Citation
Linking Autotrophic Tissue Stoichiometry in Flowing Waters to Metabolism, Structure, and Nutrient Limitation

Material Information

Title:
Linking Autotrophic Tissue Stoichiometry in Flowing Waters to Metabolism, Structure, and Nutrient Limitation
Creator:
Nifong, Rachel L
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (97 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Interdisciplinary Ecology
Committee Chair:
COHEN,MATTHEW J
Committee Co-Chair:
BRENNER,MARK
Committee Members:
MARTIN,JONATHAN BOWMAN
BROWN,MARK T
FRAZER,TOM K
Graduation Date:
5/2/2015

Subjects

Subjects / Keywords:
Algae ( jstor )
Autotrophs ( jstor )
Ecosystems ( jstor )
Homeostasis ( jstor )
Metabolism ( jstor )
Nutrients ( jstor )
Species ( jstor )
Stoichiometry ( jstor )
Supply ( jstor )
Taxa ( jstor )
Interdisciplinary Ecology -- Dissertations, Academic -- UF
carbon -- homeostasis -- nitrogen -- phosphorus -- springs
Alexander Springs ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Interdisciplinary Ecology thesis, Ph.D.

Notes

Abstract:
Ecological stoichiometry investigates ratios of elements, e.g., carbon-C, nitrogen-N, and phosphorus-P, in ecological processes. Organisms require elements in proportions different from their surroundings, creating imbalances between resource supply and biotic demand. This work examines supply and demand in Florida spring-fed rivers, where constant inputs and high productivity create model aquatic ecosystems. Theory predicts, and chemostat experiments support, that autotrophs adjust their stoichiometry to elemental supply. In light of contrary field evidence, I tested the hypothesis that tissue plasticity arises only when nutrient supply does not satisfy demand, i.e. nutrients are limiting. I observed strict homeostasis, i.e. no plasticity, in vascular plant and algal tissues across 41 springs spanning an unparalleled nutrient ratio gradient. Simulation modeling suggests high hydraulic turnover saturates biotic nutrient demand, even at low concentrations. This provides a framework for detecting nutrient limitation from tissue stoichiometry and implies hydraulics influence nutrient limitation in flowing waters. Nutrient enrichment has been linked to shifts from vascular plants to benthic filamentous algae in Florida springs. Because supply greatly exceeds demand, I hypothesized that alternative factors, flow and grazers, control autotroph competition. I implemented in situ grazer-manipulation mesocosms, spanning a supply gradient across springs, and varying flow velocity within springs. Reducing grazers had the largest effect, diminishing vascular taxa success, and tripling algal standing stocks. Flow also influenced algal standing stocks, though effects were weaker. These results support the hypothesis that top-down trophic controls dominate algal accumulation in springs. Carbon metabolism is measured in flowing waters using diel variation in dissolved oxygen, an approach recently generalized for N and P, yielding the stoichiometry of ecosystem primary production. To date, however, no rigorous test links system-scale values to the autotrophs present. Because autotrophs are divergently homeostatic, I hypothesized the stoichiometry of ecosystem metabolism is governed by autotroph tissues. Across 7 springs of varying autotrophic abundance, ecosystem C:N (C:NE) was strongly correlated with tissue C:N (C:NA). However, as vascular plant density increased, C:NE systematically overpredicted C:NA values, suggesting vascular taxa obtain ~ 30% of N from sediment sources. These results support the hypothesis that metabolism and tissue stoichiometry control coupling of element cycles across scales. ( en )
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.
Thesis:
Thesis (Ph.D.)--University of Florida, 2015.
Local:
Adviser: COHEN,MATTHEW J.
Local:
Co-adviser: BRENNER,MARK.
Statement of Responsibility:
by Rachel L Nifong.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Nifong, Rachel L. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Classification:
LD1780 2015 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

LINKING AUTOTROPHIC TISSUE STOICHIOMETRY IN FLOWING WATERS TO METABOLISM, STRUCTURE, AND NUTRIENT LIMITATION By RACHEL L. NIFONG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFI LLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2015

PAGE 2

© 2015 Rachel L. Nifong

PAGE 3

To my wonderful family : past, present, and future

PAGE 4

4 ACKNOWLEDGMENTS I thank my advis or, Dr. Matthew Cohen, for his guidance, his wisdom, his challenges, and his persistence to seek the scientific truth; I truly appreciate your efforts . Without them , I would not be where I am today. I also like thank my wonderful co chair , Dr. Mark Brenner , who has always lent a hand with a smile and kind word. I extend my gratitude also to my committee, Dr. Mark Brown, Dr. Tom Frazer, and Dr. Jon Martin for their support. I extend my sincere gratitude to Larry Korhnak whose field prowess enabled much of t his research. I thank Andrea Albertin, Brett Caudill, Dr. Jason Curtis, Chad Foster, Bobby Hensley, William Kenney, Jasmine McAdams, and Jenny McBride whose assistance in the lab and field has been invaluable. I would be remiss without acknowledging a grea t co author, Dr. Wendell P. Cropper Jr.; I truly appreciate your patience and assistance. The first three years of this work w ere generous ly supported by an assistantship through the School of Natural Resources and Environment thanks to Dr. Matt hew Cohen a nd Dr. Steve Humphrey. Later, I was humbled and grateful to be supported by a United States Environmental Protection Agency STAR Graduate Fellowship . I thank my husband, James Nifong, who has provided endless support, assistance, and encouragement. It has been a long and winding journey , but I am grateful that you have been my stubborn, funny, and intelligent partner throughout. My parents, Susen and Thomas Douglass, my sister, Abigail Douglass, my grandparents William and Janice Douglass, and the entire Wo lford clan have been wonderful sources of love and support throughout these years.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 RESEARCH BACKGROUND ................................ ................................ ................. 12 Balance Between Supply and Demand Within Ecosystems ................................ ... 12 Supply and Demand in Springs Ecosystems ................................ .......................... 12 Research Outline ................................ ................................ ................................ .... 14 2 HOMEOSTASIS AND NUTRIENT LIMITATION OF BENTHIC AUTOTROPHS IN NATURAL CHEMOSTATS ................................ ................................ ................. 16 Introduction ................................ ................................ ................................ ............. 16 Methods ................................ ................................ ................................ .................. 20 Study Sites ................................ ................................ ................................ ....... 20 Water and Tissue Collection ................................ ................................ ............. 21 Data Analysis ................................ ................................ ................................ ... 22 Model Development ................................ ................................ ......................... 23 Results ................................ ................................ ................................ .................... 26 Discussion ................................ ................................ ................................ .............. 29 3 EXPERIMENTAL EFFECTS OF FLOW AND GRAZERS ON PRIMARY PRODUCER COMMUNITY STRUCTURE IN FLORIDA SPRINGS ....................... 44 Introduction ................................ ................................ ................................ ............. 44 Methods ................................ ................................ ................................ .................. 46 Study Sites ................................ ................................ ................................ ....... 46 Experimental Design ................................ ................................ ........................ 47 Field and Laboratory Methods ................................ ................................ .......... 49 Statistical Analysis ................................ ................................ ............................ 49 Results ................................ ................................ ................................ .................... 50 Site Characteristics ................................ ................................ .......................... 50 Autotroph Characteristics by Site and Initial Ecosystem Structure ................... 50 Evaluation of Linear Mixed Effects (LME) ................................ ......................... 51 Grazer Effects ................................ ................................ ................................ .. 51 Flow Effects ................................ ................................ ................................ ...... 52 Vegetation Effects ................................ ................................ ............................ 53

PAGE 6

6 Discussion ................................ ................................ ................................ .............. 53 4 ON THE COUPLING OF AUTOTROPHIC STOICHIOMETRY TO ECOSYSTEM METABOLISM ................................ ................................ ................................ ........ 66 Introduction ................................ ................................ ................................ ............. 66 Methods ................................ ................................ ................................ .................. 67 Sensor Deployments ................................ ................................ ........................ 68 Metabolism ................................ ................................ ................................ ....... 69 Autotrophic Stoichiometry and Composition ................................ ..................... 70 Statistical Analysis ................................ ................................ ............................ 71 Results ................................ ................................ ................................ .................... 71 Discussion ................................ ................................ ................................ .............. 73 5 SYNTHESIS AND CONCLUSION ................................ ................................ .......... 85 Absence of Resource Limitation and Homeostasis ................................ ................. 85 Ecosystem Structure Implications Revealed by Experimental Mesocosms ............ 86 Lin king Ecosystem Metabolism and Autotroph Stoichiometry ................................ . 87 Overall Implications ................................ ................................ ................................ 88 LIST OF REFERENCES ................................ ................................ ............................... 89 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 97

PAGE 7

7 LIST OF TABLES Table page 2 1 Model parameters used in numerical solutions unless otherwis e noted. ............ 36 2 2 Autotrophic molar tissue ratios by species ................................ ......................... 37 2 3 Fitted logistic regression slopes for predicting species presenc e and absence based on resource chemistry predictors.. ................................ ........................... 38 3 1 Measured resource parameters by site with resource ratios calculated on a molar basis. ................................ ................................ ................................ ........ 56 3 2 Linear mixed effects model results evaluating the effects of grazers, flow, and vegetation as well as their interactive effects by taxa specific parameter. ... 57 4 1 Ecos ystem metabolism study sites ................................ ................................ ..... 7 7 4 2 Summary of metabolism and nutrient assimilation by deployment. .................... 78

PAGE 8

8 LIST OF FIGURES Figure page 2 1 Conceptual diagram relating supply (hydraulic turnover, , and input concentration, (R in,N and R in,P ) and light (PAR) to demand in a be nthic chemostat. ................................ ................................ ................................ .......... 39 2 2 Molar tissue vs. resource ratios for C:N, C:P, and N:P . . ................................ ..... 40 2 3 Factors that control tissue stoichiometry include nutrient supply (hydraulic turnover, ; input concentration, R in,N and R in,P ) and biotic demand (light, PAR; uptake half saturation, K i ) ................................ ................................ .......... 41 2 4 Effects of resource concentration and hydraulic turnover ( ) on tissue stoichiometry . ................................ ................................ ................................ ..... 42 2 5 Varying the half saturation const ant (K N , K P ) creates weak plasticity in tissue stoichiometry even when light limits growth (i.e., constant biomass across simulations). . ................................ ................................ ................................ ...... 43 3 1 Diagram of experimental mesocosms design. ................................ .................... 58 3 2 Diagram of caging treatment and cage control design. ................................ ...... 59 3 3 Effect of grazer treatment on the relative standing stock of vascular plants. . ..... 60 3 4 Interaction plot showing that grazer inclusion led to higher relative standing stocks in the vascular plant treatment relative to the mixed v egetation treatment. ................................ ................................ ................................ ........... 61 3 5 Grazer exclusion led to significantly higher relative standing stock of algal taxa. . ................................ ................................ ................................ ................... 62 3 6 Interaction plot showing grazer exclusion resulted in significantly higher relative algal standing stock in mixed and all vascular vegetation levels. ........... 63 3 7 Relative algal standing stocks in intermediate flows were significantly higher than stocks in lower or higher flow environments. . ................................ ............. 64 3 8 Interaction plot showing the effect o f flow among different vegetation types. Highest relative algal standing stocks were found at intermediate flows in 100% vascular vegetation. ................................ ................................ ................. 65 4 1 Diel signals of all deployments. ................................ ................................ .......... 79 4 2 Relationship between ecosystem metabolism C:N (C:N E ) and autotrophic tissue C:N (C:N A ). ................................ ................................ ............................... 80

PAGE 9

9 4 3 Relationship between (A) percent of algal cover and N uptake (mmol N m 2 d 1 ) as well as between (B) NO 3 concentrations and N uptake (mmol N m 2 d 1 ) .. 81 4 4 The difference ( C:N) between ecosystem C:N (C:N E ) and aggregated tissue C:N (C:N A ) ratios shown with 95% confidence intervals. ................................ .... 82 4 5 Relationship between C:N and vascular plant cover. ................................ ....... 83 4 6 Tissue C:N by vascular plant part type ................................ ............................... 84

PAGE 10

10 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 Philosophy LINKING AUTOTROPHIC TISSUE STOICHI OMETRY IN FLOWING WATERS TO METABOLISM, STRUCTURE, AND NUTRIENT LIMITATION By Rachel L. Nifong May 2015 Chair: Matthew J. Cohen Cochair: Mark Brenner Major: Interdisciplinary Ecology Ecological stoich iometry investigates ratios of elements , e.g. , carbon C, nitroge n N, and phosphorus P, in ecolog ical processes. O rganisms requ ire elements in proportions differ ent from their surro unding s, creating imbalances between res ource supply and biotic demand. This work examine s sup ply and demand spring fed rivers, where constant inputs and high productivity create model aquatic ecosystems. T heory predicts , and chemostat exper iments support, that autotrophs adjust their stoichiometry to elemental supply. In light of contrary field evidence, I tested the h ypothesis that tissue plasticity ar ises only when nutrient supply does not satisfy demand , i.e. nutrients are limiting . I observed strict h omeostasis, i.e. no plasticity , in vascular plant and algal tissues across 41 springs spanning an unparalleled nutrie nt ratio gradient . Simulation modeling suggests high hydraulic turnover saturates biotic nutrient demand, even at low concentrations. This provides a framework for detecting nutrient limitation from tissue stoichiometry and implies hydraulics influence nut rient limitation in flowing waters.

PAGE 11

11 Nutrient enrichment has been linked to shifts from vascular plants to benthic filamentous algae s upply greatly exceeds demand, I hypothesize d that alternative factors, flow and grazers, cont rol autotroph competition. I implemented in situ grazer manipulation mesocosms , spanning a supply gradient across springs, and varying flow velocity within springs . Reducing grazers had the largest effect, diminishing vascular taxa success, and tripling al gal standing stoc ks . F low also influenced algal standing stocks , though effects were weaker. These results support the hypothesis that top down trophic controls dominate algal accumulation in springs. Carbon metabolism is measured in flowing waters using d iel variation in dissolved oxygen , an approach r ecently generalized for N and P , yielding the stoichiometry of ecosystem primary production. To date, however, no rigorous test links system scale v alues to the autotrophs present. Because autotrophs are dive rgently homeostatic, I hypothesize d the stoichiometry of ecosystem metabolism is governed by autotroph tissues . Across 7 springs of varying autotrophic abundance, ecosystem C:N (C:N E ) was strongly correlated with tissue C:N (C:N A ) . However, as vascular pla nt density increased, C:N E systematically overpredicted C:N A values, suggesting vascular taxa obtain ~ 30% of N from sediment sources . These resu lts support the hypothesis that metabolism and tissue stoichiometry control coupling of element cycles across s ca les.

PAGE 12

12 CHAPTER 1 RESEARCH BACKGROUND Balance B etween Supply and Demand Within Ecosystems The function and structure of e c osystems develop and change over time . Species interactions (e.g. disease, predation, competition), alterations in patterns of p recip itation and temperature caused by climate change, and other anthropogenic impacts all influence ecosystem level changes (Crowl et al. 2008, Nelson et al. 2006, Parmesan & Yohe 2003) . The context in which these drivers act , as well as the outcomes created b y them, both individually and synergistically, are framed by the balance between resource supply and biotic demand ( sensu Schade et al. 2005) . Ecological s toichiometry is a valuable approach to investiga te the b alance between resource supply and biotic dem and . This approach involves measurement of multiple elemental ratios in organisms and their environment , e.g. , carbon C, nitrogen N, and phosphorus P, as they are involved in ecological interactions and processes (Sterner & Elser 2002). O rganisms often req uire elements in proportions that differ from the quantities available in the surrounding environment; this mismatch creates imbalances which inform a variety of ecological phenomena ranging from trophic interactions to population dynamics (Elser et al. 20 00 , Elser et al. 2010). Supply and Demand in Springs Ecosystems Springs provide ideal study system s to examine lotic ecosystem ecology and test stoichiometric theory due , in part, to exceptional water clarity which enables high autotrophic productivity an d stable abiotic conditions with regard to tem perature, nutrients, and flow . Florida is h ome to an estimated 600 springs, and as such, s prings are important hydrologic and economic resource s throughout the state. They also

PAGE 13

13 provide important habitat for man y organisms including , for example, manatees, otters, turtles, crayfish, and fish es . A cross Florida springs, there are l arge natural gradient s in carbon C, nitrogen N, and phosphorus P , wh ereas within spring variation is quite low (Heffernan et al. 2010). S prings have undergone drastic changes in autotrophic composition in recent years , as filamentous algal m ats have become ubiquitous and native macrophytes have disappeared or declined in the majority of sp rings (Stevenson et al. 2007 ). Nutrient loading in the form of nitrate N has increased in many springs and has been proposed to be the primary driver of this autotrophic shift in spr ecosystems and other freshwater ecosystems worldwide (Smith et al. 2006; Stevenson et al. 2007). Many regulatory and restorat ion efforts have focused on reducing nutrient loading from po int and nonpoint sources to spring eco system s to reverse this shift (He ffernan et al. 2010 ). Some studies, however, suggest nutrient loading is not the fundamental cause of this autotroph ic shift. For instance, Heffernan et al. (2010) showed that nitrate concentrations across springs are not significantly correlated with percent cover of algal b iomass . Similarly, g ross primary production and ecosystem respiration are not with nitrate conce ntrations (Odum 1957 ) . If spring ecosystems were responding to nitrate enrichment, a correlation between nutrient concentration and metrics for autotrophic biomass or productivity would be expected. Although increased nutrient loading to freshwater ecosyst ems can lead to changes that include increased phytoplankton and al gae biomass , changes in species composition, and reduced habitat quality for fish and macroinvertebrates , the mechanism responsible for autotrophic species shifts in Flor ida springs remains unresolved. This research aims to investigate

PAGE 14

14 the relatio nship between autotrophs and the supply of elements in Florida spring ecosystems. Research Outline To examine the relationship between the balance of resource supply and autotrophic demand using st oichiometry and investigate the potential ecological implications of this relationship for ecosystem structure and function in Florida spring ecosystems, I employed a number of field, experimental, and modeling approaches. I employed thre e lines of inquiry to test hypotheses about the relatio ns between benthic autotroph demand and resource supply , and implications for the ecosystems they inhabit : 1) the alleviation of nutrient limitation enables aquatic autotroph homeostasis, 2) under conditions of autotrop hic homeostasis, alternative factors such as flow and herbivory, play a large role in influencing ecosystem structure via autotrophic shifts , and 3) when autotrophic homeostasis prevails, C:N:P ratios of ecosystem metabolism reflect autotroph tissue ratios , whereas differences between tissue ratios and ecosystem metabolism ratios indicate daily variation in dissimilatory pathways such as denitrification , or use of alternative pathways of resource assimilation . In Chapter 2, I combine empirical measure s of a utotrophic tissue and resource nutrient ratios across 40 springs with simulation modeling to explore the hypothesis that organisms in flowing waters stabilize their stoichiometry when resource supply exceeds biotic demand . In Chapter 3, I implement multi f actorial , in situ experimental mesocosms t o assess the extent to which a physical factor ( flow ) and a biological factor ( grazing pressure ) influence ecosystem structure by altering algal and vascular standing stocks and net production across a range of nut rient concentrations and ratios . In Chapter 4, I assess the link between th e ratios of ecosystem primary production and tissues using a

PAGE 15

15 combination of high freque ncy in situ nutrient sen sors, elemental tissue analysis, and statistical modeling to elucidate how ecosystem nutrient cycling corresponds to the demand s of the biota .

PAGE 16

16 CHAPTER 2 HOMEOSTASIS AND NUTRIENT LIMITATION OF BENTHIC AUTOTROPHS IN NATURAL CHEMOSTATS Introduction Ecological stoichiometry provides a unifying framework for understanding how o rganisms grow and compete , given environmental variation in element supply . Because organisms often demand elements in proportions that differ from their availability , they can both affect and be affected by element ratios in the environment . Investigating these mism atches has yielded insights on trophic interactions and population dynamics (Sterner and Elser 2002 ). Where organisms strongly regulate their internal stoichiometry (i.e., homeostasis), differences with their environmental supply can modify ecos ystem structure and function through resource competition, nutrient cycling, and trophic transfer (Sterner and Elser 2002) . Conversely, where organismal inte rnal stoichiometry is flexible, i.e. organisms take up nutrients as a function of supply, or exhi bi , they can better contend with changing nutrient availability, but this may also induce food quality constraints for organisms at higher trophic levels (Elser et al. 2000) and modify litter decomposition rates (Enriquez et al. 1993). A centra l tenet of ecological st oichiometry theory is that whereas animals are strictly homeostatic, autotrophs adapt their internal stoichiometry to environmental conditions (Sterner and Elser 2002) . Although this is clearly supported by autotroph plasticity in e xperimental This chapter reprinted with permission from Wiley. Limnol. Oceanogr., 59(6), 2014, 2101 2111 © 2014, by the Association for the Scien ces of Limnology and Oceanography, Inc. doi:10.4319/lo.2014.59.6.2101

PAGE 17

17 chemostats (Rhee 1978) and several ecological studies (Vitousek 1984), instances of autotroph homeostasis have also been widely observed, even over large natural and experimental gradients in element supply and resource stoichiometry (Redfield 1958; Hall et al. 2005; Yu et al. 2011) . Indeed, a recent meta analysis suggests algal homeostasis is the prevailing condition , in spite of varying environmental conditions (Persson et al. 2010). Hall et al. (2005) suggest several mediating factors (herbiv ory, high hydraulic loss rates, resource competition, and episodic nutrient delivery) that may explain the discrepancy between field observations of homeostasis and lab oratory findings of plasticity. However, a general theory that explains why autotroph st oichiometry is plastic in some settings and homeostatic in others is still clearly needed. Nutrient limitation of plant production can occur when temperature is adequate, light is supplied in excess of photosynthetic requirements, and biomass loss rates ar e low (Hecky and Kilham 1988) . Limitation of growth by nutrients has been widely observed in terrestrial (Vitousek 1984), freshwater (Carpenter et al. 1998), and ocean (Howarth 1988) ecosystems . Although the original conception invoked limitation by a sing le element (Leibig 1855), with local support for this in some settings (Slavik et al. 2004), recent work highlights that a combination of temporal, spatial , or successional factors lead to co limitation or sequential limitation (Harpole et al. 2011 ). Altho ugh nutrient limitation in flowing water systems has been demonstrated in some settings (e.g., P limitation Slavik et al. 2004, N limitation Grimm and Fisher 1986), factors such as flood disturbance, shading, and nutrient supply via advection (King et al. 2014) have led to revised models of eutrophication in streams and rivers, particularly those in

PAGE 18

18 which benthic (i.e., attached to fixed substrates) productivity dominates (Hilton et al. 2006). Nutrient spiraling theory (Newbold et al. 1981) predicts str ict nutrient limitation in stre ams where environmental supply, i.e. water column concentration, approaches zero; that is, mineral nutrients in the water implies that autotrophs are released from nutrient limitation by advection and organic matter mineraliz ation. In practice, this heuristic prediction is relaxed (Grimm and Fisher 1986), in part because autotroph uptake is concentration dependent (Droop 1973), and also because boundary layers can exist around plants and biofilms , which create diffusion constr aints at low concentrations (Larned et al. 2004, though see Nishihara and Ackerman 2009). However, spiraling theory does suggest that concentration does not fully enumerate resource supply in flowing water, where hydraulic turnover is integral (Hilton et a l. 2006) . There is strong demand (Earl et al. 2006) with increasing concentration. Moreover, autotrophic uptake can be surprisingly small relative to supply (Heffernan and Cohen 2010; Cohen et al. 2013), which may imply widespread saturation of plant demand in lotic systems (King et al. 2014) . The effects of satisfying autotroph nutrient demand on tissue stoichiometry are poorly understood . Here I hypothesize that tissu e plasticity arises only when nutrient supply rates (controlled by hydraulic turnover and concentration) are insufficient to satisfy demand; by complement, homeostasis occurs when nutrients are supplied in excess of demand. Models of experimental chemostat s (Droop 1973; Klausmeier et al. 2004), which recreate observations of strict autotrophic plasticity (Rhee 1978), fail to capture two

PAGE 19

19 critical features of flowing water ecosystems . Chemostat primary production is entir ely pela gic, i.e. suspended in the wat er co lumn, whereas primar y production in short residence time rivers (Hilton et al . 2006) is principally benthic, i.e. attached . Pelagic production constrains the hydraulic turnover ( ) to values below the r elative biomass growth rate (~ 1.1 d 1 ) because b iomass declines to zero when loss rates to advection exceed growth . Under these conditions, declining biomass limits nutrient removal from the water, which notably leads to declining tissue plasticity at high turnover in chemostat models ( = 1.05 d 1 ; Kla usmeier et al. 2007). In benthic systems, in contrast, hydraulic turnover is decoupled from biomass export, at least under baseflow conditions, allowing substantially larger values without affecting biomass accrual, and with commensurate increases in nut rient supply rate. Chemostat models typically assume limitation by one or a combination of nutrients (Klausmeier et al. 2004), but not by factors such as light (though see Zonneveld et al. 1997) , which can be the dominant control on primary production rate s in streams (Odum 1957). Here, I tested the hypothesis that alleviation of nutrient limitation enables aquatic autotroph homeostasis by coupling empirical measurements and models of plant tissue stoichiometry across a large element (N and P) supply gradie nt in na tural flowing water chemostats, i.e. spr ing fed rivers of north Florida . I investigated relationships between resource and tissue stoichiometry of four dominant autotrophic species (2 vascular plants, 2 f ilamentous algae), compared taxon specific s toichiometry, and evaluated stoichiometric controls on the ir distribution across sites. I predicted autotroph homeostasis across this resource supply gradient due to continuous nutrient supply via advection. Using the Droop model, modified for benthic prod uction an d potential light

PAGE 20

20 limitation, I evaluated the role of increasing hydraulic turnover and light limitation on tissue stoichiometry, predicting homeostasis as light limitation occurs, induced by increased hydraulic turnover and associated increases i n nutrient supply . Methods Study Sites The spring fed rivers of north Florida are stable and biologically productive . Despite unparalleled within spring stability in chemistry, flow, a nd temperature, geologic and anthropogenic variation s across springs create large resource gradients . Low temporal variation makes tenable the assumption that biomass and tissue stoichiometry are close to equilibrium with the environment (Odum 1957), allow ing meaningful inference from synoptic measures of water and tissue chemistry. From a recent survey o f springs (Scott et al. 2004) I selected 41 sites of varying discharge (Q) based on variation in nutrient concentrations, secure access and protection from excess recreational disturbance. Soluble reactive phosphorus (SRP) concentrations, variation in which is principally controlled by contact with P rich geologic strata, spanned an order of magnitude , while nitrate (NO 3 ) concentrations, variation in which is principally controlled by anthropogenic contamination, spanned two orders of magnitude . Sites encompassed conditions considered b oth severely N limited (N:P < 14 ; Koerselman and Meuleman 1996) and P limited (N:P > 16 ). Although dissolved inorganic C (DI C) concentrations were not a selection criterion, DIC also varied substantially across sites.

PAGE 21

21 Water and Tissue Collection Water samples were collected from the main spring vent at each site and analyzed for NO 3 , SRP, and DIC; water column mineral nutrient concentrations are the overwhelmingly dominant N and P form in these systems (Cohen et al. 2013 ). I also measured water temperature (T), specific conductance (SpC), pH, and dissolved oxygen (DO) at a depth of 1 m using a multi parameter sonde (YSI6920, Ye llow Springs) that was calibrated each day. DIC was measured using a UIC Inc. 5011 CO 2 coulometer coupled with an AutoMate automated carbonate preparation device. SRP and NO 3 were measured using United States Environmental Protection Agency Methods 365.1 and 353.2, respectively. I sampled vascular plant and algal tissues from the benthic autot roph community at each site. The goal was to sample the same taxa of submerged vascular plants and benthic filamentous alga e at all sites. Based on prior research in (Stevenson et al. 2007), I target ed two macrophytes, tapegrass, Vallisneria americana Michaux 1803 (hereafter Vallisneria ), and springtape, Sagittaria kurziana Glück 1827 (hereafter Sagittaria ), as well as two filamentous algal species, L yngbya wollei (Farlow ex Gomont) Speziale and Dyck (hereafter Lyngbya ) and Vaucheria disperma de Candolle (hereafter Vaucheria ) . However, not all taxa were present at all sites. Where a species was not found after extensive surveying, it was considered abs ent; at least one representative of each guild (vascular plants and algae) was found at the majority of sites. I sampled tissues from open and closed canopy conditions to explore light regime effects on tissue stoichiometry (Sterner and Elser 2002) . Three tissue samples per species per light regime were collected and composited for elemental analysis . Because

PAGE 22

2 2 element composition can vary with tissue age and type I sampled only new growth (Sterner and Elser 2002). Samples were stored on ice; within 48 hours, they were triple washed with deionized water, dried at 105°C and ground. Percent C and N were measured with a Carlo Erba NA1500 CNS elemental analyzer . Percent P was measured as SRP on a Technicon Autoanalyzer II with a single channel colorimeter and elec tronic data acquisition following digestion with H 2 SO 4 and K 2 S 2 O 8 . Data Analysis To test the hypothesis that saturating autotrophic nutrient demand leads to tissue homeostasis, I first evaluated taxon specific stoichiometric signatures and then explo red wh ether these are constant, i.e. homeostatic, across the large enviro nmental gradients present in the study sites . I also evaluated whether nutrient conditions (element stoichiometry and concentrations) affect the presence or absence of different taxa across sites. Taxonomic and light regime effects on tissue stoichiometry were compared using a two way analysis of variance (ANOVA) for C:N , C:P, and N:P; post hoc significance level p < 0.05. All statistical analyses were performed using R version 3.0.1 (R Development Core Team 2013). Physiological homeostasis refers to active internal regulation, e.g. via gene r egulation (Ashworth et al. 2013), of nutrient quotients to maintain stabil ity despite external fluctuations. Although mechanisms that may allow autotrophs to exert cybernetic control are increasingly evident (S mith et al. 2009), the definition of homeostasis used here is strictly stoichiometric, based on the constancy of interna l elemental composition de spite environmental variation. As such, I do not discriminate

PAGE 23

23 between active and passive (e.g., diffusion out of the cell) mechanisms that enable intra cellular constancy . Homoeostatic regulation (H) was thus quantified as the inv erse slope of the log log association between organism and resource stoichiometry (Sterner and Elser 2002) . The statistical significance of the log log slopes was evaluated, with values near 1 indicating plasticity, and larger values indicating increasing regulation of internal tissue stoichiometry compared with environmental supply; H under strict resource (NO 3 , SRP, DIC, pH, DO, SpC, T, and Q) controls on species presence and absence were evaluated using logistic regression. Model Development I made three modifications to a Droop model that considers two potentially limiting nutrients (Klausmeier et al. 2007) to better reflect benthic conditio ns in our study sites (Fig ure 2 1). 1) I decoupled , the hydraulic turnover rate (d 1 ), from m, the biomass loss rate (d 1 ), allowing for the use of more realistic values of for a 500 m reach (~ 10 to 50 d 1 ; Hensley and Cohen 2012). Chemostat models typically assume equals m, but this fails to accura tely represent attached autotrophs in flowing systems. 2) I added potential growth lim itation by light (Odum 1957). 3) I estimated parameter values, including input nutrient concentrations (R in,P , and R in,N ), uptake half saturation constants (K i ) and reali zed maximum biomass growth rate ( max ) for spring fed systems (Table 2 1). Simulations were parameterized for a spring fed river reach 500 m long, 10 m wide and 1 m deep , comprised of Vallisneria beds; the reach length is arbitrary, selected to minimize nutrient mineralization as a fracti on of autotroph supply (which increases with longer reaches), while allowing for significant autotrophic reduction of

PAGE 24

24 nutrient inputs. The length also reflects the maximum distance downstream of the vent at which tissue samples were obtained. The mathemati cal representation of the model is presented below: (2 1 ) ( 2 2) ( 2 3) ( 2 4) ( 2 5) where R in,N and R in,P are input concentrations, and R N and R P are water column concentrat ions of N and P, respectively. Nutrient uptake for nutrient i is defined by the f ollowing equations with uptake rates, v maxh, i and v maxl, i , and half saturation constants, K i which allow for uptake to decrease as the internal quota, Q i , approaches the maximum quota, Q max, i : ( 2 6) ( 2 7) Although K i can vary with ambient nutrient concentrations (Collos et al. 2005), temperature ( Shatwell et al. 2014), and between species (Carpenter and Guillard 1971) further physiological studie s beyond the scope of this study would be required to parameterize this variability. As such, and because temperature and species composition do not vary o ver model simulations, and input nutrient concentrations are time invariant within each model run, we adopted constant K i values for the model

PAGE 25

25 species , which were modified from published va lues (Klausmeier et al. 2007). B is biomass in grams over the mode l domain, min represents the minimum of the referenced limitation equations, and Q i is the internal plant cell nutr ient (Klausmeier et al. 2007). Biomass loss rate (m) and realized maximum growth rate ( max , Cherif and Loreau 2010) were adjusted for spring s based on published net primary productivity (Heffernan and Cohen 2010) and biomass turnover rates (H auxwell et al. 2004, Table 2 1). I defined N limitation (N lim ), P limitation (P lim ), and light limitation (L lim ) as follows , based on the formulation pres ented in Cherif and Loreau (2010): ( 2 8) ( 2 9) ( 2 10) This formulation for N and P limitation includes parameter estimates f or the maximum internal cell quotas (Q max,N and Q max,P ). Eqs. 2 8 and 2 9 differ from previous model formulations (Klausmeier et al. 2007) in explicitly creating conditions where max is equal across resou rces (Cherif and Loreau 2010). However, I am unaware of any physiological studies t hat estimate these parameters. For this model, tissue stoichiometry when both cell quotas are at their maximum, i.e. Q maxN : Q maxP , equals 21.65. This value is specific to Vallisneria (derived as the 95 th percentile of observe d values), though variation in this value, e.g. using different taxa, does not a ffect general model inference. Light is represented by PAR, photosynthetically active radiat ion in watts per square meter. Maximum scaled light use, A max , is assumed to vary ex ponentially with biomass due to self shading and competition: A max = e cxB and where c is a constant (m 2 g 1 ). Because measurements of the light half saturation constant,

PAGE 26

26 PAR k , and A max were unavailable, these parameters were estimated from observations of biomass in s prings (Hauxwell et al. 2004). Rather than perform a formal sensitivity analysis, I conducted a series of numerical experiments wherein variation in tissue stoichiometry was evaluated as a function of , input concentrations and stoichiometry, light variation, and nutrient uptake parameters, K i . Model simulations were run using Python (version 2.6.6) using a Real valued Variable coefficient Ordinary Differential Equation solver method with a time step o f 0.01 d. Results Resource stoichiometry varied broadly across the 41 sites, with discharge varying from 0.01 to 17.95 m 3 s 1 . Molar C:N, C:P, and N:P ratios ranged from 5.7 to 471, 65 to 2100, and 0.28 to 89.4, respectively. Concentrations of DIC, NO 3 , a nd SRP ranged from 10 to 59 mg C L 1 , 36 to 2661 g N L 1 , and 7 to 165 g P L 1 , respectively. Whereas T (mean ± standard deviation (SD) = 21.67°C ± 1.5) and water source (the Floridan Aquifer) were consistent across sites, I observed large across site variation in other field parameters (SpC ranged fro m 17 to 1124 S cm 1 ; DO ranged from 0.3 to 9.8 mg L 1 ). Notably, for 26 springs in which water chemistry measurements were repeated over a 5 year period (2000 2005), the coefficient of variation (CV = SD / µ ) averaged 0.13, 0.10, and 0.16 for C:N, C:P, an d N:P, respectively, underscoring their unique chemical stability. Tissue stoichiometry varied signif icantly across taxa (Table 2 2). A two factor ANOVA showed a significant main effect of species on C:N (F 3,121 = 242.48, p < 0.001); all species had statis tically distinct C:N values , with higher C:N for vascular plants than for algal taxa (Table 2 2). There was no significant effect of light on tissue C:N ratios

PAGE 27

27 (F 1,121 = 0.94, p > 0.1), nor the interaction between species and l ight (F 3,121 = 1.78, p > 0.1) . In addition, although I observed a significant main effect of species for both N:P (F 3,121 = 33.74, p < 0.001), and C:P (F 3, 121 = 7.75, p < 0.001), I observed no significant light or light x species interaction effects for either C:P or N:P; all reporte d values are averaged across high and low light conditions (Table 2 2). I observed strong evidence for homeostasis acros s sites for all taxa (Figure 2 2). H values for C:N varied between 74.2 for Sagittaria to undefined for Lyngbya , Vaucheria and Vallisner ia . H values for C:P for Vallisneria and Sagittaria were 11.0 and 4.8, respectively; this evidence of weak C:P plasticity in the vascular plants was in contrast to the algal taxa for which H values were undefined. H values for N:P ranged from 7.5 for Sagit taria and 27.6 for Vallisneria , to values of 295 for Lyngbya and undefined for Vaucheria . I observed negligible effects of variation in resource stoichiometry or concentrations on species presence or absence (Table 2 3). Logistic regression results suggest a 1.32 fold increase in the odds of observing Vallisneria with one standard deviation increase in resource C:N (hereafter odds ratio (OR SD ); p < 0.001). Surprisingly, NO 3 had no effect on any species except Lyngbya , which was less likely to be present as NO 3 increased ( p = 0.014, slope = 0.0002, OR SD = 0.84). Increasing discharge increased the probability of observing Lyngbya ( p = 0.026, slope = 0.038, OR SD = 1.16) and Sagittaria ( p = 0.04, slope = 0.033, OR SD = 1.14) whereas increasing DO increased the probability of observing Vallisneria ( p < 0.00 1, slope = 0.091, OR SD = 1.24). Decreasing SpC produced a marginal increase in the probability of observing

PAGE 28

28 Sagittaria ( p = 0.05, slope = 0.0005, OR SD = 0.88). Variation in C:P, N:P, SRP, T, and pH had no sig nificant effect on the probabilit y of observing any of the taxa. Adjusting the Droop model to decouple hydraulic and biomass turnover, and to include light limitation , led to dramatic changes in how tissue stoichiometry responds to exogenous drivers. I exp lored model behavior by adjusting parameters that control nutrient supply rate (hydraulic turnover, and input concentrations, R in,N or R in,P ) and biological demand (uptake half saturation, K i and light, PAR ). Model runs using published values for K i , fix ed R in,P (0.8 mol P L 1 ) , but varying R in,N (0.008 to 40 mol N L 1 ) , yielded strongly plastic tissue stoichiometry at low , but increasingly homeostatic behavior as , and thus nutrient supply, increases (Figure 2 3 ). Increasing R in,N or R in,P also increases nutrient supply, and tissue stoichiometry showed similar patterns, with plasticity at low concentrations and homeostasis with in creasing concentrations (Figure 2 3 ). Homeostasis can also be induced by altering parameters that control nutrient demand . As light limi tation is relieved by increasing PAR, creating nutrient limiting conditions because of elevated biomass, I observed tissue plasticity that was absent under light limiting conditions (Fig ure 2 3 ). Similarly, varying the half saturation parameter (K N ), which control N uptake rates, can induce nutrient limitation, and thus plasticity, at high values, and homeostasis at low values, where g rowth is limited by light (Figure 2 3 ). When model parameters matched values expected in springs, effects on tissue stoichio metry of varying R in,N (range = 2 to 96 µ mol L 1 ) and (range = 10 50 d 1 ), which together set th e N supply rate, are clear (Figure 2 4); increases in both lead to increasing homeostasis. However, this effect was also affected by nutrient uptake rates

PAGE 29

29 (specifically, the half saturation parameter K i ). At pub lished values of K N , modest plasticity was obs erved at low resource N:P (Figure 2 4 ), but at lower values of K N (5% of published values) this plastici ty was less evident (Figure 2 4 ). Whereas varying K N can induce growt h limitation by nutrients (Figure 2 3 ), at values typical for springs, increasing K N induces small but increasing deviation f rom optimal tissue ratios (Figure 2 5) , even where biomass is constant and growth is light limited. At intermediate values of K N (2.2 to 8.7) and constant K P , H values fall between 434 and 99. Similarly, increasing K P , holding K N constant, increases tissue N:P ratios (for K P = 0.1 to 0.4, H ranges from 310 to 69). Discussion These results support the unequivocal conclusion that benthic autotrophs, both algae and vascular plants, grown in natural flowing water chemostats, exhibit constant element ratios (i .e., they are homeostatic, Figure 2 2), challenging the generalization that plants are stoichiometrically plastic (Loladze et al. 2000; Sterner and Elser 2002) . Whereas t his generalization has been increasingly scrutinized (Hall et al. 2005; Elser et al. 2010; Yu et al. 2011), measurements in these spring fed systems are particularly informative. Specifically, low temporal variation in discharge and chemistry makes tenable the assumption that biomass is at equilibrium with external supply. Whereas stoichiometric variation has been observed in response to growth rate or temperature (Chrzanowski and Grover 2008), relatively constant temperatures (CV=3%) and ecosystem primary production (Odum 1957) redu ce this source of uncertainty. Moreover, transient effects due to pulse disturbances or lags in response to time varying chemistry (Hall et al. 2005; Ã…gren 2008) are minimized. The presence of the same species across resource gra dients that span more than two orders of magnitude

PAGE 30

30 (e.g., N:P from 0.6 to 89) enables exploration of in situ tissue responses to environmental variation commensurate with chemostat experiments (Rhee 1978). Despite wide variat ion in environmental supply, my empirical results suggest distinct and consistent taxon specific stoichiometric ratios. As has been widely reported, algal species had significantly lower C:N ratios than vascular plants (Elser et. al. 2000), but higher N:P ratios (Hall et al. 2005, Table 2 2). The remarkable consis tency of taxon specific stoichiometry , despite large resource variation across springs , may indicate the presence of an evolutionary optimum (Ã…gren 2008) . I note that taxon specific optima have previously been inferred based on tissue stoichiometry under co limiting conditions (i.e., Q min,N :Q min,P ; Ã…gren 2008), reasoning that luxury uptake obscures this ratio when nutrients are supplied in excess . These results suggest, however, that under conditions of nutrient saturation, obser ved tissue ratios are highly stab le , converging on a taxon specific value that may be an evolutionary optimum (i.e., Q max,N :Q max,P ). It is unclear whether these two ratios are equal (i.e., Q min,N :Q min,P = Q max,N :Q max,P ), but I argue that the latter may be particularly informative because it represents tissue storage quotients when growth occurs without elemental constraint. This ratio is also far more likely to be observed in lotic systems because of the unlikeliness that environmental supply exactly matche s tissue demand, and because other factors (light, disturbance, and herbivory) can control autotroph growth and thus release them from nutrient limitation . The tissue ratio at nutrient saturation define s the luxury storage capacity for relatively immobile benthic autotrophs, and is a quantity that could inform competition over evolutionary time scales. Where the two ratios diverge , they may indicate different aspects of plant competitiveness, with the former (Q min,N :Q min,P ) indicating lower limits of resour ce

PAGE 31

31 acquisition and the latter (Q max,N :Q max,P ) indicati ng luxury storage capacity when nutrients do not constrain growth. The degree to which autotrophs actively regulate their internal nutrient composition remains unclear . Traditionally, autotrophs were mo deled as passive recipients of nutrients; in these conditions, saturation of internal quotients was interpreted as the result of passive diffusion losses out of the cell, or saturation of nutrient uptake carriers . However , it is becoming increasingly clear that plants actively regulate nutrient acquisition, with response s ranging from regulating of the timing of nutrient uptake (Ashwo rth et al. 2013), to adjusting the uptake kinetic parameters . Although the results presente d here do not directly inform this debate, remarkably stable autotrophic internal nutrient composition , despite enormous variation in resource concentrations, which should affect passive diffusion equilibria, suggests active regulation. Moreover, recent re sults on the timing of nutrient uptake in these spring ecosystems (Cohen et al. 2013), specifically the asynchrony of N (day) and P (night) assimilation, are strongly consistent with observations of active up and down regulation of nutrient acquisition ge nes in an aquatic autotroph (Ashworth et al. 2013). The model and empirical results together present strong evidence for the hypothesis that autotroph plasticity or homeostasis is driven by the balance between nutrient supply (concentration and hydraulic t urnover) and autotroph demand (light and nutrient u ptake saturation, Figure 2 3) . Where supply is high enough or demand low enough to induce growth limitation by light, tissue stoi chiometry is homeostatic (Figure 2 3 and Figure 2 4) . Recent work suggests n utrient supply and demand together

PAGE 32

32 describe limitation better than concentration alone, and thus may enable better predictions enrichment responses in lotic ecosystems (King et al. 2014) . New tools for measuring ecosystem level autotroph demand (Heffernan and Cohen 2010; Cohen et al. 2013) may allow experimental testing of stoichiometric effects of light vs. nutrient limitation . The central prediction that emerges from these results is that where re source supply saturates demand, i.e. induces limitation by light or some other element, autotrophs should be stoichiometrically homeostatic for non limiting nutrients. The model predicts stark differences in tissue stoichiometry between light and nutrient limited growth, but also predicts modest effects of uptake kinetics even whe n growth is lig ht limited (Figure 2 5) . The Droop model envisions intra cellular stores of nutrients, which govern growth limitation, replenished from external supplies following Michaelis Menten uptake kinetics. When the half saturation constant governing uptake kinetics (K i ) is near external concentr ations , but light limits growth, e.g. where high hydraulic turnover enable s saturation of nutrient demand , the model predicts an equilibrium nutrient quotient (Q i ) that adjusts the resulting tissue ratios; resulting H values are large, indicating homeostasis. This kinetic plasticity disappears as K i values decline far below ambient concentrations . The model does not consider variation of K i with factors such as temperature or ambient nutrient concentrations, and thus acclimation to ambient conditions by alteration of uptake parameters may contribute to the observed autotrophic response (Collos et al. 2005). However, given the range of K i values and input concentrations modeled, I would not expe ct the modeled response to be a dram atic overestimation of uptake. Relative constancy of nutrient concentrations within each system may also limit acclimation; as such, strict homeostasis ob served for

PAGE 33

33 the algal taxa (Figure 2 2 ) may indicate extremely low K i values. Likewise, modest but consistent covariation between vascular plant tissue and re source stoichiometry (Figure 2 2 ), although still indicative of homeostasis, may be consistent with higher K i values (Klausmeier et al. 2004). The model results prov ide strong evidence that tissue homeostasis arises with saturation of nutrient demand. Constant tissue ratios occur in the model only when resource supply is large rel ative to autotroph demand (Figure 2 3), and converge on the value of Q max,N :Q max,P . Herea fter I refer to this ratio as the optimal species specific tissue stoichiometry, reasoning that the relative size of internal luxury stores is a function of a long term opti mization to ecological niches. Further work to establish this optimality could supp ort inference of taxon specific nutrient limitation s tatus across multiple systems. That is, deviatio n between measured and optimal, i.e. Q max,N :Q max,P , tissue stoichiometry may be diagnostic of nutrient limitation because conditions of P limitation would result in tissue s toichiometry above the optimum, i .e. N:P > Q max,N :Q max,P , whereas N limitation yields tissue stoichiometry below the optimum, i.e., N:P < Q max,N :Q max,P . Although the empirical data provide limited support for the possibility that deviatio n from optimum indicates nutrient limitation, principally because tissue stoichiometry was consistently homeostatic and therefore always indicative of nutrient saturation, nutrient limitation of autotroph growth has been documented in other lotic systems ( Grimm and Fisher 1986; Slavik et al. 2004), and could provide an experimental venue for testing this prediction. The idea that tissue stoichiometry can be used to indicate limitation status is clearly not new (Koerselman and Meuleman 1996), but the standar d inference, wherein tissues are N limited (N:P < 14), P limited (N:P >

PAGE 34

34 16 ), or co limited, omits the possibility of limitation by any other factor, and precludes variation among taxa. Determination of species specific optimal tissue ratios could potential ly improve upon this inference, by enabling consideration of other limiting factors (the primacy of which would induce growth at optimum stoichiometry) and the presence of inter specific variation in optima, a need clearly warranted based on strong evidenc e of taxonomic variation in this study and others (Demars and Edwards 2007; Yu et al. 2011) . Although there are currently no inventories that report taxon specific optimum tissue stoichiometry, once established, this optimum (Q max,N :Q max,P ) may provide a r eference point for inference of nutrient limitation status based on measured tissu e deviation from that optimum. This would further enable assessment of nutrient limitation by other elements (e.g., Ca, Fe, K, Mg, Mn, and S) that have received relatively li mited attention for their role i n controlling primary production (Ã…gren 2008). Ecological stoichiometry uses element ratios as a unifying framework for flows across many levels of biological organization . As such, an understanding of controls on autotroph stoichiometry is integral for ecosystem studies. This work contributes to a growing body of literature that challenges the generality of autotroph plasticity . Although the disagreement between chemostat results of plasticity and environmental samples that exhibit a range of stoichiometric behaviors, including strict homeostasis, remains an area in ne ed of theoretical attention, this work offers an explanation for when and why autotrophs vary based on nutrient limitation, and more specifically the magnitude and frequency of nutrient delivery (N and P) relative to demand (controlled by light and other factors). This hypothesis may have important implications for understanding links

PAGE 35

35 between autotroph growth, their environmental supply of mineral nutrients, and the food webs they sustain.

PAGE 36

36 Table 2 1. Model parameters used in numerical solutions unless otherwise noted. Parameter Meaning (units) Value Hydraulic turnover rate (d 1 ) 0.01 50 R in, P Input P concentration ( mol P L 1 ) 0.2 6.4 R in, N Input N concentration ( mol N L 1 ) 0.008 320 v maxh, P P upper bound uptake rate ( mol P g 1 d 1 ) 2.1 v maxl,P P lower bound uptake rate ( mol P g 1 d 1 ) 0 .104 v maxh, N N upper bound uptake rate ( mol N g 1 d 1 ) 44.9 v maxl,N N lower bound uptake rate ( mol N g 1 d 1 ) 2.24 K P P half saturation constant ( mol P L 1 ) 0.2 K N N half saturation constant ( mol N L 1 ) 4.33 max Realized maximum growth rate 0.022 m Mortality rate (d 1 ) 0.0009375 Q min, P Minimum P quota ( mol P g 1 ) 2.097 Q min, N Minimum N quota ( mol N g 1 ) 45.4 Q max, P Maximum P quota ( mol P g 1 ) 94.69 Q max, N Maximum N quota ( mol N g 1 ) 2050 Q max, N :Q max,P Saturated optimal N:P ratio 21.65 PAR Light available for uptake (W m 2 ) 300 PAR k Light half saturation constant (W m 2 ) 125 Klausmeier et al. 2007 Cherif and Loreau 2010

PAGE 37

37 Table 2 2 . Autotrophic molar tissue ratios by species (me an ± 95% confidence interval). Differen t lowercase letters denote significant differences at p Honestly Significant Difference test) be tween taxa within each column. Tissue samples from sun and shade environments were not sign ificantly different for any taxon so values are comb ined. Species C:N C:P N:P Lyngbya 7.71 ± 0.56 a 387.32 ± 36.62 b 51.38 ± 5.70 a Vaucheria 9.89 ± 0.32 b 539.05 ± 58.85 a 54.35 ± 5.48 a Sagittaria 15.39 ± 0.83 c 485.24 ± 59.09 a,b 31.71 ± 3.72 b Vallisneria 18.48 ± 0.91 d 391.18 ± 32.29 b 21.65 ± 2.51 b

PAGE 38

38 Tabl e 2 3. Fitted logistic regression slopes for predicting species presence and absence based on resource chemistry predictors. Mean values for sites where taxa were present and absent, respectively, are reported in parentheses for statistically significant a ssociations. Resource ratios are reported on a molar basis. Significance codes: *** = 0.001, ** = 0.01, * = 0.05, ns = non significant. Sagittaria overall best fit model ( 2 = 1.21, degrees of freedom (df) = 3, p = 0.75); Vallisneria overall best fit model ( 2 = 4.23, df=4, p = 0.37); Lyngbya overall best fit model ( 2 = 3.79, df = 5, p = 0.58). Parameters Mean±SD Sagittaria ( n =7) Vallisneria ( n =13) Lyngbya ( n =13) Vaucheria ( n =27) Resource C:N 68±91.2 ns 0.003*** (95, 40) ns ns Resource C:P 594±489 ns ns n s ns Resource N:P 21.4±20.4 ns ns ns ns DIC (mg L 1 ) 35.9±12.9 ns ns ns ns NO 3 ( g L 1 ) 1140±853 ns ns 0.00019* (691, 1423) ns SRP ( g L 1 ) 37±33 ns n s ns ns Discharge (m 3 s 1 ) 3.1±3.9 0.033* (5.6,2.6) ns 0.038* (5.3, 2) ns DO (mg L 1 ) 3.6±2.4 ns 0. 091*** (4.8, 3) ns ns SpC ( S cm 1 ) 372.4±255.3 0.0005* (295,389) ns ns ns T (°C) 21.8±1.32 ns ns ns ns

PAGE 39

39 Figure 2 1. Conceptual diagram relating supply (hydraulic turnover, , and input concentration, ( R in,N and R in,P ) and light (PAR) to demand in a benthic chemostat.

PAGE 40

40 Fig ure 2 2. Molar tissue vs. resource ratios for C:N, C:P, and N:P for (A) Lyngbya wollei , (B) Vaucheria sp., (C) Sagittaria kurziana , and (D) Vallisneria americana . H values uniformly indicate strong homeostasis acros s all taxa an d element ratios. All log log slopes, except Vallisneria americana C:P, are not significantly different from zero.

PAGE 41

41 Fig ure 2 3. Factors that control tissue stoichiometry include nutrient supply (hydraulic turnover, ; input concentration, R in,N and R in,P ) and biotic demand (light, PAR; uptake half saturation, K i ). (A) Effects of increasing (0.05, 0.25, 50) with R in,P constant (0.8 mol L 1 ) while R in,N increases (0.008 to 40 mol L 1 ). (B) Eff ects of increasing lig ht (100 to 1060 W m 2 ) on tissue stoichiometry; note homeostasis under light limitation and plasticity under nutrient limitation. (C) Effects of varying R in,P on tissue stoichiometry; low R in,P (0.2 mol L 1 ) yields plasticity while increasing R in,P (inter mediate = 1.2 mol L 1 , high P = 6.4 mol L 1 ) yield increasing homeostasis. (D) Effects of K N (low = 0.22 mol N L 1 , intermediate = 4.33 mol L 1 , high = 21.65 mol N L 1 ) on tissue stoichiometry; increases in K N causes below optimal N:P; K P is low (0.01 ).

PAGE 42

42 Figure 2 4 . Effects of resource concentration and hydraulic turnover ( ) on tissue stoichiometry. Black lines are N:P = 40:1 at different concentrati ons; grey lines are N:P = 10:1. Tissue N:P converges on species optima with increased concentration and , but more slowly at (A) higher values of K N than at (B) lower values.

PAGE 43

43 Figure 2 5. Varying the half saturation constant (K N , K P ) creates weak plasticity in tissue stoichiometry even when light limits growth (i.e., constant biomass across simulatio ns). Lower H (increasing apparent plasticity) occurs with (A) increasing K N (at constant K P = 0.1), and (B) increasing K P (at constant K N = 2.165).

PAGE 44

44 CHAPTER 3 EXPERIMENTAL EFFECTS OF FLOW AND GRAZERS ON PRIMARY PRODUCER COMMUNITY STRUCTURE IN FLORIDA SPRI NGS Introduction Eutrophication, particularly driven by increased nitrogen and phosphorus loading, has been shown to critically impact aquatic ecosystem structure by inducing increased autotrophic production, shifts in species composition, and loss of habi tat for higher trophic levels (Duarte 1995, Carpenter et al. 1998). Resolving the thresholds of enrichment at which ecosystem structure deficits occur, particularly in flowing waters, is an active area of inquiry in water quality management (Conley et al. 2009). Further management complications may arise because recovery from eutrophication is not simple, and restoration trajectories often fail to return to the reference state, with the eutrophic state persis ting (Duarte et al. 2009). Although numerous stud ies have described the ecosystem structure impacts due to eutrophication in coastal, lak e, and wetland ecosystems ( Duarte 1995, Gaiser et al. 2006, Schindler et al. 2008), the value of autotrophic taxa, part icularly submerged vascular plants, as indicator s of ecological integrity and restoration success in flowing water systems , has only recently begun to be evaluated (Pedersen et al. 2007, Lorenz et al. 2012). Many Florida springs and spring fed rivers have exhibited autotrophic species shifts, with benthi c filamentous algal species replacing rooted vascular plants (FSTF 2000). Rising nitrate (NO 3 N) concentrations have been cited as the primary causal agent for this change (FDEP 2013). In an effort to reverse these ecological changes, specifically the nui sance growth of filamentous algae (FSTF 2000), the Florida Department of Environmental Protection (FDEP) and the US Environmental Protection Agency (USEPA) have adopted numeric nutrient criteria (NNC), with a singular

PAGE 45

45 emphasis on nitrate reduction , to conc 0.35 mg N L 1 , for water discharging from springs (FDEP 2013). The explicit hypothesis that autotrophs in these systems are N limited and that nutrient enrichment favors the overgrowth of rooted vascular plants by fast growing algae was critic ally reviewed by Heffernan et al. (2010), who concluded that NO 3 N concentration does not explain algal cover or biomass . Although nutrient concentrations are strong predictors of primary producti on in lakes and estuaries, this link is less clear in flowi ng waters, likely because nutrient resupply from advection readily satisfies biological demand. A r ecent study fed rivers, proposed that flux , not concentration, indexed to autotrophic demand, is the most effective variable for predicting growth responses (King et al. 2014). Th at work implies that high flux , even at low concentration, can satisfy autotrophic demand, a conten tion supported by the saturation kinetics of nutrient uptake observed in streams (Earl et al. 2006, Co vino et al. 2010). Moreover, evidence from these spring fed , natural flowing water chemostats suggests that algal and vascular taxa are stoichiometrically homeostatic over large gradients of nutrient ratios (Nifong et al. 2014). Simulation model ing suggest s this occurs because of rapid hydraulic turnover (~ 10 50 d 1 ) which saturates autotrophic demand at NO 3 N concentrations of ~ 0.02 mg L 1 , far below the established NNC. In short, evi dence for nutrient limitation of autotrophs in these spring fe d ecosys tems, and thus the prospect of nutrient reductions alone restoring a utotroph community structure, are limited. In the absence o f nutrient limitation, alternative factors such as flow (Rolls et al. 2012, King et al. 2014) and herbivory (L iebowitz et al. 201 4) may be more important in

PAGE 46

46 determining the outcome of autotrophic competition between vascular plants and algal taxa (Heffernan et al. 2010). There is evidence that each of these alternati ve factors may play a dominant role in determining the spatial and temporal distr ibution of i nteractions with nutrient concentrations, in controlling autotrophic competition , remain unclear. Moreover, most research has focused on the growth and density of algal taxa. Dramatic differences between vascular plants and algal species with respect to growth form and nutrient storage suggests that understanding how these multiple factors influence vascular taxa success is critical to preserving and restoring desir ed ecosystem structure. I hypothesized 1) increased flow velocity will induce decreased algal biomass , 2) higher grazer density facilitates vascular plant success and 3) flow and grazers interact to create environments that are conducive to vascular plant success over algal species. I use d an in situ experimental approach to assess the relative importance of top down and bottom up control on changes in ecosystem structure (Hilton et al. 2006, Brown et al. 2008). Specifically, I employed a multi factorial ex perimental approach wherein vascular plants, algae and grazer densi ties were manipulated within in situ mesocosms over a gradi ent of nutrient concentrations across sites and flow rate within sites . Methods Study Sites This study was conducted in three larg e spring fed river ecosystems in north central Florida: Alexander Spring Creek in Lake County, Ichetucknee River (with multiple source springs) in Columbia County, and Gum Slough in Sumter County. Like fed rivers, these ecosystems exhibit unparalleled stability in

PAGE 47

47 discharge, temperature and solute concentrations (Heffernan et al. 2010), coupled with substantial across spring resource gradients , in response to anthropogenic and geologic variation (Scott et al. 2004). Nutrient fluxes in the springs are dominated by mineral forms, and low temporal variation makes tenable the assumption that biomass and tissue stoichiometry are close to equilibrium with the environment, allowing meaningful inference from synoptic measures of water and ti ssue chemistry (Cohen et al. 2013, Nifong et al. 2014). The three study sites were selected from the gradient of nutrient concentration s and molar resource ratios. Whereas dissolved inorganic carbon (DIC) and soluble reactive phosphorus (SRP) concentration s differed only slightly among sites, nitrate (NO 3 ) concentrations, variation in which is controlled principally by anthropogenic pollution, varied enormously, spanning two orders of magnitude (Table 2 1). Variation in resource concentrations meant that s elected sites spanned a wide gradient of C:N, C:P, and N:P ratios, which is important because stoichiometry varies significantly among the dominant plant taxa (Nifong et al. 2014). Additionally, sites were selected for their diversity of autotrophic taxa , including two target macrophytes , tapegrass, Vallisneria americana Michaux 1803, hereafter Vallisneria , and springtape, Sagittaria kurziana Glück 1827, hereafter Sagittaria , and two filamentous algal species Lyngbya wollei (Farlow ex Gomont) Speziale and Dyck 1992, hereafter Lyngbya , and Vaucheria disperma de Candolle 1801, hereafter Vaucheria . Experimental D esign A 3x3x3 split block design was implemented with three qualitative levels of flow (low, medium, and high), three levels of vegetation (100% vascular plant cover, 50/50 mixture of vascular plants and algae, and 100% algae cover), and three levels of grazer densit y (absent, present, and ambient, Figure 3 1) . Cage treatments to exclude/ include

PAGE 48

48 grazers were PVC frames (30.5 x 30.5 x 61 cm, W x L x H) , covered by black pol yethylene netting (4.76 mm mesh, Figure 3 2) . Three flow velocity treatments were applied at selected locations (transects) within each site based on surface velocity va riation along the spring fed run . This within channel flow va riation was assumed to be a proxy for variation in flow regime that might result from climate variation and groundwater pumping. At each flow and vegetation level I installed an uncaged mo nitoring plot (control), a cage control (cage raised 5 cm above subs trate) to ass ess non treatment cage effects, i.e. shading, grazer refugia, etc. (cage control) , a caged grazer inclusion plot (grazer+), and a caged grazer exclusion plot (grazer ). Inclusion and exclusion cages were buried 5 cm into the substrate to preve nt grazer movement in to or out of cages. For the grazer inclusion treatment, I added 150 g wet weight m 2 of snails ; all added grazers were collected from a single on site location. Grazer presence was patchy within sites with ambient densities ranging fro m 60 to 450 g wet weight m 2 . The same grazer species was not used at all site s ; Elimia floridensis was used at Gum Slough, whereas Pomacea paludosa was utilized at Alexander Spring , and Micromenetus sp. was employed at Ichetucknee River. Although it would have been preferable to have used the same grazer taxon across sites , there is evidence that Pomacea paludosa favor s low C:N algae (Sh arfstein & Steinman 2001) and E. floridensis has been shown to control the alga Lyngbya wollei (Li ebowitz et al. 2014). A lthough there is little literature on s pecies specific dietary preferences in the genus Micromenetus , planorbid snails typically graze on epiphytic algae (Underwood & Thomas 1990).

PAGE 49

49 Field and Laboratory Methods Aboveground biomass (g wet mass m 2 ) of both a lgal and vascular plant taxa was measured at the end of four weeks . Excess water was blotted from the biomass before weighing; dry mass of algal and vascular taxa was highly correlated with wet mass (algal taxa, = 4.58, R 2 = 0.97, n = 108; vascular taxa, = 13.77, R 2 = 0.87, n = 105). From wet mass measurements, the relative standing stock was calculated as: ((measured treatment biomass)/(measur ed control biomass)) (3 1 ) The number of stems per plot was counted at the beginning and end of the experiment ; net stem production of vascular plants (# stems m 2 ) was calculated by subtracting the number of stems per plot at the beginning of the experiment from the number of stems at the end of the experiment . Beginning algal volume per plot was calculated from four measurements of algal volume (cm 3 m 2 ) within each plot; the change in algal volume, an estimate of net production, was calculated as th e change in volume of algae between the beginning and end of the experiment . Surface flow velocity at each transect was measured weekly using the float method and water samples were collected on a weekly basis from t he main spring vent of each system, and analyzed for NO 3 , SRP, and DIC. DIC was maintained in solution until analysis with HgCl 2 . DIC was measured using a UIC (Coulometrics) 5011 CO 2 coulometer coupled with an AutoMate automated carbonate preparation device. SRP and NO 3 were measured using EPA Methods 365.1 and 353.2, respectively. Statistical Analysis package (Pinheiro et al. 2014), to perform a linear mixed effects analysis of the relationship between flow and grazers on the relative standing stock and net production

PAGE 50

50 of algal and vascular taxa. As fixed effects, I entered flow rate (low, intermediate, high) and grazer treatment. Initial vegetation type was also accounted for as a fixed effect , to determine if the factors of flow and grazers interacted differently across vegetation types. Site was included as a random effect. I conducted post hoc analysis using the Results Site C haracteristics Stoichiometry and nutrient concentrations were distinct across the three sites (Table 3 1); molar resource ratios ranged considerably for C:N (26.4 1 1396.77 ) , C:P (1200.36 3514.96 ), and N:P (1.4 1 133.26 ) . Resource concentrations of DIC, NO 3 , and SR P had ranges of 19. 55 3 1.50 mg DI C L 1 , 0.0 3 1.34 mg N O 3 L 1 , and 0.02 0.05 mg P L 1 . Highest NO 3 concentrations were measured at Gum Slough. Measured flow velocities ranged from 2.9 to 15.2 cm s 1 with average flow velocity by transect ranging fr om 4.5 to 13.0 cm s 1 over the course of the experiment. Within transects, flow velocities were relatively constant over the course of the experiment, with a mean coefficient of variation (CV = SD/ ) of 0.22. A utotroph Characteristics by Site and Initial Ecosystem S tructure S tanding stocks of vascular plants at the end of the 4 week experiment were highest at Alexander Spring Creek , averaging 1716 (g wet mass m 2 ), intermediate at Ichetucknee with 1 264 (g wet mass m 2 ), and lowest at Gum Slough (1204 g wet mass m 2 ). Although initial stem densities we re similar at Gum Slough (474 stems m 2 ) an d Alexander Spring Creek (483 stems m 2 ), they were much higher at Ichetucknee (1006

PAGE 51

51 stems m 2 ); net stem pro duction mirrored final standing stock patterns , with the highest values observed a t Alexander Spring Creek (233 stems m 2 ), a slightly lo wer value at Ichetucknee (220 stems m 2 ) and the lo west value at Gum Slough (179 stems m 2 ). Not surprisingly, initial stem densities and final standing stocks of vascular plants were highest in the 100% vascular block, intermediate in the 50%/50% block, and lowest in the 100% algae block at all sites. H owever, net stem production of vascular plants was higher in the 50%/5 0% block than the 100% vascular plant block at Ichetucknee , but not at Alexander or Gum Slough. Final standing stocks of algae we re highest at Gum Slough (907 g wet mass m 2 ), int ermediate at Ichetucknee (221 g wet mass m 2 ) and lowest at Alexander (46 g w et mass m 2 ). Initial algal volume was als o largest at Gum Slough (2157 cm 3 m 2 , followed by Alexander (1946 cm 3 m 2 ), and Ichetucknee (1485 cm 3 m 2 ). Initial algal volumes were highest in the 100% algae block within each flow transect at each site; howeve r all sites experienced a loss in algal volume within the 100% algae block. Rel ative algal standing stocks ended up being highest in the mixed taxa block at both Gum Slough (highest NO 3 N) and Alexander Spring Creek (lowest NO 3 N). Evaluation of Linear M ixed Effects (LME) Interclass correlation provides no evidence of an effect of site on the relative standing stock o f vascular taxa (ICC1 = 0.03) or net stem production (ICC1 = 0.02). Similarly, relative algal standing stocks (ICC1 = 0.03) and net algal production (ICC1 = 0.12) were not influenced by site. Grazer E ffects Overall, grazers significantly influenced relative vascular plant standing stocks (Table 3 2). Relative standing stocks of vascular taxa were significantly lower in the

PAGE 52

52 grazer exclusion treatment compared to the cage control treatment ( = 0.59, SE = 0.18, p = 0.005, Figure 3 3 ). The relative standing stock of vascular plants was significantly higher in the caged control plot compared to the monitoring plot ( = 0.46, SE = 0.17, p = 0.03 ; Figure 3 3 ). Grazers also interacted with vegetation type to significantly influence the relative stand ing stock of vascular taxa (Figure 3 4 ). Grazer exclusions led to significantly higher relative algal standing stocks compared to al l other grazer trea tments (Figure 3 5 ). Additionally, there was a significant interaction of grazer treatment a nd vegetation types (Figure 3 6 ). Relative algal standing stocks in grazer exclusion plots were significantly higher in the 100% vascular vegetation type than in th e 50% vascular/50% algae vegetation type ( = 6.38, SE = 1.66, p = 0.007 ) or the 100% algae vegetation type ( = 8.80 , SE = 1.72 , p < 0.001). Flow E ffects Flow significantly influenced relative algal standing stocks (Table 3 2) , with the highest standin g stocks occurring in intermediate flow en vironments (Figure 3 7 ). The interaction of flow and vegetation type significantly influenced relative algal standing stocks (Table 3 2) , with the influence of flow on algal standing stocks being most pronounced in the 100 % vascular vegetation type (Figure 3 8 ). A significant interaction of grazers and flow showed the grazer exclusion treatment at intermediate flow velocity had significantly higher relative algal standing stocks than the grazer exclusion treatment a t both low ( = 11.41, SE = 1.60, p < 0.001) and higher flows ( = 11.70, SE = 1.72, p < 0.001). Finally, the interaction of grazers, flow and vegetation type was found to significantly influence relative algal standing stocks (Table 3 2).

PAGE 53

53 Vegetation E ff ects Vegetation type significantly influence d both the relative standing stock and net stem production of vascular plants (Table 3 2). Post hoc analysis revealed that vascular relative standing sto ck and net stem production were significantly lower in alga l plots compared to 100% vascular plots (relative standing stock, = 0.69 , SE = 0.16, p < 0.001; net stem production, = 275.23, SE = 77.67, p = 0.001) and 50%vascular/50%algae plots (relative standing stock, = 0.42 , SE = 0.15, p = 0.01; net stem production, = 218.72 , SE = 74.06, p = 0.009). Neither net stem production ( = 56.51, SE = 76.25, p = 0.74) nor relative standing stock ( = 0. 27, SE = 0.15, p = 0.18) was significantly different between 100% vascular and 50% vascular/50% algae levels. Discussion This mesocosm experiment sugges ts that al ternative fa ctors other than NO 3 N concentrations, may play important roles in maintaining and restoring autotrophic community structure in Florida spring ecosystems. Although the N limitation hypothesis in Florida springs has been challenged (Heffernan et al. 2010), faced with shifts in autotrophic species composition , and in the absence of demonstrated alternatives, management strategie s continue to rely on nutrient load red uctions. In this study , flow and herbivory influence d the outcome of competition between alga l and vascular taxa across a gradient of resource nutrient concentrations . In particular, results suggest herbivory exert s significant control over algal taxa and facilitate s the success of the submerged vascular plants , which form the autotroph foundation in these ecosystems

PAGE 54

54 and provide critical structural benefits, such as refugia for higher order consumers (Pinto et al. 2006) and increased sedimentation rates (Madsen et al. 2001). Grazers significantly influence d the outcome of competition between vascul ar and algal taxa. In flowing water , in situ mesocosms there was a significant reduction i n standing s tock of vascular taxa in grazer exclusion treatments (Figure 3 3 ), indicating t hat t he absence of grazers can be detrimental to the success of vascular ta xa. Moreover, exclusion of grazers was found to more than triple the sta nding stock of algal taxa (Figure 3 5 ). The magnitude of the effect varied across vegetation types , with 100% vascular plots experiencing the highest rela tive algal standing stocks (Fi gure 3 6 ). A large number of studies have found similar exclusion effects of grazers on periphyton (Rosemond et al. 2000) and diato ms (Power et al. 1988); however, there are few in situ experimental demonstration s of grazer control of benthic algae in spri ng ecosystems (Liebowitz et al. 2014). Algal taxa display low C:N ratios and thus may supply a substantial amount of nitrogen to grazers (Nifong et al. 2014). However, C:P and N:P ratios of algal taxa are relatively high (~50) and food low in P can limit g rowth rates of Elimia sp. (C:P ~ 203) in oligotrophic systems (Stelzer and Lamberti 2002). As such, rates of herbivory across grazer and algal taxa in springs remain an open area of inquiry. Nevertheless, these results suggest that grazers actively manage a lgal biomass through differential herbivo ry across vegetation types (Figure 3 6 ) , and facilitate vascular plant success across gradients in nutrient supply. Grazer inclusions failed to exert significant control on either vascular (Figure 3 3 ) or algal rela tive stand ing stocks (Figure 3 5 ), suggesting that grazer inclusion densities may not have been high enough to produce significance. However, relative vascular

PAGE 55

55 standing stocks were significantly hi gher in cage control plots (Figure 3 3 ) and among grazer tr eatments acros s vegetation types (Figure 3 4 ). Grazers were frequently observed attached to the outside of cages, and although grazer presence was extremely patchy, when present, they could be foun d in high density aggregations of ~200 individuals per m 2 . Little is known about controls on grazer aggregations in these spring systems. T hus, there is a need to further investigat e the threshold grazer density necessary to influence the outcome of competition between vascular plants and algae . Contrary to my hyp othesis, algal standing stocks did not exhibit a linear decrease wit h increasing velocity. I nstead, algal biomass was highest at in termediate flow velocities (Figure 3 7 ) for both 100% vascular and 50% vascular/ 50% algae vegetation types (Figure 3 8 ). T her e was , however, a small linear decrease in relative algal standing stocks from low to high flows for algal plots. Algal export , as a result of increased flow , has been observed (King 2014), a lbeit at higher flow velocities. T hus , further experimental work at higher flow velocities may be needed to observe significant algal export across vegetation types. This research explored out comes of autotrophic competition between algal and vascular taxa in Florida springs. B oth top down and bottom up factors , as well as their interactions, we re considered. Whereas nutrient reduction policies such as numeric nutrient criteria can help achieve benefits , including reduction of downstream nutrient exports (Smith et al. 2006), mitigation of potential human health imp acts ( Ward et al. 2005), and prevention of endocrine disruption in native fauna (Guillette and Edwards 2005), findings of this study indicate that other factors need to be considered when managing ecosystem structure in these flowing water systems.

PAGE 56

56 Table 3 1 . M ean resource variable s by site with resource ratios calculated on a molar basis. Site DIC (mg L 1 ) NO 3 (mg L 1 ) SRP (mg L 1 ) C:N C:P N:P Alexander 19.55 0.0 3 0.04 1396. 77 1200. 36 1.4 1 Ichetucknee 3 1.50 0. 32 0.05 122.34 1 751.38 1 5.44 Gum Slough 30.33 1. 34 0.02 26.4 1 3514. 86 133. 26

PAGE 57

57 Table 3 2. Linear mixed effects model results evaluating the effects of grazers, flow, and vegetation as well as their interactive effects by tax on specific parameter . Taxa Metric F value p value SAV Relative biomass Inter cept 349.66 <0.001 Grazers 4.42 0.01 Flow 1.97 0.15 Vegetation 10.41 <0.001 Grazers x Flow 1.21 0.31 Grazers x Vegetation 2.69 0.02 Flow x Vegetation 1.97 0.11 Grazers x Flow x Vegetation 1.12 0.36 Net stem production Intercept 50.37 <0.001 Grazers 1.70 0.17 Flow 1.42 0.25 Vegetation 7.70 0.001 Grazers x Flow 0.91 0.49 Grazers x Vegetation 0.97 0.45 Flow x Vegetation 0.52 0.72 Grazers x Flow x Vegetation 0.51 0.89 Algae Relative biomass Intercept 8.96 0.003 Gr azers 4.56 0.005 Flow 6.09 0.004 Vegetation 2.05 0.14 Grazers x Flow 5.91 <0.001 Grazers x Vegetation 3.28 0.01 Flow x Vegetation 3.41 0.01 Grazers x Flow x Vegetation 4.07 <0.001 Net Production Intercept 0.10 0.75 Grazers 1.39 0.26 Flow 0.38 0.68 Vegetation 9.90 <0.001 Grazers x Flow 1.89 0.09 Grazers x Vegetation 0.40 0.88 Flow x Vegetation 0.76 0.56 Grazers x Flow x Vegetation 0.53 0.88

PAGE 58

58 Figure 3 1. Diagram of experimental mesocosms design. Experim ental unit grazer + grazer control Transect 4 cage control Alexander Low N Ichetucknee Intermediate N Gum Slough High N Transect 1 100% Algae 50% Algae / 50% Vascular 100% Vascular Tran sect 2 Tr ansect 3 100% Algae 50% Algae / 50% Vascular 100% Vascular 100% Algae 50% Algae / 50% Vascular 100% Vascular 100% Algae 50% Algae / 50% Vascular 100% Vascular 100% Algae 50% Algae / 50% Vascular 100% Vascular 100% Algae 50% Algae / 50% Vascular 100% Vascular Transect 6 Transect 5 100% Algae 50% Algae / 50% Vascular 100% Vascular 100% Algae 50% Algae / 50% Vascular 100% Vascular Transect 8 Transect 9 Transect 7 100% Algae 50% Algae / 50% Vascular 100% Vascular

PAGE 59

59 Figure 3 2. Diagram of caging treatment and cage control design.

PAGE 60

60 Figure 3 3 . Effect of grazer treatment on the relative sta nding stock of vascular plants. Lowercase letters denote a s ignificant differenc e at p < 0.05. Bars depict 95% confidence intervals.

PAGE 61

61 Figure 3 4 . Interaction plot showing that grazer inclusion led to higher relative standing stocks in the vascular plant treatment rela tive to the mixed vegetation treatment. Vascular plant relative standing st ocks were uniformly low in the 100 % algae treatment.

PAGE 62

62 Figure 3 5 . Grazer exclusion led to significantly higher relative standing stock of algal taxa. Lowercase letters denote si gnificant differences at p < 0.05. Bars depict 95% confidence intervals.

PAGE 63

63 Figure 3 6 . Interaction plot showing grazer exclusion resulted in significantly higher relative algal standing stock in mixed and all vascular vegetation levels.

PAGE 64

64 Figure 3 7 . Relative algal standing stocks in intermediate flows were significantly higher than stocks in lower or higher flow environments. Lowercase letters denote significant differences at p < 0.05. Bars depict 95% confidence intervals.

PAGE 65

65 Figure 3 8 . Interaction plot showing the effect of flow am ong different vegetation types. Highest relative algal standing stocks were found at intermediate flows in 100 % vascular vegetation.

PAGE 66

66 CHAPTER 4 ON THE COUPLING OF AUTOTROPHIC STOICHIOMETRY TO ECOSYSTEM METABOLISM Introduc tion Ecosystem metabolism integrates carbon dynamics specifically, primary production and respiration across space and time, informing both the timing and magnitude of C uptake and release at the system scale (Young et al. 2008). Because ecosystem C fl uxes respond to multiple factors including light availability, temperature, hydrology (flow rate, turbulence) , inorganic carbon, nut rient supply and organic matter supply , metabolis m measurements yield insights into the controls and dynamics of carbon cycl ing ( Young & Huryn 1999, Mulholland et al. 2001, Tank et al. 2010) . By integrating across space, time and organisms, this measurement implicitly scales the processes of primary production and respiration occurring at the organismal level to the production and consumption of solutes ( dissolved oxygen ; Od um 1956 ) at the scale of the ecosystem . R ecent work has demonstrated the presence of coherent patterns of gross primary production both within and across sites and seasons (Mulholland et al. 2001, Valett et a l. 2008). Moreover, r ecent m ethod refinements have enabled measurements of diel signals of carbon C, nitrogen N, and phosphorus P, yielding in sight into the coupling of elemental metabolism at hourly time scales ( Heffernan & Cohen, 2010, Pellerin et al. 20 12 , Cohen et al. 2013 ) . High temporal resolution sensors provide simultaneous time series measures of multiple solutes , enabling estimation of element use at the ecosystem scale , and testing of ecological stoichiometry theory at the same daily time step s a s metabolism measurements. Furthermore, these sensors enable new application of stoichiometry scaling from the organism to the ecosystem.

PAGE 67

67 E cosystem stocks, in terms of resource and tissue ratios have been examined, but the nature of the relation between st ock tissue nutrient ratios of algae and vascular plants , and the stoichiometric ratios of the fluxes (i.e., ecosystem metabolism C:N , hereafter C:N E ) that sust ain them , remains unexplored. I seek to answer how stock tissue ratios are related to the ratio o f fluxes measured as ecosystem metabolism . Here, I examine how the stoichiometr y of autotroph tissues scales to the stoichiometry of ecosystem metabolism across a gradient of autotrophic cover type. W here autotroph tissues are plastic (i.e., demand is nutr ient limited) , ecosystem stocks ( resource C:N, C:N R ) and tissue stocks (C:N A ) are expected to align and this will be reflected in ecosystem metabolism measurements (i.e. C:N R = C:N A = C:N E ). In contrast, if tissues are homeostatic , resource and tissue rati os are unrelated and tissue ratios will differ among taxa (i.e. C:N R C:N A ). Th e rates o f element flux measured by metabolism will be stable and c orrespond to the tissue ratios of the dominant taxa , as demand and supply achieve a balance , expressed by homeostasis (C:N E = %cover taxa 1(C:N A ) + % cover taxa 2(C:N A )). The go al of this work is to test the hypothesis that the stoichiometry of plant tissues controls the stoich iometry of ecosystem metabolism. Methods To test this hypothesis, I quantified C and N metabolism in 7 spring fed rivers of north central Florida (Table 4 1), spanning a gradient in expected C:N E ratio in response to variation in site level relative abundance of taxa with different tissue C:N ratios. The spring fed rivers in North Florida emerge from the Floridan Aquifer and offer unique opportunities to stu dy e cosystem metabolism because they display relatively constant discharge, chemistry, and temperature at the vent over time, which makes downstream change s solely attributable to the bio ta (Odum 1957). E lemental concentrations and

PAGE 68

68 ratios differ among spri ngs, but within spring discharge water chemistry is stable, and nitrogen and phosphorus in all systems are domi nated by mineral forms (Cohen et al. 2013 ). The 7 spring fed rivers in north cent ral Florida (Table 4 1) that I chose to study were selected to s pan a gradient in benthic autotroph composition , ranging from algal to vascular plant dominated. The dominant algae in the selected systems are the benthic filamentous taxa Lyngbya wollei ((Farlow ex Gomont) Speziale and Dyck, hereafter Lyngbya ) and Vauche ria disperma (de Candolle, hereafter Vaucheria ), while the dominant vascular plants are Vallisneria americana ( Michaux 1803, hereafter Vallisneria ) and Sagittaria kurziana (Glück 1827, hereafter Sagittaria ). Sensor D eployments I deployed two in situ sensor s in the advective zone o f each spring fed river during 7 multi day deployments , be tw een February 2010 and March 2015 (Table 4 1) . Deployments lasted for an average of 8 days ( range = 7 11 days), t hough estimates of metabolism for the first and last day of deployments were omitted because of incomplete 24 hr measurements. Sensors were located as far from the boil as possible to maximize signal generation, up to a maximum travel time from the spring vent of 6 hours to avoid dispersion effects that mix diel signals from successive days (Hen sley 2014 ). F or the 2010 2012 deployments, I obtained filtered wate r samples from each spring vent , which were analyzed for dissolved inorganic carbon (DIC) and nitrate ( NO 3 ) using second derivative UV spectra obtained fr om an Aquamate UV Vis spectrophotometer prior to deployment (Simal et al. 1985). For the 2014 and 2015 data, I first measured DO and NO 3 at the spring vent using the sensors, and then commenced downstream deployment. Because of low GPP mea surements during rain

PAGE 69

69 events, I only utilized data from days with little to no rain and sufficient solar radiation to reflect typical autotrop hic production. During the late April early May 2012 deployment at Otter Springs, usable data were not obtained for the last 3 day s of the deployment due to tannic overflow from the Suwan n ee River which interfered with N measurements . During the 2010 2014 deployments, I measured temperature (T), specific conductance (SpC) and dissolved oxygen (DO) hourly , using a multi parameter sond e (6920 V2, YSI, Yellow Springs, OH). For the 2015 deployments , temperature (T) and dissolved oxygen (DO) wer e measured every 15 minutes using a n optical dissolved oxygen sensor (U26 001, Onset Com puter Corporation, Bourne, MA). For all deployments, NO 3 w as measured every 15 minutes with a submersible UV nitrate analyzer (SUNA, Satlantic, Halifax Nova Scotia) , operated in polled mode wherein 10 measurements were obtained and averaged for each sampling interval . The SUNA was wrapped in 100 screening to prevent biofouling during deployment. Sensors were cleaned and calibrated in the laboratory prior to the start of each deployment. Metabolism Fo r each deployment, gross primary production (GPP; g O 2 m 2 d 1 ) and ecosystem respiration (ER; g O 2 m 2 d 1 ) was calculated using the single station method (Odum 1957; Cohen et al. 2013). DO ranged from 3.99 to 11.02 mg L 1 in Ichetucknee River, 4.21 to 11.21 mg L 1 in Otter Spring, 2.74 to 8.66 mg L 1 in Silver River, 4.49 to 8.89 mg L 1 in Gi lchrist Blue Spring (2012) . There was low variation in DO in Juniper Spring (7.3 to 8.88 mg L 1 ). The re aeration rate at maximum deficit, k max (in units of g O 2 m 2 hr 1 ) was calculated as a functi on of mean velocity (u; cm s 1 ) using the floating dome met 2006) .

PAGE 70

70 k m ax =0.0604u+ 0.0929 (4 1) This value is then multiplied by the saturation deficit (1 C/Csat) to calculate the re aeration rate. To calculate net primary production (NPP; mol C m 2 d 1 ) from GPP, a photosynthetic coefficient of 1.0 (Odum 1957) and autotrophic respiration equ al to 50% of GPP ( Hall and Beaulieu 2 01 3) was assumed and mol O 2 m 2 d 1 was converted to mol C m 2 d 1 using atomic mass. I calculated nitrogen r emoval attributed to assimilation (Ua; mg N m 2 d 1 ) using the extrapolated diel method (Heffernan and Cohen 2010). In brief, sensor measurements are used to construct hourly mass balance estimates of N removal wherein assimilatory N demand is calculated u sing the integrated difference between a NO 3 baseline and observed diel variation. In this approach, daily NO 3 peaks are first determined and daily autotrophic assimilation is calculated as the discharge area weighted summation of the difference between the previous NO 3 peak and hourly NO 3 concentration over a 24 hr period . As input concentrat ion of ammonium and dissolved organic N were below detection limits, these uptakes were not calculated. C: N E ratios were calculated for each day and averaged over the deployment to obtain a single C:N ratio for each deployment. Autotrophic Stoichiometry and C omposition Each run was div ided into ten transects of approximat ely equal length and vegetation was sampled at 5 random locations along each transe ct to estimat e percent cover using a 0.5 m 2 weighted PVC quadrat. Percent cover by guild (filamentous algae, rooted vascular plants) was calculated by assessing the relative percentages of algal and vascular taxa at each transect and along the run. I sampled each autotr ophic species present at each transect. Samples were stored on ice, and within 48 hours,

PAGE 71

71 triple washed with deionized water, dried at 65°C and ground. Perc ent C and N was measured with a Carlo Erba NA1500 CNS elemental analyzer . Molar tissue C:N ratios wer e calculated f rom percent C and N by mass . I calculated the expected ecosystem scale C:N ratio from the cover weighted average tissue C:N (C:N A ) value measured for each deployment. Statistical Analysis I used linear regressi on t o test the hypothesis that the C:N ratio of ecosystem metabolism (C:N E ) reflects the C:N value of the autotrophs (C:N A ) . I evaluated the estimates of N uptake as a function of algal cover and NO 3 concentrations. Prior work (Cohen et al. 2 013 ) on ecosystem C:N ratios suggested that ecosystem values were generally larger than th e tissue stoichiometry values. As such, I quantified any difference between ecosystem metabolism stoichiometry and tissue stoichiometry by subtracting the calculated deployment tissue C:N ratio from the ecosystem C:N ratio . Results D iel changes in dissolved oxygen and nitrate were observed in all systems (Figure 4 1) . Sensors were deployed at locations that ranged from 485 m to 5000 m downstream (mean = 2500 m); deployment lengths were sufficient to enabl e the emergence of diel signals . The diel signals were relatively uniform for b oth DO and N, although the magnitude of diel variation was quite variable . Across all deployments, the timing of the diel maxima for solar radiation did not vary , peaking near n oon each day , whereas the timing of maxima for DO varied slightly; shorter runs such as Gilchrist Blue exhibited earlier peaks than did longer runs (Fig ure 4 1) . The patterns of diel minima for NO 3 were relatively consistent across the majority of deploym ents; however, Gilchrist

PAGE 72

72 Blue and Otter NO 3 minima occurred out of phase with the rest of the deployments (Fig ure 4 1). Variability of N uptake was assessed within and across deployments (Table 4 2) . N uptake across deployments average d 4.60 ± 2.86 m mol N m 2 d 1 , and ranged from 11 m mol N m 2 d 1 at Rainbow to 0. 07 m mol N m 2 d 1 at Otter Springs. As expected , the variability of N uptake across all springs was high (coefficient of variation CV = SD/ ~74 %). However, within each spring , average N uptake wa s much more tightly constrained (CV ~ 13%) . No relationship was found between NO 3 concentrations and N uptake rates ; however, N uptake decreased dramat ically as percent algal cover in creased (R 2 = 0.54 , p = 0.18 ) . Ecosystem C:N varied across deployments ( 20.64 ± 2.86 ) ; yet, c onsidering the diversity of autotrophic assemblages, the degree of variance in ecosystem C:N ratios across all deployments was low (CV ~14 %). D uring each deployment , the variance was correspondingly low. Gilchrist Blue, also the shortes t run, exhibited the lowest variance, varying Silver and Otter each showed about 25% variation in eco system C:N over the deployment. There is strong correlation between autotrophic tissue stoichiometry and the stoichiometry of ecosystem metab olism (Fig ure 4 2). This supports the hypothesis that tissue stoichiometry controls ecosystem metabolism stoichiometry with higher tissue C:N ratios corresponding to higher ecosystem metabolism C:N ratios (R 2 = 0.63 ). The algal dominated system, Ott er Sprin gs, had the lowest ecosystem C:N ratio; the highest C:N E ratio was observed in Gilchrist Blue, a vascular plant dominated system. The

PAGE 73

73 highest tissue C:N ratios were observed in Ichetucknee spring in 2010 , a vascular plant dominated system. D ifferences betw een the C:N A and C:N E E C:N A ) were clearly evident ( Fig ure 4 4 ) , though not uniform across systems, ranging from 0.11 in Otter Springs to 5.65 in Gilchrist Bl ue . The coefficient of variation for C:N across systems was 56% , indicating a relati vely high degree of variability in among systems. Discussion There was remarkable correlation between C:N E and C:N A (Fig ure 4 2). This strong correlation between flux and stock ratios (Fig ure 4 2) suggest s that the use of elements in ecosystems is sha ped by organisms, in this case the autotrophs, therein. Moreover, here C:N E acts as a linear function of C:N A . This enables predictions of coupling across scales as long as the stoichiometry and composition of the organisms at the organismal scale are know n. However, the slope of this relationship deviates fro m a simple 1:1 line, suggesting a systematic mismatch, as evidenced by the estimation of increase d as the relative abundance of vascular plants became dominant (R 2 = 0.96, p = 0.02 , Fig ure 4 calculated the daily N deficit as well as the residence time of N in vascular plants . The daily N deficit ( m mol N m 2 d 1 ) was highly variable and is characterized by the daily N flux necessary to m eet tissue C:N, averaging (0.97 ± 2.76 m mol N m 2 d 1 ) across deployments. The N deficit was highest at Silver springs and lowest at Gilchrist Blue spring. The residence time of N in vascular plants was highest at Juniper springs and lowest at Rainbow (Table 4 2). The average residence time of N in vascu lar plants was estimated to be 358 ± 207 d across deployments.

PAGE 74

74 This difference between tissue and ecosystem metabolism C:N ratio was thought to aris e from four plausible factors . First , may be a consequence of measurement error caused by hydraul ic mixing of the diel signals. Second , the assumption that denitrificat ion is constant may be invalid . Third, a biologically mediated response of internal plant recycling of N may be occurring . Lastly , alternative pathways of resource supply may be supplementing biota. Hydr aulic mixing occurs in response to solute dispersion , which leads to parcels of water passing the sensor that have been acted upon over a wide range of times . This exerts a differential effect on oxygen (which interacts with the atmosphere) and nitrate (which does not), creating potential errors when ecosyst em C:N ratios are calculated using the single station approach employed here . Despite the potential of this effect to alter the inferred C:N E C:N values observed here, the effect is most pronounced in longer residence time rivers. Hensley (2014) estimates that the maximum effect occurs when mean residence times to arrive at the sensor are longer than 12 hours downstream of the spring vent. I intentionally deployed the sensors to ensure shorter residence times ( 8 hours ) , lo wering the probability that this hydraulic effect introduces measurement error. Furthermore, the geometry of the raw sensor signals , particularly the peaks of the dissolved oxygen signals, is largely uniform and indicate s a lack of signal dispersion caused by hydraulic factors (Fig ure 4 1) . The potential for diurnal variation in denitrification was explored by evaluating the productive areas (GPP ~10 to 16g O 2 m 2 d 1 ) and provide valuable habitat for invertebrates, fish and other organisms. As these p lants tend to senesce in situ , beds

PAGE 75

75 can be a vital source of autochthonous organic carbon. Additionally, these beds act to slow water velocities, lengthening residence times and allowing for larger nutrient removal through autotrophic assimilation or denit rification. R esorption is an important nutrient conservation strategy among terrestrial autotrophs (Vitousek 1982). To investigate nutrient resorption by vascular plants, I compared tissue stoichiometry across young, mature, senescent, and root components , using the methods described above . I observed n o evidence of intern al N recycling by vascular taxa. Indeed, although there were no large differences between leaf ages, the lowest C:N values were observed in the senescent and root tissues for both Va llisneria and Sagittaria (Figure 4 6 ). If internal N recycling was occurring , C:N ratios wo uld increase as the tissue progressed from maturity to senescence. Extracellular polymeric substance production could be another rich exudates would be captured by metabolism measurements but not measured in tissue stoichiometry (Elser et al. 1996), raising C:N E . Finall y, I evaluated the po ssibility ses because rooted plants obtain nutrients in alternative forms, such as ammonium (NH 4 + ) from the bottom substrate, rather than by direct assimilation from the water column. In these systems, NH 4 + concentrations in the water column are low; however, both dominant species of vascular plants possess shallow root systems allowing for the possibility of ammonium assimilation from the sediment (Barko et al. 1991, Barko and Smart 1981) . strongly associated with increasing v ascular cover, whereas algal sit es exhibited negligible differences , possibly indicating N assimil ation from the sediment by vascular

PAGE 76

76 plants . I quantified this uptake pathway for each deployment using the following equation , which assumes that the difference between C:N E and C:N A is all attributable to alternative nutrient sources : (4 2) This indicates that N assimilation from the sediment fulfilled between 1 2 % ( Ichetucknee ) and 49 % (Gilchrist Blue ) of the vascular plant N demand . These results suggest that the stoichi ometry of ecosystem metabolism and tissue stoichiometry are closely linked. Metabolism acts to integrate the interactions of multiple inputs, processes, and organisms, and c ouple elemental cycles. Whereas e le ment cycles can be both dir ectly and indirectly coupled, these results suggest that the degree of direct coupling among cycles is strongly controlled by the elemental demand required by organisms.

PAGE 77

77 Table 4 1. Ecosystem metabolism study s ites Site Name Dep loyment date Q (m 3 s 1 ) Deployment length (m) Mean PAR (W m 2 d 1 ) Max/Min Air Temp. (C) % algal cover Otter Apr May 2012 0.009 1320 292 33.1/13.6 100% Ichetucknee Apr 2012 5.7 5000 2 41 31.5/8.2 25% Ichetucknee Jan 2010 7.1 5000 124 26.1/ 8.3 15% Jun iper Nov 2010 1.3 1700 153 29.2/4.1 10% Gilchrist Blue Apr 2012 1. 1 485 209 29.4/4.5 8.56% Silver Jul 2012 14. 7 3860 201 33.7/18.5 4% Rainbow Jan Feb 2015 1 9.0 4300 143 23.5/1.7 0%

PAGE 78

78 Table 4 2. Summary of metabolism and nutrient assimilation by dep loyment including mean ± standard deviation of gross primary production (GPP), ecosystem respiration (R), autotrophic assimilation of N (U a, N), e cosystem metabolism C:N ( C:N E ) , N deficit, N residence time in vascular taxa , as well as percent cover weighted tissue C:N (C:N A ) and the r atio of C:N A /C:N E . Site GPP (g O 2 m 2 d 1 ) R (g O 2 m 2 d 1 ) N uptake (mmol N m 2 d 1 ) C:N E C:N A N deficit ( m mol N m 2 d 1 ) N residence time (d) C:N A /C:N E Otter 0.68±0.03 0.31±0.03 0.70±0.18 16.06±4.17 16.17 0.0 4 ±0. 2 1.00 Ic hetucknee 2012 8.24 ±1. 14 15.02 ±0. 30 7.61 ± 2.07 17.02±1.14 14.94 2±0. 4 287.19±47.80 0.8 8 Ichetucknee 2010 5.35±1.29 7.63±0.45 3.96 ± 0.75 22.95±5.04 17.93 0.7±0. 7 298.93±86.14 0.78 Juniper 1.94±0.30 3.30±0.89 1.36 ±1. 08 22.26±2.65 1 7.43 0. 6 ±0. 2 775.78±44.71 0 .7 8 Gilchrist Blue 9.54±0.43 3.23±0.19 4.14 ± 0.78 33.09±0.64 16. 98 4±0. 7 296.69±65.53 0. 51 Silver 10.20±3.20 25.80±0.26 7.37 ±2 .10 22.29±5.38 17.47 2 . 5 ±1. 3 273.56±125.13 0.7 8 Rainbow 8.77 ± 0.29 10.05 ±0. 48 7.06 ± 2.75 21.26 ± 7.58 1 6.0 1 1.3 ± 2.3 217.98 ± 71.73 0. 75

PAGE 79

79 Fig ure 4 1. Diel signals of all deployments.

PAGE 80

80 Figure 4 2 . Relationship between ecosystem metabolism C:N (C:N E ) and autotrophic tissue C:N (C:N A ). y = 2.17x 15.53 R² = 0.63 p = 0.03 15 20 25 14 15 16 17 18 19 Ecosystem C:N Tissue C:N

PAGE 81

81 Figure 4 3 Relationship between (A) percent of algal cover and N uptake ( m mol N m 2 d 1 ) as we ll as between (B) NO 3 concentrations and N uptake ( m mol N m 2 d 1 ) y = 0.0055e 0.02x R² = 0.54 p = 0.18 0 1 2 3 4 5 6 7 8 9 0 20 40 60 80 100 120 N uptake (mmol N m 2 d 1 ) Percent Algal Cover y = 0.0017x + 0.0028 R² = 0.10 p = 0.49 0 1 2 3 4 5 6 7 8 9 0 0.5 1 1.5 2 NO 3 concentration (mg L 1 ) a b

PAGE 82

82 Figure 4 4 . The differenc e ( C:N) between ecosystem C:N (C:N E ) and aggregated tissue C:N (C:N A ) ratios shown with 95% confidence intervals . -4 0 4 8 12 Otter Ichetucknee 2012 Silver Juniper Ichetucknee 2010 Gilchrist Blue Rainbow C:N

PAGE 83

83 Figure 4 5 . Relationship between C:N and vascular plant cover. y = 0.1125e 0.04x R² = 0.96 p = 0.02 0 1 2 3 4 5 6 7 0 20 40 60 80 100 Percent Vascular Cover

PAGE 84

84 Figure 4 6. Tissue C:N by vascular plant part type. Different lowercase letters indicate significant differences between plant parts at p < 0.05 0 5 10 15 20 Young Mature Senescent Roots Molar tissue C:N a a b b

PAGE 85

85 CHAPTER 5 SYNTHESIS AND CONCLUSION This dissertation and the studies therein investigate d the relationship between resource (nutrient) supply and au totroph demand in Florida spring ecosystems using ecological stoichiometry , and assess ed the impact of this relationship on both ecosystem structure and function. Using spring ecosystems as model river ecosystems, thi s body of work has improved the understanding of when and why autotroph s are homeo static, identified factors that contribute to autotrophic species shifts, and revealed linkages between ecosystem metabolism and the representative autotrophs. Absence of Resource Limitation and Homeostasis In Chapter 2, t o explore how nutrient availability influence s the stoichiome try of freshwater autotrophs, I measured C:N:P ratios in algal and submerged va scular plant tissues and water from 41 north Florida spring fed rivers, which are ideal natural labo ratories because of their high temporal stability with respect to che mistry, temperature, and flow. Water chemistry across springs spanned a molar inorganic N:P gradient from 0.6 to 89.4, whereas repeated meas urements within springs indicated low variation ( CV < 20% ) over inter annual timescale s. I observed significant interspecific differences in stoichiometry , suggesting different nutrient requirements among taxa , but negligible evidence that variation in nutrient supply influ enced taxa presence or absence. Most importantly, I observed no plas ticity in t issue stoichiometry for any taxon ; homeostatic regulation coefficients (H) ranged from 4.8 to undefined (slope of log resource vs. log 0) suggesting stric t homeostasis in this setting. To explore the paradox of plasticity in chem ostat experiments vs. homeostasis in these natural chemostats, I adapted the Droop model to be tter represent benthic systems, i.e.

PAGE 86

86 where hydraulic and biomass turnover rates can be different, and to allow light limitation. Model results suggest increasing hydraulic turnover induces homeostasis and releases autotrophs from grow th limitation by nutrients, thus inducing light limitation, over a large range of nutrient concentrations and el ement ratios. This suggests that, relative to ecosystem demand, it is th e resource supply rate, not nutrient concentration s or ratios, that is the primary determinant of nutrient limitation, and thus a dominant contr ol on autotroph stoichiometry. R esults also suggest that differences between observed and taxon specific optimal stoichiometry m ay be diagnostic of nutrient limitation. Ecosystem Structure Implications Revealed by Experimental Mesocosms Many Florida s prings have experienced degradation of ecosystem structure, in the form of shifts in autotrophic species assemblages , from submerged vascular plant s to benthic filamentous algae. Increasing NO 3 N concentrations have been cited as a primary driver of this autotrophic species composition shift , with algal proliferation hypothesized to occur after alleviation of NO 3 N lim itation. Numeric nutrient criteria have been developed to address rising nitrate concentrations and restore native vascular plant dominance. However, the understanding of factors that influence the outcome of competition between algal and vascular species in these systems remains poor . Recent evidence suggests the autotrophs are not nutrient limited; part of the reason may be that nutrient resupply from advection satiates ecosystem demand, making other bottom up factors such as flow and top down effects suc h as grazing important factors in determining the outcomes of competition between native va scular plants and nuisance algae . In Chapter 3, I used in situ multi factorial experimental mesocosms to assess t he extent to which the physical factor of flow and t op down

PAGE 87

87 factor of grazing pressure influence s algal and vascular standing stocks and net production across a wide range of nitrate (NO 3 N) concentrations . This study suggested variation in resource concentration or ratios has minimal effects on the outcom e of autotrophic competition in Florida spring systems. Instead, absence of grazers was found to diminish the success of vascular taxa, and to more than triple the relative standing stock of algal taxa. Additionally, flow was found to influence relative al gal standing stocks , with both low and high flow environments resulting in smaller standing stock s. These findings indicate that factors other than NO 3 N need to be considered for managing ecosystem structure in these flowing water systems. L inking Ecosys tem Metabolism and Autotroph Stoichiometry Whereas ecological stoichiometry is a framework based on elemental ratios for understanding how organ isms interact within ecosystems; metabolism is an integrative metric of ecosystem function and energetics, synth esizing the relative contributions of multiple inputs, p rocesses, and interactions. Relating the two may potentially inform ecosystem scal e use of elements and energy. In Chapter 4, I sought to quantify the link between the stoichiometry of ecosystem metab olism, specifically the C:N ratios of integrated autotrophic assimilation and the stoichiometric tissue ratios observ ed in the dominant autotrophs. Using high frequency in situ nutrient sensors I estimated the assimilatory fluxes of C and N in spring fed F lorida rivers of varying autotrophic species composition. I measured autotroph cover in each spring river, collected vegetation samples, and e valuated t issue stoichiometry. St rong correlation was found between measured tissue stoichiometry and elemental ra tios at the ecosystem scale, suggesting that aggregated assimilatory fluxes may be useful for linking organismal nutrient content to the stoichio metry of ecosystem metabolism. When differences between tissue an d

PAGE 88

88 ecosystem metabolism C:N did occur, results suggest ed this difference was not the result of hydraulic mixing , nor internal plant recycling. Instead, high densities of vascular plants wer e observe d to be associated with increase . Potential N assimilation from sediments and denitrification were thought to be likely sources of this difference and together contribute to the ~ 30% difference in C:N. Overall Implications Aquatic e cosystems throughout the world face a multitude of synergistic drivers and stressors on an unprecedented scale. To predict and mitigate the cumulative impacts of these stressors, new theoretical frameworks are needed to better understand patterns of ecological response . I utilized the approach of ecological stoichiometry to explore the relationship between resource supply and biotic demand. Often supply and demand are not in balance; nevertheless, I demonstrate d that autotrophs can and do display homeostasis across large gradients of supply and that their tissue stoichiometry can be diagnostic of nutrient limitation. Moreover, in the absence of nutrient limitation, these result s suggest alternative factors , i.e. top down troph ic control , become dominant influences on ecosystem composition and species competition. Furthermore, I show ed that stoichiometry can be used to scale coup led net el ement utilization from the organism to the ecosystem. In summary, these results can be applied to inform predictions of ecological response s to the balance between resource supply and biotic demand.

PAGE 89

89 LIST OF REFERENCES Ã…gren, G. I. 2008. Stoichio metry and nutrition of plant growth in natural communities. Annu. Rev. Ecol. Evol. Syst. 39: 153 170, doi:10.1146/annurev.ecolsys.39.110707.173515 Ashworth, J., S. Coesel, A. Lee, E. V. Armbrust, M. V. Orellana, and N. S. Baliga. 2013. Genome wide diel gro wth state transitions in the diatom Thalassiosira pseudonana. Proc. Nat. Acad. Sci. 110: 7518 7523, doi:10.1073/pnas.1300962110 Barko, J. W., and R. M. Smart. 1981. Sediment based nutrition o f submersed macrophytes. Aquat. Bot . 10: 339 352. Barko, J. W., D . Gunnison, and S. R. Carpenter. 1991. Sediment interactions with submersed macrophyte growth and community dynamics. Aquat. Bot . 41: 41 65, doi:10.1016/0304 3770(91)90038 7 Bliese, P. 2013. Multilevel: Multilevel Functions. R package version 2.5. Brown, M . T., K. Chinners 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 on the effects o f nutrients on spring organisms and systems. Final Report. Florida Department of Environmental Protection. Tallahassee, Florida, USA. Carpenter, E. J., and R. R. L. Guillard. 1971. Intraspecific differences in nitrate half saturation constants for three sp ecies of marine phytoplankton. Ecology 52: 183 185, doi:10.2307/1934753 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. Ecol. Appl. 8: 5 59 568, doi:10.1890/1051 0761(1998)008[0559:NPOSWW]2.0.CO;2 Cherif, equations to model growth under multiple nutrient limitation. Oikos. 119: 897 907, doi:10.1111/j.1600 0706.2010.18397.x Chrzanowski, T. H., and J. P. Grover. 2008. Element conte nt of Pseudomonas fluorescens varies with growth rate and temperature: A replicated chemostat study addressing ecological stoichiometry. Limnol. Oceanogr. 53: 1242 1251, doi:10.4319/lo.2008.53.4.1242 Cohen, M. J., M. J. Kurz, J. B. Heffernan, J. B. Martin, R. L. Douglass, C. R. Foster, and R. G. Thomas. 2013. Diel phosphorus variation and the stoichiometry of

PAGE 90

90 ecosystem metabolism in a large spring fed river. Ecol. Monogr. 83: 155 176, doi:10.1890/12 1497.1 Collos, Y., A. Vaquer, and P. Souchu. 2005. Acclima tion of nitrate uptake by phytoplankton to high substrate levels. J. Phycol. 41: 466 478, doi:10.1111/j.1529 8817.2005.00067.x Conley, D. J., H. W. Paerl, R. W. Howarth, D. F. Boesch, S. P. Seitzinger, K. E. Havens, C. Lancelot, and G. E. Likens. 2009. Con trolling eutrophication: nitrogen and phosphorus. Science 323: 1014 1015. Co peland, B. J. and W. R. Duffer. 1964. Use of a clear plas tic dome to measure gaseous diffusion rates in natural waters. Limnol. Oceanogr. 9:494 499. Covino, T. P, B. L. McGlynn, an d R. A.McNamara. 2010. Tracer Additions for Spiraling Curve Characterization (TASCC): Quantifying stream nutrient uptake kinetics from ambi ent to saturation. Limnol. Oceanogr. Meth. 8: 484 498. Crowl, T. A., T. O. Crist, R. R. Parmenter, G. Belovsky, and A . E. Lugo. 2008. The spread of invasive species and infectious disease as drivers of ecosystem change. Front . Ecol . Environ. 6:238 246, doi:10.1890/070151 Demars, B. O. L., and A. C. Edwards. 2007. Tissue nutrient concentrations in freshwater aquatic macro phytes: high inter taxon differences and low phenotypic response to nutrient supply. Freshw. Bio l . 52: 2073 2086, doi:10.1111/j.1365 2427.2007.01817.x Droop, M. R. 1973. Some thoughts on nutrient limitation in algae. J. Phycol. 9: 264 272, doi:10.1111/j.15 29 8817.1973.tb04092.x Duarte, C. M. 1995. Submerged aquatic vegetation in relation to different nutrient regimes. Ophelia 41:87 112. Duarte, C. M., D. J. Conley, J. Carstensen, and M. Sanchez Camacho. 2009. Return to Neverland: Shifting baselines affect e utrophicatio n restoration targets. Estuar. Coast. 32: 29 36. Earl, S. R., H. M. Valett, and J. R. Webster. 2006. Nitrogen saturatio n in stream ecosystems. Ecology 87: 3140 3151. Elser, J. J., D. R. Dobberfuhl, N. A. MacKay, and J. H. Schampel. 1996. Organi sm size, life history, and N:P stoichiom etry. Bioscience 46: 674 684. Elser, J. J., W. F. Fagan, R. F. Denno, D. R. Dobberfuhl, A. Folarin, A. Huberty, S. Interlandi, S. S. Kilham, E. McCauley, K. L. Schulz, E. H. Siemann, and R. W. Sterner. 2000. Nutritio nal constraints in terrestrial and freshwater foodwebs. Nature 408: 578 580, doi:10.1038/35046058

PAGE 91

91 Elser, J. J., W. F. Fagan, A. J. Kerkhoff, N. G. Swenson, and B. J. Enquist. 2010. Biological stoichiometry of plant production: Metabolism, scaling and ecolo gical response to global change. New . Phytol. 186: 593 608, doi:10.1111/j.1469 8137.2010.03214.x Enriquez, S., C. M. Duarte, and K. Sand Jensen. 1993. Patterns in decomposition rates among photosynthetic organisms: The importance of de tritus C:N:P content. Oecologia 94: 457 471, doi:10.1007/BF00566960 FDEP [Florida Department of Environmental Protection]. 2013. Implementation of Protection, Tallahassee, Florida, USA. FSTF [Florida Spr restoration. Florida Department of Environmental Protection, Tallahassee, Florida, USA. Gaiser, E. E., D. L. Childers, R. D. Jones, J. H. Richards, L. J. Scinto, and J. C. Trexler. 20 06. Periphyton responses to eutrophication in the Florida Everglades: Cross system patterns of structural and comp ositional change. Limnol. Oceanogr. 51: 617 630. Grimm, N. B., and S. G. Fisher. 1986. Nitrogen limitation in a Sonoran desert stream. J. N. A m. Benthol. Soc. 5: 1 15. Guillette, Jr., L. J. and T. M. Edwards. 2005. Is nitrate an ecologically relevant endocrine disru ptor in vertebrates? Integr. Comp. Biol. 45: 19 27. Hall, S. R., V. H. Smith, D. A. Lytle, and M. A. Leibold. 2005. Constraints on p rimary producer N:P stoichiometry along N:P supply ratio gradients. Ecology 86: 1894 1904, doi:10.1890/04 1045 Hall, R. O., and J. J. Beaulieu. 2013. Estimating autotrophic respiration in streams using daily met abolism data. Freshw. Sci . 32: 507 516, doi:1 0.1899/12 147.1 Harpole, W. S., J. T. Ngai, E. E. Cleland, E. W. Seabloom, E. T. Borer, M. E. S. Bracken, J. J. Elser, D. S. Gruner, H. Hillebrand, J. B. Shurin, and J. M. Smith. 2011. Nutrient co limitation of primary producer communities. Ecol. Lett. 14: 852 862, doi:10.1111/j.1461 0248.2011.01651.x Hauxwell, J., T. K. Frazer, and C. W. Osenberg. 2004. An annual cycle of biomass and productivity of Vallisneria americana in a subtropical spring fed estuary. Aquat. Bot. 87: 61 68, doi:10.1016/j.aquabot.2007 .03.003 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. Limnol. Oceanogr. 33: 796 822, doi:10.4319/lo.1988.33.4_part_2.0796

PAGE 92

92 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. Limnol. Oceanogr. 55: 677 688, doi:10.4319/lo.2009.55.2.0677 Heffernan, J. B., D. M. Liebowitz, T. K. Frazer, J. M . Evans, and M. J. Cohen. 2010. Algal blooms and the nitrogen enrichment hypothesis in Florida springs: evidence, alternatives, and adaptive management. Ecol. Appl. 20: 816 829. Hensley, R. T., and M. J. Cohen. 2012. Controls on solute transport in large s pring fed karst rivers. Limnol. Oceanogr. 57: 912 924, doi:10.4319/lo.2012.57.4.0912 Hensley, R. T. 2014. Hydraulic controls on spatial and temporal variation in nitrogen spiraling in large rivers. University of Florida, Gainesville, FL. are, M. J. Bowes, and J. I. Jones. 2006. How green is my river? A new paradigm of eutrophication in rivers. Sci. Total. Environ. 365: 66 83, doi:10.1016/j.scitotenv.2006.02.055 Howarth, R. W. 1988. Nutrient limitation of net primary production in marine ec osystems. Annu. Rev. Ecol. Syst. 19: 89 110, doi:10.1146/annurev.es.19.110188.000513 King, S. A. 2014. Hydrodynamic control of filamentous macroalgae in a sub tropical spring fed river in Florida, USA. Hydrobiologia 734: 27 37. King, S. A., J. B. Heffernan , and M. J. Cohen. 2014. Nutrient flux, demand, and autotrophic limitation in lotic systems. Freshw. Sci. 33: 85 98, doi:10.1086/674383 Klausmeier, C. A., E. Litchman, and S. A. Levin. 2004. Phytoplankton growth and stoichiometry under multiple nutrient li mitation. Limnol. Oceanogr. 49: 1463 1470, doi:10.4319/lo.2004.49.4_part_2.1463 Klausmeier, C. A., E. Litchman, and S. A. Levin. 2007. A model of flexible uptake of two essential resources. J. Theor. Biol. 246: 278 289, doi:10.1016/j.jtbi.2006.12.032 Koers elman, W., and A. F. M. Meuleman. 1996. The vegetation N:P ratio: A new tool to detect the nature of nutrient limitation. J. Appl. Ecol. 33: 1441 1450, doi:10.2307/2404783 Larned, S. T., V. I. Nikora, and B. J. F. Biggs. 2004. Mass transfer limited nitrog en and phosphorus uptake by stream periphyton: a conceptual model and experimental evidence. Limnol. Oceanogr. 49: 1992 2000, doi:10.4319/lo.2004.49.6.1992 Lenth, R. V., and M. Herva. 2014. lsmeans: Least squares means. R package version 2.13.

PAGE 93

93 Liebig, J. 1 855. Principles of agricultural chemistry with special reference to the late researches made in England. John Wiley. Liebowitz, D. M., M. J. Cohen, J. B. Heffernan, L. V. Korhnak, and T. K. Frazer. 2014. Environmentally mediated consumer control of algal p roliferation in Flo rida springs. Freshw. Biol. 59: 2009 2023. Loladze, I., Y. Kuang, and J. J. Elser. 2000. Stoichiometry in producer grazer systems: linking energy flow with element cycling. B. Math. Biol. 62: 1137 1162, doi:10.1006/bulm.2000.0201 Lorenz, A. W., T. Korte, A. Sundermann, K. Januschke, and P. Haase. 2012. Macrophytes respond to reach scale river r estorations. J. Appl. Ecol. 49: 202 212. Madsen, J. D., P. A. Chambers, W. F. James, E. W. Koch, and D. F. Westlake. 2001. The interaction between water movement, sediment dynamics and submersed macrophytes. Hydrobiologia 444: 71 84. Mulholland, P. J., C. S. Fellows, J. L. Tank, N. B. Grimm, J. R. Webster, S. K. Hamilton, E. Marti, L. Ashkenas, W. B. Bowden, W. K. Dodds, W. H. McDowell, M. J. Paul, a nd B. J. Peterson. 2001. Inter biome comparison of factors controlling stream metabolism. Freshw. Biol. 46: 1503 1517. Munch, D.A., D.J. Toth, C. Huang, J.B. Davis, C.M. Fortich, W.L . Osburn, E.J. Phlips, E.L. Quinlan, M.S. A llen, M.J. Woods, P. Cooney, R. L. Knight, R .A. Clarke and S.L. Knight. 2006. Fifty year retrospective study of the ecology of Silver Spring s, Florida. St. Johns River Water Management District Publication SJ2007 SP4, Palatka, FL. Nelson, G. C., E. Bennett, A. A. Berhe, K. Cassman, R. De Fries, T. Dietz, A. Dobermann, A. Dobson, A. Janetos, M. Levy, D. Marco, N. Nakicenovic, B. Held, D. Ojima, P. Pingali, R. Watson, and M. Zurek. 2006. Anthropogenic drivers of ecosystem change: an overview. Ecol. Soc . 11: 29 60. nutrient spiraling in streams. Can. J. Fish. Aquat. Sci. 38: 860 863, doi:10.1139/f81 114 Nifong, R. L., M. J. Cohen, and W. P. Cropper, Jr. 2014. Homeostasis and nutrie nt limitation of benthic autotrophs in natural chemos tats. Limnol. Oceanogr. 59: 2101 2111. Nishihara, G. N., and J. D. Ackerman. 2009. Diffusive boundary layers do not limit the photosynthesis of the aquatic macrophyte Vallisneria americana at moderate fl ows and saturating light levels. Limnol. Oceanogr. 54: 1874 1882, doi:10.4319/lo.2009.54.6.1874

PAGE 94

94 saturation of N cycling in Central Plains streams: 15N experiments across a broad gradient of nitrate concentrations. Biogeochemistry 84: 31 49, doi:10.1007/s10533 007 9073 7 Odum, H. T. 1956. Primary product ion in flowing waters. Limnol. Oceanogr. 1: 102 117, doi:10.4319/lo.1956.1.2.0102 Odum, H. T. 1957. Trophic structure and productivity of Silver Springs, Florida. Ecol. Monogr. 27: 55 112, doi:10.2307/1948571 Parmesan, C., and G. Yohe. 2003. A globally coherent fingerprint of climate change impacts across natural syst ems. Nature 421:37 42, doi:10.1038/nature01286 Pedersen, M . L., N. Friberg, J. Skriver, A. Baatrup Pedersen, and S. E. Larsen. 2007. Restoration of Skjern River and its valley Short term effects on river habitats, macrophytes and macroinvertebrates. Ecol . Eng. 30: 145 156. Pellerin, B.A., J.F. Saraceno, J.B. Sha nley, S.D. Sebestyen, G.R. Aiken, W.M. Wollheim, and B.A. Bergamaschi. 2012. Taking the pulse of snowmelt: in situ sensors reveal seasonal, event and diurnal patterns of nitrate and dissolved organic matter variability in an upland forest stream. Biogeoche mistry 108:183 198. Persson, J., P. Fink, A. Goto, J. M. Hood, J. Jonas, and S. Kato. To be or not to be what you eat: Regulation of stoichiometric homeostasis among autotrophs and Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and R Core Team. 2014. nlme: Linear and nonlinear mixed effects models. R package version 3.1. Pinheiro, J., D. Bates, S. DebRoy, D. Sarkar, and R Core Team. 2014. nlme: Linear and nonlinear mixed effects models. R package version 3.1. Pinto, P., M. Morais, M. Ilheu, and L. Sandin. 2006. Relationships among biological elements (macrophytes, macroinvertebrates and ichthyofauna) for different core river types across Europe at two different spatial scales. Hydrobiologia 566: 75 90. Power, M.vE., A.vJ. Stewart, and W.vJ. Matthews. 1988. Grazer control of algae in an Ozark mountain stream: effects of short term exclusion. Ecology 69: 1894 1898. heterotrophs. Oikos 119: 741 751, doi:10.1111/j.1600 0706.2009.18545.x R Development Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. Redfield, A. C. 1958. The biological control of chemical factors in the environment. Am. Sci. 46: 205 221.

PAGE 95

95 Rhee, G. 1978. Effects of N:P atomic ratios and nitrate limitation on algal growt h, cell composition, and n itrate uptake. Limnol. Oceanogr. 23: 10 25, doi:10.4319/lo.1978.23.1.0010 Rolls, R. J., C. Leigh, and F. Sheldon. 2012. Mechanistic effects of low flow hydrology on riverine ecosystems: ecological principles and consequences o f al teration. Freshw. Sci. 4: 1163 1186. 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 lights, nut rients, a nd herbivores. Can. J. Fish. Aquat. Sci. 57: 66 75. Schade, J.D., J.F. Espeleta, C.A. Klausmeier, M.E. McGroddy, S.A. Thomas, and L.X. Zhang.2005. A conceptual framework for ecosystem stoichiometry: Balancing re source supply and demand. Oikos 109:40 51. Sc hindler, 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. Kaisan. 2008. Eutrophication of lakes cannot be controlled by reducing nitrogen input: Results of a 37 year whole ecosystem experim ent. Proc. Nat. Acad. Sci. 105: 11254 11258. Scott, T. M., G. H. Means, R. P. Means, R. C. Means, S. B. Upchurch, R. E. Copeland, J. Jones, T. Roberts, and A. Willet. 2004. Bulletin 66: Springs of Florida. Florida Geological Survey. Sharfstein, B., and A. D. Steinman. 2001. Growth and survival of the Florida apple snail (Pomacea paludosa) fed 3 naturally occurring macrophyte ass emblages. J. N. Amer. Benthol. Soc. 20: 84 95. s with phosphorus limited growth and compe tition of two diatoms. PLoS ONE 9: e102367, doi:10.1371/journal.pone.0102367 Simal, J., M. Lage, and I. Iglesias. 1985. Second derivative UV spectroscopy and sulfamic acid method for determination of nitrate in wat er. J. Anal. Chem. 68:962 964. Slavik, K., B. J. Peterson, L. A. Deegan, W. B. Bowden, A. E. Hershey, and J. E. Hobbie. 2004. Long term responses of the Kuparuk river ecosystem to ph osphorus fertilization. Ecology 85: 939 954, doi:10.1890/02 4039 Smith, V. H., S. J. Joye, and R. W. Howarth. 2006. Eutrophication of freshwater and marine ecosystems. Limnol . Oceanogr. 51: 351 355. Smith, S. L., Y. Yamanaka, M. Pahlow, and A. Osclies. 2009. Optimal uptake kinetics: physiological acclimation explains the pattern of nitrate uptake by phytoplankton in the ocean. Mar. Ecol. Prog. Ser. 384: 1 12, doi:10.1073/pnas.1300962110

PAGE 96

96 Stelzer, R. S., and G. A. Lamberti. 2002. Ecological stoichiometry in running waters: periphyton chemical composition and snail growth. Ecology 8 3:1039 1051. Sterner, R. W., and J. J. Elser. 2002. Ecological stoichiometry: The biology of elements from molecules to the biosphere. Princeton University Press. Stevenson, R. J., A. Pinowska, A. Albertin and J. O. Sickman. 2007. Ecological condition of a lgae and nutrients in Florida springs. WM858, Florida Dept. of Environmental Protection. Tank, J. L., E. J. Rosi Marshall, N. A. Griffiths, S. A. Entrekin, and M. L. Stephen. 2010. A review of allochthonous organic matter dynamics and metabolism in streams . J. N. Amer. Benthol. Soc. 29: 118 146, doi: 10.1899/08 170.1 Underwood, G. J. C., and J. D. Thomas. 1990. Grazing interactions between pulmonate snails and epiphyti c algae and bacteria. Freshw. Biol . 23: 505 522. Valett, H. M., S. A. Thomas, P. J. Mulholl and, J. R. Webster, C. N. Dahm, C. S. Fellows, C. L. Crenshaw, and C. G. Peterson. 2008. Endogenous and exogenous control of ecosystem function: N cycling in headwater streams. Ecology 89: 3515 3527. Vitousek, P. M. 1982. Nutrient cycling and nutrient use efficiency. Am. Nat. 119: 553 572 . Vitousek, P. M. 1984. Litterfall, nutrient cycling, and nutrient limitati on in tropical forests. Ecology 65: 285 298, doi:10.2307/1939481 Ward, M. H., T. M. deKok, P. Levallois, J. Brender, G. Gulis, B. T. Nolan, and J. V anDerSlice. 2005. Workgroup Report: drinking water nitrate and health: recent findings and research needs. Envir on. Health. Persp. 113: 1607 1614. Young, R. G., and A. D. Huryn. 1999. Effects of land use on stream metabolism and organic matter t urnover . Ec ol. Appl . 9: 1359 1376. Young, R. G., C. D. Matthaei, and C. R. Townsend. 2008. Organic matter breakdown and ecosystem metabolism: functional indicators for assessing river ecosystem health. J. N. Am. Benthol. Soc. 27: 605 625, doi: 10.1899/07 121.1 Yu, Q., J. J. Elser, N. He, H. Wu, Q. Chen, G. Zhang, and X. Han. 2011. Stoichiometric homeostasis of vascular plants in the Inne r Mongolia grassland. Oecologia 166: 1 10, doi:10.1007/s00442 010 1902 z Zonneveld, C., H. A. van den Berg, and S.A.L.M. Kooijman. 199 7. Modeling carbon cell quota in light limited phytoplankton. J. Theor. Biol. 188: 215 226, doi:10.1006/jtbi.1997.0470

PAGE 97

97 BIOGRAPHICAL SKETCH Rachel L. Nifong graduated from North Carolina S tate University with a B.S. in e nv ironmental t echnology in 2006. A fter graduation, she continued her studies at the Nicholas School for the Environment and Earth Sciences . During her graduate work at Duke she worked at the United States Environmental Protection Agency quantifying the costs and benefits of reducing air po llution. In 2008, she graduated with a Master of Envi ronmental Management degree in e nv ironmental economics and p olicy; her thesis examined the economic valuation of human health impacts due to air pollution from wildfires in the western United States. Aft er moving to Florida, Rachel began her doctoral work studying Florida spring ecosystems u sing ecological stoichiometry. During that time, she was awarded a US EPA STAR Graduate Fellowship; her fellowship focused on understanding the importance of flow, gra zers, and nutrients on the success of submerged vascular plants as well as the coupling of ecosystem metabolism and autotrophic stoichiometry in order to better inform spring restoration efforts. She now pla ns to continue her research career as a post doct oral research associate at the Appalachian Lab studying the intersection of stoichiometry and Amazonian land use change.