USF Libraries
USF Digital Collections

Nutrient uptake by seagrass communities and associated organisms

MISSING IMAGE

Material Information

Title:
Nutrient uptake by seagrass communities and associated organisms impact of hydrodynamic regime quantified through field measurements and use of an isotope label
Physical Description:
Book
Language:
English
Creator:
Cornelisen, Christopher David
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Seagrasses -- Ecology   ( lcsh )
Hydraulics -- Measurement   ( lcsh )
nutrient uptake
mass transfer
isotope
dissolved inorganic nitrogen
water flow
Dissertations, Academic -- Biology -- Doctoral -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Seagrass communities are composed of numerous organisms that depend on water-column nutrients for metabolic processes. The rate at which these organisms remove a nutrient from the water column can be controlled by physical factors such as hydrodynamic regime or by biological factors such as speed of enzyme reactions. The impact of hydrodynamic regime on rates of nutrient uptake for seagrass (Thalassia testudinum) communities and for organisms that comprise the community (seagrass, epiphytes, phytoplankton, and microphytobenthos) was quantified in a series of field flume experiments employing the use of 15N-labeled ammonium and nitrate. Rates of ammonium uptake for the entire community and for seagrass leaves and epiphytes were significantly dependent on bulk velocity, bottom shear stress, and the rate of turbulent energy dissipation. Relationships between uptake rates and these parameters were consistent with mass-transfer theory and suggest that the effect of water flow on ammonium uptake is the same for the benthos as a whole and for the organisms that form the canopy. In addition, epiphytes on the surface of T. testudinum leaves were shown to depress leaf uptake by an amount proportional to the area of the leaf covered by epiphytes. Water flow influenced rates of nitrate uptake for the community and the epiphytes; however, uptake rates were depressed relative to those for ammonium suggesting that uptake of nitrate was also affected by biological factors such as enzyme activity. Epiphytes reduced uptake of nitrate by the leaves; however, the amount of reduction was not proportional to the extent of epiphyte cover, which provided further evidence that nitrate uptake by T. testudinum leaves was biologically limited. As an additional component of the research, hydrodynamic regime of a mixed seagrass and coral community in Florida Bay was characterized using an acoustic Doppler velocimeter. Hydrodynamic parameters estimated from velocity data were used in mass-transfer equations to predict nutrient uptake by the benthos over a range of water velocity. Measured rates of uptake from field flume experiments conducted in the same community confirmed that hydrodynamic data could be used to accurately predict nutrient transport to the benthos under natural flow conditions.
Thesis:
Thesis (Ph.D.)--University of South Florida, 2003.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Christopher David Cornelisen.
General Note:
Includes vita.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 185 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001430561
oclc - 52314473
notis - AJL4022
usfldc doi - E14-SFE0000079
usfldc handle - e14.79
System ID:
SFS0024775:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001430561
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 031007s2003 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000079
035
(OCoLC)52314473
9
AJL4022
b SE
SFE0000079
040
FHM
c FHM
049
FHME
090
QH307.2
1 100
Cornelisen, Christopher David.
0 245
Nutrient uptake by seagrass communities and associated organisms
h [electronic resource] :
impact of hydrodynamic regime quantified through field measurements and use of an isotope label /
by Christopher David Cornelisen.
260
[Tampa, Fla.] :
University of South Florida,
2003.
502
Thesis (Ph.D.)--University of South Florida, 2003.
504
Includes bibliographical references.
500
Includes vita.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 185 pages.
520
ABSTRACT: Seagrass communities are composed of numerous organisms that depend on water-column nutrients for metabolic processes. The rate at which these organisms remove a nutrient from the water column can be controlled by physical factors such as hydrodynamic regime or by biological factors such as speed of enzyme reactions. The impact of hydrodynamic regime on rates of nutrient uptake for seagrass (Thalassia testudinum) communities and for organisms that comprise the community (seagrass, epiphytes, phytoplankton, and microphytobenthos) was quantified in a series of field flume experiments employing the use of 15N-labeled ammonium and nitrate. Rates of ammonium uptake for the entire community and for seagrass leaves and epiphytes were significantly dependent on bulk velocity, bottom shear stress, and the rate of turbulent energy dissipation. Relationships between uptake rates and these parameters were consistent with mass-transfer theory and suggest that the effect of water flow on ammonium uptake is the same for the benthos as a whole and for the organisms that form the canopy. In addition, epiphytes on the surface of T. testudinum leaves were shown to depress leaf uptake by an amount proportional to the area of the leaf covered by epiphytes. Water flow influenced rates of nitrate uptake for the community and the epiphytes; however, uptake rates were depressed relative to those for ammonium suggesting that uptake of nitrate was also affected by biological factors such as enzyme activity. Epiphytes reduced uptake of nitrate by the leaves; however, the amount of reduction was not proportional to the extent of epiphyte cover, which provided further evidence that nitrate uptake by T. testudinum leaves was biologically limited. As an additional component of the research, hydrodynamic regime of a mixed seagrass and coral community in Florida Bay was characterized using an acoustic Doppler velocimeter. Hydrodynamic parameters estimated from velocity data were used in mass-transfer equations to predict nutrient uptake by the benthos over a range of water velocity. Measured rates of uptake from field flume experiments conducted in the same community confirmed that hydrodynamic data could be used to accurately predict nutrient transport to the benthos under natural flow conditions.
590
Adviser: Thomas, Florence I.M.
653
nutrient uptake.
mass transfer.
isotope.
dissolved inorganic nitrogen.
water flow.
690
Dissertations, Academic
z USF
x Biology
Doctoral.
650
Seagrasses
Ecology.
Hydraulics
Measurement.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.79



PAGE 1

NUTRIENT UPTAKE BY SEAGRASS COMMUNITIES AND ASSOCIATED ORGANISMS: IMPACT OF HYDRODYNAMIC REGIME QUANTIFIED THROUGH FIELD MEASUREMENTS AND USE OF AN ISOTOPE LABEL by CHRISTOPHER DAVID CORNELISEN A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Biology College of Arts and Sciences University of South Florida Major Professor: Florence I.M. Thomas Ph.D. Susan S. Bell, Ph.D. Paula G. Coble, Ph.D. Peter D. Stiling, Ph.D. D ate of Approval: February 28, 2003 Keywords: mass transfer, nutrient uptake isotope, nitrogen water flow Copyright 2003, Christopher D. Cornelisen

PAGE 2

This doctoral dissertation is dedicated to my wife Amy Keeler Cornelisen, who has provided continued enthusiasm, encouragement, support, and patience throughout my pursuit of the doctoral degree

PAGE 3

ACKNOWLEDGEMENTS I would like to extend my gratitude to Dr. Flo Thomas, my major professor, for all of her support, encouragement, and friendship throughout my years as a doctoral student. This research was made possible by a National Science Foundation PECASE award to F.I.M. Thomas (OCE 9996361) I thank my committee members, Drs. Susan Bell, Peter Stiling, and Paula Coble for their advice, input, and encouragement throughout my years at USF. The participation of Dr. Mark Luther as Chair of my defense wa s also greatly appreciated. I would also like to thank Dr. Clifford Hearn for insightful discussions on the interpretation of my data. I am indebted to the n umerous students, friends, and p ostdocs that have contributed considerable time and effort to the field component of this research. These individuals included Dr. Toby Bolton, Larry Ritchie, Greg Delozier, Jen Kapp, Laura Karpalisto, Dr. Nancy Craig, Frederick Bleeker, Kirsty Wall, Karen Sharpe, Nate Smiley, Mark Driscoll, and Sean Kinane. I thank the staf f at ISO Analytical and Dr. Kent Fannings lab and staff for assistance with sample analysis. I would also like to acknowledge Dr. Margaret ( Penny ) Hall for her assistan ce with epiphyte identification Lastly, I thank my parent s, Carol and Bob Cornelisen, brother Dana, sister Erika, and wife Amy, for continued encouragement and support throughout my graduate career

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES .iv LIST OF FIGURES v ABSTRAC T ..ix CHAPTER ONE OVERVIEW OF RESEARCH ................................ ............................ 1 Research goal ................................ ................................ ................................ .......... 4 Chapter objectives ................................ ................................ ................................ ... 4 Significance of research ................................ ................................ .......................... 6 CHAPTER TWO AMMONIUM UPTAKE BY SEAGRASS EPIPHYTES: ISOLATION OF THE EFFECTS OF WATER VELOCITY USING AN ISOTOPE LABEL ................................ ................................ ................................ ................................ 8 Introduction ................................ ................................ ................................ ............. 8 Methods ................................ ................................ ................................ ............... 10 Results ................................ ................................ ................................ .................. 17 Uptake by epiphytes ................................ ................................ ................. 17 Uptake by the entire assemblage ................................ .............................. 18 Ammonium uptake by epiphytes vs. assemblage ................................ .... 21 Discussion ................................ ................................ ................................ ............ 23 CHAPTER THREE AMMONIUM AND NITRATE UPTAKE BY LEAVES OF THE SEAGRASS THALASSIA TESTUDINUM : IMPACT OF HY DRODYNAMIC REGIME AND EPIPHYTE COVER ON UPTAKE RATES ................................ ........................ 29 Introduction ................................ ................................ ................................ .......... 29 Methods ................................ ................................ ................................ ............... 33 Flume deployment ................................ ................................ .................... 33 Hydrodynamic characterization ................................ ............................... 38 Collection of seagrass and epiphyte samples ................................ ........... 40 Calculation of uptake rates ................................ ................................ ....... 42 Effects of wa ter flow on DIN up take rates ................................ .............. 44 Effects of epiphy te cover on DIN uptake rates ................................ ........ 45 Results ................................ ................................ ................................ .................. 45 Hydrodynamic characteristics ................................ ................................ .. 45 Community characteristics ................................ ................................ ....... 49 Effects of wa ter flow on DIN uptake rates ................................ .............. 51 Effects of epiphy te cover on DIN uptake rates ................................ ........ 56

PAGE 5

ii Discussion ................................ ................................ ................................ ............ 57 Ef fects of water flow on DIN uptake rates ................................ .............. 57 Effects of epiphyte cover on DIN uptake rates ................................ ........ 62 CHAPTER FOUR APPLICATION OF AN ISOTOPE LABEL FOR ISOLATING EFFECTS OF HYDRODYNAMIC REGIME ON AMMONIUM AND NITRATE UPTAKE BY MAJOR COMPON ENTS OF A SEAGRASS COMM UNITY ............... 67 Introduction ................................ ................................ ................................ .......... 67 Methods ................................ ................................ ................................ ................ 71 Flume deployment ................................ ................................ .................... 71 Measurement of hydrodynamic parame ters ................................ ............. 74 DIN upt ake by individual components ................................ .................... 76 DIN uptake by the community ................................ ................................ 80 Contributions of components to total DIN uptake by the community ..... 8 2 Results ................................ ................................ ................................ .................. 83 Ammonium upt ake by individual components ................................ ........ 84 Ammo nium uptake by the community ................................ .................... 91 Nitrate upt ake by individual components ................................ ................ 96 Nitrate uptake by the community ................................ ............................. 97 Contributions of components to total DIN uptake by the community .. 100 Discussion ................................ ................................ ................................ ......... 102 Amm onium uptak e by individual components ................................ ..... 102 Ammoni um uptake by the community ................................ ................. 105 Nitrate uptake b y individual components ................................ ............. 107 Nitra te uptake by the community ................................ .......................... 108 Contributions of components to total D IN uptake by the community .. 109 Conclusions ................................ ................................ ........................... 112 CHAPTER FIVE HYDRODYNAMIC CHARACTERIZATION OF A CARBONATE BANK IN FLORIDA BAY: IMPLICATIONS FOR NUTR IENT UPTAKE BY THE BENTHOS ................................ ................................ ................................ .................... 113 Introduction ................................ ................................ ................................ ....... 113 Methods ................................ ................................ ................................ ............. 118 Study site ................................ ................................ ............................... 118 Community composition ................................ ................................ ....... 119 Hyd rodynamic characterization ................................ ............................ 120 Estim ating nutrient uptake ................................ ................................ .... 124 Results ................................ ................................ ................................ ............... 125 Community composition ................................ ................................ ....... 125 Hydrodynamic characterization ................................ ............................ 128 Estimating nutrient uptake ................................ ................................ .... 136 Discussion ................................ ................................ ................................ ......... 139 Community composition ................................ ................................ ....... 140 Hyd rodynamic characterization ................................ ............................ 141

PAGE 6

iii Estimating nutrient uptake ................................ ................................ .... 146 C onclusions ................................ ................................ ........................... 15 3 REFERENCES CITED ................................ ................................ ................................ 15 5 APPENDICES ................................ ................................ ................................ .............. 16 5 Appendix A: Additional figures ................................ ................................ ........ 166 ABOUT THE AUTHOR ................................ ................................ ..................... End Page

PAGE 7

iv LIST OF TABLES Table 1 C haracteristics for the study sites where field flume experiments were conducted to measure uptake rate s for ammonium (left column) and nitrate (right column) ................................ ................................ ............................. 36 Table 2 Canopy characteristics at study sites for NH 4 + and NO 3 experiments ........ 73 Table 3 Values for depth averaged velocity (U b ), shear velocity (U ), and roughness length (Z o ) calculated us i ng the log prandtl equation. ................................ 85 Table 4 Ammonium uptake rate constants for the community (S) and uptake rates estimated for individual components ( r ). ................................ .................. 86 Table 5 Nitrate uptake rate constants for the community (S) and uptake rates estimated for individual components ( r ). ................................ ................... 98 Table 6 Dependence of uptake rates for the community (S) and uptake rates for benthic components ( r ) on hydrodynamic parame ters for ammonium (t op) and for nitrate (bottom). ................................ ................................ .............. 99 Table 7 Hydrodynamic parameters estimated for each of the vertical profiles co llected on Old Sweat Bank. ................................ ................................ 129

PAGE 8

v LIST OF FIGURES Figure 1 Map of the Tampa Bay area and study site at Emerson Point Pa rk ........... 12 Figure 2 (A) Rates of NH 4 + uptake by epiphytes ( r Chl ) and (B) uptake rate constants (S) for the assemblage as a function of water velocity (U b ). ...................... 19 Figure 3 Total ammonium removed during the experiments by epiphytes (A) and seagrass leaves (B) versus water velocity (U b ). ................................ .......... 20 Figure 4 Uptake rates nor malized to Chl a for epiphytes ( r Chl ) as a function of total ammon ium uptake by the assemblage. ................................ ....................... 22 Figure 5 Area map of Tampa Bay indicating locations of flume experiments for measuring rates of ammonium and nitrate uptake. ................................ ..... 35 Figure 6 Diagram of the fie ld flume used for isolating a section of the seagrass bed and c onducting uptake experiments. ................................ ........................... 37 Figure 7 Example of two velocity profiles collected during flume experiments for measuring NH 4 + uptake rates, including one conducted at low ( o ) and one at hig h ( ) flow. ................................ ................................ .............................. 46 Figure 8 Vertical profiles of (A) Reynolds stress ( ' W U ) and (B) total turbulent energy (K = 0.5[ ' U U + ' V V + ' W W ]). ................................ .................. 48 Figure 9 The relative proportion of deflected ca nopy height (h d ) represented by roughness length (Z 0 ) versus bulk velocity (U b ) for all profiles collected during the ammonium an d nitrate uptake experiments. .............................. 50 Figure 10 The rate of ammonium uptake ( r ) by seagrass leaves with ( o ) and without ( ) epiph yte cover versus (A) bottom shear stress ( t ) raised to the 0.4 power and (B) energy dissipation rate ( e ) raised to the 0.25 power. ..................... 52 Figure 11 The rate of nitrate uptake ( r ) by seagrass leaves with ( open symbols) and without (solid symbols ) epiphyte cov er versus (A) bottom shear stress ( t ) raised to the 0.4 power and (B) energy dissipation rate ( e ) raised to the 0.25 power. ................................ ................................ ................................ .......... 54

PAGE 9

vi Figure 12 (A) Rates of ammonium uptake for seagrass leaves with ( open symbols ) and without ( closed symbols ) epiphyte cover and for epiphytes ( ) versus energy dissipation rate raised to a power of 0.25. ................................ ....... 55 Figure 13 The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by PON versus bulk velocity (U b ). ................................ .......................... 87 Figure 14 The concentration of PON in the water colum n as a function of bulk velocity (U b ) during experiments for measuring ammonium (solid symbols) and n itrate (open symbols) uptake. ................................ ............................. 89 Figure 15 The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by epiphytes versus rate of turbulen t energy dissipation ( e ) to the 0.25 power. ................................ ................................ ................................ .......... 90 Figure 16 The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by seagrass leaves versus rate of turbulent energy dissipation ( e ) raised to the 0.25 power. ................................ ................................ ............................ 92 Figure 17 The rate of ammoni um (solid symbols) and nitrate (open symbols) uptake by microphytobenthos ( r ) versus rate of turbulent energy dissipation ( e ) raised to the 0.25 power. ................................ ................................ ............. 93 Figure 18 Uptake rate constants (S) for ammonium (solid symbols) and nitrate (open symbols) versus bulk velocity (U b ). ................................ ............................ 9 4 Figure 19 Measured versus predicted uptake rate constants (S) for uptake of ammoniu m (left) and nitrate (right). ................................ ........................... 95 Figure 20 Pie charts showing estimated percent contributions of individual components to total uptake by the commu nity for up take of ammonium (top) and nitrate (bottom). ................................ ................................ ........ 101 Figure 21 Regional (A) and close up (B) map show ing location of study site. ....... 117 Figure 22 Densities (individuals/m 2 ) of dominant organisms inhabiting Old Sweat Bank. ................................ ................................ ................................ ........ 126 Figure 23 Example of two velocity profiles collected on the b ank. ......................... 130 Figure 24 Vertical profiles of (A) total turbulent energy (K = 0.5[ ' U U + ' V V + ' W W ]), (B) relative turbulence intensity ( K / r U ), and (C) Reynolds stress ( ' W U ). ................................ ................................ ................................ ... 131

PAGE 10

vii Figure 25 Turbulent Reynolds stress ( ' W U ) plotted versus relative depth (1 Z/D) for the three profiles collected at site 1 (see Table 7 for statistics), wh ere D was the depth of the water and Z was the height above the bottom that the ADV measurement was taken. ................................ ................................ 133 Figure 26 Estimates of shear velocity (U ) versus depth averaged velocity (U b ). ... 134 Figure 27 Friction coefficients (c f ) for various bottom t ypes as a function of flow Reynolds numbe r. ................................ ................................ .................... 135 Figure 28 Rate of turbulent energy dissipation ( e ) versus velocity (U b ). ................. 137 Figure 29 Predicted uptake rate constants (S) for ammonium versus (A) energy dissipation rate raised to the 0.25 power, (B) bottom shear stress raised to the 0.40 power, and (C) bulk velocity raised to the 0.80 power. ............ 138 Figure 30 (A) Vertical profiles of velocity (U), (B) Reynolds stress ( ' W U ), (C) total turbulent energy (K = 0.5[ ' U U + ' V V + ' W W ]), and (D) relative turbulence intensity ( K / r U ) for a velocity profile collected on OSB (profile 2 1) and in an area of dense grass (profile DG 2). ...................... 142 Figure 31 (A) Bulk velocity over time during the period of data collection and (B) predicted flux of ammonium (in Mol NH 4 + m 2 s 1 ) o ver the same time interval. ................................ ................................ ................................ .... 149 Figure 32 Results from a series of paired 15 N labeled NH 4 + uptake experiments conducted in a fie ld flume on Old Sweat Bank. ................................ ...... 151 Figure 33 Velocity profiles collected during flume experiments for me asuring uptake rates for ammonium ................................ ................................ ................ 166 Figure 34 Velocity profiles collected during flume experiments for measuring uptake rates for nitrate. ................................ ................................ ...................... 167 Figure 35 Vertical distribution of Reynolds stress ( ' W U ) estimated from velocity profi les collected during experiments for measuring uptake rates for (left) ammonium, and (right) nitrate. ................................ ................................ 168 Figure 36 Total turbulent energy (left) and relative turbulence intensity (right) estimated from profiles collected during flume experiments for mea suring rates of ammonium uptake. ................................ ................................ ...... 169

PAGE 11

viii Figure 37 Total turbulent energy (A) and relative turbulence intensity (B) estimated from profiles collected during flume experiments for measuring rates of nitrate uptake. ................................ ................................ ........................... 170 Figure 38 Shear velocity (U ) as a fun ction of bulk velocity (U b ) for ammonium uptake experiments (solid symbols) and for nitrate uptake experiments (open symbols). ................................ ................................ ........................ 171

PAGE 12

ix NUTRIENT UPTAKE BY SEAGRASS COMMUNITIES AND ASSOCIATED ORGANISMS : I MPACT OF HYDRODYNAMIC REGIME QUANTIFIED THROUGH F IELD MEASUREMENTS AND USE OF AN ISOTOPE LABEL Christopher D. Cornelisen ABSTRACT Seagrass communities are composed of numerous organisms that depend on water column nutrients for metabolic processes. The rate at which these organisms remove a nutrient from the water column can be controlled by physical factors such as hydrodynamic regime or by biological factors such as speed of enzyme reactions. The impact of hydrodynamic regime on rates of nutrient uptake for seagrass ( Thalassia testudinum ) communiti es and for organisms that comprise the community (seagrass, epiphytes, phytoplankton, and microphytobenthos) was quantified in a series of field flume experiments employing the use of 15 N labeled ammonium and nitrate. Rates of ammonium uptake for the enti re community and for seagrass leaves and epiphytes were significantly dependent on bulk velocity, bottom shear stress, and the rate of turbulent energy dissipation. Relationships between uptake rates and these parameters were consistent with mass transfer theory and suggest that the effect of water flow on ammonium uptake is the same for the benthos as a whole and for the organisms that form the canopy. In addition, epiphytes on the surface of T. testudinum leaves were shown to depress leaf uptake by an a mount proportional to the area of the leaf covered by epiphytes. Water flow influenced rates of nitrate uptake for the community and the epiphytes;

PAGE 13

x however, uptake rates were depressed relative to those for ammonium suggesting that uptake of nitrate was a lso affected by biological factors such as enz yme activity. E piphytes reduced uptake of nitrate by the leaves; however, the amount of reduction was not proportional to the extent of epiphyte cover, which provided further evidence that nitrate uptake by T. testudinum leaves was biologically limited. As an additional component of the research, hydrodynamic regime of a mixed seagrass and coral community in Florida Bay was characterized using an acoustic Doppler velocimeter. Hydrodynamic parameters estimated from velocity data were used in mass transfer equations to predict nutrient uptake by the benthos over a range of water velocity. Measured rates of uptake from field flume experiments conducted in the same community confirmed that hydrodynamic data could be used to accurately predict nutrient transport to the benthos under natural flow conditions.

PAGE 14

1 CHAPTER ONE OVERVIEW OF RESEARCH In the marine environment, the transport of dissolved chemicals from the w ater column to the benthos is of ecological importance. For example, rates of chemical transport affects calcification in corals (Dennison and Barnes 1988), photosynthesis (Dennison and Barnes 1988, Koehl and Alberte 1988, Koch 1994), and respiration (e.g ., Patterson et. al 1991). Turbulent shear and stress can strongly mediate the thickness of the diffusive boundary layer and therefore mass transfer to and from a surface (Kays and Crawford 1993). Benthic communities create surfaces with varying degrees o f roughness that directly influence turbulence within the overlying water. This turbulence can influence diffusive boundary layer thickness and therefore the transport of chemicals (i.e., nutrients) to and from the benthos (Patterson and Sebens 1989, Thoma s and Atkinson 1997, Thomas et al. 2000). The effect of morphological characteristics on water flow and chemical transport has been demonstrated at the organismal level for cnidarian colonies (Patterson and Sebens 1989, Patterson 1992) and algal leaves (K oehl and Alberte 1988). Studies have also shown that the morphology and spatial arrangement of an assemblage of organisms largely affect flow characteristics (Eckmann 1983; Fonseca and Kenworthy 1987; Fonseca and Calahan 1992). For example, seagrasses ar e known to dampen and reduce

PAGE 15

2 wave energy, thereby influencing accumulation of sediment and organic matter within the canopy (Fonseca and Calahan, 1992). Over the past 15 years, a considerable amount of research has been conducted on the role of hydrodyna mics in nutrient uptake by benthic communities including coral reefs (e.g., Atkinson 1987; Bilger and Atkinson 1992; Thomas and Atkinson 1997; Hearn et al. 2001) and seagrass beds (Thomas et al. 2000; Thomas and Cornelisen 2003 ). These studies have demonst rated that physical factors, including water velocity and roughness of the benthos, control uptake rates for these communities. Both physical and biological factors can influence rates of nutrient uptake (Sanford and Crawford 2000). For example, if the r ate of nutrient uptake by a benthic community is limited by the rate at which the nutrient in delivered to its surface, rather than the rate at which the benthos processes the nutrient, then factors that influence thickness of diffusive boundary layers (DB L) will limit uptake rates. In this case the community is said to be mass transfer limited, and enhanced turbulence resulting from increased water flow and interaction between the rough surface and the water column will decrease DBL thickness and therefor e enhance uptake rates. Conversely, if the rate at which the benthos processes the nutrient exceeds the rate of delivery, then water flow will have little to no effect on uptake rates. In this case the community is said to be biologically limited since fa ctors such as speed of enzyme reactions or availability of active uptake sites are limiting uptake rates. Transitions between physical and biological limitation can also exist (Bilger and Atkinson 1995; Sanford and Crawford 2000). In cases where nutrient uptake was mass transfer limited, investigators have demonstrated the ability to predict rates of nutrient uptake by the benthos using

PAGE 16

3 engineering models originally intended for describing he at and mass transfer in pipes (s ee Bilger and Atkinson 1992; Tho mas et al. 2000; Kays and Crawford 1993). This approach involves the calculation of a Stanton number, which is the ratio of nutrient flux (represented as an uptake rate constant) to advection of a nutrient past the benthos (represented as the bulk velocit y). Stanton numbers, and therefore uptake rate constants, are a function of water velocity, the geometry of the benthos, and the molecular diffusivity of the nutrient. Recently, Hearn et al. (2001) derived mass transfer equations that allow analogous pre dictions of uptake rate constants based on a more direct link between hydrodynamic conditions in the marine environment and the uptake of nutrients by the benthos. Both of these approaches view the benthos as a single entity; a rough surface that removes nutrients from the water column. However, communities such as coral reefs and seagrass beds are comprised of a complex assemblage of organisms that depend on water column nutrients for metabolic processes. These organisms vary in morphology, location rela tive to the sediment water interface, and physiology and as a consequence are likely affected differently by hydrodynamics. In addition to treating the benthos as a single entity, previous studies on mass transfer in coral reefs and seagrass communities w ere also based on experiments conducted in flumes and with the communities exposed to simulated flow. However, the equations from Hearn et al. (2001) provide a means to link hydrodynamic measurements taken in situ with an ecological process, such as nutri ent uptake by the benthos.

PAGE 17

4 Research goal The overall goal of my doctoral research is to describe and quantify the impact of hydrodynamic regime on rates of nutrient uptake for benthic communities colonized by seagrass, both at the scale of the communit y as a whole and that of the organisms forming the community. For this research I employ an interdisciplinary approach and utilize several important tools, including isotope labels ( 15 NH 4 and 15 NO 3 ) field deployed flumes, and acoustic Doppler velocimeters (ADV). In the research presented in Chapters Two through Four, I utilize these tools to quantify the effects of hydrodynamics on rates of nutrient uptake for individual organisms (i.e., seagrass plants, epiphytes, microphytobenthos, phytoplankton) while t hey are situated within seagrass beds. In Chapter Five, I integrate hydrodynamic parameters measured in situ with theoretical mass transfer equations (Hearn et al. 2001) to predict the amount of nutrients (i.e., NH 4 + and PO 4 3 ) the benthos is removing fro m the water column over time. Estimates of nutrient uptake using mass transfer equations are compared to estimates based on measurements of nutrient uptake using a field flume. In addition, a series of paired flume experiments using 15 N labeled NH 4 + are conducted to evaluate the effects of water flow on organisms situated within the community. Chapter objectives In order to meet my goal, I carry out field based experiments as described in the four chapters to follow. In Chapter Two I demonstrate the app lication of isotope labels for isolating effects of water velocity on 15 NH 4 + uptake by an individual component

PAGE 18

5 (epiphytes) situated within an assemblage of seagrass leaves. Specific objectives of Chapter Two are to: (1) Establish protocols for measuring nutri ent uptake by an individual component of a seagrass community (epiphytes) using an isotope label. (2) Isolate NH 4 + uptake by epiphytes and evaluate how this uptake compares to the overall assemblage of organisms. (3) Evaluate the effects of velocity on NH 4 + uptake by epiphytes. In Chapter Three I utilize isotope labels to investigate the concomitant effects of water flow and epiphyte cover on rates of ammonium and nitrate uptake by Thalassia testudinum leaves. Experiments are conducted in a field flume deployed in a natural seagrass bed in order to: (1) Quantify the effect of epiphyte cover on NH 4 + and NO 3 uptake by seagrass leaves. (2) Assess whether or not presence of epiphytes inhibit uptake of NH 4 + and/or NO 3 by seagrass leaves and if their presence on the leaf surf ace influences the effect of velocity on uptake by seagrass leaves. In Chapter Four I expand the analysis in Chapter Three to include multiple components of the community (seagrasses, epiphytes, sediments, and phytoplankton) and investigate the contribut ions of these components to uptake by the community. Specific objectives of Chapter Four are to: (1) Isolate uptake of DIN (NH 4 + and NO 3 ) by individual components of a natural seagrass bed. (2) Isolate the effects of hydrodynamic regime on DIN uptake by the vari ous components and the community by conducting experiments over a range of velocity and measuring hydrodynamic parameters with an ADV. (3) Compare NH 4 + vs. NO 3 uptake rates among individual components and the community and assess the extent to which uptake of these nutrients is dependent on hydrodynamic regime. (4) Evaluate the application of mass transfer equations for predicting uptake by the benthic components.

PAGE 19

6 Finally, in Chapter Five I characterize water flow over a mixed hardbottom seagrass community in Fl orida Bay and integrate hydrodynamic data with theoretical mass transfer equations (Hearn et al. 2001) to predict rates of nutrient uptake by the benthos. In this chapter I aim to quantify the interactions between the benthos and water flow and demonstrate the importance of these interactions in influencing nutrient uptake by the benthos. Specific objectives of Chapter Five are to: (1) Collect data on community composition, bottom roughness, and hydrodynamics on a shallow bank in Florida Bay. (2) Using hydrodynami c data and mass transfer equations, predict rates of nutrient uptake for the benthos. (3) Compare predicted values with those measured in field flumes and with data collected using isotope labels. Significance of research By meeting the above objectives, I aim to increase our understanding of nutrient transport processes in estuarine and near shore systems, and as a result, further demonstrate the important role of hydrodynamics in ecological processes. My application of isotope labels presents a new approa ch for investigating the effects of water flow on nutrient transport at both the scale of individual organisms and the entire community. Through the use of isotope labels and field flumes, rates of nutrient uptake are quantified for organisms while they a re situated in their natural community. As a result, I demonstrate that isotope labels can be used to isolate the response of an individual component of a community to changes in an environmental factor (water velocity) and evaluate how this response compa res to the response of the community as a whole. The approach also enables me to quantify the contributions of components to

PAGE 20

7 total uptake by the community and separate uptake by the benthos from that within the water column. I can then evaluate the impli cations of uptake by phytoplankton on predictions of uptake by the benthos based on mass transfer models. Through the application of previously derived equations (Hearn et al. 2001), I link hydrodynamic data collected under natural flows to nutrient uptak e by the benthos and demonstrate the utility of using simple mass transfer models for understanding the factors controlling nutrient uptake by the benthos. The methods used in Chapter Five reveal the great potential of integrating equations, such as those derived by Hearn et al. (2001), within larger scale models in effort to better understand and predict the factors influencing water quality in estuarine and near shore waters.

PAGE 21

8 CHAPTER TWO AMMONIUM UPTAKE BY SEAGRASS EPIPHYTES: ISOLATION OF THE EFFECTS OF WATER VELOCITY USING AN ISOTOPE LABEL Introduction Rates of nutrient uptake by benthic organisms can influence important ecological processes (i.e. photosynthesis, calcification) and can be strongly mediated by water flow and boundary layer character istics adjacent to uptake surfaces (Patterson 1992). If uptake is controlled by rates of diffusion of nutrients through the diffusive boundary layer (physically limited), then enough uptake sites or metabolic enzymes are present to take up all of the nutr ients that are delivered to an organisms surface. In this case, an increase in water velocity reduces the thickness of the diffusive boundary layer leading to a higher rate of nutrient uptake. For individual organisms, water velocity is positively corre lated to nutrient uptake by algae (e.g. Gerard 1982; Hurd et al. 1996) and nutrient dependent processes including photosynthesis in algae (e.g. Wheeler 1980; Koch 1993) and seagrasses (Koch 1994) and photosynthesis and calcification in corals (Dennison and Barnes 1988; Patterson et al. 1991). Rates of nutrient uptake have also been shown to be dependent on water velocity for assemblages of algae (Larned and Atkinson 1997) and coral (e.g., Bilger and Atkinson 1992; Thomas and Atkinson 1997) and for naturally occurring seagrass beds (Thomas et al. 2000). These studies have demonstrated the importance of water flow on nutrient uptake by individuals and entire communities.

PAGE 22

9 In a recent study by Thomas et al. (2000), ammonium uptake by seagrass communities is s hown to be dependent on water velocity and the morphology of the canopy. In their experimental approach, the seagrass community is viewed as a single entity whose rough surface removes NH 4 + and influences the flow of water over the benthos. However, seagr ass communities are composed of a diverse assemblage of organisms that remove nutrients from the water column, including seagrass plants, epiphytes attached to the seagrass leaves, and phytoplankton. While Thomas et al.s data provide estimates of whole c ommunity uptake they do not provide information about the kinetics of nutrient uptake for individual components of the community. These components vary in their morphology, physiology and location relative to the canopy. Therefore, water flow may have v ariable effects on the different components of a seagrass community, which in turn may collectively contribute to the community scale response quantified in Thomas et al. (2000). In the present study, we isolate the effects of water velocity on nutrient uptake by a single component of seagrass communities (epiphytes) while it is situated within an assemblage of seagrass leaves. By utilizing an isotope labeling approach, rates of nutrient uptake for epiphytes are measured while they are attached to the ho st plant in order to obtain ecologically relevant estimates of nutrient uptake kinetics for epiphytes. Epiphytes play an integral role in the ecology of seagrass communities, including food web dynamics (e.g. Fry and Parker 1979) and nutrient cycling (e.g. Harlin 1973; McRoy and Goering 1974). In addition, epiphytes are a major contributor to the overall productivity of seagrass meadows (e.g. Moncreiff et al. 1992) and are considered an important factor influencing the distribution and abundance of seagras ses (Kuo and

PAGE 23

10 McComb 1989). Although these studies provide an extensive database describing the important roles of epiphytes in seagrass communities, little data is available on nutrient uptake kinetics of epiphytes and their contribution to total dissolve d inorganic nitrogen (DIN) inputs to seagrass communities (Hemminga et al. 1991). Our study is intended to further our understanding of the factors that influence nutrient uptake by epiphytes by describing the effects of water velocity on NH 4 + uptake by ep iphytes and how uptake by epiphytes relates to the nutrient dynamics of the community as a whole. Methods In order to separate nutrient uptake by epiphytes from other components of the community (i.e. seagrass plants and phytoplankton), we used 15 N label ed NH 4 + to trace uptake from the water column into the epiphytes attached to seagrass leaves. Earlier studies have demonstrated the effectiveness of using isotope labels in understanding relationships between epiphytes and their host plant (e.g. Harlin 19 73; McRoy and Goering 1974; Johnstone 1979). In addition, numerous studies have employed isotope labels to partition nutrient uptake among community components and understand nutrient cycling in both freshwater (e.g. Pelton et al. 1998; Eriksson 2001, Ham ilton et al. 2001) and marine systems (e.g. Winning et al. 1999; Koop et al. 2001). In our experiments, we expanded this application of isotope tracers to isolate the effects of water velocity on NH 4 + uptake by a single community component while it was sit uated in a complex assemblage of organisms. To assess the effects of water velocity on rates of NH 4 + uptake, 15 N accumulation in epiphyte tissues and the total uptake of NH 4 + from the water column by all organisms

PAGE 24

11 combined was measured in 14 flume experim ents conducted over a range of velocity (0.02 0.20 m s 1 ). This range was chosen to best represent ambient water velocity observed at the site and in seagrass beds located elsewhere (Fonseca and Kenworthy 1987; Koch and Gust 1999; Thomas and Cornelisen; unpubl. data). The flume (Vol=180 L) was of a racetrack design and transported into the field and placed on a table along the shore for experiments. An electric trolling motor housed in a drop box at one end of the flume imposed controlled, unidirectional flow. Experiments were conducted at Emerson Point Park located at the mouth of the Manatee River in Southwest Tampa Bay on the west coast of Florida (Fig. 1). Seagrass leaves ( Thalassia testudinum ), with epiphytes attached, were collected from one of tw o donor beds (sites) located close to shore and transplanted into the flume. The two sites were approximately 150 meters apart on either side of Emerson Point. Six experiments using leaves from donor site 1 were conducted on 7 8 December 1999, 29 March 2 000 and 4 April 2000 and eight experiments using transplants from donor site 2 were completed on 8 9 December 1999 and 3, 7, and 21 November 2000. On each day, experiments were completed within one hour of each other and between 1100 h and 1500 h. The spe cific velocity for each experiment and the order of low and high velocity experiments was randomized. The flume size was minimized to allow for detection of nutrient uptake from the water column and as a result only the top 10 13 cm of individual leaves w as used in experiments. Because seagrass leaves in the donor beds were only 12 to 15 cm in length, the severed leaves included most of the blade and attached epiphytes. New and senescing leaves were not used in the experiment due to limited epiphyte growt h on new leaves and the difficulty in cleanly removing epiphytes from senescing leaves. Leaves (n=150) were

PAGE 25

12 Old Tampa Bay Hillsborough Bay Tampa Bay Manatee River Emerson Pt. Park N Gulf of Mexico Tampa St. Pete 27 30 27 45 Tampa Bay Area F l o r i d a 5 km Figure 1. Map of the Tampa Bay area and study site at Emerson Point Park.

PAGE 26

13 transplanted to a removable Plexiglas floor (0.14 m 2 ) by affixing them into drilled holes with rubber stoppers. Before each experiment the flume was filled with seawater from the study site and spiked with labeled 15 N NH 4 + (as 98 atom % 15 (NH 4 ) 2 SO 4 or 15 NH 4 Cl) to achieve a final water column concentration of approximately 6 m M. The beginning concentration fluctuated between 6 and 7 m M due to background NH 4 + concentrations. Although this spike was higher than ambient NH 4 + levels at the time of the experiments (0.5 1 m M), a 6 m M spike was used in order to allow accurate detec tion of NH 4 + depletion in the water column over time and assess the effects of velocity on potential rates of NH 4 + uptake. After mixing (~3 minutes), the Plexiglas floor, with seagrass leaves attached, was placed into the flume and held in place with flow straighteners. Bulk water velocity (U b ) was estimated by timing neutrally buoyant particles over a known distance (n=20). Experiments were conducted for one hour so that a sufficient number of water and tissue samples for determining uptake rates could b e collected. Rates of NH 4 + uptake by epiphytes were determined by measuring 15 N accumulation in epiphyte tissues over time during flume experiments. For each one hour experiment, three leaves with epiphytes were randomly removed from the flume after 15 30, 45, and 60 minute intervals. Three leaves were required for each sample to ensure adequate amounts of epiphyte tissue for analysis. At the end of each 15 min interval, epiphytes (all attached organisms) were removed from the seagrass leaves by gent ly scraping the leaves with a dull edge and were pooled to represent an epiphyte sample for the interval. Epiphyte samples were briefly rinsed with DI water over a 35 m M screen to remove salt (Winning et al. 1999) and placed on ice. Samples of epiphytes fr om each

PAGE 27

14 donor site were also collected and processed for determination of ambient 15 N in the epiphyte tissues. In addition to epiphytes, whole seagrass leaves (n=3) were randomly selected from the flume at the end of each experiment, cleaned of epiphytes, pooled, and retained for 15 N analysis. All samples of epiphytes and seagrass were dried at 60 C for 24 h, weighed, homogenized, and stored in glass vials. Dry weights were used along with blade densities to estimate total bi omass (g dry wt) of epiphytes and seagrass in the flume during each experiment. Epiphyte and seagrass samples were analyzed using EA IRMS (elemental analyzer isotope ratio mass spectrometry) for determination of nitrogen content (% N) and atom % 15 N in the tissues. Specific uptake rates for epiphytes (V epi ) were calculated using the equation V epi =(da s /dt)/(a w a s ), where a s is the atom % 15 N in the epiphyte tissue, a w is the atom % 15 N of the enriched substrate, and t is time (in seconds) (Dugdale and Go ering 1 967, Iizumi and Hattori 1982). The units for V epi are g N removed (g N tissue) 1 s 1 or simply s 1 The numerator (da s /dt) was calculated as the slope of the least square regression of a s versus time. The atom % 15 N of the enriched water (a w ) wa s based on the amount of 98 atom % 15 NH 4 + added and background NH 4 + concentration (assumed to reflect 15 N concentration of atmospheric N ~0.37 atom % 15 N). To compare NH 4 + uptake by epiphytes to uptake by their host plant, uptake rates for the seagrass lea ves (V grass ) were also estimated for each of the experiments. Because seagrass leaves were retained only at the end of each experiment, V grass was based on the final excess atom % 15 N in the tissue (Dugdale and Goering 1967; Iizumi and Hattori 1982). It is noted that the use of the above equation in calculating specific uptake rates for epiphytes and seagrass leaves assumes that the atom % 15 N of the source pool did not change during the course of the

PAGE 28

15 experiment. Dilution of 15 N in the source pool result ing from inputs of non labeled NH 4 + into the water column (via regeneration, excretion) would expectedly result in underestimated uptake rates (Laws 1984). While we acknowledge this potential source of error, the short duration of these experiments along with the high concentration and atom % 15 N of the spike likely minimized dilution. Furthermore, time course measurements of 15 N accumulation in epiphytes (da s /dt) was found to be linear, suggesting that isotopic dilution of the substrate during the course of the experiments was not significant. Specific uptake rates (V epi and V grass ) were normalized to nitrogen concentration (% N) of the epiphyte and seagrass tissues to calculate an uptake rate for epiphytes ( r epi ) and se agrass leaves ( r grass ) in units g N removed (g dry wt) 1 s 1 (Dugdale and Goering 1967). These values were multiplied by the total biomass of each component in the flume to estimate the contribution of epiphytes and seagrass to the total NH 4 + removed from the water column during each experiment. Because of observed differences in the composition of epiphytes between donor sites (abundance of autotrophs vs. heterotrophs), uptake rates ( r epi ) were normalized to chlorophyll a concentration in the tissues to obtain an uptake rate that was representative of the autotrophic fraction ( r Chl = r epi Chl a 1 in units g N removed (mg Chl a) 1 s 1 ; Frankovich and Fourqurean 1997; Dickson and Wheeler 1995). Differences in the abundance of autotrophs that actively remove NH 4 + from the water column and heterotrophs (i.e. bryozoans) that do not would predictably result in misleading rates of 15 N accu mulation for the fraction of epiphytes that are actively removing NH 4 + from the water column Chlorophyll a concentrations [ mg Chl a (g dry wt) 1 ] were estimated for epiphyte samples that included all organisms attached to

PAGE 29

16 the seagrass leaves and were base d on samples collected at the end of each experiment. Chlorophyll a in these samples was estimated using spectrophotometric methods as outlined in Strickland and Parsons (1972). Rates of NH 4 + uptake for the entire assemblage of organisms (seagrass leaves, phytoplankton, and epiphytes combined) were determined by measuring the rate of decline in NH 4 + concentration in the water column over the duration (~1 h) of flume experiments. Methods of sample collection and analysis are outlined in detail in Thomas et al. (2000). In all but one experiment, uptake rates were based on a set of seven water samples collected over time. These samples were analyzed for NH 4 + concentration using an autoanalyzer to an accuracy of 0.05 m M. Only the beginning and end bottle wer e used for one experiment due to loss of water samples. Ammonium concentrations for these samples were determined using the indophenol blue method (Solorzano 1969). A first order rate constant (k) describes the decline in NH 4 + concentration in the flume over time (Rate of decline = dC/dt = k (C)), where C is the concentration of NH 4 + t is time, and k is the first order rate constant (s 1 ). This constant (k) was estimated for each experiment as the slope of the least square regression of the natural log of concentration versus time (see Bilger and Atkinson 1992, Thomas and Atkinson 1997, and Thomas et al. 2000 for discussion). Each first order rate constant was then normalized for water volume (Vol) in the flume (180 l) and the planar surface area (A) o f the bottom covered by the seagrass leaves (0.14 m 2 ) to calculate an uptake rate constant (S) in units m s 1 (S= k Vol A 1 ). Although Vol and A did not change during the experiments, this conversion was done in order to provide data in a form that is consistent (and therefore comparable) with previous studies on nutrient uptake kinetics in benthic

PAGE 30

17 communities (Thomas and Atkinson 1997, Thomas et al. 2000). In addition to S, the total NH 4 + uptake over time (in g N removed s 1 ) was calculated from the re gression of k vs. time for each experiment in order to compare uptake rates for epiphytes ( r Chl ) to the total NH 4 + removed by the assemblage in similar units. Results Uptake by epiphytes Coefficients of determination (r 2 ) for linear regressions of the atom % 15 N in epiphyte tissues vs. time (da s /dt) used in calculating specific uptake rates for epiphytes (V epi ) had a mean value of 0.85 (range=0.69 to 0.99, SD=0.11, n=14). Multiplying V epi by nitrogen content in epiphyte tissues provided NH 4 + uptake rates ( r epi ) that ranged from 0.45 10 8 to 3.3 10 8 and 0.66 10 8 to 6.8 10 8 g N remov ed (g dry wt) 1 s 1 for sites 1 and 2, respectively. Water velocity had a significant effect on r epi for epiphytes at site 1 ( r epi = (11.6 10 8 )U b 0.89 r 2 =0.80, P<0.05) and site 2 ( r epi = (20.7 10 8 )U b 0.74 r 2 =0.64, P<0.05). Although velocity had a s imilar effect on uptake by epiphytes at both sites (Homogeneity of slopes, P>0.05), estimates of r epi were significantly lower for site 1 than site 2 (ANCOVA, ln r epi vs. ln U b P<0.01). The disparity in r epi between sites can be explained by difference s in epiphyte composition. Epiphyte samples at site 1 included encrusting corallines, diatoms, attached macroalgae and a large portion (~ 30%) of epifauna (bryazoans, amphipods), while those at site 2 were primarily composed of autotrophic organisms (cora llines, macroalgae). Significantly lower chlorophyll a concentrations at site 1 (~0.57 mg Chl a (g dry wt) 1 SD=0.20, n=12) versus site 2 (~1.01 mg Chl a (g dry wt) 1 SD=0.25, n=15) reflected

PAGE 31

18 these differences in composition (t test, P<0.001, df=25). B ecause all organisms attached to the leaves were included in the analysis the abundance of epifauna composed of inactive nitrogen at site 1 would expectedly result in the lower estimates of r epi observed for these samples. This was confirmed by normalizin g uptake rates ( r epi ) to Chl a concentrations ( r Chl ). Estimates of r Chl were similar for both sites and ranged from 0.79 10 8 to 5.8 10 8 and 0.65 10 8 to 6.8 10 8 g N removed (mg Chl a) 1 s 1 for sites 1 and 2, respectively (Fig. 2A). Contrary t o estimates of r epi, there was no significant difference between sites for r Chl over the range of velocity (ANCOVA ln r chl vs ln U b P=0.14) and the dependence of r Chl on velocity for the pooled samples was on the order of 0.80 ( r chl = (18.9 10 8 )U b 0.77 r 2 =0.64, P<0.001). Although rates of uptake ( r Chl ) and epiphyte biomass were similar between sites, the epiphytes at site 2 contained a larger fraction of autotrophic epiphytes and therefore removed a greater portion of NH 4 + from the water column than epi phytes from site 1 (Fig. 3A; ANCOVA, ln Uptake (epiphytes) vs. ln U b P<0.05). Ammonium uptake by seagrass leaves was also dependent on water velocity (Fig. 3B). In general, uptake rates for the seagrass leaves ( r grass ) were lower than those for epiphyte s ( r epi ) and ranged from 0.3 2.0 10 8 g N removed (g dry wt) 1 s 1 It is noted that total uptake of NH 4 + by seagrass leaves at site 1 was significantly higher than uptake by seagrass leaves at site 2 (Fig. 3B; ANCOVA, ln Uptake (grass) vs. ln U b P<0.001 ). Uptake by the entire assemblage Linear regressions of the natural log of NH 4 + concentration vs. time used to determine first order rate constants (k) were significant for

PAGE 32

19 0 2 4 6 8 0.00 0.05 0.10 0.15 0.20 r Chl (g N (mg Chl a) -1 s -1 ) x 10 8 Epiphytes (Site 1) Epiphytes (Site 2) 0 4 8 12 16 0.00 0.05 0.10 0.15 0.20 Velocity, U b (m s -1 ) S (m s -1 ) x 10 5 Assemblage (Site 1) Assemblage (Site 2) Figure 2. (A) Rates of NH 4 + uptake by epiphytes ( r Chl ) and (B) uptake rate constants (S) for the assemblage as a function of water velocity (U b ). Estimates of ( r Chl ) were similar between sites and equally affected by water velocity (ANCOVA ln r chl vs ln U b P=0.14). The dependence of r Chl on velocity for the pooled samples is on the order of 0.80 ( r chl = (18.9)U b 0.77 r 2 =0.64, P<0.001). Uptake rate constants (S) were a function of velocity to the 0.59 power for site 1 (S=(25.3)U b 0.59 r 2 = 0.92, n=5, P<0.01) and to the 0.35 power for site 2 (S=(19.7)U b 0.3 5 r 2 = 0.62, n=7, P<0.05). B A

PAGE 33

20 -19 -18 -17 -16 -15 -14 -13 -5 -4 -3 -2 -1 ln Total NH 4 + removed (g N s -1 ) Assemblage (Site 1) Assemblage (Site 2) Epiphytes (Site 1) Epiphytes (Site 2) -19 -18 -17 -16 -15 -14 -13 -5 -4 -3 -2 -1 ln Velocity, U b (m s -1 ) ln Total NH 4 + removed (g N s -1 ) Assemblage (Site 1) Assemblage (Site 2) Grass (Site 1) Grass (Site 2) Figure 3. Total ammonium removed during the experiments by epiphytes (A) and seagrass leaves (B) versus water velocity (U b ). The total ammonium removed by the entire a ssemblage (calculated from S) is presented in both graphs. Regression statistics are as follows: Site 1 epiphytes (ln Uptake=0.72 ln U b 14.6; r 2 =0.79, P<0.05), Site 2 epiphytes (ln Uptake=0.69 ln U b 14.0; r 2 =0.56, P<0.05), Site 1 seagrass (ln Uptake=0.36 ln U b 15.4; r 2 =0.64, P=0.05), Site 2 seagrass (ln Uptake=0.64 ln U b 15.2, r 2 =0.90, P<0.001). For the entire assemblage, there was no significant difference between sites (ANCOVA, P=0.08) and a common regression line for pooled data is shown (ln Uptake=0.32 ln U b 13.3; r 2 =0.61, P<0.01). Total NH 4 + removed by epiphytes at site 1 was significantly lower than the total NH 4 + removed by epiphytes at site 2 (ANCOVA, P<0.05). The reverse was true for the seagrass leaves (ANCOVA, P<0.001). B A

PAGE 34

21 11 of the 14 flume experi ments and had a mean r 2 value of 0.93 (Range = 0.75 to 0.99, SD=0.09, n=11). Regressions from two experiments conducted at a low velocity (0.02 and 0.04 m s 1 ) were not significant and showed no NH 4 + uptake from the water column. This is likely due to sam pling periods being set too short to detect uptake at this low velocity. No regression statistics were available for the experiment conducted at 0.10 m s 1 since uptake was based on samples collected at the beginning and end of the experiment (See methods ). The uptake rate constant was a function of velocity to the 0.59 power for experiments using donor leaves from site 1 (S=(25.3 10 5 )U b 0.59 r 2 = 0.92, n=5, P<0.01) and to the 0.35 power for experiments using donor leaves from site 2 (S=(19.7 10 5 )U b 0 .35 r 2 = 0.62, n=7, P<0.05, see Fig. 2B). Ammonium uptake by epiphytes vs. assemblage The total NH 4 + removed by the assemblage ranged between 4.8 and 12.9 10 7 g N s 1 Uptake rates for epiphytes ( r Chl ) were a function of the total NH 4 + removed by the assemblage and increased as total uptake by the assemblage increased (Fig. 4; Model II regression, r Chl =(0.80 10 8 )x 2.9, r 2 =0.91, P<0.001, where x is the total NH 4 + removed by the assemblage in g N s 1 ). Unlike epiphytes, rates of uptake by seagrass leaves ( r grass ) were not dependent on the total NH 4 + uptake by the assemblage (r 2 =0.17, P=0.18). The total NH 4 + removed by individual components (epiphytes, seagrass leaves) and by the assemblage is de pendent on velocity (Fig. 3). On average, 17% (range=7 32%, SD=8, n=12) of the total NH 4 + removed from the water column was attributed to uptake by epiphytes. The relative contribution of epiphytes increased as a linear function of total NH 4 + uptake by t he assemblage (r 2 =0.57,

PAGE 35

22 Figure 4. Uptake rates normalized to Chl a for epiphytes ( r Chl ) as a function of total ammonium uptake by the assemblage (Model II regression; Sokal and Rohlf 1995 ). There is no significant difference (ANCOVA, P=0.97) in the relationship of r Chl vs. total uptake between sites a nd therefore the data is pooled The linear fit to this relationship is r Chl =0.80x 2.9, r 2 =0.91, P<0.001, where x was the total NH 4 + removed by the assemblage over time. 0 1 2 3 4 5 6 7 8 0 2 4 6 8 10 12 14 Total NH 4 + removed (g N s -1 ) x 10 7 r chl (g N [mg Chl a ] -1 s -1 ) x 10 8

PAGE 36

23 P<0.01) indicating that as total uptake by the assemblage increased, the relative contribution of epiphytes to total u ptake increased. Approximately 9% (range=4 17%, SD=4, n=12) of the total uptake during experiments was attributed to leaf uptake. Therefore, approximately 74% (range=61 83%, SD=7, n=12) of the NH 4 + was removed by other mechanisms including phytoplankton up take, adsorption of NH 4 + ions to flume walls, volatilization of ammonia gas, and nitrification (Laws 1984; Dugdale and Wilkerson 1986). Because this study was designed to investigate the effects of velocity on NH 4 + uptake, a comprehensive mass balance appr oach intended to quantify the contribution of all potential mechanisms of 15 N removal was not taken. However, the majority of the 74% is assumed to have been taken up by organisms in the water column since adsorption of NH 4 + to container walls has been sh own to be minimal (Dugdale and Wilkerson 1986; Slawyk and Raimbault 1995) and pH and nitrate concentrations remained relatively constant throughout these experiments (Cornelisen and Thomas, unpubl. data). Discussion Utilization of 15 N labeled ammonium in flume experiments has enabled us to isolate the effects of water velocity on NH 4 + uptake by epiphytes while measuring the effects of velocity on NH 4 + uptake by an assemblage of organisms that included epiphytes, seagrass leaves, and phytoplankton. The ou tcome of these experiments indicate that (1) rates of NH 4 + uptake for epiphytes are dependent on water velocity to an extent that suggests that uptake by epiphytes is physically limited (Fig. 2A, 3A), and (2) estimates of NH 4 + rates normalized to Chlorophyll a ( r Chl ) are tightly coupled to total

PAGE 37

24 NH 4 + removed from the water column by the assemblage (Fig. 4). These results demonstrate the utility of using isotope labels in flume studies to investigate the effects of water flo w on nutrient dynamics of individual components of a community without separating the component from other community members. Based on our data, rates of NH 4 + uptake for epiphytes increased by an order of magnitude over a range of water velocity (0.02 0.20 m s 1 ) commonly observed in the field (Fig. 2A; Fonseca and Kenworthy 1987; Koch and Gust 1999; Thomas and Cornelisen; unpubl. data). This result emphasizes the importance of water flow in regulating NH 4 + uptake of this integral component of seagrass systems and is consistent with previous studies that have demonstrated flow dependent nutrient uptake by individual organisms (e.g. Gerard 1982; Hurd et al. 1996) and communities (e.g. Thomas and Atkinson 1997; Thomas et al. 2000). The dependence of r Chl on velocity ( r Chl ~U b 0.8 ) was on the order of that expected for mass transfer limited uptake (Fig. 2A; Kays and Crawford 1993). Furthermore, despite differences in epiphyte composition between donor sites, rates of NH 4 + uptake normalized to Chl a ( r Chl ) were comparable between sites and similarly affected by water velocity (Fig. 2A, 4). Measurements of Chl a provided an estimate of the fraction of an epiphyte sample that is involved in photosynthesis and responds to nutrient availability (Frankovich and F ourqurean 1997). Thus normalizing to Chl a allowed for comparison among sites that differed in relative abundance of autotrophs and heterotrophs and provided estimated NH 4 + uptake rates specific to those epiphytes that actively removed NH 4 + from the water column. We recognize that organisms other than those that contain Chl a (i.e. heterotrophic bacteria) can remove ammonium from the water column (Hoch and Kirchman 1995); however,

PAGE 38

25 normalizing to Chl a in our experiments accounted for the differences in epi phytes from the two sites. Therefore it is likely that uptake rates for these epiphytes were dependent on their autotrophic fraction. Further, uptake rates were dependent on water velocity to the order expected for mass transfer limited uptake indicating that rates of NH 4 + uptake by these epiphytes were likely physically limited. Uptake rate constants (S) for the entire assemblage were also dependent on water velocity (Fig. 2B) and the relationship between S and velocity is similar to that reported for se agrass communities in the field (Thomas et al. 2000). The total N removed from the water column by the assemblages, calculated from these values of S, were compared to r chl for the epiphytes (Fig. 4). Rates of NH 4 + uptake for epiphytes ( r chl ) were a linear function (over the range of our data) of the total NH 4 + removed by the entire assemblage indicating that NH 4 + uptake by epiphytes is tightly coupled to uptake by the e ntire assemblage. It is possible that this relationship is non linear; however, we do not have data in the lowest range of NH 4 + uptake. A non linear relationship is suggested if uptake by epiphytes occurs over all ranges of uptake by the assemblage and th e intercept is zero (no uptake by epiphytes when there is no uptake by the assemblage and vice versa). Our results suggest that a non linear relationship may exist since the relative contribution of epiphytes increases with total NH 4 + removed by the assem blage. Such a relationship between uptake by epiphytes and uptake by the community would indicate that epiphytes become an increasingly important contributor to NH 4 + uptake as total NH 4 + removed from the water column increases with velocity. The close rel ationship between NH 4 + uptake by epiphytes and the assemblage also suggests that epiphyte populations may be a good predictor of community scale nutrient

PAGE 39

26 uptake (or vice versa). For example, there is a significant relationship between NH 4 + uptake rates fo r epiphytes ( r Chl ) and uptake rate constants (S) for the entire assemblage ( r Chl = 0.65 10 8 (S)+1.4, r 2 =0.90, P<0.001). Therefore, the contribution of epiphytes to uptake by a community in the field can be predicted from measurements of S, chlorophyll c oncentrations (mg Chl m 2 ) in the epiphytes, and water velocity in situ Further, uptake rate constants (S) can be predicted from equations describing heat and mass transfer (Thomas et al. 2000) and these models may prove useful for predicting uptake rate s by epiphytes. As implied by the close relationship between uptake by epiphytes and the community (Fig 4) and the high dependence of NH 4 + uptake by epiphytes on water velocity, epiphytes contribute a significant portion of the velocity dependent NH 4 + upta ke by the entire assemblage (Fig. 3A). Seagrass leaves were also positively affected by water velocity, however, contributed less than epiphytes to total uptake by the assemblage (Fig. 3B). In addition, unlike the epiphytes the rate of NH 4 + uptake for sea grass leaves was not dependent on the total NH 4 + removed by the assemblage. It is well documented that seagrass plants can remove a significant portion of their required nitrogen from the water column (e.g. Iizumi and Hattori 1982; Lee and Dunton 1999) an d it has been suggested that this uptake contributes largely to the overall input of dissolved inorganic nitrogen (DIN) to seagrass communities (Hemminga et al. 1991). Our results provide evidence that epiphytes are also an important pathway for the entry of DIN from the water column into seagrass communities. Seagrass plants utilize both water column and sediment nutrients (Iizumi and Hattori 1982; Lee and Dunton 1999), which compounds the complexity of biogeochemical processes in seagrass communities and the

PAGE 40

27 ability to estimate the relative contribution of epiphytes to total DIN inputs. Nonetheless, our data reveal that epiphytes contribute to DIN inputs and identifies water velocity as a major factor controlling inputs of DIN via both epiphyte and seagr ass uptake into seagrass communities. The reverse relationship between sites for total uptake of NH 4 + by epiphytes and seagrasses (Fig. 3) support earlier suggestions that epiphytes compete with their host plant for water column nutrients (Hemminga et al. 1991; Williams and Ruckelshaus 1993). Differences in NH 4 + uptake between sites for the seagrass may be due to variations in nutrient availability between locations and uptake affinity of the leaves (Lee and Dunton, 1999). However, close proximity of the donor beds and similar N tissue concentrations suggest that the two sites have a comparable nutrient history. If this is the case, differences between sites may be due to indirect competition for nutrients between the grass and epiphytes brought on by var iations in spatial coverage of attached organisms (Johnstone 1979) or in the depleted concentration gradient at the surface of the seagrass leaves caused by epiphyte uptake (Sand Jensen et al. 1985). In addition to the above findings, our data suggest tha t variable effects of velocity on NH 4 + uptake by different components (i.e. seagrass, epiphytes, and phytoplankton) of seagrass communities collectively contribute to the overall NH 4 + uptake response of the assemblage to changes in water velocity. For exam ple, NH 4 + uptake by benthic components (epiphytes and seagrass leaves) was more dependent on velocity than uptake by the entire assemblage (Fig. 3). The depressed relationship for the assemblage may indicate that it is in a transitional phase rather than b eing mass transfer limited (e.g. Sanford and Crawford 2000). It is also possible that the depressed relationship is due to

PAGE 41

28 bending of the canopy with flow as suggested by Thomas et al.(2000). An alternative explanation suggested by our data is that individ ual components of a community are mass transfer limited while others are not. For example, if ammonium uptake by phytoplankton of small class sizes are independent of water velocity (e.g. Karp Boss et al. 1996), while benthic components are velocity depen dent (seagrass, epiphytes), then combined effects of velocity on the whole community will be depressed relative to that of the benthos alone. Follow up research in field flumes (Thomas et al. 2000) employing a mass balance approach will provide a more in depth analysis of the relative contributions of individual components to total nutrient uptake and the dependence of uptake on velocity. The application of 15 N labeled ammonium in flume experiments has allowed us to isolate the effects of water velocity o n NH 4 + uptake by epiphytes while they were situated within a seagrass assemblage. As a result, we have been able to demonstrate the importance of water velocity in controlling rates of NH 4 + uptake by this important component of seagrass communities. Furth ermore, our data indicate that epiphytes are tightly coupled to uptake by the entire community and along with seagrass leaves contribute to DIN inputs to seagrass communities as well as the uptake response of the assemblage to water velocity. Future studie s conducted in situ through the use of field flumes and labeled nutrients will provide valuable information on the role of the physical environment on nutrient cycling processes at the scale of individual organisms (i.e. seagrass plants) and groups of orga nisms (i.e. epiphytes and phytoplankton) and how these processes are reflected in the response of the entire community.

PAGE 42

29 CHAPTER THREE AMMONIUM AND NITRATE UPTAKE BY LEAVES OF THE SEAGRASS THALASSIA TESTUDINUM : IMPACT OF HYDRODYNAMIC REGIME AND EPIPHYTE COVER ON UPTAKE RATES Introduction Seagrasses assimilate dissolved inorganic nitrogen (DIN) from both water column and sediment pools (Iizumi and Hattori 1982; Short and McRoy 1984; Stapel et al. 1996). Although seagrass roots are exposed to DIN concent rations that are an order of magnitude greater than water column concentrations, their leaves can account for a significant portion of total N acquisition (Iizumi and Hattori 1982; Short and McRoy 1984; Lee and Dunton 1999). Lee and Dunton (1999) developed an annual nitrogen budget for Thalassia testudinum that provides valuable data on DIN uptake kinetics for both T. testudinum roots and leaves. The contribution of roots and leaves to the total N budget was found to be approximately equal (52% for roots a nd 48% for leaves). In addition, uptake kinetics and affinity for the leaves, derived from a Michaelis Menten model, were shown to be significantly higher for ammonium than for nitrate (Lee and Dunton 1999). These findings were consistent with earlier stu dies that demonstrated a preference for the reduced form of nitrogen over NO 3 (e.g. Short and McRoy 1984; Terrados and Williams 1997). The available database on uptake of DIN by seagrasses is largely based on experiments involving plants isolated from o ther organisms (i.e., epiphytes) and data

PAGE 43

30 analysis using the Michaelis Menten model (e.g., Iizumi and Hattori 1982; Terrados and Williams 1997; Lee and Dunton 1999). These studies provide insight on the physiology and enzyme kinetics of seagrasses; howev er, they do not account for factors that can influence uptake by seagrass leaves in a natural canopy, such as water flow and the presence of epiphytes. If the rate at which DIN is delivered to the leaf surface exceeds the rate at which the DIN can be proce ssed by the leaf, then the uptake rate will be largely controlled by kinetic factors such as the speed of enzymatic reactions or the availability of uptake sites. Conversely, if the rate at which the leaf processes a nutrient exceeds the delivery rate, th en uptake rates can be influenced by (1) water flow, which can affect diffusive boundary layer thickness adjacent to the leaf surface (Koch 1994), and (2) epiphyte cover (Johnstone 1979), which can reduce the number of active uptake sites available for upt ake. The relative impact of these physical factors on uptake can vary over time (Koch 1994) and intermediate levels of kinetic vs. physical control can also occur (Sanford and Crawford 2000). While some data exist on the individual effects of water flow a nd epiphyte cover on DIN uptake, there is little information on how these factors collectively affect uptake by seagrass leaves in their natural environment. Seagrasses are primarily found in shallow coastal waters that are exposed to a range of hydrodyn amic conditions (Fonseca and Kenworthy 1987; Koch and Gust 1999). Increased water velocity and turbulence associated with waves and tides influence important ecological processes in seagrass communities, including nutrient uptake (Thomas et al., 2000; Tho mas and Cornelisen, 2003), photosynthesis and productivity (Fonseca and Kenworthy 1987; Koch 1994), and the dispersal of pollen (Ackerma n 1986) and seeds (Orth et al. 1994). Dense stands of seagrass shoots and associated epiphytes

PAGE 44

31 collectively form a roug h and flexible surface that strongly influences characteristics of water flow in and above the canopy. Turbulent energy and stresses are typically greatest at the top of the seagrass leaves (Gambi et al. 1990; Nepf and Vivoni 2000) and penetrate further in to the canopy as flow speed increases (Ghisalberti and Nepf 2002). The rate at which turbulent energy is dissipated at the surface of the benthos ( e ) can be correlated to diffusive boundary layer thickness at the uptake surface. Rates of nutrient uptake are expected to be proportional to e (uptake rate e 0.25 ) when they are limited by rates of delivery (Hearn et al. 2001). Further, if shear genera ted turbulence at the benthic surface is balanced by its dissipation then uptake rates should also be proportional to bottom shear stress (uptake rate t 0.4 ; Hearn et al. 2001 ) These relationships are proposed for non flexible benthic communities (e.g., coral reef flats); however, it is possible that they apply to a broad range of benthos including plant canopies. In flexible canopies, sources other than shear can generate turbulence, which leads to imbalances between production and dissipation of turbu lence within various regions of the canopy (see Raupach et al. 1996; Ghisalber ti and Nepf 2002). However, the relationships derived by Hearn et al. (2001) may be transferable to seagrass beds despite any small scale inequalities between production and dis sipation occurring among the seagrass shoots. This is especially true if estimates of e within a roughness length of the bottom even out any small deviations in the balance of turbulent energy within the canopy. Further, since uptake by the benthos as a whole is driven by the organisms that comprise the benthos, analogous relationships between hydrodynamic parameters and uptake rates for individual benthic components (seagrass plants and epiphytes) are proposed to exist.

PAGE 45

32 In addition to water flow, epiph ytes attached to the leaf surface can influence the rate of uptake by the leaves. Symbiotic interactions between seagrasses and epiphytes h ave been described (Harlin 1973; McRoy and Goering 1974; Penhale and Thayer 1980; Libes 1986); however, their relatio nship may not be entirely mutual since epiphyte cover may act as a barrier to nutrient uptake (Sand Jensen 1977; Johnstone 1979) or deplete nutrient concentrations within the diffusive boundary layer at the leaf surface (Sand Jensen et al. 1985). Conversel y, there may be some interplay between epiphytes and water flow that could enhance uptake by seagrass leaves (Koch 1994). These interactions are further complicated by the potential impact of shading by epiphytes on seagrass productivity (Sand Jensen 1977 ), which in turn could influence DIN uptake by the leaves. Data from Lee and Dunton (1999) suggest that the form of DIN being assimilated (NH 4 + vs. NO 3 ) will influence the extent to which uptake rates are affected by water flow and epiphyte cover. Michae lis Menten parameters (i.e., V max ) estimated for T. testudinum indicate that the capacity for T. testudinum leaves to take in ammonium is over twice as high as the capacity to take in nitrate (Lee and Dunton 1999). Uptake rates for NH 4 + may be more depend ent on water flow than rates of NO 3 uptake since uptake of nitrate relies heavily on physiological factors (i.e., nitrate reductase (NR) activity and the availability of stored carbohydrates; Roth and Pregnall 1988; Touchette and Burkholder 2000). Thus, rates of ammonium uptake for seagrass leaves may be influenced by water flow (Cornelisen and Thomas 2002); whereas, rates of nitrate uptake may be limited by physiological factors that could reduce or possibly eliminate enhancement of uptake rates due to t hinning of the diffusive boundary layer. If the delivery rate of NO 3 to the bare surface of the leaf exceeds the rate at which the leaf can process nitrate, then epiphyte

PAGE 46

33 cover would also be expected to have little impact on uptake rates. Conversely, epi phytes would be expected to significantly impact the rate of ammonium uptake by the leaves. In the present study, the concomitant effects of hydrodynamics and epiphyte cover on rates of DIN uptake for T. testudinum leaves are investigated while they are s ituated in natural seagrass beds. Ammonium and nitrate are used in two separate experiments in order to compare the effects of water flow and epiphyte cover on two forms of DIN that vary in their physiological requirements for uptake. To isolate effects of these factors on uptake by the leaves, 15 N labeled ammonium and 15 N labeled nitrate is applied in a field flume deployed in a seagrass bed and uptake rates are measured over a range of water velocity. Hydrodynamic parameters, including energy dissipati on rate and bottom shear stress, are estimated from velocity profiles collected with an acoustic Doppler velocimeter. Dependence of uptake rates on these parameters is then assessed for leaves both with and without epiphytes covering their surface. The pot ential implications of our findings for nutrient uptake by seagrasses in natural communities are discussed. Methods Flume deployment Isotope labels were applied in a flume deployed in natural seagrass beds in order to isolate the effects of water fl ow and epiphyte cover on the rate of NH 4 + and NO 3 uptake by Thalassia testudinum leaves. Uptake rates for ammonium and nitrate were measured between 7 and 14 June 2001 and between 17 and 22 June 2001, respectively. For measuring rates of ammonium uptake the field flume was deployed within a T. testudinum bed located in Pass A Grille Channel, which is approximately 2 km north of Fort Desoto County Park at the southernmost point of Pinellas County,

PAGE 47

34 Florida (Fig. 5). Due to limited accessibility to the me adow at the Pass A Grille site, uptake rates for nitrate were measured within the park boundaries approximately 2 km south of Pass A Grille Channel. Weather conditions and characteristics of these two sites were similar and are provided in Table 1. All r ates were measured between 1000 and 1500 h and within 100 meters of the shoreline in water depth less than the flume height (0.8 m). The field flume enclosed a 3.7 m 2 area of the seagrass bed and enabled controlled unidirectional flow over the community and measure nutrient uptake by the seagrass enclosed in the flume (Fig. 6; See Thomas et al. 2000; Thomas and Cornelisen 2003). Uptake rates for NH 4 + were determined from a series of nine flume experiments, each of which was conducted at a velocity random ly chosen from a predetermined range observed under natural flows in seagrass beds (0.02 0.18 m s 1 ). A total of seven flume experiments over a similar range of velocity (0.03 0.17 m s 1 ) were conducted for measuring rates of NO 3 uptake. The flume w as moved to a new location in the seagrass bed for the measurement of each uptake rate. After ensuring the flume was sufficiently sealed from the surrounding water, a spike of 15 NH 4 + (as 20 mmol/L 98 atom % 15 NH 4 Cl) or 15 NO 3 (as 20 mmol/L 98% 15Na NO 3 ) was added to the flume to obtain a beginning water column concentration of ~ 6 m mol/L for NH 4 + and ~ 4 m mol/L for NO 3 The spike was added slowly (over ~3 minutes) through a tube leading to the motor box and the trolling motor was used to assist in mixing the water to obtain a uniform beginning concentration. Ambient water samples were collected prior to each experiment to confirm the contribution of background nutrients to the total amount of nutrients in the flume. Several parameters were measured

PAGE 48

35 Old Tampa Bay Hillsborough Bay Tampa Bay Manatee River Fort Desoto County Park N Gulf of Mexico Tampa St. Pete NO 3 NH 4 27 30 27 45 Study sites Tampa Bay Area F l o r i d a 5 km Figure 5. Area map of Tampa Bay indicating lo cations of flume experiments for measuring rates of ammonium and nitrate uptake.

PAGE 49

36 Ammonium Nitrate Average water depth (m) 0.63 ( 0.05, 9) 0.58 ( 0.09, 7) Water temperature ( C) 32.1 ( 2.2, 9) 32.9 ( 2.1, 7) Ambient NH 4 + concentration s ( m M) 0.51 ( 0.22, 9) 0.29 ( 0.13, 7) Ambient NO 3 concentrations ( m M) 0.10 ( 0.10, 9) 0.08 ( 0.07, 7) SEAGRASS Biomass (g m 2 ) 132 ( 32, 9) 127 ( 24, 7) Shoots per m 2 430 ( 104, 45) 376 ( 119, 35) Leaves per m 2 116 6 ( 212, 9) 1171 ( 242, 7) Canopy height in still water (m) 0.24 ( 0.03, 9) 0.32 ( 0.03, 7) % N content 2.13 ( 0.21, 18) 1.98 ( 0.13, 14) EPIPHYTES Biomass (g m 2 ) 201 ( 60, 9) 142 ( 44, 7) % N content 0.98 ( 0.13, 18) 1.46 ( 0.14, 14) Chl a (mg Chl a g dry wt 1 ) 0.84 ( 0.26, 27) 1.32 ( 0.36, 21) % of leaf surface covered by epiphytes 90 ( 4, 12) 73 ( 15, 12) % Organic content 23.7 ( 1.3, 6) 24.0 ( 2.3, 6) Composition Diatoms, blue green algae, particulates, red macroalga e Diatoms, corallines, blue green algae, red macroalgae Table 1 Characteristics of the study sites where field flume experiments were conducted to measure uptake rates for ammonium (left column) and nitrate (right column). Nine experiments were conduc ted at the NH 4 + site and 7 at the NO 3 site. Numbers in bold represent the mean of n samples and numbers in parentheses are 1 S.D. and the number (n) of samples.

PAGE 50

37 2.4 m 0.8 m Trolling motor Flow straighteners ADV 1.2 m Figure 6. Diagram of the field flume used for isolat ing a section of the seagrass bed and conducting uptake experiments. The flume was constructed of an aluminum frame and clear Lexan walls. The turn sections were made from half sections of fiberglass pipes. The flume extended 5 cm into the substratum (i ndicated by the shaded region) and isolated the community from water outside the flume. The trolling motor was powered by a 12 volt marine battery and controlled using a diode system (Linear Power Systems, Clearwater, Florida, USA) for continuous unidirec tional flow over the community. Velocity profiles were collected with an ADV (Sontek, San Diego, California, USA; 10 Hz field model) at the location shown.

PAGE 51

38 during each flume deployment, including water height, temperature, and the deflected height (h d ) of the canopy in flowing water. Deflected height was based on an average of ten measurements taken manually with a meter stick while the water was flowing. Prior to placing the flume over the seagrass bed, epiphytes were completely removed from six rand om leaves within a 0.25 m 2 area in order to measure uptake rates for seagrass leaves without epiphyte cover. New and senescent leaves were not used as leaves without epiphyte cover due to the difficulty in completely removing epiphytes from senescing leav es and the virtual absence of epiphytes on the new leaves. Epiphytes were easily removed by gently rubbing and scraping the leaves between the thumbnail and index finger while snorkeling above the canopy. There were no epiphytes visible to the eye on these leaves following collection. While the effectiveness of epiphyte removal was not assessed using a microscope for this study, previous analysis of leaves that were collected from the same locations that had epiphytes removed indicated effective (> 95%) re moval using this technique (Cornelisen, unpublished data). Hydrodynamic characterization Velocity profiles were collected during each flume experiment to describe the vertical distribution of mean velocity, Reynolds stress, and turbulent energy in and above the canopy and to estimate hydrodynamic parameters ( e and t ) from the logarithmic portion of the velocity gradient. Velocity data were collected using an acoustic Doppler velocimeter (Field ADV; YSI/Sontek) that measures velocity in three dimension s: longitudinal along the main flow (U), transverse (V), and vertical (W). The probe was affixed to a metal rod that could be moved vertically to position the sensors at a specific height above the bottom. Profiles were collected in the center of the

PAGE 52

39 wor king section approximately 1.5 m downstream from the first turn (Fig. 6) and within the area where leaves were collected for determining uptake rates. For each velocity profile data were recorded at 5 Hz for 1 minute (n = 300) at each of 10 to 12 heights above the sediment water interface. Heights included ~ 4 6 measurements within the canopy and at least 5 above the seagrass up to ~ 10 cm beneath the surface of the water. Care was taken to ensure that the sampling volume was sufficiently above the bo ttom when data was collected near th e sediment water interface (Fin elli et al. 1999 ). When data were collected in the canopy, leaves that were directly in contact with the ADV sensors were trimmed to prevent interference with data collection. This techni que has been shown to have no significant effect on flow measurements taken in vegetative canopies (Ikeda and Kanazawa, 1996). Mean velocity for each of the components ( U V and W ), Reynolds stress ( ' W U ) and total turbulent energy (K=0.5[ ' U U + ' V V + ' W W ]) were calculated using the velocity data collected at each height (Denny 1988; Nikora et al. 1998; Dade et al. 2001). Signal to n oise ratios were well above the recommended 15 db level and the correlation values for the three sensors consistently ranged between 85 and 95%. The vertical gradient of mean velocity in the dominant direction of flow ( U ) was used to o btain estimates of shear velocity (U ) and roughness length (Z o ) by employing the commonly cited Karman Prandtl equation: U = U / k ln (Z/Z o ) (1)

PAGE 53

40 where U is the mean velocity at a given distance from the botto m, k is the von Karman constant (0.4), and Z is the height above the sediment water interface. This method utilized the portion of the profile in which velocity increased logarithmically with increasing height (Z) above the bottom. The roughness length ( Z o ) was estimated as the intercept of the velocity vs. ln Z plot (see Denny, 1988). Confidence limits (95%) on estimates of U were based on the number of measurements in the logarithmic portion of the vertical profile and the regression coefficient (r 2 ) and using the expression outlined in Grant et al. (1984). Estimates of bottom shear stress ( t ) were based on the relationship t = r U 2 where r is the density of seawater (Denny 1988). The rate of energy dissipation ( e ) at the height of the roughness len gth (Z o ) was estimated using the equation e =( t/r ) 3/2 / k Z o (Hearn et al., 2001). Estimates of depth averaged velocity (U b ) were also obtained from each profile using equation 1. Equation (1) can be modified to include an estimate of the displacement height (d). Values of d were on the order of 75% ( 35%) of the canopy height, which is consistent with previous studies on vegetated canopies (e.g., Nepf and Vivoni 2000; Dong et al. 2001). Application of the displacement height in equation 1 resulted in incr eased variability in estimates of Z o and U and did not improve the fit of the data to the equation and therefore d was omitted. Similar results have been noted in studies on terrestrial systems and it has been demonstrated that better estimates of U can be made by assuming d to be zero (e.g. Dong et al., 2001). Collection of seagrass and epiphyte samples Seagrass leaves with and without epiphyte cover were collected before and immediately following each flume experiment for

PAGE 54

41 determining the rate at w hich NH 4 + or NO 3 accumulated within their tissue. Epiphytes were gently scraped from the seagrass leaves (n = 10) that were covered in epiphytes and along with the seagrass leaves were retained for analysis. Epiphytes removed from the seagrass leaves we re pooled and split into two portions, with half designated for 15 N analysis and half for determination of Chlorophyll a concentration. Seagrass samples, which consisted of whole leaves, were rinsed thoroughly with filtered seawater. Epiphytes were rinsed with filtered seawater over a 35 m M screen. Following the filtered seawater rinse, epiphytes were briefly rinsed with DI water to remove salt (Winning et al. 1999). Samples of seagrass and epiphytes were also collected from an area away from the flume and processed in a separate area f or determination of ambient 15 N in the tissues. All samples were placed in separate whirl pak bags and ke pt in a seawater ice mixture (~ 0 C) until sample processing at the laboratory (within 6 hours). Samples were dried at 60 C, ground to a fine powder and placed in glass vials for transport to the Isotope Ratio Mass Spectrometry (IRMS) facility. Chlorophyll a concentration in the epiphyte tissues was estimated using spectrophotometric methods as outlined in Strickland and Parsons ( 1968 ). For each stu dy site, random segments (~2 cm in length) from twelve whole leaves with epiphytes in tact were analyzed to quantify the percentage of leaf area covered by the epiphytes. Epiphyte cover on scenescent and newly formed leaves was not quantified since these same leaves were not included in 15 N analysis. Digital images were taken of each leaf segment using a Nikon camera mounted on a dissecting microscope. Images were imported into Image Pro, where areas of bare leaf were traced and quantified using a calibra ted polygram measurement tool. Total area of bare leaf was divided by total area

PAGE 55

42 of the segment to determine the percentage of the leaf surface that was bare of epiphytes. The average of the twelve estimates was taken as an estimate of epiphyte cover for each site. Five estimates of shoot and leaf density were obtained at the site of each flume experiment using a randomly placed quadrat (0.25 m 2 ). In addition, all leaves with epiphytes within a 0.01 m 2 quadrat were collected, separated, and dried to obt ain estimates of biomass for the seagrass leaves and the epiphytes. Calculation of uptake rates Dried, homogenized samples of seagrass leaves with and without epiphyte cover and the epiphyte samples were analyzed using EA IRMS (elemental analyzer isotop e ratio mass spectrometry) for determination of nitrogen content (% N) and atom % 15 N. Specific uptake rates (V) for the samples were calculated using the following equation: V = ( d a s / d t)/(a w a s ) (2) where a s is the atom % 15 N in the components tiss ue, a w is the atom % 15 N of the enriched substrate, and t is time (in seconds) (Dugdale and Goering 1967). The units for V are g N removed (g N tissue) 1 s 1 or simply s 1 The atom % 15 N of the enriched water (a w ) was based on the amount of 98 atom % 1 5 NH 4 + or 15 NO 3 added and background DIN concentrations (assumed to reflect 15 N concentration of atmospheric N ~0.37 atom % 15 N). The numerator ( d a s / d t) was estimated as the difference in atom % 15 N between ambient samples and samples collected at the end of each experiment divided by the duration of the experiment, which ranged from 45 to 60 minutes. It is noted that the use

PAGE 56

43 of equation (2) in calculating V assumes that the atom % 15 N of the source pool did not change during the course of the experiment. Dilution of 15 N in the source pool resulting from inputs of non labeled NH 4 + (via regeneration, excretion) or NO 3 (via nitrification) into the water column would expectedly result in underestimated uptake rates (Laws 1984). While this is a potential sou rce of error, the short duration of these experiments along with the high concentration and atom % 15 N of the spike likely minimized dilution. Each specific uptake rate (V) was normalized to the nitrogen concentration (% N) of the tissue to calculate an uptake rate ( r ) in units g N removed (g dry wt) 1 s 1 (Dugdale and Goering 1967). For the epiphytes, uptake rates ( r ) were also normalized to chlorophyll a concentration in the sample to obtain an uptake rate that was representative of the autotrophic fraction ( r Chl = r Chl a 1 in units g N removed (mg Chl a) 1 s 1 ; Dickson and Wheeler 1995; Frankovich and Fourqurean 1997; Cornelisen and Thomas 2002). Since the decline in ammonium or nitrate concentration in the water column is a first order decline (See Thomas et al. 2000), the rate of 15 N accumulation in the seagrass leaves is assumed to be first ord er. Equation 2 assumes linear uptake and any change in water column concentration over time will influence the calculation of uptake rates. Therefore, uptake rates ( r ) must be multiplied by a correction term ( a ) to compensate for changes in water column concentration. For instance, water column concentration during measurement of uptake rates at a high velocity was depleted more than during measurements at low velocity. As a result, r would be underestimated if concentration were assumed to remain consta nt. The correction term was calculated as the average value of 1/e kt over the course of the experiment, where k is the first order rate of decline in concentration and t is time. Water samples were collected over the duration of each

PAGE 57

44 experiment and valu es of k were estimated as the slope of the natural log of concentration in these samples versus time (see Thomas et al. 2000 for details). Samples were analyzed with an autoanalyzer to determine concentration (to an accuracy of 0.1 m mol/L) of ammonium and nitrate. Uptake rates corrected for concentration change ( r a ) are considered potential uptake rates for the beginning water column concentration (6 m M for NH 4 + experiments and 4 m M for NO 3 experiments) and are used here to asse ss the effects of water flow and epiphyte cover on uptake by seagrass leaves. Effects of water flow on DIN uptake rates If rates of ammonium or nitrate uptake for the seagrass leaves are limited by the rate at which the nutrient is delivered to the le af surface, then uptake rates will be dependent on factors that influence the thickness of diffusive boundary layers. Several hydrodynamic parameters, including e and t can be correlated to diffusive boundary layer thickness and therefore uptake rates (see Hearn et al. 2001). If uptake rates for the seagrass leaves were limited by the rate of delivery, r would be expected to be proportional to e to the quarter po wer ( r @ e 0.25 ) and t to the 0.4 power ( r @ t 0.4 ) (see Hearn et al. 2001). The dependence of uptake rates ( r ) on e and t was evaluated using regression and comparing the data to the expected rela tionship. Because there was error in estimating both the dependent and independent variables, model II regressions (geometric mean) were used to calculate the slope and con fidence limits (Sokal and Rohlf 1995). Correlations between uptake rates and hydro dynamic parameters were also estimated using the product moment correlation coefficient (r).

PAGE 58

45 Effects of epiphyte cover on DIN uptake rates To assess the effects of epiphytes on uptake, uptake rates ( r ) for seagrass leaves with and without epiphyte cover were normalized to the average proportion of the leaf area that was not covered by epiphytes. If rates of uptake were limited by the rate of delivery to bare areas of the seagrass leaf, then uptake ra tes normalized to the proportion of bare area would be expected to be similar between leaves with and without epiphyte cover. Furthermore, if epiphytes enhanced rates of delivery, then uptake rates per bare leaf area for leaves with cover would be higher than those for leaves without cover. Uptake rates for leaves with epiphytes could also exceed those for leaves without epiphytes if uptake rates are controlled by the ability of the leaf to process the nutrient rather than the rate of delivery to the bare surface. For leaves covered in epiphytes, uptake rates were normalized to the proportion of the total surface area of the leaf that was not covered by epiphytes to obtain an uptake rate in g N removed (m 2 leaf area) time 1 This uptake rate was calcu lated as r proportion bare area g tissue per m 2 of leaf area, where leaf area is equal to mean overall leaf length times the mean leaf width. Uptake rates for leaves without epiphytes were normalized to total leaf area. For comparison purposes, uptake rates f or epiphytes were normalized to the proportion of the leaf area that they covered. Results Hydrodynamic characteristics Vertical profiles of mean velocity ( U V and W ) reveal that water flow in the flume was highly unidirectional (Fig. 7). Velocity profiles closely resemble those described in natural tide driven flows in seagrass beds (e.g., Koch and

PAGE 59

0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 0.00 0.05 0.10 0.15 0.20 0.25 Velocity, U (m s -1 ) Z/h d U b = 0.17 m s -1 U b = 0.04 m s -1 -0.05 0.00 0.05 Velocity, V (m s -1 ) -0.05 0.00 0.05 Velocity, W (m s -1 ) Figure 7 Example of two velocity profiles collected during flume experiments for measuring NH 4 + uptake rates, including one conducted at low ( o ) and one at high ( ) flow. Profiles of mean velocity in the main flow (U), transverse (V), and vertical (W) directions are shown for heights above the bottom (Z) normalized to deflected canopy height (h d ). Mean velocity at each height was based on the average of 300 measurements (Collected at 5 Hz for 1 minute with ADV). Bars represent 1 standard deviation (equivalent to rms and turbulence). The dotted line indicates the height of the canopy (Z/h d = 1). Depth averaged velocity (U b ) was 0.04 and 0.17 m s 1 shear velocity (U ) was 0.012 and 0.034 m s 1 and roughness length (Z 0 ) was 0.086 and 0.043 cm for the profile c ollected at low and high flow, respectively. Actual measured canopy height during these profiles was 0.19 cm for the low flow experiment and 0.14 cm for the high flow experiment. The above profiles were typical of those collected for the study. 46

PAGE 60

47 Gust 199 9). Flow was attenuated within the canopy; however, at high velocity flow penetrated deeper into the canopy and attenuation was less pronounced (see Fig. 7). Depth averaged velocity (U b ) ranged between 0.03 and 0.17 m s 1 and 0.03 and 0.18 m s 1 in flum e experiments for measurement of NH 4 + and NO 3 uptake rates, respectively. Distributions of Reynolds stress ( ' W U ) and turbulent energy (K) indicated that shear and turbulence was highest at the top of the canopy and decreased as the sedi ment water interface was approached (Fig. 8). At high velocity (U b ), estimates of Reynolds stress and energy were up to 5 times as high at the top of the canopy as those estimated from velocity profiles collected at low U b Furthermore, Reynolds stress a nd energy penetrated deeper into the canopy as U b was increased (see Fig. 8). Flow above the deflected canopy increased logarithmically with height and fits of the velocity ( U ) data to the Karman Prandtle equation were significant (P < 0.05, range of r 2 = 0.90 0.99, mean r 2 = 0.95) for all profiles collected during NH 4 + uptake experiments and for six of the seven profiles collected during NO 3 uptake experiments. During one flume experiment for nitrate, a full profile was not obtained due to equipment fa ilure and data was limited to an estimate of depth averaged velocity (U b ) based on measurements taken at a height equidistant between the canopy and the waters surface. Estimates of shear velocity (U ) ranged between 0.010 and 0.034 m s 1 (95% CLs range = 0.002 to 0.011 m s 1 ) during NH 4 + uptake experiments and between 0.021 and 0.050 m s 1 (95% CLs range = 0.006 to 0.017 m s 1 ) during NO 3 uptake experiments. Roughness length (Z o ) ranged between 0.03 and 0.17 m for NH 4 + uptake experiments and bet ween 0.09 and 0.19 m for NO 3 uptake experiments. As U b increased, the canopy deflected and Z o decreased in experiments involving NH 4 + (Z o = 0.01U b 0.76

PAGE 61

48 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 -5 5 15 25 35 45 Reynolds stress, -U'W' (m s -1 ) 2 x 10 5 Z/h d U b = 0.17 m s -1 U b = 0.04 m s -1 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 3 6 9 12 Total turbulent energy, K (m s -1 ) 2 x 10 4 Z/h d U b = 0.17 m s -1 U b = 0.04 m s -1 Figure 8. Vertical profiles of (A) Reyno lds stress ( ' W U ) and (B) total turbulent energy (K = 0.5[ ' U U + ' V V + ' W W ]). Parameters were calculated from the same velocity data used in Fig. 7. Depth above the bottom (Z) was normal ized to deflected canopy height (h d ). Note the enhanced Reynolds stress and turbulence at the top of the canopy (represented by the dotted line) during both experiments and the greater penetration into the canopy during the higher velocity (U b ) experiment.

PAGE 62

49 r 2 = 0.70, P<0.01) and those involving NO 3 (Z o = 0.04U b 0.50 r 2 = 0.67, P<0.05). The proportion of the deflected canopy (h d ) that was penetrated by water flow (represented as 1 Z o /h d ) increased from 20 to 80 % over the range of U b at both sites (see Fig. 9; (1 Z o /h d ) = 1.66U b 0.48 r 2 = 0.71, P<0.01). Bottom shear stress ( t ) ranged between 0.09 and 1.20 N m 2 during measurement of NH 4 + uptake rates and between 0.49 and 2.55 N m 2 during measurement of NO 3 uptake rates. Rates of turbulent energy dissipation ( e ) within the roughness length (Z o ) of the bottom ranged between 0. 02 10 3 and 2.25 10 3 m 2 s 3 during measurement of NH 4 + uptake rates and between 0.15 10 3 and 3.21 10 3 m 2 s 3 for experiments involving nitrate. Canopy characteristics A summary of data on canopy characteristics is provided in Table 1. Water temperatures at the two sites were similar and averaged around 32.5 C. Ambient nutrient concentrations collected during the time of the experiments indicate that ammonium concentrations were higher (~ 0.40 m M) than nitrate concentrations (~ 0.09 m M) and that the site in Pass A Grille Channel had slightly higher NH 4 + concentrations than the site within the park boundaries (Table 1). Shoot density of the seagrass beds at both sites was relatively low (< 400 sh oots m 2 ) in comparison to densities reported elsewhere (e.g., Thomas et al., 2000). The canopy height (without flow) at the site for NO 3 uptake experiments was ~ 8 cm taller than the canopy height at the site for NH 4 + uptake experiments. A quantitative analysis using the dissecting scope indicated that approximately 90% ( 4%) and 73% ( 15%) of the leaf surface was covered with epiphytes during measurement of NH 4 + and NO 3 uptake, respectively.

PAGE 63

50 y = 1.66x 0.48 R 2 = 0.71 0.0 0.2 0.4 0.6 0.8 1.0 0.00 0.05 0.10 0.15 0.20 U b (m s -1 ) 1 Z o /h d Figure 9. The relative proportion of deflected canopy height (h d ) represented by roughness length (Z 0 ) versus bulk velocity (U b ) for all profiles collected during the ammonium and nitrate uptake experiments. Values were subtracted from 1 to provide an indication of the relativ e proportion of the canopy that flow penetrated. For example, the height above the bottom at which the logarithmic portion of the velocity profile extrapolated to 0 was at approximately 20% of the deflected canopy height at low velocity (~ 0.03 m s 1 ) and approximately 70% of the canopy height at high velocity (~ 0.17 m s 1 ).

PAGE 64

51 Differences in percent cover between the sites are consistent with differences in epiphyte biomass (Table 1). Epiphytes at the NH 4 + site were primarily composed of benthic diato ms and blue green algae mixed together within a gelatinous mass covering the leaf surface. There were also a few red macroalgae (i.e., Ceramium sp.) and fine sediments and particulate matter mixed in with the diatom assemblage. At the NO 3 site, epiphyte samples were also largely composed of diatoms and blue green algae, but also contained an abundance of crustose corralines (i.e., Fosliella sp.). Less common but present were encrusting brown algae (i.e., Myrionema sp.) and some red macroalgae. Polychae tes were also attached to the leaf surface. Effects of water flow on DIN uptake rates The rate of NH 4 + uptake by seagrass leaves with epiphyte cover was dependent on hydrodynamic conditions and ranged from 0.10 10 8 to 0.44 10 8 g NH 4 + N (g dry wt ) 1 s 1 (Fig. 10). One of the calculated rates for these leaves was determined to be a significant outlier (P < 0.01, Grubbs test; Sokal and Rohlf 1995) and was removed from data analysis. Based on Model II regressions, the rates of NH 4 + uptake for the lea ves with epiphyte cover were significantly dependent on shear stress ( r = [4.07 10 9 ] t 0.66 95% CLs on slope = 0.28 and 1.03, r = 0.80, P<0.05) and energy dissipation rate ( r = [2.92 10 8 ] e 0.31 95% CLs on slope = 0.17 and 0.45, r = 0.88, P<0.01). Uptake rates for the leaves without epiphyte cover were approximately 9 times those of leaves with epiphyte cover and ranged between 1.00 10 8 and 3.89 10 8 g

PAGE 65

A y = 2.98x R 2 = 0.67 y = 0.34x R 2 = 0.59 0 1 2 3 4 5 0.0 0.4 0.8 1.2 t 0.4 (N m 2 ) r (g NH 4 -N (g tissue) -1 s -1 )x10 8 Leaves without epiphyte cover Leaves with epiphyte cover B y = 15.48x R 2 = 0.73 y = 1.78x R 2 = 0.77 0 1 2 3 4 5 0.0 0.1 0.2 0.3 e 0.25 (m 2 s -3 ) r (g NH 4 -N (g tissue) -1 s -1 )x10 8 Leaves without epiphyte cover Leaves with epiphyte cover Figure 10. The rate of ammonium uptake ( r ) by seagrass leaves with ( o ) and without ( ) epiphyte cover versus (A) bottom shear s tress ( t ) raised to the 0.4 power and (B) energy dissipation rate ( e ) raised to the 0.25 power. Lines represent the best linear fit for the data with an intercept set at 0. Regressions are significant (P<0.01) and demonstrate that rates of NH 4 + uptake fo r seagrass leaves were limited by the rate of delivery. Uptake rates for leaves without epiphyte cover were approximately 9 times higher than those for leaves with epiphyte cover. 52

PAGE 66

53 NH 4 + N (g dry wt) 1 s 1 (Fig. 10). They were also significantly dependent on bottom shear stress ( r = [3.36 10 8 ] t 0.60 95% CLs on slope = 0.29 and 0.92, r = 0.81, P<0.05) and energy dissipation rate ( r = [2.05 10 7 ] e 0.29 95% CLs on slope = 0.16 and 0.42, r = 0.86, P<0.01). Slopes were slightly higher than those expected for r vs. t (0.4) and r vs. e (0.25) (Hearn et al. 2001). This result was partly explained by the application of using a geometric mean regression, which always results in greater slopes than would be provided by usual linear regression (Sokal and Rohlf 1995). Nonetheless, the expected slopes fall within the 95% confidence intervals and the data showed good agreement with the expected relationships (Fig. 10). Rates of ammonium uptake ( r ) for epiphytes normalized to Chl a concentrations ( r Chl ) ranged between 1.26 10 8 and 4 .32 10 8 g NH 4 + N (mg Chl a ) 1 s 1 and were significantly dependent on bottom shear stress ( r = [3.82 10 8 ] t 0.55 95% CLs on slope = 0.31 and 0.80, r = 0.87, P<0.01) and energy dissipation rate ( r = [2.01 10 7 ] e 0.26 95% CLs on slope = 0.16 and 0.36 r = 0.91, P<0.01). As was the case for seagrass leaves, slopes fell within the range of those expected and indicated that uptake rates were limited by the rate of delivery. The rate of nitrate uptake by seagrass leaves with epiphyte cover ranged betwe en 0.046 10 8 and 0.087 10 8 g NO 3 N (g dry wt) 1 s 1 and was not dependent on hydrodynamic parameters (P>>0.05, Fig. 11). Rates of NO 3 uptake for seagrass leaves without epi phyte cover ranged between 0.07 10 8 and 0.14 10 8 g NO 3 N (g dry wt) 1 s 1 and were also not affected by water flow. Rates of NO 3 uptake ( r ) for epiphytes normalized to Chl a concentration ranged between 0.37 10 8 and 0.89 10 8 g NO 3 N

PAGE 67

A 0.00 0.05 0.10 0.15 0.20 0.0 0.5 1.0 1.5 2.0 t 0.4 (N m 2 ) r (g NO 3 -N (g tissue) -1 s -1 )x10 8 Leaves without epiphyte cover Leaves with epiphyte cover B 0.00 0.05 0.10 0.15 0.20 0.0 0.1 0.2 0.3 e 0.25 (m 2 s -3 ) r (g NO 3 -N (g tissue) -1 s -1 )x10 8 Leaves without epiphyte cover Leaves with epiphyte cover Figure 11. The rate of nitrate uptake ( r ) by seagrass leaves with ( open symbols) and without (closed symbols ) epiphyte cover versus (A) bottom shear stress ( t ) raised to the 0.4 power and (B) energy dissipation rate ( e ) raised to the 0.25 power. A velocity profile was not collected during the experiment indicated by the diamond and therefore t and e for these points were based on the relationship between these parameters and bulk velocity (U b ). ( t = 17.3U b 0.04, r 2 = 0.89; e = 0.33U b 2.4 r 2 = 0.91). There were no significant regressions between r and hydrodynamic param eters; however, uptake rates for leaves without epiphyte cover were significantly higher than those for leaves that were covered (paired t test, P<0.01, df = 6). 54

PAGE 68

Ammonium y = 325.49x R 2 = 0.73 y = 511.05x R 2 = 0.80 0 50 100 150 0.00 0.10 0.20 0.30 e 0.25 (m 2 s -3 ) r (g NH 4 -N (m -2 ) s -1 ) x 10 8 Leaves without epiphyte cover Leaves with epiphyte cover Epiphytes Nitrate y = 62.03x R 2 = 0.61 0 5 10 15 20 0.00 0.10 0.20 0.30 e 0.25 (m 2 s -3 ) r (g NO 3 -N (m -2 ) s -1 ) x 10 8 Leaves without epiphyte cover Leaves with epiphyte cover Epiphytes Figure 12. (A) Rates of ammonium uptake for seagrass leaves with ( open sy mbols ) and without ( solid symbols ) epiphyte cover and for epiphytes ( ) versus energy dissipation rate raised to a power of 0.25. (B) Rates of nitrate uptake rates for seagrass leaves with ( open symbols) and without (closed symbols ) epiphyte cover and for epiphytes ( ) versus energy dissipation rate raised to a power of 0.25. Uptake rates are based on a percentage of bare leaf area that leaves or epiphytes were exposed to the water column. For instance, during NH 4 + uptake experiments, epiphytes covered a pproximately 90% of the leaf area. Therefore, uptake rates for epiphytes were normalized to 90% of the total leaf area and seagrass leaves with epiphyte cover were normalized to 10% of the total leaf area. Lines represent the best linear fit for the data with an intercept set at 0. NH 4 + uptake rates for seagrass leaves with and without epiphyte cover were not significantly different (ANCOVA based on ln r vs. ln e P = 0.58) and the line is for the pooled data. Rates of NH 4 + uptake for the epiphytes were higher than those for seagrass leaves (ANCOVA based on ln r vs. ln e P<0.01). There was no significant regression for NO 3 uptake rates for seagrass leaves. Rates of NO 3 uptake for the epiphytes were significantly dependent on e Data includes points (i ndicated by + and diamonds for seagrass) based on the relationship between e and bulk velocity (U b ). 55

PAGE 69

56 (mg Chl a ) 1 s 1 Unlike the seagrass leaves, uptake rates for epiphytes were dependent on bottom shear stress ( r = [4.92 10 9 ] t 0.54 95% CLs on slope = 0.14 and 0.93, r = 0.77, P<0.05) and energy dissipation rate ( r = [4.76 10 8 ] e 0.30 95% CLs on slope = 0.09 and 0.51, r = 0.78, P<0.05). Slopes were in the same range as those expected for uptake rates that are delivery limited (Hearn et al. 2001). Effects of epiphyte cover on DIN uptake rates Rates of ammonium uptake for leaves without epiphyte cover were on average 90% (range = 86 93%) higher than those for seagrass leaves with epiphyte cover (Fig. 10). The extent of this difference in uptake rates was directly proportional to the extent of epiphyte cover at the site (~90%). Uptake rates normalized to bare surface area of the leaf for leaves with and without epiphyte cover were within the same range (19.6 10 8 to 86.8 10 8 g NH 4 N m 2 s 1 ) were not significantly different (ANCOVA based on ln r vs. ln e P = 0.58) and were proportional to e to the 0.25 root (Fig. 12). These data also suggest that uptake rates for the leaves with cover were not enhanced by the presence of epiphytes. Rates of NH 4 + uptake for epiphytes normalized to the area of the leaf surface covered by the epiphytes were higher than those for seagrass leaves and ranged between 35.2 and 120.4 10 8 g NH 4 N m 2 s 1 (ANCOVA based on ln r vs. ln e P < 0.001). Like the seagras s leaves, uptake rates fit the theoretical relationship between uptake rates and e (Fig. 12). In all but one case, seagrass leaves without epiphyte cover removed nitrate at a rate ~36% (range = 0 58%) times as high as the rate for seagrass leaves with epiphyte cover (Fig. 11). This difference between uptake rates was significant based on a paired

PAGE 70

57 t test (P<0.001, df = 6). Uptake rates normalized to bare leaf area for leaves with epiphyte cover were ~ 40% higher than rates for leaves without epiphyte cover (Fig. 12; paired t test, P < 0.01, df = 6), which was less than the extent to which epiphytes covered the surface of the leaves (~ 73%). Uptake rates for epiphytes normalized to the planar area of the leaf surface that the epiphytes covered were pro portional to e to the 0.25 root and were higher than uptake rates for the seagrass leaves (Fig. 12). Discussion Results indicate that rates of ammonium uptake for T. testudinum leaves were limited by the rate of delivery to the leaf surface and greatly r educed by epiphyte cover. Conversely, rates of nitrate uptake were limited by the rate at which the leaves could process nitrate rather than the delivery rate. These findings quantitatively demonstrate the potential impact of hydrodynamics and epiphyte c over on rates of DIN uptake for seagrass leaves and indicate that the importance of these factors in influencing uptake rates can vary depending on the form of DIN being assimilated. Effects of water flow on DIN uptake rates The rate of ammonium uptake by seagrass leaves with and without epiphyte cover was influenced by the rate at which ammonium was delivered to the surface of the leaves (Fig. 10). Hydrodynamic characteristics within and above the seagrass canopy were similar to those described for uni directional flows in natural (Koch and Gust 19 99) and simulated (Gambi et al. 1990; Nepf and Vivoni 2000) canopies (Fig. 7 and 8). Further, the correlation between U b and the relative proportion of the canopy height represented by the roughness length (Z 0 ) revealed that shear

PAGE 71

58 penetrated deep within the seagrass canopy when velocity was increased (Fig. 9). If rates of DIN uptake by the seagrass leaves were dependent on the rate of delivery to the surface of the leaves, then the interactions between the can opy and water flow would be expected to increase uptake rates due to thinning of diffusive boundary layers at the uptake surface. Relationships between rates of NH 4 + uptake for the seagrass le aves that form the canopy (Fig. 10 ) and hydrodynamic parameters were consistent with those expected for the b enthos as a whole (Hearn et al. 2001). These results provide further evidence that the response of an individual component of a benthic community to changes in hydrodynamic regime is consistent with that of th e whole community (Cornelisen and Thomas 2002). Hydrodynamic parameters were based on the logarithmic portion of the velocity profile in order to test previously derived relationships between parameters measured from boundary layer conditions and nutrient uptake rates for the ben thos (e.g., Bilger and Atkinson 1992; Hearn et al. 2001). We recognize that velocity profiles measured in terrestrial and aquatic canopies may exhibit characteristics of a mixing layer rather than perturbed boun dary layer flow (Ra upach et al. 1996; Ghisalberti and Nepf 2002). Velocity profiles, particularly at low flows, had characteristics similar to those described in Ghisalberti and Nepf (2002) and it is possible that these data could be interpreted using the mixing layer anal ogy. Characteristics of a mixing layer indicate that penetration of tu rbulent stress into the canopy wa s enhanced under high flows (see Ghisalberti and Nepf 2002). Observations from this study provided similar information as to the amount of shear and tu rbulence imposed on the canopy (estimated as t ) and the depth to which turbulent str ess penetrated the canopy (Fig. 9 ). Several mechanisms

PAGE 72

59 ( e.g., shear, wake, transport) can produce turbulence within plant canopies and as a result there is not always a balance between turbulence generated by shear and the dissipation of turbulence (see Raupach et al. 1996; Ghisalberti and Nepf 2002). Despite these potential inequalities, data were consistent with the relationships outlined by Hearn et al. (2001). These results suggest that variability in small sca le hydrodynamic cha racteristics within the canopy wa s, in effect, averaged when the estimate of e incorporates the roughness length (Z o ). Further research may elucidate how rates of nutrient uptake relate to parameters of the mixing layer analogy as opposed to those estimated from the logarithmic region of the velocity profile. Results from previous studies have demonstrated a similar effect of water flow on NH 4 + uptake by coral assemblages (e.g., Baird and Atkinson 1997; Thomas and Atkinson 1997) a nd seagrass beds (Thomas et al. 2000; Thomas and Cornelisen 2003). In these studies, nutrient flux int o the benthos is described using a simple equation where total flux (m) is equal to an uptake rate constant (S) times the concentration gradient of the nutrien t at the uptake surface (m = S (C b C w ), where C b is the concentration in the bulk fluid and C w is the concentration at the uptake surface). Therefore, values of flux (m) are maximal when the concentration gradient is greatest (C b >> C w ) and are dependent on hydrodynamic parameters as long as C b and C w are not equivalent. The uptake of ammonium in t he present study was dependent on hydrodynamic parameters, thus there was a concentration gradient at the surface of seagrass leaves during flume experiments. It is possible that the rates of ammonium uptake were not maximal and that uptake rates were aff ected by, rather than controlled by, the rate of delivery because the concentration gradient was n ot maximal (Bilger and Atkinson 1995; Sanford and Crawford 2000).

PAGE 73

60 Regardless, the results demonstrate the importance of hydrodynamics in ammonium uptake and that the rate of NH 4 + uptake by the leaves is directly proportional to the rate of energy dissipation (Fig. 10). The effect of water flow on NH 4 + uptake has important implications to nutrient dynamics of seagrasses in their natural setting. For instance, data suggest that the rate of NH 4 + uptake by seagrass leaves would decline during slack tide and increase during ebbing and flooding tides and/or periods of high wave energy. Sediment resuspension and flux of ammonium into the water column also increases during periods of high c urrents and waves (Cowan et al. 1996), which would further fuel seagrasses when their uptake rates are enhanced due to water flow. The effect of water flow on NH 4 + uptake by seagrass leaves is also important to consider when develo ping nutrient budgets for seagrasses and the communi ties they form (Hemminga et al. 1991; Erftemeier and Middleburg 1995; Lee and Dunton 1999; Hansen et al. 2000). The significant effect of water flow would presumably influence Michaelis Menton (M M) para meters estimated for T. testudinum leaves (S anford and Crawford 2000; Smit 2002). Uptake rates measured at low velocity (U b < 0.05 m s 1 ) for leaves that had epiphytes removed prior to flume experiments were very similar to those estimated by parameters o f the M M model (at a concentration of 6 m M) in Lee and Dunton (1999). However, because uptake rates increased four fold over a range of U b commonly observed under field conditions it is probable that values of K m and V max approximated in Lee and Dunton ( 1999) and similar studies underestimate the potential uptake under hig her flows (Sanford and Crawford 2000; Smit 2002). Conversely, if M M parameters were calculated under conditions with

PAGE 74

61 no diffusion limitation, the model could overestimate uptake under field conditions since flow would vary over a range of velocity. Rates of NO 3 uptake for seagrass leaves were not dependent on hydrodynamic parameters (Fig. 11) and were over 10 times lower than those for ammonium. Depressed uptake of nitrate may have resulted from the lower concentration used in nitrate experiments (4 m M) versus ammonium experiments (6 m M). However, based on the M M model from Lee and Dunton, uptake rates for the seagrass leaves would be expected to decrease by ~ 20% if NO 3 concentrat ion was reduced by 2 m M; a magnitude of change far lower than the observed difference. Previous studies have demonstrated a higher uptake affinity for ammonium than nitrate for s eagrass leaves (Short and McRoy 1984; Terrados and Williams 1997; Lee and Dun ton 1999), which has been attributed to physiological demands associated with taking in nitrate (Roth an d Pregnall 1988; Turpin et al. 1991; Touchette and Burkholder 2000). The results provide further evidence that the form of DIN affects uptake rates and that it also influences the relative impact water flow will have on uptake. The fact that uptake rates for nitrate were not influenced by hydrodynamics suggests that the NO 3 concentration gradient at the surface of the leaves was minimal (i.e., C b @ C w ). A minimal concentration gradient would occur when the rate at which the leaves process the nitrate (i.e. enzyme kinetics) is much lower than the rate of nitrate delivery to the surface of the leaves. At ambient levels of nitrate it is possible that up take rates are affected by hydrodynamics (i.e., C b >> C w ) and that increased water flow enhances uptake. Application of tracer levels of labeled nutrients in field flumes

PAGE 75

62 deployed for longer periods of time will provide insight on delivery versus physiolo gical limitation of nitrate uptake by seagrass leaves. Effects of epiphyte cover on DIN uptake rates Epiphyte cover significantly inhibited uptake by T. testudinum leaves. Previous studies have provided evidence that epiphytes can reduce phosphate uptak e by covering the leaf (Johnstone 1979) and deplete nutrient concentrations (O 2 ) in the diffusive boundary layer adjacent to the leaf surface (Sand Jensen et al. 1985). Inhibition of NH 4 + uptake was directly proportional to the extent of epiphyte cover im plying that the epiphytes created a barrier between the available NH 4 + and active uptake sites on the leaf surface. Epiphytes may have superior uptake kinetics that allow them to out compete seagrasses for water column nutrients (Sand Jensen 1977; Wallent inus 1984; Sand Jensen et al. 1985; Cornelisen and Thomas 2002). For ammonium, the rate of delivery rather than physiological characteristics limited the rate of uptake by epiphytes (Fig. 12). Therefore, the most significant factor benefiting epiphytes w as their placement between the nutrient source and the leaf surface, which allowed the epiphytes to intercept a greater portion of water column nutrients relative to the leaves they covered. Koch (1994) suggested that a moderate amount of epiphytes add r oughness to the surface of seagrass leaves and as a result increase turbulence and thin the diffusive boundary layer adjacent to the leaves. As indicated by Koch (1994), this effect would be more pronounced for seagrass leaves that are colonized by rough (i.e., fibrous or branching) rather than by low profile (i.e., < 1 mm thick) epiphytes. The presence of epiphytes did not enhance the rate of ammonium uptake by seagrass leaves, which was

PAGE 76

63 likely because epiphytes were composed of a thin gelatinous layer o f diatoms and blue green algae that covered 90% of the leaf surface. A plot of NH 4 + uptake rates normalized to bare area of the leaf demonstrates the close similarity between uptake rates for covered and uncovered leaves (Fig. 12). Uptake by the epiphyte s was slightly enhanced compared to the seagrass leaves (Fig. 12), which may have resulted from small scale effects of roughness on diffusive boundary thickness at the surface of the epiphytes. As was the case for water flow, the relative impact of epiphyt es on uptake rates depends on the form of DIN being assimilated by the seagrass leaves. Epiphytes reduced the rate at which the leaves took in nitrate; however, the extent of epiphyte cover was not consistent with the degree to which uptake rates were red uced. Based on rates normalized to the proportion of bare area of the leaf, leaves with epiphyte cover had higher uptake rates than those for leaves without epiphyte cover (Fig. 12). This result could be interpreted as enhanced uptake rates due to the pr esence of epiphytes and their effect on water flow characteristics adjacent to the leaf (Koch 1994). However, uptake rates were depressed and not dependent on hydrodynamics for both the covered and uncovered leaves (Fig. 11), which implies that physiologi cal limits of the seagrass plants minimized the amount of additional NO 3 that could be removed by the leaf following removal of the epiphytes. The rate of NO 3 uptake by epiphytes was proportional to e suggesting that uptake was affected by the rate of nitrate delivery to the epiphytes. However, uptake rates for nitrate were significantly depressed (by a factor of 4) relative to those for ammonium. This difference in uptake rates is far greater than would be expected due to differences in concentration during the experiments. As a result, uptake rates were dependent on

PAGE 77

64 hydrodynamic parameters that control rates of delivery, but were depressed because the concentration gradient was not at a maximum due to physiological limitations. Epiphytes were theref ore in an intermediate range where both physical and biological factors were influencing uptake rates (Bilger and Atkinson 1995; Sanford and Crawford 2000). The potential impact of epiphytes on leaf uptake is an important observation since seagrasses hav e long been known to utilize both sediment and water column nutrients (Iizumi and Hattori 1982; Thursby and Harlin 1984; Short and McRoy 1984; Pedersen et al. 1997; Lee and Dunton 1999). The available database on contributions of leaves versus roots in me eting seagrass nutrient requirements is largely based on field and lab experiments that did not incorporate the effects of epiphytes or water flow (Iizumi and Hattori 1982; Short and McRoy 1984; Pedersen and Borum 1992; Lee and Dunton 1999). Findings from the current study suggest that uptake rates measured for seagrass leaves without epiphyte cover were likely overestimated. For instance, what has previously been considered luxury uptake (beyond the plants requirements) during experiments in which epiphy tes were removed may simply have been enhanced leaf uptake due to increased availability of uptake sites (e.g., Lee and Dunton 1999). Lee and Dunton (1999) demonstrated that uptake by leaves and roots contribute equally (~ 50/50) to total N acquisition by the plant However, their contributions to total N uptake are based on laboratory experiments in which the epiphytes were removed from the leaf surface. In a natural seagrass bed the relative contribution of roots versus leaves to meeting nutrient requir ements may be influenced by the degree of epiphyte cover. For instance, in beds with dense epiphyte cover, seagrasses may rely more on uptake by the roots than by the leaves. Similarly, the extent of N conservation within the plant through translocation of

PAGE 78

65 N from old to young tissues may also vary depending on epiphyte abundance (Pedersen and Borum 1992). In addition to acting as a barrier to nutrient uptake, epiphytes can reduce the amount of light reachin g the leaf surface (Sand Jensen 1977). When li ght is sufficient (i.e. above saturation), epiphytes could limit photosynthesis due to inhibition of nutrient uptake; whereas, when light is below saturation, effects of shading may limit productivity (Sand Jensen 1977). As suggested by Sand Jensen, the t hickness of the epiphyte cover influences the amount of light reaching the leaf surface; however, thickness may have little effect on the extent to which epiphytes act as a barrier to nutrient uptake. Quantifying the relative contribution of these effects on seagrass productivity is difficult since seagrass leaves can receive nutrients via translocation from the roots (e.g., Iizumi and Hattori 1982) or even from the epi phytes (Harlin 1973; McRoy and Goering 1974). Interpretation of the results is limited to short term uptake rates and it is possible that epiphytes eventually released some of the acquired N over time, which was subsequently removed by seagrass leaves. Furthermore, data on the amount of light attenuated by the epiphytes was not collected. E xperiments of longer duration are necessary to more fully understand the complex interactions between seagrasses and epiphytes and the effect of these interactions on seagrass productivity. Nonetheless, the data raise important considerations that must be addressed to more fully understand the nutrient dynamics of seagrass communities. Our study highlights the potential impact of water flow and epiphytes on the rate of DIN uptake by T. testudinum leaves in the field. The relative impact of these factors will be largely dependent on the form of DIN being assimilated and the physiological

PAGE 79

66 state of the seagrass plants. Seagrasses inhabit a physically dynamic environment that consists of numerous organisms packed closely together that are vying for water col umn nutrients. It is only appropriate then, to study the nutrient dynamics of seagrasses and the complex communities they form within their natural setting to more accurately describe nutrient cycling processes occurring within seagrass beds (Hemminga et al. 1991). The application of labeled nutrients and deployment of field based flumes affords the opportunity to study the complex interactions between physics and biology and how these interactions influence nutrient transport in complex systems.

PAGE 80

67 CH APTER FOUR APPLICATION OF AN ISOTOPE LABEL FOR ISOLATING EFFECTS OF HYDRODYNAMIC REGIME ON AMMONIUM AND NITRATE UPTAKE BY MAJOR COMPONENTS OF A SEAGRASS COMMUNITY Introduction Seagrass communities inhabit shallow estuarine and nearshore waters that are physically dynamic and experience rapid fluctuations in salinity, turbidity, water flow, and temperature. The canopies formed by seagrass plants serve many important ecological functions in the marine environment, including attenuation of water flow, whi ch enhances the settlement of particles (Fonseca and Kenworthy 1987; Koch 1999) and the transport and recruitment of larvae (Eckman 1987). The interaction between seagrasses and the water column influences nutrient delivery to the community as a whole (Th omas et al. 2000) and seagrass communities efficiently filter nutrients from the water column, thereby maintaining water quality for neighboring benthic communities such as coral reefs (Short and Short 1984). Much is known about nutrient uptake by seagras s plants and the role of leaves and roots in nutrient acquisition (e.g. Iizumi and Hatorri 1982; Short and McRoy 1984; Lee and Dunton 1999). Previous studies have investigated nutrient uptake by epiphytes and potential interactions between epiphytes and th eir host plant (Harlin 1973; McRoy and Goering 1974; Johnstone 1979). In addition, there have been numerous studies on the uptake kinetics of estuarine phytoplankton ( Wheeler et al. 1982) and microphytobenthos

PAGE 81

68 that inhabit estuarine sediments (Hansen et a l. 2000). The existing database, while informative, does not provide estimates of nutrient uptake rates for these organisms while they were situated within a natural seagrass community. Furthermore, there is little information on the role of hydrodynamic regime on nutrient dynamics at both the scale of individual organisms and the community as a whole. Enrichment of water column nutrients promotes increased biomass of phytoplankton and epiphytes, subsequent light attenuation, and decreased seagrass produc tivity (Tomasko and Lapointe 1991; Neckles et al. 1993; Short et al. 1995). This implies that epiphytes and phytoplankton have a high affinity for nutrient uptake and capitalize on water column nutrients. Biological and/or physical factors influence the ra te that these organisms, as well as the community as a whole, take in nutrients. For example, biological factors such as availability of uptake sites or enzymes can control uptake rates (Button 1991; Galvan et al. 1992). Alternatively, if the community o r organism is able to process a nutrient at a faster rate than it can be delivered to its surface, uptake rates are controlled by the rate of delivery (Bilger and Atkinson 1992; Sanford and Crawford 1999). In this case, uptake rates are limited by rates o f molecular diffusion across a concentration gradient (diffusive boundary layer, DBL) at the uptake surface. Engineers refer to these physically limited rates as being mass transfer limited. The concentration gradient that forms the DBL represents the a rea nearest an uptake surface (i.e., seagrass leaf) where transport is controlled by diffusion (Vogel 1994). As the gradient in velocity above the benthos becomes steeper (bottom shear increases) the concentration gradient within the DBL will also grow st eeper, thereby increasing the amount of nutrient reaching the uptake surface. In this case characteristics

PAGE 82

69 of the benthic boundary layer (i.e., bottom shear, dissipation of turbulent energy) are directly correlated to DBL thickness, which in turn is prop ortional to the rate of nutrient uptake (see Hearn et al. 2001). For a seagrass bed, it is assumed that analogous relationships between characteristics of the benthic boundary layer and the DBL will exist at both the scale of the canopy and the individual components that comprise the canopy (i.e., seagrass plants). Hearn et al. (2001) demonstrated through the derivation of mass transfer equations that rates of nutrient uptake for coral reef flats were proportional to the rate at which turbulent energy is d issipated near the benthos ( e ). Although e is largely dependent on U b in unidirectional flow, the relationship between energy dissipation rate and DBL thickness provides a theoretical foundation for linking physical processes to nutrient transport. This is because the rate at which turbulent energy is dissipated by a benthic surface (i.e., seagrass canopy) is directly proportional to the thickness of the DBL (see Hearn et al. 2001 for discussion). If nutrient uptake by the benthos is mass transfer limited, then rates of nutrient up take will be proportional to e 0.25 Further, because e is dependent on water velocity (U b ) and shear stress ( t ), similar relationships can be derived for these hydrodynamic parameters (see Hearn et al. 2001). The rate at which turbulent energy is dissipated at a benthic surface is l argely dependent on the morphology and arrangement of the roughness elements (Hearn et al. 2001). In seagrass beds, dense stands of seagrass shoots and associated epiphytes collectively form a rough and flexible surface that strongly influences water flow within and above the canopy (e.g., Koch and Gust 1999; Nepf and Vivoni 2000). While the hydrodynamics of seagrass canopies are generally understood, there is little known about

PAGE 83

70 the effect of hydrodynamic regime on nutrient uptake by the individual organi sms that inhabit seagrass communit ies Seagrass communities are comprised of many types of organisms that assimilate nutrients from the water column, including the seagrass plants, epiphytes, phytoplankton, and organisms at the sediment water interface. T hese organisms are very different in terms of physiology, morphology and location relative to the canopy and as a result, the relationship between nutrient uptake by these organisms and e may vary. Components that are attached to the bottom experience shear as water passes over their surface; therefore, rates of nutrient uptake for organisms such as the seagrasses and attached epiphytes may be dependent on hydrodynamic regime. Conversel y, uptake rates for small celled organisms suspended in the water column (i.e., picophytoplankton) may not be affected by hydrodynamics since they are moving with the water. The presence of the seagrass canopy can greatly attenuate water flow at the sedime nt water interface (Koch and Gust 1999). As a result, uptake rates for microphytobenthos may be relatively constant due to a limited range of flow conditions. In this chapter, I investigated the effects of hydrodynamic regime on uptake of dissolved inorga nic nitrogen (DIN) by individual components (phytoplankton, epiphytes, seagrass leaves, and microphytobenthos) while they were situated within a seagrass community. Different forms of DIN (NH 4 + and NO 3 ) that vary in physiological requirements for uptake were used in two separate series of experiments. In order to isolate nutrient uptake rates for individual components in the seagrass community, labeled DIN ( 15 NH 4 + or 15 NO 3 ) was applied in a field flume deployed in natural seagrass beds. Uptake rates wer e quantified over a range of water velocity (0.02 0.18 m s 1 ) and an Acoustic Doppler Velocimeter (ADV) was used to collect velocity profiles during

PAGE 84

71 experiments. Velocity profiles were used to calculate hydrodynamic parameters, including the rate of tur bulent energy dissipation ( e ) and bottom shear stress ( t ). The dependence of uptake rates on these hydrodynamic parameters was then evaluated and compared to what was expected for mass transfer limited uptake by the benthos (Bilger and Atkinson 1992; Hearn et al. 2001). Our data wa s also used to estimate the relative contributions of individual components to total DIN uptake by the community as a whole and to evaluate the application of mass transfer equations for predicting uptake by the benthos (Hearn et al. 2001). Methods Flum e deployment To isolate the effects of water flow on DIN uptake by individual components of a seagrass community, a spike of 15 N labeled DIN was added to a field flume deployed in natural seagrass beds (see Fig. 6). Two separate sets of flume experiment s were conducted, one using ammonium ( 15 NH 4 + ) and one using nitrate ( 15 NO 3 ) as the labeled DIN source. Uptake rates for ammonium were measured between 7 and 14 June 2001 and within a T. testudinum bed located in Pass A Grille Channel, which is approximat ely 2 km north of Fort Desoto County Park at the southernmost point of Pinnelas County, Florida (Fig 5). Uptake rates for nitrate were measured between 17 and 22 June 2001 and within the park boundaries approximately 2 km south of Pass A Grille Channel. Characteristics of these two sites were similar and are provided in Table 2. Flume experiments for estimating uptake rates were conducted between 1000 and 1500 h and within 100 meters of the shoreline in water depth less than the flume height (0.8 m).

PAGE 85

72 Th e field flume enclosed a 3.7 m 2 area of the seagrass bed and allowed us to impose controlled unidirectional flow over the community and measure nutrient uptake by the seagrass enclosed in the flume (Fig 6; see Thomas et al. 2000; Thomas and Cornelisen 2003 ). A unidirectional current was imposed over the seagrass community using a 12 volt electric trolling motor. The motor was housed in a motor box that was tapered at the entry and exit and included a series of flow straighteners (composed of plastic grati ng used for fluorescent lighting) within the exit section to minimize turbulence created by the motor (Fig. 6). Uptake rates for NH 4 + were determined from a series of nine flume experiments, each of which was conducted at a velocity randomly chosen from a predetermined range observed under natural flows in seagrass beds (0.02 0.18 m s 1 ). A total of seven flume experiments over a similar range of velocity (0.03 0.17 m s 1 ) were conducted for measuring rates of NO 3 uptake. The flume was moved to a ne w location in the seagrass bed for the measurement of each uptake rate and provide replication. After ensuring the flume was sufficiently sealed from the surrounding water, a spike of DIN, either as 20 mmol/L 98 atom % 15 NH 4 Cl for ammonium uptake experi ments or Na 15 NO 3 for nitrate uptake experiments was added to the flume. The spike was added slowly (over ~3 minutes) through a tube leading to the motor box and the trolling motor was used to assist in mixing the water to obtain a uniform beginning conce ntration. Ambient water samples were collected prior to each experiment to confirm the contribution of background nutrients to the total amount of nutrients in the flume. For the ammonium uptake experiments, the beginning concentration following the addit ion of the spike ranged from 5 to 7 m mol/L for NH 4 + For the flume experiments in which

PAGE 86

Ammonium experiments Nitrate experiments Canopy characteristics Mean S.D. n Mean S.D. n Shoot density (No. m 2 ) 430 104 36 376 119 28 Leaf density (No. m 2 ) 1166 212 36 1171 242 28 Leaf biomass (g dry wt m 2 ) Canopy height in still water (m) 132 0.24 32 0.03 9 45 127 0.32 24 0.03 7 35 Epiphyte biomass (g dry wt m 2 ) Small (<35 m M) Large (>35 m M) Epiphyte Chl a (mg Chl a (g dry wt) 1 ) 201 10 191 0.84 60 3 57 0.26 9 9 9 27 142 10 131 1.32 44 5 42 0.36 7 7 7 21 TSS (g liter 1 ) PON (g N liter 1 )* 0.10 3.4 x 10 4 0.05 7.8 x 10 5 9 9 0.08 3.4 x 10 4 0.02 6.7 x 10 5 7 7 Sediment (g N 0.005 m 3 )** Sediment Chl a ( m g Chl a (g dry wt) 1 ) 24 6.8 9 1.9 7 21 18 14.4 7 3.2 7 21 Seagrass % N Content Epiphytes TSS Sediments 2.13 0.98 0.37 0.12 0.21 0.13 0.17 0.04 9 9 9 7 1.98 1.46 0.46 0.10 0.13 0.14 0.09 0.04 7 7 7 7 Table 2. Canopy characteristics at study sites for NH 4 + and NO 3 experiments.**0.005m 3 represents a m 2 planar area by ~0.005 m depth. Canopy height is based on measurements in still water. 73

PAGE 87

74 nitrate was used as labeled form of DIN the beginning concentration ranged between 4 and 5 m mol/L. During each flume experiment, several parameters were measured including water height, temperature, and the deflected height (h d ) of the canopy in flowing water. Deflected height was based on an average of ten measurements taken manually with a meter stick while the water was flowing. The height of the canopy in still water (with the motor turned off) was also recorded. Measurement of hydrodynamic parameters A velocity profile was collected during each flume experiment in order to calculate hy drodynamic parameters, including bulk velocity (U b ), bottom shear stress ( t ), and the rate of turbulent energy dissipation ( e). Velocity data for constructing profiles were collected using an acoustic Doppler velocimeter (Field ADV; YSI/Sontek) that measures velocity in three dimensions: longitudinal along the main flow (U), tr ansverse (V), and vertical (W). The probe was affixed to a metal rod that could be moved vertically to position the sensors at a specific height above the bottom. Profiles were collected in the center of the working section approximately 1.5 m downstream from the first turn (Fig. 6) and within the area where leaves were collected for determining uptake rates. For each velocity profile, data were recorded at 5 Hz for 1 minute (n = 300) at each of 10 to 12 heights above the sediment water interface. Heigh ts included ~ 4 6 measurements within the canopy and at least 5 above the seagrass up to ~ 10 cm beneath the surface of the water. Care was taken to ensure that the sampling volume was sufficiently above the bottom when data were collected near the sedi ment water interface

PAGE 88

75 (Finelli et al. 1999 ). When data were collected in the canopy, leaves that were directly in contact with the ADV sensors were trimmed to prevent interference with data collection. This technique has been shown to have no significant effect on flow measurements taken in vegetated canopies (Ikeda and Kanazawa, 1996). Signal to noise ratios were well above the recommended 15 db level and the correlation values for the three sensors consistently ranged between 85 and 95%. Bulk velocity (U b ) was estimated as the depth averaged velocity within each profile. The vertical gradient of mean velocity in the dominant direction of flow ( U ) was also used to obtain estimates of shear velocity (U ) and roughness length (Z o ), which in turn were used to calculate bottom shear stress ( t ) and rates of turbulent energy dissipation ( e ). Shear velocity (U ) is not a true velocity measurement but rather an estimate of shear stress and can be converted to t using the equation t = r U 2 (Denn y 1988), where r is the density of seawater. The rate of turbulent energy dissipation for a bottom with roughness length (Z o ) was estimated using the equation e =( t/r ) 3/2 / k Z o where k is the von Karman constant (0.4) (Hearn et al., 2001). The commonly cit ed Karman Prandtl equation was used to obtain estimates of shear velocity (U ): U = U / k ln (Z/Z o ) (1) where U is the mean velocity at a given distance from the bottom and Z is the height above the sediment water interface. This method utilized the portion of the profile in which velocity increased logarithmically with increasing height (Z) above the bottom.

PAGE 89

76 The roughness length (Z o ) was estimated as the intercept of the velocity vs. ln Z plot (see Denny, 1988). Confidence limits (95%) on estimates of U were calculated based on the number of measurements in the logarithmic portion of the vertical profile and the regression coefficient (r 2 ) and using the expression outlined in Grant et al. (1984). DIN uptake b y individual components Rates of DIN (NH 4 + or NO 3 ) uptake were determined for individual components of the seagrass community, including phytoplankton, epiphytes attached to seagrass leaves, seagrass leaves, and microphytobenthos at the sediment water i nterface. These components utilize DIN in the water column and are the major contributors to primary productivity in seagrass communities (Moncreiff et al. 1992). Samples for each of these components were collected prior to experiments for determination of ambient isotope ratios and at the end of each experiment to determine uptake rates based on 15 N accumulation over time. Previous experiments (Chapter Two) showed that rates based on samples collected at the beginning and end of the experiment were the s ame as rates based on samples collected over the duration of the experiment. Particulate organic nitrogen (PON) filtered from the water column was used as an estimate of the phytoplankton fraction. Immediately upon completion of each experiment, a 1 L s ample of water was collected by filling a 1 L bottle (Nalgene) at mid water depth within the flume. The liter of water was filtered through a pre weighed and combusted 0.7 m M filter (Whatman) using a hand pump (Nalgene) and a 2 L filter flask (Pyrex).

PAGE 90

77 Following each experiment approximately 20 whole seagrass leaves were quickly removed from random locations within the flume. Epiphytes (all attached material) were separate d from the leaves by gently scraping the leaves with the edge of a microscope slide. All the epiphyte material removed from these leaves was pooled to represent the epiphyte sample. The pooled sample was split into two portions; half designated for 15 N an alysis and half for determining Chlorophyll a concentration. Epiphytes for 15 N analysis were rinsed with filtered seawater over a 35 m M screen stacked over a 0.7 m M pre weighed and combusted filter to partition the epiphytic material into two size classes (> 35 m M and between 0.7 m M and 35 m M) and to minimize loss of small cells during rinsing (Cornelisen and Thomas 2002). Following the filtered seawater rinse, epiphytes were briefly rinsed with DI water to remove salt (Winning et al. 1999). The seagrass leaves that had epiphytes removed from their surface were rinsed and pooled to represent the seagrass sample. Heavily senescent leav es were not retained for 15 N analysis due to the difficulty in completely removing all epiphytes from the leaf surface. In addition to the seagrass leaves that had their epiphytes removed, a sample of leaves with epiphytes in tact were also collected afte r each experiment and retained for 15 N analysis. In order to assess the potential loss of 15 NH 4 + during the separation and rinsing process, the leaves and epiphytes in these samples were not separated and were not rinsed. Samples of sediment at the sedim ent water interface were collected to assess the effects of water velocity on DIN uptake by microphytobenthos. A microscope slide was used to scrape the top ~ 0.5 cm of sediment from several random locations in the flume

PAGE 91

78 into a whirl pak bag. Samples were split, with half designated for 15 N analysis and half for determining Chlorophyll a concentration. Following collection, samples of PON and the smaller fraction of epiphytes collected on 0.7 m M filters were wrapped in precumbusted tinfoil and dried at 60 C for 24 h. Samples of the larger fraction of epiphytes, seagrass leaves with epiphytes removed, seagrass leaves with epiphytes still attached, and sediments were dried at 60 C for 24 h, homoge nized to a fine powder with a mortar and pestal, and stored in individual glass vials. The portion of epiphytes and sediments designated for Chl a analysis were frozen ( 80 C) until later analysis. Previous work demonstrated that normalizing uptake rate s to Chl a corrects for the presence of heterotrophic organisms within the sample biomass that do not actively remove ammonium from the water column (Cornelisen and Thomas 2002). Chlorophyll a concentrations (mg Chl a (g dry wt) 1 ) were analyzed using spec trophotometric methods as outlined in Strickland and Parsons (1968 ). All dried samples of PON, epiphytes, seagrass, seagrass with epiphytes, and sediments were analyzed using an elemental analyzer coupled with an isotope ratio mass spectrometer (EA IRMS : elemental analyzer isotope ratio mass spectrometry) for determination of nitrogen content (% N) and atom % 15 N. Specific uptake rates (V) for the samples were calculated using the following equation: V = ( d a s / d t)/(a w a s ) (2)

PAGE 92

79 where a s is the atom % 1 5 N in the components tissue, a w is the atom % 15 N of the enriched substrate, and t is time (in seconds) (Dugdale and Goering, 1967). The units for V are g N removed (g N tissue) 1 s 1 or simply s 1 The atom % 15 N of the enriched water (a w ) was based o n the amount of 98 atom % 15 NH 4 + or 15 NO 3 added and background DIN concentrations (assumed to reflect 15 N concentration of atmospheric N ~0.37 atom % 15 N). The numerator ( d a s / d t) was estimated as the difference in atom % 15 N between ambient samples and s amples collected at the end of each experiment divided by the duration of the experiment (~ 40 to 60 minutes). It is noted that the use of equation (2) in calculating V assumes that the atom % 15 N of the source pool did not change during the course of the experiment. Dilution of 15 N in the source pool resulting from inputs of non labeled NH 4 + (via regeneration, excretion) or NO 3 (via nitrification) into the water column would expectedly result in un derestimated uptake rates (Laws 1984). While this is a p otential source of error, the short duration of these experiments along with the high concentration and atom % 15 N of the spike likely minimized dilution. Specific uptake rates (V) were normalized to the nitrogen concentration (% N) of each component to c alculate uptake rates ( r ) in units g N removed (g dry wt) 1 s 1 (Dugdale and Goering 1967). For plankton, V was multiplied by g PON L 1 to calculate an uptake rate ( r ) in g N removed (liter) 1 s 1 For epiphytes and sediments uptake rates were also normalized to chlorophyll a concentration in the sample to obtain an uptake rate that was representative of the autotrophic fraction ( r Chl = r Chl a 1 in units g N removed (mg Chl a) 1 s 1 ; Frankovich and Fourqurean 1997; Dickson and Wheeler 1995). Since the decline in ammonium or nitrate concentration in the water column is first order, the rate of 15 N accumulation in the seagrass leave s is assumed to be first

PAGE 93

80 order. Equation (2) assumes linear uptake and any change in water column concentration over time will influence the calculation of uptake rates. Therefore, uptake rates ( r ) must be multiplied by a correction term ( a ) to compensat e for changes in water column concentration over time. For instance, water column concentration during measurement of uptake rates at a high velocity was depleted more than during measurements at low velocity. As a result, r would be underestimated if con centration were assumed to remain constant. The correction term was calculated as the average value of 1/e kt over the course of the experiment, where k is the first order rate of decline in concentration and t is time (see DIN uptake by the community). Uptake rates corrected for concentration change ( r a ) represent uptake rates for the beginning water column concentration (~ 6 m M for NH 4 + experiments and ~ 4 m M for NO 3 experiments) and were used to assess the effects of water flow on DIN uptake by the individual components. Dependence of r on hydro dynamic parameters (U b t and e ) was evaluated using regression and fitting data to the expected relationship as outlined by Hearn et al. (2001). Model II regressions (geometric mean) were used since there was error in estimating both the dependent and i ndepen dent variables (Sokal and Rohlf 1995). Correlations between uptake rates and hydrodynamic parameters were also estimated using the product moment correlation coefficient (r). DIN uptake by the community Uptake of DIN by the entire community was ba sed on the rate at which DIN (either as NH 4 + or NO 3 ) was depleted from the water column. Methods for water collection are explained in detail in Thomas et al. (2000) and are only briefly discussed here. Over the duration of each flume experiment, a subme rsed pump

PAGE 94

81 placed at mid water height above the seagrass canopy was used to continually pump water from the flume into 1 L Nalgene bottles. Water was pumped from the flume at a rate that filled a 1 L bottle over a 5 to 7 minute sampling period. There were seven sampling periods (seven 1 L bottles filled) over the 40 to 60 minute duration of each experiment. The duration of experiments was based on previous results that indicate rapid uptake by seagrass communities (Thomas et al. 2000) and allowed sufficien t uptake without depleting the concentration appreciably. Immediately following each sampling period, duplicate samples of water were drawn from each of the 1 L bottles with a large syringe and filtered through an in line 0.7 m M filter into 30 ml bottles ( Nalgene). These water samples were then placed in a seawater ice mixture (~0 C) and stored in a freezer ( 80 C) upon return to the laboratory and until samples could be analyzed. Samples were analyzed with an autoanalyzer to determine concentration (to an accuracy of 0.1 m mol/L) of ammonium and nitrate. Both nutrients were measured in both series of experiments to assess any significant contributions of nitrification to the decline in ammonium or nitrate, respectively. Decline in nutrient concentration over time in the flume is a first order relationship. A first order rate constant (k) was estimated as the slope of the natural log of concentration versus time (see Thomas et al. 2000). Each first order rate constant (k) was normalized for water volume in the flume (Vo l) and planar surface area of the benthos enclosed by the flume (A) to estimate an uptake rate constant (S; S = k Vol / A ). An estimate of the total N removed over time (in g NH 4 + N or g NO 3 N removed s 1 ) was also calculated from the regression of k vs. tim e for each experiment.

PAGE 95

82 Dependence of S on hydrodynamic parameters (U b t and e ) was evaluated using the same methods described for individual components. Measured uptake rate constants (S) for the community were also compared to those expected based o n the relationship between S and the rate of turbulent energy dissipation ( e ) (Hearn et al. 2001). Uptake rate constants were predicted based on estimates of e and using the following equation identified by Hearn et al. (2001) as the e 1/4 law: S = (1/ a) ( D 2 e / v ) 1/4 (3) Equation (3) is based on the relationship between S and the thickness of the diffusive boundary layer ( d ), which in turn is dependent on e (see Hearn et al. 2001). As outlined by Hearn et al., S is equivalent to the ratio of mole cular diffusivity of the nutrient (D) to the thickness of the diffusive boundary layer ( d ). The molecular diffusivity of a nutrient (D), such as ammonium, is several orders of magnitude lower than diffusivity of momentum (v) and therefore d must be placed within the Batchelor length scale (Batchelor 1959) as d = a (vD 2 / e ) 1/4 A constant ( a ) is added to the equation in order to prevent overestimating turbulent mixing (see Lazier and Mann 1989). A value of 3 was chosen for a since it provided estimates of S that were consistent with those obtained using empirically derived equations (Bilger and Atkinson 1992). Values of D and v used for calculating values of S were based on water temperatures during flume experiments (Table 1) and a salinity of 35 0 / 00 (Li and Gregory 1974).

PAGE 96

83 Contribution of components to total DIN uptake by the community Uptake rates ( r ) were multiplied by the total biomass of each component to estimate the amount of NH 4 + or NO 3 removed by each component. Biomass of PON in the water column (g PON L 1 ) was based on the concentration of PON and the total volume of water in the flume. Est imates of shoot density (n=5) and biomass of seagrass leaves and epiphytes (n=1) were collected using randomly placed quadrats (0.01 m 2 ). Leaves and epiphytes within a quadrat were separated, dried (60 C) and then weighed to determine biomass of these components in g m 2 Biomass of sediment was based on dried cm 3 samples. Uptake rates ( r ) for seagrass leaves, epiphytes, and sediments were multiplied by their total biomass in the flume ( r g dry wt) to estimate the total amount of NH 4 + or NO 3 removed by each these components over the course of each flume experiment. The contribution of each component to total uptake by the community was estimated as the total amount of NH 4 + or NO 3 removed by each component divided by the total amount removed by the community as a whole (see DIN uptake by the community ). Results Vertical profiles of velocity components (U, V, and W) revealed that water flow in the flume was highly unidirectional (see Fig. 7; A ppendices ). A comprehensive characterization of water flow during flume experiments is provided in the appendices. Depth averaged velocities (U b ) ranged between 0.03 and 0.17 m s 1 and 0.03 and 0.18 m s 1 for the NH 4 + and NO 3 experiments, respectively (Table 3). Fits of velocity data to Equation (1) were significant for all profiles a nd revealed that velocity increased

PAGE 97

84 logarithmically with height above the seagrass canopy. Coefficients of determination (r 2 ) for U vs. ln Z plots ranged from 0.92 to 0.99 (mean = 0.95, SD = 0.03; Table 3). Estimates of shear velocity (U ) ranged between 0.010 and 0.034 m s 1 (95% CLs range = 0.002 to 0.011 m s 1 ) during NH 4 + uptake experiments and between 0.021 and 0.050 m s 1 (95% CLs range = 0.006 to 0.017 m s 1 ) during NO 3 uptake experiments. Bottom shear stress ( t ) estimated from U ranged between 0.09 and 1.20 N m 2 and between 0.49 and 2.55 N m 2 for the NH 4 + and NO 3 experiments, respectively. Estimates of energy dissipation rate ( e ) ranged between 0.02 10 3 and 2.25 10 3 m 2 s 3 for the NH 4 + uptake experiments and between 0.15 10 3 an d 3.21 10 3 m 2 s 3 for the NO 3 uptake experiments. Ammonium uptake by individual components Labeled ammonium ( 15 NH 4 + ) was recovered in all components, including PON, epiphytes (both large and small fractions), seagrass leaves, and microphytobenthos Uptake rates ( r ) based on the rate of 15 N incorporation and corrected for decline in water column NH 4 + concentration (see methods) are listed in Table 4. Ammonium uptake rates ( r ) for PON ranged from 0.28 10 8 to 1.55 10 8 g NH 4 N (liter) 1 s 1 Although suspended w ithin the water column, uptake rates were significantly dependent on hydrodynamic parameters, including bulk velocity ( r = [5.99 10 8 ] U b 0.85 95% CLs on slope = 0.59 and 1.10, r = 0.91, P<0.05) (Fig. 13), bottom shear stress ( r = [1.30 10 8 ] t 0.76 95% CLs on slope = 0.52 and 0.99, r = 0.80, P<0.05) and energy dissipation rate ( r = [12.4 10 8 ] e 0.36 95% CLs on slope = 0.12 and 0.60, r = 0.88, P<0.05). However, the amount of PON in the water column

PAGE 98

85 AMMONIUM UPTAKE EXPERIMENTS U=U /k ln (Z/Z o ) Exp. U b (m/s) Grass he ight (m) U (m/s) Z 0 (m) r 2 1 0.026 0.20 0.012 0.146 0.94 2 0.136 0.13 0.026 0.039 0.98 3 0.047 0.19 0.013 0.086 0.99 4 0.074 0.17 0.021 0.094 0.92 5 0.172 0.14 0.028 0.027 0.97 6 0.043 0.19 0.010 0.054 0.97 7 0.075 0.20 0.025 0.110 0.99 8 0.031 0.22 0.016 0.171 0.92 9 0.169 0.14 0.034 0.043 0.95 NITRATE UPTAKE EXPERIMENTS U=U /k ln (Z/Z o ) Exp. U b (m/s) Grass height (m) U (m/s) Z 0 (m) r 2 1 0.063 0.28 0.039 0.191 0.97 3 0.104 0.24 0.042 0.125 0.95 4 0.055 0.29 0.027 0.146 0.95 5 0.090 0.24 0.043 0.119 0.96 6 0.052 0.27 0.022 0.163 0.98 7 0.154 0.19 0.050 0.095 0.91 Table 3. Values for depth averaged velocity (U b ), shear velocity (U ), and roughness length (Z o ) calculated using the log prandtl equati on. Also provided are grass heights measured in the flowing water. Coefficients of determination (r 2 ) are for the fit of U b vs. ln height plots. Vertical profiles of velocity were not collected during experiment 2 for NO 3 due to equipment failure. No r esults are provided if the inclusion of d in the calculation did not improve the fit to the law of the wall equation.

PAGE 99

AMMONIUM UPTAKE RATES Community S eagrass Epiphytes Sediments PON ( r E and r Chl ) (> 0.35 m m) (< 0.35 m m) Exp Water height (m) Vol (l) (S) m s 1 ( 10 5 ) (r G ) g NH 4 N g tissue 1 s 1 ( 10 8 ) g NH 4 N g tissue 1 s 1 ( 10 8 ) g NH 4 N mg Chl a 1 s 1 ( 10 8 ) g NH 4 N g tissue 1 s 1 ( 10 8 ) g NH 4 N mg Chl a 1 s 1 ( 10 8 ) (r S ) g NH 4 N m g Chl a 1 s 1 ( 10 10 ) (r P ) g NH 4 N liter -1 s 1 ( 10 8 ) 1 0.56 2079 7.5 0.29* 1.87 1.77 1.22 1.16 NA 0.30 2 0.63 2339 28.7 0.44 2.47 4.29 1.75 3.04 NA 1.37 3 0.70 2599 9.8 0.21 0.97 1.64 0.84 1.43 0.26 0.49 4 0.71 2636 19.4 0.16 1.79 1.87 1.96 2.04 0.30 0.65 5 0.65 2413 15.3 0.33 1.65 3.03 2.61 4.79 0.62 0.77 6 0.61 2246 8.4 0.10 1.40 1.26 1.94 1.75 0.23 0.28 7 0.63 2339 13.9 0.24 2.08 2.25 2.08 2.25 0.23 0.41 8 0.61 2265 11.0 0.11 1.27 1.61 2.14 2.71 0.21 0.74 9 0.61 2265 27.1 0.38 4.48 4.32 4.60 4.44 0.23 1.55 Table 4. Ammonium uptake rate constants for the community (S) and uptake rates estimated for individual components ( r ). Uptake rate constants are first order rate constants ( k ) normalized to planar surface area available for uptake (S = k V/A) *The sample of grass in experiment 1 was determined to be an outlier, most likely due to contamination. Sediment samples were not collected during experiments 1 and 2. Duration of experiments ranged from 40 to 63 minutes (mean = 47). Mean water temperat ure during experiments was 32C ( 2C). 86

PAGE 100

87 PON 0.0 0.5 1.0 1.5 2.0 0.00 0.05 0.10 0.15 0.20 0.25 U b (m s -1 ) r (g NH 4 -N (liter) -1 s -1 ) x 10 8 Ammonium Nitrate Figure 13. The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by PON versus bulk velocity (U b ). Ammonium uptake rates were significantly dependent on bulk velocity (Model II regressio n; r = [5.99 10 8 ] U b 0.85 95% CLs on slope = 0.59 and 1.10, r = 0.91, P<0.05). Although rates of NO 3 uptake also tended to increase as water velocity increased, there was no significant dependence of NO 3 uptake rates on U b

PAGE 101

88 during NH 4 + experiments ran ged between 1.7 10 4 and 4.4 x 10 4 g N (liter) 1 and was also dependent on water flow (Fig. 14; mg PON = [0.91]U b 0.36 C.L.s on slope = 0.14 and 0.58, r = 0.83, P < 0.01). Ammonium uptake rates ( r ) for the large sized fraction of epiphytes (> 35 m m) ra nged from 0.97 10 8 to 4.48 10 8 g NH 4 N (g tissue) 1 s 1 (Table 4). Uptake rates normalized to Chl a concentrations ( r Chl ) ranged between 1.26 10 8 and 4.32 10 8 g NH 4 N (mg Chl a ) 1 s 1 and were significantly dependent on bulk velocity ( r Chl = [1. 17 10 7 ] U b 0.62 95% CLs on slope = 0.43 and 0.81, r = 0.87, P<0.01), bottom shear stress ( r Chl = [3.82 10 8 ] t 0.55 95% CLs on slope = 0.37 and 0.74, r = 0.87, P<0.01) and energy dissipation rate (see Fig. 15; r Chl = [2.01 10 7 ] e 0.26 95% CLs on sl ope = 0.10 and 0.42, r = 0.91, P<0.001). Rates of ammonium uptake for the small sized fraction of epiphytes (< 0.35 m m) were within the same range as the larger epiphyte fraction (paired t test, P = 0.58) and were significantly dependent on bulk velocity ( r Chl = [1.44 10 7 ] U b 0.67 95% CLs on slope = 0.47 and 0.88, r = 0.84, P<0.01), bottom shear stress ( r Chl = [4.27 10 8 ] t 0.60 95% CLs on slope = 0.41 and 0.80, r = 0.86, P<0.01) and energy dissipation rate ( r Chl = [2.58 10 7 ] e 0.29 95% CLs on slope = 0.11 and 0.46, r = 0.88, P<0.01) (see Fig. 15; Table 4). Rates of NH 4 + uptake for seagrass leaves were approximately 90% lower than those for epiphytes and ranged from 0.10 10 8 to 0.44 10 8 g NH 4 N (g dry wt) 1 s 1 (Table 4). One of the calculated rates for these leaves (experiment 1) was determined to be a highly significant outlier (P < 0.001, Grubbs test; Sokal and Rohlf 1995) and was removed from data analysis. Based on Model II regressions, rates of NH 4 + uptake for the

PAGE 102

89 0 0.1 0.2 0.3 0.4 0.5 0.6 0.00 0.05 0.10 0.15 0.20 0.25 U b (m s -1 ) mg PON L -1 Ammonium experiments Nitrate experiments Site in Florida Keys Figure 14. The concentration of PON in the water column as a function of bulk velocity (U b ) during experiments for measuring ammonium (solid symbols) and nitrate (open symbols) uptake. The line represents a significant regression for data co llected during NH 4 + uptake experiments (statistics). Also shown on the graph are estimates of PON measured during flume experiments conducted on a shallow seagrass/coral bank in the Florida Keys. As demonstrated in Chapter 4, the site in the Florida Keys is regularly exposed to high velocity (range = 0.10 0.50 m s 1 ).

PAGE 103

Epiphytes (> 35 m m) y = 18.37x R 2 = 0.80 y = 3.42x R 2 = 0.61 0.0 1.0 2.0 3.0 4.0 5.0 0.00 0.10 0.20 0.30 e 0.25 (m 2 s -3 ) r (g NH 4 -N (mg chl a) -1 s -1 ) x 10 8 Ammonium Nitrate Epiphytes (< 35 m m) y = 19.76x R 2 = 0.81 y = 5.86x R 2 = 0.92 0.0 1.0 2.0 3.0 4.0 5.0 0.00 0.10 0.20 0.30 e 0.25 (m 2 s -3 ) r (g NH 4 -N (mg Chl a) -1 s -1 )x10 8 Ammonium Nitrate Figure 15. The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by epiphytes ( r Chl ) versus rate of turbulent energy dissipation ( e ) to the 0.25 powe r. Lines represent a significant (P<0.05) linear fit with an intercept set at 0. The shaded diamond refers to a value of e that is based on the regression of e vs. U b 90

PAGE 104

91 leaves were significantly dependent on bulk velocity ( r = [1.54 10 8 ] U b 0.73 95% CLs on slope = 0.56 and 0.91, r = 0.91, P<0.01), bottom shear stress ( r = [0.41 10 8 ] t 0.66 95% CLs on slope = 0.28 and 1.03, r = 0.80, P<0.05) and dissipation rate ( r = [2.92 10 8 ] e 0.31 95% CLs on slope = 0.17 and 0.45, r = 0.88, P<0.01) (see Fig. 16). Following each experiment, samples of grass with epiphytes still attached were retained for analysis. These samples were not rinsed to assess the potential loss of 15 N due to rinsing of epiphytes. Rates of NH 4 + uptake for the seagrass and epiphyte s combined ranged from 1.09 10 8 to 4.01 10 8 g NH 4 N (g tissue) 1 s 1 and were within the same range as uptake rates estimated for the epiphytes alone, which suggests that an appreciable amount of 15 N was lost when epiphytes were separated from the seag rass leaves and were rinsed. Uptake rates ( r ) for sediments, determined for seven of the nine NH 4 + experiments ranged from 0.21 10 10 to 0.62 10 10 g NH 4 N ( m g Chl a ) 1 s 1 (Table 4) and were not dependent on hydrodynamic parameters (Fig. 17). Ammo nium uptake by the community For all nine NH 4 + uptake experiments there was a significant first order decline in ammonium concentration in the water column over time (Table 4). Uptake rate constants (S) calculated from k (see methods) ranged between 7.5 10 5 to 28.7 10 5 s 1 and were dependent on bulk velocity (Fig. 18; S = [8.76 10 4 ] U b 0.68 95% CLs on slope = 0.49 and 0.88, r = 0.85, P<0.01 ), shear stress ( S = [2.56 10 4 ] t 0.61 95% CLs on slope = 0.34 and 0.79, r = 0.88, P<0.01 ), and energy dissipation ( S = [1.58 10 3 ] e 0.29 95% CLs on slope = 0.10 and 0.48, r = 0.87, P<0.01 ). The total ammonium removed by the community enclosed by the flume ranged from 0.082 to

PAGE 105

92 Seagrass leaves y = 1.78x R 2 = 0.77 0 0.1 0.2 0.3 0.4 0.5 0.00 0.10 0.20 0.30 e 0.25 (m 2 s -3 ) r (gN taken up (g tissue) -1 s -1 ) x 10 8 Ammonium Nitrate Figure 16. The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by seagrass leaves versus rate of turbulent energy dissipation ( e ) raised to the 0.25 power. Lines represent a significant (P<0.05) linear fit with an intercept set at 0. The shaded circle refers to a value of e that is based on the regression of e vs. U b since no profile was collected during this experiment.

PAGE 106

93 Sediments 0.00 0.02 0.04 0.06 0.00 0.10 0.20 0.30 e 0.25 (m 2 s -3 ) r (gN taken up ( m g chl a) -1 s -1 ) x 10 10 Ammonium Nitrate Figure 17. The rate of ammonium (solid symbols) and nitrate (open symbols) uptake by sediments (microphyt obenthos) versus rate of turbulent energy dissipation ( e ) raised to the 0.25 power. There was no significant relationship between uptake by sediments and e The shaded triangle refers to a value of e that is based on the regression of e vs. U b since no pr ofile was collected during this experiment.

PAGE 107

94 0 10 20 30 40 0.00 0.05 0.10 0.15 0.20 Ub (m s -1 ) Measured S (m s -1 ) x 10 5 Ammonium Nitrate Figure 18. Uptake rate constants (S) for ammonium (solid symbols) and nitrate (open symbols) versus bulk velocity (U b ). Model II regression results for ammonium were as foll ows: ( S = [8.76 10 4 ] U b 0.68 95% CLs on slope = 0.49 and 0.88, r = 0.85, P<0.01 ). Values of S for nitrate were less dependent on velocity: ( S = [2.67 10 4 ] U b 0.40 95% CLs on slope = 0.23 and 0.57, r = 0.93, P<0.01 ).

PAGE 108

0 10 20 30 40 0 5 10 15 Expected S (m s -1 ) x 10 5 Measured S (m s -1 ) x 10 5 Total uptake by community Uptake by benthic components (1:1) 0 5 10 15 20 0 5 10 15 Expected S (m s -1 ) x 10 5 Measured S (m s -1 ) x 10 5 Total uptake by community Uptake by benthic components (1:1) Figure 19. Measured versus expected uptake rate constan ts (S) for uptake of ammonium (left) and nitrate (r ight). Measured values are calculated as the first order decline in nutrient concentration over time normalized to pla nar surface area of the benthos and volume of water in the flume (S = k V / A ). Also included are measured values of S B based on uptake by the benthic components only. The line represents the 1:1 ratio. Data suggests that uptake of ammonium by the benthos is limited by the rate of delivery; whereas, uptake of nitrate is depressed and limited by the rate at which the benthos could process the nutrient. 95

PAGE 109

96 0.184 g NH 4 N hr 1 Predicted values of S based on Equation (2) and estimates of energy dissipation rate we re approximately 50% lower than those that were measured (Fig. 19). Nitrate uptake by individual components Labeled nitrate ( 15 NO 3 ) was recovered in all components, including PON, epiphytes (both large and small fractions), seagrass leaves, and microph ytobenthos. Uptake rates ( r ) based on the rate of 15 N incorporation and corrected for decline in water column NO 3 concentration (see methods) are listed in Table 5. Rates of nitrate uptake for the PON in NO 3 experiments ranged between 0.10 10 8 and 0. 43 10 8 g NO 3 N (liter) 1 s 1 Although rates tended to increase with increased water flow, dependence of uptake on hydrodynamic parameters was not significant (Fig. 13). The amount of PON in the water column during NO 3 experiments was within the same range as NH 4 + experiments; however, there was no significant correlation between PON and bulk velocity (Fig. 14). Nitrate uptake rates ( r ) for epiphytes ranged from 0.43 10 8 to 1.12 10 8 g NO 3 N (g tissue) 1 s 1 (Table 5). Uptake rates normalized to Chl a ranged between 0.37 10 8 to 0.89 10 8 g NO 3 N (mg Chl a ) 1 s 1 Based on Model II regressions, r Chl was not significantly depend ent on U b but was dependent on bottom shear stress ( r Chl = [4.92 10 9 ] t 0.54 95% CLs on slope = 0.25 and 0.83, r = 0.77, P<0.05) and energy dissipation rate ( r Chl = [4.76 10 8 ] e 0.30 95% CLs on slope = 0.02 and 0.58, r = 0.78, P<0.05) (see Fig. 15; Table 5) Epiphytes consisting of small cells (< 0.35 m m) exhibited higher uptake rates than the larger epiphyte fraction (paired t test, P<0.01) and had a significant dependence on hydrodynamic parameters (Fig. 15; Table 6), including bulk velocity

PAGE 110

97 ( r Chl = [3.67 10 8 ] U b 0.71 95% CLs on slope = 0.54 and 0.88, r = 0.92, P<0.01), bottom shear stress ( r Chl = [8.61 10 9 ] t 0.51 95% CLs on slope = 0.35 and 0.66, r = 0.94, P<0.01) and energy dissipation rate ( r Chl = [7.18 10 8 ] e 0.28 95% CLs on slope = 0 .17 and 0.39, r = 0.97, P<0.001) The rate of nitrate uptake by seagrass leaves ranged between 0.046 10 8 and 0.087 10 8 g NO 3 N (g dry wt) 1 s 1 Uptake of nitrate by seagrass leaves was not dependent on hydrodynamic parameters (Fig. 16). As was the case during ammonium uptake experiments, uptake rates for the combined epiphytes and seagrass leaves were also similar to NO 3 uptake rates estimated for epiphytes alone [range = 0.46 10 8 to 1.13 10 8 g NO 3 N (g tissue) 1 s 1 ], indicating that 15 N was lost when epiphytes were removed from the seagrass leaves and were rinsed. Uptake rates ( r ) for sediments ranged from 0.06 10 10 to 0.30 10 10 g NO 3 N ( m g Chl a) 1 s 1 (Table 5) and were not dependent on hydrodynamic parameters (Fig. 17). Nitrat e uptake by the community Uptake rate constants (S) for nitrate ranged between 7.8 x 10 5 to 12.7 x 10 5 s 1 and were dependent on bulk velocity (Fig. 18; S = [2.67 10 4 ] U b 0.40 95% CLs on slope = 0.23 and 0.57, r = 0.93, P<0.01 ), shear stress ( S = [9 .05 10 5 ] t 0.29 95% CLs on slope = 0.07 and 0.50, r = 0.88, P<0.01 ), and energy dissipation ( S = [3.04 10 4 ] e 0.16 95% CLs on slope = 0.02 and 0.34, r = 0.92, P<0.01 ). During NO 3 uptake experiments, the community removed a total of 0.029 to 0.067 g NO 3 N hr 1 Measured values of S were in the same range as those that were predicted (Fig. 19).

PAGE 111

NITRATE UPTAKE RATES Community Seagrass Epiphytes Sediments PON ( r E and r Chl ) (> 0.35 m m) (< 0.35 m m) Exp Water height (m) Vol (l) (S) m s 1 ( 10 5 ) (r G ) g NO 3 N g tissue 1 s 1 ( 10 8 ) g NO 3 N g tissue 1 s 1 ( 10 8 ) g NO 3 N mg Chl a 1 s 1 ( 10 8 ) g NO 3 N g tissue 1 s 1 ( 10 8 ) g NO 3 N mg Chl a 1 s 1 ( 10 8 ) (r S ) g NO 3 N m g Chl a 1 s 1 ( 10 10 ) (r P ) g NO 3 N liter -1 s 1 ( 10 8 ) 1 0.62 2311 9.1 0.046 0.67 0.57 1.05 0.89 0.06 0.19 2 0.46 1698 12.1 0.074 1.12 0.88 1.97 1.53 0.09 0.43 3 0.56 2075 10.1 0.050 0.43 0.43 1.02 1.03 0.29 0.17 4 0.53 1981 7.8 0.079 0.71 0.52 1.14 0.84 0.11 0.16 5 0.44 1650 11.4 0.075 0.83 0.89 1.17 1.25 0.30 0.10 6 0.71 2641 8.5 0.054 0.47 0.37 0.79 0.61 0.15 0.13 7 0.64 2358 12.7 0.087 0.93 0.76 1.65 1.36 0.09 0.26 Table 5. Nitrate uptake rate constants for the community (S) and uptake rates estimated for individual components ( r ). Uptake rate constants are first order rate constants ( k ) normalized to planar surface area available for uptake (S = k V/A). Duration of experiments ranged from 32 to 44 minutes (mean = 35). Mean water temperature during experiments was 33C ( 2C). 98

PAGE 112

99 Dependence of NH 4 + uptake rates on hydrodynamic parameters U b t e Community (S) 0.68 ( 0.20) 0.61 ( 0.18) 0.29 ( 0.19) Seagrass leaves 0.73 ( 0.17) 0.66 ( 0.24) 0.31 ( 0.19) Epiphytes > 0.35 m m 0.62 ( 0.19) 0.55 ( 0.18) 0.26 ( 0.16) Epiphytes < 0.35 m m 0.67 ( 0.21) 0.60 ( 0.19) 0.29 ( 0.18) Dependence of NO 3 uptake rates on hydrodynamic parameters U b t e Community (S) 0.40 ( 0.17) 0.29 ( 0.22) 0.16 ( 0.18) Epiphytes > 0.35 m m NS 0.54 ( 0.29) 0.30 ( 0.28) Epiphytes < 0.35 m m 0.71 ( 0.17) 0.51 ( 0.15) 0.28 ( 0.11) Table 6. Dependence of uptake rates for the community (S) and uptake rates for benthic components ( r ) on hydrodynamic parameters f or ammonium (top) and nitrate (b ottom). Bold num bers represent the slopes for a model II regression (geom etric mean; Sokal and Rohlf 1995 ). All slopes provided are significant (see results for regression statistics) unless indicated by NS. The expected slopes for the relationship between uptake rates an d hydrod ynamic parameters are 0.80, 0.40 and 0.25 for U b t and e respectively (Hearn et al. 2001). Also shown is the 95% confidence interval for the slope. Data for epiphytes are based on uptake rates normalized to Chl a concentration. Regressions for rates of ammonium and nitrate uptake for sediments vs. hy drodynamic parameters were not significant. Uptake of nitrate by seagrass leaves was also not dependent on water flow.

PAGE 113

100 Contribution s of components to total DIN uptake by the community Phytoplankton, represented by PON, contributed to approximately 29. 9 % (S.D. = 8.7 %) and 24.8 % (S.D. = 14.4 %) of the total ammonium and nitrate removed by the community, respectively (Fig. 20). Ammonium uptake by the large sized and small sized fraction of epiphytes accounted for approximately 26.4 % (S.D. = 9.2 %) and 1.5 % (S.D. = 0.6 %) of the total NH 4 + removed by the community, respectively. Contributions of epiphytes to NO 3 uptake were similar to those for NH 4 + and were approximately 21.4 % (S.D. = 7.2 %) for the large sized fraction and 2.4 % (S.D. = 0.6 %) for the small sized fraction. Seagrass leaves, although a dominant feature of the community, only contributed to about 2.4 % (S.D. = 1.2 %) and 1.6 % (S.D. = 0.4 %) of the total uptake of ammonium and nitrate by the community (Fig. 20). Sed iment microflora contributed to approximately 3.5 % (S.D. = 2.3 %) and 5.9 % (S.D. = 3.0 %) of the total ammonium and nitrate removed by the community, respectively. S amples of seagrass leaves with epiphytes attached that were unrinsed accounted for approximately 51.1 % (S.D. = 18.3 %) and 40.8 % (S.D. = 8.8 %) of the total ammonium and nitrate removed by the community, respectively. Therefore, approximately 20.8 % [ = 51.1 % (26.4 % + 1.5 % + 2.4 %)] of the total NH 4 + estimated to have been re moved from the water column was attributed to NH 4 + lost during the separation and rinsing process. Similarly, an estimated 14.4 % [ = 40 .8 % (21.4 % + 2.4 % + 1.6 %)] was lost while separating and rinsing seagrass and epiphyte samples during NO 3 uptake experiments.

PAGE 114

101 Ammonium PON 29.9% Epiphytes 27.9% Seagrass 2.4% Sediments 3.5% Loss due to rinsing* 20.8% Other sinks 15.5% Nitrate PON 24.8% Epiphytes 23.8% Loss due to rinsing* 14.4% Other sinks 29.5% Sediments 5.9% Seagrass 1.6% Figure 20. Pie charts showing estimated percent contributions of individual components to total uptake by the community f or ammonium (top) and nitrate (b ottom). Data for epiphytes are for both size cla sses combined. Also shown is the estimated loss of 15 N due to rinsing of epiphytes (see text). Other sinks represents the portion of 15 N that was not accounted for by the components collected during experiments.

PAGE 115

102 Discussion Deployment of a field flume an d application of labeled nutrients ( 15 NH 4 + and 15 NO 3 ) enabled the isolation of DIN uptake by individual components situated in natural seagrass beds. Rates of ammonium uptake for the community as a whole and epiphytes and seagrass leaves were significant ly dependent on hydrodynamic parameters. Uptake of NH 4 + by PON in the water column was also dependent on water flow; however, this effect was largely due to flow dependent resuspension of epiphytes and/or microphytobenthos. The effect of water flow on rat es of nitrate uptake was less pronounced for the community as a whole and only epiphytes experienced enhancement of NO 3 uptake rates with increased water velocity. Results demonstrate mass transfer limitation of ammonium uptake rates and suggest that upt ake rates for nitrate were largely controlled by physiological factors such as availability of carbohydrates, nitrate reductase activity, or availability of active uptake sites (see Touchette and Burkholder 2000 for review). Data also indicate that epiphy tes are a significant contributor to total uptake by the community and that high densities of epiphytes may require seagrasses to rely largely on pore water DIN. Results also demonstrate that uptake by PON in the water column can influence the ability to predict uptake rates based on relationships between hydrodynamic parameters and the transport of nutrients to the benthos. Ammonium uptake by individual components Rates of ammonium uptake for all components, with the exception of microphytobenthos, w ere dependent on hydrodynamics. An unexpected outcome from the data was flow dependent NH 4 + uptake by the PON fraction in the water column (Fig. 14). With the exception of large

PAGE 116

103 phytoplankton cells, chains and/or filaments, the effect of turbulence and sh ear on mass transfer is expected to be minimal (Karp Boss et al. 1997). The community composition and size fractions of the phytoplankton were not determined and as a result I am unable to speculate on the potential effects of shear on nutrient uptake by t he PON. However, the actual amount of PON in the water column was correlated to water velocity suggesting that the dependence of uptake rates on hydrodynamic parameters for PON was in part due to the resuspension of microphytobenthos and/or epiphytes. Th is result has important implications to the coupling of nutrient cycling and resuspension events in estuarine and nearshore systems. Ammonium concentrations in the water column can become elevated during resuspension events (Cowan et al. 1996). A rapid i ncrease in NH 4 + uptake rates in the water column would be expected to accompany elevated concentrations, in part due to increased concentration, but also the resuspension of organisms into the water column that in turn remove the ammonium at a higher rate than the benthic components. It is expected that the effect of resuspension on uptake rates in the water column will be dependent on several characteristics of the site, including the hydrodynamic regime, pore water concentrations, seagrass shoot density, and the composition of the epiphytic community (Koch 1999). For instance, resuspension of PON and subsequent uptake would likely be minimal in an area regularly exposed to high wave and/or tide driven currents (see Fig. 14 ). Rates of ammonium uptake for epiphytes and seagrass leaves were significantly dependent on water velocity (U b ). Uptake rates were also proportional to bottom shear stress ( t ) and rate of energy dissipation ( e ) since these parameters are dependent on U b (Table 6). Relationships betwe en hydrodynamic parameters and uptake rates for the

PAGE 117

104 benthos (represented as S) have been derived for transport processes occurring over an area of the benthos (see Bilger and Atkinson 1992; Hearn et al. 2001). It is assumed that analogous proportionalities between hydrodynamic parameters and uptake rates for individual benthic components of the community should also apply if hydrodynamic processes occurring at the scale of the entire benthic surface are driving uptake by the organisms comprising the surface The rate at which seagrasses and epiphytes removed NH 4 + uptake was dependent on hydrodynamic parameters within the range of that expected from the relationships derived by Hearn et al. (2001) for coral reef flats. This is an important result since it r eveals that hydrodynamic parameters that are dependent on water velocity and the morphology of the canopy have a similar effect on uptake at both the scale of the community and the components that form the canopy. Of the three hydrodynamic parameters, ene rgy dissipation rate ( e ) provided the closest fit to the laws outlined by Hearn et al., which provides evidence of the close relationship between diffusive boundary layer thickness and the rate of energy dissipation. In addition, the method used in estima ting e incorporates both a measure of bottom shear stress (U ) and roughness length in its calculation and therefore is normalized to bending of the canopy with increased flow. Rates of NH 4 + uptake for both size classes of epiphytes were similar and equa lly dependent on hydrodynamic parameters, which provides further evidence that the rate of NH 4 + uptake by epiphytes was limited by the rate of delivery to their surface rather than physiological differences among epiphytes. Turbulent energy and shear was g reatest near the top of the canopy and decreased as the sediment water interface was approached (see Fig. 8; Appendices); therefore, it is likely that uptake rates for epiphytes varied according

PAGE 118

105 to their location along the leaf surface. An evaluation of th e effect of location on uptake rates cannot be made from these data since all epiphytes covering the entire leaf were pooled in our analysis. Finer scale studies utilizing isotope labels and involving vertical segregation of epiphytes within the seagrass canopy will provide information on the effects of location on uptake kinetics. Unlike the seagrass and epiphytes, the rate of NH 4 + uptake for microphytobenthos at the sediment water interface was not influenced by hydrodynamic parameters (Fig. 17). It is possible that a biological factor was limiting uptake rates, such as availability of active uptake sites. Another possibility is that the hydrodynamic conditions near the sediment surface were relatively constant due to attenuation of flow by the canopy (see Figures 7 and 8; Appendices) and therefore uptake rates were minimally affected by changes in hydrodynamic regime. Ammonium uptake by the community Uptake rate constants (S) for ammonium were within the same range as those measured in a previous st udy conducted in Thalassia testudinum beds (Thomas et al. 2000). The rate at which the community removed ammonium was also significantly dependent on hydrodynamic parameters. Studies on the effects of water velocity on mass transfer of nutrients to benth ic communities have focused on the relationship between U b and S (e.g., Bilger and Atkinson 1992; Thomas and Atkinson 1997; Thomas et al. 2000). In theory, the best correlate to nutrient uptake rates should be e since diffusive boundary layer thickness can be correlated to the rate at which turbulent energy is broken down into smaller scales (Richardson cascade) and dissipated as heat (Richardson 1922; Kolmogorov 1962; Hearn et al. 2001). The

PAGE 119

106 calculation of e i ncorporates both bottom shear stress ( t ) and an empirical measure of the roughness of the benthos (Z 0 ). Seagrass canopies bend with increased flow thereby reducing the roughness and associated friction imposed by the bottom on the water column (Thomas et al. 2000). Thus e is dependent on bending of the canopy and accounts for this change in morphology with increased flow. The relationship between S and e (Table 6) is consistent with the relationships derived in Hearn et al. (2001) and provides strong evi dence that rates of ammonium uptake for the seagrass communities used in our study were controlled by the rate at which ammonium was delivered to uptake surfaces. If rates of ammonium uptake for the benthos were mass transfer limited than the measured val ues of S based on a first order decline in nutrient concentrations over time should be in close agreement with those predicted from equation (2). However, measured values of S were nearly twice as high as those that were predicted (Fig. 19). The equation s derived by Hearn et al. (2001) relate hydrodynamic parameters to transport processes occurring at the benthic surface (i.e., the seagrass canopy) and do not account for removal of a nutrient by organisms suspended within the moving fluid. During flume e xperiments there was flow dependent resuspension of epiphytes and/or microphytobenthos that removed a significant portion (~ 30 %) of the total NH 4 + from the water column (Fig. 20). If measured values of S are calculated based on the proportion of nutrien ts removed by benthic components only, then estimates are in closer agreement (Fig. 19). This result suggests that rates of NH 4 + uptake for the benthos were in fact mass transfer limited for the seagrass bed used in this study and that uptake by suspended organisms in the water column resulted in enhanced uptake relative to that expected for

PAGE 120

107 the benthos alone. Results also emphasize the importance of considering uptake in the water column when applying mass transfer equations and models to field situation s. Nitrate uptake by individual components With the exception of epiphytes, water flow had no effect on uptake rates for components of the seagrass community. In addition, the rate at which all components removed nitrate was depressed relative to upta ke rates for ammonium (Figures 13 17). Separate study sites were used for the two datasets; therefore, comparisons between NO 3 uptake rates and NH 4 + uptake rates are limited since sites may have varied in terms of nutrient history or physiological aspect s of the organisms within the community. Differences in results among the datasets could also be a consequence of the lower beginning nutrient concentration used in measuring rates of NO 3 uptake (~ 4 m M) versus NH 4 + uptake (~ 6 m M). Michaelis Menten mod els from previous studies conducted on uptake kinetics in seagrasses suggest a 20% decrease in uptake rates with a decline in concentration from 6 to 4 m M (Lee and Dunton 1999). A similar decrease (~ 30%) in the total amount of a nutrient removed by the e ntire seagrass community can be approximated from the uptake rate constant (S). However, NO 3 uptake rates for the components were up to four times lower than those for ammonium, which suggests that some biological factor was limiting the rate at which ni trate could be processed by the organisms within the community. The absence of an effect of increased water flow on uptake rates for the seagrass leaves also suggests that rates of NO 3 uptake for the seagrasses were limited by a biological factor such as availability of carbohydrates and/or speed of enzyme reactions (Touchette and Burkholder 2000). Previous studies conducted on uptake kinetics in

PAGE 121

108 seagrasses have demonstrated preferential uptake of ammonium (Short and McRoy 1984; Terrados and Williams 1 997; Lee and Dunton 1999), which has been attributed to the physiological demands associated with assimilating nitrate. Ambient nitrate concentrations at our study site were low (~ 0.10 m M), so it is likely that components in the seagrass community lacked the enzymes to assimilate the nitrate within the short time period they were exposed to the elevated concentration in the flume. Future experiments involving trace levels of 15 NO 3 will assist in determining the relative role of hydrodynamics in uptake kinetics of nitrate at ambient concentrations. Nitrate uptake by the community Uptake rate constants (S) for nitrate were weakly dependent on hydrodynamic parameters (Table 6). The d epressed relationship between S and hydrodynamic parameters may have resulted from a reduction in friction imposed on the water column as the canopy bends with increased water flow, as was demonstrated in Thomas et al. (2000). Close agreement between meas ured and predicted values of S also suggests that rates of nitrate uptake by the community were in fact controlled by the rate of delivery (Fig. 19). However, PON represented a significant portion of the total uptake by the community (Fig. 20). Values of S based on uptake by the benthic components were depressed relative to what would be expected according to the relationship between S a nd rates of energy dissipation (Hearn et al. 2001 ) Thus the rate of nitrate uptake by the community was affected by water flow, but was probably depressed due to physiological limitations of organisms within the community. Despite the elevated concentration and apparent physiological limitation of upta ke rates, there was still a

PAGE 122

109 ~ 60% increase in the uptake rate constant (S) over the range of water ve locity. Therefore, hydrodynamics influenced rates of NO 3 uptake in these seagrass beds It is clear from the results that nutrient uptake by PON in the w ater column has important implications for the application of mass transfer models to field situations. The extent of the effect of phytoplankton uptake will depend on the physical environment and characteristics of the epiphyte and plankton communities. For instance, there were low amounts of PON in the water column for flume experiments conducted in St. Jo s e ph Bay (Thomas et al., unpubl. data) and from sites in the Florida Keys (Fig. 14). Lower uptake in the water column during these other studies resu lted in measured values of S based the whole community that were in close agreement with those predicted using mass transfer equations. However, if these same equations are used to predict uptake by the benthos in an area with dense populations of phytopl ankton or in areas that are susceptible to resuspension, then predicted values will greatly underestimate the total assimilation occurring in the community. The results identify a need to incorporate uptake by phytoplankton within mass transfer models bef ore they can be appropriately applied to field situations. Contributions of components to DIN uptake by the community Epiphytes and PON were the dominant sinks for both ammonium and nitrate (Fig. 20). The contribution of epiphytes to total NH 4 + and N O 3 uptake was probably larger than the estimated amount since there appears to have been considerable loss of 15 N during the rinsing process, which was demonstrated by the higher amount of 15 N recovered in the combined seagrass and epiphyte samples that w ere not rinsed. If the difference between rinsed and unrinsed

PAGE 123

110 samples is assumed to be equivalent to the amount lost during processing of samples, then the contribution of epiphytes to total nutrient uptake was much higher and on the order of 49% and 39% for ammonium and nitrate, respectively. It is possible that some of the nutrients had adsorbed to the surface of the epiphytes or gelatinous secretions of the diatoms but was not assimilated, and as a result was rinsed away. In addition, 15 N may have bee n lost when small cells (i.e., cyanobacteria) passed through the filters or were lysed when rinsed with distilled water. Despite potential errors in estimating contributions, our data provide evidence that epiphytes and phytoplankton are the primary compon ents removing nutrients from the water column. These results are consistent with studies that have demonstrated increased phytoplankton and epiphyte growth with elevated water column nutrients (e.g., Tomasko and Lapointe 1991; Neckles et al. 1993; Neundor fer and Kemp 1993; Short et al. 1995; Taylor et al. 1995). Seag rass leaves contributed to only 2 % of the total ammonium and nitrate removed from the water column, which suggests that epiphytes and phytoplankton out competed the seagrass for water column nutrients. However, it is important to consider that epiphytes and phytoplankton rely on nutrient pools in the water column whereas seagrass plants can utilize both water column and pore water nutrients (e.g., Iizumi and Hatorri 1982; Lee and Dunton 1999) It is not surprising that phytoplankton and epiphytes contributed the most to uptake from the water column since they are in an opportunistic location for maximizing nutrient concentrations adjacent to their surface. Furthermore, epiphytes were the most dominant component in terms of total biomass and their location on the top of seagrass leaves exposed the epiphytes to higher effective concentrations than are experienced by the leaves beneath. In Chapter Three, epiphytes

PAGE 124

111 covering seagrass leaves were s hown to inhibit NH 4 + and NO 3 uptake by up to 90% and 73 %, respectively. Therefore, the low contribution of seagrass leaves to total uptake by the community was largely due to epiphyte cover. A mass balance of the 15 N removed during experiments cannot be completed based on my data and I am unable to account for approximately 15% of the ammonium and 28% of the nitrate removed from the water column. A portion of this DIN may have been lost during filtering and rinsing of PON samples, a common artifact identi fied in previous studies (see Laws 1984). Short shoots at the base of seagrass plants were not included in the analysis and may have accounted for some uptake, especially since they were covered with epiphytic organisms. Uptake rates for individual compone nts may have been underestimated if the atom % 15 N in the water column declined appreciably during experiments. Inputs of non labeled NH 4 + through regeneration or release from the sediments and inputs of non labeled NO 3 via nitrification would reduce the atom % 15 N concentration of 15 N during experiments. This in turn would cause uptake rates for the components to be underestimated. However, it is likely that at the high concentration and atom % 15 N used in our experiments minimized dilution. Field flu me experiments conducted without a nutrient spike have demonstrated that ammonium and nitrate concentrations in the water column do not increase over the course of an experiment (Thomas, unpublished data). The amount of NO 3 unaccounted for was nearly twi ce as high as the amount for NH 4 + This may indicate that a major sink for the NO 3 were small sized cells that were lost during rinsing or perhaps there were high rates of denitrification in the upper layers of the sediments or within the epiphyte assemb lage (Kaspar 1983). Rates of denitrification have been shown to increase with enhanced

PAGE 125

112 enrichment of NO 3 despite the presence of O 2 (Kana et al. 1998). Future analysis of water samples collected during experiments will reveal whether changes in atom % 1 5 N in the water occurred and provide further insight on unexplained sinks of DIN Conclusions The deployment of a field flume and application of isotope labels in natural seagrass beds allowed for the isolation of the effects of hydrodynamics on rates of ammonium and nitrate uptake for the major photosynthetic components of a seagrass community. Results demonstrated that hydrodynamic regime plays an important role in the delivery and uptake of DIN in seagrass communities and that the relative importan ce of hydrodynamics can vary depending on the form of DIN being assimilated. Seagrass communities exist in physically dynamic environments that are exposed to a range of hydrodynamic conditions. With increasing human pressures along the coast, adverse ef fects of nutrient loading on seagrass distribution and abundance are a serious concern. The data presented here demonstrate the important role of hydrodynamics in influencing rates of uptake for two important forms of nitrogen. The interactions between b iological and physical processes in seagrass communities and the effects of these interactions on nutrient cycling processes must be studied in order to determine the fate of nutrients derived from anthropogenic sources. Furthermore, in order to develop a ppropriate models that can be used for estimating assimilative capacities of the benthos data on hydrodynamic regime must be incorporated along with the effects of uptake by phytoplankton in the water column. Future research involving application of isoto pe labels in field based studies will provide further insight into the mechanisms controlling nutrient transport in near shore and estuarine systems.

PAGE 126

113 CHAPTER FIVE HYDRODYNAMIC CHARACTERIZATION OF A CARBONATE BANK IN FLORIDA BAY: IMPLICATIONS FOR NUTR IENT UPTAKE BY THE BENTHOS Introduction The intensity and distribution of turbulent energy and stresses near the benthos is largely dependent on mean velocity and roughness of the bottom (Dade et al. 2001). Physically formed structures (i.e., sand ripp les and mounds) and biological roughness elements such as seagrasses and corals contribute to bottom roughness. The interaction between water flow and the benthos plays an important role in ecological processes, including sediment and larval transport (Ec kman et al. 1981), photosynthesis (Koch 1994; Fonseca and Kenworthy 1987), respiration (Patterson et al. 1992) and nutrient uptake (e.g., Bilger and Atkinson 1992; Thomas and Atkinson 1997; Thomas et al. 2000). There is an extensive database on the effec ts of water flow and bottom roughness on the exchange of chemicals between benthic surfaces and the water column (e.g., Boudreau and Scott 1978; Riber and Wetzel 1987; Jorgensen and Des Marais 1990). Over the past decade, a significant portion of this lit erature has investigated the effects of water flow on rates of nutrient uptake for coral assemblages (e.g., Bilger and Atkinson 1992; Bilger and Atkinson 1995; Baird and Atkinson 1997; Thomas and Atkinson 1997; Atkinson et al. 2001) and seagrass beds (Thom as et al. 2 000; Thomas and Cornelisen, 2003 ). These studies demonstrated that rates of nutrient uptake for these communities

PAGE 127

114 were limited by the rate at which the nutrient was delivered to the uptake surface. The rate of delivery, and therefore uptake, w as in turn controlled by physical factors including water flow and the roughness of the benthos. Coral reefs and seagrass beds are composed of a diverse assemblage of organisms that utilize nutrients from the water column for metabolic processes. The ra te at which the benthos removes a nutrient from the water column may be limited by the rate at which it can process the nutrient or by the rate of nutrient delivery to the benthic surface or a combination of both physiological and physical factors (e.g., B ilger and Atkinson 1992; Bilger and Atkinson 1995; Sanford and Crawford 2000). The flux ( m ) of a nutrient can be described using the commonly cited equation: m = b (C b C w ) where b is the mass transfer coefficient, and (C b C w ) represents the gradien t between the concentration in the water column (C b ) and concentration at the benthic surface (C w ) (Bilger and Atkinson 1992; Dade 1993 ). If the delivery rate exceeds the rate at which the benthic surface processes the nutrient then the concentration grad ient will be minimal (C b @ C w ). In this case biological factors such as enzyme kinetics or availability of active uptake sites are limiting uptake and b will remain relatively constant (Bilger and Atkinson 1995; Sanford and Crawford 2000). Conversely, if the rate of processing exceeds the rate of nutrient delivery (C b >> C w ), then physical factors that affect the rate of mass transfer ( b ) will in turn in fluence the flux into the benthos. Uptake by the

PAGE 128

115 benthos can also be in a transitional phase with both physical and biological factors influencing uptake rates (Bilger and Atkinson 1995; Sanford and Crawford 2000). In the studies conducted on nutrient up take by coral assemblages and seagrass beds the mass transfer coefficient ( b ) was influenced by physical factors including water velocity and bottom roughness (e.g., Baird and Atkinson 1997; Thomas and Atkinson 1997; Thomas et al. 2000). In these studies b was expressed as an uptake rate constant (S) and was calculated as the fir st order decline in water column concentration over time normalized to water volume and square area of the benthos. Uptake rate constants were measured in flume experiments and then compared to values predicted using empirically derived equations original ly intended for modeling heat and mass transfer in pipes (Dipprey and Sabersky 1963; Bilger and Atkinson 1992; Kays and Crawford 1993). Application of these models involves the prediction of a Stanton number, which represents the ratio of flux to advection over the benthic surface (S/U b where U b is the bulk velocity) (Bilger and Atkiinson 1992; Atkinson and Bilger 1992; Baird and Atkinson 1997; Thomas and Atkinson 1997; Larned and Atkinson 1997; Thomas et al. 2000). While these studies have been instrumen tal in describing the processes that control nutrient transport in benthic communities, they are largely based on bulk flow measurements and models intended for non biological roughness elements that are much smaller in scale (i.e., roughness of sand grain s) than the roughness elements in coral and seagrass communities. Hearn et al. (2001) derived equations for estimating S based on the rate at which turbulent energy is dissipated near the benthos ( e) If the rate of nutrient uptake for a square area of the benthic surface is limited by the rate of delivery to the area, then S will

PAGE 129

116 be dependent on diffusive boundary layer thickness and therefore proportional to e (S e 0.25 ; Hearn et al. 2001). Further, if the production and dissipation of turbulence in the system is balanced then S will also be proportional to bottom shear stress (S t 0. 40 ; Hearn et al. 2001 ) The relationships derived by Hearn et al. (2001) provide a means to predict rates of nutrient uptake by natural benthic communities and directly link uptake to hydrodynamic parameters that can be measured in situ While descriptive studies on the effect of biological roughness elements on hydrodynamics have been conducted (Koch and Gust 1999; Ackerman and Okubo 1993; Shashar et al. 1996; Sand Jen sen and Mebus 1996), few studies have quantitatively linked the consequence of these effects on ecological processes such as rates of nutrient removal by the benthos. In this study, a series of velocity profiles are collected over a shallow community comp osed of a patchy matrix of seagrasses, non reef building corals, sponges, and macroalgae. Data from the velocity profiles are used to describe the structure of natural tide driven flow over the benthos and estimate hydrodynamic parameters including bulk v elocity (U b ), bottom shear stress ( t ), and rate of energy dissipation ( e ). The benthic community is located in Long Key Channel (Fig. 21), which represents a major corridor for water exchange between the Western portion of Florida Bay and the Reef Track. The area is exposed to large range of velocity (0.10 and 0.50 m s 1 ) and a significant portion of the water from the everglades flows in a southeast direction and exists the bay through Long Key Channel (Wang 1998). As water from Florida Bay passes through it becomes increasingly depleted in nutrients as the benthos removes water column nutrients (Lapointe and Clar k 1992). Thus the site provides an ideal location for studying the interaction between the benthos and water flow and the potential impact of

PAGE 130

117 20 m 20 m 20 m Florida Bay Reef Track Florida Peninsula Everglades Marquesas Dry Tortugas Gulf of Mexico Hawk Channel Old Sweat Bank 0 10 km <1 m depth 1 2 4 5 3 Daybeacon Tide channels N 0 100 m Flood Ebb Old Sweat Bank Daybeacon Figure 21. Regional (A) and close up (B) map showing location of study site. Arrows indicate the dominant directional of water flow. Data were collected at the five locations indicated on Old Sweat Bank. Water depths on the bank were approximately 1 met er or less, with deeper areas within the tide channels.

PAGE 131

118 hydrodynamics on nutrient transport. Using estimates of hydrodynamic parameters and equations derived by Hearn et al. (2001), a series of uptake rate constants (S) are predicted for ammonium and pho sphate. Based on ambient nutrient concentrations, an estimate of flux ( m ) to the benthos is then predicted over a time interval. Many research programs have focused on understanding the factors controlling nutrient cycling processes within Florida Bay and throughout the Keys in an effort to improve water quality and the health of the Florida reef track and associated habitats (LaPointe and Clark 1992; Hall et al. 1999). The research presented here aims to contribute to this knowledge base by quantitatively describing the interaction between the benthos and water flow and how this interaction can potentially influence nutrient uptake by the benthos. Methods Study Site The study site was Old Sweat Bank, which is located in Long Key Channel on the Florida Bay side of the Keys, approximately 2 km northwest of US Highway 1 (Fig. 21). Benthic communities in the waters surrounding the Florida Keys include coral reefs, seagrass beds, hardbottom, and areas of bare sediment. Seagrass beds are most dominant and comprise ~ 70% of the area within the Florida Keys National Marine Sanctuary (FMRI 1998). A portion of this habitat includes shallow (~1 m depth) carbonate banks, such as the study site, that are composed of seagrasses (primarily Thalassia testudinum and Syringodium filiforme ) intermittently mixed with hardbottom organisms, including ahermatypic corals (i.e., Porites sp.), calcareous algae (i.e., Halimeda sp.), and sponges (i.e., Chondrilla sp.). The study site is exposed to semi

PAGE 132

119 diurnal tides that link W estern Florida Bay and the Atlantic Ocean. Data were collected at five locations on Old Sweat Bank that were ~100 m apart from one another in order to better characterize the community as a whole and to assess any potential effects of patchiness of organ isms and local topography on hydrodynamics (Fig. 21). Data were collected at sites one through four between 11 and 15 September 2000 and at site five on 14 December 2001. Community composition In order to describe the dominant benthic organisms contr ibuting to bottom roughness, between 6 and 12 quadrats (0.25 m 2 ) were collected at each of the five locations on Old Sweat Bank (Fig. 21). Quadrats were randomly placed within a 3 m radius of the site where hydrodynamic data were collected. The total numb er of seagrass shoots, macroalgae, coral colonies, and sponges were recorded for each quadrat. Macroalgae, corals, and sponges were identified to the genus level. In addition to identifying the dominant benthic organisms, a visual estimate of bare sedime nt (as % area not colonized) was made for each quadrat. During collection of hydrodynamic data, approximately 12 estimates of grass height (actual height and deflected height (h d ) in flowing water) were made using a meter stick. Heights of other roughne ss elements (i.e. corals, algae, sponges) were only collected on the final field day and it is assumed that the height of these roughness elements were relatively constant over the bank. A topographic, non dimensional index (measured as relief using a cha in) has also been shown to correlate with bottom friction (Thomas and Atkinson 1997). Six to 10 topographic indices at each site were estimated as the ratio of the overall length of the chain (2.5 m; link size ~ 0.05 cm) to the length of

PAGE 133

120 the chain conform ed to the benthos (not including the grass). A larger index indicates a greater amount of vertical relief. Hydrodynamic characterization Velocity data were collected using an acoustic Doppler velocimeter (Field ADV, YSI/Sontek) that measures the veloc ity of particles in three dimensions: longitudinal along the main flow (U), transverse (V), and vertical (W). The acoustic transmitting sensor was mounted to the probe housing on a flexible cable that enabled the sensor to be faced down toward the bottom o r up toward the surface of the water. The sensor was affixed to a movable arm that extended ~ 0.4 m from a vertical pole attached to a flat weighted base, which allowed the sensor to be fixed at various heights above the bottom and prevented any movement o f the probe during measurements. All measurements, with the exception of some high measurements (> 50 cm from the bottom), were made with the sensor down looking and with the X axis (U) sensing element aligned parallel to the main flow. A seri es of veloc ity profiles (n = 3 to 7) were collected at each of the five locations (Fig. 21) in order to assess the effects of the benthos on water flow characteristics near the bed (Nikora et al. 1998). Data from the logarithmic portion of the profiles were used to obtain estimates of hydrodynamic parameters that describe the structure of the boundary layer. For each profile, velocity data were recorded at 10 Hz for 1 minute (n = 600) at each of 9 to 12 heights above the sediment water interface. In order to charac terize the effects of the roughness elements on water flow characteristics, heights were closely spaced near the benthos and more widely spaced with increasing distance from the bottom. Heights included ~ 5 measurements beneath the height of the

PAGE 134

121 seagrass and ~ 5 above the seagrass up to a maximum of 70 cm above the bottom. When measurements were taken beneath the height of the grass, leaves that were directly in contact with the ADV sensors were trimmed to prevent interference with data collection. The r emoval of a small number of leaves has been shown to have no significant effect on flow measurements taken in vegetative canopies (Ikeda and Kanazawa 1996). The duration of data collection for each file was relatively short (~ 1 min), and the order in whi ch heights were used was randomized to minimize the effects of changing flow intensity during the tide cycle. Replicate measurements of velocity were collected at the beginning and end of the profile at the tallest height above the bottom in order to asse ss whether the tidal flow changed appreciably during data collection. Profiles were collected during various stages of flooding or ebbing tides. Velocity data were used to calculate hydrodynamic parameters that describe the structure of flow over the be nthos, including mean velocity for each of the components ( U V and W ), total turbulent energy (K=0.5[ ' U U + ' V V + ' W W ]) relative turbulence intens ity, calculated as K / r U where r U is the total mean velocity ( U + V + W ), and Reynolds stress ( ' W U ) (Denn y 1988; Nikora et al. 1998; Dade et al. 2001). The commonly cited Karman Prandtl equation was used to obtain estimates of shear velocity (U ) and roughness length (Z o ): U = U /k ln (Z/Z o ) (1)

PAGE 135

122 Where, U is the mea n velocity at a given distance (Z) from the bottom and k is the von Karman constant (k = 0.4). This method utilizes the portion of the profile in which velocity increases logarithmically with increasing height above the bottom. The roughness length (Z o ) was estimated as the intercept of the velocity vs. ln Z plot and is a measure of the roughness imposed by the benthos on the flowing water. For each profile an estimate of depth averaged velocity (U b ) was calculated as the velocity at the average height wi thin the logarithmic layer of the velocity gradient. Equation (1) can be modified to include an estimate of the displacement height (d): U = U /k ln (Z d/Z o ) (2) The displacement height (d) represents the height above the bottom that the logarithmi c profile extrapolates to zero velocity due to the presence of the roughness elements (Denny 1988). The inclusion of d provided little improvement in the fit of U against ln Z and therefore was assumed to be negligible. Similar results have been noted in studies on terrestrial canopies and it has been demonstrated that better estimates of U can be made by assuming d to be a constant such as zero (e.g. Dong et al. 2001). For this study, estimates of U were made using Equation (1) and assuming d=0, which results in estimates of roughness length (Z 0 ) equivalent to Z 0 +d. Only those profiles that fit the equation with an r 2 > 0.90 were retained for estimating U Confidence limits (95%) on estimates of U were calculated based on the number of measurements in the vertical profile and the regression coefficient (r 2 ) using the following expression:

PAGE 136

123 U (1 e) U U (1+e), where e is calculated as (3) e = (t a /2,n 2 )[1/n 2(1 r 2 /r 2 )] 0.5 where t is the Students t distribution with n 2 degrees of freedom for a confidence interval of (1 a ), n is the number of vertical heights within the logarithmic layer that data was collected, and r 2 is the coefficient of determination from the ln height vs. velocity regression (Grant et al. 1984). Estimates of friction coef ficient (c f ) for each vertical profile were estimated using the equation c f = 2(U 2 /U b 2 ). The friction coefficient represents the drag imposed by the benthos on the overlying water and is influenced by the roughness of the bottom and water depth as well a s the ratio of the two. Values of c f were estimated for the community in order to compare the drag imposed by the benthos on the bank to that measured for other communities including seagrass beds (Thomas et al. 2000) and coral assemblages (Thomas and At kinson 1997). Although U has the same units as velocity (m s 1 ), it is a measure of shear stress ( t ) at the benthic surface ( t = U 2 / r ). The rate of turbulent energy dissipation ( e in units m 2 s 3 ) for a bottom with roughness length (Z o ) incorporates this me asure of stress at the bottom in the following equation : e = ( t/r ) 3/2 /kZ o (4) Where k is von Karmans constant (0.4) and r is the density of seawater. This value of e represents the rate of energy dissipation within the region near the benthos. Essentially, e

PAGE 137

124 is the rate at which larger turbulent eddies are broken down into smaller and sma ller eddies until energy is given off as heat (referred to as the Richardson cascade; see Hearn et al. 2001). Estimating nutrient uptake If the rate at which the benthos removes a nutrient from the water column exceeds the rate of delivery to the upta ke surface, then S will be dependent on the molecular diffusivity of the nutrient ( D ) and the thickness of the diffusive boundary layer ( d ) (S = D / d ). The molecular diffusivity of a nutrient such as ammonium and phosphate, is several orders of magnitude lower than the diffusivity of momentum ( v ) and therefore d must be placed within the Batchelor length scale as d = a ( v D 2 / e ) 1/4 where a represents a constant that prevents overestimating turbulent mixing, and ther efore S, at such small scales (s ee Lazier and Mann 1989; Hearn et al. 2001). A value of 3 for a was chosen since it provided estimates of S that were consistent with those obtain ed through prediction of a Stanton number (s ee Bilger and Atkinson 1992). By placing uptake within a scale appropriate to diffusivity of nutrients, uptake rate constants can be predicted based on estimates of e using the equation : S = 1/ a ( D 2 e / v ) 1/4 (5) Equation 5 indicates that S will be proportional to e raised to the 1/4 power and is defined by Hearn et al. as the e 1/4 law. Because e is dependent on U b and t S will also be proportional to these parameters; U b to the 0.80 power and t to the 0. 40 power (Hearn et

PAGE 138

125 al. 2001). It is important to note that D and v, and therefore S, will vary with changes in water temperature (Li and Gregory 1974). Values of D and v used for calculating values of S were based on an average water temperature of 31 C, which was observed during data collection at locations 1 4. At 31 C and typical salinity (35 0 / 00 ) the diffusivity of ammonium and phosphate used was 2.2 10 9 and 8.3 10 10 respectively. The value of v used at these conditions was 8.2 10 7 The rate of nutrient flux ( m ) in units Moles NH 4 + or PO 4 3 removed m 2 time 1 was estimated by multiplying S by the nutrient concentration in the water column (C b ). This calculation is based on the assumption that rates of nutrient uptake for the community ar e mass transfer limited and that the concentration gradient at the uptake surface is maximal (C b >> C w where C w is the concentration at the uptake surface ) In order to calculate m we measured ambient nutrient concentrations in filtered water samples (n = 43) that were obtained at the field site during data co llection at locations 1 4 Concentrations for ammonium and phosphate (SRP) were determined using an autoanalyzer (Technicon) to within 0.05 m M. Results Community composition The benthos on Old Sweat Bank was composed of a diverse assemblage of organisms, including seagrasses, macroalgae, corals, and sponges (Fig. 22). Mean densities of these organisms, in number of individuals (shoots, colonies, etc.) per m 2 were based on a total of 44 qua drats collected at the five sampling locations. The distribution of dominant organisms over the five sites was fairly uniform, however,

PAGE 139

126 0 50 100 150 200 250 300 T. testudinum S. filiforme Halimeda sp. Sponges Corals Laurencia sp. Penicilus sp. Number/m 2 Site 1 Site 2 Site 3 Site 4 Site 5 A 0 50 100 150 200 250 300 T. testudinum S. filiforme Halimeda sp. Sponges Corals Laurencia sp. Penicilus sp. Number/m 2 B Figure 22. Densities (individuals per m 2 ) of common organ isms inhabiting Old Sweat Bank. A shoot, clump, or colony represents an individual for seagrass, algae, and coral, respectively. Results from randomly placed quadrats collected at the five sampling locations (A). Six to 12 quadrats were collected at e ach site and mean densities are shown with error bars equaling one standard deviation. Panel (B) represents the pooled data (n = 44 quadrats).

PAGE 140

127 densities of Syringodium filiforme sponges, and Laurencia sp. were variable between a few of the sites (prim arily sites 3 and 4). Seagrasses included Thalassia testudinum which had the highest mean shoot density (mean density = 176 shoots m 2 S.D. = 53), and S. filiforme (mean density = 96 shoots m 2 SD = 72), which had a lower and more variable density than T. testudinum (Fig. 22). The dominant macroalgae on the bank (based on numbers of individuals) were species of Halimeda (mean density = 88 individuals m 2 S.D. = 30). Other dominant algae on the bank were unattached clumps (5 to 10 cm diameter) o f Laurencia sp. (mean density = 25 clumps m 2 S.D. = 17) and highly variable amounts of Penicilus sp. (mean density = 10 individuals m 2 S.D. = 20). Less frequent types of macroalgae observed on the bank(< 5 individuals m 2 ) included species of the genera Caulerpa Udotea Rhipocephalus Neogoniolithon and Dictiospheria The dominant sponges were encrusting forms, including chicken liver sponge ( Chondrilla sp. ) and fire sponge ( Tedania sp. ) (mean density = 37 sponges m 2 S.D. = 22). Corals wer e also abundant and were dominated by colonies of Porites porites (forma divaricata ), Manicena aereolata and Siderastrea radians (mean density = 31 colonies m 2 S.D. = 22). Although coral density was lower than other benthic organisms (i.e. seagrasses a nd macroalgae), their relatively large size (diameter ~ 5 to12 cm) often covered a significant portion of the bottom. Visual estimates of the proportion of the bottom that was not colonized (bare sediment/limestone) were similar for all five sites and rang ed between 31 and 39 % (mean = 35 %, S.D. = 12%). During collection of hydrodynamic data, water depth ranged between 0.39 and 0.99 m and deflected height (h d ) of seagrass plants (for T. testudinum and S. filiforme

PAGE 141

128 combined) ranged between 0.10 and 0.16 m (Table 7). The overall height of the seagrass leaves ranged between 0.14 and 0.23 m. Heights of other roughness elements (i.e. coral, algae, and sponges) were lower than the seagrass heights and ranged between 0.02 and 0.12 m (mean = 0.048 m, S.D. = 0. 019, n = 53). Mean topographic index based on the chain method (length of overall chain/length of chain conformed to bottom) were 1.47 ( 0.12), 1.36 ( 0.06), 1.42 ( 0.06), 1.31 ( 0.04), and 1.44 ( 0.15) for sites 1 through 5, respectively. Hydrody namic characterization During collection of velocity data, signal to noise ratios were well above the recommended 15 db level and the correlation values for the three sensors consistently ranged between 85 and 95% (Sontek Manual). A total of 19 vertical profiles exhibited a significant fit (r 2 <0.90) to Equation (1) Two vertical profiles of mean longitudinal ( U ), transverse ( V ), and vertical ( W ) velocity are shown in Figure 23. These profiles wer e typical of the data collected on the bank and indicate that the flow is mainly unidirectional and dominant along the longitudinal axis ( U ). Velocity was low (< 0.03 m s 1 ) in both the transverse ( V ) and vertical ( W ) directions (Fig. 23). Turbulent kine tic energy (Fig. 24A) peaked near the top region of the seagrass plants. Relative turbulence intensities ranged from 0.2 to 0.8 and increased steadily as the bottom was approached, becoming slightly more scattered near the benthos (Fig 24B). Visual anal ysis of vertical distribution of Reynolds stress ( ' W U ) reveals the presence of the logarithmic layer and a roughness sublayer (Fig 24C), similar to that described by Nikora et al. (1998). Reynolds stress was highest at the top of the ro ughness elements (~0.12 14 m). Beneath this height, Reynolds stress was strongly

PAGE 142

129 Site Profile Water depth (m) Grass height (m) U b (m s 1 ) U (m s 1 ) Z o (m) C f t (N m 2 ) e (m 2 s 3 ) 10 3 1 1 0.67 0.16 0.187 0.041 0.056 0.096 1.71 3.07 1 2 0.60 0. 15 0.216 0.047 0.049 0.095 2.25 5.24 1 3 0.53 0.16 0.198 0.041 0.040 0.088 1.76 4.45 2 1 0.84 0.13 0.386 0.066 0.039 0.058 4.44 18.09 2 2 0.81 0.13 0.320 0.053 0.035 0.054 2.84 10.53 2 3 0.77 0.14 0.193 0.031 0.032 0.052 0.99 2.38 2 4 0.70 0.13 0.218 0.046 0.053 0.087 2.12 4.48 2 5 0.65 0.13 0.312 0.065 0.048 0.086 4.28 13.97 2 6 0.64 0.13 0.326 0.060 0.037 0.069 3.74 14.90 3 1 0.86 0.12 0.354 0.047 0.020 0.035 2.26 12.89 3 2 0.86 0.14 0.217 0.031 0.024 0.040 0.96 2.98 3 3 0.87 0.14 0.250 0.039 0. 033 0.049 1.58 4.54 4 1 0.87 0.12 0.168 0.027 0.035 0.051 0.74 1.39 4 2 0.93 0.11 0.235 0.032 0.025 0.038 1.08 3.46 4 3 0.99 0.11 0.306 0.039 0.020 0.032 1.54 7.39 4 4 0.95 0.11 0.273 0.032 0.014 0.027 1.02 5.63 5 1 0.48 0.13 0.160 0.036 0.041 0.099 1 .30 2.73 5 2 0.44 0.10 0.161 0.026 0.019 0.053 0.70 2.30 5 3 0.39 0.10 0.081 0.015 0.021 0.064 0.22 0.37 DG 1 0.49 0.21 0.209 0.114 0.105 0.597 13.36 35.47 DG 2 0.49 0.21 0.266 0.124 0.094 0.432 15.65 50.06 Table 7. Hydrodynamic parameters estimated for each of the vertical profiles collected on Old Sweat Bank. Included are data for the 19 profiles that had the best fit (r 2 > 0.90) to Equation (1). The last two profiles (DG 1 and 2) were collected in a small area (~ 50 m 2 ) of the bank that was coloni zed by a dense stand of T. testudinum Grass height ( deflected ) is the mean of 6 to 10 measurements taken during the profile. Parameters included depth averaged velocity ( U b ), shear velocity (U ), roughness length (Z o ), friction coefficient (c f ), bottom shear stress ( t ) and the rate of energy dissipation ( e ).

PAGE 143

0 1 2 3 4 5 6 7 0 0.1 0.2 0.3 0.4 Longitudinal velocity, U (m s -1 ) Z/h d Profile 2-6 Profile 4-3 -0.05 0.00 0.05 Transverse velocity, V (m s -1 ) -0.05 0.00 0.05 Vertical velocity, W (m s -1 ) Figure 23. Example of two velocity profiles collected on the bank. Statistics for these profiles are provided in Table 7. Profiles show mean velocity in the longitudinal (U), transverse (V), and vertical (W) directions for heights above the bottom (Z) normalized to deflected canopy height (h d ). The line indicates the height of the canopy (Z/h d = 1). Mean velocity at each height was based on the average of ~ 600 measurements (Collected at 10 Hz for 1 minute with an ADV). The above profiles were typical of those collected for the study and indicate that water flow over the community was unidirectional and tide drive n during data collection. 130

PAGE 144

0 1 2 3 4 5 6 7 0 1 2 3 4 5 Turbulent energy, K (m s -1 ) 2 x 10 3 Z/h d 0 1 2 3 4 5 6 7 0 0.2 0.4 0.6 0.8 1 Turbulence Intensity, K0.5/U Z/h d 0 1 2 3 4 5 6 7 -0.5 0 0.5 1 1.5 Reynolds stress, -U'W' (m s -1 ) 2 x 10 3 Z/h d Profile 2-6 Profile 4-3 Figure 24. Vertical profiles of (A) total turbulent energy (K = 0.5[ ' U U + ' V V + ' W W ]), (B) relative turbulence intensity ( K / r U ), and (C) Reynolds stress ( ' W U ) Parameters were calculated from the data collected for profiles in Fig. 23. Depth above the bottom (Z) w as normalized to deflected canopy height (h d ). The line indicates the height of the canopy (Z/h d = 1) A B C 131

PAGE 145

132 influenced by the benthos and decre ased toward the bottom In the logarithmic layer, Reynolds stress decreased toward the waters surface, where it is a ssumed to be 0 in the absence of wind stress and waves A graph of Reynolds stress vs. relative depth shows the distinct layers within the boundary layer (Fig 25). Hydrodynamic parameters estimated from each of the velocity profiles, including depth aver aged velocity (U b ), shear velocity (U ), roughness length (Z o ), friction coefficient ( c f ), bottom shear stress ( t ), and rate of energy dissipation ( e ) are provided in Table 7. Depth averaged velocity (U b ) on the bank ranged between 0.08 and 0.39 m s 1 E stimates of shear velocity (U ) ranged between 0.016 and 0.066 m s 1 and were positively correlated to U b (Fig. 26). Regressions for determining U had a mean r 2 of 0.96 (S.D. = 0.03), suggesting a good fit to equation 1. Due to the small number of heig hts (n = 4 to 8) used in estimating the slope, the 95% confidence intervals were relatively high (~ 30%, 12%) in comparison to other boundary layer studies (e.g., < 22 % in Cheng et al. 1999). However, setting a high r 2 threshold (>0.90 vs. >0.80 in Che ng et al. 1999) minimized potential error in our estimates. Roughness length (Z o ) ranged between 0.013 and 0.062 m. Grass height was positively correlated to Z o and explains a portion (~46%) of the variability in the roughess length (Z o = 0.46 (grass he ight) 0.02, r 2 = 0.46, p<0.001, n = 19). Roughness length represented approximately 13 to 40% of the grass height (mean = 26%, S.D. = 0.08, n = 19). Friction coefficients for the bank (range = 0.020 and 0.096, mean = 0.062, SD = 0.023, n = 19) were with in the range of those for low and high relief coral rubble (Thomas and Atkinson 1997), and were considerably lower than those estimated for dense stands of T. testudinum (Fig. 27) and were dependent on the ratio of grass height to

PAGE 146

133 -0.2 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 Relative depth (1-Z/D) Reynolds stress (m s -1 ) 2 x 10 3 Roughness sublayer Logarithmic layer -U'W'=U 2 (1-Z/D) U =0.03 m s -1 Figure 25. Turbulent Reynolds stress ( ' W U ) plotted versus relative depth (1 Z/D) for the three profiles collected at site 1 (see Table 7 for statistics), where D was the depth of the water and Z was the height above the botto m that the ADV measurement was taken. The plot shows the logarithmic layer and the roughness sub layer of the benthic boundary layer. As demonstrated in Nikora et al. (1998), if the data fit the logarithmic law, then the slope of the decline in Reynolds stress between the roughness sublayer and the surface of the water (1 Z/D = 0) can be used to approximate U Here the estimate of U = 0.03 m s 1 is slightly lower than values obtained using Equation 1 for the three profiles collected at site 1 (See Ta ble 7). The reduction of ' W U to near 0 at the surface of the water reveals that there was little affect of wind stress/waves on water flow.

PAGE 147

134 0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.10 0.20 0.30 0.40 0.50 U b (m s -1 ) U (m s -1 ) Figure 26. Estimates of shear velocity (U ) versus depth ave raged velocity (U b ). Error bars represent 95% confidence limits based on the expression described in Equation 3 in the text (Grant et al. 1984). Linear regression results: U = 0.14U b + 0.01, r 2 = 0.66, P < 0.001, 95% C .L. s on slope = 0.09 and 0.19.

PAGE 148

135 0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 150 200 250 300 350 Re (x 10 3 ) C f Five sites from this study Area of dense T. testudinum T. testudinum (Thomas et al. 2000) Low relief coral rubble (Thomas and Atikinson 1997) High relief coral rubble (Thomas and Atkinson 1997) Figure 27. Friction coefficients (c f ) for various bottom types as a function of flow Reynolds number (Re = U b h/ v, where h is the height of the water). Included in the graph are data collected in an area of dense seagrass on the bank and from T. testduninum beds elsewhere (Thomas et al. 2000). Friction coefficients for low and high relief coral rubble from Thomas and Atkinson (1997) are also shown for comparison. For this study c f was estimated as 2(U 2 /U b 2 ), whereas, th ose for Thomas et al. and Thomas and Atkinson were estimated as 2ghs/U b 2 where g is gravity, h is the height of the water, and s is the slope of the waters surface. Theoretically, these methods provide comparable estimates of bottom shear stress (e.g., U = t r / = U b cf / 2 ), as demonstrated by the close agreement between data points for dense beds of seagrass.

PAGE 149

136 water depth ( c f = 0.38(h d /depth) 1.10 r 2 = 0.71, p<0.001). There was no dependence of c f on U b ; howe ver, there was a weak negative dependence of c f on the flow Reynolds number ( c f = 0.15 3 (Re) + 0.09, r 2 = 0.33, P<0.05; Fig. 27). Values of bottom shear stress ( t ), estimated from U were between 0.22 and 4.44 N m 2 and rates of energy dissipation ( e ) were between 0.37 and 18.09 10 3 m 2 s 3 over the range of U b (Table 7). Rates of energy dissipation were dependent on U b ( e = 0.17U b 2.46 r 2 = 0.92, P < 0.001, 95% C.L.s on slope 2.10 = 2.81; Fig. 28). and t ( e = 0.0027 t 1.21 r 2 = 0.87, p<0.001, 95% C .L.s on slope 0.97 = 1.44). Estimating nutrient uptake Estimates of energy dissipation rate ( e ) were used in Equation (5) to predict a series of uptake rate constants (S) for ammonium and phosphate. Based on field conditions during data collection (wate r column temperature of 31 C), predicted values of S for ammonium ranged between 7.1 and 18.9 10 5 m s 1 Predicted values were in the same range as measured values of S obtained from field flume experiments conducted on the bank (Fig 29). In addition t o the expected relationship for S vs. e both predicted and measured uptake rate constants were in close agreement with the expected relationships for S vs. t and S vs. U b (See Fig. 29; Hearn et al. 2001). Due to a lower diffusivity than NH 4 + predicted values of S for phosphate were lower than those for ammonium and ranged between 4.3 and 11.7 10 5 m s 1 over the range of e Based on the assumption that nutrient uptake by the benthos was limited by the rate of delivery to uptake surfaces, the flux ( m ) of a nutrient into the benthos was p redicted using estimates of S and the ambient water column concentration of the

PAGE 150

137 y = 0.17x 2.46 R 2 = 0.92 0.000 0.004 0.008 0.012 0.016 0.020 0.00 0.10 0.20 0.30 0.40 0.50 U b (m s -1 ) e (m 2 s -3 ) Figure 28. Rate of turbulent energy dissipation ( e ) versus velocity (U b ). For flow over the bank, e is proportional to U b raised to the 2. 5 power ( e = 0.17U b 2.46 r 2 = 0.92, P < 0.001, 95% CL s on slope 2.10 = 2.81).

PAGE 151

0 5 10 15 20 0 0.1 0.2 0.3 0.4 e 0.25 S (m s -1 ) x 10 5 Predicted Measured (Thomas et al., in prep) 0 5 10 15 20 0.0 0.5 1.0 1.5 2.0 t 0.40 S (m s -1 ) x 10 5 0 5 10 15 20 0.0 0.1 0.2 0.3 0.4 0.5 U b 0.8 (m s -1 ) S (m s -1 ) x 10 5 Figure 29. Predicted uptake rate constants (S) for ammonium versus (A) energy di ssipation rate raised to the 0.25 power (B) bottom shear stress raised to the 0.40 power, and (C) bulk velo city raised to the 0.80 power Plots are based on the expected relationships between S and hydrodynamic parameters (Hearn et al. 2001). Lines rep resent the best linear fit for the data with an intercept set at 0. Also included in the plots are estimates of S that are based on ammonium uptake measured during field flume experiments conducted at the same field site (Thomas et al., in prep ). Measured S is calculated as k V / A which is the first order decline (k) of NH 4 + in the water column over time normalized to volume/bott om area enclosed by the flume (s ee Tho mas et al. 2000 ). Hydrodynamic parameters for measured S were based on profiles collected in the field flume during uptake experiments. A B C 138

PAGE 152

139 nutrient ( m = S C b ). Water column samples were collected during this study and analyzed for NH 4 + and PO 4 3 Mean concentrations for 43 samples were 0.31 (SD = 0.10) and 0.04 (SD = 0.03) m Mol/L for NH 4 + and PO 4 3 (SRP), respectively. Over the range of flow conditions and these average nutrient concentrations the benthos is expected to remove between 2.2 and 5.9 10 7 m Mol NH 4 + m 2 s 1 and between 0.17 and 0.46 10 7 m Mol PO 4 3 m 2 s 1 Discussion The benthos on Old Sweat Bank (OSB) is comprised of a diverse assemblage of seagrasses, corals, macroalgae, and sponges (Fig. 22). These organisms collectively form a rough surface that influences boundary layer characteristics. Although the presence of the seagrass plants contributed to bottom roughness, hydrodynamic characteristics more closely resembled those of moss covered cobbles (Nikora et al. 1998) or low relief coral assemblages (e.g., Thomas et al. 1997; Gardella and Edmunds 2001 ) than those of seagrass beds (e.g., Gambi et al. 1990; Koch and Gust 1999; Thomas et al. 2000). For ammonium, predicted uptake rate constants (S) for the bank were approximately one half those measured and predicted for uniform seagrass beds (Thomas et al 2000; Thomas and Cornelisen 2003 ). However, predicted values of S for ammonium and phosphate were similar to those predicted and measured for coral assemblages with similar degrees of roughness (see Thomas and Atkinson 1997; Hearn et al. 2001). If the rate of nutrient uptake by the benthos is limited by the rate at which the nutrient is delivered to the benthic surface, then uptake rates will be strongly influenced by physical factors such as water flow and roughness of the bottom (e.g., Bilger and Atki nson 1992). In this case

PAGE 153

140 rates of nutrient uptake for the community on the bank will continually fluctuate as flow intensity changes over a tide cycle. In addition, changes in nutrient concentrations and water temperature, or roughness with changeover in community composition, will also influence the flux of nutrients into the benthos. Community composition In a benthic survey of the waters surrounding the Florida Keys, OSB was described as an area of patchy seagrass made up of discontinuous beds of m oderate to high densities (FMRI 1998). While the bank had a few small areas (~50 m 2 ) of dense seagrass along the edges and adjacent to tide channels, the majority of the benthos was a mixture between a seagrass and hardbottom community, with seagrasses spa rsely mixed among macroalgae, solitary colonies of ahermatypic corals, and sponges (Fig. 22). Approximately 1/3 of the bottom was not colonized by these organisms but rather covered by layers of coarse, calcareous sand over Key Largo limestone. In additi on, low and highly variable shoot densities in comparison to T. testudinum beds surveyed elsewhere (Fonseca and Cahalan 1992; Thomas et al. 2000; Koch and Gust 1999) suggest that the seagrass was intermittently mixed and did not form a uniform canopy of se agrass shoots. Based on numbers of individuals per square area, seagrasses are the dominant benthic feature on the bank (Fig. 22). However, less abundant organisms such as colonies of Porites and clumps of Laurencia were ~ 5 to 15 cm in diameter and oft en covered a larger area of the bottom than the seagrass plants. Heights of these and other roughness elements were approximately 1/3 the height of the seagrass leaves. In addition, estimates of bottom roughness using the topographic index were within the same range as those

PAGE 154

141 estimated for low relief assemblages of coral (Thomas and Atkinson 1997). Due to the sparse and patchy distribution of seagrasses, these organisms likely made a significant contribution to the overall roughness of the benthos and there fore contributed to the effects of the bottom on water flow. More detailed surveys of these diverse communities are needed to better quantify the relative contributions of different types of organisms to overall bottom roughness. Hydrodynamic characteriz ation The composition and patchiness of the benthos resulted in hydrodynamic conditions that were different than those des cribed for seagrass beds ( Gambi et al. 1990; Koch and Gust 1999). To provide a visual comparison of flow over the bank to flow over a typical seagrass bed, two velocity profiles were collected within a small area (< 50 m 2 ) on the edge of the bank that was colonized by a more dense (560 shoots m 2 ) and uniform stand of T. testudinum (Fig. 30; Table 7). The shape of velocity profiles c ollected on the bank was characteristic of perturbed boundary layer flow (Figures 23A and 30A ), whereas profiles in the dense stands of seagrass resembled a hyperbolic tangent (Fig. 30 A ) that is more indicative of a mixing layer (Raupach et al., 1996; Ghis alberti and Nepf, 2002). Attenuation of velocity beneath the height of the seagrass for profiles collected on the bank was not evident in comparison to profiles collected in de nse seagrass In a dense and evenly distributed stand of seagrass, the flexibl e plants form a semi sealed canopy that significantly reduces shear stresses on the bottom relative to above the canopy, which in turn promotes particle retention within the meadow (Fonseca and Fisher 1986). Reduced flow attenuation on the bank in compari son to typical seagrass beds suggests that import and accrual of sediments on the

PAGE 155

142 0.0 0.5 1.0 1.5 2.0 0.0 0.1 0.2 0.3 0.4 0.5 Velocity, U (m s -1 ) Z/h d 0.0 0.5 1.0 1.5 2.0 -1 0 1 2 3 -U'W' (m s -1 ) 2 x10 3 Z/h d 0.0 0.5 1.0 1.5 2.0 0 10 20 30 40 K (m s -1 ) 2 x10 3 Z/h d 0.0 0.5 1.0 1.5 2.0 0.0 0.2 0.4 0.6 0.8 1.0 Turbulence intensity, K 0.5 / U Z/h d Figure 30. (A) Vertical profiles of velocity (U ), (B) Reynolds stress ( ' W U ), (C) total turbulent energy (K = 0.5[ ' U U + ' V V + ' W W ]), and (D) relative turbulence intensity ( K / r U ) for a velocity profile collected on OSB (profile 2 1 indicated by the open circles ) and in an area of dense grass (profile DG 2 indicated by the closed circles ). The y axis is depth above the bottom (Z) normalized to deflected canopy height (h d ) and the dotted line represents the canopy height. See Table 7 for details on these profiles. A B C D

PAGE 156

143 bank may be restricted to areas of dense seagrass and during brief periods of slack tide. Furthermore, currents remained strong (U b ~ 0.20 to 0.40 m s 1 ) during ebbing and floo ding tides and the direction of tides would shift within a matter of minutes, which indicates that the benthos on the bank is almost always exposed to high currents and, therefore, high shear stresses. As a result of these conditions, existing sediments o n the bank are likely supplied by locally produced s ources, such as calcareous macroalgae Limited import and accumulation of sediments would in turn limit seagrass density and distribution. Furthermore, the physically dynamic nature of the bank may coun teract any enhanced growth of seagrasses that may result from increased nutrient delivery associated with increased water flow (Fonseca and Kenworthy 1987). Turbulent energy (K) and stress ( ' W U ) was enhanced near the top of the seagrass l eaves on the bank; however, considerably less so than in dense stands of seagrass (Fig. 30; Verduin and Backhaus 2000). In the dense bed, synchronous waving of the canopy (monami) was visually observed and likely contributed to the high energy toward the top of the canopy (s ee Ghisalberti and Nepf, 2002). Lower turbulent energy and stress in profiles collected on the bank in comparison to the dense bed is likely attributed to the patchy distribution and low shoot density of the seagrass, which resulted in less interaction between flapping leaves and the water column. The vertical distribution of Reynolds stress reveals both the logarithmic layer and roughness sublayer within the benthic boundary layer and resembles the range and distribution of ' W U measured ov er moss covered cobbles (Fig. 25 ; Nikora et al. 1998). In addition, the steady decline of Reynolds stress toward 0 at the surface of the water indicates that flow was primarily tide generated rather than influenced by wind shear and s urface waves.

PAGE 157

144 Turbulence intensity was estimated as turbulent energy normalized to mean total velocity. In aquatic vegetation, turbulence intensity typically peaks at the top of the canopy and declines within the canopy as the bottom is approached ( Fig. 30D; Gambi et al. 1990; Nepf and Vivoni 2000). Although the presence of the grass on the bank may enhance turbulent energy, the vertical distribution of relative turbulence intensity near the benthos is not evocative of a dense canopy. The uniform decline in turbulence intensity with depth is more typical of non vegetated bottoms and is the product of the low energies at the five sites (in comparison to the dense grass) and the virtual absence of flow attenua tion within the canopy (Fig. 30A ). Several hy drodynamic parameters, including shear velocity (U ) and stress ( t ), friction coefficient ( c f ), and rate of turbulent energy dissipation ( e ), were estimated in order to describe the effects of bottom roughness on boundary layer flow and the potential implications of these effects on nutrient uptake. Shear velocity at co mparable U b were within the range of those in Nikora et al. (1998) and were ~50% lower than U in more dense stands of seagrass (Table 7; Gambi et al. 1990). Increased bottom roughness enhances bottom shear, and it is apparent from these results that the roughness elements on the bank collectively form a surface that has less of an effect on the overlying water than a typical seagrass bed. The effect of the benthos on U was variable over the range of velocity and among the five locations on the bank (Tab l e 7; Fig. 26 ). Over a uniform rough surface (i.e., a sand bottom) U should be directly correlated to U b (Cheng et al. 1999). Part of the dissociation between U and U b is probably due to error in determining U However, dissimilarities in grass height s and water depths among sampling locations likely contributed to a significant portion of the variability in the relationship between U

PAGE 158

145 and U b For instance, those sites with the shortest grass and the lowest topographic index (i.e., sites 4 and 5) have the lowest values of U and bottom shear stress ( t ) (Table 7). Furthermore, the friction coefficient, which is essentially the ratio of U to U b also varied among sites in part due to differences in grass height and water depth. So although hydrodynami c parameters measured on the bank were different than those estimated in previous studies conducted on aquatic vegetation, the presence and height of the seagrass still had some influence on water flow characteristics. Despite some dependence on grass he ight, friction coefficients for the bank were ~ 5 times lower than those estimated for the area of dense T. testudinum and estimates from Thomas et al. (2000) over a similar range of flow Reynolds number (Fig. 27). Friction imposed by the benthos inhabiti ng the bank was equivalent to that estimated for assemblages of coral rubble with similar topographic relief ( see Thomas and Atkinson 1997). Therefore, the total drag imposed by the bottom remained within the same range as those measured in Thomas and Atk inson (1997) despite contributions of seagrass to bottom roughness and c f Another important observation is that the c f for the five locations remained relatively constant over a range of Re despite the presence of flexible seagrass leaves. In beds of aqu atic vegetation, the friction coefficient is dependent on water velocity due to the bending of the grass with increased flow (Thomas et al. 2000). A weak relationship between c f and Re for our data suggests that shoot density and plant heights were too lo w to create a sealed canopy which would have reduced friction as the canopy deflected with increased flow. In addition, there was variability in overall grass height among the sites resulting in no correlation between deflection of the canopy (h d ) and water velocity. Clearly, several factors influence the friction coefficient and no

PAGE 159

146 one parameter measured at the site is a good predictor. This is largely due to the patchy nature of the system and the variable grass heights, water depths, and topographic relief. Variability in hydrodynamic parameters among locations on the bank provides evidence that small scale variation in local topography (on the order of meters) and bottom roughness influences the amount of drag experienced by the benthos. As a resul t, those organisms surrounded by taller roughness elements and in areas of greater topographic relief will be exposed to enhanced shear energy and stress relative to organisms located within areas of the bank that are surrounded by lower profile roughness elements. Such effects of local topography on flux of momentum would in turn affect flux of nutrients and particles (i.e., larvae, food) to benthic organisms. Estimating nutrient uptake Estimation of energy dissipation rates ( e ) allowed us to use the e 1/4 law equation to calculate uptake rate constants (S) for the benthic community on Old Sweat Bank. Predicted values of S were based on the assumption that the rate of nutrient uptake by the benthos is limited by the rate at whi ch the nutrient is delivered to uptake surfaces (mass transfer limited), rather than by a biological factor such as availability of active uptake sites or enzyme kinetics. Assuming mass transfer limitation, uptake rate constants for ammonium and phosphate were in the same range as those measured and predicted for coral reef assemblages of similar scale roughness (Baird and Atkinson 1997; Thomas et al. 1997; Hearn et al. 2001). This result implies that the organisms on OSB collectively form a rough surface that dissipates energy at a similar rate as the coral assemblages in these studies, which is supported by similar friction coefficients for the two types of communities (Fig. 27).

PAGE 160

147 Despite the presence of seagrass plants on the bank, predicted values of S f or ammonium were approximately half those measured and predicted for typical seagrass beds (Thomas et al. 2000; Thomas and Cornelisen 2003 ). A dense and uniform distribution of seagrass plants imposes a greater friction on the water column (Fig. 27) resul ting in a higher rate of dissipation in the system and in turn, higher uptake rates T he sparse and patchy distribution of seagrass on the bank may not affect bottom roughness enough to enhance uptake rates to the level of observed in dense beds of seagr ass. Under similar intensities of velocity it is likely that areas of dense grass along the edges of the bank remove a similar amount of nutrients over time as the seagrass beds i n Thomas et al. (2000) Assuming rates of ammonium and phosphate upta ke fo r the benthos are maximal flux ( m ) can be estimated by multiplying values of S by the bulk concentration in the water column (C b ) Estimates of flux based on ambient concentrations were higher for ammonium than for phosphate, which was a consequence of t he higher molecular diffusivity of ammonium and lower concentrations of phosphate versus ammonium in the water column. While estimates of m are somewhat informative by providing an estimate of the capacity of the benthos to remove a nutrient over a range of hydrodynamic conditions, they do not account for the changes in flow intensity that occurs over a tide cycle. A simple model using data acquired in this study can be constructed t o demonstrate the potential effect of changes in flow intensity on nutrie nt upta ke by the benthos. First, values of U b from the profiles and all estimates of velocity collected at a height above of the bottom of approximately 35 cm (roughly mid water depth) were plotted versus the time of data collection (Fig 31A ). The additi on of velocity data

PAGE 161

148 collected at 35 cm created a more robust dataset, which in turn allowed for the data to be fit to a modified sine wave for approximating the pattern of change in U b over time. Using the relationship between e and U b (Fig. 28), values o f S were predicted using equation 5 and were then multiplied by ambient concentration of ammonium (S (m s 1 ) 0.31 m Mol NH 4 + L 1 10 6 L m 3 ) to estimate flux to a square area of the benthos over time (Fig. 31B ). This model demonstrates that uptake is lowe st near slack tide (when U b is lowest) and highest during the peak of ebbing or flooding tides (when U b is highest). If we take the in tegral of the plot in Figure 31B the total ammonium removed by the benthos over a 12 hr period is approximately 0.0015 M oles m 2 For phosphate, the total flux over this same time interval would be approximately 0.0002 Moles m 2 which is mu ch lower due to the lower water column concentration and lower diffusivity of phosphate. The model presented in Figure 31 is in its simplest form using the data currently available. Several environmental factors will contribute to variations in nutrient flux, including fluctuations in nutrient concentration and water temperature. For instance, the total flux would increase by an amou nt directly proportional to an increase in nutrient concentration. Water temperature influences molecular diffusivity ( D ) and kinematic viscosity ( v ), which are both important variables in equation 5. The community on Old Sweat Bank is exposed to a tempe rature range of approximately 10 C throughout the year (~ 22 C in winter up to ~ 32 C in summer). Based on these extremes, total flux will increase by approximately 15 to 20% during the summer months in comparison to winter months. In addition to concent ration and temperature, changes in community composition that result in a change in bottom roughness will also impact flux. For

PAGE 162

149 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0 10 20 30 40 50 60 Time (hr) Velocity (m s -1 ) 35 cm height Depth averaged 0 0.01 0.02 0.03 0.04 0.05 0 10 20 30 40 50 60 Time (hr) Nutrient uptake ( m mol m -2 s -1 ) Ammonium Phosphate Figure 31. (A) Bulk velocity over time during the period of data collection and (B) predic ted flux of ammonium and phosphate over the same time interval. Values of U b include those estimated from the velocity profile and values of U from measurements taken at ~ 35 cm above the sediment water interface. A modified sine wave is fitted to the velocity data. Predicted flux was calculated as S (m s 1 ) C b where S was based on equation ( 5 ) and the relationship between e and U b (Fig. 28) and C b was the average concentration of NH 4 + (~ 0 .31 m Mol per 0.001 m 3 ) and PO 4 3 (~ 0.04 m Mol per 0.001 m 3 ) A B

PAGE 163

150 instance, if the community was to be replaced by a typical T. testudinum bed, nutrient uptake would nearly double due to the substantial increase in friction (Fig. 27) and subsequent increase in rates of energy dissipation. The opposite would occur with a conversion of the benthic community on Old Sweat Bank to bare substratum. Finally, as suggested by Hearn et al. (2001) and d emonstrated by Thomas and Cornelisen (200 3 ), the presence of waves could enhance flux to the benthos by as much as 20% to 50% Predicted uptake for the benthos is based on assumptions of mass transfer limitation and it is possible that uptake rates for t he community on the bank are not mass transfer limited, but rather biologically limited or within a transitional phase where uptake is influenced by both physical and biological factors (Sanford and Crawford 2000). Previous studies on coral and seagrass c ommunities have demonstrated close agreement between predicted values of S and those measured in flume experiments; thus nutrient uptake by these communities was considered mass transfer limited (Atkinson and Bilger 1992; Baird and Atkinson 1997; Thomas an d Atkinson 1997; Thomas et al. 2000; Thomas and Cornelisen 2003 ; Hearn et al. 2001; Atkinson et al. 2001). In another study conducted on Old Sweat Bank, values of S for ammonium were measured for the benthos using a field deployed flume (Thomas et al., in prep). Uptake rate constants were based on the first order rate of decline in ammonium concentration measured over a range of water velocity. Measured values of S from Thomas et al. (in prep) are within the same range as those predicted in this study (F ig. 29 ). Additional flume experiments have been conducted at the field site using 15 N labelled ammonium, which allowed for the assessment of flow effects on nutrient uptake by specific components of the community (Fig. 32) Results from these experiments

PAGE 164

151 Thalassia testudinum 0.0 0.5 1.0 1.5 Low flow High flow r (gNH 4 -N (g) -1 s -1 ) x 10 8 P < 0.02 Epiphytes ( T. testudinum ) 0.0 0.5 1.0 Low flow High flow r (gNH 4 -N(g) -1 s -1 ) x 10 8 P < 0.01 Syringodium filiforme 0.0 0.5 1.0 1.5 Low flow High flow r (gNH 4 -N(g) -1 s -1 ) x 10 8 P < 0.05 Epiphytes ( S. filiforme ) 0.0 1.0 2.0 3.0 Low flow High flow r (gNH 4 -N (g) -1 s -1 ) x 10 8 P < 0.02 Figure 32. Results from a series of paired 15 N labeled NH 4 + uptake experiments conducted in a field flume on Old Sweat Bank. Paired experiments (n = 7) were conducted at low (U b = 0.03 to 0.07 m s 1 ) and high (U b = 0.10 0.30 m s 1 ) fl ow. Comparisons are shown for T. testudinum epiphytes of T. testudinum S. filiforme epiphytes of S. filiforme Halimeda monile Laurencia papillosa Porites porites and PON (phytoplankton). Some organisms were not present during some experiments. The c ommunity was exposed to a spike (4 m M) of 15 NH 4 + for one half hour. See Chapter Two for methods of collection and calculation of uptake rates ( r ). Data were analyzed using a paired t test (for data paired with lines only) and significa nce is expressed as P < 0.05, P < 0.01 P < 0.0 01 or no t significant (NS)

PAGE 165

152 Figure 32: (Continued) Halimeda monile 0 0.1 0.2 0.3 0.4 0.5 Low flow High flow r (gNH 4 -N (g) -1 s -1 ) x 10 8 P < 0.05 Laurencia papillosa 0.0 0.5 1.0 1.5 Low flow High flow r (gNH 4 -N(g) -1 s -1 ) x 10 8 P < 0.001 Porites sp 0.00 0.05 0.10 0.15 Low flow High flow r (gNH 4 -N(g) -1 s -1 ) x 10 8 P < 0.05 PON 0 0.2 0.4 0.6 0.8 Low flow High flow r (gNH 4 -N (liter) -1 s -1 ) x10 9 Not Significant

PAGE 166

153 indicated that ammonium uptake by many of the benthic organisms, including seagrasses and their associated epiphytes, corals, and macroalgae is dependent on hydrodynamic conditions (Fig. 32). These data along with results from Thomas et al. (in prep) provide evidence that rates of ammonium uptake for benthic components and the community as a whole are mass transfer limited; therefore, application of mass transfer equations from Hearn et al. (2001) provid e reasonable estimates of ammonium uptake by the benthos (Fig. 29). Future flume experiments for other nutrients such as phosphate and nitrate will assist in determining the applicability of mass transfer equations for estimating uptake of these nutrients by the benthos. Conclusions The benthic composition of Old Sweat Bank is composed of a diverse assemblage of organisms that collectively form a rough surface. De spite the presence of seagrasses near bed flows and boundary layer characteristics for th e bank were different than those described within and above dense stands of seagrass. As a result, uptake rates for the benthos are expected to be lower than seagrass beds and within the same range as those measured and estimated for coral assemblages (e.g ., Thomas and Atkinson 1997; Hear n et al. 2001), which exhibit similar degrees of roughness and friction (Fig. 27). The construction of a simple model based on available data demonstrates the potential of predicting nutrient uptake by the benthos over tim e by integrating hydrodynamic data within larger physical models. With more continuous periods of data collection and addition of added variables (i.e. fluctuations in water temperature and nutrient concentrations), more robust hydrodynamic and circulation models could be developed for predicting nutrient transport over larger spatial and temporal scales. This

PAGE 167

154 would require long term deployment of sophisticated instruments that can continuously collect both water quality and hydrodynamic data. However, su ch efforts would prove beneficial for the monitoring and management of water bodies that are continually undergoing changes in water quality such as Florida Bay

PAGE 168

155 REFERENCES CITED Ackerman, J. D. 1986. Mechanistic implications for pollination in the mari ne angiosperm Zostera marina. Aquatic Botany. 24: 343 353. Ackerman, J. D. and Okubo, A. 1993. Reduced mixing in a marine macrophyte canopy. Functional Ecology. 7: 305 309. Atkinson, M. J. 1987. Rates of phosphate uptake by coral reef flat communities. L imnology and Oceanography. 32: 426 435. Atkinson, M. J ., Falter, J. L. and Hearn, C. J. 2001. Nutrient dynamics in the Biosphere 2 coral reef mesocosm: water velocity controls NH 4 and PO 4 uptake. Coral Reefs. 20: 341 346. Baird, M. and Atkinson, M. J. 19 97. Measurement and prediction of mass transfer to coral reefs. Limnology and Oceanography. 42: 1685 1693. Bilger, R. W. and Atkinson, M. J. 1992. Anomalous mass transfer of phosphate on coral reef flats. Limnology and Oceanography. 37: 261 272. Bilger, R. W. and Atkinson, M. J. 1995. Effects of nutrient loading on mass transfer rates to a coral reef communtiy. Limnology and Oceanography. 40: 279 289. Boudreau, B. P. and Scott, M. R. 1978. A model for the diffusion controlled growth of deep sea manganese nodules. American Journal of Science. 278: 903 929. Button, D. K. 1991. Biochemical basis for whole cell uptake kinetics: Specific affinity, oligotrophic capacity and the meaning of the Michaelis constant. Applied and Environmental Microbiology. 57: 2033 2038. Cheng, R. T., Ling, C H., and Gartner, J.W. 1999. Estimates of bottom roughness length and bottom shear stress in South San Francisco Bay, California. Journal of Geophysical Research. 104: 7715 7728. Cornelisen, C. D. and Thomas, F.I.M. 2002. Ammo nium uptake by seagrass epiphytes: Isolation of the effects of water velocity using an isotope label. Limnology and Oceanography. 47: 1223 1229.

PAGE 169

156 Cowan, J. L., Pennock, J. R. and Boynton, W. R. 1996. Seasonal and interacnnual patterns of sediment water nu trient and oxygen fluxes in Mobile Bay, Alabama (USA): Regulating factors and ecological significance. Marine Ecology Progress Series. 141: 229 245. Dade, W. B. 1993. Near bed turbulence and hydrodynamic control of diffusional mass transfer at the sea flo or. Limnology and Oceanography. 38: 52 69. Dade, W. B., Hogg A. J. and Boudreau, B. P. 2001. Physics of flow above the sediment water interface. In: B.B. Jorgensen and B.P. Boudreau (Editors), The Benthic Boundary Layer. Transport Processes and Biogeoche mistry. Oxford University Press, New York, pp. 4 43. Dennison, W. C. and Barnes, D. J. 1988. Effect of water motion on coral photosynthesis and calcification. Journal of Experimental Marine Biology and Ecology. 115: 67 77. Denny, M. W. 1988. Biology and the Mechanics of the Wave Swept Environment. Princeton, NJ: Princeton University Press. 329 p. Dickson, M. L. and Wheeler, P. A. 1995. Nitrate uptake rates in a coastal upwelling regime: A comparison of PN specific, absolute, and Chl a specific rates. Lim nology and Oceanography. 40: 533 543. Dipprey, D. F. and Sabersky, R. H. 1963. Heat and momentum transfer in smooth and rough tubes at various prandtl numbers. Journal of Heat and Mass Transfer. 6: 329 353. Dong, Z., Gao S. and Fryrear, D.W. 2001. Drag C oefficients, roughness length and zero plane displacement height as disturbed by artificial standing vegetation. Journal of Arid Environments. 49: 485 505. Dugdale, R. C. and Wilkerson, F. P. 1986. The use of 15 N to measure nitrogen uptake in eutrophic oc eans: experimental considerations. Limnology and Oceanography. 31: 673 689. Dugdale, R. C. and Goering, J. J. 1967. Uptake of new and regenerated forms of nitrogen in primary productivity. Limnology and Oceanography. 12: 196 206. Eckman, J. E., Nowell, A R. M. and Jumars, P. A. 1981. Sediment destabilization by animal tubes. Journal of Marine Research. 39: 361 374. Eckman, J. E. 1983. Hydrodynamic processes affecting benthic recruitment. Limnology and Oceanography. 28: 241 257.

PAGE 170

157 Eckman, J. E. 1987. Th e role of hydrodynamics in recruitment, growth, and survival of Argopecten irradians (L.) and Anomia simplex (D'Orbigny) within eelgrass meadows. Journal of Experimental Marine Biology and Ecology. 106: 165 191. Erftemeijer, P. L. A. and Middelburg, J. J. 1995. Mass balance constraints on nutrient cycling in tropical seagrass beds. Aquatic Botany. 50: 21 36. Eriksson, P. G. 2001. Interaction effects of flow velocity and oxygen metabolism on nitrification and denitrification in biofilms on submersed macrop hytes. Biogeochemistry. 55: 29 44. Finelli, C. M., Hart, D. D. and Fonseca, D. M. 1999. Evaluating the spatial resolution of an acoustic Doppler velocimeter and the consequences for measuring near bed flows. Limnology and Oceanography. 44: 1793 1801. Flo rida Marine Research Institute (FMRI) (1998). Benthic habitats of the Florida Keys. St. Petersburg. Florida Department of Environmental Protection. Fonseca, M. S. and Callahan, J. A. 1992. A preliminary evaluation of wave attenuation by four species of se agrass. Estuarine, Coastal and Shelf Science. 35: 565 576. Fonseca, M. S. and Fisher, J. S. 1986. A comparison of canopy friction and sediment movement between four species of seagrass with reference to their ecology and restoration. Marine Ecology Progre ss Series. 29: 15 22. Fonseca, M. S. and Kenworthy, J. 1987. Effects of current on photosynthesis and distribution of seagrass. Aquatic Botany. 27: 59 78. Fonseca, M. S. and Kenworthy, J. 1987. Effects of current on photosynthesis and distribution of sea grass. Aquatic Botany. 27: 59 78. Frankovich, T. A. and Fourqurean, J. W. 1997. Seagrass epiphyte loads along a nutrient availability gradient, Florida Bay, USA. Marine Ecology Progress Series. 159: 37 50. Fry, B. and Parker, P. L. 1979. Animal diet in T exas seagrass meadows: Delta13C evidence for the importance of benthic plants. Estuarine, Coastal and Shelf Science. 8: 499 509. Galvan, A., Cardenas, J. and Fernandez, E. 1992. Nitrate reductase regulates expression of nitrate uptake and nitrate reductas e activities in Chlamydomonas reinhardrii. Plant Physiology. 98: 422 426. Gambi, M. C., Nowell, A. R. M. and Jumars, P. A. 1990. Flume observations on flow dynamics in Zostera marina (eelgrass) beds. Marine Ecology Progress Series. 61: 159 169.

PAGE 171

158 Gardella, D. J. and Edmunds, P. J. 2001. The effect of flow and morphology on boundary layers in the scleractinians Dichocoenia stokesii (Milne Edwards and Haime) and Stephanocoenia michilini (Milne Edwards and Haime). Journal of Experimental Marine Biology and Eco logy. 256: 279 289. Gerard, V. A. 1982. In situ water motion and nutrient uptake by the giant kelp Macrocystis pyrifera Marine Biology. 69: 51 54. Ghisalberti, M. and Nepf, H. M. 2002. Mixing layers and coherent structures in vegetated aquatic flows. Jo urnal of Geophysical Research. 107: 1 11. Grant, W. D., Williams III, A. J. and Glenn, S. M. 1984. Bottom stress estimates and their prediction on the northern California continental shelf during CODE 1: The importance of wave current interaction. Journal of Physical Oceanography 14: 506 527. Hall, M., Durako, M., Fourqurean, J. and Zieman, J. 1999. Decadal Changes in Seagrass Distribution and Abundance in Florida Bay. Estuaries. 22: 445 459. Hamilton, S. K., Tank, J. L., Raikow, D. F., Wollheim, W. M., Peterson, B. J. and Webster, J. R. 2001. Nitrogen uptake and transformation in a midwestern U.S. stream: A stable isotope enrichment study. Biogeochemistry. 54: 297 340. Hansen, J. W., Pedersen, A. U., Berntsen, J., Ronbog, I. S., Hansen, L. S. and Lomste in, B. A. 2000. Photosynthesis, respiration, and nitrogen uptake by different compartments of a Zostera marina community. Aquatic botany. 66: 281 295. Harlin, M. M. 1973. Transfer of products between epiphytic marine algae and host plants. Journal of Phyc ology. 9: 243 248. Hearn, C. J., Atkinson, M. J. and Falter, J. L. 2001. A physical derivation of nutrient uptake rates in coral reefs: effects of roughness and waves. Coral Reefs. 20: 347 356. Hemminga, M. A., Harrison, P. G. and van Lent, F. 1991. The balance of nutrient losses and gains in seagrass meadows. Marine Ecology Progress Series. 71: 85 96. Hoch, M. P. and Kirchman, D. L. 1995. Ammonium uptake by heterotrophic bacteria in the Delaware estuary and adjacent coastal waters. Limnology and Oceanog raphy. 40: 886 897. Hurd, C. L., Ha rrison, P. J. and Druehl, L. D. 1996. Effect of seawater velocity on inorganic nitrogen uptake by morphologically distinct forms of Macrocystis integrifolia from wave sheltered and exposed sites. Marine Biology. 126: 205 214.

PAGE 172

159 Iizumi, H. and Hattori, A. 1982. Growth and organic production of eelgrass ( Zostera Marina L.) in temperate waters of the Pacific coast of Japan. III. The kinetics of nitrogen uptake. Aquatic Botany. 12: 245 256. Ikeda, S. and Kanazawa, M. 1996. Three dimensional organized vortices above flexible water plants. Journal of hydraulic engineering. 122: 634 640. Johnstone, I. M. 1979. Papua New Guinea seagrasses and aspects of the biology and growth of Enhalus acoroides (L.F.) Royle. Aquatic Botany. 7 : 197 208. Jorgensen, B. B., and Des Marais, D. J. 1990. The diffusive boundary layer of sediments: oxygen microgradients over a microbial mat. Limnology and Oceanography 35: 1343 1355. Kana, T. M., Sullivan, M. B., Cornwell, J. C. and Groszkowski, K. 19 98. Denitrification in estuarine sediments determined by membrane inlet mass spectrometry. Limnology and Oceanography. 43: 334 339. Karp Boss, L., Boss, E. and Jumars, P.A.. 1996. Nutrient fluxes to planktonic osmotrophs in the presence of fluid motion. I n A.D. Ansell, R.N. Gibson, and M.B. Barnes [eds.], Oceanography and Marine Biology: an Annual Review. UCL Press. 34: 71 107. Kaspar, H. F. 1983. Denitrification, nitrate reduction to ammonium, and inorganic nitrogen pools in intertidal sediments. Marine Biology. 74: 133 139. Kays, W. M. and Crawford, M. E. 1993. Convective heat and mass transfer, McGraw Hill. Koch, E. W. 1999. Sediment resuspension in a shallow Thalassia testudinum banks ex Konig bed. Aquatic Botany. 65: 269 280. Koch, E. W. 1993. The effect of water flow on photosynthetic processes of the alga Ulva lactuca L. Hydrobiologia. 260/261: 457 462. Koch, E. W. 1994. Hydrodynamics, diffusion boundary layers and photosynthesis of the seagrasses Thalassia testudinum and Cymodocea nodosa Mar. B iol. 118: 767 776. Koch, E. W. and Gust, G. 1999. Water flow in tide and wave dominated beds of the seagrass Thalassia testudinum Marine Ecology Progress Series. 184: 63 72. Koehl, M. A. R. and Alberte, R. S. 1988. Flow, flapping, and photosynthesis of Nereocystis luetkeana : a functional comparison of undulate and flat blade morphologies. Marine Biology. 99: 435 444.

PAGE 173

160 Kolmogorov, A. N. 1962. The local structure of turbulence in incompressible viscous fluid flow for very large Reynolds numbers. Proceedi ngs of the Royal Society of London A. 434: 9 13. Koop, K., D. Booth, Broadbents, A. and others. 2001. ENCORE: The effect of nutrient enrichment on coral reefs. Synthesis of results and conclusions. Marine Pollution Bulletin. 42: 91 120. Kuo, J. and McCo mb, A. J. 1989. Seagrass taxonomy, structure and development. p. 6 73. In A.W.D. Larkum [eds.], Biology of seagrasses: a treatise on the biology of seagrasses with special reference to the Australian region, Elsevier. Lapointe, B. E. and Clark, M. W. 199 2. Nutrient inputs from the watershed and coastal eutrophication of the Florida Keys. Estuaries. 15: 465 476. Larned, S. T. and Atkinson, M .J. 1997. Effects of water velocity on NH 4 and PO 4 uptake and nutrient limited growth in the macroalga Dictyosphaer ia cavernosa Marine Ecology Progress Series. 157: 295 302. Laws, E. 1984. Isotope dilution models and the mystery of the vanishing 15 N. Limnology and Oceanography. 29: 379 386. Lazier, J. R. N. and Mann, K. H. 1989. Turbulence and the diffusive layers around small organisms. Deep Sea Research. 36: 1721 1733. Lee, K. S. and Dunton, K. H. 1999. Inorganic nitrogen acquisition in the seagrass Thalassia testudinum : Development of a whole plant budget. Limnology and Oceanography. 44: 1204 1215. Li, Y. H. an d Gregory, S. 1974. Diffusion of ions in Sea Water and Deep Sea Sediments. Geochimica et Cosmochimica acta. 38: 703 714. Libes, M., 1986. Productivity Irradiance relationship of Posidonia oceanica and its epiphytes. Aquatic Botany. 26: 285 306. McRoy, C. P. and Goering, J. J. 1974. Nutrient transfer between the seagrass Zostera marina and its epiphytes. Nature. 248: 173 174. Moncreiff, C. A., Sullivan, M. J. and Daehnick, A. E. 1992. Primary production dynamics in seagrass beds of Mississippi Sound: the contributions of seagrass, epiphytic algae, sand microflora, and phytoplankton. Marine Ecology Progress Series. 87: 161 171. Neckles, H. A., Wetzel, R. L. and Orth, R. J. 1993. Relative effects of nutrient enrichment and grazing on epiphyte macrophyte ( Zo stera marina L.) dynamics. Oecologia. 93: 285 295.

PAGE 174

161 Nepf, H. M. and Vivoni, E. R. 2000. Flow structure in depth limited, vegetated flow. Journal of Geophysical Research. 105: 28,547 28,557. Neundorfer, J. V. and Kemp, W. M. 1993. Nitrogen versus phosphorus enrichment of brackish waters: responses of the submersed plant Potamogeton perfoliatus and its associated algal community. Marine Ecology Progress Series. 94: 71 82. Nikora, V. I., Suren, A. M., Brown, S. L. R. and Biggs, B. J. F. 1998. The effects of t he moss Fissidens rigidulus (Fissidentaceae: Musci) on near bed flow structure in an experimental cobble bed flume. Limnology and Oceanography. 43: 1321 1331. Orth, R. J., Luckenbach, M. and Moore, K. A. 1994. Seed dispersal in a marine macrophyte: Implic ations for colonizati on and restoration. Ecology. 75: 1927 1939. Patterson, M. R. 1992. A Chemical Engineering View of Cnidarian Symbioses. American Zoologist. 32: 566 582. Patterson, M. R. and Sebens, K. P. 1989. Forced convection modulates gas exchange in cnidarians. In Proceedings of the National Cacademy of Science vol. 86, pp. 8833 8836. USA. Patterson, M. R., Sebens K. P. and Olson, R. R. 1991. In situ measurements of flow effects on primary production and dark respiration in reef corals. Limnolo gy and Oceanography. 36: 936 948. Pedersen, M. F. and Borum, J. 1992. Nitrogen dynamics of eelgrass Zostera marina during a late summer period of high growth and low nutrient availability. Marine Ecology Progress Series. 80: 65 73. Pedersen, M. F., Palin g, E. I. and Walker, D. I. 1997. Nitrogen uptake and allocation in the seagrass Amphibolis antarctica Aquatic Botany. 56: 105 117. Pelton, D. K., Levine, S. N. and Braner, M. 1998. Measurements of phosporus uptake by macrophytes and epiphytes from the La Platte River (VT) using 32 P in stream microcosms. Freshwater Biology. 39: 285 299. Penhale, P. A. and Thayer, G. W. 1980. Uptake and transfer of carbon and phosphorus by eelgrass ( Zostera marina L. ) and its epiphytes. Journal of Experimental Marine Biolog y and Ecology. 42: 113 123. Raupach, M. R., Finnigan, J. J. and Brunet, Y. 1996. Coherent eddies and turbulence in vegetation canopies: The mixing layer analogy. Boundary Layer Meteorology. 78: 351 382.

PAGE 175

162 Riber, H. H. and Wetzel, R.G. 1987. Boundary layer and internal diffusion effects on phosphorus fluxes in lake periphyton. Limnology and Oceanography. 32: 1181 1194. Richardson, L. F. 1922. Weather prediction by numerical process. Cambridge: Cambridge University Press. Roth, N. C. and Pregnall, A. M. 19 88. Nitrate reductase activity in Zostera marina Marine Biology. 99: 457 463. Sand Jensen, K. 1977. Effect of epiphytes on eelgrass photosynthesis. Aquatic Botany. 3: 55 63. Sand Jensen, K. and Mebus, J. R. 1996. Fine scale patterns of water velocity wi thin macrophyte patches in streams. Oikos. 76: 169 180. Sand Jensen, K., Revsbach, N. P., and Jorgensen, B. B. 1985. Microprofiles of oxygen in epiphyte communities on submerged macrophytes. Marine Biology. 89: 55 62. Sanford, L. P. and Crawford, S. M.. 2000. Mass transfer versus kinetic control of uptake across solid water boundaries. Limnology and Oceanography. 45: 1180 1186. Shashar, N., Kinane, S., Jokiel, P. L. and Patterson, M. R. 1996. Hydromechanical Boundary Layers Over a Coral Reef. Journal of Experimental Marine Biology and Ecology. 199: 17 28. Short, F. T. and McRoy, C. P. 1984. Nitrogen uptake by leaves and roots of the seagrass Zostera marina L. Botanica Marina. 27: 547 555. Short, F. T. and Short, C. A. 1984. The seagrass filter: Purific ation of estuarine and coastal waters. The Estuary as a Filter. 395 409. Short, F. T., Burdick, D. M. and Kaldy, J. E. 1995. Mesocosm experiments quantify the effects of eutrophication on eelgrass, Zostera marina Limnology and Oceanography. 40: 740 749. Slawyk, G. and Raimbault, P. 1995. Simple procedure for simultaneous recovery of dissolved inorganic and organic nitrogen in 15N tracer experiments and improving the isotopic mass balance. Marine Ecology Progress Series. 124: 289 299. Smit, A. J. 2002. Nitrogen uptake by Gracilaria gracilis (Rhodophyta): Adaptations to a temporally variable nitrogen environment. Botanica Marina. 45: 196 209. Sokal, R. R. and Rohlf, F. J. 1995. Biometry. 3 rd edition. New York: W.H. Freeman and Company. 859 pp.

PAGE 176

163 Solorzano L. 1969. Determination of ammonium in natural waters by the phenolhypochlorite method. Limnology and Oceanography. 14: 799 801. Stapel, J., Aarts, T. L., van Duynhoven, B. H. M., de Groot, J. D., van den Hoogen, P. H. W. and Hemminga, M. A. 1996. Nutrie nt uptake by leaves and roots of the seagrass Thalassia hemprichii in the Spermonde Archipelago, Indonesia. Marine Ecology Progress Series. 134: 195 206. Strickland, J.D.H. and Parsons, T.R. 1968. A practical hand book of seawater analysis. Fisheries Rese arch Board of Canada, Ottawa, Canada. Taylor, D., Nixon, S., Granger, S. and Buckley, B. 1995. Nutrient limitation and the eutrophication of coastal lagoons. Marine Ecology Progress Series. 127: 235 244. Terrados, J. and Williams, S. L. 1997. Leaf versus root nitrogen uptake by the surfgrass Phyllospadix torreyi Marine Ecology Progress Series. 149: 267 277. Thomas, F. I. M. and Atkinson, M. J. 1997. Ammonium uptake by coral reefs: Effects of water velocity and surface roughness on mass transfer. Limnol ogy and Oceanography. 42: 81 88. Thomas, F. I. M., Cornelisen, C. D. and Zande, J. M. 2000. Effects of water velocity and canopy morphology on ammonium uptake by seagrass communities. Ecology. 81: 2704 2713. Thomas, F. I. M. and Cornelisen, C. D. 2003. A mmonium uptake by seagrass communities: Effects of oscillatory versus unidirectional flow. Marine Ecology Progress Series. 247: 51 57. Thursby, G. B. and Harlin, M. M. 1984. Leaf root interaction in the uptake of ammonia by Zostera marina. Marine Biology. 72: 109 112. Tomasko, D. A. and Lapointe, B. E. 1991. Productivity and biomass of Thalassia testudinum as related to water column nutrient availability and epiphyte levels: field observations and experimental studies. Marine Ecology Progress Series. 75: 9 17. Touchette, B. W. and Burkholder, J. M. 2000. Review of nitrogen and phosphorus metabolism in seagrasses. Journal of Experimental Marine Biology and Ecology. 250: 133 167. Turpin, D. H., Vanlerberghe, G. C., Amory, A. M., and Guy, R. D. 1991. The i norganic carbon requirements for nitrogen assimilation. Canadian Journal of Botany. 69: 1139 1145.

PAGE 177

164 Verduin, J. J. and Backhaus, J. O. 2000. Dynamics of plant flow interactions for the seagrass Amphibolis antarctica: Field observations and model simulatio ns. Estuarine, Coastal and Shelf Science. 50: 185 204. Vogel, S. V. 1994. Life in Moving Fluids: The Physical Biology of Flow, Princeton University Press. Wallentinus, I. 1984. Comparisons of nutrient uptake rates for Baltic macroalgae with different tha llus morphologies. Marine Biology. 80: 215 225. Wang, J. D. 1998. Subtidal flow patterns in Western Florida Bay. Estuarine, Coastal and Shelf Science. 46: 901 915. Wheeler, P. A., Glibert, P. M. and McCarthy, J. J. 1982. Ammonium uptake and incorporation by Chesapeake Bay phytoplankton: Short term uptake kinetics. Limnology and Oceanography. 27: 1113 1128. Wheeler, W. N. 1980. Effect of boundary layer transport on the fixation of carbon by the giant kelp Macrocystis pyrifera Marine Biology. 56: 103 110. Williams, S. L. and Ruckelshaus, M. H. 1993. Effects of nitrogen availability and herbivory on eelgrass ( Zostera marina ) and epiphytes. Ecology. 74: 904 918. Winning, M. A., Connolly, R. M., Loneragan, N. R. and Bunn, S. E. 1999. 15N enrichment as a met hod of separating the isotopic signatures of seagrass and its epiphytes for food web analysis. Marine Ecology Progress Series. 189: 289 294.

PAGE 178

165 APPENDI CES

PAGE 179

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.00 0.05 0.10 0.15 0.20 0.25 U (m s -1 ) z/h d 1 2 3 4 5 6 7 8 9 -0.05 0.00 0.05 V (m s -1 ) -0.05 0.00 0.05 W (m s -1 ) Figure 33 Velocity profiles collected during flume experiments for measuring uptake rates for ammonium. Profiles of mean velocity in the main flow (U), transverse (V), and vertical (W) directions are shown for heights above the bottom (Z) normali zed to deflected canopy height (h d ). Mean velocity at each height was based on the average of 300 measurements (Collected at 5 Hz for 1 minute with ADV). The horizontal line indicates the height of the canopy (Z/h d = 1). Values of U b U and Z o calcula ted from each of the profiles are provided Table 7. 166 Appendix A : Additional figures

PAGE 180

0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.00 0.05 0.10 0.15 0.20 U (m s -1 ) z/h d 1 3 4 5 7 8 -0.05 0.00 0.05 V (m s -1 ) -0.05 0.00 0.05 W (m s -1 ) Figure 34. Velocity profiles collected during flume experiments for measuring uptake rates for nitrate. Prof iles of mean velocity in the main flow (U), transverse (V), and vertical (W) directions are shown for heights above the bottom (Z) normalized to deflected canopy height (h d ). Mean velocity at each height was based on the average of 300 measurements (Colle cted at 5 Hz for 1 minute with ADV). The horizontal line indicates the height of the canopy (Z/h d = 1). Values of U b U and Z o calculated from each of the profiles are provided Table 7. 167 Appendix A (Continued)

PAGE 181

0.0 1.0 2.0 3.0 4.0 -10 10 30 50 -U'W' (m s -1 ) 2 x 10 5 z/h d 1 2 3 4 5 6 7 8 9 0.0 1.0 2.0 3.0 4.0 -10 10 30 50 -U'W' (m s -1 ) 2 x 10 5 z/h d 1 3 4 5 6 7 Fi gure 35. Vertical distribution of Reynolds stress ( ' W U ) estimated from velocity profiles collected during experiments for measuring uptake rates for (left) ammonium, and (right) nitrate. Depth above the bottom (Z) was normalized to deflected canopy height (h d ). Note the enhanced Reynolds stress at the canopy height (indicated by the line where Z/h d = 1 ) and greater penetration into the canopy during the higher velocity (U b ) experiments. Canopy heights were different between the two sets of experiments and therefore the distribution along the Y axis varies be tween the two graphs. 168 Appendix A (Continued)

PAGE 182

0.0 1.0 2.0 3.0 4.0 0 50 100 150 Total turbulent energy, K (m 2 s -2 ) x 10 5 z/h d 1 2 3 4 5 6 7 8 9 0.0 1.0 2.0 3.0 4.0 0.00 0.20 0.40 0.60 0.80 1.00 Relative turbulence intensity (K/ U 0.5 ) z/h d 1 2 3 4 5 6 7 8 9 Figure 36. Total turbulent energy (left) and relative turbulence intensity (right) estimated from profiles collected during flume experiments for measuring rates of ammon ium uptake. Depth above the bottom (Z) was normalized to deflected canopy height (h d ). The line represents the canopy height (Z/h d = 1). 169 Appendix A (Continued)

PAGE 183

0.0 1.0 2.0 3.0 0 20 40 60 80 K (m 2 s -3 ) x 10 5 z/h d 1 3 4 5 6 7 0.0 1.0 2.0 3.0 0.00 0.50 1.00 1.50 Relative turbulence intensity z/h d 1 3 4 5 6 7 Figure 37. Total turbulent energy (A) and relative tu rbulence intensity (B) estimated from profiles collected during flume experiments for measuring rates of nitrate uptake. Depth above the bottom (Z) was normalized to deflected canopy height (h d ). The line represents the canopy height (Z/h d = 1) 170 Appendix A (Continued)

PAGE 184

171 Appe ndix A (Continued) 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.00 0.05 0.10 0.15 0.20 U b (m s -1 ) U (m s -1 ) Ammonium Nitrate Figure 38. Shear velocity (U ) as a function of bulk velocity (U b ) for ammonium uptake experiments (solid symbols) and for nitrate uptake experiments (open symbols). Error bars represent the 95% con fidence limits using t he expression from Grant et al. (1984).

PAGE 185

ABOUT THE AUTHOR Christopher David Cornelisen was born and raised in Park Ridge, Illinois. In 1991, he co mpleted a bachelors degree in b iology and a minor in business from Drake Uni versity in Des Moines, Iowa Chris was a science instructor for a year at the Newfound Harbor Marine Institute on Big Pine Key, Florida, and for two years at the Shedd Aquarium in Chicago. In 1996, he graduated from Florida Institute of Technology (FIT) wi th an M S in Oceanography/Coastal Zone Management. His thesis focused on the effects of beach renourishment on physical attributes of sea turtle nesting beaches. After graduating from FIT, Chris was awarded a 2 year Coastal Management Fellowship from NOA As Coastal Services Center during which he conducted a regional assessment of coastal habitat and species restoration throughout the Gulf of Maine. In 1998 Chris began a Ph.D. program in Marine Science under the mentorship of Dr. Flo Thomas at Dauphin Is land Sea Lab. The following year, both Dr. Thomas and Chris transferred to the Department of Biology at USF, where he then completed his doctoral degree