 Western Florida Bay Mesocosms Ammonium 40

30 1

- 25 C N

M L M 20 NP  P 15

0 012403 012503 013003

Figure 4.24. Average ammonium results from replicate samples of triplicate nutrient addition treatments (N, P, N and P, and Control) in mesocosms using sediment from western Florida Bay (Carl Ross Key) in winter. The sample dates appear on the x axis. The error bars represent the standard error, and are only shown in one direction to facilitate viewing.

Chlorophyll concentration A positive response to nutrient additions was seen in both the water column and the sediment microalgal communities in central Florida Bay in the summer mesocosm experiment. Both responded positively to the combined NP addition with an increase in chlorophyll a standing stock. However, the communities responded differently when either N or P was added. Based on chlorophyll a standing stock, the central Florida Bay water column community was initially P limited and then by August 6th was N limited, while P additions still elicited a significant difference from controls (Fig. 4.25). This condition persisted throughout the remainder of the experiment even after a rainstorm dramatically dropped the salinity in the mesocosms and introduced an unknown amount of N and P into all the mesocosms. 98 On August 5th, N addition treatments in the sediment community were significantly lower in chlorophyll a standing stock than control treatments (Fig.4.26). P additions lead to greater chlorophyll a standing stock, but the results were not significantly different from controls (Fig. 4.26). On August 7th both the N and P treatments remained significantly lower in concentrations than controls. The NP treatments were not significantly different than controls on the 7th. By August 8th, NP and P additions, respectively, were now greater in chlorophyll a concentration than either controls or N treatments. Control and N treatments were now not different from each other. These differences noted on the 8th were significant throughout the remainder of the experiment. There was no apparent biomass response in the benthos to the rainfall event that lowered salinity by ~50%, but there could have been unmeasured changes in community composition. The increase in chlorophyll a concentration could be the result of higher concentration nutrient additions that were accidently applied to some unidentified mesocosms on the 7th.

August 2002 Central Bay Mesocosm

8 Water Column Chl a

7 C N 6

1 NP - 5

g l g P 

a 4

Chl 3

Figure 4.25. Average water column chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during summer 2002. The sample dates appear on the x axis. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily.

100 August 2002 Central Florida Bay Mesocosm

30 Sediment Chl a

2 C - 20 N NP mg m mg 15 a P

Chl 10

Figure 4.26. Average sediment chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during summer 2002. The sample dates appear on the x axis. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily.

Like the summer, the central Florida Bay water column community initially responded to NP additions (Fig. 4.27) in winter by increasing chlorophyll a standing stock. N alone reduced chlorophyll a standing stock, in contrast to the summer experiment at the same location. By January 25th the water column chlorophyll a standing stock response to NP and P additions were significantly different from each other and also from N additions and controls. By January 28th water column chlorophyll a concentration decreased dramatically in the NP treatments and was not significantly different from controls or N addition treatments. Phosphorus additions on January 26th were significantly different from all other treatments and the control, suggesting a shift to P limitation of the water column community. Phosphorus limitation persisted from January 28th through the 30th. On January 28th NP and N treatments had reduced 101 chlorophyll a standing stocks that were significantly different from both controls and each other. The NP treatments rebounded on the final day of the experiment to again be greater in chlorophyll a concentration than controls. All treatments and controls were significantly different from each other on the final day of the experiment.

January 2003 Central Florida Bay Mesocosms Water Column Chl a 1.4

1.2

1.0

1 0.8

- C g l g

 N

a 0.6

NP Chl 0.4 P

0.2

0.0

Figure 4.27. Average water column chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. The sample dates appear on the x axis. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily.

102 January 2003 Central Florida Bay Mesocosms Sediment Chl a 25

2 20 -

mg m mg 15 a

C Chl Chl 10 N NP P 5

Figure 4.28. Average sediment chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. The sample dates appear on the x axis. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily.

During the winter, sediment chlorophyll a standing stock response differed from water column results (Fig. 4.28). The sediment community from central Florida Bay was P limited at least during the latter part of the experiment. P addition treatments showed the greatest increase in benthic microalgal chlorophyll a concentrations, followed by N and NP which were not significantly different from each other except on the last three days of the experiment. However, N additions did not reduce chlorophyll a standing stock as it did in the summer at this location. On January 28th N addition treatments showed the greatest average chlorophyll a concentration, but this was not significantly different from P treatments. On January 29th, P addition treatments were the greatest average chlorophyll a concentration, while the N only treatment lost a significant amount

103 of chlorophyll a and was actually below control treatments for biomass. On January 30th the P treatment had now dropped dramatically while the N only treatment had rebounded. The NP treatment, which during the summer at this location had the greatest average chlorophyll a concentration, was greater than control treatments, but not as high in concentration as the P alone treatments. The NP treatments did not differ significantly from N treatments for January 23-27th, but did show an increase in chlorophyll a standing stock on the final day of the experiment, although not significantly different from either control or N alone.

January 2003 Western Florida Bay Mesocosms Sediment Chl a 25

2 - 15

mg m mg C a 10 N

NP Chl P 5

Figure 4.29. Average sediment chlorophyll a (g L-1) in western Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. The sample dates appear on the x axis. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily.

104 In the winter in western Florida Bay, the mesocosm experiment indicated the water column and sediment were more similar to each other in response to nutrient additions, but different than the central Florida Bay. In both water column and sediment, nutrient additions resulted in decreased chlorophyll a concentrations throughout the experiment and these differences were significant by January 25th (Figs. 4.29 and 4.30). Initially in western Florida Bay, control treatments had the greatest sediment chlorophyll a concentration followed by N> NP > P additions (Fig. 4.29). By January 28th, the P and NP treatments were not statistically different from each other, but did differ from the P and Control. On the final day the order of sediment chlorophyll a concentration from greatest to least was Control > N> P>NP – all were significantly different from each other. Since all the treatments and controls declined in chlorophyll a concentration, some factor other than nutrients must have also been limiting, such as light or temperature.

105

January 2003 Western Florida Bay Mesocosms Water Column Chl a 4.0

3.5

3.0

2.5 1

- C 2.0

g l g N 

a NP 1.5 P Chl Chl 1.0

0.5

0.0

Figure 4.30. Average water column chlorophyll a (g L-1) in central Florida Bay mesocosms by nutrient addition treatment (control, N, P, and NP) during winter 2003. The sample dates appear on the x axis. Values are the average of three replicates of treatment mesocosms in a randomized block design sampled daily.

Similarly, in the water column in western Florida Bay, all nutrient addition treatments showed a loss of chlorophyll a (Fig. 4.30). On January 25th water column chlorophyll a concentration was greatest in controls followed by N> NP > P. All treatment additions were significantly different from controls, but not from each other on January 25th and 26th. January 27th and 28th, none of the treatments or controls were significantly different from each other. By January 29th, controls again had the greatest water column chlorophyll a concentration. The NP treatments were slightly lower than controls> P> N addition treatments. Only the control and the N treatments were significantly different from each other on January 29th. On the final day of the experiment phytoplankton biomass increased significantly in the water column, again N 106 was less than control in chlorophyll a concentration, followed by NP, P and Controls. Control was significantly different from N and NP/P treatments, and N was significantly different from NP/P treatments too on the final day. The NP and P treatments were not significantly different from each other on January 30th. A summary table of the significant differences is presented in Appendix A. Discussion I compared two methods of assessing phytoplankton response to nutrient limitation in these experiments: chlorophyll a standing stock and water column dissolved nutrient concentration. The latter has been commonly measured temporally and spatially in the bay since the early 1990‟s assuming that the dissolved nutrient available in least supply in the water column must then be limiting phytoplankton biomass and that nutrient found in high concentrations must not be limiting phytoplankton growth. Benthic chlorophyll a standing stock is not commonly measured temporally and spatially, and sediment bioassays are not a part of the existing Florida Bay monitoring program. Increase in chlorophyll a concentration over time in response to nutrient addition indicates phytoplankton growth in response to nutrient additions, suggesting the absence of that nutrient was what limited growth. Numerous investigators have highlighted the disparity between dissolved nutrient availability and nutrient limitation of biomass. Another hindrance to its use is that dissolved nutrient availability does not indicate benthic or pelagic community uptake or some combination thereof. Nutrient concentrations Central Florida Bay, during both summer and winter experiments, was initially very low in all species of dissolved phosphate (TDP and SRP) suggesting P could be limiting to phytoplankton and benthic microalgae. Daily samples were variable, even within treatments. However, no significant differences in the concentration of any forms of P were found between nutrient addition treatments or controls in either of these experiments even over time. SRP exhibited a decreased over time in all nutrient addition treatments and controls for both experiments by the second or third day. This evidence supports P uptake by the sediment or benthic microalgae, but I could not distinguish them in this experimental design.

107 Dissolved nitrogen exceeded the Redfield Ratio compared to dissolved phosphate. Differences in dissolved TDN concentration in mesocosms hint at colimitation of N and P temporally. Nitrate+nitrite and nitrite concentrations suggest differential uptake of nitrogen species in response to nutrient addition. Ammonium availability also supports differential uptake of nitrogen but only for the summer experiment in central Florida Bay. I cannot determine this during the winter since ammonium measurements were at or above the upper limit of detection. Based upon dissolved nutrient availability in central Florida Bay, P would be the limiting nutrient for benthic and pelagic phytoplankton. Colimitation by N would be possible if sufficient P is regenerated or newly introduced to the system. Western Florida Bay was neither N, nor P, limited based upon dissolved nutrient availability in one winter mesocosm experiment. Mesocosms were initially greater in dissolved phosphate concentration and showed clear differences in TDP and SRP concentration between both P addition treatments and N only additions and controls. The magnitude of these differences increased throughout the course of the experiment. This is probably a combination of thermally depressed phytoplankton production limiting uptake of available P, allowing diffusion of the P addition from the sediment to the water column, but also suggests P sufficiency and N limitation. TDN:TDP again exceeded the Redfield Ratio in the winter experiment. Ammonium concentration (at least 31 M) was offscale at the upper limit of detection for all treatments. As in central Florida bay, nitrite was near the limits of detection and much less than nitrate concentration, but spatial differences between central and western Florida Bay were not apparent from the experiment. Chlorophyll concentration When change in chlorophyll a concentration was used to assess response to nutrient availability, temporal and spatial differences were found between pelagic and benthic communities. Summer mesocosm experiments revealed phytoplankton biomass (chlorophyll a) in the water column increased more quickly in response to sediment nutrient additions than benthic microalgal biomass (chlorophyll a). However, by the termination of the experiment, a similar trend in biomass was apparent in the benthic

108 microalgal community. Winter trends were not as clear, perhaps due to temperature or light limiting biomass more than nutrients.

August 2002 Central Florida Bay Mesocosm Temperature and Salinity 36 30

35 25

34 20

C C 15 

PSU Temp 32 Sal 10 31

5 30

29 0

Figure 4.31. Temperature (left y-axis) and salinity (right y-axis) in August 2002 for the central Florida Bay mesocosm experiment.

109

January 2003 Western Florida Bay Temperature and Salinity 20 45

18 40

16 35 14 30 12 25

C C 10  20 PSU Temp 8 Sal 15 6

4 10

2 5

0 0

Figure 4.32. Temperature (left y-axis) and salinity (right y-axis) in January 2003 for the western Florida Bay mesocosm experiment.

110 January 2003 Central Florida Bay Mesocosm Temperature and Salinity 20 38.5

18 38

16 37.5

14 37

12 36.5

C C 10 36  PSU Temp 8 35.5 Sal 6 35

4 34.5

2 34

0 33.5

Figure 4.33. Temperature (left y-axis) and salinity (right y-axis) in January 2003 for the central Florida Bay mesocosm experiment.

Chlorophyll a concentration was more controlled by temperature during the winter experiments than nutrient limited, as evidenced by the lack of significant differences between treatments and controls and what seems to be a release of temperature limitation on the final day of the experiment (Fig. 4.32 and 4.33) . Chlorophyll a changes in both the water column and sediment communities were negligible for the duration of the winter experiments in both locations, despite nutrient additions. The water column was always quicker to respond. Only on the final day of the experiment did growth increase appreciably (doubling), but there was no advantage to communities with nutrient additions. A longer experimental duration may have elucidated nutrient limitation at this location in the winter, particularly after passage of the cold front.

111 Based on chlorophyll a concentration the mesocosm bioassays during both summer 2002 and winter 2003 indicate nutrient limitation in benthic microalgal communities from central Florida Bay. In central Florida Bay, the limiting nutrient for the benthic microalgal community varied from NP and P limitation in the summer to P and shorter term N limitation in the winter. The water column in central Florida Bay was also nutrient limited in both experiments but N limited in central Florida Bay in the summer and P limited in the winter. Ultimately N and P were temporally colimiting, on the order of days, to the water column population in central Florida Bay. In contrast to central Florida Bay, nutrient limitation was not demonstrated in the sediment community in western Florida Bay during the winter. The water column community, like that of the sediment, did not show a significant response to nutrient additions versus controls for more than a day during the experiment. Those differences attributable to treatment effect were negative biomass responses rather than positive ones. Summary It is common to measure dissolved nutrient availability in the Bay and uncommon to conduct water column bioassays. Sediment bioassays in the Bay are very rare. Using dissolved nutrient availability as a predictor of phytoplankton nutrient limitation would have only been accurate for the sediment community in central Florida Bay, but not the water column community. Western Florida Bay was much more complicated, however the loss of chlorophyll a in both water column and sediment phytoplankton community in response to nutrient addition would not have been predicted based solely on dissolved nutrient availability. Two bioassays would statistically be considered coincidence and more experiments are necessary, however my findings suggest it would be perilous to assume the pelagic and benthic communities respond in unison to perceived nutrient limitation based upon dissolved nutrient availability. Colimiting nutrients and ephemeral short-term community response to nutrient limitation would not be evident from dissolved nutrient availability or from infrequent bioassays whether measuring the water column or sediment community. Tomas et al. (1999) highlighted the disparity between nutrient availability and nutrient limitation in natural Florida bay water column phytoplankton populations. They

112 point out that dissolved nutrient availability arises dually from nutrient uptake/ assimilation and autochthonous/allochthonous supply. Benthic-pelagic coupling of these dual processes, combined with phytoplankton community composition and response rates further complicate relationships between nutrients and chlorophyll a standing stock. Bioassays are a much better method than measuring dissolved nutrient availability to gauge phytoplankton response to nutrient additions. Experimental bioassay findings of nutrient colimitation in both the sediment and water column in this study, and in the water column in other studies (Tomas et. 1999; Jurado and Hitchcock, 2001; Vargo et al. 2001a), do not consistently support temporal and spatial trends in nutrient supply and availability in Florida Bay.

113

Chapter 5. Short term response of benthic microalgae to 33P additions to the sediment. Introduction Benthic microalgae (BMA) contribute to ecosystem primary productivity in regions where sufficient sunlight reaches the benthos (MacIntyre et al. 1996; Cahoon, 1992). In Florida, these areas include the West Florida Shelf, and those portions of coastal embayments, such as Florida Bay, that are not light limited (Kelble, et al. 2005). Florida Bay is a large triangular shaped embayment located at the southern tip of peninsular Florida. Part of Everglades National Park, it is a negative estuary approximately 2200 km2 in size (Schomer and Drew, 1982; Fourqurean et al. 1992). It is bounded to the east and south by the Florida Keys and the reef tract, and to the west by the Gulf of Mexico. To the north are the Everglades from which it receives freshwater inflow, mainly through Taylor Slough and the C111 canal system (Schomer and Drew 1982; Rudnick et al. 1999). Florida Bay has been intensively investigated since the late 1980‟s due to drastic ecosystem changes like seagrass and sponge dieoffs, and persistent phytoplankton blooms (Carlson et al. 1999; Butler et al. 1995; Phlips et al. 1995). This combined with Everglades restoration efforts, has spurred interest in the development of a comprehensive ecosystem model (Madden et al. 2005, 2007). Nutrient budgets are essential to modeling primary productivity and must include benthic flux (Madden et al. 2005). Phosphate (P) is most often implicated as the nutrient limiting primary production in Florida Bay (Fourqurean et al. 1992; Tomas et al. 1999). Biology, geology, chemistry and physics can affect benthic P flux in Florida Bay. In situ experiments provide net nutrient flux from the benthic community (Yarbro and Carlson, 2008) and problems related to geology and chemistry can be avoided. Multiple investigators have shown that biology, specifically the microbial web within BMA vegetated sediment, is important to benthic nutrient flux (Admiraal, 1977; Nilsson and

114 Sundback, 1993; Sundback and McGlathery, 2005). However it remains difficult to measure specific components of the benthic community in situ and determine what nutrients they may be using (Sundback and McGlathery, 2005). Thus controlled laboratory experiments using intact sediment cores can be useful to examine the microbial-web related benthic nutrient flux (Nilsson and Sundback, 1993; Cornwell et al. 2001; Sundback and McGlathery, 2005). BMA coexist within seagrass beds, on the sediment and some species are epiphytic on seagrass blades (MacIntyre et al. 1996; Cahoon, 1992). In competition for nutrient resources with seagrasses, BMA are dwarfed in the volume of phosphate uptake and mobilization (Nielsen et al. 2005). However, due to the variety of components in the BMA community, they may be capable of utilizing phosphate both from porewaters and loosely bound to the sediments within and below the redox zone (Sundback and McGlathery, 2005). Seagrasses and the symbiotic cyanobacteria Klebsiella sp., which is present in some seagrass species, can alter the oxic conditions of the sediment surrounding the seagrass roots and rhizomes (Fourqurean and Zieman, 1991; Kirchstein et al. 1993; Nielsen et al. 2005). Sediments can release P based upon the redox condition of the sediment (Sundback and McGlathery, 2005). Anoxic conditions can mobilize iron bound P and that loosely adsorbed onto sediment grains, particularly clays, into the porewaters. Phosphorus can be sorbed onto carbonate sediments commonly found in Florida Bay (Millero, 2000). Benthic microalgae are an important food source to higher trophic levels particularly juvenile fish and epifaunal gastropods and infaunal filter feeding bivalves that can regenerate P into the water or sediment (Fenchel and Kofoed, 1976). This community is comprised of free-living and attached organisms (MacIntyre et al. 1996), some which secrete a mucilaginous matrix that stabilize sediments against erosion (Admiraal, 1977; Huettel and Gust, 1992) and inhibit diffusion of nutrients in porewaters (Nilsson and Sundback, 1991; Sundback and McGlathery, 2005). Uptake and release of nutrients by BMA has been demonstrated both to and from the sediment and the water column– benthic pelagic coupling (Sundback and Graneli,

115 1988; Nilsson and Sundback, 1991; Sundback and McGlathery, 2005). This process would be an important source and sink of nutrients in coastal systems and estuaries. The ability to trace nutrients injected into one part of a benthic-pelagic system would more directly implicate the benthic microalgal community in the mobilization and delivery of nutrients to other parts of the system. This has been demonstrated through the mass balance measurement of dissolved and particulate nutrients within the sediment, in porewaters, and in the overlying water column (Robson et al. 2008). Isotopic ratios can be useful tracers (both stable and radioactive) and have been used in pelagic systems for many years (Steeman-Nielsen, 1951; Hoare et al. 2005; Kroeger et al. 2006). Sediment porewater is much more difficult to label with a tracer than the water column and radioactive elements require controlled use to ensure proper containment, storage and disposal. This is compounded in Florida Bay by the geochemical sorption of P onto carbonate sediments (Millero, 2001). The radiolabelled P tracer is ineffectively mixed into the sediment (Jensen et al. 1998). Measuring the pools of available dissolved P within the sediment and the water column fractions over time, while unable to overcome this obstacle, can be used to evaluate the transport of labeled P isotopes in the presence and absence of BMA. Evaluating the potential for benthic microalgae to utilize benthic P was the purpose of this study. The objectives were to determine if flux of 33P from the sediment occurs at a statistically significant level for areas with, or without, a benthic microalgal layer, and to use changes in the measured nutrient pools over time to support benthic microalgal mediation of 33P flux. Materials and Methods To test the flux of radiolabelled 33P through the BMA I used the method of Sororkin (1992). Briefly, the experiments included an abiotic control and a treatment control (core without the addition of 33P). The abiotic, or killed, control accounted for physical adsorption of the 33P to the carbonate sediments and represented Fickian diffusion across the sediment-water interface. Cores with microalgal layers (live), but with nonradioactive PO4 added served as treatment controls. Some cores were incubated in the dark to evaluate the effect of light on flux. The overlying water column was

116 carefully decanted and replaced with filtered water from the sampling site without 33P added. The water column was subsampled over time. The amount of 33P in the overlying water column was then measured as a function of time and presence/absence of microalgal layer. In the first experiment, I added a large amount of P (~37M) to the water column to evaluate the effect on 33P flux. In subsequent experiments I added light and dark P enhanced treatments.

Figure 5.1. The locations where field cores and water were collected in Florida Bay for all the experiments.

These experiments were conducted using sediment from End Key in central Florida Bay and Carl Ross Key in western Florida Bay (Fig. 5.1). Physical parameters were recorded at the field collection sites and temperature was recorded at sampling

117 intervals throughout the laboratory experiment. Cores (2.5cm diam.) were removed from areas of the sediment devoid of above and belowground seagrass biomass. Sediment cores were taken with 60cc syringes with the end removed. As the cores were inserted into the sediment the plunger was simultaneously retracted until the core reached a depth of ~10cm. Cores were capped with #7 rubber stoppers on the bottom to ensure that sediment profiles remained intact. Cores were immediately transported to the laboratory submerged in ambient seawater from the site. Upon return to the laboratory, damaged or disturbed cores were removed from the experiment. The sides of the core sleeves were pierced at a depth of 2cm below the sediment water interface with a metal probe heated on a Bunsen burner. This depth was determined based upon preliminary experiments. The hole was plugged with a piece of modeling clay to ensure no porewater escaped. Cores were haphazardly arranged in 3 or 5 replicates of treatments. The treatments consisted of light and dark replicates of: killed + 33P, live+ 33P, live control (Fig. 5.2).

118

remove water +NaN3 (killed only)

Filt. 33P

H2O 33P H2O 33 2M P or P 33P (100L of 20µM) 33P 2 cm 33P 33P 33P 33P 33P 33P

The water column was carefully siphoned off each core. Sodium azide (0.75M) was added to the surface of cores to kill the BMA layer in „killed‟ treatments. Additional water from the field site was collected in 2L brown Nalgene bottles. This water was filtered through 0.7µ GF/F filters. Twenty milliliters of filtered water from the sample site was added to the top of all cores. The following morning 0.1ml of carrier free 33P, or low nutrient seawater (P <0.001µM) at the appropriate concentration, was added to the sediment through the hole at the 2 cm depth with a syringe. Only a small amount of liquid was added to the sediment in order to minimize the effects of porewater 119 additions. The hole was then recapped with the clay to prevent porewater from leaking out. 33P additions were made to the sediment not the overlying water column. P- enhanced additions were made to the water column by the addition of 0.5ml of 20µM orthophosphate standard. Total dissolved P (TDP) and soluble reactive P (SRP) was measured in the initial field filtered seawater and the light and dark controls at the conclusion of the experiment.

Initial (t0) measurements were taken from the overlying water column from each treatment immediately following the addition of the 33P or ambient water. Samples were subsequently collected from the water column within each core at various intervals for up to 24 hours. Dark samples were kept under aluminum foil covers except when samples were being collected. Light samples were exposed to dim natural and fluorescent at 14h light: 10h dark. Samples were held at ambient room temperature, which varied from 20- 23 C. Samples (100µL) were collected from each core using a pipettor. Tips were changed between samples. Samples were placed into labeled glass scant vials with 10mls of scintillation liquid and radioactivity was measured the following day on a scintillation counter. Activity was corrected for decay and the concentration per L after correction for removal of the overlying water column due to subsampling. Preliminary experiments were conducted to determine the best way to kill the microalgal layer. Potential problems of removing the microalgae included minimizing disturbance of the sediments and not altering the microzonation of the nutrients within the porewater/sediment complex. The methods I tested were physical removal, chemical killing (treating the sediment with copper sulfate, bleach, or sodium azide) or heat killing (pasteurization). Chemical killing methods were the most efficient and maintained the polysaccharide matrix around the cells, but generated hazardous waste handling concerns and some methods affected the measurements of phosphorus either by affecting the redox potential of the sediment or by introducing color that confounded the colorimetric measurement. Of the chemical agents I tested, sodium azide was the most effective with the least pH interference. Heat killing altered the sediment/porewater complex and evaporated porewaters thus changing the concentration. Another option was the inclusion of a competitive inhibitor to autotrophic phosphate uptake (AsO3) in abiotic

120 controls, but again there were hazardous waste concerns and heterotrophic activity might confound results. Long term dark exposure for the abiotic controls was determined to be too time consuming for practical use. A preliminary time-course experiment was undertaken to determine the amount of time it takes 33P to enter the water column in live samples, to maximize 33P uptake while minimizing release in killed samples. I also evaluated the depth necessary for reliable insertion of the 33P without immediate transport to the overlying water column due to porewater disturbance. Release or uptake by the sediment in live samples that deviated from controls or killed samples was deemed biologically produced. Although dominated by BMA, I could not determine what element of the benthic community might be responsible for this release or uptake. Because of the problems associated with P radioisotopes in carbonate sediments mentioned in the introduction, change in the nutrient concentrations over time was used to show the flux of P between the sediment and water column pool. I used pre- and post-experiment water column TDP measurements from control samples in comparison to the P added to the sediment and after correcting for the removal of water as a result of subsampling. I assumed evaporation during the experiment was negligible. Results Preliminary experiments The first preliminary experiment conducted was to test the minimum depth below the sediment –water interface needed to ensure the injection process did not introduce 33P into the overlying water column. This experiment was conducted using only central Florida Bay sediment, because this location had the least consolidated sediments. The results of this experiment are shown in Figure 5.3, and indicate that a minimum of 4mm was necessary to maintain the integrity of the sediment-water interface for central Florida Bay sediments. Despite this result a standard injection depth of 2cm was used during future experiments to ensure water column contamination was not an issue. The second preliminary experiment was to determine the minimum sampling interval to measure P flux to the overlying water column. P flux is almost immediate in live samples, so a 5 minute interval was chosen for the initial sampling. Practically, this interval was not able

121 to be sampled except initially, so longer intervals were used throughout the rest of the experiment.

33P Depth of Injection 10000 9000 4mm 8000 3mm 4mm

1 7000 - 3mm 6000 2mm 5000 4000

Counts min Counts 3000 2000 1000 0 initial 5 10 20 40 80 160 320 time (mins)

Laboratory experiments The first laboratory experiment was conducted in August 2003 using sediment from western Florida Bay. I did not use P enhanced treatments in this first experiment, but instead the live samples were P enhanced partway through the experiment. Due to limited cores, only 3 replicates were used for each treatment. Figure 5.4 shows the results of this experiment and the statistical results are presented in Appendix B. Although variable in concentration among replicates, 33P immediately went into the water column from the sediments in the live samples. Killed samples showed 33P in the water column by the first ½ hour, but accumulation did not seem apparent until hour 4. 122 Controls and killed samples were significantly different from each other several times early in the experiment and consistently different after 4 hours. But, there was not a significant difference between killed and live samples after 4 hours. Dark and light live samples both indicate 33P release from the sediment initially followed by uptake by the sediment or the vessel wall, although not statistically different from each other (n=3). I did not find any effect from P enhancement of the water column with nonradiolabelled

PO4.

August 2003 9 Western Bay 8 Sediment 33P Flux Control 7 Killed Dark

) 6 1 - Light

h 5

1 - 4 PO4 Enhancement

3 (fmol l (fmol

4 2

PO 1 33 0 -1 30 1 hr 2 hr 3 hr 4 hr 5 hr 6 hr 7 hr 8 hr 20hr 24hr min -2 Time

Figure 5.4. Average benthic flux of 33P (femtomoles L-1 h-1) from the sediment in cores taken in August 2003 from western Florida Bay. Error bars represent standard error of 5 33 replicates. Carrierfree PO4 in low nutrient seawater was injected into the sediment at 2cm below the sediment-water interface. An enhancement dose of 37 M of P was added to the water column in the live and killed treatments after 8 hours, but not controls.

During the second experiment in May 2004, sediment from both sites in Florida Bay was used, with 5 replicate cores per treatment. P enhanced treatments were also

123 used for direct comparison with treatments, instead of enhancing the cores mid- experiment as had been done previously. Figure 5.5 shows the results of this experiment for western Florida Bay. As in the previous experiment, 33P immediately entered the water column in live treatments. Replicates were variable and treatments were not significantly different from each other, however trends are evident related to live v. killed samples and enhanced v. live samples. Controls and killed samples were significantly different from each other, but light and dark controls did not differ from each other. The live dark samples averaged twice the 33P concentration of light samples. P enhanced live dark samples were also greater in 33P flux than P enhanced light samples in contrast to the August 2003 experiment. The amount of 33P released to the water column was an order of magnitude greater for western Florida Bay than central Florida Bay. This supports field measurements of greater dissolved P concentration in the western Bay water column made at other times, however flux measurements were similar between locations. In central Florida Bay, P enhanced treatments showed the greatest flux of 33P from the sediment, followed by killed samples, live samples, and controls, respectively, (Fig. 5.6). The dark P enhanced cores had greater average flux than light treatments, similar to western Florida Bay. There were no significant differences in flux between dark and light P enhanced treatments, or between any other treatments due to high variability. Killed treatments were significantly different from controls and live treatments after the first hour. There were no significant differences between live and control samples during the experiment, although the dark live samples were greater in average 33P flux, followed by the light live samples and controls.

124

May 2004

Western Bay Control Dark 1.0 33 Sediment P Flux Control Light 0.9 Live Dark 0.8

Live Light

) 1

- 0.7 Enh. Live Dark

h 1 1 - 0.6 Enh. Live Light

0.5 Killed (fmol l (fmol

4 0.4

PO 0.3 33 0.2 0.1 0.0 30mins 1hr 2hr 3 hr 19.5hr Time

Figure 5.5. The average 33P flux (femtomoles L-1 h-1) from the sediment in cores taken from western Florida Bay, May 2004. Error bars represent standard error of 5 replicates per sampling interval.

125 May 2004 Control Dark 0.20 Central Bay Control Light Sediment 33P Flux Killed 0.15 Live Dark

0.10 Live Light )

1 Enh. Live Dark -

h 0.05

1 1 Enh. Live Light - 0.00

30mins 1hr 2hr 3 hr 19.5hr (fmol l (fmol

4 -0.05

PO -0.10 33

-0.15 Time

-0.20

Sediment-water column P pools In support of my experimental evidence of 33P as a tracer between sediment and water column in control treatments, I measured the amount of TDP introduced into the sediment pool and available in the water column pool at the beginning (t0) and end (tf) of the experiments. My dissolved phosphate pool results show the net direction of TDP flux was always from the sediment to the water column during the August 2003 (Fig. 5.7) and May 2004 (Figs. 5.8 and 5.9) experiments. These figures also indicate that light treatments accumulated more TDP from the sediment, however not significantly different from dark treatments. The magnitude of the flux differed between the two sites. Although sediment porewaters were not measured this probably reflects differences in original P concentration and possibly sediment sorptive potential.

126 August 2003 Western Florida Bay

Change in PO4 of the Water Column 0.35

0.3

0.25

0.2

uM PO4 PO4 uM 0.15  0.1

0.05

0 Dark 1 Dark 2 Dark 3 Light 1 Light 2 Light 3 Treatment and Replicate

Figure 5.7. Daily flux of TDP to the overlying water column from the western Florida Bay sediment during the August 2003 experiment. All results indicate flux from the sediment to the water column.

127

May2004 Western Florida Bay Change of PO4 in the Water Column 0.08

0.07

0.06

0.05

0.04 uM PO4 uM

 0.03

0.02

0.01

0 Dark 1 Dark 2 Dark 3 Dark 4 Dark 5 Light 1 Light 2 Light 3 Light 4 Light 5 Treatment and Replicate

128

May 2004 Central Florida Bay Change of PO4 to the Water Column 0.03

0.025

0.02

0.015 uM PO4 uM

 0.01

0.005

0 Dark 1 Dark 2 Dark 3 Dark 4 Dark 5 Light 1 Light 2 Light 3 Light 4 Light 5 Treatments and Replicate

Discussion 33P flux to the water column was immediate in the live samples and they were significantly different than killed in both locations. This implies biological mediation in P flux. TDP flux from the sediment to the water column pools in control samples supports this finding. In western Florida Bay, live samples showed the greatest average 33P flux followed by killed samples and then controls in the first experiment. Live unenhanced treatments again were the greatest average 33P flux in the second western Florida Bay experiment, followed by P enhanced live treatments, killed and controls. This finding is not easily explained, but it is possible that P enhancement could have triggered luxury P consumption in BMA. In central Florida Bay, P enhanced treatments showed the greatest average 33P flux followed by killed samples, live samples and then controls. This order may reflect the BMA community at this location is comprised of species unable to store excess P. 129 That live unenhanced samples released less 33P to the water column than the killed samples suggests P uptake from the sediment pore waters by BMA. It is important to remember the amount of 33P accumulated in the water column was very small in comparison to the western Florida Bay experimental cores and the variability within treatments was high. Differences in 33P flux magnitude between the two field locations mirror TDP concentration in situ. In the May experiments, dark live treatments were greater in average 33P flux than light live treatments for both ambient and P enhanced treatments at both locations. The August experiment had the opposite result with light live treatments exceeding dark live treatments in 33P flux. There were fewer replicates during this first experiment and light and dark treatments were more similar to each other following P enhancement of the water column. Chlorophyll a concentrations were not measured on the samples, but patchiness could have led to variability in the BMA layer among cores. Although inconsistent results were found among experiments, this suggests that BMA light dependent reactions somehow affect 33P flux. I expected the presence of live BMA to help or hinder the 33P flux to the water column in comparison to killed controls – depending upon nutrient concentrations and diffusion gradients. The accumulation of 33P in the water column over the killed BMA samples was not a linear process. Reaching initial equilibrium in the killed treatments took between 1 and 5 hours and concentrations did not remain stable indicating uptake by the sediment or vessel wall, or sampling error. Sampling from slightly different locations within the tubes could account for this or perhaps there was minor sediment-water interface disturbance related to the sampling that affected diffusion. Summary In Florida Bay, the presence of a living benthic microalgal layer quickly affects the release of 33P from the sediment to the overlying water column beyond Fickian diffusion. This finding is supported by changes in the sediment and water dissolved P pool. Differences between light and dark treatments suggest differential release of 33P by BMA during light dependent reactions. The results suggest BMA involvement in the mediation of P flux from the sediment to the water column.

130

Chapter 6. The use of dissolved oxygen probes and Pulse Amplitude Modulated (PAM) fluorometry in the evaluation of benthic microalgal primary production in Florida Bay Introduction Quantifying primary production is important for examining ecosystem health and estimating the energy available for transfer into biomass at higher trophic levels. Primary production is the rate per unit area, or per unit volume, at which carbon is fixed into biomass by producers (Lawlor, 2001). In aquatic systems, primary production is typically measured by determining the oxygen concentration within a closed system, by measuring the amount of inorganic 14C incorporated into organic matter, or by fluorescence techniques like pulse amplitude modulated (PAM) fluorometry. These techniques measure net primary production, rather than gross, because all the methods are indirect. The O2 method, while taking into account respiration when dark bottles are used, does not account for the non-light requiring (dark) reactions of photosynthesis (Revsbech et al. 1981). Nor is O2 produced by algae only under aerobic conditions, but can be produced by algae containing hydrogenase under anaerobic conditions using water as the reductant (Lawlor, 2001). 14C measurements do not account for heterotrophy and remineralization of carbon, even if predators are excluded and incubations are short (Hancke et al. 2008). Finally, fluorescence techniques measure the theoretical limits of primary production, rather than the actual fixation of carbon (Walz, 1998). Most methods of measuring primary production can be difficult to carry out in the sediments. 14C measurement can be complicated by heterogeneous labeling of the sediments, which can lead to underestimates of the carbon fixation rate (Revsbech et al. 1981). In addition, it can be difficult to determine the total inorganic carbon at the sediment/water interface (Revsbech et al. 1981).

131 Despite some sources of error, O2 measurement has been a useful field method for measuring net primary production accurately and rapidly (Revsbeck, et al. 1981). O2 is evolved during photosynthesis in plants when a water molecule is split in the presence of light, CO2 and chlorophyll. Plants fix the carbon from carbon dioxide and produce carbohydrates. For each ml of oxygen produced, 0.536 mg of carbon has been fixed during photosynthesis (Lawlor, 2001). Thus the basic photosynthesis equation is:

Eq. 6.1 6CO2 + 6H2O  C6H12O6 + 6O2 Oxygen probes Commonly, dissolved oxygen (DO) measurements involve either the Winkler titration method or microelectrodes. The latter can be inserted into the sediment or placed within an incubation chamber on the bottom. They have the advantage of minimal disruption of the physical and chemical gradients in the sediment (Revsbeck, et al. 1981; MacIntyre et al. 1996). Incubation chambers are the best field method available for measuring sediment oxygen evolution, since a semi-closed system is required. However, a major disadvantage of using incubation enclosures is they can disrupt mixing of the water column and surface sediments by eliminating wave and current effects (Malan and McLachlan, 1991). Encapsulating the bottom may alter the vertical distribution of microorganisms within the benthos. Additionally, decreased diffusion rates of gases into and out of the sediments in the semi-closed system can also occur, limiting carbon available for photosynthesis (MacIntyre and Cullen, 1995).

The Winkler method has been used to measure O2 since the late nineteenth century and most recently adapted for seawater use in 1968 (Strickland and Parsons). Clarke-type oxygen electrodes have been in use since the 1960‟s, and commercially available since the 1970‟s. Benthic chambers have been successfully used to determine benthic O2 evolution in North Carolina (Cahoon and Cooke, 1992) and Florida Bay

(Yarbro and Carlson, 2008; Yates and Halley, 2001), among other places. PAM fluorometry The PAM fluorometer has the following advantages over traditional methods of measuring primary production: instantaneous readings, minimal disturbance of the benthos, and accuracy. The PAM derives the effective quantum Yield (Fv/Fm) of

132 photosystem II during photosynthesis, but actually measures the photosynthetic Yield by applying a saturating light pulse to the sediment microalgal layer inducing maximum chlorophyll fluorescence Yield. The photochemical Yield is then calculated, which has been shown to be closely correlated with the effective quantum Yield of photosynthesis in macroalgae and some seagrasses (Beer and Bjork, 2000). The PAM also has a 2 quantum sensor to measure photosynthetically active radiation (PAR), which allows calculation of the electron transport rate. Rapid light curves and induction curves are also possible using this instrument in conjunction with the application of continuous actinic light (Walz, 1998). When PAR is known, the apparent relative rate of electron transport (ETR, M e- m-2 s-1) is calculated. Rapid light curves (RLC‟s) demonstrate the photochemical response to increasing light resulting in O2 production. O2 production is assumed to be linearly related to ETR (Walz, 1998), however, species specific exceptions to this assumption have been found (Masojidek et al. 2001) that are linear only at low light levels. The orientation of the light source to the photosynthetic target is perhaps the most important aspect of PAM fluorometry, for reliability and comparison of repeated measurements. Water temperature is the second most important consideration for comparison of measurements. ETR is enzyme-dependent (photochemical Yield) and requires an optimal temperature range. High temperatures (Necchi, 2004), due to high light ,can cause photoinhibition that decreases fluorescence Yield (Walz, 1998). Migration within the sediment is often invoked as a hindrance to accurate light curve readings. The application of light in instantaneous measurements or in rapid light curves is so short in duration that this can be eliminated as a confounding factor in the use of the PAM (Serodio et al. 2005). Both oxygen electrodes and PAM fluorometry have the benefit of reliable field measurements that lead to an estimation of primary production. Electrodes are cost effective but require maintenance and calibration for reliable results. PAM fluorometers are costly but require little field preparation and maintenance. Oxygen electrodes are most efficient at measuring relatively high oxygen concentrations (oxic conditions) and are limited in accuracy by the sensitivity of the electrode (Revsbech et al. 1981). Oxygen

133 techniques measure net benthic production (Cahoon and Cooke, 1992), whereas fluorescence techniques measure only the photosynthetic components of the benthic microalgal community (Walz, 1998). PAM may overestimate photosynthetic capacity in well-established populations (Schmitt-Jansen and Altenburger, 2007), and its accuracy can be adversely affected by refractive fluorescent compounds in sediments (Serodio et al. 1997). Materials and Methods Dissolved Oxygen The study sites were End Key in central Florida bay and Carl Ross in western Florida bay (Fig. 6.1). Sampling occurred between November 2000 and August 2002 and measurements were concurrent with benthic nutrient flux measurements. I measured O2 production during a short incubation period (less than 10 hours) in both light and dark chambers of a known volume placed over sediment devoid of seagrasses. I placed, recovered and sampled the chambers by wading. To ensure a good seal, the acrylic chambers were inserted into the sediment. These chambers contain both a gentle water circulation device and water replacement mechanism (attached collapsible IV bag with water) and are similar to those used by Cahoon and Cooke (1992). An oxygen microelectrode was inserted through a port in the chamber wall prior to deployment.

134

Figure 6.1. Station map indicating the locations were DO in benthic chambers (Carl Ross Key and End Key only) and PAM fluorometry of benthic microalgae were measured.

Dissolved oxygen production was measured in millivolts throughout the incubation with a Lazar DO-166MT-1 microelectrode zeroed in 1M NaHSO3. The electrode was connected to a battery operated pH meter for readout. Electrodes were calibrated in air at ambient temperature and pressure. Microelectrodes were recalibrated daily prior to sampling. These microelectrodes do not require stirring at the membrane water interface and can measure DO between 0.1 and 20 ppm (mg L-1), with 0.01ppm level of accuracy. Because respiration of the entire benthos enclosed beneath the chamber occurs (both microalgal and animal), this method only measures net benthic primary production (NBPP) in the clear chambers. Dark chambers were used to measure sediment respiration (including dark photosynthetic reactions). The two values were added to 135 calculate gross benthic production (GBP) for the incubation period. The equations used to derive these measurements are set forth in Appendix C. Despite the flaws inherent in this method, GBP was found to be a better estimate of microalgal production than NBPP in Onslow Bay, NC (Cahoon and Cooke, 1992). Areal net primary productivity (ANPP) estimates were made from the GBP extrapolated to 733.33 km-2 designated as bare bottom baywide by Prager and Halley (1997) plus that 40 km-2 of seagrass lost to dieoff since 1992. Concurrent with most of the field incubations, ambient PAR was measured throughout the incubation using a Licor 2quantum sensor and light meter or the 2sensor on the PAM fluorometer. Temperature, salinity and initial and final secchi depth were recorded as well. Time, tide, cloud cover, wind direction and speed estimates were also made and can be compared with measurements obtained from nearby monitoring stations in the SEAKEYS data buoy system (data courtesy of the Florida Institute of Oceanography and NOAA). PAM Fluorometry The study sites were End Key in central Florida bay, Arsnicker Key in southeastern Florida Bay, and Carl Ross Key in western Florida bay. End and Arsnicker Keys were always submerged, but Carl Ross Key is intertidal on a semidiurnal basis. The chambers were deployed in areas devoid of seagrass and I was careful not to disturb the area where the PAM fluorometry measurements were made. Instantaneous measurements and rapid light curves were taken throughout the sampling period adjacent to the benthic chambers using a Walz DIVING-PAM fluorometer. This resulted in quantum Yield and ETR values as well as other photosynthetic parameters. The actual absorption of light by the microphytobenthos (AF) must be determined to obtain absolute rates of electron transport (ETR) for use in the following equation: Eq. 6.2 ETR = Yield * PAR * 0.5 * AF Yield and PAR are measured by the instrument. The value of 0.5 refers to half the photons being absorbed by Photosystem II. The absorption factor (AF) is the fraction of incident light absorbed by the microphytobenthos as determined from the average of several measurements from a controlled light source both with, and without, a

136 microphytobenthic community covering the PAR sensor. Erroneous data that were the source of equipment malfunction were removed from the dataset and differences in ETR or Yield by location, time of day, or season were evaluated by linear regression. Results DO Electrodes The range of DO I measured varied from 5.80 To 8.93 mg L-1. The lowest DO values were found during the spring of 2001 and 2002 (Figs.6.7-6.9) when only the western Bay location was sampled and August 2002 (Figs. 6.3 and 6.12) when both the western and central bay were sampled. The greatest values were found in central Florida Bay in July 2002 (Fig. 6.2) and November 2000 (Fig. 6.5) when only the western bay was sampled. In general, at the time of initial electrode deployment both chambers were similar in DO concentration. Despite some variability, DO in the light chambers tended to increase with increasing PAR. DO usually peaked in the afternoon, but in three sampling periods the peak was at noon. As PAR decreased in the afternoon, light chamber DO continued to increase or leveled off. PAR measurements were much more variable than DO due to cloudiness, although averaged for the minute preceding the measurement. Dissolved oxygen flux in the dark chambers, generally, remained level or decreased throughout the incubation period. One time, in July 2002, I deployed oxygen sensors in two light chambers and one dark chamber (Fig. 6.2). Initially one light chamber had a lower DO concentration than the dark, but the other was higher in concentration. The difference among chambers was 0.3mg L-1 – high in comparison to other sampling periods and this may indicate the initial calibration was flawed. Despite the quantitative difference, both light chambers had similar qualitative DO results. The other instance when DO in the light chamber dipped lower than the dark one was in August 2002 during one sampling period at mid-day, but later rebounded (Fig. 6.12). In a few instances DO in light chambers decreased in late afternoon, coincident with decreasing PAR (Figs. 6.4, 6.5, 6.9 and 6.10). In two instances DO increased again following a mid-afternoon drop, however the overall changes to DO concentration were <0.01 mg L-1 (Figs. 6.4 and6.11). I cannot determine if these atypical fluctuations in DO

137 represents variation in respiration or photosynthesis because dark chambers were not able to be measured during either of these incubations as a result of a broken electrode, and PAR measurements for the afternoon were not available for one of the two sampling periods. Over time, DO dropped in dark chambers in comparison to light chambers. There were two exceptions in November 2000 and March 2001 in western Florida Bay (Figs.6.6 and 6.9), when single measurements showed DO increases of 0.1 and 0.2 mg L-1 h-1, respectively within the chamber. However, in both instances the overall trend in the dark chambers was decreasing DO.

Table 6.1 Net benthic primary productivity (NBPP), sediment respiration (SR) and gross benthic primary production (GBP) – see Appendix C for calculations.

Net O2 Net O2 (light) NBPP (dark) Sed Resp GBP Hour mg C m-2 mg C m-2 mg C m-2 Date s mmol L-1 h-1 mmol L-1 h-1 h-1 9/11/2000 6 -0.12 -16.18 ND ND -16.18 11/13/2000 4.5 0.20 36.33 0.30 65.40 101.73 11/14/2000 6 0.45 61.31 -0.20 -32.70 28.61 3/13/2001 3 0.10 27.25 -0.15 -49.05 -21.80 3/15/2001 6 0.45 61.31 -0.40 -65.40 -4.09 3/16/2001 6.3 0.10 12.98 -0.10 -15.57 -2.60 5/16/2001 9 0.40 36.33 -0.63 -68.67 -32.33 5/7/2002 4.5 0.13 23.23 ND ND 23.23 7/17/2002 2.5 -0.01 -2.59 0.00 0.31 -2.28 8/2/2002 3.6 0.00 -1.01 0.00 -0.92 -1.93 8/8/2002 3 0.04 11.51 -0.12 -40.22 -28.72 ND = not determined

The production values for each sampling date are presented in Table 6.1. Positive values in the net O2 columns indicate a net increased in DO throughout the incubation, negative readings indicate a net decrease in DO throughout the incubation. Net Benthic Primary Production typically exceeded respiration, resulting in a net autotrophy in the benthos. There were three sample periods where NBPP was heterotrophic. The latter two periods were only weakly heterotrophic and the first sampling date may have been 138 affected by lack of mixing within the chamber. Based upon an average NBPP of 21.86 mg C m-2 h-1, benthic microalgae inhabiting otherwise unvegetated bottom would contribute ~406000 kg C (406 metric tons) to Florida Bay daily.

Central Florida Bay July 17, 2002 9 1000

900 8.9 800

8.8 700

s -

600 2 8.7 - light 500 E m E light 8.6  dark

DO mg L mg DO 400

PAR PAR PAR 8.5 300

200 8.4 100

8.3 0 10:00 11:00 11:30 12:30

Figure 6.2. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on July 17, 2002 in central Florida Bay. PAR was not measured at the final sampling period.

139 Central Florida Bay August 2, 2002 7 1000

900 6.95 800

700

6.9 1

1 -

600 2 - Light

6.85 500 E m E

 Dark

DO mg L mg DO 400 PAR

6.8 PAR 300

200 6.75 100

6.7 0 10:24 10:55 11:30 13:00

140 Western Florida Bay September 11, 2000 9 2020

2000 8.9

1980

8.8 1

- 2 1960 - 8.7 E m E light

1940  PAR DO mg L mg DO

8.6 PAR 1920

8.5 1900

8.4 1880 9:45 11:00 11:45 12:15 12:30 12:45 13:50 14:15 14:45 15:45

141 Western Florida Bay November 13, 2000 9 1600

1400 8.8 1200

8.6 1 -

1000 s

2 - light

8.4 800 E m E

 Dark

DO mg L mg DO 600 PAR 8.2 PAR 400 8 200

7.8 0 12:40 14:40 15:40 17:10

Figure 6.5. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on November 13, 2000 in western Florida Bay.

142 Western Florida Bay November 14, 2000 8.6 2000

1800 8.4 1600

1400

8.2

1 s

- 1200 2 - Light 8 1000

E m E dark 

DO mg L mg DO 800 PAR 7.8 600 PAR

400 7.6 200

7.4 0 7:30 9:00 10:30 12:00 13:30

Figure 6.6. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on November 14, 2000 in western Florida Bay.

143 Western Florida Bay March 13, 2001 6.65 2000

6.6 1800

6.55 1600

6.5 1400

1 -

6.45 1200 2 - light

6.4 1000 E m E

 dark

DO mg L mg DO 6.35 800 PAR PAR PAR 6.3 600

6.25 400

6.2 200

6.15 0 13:15 16:15

Figure 6.7. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on March13, 2001 in western Florida Bay.

144 Western Florida Bay March 15, 2001 7 2000

1800 6.8 1600

6.6 1400

1 -

1200 2 6.4 - Light

1000 E m E

 dark 6.2

DO mg L mg DO 800 PAR PAR PAR 6 600

400 5.8 200

5.6 0 9:00 10:00 11:00 12:00 13:00 14:00 15:00

Figure 6.8. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on March 15, 2001 in western Florida Bay.

145 Western Florida Bay March 16, 2001 7 1800

6.9 1600

1400 6.8

1200 1 -

6.7 s

- 2 1000 - light

6.6 E m E

800  dark

DO mg L mg DO 6.5 PAR 600 PAR 6.4 400

6.3 200

6.2 0 9:00 10:00 10:20 11:50 14:00 15:21

Figure 6.9. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on March 16, 2001 in western Florida Bay.

146 Western Florida Bay May 16, 2001 9 2500

8.8

8.6 2000

8.4

1 -

8.2 1500 2 - light

8 E m E

 dark

DO mg L mg DO 7.8 1000 PAR PAR PAR 7.6

7.4 500

7.2

7 0 9:15 10:45 12:15 13:45 16:45 18:15

Figure 6.10. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on May 16, 2001 in western Florida Bay.

147 Western Florida Bay May 7, 2002 6.05 1400

6 1200

5.95

1000

1 2

- 5.9 - 800

5.85 m E light 600  PAR

DO mg L mg DO 5.8 PAR PAR 400 5.75

5.7 200

5.65 0 10:15 11:15 12:15 13:05 14:15 14:45

Figure 6.11. DO concentration (left y-axis) within a light benthic chamber and ambient PAR values (right y-axis) over time on May 7, 2002 in western Florida Bay.

148 Western Florida Bay August 8, 2002 6.6 1000

900 6.55 800 6.5

700 1

6.45 -

s 1

- 600

2 - Light 6.4 500

E m E dark 

DO mg L mg DO 400 6.35 PAR 300 PAR 6.3 200 6.25 100

6.2 0 9:30 11:00 12:30

Figure 6.12. DO concentration (left y-axis) within light and dark benthic chambers and ambient PAR values (right y-axis) over time on August 8, 2002 in western Florida Bay.

PAM Fluorometry The PAM fluorometry results are presented in Figures 6.13-6.21. The instantaneous Yield measurements, those taken from one pulse of actinic light applied to the benthic microalgal community under ambient PAR conditions, indicate differences in the Yield and ETR by location/season and time of day. The Yield calculation is independent of ambient PAR, but ETR is derived from multiplying Yield with PAR (see equation 6.2). When plotted by time of day, the highest Yield and ETR values were found in the morning samples. Lower Yield, and therefore lower ETR‟s, were found in the afternoon, despite abundant light (Fig. 6.13).

149 Yield vs Electron Transport Rate by Time of Day 1.0

0.9

0.8

0.7

0.6

0.5

Yield pm 0.4 am

0.3

0.2

0.1

0.0 0 50 100 150 200 250 300 ETR

Although there appears to be a seasonal signal to Yield (Fig. 6.14), this is confounded by location (Fig. 6.15) since the southeastern site was only sampled one day, in the morning, and represented the majority of wet season samples analyzed. All the southeastern site samples were characterized by high instantaneous Yield and low ETR. The central and western bay sites were variable in instantaneous Yield and ETR, however there was a positive linear correlation between the two parameters. Most of the central Florida Bay measurements were clustered near <0.4 Yield/<100 ETR, while some central Florida Bay measurements, and nearly all western Florida Bay measurements, were >0.4 Yield/>100 ETR. Temperature differences on sampling days varied from 3 to 4.5ºC between morning and afternoon measurements. Serodio et al (1997) found major

150 differences in temperature (<15ºC, or 30ºC>) affected Yield measurements in intertidal communities in Portugal, but that between these values measures were comparable. While the magnitude and range of temperature differences would be different in the Florida Bay system, the temperatures values during my experiments were within the normal temperature distribution for the region and thus not expected to adversely impact my comparisons.

Yield vs Electron Transport Rate by Season 1

0.9

0.8

0.7

0.6

0.5

Yield Wet 0.4 Dry

0.3

0.2

0.1

0 0 50 100 150 200 250 300 ETR

151 Yield vs Electron Transport Rate by Region 1

0.9

0.8

0.7

0.6

0.5

Central Yield 0.4 Southeast Western 0.3

0.2

0.1

0 0 50 100 150 200 250 300 ETR

The results of the rapid light curves of benthic microalgae outside the incubation chambers over the course of a sampling day are presented in Figures 6.16-6.21. Rapid light curves are generated from actinic light throughout a spectrum (low to high) applied to a sample that has not been dark adapted. The maximum ETR, measured during the higher light intensities at the end of a curve, tended to be low early in the morning and high in the afternoon (Figs.6.16, 6.17, 6.19). One exception was in July 2002 in central Florida Bay when the 1:00pm sampling event was lowest overall (Fig. 6.18). The southeastern Florida Bay, although only sampled once, had the highest ETR from rapid light curves (Fig. 6.16). This is in contrast to the instantaneous samples where the Southeast had low ETR coincident with high Yield. Western Florida bay also had

152 high ETR‟s for rapid light curves, but there was more variability between sample intervals and among replicates (Figs. 6.19-6.21). The lowest ETR‟s from rapid light curves were consistently in central Florida bay and were about 1/5 that found at the other two locations in the bay (Figs.6.17 and 6.18). The shape of the curves from the southeastern and western Florida bay stations were nearly linear. The central bay curves were more parabolic.

Electron Transport Rate Southeastern Florida Bay May 2, 2002 160

140

120

100

80 10:00am 60 12:00pm

Electron Transport Electron Rate Transport 40

0 2 29 98 216 333 514 725 1096 1496 PAR

Figure 6.16. Electron transport rate (ETR) from triplicate rapid light curves in southeastern Florida Bay on May 2, 2002 at the 10:00 am and 12:00pm sampling interval.

153 Electron Transport Rate Central Florida Bay April 4, 2002 35

10:00am 15 12:00pm 1:30pm 10

5 Electron Transport Rate Transport Electron

0 0 155 199 264 362 506 718 1019 1528 -5 PAR

Figure 6.17. Electron transport rate (ETR) from duplicate rapid light curves in central Florida Bay on April 4, 2002 at the 10:00 am, 12:00pm and 1:30pm sampling interval.

154 Electron Transport Rate Central Florida Bay July 17, 2002 90

40 9:00am 11:00am 30 1:00pm 20 3:00pm

10 Electron Transport Rate Transport Electron 0 1 29 92 206 324 505 715 1084 1487 -10

-20 PAR

Figure 6.18. Electron transport rate (ETR) from duplicate rapid light curves in central Florida Bay on July 17, 2002 at the 9:00 and 11:00 am and 1:00 and 3:00 pm sampling interval.

155 Electron Transport Rate Western Florida Bay March 15, 2001 200

180

160

140

120

100 9:00am 11:30am

80 1:15pm Electron Transport Rate

0 3 50 76 111 159 325 499 856 1250 PAR

Figure 6.19. Electron transport rate (ETR) from triplicate rapid light curves in western Florida Bay on March 15, 2001 at the 9:00 and 11:30 am and 1:15pm sampling interval.

156 Electron Transport Rate Western Florida Bay March 16, 2001 140

120

100

60 10:00am 11:45am 40

20 Electron Transport Rate Transport Electron

0 3 50 76 111 159 325 499 856 1250 -20 PAR

Figure 6.20. Electron transport rate (ETR) from duplicate rapid light curves in western Florida Bay on March 16, 2001 at the 10:00 and 11:45 am sampling interval.

157 Electron Transport Rate Western Florida Bay May 4, 2002 100

40 3:00pm

Electron Transport Rate Transport Electron 20

0 0 13 56 119 213 334 475 723 993 PAR

Figure 6.21. Electron transport rate (ETR) from triplicate rapid light curves in western Florida Bay on May 4, 2002 at the 3:00 pm sampling interval.

Discussion Dissolved oxygen trends were usually as expected, with light chambers increasing in DO throughout the incubation as a result of benthic community photosynthesis and dark chambers remaining stable or decreasing as a result of benthic community respiration. DO in light chambers usually followed PAR values, with a few exceptions in the late afternoons. Photoinhibition or increased benthic respiration caused by temperature stress might explain those instances where DO dropped and then rebounded. The oxygen electrodes gave reasonable measurements, but were difficult to use in the field in remote parts of Florida Bay because of power supply issues.

158 The areal baywide productivity estimated from the dissolved oxygen method in my study was -0.016 - 0.061 g C m-2 h-1 compared to a net loss of ~ -0.25 g C m-2 h-1over mud bottom in Manatee Basin, Florida Bay (Yates and Halley, 2000). Their SHARQ deployment assembly is much larger, is not inserted into the sediment and must be deployed in deeper water in a different area, but benthic microalgal biomass would still be expected to be abundant over unvegetated bottom anywhere in Florida Bay. It is possible that their method was less sensitive and underestimated production, or that my method underestimated benthic respiration because night measurements were not made. Based upon my results, and the results of other researchers, areal productivity of areas colonized by benthic microalgae is significant compared to the water column (Cahoon, 2005) while much less than seagrass beds (Nielsen et al. 2007; Yarbro and Carlson, 2008).

The PAM fluorometer proved to be much easier to use in the field than the O2 electrodes, despite finicky controls. It provides much more information on the physiology of the algal community, beyond deriving the O2 production, and the measurements are non-destructive. However, results for sediment microalgal communities must be tempered with overestimation of fluorescence values as a result of the presence of refractory components which also fluoresce. Extracted chlorophyll measurements have shown that phaeopigments exceed chlorophyll by as much as 5:1 at these locations in Florida Bay (Chapter 3 in this dissertation). In addition, PAM fluorometry is relatively new and few measurements of benthic microalgae exist worldwide for comparison. The two methods should not necessarily be considered comparable due to the whole benthic community being measured by O2 probes while only photoautotrophs are measured by PAM. Instantaneous PAM measurements provide insight into the photosynthetic efficiency (capacity) of the algal community at the light history it was experiencing at the time of the measurement. The PAM fluorometer eliminates the nonphotochemical aspects of production and measures only the photochemical changes in the plant. (Schreiber et al. 1996) Yield (Y=Fv/Fm, where Fm is the maximal fluorescence Yield, and Fv is the maximal variable fluorescence), is a reliable measure of the potential

159 quantum Yield of photosystem II (Schreiber et al, 1996; Saroussi and Beer, 2007). Changes in fluorescence Yield reflect changes in photochemical efficiency or increase of heat dissipation - like photoinhibition (Walz, 1998). In my study, the time of day differences in ETR and Yield for the instantaneous measurements (Fig. 6.14) was probably due to the light history prior to the measurements or photosynthetic efficiency differences. However it may also be the result of high temperature or photoinhibition of the benthic microalgal community. These differences may reflect differing species composition, although I did not include this as part of my investigation. Diel differences in ETR in rapid light curves and instantaneous measurements have been reported in microalgae by other researchers (Masojidek, et al. 2001). Serodio et al. (2005), found three processes work simultaneously in BMA to respond to high light conditions that might be experienced in the intertidal area: decreased photosynthetic efficiency (down-regulation of photosynthesis through energy-dissipating mechanisms), increased carbon metabolism activity, leading to an increase in ETR; and downward migration in the sediments. I based my areal productivity calculations upon the assumption that differences in

ETR are linearly positively correlated to O2 production from photosynthesis at the same irradiance level since they are mathematically related via equation 6.2 above (Saroussi and Beer, 2007). Rapid light curves minimize ETR differences associated with variable -1 -1 light intensity, and O2 production (M 02 mg Chl h ) for this method is ~three times

ETR. Since the relationship between O2 production and ETR is roughly linear at low irradiance and BMA would likely experience light in this range I used the average ETR from all the light curves at irradiances between 200 and 500 E m-2 s-1 to calculate the areal O2 production. The average O2 production measured by PAM fluorometry would result in the production of ~817000 kg C (817 metric tons) to Florida Bay daily. Linearity in the rapid light curve suggests that the benthic microalgal community in the Southeast was not reaching the maximum photosynthetic capacity even at high light, or that some other factor was limting ETR. Later in the day the ETR was similar at the two highest light intensities, indicating the maximum production rate was achieved. In the western bay, and more so in the central bay, maximum ETR in rapid light curves

160 was achieved at lower light intensity. The higher light intensity end of the curve and sometimes those curves performed at the midday sampling periods when ambient PAR was high in those parts of the bay, reveal depressed ETR. So production rates in western and central Florida Bay suggest the benthic microalgae were adapted to lower light conditions than the Southeast. Schmitt-Jansen and Altenburger (2007) found PAM fluorometry was a good indicator of production in response to pollution in pelagic algae, but only for relatively young blooms (<9 weeks in their system). Stable communities did not show discernable changes in production in response to various factors. I did not measure benthic community composition during, so I cannot evaluate whether community stability might be a factor in production for Florida Bay. Diel differences in ETR for the rapid light curves indicate that as enzymes were activated and irradiance intensified, O2 production increased - except when inhibited by high light conditions. This result is similar to findings in freshwater cyanobacteria (Masojidek, et al. 2001) and BMA off Portugal (Serodio et al., 2006). The light intensities at the lower end of the curve are much more realistic for the benthic microalgae in Florida Bay, however, the light intensities achieved at the higher end of the light curves are unlikely to be realized by benthic microalgae unless exposed at low tide. This is possible at Carl Ross Key and at most of the benthic microalgal habitat in western Florida Bay. Despite being intertidal, northwestern Florida Bay experiences the most light limitation , except for Rankin Basin (Kelble et al, 2005). The central bay has episodic pelagic phytoplankton blooms that cause light limitation. The Southeast is probably never light limited, but production may be nutrient limited. Summary Both PAM fluorometry and oxygen production are methods of measuring primary production comparable among systems over time. Since PAM fluorometry is new technology and has not been widely used to examine benthic microalgae, the historical measurement of oxygen production and 14C are still probably most useful to managers to assess long term ecosystem response, but PAM fluorometry will be of future use when more measurements have been collected for comparison. The ease of use, repeatability

161 and vast amount of information achieved by using PAM fluorometry argue for its application in monitoring programs and a variety of instruments using this technology are available to best measure different primary producers in the lab and field. Based on both DO and PAM fluorometry in this study, benthic microalgae in Florida Bay are a significant source of primary production in this system and should be included in modeling management tools. Parts of Florida Bay that are vegetated by benthic microalgae are primarily autotrophic, with temporal and spatial heterotrophy – supporting Florida Bay as a net autotrophic system. Areal benthic microalgal production rates based upon oxygen evolution were similar to those reported for other areas around the world (Cahoon, 2005), but more than that reported for mud bottom in Florida Bay by Yates and Halley (2001). PAM fluorometry rapid light curve data suggests benthic microalgae are adapted for lower light intensities in Florida Bay and primary production was most efficient at 200-500 E m-2 s-1. This would approximate the PAR reaching the submerged bottom throughout most of the day in Florida Bay. Benthic microalgae production was hampered by high light, at least in central and western Florida Bay at some time periods during the day, as a result of photoinhibition or downregulation of ETR to protect photosystem II. This could be a consideration for areas where benthic microalgae are exposed during low tide. Instantaneous PAM measurements suggest differences in the photosynthetic efficiency of at the different locations.. Dissolved oxygen curves and PAM fluorometry support the potential for photoinhibition or heat losses in the afternoon, and that benthic microalgal production tends to increase with greater light intensities to a maximum efficiency at ~20% of peak surface PAR irradiance.

162

Chapter 7. Conclusion Benthic microalgae are a fascinating community underappreciated in terms of ecological importance and primary productivity throughout the world, and in Florida Bay. This community is one of the most important aspects of benthic-pelagic coupling to understand, as anthropogenic nutrients continues to be a problem in estuarine and coastal ecosystems. I hope to continue working with BMA in my career and hope it is someday included in monitoring programs as a matter of practice. This conclusion represents a synthesis of my findings on BMA and sediment nutrient flux in Florida Bay. Sediment Phosphate Flux The three regions of Florida Bay where I performed in situ experiments proved to be variable temporally and spatially in terms of P flux from the sediment. My findings are similar to others in this same system within seagrass beds and in carbonate systems elsewhere. Phosphate flux magnitude was least in the central and southeastern sites, which have the lowest available dissolved P concentrations in the water column and also the lowest P supply, based upon Fourqurean et al. (1992). Phosphate flux in the western bay was greater in magnitude. This region of the bay has the highest P supply from both surface (Fourqurean et al. 1992) and groundwater sources (Price, 2005) and greater dissolved P concentration in the water column. Florida Bay sediment, and the associated BMA community, acts as both a source and a sink of phosphorus in this system. Soluble reactive phosphate flux was quite low, at the limits of detection no matter the location. In the afternoons, SRP flux was only out of the sediment, although mornings were variable. This was the only significant difference in time of day that was noted. There was no difference in flux between light and dark treatments, perhaps because dark conditions were simulated during daylight hours ignoring any diel cues (Serodio et al. 2005) in the BMA community that may affect P flux if it is, indeed, mediated by the BMA. Dissolved organic phosphate was the dominant component of TDP in most of Florida Bay. This is in contrast to most systems, but in character for a bay with high 163 autochthonous production and little inorganic P input that comes from diverse sources (Gulf of Mexico, C111 canal, etc). The relative contribution of organic P to the total dissolved pool varied across my three sites according to the total available P, with the highest DOP in the central bay, and lowest in the southeast bay. Total dissolved nitrogen flux from the sediment was also variable temporally and spatially across the bay. In all experiments, TDN was much more abundant than TDP. The total dissolved benthic nutrient fluxes I measured could account for 100% of the nitrogen need and 6-41% of the phosphorus need of the water column during non-bloom conditions. However, this calculation ignores community composition and preferential uptake of nutrient species by phytoplankton. Based upon my average measured flux rates (positive) and the areas of Florida Bay that can be considered BMA habitat, ~1600 metric tons of P would enter the system from the sediment each year. Actual flux rates, as shown in this dissertation, vary into and out of the sediment which would balance any net loss of P from the sediment presumed from my extrapolation bay-wide. This would probably be much larger if the nuances of anoxia and hypoxia of the sediments were better understood. The semidiurnal pulses in sediment P flux, driven by anoxia, that have been reported by other researchers would probably occur more frequently in Florida Bay. However, sediment P flux would be an important source of P to the Florida Bay system. Only by more widespread in situ measurements of benthic nutrient flux will the potential for BMA to act as a nutrient source or sink be better understood. Benthic microalgal chlorophyll a standing stock I was surprised that chlorophyll a standing stock was similar among the three locations I measured, despite differences in sediment grain size, tidal range and nutrient supply. Seasonal changes were apparent in western Florida Bay, with a spring and fall increase in chlorophyll a standing stock. The lowest concentrations were in the wintertime, coincident with peaks in water column chlorophyll a standing stock which probably shaded the BMA community, or if winter storms resuspended BMA into the water column. This relationship between benthic and pelagic chlorophyll a is an important consideration in light limitation of BMA and macrophytes. While seasonal

164 trends in chlorophyll a standing stock were not as clear at the other two locations, greater sediment chlorophyll a standing stock was associated with spring/summer at the central site. My measurements of chlorophyll a standing stock are similar to others from Florida Bay, but are generally less than those in temperate systems. Pigments in the sediment are primarily phaeopigment, the dead and decomposing macrophytes and microalgae both contributing to this signature. Despite this, the areal contribution of ~700 kg Chl a day-1 from BMA in otherwise unvegetated bottom is significant in Florida Bay. Benthic microalgal chlorophyll a standing stock, using a method that removes chlorophyllide a interferences, should be a component of ecosystem monitoring. Mesocosms as a tool to measure limiting nutrients in benthic microalgae I investigated two methods of evaluating potential nutrient limitation in BMA: dissolved nutrient availability in the water column and sediment nutrient addition bioassays. Dissolved nutrient availability was more representative of potential limiting nutrients in central Florida Bay – where dissolved P is nearly always less abundant than N and most limiting. Despite removing the pelagic phytoplankton community, it quickly reformed. The water column community responded more quickly to nutrient additions to the sediment than the BMA community did in the summer experiment. This water column community was at first P limited and then shifted to N limited, as more P was added to the sediment and presumeably made available to the water column by benthic flux. However, P flux to the water column was not detectable at a level significantly different among treatments in these mesocosms from central Florida Bay either in the summer or the winter. The BMA community was always P limited in central Florida Bay. Total dissolved nutrient availability was not indicative of the potential for nutrient limitation in western Florida Bay. Total dissolved nitrogen and NH4 were abundant in comparison to P. Although measureable NO3 was absent, all forms of P were present in low but measureable amounts. Nutrients seemed to not be affecting western Florida Bay BMA chlorophyll a concentrations so much as some other untested constituent, like light or temperature, in the winter experiment. Nitrogen addition treatments suffered less

165 chlorophyll a loss than P addition treatments, which may support N limitation as a cofactor. Bioassays using natural populations of BMA would be an important tool to include in monitoring programs and would better predict nutrient limitation of the BMA community than dissolved nutrient availability. Nutrient preferences in both pelagic and benthic microalgae discount the exclusive use of dissolved nutrient availability in determining potential nutrient limitation of a system. The results of my benthic nutrient flux measurements indicate the sediment and the associated BMA can be both source and sink for dissolved nutrients in the system. Radiolabelled tracers of P flux through benthic microalgae The use of 33P was useful in tracing P flux trends from the sediment to the water column. Fickian diffusion across a killed BMA layer took between 1 and 5 hours to complete equilibrium. The trend of BMA enhancing P flux to the water column beyond simple Fickian diffusion was clear in live versus killed treatments, although the mechanism is unknown. P flux from the sediment occurred almost immediately and there was sometimes subsequent loss of P to either the vessel walls or the sediment/BMA complex . There were differences among light and dark treatments suggesting that light- dependent reactions somehow affect 33P flux. The magnitude of P flux was different among the two field locations, but was consistent with dissolved P concentration at the sites as measured during the sediment flux studies (i.e., more 33P flux in the western site than in the central location). This method posed a hindrance to drawing conclusions on BMA mediation of P flux from the sediment due to the lack of a total P measurement in the sediment and water column pools. However, the positive trend in the total dissolved P pools between the sediment and the water column, which I did measure, supports BMA influence on P flux from the sediment. Measuring primary production in benthic microalgae

My measurements of primary production, using dissolved O2 evolution and PAM fluorometry indicate that BMA are a source of primary production to the Florida Bay

166 system, although far less than seagrass. Areal production estimates range from ~400 and -1 800 metric tons of C day , based upon O2 evolution and PAM fluorometry, respectively. These results are similar to BMA production estimates from other parts of the world, but less than muddy sediment in Florida Bay. Those areas of Florida Bay with BMA habitat would be net autotrophic based upon my measurements, however, I may have underestimated respiration. Both methods were measured during sunlight hours, which would not take in to account diel changes in production and respiration due to physiological differences. Based upon PAM fluorometry results, BMA in the western and central Florida Bay are low light adapted. Peak production (ETR) occurred between 200 and 500 Einsteins m-2 s-1, or about 20% of surface irradiance. However, diel differences were found in production, with either photoinhibition or downregulation of ETR (in response to high light) occurring in the mid-day at all locations. The southeastern site, although only sampled a few times, appeared not to reach maximum photosynthetic efficiency, even when exposed to high light, suggesting BMA were adapted to high light conditions or some other factor was limiting ETR at this location. PAM fluorometry was much easier and provided much more information on the

BMA photosynthetic capacity than O2 evolution measurements. The ability to avoid problems with migration and photoadaptation that are associated with traditional light curves and its portability and lack of preparation make PAM very attractive for field use. The newness of the technology, difficulty interpreting the information from the instrument, and the potential for species specific differences currently hinder the widespread use of PAM fluorometry as a means of measuring production. Comparisons 14 with traditional, long-used C and O2 evolution methods are only now being undertaken, so this method will probably not be useful for making ecosystem management decisions for several more years. Everglades restoration efforts If Everglades restoration were to be completed there would be changes to the Florida Bay system as a result. With an increase in freshwater would come the delivery of more nutrients, lowered salinity, and decreased residence time within the mud basins.

167 The decreased residence time might result in greater exports of production out of the Florida Bay system. Monitoring of the microphytobenthos would enable managers to evaluate the consequences of those changes upon the benthic community production, biomass and sediment nutrient flux. Such changes might be anticipated to include uptake of the nutrients by the BMA and perhaps a shift away from net autotrophy to net heterotrophy in more basins. However, if the water column biomass were to increase as a result of increased nutrient flux, there might ultimately be decreased BMA production as a result of light limitation. This could ultimately end with a complete decoupling of the benthic – pelagic interchange of production and nutrient exchange in favor of phytoplankton dominated system. Future work In the relatively short period of time that Florida Bay has been investigated, researchers have developed a better understanding of the oceanographic processes. As with any problem or question, the investigation leads us to more questions. My own future questions that have evolved from my work include conducting more sediment mesocosm bioassays in the bay. I would also like to develop a better tracer of nutrients through the sediment porewater/BMA/water column pools. Those BMA living in the hypersaline lakes within some of the mangrove islands in Florida Bay have not been investigated for biomass, production to higher trophic levels living within the pools, or community structure. These could be the most extreme BMA environments found in Florida Bay. Most interesting to me is the influence of wave and tidal pumping and sediment microtopography on the concentration or advection of nutrient porewaters and probably BMA patchiness. This final topic would perhaps be the most applicable, and potentially fundable, aspect of the coastal nutrification problem I would like to pursue.

168

Literature Cited

Admiraal, W. 1977. Experiments with mixed populations of benthic estuarine diatoms in laboratory microecosystems. Botanica Marina. 20: 479-485.

Aller, R.C., Y.Y. Aller and P.F. Kemp. 2001. Effects of particle and solute transport on rates and extent of remineralization in bioturbated sediments. pP 315-333 In: Organism – Sediment Interactions. Eds J. Y. Aller, S.A. Woodin, and R.C. Aller. University of South Carolina Press.

Amspoker, M. C., and C. D. McIntire. 1978. Distribution of intertidal diatoms associated with sediments in Yaquina estuary, Oregon. Journal of Phycology 14: 387-395.

Baillie, P. W. and B. L. Welsh. 1980. The effect of tidal resuspension on the distribution of intertidal epipelic algae in an estuary. Estuarine and Coastal Marine Science 10: 165-180.

Beer, S. and M. Bjork. 2000. Measuring rates of photosynthesis of two tropical seagrasses by pulse amplitude modulated (PAM) fluorometry. Aquatic Botany. 66:69-76.

Brand, L.E, and M.S. Ferro. 2001. Nutrient ratios and the eutrophication of Florida Bay. Abstract In 2001 Florida Bay Science Conference, April 23-26, 2001, Key Largo, FL.

Butler, M.J. IV, J.H. Hunt, W.F. Herrnkind, T. Matthews, M. Childress, R.Bertelsen, W. Sharp, J.M. Field, and H. Marshall. 1995. Cascading disturbances in Florida Bay, USA: cyanobacteria blooms, sponge mortality, and implications for juvenile spiny lobster Panulirus argus. Marine Ecology Progress Series 129: 119-125

Cahoon, L. B. 1992. Total chlorophyll in Onslow Bay, North Carolina: field observations vs. predictions of a total chlorophyll algorithm. Journal of the Elisha Mitchell Scientific Society 108: 91-101.

Cahoon, L. B., and J. E. Cooke. 1992. Benthic microalgal production in Onslow Bay, North Carolina, USA. Marine Ecology Progress Series 84: 185-196.

Cahoon, L.B. 1999. The role of benthic microalgae in neritic ecosystems. Oceanography and Marine Biology Annual Review. 37:47-86.

169 Cahoon. L.B. 2006. Upscaling primary production estimates: Regional and global scale estimates of microphytobenthos production. Pp 99-108. In: Functioning of microphytobenthos in estuaries. J.C. Kromkamp, J.F.C. De Brouwer, G. F. Blanchard, R.M. Forster, and V. Creach (eds). Royal Netherlands Academy of Arts and Science.

Caraco, N. 1988. What is the mechanism behind the seasonal switch between N and P limitation in estuaries? Canadian Journal of Fisheries and Aquatic Sciences 45: 381-382

Caraco, N.F. J. J. Cole, and G. E. Likens. 1992. New and recycled primary production in an oligotrophic lake: Insights for summer phosphorus dynamics. Limnology and Oceanography. 37(3), 1992, 590-602

Carlson, P. R., Jr., Durako, M. J., Barber, T. R., Yarbro, L. A., deLama, Y., and Hedin, B. 1990. Catastrophic mortality of the seagrass Thalassia testudinum in Florida Bay. Annual project report. St. Petersburg, Florida: Florida Department of Natural Resources.

Charphy- Roubaud C.J. and A. Sournia. 1990. The comparative estimation of phytoplanktonic microphytobenthic production in the oceans. Marine Microbial Food Webs 4:31-51

Charpy, L., and C. J. Charpy-Roubaud. 1990. A model of the relationship between light and primary production in an atoll lagoon. Journal of the Marine Biological Association of the United Kingdom 70: 357-369.

Colijn, F. a. G. van Buurt. 1975. Influence of light and temperature on the photosynthetic rate of marine benthic diatoms. Marine Biology 31: 209-214.

Colijn, F., and V. N. de Jonge. 1984. Primary production of the microphytobenthos in the Ems-Dollard estuary. Marine Ecology Progress Series. 14: 185-196.

Cooke, J. E., 1991. The effects of ultraviolet radiation B on the diatom Nitschia closterium. M. Sc. Thesis. University of North Carolina, Wilmington.

Cornwell, J.C., M. Owens and W.M. Kemp. 2001 Nitrogen cycling in Florida Bay sediments. Abstract In: 2001 Florida Bay Conference, Key Largo FL, Spring 2001

Darrow, B.P., J.J. Walsh, G.A. Vargo, R.T. Masserini jr., K.A. Fanning, and J.Z. Zhang. 2003 A simulation study of the growth of benthic microalgae following the decline of a surface phytoplankton bloom. Continental Shelf Research 23: 1265- 1283 de Jonge, V. N. 1985. The occurrence of epipsammic diatom populations: A result of interaction between physical sorting of sediment and certain properties of diatom species. Estuarine Coastal and Shelf Science. 21:607-622. 170 D'Elia, C. F., J. G. Sanders, and W. R. Boynton. 1986. Nutrient enrichment studies in a coastal plain estuary: phytoplankton growth in large-scale continuous cultures. Canadian Journal of Fisheries and Aquatic Sciences 43: 397-406

Durako, M. and J. Zieman. 2007. Chapter 7: Seagrass Ecology. Pp 92-109. In: Florida Bay Science Program: A Synthesis of Research on FloridaBay. Hunt J.H., Nuttle, W. (eds.), Fish Wildlife Res. Inst. Tech Rept. TR-11, 148 pp.

Elser, J. J., M. M. Elser, N. A. MacKay, and S. R. Carpenter. 1988. Zooplankton- mediated transitions between N- and P-limited algal growth. Limnology and Oceanography 33:1-14

Fenchel, T. and L. H. Kofoed. 1976. Evidence for Exploitative Interspecific Competition in Mud Snails (Hydrobiidae). Oikos, 27(3): 367-376.

Forster, R.M. and J.C. Kromkamp. 2006. Estimating benthic primary production: scaling up from point measurements to the whole estuary. Pp 108-120. In: Functioning of microphytobenthos in estuaries. J.C. Kromkamp, J.F.C. De Brouwer, G. F. Blanchard, R.M. Forster, and V. Creach (eds). Royal Netherlands Academy of Arts and Science.

Fourqurean, J. W., Jones, R. D., and Zieman, J. C. 1993. Processes influencing water column nutrient characteristics and phosphorus limitation of phytoplankton biomass in Florida Bay, Florida, USA: inferences from spatial distributions. Estuarine, Coastal and Shelf Science. 36:295-314.

Fourqurean, J. W., Zieman, J. C., and Powell, G. V. N. 1992. Phosphorus limitation of primary production in Florida Bay: evidence from C:N:P ratios of the dominant seagrass Thalassia testudinum. Limnology and Oceanography. 37(1):162-171.

Gordon, L. I., Jennings, J. C. Jr., Ross, A. A., and Krest, J. M. 1993. A suggested Protocol For Continuous Flow Automated Analysis of Seawater Nutrients. In: WOCE Operations Manual. WHP office Report 90-1, WOCE report 77 No. 68/91. 1-52.

Grant, J. and G. Gust. 1987. Prediction of coastal stabilization from photopigment content of mats of purple sulfur bacteria. Nature 330: 244-246.

Halley, R. B, E.J. Prager, R.P Stumpf, K. K. Yaters and C. H. Holmes, 1999. Sea-level rise and the future of Florida Bay in the next century. Abstract In Proceedings of the South Florida Restoration Science Forum, May 17-19, 1999, Boca Raton, FL.

Hancke, K., T.B. Hancke, L. M. Olsen, G. Johnsen, R.N. Glud, 2008. Temperature effects on microalgal photosynthesis-light responses measured by O2 production, pulse-amplitude-modulated fluorescence, and 14C assimilation. Journal of Phycology. 44(2): 501-514.

171 Harrison, W. D., D. Musgrave and W.S. Reeburgh. 1983. A wave-induced transport process in marine sediments. Journal of Geophysical Research. 88(c12): 7617- 7622.

Heil, C.A. K. Chaston, S. Costanza, B. Longstaff, and W.C. Dennison. 2004. Benthic microalgae in coral reef sediments of the southern Great Barrier Reef, Australia. Coral Reefs. 23:336-343.

Hoare, A. M. D. J. Hollander, C. A. Heil, S. Murasko, P. M. Glibert, M. Revilla, and J. Alexander. 2005. Isotopic fingerprinting of nutrient sources and biological sinks in Florida Bay: A geochemical tool for evaluating ecosystem response to changing nutrient inputs . Abstract in 2005 Florida Bay and Adjacent Marine Systems Science Conference, Duck Key, FL.

Howarth, R. W. 1988. Nutrient limitation of net primary production in marine ecosystems. Annual Review of Ecology 19: 89-110

Huettel, M. and G. Gust. 1992. Impact of bioroughness on interfacial solute exchange in permeable sediments. Marine Ecology Progress Series. 89:253-267.

Jensen, HS, KJ McGlathery, R Marino & RW Howarth. 1998. Forms and availability of sediment phosphorus in carbonate sand of Bermuda seagrass beds. Limnology and. Oceanography. 43(5), 799-810

Jurado, J. and G.L. Hitchcock. 2001. Development of a silica budget for northwestern Florida Bay. Abstract in 2001 Florida Bay Science Conference, April 23-26, 2001, Key Largo, FL.

Keizer, P. D., B. T. Hargrave, and J. Gordon, D. C. 1989. Sediment-water exchange of dissolved nutrients at an intertidal site in the upper reaches of the Bay of Fundy. Estuaries 12: 1-12.

Kelble, C.R., Ortner, P.B., Hitchcock, G.L. and Boyer, J.N. 2005. Attenuation of photosynthetically available radiation (PAR) in Florida Bay: Potential for light- limitation of primary producers. Estuaries. 28(4):560-571.

Kirchstein, J.D., J.P. Zehr, H. Paerl. 1993. Determination of Nitrogen Fixation Potential in the marine environment: application of the polymerase chain reaction. Marine Ecology Progress Series, 95:305-309.

Koch, M. S. Schopmeyer. C. Khyhn-Hansen, O. Nielsen, C. Madden. 2005. A conceptual model for seagrass die-off in Florida Bay based upon mesocosm and field experiments. Abstract in: Florida Bay and Adjacent Marine Systems Science Conference, Dec. 11-14, 2005, Duck Key, FL

172 Kroeger, K.D., P.W. Swarzenski, W.J. Greenwood, and C. Reich. 2007. Submarine groundwater discharge to Tampa Bay: Nutrient fluxes and biogeochemistry of the coastal aquifer. Marine Chemistry, 104(1-2):85-97.

Krom, M. D. and R. A. Berner. 1981. The diagenesis of phosphorus in a nearshore marine sediment. Geochimica Cosmochima Acta 45: 207-216

Lapointe, B. E. 1989. Macroalgal production and nutrient relations in oligotrophic areas of Florida Bay. Bulletin of Marine Science. 44(1):312-323.

Lawlor, David W. 2001. Photosynthesis. Springer- Verlag, New York. 386p.

MacIntyre, H. I., R. J. Geider, and D. G. Miller. 1996. Microphytobenthos: The ecological role of the 'secret garden' of unvegetated, shallow-water marine habitats. I. Distribution, abundance and primary production. Estuaries 19: 186- 201.

MacIntyre, H. L., and J. J. Cullen. 1995. Fine-scale vertical resolution of chlorophyll and photosynthetic parameters in shallow-water benthos. Marine Ecology Progress Series 122: 227-237.

Madden, C. A. MacDonald, 2007. A modeling study of alternate stable states in Florida Bay. Abstract in Proceeding of ERF 2007, Nov 3-9, 2007 Providence RI.

Madden, Christopher J.and Amanda A. McDonald. 2005. The Florida Bay Seagrass Model: Examination of Fresh Water Effects on Seagrass Ecological Processes, Community Dynamics and SeagrassDie-off. Abstract in Florida Bay Science Conference, Duck Key, FL

Malan, D. E. and A. McLachlan. 1991. In situ benthic oxygen fluxes in a near shore coastal marine system: a new approach to quantify the effect of wave action. Marine Ecology Progress Series. 73:69-81.

Malone, T. C., C. Garside, and P. J. Neale. 1980. Effects of silicate depletion on photosynthesis by diatoms in the plume of the Hudson River. Marine Biology 58: 197-204

Masojidek, J. J. U. Grobbelaar, L. Pechar, and m. Koblizek. 2001. Photosystem II electron transport rates and oxygen production in natural waterblooms of freshwater cyanobacteria during a diel cycle. Journal of Plankton Research 23(1): 57-66.

Millero, FJ. 2000 The carbonate system in seawater. In Chemical Processes in Marine Environments, A. Gianguzza, E. Pellizzetti and S. Sammartano (eds.), Springer- Verlag, Berlin, Ch. 1, 9-41.

173 Mills, D. K. and. M. Wilkinson. 1986. Photosynthesis and light in estuarine benthic microalgae. Botanica Marina 29: 125-129.

Moncreiff, C. A. 1992. Primary production dynamics in seagrass beds of Mississippi Sound: the contribution of seagrass, epiphytic algae, sand microflora, and phytoplankton. Marine Ecology Progress Series 87: 167-171.

Necchi, O. 2004. Photosynthetic responses to temperature in tropical lotic macroalgae. Phycological Research, 52(2):140-148.

Nelson, J. R., J. E. Eckman, C. Y. Robertson, R. L. Marinelli, and R. A. Jahnke. 1999. Benthic microalgal biomass and irradiance at the sea floor on the continental shelf of the South Atlantic Bight: spatial and temporal variability and storm effects. Continental Shelf Research 19: 477-505.

Nielsen, O., M. Koch,H.S. Jensen and C.J. Madden, 2005, Inorganic Phosphate Kinetics in Florida Bay Seagrass Ecosystems:33P Experiments and In Situ Chamber Studies. Abstract in Florida Bay Science Conference, Duck Key, FL

Nilsson, C., and K. Sundback. 1991. Growth and nutrient uptake studied in sand-agar microphytobenthic communities. Journal of Experimental Marine Biology and Ecology. 153: 207-226.

Nixon, S. W., J. R. Kelly, B. N. Furnas, C. A. Oviatt, and S. S. Hale. 1980. Phosphorus regeneration and the metabolism of coastal marine bottom communities, pps. 219- 242, In Marine Benthic Dynamics. K. R. Tenore and B. C. Coull (eds), Univ. of South Carolina.

Officer, C. B. and J. H. Ryther. 1980. The possible importance of silicon in marine eutrophication. Marine Ecology Progress Series 3: 83-91

Pennock, J. R. and J. H. Sharp. 1994. Temporal alternation between light- and nutrient- limitation of phytoplankton production in a coastal plain estuary. Marine Ecology Progress Series 111: 275-288

Phlips, E. J., and Badylak, S. 1996. Spatial variability in phytoplankton standing crop and composition in a shallow inner-shelf lagoon, Florida Bay, Florida. Bulletin of Marine Science. 58(1):203-216.

Phlips, E. J., Lynch, T. C., and Badylak, S. 1995. Chlorophyll a, tripton, color, and light availability in a shallow tropical inner-shelf lagoon, Florida Bay. Marine Ecology Progress Series. 127:223-234.

Pinckney JL and Zingmark RG (1993) Modeling the annual production of intertidal benthic microalgae in estuarine ecosystems. Journal of Phycology 29: 396 –407

174 Pinckney, J., H. W. Paerl, M. Fitzpatrick. 1995. Impacts of seasonality and nutrients on microbial mat community structure and function. Marine Ecology Progress Series 123: 207-216.

Prager E. J. and R. B. Halley. 1997. Bottom types of Florida Bay: U.S. Geological Survey Open-File Report 97-256.

Price, R. M., P. K. Swart and J. W. Fourqurean 2005. Terrestrial Brackish Groundwater Discharge – a Significant Flux of Phosphorus to the Mangrove-Estuarine Transition Zone of the Southern Everglades. Abstract in 2005 Florida Bay and Adjacent Marine Systems Science Conference, Duck Key, FL.

Revsbech, N. P., B. B. Jorgensen, and O. Brix. 1981. Primary production of microalgae 14 in sediments measured by oxygen microprofile, H CO3-fixation, and oxygen exchange methods. Limnology and Oceanography. 266(4):717-730.

Rizzo, W. M. 1990. Nutrient exchanges between the water column and a subtidal benthic microalgal community. Estuaries 13: 219-226

Rizzo, W. M., and R. L. Wetzel. 1985. Intertidal and shoal benthic community metabolism in a temperate estuary: studies of spatial and temporal scales of variability. Estuaries 8: 342-351.

Robson, B. J., P.A. Bukaveckas, and D.P. Hamilton. 2008, Modeling and mass balance assessments of nutrient retention in a seasonally-flowing estuary (Swan River Estuary, Western Australia) Estuarine, coastal and shelf science, 76(2): 282-292.

Rudnick, D., S. Kelly, C. Donovan, K. Picard, J. Cornwell, and M. Owens. 1999. Benthic nutrient fluxes near Florida Bay's mangrove ecotone. Abstract in: Proceedings of the 1999 Florida Bay and Adjacent Marine Systems Science Conference, November 1-5, 1999, Key Largo, FL.

Ryther, J.H. and W.M. Dunstan. 1971. Nitrogen, phosphorus and eutrophication in the coastal marine environment. Science 171: 1008-1013.

Sndback K. and K. McGlathery 2005. Interactions between benthic macroalgal and microalgal mats, pp 7-30, In Coastal and Estuarine Studies: Interactions between macro- and microorganisms in marine sediments, E. Kristensen, RR. Haese, and J.E. Kostka (eds.) American Geophysical Union.

Sndback, K., and W. Graneli. 1988. Influence of microphytobenthos on the nutrient flux between sediment and water: a laboratory study. Marine Ecology Progress Series 43: 63-69.

Saroussi, S. and S. Beer. 2007. Alpha and quantum yield of aquatic plants derived from PAM fluorometry: Uses and misuses. Aquatic Botany. 86:89-92

175 Schmitt-Jansen, M. and R. AltenBurger. 2008. Community-level microalgal toxicity assessment by multiwavelength-excitation PAM fluorometry. Aquatic Toxicology. 86(1): 49-58.

Schomer, N. S., and Drew, R. D. 1982. An ecological characterization of the lower Everglades, Florida Bay and the Florida Keys. Washington, D. C.: U. S. Fish and Wildlife Service, Office of Biological Services.

Seitzinger, S. P. 1988. Denitrification in freshwater and coastal marine ecosystems: ecological and geochemical significance. Limnology and Oceanography 33: 702- 724

Serôdio J, Marques da Silva J and Catarino F (1997) Nondestructive tracing of migratory rhythms of intertidal benthic microalgae using in vivo chlorophyll a fluorescence. Journal of Phycology 33: 542–553.

Serodio, J. S. Vieira, S. Cruz and F. Barrosso. 2005. Short-term variability in the photosynthetic activity of microphytobenthos as detected by measuring rapid light curves using variable fluorescence. Marine Biology. 146:903-914.

Serodio, J., S. Vieira, S. Cruz and H. Coelho. 2006. Rapid light-response curves of chlorophyll fluorescence in microalgae: relationship to steady-state light curves and non-photochemical quenching in benthic diatom-dominated assemblages. Photosynthesis Research. 90:29-43.

Shoaf, W.T. and B.W. Lium. 1976. Improved extraction of chlorophyll a and b from algae using dimethyl sulfoxide. Limnology and Oceanography. 21:926-928.

Smith, S. M. and G. L. Hitchcock. 1994. Nutrient enrichments and phytoplankton growth in the surface waters of the Louisiana Bight. Estuaries. 17(4):740-753.

Smith, T. J., III, Hudson, J. H., Robblee, M. B., Powell, G. V. N., and Isdale, P. J. 1989. Freshwater flow from the Everglades to Florida Bay: a historical reconstruction based on fluorescent banding in the coral Solenastrea bournoni. Bulletin of Marine Science. 44(1):274-282.

Solorzano, L. and J. H. Sharp. 1980. Determination of total dissolved phosphorus and particulate phosphorus in natural waters. Limnology and Oceanography. 25(4): 754-758.

Sorokin, Y. I. 1992. Phosphorus metabolism in coral reef communities: exchange between the water column and bottom biotopes. Hydrobiologia 242: 105-114.

Steeman-Nielsen, E. (1951) Measurement of production of organic matter in sea by means of carbon-14. Nature 267, 684–685.

176 Steidinger, K. A. S. Lukas-Black, S. Richards, B. Richardson, and G. McRae 1998. Florida Bay Microalgae. Abstract in Proceedings of the 1998 Florida Bay Science Conference. Miami, FL.

Steidinger, K. A., B. Richardson, M.B. Neely, G .McRae, S. Richards, R. Bray, T.H. Perkins, C. Tomas. 2001. Florida Bay Microaglal Blooms. Abstract in Florida Bay Science Conference, April 23-26, 2001, Key Largo, FL.

Sundback, K., Enoksson, V., Graneli, W., Pettersson, K. 1991. Influence of sublittoral rnicrophytobenthos on the oxygen and nutrient flux between sedunent and water: a laboratory continuous-flow study. Marine Ecology Progress Series 74: 263-279

Thibodeaux, L. J. and J. D. Boyle. 1987. Bedform-generated convective transport in bottom sediment. Nature. 325 : 341-343.

Tomas C.R., B. Bendis and K. Johns. 1999. Role of nutrients in regulating plankton blooms in Florida Bay. 323-337 In Gulf of Mexico Large Marine Ecosystem, H. Kumpf. K. Steidinger, and K. Sherman (eds) Blackwell Science, 703p.

Tomas, C.R. and B. Bendis. 1995. Phytoplankton dynamics studies in Florida Bay. 1995. Abstract in Florida Bay Science Conference, October 17 & 18, 1995, Gainesville, FL.

Ulthicke, S., and D. W. Klumpp. 1997. Ammonium excretion by holothurians enhances production and turnover of benthic diatom communities. Abstract in Proceedings of the 8th International Coral Reef Symposium I: 873-876.

Underwood. G. C. and M. Barnett. 2006. What determines species composition in microphytobenthic biofilms? pp 123-140. In: Functioning of microphytobenthos in estuaries. J.C. Kromkamp, J.F.C. De Brouwer, G. F. Blanchard, R.M. Forster, and V. Creach (eds). Royal Netherlands Academy of Arts and Science.

Valderama, J. C. (1981). The simultaneous analysis of total nitrogen and total phosphorus in natural waters. Marine Chemistry 16: 109-122.

Vargo, G. A., M.B. Neely, K.L. Lester, G. L. Hitchcock and J. Jurado. 2001b. Spatial and temporal changes in phytoplankton biomass, proximate composition, and total dissolved nitrogen and phosphorus based on bimonthly cruises throughout western Florida Bay. Abstract in 2001 Florida Bay Science Conference, April 23- 26, 2001, Key Largo, FL.

Vargo, G.,G. Hitchcock, M. B. Neely, J. Jurado, L. Melahn and D. Mir. 1999. Estimates of phytoplankton growth, production, and nutrient requirements based on a drifter tracked water parcel in western Florida Bay, USA. Abstract in ASLO 1999, Santa Fe NM.

177 Vargo. G.A., M.B. Neely, G.L. Hitchcock and J. Jurado. 2001a. A summary of results from drifter and fixed location house boat based studies: growth rates, production, proximate composition and nutrient requirements of phytoplankton populations during blooms in Northwestern and South-central Florida Bay. Abstract in 2001 Florida Bay Science Conference, April 23-26, 2001, Key Largo, FL.

Walz, H. 1998. Underwater fluorometer DIVING-PAM handbook of operation. Germany. 101p.

Welsh, B. L. 1980. Comparative nutrient dynamics of a marsh-mudflat ecosystem. Estuarine and Coastal Marine Sciences 10: 143-164.

Whitney, D. E. and W. M. Darley. 1978. A method for the determination of chlorophyll a in samples containing degradation products. Limnology and Oceanography 24: 183-186.

Yallop, M. L., B. de Winder, D. M. Paterson, and L. J. Stal. 1994. Comparative structure, primary production and biogenic stabilization of cohesive and non-cohesive marine sediments inhabited by microphytobenthos. Estuarine, Coastal and Shelf Science. 39: 565-582.

Yarbro L.A. and P.R.Carlson. 2008 . Benthic oxygen and nutrient fluxes in seagrass communities of Florida Bay, USA. Estuarine, Coastal and Shelf Science

Yates, K. and B. Halley. 2001. Geochemical Monitoring of Productivity in Florida Bay, Abstract in Proceedings of the 2001 Florida Bay Science Conference, April 23- 26, 2001. Westin Beach Resort. Key Largo, Florida.

Zeitzschel, B. 1980. Sediment-water interactions in nutrient dynamics. p. 195-218. In K. R. Tenore and B. C. Coull (eds.), Marine Benthic Dynamics. University of South Carolina Press, Columbia.

178

Bibliography

Cahoon, L. B., R. A. Laws, and T. W. Savidge. 1992. Characteristics of benthic microalgae from the North Carolina outer continental shelf and slope: preliminary results. Diving for Science. : 61-68.

Cosgrove, J. and M. Borowitzka. 2006. Applying Pulse Amplitude Modulation (PAM) fluorometry to microalgae suspensions: stirring potentially impacts fluorescence. Photosynthesis Research, 88(3):343-350.

Dring, M. J., and K. Lning. 1994. Influence of spring-neap tidal cycles on the light available for photosynthesis by benthic marine plants. Marine Ecology Progress Series 104: 131-137.

Eijsink, L. M., M. D. Krom, and G. J. de Lange. 1997. The use of sequential extraction techniques for sedimentary phosphorus in eastern Mediterranean sediments. Marine Geology 139: 147-155.

Eppley, R. W. 1972. Temperature and phytoplankton growth in the sea. Fishery Bulletin 70:1063-1085

Falkowski, P.G., R. Geider and J. A. Raven. 2004. The role of aquatic photosynthesis in solar energy conversion: a geoevolutionary approach. P 287-322. In; Molecular to Global Photosynthesis. Mary D. Archer, James Barber, (eds.), Imperial College Press.

Fenchel, T. 1988. Marine plankton food chains. Annual Review of Ecological Systems. 19: 19-38.

Fisher, T. R., A. B. Gustafson, K. Sellner, R. Lacuture, L. W. Haas, R. Magnien, and B. Michael. 1999. Spatial and temporal variation in resource limitation in Chesapeake Bay. Marine Biology 133: 763-778

Fisher, T. R., E. R. Peele, J. W. Ammerman, and L. W. Harding. 1992a. Nutrient limitation of phytoplankton in Chesapeake Bay. Marine Ecology Progress Series 82: 51-63

Fourqurean, J.W. and J.C. Zieman. 1991. Photosynthesis, respiration and whole plant carbon budgets of Thalassia testudinum, Halodule wrightii and Syringodium filiforme. pp 59-70. In The light requirements of seagrasses: proceedings of a workshop to examine the capability of water quality criteria, standards and monitoring programs to protect seagrasses. Kenworthy, W.J. and D.E. Haunert (eds.) NOAA Technical Memorandum NMFS-SEFC-287. 179 Harrison, P. J., M. H. Hu, Y. P. Yang, and X. Lu. 1990. Phosphate limitation in estuarine and coastal waters of China. Journal of Experimental Marine Biology and Ecology 140: 79-87

Hillebrand, H. and U. Sommer. 1997. Response of epilithic microphytobenthos of the Western Baltic Sea to in situ experiments with nutrient enrichment. Marine Ecology Progress Series. 160:35-46.

John H. Paul, Molly R. McLaughlin, Dale W. Griffin, Erin K. Lipp, Rodger Stokes, and Joan B. Rose, 2000. Rapid Movement of Wastewater from On-site Disposal Systems into Surface Waters in the Lower Florida Keys. Estuaries, 23(5):662- 668.

Juneau P. and P. J. Harrison, 2008, Comparison by PAM Fluorometry of Photosynthetic Activity of Nine Marine Phytoplankton Grown Under Identical Conditions Photochemistry and Photobiology: 649-653.

Karl, D. M. a. G. T. 1992. MAGIC: a sensitive and precise method for measuring dissolved phosphorus in aquatic environments. Limnology and Oceanography 37: 105-116.

Kelble, C.R., G.L. Hitchcock, P.B. Ortner, J.N. Boyer, 2003. A recent study of the light environment in Florida Bay. Abstract in 2003 Joint Conference on the Science and Restoration of the Greater Everglades and Florida Bay Ecosystems. April 13- 18, 2003, Palm Harbor, FL

Kromkamp, J., J. Peene, P. van Rijswijk, and N. Goosen. 1995. Nutrients, light and primary production by phytoplantkon and microphytobenthos in the eutrophic, turbid Westerschelde estuary (The Netherlands). Hydrobiologia 311: 9-19.

MacIntyre, H. L., and J. J. Cullen. 1996. Primary production by suspended and benthic microalgae in a turbid estuary: time-scales of variability in San Antonio Bay, Texas. Marine Ecology Progress Series 145: 245-268.

Naoshige G., M. Kihira and N. Ishida. 2008. Seasonal distribution of photosynthetically active phytoplankton using pulse amplitude modulated fluorometry in the large monomictic Lake Biwa, Japan. Journal of Plankton Research, 30(10):1169-1177.

Owens, M. S. and J. C. Cornwell. 2001. Nitrogen cycling in Florida Bay mangrove environments: Sediment-water exchange and denitrification. p. 90. Abstract in: Proceedings of the 2001 Florida Bay Science Conference, April 23-26, 2001, Key Largo, FL.

Pichard SL, L. Campbell, J.B.Kang, F.R. Tabita, and J. H. Paul. 1996. Regulation of ribulose bisphosphate carboxilase gene expression in natural phytoplankton communities. I. Diel rhythms. Marine Ecology Progress Series. 139:257–265.

180 Riaux-Gobin, C., and B. Klein. 1993. Microphytobenthic biomass measurement using HPLC and conventional pigment analysis. Handbook of Methods in Aquatic Microbial Ecology. : 369-376.

Rougerie, F., C. Jehl, and J. Trichet. 1997. Phosphorus pathways in atolls: interstitial nutrient pool, cyanobacterial accumulation and Carbonate-Fluoro-Apatite (CFA) precipitation. Marine Geology 139: 201-217.

Sorokin, Y. I. 1991. Parameters of productivity and metabolism of coral reef ecosystems off central Vietnam. Estuarine, Coastal and Shelf Science 33: 259-280.

Strickland, J.D.H., and Parsons, T.R. (1968). A Practical Handbook of Seawater Analysis. Fisheries Research Board of Canada Bulletin, 167p.

Stumpf, R. P. a. M. L. F. 1997. Temporal and spatial change in coastal ecosystems using remote sensing. Example with Florida Bay, USA, emphasizing AVHRR imagery. Pages 65-73. In The Fourth International Conference on Remote Sensing for Marine and Coastal Environments. Environmental Research Institute of Michigan, Orlando, Florida, USA.

Webb, K. L. 1988. Comment on "Nutrient limitation of phytoplankton growth in brackish coastal ponds" by Caraco, Tamse, Boutros, and Valiela (1987). Canadian. Journal of Fisheries and Aquatic Science. 45: 380-381

Yarbro, L. A. and P. R. Carlson. 1999. Nutrient flux at the sediment – water interface in Florida Bay. p. 106-107. In Proceedings of the 1999 Florida Bay and Adjacent Marine Systems Science Conference, Miami, FL.

Yates, K. and B. Halley. 2000. Geochemical Monitoring of Productivity in Florida Bay, In Proceedings of the 2000 Greater Everglades Environmental Restoration Conference, April 23-26, 2001. Innisbrook Resort. Tarpon Springs, Florida.

181

Appendices

182 Appendix A ANOVA F-test results from the mesocosm experiments by date.

Central TDP (M) SRP (M) TDN (M) NO3+NO2 (M) NO2 (M) NH4 (M) SIL (M) PO4 (M) Benthic Chl a Water Chl a 8/4/2002 0.113 0.011* 0.641 0.51 0.52 0.801 0.477 0.558 0.215 ND 8/5/2002 0.619 0.444 0.643 0.778 0.789 0.565 0.87 0.785 0.311 0.398 8/6/2002 0.126 0.004** 0.16 0.013* 0.11 0.002** 0.148 0.388 0.477 0.297 8/7/2002 0.251 0.003** 0.004 ** 0.344 0.636 0.0003*** 0.256 0.317 0.032* 0.139 8/8/2002 0.796 0.022* 0.008** 0.005** 0.008** 0.002** 0.256 0.455 0.09+ 0.1456 8/9/2002 ND 0.133 0.0002*** 0.39 0.427 0.175 0.42 0.728 0.193 0.866 8/10/2002 ND 0.003** ND ND ND ND ND ND 0.328 0.065+ 8/11/2002 ND ND ND ND ND ND ND ND NV 0.013 8/12/2002 ND ND ND ND ND ND ND ND NV 0.065 Western 1/25/2003 0.32 0.135 0.934 0.288 0.475 0.961 0.741 0.365 2.3 x 10 -9*** 0.586 1/26/2003 3.18 x 10-5*** 1.16 x 10-8*** ND ND ND ND ND ND 3.83 x 10-15*** 0.102+ 1/27/2003 4.42 x 10-5*** 8.38 x 10-8*** ND ND ND ND ND ND 3.98 x 10-8*** 0.399 1/28/2003 0.000145*** 3.27 x 10-8*** ND ND ND ND ND ND 4.58 x 10-13*** 0.345 1/29/2003 0.02** 1.87 x 10-8*** ND ND ND ND ND ND NV 0.71 1/30/2003 ND 1.31 x 10-7*** 0.068+ 0.028* 0.776 0.998 0.85 0.001*** NV 0.55 Central 1/25/2003 0.206 0.116 0.849 0.525 0.128 0.826 0.433 0.378 0.083+ 0.848 1/26/2003 0.221 0.034* ND ND ND ND ND ND 3.08 x 10-9*** 0.669 1/27/2003 0.291 0.202 ND ND ND ND ND ND 7.27 x 10-7*** 0.099+ 1/28/2003 0.667 0.067+ ND ND ND ND ND ND NV 0.166 1/29/2003 0.382 0.246 ND ND ND ND ND ND NV 0.328 1/30/2003 ND 0.29 0.091+ 0.432 0.389 0.578 0.689 0.256 NV 0.108+ +p=0.1 *p=0.05 **p=0.01 ***p=0.001 NV =no value determined, too few replicates, ND = not determined

183 Appendix B ANOVA f-test results from the 33P flux experiments by time.

August 2003 initial T1 T2 T3 T4 T5 T6 T7 T8 T9 Sandy Key 0.09+ 0.11+ 0.16 0.12+ 0.249 0.164 0.013* 0.053* 0.057* 0.044*

May 2004 Sandy Key 0.522 0.087+ 0.08+8 0.195 0.087+ 0.198 ND ND ND ND End Key 0.599 0.057+ 0.047* 0.084+ 0.069+ 0. 096+ ND ND ND ND +p=0.1, *p=0.05

184 Appendix C (from Cahoon and Cooke, 1992).

Net Benthic Primary Production = (([DO]t2 – [DO]t1 x V x 12)/(PQ x H x A)

[DO]t2 is the concentration of dissolved oxygen in mmol/liter at the end of the incubation.

[DO]t1 is the concentration of dissolved oxygen in mmol/liter at the start of the incubation. V is the volume of the chamber in liters. 12 is the atomic weight of carbon.

A Photosynthetic Quotient of 1.2 is used to represent 1.2 moles of O2 evolved per mole of carbon fixed. H is the hours of incubation. A is the area of sediment enclosed beneath the chamber in square meters.

Sediment Respiration = (([DO]t2 – [DO]t1 x V x 12 x RQ)/(H x A) A Respiratory Quotient of 1.0 is used to represent 1.0 moles of carbon respired per 1 mole of O2 consumed.

Gross Benthic Production =  Net Benthic Primary Production + Sediment Respiration

185

About the Author

Merrie Beth Neely is a native of Slippery Rock, PA. She received her Bachelor of Science degree from the University of Tampa, where she was an honors program graduate, a member of BBB Biological Society and President‟s Scholarship recipient. She received her Masters of Science degree from the University of South Florida. She is a proud member of Kappa Delta Sorority since 1989 and has been a local and national volunteer with that organization since 1993. She is also a proud member and former officer of the St. Petersburg Alumnae Panhellenic Association since 1995. She has worked for Pinellas County, the University of South Florida and the State of Florida during her professional career.