<%BANNER%>

Effects of Flow on Filamentous Algae and Nutrient Limitation in Lotic Systems

Permanent Link: http://ufdc.ufl.edu/UFE0044557/00001

Material Information

Title: Effects of Flow on Filamentous Algae and Nutrient Limitation in Lotic Systems
Physical Description: 1 online resource (162 p.)
Language: english
Creator: King, Sean A
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: algae -- autotroph -- current -- drag -- filamentous -- flow -- limitation -- lyngbya -- metabolism -- nitrate -- nutrient -- river -- springs -- velocity
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This research examined the dual role of flow as a control on filamentous algae in lotic systems and its effect on nutrient limitation, with application to Florida spring-fed rivers.  The primary goal was to understand how flow influences both algal growth by regulating nutrient supply and algal abundance by dictating drag forces.  Four different approaches were used to address the research objectives: statistical analysis of North American stream datasets, a field survey and in situ experiment, laboratory stream channel experiments, and an ecological simulation model.  The statistical analysis of North American streams utilized the LINX (Lotic Intersite Nitrogen eXperiment) datasets, and resulted in the development of a new metric, the autotrophic uptake length.  The autotrophic uptake length accounts for nutrient concentration, discharge rate, and autotrophic metabolism and was a better predictor of nitrogen limitation than nitrogen concentration alone.  A field survey and experiment at the Gum Slough spring system determined that increased filamentous algal abundance was related to declining discharge, and identified a flow velocity threshold of 35 cm/s above which algal abundance was minimal.  The laboratory stream channel experiments tested the effect of flow velocity and nutrient concentration on the metabolism of the filamentous alga Lyngbya wollei, whose response was measured by diel changes in pH.  At low nitrate concentration, metabolism was stimulated by an increase in flow velocity from 1 to 5 cm/s, whereas at high nitrate concentration metabolic rates were similar at these two velocities, suggesting that the increase in velocity decreased nutrient limitation at low concentrations.  Algal metabolism was lower at higher velocity (10 cm/s) regardless of nitrate concentration.  Finally, a simulation model was created to help explain the results of the laboratory experiment.  The model indicated that filamentous algae and flow velocity display a subsidy-stress relationship when nutrients are limiting, but velocity is only a stress if nutrients are readily available.  In combination, these approaches demonstrate that flow has significant effects on algal growth, abundance, and autotrophic nutrient limitation in lotic systems.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sean A King.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Brown, Mark T.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044557:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044557/00001

Material Information

Title: Effects of Flow on Filamentous Algae and Nutrient Limitation in Lotic Systems
Physical Description: 1 online resource (162 p.)
Language: english
Creator: King, Sean A
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: algae -- autotroph -- current -- drag -- filamentous -- flow -- limitation -- lyngbya -- metabolism -- nitrate -- nutrient -- river -- springs -- velocity
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: This research examined the dual role of flow as a control on filamentous algae in lotic systems and its effect on nutrient limitation, with application to Florida spring-fed rivers.  The primary goal was to understand how flow influences both algal growth by regulating nutrient supply and algal abundance by dictating drag forces.  Four different approaches were used to address the research objectives: statistical analysis of North American stream datasets, a field survey and in situ experiment, laboratory stream channel experiments, and an ecological simulation model.  The statistical analysis of North American streams utilized the LINX (Lotic Intersite Nitrogen eXperiment) datasets, and resulted in the development of a new metric, the autotrophic uptake length.  The autotrophic uptake length accounts for nutrient concentration, discharge rate, and autotrophic metabolism and was a better predictor of nitrogen limitation than nitrogen concentration alone.  A field survey and experiment at the Gum Slough spring system determined that increased filamentous algal abundance was related to declining discharge, and identified a flow velocity threshold of 35 cm/s above which algal abundance was minimal.  The laboratory stream channel experiments tested the effect of flow velocity and nutrient concentration on the metabolism of the filamentous alga Lyngbya wollei, whose response was measured by diel changes in pH.  At low nitrate concentration, metabolism was stimulated by an increase in flow velocity from 1 to 5 cm/s, whereas at high nitrate concentration metabolic rates were similar at these two velocities, suggesting that the increase in velocity decreased nutrient limitation at low concentrations.  Algal metabolism was lower at higher velocity (10 cm/s) regardless of nitrate concentration.  Finally, a simulation model was created to help explain the results of the laboratory experiment.  The model indicated that filamentous algae and flow velocity display a subsidy-stress relationship when nutrients are limiting, but velocity is only a stress if nutrients are readily available.  In combination, these approaches demonstrate that flow has significant effects on algal growth, abundance, and autotrophic nutrient limitation in lotic systems.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sean A King.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Brown, Mark T.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044557:00001


This item has the following downloads:


Full Text

PAGE 1

1 EFFECTS OF FLOW ON F ILAMENTOUS ALGAE AND NUTRIENT LIMITATION IN LOTIC SYSTEMS By SEAN ALAN KING A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR T HE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Sean Alan King

PAGE 3

3 To my parents Alan and Lauren and my wife Shannon

PAGE 4

4 ACKNOWLEDGMENTS I give special thanks to Dr. Mark Brown for his support and guidance as my graduate advis or and to my committee members Drs. Matt Cohen, Tom Frazer, and Joe Delfino. Each of these individuals greatly shaped the course of my graduate studies and pushed me to take my research to the highest level. My experience collaborating with Matt Cohen an d Jim Heffernan on the autotrophic uptake length manuscript was invaluable. This research would not have been possible without financial support from the University of Florida Department of Environmental Engineering Sciences. I would also like to thank Judy Smith for providing access, support, and perspective for my research at the Gum Slough spring system. I am especially appreciative of the many people who helped me with my field and laboratory work including Jason Evans, Jenet Dooley, Valerie Burkett Heather Rothrock, Carrie Boyd, Chris Moody, Chris Cattau, Jim Surdick, Robert Compton, Rachel King, and Shannon McMorrow. Finally, I am grateful for the unwavering support from my family, particularly my best friend and wife Shannon whose encouragement a nd patience allowed me to complete this endeavor.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 OVERVIEW ................................ ................................ ................................ ............ 13 Statement of the Problem ................................ ................................ ....................... 13 Research Objectives ................................ ................................ ............................... 14 2 LITERATURE REVIEW ................................ ................................ .......................... 16 Nutrient Limitation in Aquatic Systems ................................ ................................ .... 16 Nutrient Limitation in Lotic Systems ................................ ................................ 17 Nutrient Spiraling ................................ ................................ .............................. 21 Effects of Flow on Lotic Algae ................................ ................................ ................. 22 Flow as a Subsidy to Algae ................................ ................................ .............. 23 Flow as a Stress to Algae ................................ ................................ ................. 26 The Subsidy Stress Relationship between Flow and Algae ............................. 28 Florida Springs ................................ ................................ ................................ ........ 29 Springs Ecology and Nutrient Limitation ................................ ........................... 29 Autotrophic Community Shifts ................................ ................................ .......... 32 Potential Causes o f Filamentous Algal Proliferation ................................ ......... 34 Flow Velocity Control of Filamentous Algae ................................ ..................... 38 Synthesis ................................ ................................ ................................ ................ 40 3 NUTRIENT FLUX, DEMAN D, AND AUTOTROPHIC L IMITATION IN LOTIC SYSTEMS ................................ ................................ ................................ ............... 44 Introduction ................................ ................................ ................................ ............. 44 Nutrient Limitation in Lotic Systems ................................ ................................ 45 Autotrophic Uptake Length ................................ ................................ ............... 49 Methods ................................ ................................ ................................ .................. 50 Data Sources ................................ ................................ ................................ .... 50 Uptake Length Calculations ................................ ................................ .............. 52 Statistical Analysis ................................ ................................ ............................ 52 Results ................................ ................................ ................................ .................... 53 LINX I Analysis ................................ ................................ ................................ 53 LINX II Analysis ................................ ................................ ................................ 54

PAGE 6

6 Global Distribution of S w,a ................................ ................................ ................. 56 Discussion ................................ ................................ ................................ .............. 57 Nutrient Limitation and Spiraling Metrics ................................ .......................... 57 Assessing Autotrophic Nutrient Demand and Limitation ................................ ... 59 Effects of Land Use and Region ................................ ................................ ....... 62 Nutrients an d Alternative Controls ................................ ................................ .... 63 4 FLOW VELOCITY CONTRO L OF FILAMENTOUS ALG AE IN A SUB TROPICAL SPRING FED RIVER ................................ ................................ ........... 72 Introduction ................................ ................................ ................................ ............. 72 Flow Velocity and Algae ................................ ................................ ................... 73 Factors Affecting Flow Velocity ................................ ................................ ........ 74 Sprin g fed Rivers in Florida ................................ ................................ .............. 75 Methods ................................ ................................ ................................ .................. 76 Site Description ................................ ................................ ................................ 76 Field Surveys ................................ ................................ ................................ .... 78 Experimental Manipulation of Flow Velocity ................................ ..................... 79 Statistical Analysis ................................ ................................ ............................ 81 Results ................................ ................................ ................................ .................... 81 Field Surveys ................................ ................................ ................................ .... 81 Experimental Manipulation of Flow Velocity ................................ ..................... 84 Discussion ................................ ................................ ................................ .............. 85 Flow Velocity and Filamentous Algae ................................ ............................... 85 The Relationship between SAV and Filamentous Alg ae ................................ .. 86 Management Implications ................................ ................................ ................. 87 5 EFFECTS OF FLOW VELO CITY AND NUTRIENTS O N THE FILAMENTOUS CYANOPHYTE LYNGBYA WOLLEI IN LABORA TORY STREAM CHANNELS .... 96 Introduction ................................ ................................ ................................ ............. 96 Positive Effects of Flow Velocity on Algae ................................ ........................ 96 Filamentous Algae ................................ ................................ ............................ 97 Methods ................................ ................................ ................................ .................. 99 Stream Channel Design ................................ ................................ ................... 99 Experimental Conditions ................................ ................................ ................. 101 Algal Collection and Processing ................................ ................................ ..... 103 Metabolism Measurements ................................ ................................ ............ 103 Statistical Analysis ................................ ................................ .......................... 106 Results ................................ ................................ ................................ .................. 106 CO 2 and DO Reaeration ................................ ................................ ................. 106 Algal Biomass and Metabolism ................................ ................................ ...... 107 Discussion ................................ ................................ ................................ ............ 109 Algal Metabolism Response to Flow Velo city and Nutrients ........................... 109 Algal Biomass Response to Flow Velocity and Nutrients ............................... 111 Determining Nutrient Response of Attached Algae ................................ ........ 113

PAGE 7

7 6 A SIMULATION MODEL F OR FLOW AND NUTRIENT EFFECTS ON FILAMENTOUS ALGAE AN D BACTERIA IN LABORA TORY STREAM CHANNELS ................................ ................................ ................................ .......... 126 Introd uction ................................ ................................ ................................ ........... 126 Methods ................................ ................................ ................................ ................ 127 Results ................................ ................................ ................................ .................. 129 Discussion ................................ ................................ ................................ ............ 130 7 CONCLUSION S ................................ ................................ ................................ ... 143 LIST OF REFERENCES ................................ ................................ ............................. 146 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 162

PAGE 8

8 LIST OF TABLES Table page 3 1 Global anal ysis datasets ................................ ................................ ..................... 70 3 2 Summary of the S w,a,N regression results. ................................ ........................... 71 4 1 Average water chemistry for the main spring at Gum Slough during the study period ................................ ................................ ................................ ................. 95 4 2 Results of multiple linear regression models for each transect, using maximum flow velocity and the day of the year to predict filamentous algal cover ................................ ................................ ................................ ................... 95 5 1 Chemistry of fresh tap wat er prior to nutrient addition ................................ ...... 124 5 2 CO 2 reaeration coefficients estimated from the five trials without algae. .......... 124 5 3 Algal biomass and metabolism results from the experimental trials comparing the effect of flow velocity at high and low nitrate concentration. ....................... 125 5 4 Algal biomass and metabolism results from the 19 day experiment with a series of nutrient additions. ................................ ................................ ............... 125 6 1 Sources and storages used as varia bles in the simulation model. ................... 140 6 2 Model equations and coefficients. ................................ ................................ .... 141 6 3 Comparison of observed experimental results and modeled results for metabolism of filamentous algae and bacteria in the six scenarios simulated. 142

PAGE 9

9 LI ST OF FIGURES Figure page 3 1 Comparison of autotrophic uptake lengths (S w,a ) and nutrient concentrations in relation to the NLA response ratios (RR x ) from the LINX I dat aset ................. 65 3 2 Comparison of autotrophic uptake lengths (S w ,a ) and nutrient concentrations in relation to the NLA response ratios (RR x ) from the LINX II dataset ................ 66 3 3 Comparison of autotrophic N uptake length (S w,a,N ) and DIN concentration for the three land use types in relation to RR N for individual str eams from the LINX II dataset ................................ ................................ ................................ .... 67 3 4 Comparison of autotrophic N uptake length (S w,a,N ) and DIN concentration in relation to the RR N from the regionally averaged LINX II dataset (A and B) and for the regionally averaged reference sites only (C and D) .......................... 68 3 5 Normal probability distributions for S w,a, N and S w,a,P of the three groups of datasets fit to their rel ative frequency of occu rrence ................................ .......... 69 4 1 Aerial view of Gum Slough ................................ ................................ ................. 90 4 2 --) for the period of record (October 2003 to December 2011) from the USGS gage do wnstream of the spring inputs ........... 91 4 3 Picture and diagra m of the experimental baffles ................................ ................. 91 4 4 Changes in flow velocity and filamentous algal cover over time at three transects with different morphology. ................................ ................................ ... 92 4 5 Filamentous algal cover in relation to average and maximum flow velocity for individual samplin g points across all transects ................................ ................... 93 4 6 Baffle experiment results ................................ ................................ .................... 94 5 1 Diagram of the experimental laboratory stream channels. ............................... 115 5 2 Initial pH change in fresh tap water at the three flow velocities due to CO 2 reaeration. ................................ ................................ ................................ ........ 116 5 3 CO 2 reaeration rate estimation based on total CO 2 change in fresh tap water from 9/19/11 to 9/22/11. ................................ ................................ ................... 117 5 4 Example of diel pH changes for each velocity from the trial beginning on 1/19/12. ................................ ................................ ................................ ............ 118 5 5 2 ], reaeration rate, and NEP for each velocity from a high nitrate trial. ................................ ................................ ................................ 119

PAGE 10

10 5 6 Boxplot comparison of CR and GPP results from the experiment testing flow velocity effects at low and high nitrate concentration. ................................ ....... 120 5 7 Nitrate concentration ch ange over time at high flow velocity for both low and high initial nitrate concentration. ................................ ................................ ....... 121 5 8 Diel pH changes in the 1 9 day experiment with sequential nutrient additions. 122 5 9 Diel NEP changes in the 19 day experiment with sequential nutrient additions. ................................ ................................ ................................ .......... 123 6 1 Systems diagram of the simulation model for nitrate and flow velocity effect s on algae and bacteria. ................................ ................................ ...................... 134 6 2 Response of algal production to flow velocity at two nitrate concentrations (1.0 and 0.0 5 mg N/L). ................................ ................................ ...................... 135 6 3 Relationship between nitrate uptake and nitrate concentration for bacteria and algae at different flow velocities. ................................ ................................ 135 6 4 Simulation model results of storages and metabolism at low initial nitrate concentration (0.05 mg N/ L) and 1 cm/s flow velocity. ................................ ..... 136 6 5 Simulation model results of storages and metabolism at low initial nitrate concentration (0.05 mg N/L) and 10 cm/s flow velocity. ................................ ... 137 6 6 Simulation model results of storages and metabolism at high initial nitrate concentration (1.0 mg N/L) and 1 cm/s flow velocity. ................................ ....... 138 6 7 Simulation model results of storages and metabolism at high initial nitrate concentration (1.0 mg N/L) and 10 cm/s flow velocity. ................................ ..... 139

PAGE 11

11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosoph y EFFECTS OF FLOW ON F ILAMENTOUS ALGAE AND NUTRIENT LIMITATION IN LOTIC SYSTEMS By Sean Alan King August 2012 Chair: Mark T. Brown Major: Environmental Engineering Sciences This research examined the dual role of flow as a control on filamentous algae in lotic systems and its effect on nutrient limitation, with application to Florida spring fed rivers. The primary goal was to understand how fl ow influences both algal growth by regulating nutrient supply and algal abundance by dictating drag forces. F our different appro aches were used to address the research objectives: statistical analysis of North American stream dataset s a field survey and in situ experiment, laboratory stream channel experiments, and an ecological simulation model. The statistica l analysis of North American streams utilized the LINX (Lotic Intersite Nitrogen eXperiment) datasets, and resulted in the development of a new metric, the autotrophic uptake length. The autotrophic uptake length accounts for nutrient concentration, disch arge rate, and autotrophic metabolism and was a better predictor of nitrogen limitation than nitrogen concentration alone. A field survey and experiment at the Gum Slough spring system determined that increased filamentous algal abundance was related to d eclining discharge and identified a flow velocity threshold of 3 5 cm/s above which algal abundance was minimal. The laboratory stream channel experiments tested the effect of flow velocity and nutrient concentration on the metabolism of the filamentous a lga

PAGE 12

12 Lyngbya wollei whose response was measured by diel changes in pH. At low nitrate concentration, metabolism was stimulated by an increase in flow velocity from 1 to 5 cm/s, whereas at high nitrate concentration metabolic rates were similar at these tw o velocities, suggesting that the increase in velocity decreased nutrient limitation at low concentrations. Algal metabolism was lower at higher velocity (10 cm/s) regardless of nitrate concent ration Finally, a simulation model was created to help expla in the results of the laboratory experiment. The model indicated that filamentous algae and flow velocity display a subsidy stress relationship when nutrients are limiting, but velocity is only a stress if nutrients are readily available. In combination, these approaches demonstrate that flow has significant effects on algal growth, abundance, and autotrophic nutrient limitation in lotic systems.

PAGE 13

13 CHAPTER 1 OVERVIEW Statement of the Problem Aquatic ecosystems throughout the world are experiencing dramati c changes as the result of anthropogenic activities. An increasingly common example is the proliferation of algae and subsequent shifts in ecosystem structure and function, which is often attributed to addition of nutrients that previously limited algal g rowth rates (Smith 2003). Florida springs and spring fed rivers provide a recent example of algal proliferation and ecosystem change, yet the cause is still unclear (Brown et al. 2008; Heffernan et al. 2010b). Elevated nitrate concentrations have been me asured in many springs (Scott et al. 2004), and therefore alleviation of nitrogen limitation has been suggested as the main driver of increased algal abundance. However, nutrient limitation was not detected in springs historically when nitrate concentrati ons were lower than at present (Odum 1957a; Duarte and Canfield 1990), and more recent studies of filamentous algae in springs also provide little evidence for the N enrichment hypothesis (Stevenson et al. 2004, 2007) In lotic systems (e.g. rivers and s treams), nutrient availability is positively related to flow velocity (Stevenson 1996); therefore nutrient limitation may only occur where both velocity and nutrient concentration are low enough to significantly reduce nutrient availability. Discharge rat es and flow velocities have been declining in most Florida springs over recent decades due to a combination of less rainfall and increased groundwater pumping (Copeland et al. 2009), which may increase the likelihood of nutrient limitation in low nutrient systems. However, flow velocity also dictates the drag force on algae and is the main determinant of both colonization and detachment (i.e.

PAGE 14

14 export) rates (Stevenson 1996). In Florida springs, flow velocity is likely a major control on filamentous algal a bundance, and declining flow velocities could reduce algal export and lead to algal proliferation. This research focuses on the dual role of flow on filamentous algae, as it acts both as a subsidy to algal growth by increasing nutrient availability and a stress on algal abundance by increasing drag force Research Objectives This research examines the dual role of flow as a control on filamentous algae in lotic systems and how this relates to nutrient limitation, with application to Florida spring fed rive rs. The primary goal was to understand how flow influences both algal growth by regulating nutrient supply and algal abundance by dictating drag forces. Much of this research focused on the filamentous alga Lyngbya wollei since it is one of the main alg al species proliferating in Florida spring fed rivers. The research objectives were to: Determine how flow influences autotrophic nutrient limitation in lotic systems broadly Identify flow velocity thresholds which inhibit algal proliferation in Florida s pring fed rivers Quantify the interactive effects of flow velocity and n utrient concentration on Lyngbya wollei Four different, yet comple mentary, approaches were used to address these objectives: statistical analysis of North American stream dataset s a f ield survey and in situ experiment, laboratory experiments, and a simulation model. The analysis of North American streams (Chapter 3) used the LINX (Lotic Intersite Nitrogen eXperiment) datasets to assess metrics for predicting nutrient limitation. Nutr ient concentration was compared to autotrophic uptake length (S w,a ) which is a new metric that takes into

PAGE 15

15 account nutrient concentration, discharge rate, and autotrophic nutrient demand to better characterize nutrient limitation in lotic systems. A field survey and experiment (Chapter 4) were conducted at the Gum Slough spring system in order to determine the effect of flow velocity on filamentous algae in situ. The field survey consisted of detailed flow velocity and algal cover measurements at three tra nsects every two months. The field experiment utilized plastic baffles to create high and low flow channels in which algal biomass was harvested on a periodic basis. Experimental stream channels (Chapter 5) were created in a greenhouse laboratory to test the effects of flow velocity and n utrient concentration on the metabolism and biomass of the filamentous alga Lyngbya wollei Three velocities were tested and the algal response was measured by diel changes in pH, which were converted into metabolic rate s. Finally, a simulation model (Chapter 6) was setup to explore the specific mechanisms that could explain the results of the laboratory experiment. In combination, these approaches generated multiple lines of evidence to fulfill the research objectives.

PAGE 16

16 CHAPTER 2 LITERATURE REVIEW Nutrient Limitation in Aquatic Systems Eutrophication of water bodies is a global issue originating from increased nutrient loading, mostly as a result of intensive agricultural and urban development (Carpent er et al. 1998). Eutrophication is a process which occurs when the addition of limiting nutrients causes increased autotrophic productivity due to stimulation of algae and other aquatic plant growth. Since there are often substantial negative effects of eutrophication on ecosystem structure and function, recreation, aesthetics, and even human health (Lembi 2003) this phenomenon has received much attention in the realm of environmental management (USEPA 1998; USEPA 2001). In response to eutrophication, m anagers often focus on reducing nutrient loading to the impacted systems by targeting the sources of nutrients, although there are a variety of other techniques such as directly removing algae and nutrients from the systems themselves through harvesting of biomass or sediment removal (Cooke et al. 1993; Lembi 2003). Nitrogen (N) and phosphorus (P) are the main nutrients of concern due to their low availability relative to their biotic demand; although other nutrients may also limit autotrophic productivity such as has been documented with diatoms and silica (Hecky and Kilham 1988). The particular form of nutrient is also important as dissolved inorganic forms of nitrogen (DIN) and phosphorus (DIP) are most readily utilized by autotrophs. The N:P ratio is often used as indicator of nutrient limitation status; ratios of 16:1 based on molar concentrations (the Redfield ratio for marine phytoplankton) are considered balanced The optimal N:P ratio for algae varies by species but generally ratio s > 20:1 sugge st P limitation and ratios < 10: 1 suggest N limitation with

PAGE 17

17 intermediate ratios possibly indicating co limitation by N and P ( S ch anz and Juon 1983 ). The N:P ratio is not always indicative of nutrient limitation as both nutrients may be available in suffi cient quantities to saturate autotrophic nutrient demand ( Snyder et al. 2002 ). Co limitation by N and P is a common phenomenon in which the addition of one nutrient subsequently causes limitation by the other, and thus addition of both can have substantia l effects. Regional factors such as geologic setting can exert strong influence on N and P availability and in some areas aquatic systems are naturally eutrophic (i.e. high nutrient levels and productivity). Based on early studies conducted in mostly t emperate climates it was concluded that typically freshwater systems are P limited whereas estuarine systems are N limited (Schindler 1977; Howarth 1988). However recent research shows that eutrophication is a complex process which warrants the considera tion of both N and P control for all types of aquatic systems (Conley 2000; Howarth and Marino 2006; Elser et al. 2007; Lewis and Wurtsbaugh 2008). Nutrient Limitation in Lotic Systems In lotic (i.e. flowing aquatic) systems, the effects of nutrient enri chment may genera lly be less dramatic in comparison to lentic systems. This is partially due to the influence of other factors that constrain autotrophs, such as light reduction by the riparian tree canopy and velocity induced export, but more importantly with regard to nutrient lim itation is the role of flow Stream discharge is inversely proportional to the hydraulic residence time of water, which determines the rate of nutrient replenishment in the water column and the type of algae that can persist. Regarding the later, there is a distinct difference between lotic systems with residence times long enough to allow free floating algae (i.e. phytoplankton) to persist and those with shorter residence times in

PAGE 18

18 which only algae attached to a substrate (i.e. periphyton) are viable (Hilton et al. 2006). In slow flowing lotic systems dominated by phytoplankton nutrient limitation may be stronger and the effects of nutrient enrichment may be more similar to lentic systems. In lotic systems dominated by attac hed algae nutrient limitation may only occur at relatively lo w nutrient concentrations since flow generates a continuous flux of nutrients which can saturate nutrient demand (sensu Odum 1957a). Francoeur (2001) conducted a meta analysis of 237 nutrient enrichment experiments in streams and determined that 43% did not show a significant response to either N or P additions. Six of ten streams assayed in Oklahoma (Ludwig et al. 2008) and 32% of stream assays in Idaho (Sanderson et al. 2009) did not indicat e nutrient limitation. Studies of streams across North America spanning a wide range of biomes found that in most cases algal production and biomass did not exhibit nutrient limitation (Tank and Dodds 2003; Johnson et al. 2009). A study of six streams in Washington did not find a relationship between DIP and algal biomass despite N:P > 30 (Welch et al. 1988). There are many lotic systems which have displayed algal nutrient limitation; however in each case nutrient levels were extremely low (DIN < 55 g /L: Triska et al. 1983; Grimm and Fisher 1986; Hill and Knight 1988; Rosemond et al. 1993; DIP < 6 g/L: Elwood et al. 1981; Peterson et al. 1983; Bothwell 1985; Pringle 1987; Bothwell 1988; Rosemond et al. 1993). Streams which did not indicate a lgal nutr ient limitation had higher nutrient levels (DIN > 138 g/L and DIP > 12 g/L: Pringle et al. 1986; von Schiller et al. 2007). There are other nutrients which may limit algal growth rates, particularly when N and P are sufficiently available, such as certa in micronutrients (e.g. Fe, Mn, Co) (Pringle et al. 1986). Light levels may also limit a lgal growth; therefore

PAGE 19

19 a lgal nutrient limitation is more prevalent in open canopy systems where light limitation is minimal (Grimm and Fisher 1986; Hill and Knight 19 88; Rosemond et al. 2000; Tank and Dodds 2003; Johnson et al. 2009). In summary, nutrient saturation is common in lotic systems and nutrient limitation is more likely in open canopy environments. The amount of algae may affect the nutrient concentration s required for nutrient limitation to occur. As algal biomass increases, nutrient transfer into the algal layer becomes increasingly diffusion limited and thus either higher nutrient concentrations or flow velocities may be necessary to inhibit nutrient l imitation (Bothwell 1989). In lotic systems with substantial algal biomass the threshold for limitation of peak biomass occurred at higher nutrient levels due to diffusion limitation (DIP < 25 g/L: Horner et al. 1983; Bothwell 1989; DIP < 38 g/L and DI N < 308 g/L: Rier and Stevenson 2006). Bothwell (1989) proposed that as nutrient levels increase there are three phases to algal nutrient limitation: phase 1 involves cellular uptake/growth rate saturation and displays Monod kinetics; phase 2 occurs as d iffusion limits nutrient transfer into the algal layer which constrains total algal biomass and is a linear process; and phase 3 displays nutrient saturation in which there is no further positive effect of nutrient increases on algal growth or biomass. Th is conceptual model explains why Horner et al. (1983) found rapid increases in biomass up to 25 g/L DIP (phase 1) above which the biomass increased at a much slower rate (phase 2). Therefore, while algal growth rates may be nutrient saturated at low conc entrations, the total amount of algal biomass may continue to respond to nutrient enrichment if nutrient transfer into the algal layer is limited by diffusion rates.

PAGE 20

20 A general issue regarding nutrient limitation is whether the focus is on the entire syst em, a community, or a species. Methods for detecting nutrient limitation usually only focus on one of these aspects. For example, short term nutrient enrichment of a stream reach measures the combined response of all organisms residing therein, whereas n utrient diffusing substrata assay the algal community that prefers the substrate type. Although these types of nutrient limitation assays may indicate that N or P is the primary limiting nutrient, results do not preclude the possibility that an individual species is limited by an alternative nutrient (Franceour 2001). Changes in both the total amount and relative abundance of nutrients often lead to altered algal community structure indicating that some species have strong preferenc es for particular nutri ent s (Stelzer and Lamberti 2001). One common occurrence is the dominance by cyanobacteria when N is limiting and N:P ratios are low, due to the ability of many cyanobacteria species to fix atmospheric N (Peterson and Grimm 1992; Schindler et al. 2008). Regardless of the degree of nutrient limitation, there may be other factors that exert strong control on algae, particularly in lotic systems. For example, l ocal flow velocities and the intensity and frequency of flood events can limit the amount of algal biomass in a given area (Biggs and Close 1989; Biggs 1996; Biggs et al. 1998). The type and amount of algal grazers such as snails and other invertebrates can also reduce algal biomass (Mulholland et al. 1991; Rosemond et al. 1993; Feminella and Hawkins 1995; Rosemond et al. 2000), through both consumptive and non consumptive losses (Scrimgeour et al. 1991). Biotic controls on algae may be complex, since alteration of the density of one species in an interconnected food web can lead to

PAGE 21

21 significant altera tions of the other species (Carpenter et al 1985). In shallow lakes in the United Kingdom, aquatic plants were ultimately controlled by fish densities rather than nutrients, since fish consumed invertebrates which consumed epiphytic algae which grew on t he plants (Jones and Sayer 2003). Nutrient Spiraling The concept of nutrient spiraling is the basis of a fundamental theoretical framework for analyzing nutrient cycling in lotic systems (Newbold et al. 1981). Nutrient spiraling describes how nutrients are cycled between inorganic and organic forms as they are displaced some distance downstream by the advective flow of water. Novel measurement methods created to test this theory involve adding nutrients to a stream, sometimes as isotopic or radioactive tracers, and measuring the decline in the added material with distance downstream. This rate of downstream decline can then be used to calculate the uptake length, S w which is the average distance downstream a dissolved inorganic nutrient molecule travel s before being removed from the water by biotic and abiotic processes Since S w is dependent on the size of the system, it is often converted to an uptake velocity, v f or areal uptake rate, U (SSW 1990). The uptake velocity is also known as the mass tra nsfer coefficient and is considered a measure of the biotic and abiotic demand for a nutrient relative to its water column concentration (Hall and Tank 2003). Nutrient spiraling theory describes nutrient limitation in lotic systems as both uptake length and uptake velocity are proportional to the degree of nutrient limitation. With decreasing n utrient limitation, nutrient molecules travel farther before being utilized and thus the uptake length increases and the uptake velocity decreases. This theory ha s been applied to a model including different functional groups of invertebrates which

PAGE 22

22 suggests that as nutrient availability varies, components of lotic systems adjust their cycling (i.e. uptake and remineralization) rates to compensate (Newbold et al. 19 82). This tendency of the biota to self organize to alleviate nutrient limitation, combined with the continuous replenishment of nutrients from upstream, indicates that nutrient limitation in lotic systems is most likely when nutrient levels are extremely low. Nutrient spiraling measurements have been used to characterize nutrient limitation and saturation levels in lotic systems. Two studies of individual streams used N additions to examine relationships between spiraling metrics and N concentration ba sed on Michaelis Menton kinetics (Earl et al. 2006; Covino et al. 2010 a ). Another study determined spiraling metrics for a suite of streams spanning a wide range of N concentrations and found an efficiency 2007 ). Since nutrient spiraling is currently the accepted framework for understanding nutrient cycling in lotic systems, nutrient spiraling metrics c ould be useful for setting nutrient criteria whose goal is to maintain ecosystem integrity. Effects of Flow on Lotic Algae The flow of water is the distinguishing characteristic of lotic systems. Within lotic ecosystems, both the magnitude and variation of flow largely determine the distribution of species and the rates of ecological processes. There are many as pects of flow and methods for its quantification; however the overall discharge rate and local flow velocities are the primary factors affecting lotic algae. Temporal changes in discharge alter water depths and velocities, and sharp increases may cause s couring events. With regard to nutrient dynamics, discharge is directly related to the downstream advective nutrient flux in the water column which controls the rate of nutrient replenishment. At a smaller scale, local flow velocities are a dominant dri ver of algal

PAGE 23

23 community structure as some species prefer a certain velocity range (Whitford 1956; McIntire 1968; Traaen and Lindstrom 1983; Ghosh and Gaur 1991). A compilation of evidence from studies of lotic systems suggests that generally a subsidy stre ss relationship exists between flow velocity and attached algae (Horner and Welch 1981; Stevenson 1996; Biggs et al. 1998); where velocity can aid algal growth by increasing nutrient uptake but also constrains algal abundance due to drag. These opposing influences make flow a major determinant of both algal growth rates and the ultimate amount of algal biomass that can persist in a given area of a lotic system. Flow as a Subsidy to Algae The subsidy effects of flow on attached algae in lotic systems lar gely involve enhancement of nutrient availability. There are two general mechanisms by which flow influences nutrient supply to algae: flow velocity regulation of diffusive boundary layer (DBL) thickness and the contribution of overall discharge to the ad vective flux of nutrients through the water column. These two processes act as controls on the rate of nutrient transfer from the water column to algal cells and the rate of nutrient replenishment in the water column, respectively. The magnitude of the s ubsidy effect is dependent on the degree to which algae are nutrient limited, as higher flow would not aid algal growth when nutrients are readily available (although see Mass et al. 2010). The DBL is the stagnant layer of water that surrounds all submer ged surfaces in aquatic systems, through which nutrients and other dissolved ions and gases must diffuse to reach cell surfaces. The mass transfer rate of nutrients due to molecular diffusion is proportional to the thickness of the DBL (i.e. the diffusion distance) and the concentration gradient (i.e. the difference in nutrient concentration on either side of the DBL). Larger algal species may protrude through the substratum DBL but nutrient

PAGE 24

24 transfer may remain diffusion limited due to the DBL surrounding the algal cells themselves (Larned et al. 2004). The thickness of a DBL is inversely related to the local flow velocity, and as a DBL becomes thinner with increasing flow velocity the transfer rate of nutrients from the water column to algal cells increa ses (Jorgensen and Revsbech 1985; Gundersen and Jorgensen 1990; Hurd 2000). For mat forming filamentous algae, higher velocities also increase diffusion of nutrients into the algal matrix (Stevenson and Glover 1993). As flow velocity increases, the DBL e ventually becomes thin enough to no longer limit nutrient transfer, and instead cellular uptake rates (i.e. kinetics) control transfer rates (Sanford and Crawford 2000; Nishihara and Ackerman 2009). When diffusive mass transfer does not limit nutrient upt ake, the concentration of nutrients in the water column may not be a good predictor of nutrient availability to algal cells since molecular diffusion is no longer the driving process. For lotic systems, the rate at which nutrients are replenished throughou t the water column is largely dependent on the discharge rate (advection) and turbulent mixing (dispersion). The advective nutrient flux is equivalent to the mass of nutrients flowing past a point per time and is derived from the product of discharge and nutrient concentration. Without inputs from groundwater or tributaries, nutrient concentrations typically decrease with distance downstream due to either biotic assimilation or dissimilation (e.g. denitrification). This rate of decline is inversely propo rtional to the advective nutrient flux, assuming that biotic demand remains constant. Therefore nutrient flux is related to nutrient supply to benthic algae, particularly when water column nutrient depletion is significant. Under conditions in which the water column is well mixed and diffusion through the DBL is not the limiting factor, the advective nutrient

PAGE 25

25 flux may be a better representation of nutrient availability to algal cells than the water column nutrient concentration. No studies have explicitl y focused on the role of flow in relation to the advective nutrient flux in the water column, although some authors have acknowledged the potential importance of this factor (Allan 1995; Borchardt 1996; Tank and Dodds 2003). There is abundant evidence that flow velocity can subsidize algal nutrient uptake and biomass. A positive relationship was documented in laboratory streams for flow velocity and algal nutrient uptake (Whitford and Schumacher 1961, 1964; Lock and John 1979; Horner et al. 1990; Larned et al. 2004) and algal biomass (McIntire 1966; Horner et al. 1983; Biggs and Hickey 1994). Welch et al. (1988) found a positive relationship between algal biomass and velocity (9 22 cm/s) in Washington streams. While most studies of the subsidy effect of flow velocity on algal growth focus on nutrient uptake an equally important mechanism may be velocity enhancement of waste efflux. There is evidence that velocity induced efflux of oxygen (a byproduct of photosynthesis) away from cells has a strong infl uence on autotrophic production rates (Mass et al. 2010). Some studies have found that the subsidy effect of flow on algae only occurred when nutrient concentrations were above a certain threshold (Horner and Welch 1981; Horner et al. 1983). This seems counterintuitive since the subsidy effect of flow should be greater when nutrient availability is less; however it is hypothesized that in some situations algae require sufficient nutrients to allow high enough growth rates to counteract the drag force ind uced by higher velocities (Horner and Welch 1981). Therefore a sufficient nutrient supply may be a prerequisite for some algae to become

PAGE 26

26 established in a high flow environment. Also, if internal recycling within the DBL is a significant nutrient source ( at low nutrient levels) then this recycling may be impeded with higher flow velocities resulting in lower algal growth rates. Indirectly flow velocity can subsidize algae by increasing the drag on algal grazers and thus reducing their consumption rates (De Nicola and McIntire 1991; Poff and Ward 1992; Poff and Ward 1995; Opsahl et al. 2003). This relationship may be complex since each grazer species may have an optimal velocity range for feeding or may prefer certain algal species that have their own optima l velocity range. Also in some cases grazers can actually aid algal growth by excreting nutrients and removing overlying or senescent cells (McCormick and Stevenson 1991). When increasing velocity causes algal grazing rates to decrease then this is anoth er mechanism that can lead to increased algal biomass, and may confound studies examining the direct relationship between flow velocity and algal biomass (e.g. Opsahl et al. 2003). Flow as a Stress to Algae The influence of flow as a stressor to attached algae is more straightforward since only one mechanism is typically involved. Essentially with increasing flow velocity there is also an increase in drag force on algal cells which both inhibits initial colonization and creates a sloughing effect whereb y algae are detached from their substrate and exported downstream (Stevenson 1996). Drag force increases at a rate proportional to the square of the flow velocity, so that small changes in velocity can potentially lead to significant changes in algal expo rt rates and community structure. Drag force has two components: viscous drag due to the shear force of water moving over a surface and form drag due to pressure variation over a surface (Larned 2010).

PAGE 27

27 The sloughing effect is highly dependent on the phy siognomic form of the algae as the various forms are suited to particular hydraulic conditions (Biggs and Thomsen 1995; Biggs et al. 1998). Smaller algal species that adhere closely to a substrate are less likely to slough and may be sheltered within the boundary layer of the substrate. Some of these small species have strong attachment mechanisms and are therefo r e suited to high flow velocity environments At the other end of the spectrum are some larger filamentous alga l species which typically do not have strong attachment mechanisms and often form thick mats that consist of layers of intertwined filaments. These larger filamentous species usually penetrate the substrate boundary layer where they are exposed to higher velocities and drag. These feat ures of mat forming filamentous algae generally make them more suitable for lower velocity environments Some studies have shown that algal biomass (Poff et al. 1990; Biggs and Gerbeaux 1993) and algal export rates (McIntire 1966; Biggs and Thomsen 1995) a re inversely related to flow velocity. High velocities may also reduce algal immigration and subsequent colonization rates (Stevenson 1983; Stevenson and Peterson 1989). One study did not detect increased export over a range of velocities ( Horner et al 1983 ), which may indicate that sudden increases in velocity are more important than the magnitude itself (Horner et al. 1990). At the ecosystem level, the amount of autotrophic production and biomass in a stream reach is strongly dependent on the intens ity of and time since the last flood event for streams which have high flow variability (Tett et al. 1978; Fisher et al. 1982; Biggs and Close 1989; Young and Huryn 1996; Biggs 2000), highlighting the temporal aspect of flow velocity control on algae. Flo w also affects the

PAGE 28

28 type of substrate present and since most algae prefer certain substrates this could have significant effects on algal community structure (Tett et al. 1978). Flow may also act as a stress unrelated to drag induced export in some situati ons although the mechanisms involved have not been determined Humphrey and Stevenson (1992) showed negative effects of flow velocity on algal growth rates in low nutrient conditions, but no effect when nutrients were high. Likewise, Borchardt et al (1 994) found flow velocity was negatively related to P uptake and algal production rates at high current velo cities, unrelated to sloughing Various hypotheses have been posi ted to explain this stress effect of flow, such as compression of algal filaments, alteration of the microenvironment surrounding cells stretching of cellular ion channels, and i ncreased efflux of nutrients from cells (Borchardt et al. 1994) The Subsidy Stress Relationship between Flow and Algae While the subsidy and stress mechanism s involving flow effects on algae are well documented, there are few studies that have specified the velocity range for both the subsidy and stress responses, and few optimal intermediate velocities have been reported A subsidy stress relationship has be en found between flow velocity and algal biomass in both natural and laboratory streams with a peak biomass in the range of 50 70 cm/s (Horner and Welch 1981; Horner et al. 1990; Biggs and Stokseth 1996). Biggs et al. (1998) found all three response types (subsidy, stress, and subsidy stress) when examining algal species with different growth forms in relation to flow regime. In one study a stress effect of flow velocity was found for certain species, but velocity was determined to be a subsidy effect for the total amount of algal biomass (McIntire 1968). Overall, the effects of flow on algae are difficult to predict due to the variety of

PAGE 29

29 mechanisms involved and the different responses of individual algal species, as well as interactions with other factor s such as nutrients, light, and grazers Florida Springs Although the theoretical principles examined in this research regarding effects of flow on filamentous algae and autotrophic nutrient limitation are applicable to lotic systems in general, the goal w as to apply this knowledge to artesian springs and spring fed rivers in Florida. Florida springs as a whole constitute ecosystem s at risk, as there have been significant changes in system structure and function over recent decades, particularly proliferat ion of filamentous algae ( e.g. Munch et al. 2006). Algal proliferation has largely been attributed to increasing nitrate concentrations as consistent with the eutrophication paradigm; however other factors have been suggest ed such as loss of algal consum ers, declining discharge rates, and increased human disturbance. The lack of clear evidence that nitrate enrichment is the primary cause of algal proliferation indicates that changes in other factors are likely contributing to the degradation of these sys tems (Heffernan et al. 2010 b ). Springs Ecology and Nutrient Limitation fed rivers have a long history of human use; however their ecology did not receive much attention until they became the focus of research in the 1950s. Florida springs were ideal settings to test innovative methods for measuring whole ecos ystem structure and metabolism (Odum 1957a, 1957b). These studies provided a wealth of knowledge, including the determination that these springs are amon g the most productive ecosystems in the world due to their extremely clear water and relatively stable temperature, chemistry, and hydrology. Odum (1957a) also demonstrated that these spring ecosystems were highly tuned to light levels with

PAGE 30

30 no evidence o limit production under transient conditions are not important in steady state seems reasonable since the large and continuous flows are continually renewing the aquatic medium so tha At this same time the algal community was studied in detail and it was found that flow velocity was a primary control on algal distribution patterns (Whitford 1956). Decades later was replicated by one of his students (Knight 1980) and no ecological changes were detected except for a decline in fish biomass which was attributed to construction of the Rodman Dam downstream. By the 1990s iconic springs such as Ichetucknee, Wakulla, Wekiwa, and Silver were showing signs of degradation, mainly in the form of increasing filamentous algae. These issues led to the formation of the Florida Springs Task Force, a multi agency group charged discharge rates were highlighted as potential causes of ecological degradation. At the turn of the century two major ecological stu dies were conducted: a third follow up to the Silver Springs studies (Munch et al. 2006) and a comprehensive study of the relationship between nutrients and algal proliferation (Stevenson et al. 2004, 2007). The Silver Springs study found a further declin e in fish biomass and substantially more epiphytic algae and benthic algal mat biomass than in the 1950s. Ecosystem production and respiration rates had also declined as had discharge rates, and nitrate concentrations exceeded 1 mg N /L. The comprehensive study of algae and nutrients involved many parts including a survey of algal cover and water quality metrics

PAGE 31

31 in 29 springs and various laboratory experiments testing the response of common filamentous algal species to different nutrient concentration s. T he survey determined that Vaucheria was positively related to both N and P concentration s, but the other common alga l species Lyngbya did not correlate with either nutrient Among the springs studied there was no indication of the expected positive rela tionship between total algal cover and nitrate concentration, as there were multiple springs with high nitrate and low algae levels and vice versa. The experimental results were highly variable, as some found a relationship between filamentous algae and n utrient concentration and others did not. Multiple other studies have focused on drivers of the autotrophic community, particularly filamentous algae; however much uncertainty remains regarding the role of nutrients. A survey of 31 spring fed rivers did not find a relationship between either autotrophic biomass or productivity and nutrient concentration but did find an inverse relationship with canopy cover suggesting that these systems were limite d by light rather than nutrient s (Duarte and Canfield 19 90). An experiment involving nutrient additions to in situ tubes in the spring fed Chassahowitzka River found that algal growth on glass slides was stimulated by P but not N, suggesting P limitation (Notestein et al. 2003). A laboratory experiment tested various nitrate concentrations on the growth of Lyngbya wollei and found a substantial increase in growth rates from 0.3 to 0.6 mg N/L (Cowell and Dawes 2004), contrary to other studies (e.g. Stevenson et al. 2004, 2007). One explanation for the lack of correspondence between the results of these studies is the difference in the flow environment The only experiments with flow representative of lotic systems were those which used semi recirculating channels with

PAGE 32

32 a moderate flow velocity of 25 cm/s (Stev enson et al. 2004, 2007). For these particular experiments, most did not detect nutrient limitation for either Vaucheria or Lyngbya ; however one found that Lyngbya growth rates were saturated at a nitrate concentration of 0.11 mg N/L. In comparison to s tudies which found higher thresholds for N saturation, these results may indicate that higher flow velocities can reduce the nutrient concentrations required to achieve saturation. High rates of water exchange utilized in other experiments likely did not compensate for effects of flow velocity on nutrient transfer through boundary layers (and potentially other effects), and therefore higher nutrient concentration s were required to alleviate nutrient limitation. Autotrophic Community Shifts The autotrophic (i.e. photosynthetic) community in Florida springs wa s characterized by dense beds of submerged aquatic vegetation ( SAV ) primarily the grasses Sagittaria kurziana and Vallisneria americana and a diverse assemblage of epiphytic and benthic algae (Whitford 1956). More recently large filamentous algal mats have proliferated in many springs, and are likely competing with the epiphytic algae and SAV that define d the historic autotrophic community. Severe algal proliferation may cause shifts from SAV to alga l dominance due to competition for light and nutrients ( Duarte 1995; Hilton et al. 2006 ), which may explain the observed decline in SAV in some springs. Alternatively, SAV loss may be mostly due to increased uprooting resulting in part, from either nutri ent enrichment which can cause rooted SAV to develop longer leaves and less below ground material (Nixon et al. 2001) or shifts in sediment composition to finer particles with less cohesion (personal observation). Human activities such as wading and boati ng are also a direct cause of SAV loss and could exacerbate uprooting.

PAGE 33

33 A recent survey of Florida springs identified 23 species of macroalgae, of which Lyngbya wollei and Vaucheria spp. were the two most common (Stevenson et al. 2007). These species were present in the past in smaller amounts, but now cover 50% of the benthic surface on average (Stevenson et al. 2007) and have surpassed SAV biomass in some systems ( e.g. Munch et al. 2006). Although the classification of these species has varied over time, it appears that they are native, rather than invasive exotics, since they were found in Silver Springs in 1950s ( Whitford 1956 ); therefore a significant alteration in their environmental controls must have occurred to allow their proliferation. Lyngbya spp. is a particularly problematic alga which has received considerable attention globally over recent decades due to its ability to form expansive, resilient mats and its potential to produce toxins (Carmichael et al. 1997; Ahern et al. 2006). Lyngby a wo llei is a filamentous cyanobacterium with multiple characteristics that give it the potential to proliferate in aquatic systems: adaptation to low light levels (Speziale et al. 1991), ability to fix atmospheric nitrogen (Phlips et al. 1992), and a thick sh eath that serves to inhibit grazing (Camacho and Thacker 2006). Lyngbya wollei has been the focus of management efforts throughout the southeastern U.S., particularly in shallow lakes and reservoirs (Speziale et al. 1991; Phlips et al. 1992; Doyle and Sma rt 1998). Lyngbya wollei within Florida springs may actually consist of two or more species based on molecular and morphological data, and each species may have different water chemistry preferences (Joyner et al 2008). Despite the potential for some a lgal species to proliferate, algae are an important component of Florida springs ecosystems. Even problematic filamentous algae such as Lyngbya provide habitat for a large amount and diversity of invertebrates, especially

PAGE 34

34 amphipods and gastropods (persona l observation), and are likely significant contributors to chemical cycling through alteration of oxygen, carbon, and nutrient levels. Large scale eradication techniques such as intensive herbicide application and use of mechanical harvesters should be gi ven careful consideration as they may lead to further ecological degradation Potential Causes of Filamentous Algal Proliferation Florida springs are changing in a variety of ways, which makes it difficult to determine the specific causes and effects among drivers and components of the ecosystem. Overall there is little historical ecological data available with which to examine changes over time, with Silver Springs having the best long term record. In fact, much of the evidence for algal proliferation i s based on anecdotes and comparisons with historical photographs; however the current amount of filamentous algae in many springs appears to be unprecedented as are the changes in other ecological attributes (e.g. declines in SAV and fish biomass ) The a vailable long term data consists mostly of discharge rates and water chemistry parameters, which generally show that discharge rates are declining and nitrate concentration s are increasing Since the available data do not indicate a single definitive caus e of filamentous algal proliferation in Florida spring systems, all potential factors should be given consideration including increased nitrate concentrations, loss of algal consumers, decreased discharge increased human recreation, and other direct distu rbances such as aquatic plant management (Heffernan et al. 2010b) In many aquatic systems, nutrient enrichment is the primary cause of excessive aquatic plant and algal growth. Nutrients (primarily N and P ) often limit algal growth because of their relat ively low availability in relation to that required by algae; however

PAGE 35

35 once nutrients become readily available then algal growth is nutrient saturated and further increases in nutrients will have no additional effect on algal growth. For the main filament ous algal species proliferating in Florida springs the saturating N concentrations were estimated to be 0.59 mg/L and 0.25 mg/L for Vaucheria and Lyngbya respectively (Stevenson et al. 2007). It would take reductions in N below the saturating concentrat ions to begin constraining the growth of these algal species, and possibly much more to achieve a significant reduction in algal biomass. Historically N concentrations in springs were likely 0.05 0.1 mg/L and today the average concentration is > 1 mg/L (Heffernan et al. 2010b) ; therefore it is feasible that N enrichment has increased algal growth rates. However there are multiple studies which have not found a strong relationship between nutrient concentrations and autotrophic biomass and productivit y (including filamentous algae) within Florida spring systems (Odum 1957a; Duarte and Canfield 1990; Cowell and Botts 1994; Terrell and Canfield 1996) and the overall narrative that N enrichment is the primary cause of algal proliferation has been question ed (Heffernan et al. 2010 b ). Saturation concentrations for P have also been estimated (26 g/L and 33 g/L for Vaucheria and Lyngbya respectively (Stevenson et al. 2007)); however P enrichment in springs is uncommon since P is readily retained in the ca lcium carbonate matrix of the aquifer. An alternative mechanism which could lead to filamentous algal proliferation is the loss of organisms that consume algae (i.e. algal grazers) such as certain snails and other invertebrates as well as some fish species The magnitude of consumer loss and their effect on algae is currently under investigation; however there is preliminary evidence that this could be a substantial cause of algal proliferation (Heffernan et al.

PAGE 36

36 2010 b ). Consumer loss may be primarily due to declines in the dissolved oxygen (DO) content of spring water, which stresses the organisms and can lead to reduced algal consumption rates. Other potential drivers of reduced algal consumption could be loss of habitat such as SAV beds, and the presen ce of toxins such as pesticides and other man made chemicals At high enough concentrations nitrate may also act as a toxin to certain aquatic organisms (> 2 mg N/L: Camargo et al. 2005 ), and these concentrations have been observed in some springs and ma y occur in many others in the near future. Declining discharge is often mentioned as a cause of springs degradation; however this is usually in reference to substantial declines in flow that lead to complete loss of a spring such as has occurred for Kensi ngton, White, and Worthington Springs. The effect o f smaller declines in discharge on algal control mechanisms has not received much attention. This may partially be due to the difficulty in determining whether the discharge for a given spring is actuall y declining, as there is inherent variation on a monthly to annual timescale. This variation is often correlated with recent rainfall in the spring recharge area; however there are long term declines in some springs which could be due to consumptive use of groundwater by humans (Weber and Perry 2006; Williams 2006). Alteration of the landscape in a way that lowers the permeability of the land surface would also lead to less recharge to the aquifer. The potential for groundwater pumping to affect aquifer levels regionally is only now becoming evident, as it has been shown that the groundwater divide in north central Florida has shifted substantially over recent decades due to pumping in the Jacksonville area (Grubbs 2011).

PAGE 37

37 Declining discharge rates could be involved in multiple mechanisms which result in filamentous algal proliferation including less algal export, decreased DO levels, and increased nutrient concentration s. In most spring systems, discharge rates are proportional to flow velocities through out the channel. Since flow velocity is related to drag on filamentous algae, declining discharge rates could lead to less export and increased algal biomass. Lower DO levels observed in many springs could be related to lower discharge rates due to older deeper groundwater making up a greater proportion of spring discharge ( Katz et al. 1999 ), with subsequent reduction of algal consumption. In some springs, nitrate concentrations are inversely related to discharge (Katz et al. 1999). Thus, declining dis charge may increase the potential for nutrient enrichment and stimulation of algal growth rates Some human activities directly disturb the ecosystem including increased recreational use, channel alterations, and aquatic plant management. Springs are enjo yed by many people for recreational purposes such as swimming, wading, tubing, snorkeling, SCUBA diving, paddling, boating, and fishing. While these activities likely have minimal effect s on spring ecosystems when few people are involved, a large number o f users can lead to significant impacts. Recreational impacts have received little attention by scientists; however one study of the Ichetucknee River found that SAV damage and export increased with the amount of people tubing and SCUBA diving (DuToit 19 79). As a result of this study the amount of tubers and divers on the Ichetucknee River are now regulated to minimize these impacts, but degradation related to recreation is still a problem for this spring system and many others.

PAGE 38

38 Other impacts related t o human activities are the result of channel alterations such as damming and dredging and large scale aquatic plant management. Where these alterations lower flow velocities, conditions may favor filamentous algal proliferation. In some cases migration o f manatees and fish such as mullet and catfish are also impeded, and the loss of their top down control on spring food webs could have significant ramifications. Aquatic plant management is potentially a major disturbance to spring ecosystems where intens ive methods such as widespread herbicide application and large mechanical harvesters are used. Exotic invasive plants such as hydrilla and water hyacinth became problematic in some spring systems, notably Kings Bay and Wakulla Springs, resulting in manage ment plans for their eradication. In most cases this involved substantial application of various herbicides, which were largely unsuccessful at eliminating the target plant species and may have further contributed to ecosystem degradation. In Kings Bay, many of the nuisance plants eventually declined (mostly unrelated to management efforts) and were replaced by filamentous algae, primarily Lyngbya (Evans et al. 2007). When chemical application failed to control the algae, then mechanical harvesters were used to directly remove algae from the bottom of the bay, a practice which continues today. While control of invasive exotic plants is a laudable goal, the consequences of these efforts may actually promote filamentous algal proliferation by increasing al gal dispersal and removing competitors for light and nutrients Flow Velocity Control of Filamentous Algae There is some evidence that the drag created by flow velocity is a significant control on filamentous algal abundance in Florida spring systems. The re are observations that filamentous algal biomass is higher in low velocity areas such as

PAGE 39

39 topographic depressions (Munch et al. 2006; Quinlan et al. 2008), as well as along channel margins and in the vicinity of spring boils where velocities are typically lower (personal observation). SAV and other stable substrates may be necessary to support substantial filamentous algal biomass in higher velocity areas because the algae can wrap around plant stems in order to stay in place and SAV beds may provide lowe r velocity areas where benthic algal mats may form. In Florida springs the effect of flow velocity on filamentous algae has not received much attention; however a study of three coastal systems found that algal biomass was minimal where velocities were > 25 cm/s (Hoyer et al. 2004). The available evidence suggests a negative (i.e. stress) relationship between flow velocity and filamentous algae in springs. Overall, this relationship may explain the spatial heterogeneity of algal biomass within a given spring more than it explains widespread algal proliferation. However, since flow velocity can have a large effect on algal export a decline in velocity over time could be a mechanism for increased algal biomass in lotic system s. Although declining spri ng discharge is not necessarily related to declining flow velocity in a given area, substantially lower discharge would have less kinetic energy and therefore velocities would be lower overall throughout the system. The influence of SAV beds on local flow velocities may be important due to their role in constraining the channel area, which causes faster velocities above the beds and lower velocities within them. Additionally, since flow velocity is positively related to nutrient availability, lower veloci ties could increase the likelihood of nutrient limitation where nutrient concentrations are low enough.

PAGE 40

40 Synthesis This research was inspired by the need to delve deeper into the causes of filamentous algal proliferation in Florida springs. There were appa rent contradictions between available evidence and the narrative that nitrate enrichment and the alleviation of algal N limitation (i.e. eutrophication) was the predominant cause. Lotic systems are fundamentally different than other aquatic systems and pa radigms from these other systems should be applied with caution in the absence of sufficient evidence demonstrating their effectiveness. The role of flow as a primary control on nutrient availability should be considered when determining whether algae and other autotrophs are nutrient limited. Nutrient concentration may not be the best descriptor of nutrient availability in lotic systems, and instead nutrient flux should be assessed as a metric of nutrient limitation. In addition local flow velocities a lso dictate whether algae can even begin to colonize an area due to drag forces, and are a strong determinant of the ultimate amount of algal biomass that can persist in an area. The dual rol e of flow as a control on algal abundance in lotic systems is an example of a subsidy stress relationship, since some amount of flow positively affects the algae until a threshold is reached past which higher flow is increasingly a negative influence. There are many other factors that are also important when consideri ng what controls algal abundance in a specific area: such as the amount of light, nutrients, algal consumers, the type of substrate, and the specific response of each algal species to all of these factors. However the dominant role of flow must be account ed for in order to determine whether nutrients are a limiting factor and whether drag forces are a significant constraint on algal abundance in a lotic system.

PAGE 41

41 Lo tic systems are expected to respond differently to nutrient enrichment than lentic systems, si nce in the latter f low is not a dominant factor and algal communities often consist primarily of free floating phytoplankton rather than attached algae. In lentic systems the major controls on algal biomass are nutrients, consumers, and light availability Algal consumers may exert significant control over algal biomass and autotrophic community shifts (Carpenter et al. 1985; Jones and Sayer 2003), but their influence may become less significant with increasing nutrient levels. Therefore, when nutrients are readily available algal biomass may accumulate until light penetration into the water column becomes the limiting factor for further growth. However in lotic systems with sufficient flow it is unlikely that nutrients will be severely limiting to atta ched algae, so there must be other controls that prevent algae from covering the entire surface area Local flow velocities and algal consumers constrain the amount of attached algae that can persist in an area, and therefore factors which reduce the stre ngth of these mechanisms could lead to algal proliferation. The influence of nutrients on the ultimate (i.e. steady state) amount of algal biomass in lotic systems remains uncertain. During algal succession, there is often an accrual phase that leads t o peak biomass, which may be followed by loss of biomass due to sloughing that eventually levels off at an ultimate biomass (e.g. Rier and Stevenson 2006). Grimm and Fisher (1986) hypothesized that nutrients should not control the ultimate amount of algal biomass in streams, since eventually any necessary nutrients will flow by and become available. Therefore under conditions of nutrient limitation it may take longer for algae to reach the ultimate biomass due to lower growth rates, but over time the ulti mate biomass would be equivalent regardless of nutrient

PAGE 42

42 concentration. There is some support for this conceptual model as the ultimate biomass was similar for different nutrient concentrations in stream mesocosms, although peak biomass was correlated with nutrient concentration (Rier and Stevenson 2006). T his model however, omits the direct effects of flow velocity which can have both positive and negative effects on the ultimate algal biomass. Higher velocity may result in greater algal biomass (McInti re 1966) or it can prevent algal biomass from achieving the same level as in a lower velocity (Horner and Welch 1981; Poff et al. 1990; Peterson and Stevenson 1990). There are also complex temporal dynamics as peak biomass may increase with velocity but subsequent drag may reduce the ultimate biomass to a lower level than would occur in a lower velocity environment (Biggs and Stokseth 1996). While differences in nutrient concentrations may not substantially affect the ultimate amount of algal biomass in lotic systems, differences in flow velocity can lead to a variety of possible outcomes resulting from the opposing effects of nutrient availability and drag during algal colonization and succession. A more advanced conceptual model separates the subsidy and stress roles of flow. The subsidy effect on nutrient supply increases algal growth rates and peak biomass, but the stress effect of drag determines the ultimate amount of algal biomass in conjunction with nutrient concentration This model is similar to the habitat suitability respectively. If a portion of an algal population is continuously exported downst ream due to drag forces, then a species with a higher growth rate should be able to achieve higher ultimate biomass through faster replacement of exported material. Algal growth rates may be

PAGE 43

43 stimulated by increased nutrient availability due to either incr eased nutrient concentration or velocity, which could shift the balance between growth rates and export rates. This model highlights the importance of factors that affect both growth and export rates and therefore suggests a mass balance approach would be beneficial for studying algal dynamics in lotic systems (e.g. Hecky and Kilham 1989). In order to apply a mass balance approach to filamentous algal proliferation in Florida springs, the influence of flow velocity on algal growth and export rates must b e quantified. In all but the lowest nutrient springs, nutrient availability is likely high enough for algae to grow at maximum rates (i.e. nutrient saturation); however declining discharge may lead to lower velocities and less algal export which could be a significant cause of algal proliferation. Lower velocity areas within spring systems, such as near spring boils, may contain favorable conditions for algal proliferation and are where algae are most likely to respond to nutrient enrichment. Determining the response of filamentous algae, such as Lyngbya wollei to a range of n utrient concentrations and flow velocities will improve understanding of the causes of algal proliferation and nutrient limitation in Florida springs and other lotic systems.

PAGE 44

44 CHAP TER 3 NUTRIENT FLUX, DEMAN D, AND AUTOTROPHIC LIMITATION IN LOTIC SYSTEMS Introduction Nutrient enrichment is often identified as the primary cause of changes in aquatic ecosystem productivity and community composition (Carpenter et al 1998 ; Elser et al 2 007) Prior to enrichment the availability of a nutrient limits autotrophic production, and alleviation of nutrient limitation permits additional production, with a suite of potentially undesirable consequences. Typically, nutrient availability is inferr ed from water column concentration, with limitation assessed based on the concentration alone or from the ratio with other potentially limiting nutrients. This approach has been useful for diagnosing nutrient limitation in lakes, estuaries and reservoirs, but has been less successful for lotic systems (i.e. rivers and streams), where nutrient concentrations alone fail to reliably predict autotrophic biomass ( Dodds et al 2002 b). Several plausible reasons might account for the differential utility of conc entration as a predictor of nutrient limitation in lentic and lotic systems. First, compared to lakes, lotic systems are more likely to be limited by factors other than nutrients such as flood disturbances and riparian shading (Biggs 2000; Rosemond et al. 2000). Perhaps more important, however, are the differences between lakes and streams with regard to hydraulic residence time. The initial development of models for predicting nutrient limitation in lakes explicitly considered system hydraulics (i.e. fl ushing rates; Vollenweider 1982). Despite substantially shorter residence times in streams than in lakes, hydraulics are generally not considered in models of stream and river eutrophication ( Hilton et al 2006 ). This study argues that the hydraulic prop erties of lotic systems limit the utility of concentration as a predictor of autotrophic nutrient

PAGE 45

45 limitation, specifically because concentration alone does not take into account continuous nutrient resupply from advection and (re)mineralization. Despite the centrality of nutrient spiraling theory (Newbold et al. 1982) to stream ecology and its explicit consideration of system hydraulics; metrics derived from that body of theory have yet to be incorporated broadly into predictions of lotic nutrient limitat ion. This study explores a new metric of lotic ecosystem nutrient limitation called the autotrophic uptake length ( S w,a ). S w,a is based, in part, on nutrient flux, and thus takes residence time into account to better represent nutrient availability to th e biota. The metric further considers whether the nutrient flux is sufficient to saturate autotrophic nutrient demand (which may be limited by light or post flood successional status); since where fluxes are in excess of demand, nutrient enrichment may ha ve little effect, even if concentrations are relatively low. Nutrient Limitation in Lotic Systems Evidence abounds that nutrient concentrations predict autotrophic productivity and biomass more strongly for lentic than lotic systems. Dodds et al. (2002b ) showed that total nitrogen (N) and total phosphorus (P) combined explained less than 40% of the variation in chlorophyll a for benthic algae in streams, whereas these variables explain more than 60% of the variation in planktonic algae in lakes. Likewis e, ammonium uptake was only weakly linked ( R 2 = 0.41) to concentration in a study of headwater streams across biomes (Dodds et al. 2002a). A study of 30 streams also in multiple biomes (Lamberti and Steinman 1997) concluded that annual gross primary prod uction (GPP) was only weakly explained by soluble reactive phosphorus (SRP; R 2 = 0.38) and that total P, NH 4 and NO 3 were not significant predictors. Several cross system studies found that light rather than nutrients was the dominant control on autotro phic

PAGE 46

46 production (North American streams Mulholland et al. 2001; Florida spring fed rivers Odum 1957a; a New Zealand river continuum Young and Huryn 1996). Finally, in a meta analysis of nutrient diffusing substrata in streams, nearly half the experi ments (42.6%) did not show a significant response to nutrient addition (Franceour 2001). Regional characteristics may influence this association, as a few studies conducted within a single region yielded higher correlations between algal biomass and nutri ents (0.63 > R 2 > 0.46) (Biggs and Close 1989 ; Lohman et al 1992 ; Chetelat et al. 1999 ; Ludwig et al. 2008), but not in all cases ( R 2 < 0.13; Sanderson et al. 2009). The much higher predictive power (0.95 > R 2 > 0.49) (Dillon and Rigler 1974 ; Jones and B achmann 1976 ; Smith 1982 ; Vollenweider 1982 ; Brown et al. 2000) of lentic eutrophication models based on nutrient concentrations alone suggests some crucial knowledge gaps regarding how nutrients affect autotrophs in lotic systems. Conceptual models of nut rient limitation in aquatic systems typically rely on water column nutrient concentrations and omit the defining influence of flow in lotic systems. Flow may influence nutrient limitation through effects on diffusive boundary layers (DBLs) and mass trans fer, the residence time of water, ecosystem scale nutrient fluxes, and the identity and abundance of resident organisms. Nutrient transfer is diffusion limited when biotic uptake rates exceed rates of diffusion from the water column through a DBL, leading to a nutrient depleted layer surrounding the biota (Borchardt 1996). Water column nutrient concentrations are theoretically most important when diffusion limitation occurs, but the prevalence of diffusion limitation in aquatic systems is currently uncert ain. Diffusion limitation is more likely under relatively quiescent conditions in lentic systems, whereas more turbulent conditions in lotic systems minimize diffusion

PAGE 47

47 lim itation by reducing the size of DBL s In one laboratory study, nutrient uptake was limited by diffusion rates through a DBL over a wide range of velocities (1 50 cm/s) (Larned et al. 2004 ). In contrast, however, Nishihara and Ackerman ( 2009 ) reported that concentration gradients were large enough to generate diffusion rates in excess of uptake rates even at low velocities (< 0.5 cm/s). When diffusion exceeds uptake then the water column nutrient concentration may cease to be a good predictor of nutrient availability, which may partially explain why lotic studies show highly variable c orrelations between nutrient concentrations and au totrophic metrics (i.e. productivity and biomass) Advective nutrient flux based on water residence times may be a better predictor of autotrophic metrics, particularly when diffusion limitation is minimal (i.e. concentration gradients are of minor importance). Moreover, when autotrophic uptake is large enough to reduce water column concentrations, nutrient flux controls the replenishment rate of nutrients in the water column and therefore the potential fo r nutrient limitation. Few studies have explicitly focused on the role of flow in relation to the advective nutrient flux in the water column, although several authors have acknowledged the potential importance of this factor (Allan 1995 ; Borchardt 1996 ; Tank and Dodds 2003) When considering the influence of advective nutrient flux on autotrophic demand in lotic systems it is important to distinguish between systems which are dominated by autotrophs attached to the benthos versus those dominated by free f loating phytoplankton. The determining factor is whether the residence time of the water is long enough to allow phytoplankton to reproduce at rates that maintain a viable population (Hilton et al. 2006). For lotic systems with shorter residence times, o nly

PAGE 48

48 autotrophs which remain attached to a substrate can persist. Attached benthic autotrophs are Eulerian (Doyle and Ensign 2009) in the sense that they remain in place while the nutrient molecules move past them due to advection, as quantified by the adv ective nutrient flux. In contrast, free floating phytoplankton move with the water in a Lagrangian sense, occupying a single parcel of water where diffusive processes are likely to control nutrient availability. In addition to system hydraulics, the influ ence of autotrophic demand on nutrient availability and limitation is also unaccounted for by water column nutrient concentrations alone. As autotrophic production increases so does the demand for nutrients, which may lead to a significant transfer of nut rients from the water column to autotrophic biomass (Mulholland et al. 2006 ; Roberts and Mulholland 2007 ; Hall et al. 2009 ; Heffernan and Cohen 2010 ). Nutrient spiraling theory supports this concept; since biotic uptake converts nutrients from dissolved i norganic forms to particulate organic forms, the ratio of inorganic to organic nutrients is indicative of nutrient limitation (Newbold et al. 1982). Autotrophic nutrient demand eventually becomes saturated as nutrient availability increases (Earl et al. 2 006 ; Covino et al. 2010a), and past this saturation threshold nutrients do not limit autotrophic production even if concentrations are low. A single metric that relates nutrient flux to nutrient demand would include multiple reach scale factors (nutrient s, discharge, and metabolism) and could be useful in characterizing the nutrient status of lotic systems. For lotic systems dominated by attached autotrophs, the ratio of advective flux to autotrophic demand for any nutrient is expected to correlate with the degree of nutrient limitation. Enumerating thresholds at

PAGE 49

49 which nutrient limitation typically occurs would provide an a priori screening tool that could be used to prioritize more detailed studies and help set protective regulatory criteria. Autotrophi c U ptake L ength This study adapt s the concept of the nutrient uptake length ( S w ) from nutrient spiraling theory (SSW 1990 ; Dodds et al. 2002a) to create a new metric that quantifies the relative magnitudes of nutrient flux and autotrophic demand. The prop osed metric, the autotrophic uptake length ( S w,a ), is obtained from a calculation that proceeds opposite the typical direction where S w is measured and the uptake rate (U) is inferred. Specifically, S w,a is computed as fo llows: where Q is the stream dis charge (L 3 T 1 ), C is the nutrient concentration (M L 3 ), w is the stream width (L), and U a is the areal uptake rate (e.g. nutrient demand; M L 2 T 1 ) due to autotrophic production onl y. In this study, t he autotrophic uptake rate U a is estim ated based on stream metabolism and autotroph stoichiometry. S w,a is equivalent to the more typical uptake length, S w but excludes removal via other biotic and abiotic pathways. S w,a,N and S w,a,P refer to the autotrophic uptake length for N and P, respectively. Ultimately, the purpose of this analysis was to advance understanding of conditions under which nutrient limitation occurs and to compare metrics for assessing nutrient limitation across a range of lotic systems. The specific objecti ve was to test the hypothesis that flow mediates nutrient concentration effects on autotrophic nutrient limitation. It is expected that autotrophic uptake length, S w,a would be a better predictor

PAGE 50

50 of autotrophic N and P limitation in lotic systems than wa ter column nutrient concentrations. This study also sought to identify critical values of S w,a below which nutrient limitation of autotrophic production is likely, and extrapolate that value to a global distribution of S w,a,N and S w,a,P for lo tic systems obtained from the literature. Methods Data S ources The two primary datasets for this analysis are the Lotic Intersite Nitrogen eXperiment (LINX) I and II studies, which combined include 82 stream assessments from across North America (Tank and Dodds 2003 ; Johnson et al. 2009). These studies utilized both inorganic (glass) and organic (cellulose) nutrient diffusing substrata (NDS) to deliver nutrients for nutrient limitation assays (NLAs). For each substrate type there were three nutrient t reatments (N, P, and N+P) and a control. The response variable was attached biofilm biomass (as chlorophyll a) after 17 22 days. Tank and Dodds (2003) was the source of the majority of the LINX I data, including NLA data (extracted from graphs using comp uterized analysis). This study assessed 10 streams resulting in 10 inorganic and 9 organic substrate NLAs total. This dataset was supplemented with stream width data (Dodds et al. 2002 a ; Mulholland et al. 2001) The LINX II data primarily came from John son et al. (2009), and the NLA data was obtained directly from the authors. Of the 72 streams assayed, 69 inorganic and 47 organic substrate assays were successfully completed. This dataset was supplemented with stream width (Mulholland et al. 2008) and gross primary production (GPP) data (Bernot et al. 2010) Mulholland et al. (2008) also contained NO 3 uptake lengths ( S w ) using the traditional spiraling methodology with a 15 NO 3 tracer which is compared in this study to the calculated S w,a,N values for the same LINX II streams.

PAGE 51

51 Nutrient concentration data was based on dissolved inorganic nitrogen (DIN) and soluble reactive phosphorus (SRP), since these inorganic forms are generally used by autotrophs. Combined, these sources provided the req uisite data (DIN and SRP concentrations NLA results, discharge, stream width, GPP) for 80 stream segments, including 79 inorganic and 56 organic substrate NLAs. The consistent methodology and mostly contemporaneous measurement of the variables in each st ream make this dataset particularly appropriate for this analysis. A literature search did not reveal any additional studies which contained the necessary data measured contemporaneously. A larger dataset was compiled from the literature to determine th e broader distribution of S w,a,N and S w,a,P across diverse streams and rivers (Table 3 1). The required data were the same as above, except for the omission of NLA results. There were many datasets ( n = 15) that fit th ese criteria, with the limiting factor usually being contemporaneous GPP and nutrient concentration measurements. In some cases, inclusion of studies in this broader dataset required assumptions about some variables. For example, Ortiz Zayas et al. (2005 ) included discharge at only 1 of the 3 stream reaches, so ratios of contributing areas were used to estimate discharge for the other 2 sites. In another case, total N and P data were used in lieu of inorganic N and P data (Duarte and Canfield 1990) alth ough organic N and P are typically very low in these systems Most studies used a 2 station method to measure GPP; however some used a 1 station method or empirically derived relationships based on stream characteristics such as autotrophic biomass, nutr ient levels, and PAR (Table 3 1) This analysis assumes these methods produce comparable results

PAGE 52

52 fed rivers are prominent in the broader S w,a dataset as they are fairly unusual lotic systems due to their discharge and chemical stability and high autotrophic productivity and their potential for nutrient limitation is uncertain Previous work on these systems (Odum 1957 a ) concluded that they were light rather than nutrient limited; however subsequen t N enrichment has been proposed as the cause for recent ecological changes (i.e. proliferation of benthic filamentous algae) via alleviation of N limitation. While this hypothesis for algal proliferation has been challenged (Heffernan et al. 2010 b ), unde rstanding nutrient limitation in these systems is critical to their management. Uptake L ength C alculations Calculations of S w,a for each stream reach, required contemporaneous measurements of Q, C, w, and U a Autotrophic nutrient demand (U a ) was determine d by converting GPP data to NPP assuming 50% of GPP was consumed by autotrophic respiration (e.g. Odum 1957a ; Hall and Tank 2003). NPP was converted from O 2 to C using a photosynthetic quotient of 1.00, then converted from C to either N or P using molecul ar ratios of C:N = 12 and C:P = 200 (approximate median values from Stelzer and Lamberti 2001). These stoichiometric assumptions for autotrophic production and biomass allowed the use of GPP data to approximate daily autotrophic N and P demand on an area basis. Nutrient flux alone (S w,a excluding autotrophic demand, U a ) was also calculated for each site as a predictor of nutrient limitation. In order to put the nutrient flux on an area basis, flux was divided by an arbitrary reach length of 100 m Statis tical A nalysis In all cases, the regressions fit a power function to the data where the NLA response ratio was the dependent variable, as this is a measure of the degree of

PAGE 53

53 nutrient limitation. NLA response ratios were calculated by dividing the average c hlorophyll a of each nutrient treatment by that of the control (RR N and RR P for N and P diffusing substrata respectively). A site was considered nutrient limited if RR was greater than 1. For each regression model, the S w,a threshold indicative of nutrie nt limitation was determined by calculating S w,a when the fitted line crossed RR = 1. For significant models (i.e. p < 0.05) the 90% confidence interval of this S w,a threshold was also calculated by determining where the confidence interval lines crossed RR = 1. Inorganic and organic substrate response ratios were analyzed separately for each stream (if applicable) due to substantial differences in their responses. Prior to statistical analysis, all values of nutrient concentrations, S w,a and RR were lo g transformed to meet the model assumptions of normality and homogeneity of variance. The LINX II dataset included eight regions across North America in which three land use types (reference (REF), agricultural (AGR), and urban (URB)) were evenly sampled allowing analysis by region or land use type. For the regional average analyses the values of nutrient concentration, S w,a and RR were averaged for the sites within each region and then follow ed the same regression procedure de scribed above. The regio nal average analyses consisted of one which included all sites and another which only included reference sites. All statistical analyses were conducted using the software package R version 2.11.1 (R Foundation for Statistical Computing). Results LINX I A nalysis For the LINX I analysis S w,a,N was a much better predictor of RR N than DIN concentration (Fig ure 3 1). The relationship between S w,a,N and RR N was significant for both NDS substrate types ( Figure 3 1A Organic: R 2 = 0.85, p < 0.001; Figure 3 1 A

PAGE 54

54 Inorganic: R 2 = 0.46, p = 0.03), but not for DIN concentration and RR N ( Figure 3 1B Organic: R 2 = 0.19, p = 0.25; Figure 3 1B Inorganic: R 2 = 0.09, p = 0.35). The N limitation threshold for S w,a,N on organic NDS was 2410 m (90% CI : 1210 5250 m). D IN flux alone was also a significant predictor of RR N (Organic: R 2 = 0.54, p = 0.02; Inorganic: R 2 = 0.52, p = 0.02). Neither S w,a,P nor SRP concentration were a significant predictor of RR P for any substrate LINX II A nalysis For the LINX II analysi s S w,a,N and DIN concentration were both significant predictors of RR N ; however, despite the increased statistical power (47 organic NDS and 69 inorganic NDS ), neither explained as much variation as S w,a,N did for the LINX I data (Figure 3 2A Organic: R 2 = 0.11, p = 0.02; Figure 3 2A Inorganic: R 2 = 0.07, p = 0.03; Figure 3 2B Organic: R 2 = 0.15, p = 0.007 ; Figure 3 2B Inorganic: R 2 = 0.07, p = 0.03). The N limitation threshold was beyond a S w,a,N of 18,700 m (90% CI: 2510 2,930,000 m) on organic ND S. Compared to the LINX I S w,a,N regressions, the slopes were much less steep, the intercepts were smaller, and both variables spanned a much larger range. As with LINX I, neither S w,a,P nor SRP concentration explained significant variation in RR P A co mparison of S w,a,N values to S w values for NO 3 addition measured as part of the LINX II study found a weak but significant correlation ( R 2 = 0.24; p < 0.001); however, these isotope tracer derived S w values were no more effective at predicting RR N than was S w,a,N ( R 2 < 0.1 for both substrates). To explore the dramatic differences in explanatory power of S w,a,N between the LINX I and LINX II datasets land use effects were examined, since LINX I streams were all characterized as reference sites and LINX II in cluded three sites of each land use type (reference, agricultural, and urban) within each region. When separated by land

PAGE 55

55 use (Figure 3 3), S w,a,N and DIN concentration were fairly poor predictors for all three land use types, although DIN concentration wa s significant in two instances. While significant scatter occurred below a S w,a,N of approximately 1000 m for reference sites and 10,000 m for agricultural and urban sites, above these thresholds the points were tightly clustered together around RR N = 1 i ndicating a lack of nutrient limitation. The models for LINX II reference sites had similar intercepts and slopes as LINX I inorganic NDS, but the LINX I organic NDS had much higher values. For the LINX II reference sites, three streams had extremely low DIN flux and unexpectedly low RR N ; when omitted, regressions improved markedly and more closely resembled LINX I. Upon further investigation these streams had high NH 4 :NO 3 ratios, which may explain why they did not respond to the NO 3 diffusing NDS. Ove rall, regressions of S w,a,N and RR N within the eight individual regions were inconsistent and not statistically significant. However, analysis of LINX II data on a regional average basis substantially improved the explanatory power of both S w,a,N and DIN concentration (Fig ure 3 4), though not S w,a,P and SRP concentration. As with LINX I, S w,a,N was a significantly better predictor of R R N than DIN concentration ( Figure 3 4A Organic: R 2 = 0.80, p = 0.003; Figure 3 4A Inorganic: R 2 = 0.38, p = 0.11; Figure 3 4B Organic: R 2 = 0.37, p = 0.11; Figure 3 4B Inorganic: R 2 = 0.04, p = 0.62). In this regional analysis, the N limitation threshold for S w,a,N on organic NDS was 24,600 m (90% CI: 6180 766,000 m). Compared to the LINX I S w,a,N regressions the slop es were similar, but the intercepts were larger and overall the S w,a,N values were higher. Because LINX II included agricultural and urban sites, regional average analyses conducted for reference sites only ar e also reported (Figure 3 4), for which S w,a,N and

PAGE 56

56 DIN concentration were both highly significant and each explained over 60% of the variation in RR N on both organic and inorganic NDS. Overall 10 different regression models were created with S w,a,N as the explanatory variable for RR N (Table 3 2). The LINX I and LINX II regionally averaged models for organic NDS were the best models for predicting RR N ( R 2 > 0.80). In general S w,a,N was a better predictor of RR N on organic than inorganic NDS. While some LINX II analyses using individual sites (rather than averaged values) were significant, none explained much variation in RR N even for reference sites only. The threshold S w,a,N values that may delimit nutrient limitation varied substantially across models. Most suggested the limitation threshold was at S w,a,N values ca. 10,000 60,000 m, although the best model (LINX I Organic NDS) had the lowest value of 2410 m. The 90% confidence intervals for these values also span a wide range, with some of the upper limits on the order of 1 x 10 7 m; however t his is likely due to small slopes of these models and high uncertainty in this data range Global D istribution of S w,a The global analysis for both S w,a,N and S w,a,P yielded median values of 12,500 m and 10,100 m, re spectively (Figure 3 5). This S w,a,N va lue is above the estimated threshold of 2410 m from the LINX I organic NDS analysis but within the range of thresholds from the LINX II analyses. Median S w,a,N and S w,a,P values for the LINX streams (5260 m and 1760 m, respectively) were lower than the me dian of all streams combined. Florida springs had higher median values (S w,a,N = 72,300 m, S w,a,P = 39,900 m), although three systems had S w,a,N values less than the LINX I threshold indicating potential N limitation. When Florida springs were analyzed a ssuming historic DIN concentrations of 0.05 mg N/L (keeping GPP, discharge, and width the same as in the

PAGE 57

57 previous analysis), the median S w,a,N value was 5480 m and nine additional systems were below the 2410 m S w,a,N value indicative of N limitation from t he LINX I dataset. For the entire global analysis, only 89 out of 276 (32%) stream reaches were below this threshold, although many of these systems may be nutrient enriched and likely had lower S w,a,N valu es in the past. Discussion Nutrient L imitation an d S piraling M etrics While the core conceptual model of nutrient limitation in lotic systems continues to require refinement, this analysis provides evidence that metrics of lotic nutrient limitation derived from the conceptual framework of nutrient spirali ng may outperform use of nutrient concentration alone. The results generally support the hypothesis that the autotrophic uptake length ( S w,a ) indicates the severity of N limitation in lotic systems; however, P limitation was both rare and poorly predicted by all metrics. The two clearest examples of superior predictive power using S w,a,N as compared to DIN concentration are for the LINX I and regionally averaged LINX II analyse s (Fig ure s 3 1 and 3 4). In both cases S w,a,N was highly significant and expla ined app roximately 80% of the response ratio variation for organic NDS, whereas DIN concentration was a non significant predictor. It appears that no other studies have examined nutrient spiraling metrics in relation to nutrient limitation as measured by NLAs; however, there are several studies which directly used spiraling metrics to assess nutrient limitation. Two studies used nutrient additions to characterize nutrient saturation (i.e. Michaelis Menton uptake kinetics) for individual streams based on relationships between nutrient concentration and spiraling metrics (Earl et al. 2006 ; Covino et al. 2010 a ). Another study determined spiraling

PAGE 58

58 metrics for a suite of streams spanning a wide range of NO 3 concentrations and found a response characteristic o f an efficiency 2007). In general, short S w and high v f are likely to be indicative of lotic nutrient limitation. Considering the utility of nutrient spiraling theory with regard to nutrient transport in lotic s ystems, a conceptual model of lotic nutrient limitation should be based on nutrient spiraling metrics rather than nutrient concentrations and ratios. Three aspects of the autotrophic uptake length, S w,a make it a theoretically compelling metric: 1) it ha s a basis in spiraling theory, 2) it integrates information about stream GPP, which can vary for reasons other than nutrient availability, and 3) it is scale invariant, and therefore applicable across drainage networks. S w,a explicitly assesses autotroph ic nutrient limitation within the nutrient spiraling framework, and could potentially improve our ability to predict the effects of nutrient enrichment on autotrophic communities. S w is a measure of advective nutrient flux per the rate of nutrient removal from the water column, which is useful but does not differentiate between autotrophic uptake and other removal processes (e.g. denitrification for N and sorption for P). S w,a represents only the autotrophic portion of S w while retaining other reach chara cteristics (i.e. discharge, width, GPP, and nutrient concentration) which affect nutrient availability and the degree of autotrophic limitation. A measure of autotrophic nutrient limitation is important since concern for nutrient dynamics in aquatic syst ems is often related to overgrowth of algae and macrophytes (i.e. eutrophication). A metric which integrates autotrophic production (i.e. GPP), indirectly accounts for other factors such as light levels, scouring, and grazing which

PAGE 59

59 may constrain productio n and nutrient demand. For example, in shaded lotic systems lower nutrient concentrations are required to saturate autotrophic demand since the low light levels limit the amount of autotrophic biomass (Hill and Knight 1988 ; Rosemond et al. 2000). Anothe r useful feature of S w,a is that theoretically it is scale invariant, meaning that the ability of S w,a to predict autotrophic nutrient limitation should apply to lotic systems of all sizes, as long as benthic production is dominant. This relationship may be different and the Mississippi River), since these organisms are suspended in the water column and nutrient transfer is more dependent on diffusive processes. Ligh t intensity is a strong predictor of autotrophic nutrient limitation (Tank and Dodds 2003) across small streams, largely due to heavy canopy cover in many of these systems. For larger systems canopy cover does not substantially limit light availability, a nd therefore incident light intensity would not be able to explain differences in nutrient limitation between large river systems whereas S w,a should retain predictive power. Since it is difficult to assess autotrophic nutrient limitation and spiraling me trics in larger lotic systems due to methodological constraints, a scale invariant metric such as S w,a could be useful in under studied (Ensign and Doyle 2006) large systems. Assessing A utotrophic N utrient D emand and L imitation Our ability to assess autotr ophic nutrient demand at the scale of whole ecosystems is currently limited by available measurement techniques. In this analysis it was assumed that there were consistent C:N, C:P, GPP:NPP, and photosynthetic quotients among all systems in order to conve rt GPP values (based on oxygen changes) into autotrophic nutrient demand (U a ). Calculations of U a do not presently

PAGE 60

60 account for variation in these values, and the effects of this omission on precision and bias of S w,a are not known. The measurement of die l variations in N and P using improved in situ sensors may obviate these stoichi ometric and physiological assumptions by providing a direct measure of nutrient demand (Heffernan and Cohen 2010 ). Nutrient diffusing substrata (NDS) of various types have be en the primary method for characterizing the degree of nutrient limitation in lotic systems to date, and their utility is apparent based on extensive use in many settings. However NDS are limited in their ability to assess reach scale autotrophic nutrient limitation, since they generally target biofilms which may not be representative of the autotrophic community within that reach. Larger autotrophs (particularly rooted macrophytes) would not be captured by most NDS, whereas their production could be a si gnificant component of the total amount of autotrophic production in a reach. Rooted macrophytes may also experience different nutrient regimes than autotrophs in benthic biofilms due to their ability to acquire resources directly from sediment porewaters Results of NDS assays are also influenced by temporal attributes of a system such as algal immigration rates and storm events. In this analysis, some of the streams with the lowest DIN flux had a minimal response to the N enriched NDS despite moderate l ight and GPP levels, which could reflect such limitations of NDS There were also different relationships for the two substrate types used in LINX (e.g. organic and inorganic), and S w,a,N was a better predictor of organic NDS response ratios overall. Rec ently developed techniques which measure the degree of nutrient limitation for pre existing autotrophic communities in situ, such as the tracer additions for spiraling curve

PAGE 61

61 characterization (TASCC) methodology recently proposed by Covino et al. (2010b), m ay provide better metrics of the severity of nutrient limitation. Evidence is accumulating that NDS are particularly poor assays of autotrophic P limitation. In several cases P enriched NDS have shown less algal accrual than controls indicating that P e nrichment may have inhibited algal growth (Hill and Knight 1988 ; Tank and Dodds 2003 ; Sanderson et al. 2009). Although the mechanism behind this phenomenon and its significance are unknown, it is likely related to the extremely low P concentrations requir ed for growth rate saturation (< 4 g/L: Bothwell 1985) and that P cycling in the benthos may be sufficient to alleviate any substantial P limitation. In this analysis, both SRP concentration and S w,a,P were unable to explain any variation in the P respon se ratios, and the majority of these response ratios were less than 1 (i.e. less algae grew on the P enriched NDS than on the un enriched controls). A general issue with nutrient limitation assays is the distinction between instantaneous and chronic nutr ient enrichment effects. Short term assay techniques may not capture effects of long term chronic nutrient enrichment which can cause substantial changes in autotrophic community composition with altered nutrient cycling and potentially higher nutrient de saturation in biotic demand among streams spanning a range N concentrations due to chronic N enrichment, in contrast to the findings of two studies which assessed nutrient limitation using instantaneous N additions (Earl et al. 2006 ; Covino et al. 2010a). One mechanism explaining this distinction is the formation of large algal mats under chronic nutrient enrichment, and the difference between nutrient limitation of peak algal biomass and cellular growt h rate. Peak algal biomass requires time to develop and nutrient

PAGE 62

62 saturation occurs at higher levels than for cellular growth rate (Bothwell 1989). NDS would mostly capture growth rate nutrient limitation (since they usually collect a thin layer of algae) and are therefore unlikely to characterize nutrient limitation for systems with substantial algal mats. This could explain why some streams with lower S w,a values did not exhibit nutrient limitation, as GPP may have been high due to thick algal layers t hroughout the reach but nutrient limitation was not evident due to thin algal growth on NDS. Overall, the extent to which these results reflect the idiosyncrasies of NDS (as opposed to limitations of the S w,a metric) is not know n. Effects of L and U se and R egion For the LINX II dataset, it was expected that S w,a would be a good predictor for each land use type since it accounts for any major differences related to autotrophic nutrient limitation. However both S w,a,N and DIN concentration were fairly poor predictors of RR N when separated by land use, which may indicate that the differences between each land use type were minimal. Although DIN concentration was a significant predictor of RR N in two instances, the results overall are inconsistent and there a re no clear trends related to land use effects. It was also expected that S w,a would be a good predictor of nutrient limitation both among and within regions. Dodds et al. (2002b) suggest that relationships between nutrient concentrations and algal biom ass within a region should be stronger than among regions, as found in some studies (Biggs and Close 1989 ; Lohman et al. 1992 ; Chetelat et al. 1999 ; Ludwig et al. 2008). Dodds and Welch (2000) describe how region specific data have led to more accurate pr edictions of algal biomass. While they caution that it is possible to have high variation both temporally and spatially at the stream reach scale, they conclude that if conditions are similar (i.e. light, grazers, etc.)

PAGE 63

63 then the nutrient concentration rel ationship with algal biomass should be valid among regions. While nutrient concentration may not be the best predictor of autotrophic biomass, there is nothing inherent in a region or individual stream that should affect the relationships discussed here. Differences in factors such as light intensity, grazer abundance, sediment or dissolved organic particle effects on light availability, and toxins in runoff should all be indirectly accounted for in the autotrophic demand component of S w,a since this ref lects the current level of primary production. In contrast to other studies of lotic systems which indicated that stronger relationships between nutrient levels and autotrophic biomass occur within a single region than among regions, S w,a was not a very good predictor within any of the eight regions covered by the LINX II dataset. One partial explanation is that there was not a large range of both S w,a and response ratio values within most of the regions. Overall, it is unclear why regression models for S w,a on organic NDS are significant when using regional averages (i.e. among regions) and not otherwise, but it may point to some unaccounted for regional differences. The fact that the LINX I dataset was compiled from individual streams in similar regio ns as for LINX II and shows fair model fits indicates that the S w,a metric is a good predict or of nutrient limitation among regions. Nutrients and A lternative C ontrols Nutrients are one of many factors which can strongly affect autotrophic production a nd biomass accrual in aquatic systems. In lotic systems, top down controls and velocity induced sloughing may limit algal biomass as much as bottom up controls (e.g. nutrients and light ) Recent research indicates that a preponderance of lotic systems do not display nutr ient limitation (Francoeur 2001 ; Tank and D odds 2003 ; Johnson et al. 2009), suggesting that these other factors dominate. The global analysis of N limitation

PAGE 64

64 furthers supports this finding as 68% of the systems were above the provisional S w,a,N threshold of 2410 m indicative of N limitation. This finding does not diminish the need for nutrient reduction in anthropogenically enriched systems, but does indicate that in many lotic systems nutrients are not the primary driver of autotrophic production. Due to the difficulty of decreasing elevated nutrient levels combined with the potential for hysteresis in degraded ecosystems, management actions that go beyond nutrient control and reduction appear warranted. Further research that focuses o n the full suite of controls on autotrophic production and biomass would complement efforts to reduce the potential effects of nutrient enrichment by providing additional techniques for effective ecosystem management

PAGE 65

65 Figure 3 1. Comparison of autotroph ic uptake lengths (S w,a ) and nutrient concentrations in relation to the NLA response ratios (RR x ) from the LINX I dataset. Black indicates organic NDS and gray indicates inorganic NDS.

PAGE 66

66 Figure 3 2. Comparison of autotrophic uptake lengths (S w,a ) and nut rient concentrations in relation to the NLA response ratios (RR x ) from the LINX II dataset. Black indicates organic NDS and gray indicates inorganic NDS.

PAGE 67

67 Figure 3 3 Comparison of autotrophic N uptake length (S w,a,N ) and DIN concentrat ion for the three land use types in relation to RR N for individual streams from the LINX II dataset. REF reference, AGR agricultural, URB urban. Black indicates organic NDS and gray indicates inorganic NDS.

PAGE 68

68 Figure 3 4. Comparison of autotrophic N uptake length ( S w,a,N ) and DIN concentration in relation to the RR N from the regionally averaged LINX II dataset (A and B) and for the regionally averaged reference sites only (C and D). Black indicates organic NDS and gray indicates inorganic NDS.

PAGE 69

69 Figure 3 5. Normal probability distributions for S w,a,N and S w,a,P of the three groups of datasets fit to their relative frequency of occurrence. Florida spring fed rivers were included separately to demonstrate where they are located in terms of expected nutrient limitati on. The S w,a,N value at which RR N = 1 is shown (vertical black line) with the 90% confidence interval (gray bar) based on the LINX I organic NDS analysis

PAGE 70

70 Table 3 1. Global analysis datasets. In some studies a stream reach was sampled on multiple occas ions and these values were included as separate points. Florida springs systems are listed separately at the bottom of the table. Reference Location Reaches sampled Total data points GPP methodology Elosequi and Pozo 1998 a Spain 2 8 diel DO change Fello ws et al 2006 New Mexico, USA 2 2 2 station DO change Tennessee, USA 1 1 2 station DO change North Carolina, USA 1 1 2 station DO change Fisher et al 1982 Arizona, USA 1 7 empirical equations Grimm 1987 Arizona, USA 1 5 in situ chambers Arizona, USA 1 1 diel DO change Hall and Tank 2003 Wyoming, USA 11 11 2 station DO change Hall et al 2003 Wyoming, USA 1 3 2 station DO change Meyer et al 2005 Georgia, USA 6 6 diel DO change Mulholland et al 1997 Tennessee, USA 1 1 2 station DO change Nor th Carolina, USA 1 1 2 station DO change O'Brien et al 2007 Kansas, USA 9 9 diel DO change Ortiz Zayas et al 2005 b Puerto Rico 3 28 diel DO change Wright and Mills 1967 Wyoming, USA 5 5 2 station pH change Young and Huryn 1999 New Zealand 5 5 2 stati on DO change Duarte and Canfield 1990 c Florida, USA 32 32 empirical equations WSI 2008 Florida, USA 6 24 2 station DO change WSI 2010 d Florida, USA 12 12 2 station DO change a nutrients and discharge values are annual averages b discharge estimate d for two of the streams, based on the discharge of the main stream and the basin area ratio c widths were estimated by dividing discharge by velocity and depth; TN and TP data used instead of DIN and SRP d maximum widths rather than average

PAGE 71

71 Table 3 2. Summary of the S w,a,N regression results. The S w,a,N indicative of N limitation (i.e. S w,a,N when RR N = 1) is also included along with the 90% confidence interval for significant models. Model Limiting 90% Confidence Interval M odel Parameter Value SE p R 2 p S w,a,N (m) Lower Upper 1) LINX I Organic NDS Intercept 1.12 0.17 <0.001 0.85 <0.001 2.41E+03 1.21E+03 5.25E+03 S w,a,N 0.33 0.05 <0.001 2) LINX I Inorganic NDS Intercept 0.53 0.15 0.009 0.46 0.031 2.34E+0 4 4.25E+03 3.03E+07 S w,a,N 0.12 0.05 0.031 3) LINX II Organic NDS Intercept 0.47 0.18 0.011 0.11 0.023 1.87E+04 2.51E+03 2.94E+06 all sites S w,a,N 0.11 0.05 0.023 4) LINX II Inorganic NDS Intercept 0.31 0.11 0.006 0.07 0 .030 2.68E+04 3.24E+03 2.33E+07 all sites S w,a,N 0.07 0.03 0.030 5) LINX II Organic NDS Intercept 0.64 0.25 0.023 0.23 0.070 3.73E+04 N/A N/A reference sites S w,a,N 0.14 0.07 0.070 6) LINX II Inorganic NDS Intercept 0.51 0 .22 0.029 0.17 0.055 8.38E+03 N/A N/A reference sites S w,a,N 0.13 0.06 0.055 7) LINX II Organic NDS Intercept 1.43 0.25 0.001 0.80 0.003 5.84E+04 2.49E+04 2.56E+05 regional averages S w,a,N 0.30 0.06 0.003 8) LINX II Inorg anic NDS Intercept 0.74 0.30 0.048 0.38 0.106 1.93E+05 N/A N/A regional averages S w,a,N 0.14 0.07 0.106 9) LINX II Organic NDS Intercept 1.01 0.26 0.008 0.65 0.016 2.46E+04 6.18E+03 7.66E+05 regional avg reference S w,a,N 0.23 0.07 0. 016 10) LINX II Inorganic NDS Intercept 1.07 0.23 0.004 0.73 0.007 1.30E+04 4.44E+03 7.98E+04 regional avg reference S w,a,N 0.26 0.06 0.007 *all variables log10 transformed

PAGE 72

72 CHAPTER 4 FLOW VELOCITY CONTRO L OF FILAMENTOUS ALGAE IN A SUB TROPICAL SPRING FED RIVER Introduction Algae are an integral component of most aquatic systems due to their influence on oxygen and nutrient dynamics and their role as a food source for grazing organisms that, in turn, support the production of higher trophic levels. However, when algal abundance increases rapidly (i .e. proliferates) there are often a broad suite of negative consequences such as degraded ecosystem structure and function, impaired recreational use and reduction in aesthetic quality and even human health impacts (Lembi 2003; Smith 2003). Algal proliferation is typically attributed to nutrient enrichment (i.e. eutrophication ) which may stimulate the growth rate of nutrient limited algae, and for systems initiall y dominated by submerged vascular plants this can cause a shift to attached algae and then potentially to free floating planktonic algae (Duarte 1995; Valiela et al. 1997). While nutrient enrichment is often the primary cause of algal proliferation in aqu atic systems in general there are other possible explanations such as loss of algal consumers (Jones and Sayer 2003) or changing flow conditions (Wade et al. 2002). In lotic systems (i.e. rivers and streams), flow is known to have a significant influenc e on algal metabolism, abundance, and community structure (Biggs 1996; Stevenson 1996), but when algal proliferation occurs flow conditions are rarely considered as a regulating factor. Most of the focus on flow has been in relation to periodic flood even ts which scour channel s and dislodge attached algae (Biggs 2000). However in lotic systems where scouring events are infrequent, attached algae may accumulate over inter annual time periods, and algal abundance may ultimately be

PAGE 73

73 controlled by long er ter m hydrologic fluctuations This study examined a sub tropical spring fed river with minimal scouring events to assess whether long term declines in flow might account for observed filamentous algal proliferation. Flow V elocity and A lgae The flow of wate r is involved in multiple mechanisms which regulate algal abundance and distribution in lotic systems (Biggs 1996; Stevenson 1996). Flow velocity has been shown to have a positive effect on algal growth due to increased transfer of nutrients to cells (Whi tford and Schumacher 1961, 1964; McIntire 1966; Lock and John 1979; Horner et al. 1990) and waste products away from cells (Mass et al. 2010). However, with increasing flow velocity, the negative effects of drag force will eventually override the positive effects on chemical transfer, creating relationship (Odum et al. 1979) between flow and biomass accumulation. This has been demonstrated over a wide velocity range (Horner and Welch 1981; Horner et al. 1990 ), but a synthesis indicates that the strength and direction of the flow velocity algae relationship varies widely (see Table 1 in Stevenson 1996), making it difficult to predict with certainty how algae will respond to changing flow conditions. While some of the variation in the flo w velocity algae relationship can be attributed to interactions with other key environmental variables such as nutrients, light intensity, and algal grazers; much is due to differential response among algal species that exhibit a wide range of morphologica l characteristics and attachment capabilities. Particularly important is the growth form (i.e. physiognomy) of the algal species as small, closely adhering species are more likely to show a positive relationship with velocity due to increased chemical tra nsfer whereas larger filamentous forms of algae are likely to show a negative relationship due to increased drag force (Biggs et al. 1998).

PAGE 74

74 In lotic systems, the drag force due to flow velocity can have two main effects on algae: 1) inhibiting establishme nt and 2) dictating sloughing (i.e. export) rates. High velocities can inhibit colonization for some species of algae (Stevenson 1983; Stevenson and Peterson 1989). Once algal colonization has occurred, velocity controls the sloughing rate by which algal cells are detached from their substrate and exported downstream, thus contributing to the overall loss rate of algal biomass (McIntire 1966; Biggs and Thomsen 1995). Drag force due to friction increases at a rate proportional to the square of the flow ve locity, so that small changes in velocity can potentially lead to significant changes in algal export rates. Algal cells may be sheltered from drag force by residing within the boundary layers that surround submerged surfaces. Local flow velocity control s the thickness of the boundary layer, and therefore can alter the hydraulics of the near surface environment in which attached algae reside. While most field studies of flow velocity effects focus on the role of flood (i.e. scouring) events (Tett et al. 1978; Biggs and Close 1989; Peterson and Stevenson 1992), the inhibitory effect of velocity has also been recorded in lotic systems with relatively stable hydrology (e.g., low gradient systems fed by groundwater) (Poff et al. 1990; Ghosh and Gaur 1991). T his study is one of the first to examine the effect of declining flow velocity on algal abundance over an inter annual time period in a system with low hydrologic variation. Factors A ffecting Flow V elocity The magnitude and distribution of flow velocity wi thin a lotic system are largely controlled by the overall rate of discharge and the geomorphology of the channel, although woody debris and submerged aquatic vegetation (SAV) may have significant influence on velocity at smaller spatial scales. Due to fri ction, velocities are typically highest in the center of the channel near the water surface and lower at the channel

PAGE 75

75 margins and near the sediment/water interface. Submerged aquatic vegetation, particularly dense monotypic patches (i.e. beds), obstruct th e flow of water and alter the distribution of flow velocities such that velocity is lower within the beds, but higher throughout the rest of the water column (Sand Jensen and Mebus 1996). Algal mats may significantly alter the flow velocity distribution i n a similar manner (Dodds 1991), though algae are sloughed downstream more easily than rooted vascular plants due to the difference in substrate attachment capability. It is hypothesized that the biomass of filamentous algae and SAV are positively related since SAV may decrease velocities locally and provide stable substrate for algal colonization; however over longer time periods algae may lead to the decline of SAV due to factors such as light competition. Spring fed R ivers in Florida Florida contains a mong the highest density of first magnitude (> 2.8 m 3 /s; Copeland 2003 ) artesian springs in the world as a result of unique karst geology and a highly productive aquifer system ( Scott et al. 2004 ). Many of these springs form spring fed rivers, which are c haracterized by high water clarity, stable hydrology, chemistry, and temperature, and dense SAV beds (Odum 1957b; Canfield and Hoyer 1988). These spring fed rivers were the focus of early ecological studies which determined that flow velocity and light av ailability were the primary drivers of algal community structure and autotrophic production rates, rather than nutrients (Whitford 1956; Odum 1957a). Over recent decades, filamentous algae such as the cyanophyte Lyngbya wollei and the xanthophyte Vaucheri a (Stevenson et al. 2004; Stevenson et al. 2007). Anthropogenic nitrogen enrichment is widely considered to be the cause of algal proliferation, but a recent synthesis of the evidence for this con tention suggests that nutrient enrichment is not likely the primary

PAGE 76

76 causal factor and that alternative hypotheses should be evaluated (Heffernan et al. 2010b). The influence of flow conditions and hydraulics on filamentous algal abundance in Florida spri ng fed rivers has not been adequately investigated. These rivers typically respond to rainfall over a period of months (e.g., Florea and Vacher 2006; Heffernan et al. 2010 a) and do not exhibit scouring events which limit algal abundance in many lotic syst ems. Many spring fed rivers have experienced declining discharge over the last few decades (Munch et al. 2006; Weber and Perry 2006; Copeland et al. 2009 ; Grubbs 2011), which would generally lead to lower velocities and less drag on algae. It is hypothes ized that lower flow velocities allow greater amounts of filamentous algae to accumulate in Florida spring systems, and there is a velocity threshold above which algal abundance is minimal. Methods Site D escription The Gum Slough spring system is a second magnitude springs complex located near the central west coast of peninsular Florida in Sumter County (Figure 4 1). Six larger and numerous smaller springs combine to form an approximately 8 km river that discharges into the Withlacoochee River. The large r springs are all within the upper 2 km of the river where the water is extremely clear and tree canopy cover is moderate, thus providing sufficient light for high rates of autotrophic production. Further downstream, the river becomes more forested and br aided (i.e. a slough) with less autotrophic biomass. Most of the adjacent land is undeveloped and the river receives sparse human traffic due to its remote location and relative inaccessibility.

PAGE 77

77 Discharge downstream of the main springs complex averaged 2.62 m 3 /s from October 2003 to December 2011 (the period of record at USGS gage #02312764), fitting the classification as a large second magnitude spring system (Figure 4 2). The period of record for discharge at Gum Slough is not sufficient to assess lon g term flow conditions; however, it is likely that over the past decade discharge was lower than for previous decades due to substantially less rainfall as observed for other spring systems ( Copeland et al. 2009 ). Over the period of record, Gum Slough was subject to the effects of multiple tropical storms followed by low flow periods in which sections of the upper river ceased flowing due to minimal discharge from the upstream springs. During this study, discharge decreased from moderate to extremely low levels, which provided an opportunity to examine the effects of declining discharge on flow velocity and filamentous algal cover. Gum Slough water chemistry is typical of a hard freshwater spring (e.g., Whitford 1956) with higher sulfate concentrations t han many similar inland springs (Champion and Starks 2001) (Table 4 1). Nitrate concentrations have increased from 0.5 mg N/L in 1972 (Scott et al. 2004) to a high of 1.53 mg N/L in 2011, and were likely less than 0.1 mg N/L prior to human settlement (Hef fernan et al. 2010b). The increased nitrate concentrations in many Florida springs are due to a combination of agricultural and residential fertilizer application, intensive animal operations, land application of wastewater effluent and residuals, and sep tic tank discharge within the recharge area ( Katz 2004; Phelps 2004 ). The autotrophic community is dominated by the rooted macrophytes Sagittaria kurziana and Vallisneria americana which shift in dominance from the former to the

PAGE 78

78 latter with distance dow nstream from spring inpu ts. A variety of epiphytic algal species contribute significantly to overall primary production in spring systems (Whitford 1956; Odum 1957b). Over the past decade, filamentous algae, primarily the cyanophyte Lyngbya wollei has i ncreased in abundance and formed persistent mats which occasionally proliferate to extreme nuisance levels, especially near the upper most spring vents. Despite the shift towards filamentous algal dominance, there remains a large and diverse fish populati on dominated by large mouth bass ( Micropterus salmoides ) and various sunfish ( Lepomis spp.) ( WSI 2011 ). There is also a large snail population comprised primarily of Elimia spp. but also including apple snails ( Pomacea spp.) and other species (personal ob servation). These snails likely have a significant influence on algal dynamics in their role as algal grazers (Heffernan et al. 2010b). Field Surveys Field data were collected at Gum Slough at three transects perpendicular to the river channel, which were sampled approximately every two months ( n = 10) (Figure 4 1). The three sites were chosen based on river morphology and relative flow velocities; T1 is wide and deep with low velocity (width = 30 m; depth = 1.2 m), T2 is narrow and shallow with high velo city (width = 13 m; depth = 0.6 m), and T3 is deep with moderately high velocity (width = 18 m; depth = 1.2 m). Sampling began on 6/21/10 and ended on 12/14/11, and over this time period discharge (and water levels) d ecreased substantially below the longe r term average All discharge rates and stage levels refer to measurements at the USGS gage located just downstream of all spring inputs (Figure 4 1), since discharge rates calculated from the velocity data for each transect were all linearly related to t he USGS gage discharge rate (R 2 > 0.98; p < 0.001).

PAGE 79

79 Flow velocity with depth and filamentous algal cover were measured at 2 m intervals along each transect. Flow velocity was measured with a portable velocity meter (Model 2000; Marsh McBirney Inc., Fred erick, Maryland), using a 5 second average. At each sampling point vertical variation in f low velocity was measured at 10 cm increments starting just below the water surface and ending at the sediment surface. T1 was sampled in less vertical detail (20 c m depth increments) due to its width and low variation in velocity. At each transect sampling point the areal percent cover of filamentous algae was visually estimated with a 0.25 m 2 quadrat. When a thick algal mat or SAV bed was present, the depth at th e top of the mat or bed was recorded. Transect data were analyzed in two ways: for transects as a whole and for individual sampling points. The analysis of transects as a whole used the average (v avg) and maximum (v max) of all velocities measured with d epth across each transect, and averaged the filamentous algal cover recorded at each sampling point. The analysis of individual sampling points along a given transect also used average and maximum velocities, but in these cases the average and maximum vel ocities refer to the measurements with depth for each sampling point. E xperiment al Manipulation of Flow Velocity Plastic baffles were used to construct two side by side channels with different flow conditions to experimentally test the effect of flow vel ocity on both filamentous algae and SAV abundance as indicated by biomass (Figure 4 3). Three clear plastic sheets (0.65 m wide and 3 m long) were used to create the baffles. Each sheet was trimmed to 2.5 m to make the two channels. The 0.5 m trimmed se ctions were used as inlet control structures with one side open (i.e. \ /) and the other semi closed (i.e. / \ ), thus forming high and low flow channels, respectively. Each section of plastic sheet was attached to

PAGE 80

80 wooden stakes which were inserted into th e sediment so that the bottom of the sheet was flush with the sediment surface. The baffles were placed in three different locations in the upper section of the river where SAV density was homogen e ous and water depths were < 0.65 m. The first two locatio ns were only harvested once while the third location was harvested three times, resulting in five sampling events. At this third location, there were two harvests when samples were taken in previously un harvested areas, then the entire baffle area was cl eared of algae and SAV to test the effect of SAV absence. Additional sampling was precluded by declining discharge which eliminated appropriate installation locations. The baffles were installed for approximately 7 weeks (46 to 52 days) to allow for filam entous algal colonization prior to harvesting (except for the one instance when the baffles were harvested a second time, which occurred 28 days after the first harvest). After the colonization period, a metal cylinder (0.05 m 2 ) was used to harvest all au totrophic biomass from three equidistant points within the center of each channel. This material was transported on ice back to the laboratory where it was separated into filamentous algae and SAV components, and rinsed of other materials (e.g., sediment, leaves, invertebrates). The separated sample material was then weighed and dried in an oven at 70 C until constant weight to determine the dry weight, after which the dried samples were ashed in a muffle furnace at 440 C for four hours to determine the ash free dry mass (AFDM). Just prior to biomass harvest, flow velocity measurements were taken at three equidistant points in the center of each channel, at depth increments of 10 cm throughout the water column.

PAGE 81

81 Statistical A nalys i s Linear regression was used to assess the relationships between discharge and flow velocity, and between flow velocity and filamentous algal cover. Temporal variation in filamentous algal cover was investigated by numbering each day of a year (1 365) starting on 10/15, si nce this date was approximately the end of the growing season. The effects of both flow velocity and temporal variation (i.e. day of the year) on filamentous algal cover were assessed with additive multiple linear regression. Linear regression and a t te st were used to assess the effects of experimental manipulation of flow velocity on filamentous algae and SAV biomass (omitting the last harvest event when SAV was absent). All statistical analyses were conducted using the software package R version 2.11 .1 (R Foundation for Statistical Computing). Results Field Surveys During the course of this study discharge decreased from a high of 2.98 m 3 /s on 6/21/10 to a low of 0.94 m 3 /s on 6/29/11 (Figure 4 2). Subsequently, discharge increased, but never exceed ed 1.70 m 3 /s. Discharge for each sampling event was highly correlated with both the average velocity (R 2 > 0.89; p < 0.001) and maximum velocity (R 2 > 0.85; p < 0.001) measured for each transect. Overall the sampling dates fell into three distinct groups based on discharge; high flow: > 2.75 m 3 /s (6/21/10 10/8/10); medium flow: 1.13 to 1.64 m 3 /s (1/20/11 5/9/11 and 8/16/11 12/14/11); and low flow: < 1.0 m 3 /s (6/29/11). For each transect, the distribution of flow velocity was generally highest near t he top of the water column in the center of the channel and decreased to zero near the sediment surface. Where SAV beds were present (for T2 and T3) there was a sharp

PAGE 82

82 change in velocity with depth, and velocities within SAV beds were typically < 10 cm/s. Large fluctuations in velocity ( 5 cm/s) occurred just above SAV beds due to turbulence, particularly at high discharge. SAV beds were absent at T1 throughout the study, and while velocity decreased near the sediment surface the change with depth was le ss pronounced. T entire study period (Figure 4 4). When flow velocity was the lowest (on 6/29/11) algal cover was greatest; although on 1/20/11 the algal cover was also high despite slightly greater flow velocity. Large floating algal mats were present at the water surface towards the transect edges throughout the study period; flow velocity at these edges was always < 3 cm/s. On the lowest flow date (6/29/11), when velocity was barely detec table, floating algal mats were present across the entire transect. The other two transects (T2 and T3) had much higher flow velocities throughout the study period and filamentous algal cover was generally low except at low velocities, suggestive of a thre it increased to 85% during the summer months. This increase in algal cover coincided with a decline in average flow velocity from > 20 cm/s to a low of 5 cm/s on 6/29/11. For T3, algal cover was < 6% until 3/9/11 and also reached 80% by the end of the summer with similar changes in flow velocity. Overall, the highest values of filamentous algal cover occurred below 15 cm/s average velocity (22 cm/s maximum velocity), which corresponded to a discharge rate of approximately 1.5 m 3 /s at the downstream USGS gage.

PAGE 83

83 A seasonal pattern of filamentous algal cover was apparent, as winter values were low both years but summer values were only low when flow velocity was high. During two winter samp ling events (1/20/11 and 12/14/11), algal cover was < 15% despite low flow velocity for the two higher velocity transects (T2 and T3). However, algal cover remained > 50% for the low velocity transect (T1). Filamentous algal cover substantially increased to > 80% from 3/9/11 to 8/16/11 for T2 and T3; however measurements from the previous summer were < 20% when flow velocities were much higher. This suggests that there is potential for filamentous algal proliferation in the spring and summer months, but if the average velocity is above the 15 cm/s threshold then algal accrual is inhibited. The linear relationship between flow velocity (both average and maximum) and filamentous algal cover was weak and insignificant for transects T1 and T2 (R 2 < 0.30, p > 0.1); however for T3 maximum velocity was a significant predictor (R 2 = 0.41, p = 0.046) and average velocity was borderline (R 2 = 0.37, p = 0.059). When a seasonal effect was included in these regression models by accounting for the day of the year, t hen the models for T2 and T3 improved substantially and were significantly related to maximum velocity (Table 4 2). Inclusion of both the seasonal effect and maximum flow velocity did not generate a significant model for T1, likely because the velocities were always below a threshold value that might otherwise inhibit algal accrual. Analysis of individual sampling points along transects yielded an average velocity threshold of 25 cm/s and a maximum velocity threshold of 35 cm/s beyond which algal cover was < 5% (Figure 4 50% at multiple locations, suggesting that velocity induced drag is a significant

PAGE 84

84 constraint on filamentous algal accrual. These apparent threshold values are larger than those de rived from average transect values, since there were individual sampling stations with moderate algal cover despite higher velocities. Overall, many low flow velocity stations contained low filamentous algal cover, which indicates that factors other than velocity also limit algal cover. The primary role of local flow velocities was to create an upper bound that constrained the maximum potential algal cover, unless velocities were < 5 cm/s at which point 100% algal cover was frequent. E xperiment al Manipula tion of Flow Velocity The baffle experiment results generally supported the survey results, as filamentous algal biomass was negatively correlated with both average and maximum flow velocity (Figure 4 6). The average algal biomass (AFDM) was 216.3 g/m 2 an d 137.7 g/m 2 in the low and high flow channels, respectively, although values varied widely depending on baffle location and associated flow velocity differences. The t test did not find a significant difference in algal biomass between the low and high f low channels, likely due to a combination of relatively small differences in flow velocity and low statistical power ( 4 samples total ). The flow velocity threshold that inhibited substantial algal biomass was approximately 10 cm/s. There was a positive c orrelation between filamentous algal biomass and SAV biomass (Figure 4 6), which suggests that filamentous algae may increase with SAV density. However, when the baffles were left at the third location for three harvest events, the resulting algal biomass for each channel was similar despite the absence of S AV for the final harvest (208 230 g/m 2 in the low flow channel; 128 153 g/m 2 in the high flow channel).

PAGE 85

85 Discussion Flow V elocity and F ilamentous A lgae This study involved both in situ field surveys an d an experimental manipulation, and the combined results suggest that flow velocity constrains filamentous algal abundance in the Gum Slough spring system. The early part of the study period was ideal for testing the effect of declining discharge and velo city on filamentous algal cover, and also allowed comparison of two growing seasons with different flow conditions. Over this period, filamentous algal cover increased substantially at the two transects with initial velocities above the implied threshold of 25 cm/s depth averaged velocity (35 cm/s maximum velocity), whereas the low velocity transect was below the velocity threshold for the entire study period and always had > 50% algal cover. This flow velocity threshold is consistent with findings by Hoy er et al. (2004) who found that above 25 cm/s filamentous algal biomass was minimal in three coastal Florida spring fed rivers, which suggests that this velocity threshold applies to lotic systems with similar characteristics. While a few other studies of Florida springs have anecdotally mentioned that less algae occurs where velocity is greater (Munch et al. 2006; Quinlan et al. 2008), this study and that by Hoyer et al. (2004) are the only ones which have quantified the relationship between flow velocity and filamentous algae in sufficient detail to specify a velocity threshold that likely inhibits algal proliferation. Fe w studies have focused on how declining discharge rates in lotic systems may contribute to attached algal proliferation. Since flow vel ocity is directly related to the drag force exerted on attached algae, declines in flow velocity should allow more algae to establish. This process, however, may be non linear due to the presence of velocity thresholds which substantially constrain algal abundance. Lower velocities could also

PAGE 86

86 exacerbate the effects of nutrient enrichment by reducing the export rate of attached algae. Wade et al. (2002) created a simulation model to examine effects of nutrient enrichment and flow declines in an English ri ver, and found that lower flow reduced the loss of epiphytic algae which, in turn, reduced SAV biomass. The concept that declining discharge is as important a contributor to algal proliferation as nutrient enrichment should be more fully explored since fl ow is expected to change in many lotic systems globally due to climate change effects on rainfall and increasing water withdrawal by humans. T he Re lationship between SAV and F ilamentous A lgae In many Florida spring fed rivers, filamentous algal abundance i s increasing as SAV abundance is decreasing (e.g., Quinlan et al. 2008), but it is currently unknown if these two trends are directly related. The results of the baffle experiments suggest that filamentous algae may utilize SAV as a substrate, but SAV pre sence is not necessary if flow velocities are sufficiently low. This study also found evidence that SAV beds decrease flow velocities near the sediment surface and increase turbulence just above SAV beds. Overall, it is unclear whether SAV beds promote f ilamentous algal proliferation by lowering velocities within the beds and providing substrate for attachment, or inhibit proliferation by creating higher turbulence above the beds, reducing light intensity within the beds, and supplying habitat for algal c onsumers. Conversely, it is likely that filamentous algae have contributed to the loss of SAV, since it has been shown that vascular macrophytes are often replaced by filamentous algae due to competition for light (Duarte 1995; Valiela et al. 1997). Loss of SAV could ultimately lead to less filamentous algal abundance (Frazer et al. 2006), since this algae appears to use SAV as a substrate, particularly in higher velocities.

PAGE 87

87 Management Implications Maintenance of flow velocities above identified thre sholds may be necessary to minimize filamentous algal proliferation and associated impacts in many lotic systems. Typically, variation in flow velocities over time is driven by fluctuations in overall discharge, which are largely related to rainfall patt erns. However, where consumptive use or other factors have lowered discharge, there are opportunities to ameliorate these effects by reducing water use and restoring landscapes to promote pre development hydrologic conditions. In addition, there may be m ore direct methods to prevent decreased flow velocities, such as minimizing dredging and damming activities which result in lower velocities. Intensive aquatic plant management may also be problematic as removal of substantial amounts of vegetation may al ter the distribution of flow velocities in the water column, which in addition to the effects of disturbance and increased light availability, could create conditions that favor a shift to algal dominance. Finally, while managers do not usually have direc t control over river discharge; in regulated rivers, controlled flow releases can be used to suppress or reduce filamentous algal blooms (Mitrovic et al. 2011 ). In Florida, many spring fed rivers have exhibited reduced discharge over recent decades due to a combination of less rainfall and increased groundwater pumping which, in turn, has lowered aquifer levels and associated springs discharge (Munch et al. 2006; Weber and Perry 2006; Copeland et al. 2009; Grubbs 2011). Until the late 1990s, water policy i n Florida focused more on water quality than quantity; however, dwindling water supplies have increased the attention to water quantity and spurred the implementation of minimum flows and levels (MFLs) for Florida water bodies (Munson et al. 2005). To dat e, MFLs have been based on factors such as fish passage, aquatic

PAGE 88

88 habitat preservation, and floodplain inundation, but MFLs could also account for filamentous algal proliferation by setting a minimum flow that maintains velocities above the thresholds ident ified in this study. Reducing nitrate loading to groundwater has been the primary focus for springs management to date, but while nutrient enrichment may increase algal growth rates, the flow conditions in these systems have been shown here to play an im portant role in controlling algal abundance. Nitrate concentrations and loading rates may have to be reduced substantially to inhibit algal proliferation, which is expected to require long time periods and may not reduce already established algal mats. W hile decreased flow velocities are likely to contribute to algal proliferation in Florida spring fed rivers, the importance of this factor in comparison to other changes such as nutrient enrichment and loss of algal consumers is currently unknown. The ne gative relationship between flow velocity and filamentous algae may, in part, explain why algal proliferation frequently occurs near spring vents (i.e. locations of discharge from the aquifer). Spring vent areas tend to resemble pools more than rivers, wi th lower flow velocities except near the vent itself. Downstream of vent areas, spring systems usually become more channelized with higher velocities, which might reduce the potential for algal accumulation. Nutrient enrichment may also have a greater ef fect where velocities are lower, since nutrient availability becomes more limited with increasing water residence time. Other factors that promote filamentous algae such as human recreation disturbance and loss of algal consumers may also have more influe nce in the vicinity of spring vents (DuToit 1979; Heffernan et al. 2010 b ).

PAGE 89

89 When examining the causes of algal proliferation in lotic systems, factors affecting both the growth of algae (e .g., nutrients, light) and the loss of algae (e.g., drag induced ex port, consumption) should be considered. A multi faceted approach would draw attention to the full suite of factors potentially controlling algal abundance in a given system. For Florida spring fed rivers, filamentous algal proliferation is likely a mult i causal response to the combined effects of lower flow velocities, less algal consumption, and increased nutrient concentrations, as well as increased disturbance in general. A management strategy that acknowledges and incorporates a broad suite of poten tial drivers of algal proliferation will provide additional options and techniques for maintaining the ecological integrity of Florida springs

PAGE 90

90 Figure 4 1. Aerial view of Gum Slough. The spring fed river begins in the northeast and flows towards th e southwest. The majority of the water comes from six large springs (circles), whose combined discharge rate is recorded at a USGS gage downstream (pentagon in the lower left). The three transects sampled every two months (solid lines) are shown.

PAGE 91

91 Fig ure 4 --) for the period of record ( October 2003 to December 2011 ) from the USGS gage downstream of the spring inputs. The study period is highlighted in gray. Figure 4 3. Picture and diagram of the experimental baffles. Circl es indicate the locations of flow velocity measurements and algae/SAV sampling.

PAGE 92

92 Figure 4 4. Changes in flow velocity and filamentous algal cover over time at three transects with different morphology.

PAGE 93

93 Figure 4 5. Filamentous algal cover in relation t o average and maximum flow velocity for individual sampling points across all transects. High Flow: n = 3; Med Flow: n = 6; Low Flow: n = 1; where n is the number of sampling dates in that flow category.

PAGE 94

94 Figure 4 6. Baffle experiment results. Filament ous algal biomass (as ash free dry mass (AFDM)) in relation to average and maximum flow velocity, and SAV biomass.

PAGE 95

95 Table 4 1. Average water chemistry for the main spring at Gum Slough during the study period (data from quarterly monitoring by the Southwes t Florida Water Management District). Temp (C) Sp. Cond (S/cm) Alkalinity (mg/L CaCO 3 ) pH DO (mg/L) NO 3 +NO 2 (mg N/L) Ortho P (mg P/L) Ca (mg/L) Cl (mg/L) Fl (mg/L) K (mg/L) Mg (mg/L) Na (mg/L) SO 4 (mg/L) 23.04 332 125.1 7.73 3.96 1.45 0.028 54.7 7.5 0.1 3 0.47 7.03 4.17 27.6 Table 4 2. Results of multiple linear regression models for each transect, using maximum flow velocity (v max) and the day of the year (day) to predict filamentous algal cover. Model Transect Parameter Value SE p R 2 p T1 Intercept 93.5 8.9 <0.001 0.41 0.156 v max 1.81 1.85 0.360 day 0.05 0.04 0.240 T2 Intercept 29.0 9.4 0.018 0.86 0.001 v max 1.68 0.31 0.001 day 0.25 0.04 <0.001 T3 Intercept 111.4 30.9 0.008 0.62 0.034 v max 4.33 1.31 0.013 day 0.13 0.07 0.093

PAGE 96

96 CHAPTER 5 EFFECTS OF FLOW VELO CITY AND NUTRIENTS O N THE FILAMENTOUS CYANOPHYTE LYNGBYA WOLLEI IN LABORATORY STREAM CH ANNEL S Introduction In lotic systems, hydraulic conditions are known to be a major determinant of autotrop hic structure and function (Biggs 1996; Stevenson 1996), but are usually not explicitly considered when assessing the conditions under which autotrophs might be nutrient limited. Nutrient availability (primarily N and P) often limits autotrophic productio n, particularly under conditions of low nutrient concentration and high light availability (Grimm and Fisher 1986; Hill and Knight 1988; Rosemond et al. 2000). However, there are many lotic systems which do not display nutrient limitation (Franceour 2001; Tank and Dodds 2003; Johnson et al. 2009) and nutrient concentrations do not correlate with algal biomass to the same degree as for lentic systems (Dodds et al. 2002 b ). Part of the difficulty in assessing nutrient limitation in lotic systems is due to th e influence of hydraulic conditions on benthic autotrophs. Many studies have found that flow velocity substantially affects algal nutrient uptake, production, and biomass (see Stevenson 1996 for a review); although the specific response varies greatly due to factors such as algal grazing (Poff and Ward 1995; Opsahl et al. 2003) and algal growth form (Biggs et al. 1998). This study examined the response of the filamentous cyanophyte Lyngbya wollei to nutrients under different flow velocities, to better und erstand how flow mediates nutrient availability and limitation for this species. Positive Effects of Flow Velocity on Algae Assessments of hydraulic conditions in lotic systems often focus on flow velocity (i.e. current), which may affect nutrient availabi lity to benthic autotrophs by regulating

PAGE 97

97 the thickness of diffusive boundary layer s (DBL s ). A DBL is a stagnant layer of water that surrounds submerged surfaces in aquatic environments, and may limit the mass transfer of nutrients and other dissolved ions and gases to cell surfaces. A DBL becomes thinner with increasing flow velocity, thus increasing the diffusive mass transfer rate of nutrients from the water column to cellular uptake sites (Jorgensen and Revsbech 1985; Gundersen and Jorgensen 1990; Hurd 2000). As flow velocity increases, DBL s may become thin enough to no longer limit nutrient transfer, and instead cellular uptake rates become the limiting factor (Sanford and Crawford 2000; Nishihara and Ackerman 2009). In addition, flow velocity has re cently been found to enhance waste efflux (such as oxygen) away from autotroph cells with a corresponding increase in production rates (Mass et al. 2010). This latter mechanism could lead to a positive relationship between flow velocity and algae independ ent of nutrient availability. The effects of flow velocity on DBL s and waste efflux could have significant implications for assessing nutrient limitation in lotic systems. Filamentous A lgae The proliferation of filamentous algal mats is a growing proble m globally, and often results in degraded ecosystem structure and function as well as diminished aesthetics and recreational opportunities (Lembi 2003; Smith 2003 ). Flow velocity can regulate nutrient transfer to algal mats as previously described, and ad ditionally stimulates nutrient flux into the mats (Stevenson and Glover 1993). Internal nutrient cycling within filamentous algal mats can be substantial, although its significance as a nutrient source is dependent on external nutrient concentration (Mulh olland et al. 1994; Sickman et al. 2009). In this study, portions of filamentous algal mats (i.e. clumps) were used, rather

PAGE 98

98 than single filaments or a thin biofilm, since the goal was to determine the response of algal mats to flow velocity and nutrients. The test organism in this study was the filamentous cyanophyte Lyngbya wollei which has received much attention globally due to its tendency to form prolific floating mats of high biomass and its potential to generate neurotoxins (Carmichael et al. 1997). Lyngbya often forms dense mats on the benthic surface of freshwater systems which may detach from the benthos and become floating mats due to buoyancy from gas bubbles. Lyngbya is shade adapted with a low light compensation point of 32 mol/m 2 /s (Spezia le et al. 1991) and has a thick sheath which may make it unpalatable to most algal grazers (Camacho and Thacker 2006). Although this study is focused on a single filamentous algal species, Lyngbya wollei the general principles involved are presumed to ap ply to o ther mat forming species Lyngby a wollei is one of the primary algal species currently proliferating in Florida springs (Stevenson et al. 2004) and may actually consist of two or more species based on molecular and morphological data (Joyner et al. 2008). Attempts to find a relationship between Lyngbya and nutrient (N and P) concentrations within Florida springs have been largely inconclusive (Cowell and Botts 1994; Stevenson et al. 2004) and most laboratory experiments which found a relationship w ere conducted in relatively static hydraulic conditions not representative of lotic systems (Cowell and Dawes 2004; Stevenson et al. 2007). One laboratory study using recirculating channels found some evidence that nitrate concentration was related to Lyn gbya growth rates, although the threshold which would saturate Lyngbya growth was relatively low (< 110 g N/L) in comparison to the levels currently found in most springs (Stevenson et al.

PAGE 99

99 2007). The uncertainty associated with the nitrate relationship m ay be partially due to Lyngbya being a nonheterocystous cyanobacterium which has the ability to fix atmospheric nitrogen (N 2 ) under certain conditions (Phlips et al. 1992). An alternative hypothesis tested in this study is that flow velocity is a signific ant determinant of Lyngbya metabolism and biomass. This study tested the effects of flow velocity on clumps of the filamentous alga Lyngbya wollei at different nutrient concentrations. Algal response was assessed based on metabolic rates derived from di el changes in pH and DO, and changes in algal biomass over time. The first experiment examined algal response to flow velocity at high (1.00 mg N/L) and low (< 0.05 mg N/L) nitrate concentrations. It was hypothesized that algal metabolism and biomass wou ld increase with flow velocity at low nitrate concentration; but would be relatively similar among velocities at the high nitrate concentration due to nutrient saturation. A second experiment involved a series of nutrient additions, beginning with amended deionized (DI) water, followed by P and micronutrient addition, and finally N addition. For this experiment it was expected that algal metabolism and biomass would increase with each nutrient addition. The combined results of both experiments were evalu ated to determine how flow velocity influences the metabolism of filamentous algal clumps in a variety of nutrient conditions ranging from severe nutrient limitation to nutrient saturation. Methods Stream C hannel D esign T hree laboratory stream channels w ere constructed out of a 3 m section of 15 cm diameter PVC pipe with the top quarter cut out, creating a channel open to light and gas exchange (Figure 5 1). A 1.9 cm diameter PVC pipe was inserted perpendicularly

PAGE 100

100 within the channel near the inlet and out let in order to brace the pipe open, as the pipe would slightly constrict with the top cut out. PVC caps were fitted to the ends of each channel with a 3.2 cm hole cut near the top of the upstream cap for the inlet pipe and a 5 cm hole cut near the bottom of the downstream cap for the outlet pipe. Each channel was connected to separate 190 L plastic tanks from which the water was recirculated to the channel inlet using a pump. Each pump moved water to the channel inlet through a short section of flexibl e plastic tubing attached to 3.2 cm diameter PVC pipe configured to enter through the hole in the upstream end cap. Near the inlet, 1.9 cm diameter PVC was fitted into the channel perpendicular to the flow direction in order to diffuse the inflowing water thus generating more uniform flow through the channel. A 5 cm diameter PVC outlet pipe was fitted into the hole in the channel outlet cap and ran directly into the plastic water tank. The outflow rate and channel water level were controlled with a gate valve on the outlet pipe. This configuration resulted in three identical artificial stream channels with adjustable flow rates (as dictated by pump choice). The three pumps were rated at 6.3, 31.5, and 75.7 L/min and generated average flow velocities of 1, 5, and 10 cm/s, respectively, with a water depth of 10 cm. Each channel contained a rack to hold the filamentous algal clumps in place, which was made using 2 m of the cut out top section of the PVC channel into which 1 cm diameter wooden dowels (made from poplar) were inserted perpendicular to the water flow. This created a substrate for the algae similar to macrophyte stems on which they were observed to grow in the spring fed river where collected. Fine mesh was placed over the channel outlets in o rder to capture export of filamentous algae. Coarse

PAGE 101

101 mesh (2 cm x 2 cm spacing) was placed just below the water surface to prevent the algae from floating to the surface, which ensured that gas exchange resulting from algal metabolism occurred in the water and not directly with the atmosphere. Experimental C onditions The experiments were conducted within an air conditioned greenhouse laboratory at the University of Florida in Gainesville. Photosynthetically active radiation (PAR) was maintained at < 700 mol/m 2 /s (with shade cloth if necessary), since Lyngbya wolle i is light saturated at < 664 mol/m 2 /s (Speziale et al. 1991) and an unidentified small green alga occasionally became abundant when light levels were higher. PAR was measured using a LI COR m eter (LI 1400 model) with a PAR sensor and averaged 7.42 mol/m 2 /d (SD: 2.84 mol/m 2 /d). Water temperature was recorded every 15 min with a YSI 556 MPS, and submersible heaters were used to help maintain the desired water temperature of 24 C ( 6 C). Tem perature varied slightly among the channels ( < 1 C ) due to differential heat from the pumps. Flow velocity was measured in the middle of the water column within each channel downstream of the algae using a Marsh McBirney Inc. Flo Mate (Model 2000) port able flow meter. The three flow rates were randomly assigned to each channel prior to each trial and the velocity in the center of the channel averaged either 1 cm/s (low v), 5 cm/s (med v), or 10 cm/s (high v). Two experiments were conducted which diffe red in their duration, water source, and nutrient conditions. Both experiments utilized the three laboratory stream channels which were identical except for the flow velocity (1, 5, and 10 cm/s). The first experiment tested two nitrate concentrations (1. 0 and < 0.05 mg/L) using amended tap water, and consisted of five short (5 6 day) trial runs at each nitrate concentration. The second experiment tested the effect of sequential nutrient additions over a 19 day

PAGE 102

102 period, beginning with amended deionized wat er without N and P, addition of P and micronutrients after six days, and addition of N after twelve days. Prior to each trial, all channels and tanks were scrubbed clean, filled with tap water, dosed with 400 mL of bleach, and circulated for 48 hrs. For t he first experiment, the water was then replaced with fresh tap water and circulated for 48 hrs in order for residual chlorine to diffuse out of the water and for gas equilibration with the atmosphere, as initially CO 2 and O 2 were undersaturated. Tap wate r was analyzed for cations and anions by ion chromatography (Dionex ICS 3000) (Table 5 1 ). Specific conductivity ranged from 100 to 400 S/cm. An in situ submersible ultraviolet nitrate analyzer (SUNA; Satlantic) was placed in the tank of the high flow c hannel for multiple trials and recorded nitrate concentration every 15 min. Tap water initially contained some nitrate (0.02 0.05 mg N/L) and NaNO 3 was added to generate the higher NO 3 concentration tested (1.0 mg N/L). In each trial, phosphorus was elev ated to 1.0 mg P/L using KH 2 PO 4 and this concentration was assumed to saturate biotic P demand. These were the initial N and P target concentrations and there was no effort to maintain concentrations due to the short duration of each trial. The second e xperiment used deionized water instead of tap water and consisted of a series of nutrient additions. Initially each tank was filled with 150 L of DI water and amended with 10 g CaCl 2 1 g MgSO 4 and 8 g of NaHCO 3 in order to raise conductivity, pH, and a lkalinity to levels resembling the tap water. After six days, 10 mL of a micronutrient solution (Brightwell Aquatics FlorinMulti) was added to each tank and phosphorus was elevated to 1.0 mg P/L using KH 2 PO 4 After six more days, nitrogen

PAGE 103

103 was added for t he first time using NaNO 3 to generate 1.0 mg N/L. Prior to each nutrient addition, the tanks were refilled to 150 L with DI water to replenish evaporative loss. Algal C ollection and P rocessing Samples of fresh filamentous algal mats dominated by the cyano phyte Lyngbya wollei were collected in the vicinity of the main spring vent at the Gum Slough spring system in Sumter County, Florida, USA. The algae samples were placed in plastic containers and transported on ice back to the greenhouse. The algae were washed with deionized water and all macroscopic objects such as leaves and invertebrates were removed. The algae samples were then acclimated for at least two days in a tank which contained water with the same nutrient concentrations and chemistry as the intended trial. To initiate a trial, algae samples were spun dry and separated into 10 g wet weight ( 1 g) clumps, and seven clumps were added to each channel (six clumps in the DI water experiment). At the end of each trial the algae was collected, spu n dry, and weighed to determine the change in wet weight over the course of the trial. Metabolism M easurements T o measure the filamentous algal response to each combination of flow velocity and nutrients, diel pH changes were recorded and converted into CO 2 changes using the equilibrium equation for total CO 2 and alkalinity measurements. Downstream of the algae in each channel, pH was measured every minute using a pH meter (Artisan PH2000), equipped with a double junction pH probe (Oakton) and an automat ic temperature compensation (ATC) probe. These pH meters were calibrated by three point calibration prior to each trial, and when checked for drift at the end of each trial were always within 0.08 pH units. Carbonate alkalinity ( CA ; meq/L) was measured e very 1 2 days in each channel by titrating a 100 mL sample with 0.2 N H 2 SO 4 until the

PAGE 104

104 pH was 4.5. Initial carbonate alkalinity varied from 0.59 to 0.80 meq/L in tap water and was 0.5 meq/L in the amended DI water. Alkalinity increased linearly over the c ourse of each trial, so a linear relationship was developed for each channel to account for changing alkalinity over time. Hourly averages of the pH were converted to total CO 2 and [CO 2 ] (i.e. free CO 2 ) using the following equilibrium equations modified to include carbonate alkalinity: w here { H + } is the activity of the hydrogen ion derived from the pH, and K 1 and K 2 are equilibrium coefficients (from Stumm and Morgan 1996: pK 1 = 6.35 at 25C and pK 2 = 10.33 at 25C; note that these K values are given as pK). The contribution of CO 2 reaeration (i.e. atmospheric gas exchange) to changes in total CO 2 was determined by setting the clean channels at the target velocities (1, 5, or 10 cm/s) without algae and measuring the change in pH in fresh tap water (initi ally undersaturated with CO 2 ) for five trials. An additional trial was conducted for the amended deionized water prior to adding algae. The results of these trials were used to calculate an average CO 2 reaeration coefficient ( k CO2 ; m/hr) for each of the three target velocities. It was assumed that the change in total CO 2 was due only to CO 2 diffusion into and out of the water in each channel; therefore CO 2 reaeration rates ( D CO2 ; mmol/L/hr) were estimated by adjusting k CO2 to best match the measured tota l CO 2 change. D CO2 was estimated using the following model:

PAGE 105

105 where A is the channel surface area (m 2 ), V is the total system volume (m 3 ), K H is the 2 (mmol/L/atm), and p CO2 is the partial pressure of CO 2 in the atmosphere (3.91E 4 atm from NOAA/ESRL ( www.esrl.noaa.gov/gmd/ccgg/trends) ). The effect of temperature variation on K H Hoff equation: where K H ,T C is an enthalpy coefficient (2400 K), T is the water temperature (K), and T is standard temperature (298 K). In order to match the observed reaeration rates, it was necessary to multiply the temperature adjusted K H values by a constant amount ( C 2 ) depending on the channel velocity. It is currently unknown what caused the deviation in K H from typical values, but this effect was consistent and may be related to conditions in the greenhouse or water chemistry. For the trials with algae, the CO 2 reaeration rate was calculated on an hourly basis and subtracted from the total CO 2 change. This diffusion corrected total CO 2 change was then used to calculate the rate of net ecosystem production pe r algal biomass (NEP; mg C/g biomass/hr). Algal biomass was assumed to change linearly over the course of each trial based on the initial and final weights. Daily NEP (mg C/g biomass/d) was determined by summing the hourly NEP for each day. Daily commun ity respiration (CR) was estimated by averaging the hourly NEP values over the previous night, and assuming that this average hourly va lue applied throughout the day. Daily CR was then added to NEP to determine gross primary production (GPP) on a daily ba sis due to the filamentous algae.

PAGE 106

106 Diel changes in dissolved oxygen (DO) were also used to estimate filamentous algal metabolism using a similar methodology to the pH changes just described. Data sondes equipped with optical DO probes ( YSI 6 600) were pla ced in the tank of each channel where they recorded DO concentration, DO saturation, specific conductivity, and temperature every 15 min. Oxygen reaeration was estimated for each velocity based on changes in DO saturation in the clean channels without alg ae. Metabolism results were calculated from diel changes in DO concentration corrected for reaeration, as described for pH. Statistical Analysis Daily values of NEP, CR, and GPP were averaged for each trial. Analysis of variance (ANOVA) was used to test for the separate effects of nitrate concentration and flow velocity on the overall change in algal biomass and daily average NEP, CR, and GPP. For significant ANOVA models, post (HSD) was used to examine the signi ficance of differences between means of each level (e.g. velocity). All statistical analyses were conducted using the software package R version 2.11.1 (R Foundation for Statistical Computing). Results CO 2 and DO Reaeration For the five reaeration trial s in tap water without algae, initially the fresh water was undersaturated with CO 2 and therefore had an elevated pH that declined with time at a rate dependent on the channel velocity. The time required to reach CO 2 equilibrium was approximately 1 day at 10 cm/s, 2 days at 5 cm/s, and 4 days at 1 cm/s (Figure 5 2). After reaching CO 2 equilibrium, total CO 2 varied slightly due to temperature variation, and the entire trial period was used to calculate the k CO2 and C 2 coefficients for each

PAGE 107

107 velocity for eac h of the five trials in tap water (Figure 5 3; Table 5 2). For the experiment in deionized water, the same k CO2 values were used, but C 2 was adjusted to 1.80 for all three velocities. Dissolved oxygen was also undersaturated initially in the tap water b ut equilibrated much faster than CO 2 due to reaeration rates which were approximately two orders of magnitude greater. Due to the large influence of reaeration on diel DO changes, DO was deemed unsuitable for algal metabolism calculations using these labo ratory stream channels. Algal Biomass and Metabolism For the experiments with filamentous algae, diel pH changes were dictated by a combination of CO 2 reaeration and biotic metabolism associated with the filamentous algal clumps (Figure 5 4). In the t rials of the first experiment pH ranged from 7.6 to 9.0 (usually not exceeding 8.6), but reached 9.4 towards the end of the longer experiment. In the low velocity channel, pH routinely spiked which appeared to be related to algal metabolism rather than po or mixing or measurement error. Generally, the amplitude of the diel pH changes was inversely related to flow velocity, which illustrated that flow velocity increased CO 2 reaeration and kept the higher velocity channel closer to CO 2 equilibrium with the a tmosphere. Water in the high velocity channel routinely reached CO 2 equilibrium with the atmosphere at night whereas the two channels with lower velocity did not completely equilibrate during each trial. Despite the apparent influence of CO 2 reaeration o n pH changes, changes in total CO 2 were mostly due to biotic production and respiration rather than reaeration (Figure 5 5). In the first experiment, biomass of the filamentous algal clumps declined in almost all of the trials and was not significantly aff ected by either nitrate concentration or flow velocity; although the largest average decrease was at high velocity (Table 5 3). Algal

PAGE 108

108 export was found to be minimal over the course of a trial (< 2 g wet weight), but slightly increased with flow velocity. Net ecosystem production was negative for most trials and was significantly lower with high nitrate concentration ( p = 0.0008), but not strongly affected by flow velocity. Both flow velocity and nitrate concentration affected CR and GPP of the filamentou s algal clumps (Figure 5 6). At low nitrate concentration, CR and GPP were much higher at a flow velocity of 5 cm/s than 1 cm/s, whereas at 10 cm/s both were intermediate and highly variable. At high nitrate concentration, CR and GPP were similar at 1 an d 5 cm/s, but much lower at 10 cm/s. The only significant flow velocity relationship was for GPP at high nitrate concentration ( p = 0.028), which confirmed that at 10 cm/s GPP was lower than at 1 and 5 cm/s. Nitrate concentration declined substantially o ver the duration of the high nitrate trials and leveled at 0.5 mg/L; in contrast nitrate concentration was relatively stable in the low nitrate trials (Figure 5 7). In the second experiment, the filamentous algal clumps gained substantial biomass over the 19 day duration although the timing of growth was highly variable between the three flow velocities (Table 5 4). Algal biomass in the low velocity channel increased from 60 to 80 g wet weight over the first six days without any N, P, or micronutrient addi tion, but then declined slightly once these nutrients were added. Algal biomass in the medium velocity channel increased by 11.7 g over the first six days, and then remained steady until growing another 11.1 g once N was added. Algal biomass in the high velocity channel was steady except for an increase of 11.8 g during the six day period after P was added. Overall, metabolism of the filamentous algal clumps increased with time and nutrient additions as shown by the amplitude of the diel pH and NEP chan ges over the

PAGE 109

109 19 day experiment (Figures 5 8 and 5 9). Net ecosystem production was uniformly negative over the first six days in the amended DI water, despite the steady or increasing amounts of algal biomass over the same time period. Once nutrients wer e added then NEP increased in each channel, and was positive after N addition. In all channels, CR and GPP increased with nutrient additions over the course of the experiment, except for CR in the medium velocity channel which oscillated around an average value of 1.72 mg C/g biomass/d. CR and GPP were highest in the medium velocity channel initially, but by the end of the experiment the low and high velocity channels were greater. On day 16 of this experiment, pH and NEP increased substantially which co incided with the end of a storm system that lowered the light levels for the first few days after N addition. Nitrate addition also stimulated the metabolism of organisms not associated with the filamentous algal clumps as the water in all channels became visibly green and cloudy, and a biofilm was apparent on most surfaces exposed to light. Discussion Algal Metabolism Response to Flow Velocity and Nutrients The two experiments in this study indicate that flow velocity and nutrient concentration influence the metabolism of filamentous algal mats dominated by Lyngbya wollei In the first experiment, higher flow velocity appeared to stimulate algal metabolism in the low nitrate concentration trials, whereas at high nitrate concentration the higher velocity h ad an inhibitory effect. The positive effect of flow velocity on algal metabolism at low nitrate concentration only, suggests that velocity increased N availability and therefore decreased N limitation. This effect has been observed in other similar stud ies (Whitford and Schumacher 1961, 1964; Lock and John 1979; Horner et

PAGE 110

11 0 al. 1990) and indicates that flow velocity can be an important driver of algal metabolism, particularly when nutrient availability is limited. Additionally, the lack of a positive effe ct of flow velocity at high nitrate concentration suggests that the main effect of flow velocity was to increase nutrient availability rather than waste efflux (e.g. Mass et al. 2010). The decrease in filamentous algal metabolism at high flow velocity in t he high nitrate concentration trials is more difficult to explain, particularly since a similar effect was not observed in the deionized water experiment. With regard to nutrient availability, N saturation was expected due to the high initial nitrate conc entration (1 mg N/L). If N is saturating then the stimulatory effect of flow velocity on nutrient availability would not be observed, and algal metabolism should remain relatively constant across this range of flow velocities. Nitrogen saturation was lik ely based on the similar magnitude of CR and GPP at low and medium flow velocity at high nitrate concentration, which were also comparable to CR and GPP at high and medium flow velocity at low nitrate concentration. One possibility is that the high veloci ty (10 cm/s) negatively affected algal metabolism as occurred in a similar study (Borchardt et al. 1994); although the specific mechanism is unknown. This unknown mechanism may also explain why at low nitrate concentration algal metabolism at high flow ve locity was highly variable and slightly lower than at medium velocity. An alternative explanation for the decrease in algal metabolism at high flow velocity and high nitrate concentration is that another organism interfered with the filamentous algae in th ese conditions. Two observations suggest that a species of bacteria with high N demand thrived in the high nitrate concentration trials. First, nitrate concentrations declined asymptotically to 0.5 mg N/L and this could not be accounted

PAGE 111

111 for by algal upta ke alone since there was no net accrual of algal biomass and nitrate declined regardless of light availability (i.e. at night). Second, CR was higher and NEP was much more negative in the high nitrate trials, even though GPP was relatively similar for bot h nitrate concentrations. Denitrification is suspected to be the mechanism of nitrate loss, which may seem unlikely in these well oxygenated stream channels but has been shown to occur within filamentous algal mats if nitrate concentration is sufficiently high ( Triska and Oremland 1981; Duff et al. 1984). It is possible that denitrifying bacteria negatively affected filamentous algal metabolism in the higher flow velocity when nitrate concentration was elevated above 0.5 mg N/L. Algal Biomass Response t o Flow Velocity and Nutrients In this study, the effect of flow velocity and nutrients on filamentous algal biomass is unclear. In the first experiment algal biomass declined in almost all trials, largely independent of flow velocity or nitrate concentrat ion. However in the second experiment algal biomass increased in all velocities, and the influence of nutrient additions was highly unpredictable. The largest increase in algal biomass (20 g wet weight) occurred over the firs t six days at low flow veloc ity in the amended deionized water without any addition of N, P, or the micronutrient solution. Overall there was no clear trend in biomass changes related to nutrient additions, even though the additions did seem to consistently stimulate algal metaboli sm. The decrease in algal biomass in the first experiment has many possible explanations. Algal export from the channel could only account for a small fraction of this decline, which indicates that most of this loss was due to mortality or consumption. S mall invertebrates were observed in the algal clumps, and consisted mostly of amphipods which are able to consume Lyngbya though they do not prefer it (Camacho

PAGE 112

112 and Thacker 2006). The influence of these invertebrates on algal consumption is unknown but exp ected to be minimal as most were removed prior to each trial. Both export and consumption are likely minor sources of biomass loss, since these would have occurred in the deionized water experiment as well. A likely explanation for the loss of algal biom ass in the first experiment is that the chemistry of the tap water was unsuitable, which caused the algae to slowly decay. There are two possible sources of negative water chemistry effects: the presence of a toxin such as chlorine or copper, or the chemi stry was simply not appropriate for Lyngbya Tap water is typically chlorinated, but no chlorine was detected in the fresh tap water and the water was aerated for 48 hours prior to adding algae. Overall the presence of a toxin seems unlikely since algal production was measurable, and the algae persisted in other tanks with tap water for over a year. The chemistry of the tap water was comparable to that of most Florida springs and originated from the same source (the Floridan aquifer), but still may not h ave been ideal for Lyngbya Since the algae did accrue biomass in the deionized water experiment, there was apparently a key difference in water chemistry. Although the deionized water was amended to resemble the tap water, the calcium concentration was lower ( 16 mg/L versus 30 mg/L ) and bicarbonate was added directly. Calcium can affect Lyngbya growth over this range of concentrations, but the effect may depend on poorly understood interactions with N and P (Cowell and Botts 1994). Lyngbya is known to prefer bicarbonate as a carbon source (Beer et al. 1992); however it was expected that bicarbonate was the dominant form of carbonate in the tap water as well since bicarbonate dominates in the pH range of 6.3 to 10.3. There are likely other relevant

PAGE 113

113 diff erences in water chemistry such as iron (Ahern et al. 2006), but without further experimentation it is difficult to specify the reason why algal biomass increased in deionized water and not tap water. Determining Nutrient Response of Attached Algae A comm on goal of many laboratory experiments with algae is to determine algal response over a range of nutrient concentrations, which can help identify nutrient criteria that inhibit algal proliferation. The results of this study do not allow a specification of the range of N concentrations that stimulate Lyngbya wollei H owever, there is indication that N demand is saturated at relatively low concentrations as shown by the similar production rates in the low and high nitrate concentrations tested. Additionall y, while there was less algal production in low velocity and low nitrate concentration, a modest velocity increase to 5 cm/s appeared to generate the maximum production rate. Other studies which have examined the relationship between Lyngbya and nitrate c oncentration did not explicitly account for the influence of flow velocity, which may explain why their results differed. One Lyngbya experiment used small flasks in which the water was continuously replenished but with little flow velocity, and found tha t growth rates were stimulated when nitrate increased from 0.3 to 0.6 mg N/L (Cowell and Dawes 2004). In contrast, a study that used semi recirculating channels with a velocity of 25 cm/s found that Lyngbya growth rates were N saturated at a much lower ni trate concentration of 0.11 mg N/L (Stevenson et al. 2007). A likely explanation for the different saturating nitrate concentrations is that the higher velocity increased nutrient availability, and therefore lower nitrate concentrations were required to s aturate N demand and inhibit N limitation. The substantial effect of flow velocity on algae in general (Biggs 1996; Stevenson 1996) implies that experiments meant to assess the

PAGE 114

114 range of nutrient concentrations that stimulate a particular algal species or community in a lotic environment must consider hydraulic conditions. In conclusion, flow velocity and nutrient concentration interact to determine nutrient availability to the filamentous cyanophyte Lyngbya wollei but flow velocity may also negatively a ffect this species in other ways. Since flow velocity can increase nutrient availability and decrease nutrient limitation for filamentous algae, flow velocity should be considered in conjunction with the nutrient concentration when assessing the potential for algal nutrient limitation. Laboratory stream channel experiments may offer a controlled approach to determining nutrient limitation thresholds for lotic algae by examining the species specific relationship with flow velocity in conjunction with nutri ent concentrations, as different species have been shown to have wide variation in responses. The degree to which flow affects nutrient availability deserves further attention, since flow may confound the use of nutrient concentration alone as a predictor of nutrient limitation in lotic systems.

PAGE 115

115 Figure 5 1. Diagram of the experimental laboratory stream channels.

PAGE 116

116 Figure 5 2. Initial pH change in fresh tap water at the three flow velocities due to CO 2 reaeration.

PAGE 117

117 Figure 5 3. CO 2 reaeration r ate estimation based on total CO 2 change in fresh tap water from 9/19/11 to 9/22/11.

PAGE 118

118 Figure 5 4. Example of diel pH changes for each velocity from the trial beginning on 1/19/12

PAGE 119

119 Figure 5 5. Example 2 ], reaeration rate, and NEP for each velocity from a high nitrate trial.

PAGE 120

120 Figure 5 6. Boxplot comparison of CR and GPP results from the experiment testing flow velocity effects at low and high nitrate concentration.

PAGE 121

121 Figure 5 7. Nitrat e concentration change over time at high flow velocity for both low and high initial nitrate concentration.

PAGE 122

122 Figure 5 8. Diel pH changes in the 19 day experiment with sequential nutrient additions.

PAGE 123

123 Figure 5 9. Diel NEP changes in the 19 day experiment with sequential nutrient additions.

PAGE 124

124 Table 5 1. Chemistry of fresh tap water prior to nutrient addition. BDL below detection limit. Cations mg/L Anions mg/L Ca 30.37 Br BDL K 1.29 Cl 23.3 Li BDL I BDL Mg 18.28 SO 4 62.6 Na 10.79 NH 4 BDL Table 5 2. CO 2 reaeration coefficients estimated from the five trials without algae. k CO2 (m/hr) C 2 Low v Med v High v Low v Med v High v Average 1.15E 02 2.53E 02 3.76E 02 1.55 1.56 1.50 Max 1.40E 02 4.20E 02 4.71E 02 1.77 1.85 1.67 Min 6 .72E 03 1.52E 02 2.81E 02 1.30 1.35 1.29

PAGE 125

125 Table 5 3. Algal biomass and metabolism results from the experimental trials comparing the effect of flow velocity at high and low nitrate concentration. wet wt.) NEP avg (mg C/g biomass/d) CR avg (mg C/g biomass/d) GPP avg (mg C/g biomass/d) Trial Low v Med v High v Low v Med v High v Low v Med v High v Low v Med v High v High N1 0.2 18.1 14.3 0.14 0.16 0.12 1.18 1.31 0.76 1.05 1.16 0.65 High N2 5.5 6.0 11.1 0.54 1.03 0.90 1.28 2.00 1.06 0.74 0.97 0.22 High N3 4.7 0.0 4.0 0.56 0.44 0.59 1.21 1.33 1.38 0.65 0.89 0.78 High N4 28.9 12.4 21.9 0.38 0.52 0.47 1.77 1.11 0.86 1.39 0.59 0.38 High N5 16.9 18.5 19.0 0.36 0.46 0.48 1.51 1.46 1.13 1.15 1.00 0.65 Average 11.2 8.6 14.1 0.40 0.52 0.51 1.39 1.44 1.04 1.00 0.92 0.54 SD 11.6 11.0 7.0 0.17 0.32 0.28 0.25 0.34 0.24 0.30 0.21 0.23 Low N1 5.7 14.9 9.0 0.19 0.33 0.52 0.92 0.81 1.05 0.72 0 .48 0.53 Low N2 10.2 17.7 22.4 0.19 0.25 0.28 0.99 1.44 0.85 0.81 1.19 0.57 Low N3 15.4 12.0 14.0 0.05 0.23 0.68 0.89 1.74 0.71 0.85 1.51 1.39 Low N4 22.9 19.7 20.6 0.13 0.33 0.23 0.94 1.00 1.54 1.07 1.33 1.31 Low N5 8.3 5.7 9.9 0.08 0.02 0.39 0.60 1.38 1.47 0.68 1.40 1.07 Average 12.5 14.0 15.2 0.04 0.09 0.15 0.87 1.27 1.12 0.83 1.18 0.97 SD 6.8 5.5 6.1 0.15 0.27 0.48 0.15 0.37 0.37 0.15 0.41 0.41 Table 5 4. Algal biomass and metabolism results from the 19 day experiment with a series of nutrient additions. wet wt.) NEP avg (mg C/g biomass/d) CR avg (mg C/g biomass/d) GPP avg (mg C/g biomass/d) Days Nutrients Low v Med v High v Low v Med v High v Low v Med v High v Low v Med v High v 0 6 ----20.0 11 .7 0.6 0.57 0.73 0.60 1.31 1.92 1.46 0.74 1.19 0.86 6 12 P, micro 5.9 0.3 11.8 0.21 0.06 0.08 1.97 1.46 1.94 1.75 1.52 2.01 12 19 N, P, micro 3.8 11.1 0.6 0.33 0.64 0.69 2.85 1.77 2.49 3.18 2.41 3.19

PAGE 126

126 CHAPTER 6 A SIMULATION MODEL F OR FLOW AND NUTRIENT EFFECTS ON FILAMENTOUS ALGAE AND BACTERIA IN LABORATORY STREAM CHANNELS Introduction In lotic systems, two of the primary factors which control attached benthic algae are nutrients and flow velocity (Borchardt 1996; Stevenson 1996). Algal produc tion (i.e. growth) may be stimulated by increased nutrient concentrations or flow velocity due to greater nutrient availability, until nutrient demand is saturated and further increases no longer have a stimulatory effect. Flow velocity can also negativel y affect algal colonization and biomass accrual due to drag, and therefore an intermediate flow velocity is usually optimal which varies by algal species ( Biggs 1996 ). When managing aquatic systems, a common goal is to minimize nutrient enrichment so tha t algae are not overly stimulated and the formation of large colonies (i.e. blooms or mats) is inhibited. This approach often relies on a target nutrient concentration that if not exceeded will prevent the overstimulation of algae and maintain the integri ty of the aquatic ecosystem. This approach has proven useful when managing algae in lakes and slow moving rivers in which free floating phytoplankton is the dominant form of algae; however it has been more difficult to specify a relationship between nutr ient concentration and attached benthic algae found in most streams and rivers (Dodds et al. 2002b). This is partially due to the effect of flow velocity on both nutrient availability and drag, and therefore it is necessary to consider nutrient algae rela tionships in the context of flow conditions. The simulation model described here specifies the effects of nitrate concentration and flow velocity on the filamentous cyanophyte Lyngbya wollei The relationships simulated were derived from an experiment usi ng laboratory stream channels in which

PAGE 127

127 nitrate concentration and flow velocity were manipulated to test the response of algal metabolism and biomass (see Chapter 5). The simulation model was used to help explain the experimental results by examining multi ple hypotheses regarding the mechanistic relationships between the variables measured. Methods The simulation model was designed to determine the potential mechanisms that could lead to the observed changes in algal biomass and metabolism in a laboratory stream channel experiment. Metabolism was measured as gross primary production (GPP), community respiration (CR), and their difference, net ecosystem production (NEP). The experiment consisted of six scenarios: two nitrate concentrations (0.05 and 1.0 m g N/L) at three flow velocities (1, 5, and 10 cm/s). The overall performance of the model was judged on its ability to recreate five observed results for each scenario: overall change in algal biomass and nitrate concentration, and average daily GPP, CR, and NEP. After multiple rounds of testing different models, only one model was able to adequately approximate the experimental results (Figure 6 1). The model explicitly involves two organisms, filamentous algae and bacteria, whose nitrate uptake rates follow a Michaelis Menton saturation function. During the simulation, a portion of the algae biomass senesces and goes to a detritus storage from which remineralized nitrate flows back into the nitrate storage, completing a closed cycle. At high nitrate concentrations (> 0.5 mg N/L), the bacteria becomes active and converts nitrate into nitrogen gas through the process of denitrification, which causes removal of nitrate from the system. A unique aspect of this model is that nitrate availability to the al gae

PAGE 128

128 increases with flow velocity, but flow velocity also negatively affects algal production, and therefore peak algal production occurs at an intermediate velocity. The simulation model was created in the energy circuit language (Odum 1994), which can b e used for any systems containing storages of material (or energy) and the flows between them. The model consists of sets of differential equations that represent changes in the quantity of stored material over time (Tables 6 1 and 6 2). The model was ru n with a one minute time step for five days, and was calibrated by adjusting coefficients in the model to match the results of a laboratory stream channel experiment (Table 6 2). The estimated turnover time of each storage was used to constrain the coeffi cients and aid calibration, by reducing the number of unknown variables. Light (based on photosynthetically active radiation (PAR)) was modeled with a sin function that initiated at 7am, peaked at 270 mol/m 2 /s at 1pm, and ended at 7pm. Metabolism values were summed up for each day of the simulation model and divided by the average algal biomass for that day. This resulted in five daily average metabolism values on a biomass basis (mg C/g biomass/d), which were then averaged to yield average daily metabo lism for each scenario, comparable to the observed experimental values. Initially the model was run at steady state to ensure that it was parameterized correctly (i.e. steady state was achieved). The steady state simulation was setup for the base conditi ons of low flow velocity (1 cm/s) and low nitrate concentration (0.05 mg N/L). Other initial conditions were set based on measured experimental conditions and estimated turnover times of each storage (Table 6 1). To use the steady state approach when a s torage had multiple inflows or outflows, the flows were partitioned. This was

PAGE 129

129 necessary for the two outflows from nitrate to algae and bacteria. At low nitrate concentration, only the algae was involved in nitrate uptake, whereas at high nitrate concentr ation (1.0 mg N/L) uptake was split evenly between the algae and bacteria. After checking for steady state, it was necessary to increase the mortality rate of the algae to match the observed results since the actual system was not in steady state initiall y Results The simulation model was able to approximate the results observed in the laboratory stream channel experiment for the six combinations of nitrate concentration and flow velocity tested (Table 6 3). Only for a small range of nitrate uptake and v elocity coefficients was the model able to reproduce the peak in algal production observed at low nitrate concentration and 5 cm/s flow velocity (Figure 6 2). For the modeled experimental conditions, the highest algal production occurred at 3 cm/s flow ve locity when nitrate concentration was 0.05 mg/L, but when nitrate was readily available at 1.00 mg/L velocity only had a negative effect. Similarly, algae responded to increased nitrate concentration at low velocity, but at higher velocity nitrate uptake was constant with a lower saturation concentration due to the inhibitory effect of velocity (Figure 6 3). Bacteria required higher nitrate concentration for saturation than the algae, and was not directly affected by flow velocity. The main goal of the mo del was to reproduce the observed algal biomass and metabolism results. In each scenario, algal biomass declined by 9 to 13 g wet weight, similar to experimental results (Figures 6 4 to 6 7). At low initial nitrate concentration, the peak in GPP at mediu m flow velocity was captured, although GPP was not quite as large as measured. CR also peaked at medium flow velocity, and was higher than GPP

PAGE 130

130 for all three velocities tested, resulting in slightly negative NEP as observed. At high initial nitrate concen tration, GPP decreased with increasing flow velocity, but all values were higher than observed, especially at 10 cm/s. CR increased due to the stimulation of the bacteria, which caused lower NEP. CR was also much higher than observed at 10 cm/s. A seco ndary goal of the model was to simulate the observed fluctuations in nitrate concentration. The modeled nitrate dynamics at both low and high initial nitrate concentration did correspond with the experimental measurements. At low initial concentration, n itrate oscillated slightly around the initial value due to uptake at day, followed by remineralization at night. At high initial concentration, there was an asymptotic decline in nitrate due to the denitrifying bacteria, which leveled off at about 0.55 mg N/L. Discussion A simulation model of filamentous algae under different nitrate concentrations and flow velocities that included bacteria was the only model proposed which was able to approximate the observed results of the laboratory stream channel expe riment. Within a specific range of model coefficients, the subsidy stress relationship ( Odum et al. 1979 ) between algae and flow velocity was reproduced. When nitrogen availability was limited, higher flow velocity subsidized algal production by increasi ng nitrate availability. With increasing velocity, eventually the negative effect of velocity became greater than the subsidy effect, resulting in peak algal production at an intermediate velocity. Typically the negative effect of flow velocity on attac hed algae is associated with physical drag; however drag would affect algal biomass rather than production rates. A mechanism by which flow velocity negatively affects algal production has not been

PAGE 131

131 determined, but this effect has been observed in another study and may be related to compression of algal filaments or disruption of cellular uptake processes ( Borchardt et al. 1994 ). One observed result which was not captured by the model, was the substantial decline in GPP and CR at 1.0 mg N/L and 10 cm/s v elo city. It is possible that the negative effect of velocity was stronger than modeled, but if this velocity effect is increased then metabolism also declines substantially at low nitrate concentration, which was not observed. There appears to be an unkn own mechanism involved, which caused lower overall metabolism only when both nitrate concentration and flow velocity were elevated. The decline of CR in these conditions could partially be attributed to the decline in algal production and associated respi ration; however lower algal respiration cannot account for this entire decline. This indicates that bacterial respiration was also lower in these conditions, and may be dependent on algal production to provide the labile carbon necessary for denitrificati on. Bacteria was not directly quantified in the laboratory experiment; however evidence of increased heterotrophy and declining nitrate concentration in the high nitrate trials as com pared to the low nitrate trials suggests that a denitrifying bacteria wa s present (see Chapter 5). Inclusion of bacteria in the model was necessary to simulate CR and NEP accurately which is further evidence of the influence of bacteria in the experiment. In order to reproduce the observed asymptotic decline in nitrate at hi gh initial concentration, the Michaelis Menton uptake function was altered so that uptake did not occur unless nitrate concentration was greater than 0.5 mg N/L. While this is

PAGE 132

132 atypical, it is feasible that a threshold concentration was required in order f or the bacteria to become sufficiently active for substantial denitrification. One purpose of this experiment and model was to determine the nitrate concentration that limits Lyngbya growth, and how this threshold concentration may change due to flow vel ocity. The half saturation constant (k A Half ) of 0.2 mg N/L used in the Michaelis Menton function indicates that nitrate uptake by Lyngbya would be saturated at 0.4 mg N/L. This saturation concentration lies between the values found in similar Lyngbya st udies (0.60 mg/L : Cowell and Dawes 2004; 0.11 mg/L: Stevenson et al. 2007) Since flow velocity is known to increase nutrient availability, it seems appropriate that the saturation concentration found in this study would be intermediate between these othe r two studies because the velocities tested were also intermediate (1 to 10 cm/s: this study; approximately 0 cm/s : Cowell and Dawes 2004 ; 25 cm/s : Stevenson et al. 2007 ). However, the inclusion of flow velocity effects on nitrate uptake in the model sub stantially alters the saturation concentration. The simulated response of the algae to nitrate concentration at a velocity of 1 cm/s suggests that saturation would not occur until the concentration was > 1.0 mg N/L. Furthermore, the saturation concentrat ion becomes irrelevant at flow velocity > 5 cm/s, since nitrate uptake is no longer related to concentration and is instead controlled by flow velocity alone. Nitrate concentration did appear to affect Lyngbya growth in a similar study at a flow velocity of 25 cm/s (Stevenson et al. 2007), so the model may require further adjustment to accurately represent the effects of both nitrate concentration and flow velocity on Lyngbya growth and nitrate uptake. While models such as this can be useful for mechanist ic description

PAGE 133

133 of experimental outcomes, the specific model equations and parameters should be regarded with caution as the model itself is a hypothesis and does not include all of the components and processes occurring in the actual system.

PAGE 134

134 Figure 6 1. Systems diagram of the simulation model for nitr ate and flow velocity effects on algae and bacteria.

PAGE 135

135 Figure 6 2. Response of algal production to flow velocity at two nitr ate concentrations (1.0 and 0.05 mg N /L). Figure 6 3. Relationship between nit rate uptake and nitrate concentration for bacteria and algae at different flow velocities.

PAGE 136

136 Figure 6 4 Simulation model results of storages and metabolism at low initial nitrate concentration (0.05 mg N/L) and 1 cm/s flow velocity

PAGE 137

137 Figure 6 5 Simula tion model results of storages and metabolism at low initial nitrate concentration (0.05 mg N/L) and 10 cm/s flow velocity

PAGE 138

138 Figure 6 6 Simulation model results of storages and metabolism at high initial nitrate concentration (1.0 mg N/L) and 1 cm/s flo w velocity

PAGE 139

139 Figure 6 7 Simulation model results of storages and metabolism at high initial nitrate concentration (1.0 mg N/L) and 10 cm/s flow velocity

PAGE 140

140 Table 6 1. Sources and storages used as variables in the simulation model. All storage turnover times were estimated. Sources/Storages Variable Initial Value Units Source Light L 86 mol/m 2 /s Measured Available Light R L 43 mol/m 2 /s 50% of L Estimated Velocity V 1 cm/s Measured Algae (30 days) A 68 g wet weight Measured Bacteria (10 days ) B 0.1 g wet weight Estimated Detritus (5 days) D 2 g wet weight Estimated Nitrogen (2 days) N 0.05 mg/L Measured

PAGE 141

141 Table 6 2. Model equations and coefficients. Name Variable Equation Coefficients Algae A k 1 = 5.02E 2 k 2 = 5.56E 5 k VA = 0.4 Bacte ria B k 3 = 2.08 k 4 = 6.94E 5 Nitrate N k 5 = 8.68E 6 k 6 = 5.54E 4 k 7 = 36.46 Detritus D k 8 = 4.09E 6 k 9 = 1.39E 4 Available Light R L k RL = 1.37E3 Alga l N Uptake A k A Max = 5.0E 5 k A Half = 0.2 k V1 = 0.1 k V2 = 2.5 Bacteria l N Upta ke B k B Max = 1.0E 4 k B Half = 1.0 Gross Primary Production GPP k A = 23.7 Community Respiration CR k B = 5000 k D = 92 Net Ecosystem Production NEP NEP = GPP CR

PAGE 142

142 Table 6 3. Comparison of observed experimental results and modeled results for metabolism of filamentous algae and bacteria in the six scenarios simulated Initial Conditions Metabolism Results (mg C/g biomass/d) Nitrogen (mg/L) Velocity (cm/s) GPP CR NEP 0.05 1 Observed 0.83 0.87 0.04 5 1.18 1.27 0.09 10 0.97 1.12 0 .15 0.05 1 Modeled 0.84 1.01 0.17 5 1.02 1.09 0.07 10 0.95 1.06 0.11 1.0 1 Observed 1.00 1.39 0.39 5 0.92 1.44 0.52 10 0.54 1.04 0.50 1.0 1 Modeled 1.09 1.49 0.40 5 1.02 1.47 0.45 10 0.95 1.44 0.49

PAGE 143

143 CHAPTER 7 CONCLU SION S The purpose of this research was to examine the effects of flow on filamentous algae and nutrient limitation in lotic systems, with application to algal proliferation in Florida springs and spring fed rivers. The first task was to assess predictors of nutrient limitation in lotic systems broadly by comparing nutrient concentration to a new metric, the autotrophic uptake length, which also accounts for discharge rate and autotrophic metabolism. Autotrophic uptake length was a better predictor of auto trophic N limitation, suggesting that flow d oes affect nutrient lim itation in lotic systems An autotrophic uptake length threshold was quantified above which N limitation is unlikely, and the majority of Florida springs were above this threshold value in dicating that N limitation is minimal in all but a few spring systems. Based on a literature review, a subsidy stress relationship between flow velocity and filamentous algae was hypothesized for Florida springs. The laboratory experiment found that at lo w nitrate concentration (< 0.05 mg N/L), a flow velocity increase from 1 to 5 cm/s stimulated production rates of the filamentous cyanophyte Lyngbya wollei At high nitrate concentration (1.0 mg N/L), no stimulatory effect of flow velocity was observed. This indicates that flow velocity can control nutrient availability to algae, and algal N demand may be saturated even at fairly low concentrations with moderate velocity (at least for Lyngbya wollei ). An inhibitory effect of flow velocity on algal produc tion was observed at 10 cm/s, which is relatively low in comparison to velocities in many spring fed rivers and was apparently unrelated to export The field survey and experiment at Gum Slough found that filamentous algal abundance, composed mostly of Ly ngbya wollei was negatively affected across a range of flow velocities from 5 35

PAGE 144

144 cm/s, with very little filamentous algae above 35 cm/s and the potential for large algal mats below 5 cm/s Combined, these results indicate that filamentous algae in Flo rida springs are sensitive to flow velocity, and the presence of a subsidy effect is dependent on N concentration. The velocity threshold for a shift from a subsidy to a stress effect may occur at about 5 cm/s when N concentration is low (< 0.05 mg/L), bu t when N concentration is higher then any subsidy effect would be minimal since N availability is no longer a limiting factor Overall this suggests that filamentous algae is likely to increase substantially as flow velocity decreases below 35 cm/s regar dless of N concentration, until a threshold is reach ed around 5 cm/s below which N concentration may become important. S ince declining discharge usually lead s to lower velocities, this could be a primary cause of filamentous algal proliferation in some sp ring systems. For Florida springs, the pertinent question is not whether one of the proposed sources of degradation (e.g. N enrichment, declining flow velocity, loss of consumers) is an issue but rather to what degree. Currently there is no indication tha t a single cause is responsible for observed ecological degradation but rather a plurality of causes (i.e. multiple stressors) which may have complex interactions. This situation is particularly difficult for managers as it requires a more holistic approa ch in order to maintain or restore the ecology of each system, and each spring or region may have its own particular suite of challenges and solutions. Unfortunately it is more difficult to communicate complex issues with multiple causes and high uncertai nty to the public, whose participation is often necessary to resolve local and regional problems. The simplicity of the nutrient enrichment narrative is likely the reason it has been so widely

PAGE 145

145 adopted; however nutrient reduction alone is unlikely to inhi bit filamentous algal proliferation in most Florida springs Although maintenance (or restoration) of historic nutrient levels is a desirable goal, in many cases it would be appropriate to adopt a more holistic approach that considers alternative manageme nt options in addition to nutrient control.

PAGE 146

146 LIST OF REFERENCES Ahern, K. S., J. M. O'Neil, J. W. Udy and S. Albert. 2006. Effects of iron additions on filament growth and productivity of the cyanobacterium Lyngbya majuscula Marine and Freshwater Researc h 57:167 176. Allan, J. D. 1995. Stream ecology: structure and function of running waters. Kluwer Academic Publisher, Dordrecht, Netherlands. Beer, S., W. E. Spencer and G. Bowes. 1992. HCO 3 use and evidence for a carbon concentrating process in the mat f orming cyanophyte Lyngbya birgei G.M. Smith. Aquatic Botany 42:159 171. Bernot, M. J., D. J. Sobota, R. O. Hall, P. J. Mulholland, W. K. Dodds, J. R. Webster, J. L. Tank, L. R. Ashkenas, L. W. Cooper, C. N. Dahm, S. V. Gregory, N. B. Grimm, S. K. Hamilton, S. L. Johnson, W. H. McDowell, J. L. Meyer, B. Peterson, G. C. Poole, H. M. Valett, C. Arango, J. J. Beaulieu, A. J. Burgin, C. Crenshaw, A. M. W. Sheibley, S. M. Thomas a nd K. Y. M. Wilson. 2010. Inter regional comparison of land use effects on stream metabolism. Freshwater Biology 55:1874 1890. Biggs, B. J. F. 1996. Hydraulic habitat of plants in streams. Regulated Rivers: Research & Management 12:131 144. Biggs, B. J. F. 2000. Eutrophication of streams and rivers: dissolved nutrient chlorophyll relationships for benthic algae. Journal of the North American Benthological Society 19:17 31. Biggs, B. J. F. and M. E. Close. 1989. Periphyton biomass dynamics in gravel bed rive rs: the relative effects of flows and nutrients. Freshwater Biology 22:209 231. Biggs, B. J. F. and P. Gerbeaux. 1993. Periphyton development in relation to macro scale (geology) and micro scale (velocity) limiters in two gravel bed rivers, New Zealand. Ne w Zealand Journal of Marine and Freshwater Research 27:39 53. Biggs, B. J. F., D. G. Goring and V. I. Nikora. 1998. Subsidy and stress responses of stream periphyton to gradients in water velocity as a function of community growth form. Journal of Phycolog y 34:598 607. Biggs, B. J. F. and C. W. Hickey. 1994. Periphyton responses to a hydraulic gradient in a regulated river in New Zealand. Freshwater Biology 32:49 59. Biggs, B. J. F. and S. Stokseth. 1996. Hydraulic habitat suitability for periphyton in rive rs. Regulated Rivers: Research & Management 12:251 261. Biggs, B. J. F. and H. A. Thomsen. 1995. Disturbance of stream periphyton by perturbations in shear stress: time to structural failure and differences in community resistance. Journal of Phycology 31: 233 241.

PAGE 147

147 Borchardt, M. A. 1996. Nutrients. Pages 183 227 in R.J. Stevenson, M.L. Bothwell and R.L. Lowe (editors). Algal ecology: freshwater benthic ecosystems. Academic Press, Inc., San Diego, California, USA. Borchardt, M. A., J. P. Hoffmann and P. W. Co ok. 1994. Phosphorus uptake kinetics of Spirogyra fluviatis (Charophyceae) in flowing water. Journal of Phycology 30:403 417. Bothwell, M. L. 1985. Phosphorus limitation of lotic periphyton growth rates: an intersite comparison using continuous flow trough s (Thompson River System, British Columbia). Limnology and Oceanography 30:527 542. Bothwell, M. L. 1988. Growth rate responses of lotic periphytic diatoms to experimental phosphorus enrichment: the influence of temperature and light. Canadian Journal of F isheries and Aquatic Sciences 45:261 270. Bothwell, M. L. 1989. Phosphorus limited growth dynamics of lotic periphyton diatom communities: areal biomass and cellular growth rate responses. Canadian Journal of Fisheries and Aquatic Sciences 46:1293 1301. Br own, C. D., M. V. Hoyer, R. W. Bachmann and D. E. Canfield Jr. 2000. Nutrient chlorophyll relationships: an evaluation of empirical nutrient chlorophyll models using Florida and north temperate lake data. Canadian Journal of Fisheries and Aquatic Sciences 57:1574 1583. Brown, M. T., K. C. Reiss, M. J. Cohen, J. M. Evans, K. R. Reddy, P. W. Inglett, K. S. Inglett, T. K. Frazer, C. A. Jacoby, E. J. Phlips, R. L. Knight, S. K. Notestein and K. A. McKee. 2008. Summary and synthesis of the available literature o n the effects of nutrients on spring organisms and systems. University of Florida Water Institute, Gainesville, Florida, USA. Camacho, F. A. and R. W. Thacker. 2006. Amphipod herbivory on the freshwater cyanobacterium Lyngbya wollei : chemical stimulants an d morphological defenses. Limnology and Oceanography 51:1870 1875. Camargo, J. A., A. Alonso and A. Salamanca. 2005. Nitrate toxicity to aquatic animals: a review with new data for freshwater invertebrates. Chemosphere 58:1255 1267. Canfield, D. E. and M. V. Hoyer. 1988. Influence of nutrient enrichment and light availability on the abundance of aquatic macrophytes in Florida streams. Canadian Journal of Fisheries and Aquatic Sciences 45:1467 1472. Carmichael, W. W., W. R. Evans, Q. Q. Yin, P. Bell and E. M oczydlowski. 1997. Evidence for paralytic shellfish poisons in the freshwater cyanobacterium Lyngbya wollei (Farlow ex Gomont) comb. nov. Applied and Environmental Microbiology 63:3104 3110.

PAGE 148

148 Carpenter, S. R., N. F. Caraco, D. L. Correll, R. W. Howarth, A. N. Sharpley and V. H. Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8:559 568. Carpenter, S. R., J. F. Kitchell and J. R. Hodgson. 1985. Cascading trophic interactions and lake productivity. BioScie nce 35:634 639. Champion, K. M. and R. Starks. 2001. The hydrology and water quality of springs in west central Florida. Southwest Florida Water Management District, Brooksville, Florida, USA. Chetelat, J., F. R. Pick, A. Morin and P. B. Hamilton. 1999. Pe riphyton biomass and community composition in rivers of different nutrient status. Canadian Journal of Fisheries and Aquatic Sciences 56:560 569. Conley, D. J. 2000. Biogeochemical nutrient cycles and nutrient management strategies. Hydrobiologia 410:87 96 Cooke, G. D., E. B. Welch, S. A. Peterson and P. R. Newroth. 1993. Restoration and management of lakes and reservoirs. Lewis Publishers, Boca Raton, Florida, USA. Copeland, R. 2003. Florida spring classification system and spring glossary. Special Public ation No. 52. Florida Geological Survey, Tallahassee, Florida, USA. Copeland, R., N. A. Doran, A. J. White and S. B. Upchurch. 2009. Regional and statewide trends in Florida's spring and well groundwater quality (1991 2003). Bulletin No. 69. Florida Geolog ical Survey, Tallahassee, Florida, USA. Covino, T., B. McGlynn and M. Baker. 2010a. Separating physical and biological nutrient retention and quantifying uptake kinetics from ambient to saturation in successive mountain stream reaches. Journal of Geophysic al Research 115:G04010. Covino, T. P., B. L. McGlynn and R. A. McNamara. 2010b. Tracer additions for spiraling curve characterization (TASCC): quantifying stream nutrient uptake kinetics from ambient to saturation. Limnology and Oceanography: Methods 8:484 498. Cowell, B. C. and P. S. Botts. 1994. Factors influencing the distribution, abundance and growth of Lyngbya wollei in central Florida. Aquatic Botany 49:1 17. Cowell, B. C. and C. J. Dawes. 2004. Growth and nitrate nitrogen uptake by the cyanobacteriu m Lyngbya wollei Journal of Aquatic Plant Management 42:69 71. DeNicola, D. M. and C. D. McIntire. 1991. Effects of hydraulic refuge and irradiance on grazer periphyton interactions in laboratory streams. Journal of the North American Benthological Societ y 10:251 262.

PAGE 149

149 Dillon, P. J. and F. H. Rigler. 1974. The phosphorus chlorophyll relationship in lakes. Limnology and Oceanography 19:767 773. Dodds, W. K. 1991. Micro environmental characteristics of filamentous algal communities in flowing freshwaters. Fre shwater Biology 25:199 209. Dodds, W. K., A. J. Lopez, W. B. Bowden, S. Gregory, N. B. Grimm, S. K. Hamilton, A. E. Hershey, E. Mart, W. H. McDowell, J. L. Meyer, D. Morral, P. J. Mulholland, B. J. Peterson, J. L. Tank, M. Valett, J. R. Webster and W. Wol lheim. 2002a. N uptake as a function of concentration in streams. Journal of the North American Benthological Society 21:206 220. Dodds, W. K., V. H. Smith and K. Lohman. 2002b. Nitrogen and phosphorus relationships to benthic algal biomass in temperate st reams. Canadian Journal of Fisheries and Aquatic Sciences 59:865 874. Dodds, W. K. and E. B. Welch. 2000. Establishing nutrient criteria in streams. Journal of the North American Benthological Society 19:186 196. Doyle, M. W. and S. H. Ensign. 2009. Altern ative reference frames in river system science. BioScience 59:499 510. Doyle, R. D. and R. M. Smart. 1998. Competitive reduction of noxious Lyngbya wollei mats by rooted aquatic plants. Aquatic Botany 61:17 32. Duarte, C. M. 1995. Submerged aquatic vegetat ion in relation to different nutrient regimes. Ophelia 41:87 112. Duarte, C. M. and D. E. Canfield. 1990. Macrophyte standing crop and primary productivity in some Florida spring runs. Water Resources Bulletin 26:927 934. Duff, J. H., F. J. Triska and R. S Oremland. 1984. Denitrification associated with stream periphyton: chamber estimates from undisrupted communities. Journal of Environmental Quality 13:514 518. DuToit, C. H., 1979. The carrying capacity of the Ichetucknee Springs and River. M.S. thesis, University of Florida, Gainesville, Florida, USA, 176 pp. Earl, S. R., M. Valett and J. R. Webster. 2006. Nitrogen saturation in stream ecosystems. Ecology 87:3140 3151. Elosequi, A. and J. Pozo. 1998. Epilithic biomass and metabolism in a north Iberian st ream. Aquatic Sciences 60:1 16. Elser, J. J., M. E. S. Bracken, E. E. Cleland, D. S. Gruner, W. S. Harpole, H. Hillebrand, J. T. Ngai, E. W. Seabloom, J. B. Shurin and J. E. Smith. 2007. Global analysis of nitrogen and phosphorus limitation of primary prod ucers in freshwater, marine and terrestrial ecosystems. Ecology Letters 10:1135 1142.

PAGE 150

150 Elwood, J. W., J. D. Newbold, A. F. Trimble and R. W. Stark. 1981. The limiting role of phosphorus in a woodland stream ecosystem: effects of P enrichment on leaf decompo sition and primary producers. Ecology 62:146 158. Ensign, S. H. and M. W. Doyle. 2006. Nutrient spiraling in streams and river networks. Journal of Geophysical Research 111:G04009. Evans, J. M., A. C. Wilkie, J. Burkhardt and R. P. Haynes. 2007. Rethinking exotic plants: using citizen observations in a restoration proposal for Kings Bay, Florida. Ecological Restoration 25:199 210. Fellows, C. S., H. M. Valett, C. N. Dahm, P. J. Mulholland and S. A. Thomas. 2006. Coupling nutrient uptake and energy flow in h eadwater streams. Ecosystems 9:788 804. Feminella, J. W. and C. P. Hawkins. 1995. Interactions between stream herbivores and periphyton: a quantitative analysis of past experiments. Journal of the North American Benthological Society 14:465 509. Fisher, S. G., L. J. Gray, N. B. Grimm and D. E. Busch. 1982. Temporal succession in a desert stream ecosystem following flash flooding. Ecological Monographs 52:93 110. Florea, L. J. and H. L. Vacher. 2006. Springflow hydrographs: eogenetic vs. telogenetic karst. G round Water 44:352 361. Franceour, S. N. 2001. Meta analysis of lotic nutrient enrichment experiments: detecting and quantifying subtle responses. Journal of the North American Benthological Society 20:358 368. Frazer, T. K., S. K. Notestein and W. E. Pine 2006. Changes in the physical, chemical and vegetative characteristics of the Homosassa, Chassahowitzka and Weeki Wachee Rivers. 03CON000038. Southwest Florida Water Management District, Tampa, Florida, USA. FSTF [Florida Springs Task Force]. 2000. Flori da's springs: strategies for protection and restoration. Florida Department of Environmental Protection, Tallahassee, Florida, USA. Ghosh, M. and J. P. Gaur. 1991. Regulatory influence of water current on algal colonization in an unshaded stream at Shillon g (Megahalaya, India). Aquatic Botany 40:37 46. Grimm, N. B. 1987. Nitrogen dynamics during succession in a desert stream. Ecology 68:1157 1170. Grimm, N. B. and S. G. Fisher. 1986. Nitrogen limitation in a Sonoran Desert stream. Journal of the North Ameri can Benthological Society 5:2 15.

PAGE 151

151 Grubbs, J. W. 2011. Analysis of long term trends in flow from a large spring complex in northern Florida. Scientific Investigations Report 2011 5031. USGS, Reston, Virginia, USA. Gundersen, J. K. and B. B. Jorgensen. 1990. Microstructure of diffusive boundary layers and the oxygen uptake of the sea floor. Nature 345:604 607. Hall, R. O. and J. L. Tank. 2003. Ecosystem metabolism controls nitrogen uptake in streams in Grand Teton National Park, Wyoming. Limnology and Oceanog raphy 48:1120 1128. Hall, R. O., J. L. Tank and M. F. Dybdahl. 2003. Exotic snails dominate nitrogen and carbon cycling in a highly productive stream. Frontiers in Ecology and the Environment 1:407 411. Hall, R. O., J. L. Tank, D. J. Sobota, P. J. Mulholla nd, J. M. O'Brien, W. K. Dodds, J. R. Webster, H. M. Valett, G. C. Poole, B. J. Peterson, J. L. Meyer, W. H. McDowell, S. L. Johnson, S. K. Hamilton, N. B. Grimm, S. V. Gregory, C. N. Dahm, L. W. Cooper, L. Ashkenas, S. M. Thomas, R. W. Sheibley, J. D. Pot ter, B. R. Niederlehner, L. T. Johnson, A. M. Helton, C. M. Crenshaw, A. J. Burgin, M. J. Bernot, J. J. Beaulieu and C. P. Arango. 2009. Nitrate removal in stream ecosystems measured by 15 N addition experiments: total uptake. Limnology and Oceanography 54: 653 665. Hecky, R. E. and P. Kilham. 1988. Nutrient limitation of phytoplankton in freshwater and marine environments: a review of recent evidence on the effects of enrichment. Limnology and Oceanography 33:796 822. Heffernan, J. B. and M. J. Cohen. 2010. Direct and indirect coupling of primary production and diel nitrate dynamics in a subtropical spring fed river. Limnology and Oceanography 55:677 688. Heffernan, J. B., M. J. Cohen, T. K. Frazer, R. G. Thomas, T. J. Rayfield, J. Gulley, J. B. Martin, J. J. Delfino and W. D. Graham. 2010a. Hydrologic and biotic influences on nitrate removal in a subtropical spring fed river. Limnology and Oceanography 55:249 263. Heffernan, J. B., D. M. Liebowitz, T. K. Frazer, J. M. Evans and M. J. Cohen. 2010b. Algal bloom s and the nitrogen enrichment hypothesis in Florida springs: evidence, alternatives, and adaptive management. Ecological Applications 20:816 829. Hill, W. R. and A. W. Knight. 1988. Nutrient and light limitation of algae in two northern California streams. Journal of Phycology 24:125 132. Hilton, J., M. O'Hare, M. J. Bowes and J. I. Jones. 2006. How green is my river? A new paradigm of eutrophication in rivers. Science of the Total Environment 365:66 83.

PAGE 152

152 Horner, R. R. and E. B. Welch. 1981. Stream periphyto n development in relation to current velocity and nutrients. Canadian Journal of Fisheries and Aquatic Sciences 38:449 457. Horner, R. R., E. B. Welch, M. R. Seeley and J. M. Jacoby. 1990. Responses of periphyton to changes in current velocity, suspended s ediment and phosphorus concentration. Freshwater Biology 24:215 232. Horner, R. R., E. B. Welch and R. B. Veenstra. 1983. Development of nuisance periphytic algae in laboratory streams in relation to enrichment and velocity. Pages 121 134 in R.G. Wetzel (e ditor). Periphyton of freshwater ecosystems. Dr. W. Junk Publishers, Hague, Netherlands. Howarth, R. W. 1988. Nutrient limitation of net primary production in marine ecosystems. Annual Review of Ecology and Systematics 19:89 110. Howarth, R. W. and R. Mari no. 2006. Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: evolving views over three decades. Limnology and Oceanography 51:364 376. Hoyer, M. V., T. K. Frazer, S. K. Notestein and D. E. Canfield Jr. 2004. Vegetative chara cteristics of three low lying Florida coastal rivers in relation to flow, light, salinity and nutrients. Hydrobiologia 528:31 43. Humphrey, K. P. and R. J. Stevenson. 1992. Responses of benthic algae to pulses in current and nutrients during simulations of subscouring spates. Journal of the North American Benthological Society 11:37 48. Hurd, C. L. 2000. Water motion, marine macroalgal physiology, and production. Journal of Phycology 36:453 472. Johnson, L. T., J. L. Tank and W. K. Dodds. 2009. The influenc e of land use on stream biofilm nutrient limitation across eight North American ecoregions. Canadian Journal of Fisheries and Aquatic Sciences 66:1081 1094. Jones, J. I. and C. D. Sayer. 2003. Does the fish invertebrate periphyton cascade precipitate plant loss in shallow lakes? Ecology 84:2155 2167. Jones, J. R. and R. W. Bachmann. 1976. Prediction of phosphorus and chlorophyll levels in lakes. Journal of the Water Pollution Control Federation 48:2176 2182. Jorgensen, B. B. and N. P. Revsbech. 1985. Diffus ive boundary layers and the oxygen uptake of sediments and detritus. Limnology and Oceanography 30:111 122. Joyner, J. J., R. W. Litaker and H. W. Paerl. 2008. Morphological and genetic evidence that the cyanobacterium Lyngbya wollei (Farlow ex Gomont) Spe ziale and Dyck encompasses at least two species. Applied and Environmental Microbiology 74:3710 3717.

PAGE 153

153 Katz, B. G. 2004. Sources of nitrate contamination and age of water in large karstic springs of Florida. Environmental Geology 46:689 706. Katz, B. G., H. D. Hornsby, J. F. Bohlke and M. F. Mokray. 1999. Sources and chronology of nitrate contamination in spring waters, Suwannee River Basin, Florida. Water Resources Investigations Report 99 4252. USGS, Tallahassee, Florida, USA. Knight, R. L., 1980. Energy b asis of control in aquatic ecosystems. Ph.D. dissertation, University of Florida, Gainesville, Florida, USA, 200 pp. Lamberti, G. A. and A. D. Steinman. 1997. A comparison of primary production in stream ecosystems. Journal of the North American Benthologi cal Society 16:95 104. Larned, S. T. 2010. A prospectus for periphyton: recent and future ecological research. Journal of the North American Benthological Society 29:182 206. Larned, S. T., V. I. Nikora and B. J. F. Biggs. 2004. Mass transfer limited nitro gen and phosphorus uptake by stream periphyton: a conceptual model and experimental evidence. Limnology and Oceanography 49:1992 2000. Lembi, C. A. 2003. Control of nuisance algae. Pages 805 834 in J.D. Wehr and R.G. Sheath (editors). Freshwater algae of N orth America: ecology and classification. Academic Press, New York, New York, USA. Lewis, W. M. and W. A. Wurtsbaugh. 2008. Control of lacustrine phytoplankton by nutrients: erosion of the phosphorus paradigm. International Review of Hydrobiology 93:446 46 5. Lock, M. A. and P. H. John. 1979. The effect of flow patterns on uptake of phosphorus by river periphyton. Limnology and Oceanography 24:376 383. Lohman, K., J. R. Jones and B. D. Perkins. 1992. Effects of nutrient enrichment and flood frequency on peri phyton biomass in northern Ozark streams. Canadian Journal of Fisheries and Aquatic Sciences 49:1198 1205. Ludwig, A., M. Matlock, B. E. Haggard, M. Matlock and E. Cummings. 2008. Identification and evaluation of nutrient limitation on periphyton growth in headwater streams in the Pawnee Nation, Oklahoma. Ecological Engineering 32:178 186. Mass, T., A. Genin, U. Shavit, M. Grinstein and D. Tchernov. 2010. Flow enhances photosynthesis in marine benthic autotrophs by increasing the efflux of oxygen from the o rganism to the water. Proceedings of the National Academy of Sciences 107:2527 2531.

PAGE 154

154 McCormick, P. V. and R. J. Stevenson. 1991. Grazer control of nutrient availability in the periphyton. Oecologia 86:287 291. McIntire, C. D. 1966. Some effects of current velocity on periphyton communities in laboratory streams. Hydrobiologia 27:559 570. McIntire, C. D. 1968. Structural characteristics of benthic algal communities in laboratory streams. Ecology 49:520 537. Meyer, J. L., M. J. Paul and W. K. Taulbee. 2005. S tream ecosystem function in urbanizing landscapes. Journal of the North American Benthological Society 24:602 612. Mitrovic, S. M., L. Hardwick and F. Dorani. 2011. Use of flow management to mitigate cyanobacterial blooms in the Lower Darling River, Austra lia. Journal of Plankton Research 33:229 241. Mulholland, P. J., C. S. Fellows, J. L. Tank, N. B. Grimm, J. R. Webster, S. K. Hamilton, E. Mart, L. Ashkenas, W. B. Bowden, W. K. Dodds, W. H. McDowell, M. J. Paul and B. J. Peterson. 2001. Inter biome compa rison of factors controlling stream metabolism. Freshwater Biology 46:1503 1517. Mulholland, P. J., A. M. Helton, G. C. Poole, R. O. Hall, S. K. Hamilton, B. J. Peterson, J. L. Tank, L. R. Ashkenas, L. W. Cooper, C. N. Dahm, W. K. Dodds, S. E. G. Findlay, S. V. Gregory, N. B. Grimm, S. L. Johnson, W. H. McDowell, J. L. Meyer, H. M. Valett, J. R. Webster, C. P. Arango, J. J. Beaulieu, M. J. Bernot, A. J. Burgin, C. L. Crenshaw, L. T. Johnson, B. R. Niederlehner, J. M. O'Brien, J. D. Potter, R. W. Sheibley, D J. Sobota and S. M. Thomas. 2008. Stream denitrification across biomes and its response to anthropogenic nitrate loading. Nature 452:202 205. Mulholland, P. J., E. R. Marzolf, J. R. Webster, D. R. Hart and S. P. Hendricks. 1997. Evidence that hyporheic z ones increase heterotrophic metabolism and phosphorus uptake in forest streams. Limnology and Oceanography 42:443 451. Mulholland, P. J., A. D. Steinman, E. R. Marzolf, D. R. Hart and D. L. DeAngelis. 1994. Effect of periphyton biomass on hydraulic charact eristics and nutrient cycling in streams. Oecologia 98:40 47. Mulholland, P. J., A. D. Steinman, A. V. Palumbo, J. W. Elwood and D. B. Kirschtel. 1991. Role of nutrient cycling and herbivory in regulating periphyton communities in laboratory streams. Ecolo gy 72:966 982. Mulholland, P. J., S. A. Thomas, H. M. Valett, J. R. Webster and J. J. Beaulieu. 2006. Effects of light on NO 3 uptake in small forested streams: diurnal and day to day variations. Journal of the North American Benthological Society 25:583 5 95.

PAGE 155

155 Munch, D. A., E. J. Phlips, M. S. Allen, R. L. Knight, D. J. Toth, C. Huang, J. B. Davis, C. M. Fortich, W. L. Osburn, E. L. Quinlan, M. J. Woods, P. Cooney, R. A. Clarke and S. L. Knight. 2006. Fifty year retrospective study of the ecology of Silver S prings, Florida. SJ2007 SP4. St. Johns River Water Management District, Palatka, Florida, USA. Munson, A. B., J. J. Delfino and D. A. Leeper. 2005. Determining minimum flows and levels: the Florida experience. Journal of the American Water Resources Associ ation 41:1 10. Newbold, J. D., J. W. Elwood, R. V. O'Neill and W. Van Winkle. 1981. Measuring nutrient spiralling in streams. Canadian Journal of Fisheries and Aquatic Sciences 38:860 863. Newbold, J. D., R. V. O'Neill, J. W. Elwood and W. Van Winkle. 1982 Nutrient spiralling in streams: implications for nutrient limitation and invertebrate activity. The American Naturalist 120:628 652. Nishihara, G. N. and J. D. Ackerman. 2009. Diffusive boundary layers do not limit the photosynthesis of the aquatic macro phyte Vallisneria americana at moderate flows and saturating light levels. Limnology and Oceanography 54:1874 1882. Nixon, S., B. Buckley, S. Granger and J. Bintz. 2001. Responses of very shallow marine ecosystems to nutrient enrichment. Human and Ecologic al Risk Assessment 7:1457 1481. Notestein, S. K., T. K. Frazer, M. V. Hoyer and D. E. Canfield. 2003. Nutrient limitation of periphyton in a spring fed, coastal stream in Florida, USA. Journal of Aquatic Plant Management 41:57 60. O'Brien, J. M., W. K. Dod ds, K. C. Wilson, M. J. N. and J. Eichmiller. 2007. The saturation of N cycling in Central Plains streams: 15 N experiments across a broad gradient of nitrate concentrations. Biogeochemistry 84:31 49. Odum, E. P., J. T. Finn and E. H. Franz. 1979. Perturbat ion theory and the subsidy stress gradient. BioScience 29:349 352. Odum, H. T. 1957a. Primary production measurements in eleven Florida springs and a marine turtle grass community. Limnology and Oceanography 2:85 97. Odum, H. T. 1957b. Trophic structure an d productivity of Silver Springs, Florida. Ecological Monographs 27:55 112. Odum, H. T. 1994. Ecological and general systems: an introduction to systems ecology. University Press of Colorado, Niwot, Colorado, USA, 644 pp.

PAGE 156

156 Opsahl, R. W., T. Wellnitz and N. L. Poff. 2003. Current velocity and invertebrate grazing regulate stream algae: results of an in situ electrical exclusion. Hydrobiologia 499:135 145. Ortiz Zayas, J. R., W. M. Lewis Jr., J. F. Saunders III, J. H. McCutchan Jr. and F. N. Scatena. 2005. Met abolism of a tropical rainforest stream. Journal of the North American Benthological Society 24:769 783. Peterson, B. J., J. E. Hobbie and T. L. Corliss. 1983. A continuous flow periphyton bioassay: tests of nutrient limitation in a tundra stream. Limnolog y and Oceanography 28:583 591. Peterson, C. G. and N. B. Grimm. 1992. Temporal variation in enrichment effects during periphyton succession in a nitrogen limited desert stream ecosystem. Journal of the North American Benthological Society 11:20 36. Peterso n, C. G. and R. J. Stevenson. 1990. Post spate development of epilithic algal communities in different current environments. Canadian Journal of Botany 68:2092 2102. Peterson, C. G. and R. J. Stevenson. 1992. Resistance and resilience of lotic algal commun ities: importance of disturbance timing and current. Ecology 73:1445 1461. Phelps, G. G. 2004. Chemistry of ground water in the Silver Springs basin, Florida, with an emphasis on nitrate. Scientific Investigations Report 2004 5144. USGS, Reston, Virginia, USA. Phlips, E. J., J. Ihnat and M. Conroy. 1992. Nitrogen fixation by the benthic freshwater cyanobacterium Lyngbya wollei Hydrobiologia 234:59 64. Poff, N. L., N. J. Voelz, J. V. Ward and R. E. Lee. 1990. Algal colonization under four experimentally con trolled current regimes in a high mountain stream. Journal of the North American Benthological Society 9:303 318. Poff, N. L. and J. V. Ward. 1992. Heterogeneous currents and algal resources mediate in situ foraging activity of a mobile stream grazer. Oiko s 65:465 478. Poff, N. L. and J. V. Ward. 1995. Herbivory under different flow regimes: a field experiment and test of a model with a benthic stream insect. Oikos 72:179 188. Pringle, C. M. 1987. Effects of water and substratum nutrient supplies on lotic p eriphyton growth: an integrated bioassay. Canadian Journal of Fisheries and Aquatic Sciences 44:619 629. Pringle, C. M., P. Paaby Hansen, P. D. Vaux and C. R. Goldman. 1986. In situ nutrient assays of periphyton growth in a lowland Costa Rican stream. Hydr obiologia 134:207 213.

PAGE 157

157 Quinlan, E. L., E. J. Phlips, K. A. Donnelly, C. H. Jett, P. Sleszynski and S. Keller. 2008. Primary producers and nutrient loading in Silver Springs, FL, USA. Aquatic Botany 88:247 255. Rier, S. T. and R. J. Stevenson. 2006. Respons e of periphytic algae to gradients in nitrogen and phosphorus in streamside mesocosms. Hydrobiologia 561:131 147. Roberts, B. J. and P. J. Mulholland. 2007. In stream biotic control on nutrient biogeochemistry in a forested stream, West Fork of Walker Bran ch. Journal of Geophysical Research 112:G04002. Rosemond, A. D., P. J. Mulholland and S. H. Brawley. 2000. Seasonally shifting limitation of stream periphyton: response of algal populations and assemblage biomass and productivity to variation in light, nut rients, and herbivores. Canadian Journal of Fisheries and Aquatic Sciences 57:66 75. Rosemond, A. D., P. J. Mulholland and J. W. Elwood. 1993. Top down and bottom up control of stream periphyton: effects of nutrients and herbivores. Ecology 74:1264 1280. S and Jensen, K. and J. R. Mebus. 1996. Fine scale patterns of water velocity within macrophyte patches in streams. Oikos 76:169 180. Sanderson, B. L., H. J. Coe, C. J. Tran, K. H. Macneale, D. L. Harstad and A. B. Goodwin. 2009. Nutrient limitation of perip hyton in Idaho streams: results from nutrient diffusing substrate experiments. Journal of the North American Benthological Society 28:832 845. Sanford, L. P. and S. M. Crawford. 2000. Mass transfer versus kinetic control of uptake across solid water bounda ries. Limnology and Oceanography 45:1180 1186. Schanz, F. and H. Juon. 1983. Two different methods of evaluating nutrient limitations of periphyton bioassays, using water from the River Rhine and eight of its tributaries. Hydrobiologia 102:187 195. Schindl er, D. W. 1977. Evolution of phosphorus limitation in lakes. Science 195:260 262. Schindler, D. W., R. E. Hecky, D. L. Findlay, M. P. Stainton, B. R. Parker, M. J. Paterson, K. G. Beaty, M. Lyng and S. E. M. Kasian. 2008. Eutrophication of lakes cannot be controlled by reducing nitrogen input: results of 37 year whole ecosystem experiment. Proceedings of the National Academy of Sciences 105:11254 11258. Scott, T. M., G. H. Means, R. P. Meegan, R. C. Means, S. B. Upchurch, R. E. Copeland, J. Jones, T. Robert s and A. Willet. 2004. Springs of Florida. Bulletin 66. Florida Geological Survey, Tallahassee, Florida, USA.

PAGE 158

158 Scrimgeour, G. J., J. M. Culp, M. L. Bothwell, F. J. Wrona and M. H. McKee. 1991. Mechanisms of algal patch depletion: importance of consumptive a nd non consumptive losses in mayfly diatom systems. Oecologia 85:343 348. Sickman, J. O., A. Albertin, M. W. Anderson, A. Pinowska and R. J. Stevenson. 2009. A comparison of internal and external supply of nutrients to algal mats in two first magnitude spr ings in Florida. Journal of Aquatic Plant Management 47:135 144. Smith, V. H. 1982. The nitrogen and phosphorus dependence of algal biomass in lakes: an empirical and theoretical analysis. Limnology and Oceanography 27:1101 1112. Smith, V. H. 2003. Eutroph ication of freshwater and coastal marine ecosystems: a global problem. Environmental Science and Pollution Research 10:126 139. Snyder, E. B., C. T. Robinson, G. W. Minshall and S. R. Rushforth. 2002. Regional patterns in periphyton accrual and diatom asse mblage structure in a heterogeneous nutrient landscape. Canadian Journal of Fisheries and Aquatic Sciences 59:564 577. Speziale, B. J., E. G. Turner and L. A. Dyck. 1991. Physiological characteristics of vertically stratified Lyngbya wollei mats. Lake and Reservoir Management 7:107 114. SSW [Stream Solute Workshop]. 1990. Concepts and methods for assessing solute dynamics in stream ecosystems. Journal of the North American Benthological Society 9:95 119. Stelzer, R. S. and G. A. Lamberti. 2001. Effects of N :P ratio and total nutrient concentration on stream periphyton community structure, biomass, and elemental composition. Limnology and Oceanography 46:356 367. Stevenson, R. J. 1983. Effects of current and conditions simulating autogenically changing microh abitats on benthic diatom immigration. Ecology 64:1514 1524. Stevenson, R. J. 1996. The stimulation and drag of current. Pages 321 340 in R.J. Stevenson, M.L. Bothwell and R.L. Lowe (editors). Algal ecology: freshwater benthic ecosystems. Academic Press, I nc., San Diego, CA, USA. Stevenson, R. J. and R. Glover. 1993. Effects of algal density and current on ion transport through periphyton communities. Limnology and Oceanography 38:1276 1281. Stevenson, R. J. and C. G. Peterson. 1989. Variation in benthic di atom (bacillariophyceae) immigration with habitat characteristics and cell morphology. Journal of Phycology 25:120 129.

PAGE 159

159 Stevenson, R. J., A. Pinowska, A. Albertin and J. O. Sickman. 2007. Ecological condition of algae and nutrients in Florida springs: the synthesis report. S0291. Florida Department of Environmental Protection, Tallahassee, Florida, USA. Stevenson, R. J., A. Pinowska and Y. K. Wang. 2004. Ecological condition of algae and nutrients in Florida springs. WM858. Florida Department of Environment al Protection, Tallahassee, Florida, USA. Stumm, W. and J. J. Morgan. 1996. Aquatic chemistry: Chemical equilibria and rates in natural waters. John Wiley & Sons, Inc., New York, New York, USA, 1040 pp. Tank, J. L. and W. K. Dodds. 2003. Nutrient limitatio n of epilithic and epixylic biofilms in ten North American streams. Freshwater Biology 48:1031 1049. Terrell, J. B. and D. E. Canfield. 1996. Evaluation of the effects of nutrient removal and the "storm of the century" on submersed vegetation in Kings Bay Crystal River, Florida. Lake and Reservoir Management 12:394 403. Tett, P., C. Gallegos, M. G. Kelly, G. M. Hornberger and B. J. Cosby. 1978. Relationships among substrate, flow, and benthic microalgal pigment density in the Mechums River, Virginia. Limn ology and Oceanography 23:785 797. Traaen, T. S. and E. A. Lindstrom. 1983. Influence of current velocity on periphyton distribution. Pages 97 99 in R.G. Wetzel (editor). Periphyton of freshwater ecosystems. Dr. W. Junk Publishers, Hague, Netherlands. Tris ka, F. J., V. C. Kennedy, R. J. Avanzino and B. N. Reilly. 1983. Effect of simulated canopy cover on regulation of nitrate uptake and primary production by natural periphyton assemblages. Pages 129 159 in T.D. Fontaine and S.M. Bartell (editors). Dynamics of lotic systems. Ann Arbor Science Publishers, Ann Arbor, Michigan, USA. Triska, F. J. and R. S. Oremland. 1981. Denitrification associated with periphyton communities. Applied and Environmental Microbiology 42:745 748. USEPA [U. S. Environmental Protecti on Agency]. 1998. National strategy for the development of regional nutrient criteria. EPA 822 R 98 002. U.S. Environmental Protection Agency, Washington, DC, USA. USEPA [U. S. Environmental Protection Agency]. 2001. Nutrient criteria technical guidance ma nual estuarine and coastal marine waters. EPA 822 B 01 003. U.S. Environmental Protection Agency, Washington, DC, USA. Valiela, I., J. McClelland, J. Hauxwell, P. J. Behr, D. Hersh and K. Foreman. 1997. Macroalgal blooms in shallow estuaries: controls an d ecophysiological and ecosystem consequences. Limnology and Oceanography 42:1105 1118.

PAGE 160

160 Vollenweider, R. A. 1982. Eutrophication of waters: monitoring, assessment, and control. OECD, Paris, France. von Schiller, D., E. Mart, J. L. Riera and S. Sabater. 20 07. Effects of nutrients and light on periphyton biomass and nitrogen uptake in Mediterranean streams with constrasting land uses. Freshwater Biology 52:891 906. Wade, A. J., P. G. Whitehead, G. M. Hornberger and D. L. Snook. 2002. On modelling the flow co ntrols on macrophyte and epiphyte dynamics in a lowland permeable catchment: the River Kennet, southern England. The Science of the Total Environment 282 283:375 393. Weber, K. A. and R. G. Perry. 2006. Groundwater abstraction impacts on spring flo w and base flow in the Hillsborough River Basin, Florida, USA. Hydrogeology Journal 14:1252 1264. Welch, E. B., J. M. Jacoby, R. R. Horner and M. R. Seeley. 1988. Nuisance biomass levels of periphytic algae in streams. Hydrobiologia 157:161 168. Whitford, L. A. 1956. The communities of algae in the springs and spring streams of Florida. Ecology 37:433 442. Whitford, L. A. and G. J. Schumacher. 1961. Effect of current on mineral uptake and respiration by a fresh water alga. Limnology and Oceanography 6:432 4 25. Whitford, L. A. and G. J. Schumacher. 1964. Effect of current on respiration and mineral uptake in Spirogyra and Oedogonium Ecology 45:168 170. Williams, S. A. 2006. Simulation of the effects of groundwater withdrawals from the Floridan aquifer system in Volusia County and vicinity. SJ2006 4. St. Johns River Water Management District, Palatka, Florida, USA. Wright, J. C. and I. K. Mills. 1967. Productivity studies on the Madison River, Yellowstone National Park. Limnology and Oceanography 12:568 577. W SI [Wetland Solutions Inc.]. 2008. Pollutant load reduction goal (PLRG) analysis for the Wekiva River and Rock Springs Run, Florida 2006 2007. SJ2008 SP5. St. Johns River Water Management District, Palatka, Florida, USA. WSI [Wetland Solutions Inc.]. 2010. An ecosystem level study of Florida's springs. 08010. Florida Fish and Wildlife Conservation Commission, Tallahassee, Florida, USA. WSI [Wetland Solutions Inc.]. 2011. An ecosystem level study of Florida's springs Part II -Gum Slough Springs ecosystem characterization. Florida Fish and Wildlife Conservation Commission, Tallahassee, Florida, USA.

PAGE 161

161 Young, R. G. and A. D. Huryn. 1996. Interannual variation in discharge controls ecosystem metabolism along a grassland river continuum. Canadian Journal of Fis heries and Aquatic Sciences 53:2199 2211. Young, R. G. and A. D. Huryn. 1999. Effects of land use on stream metabolism and organic matter turnover. Ecological Applications 9:1359 1376.

PAGE 162

162 BIOGRAPHICAL SKETCH Sean Alan King was born in 1983 in Columbia, So uth Carolina where he lived for eighteen years In 2001, he moved to Gainesville, Florida to begin college at the University of Florida where he met his wife, Shannon McMorrow He received a Bachelor of Science degree in e nvironmental e ngineering in 2005 During the final year of his undergraduate studies, he began graduate work at the University of Florida with his advisor Mark T. Brown, who introduced him to the concepts of systems ecology and ydrologic modeling for wetland restoration on phosphate mining land, and he received a Master of Engineering degree in 2007. Th at year he received a fellowship to begin his doctorate research at the University of Florida under the same advisor. He was al so accepted as an associate member of a NSF Interdisciplinary Graduate Education and Research Traineeship (IGERT) program, which was focused on adaptive management of water resources and involved trips to the Florida Everglades and the Okavango Delta in Bo tswana, Africa. As a result of this program, he became interested in issues involving the ecology of Florida springs and rivers, and decided to focus on this topic for the next four years of his life. Sean now considers himself both an engineer and a sci entist, and plans to use what he has learned to continue working at the interface of human and natural systems.