CHARACTERIZATION, MOBILITY, AN D FATE OF DISSOLVED ORGANIC CARBON IN A WETLAND ECOSYSTEM By TODD Z. OSBORNE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005
Copyright 2005 by Todd Z. Osborne
To Zack and Edna Osborne
iv ACKNOWLEDGMENTS No student attaining graduate degree can take sole credit for the work. Often it is an effort of many to prepare a student for grad uation. It is with mu ch appreciation that I acknowledge the efforts and mentorship of my committee members (Dr. Mark W. Clark, Dr. Thomas L. Crisman, Dr. Joseph J. Delfino, and Dr. Willie G. Harris). Dr.Thomas Crisman has been greatly involve d in my graduate study since I began, for that I am especially gratef ul. Special thanks go to my supervisory committee chair and friend, Dr. K. Ramesh Reddy, who through a stea dy stream of Â“opportunitiesÂ”, has given me the tools to be successful and helped me more than he knows. There are many others to whom I owe tha nks: Dr. Susan Newman, for insight into the ecology of the Everglades; Yu Wang, Gavin Wilson, and Ron Elliot for their assistance in the laboratory; Patrick Inglet t and Rex Ellis for valuable insights; Dr. Kanika S. Inglett for assistance formatting this dissertation, and the rest of the students, faculty, and staff of the Wetland Biogeochem istry Laboratory, who, in one fashion or another, have all assisted me along the wa y. I owe special thanks to Dr. E. Lloyd Dunn for helping me see the subtle balance in natu re and the value of keeping that balance in life. To my family and friends I owe many thanks. Your network of support has enabled me to get this far. My beautiful wife, Ann, sacrificed gr eatly and supported me wholeheartedly in this effort. I love and thank her sincerely. I dedicate this work to my
v parents (Zack and Edna Osborne) because w ithout their tireless love, support, and encouragement, none of this would be possible. This research was funded in part by the USDA National Needs Fellowship Program, the South Florida Water Management District, and the National Science Foundation grant DEB-0078368.
vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES...........................................................................................................10 ABSTRACT......................................................................................................................x ii CHAPTER 1 INTRODUCTION........................................................................................................1 Problem Statement........................................................................................................1 Need For Research........................................................................................................3 Hypothesis....................................................................................................................4 Research Objectives......................................................................................................5 Background Information...............................................................................................5 Ecological Significance.........................................................................................6 Abiotic Degradation..............................................................................................7 Biotic Degradation...............................................................................................12 Chemical Composition........................................................................................16 Methods of Characterization...............................................................................19 Separation.....................................................................................................19 Quantification...............................................................................................20 Elemental composition and stable isotopes..................................................21 Ultraviolet, visible and infrared spectroscopy..............................................22 Mass spectrometry........................................................................................23 Nuclear magnetic resonance.........................................................................24 Dissertation Format....................................................................................................26 2 LINKAGES BETWEEN PARTICULATE AND DISSOLVED ORGANIC MATTER IN WETLAND ECOSYSTEMS: THE ROLE OF VEGETATION TYPE..........................................................................................................................3 1 Introduction.................................................................................................................31 Methods......................................................................................................................32 Plant Species and Sample Locations...................................................................32
vii Particulate Organic Matt er Nutrient Analysis.....................................................34 Particulate Organic Ma tter Fiber Analysis..........................................................34 Particulate Organi c Matter E4/E6.......................................................................35 Dissolved Organic Matter Production.................................................................35 Ultra Filtration.....................................................................................................36 Dissolved Organic Matter Nutrient Analysis......................................................37 Dissolved Organic Matter Total Carbohydrate Content......................................37 Dissolved Organic Matter Total Phenolics Content............................................37 Dissolved Organic Matter Specific Absorbance and E4/E6................................38 Linkages between Particulate a nd Dissolved Organic Matter.............................38 Results and Discussion...............................................................................................38 Particulate and Dissolved Orga nic Matter Nutrient Analysis.............................38 Particulate Organic Ma tter Fiber Content...........................................................41 Particulate and Dissolved Organic Matter E4/E6................................................42 Isotopic Analysis.................................................................................................43 Dissolved Organic Matter Mol ecular Weight Fractionation...............................45 Dissolved Organic Matter Total Carbohydrate...................................................46 Dissolved Organic Matter Total Phenolics Content............................................47 Principle Components Analysis..........................................................................48 Summary and Conclusions.........................................................................................49 3 DECOMPOSITION AND MICROBIAL UTILIZATION OF DISSOLVED ORGANIC MATTER DERIVED FROM MA CROPHYTES: A COMPARISON OF PLANT SPECIES FROM TH E FLORIDA EVERGLADES..............................63 Introduction.................................................................................................................63 Materials and Methods...............................................................................................66 Plant Species and Sample Locations...................................................................66 Dissolved Organic Matter Produc tion and Characterization...............................67 Experimental Design...........................................................................................68 Bacterial Growth Efficiency and Decomposition Rates......................................71 Results and Discussion...............................................................................................71 Dissolved Organic Matter Characterization........................................................71 Decomposition.....................................................................................................74 Linkages..............................................................................................................79 Ecological Significance.......................................................................................81 Summary and Conclusions.........................................................................................83 4 PHOTOLYTIC MINERALIZATION AND PHOTO-BLEACHING OF DISSOLVED ORGANIC MATTER (DOM ) DERIVED FROM VEGETATION TYPES OF THE FLORIDA EVERGLADES............................................................93 Introduction.................................................................................................................93 Materials and Methods...............................................................................................96 Plant Species and Sample Locations...................................................................96 Dissolved Organic Matter Produc tion and Characterization...............................97 Experimental Design...........................................................................................97
viii Natural Sunlight Exposure..................................................................................98 Quantification of Photo-Mine ralization and Photobleaching..............................99 Results and Discussion.............................................................................................100 Characterization of Diss olved Organic Matter..................................................100 Photolytic Decomposition and Mineralization..................................................101 Photo-Bleaching................................................................................................105 Ecological Significance.....................................................................................107 Summary and Conclusions.......................................................................................109 5 SPATIAL DISTRIBUTION AND VEG ETATION EFFECTS ON DISSOLVED ORGANIC CARBON IN THE GREA TER EVERGLADES ECOSYSTEM.........117 Introduction...............................................................................................................117 Methods....................................................................................................................119 Study Site...........................................................................................................119 Sampling............................................................................................................120 Data Analysis.....................................................................................................121 Results and Discussion.............................................................................................122 Hydrologic Units...............................................................................................122 Spatial Distribution............................................................................................124 Vegetation..........................................................................................................127 Ecotypes............................................................................................................131 Conclusions...............................................................................................................132 6 SUMMARY AND SYNTHESIS.............................................................................141 Review of Objectives................................................................................................141 Synthesis...................................................................................................................145 Conclusions...............................................................................................................147 Future Research Needs.............................................................................................148 LIST OF REFERENCES.................................................................................................151 BIOGRAPHICAL SKETCH...........................................................................................166
ix LIST OF TABLES Table page 1-1 General chemistry of subs tances isolated from DOM..............................................27 1-2 Percent elemental composition of humic and fulvic acids.......................................27 2-1 Nutrient analysis for all species bulk tissue samples...............................................60 2-2 Nutrient analysis of leachate samples......................................................................61 2-3 Tissue NaOH extraction E4/E6 analysis and leachate E4/E6 analysis....................61 2-4 Isotopic analysis of nitrogen and carbon..................................................................62 3-1 Nutrient analysis of leachate samples.....................................................................90 3-2 Nutrient analysis of standardized DOM...................................................................90 3-3 Characterization analys is of standardized DOM......................................................91 3-4 Results of standardized DOM sa mple decomposition experiments........................91 3-5 Calculated values of carbon transfer for five pathways...........................................92 4-1 Characterization analysis of st andardized DOM sample prior to experimentation......................................................................................................116 4-2 Summary of results of photolys is and photobleaching experiments......................116 5-1 Summary of dissolved organic carb on (DOC) measurements throughout the greater Everglades basin.........................................................................................138 5-2 Summary of plant community diss olved organic carbon (DOC) comparisons throughout the greater Everglades basin................................................................139 5-3 Summary of ecotype comparisons w ith respect to DOC concentration.................140 6-1 Summary of potential car bon storage and losses in th e dissolved organic carbon cycle of a wetland ecosystem.................................................................................150
10 LIST OF FIGURES Figure page 1-1 Conceptual model of the carbon cycle.....................................................................28 1-2 Conceptual model of the microbial loop..................................................................29 1-3 Conceptual model of DOM production....................................................................29 1-4 Flow chart of decomposition of lignin and cellulose substrates..............................30 2-1 Map of areas where plant samples were obtained in 2002 and 2003.......................51 2-2 Results of a 24 hour water extraction.......................................................................52 2-3 Regression analysis of DOM and POM...................................................................52 2-4 Analysis of soluble fiber frac tion of senescent plant tissuess..................................53 2-5 Analysis of hemicellulose fraction of senescent plant tissues..................................53 2-6 Analysis of cellulose fracti on of senescent plant tissues..........................................54 2-7 Analysis of lignin fraction of senescent plant tissues..............................................54 2-8 Regression analysis of DOM total carbon................................................................55 2-9 Molecular weight fractionation analysis <1 KDa....................................................55 2-10 Molecular weight fractio nation analysis 1-3 KDa...................................................56 2-11 Molecular weight fractio nation analysis 3-10 KDa.................................................56 2-12 Molecular weight fractionation analysis >10 KDa..................................................57 2-13 Analysis of leachate total carbohydrate content.......................................................57 2-14 Total phenolics content analysis..............................................................................58 2-15 Scree plot showing th e proportion of variance.........................................................58 2-16 Location of individual samples within PC space.....................................................59
11 2-17 Individual samples projected into principal components space...............................60 3-1 Map of areas where plant samples were obtained....................................................85 3-2 Results of Phase 1 decomposition as loss of carbon over 8 days for Eleocharis , Typha , and Cladium .................................................................................................86 3-3 Results of Phase 2 decomposition as loss of carbon over 32 more days for Eleocharis , Typha , and Cladium ..............................................................................86 3-4 Results of Phase 1 decomposition as loss of carbon over 8 days for Spartina , Thalia , and Nuphar ..................................................................................................87 3-5 Results of Phase 2 decomposition as loss of carbon over 32 more days for Spartina , Thalia , and Nuphar ...................................................................................87 3-6 Results of Phase 1 decomposition as loss of carbon over 8 days for Nymphea , Panicum , Taxodium , and Glucose............................................................................88 3-7 Results of Phase 2 decomposition as loss of carbon over 32 more days for Nymphea , Panicum , Taxodium , and Glucose...........................................................88 3-8 Conceptual diagram of biotic degradation...............................................................89 4-1 Map of areas where plant sample s were obtained in 2002 and 2003.Water Conservation Area (WCA-1).................................................................................111 4-2 Diagram of experimental photolysis exposure chamber........................................112 4-3 Loss of dissolved organic carbon over 9 days of exposure to natural sunlight...................................................................................................................113 4-4 Loss of dissolved organic carbon based upon quantified UVA and UVB exposure.................................................................................................................114 4-5 Linear regression analysis of cha nge in specific absorbance at 254 nm................115 4-6 Linear regression analysis of cha nge in specific absorbance at 325 nm................115 5-1 Map of area where wate r samples were obtained..................................................134 5-2 Locations of all water sampling locations within the greater Everglades basin.....................................................................................................135 5-3 Spatial distribution of DOC in the greater Everglades basin.................................136 5-4 Satellite image of the greater Everglades basin study area....................................137 6-1 Amended conceptual model of dissolved organic carbon......................................149
xii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHARACTERIZATION, MOBILITY, AN D FATE OF DISSOLVED ORGANIC CARBON IN A WETLAND ECOSYSTEM By Todd Z. Osborne May, 2005 Chair: K. R. Reddy Major Department: Soil and Water Science Many investigations into the function of dissolved organic matter (DOM) in wetland and aquatic ecosystems have demons trated that it is of great ecological significance. However, investigations of DOM derived from wetland vegetation have received little attention, even though these plant communities can be a dominant source of DOM. The goals of this res earch were to investigate DOM derived from different plant species commonly found in wetlands to determ ine if differences existed among species in the characteristics of DOM and its reactivity to abiotic and biotic degradation, and if plant communities influence dissolved or ganic carbon (DOC) at the field scale. Chemical and physical characterization of both plant tissue and DOM from these detrital materials revealed significant diffe rences among species with respect to DOM nutrient content, molecular weight fracti onation, carbohydrates, polyphenols, protein, and leaching potential. Fiber fractions and ratios of nutrients were correlated between plant tissue and derived DOM.
xiii Microbial decomposition experiments result ed in significantly different rates and total losses of carbon from the DOM samples. Results suggest that DOM, derived from different plants reacts diffe rently to microbial degradat ion. Total phenolic content and DOM in the molecular weight fraction of one to three kilodaltons explained 86% of the variability (multiple regression analysis ) in the observed loss of carbon during decomposition. Photolytic mineralization of DOC was found to be si gnificantly different among groups of species in UV exposure experiment s. Phenolic content was correlated with changes in specific absorbance at 254 nm, indicating loss of aromatic and hydrophobic structure. Photo-bleaching, measured by lo ss of absorbance at 325 nm, was correlated (r2=0.65) with phenolic compound content in DOM samples, but results suggest that other compounds are contributing to photolysis losses of carbon. Surface water samples collected from 1283 s ites in the Everglades were analyzed for DOC, with values ranging from 4.4 to 83.8 mg C L-1. Results of spatial analyses suggest an influence of canal inputs of DOC into the north ern Everglades. The observed gradient in DOC suggests that most DOC is consumed within the Everglades. Comparisons of mean DOC concentrations within plant communities and ecotypes were significantly different sugges ting that these inputs influe nce DOC dynamics. Results of all experiments suggest that DOM characte ristics (and thus reac tivity) can be source species dependant. Further, the role of plant communities in DOM cycling warrants further investigation.
1 CHAPTER 1 INTRODUCTION Problem Statement Organic matter is a key source of energy in the heterotrophic food web of wetlands and other aquatic ecosystems. This organic ma terial is often derived from autochthonous primary production of algae in open-water systems or the autochthonous and allochthanous production of macrophytes and herbaceous plant species. Whatever the source, this organic material is a vital energy source to the microbial food web (basis of heterotrophic production) and thus is a defining charac teristic of many aquatic ecosystems (Jones, 1992; Wetzel, 1992; Amon and Brenner, 1996). Dissolved organic matter (DOM) (that fracti on operationally defined as less than 0.45 um in size) has received much scrutiny in the last 25 years, as it is often the most labile source of carbon in we tlands and other aquatic ecosys tems and has many effects on the chemical and physical characteristics of the water column as well as the soils of these ecosystems (Wetzel, 1992; Stevenson, 1994; Benke et al., 2000; Ivanoff et al., 1998). This DOM in natural waters is the least understood of the organic matter fractions because of its relatively unknow n chemical structure and th e factors involved in its production, fate, and transport (Fievre et al., 1997; Sachse et al., 2001). Because of the nature of wetland ecosystem s (low oxygen availability in the water column often coupled with low decomposition rates in relation to primary productivity and low redox potential in soils), these sy stems often contain high amounts of organic matter (Reddy and Graetz, 1988; DeBusk et al., 2001). There are many modes of DOM
2 production in wetlands. Decomposition and leachi ng of standing dead biomass, detritus, peat, and the influx of allo chthonous DOM from point or non point sources in the watershed are some of the major pathwa ys (Figure 1-1). Wetland systems present themselves as sites of unique opportunity for the study of organic matter dynamics in aquatic ecosystems due to the amplified rela tionship of organic matter dynamics within the system. Many studies have documente d high organic matter (both DOM and POM) export from these systems (Mulholland, 1981; Klavins, 1997); yet no detailed DOM budgets exist for wetlands. Often, black wa ter rivers and dystrophi c lakes are strongly associated with wetlands, which contribute large portions of DOM to these systems (Mulholland, 1981; Moran and Hods on, 1994; Benke et al., 2000). To date, most studies attempting to ma ke connections between plant communities and DOM exports have centered on soil (Kaise r et al., 2001; Strobe l et al., 2001) and stream (McDowell and Likens, 1988) exports of DOM from watersheds dominated by various herbaceous species. With the exception of a few studies, the significance in terms of microbial degradability of the DOM from various plant materials has yet to be fully investigated (Benner et al., 1984; Mora n and Hodson, 1994; Mann and Wetzel, 1996). Technological advances in rece nt years have enabled inve stigation of the dynamics of DOM with much more scrutiny. Carbon nuc lear magnetic resonance, stable isotope mass spectrometry, and infrared spectroscopy ar e just a few of the tools being employed to delve deeper into the functional role of DOM in terrestrial and aquatic systems at the molecular level. These studies uncovered many new interesting facets of the reactivity and fate of DOM and have revived the interest of those working in aquatic ecosystems. Inhibition of enzyme activities by DOM (B oavida and Wetzel, 1998), UV photolysis and
3 microbial utilization of byproducts (Wetzel et al., 1995; Vahata lo and Wetzel, 2004), formation of disinfection byproducts when tr eating natural waters for municipal supply (Betts, 1998; Weinberg, 1999; Richardson et al., 2002), and variati on of DOM products and nutrient contents from similar tree speci es (Driebe and Whitman, 2000) are just a few of the new avenues of research in DOM ecol ogical functions. However, connectivity is lacking in understanding relationships among sources of DOM (herbaceous and macrophytic vegetation, soils, etc.) and the hi ghly variable char acteristics of DOM observed in recent studies (Sun et al., 1997; Opsahl and Benner, 1999; Amon et al., 2001). Further, many questions remain as to why different types of DOM are found in different aquatic ecosystems, and what f actors affect their production and fate. Need For Research Much is known about the chemistry and ro le of DOM in aquatic ecosystems, and it is obvious that DOM plays a crit ical role in aquatic ecosystem energetics and chemistry; but there is still much to learn about the true nature of this material. Little is known about how changes in DOM quantity and quality a ffect microbial communities and ecosystem function. Further, our understa nding of sources of DOM in aquatic systems (especially wetlands) has historically placed limited valu e on macrophytes, and little is known about the relationships among plant sources of DOM and the role of various plant types in DOM composition and dynamics. Most studies have investigated allochthonous inputs and assumed much about the fluxes from the detrital pool. In most cases, vegetation communities are grouped together as a single and constant source of DOM. In this research, many different methods of characterization were used to quantitatively and qualitatively assess linka ges between DOM and its source material. Evaluation of the bioavailability of DOM from different plant species and their
4 susceptibility to UV degradation help eluc idate the interactivity of different DOM sources within a wetland system. The informa tion gained from these experiments help to better understand the ecological significance of various wetland plant communities and further understand of the reactivity and fate of DOM in wetland ecosystems. The intent of this research is to co ntribute to the overall understanding of the regulators of DOM production, and hence the char acteristics of this dissolved fraction, by investigating the characteristics of the source material (particulate organic matter) and corresponding characteristics of the produced DOM . I also investigat e the degradation of this material under aerobic conditions to understand better the links between DOM composition and degradability. Many factors of interest, including nutr ient content of the parent material, structural and elemental composition, a nd the resulting degradation rates were investigated. Also, the role of environmen tal factors such as UV light exposure were evaluated to determine the extent that these factors affect the characteristics and mobility of this OM fraction in wetland systems. Fina lly, a field scale measur ement of DOC in the Everglades wetland ecosystem was used to investigate relationships between plant communities and DOC production and export. Hypothesis The central hypothesis of this research is th at vegetation type is a significant factor in formation, characteristics, and reactivit y of DOM in wetlands. Specific hypotheses are: Hypothesis 1: DOM from different vegetati on types is characteristically different and thus the func tional role of this material in wetland systems is variable.
5 Hypothesis 2: DOM from different plant sources will react differently to microbial degradation. Hypothesis 3: DOM of different plant origin will vary in its susceptibility to UV photolysis. Hypothesis 4: Spatial distribution of DOM in large wetland systems is dependant on vegetation communities. Research Objectives Objective 1: Characterize both POM and DOM from a variety of dominant wetland vegetation types to determine if differences in DOM exist and if there are significant links between PO M characteristics and resulting DOM characteristics (both chemical and physical). Objective 2: Determine the biodegradability (relative amount of microbial utilization) of DOM produced by di fferent dominant wetland vegetation types and relate this to the characte ristics (physical and chemical) of DOM from each vegetation type Objective 3: Determine the effects of abiotic degradation via UV photolysis and photo-bleaching on plan t derived DOM and determine the potential amount of DOC mineraliz ed from UV light exposure. Objective 4: Investigate the spatial distri bution of DOM and the functional role of vegetation types in various hydrologic units of the Everglades wetland ecosystem. Background Information For over a century, the ecological functi on of DOM in aquatic and terrestrial ecosystems has been a topic of great interest. This stems from the fact that DOM has varying effects on the environment where it ex ists and little is know n about its chemistry and structure. As technological advances in analytical chemistry have been applied to researching this material, more facets of its chemistry and function become apparent, revealing even more questions as to the char acter of this material. The great level of variability in the chemical st ructure of DOM confounds attempts to apply general rules to the reactivity and function of DOM in aquatic ecosystems (Kronberg, 1999).
6 Ecological Significance Dissolved organic matter (DOM) is a ubiqu itous fraction of organic matter found in varying degrees in aquatic and terrestrial eco systems. This form of organic matter is operationally defined as that material passi ng through a 0.45 um filter. This size class includes material from particles in a co lloidal suspension down to small organic molecules less than 100 Da. This DOM is f ound both in the water column and sediment pore waters of aquatic ecosystems, including wetlands, and in interstitial waters in terrestrial soils (W etzel, 2001; Tan, 2003). Interest in DOM began in 1786, Archard firs t described humic substances extracted from soils (Stevenson, 1994) and later, in the 1930Â’s when Krogh and Keys produced a method for quantifying DOM in marine systems (Hansell and Carlson, 2002). Also in the 1930Â’s, Birge and Juday first quantified DOM in freshwater systems while undertaking a limnological study. Since those early studies , much research has been conducted on DOM and several relationships between DOM and the environment have been discovered. The most important of these relationships , from an ecological standpoint, is that DOM is a major carbon source for heterotroph ic production in wetlands, other aquatic, and terrestrial ecosystems (i.e. the base of the heterotrophic food we b). Wetzel (1992) found that up to 90% of the heterotrophi c bacterial productivity of some aquatic ecosystems relies on DOM. The transfer of carbon from the DOM pool to higher trophic levels depends on use of DOM by bacteria to move this carbon source up the food web (Figure 1-2). While often fueling the base of the food web, DOM in aquatic systems can also be a source of nutrients (Qualls and Haines, 1991), sink for toxic elements such as aluminum
7 and anthropogenic organics (Ryan et al ., 1996) (Chefetz et al., 2000), and provide protection to biota from damaging UV light (S chindler and Curtis, 1997). In soils and sediments, DOM is also an energy source for bacteria and can become a source and sink for various elements and compounds through cation exchange and sorption. In both terrestrial and aquatic systems, DOM can have great impacts on water chemistry by buffering pH and controlling nutrie nt availability. The presen ce of this material is not always beneficial though; high c oncentrations of DOM can be inhibitive to biota (Wetzel, 1993), create oxygen poor environments, pose a human health risk by disinfection byproduct formation (Betts, 1998), and limit light availability in aquatic environments (Wetzel, 2001). Due to its hydrophobic nature, DOM can also be a mode of transfer for xenobiotics and toxic elements from one syst em to another via movement (export) with water (Schwarzenbach et al., 2003). Dissolved organic matter is composed of decomposition / metabolic products of both primary (algae and higher plant tissu es) and secondary production (bacterial and higher organism tissues). Aquatic systems tend to accumulate DOM from terrestrial sources through soil and landscape drainage , which is often termed allochthonous because it is imported from outside the system. The DOM produced within the system by decomposition of algae or detrital organi c matter is termed autochthonous. Both autochthonous and allochthonous DOM are com posed of a mixture of both labile and refractory organic compounds. Abiotic Degradation The scientific community has long been aw are of the effects of solar radiation on aquatic ecosystems. The historical viewpoint was one of the biological effects of light on such processes as photosynthesis, photot axis, and biorhythms by photosynthetically
8 active radiation (PAR 400-750 nm). This unde rstanding included biological effects of ultraviolet radiation (UV 200-400 nm) as it pe rtains to the damaging and inhibitory effects on aquatic biota (Zafiriou, 2002; Khan and Wetzel, 1999). However, relatively recent discoveries of the abiotic effects of so lar radiation, particular ly UV, have greatly broadened understanding of th e functional role of light in the chemistry and energy dynamics of aquatic ecosystems. Nowhere is the importance of light better demonstrated than in the dynamics of dissolved organic matter (DOM). It is well known that DOM in aquatic systems can have great and far reaching effects on heterotrophic production and carbon cy cling, metal and nutri ent availability, water chemistry, and water column light attenuation (Wetzel, 1992; Scully and Lean, 1994). Recent investigations have revealed that photolysis, the lysing of carbon to carbon bonds in DOM, and photo-bleaching, the destruction of chro mophores within DOM, can influence the effects of DOM on aquatic systems. Photolysis, often referred to as photo-oxidation, photo-mineralization, photochemical degradation, etc., is an abioti c process where high energy solar radiation (UV light) breaks down or alters DOM in solu tion. Ultraviolet light is highly energetic and is defined in three categories, UVA (320-400 nm), UVB (280-320 nm) and UVC (200-280 nm), by wavelength. UVA and UVB are the most pertinent for photolysis reactions due to the very low transmittance of UVC through the atmosphere. There is some evidence that PAR may also have some influence on photolysis, but the degree to which it is effective is not fully underst ood (Wetzel, 2001). When DOM is exposed to UVA and UVB light, the bonds between carbon molecules absorb this energy, which energizes the electrons associ ated with the bonds to higher (less stable) orbitals. The
9 absorbed energy must then dissipate to bring those electrons back to stable orbitals. Dissipation of energy can be achieved by reradi ating this energy as light (the basis for fluorescence spectroscopy) and is achiev ed by chromophoric DOM or by passing the energy to another molecule ( photosensitization) such as oxy gen (Schwartzenbach et al., 2003; Wetzel, 2001). In the event that th e energy absorbed exceeds the energy of formation for these chemical bonds, the bonds are broken. Destruction of chemical bonds in DOM can result in formation of several products such as smaller organic molecules (Keiber et al., 1990) and reactive species such as hydroxyl radicals, hydrogen peroxide (Obernosterer et al., 2001), fr ee oxygen radicals, and other carbonaceous reactive moieties. These radical species are highly reactive and can subsequently further the decomposition of surrounding DOM. There is some evidence that nitrate and organically bound iron can enhance the activity of UV irradiation on DOM (Sharpless et. al., 2003). While incomplete photolysis is comm on in aquatic ecosystems, complete photolysis (direct mineraliza tion of DOM to CO or CO2) is also possible (Valentine and Zepp, 1993; Zou and Jones, 1997). Mopper et al. (1991) report ed that CO was the major product (approximately 90%) of photolysis of DOM in the upper 20 m of the Sargasso Sea. This result is due in part to the fact that most natural DOM in aquatic systems is humic in origin. Humic materials are often composed of organic functional groups such as phenolics, aromatics, and carboxylics that have double C bonds (unsaturated) (Thurman, 1985). These carbonyl structures are chromophoric in nature and are most susceptible to UV photolysis and photoblea ching (Zepp, 1988). Because of the highly variable nature of humic material composition, humic por tions of DOM can react to
10 photolysis at different levels (Clair and Sayer, 1997); how ever, aliphatic moieties of DOM are not known to react substantially to UV photodegradation due to their low occurrence of conjugate d (double) bonds. Although the full ecological significance of photochemical reactions in natural waters has not been fully realized, much research has been conducted on the possible effects of photolysis on aquatic ecosystem s. Because most of DOM in aquatic ecosystems is humic in nature, it is often recalcitrant to microbial utilization, and therefore, photolysis could potentially provi de a mechanism to breakdown the recalcitrant portions of natural DOM and make it availabl e once again for bacter ial utilization. This would allow DOM that is largely sequestered from aquatic ecosystem energetics to be mobilized. Wetzel et al. (1995) found that exposure to UV light produced numerous smaller organics from humic and fulvic acid fractions of DOM and that the small organic fraction stimulated bacterial gr owth. Similar results have be en found in rivers, wetlands, lakes, and marine systems (Engelhaupt et al., 2003; Reitner et al., 1997; Gellar, 1986). Lehto et al. (2003) found that solutions c ontaining polycyclic aromatic hydrocarbons (PAH) generated by industrial processes (a nd often found in aquatic systems) also became more susceptible to bacterial deco mposition after exposure to UV radiation. Photochemical alteration and subsequent increased decomposition of other harmful organics (by way of indirect photolysis) found in natura l waters have also been investigated with promising results (Hapeman et al., 1998). As well as potentially increasing microbial access to refractory DOM, nutrients bound within DOM can also be accessed. Wang et al. (2000) reporte d production of up to 1.9 uM ammonium per hour by photolysis of natural DOM from various sources.
11 These findings agree with Bushaw et al. (1996) who reported ammonium production of 86 uM per day from photochemical deco mposition of whole DOM samples from Okeefenokee Swamp, GA. Photolytic productio n of ammonium has great implications especially in nitrogen poor aquatic systems. Further, the release of organically bound nitrogen and phosphorus can be a significan t pathway in oligotrophic systems (Wetzel, 2001). Boavida and Wetzel (1998) reported the reactivation of phosphatase enzymes that had been inactivated due to complexation w ith humic compounds. These findings have important implications for nutrient cycling in aquatic systems. While such findings have initiated much interest and research, ther e has been concern forthcoming as well. With the increase in UV radiation with ozone depletion in our atmosphere, many researchers are unsure and apprehensive as to how aquatic systems will respond (Schnider and Curtis, 1997). As previously mentioned, UV radiation can be harmful to aquatic biota. The presence of DOM can func tion to attenuate UV radiation in the photic zone and thus protect biot a from UV damage. Because chromophoric DOM tends to breakdown under exposure to UV light (photobleach ing) the absorbing characteristics of this DOM are gradually lost (Osburn et al ., 2001). Increased UV exposure could also change the carbon budgets (possibly catastr ophically) of many aquatic systems and therefore warrant special atte ntion (Morris and Hargraves, 1997; Shindler et al., 1997). Through recent research it is clear that photolysis, either direct or indirect, has broad reaching implications for the chemis try, organic matter dynamics, and biological processes in aquatic ecosystems.
12 Biotic Degradation As discussed previously, DOM is important as a carbon source for biosynthesis and as an energy source for bacterial metabolism. Many studies have examined the role of DOM in aquatic ecosystem metabolism and secondary production. Due to heterogeneous nature of DOM constituents, consideration ha s been given to studying separated portions of the DOM pool such as high and low mo lecular weight fractions, monomers and polyermers, and DOM from different sources (Findlay and Sinsabaugh, 2003). To date, most of the investigations have focused on the uptake of low molecular weight (LMW) monomers such as amino acids and free sugars (Kirchman, 2003). Due to their relatively small size, these molecules are taken into ba cterial cells easily through transport across cellular membranes. Amino acids in the DOM pool are rapidly ut ilized by bacterial communities because of their rapid uptake. Am ino acids can represent a large portion of the labile DOM fraction and support a grea t deal of bacteria l growth (Hoch and Kirchman, 1995; Tranvik and Jorgensen, 1995; Rosenstock and Simon, 2001). Amino acids are often abundant in aquatic systems in the form of protein. An example of that abundance can be found in the planktonic community where on average, 50% of the biomass (dry weight) is protein. In a review of amino acid utilization in aquatic systems, Kirchman (2003) found roughly 20% of bact erial production was due to amino acid uptake. In a study of coastal waters in Delaware, Hoch and Kirchman (1995) observed up to 100% of bacterial productivity depende d upon amino acids. Protein is thus understandably the most examined constituent of the DOM pool. Monosaccharides, another group of LMW compounds, comprised of single sugars such as glucose, are also very important por tions of the DOM pool that are degradable by bacteria. These monosaccharides are utilized quickly due to their ease of uptake and
13 readily degradable stores of energy for metabolism. Like amino acids, glucose can account for 20-30 % of bacterial production and up to 97% in extreme cases (Rich et al., 1997). Glucose and fructose are the two mo st abundant monosaccharides in aquatic environments (Kirchman, 2003). Most resear ch concerning monosacc haride utilization by bacteria is found in comp arison studies between monosaccharides and amino acids. Dissolved amino acids (DAA) comparison st udies found amino acids more important than glucose for growth and metabolism (Tra nvik and Jorgensen, 1995). This is likely due to the fact that DAA can be used fo r both energy production and biosynthesis, while glucose and fructose are used exclusively for energy production. Combinations of sugars and proteins, such as DNA and RNA (pyrimid ine bases and pentos e sugars) are also found in abundance (up to 20% of the DOM pool ) in some aquatic systems dominated by phytoplankton (Kirchman, 2000). Low molecular weight organic acids are also very bioavailable for bacterial uptake. Pyruvate, acetate, formate, and malonate are all LMW organic acids found in the DOM pool. They are, however, rapidly utilized by bacteria due to the ease of uptake and propensity to support significant bacterial growth (Bertilsson and Tranvik, 1998). Respiration of malonic and acet ic acids were found to be significant (approx. 20% of bulk additions) in studies by Bertilsson a nd Tranvik (1998) who also measured 98% uptake and mineralization of formic acid by ba cterial communities. The different levels of utilization of these organi c acids suggests that differe nt portions of the DOM pool degrade at different rates and therefore propor tions of LMW organics, as well as sugars and DFAA, could predict bioavailability and degr adability. It is al so important to note that photolysis products of DOM often include LMW organic acids such as those
14 mention above, and thus explai n why increased bacterial acti vity is observed in many photolysis experiments with DOM. Unlike monomers, polymers (polysaccharides , proteins) and high molecular weight (HMW) organic molecules (humic, fulvic ac ids) in the DOM pool are less readily available to bacteria for biosynt hesis or metabolism. This is due to size restrictions on uptake across the cell membrane, which in turn, necessitates the use of exocellular enzymes. These enzymes function to hydroly ze or cleave portions of large organic molecules into smaller pieces to facilitate uptake by bacteria (Moran and Hodson, 1989). This process requires energy on the part of bacteria to produce exocellular enzymes and also inherits a time component (Ljungdahl and Eriksson, 1985). Hence, large polymers such as lignin and cellulose degrade slowly and DOM of ligno-cellulose origin is more refractory to microbial utili zation (Hoppe, 1983). While ligni n and cellulose are the two most ubiquitous biopolymers on earth, our understanding of lingo-cellulose derived DOM degradation by bacteria is limited (Benner and Hodson, 1985; Moran and Hodson, 1989). All cleavage products of exocellular enzyme activity are not utilized by bacteria and therefore are incorporated into the DOM pool (Benner et al., 1984, 1986). Lignin is very high in aromatic content which transl ates to high aromatic content of the DOM produced in its decomposition. Phenolics are known to ha ve a range of effects on microbial communities from stimulation of re spiration to antimicrobial and inhibiting effects (Puupponen-Pimia et al., 2001). It is recognized th at as organic matter goes through the decay continuum from fresh plant li tter to refractory peat material, the DOM produced at each stage is less bioavailable and hence, more recalcitrant (Figure 1-3). This is likely due to the increased proportion of lingo-cellulose-derived DOM (Moran and
15 Hodson, 1989). Ligno-cellulose derived DOM c ontains both polysaccharides, which can be broken down into monomeric sugars for ra pid bacterial utiliza tion, and lignin derived phenolics compounds which likely build up in the DOM pool and contribute to the formation of refractory aquatic humic subs tances (Thurman, 1985; Fustec et al., 1988; Hessen and Tranvik, 1998; Tan, 2003) These humic substances can have detrimental effects on decomposition as they can complex with certain bacterial exocelluar enzymes and deactivate them (Boavida and We tzel, 1998; Espeland and Wetzel, 2001). Whether constituents of the DOM pool ar e LMW or HMW, derivatives of large biopolymers such as lignin or simple mono meric sugars and amino acids, these organic materials are critical to the metabolism and growth of the aquatic microbial community. Often different groups of bacteria utilize different compounds within DOM pool. Hence, the relative proportion of di fferent compounds in the DOM pool can regulate bacterial growth based upon the supply of these compoun ds and the bacterial assemblages present (Williams, 2000). Other than the molecular size and form of DOM constituents, other factors can regulate bioavailability of DOM as well. Th e limitation of macro and micro nutrients can affect the ability of microbial communities to utilize DOM. Concentration can also be a factor in that very high concentrations of DOM can alter pH significantly. Enzymatic activity of microbial communities have been shown to decrease in the presence of high amounts of humic substances, presumably by binding to the proteins and denaturing or deactivating them (Wetzel, 2001). Environmental factors such as temperature and water quality factors such as dissolved oxygen content, pH, and conduc tivity can influence microbial activity and thus biot ic degradation rates of DOM.
16 Literally hundreds of studi es of the bioavailability of DOM in various aquatic environments have shown that variable propor tions of the DOM pool are utilizable by the microbial assemblages present to some degree (Findlay and Sinsabaugh, 2003). These studies suggest that the chemical character of the DOM and the source material from which it is derived are determining factor s in its decomposition. Further, these investigations suggest that the chemical nature of the DOM and the microbial communities present will both be important in the determination of the degree to which a particular pool of DOM will be utilized. Chemical Composition The mixture of organic materials in DOM can be separated into two very general categories; humic and non-humic substances (Aiken et al., 1985). The non-humic substances contain simple organic compounds such as proteins, carbohydrates, lipids, waxes, and low molecular weight organic ac ids. These highly labile materials are generally of low molecular weight (<500 Da) and in many cases, because they are so readily degraded, only make up a smaller perc entage of bulk DOM averaged over time. The higher molecular weight, more recalcitrant , materials that make up the majority of DOM are the humic substances. It has b een shown that humics make up at least 70 percent of the aquatic DOM and in many cas es more (Thurman, 1985; Qualls and Haines, 1991) (see Figure 1-1). These humic substances are resistant to microbial decomposition in part due to their size and chemical structure. Humic and Fulvic acids make up what is collectively referred to as humic substances. Humic substances are mostly de rived from partial decomposition of plant polymers such as cellulose and lignin. These humics are characterized as biogenic heterogeneous organic materials. They are of high to very high mole cular weights (500 to
17 over 100,000 Da) and often impart a yellow to black color to natu ral waters. Most humics are chromophoric in that they ab sorb UV radiation and can undergo photolytic degradation (Bertilsson and Tranvik, 2000). There are currently two theories of humi c substance formation or humification. One concept is that humics are formed directly from cellulose and lignin decomposition, termed the polyphenol theory (Figure 1-4) , and are only mildly transformed through microbial processes and abiotic processes su ch as photolysis and sorption (Stevenson, 1994; Wershaw, 2004). The other theory is that low and medium molecular weight organic molecules, especial ly carbohydrates and protei ns, condense through a complex set of reactions to form high molecular we ight, less bioavailable substances (Hayes, 1998). There is evidence that the process of humification is likely to be a combination of both theories, depending upon the source organi c materials and the level of biotic and abiotic transformations that have occurred (McKnight and Aiken, 1998). Humic substances in the DOM pool are separated into two classes based upon solubility, with humic acid being insoluble at low pH (2) and fulvics being soluble at all pH values (Stevenson, 1994). Humic acids ar e characterized as having higher molecular weights (5000 to over 100,000 Da), high aromatic carbon content, decreased mobility and diffusion, high uptake potential for hydrophobic compounds and proteins (Knicker and Hatcher, 1997), and lower oxygen functional group content (Wetzel , 2001). These humic acids are less susceptible to microbial degrad ation than their fulvic acid counterparts, even though they exhibit higher carbon and ni trogen contents (Stevenson, 1994; Wetzel et al., 1995). Often, humic acids impart a da rk brown or black colo r to natural waters.
18 Fulvic acids are generally of lower mo lecular weight (500-5000 Da) than humic acids and often make up a majority of the humic substances present in DOM. They contain more oxygen functional groups, are the least polymerize d, and exhibit lower nitrogen content relative to humic acids. Fulv ic acids generally impart a yellow to light brown color to natural waters and are more susceptible to microbial degradation. Aiken et al. (1985) presents a gra phical representation of gene ral humic and fulvic acid structures. Humic and fulvic acids in DOM are heterogeneous mixtures of many different organic moieties (Table 1-1). Th ese are only general compound classes; the possible individual analytes are infinite. The chemical and structural complexity of natural DOM has introduced great difficulty to researchers attempting to better und erstand this material. Characterization of any DOM sample is often the first step in understanding its chemical reactivity and overall ecological role. This characterization is made di fficult because DOM is often found associated with metals, detrital material of various origins, and the simple fact that these materials are such an amalgamation of so many organic moieties. Separation of this matrix of organics to investigate a particular class of compounds can be exhaustive and in some cases impossible without changi ng the chemical nature of the sample. Further, due to the aggregation of many biopol ymers and other organic molecules, signals of particular structures are often masked (Hay es, 1998). Separation of this material to its original fragments is curren tly elusive and therefore the actual modes of formation, or humification, are not clear. While the high co st of various methods of characterization is also inhibitive to this area of research, the study of DOM is widespread. Technological
19 advances in recent years have opened new door s to investigations into the nature and properties of natural DOM. Methods of Characterization Historically, the first methods of charact erization of DOM were based upon simple separation of DOM into fractions based upon solubility. Then th e quantification of carbon within these fractions made genera l comparisons of DOM between samples possible. Stable isotope analysis and elemen tal analysis helped derive origins of DOM while UV/VIS absorbance analysis made iden tification of very general compound groups possible. In the mid 1900Â’s the advent of in frared spectroscopy (IR), gas chromatography / mass spectrometry, and 13C nuclear magnetic resonance (13C NMR) among others, brought more powerful tools to the forefront of DOM research. Although there are a host of very powerful methods of molecular analysis of DOM, this study will focus on the use of filtration separation techniques, elemental composition of carbon, nitrogen, and phosphorus, and some speciation of these elements, UV Â–Vis spectrophotometry, and wet chemistry char acterization techniques. The following discussion of methodology is meant to exemplif y the broad range of techniques available to study DOM. Separation As discussed previously, the separation of DOM into humics and fulvics is based on simple pH manipulation. Extraction met hods for attaining proteins and hydrophobic fractions are in common use today and em ploy apolar solvents to extract these compounds form the bulk material (Parish, 1999 ). More advanced methods are also being employed now to separate DOM based upon size and functionality. Many molecular weight fractions can be separated using ultra-filtration, di alysis, gel permeation
20 chromatography (GPC), ultracentrifugation, and electrophoresis (Wetzel, 2001). Ultrafiltration and dialysis use molecular sieves to separate DOM into molecular size fractions using pressure and diffusion, respectively (Amon and Brenner, 1996; Hoffman et al., 2000). GPC and electrophoresis use charge gr adients to separate samples based upon the polarity of the DOM constituents while retainin g less polar entities within the gel matrix. Sachse et al. (2001) were successful in employing size exclusion chromatography (SEC) to fractionate DOM from dystrophic lakes in to polysaccharides, low molecular weight organic acids, and humic substances. Orga nic resin columns (XAD-4, XAD-8) have also been used extensively to separate polar and neutral species, as well as, humic and fulvic acids from bulk DOM (Waiser and Robarts, 2000; Fukushima et al., 2001). Using macroporous absorbent resins, Imai et al. (2001) fractioned DOM into five categories for characterization; aquatic humics, hydrophobic neutrals, hydr ophilic acids, bases, and hydrophilic neutrals. These resins retain target molecule type s and then can be eluted to extract the target molecules in the eluent. They are relatively inexpensive compared to the required hardware for SEC, GPC and ultrafiltration. Quantification The quantification of carbon in DOM has not changed very much in the last few decades. In the early years of quantification of carbon, ther e were bench top methods of wet oxidation using high heat coupled with persulfate or concentrated chromic acid to oxidize organic matter. To avoid the use of such strong chemicals, most researchers employ an organic carbon analyzer that uses either strong UV light or high temperature combustion furnaces to oxidize organic samples. These oxidizing methods are coupled with infrared (IR) carbon detectors and can quickly and easily meas ure levels of carbon in bulk or fractioned samples. It should be not ed here that some of the more advanced
21 techniques (e.g. nuclear magnetic resonance, mass spectrometry) that produce integrated spectra could be used to estimate carbon content as well. Elemental composition and stable isotopes The use of elemental composition has been employed for many years in studies that compare pre and post exposure of DOM to microbial degradation. Elemental composition also gives information relative to levels of saturation (C:H indicat ive of aromaticity and fatty acid content) and levels of humificati on. Kracht and Glexnir (2000) used elemental analysis effectively in differentiating sources of DOM in a stream and adjacent peat bog. Likewise, Santos and Duarte (1998) used elem ental analysis to el ucidate the level of incorporation of paper mill effluent into humic and fulvic acids of a nearby stream. An exceptional model (using elemental analysis) is presented by Sun et al. (1997) who used the following equation with much success to predict the bioavailibility of DOM (r2=0.93). Bioavailability = a + a1(H:C) + a2(O:C) + a3(N:C) Elemental analysis is simple given the availa bility of an elemental analyzer that can measure C, H, N, and S. Oxygen is normally quantified by subtraction. Some elemental analyzers only measure C and N, but ofte n this is enough information to estimate bacterial growth efficiency (Kroer, 1993). Measuring levels of each major element in the DOM can give many clues into origin, level of humification, and potential fate of DOM (Table 1-2). The use of stable isotopes, such as 13C and 15N, give researchers the unique opportunity to directly identif y a molecular fingerprint in bulk DOM. Stable isotopes can be used to differentiate between marine vs . terrestrial derived DOM, allochthanous vs. autochthonous sources, algal vs. macrophyte derived DOM, and even between different
22 higher plant sources (Hansell and Carlson, 2002) . Wang et al. (2002) successfully used 13C to differentiate the relative contribution of DOM from sugar plantations vs. that found naturally in the Everglades. Because sugar cane and wetland vegetation of the Everglades have different biochemical pathways of car bon fixation (C4 and C3 respectively), the per mil difference in 13C enrichment can be used to elucid ate the contributi ons or source of the DOM. Kracht and Gleixner (2000) used 13C to identify fractions of DOM that bacteria communities fed upon by measuring b acteria tissue enrichment. The stable isotope of nitrogen, 15N, is commonly used in tracing food web dynamics and trophic relationships with DOM due to its high rate of incorporation into biomass of low order organisms. The measurement of stable isot opes requires the use of a mass spectrometer, and although this equipment is expensive, st able isotopes remain one of the most popular tools for evaluation of DOM dynamics in many ecosystems. Ultraviolet, visible and infrared spectroscopy The principle that light energy (photons) can be absorbed, and thus cause excitation and/ or vibrational changes in organic bonds, is the basis for the use of UV/VIS and IR spectroscopy. These are the simplest and fi rst two mechanisms for modern analytic structure determination. Ultraviolet light is more energetic than th e visible, but both can be used to excite organic chemical bonds thus causing and absorbance of energy at particular wavelengths. These bonds have indi vidual areas of the spectrum at which they express peak absorbance and thus samples can be scanned for presence of functional groups (bond types)(Crews et al., 1998). Absorbance at specific wavelengths can be useful information as well. The E2/E4 and E4 /E6 ratios are often used in describing the ratios of humic/fulvic acids and the intensity of color in DOM samples (Stevenson, 1994; McKnight and Aiken, 1998). These ratios ar e comparisons of absorbance values at
23 265,465, and 665 nm. These are often used as a quick reference when more detailed structural information is not needed or general comparis on is being made (Pempkowiak et al., 1998). Before the advent NMR, IR was the single most important analyt ical tool for the elucidation of organic chemical structure. Infrared spectroscopy is based on similar principles of UV/VIS. The major differen ce is that bonds emit vibrational energy when exposed to energy in the 25um-2.5um wavele ngths (mid IR) and these vibrational changes in energy state correspond very well to bond energies in organic chemical bonds. The advantage of this technique over UV/VIS is that the frequency of bond vibration is more discrete and thus better interpretations of the spectra are possible. Unfortunately, IR still is not the best technique for evaluati ng organics. The interf erence of solvents and overlapping of signals can often mask the tr ue identity of func tional groups. With Fourier Transformed IR (FTI R) some of the interferen ces can be eliminated via integration of multiple scans and the remova l of solvent spectra. Johnson et al. (2001) successfully used FTIR to identify loss of functional groups between soluble and insoluble peat material. Although IR is no l onger used exclusively due to the relative high level of uncertainty, most research toda y does use it as a backup for more advanced analytical tools (Johnson et al ., 2001; Pempkowiak et al., 1998). Mass spectrometry Unlike UV/VIS and IR, mass spectrometry (MS) does not entail absorption or emission of energy. Instead, organic species are ionized and separated into charged fragments that are then detected. The mass to charge ratio of these ionized fragments is plotted against their relative abundance. By creating a library of known fragmentation spectra, one can easily identify fragmentati on patterns based upon the mass to charge
24 ratio and relative abundance to propose a ch emical structure of the unknown sample. With DOM, the sheer amount of material and species within the sample complicate the elucidation of simple structur es (Crews et al., 1998). While the fragmentation patterns of known compounds are present, the interferen ce of such a large number of fragments limits the usefulness of this method in a pplications of DOM (Hayes, 1998). Many different types of MS are used today in conjunction with othe r analytical instrumentation. Most of the different forms of MS arise from the method in which the sample is ionized and fragmented. Often, electron ionization ch amber (EI), a pyrolysis chamber, or laser ionization beams can be used to ionize a nd gas chromatography (GC), to separate a sample into size fragments before it enters the MS. Pyrolysis-GC/MS is currently the most popular MS setup used when investig ating natural DOM. Zegouagh et al. (1999) had success with this method in identifying alkane, alkene, organos ulfur, and phenolic compounds in marine sediment DOM. Like wise, Kracht and Gleixner (2000) combined pyrolysis-GCMS with isotopic data to furt her elucidate information on biogeochemical processes of DOM degradation, bacteria l growth efficiency, and preservation. It is common to see MS used in conjunc tion with other analyt ical tools (Stevenson, 1994; Gjessing et al., 1998; Dai et al., 2002). Nuclear magnetic resonance A very powerful tool to DOM research is the use of nuclear magnetic resonance (NMR). Beginning in the 1950Â’s, the use of NMR has fast become the standard with which to determine organic structures and functional group composition (Crews et al., 1998). The guiding principle behind NMR is that many natural elements have a significant natural magnetic force around th eir nuclei. Different electron bonding scenarios affect this magnetic field and thus by measuring changes in these discrete fields
25 when a strong magnetic source is applied, one may determine what bonding types are present. Because the magnetics of stable isotopes are somewhat different than their common counterparts (13C vs. 12C), the NMR is directed towa rd looking at isotopes more often. This is the case for 13C and 15N NMR. The simplest of NMR techniques involves looking at hydrogen (1H NMR), a common element in organics. Because deuterium and tritium have very low natural abundance, deuter ated water is often us ed as a solvent and natural H is focused upon. This enables the re moval of interference of natural water. When analyzing a natural DOM sample , lyophilization and dissolution into D2O is often required (Pempkowiak et al., 1998). 13C NMR can also be performed to a less accurate degree on solids (Solid State NMR). The sp ectra of solid state NMR can strongly overestimate the abundance of a liphatic carbons without hete rosubstituents (Zegouagh et al., 1999). Recent advances in the manipulation of 13C NMR by Preston (1996) have increased the coupled use of 13C NMR with other analysis to attain more definite results when working with organic matter. Pres ton (2001) emphasizes the importance of collaboration with experienced NMR speci alists when drawing conclusions about samples in the absence of secondary analyses such as FTIR or MS. Basic information provided by analysis such as UV/VIS and IR could be greatly enhanced by coupling with 13C NMR. Many recent studies have been su ccessful when coupling NMR with other techniques. Thomsen et al. (2002) used NM R coupled with UV to differentiate sources of DOM from various ecosystems. NMR wa s the key to a broad characterization of organic matter by Gjessing et al. (1998) w ho also employed UV/VIS, GC/MS, and HPLC to characterize the DOM. Santos and Duarte resolved the presence of lignosulfonic acids
26 with NMR when FTIR and MS could not do so during an investiga tion of pulp and paper mill effluents. Many other researchers have used this technique to better characterize their organic matter samples and further resolve independent analysis problems (Zegouagh et al., 1999; Wu et al., 2000; Johnson et al., 2001; Dai et al., 2001). The use of 15N NMR and others (P and S) is currently less popular, probab ly due to the high cost and low amount of comparable data. A thorough re view of the techniques commonly used in NMR is provided by Crews et al. (1998). Dissertation Format This dissertation begins w ith Chapter 1 where the problem statement, need for research, hypotheses, objectives and releva nt background information are presented. The next chapter (Chapter 2) consists of a char acterization of plant ma terial and resulting DOM derived from this material, as well as , analysis of linkage s between the two. Extensive characterization results are used in subsequent chapte rs for comparative analyses. Chapter 3 begins the experiment al phase with invest igations into the bioavailability and degradability of DOM derive d from different plant species. Chapter 4 follows up with experiments designed to determine abiotic degradation rates and potentials of these same DOM samples. Results of Chapters 3 and 4 are integrated into models of DOM lability and export. Chapte r 5 presents a large scale modeling of the Everglades ecosystem and suggests influen ce of agricultural drai nage to DOC dynamics in the system. Analysis of plant communities and ecotypes by comparison of mean DOC values is used to investigat e the potential role of vegeta tion in DOC dynamics in a large wetland system. Chapter 6, the final chapter, is a summary and synthesis of the results of the dissertation and a proposal fo r further research directions.
27 Table 1-1. General chemistry of subs tances isolated from DOM by sorption chromatography using nonionic and ioni c exchange resins. Adapted from Herbert and Bertsch (1995) fr om Qualls and Haines, (1991). Fraction Compounds Hydrophobic Neutrals Hydrocarbons Chlorophyll Cartenoids Phospholipids Humics with less than 1 ionic or Phenolic group per 13 carbons Weak (Phenolic) hydrophobic acids Tannins Flavonoids Other polyphenols Vanillin Strong (carboxylic) hydrophobic acids Fulvic and humic acid Humic bound amino acids and peptides Humic bound carbohydrates Aromatic acids (Phenol/carboxyl) Oxidized Polyphenols Long chain fatty acids Hydrophilic acids Humic like w ith lower mol size and higher COOH/C ratios Oxidized carbohydrates with COOH groups Hydrophilic neutrals Small carboxylic acids Inositol and other sugar phosphates Simple neutral sugars Non-humic bound polysaccharides Alcohols (
28 LitterMicrobial biomass DOC HCO3 PeatMicrobial biomass DOC HCO3 -CH4 CO2 CO2 Decomposition/leaching Decomposition/leaching Decomposition leaching Decomposition/leachingCH4 UV Import Export Figure 1-1. Conceptual model of the carbon cy cle in wetlands and littoral zones of a quatic ecosystems. Import arrows designat e allochthonous sources of POM/DOM from ups tream or terrestrial sources. UV represen ts the action of u ltraviolet light on DOM pool. Depictions of pools of carbon storage do not repr esent relative size or degree of ecological significance
29 POM POM Zooplankton Zooplankton DOM DOM Microbial Biomass Microbial Biomass Fish & Invertebrates Fish & Invertebrates Figure 1-2. Conceptual model of the microbi al loop. The process of transformation of POM to DOM and incorporation of DOM carbon into microbial biomass is termed the microbial loop. This transf ormation is integral to organic carbon availability to higher tr ophic level consumers. Live plant Plant standing dead Litter layer Surface peat Buried peat DOM CO2CH4 Figure 1-3. Conceptual model of DOM produc tion along the decay continuum from fresh plant material to highly decomposed peat material. Note that as the bulk POM is decomposed, the resulting DOM produced from these materials is less bioavailable.
30 LigninCelluloseand other non-lignin substrates Phenolic aldehydes and acids Microbial utilization and oxidationPolyphenols Microbial utilization Quinones Humic AcidsFulvic Acids LigninCelluloseand other non-lignin substrates Phenolic aldehydes and acids Microbial utilization and oxidationPolyphenols Microbial utilization Quinones Humic AcidsFulvic Acids Figure 1-4. Flow chart of deco mposition of lignin and cellulose substrates as described by the Â“polyphenol theory of humifica tionÂ” (Stevenson, 1994). Products of decomposition end up in the humic acid pool, the most recalcitrant pool of DOM found in aquatic systems.
31 CHAPTER 2 LINKAGES BETWEEN PARTICULATE AND DISSOLVED ORGANIC MATTER IN WETLAND ECOSYSTEMS: THE ROLE OF VEGETATION TYPE Introduction Dissolved organic matter (DOM) represen ts a majority of the organic carbon exported from terrestrial ecosystems into aquatic habitats (Wetzel and Manny, 1977; Wetzel, 1992). This DOM forms the basis fo r heterotrophic production and thus is very important in the transfer of energy in aquatic food webs (Kap lan et al., 1980; Fisher and Likens, 1973; Moran and Hodson, 1990; Tranvi k, 1992). It also has been found to be a significant source and sink of essential nutrien ts such as nitrogen and phosphorus (Lewis, 2002; Qualls et al., 2002; Qualls and Richardson, 2003). The ch emical characteristics of the DOM pool can control the availability and la bility of metals and toxic organics in the aquatic ecosystem as well as affect water ch emistry and light availa bility (McKnight et al, 1985; Strober et al., 1995). A review of the literature finds hundreds of studies that use a variety of techniques, both qualitative and quantita tive, to characterize the DOM fraction (see Findlay and Sinsibaugh, 2003, for a thorough review). There ex ists a great range in values of these characteristics for both lentic and lotic waters gl obally. Further, it is often unclear as to the origin of the DOM, whether it is auto chthonous or allochthonous, terrestrial or aquatic in origin. While it is commonly asse rted that most DOM in freshwater wetlands originates from terrestrial or aquatic macrophytic or woody species, little attention has been given to determination of the direct sour ces of this material, i. e. what plant species
32 are responsible for the wide range of charac teristics found in different wetland systems. This is partially due to the fact that DOM in aquatic systems undergoes constant decomposition and alteration by microbial communities and UV photolysis (Hongave, 1994; Wetzel et al., 1995; Amon et al., 2001; Pull in et al., 2004). Therefore, identifying a single source of DOM can be extremely cha llenging, especially in the presence of many plant or algal sources and an unknown age of the material. Thus, the role of different plant communities in the produc tion of DOM is relatively un known. Further, it has yet to be shown that different species of macrophyt es contribute charact eristically different portions to the DOM pool. Understanding the linkage between chemical characteristics of macrophyte and woody plant parent material and the chemical characteristics of the resulting DOM produced can be a powerful tool in determining the role of vegetation in DOM dynamics and the associated effects that DOM has in the aquatic ecosystem. The focus of the research discussed in this chapter is to determine: 1) if different wetland vegetation types produce significantl y different forms of DOM and 2) the linkages between the chemical composition of parent plant material s and the nature of DOM produced from these materials. The hypot hesis for this study was that different wetland vegetation types will produce characteristically different DOM. Further, the chemical characteristics of the parent mate rial, or POM, will predict the chemical and physical characteristics of the DOM that originates from this material. Methods Plant Species and Sample Locations To investigate the relationships between plant derived organic matter and DOM in wetlands, selected species of dominant wetl and vegetation were c hosen and the standing dead biomass collected from sites within the Florida Everglades. Specific vegetation
33 types chosen for this study were common to the greater Everglades wetland ecosystem and included Typha domingensis Pers., Cladium jamaicense Crantz, Panicum hemitomon Schult. , Spartina bakerii Merr., Eleocharis interstincta (Vahl) R.and S., Thalia geniculata L., Taxodium disticum (L.) Rich., Nymphea odorata Ait., and Nuphar luteum (L.) Sibth. and S. All plant material was colle cted above the current water level to avoid possible pre leached material. Stems, leav es, and inflorescence were collected and combined. Collection of senescent Nymphea and Nuphar that was not in the water column was accomplished by locating sites of water drawdown induced senescence where the plant material was exposed above th e saturated soil surface. In the case of Taxodium , the only woody species collected, only leav es were used in tissue and leachate experiments. Sites of collection of these plants in cluded the Loxahatchee National Wildlife Refuge (WCA-1), Water Conservation Area (WCA) 2A, 2B, 3A, 3B , the Miccosukee and Seminole Indian Reservations, Big Cypre ss National Preserve, and the Everglades National Park (Figure 2-1). Collection of samples occurred during the months of November Â– December 2002 and in 2003 duri ng months of natura l senescence. Plant tissue collection consisted of harvest of recently senesced above ground biomass. For each species, a minimum of 10 samples were collected, returned to the laboratory and immediately dried at 55 ÂºC. The temperature of 55Âº C was chosen to avoid artificial lignification of the tissues observed with higher drying temper atures (Roberts and Rowland 1998). Tissue samples were subseque ntly ground in a large Wiley mill to 1mm and thoroughly mixed to produce 3 aggregate samples. Sub samples of the ground tissue were ball milled as necessary for nutrient analysis.
34 Particulate Organic Matter Nutrient Analysis Analysis of plant detrital material total nitrogen (TN) and total carbon (TC) was conducted on a Carlo-Erba CN analyzer, tota l phosphorus (TP) was analyzed by ashing, acid digestion, and colorimetric analysis on an auto analyzer (Bran Leubbe Auto Analyzer 3 with digital colorimeter), 13C and 15N signatures were analyzed via Finnigan Mat Delta plus XL mass Spectrometer. Particulate Organic Matter Fiber Analysis Soluble tissue fiber (lipids, proteins, cellular constituents etc.), hemicellulose, cellulose, and lignin fiber fractions were quantified by a modified sequential fiber extraction method (Ankom Technology, Fairport, NY) modified from the feed and forage analysis by Van Soest (1970). This method has been semi-automated by employing an Ankom Total Fiber Analyzer 200 which consis ts of a contained sample carousel with adjustable heating and agitati on. This methodology has been shown to be repeatable and consistent for a given plant type (Roberts a nd Rowland, 1998). This analysis required the use of a sequential extraction process and calcu lation of fiber fractions by mass loss after each extraction. Briefly, approximately 500 mg of a sample of ground tissue (1mm) was placed in a permeable high density polyethylene envelope and heat sealed, it was then extracted with a neutral dete rgent solution ( 30 g sodium lauryl sulfate, USP; 18.61 g ethylenediaminetetraacetic disodium sa lt, dehydrate; 6.81 g sodium tetraborate decahydrate; 4.56 g sodium phosphate dibasic, anhydrous; and 10 ml triethylene glycol in 1 L distilled water) at 100ÂºC w ith agitation, 20 g sodium sulfite , and in the presence (4 ml per 2 L detergent solution) of heat stable alpha amylase. After 75 min. of washing, samples were removed and washed 3 times w ith boiling water, rinsed with HPLC grade acetone for 3 minutes and dried again at 55Âº C. After drying, samples were weighed
35 again and mass loss was calculated. This mass re presents the soluble plant tissue fraction (waxes, lipids, proteins, and ot her cellular constituents) which is equivalent to the neutral detergent soluble fraction described in the fo rage analysis. To determine hemicellulose content, the samples were further extracted with an acid detergent solution (30 g cetyltrimethylammonium bromide 1 L 1 N sulfuric acid for 1 hour at 100Âº C. The samples were then washed with boiling water and acetone and dried again for reweighing as described previously. The difference in mass was calculated as the loss of hemicellulose (the ADF fraction of the forage analysis). To determine cellulose, the samples were digested in 24 N sulfuric acid for 3 hours at room temperature with agitation every half hour. Samples were again washed and dried and reweighed. The mass loss was calculated as cellulose content, and the remain ing material calculated as lignin and ash. Samples were subsequently combusted at 550Âº C for 4 hours to determine mineral content. A complete description of material s and equipment for this method can be found at the Ankom company website : http://www.ankom.com . Particulate Organic Matter E4/E6 To obtain E4/E6 measurements on plan t detrital material, 3g ground sample material was shaken overnight with 25 ml s of 0.1 M NaOH. The extracts were then diluted to standard concentration of 25 mg /l carbon and analyzed for absorbance at 465 and 665 nm on a Shimadzu UV-160 UV-Vis sp ectrophotometer (Dilling and Kaiser, 2002). Dissolved Organic Matter Production To determine the potential for detrital material to produce DOM, ground senescent plant tissue (5 g) was extracted in room te mperature DDI water (300 ml) for three hours with continuous stirring. Afte r three hours, extracts were sequentially filtered through
36 Whatman 934-AH 90mm glass microfiber filte rs, Whatman GF/F (0.7 Âµm) 47mm glass fiber filters, and Pall Supor-450 (0.45 Âµm) 47mm membrane filters to remove particulate material from the extracts. A 24 hour extr action side experiment was performed to determine if three hours was sufficient to maximize release of DOM from the ground material (see Figure 2-2). Potential extr actable DOM was quantified and used as a descriptor for POM as well. DOC concentratio ns in each extract were standardized based upon carbon content requirements for each ch emical analysis. DOC concentration was determined by high temperature oxidation of DOM to CO2 coupled with IR detection of CO2 in a Shimadzu TOC-5050 (Columbia, Mary land). Bulk DOM for all other analysis was created by leaching excessive amounts (approximately 200g) of ground plant material in 1 liter of DDI water to achiev e very high concentrations of DOC. These extracts were used as stoc k solutions for making DOM of various concentrations for separate analyses. Ultra Filtration Molecular size fractions of leachates were determined by ultra filtration techniques utilizing Amicon 8500 continuously stirred ultra filtration cell s. Millipore regenerated cellulose filter membranes (76 mm diamete r) of nominal molecu lar weight limit 1000 (YM1), 3000 (YM3), and 10,000 (Y M10) Daltons were used under 55 psi of ultra pure nitrogen gas in continually stirred cells. Filtrate volume of 160 mls out of 200 mls samples (80%) were used to calculate DOC in size fraction ranges (Tadanier et al., 2000). Filtrate and retentate were analyzed for DOC on a Shimadzu TOC-5050 for mass balance calculations.
37 Dissolved Organic Matter Nutrient Analysis Total Carbon (TC) of DOM was determined by diluting samples 10:1 with acidified DI water and an alyzing for DOC on a Shimadzu TOC-5050. Total Nitrogen (TN) analysis was performed following met hods for Total Kjeldahl Nitrogen (TKN) which entailed acid digestion and autoclaving of samples to digest organic matrix and release nitrogen for colorimetric analys is (EPA method 351.2). Total Phosphorus (TP) analysis also employed a TKN digestion and colorimetric analysis for P (EPA method 351.2). Nitrate content was determined on diluted samples by Alpkem Rapid Flow Analyzer with coupled cadmium reducti on column (EPA met hod 353.2) and Ammonium by colorimetric analysis on an Alpkem 300 Se ries auto analyzer (EPA method 351.2). Soluble Reactive Phosphorus (SRP) was determ ined on diluted samples colorimetrically using a Bran Leubbe AA3 with digi tal colorimeter (EPA method 365.1). Dissolved Organic Matter Total Carbohydrate Content To determine total carbohydrat e content in the leachates, a modified version of the Phenol-Sulfuric Acid Method was used (Dubois et al., 1956; Liu et al ., 1973). Briefly, 1ml of sample was added to 1 ml of phenol re agent (10% phenol / DDI water v/v) in a 30 ml Pyrex test tube. Then 5 ml of concentrat ed sulfuric acid was added rapidly by pipette (heat generated speeds color development reaction). The samples were allowed to sit for 10 min to allow for complete color developm ent and then absorbance recorded at 485 nm on a Hach DR4001 UV-Vis spectrophotometer. Solutions of dextrose at 25, 50, 100, and 200 ummol/ml were used for prepara tions of the standard curve. Dissolved Organic Matter Total Phenolics Content Total phenolics content, content of bo th mono and poly phenolics compounds, was determined following the colorimetric method of Price and Butler (1977). Briefly, 250 ul
38 of sample was added to 25 mls of DI water and 3 mls of ferric chloride reagent (0.1M solution of ferric chloride, FeCl3, in 0.1M hydrochloric acid. After 3 minutes, 3 ml of potassium ferricyanid e reagent (0.008M K3Fe(CN)6 in DI water) was added and mixed. The sample was allowed to sit for 15 minutes at room temperature for color development. Sample was then measured for absorbance at 720 nm on a Shimadzu UV-160 UV-Vis spectrophotometer. Analytical grade tannic acid, C76H52O46, (Fisher Scientific A310500) was used as a standard. Dissolved Organic Matter Specific Absorbance and E4/E6 Specific UV absorbance (SUVA) was obtained by measuring absorbance of pure sample at 254 nm on a Shimadzu UV-160 UV-Vis spectrophotometer and dividing by carbon content. This ratio is used to estim ate the aromatic content of the DOM sample (Dilling and Kaiser, 2002). Similarly, pure samp les of leachates standardized to 25 mg C L-1 were measured for absorbance at 465 and 665 nm to determine E4/E6 ratios. This ratio is used to estimate the level of humic like substances in the leachates. Linkages between Particulate and Dissolved Organic Matter To determine if any characteristics of the POM of senescent plant tissues contributed to the characteristics of the leachate DOM, regression analyses were performed on the results of the chemical characterization data of POM vs. DOM. Statistical analyses of all data were performed using NCSS software package (Number Cruncher Statistical System, East Kaysville, Utah). Results and Discussion Particulate and Dissolved Organic Matter Nutrient Analysis Analysis of plant detrital mate rial revealed that TC values ranged from 417-479 g kg-1 ( Thalia and Taxodium respectively) with a mean value of 445 g kg-1 or 45% of the
39 bulk material (Table 2-1). Total Nitr ogen (TN) values ranged from 7.6-19 g kg-1 ( Cladium and Nuphar respectively) with a mean value of 12 g kg-1 or 1.2% of the bulk material. Total Phosphorus ranged from 0.342-1.808 g kg-1 ( Typha and Taxodium respectively) with a mean value of 0.841 g kg-1 or 0.84% of the bulk material. Ratios of C to N ranged form 23.2 to 59.6 with Nuphar representing the lowest and Typha representing the highest value observed. Ratios of C to P ranged from 264.8 ( Taxodium ) to 1342 ( Typha ) and ratios of N to P ranged from a low of 9.6 ( Taxodium ) to a high value of 26.4 ( Eleocharis ). Analysis of DOM for nutrient content are e xpressed as g of mate rial leached from 1 kg of senescent tissue (Table 2-2). Values for TC ranged from a low of 21.5 g kg-1 ( Cladium ) to a high of 219.6 g kg-1 ( Nymphea ). Nuphar and Nymphea were both significantly higher in leachable carbon than the remaini ng species. With respect to TN, again Cladium (0.62 g kg-1) expressed the lowest value and Nuphar the highest (6.55 g kg-1) with an order of magnitude greater value. The patter n was observed again for TP with Cladium having the lowest value (0.11 g kg-1) and Thalia having the highest value (1.96 g kg-1). Again the range had more than an order of magnitude difference between the lowest and highest values. Ratios of C:N ranged from a low of 10.1 ( Spartina ) to a high of 51.5 ( Taxodium ). Similarly, Spartina had the lowest C:P ratio (33) and Typha the highest (294). N:P ratios ha d similar range as bulk tissu e nutrients with a low of 1.7 ( Taxodium ) and a high of 14.8 ( Eleocharis ). Soluble Reactive Phosphorus (SRP) exhibited quite a range with Typha being the low value of 0.014 g kg-1 and Thalia the high value of 1.432, a difference of two orders of magnitude. While nearly half the species did not produce measurable ammoni um, there was a two order of magnitude
40 difference between the five specie s that did produce ammonium. Nuphar produced the lowest measured value of 0.02 g kg-1 and Spartina the highest with 1.38 g kg-1. Unlike ammonium, nitrate was detected in all leachates with Cladium producing the least nitrate (0.006 g kg-1) and Nuphar producing the most (0.161 g kg-1). Tissue levels of TC,TN, and TP were found to be quite variable in plants growing the Everglades and the dominate species such as Cladium , Eleocharis , and Typha contained levels of nutrients close to thos e observed in Everglades soils (DeBusk and Reddy, 2003; DeBusk et al., 2001; DeBusk and Reddy, 1998). Similar senescent tissue values are found in the literature for Typha and Cladium (Davis, 1991). The high levels of N and P found in some of the other species are interesting in that these plants may be significant sources of nutri ents to the system during decomposition. Nutrient concentrations of the resulting DOM produced from these tissues was not found to be predictable based upon tissue conc entration of nutrients and was also highly variable. Similar values were found in a study Everglades of Typha and Cladium leachates (Qualls and Richardson, 2003). These findi ngs suggest that different species allocate N and P differently in their structural tissues and therefore DOM from these species can be quite different in respect to nutrient content (Sa lisbury and Ross 1992; Davis and Van der Valk 1993). At the broad level of tissue fractio nation and nutrient analysis employed, the relationships of storage could not be determined. It is also im portant to note that different species of aquatic plants will reallocate nutrients to rhizome and other structures upon senescence, therefore adding a nother level of variability to the DOM produced from these materials (Twilley et al., 1977). To better understand this process, more in-depth research into the structural allocati on of N and P would be required. The results of these
41 experiments show that leachate levels of N a nd P are significantly different and therefore species can play a role in determination of DOM levels of N and P. Furthermore, because no relationship was found between tissue and leachate concentrations of N and P, there exists no strong linkage between species tissues and the resu lting leachates with respect to N and P. However, a significant re lationship between N:P ratios of DOM and POM was observed suggesting similar allocation strategy within soluble fiber among the different species with respect to these nutrien ts (Figure 2-3). The tissue levels of TN did show a moderate correlation to the TC of the leachates (r2=0.49, P< 0.001), likely due to growing conditions of these plants and the amount of soluble fiber. Particulate Organic Matter Fiber Content Analysis of the senescent pl ant tissue for fiber fractions revealed a wide range of values for the soluble fiber fraction (Figure 2-4 to 2-7) Spartina represented the lowest value of soluble tissue content at 11% while Nuphar represented the highest at 56%. Nymphea (51%) and Taxodium (46%) were also very high in relation to the other species. Reciprocally, Taxodium had the lowest value of hemicellulose (13%) and Spartina had the highest (40%) (Figure 2-5). Tissue cellulo se values (Figure 2-6) where less variable with Nuphar having the lowest (18%) cellulose content and Eleocharis having the highest (40%). A majority of the species had cellulose content between 37-40%. Lignin, the most structurally significant component of tissue fiber, and the most recalcitrant to decomposition was found to be relativel y low in abundance in all species but Taxodium , a woody species capable of higher lignin production. Values ranged from 6% ( Panicum ) to 17% ( Taxodium ), with most species cont aining 6-10% lignin in thei r tissues (Figure 2-7). The relationship between the soluble fibe r plus hemicellulose fractions and the leachate TC was significant (r2=0.82, P<0.01). This is likely due to the water solubility
42 of these two tissue fractions. It does provide a useful tool in pr edicting the amount of leachable carbon with respect to plant spec ies (Figure 2-8). In wetland and aquatic systems, it is accepted that nutrient availability can affect the partitioning of tissue fiber produced in a particular plant with respect to other fiber types. The relationship observed between solubl e fiber content in the POM and DOM nitrate concentration (r2=0.73, P<0.01) suggests similar alloca tion of nitrate by the plants. Because of the manner in which nitrate is us ed by plants, it is necessary for it to be located in the plant cell. Because the leaching process removes cellular constituents, the level of soluble fiber in the tissues should correlate to the amount of nitrate and other N forms as well as SRP. In the case of a mmonium and SRP, no relationship was found. This disparity suggests that the species used in this experiment di fferentially reallocate SRP and possibly free ammonium upon senescence. Particulate and Dissolved Organic Matter E4/E6 Extraction of POM with weak NaOH result ed in a wide range E4/E6 ratio values (Table 2-3). Eleocharis had the highest value (11.5) and Nymphea the lowest (3.2). Ratios above 5 are considered to indicate more fulvic acid like materials while values below five are indicative of humic acid materials. In the case of plant extracts, E4/E6 ra tios are likely indicatin g the presence of humic or fulvic acid precursors and some in teraction of molecular weight distribution of the extractable materials. Investigation of E4/E6 ratios for leachates found a similar range of values as those of E4/E6 ratios of the tissue NaOH extr action. As in the analysis of POM, Nymphea had the lowest observed value, 1.9. Unlike the tissue analysis, Typha had the highest value of 11.1. In five of the nine species, DOM E4 /E6 ratio increased in relation to the
43 POME4/E6, in two species, the value decr eased in the DOM, and two species E4/E6 ratios did not change significantly. The E4/E6 spectrophotometric analysis re vealed little about the POM or DOM and proved to be unsatisfactory as a predictor of leachate characteristics. Some species POM values were higher than leachat es and vice versa. This suggests that some other factors outside of the ones measured in this study ar e controlling this para meter. Often employed to characterize DOM in bulk DOM studies, the E4/E6 ratio, in this case, could only be used to further the argument that the charac teristics of different species leachates are different and that the POM content of humic and fulvic acid precursors does not accurately reflect the nature of the DOM fo rmed from leaching these materials. The possibility of a methodological problem also exis ts in that Lobartini et al. (1991) report that NaOH extraction of organic matter raised the relative percent of fulvic like materials and therefore raised the E4/E6 ratio to refl ect more fulvic acid presence. This was possible in this extraction, although some E4/E6 ratios were observed to increase in the DOM fraction. Isotopic Analysis Analysis of POM concentrations of 15N and 13C (Table 2-4) revealed a broad range of values for 15N with the lowest value being Spartina (0.11Â‰) and the highest, Thalia (4.00Â‰). In the analysis of 13C, less variable results were observed. With the exception of the monocot grass Spartina (-15.24Â‰), the other species exhibited values in the expected range of -25.44Â‰ to -29.96Â‰, Nuphar and Eleocharis respectively. Results of the 15N and 13C analysis revealed some unexpected changes for the DOM samples. The values for 15N decreased in all but two species ( Spartina +1.1Â‰ and Thalia + 3.4Â‰). Decreases in pe r mil concentration of 15N ranged from 0.5 to 5Â‰.
44 Changes in per mil concentration of 13C were less dramatic with 3 species decreasing ( Typha , Spartina , and Thalia ), two increasing ( Eleocharis and Panicum ), and four not changing significantly ( Cladium , Nuphar , Nymphea , Taxodium ). Magnitude of change in per mil concentrations of 13C in leachates ranged from -1.5 to +2.0Â‰. POM levels of 13C and 15N were not reflected in the resulting DOM which raised the question as to how the indi vidual species of plants alloca te C and N to their tissues. In the case of 13C, unidirectional changes in depletion are expected (Huller et al., 1996; Dawson et al., 2002), but not all sp ecies exhibited this. This re sult suggests that there is some differential allocation to 13C in tissues that would require analysis of fiber fractions to elucidate the relations hips. It also suggests that not all species do this in the same way hence the differences in leachat es observed with respect to 13C (Kracht and Glexiner, 2002). This phenomenon was observed even mo re drastically in the 15N fraction of POM and DOM. There were large changes in the per mil concentr ation of 15N within some species and not others. These findings su ggest that there is even more preferential allocation of 15N within plant tissues. Th e literature supports this in that various N pools (TN, amino acids, NH4+) in soil organic matte r can have drastically different values for 15N (Griffiths, 1998) when the plant source is the same. Values of 13C measured for POM was within the range of those reported for plants in the northern Everglades by (Wang et al., 2002), and while there was no significant predictability between the POM values and the resulting DOM values for either 13C or 15N, the data do suggest that different sp ecies produce isotopically different DOM, especially with respect to 15N.
45 Dissolved Organic Matter Mole cular Weight Fractionation Fractionation of DOM by molecular we ight revealed three groupings in the smallest molecular weight cutoff of 1KDa (Figure 2-9). Eleocharis and Nuphar partitioned 4.3% and 6.5% of their carbon content respectively, in this fraction. Spartina and Thalia contained 29% and 32% re spectively in this size fraction, representing the highest content of this fraction, while Typha , Cladium , Nymphea , Panicum and Taxodium were in the range of 15% to 20% These th ree groupings were f ound to be significantly different from each other by ANOVA with a D uncanÂ’s Multiple Range post hoc test (n=4, =0.05, P<0.01). The 1KDa to 3KDa fr action (Figure 2-10) exhibited similar separation of distinct groupings with Cladium , Spartina , and Thalia representing the low values in the range of 2%-3%. Nymphea was significantly different from the others at 8%. Similarly, Typha was significantly different fr om the other species at 12%. Taxodium , Eleocharis , and Panicum partitioned between 14% and 16% of their carbon into this fraction and Nuphar was significantly higher than all other species at 30%. The 3KDa to 10KDa fraction (Figure 2-11) exhibited less variation in that Thalia alone represented the lowest values at 2% of its to tal carbon being partitioned into this fraction. Taxodium , Panicum , and Spartina made up the next separate group with values ranging from 7%-10%. Nuphar , Typha , and Cladium were grouped into the range of 12% 14%. Nymphea was alone at 18% at the higher end of the partition values, and Eleocharis contained the highest amount of car bon in this size class at 28 %. The least variable size fraction analyzed was the high molecular we ight partition >10KDa, which ranged from 51% ( Nuphar ) to 64% ( Thalia ) (Figure 2-12). All species ex hibited a majority of carbon content in this size fraction. Nuphar and Eleocharis represented the low end grouping of the range (51%-52%) and all other species fe ll into the high end grouping of 57% to 64%.
46 The great diversity in molecular weight fr actions present in di fferent plant species was expected to relate to the bulk tis sue fiber and nutrient status; however, no relationship was found. Studies of bulk DOM from freshwaters and marine ecosystems report higher values for the smaller size fract ions. Burdige and Gardener (1998) report 60-90% of the DOM in an estuarine study to be in the fraction less than 3 KDa and 73% of lake bulk DOM was found to be less than 1 KDa by Waiser and Robarts (2000). In the case of this study, the larg est fraction of DOM was found in the > 10KDa fraction, suggesting that this fraction will undergo much decomposition before finally approaching the values reported for the No rthern Everglades by Wang et al. (2002) who found that on average, 50% of the bulk DOM was in the <1 KDa fraction. While molecular fractionation schemes vary by investigator, the use of four fractions was believed to be adequate to reveal any relati onships present. While no linka ges to tissue parameters were observed, the fractionation scheme did provi de ample informati on to support the hypothesis that there is signifi cant differences present in the DOM produced by different species of plants. Dissolved Organic Matter Total Carbohydrate Analysis of DOM total carbohydrate content is presented as percent of total DOM carbon in the form of ca rbohydrate (Figure 2-13). Cladium , Spartina , and Nuphar represent the significantly lo wer grouping with carbohydrate contents ranging from 29%32% (ANOVA =0.05, P< 0.01) with Duncans Multiple Range post hoc test, Eleocharis , Nymphea , and Panicum represent another signifi cantly different grouping with median values of 36%-40%. Thalia and Taxodium were not significantly different from each other with values of 43% and 45% carbohydrate content respectively, but were
47 different from the two lower groups. Thalia was not found to be significantly different from Typha , which had the highest ca rbohydrate content at 49%. It was suspected that soluble fiber and hemicellulose would correlate with leachate carbohydrate content based upon the simplicity of the hemicellulose structure (as opposed to cellulose and lignin), but the tissu e levels of simple sugars is likely the parameter that would reveal be tter any relationship between th e two. Similar values have been reported for the species Taxodium (29% vs. 42% reported here) by Opsahl and Benner (1999). Other studies have shown that similar levels of carbohydrate in the bulk DOM pool can be found in lakes and streams as well (Volk et al., 1997; Dai et al., 2001) The analysis of total carbohydr ate in the leachate DOM did, however, reveal that there is a large amount of variabili ty among DOM samples which support the hypot hesis that different species produce characterist ically different leachable DOM. Dissolved Organic Matter Total Phenolics Content The results of the total phenolics content analysis revealed similar results as the total carbohydrate analysis as there were 4 distinct groups of values, yet the species composition of those groups was not the same (Figure 2-14). Eleocharis , Spartina , Panicum , and Taxodium made up the group with the lowest values (4.7%-7.5%). The next significantly different group included Cladium and Thalia with a range of 11% to 12%. Nymphea and Nuphar were similar with values of 27% and 30% respectively, and Typha and Nuphar represented the grouping with the highest values at 37% and 30% respectively. Groupings were determ ined as before with ANOVA (n=5, =0.05, P<0.01) and DuncanÂ’s Multiple Range post hoc test. It was expected that factors such as li gnin content would influence the total phenol content to a large degree due to the aromatic nature of the structural material, but no
48 relationship was found with regres sion analysis. It is likely that different species of plants have adapted different chemical defe nse strategies (secondary plant metabolites) which contain various levels of phenolic compounds. These compounds are likely contained mostly in the leaves and used to defend against herbivory. This is well documented for terrestrial plants (Waterman and Mole, 1994). Typha , the species with the highest phenolic content ha s been shown to have allelopa thic interactions with other plants through production and exudation of phenolic and volatile fatty acid exudates (Ervin and Wetzel, 2003). Richardson et al . (1999) reported that phenolic content of Cladium decreased with increased P availabil ity and argued that under conditions of elevated limiting nutrients, Cladium and other plants would a llocate more energy into growth and less into anti-herbivory compounds. In this study, comparisons of tissue TN and TP found no relationship with total phenol ic content. Comparisons of total phenolic content measurements with all tissue pa rameters did not expose any valuable relationships. The analysis for total phenolic content did, however, suggest another level of separation of species leachate character istics and therefore supported the hypothesis that these leachates are significantly different. Principle Components Analysis Although no two leachates were ever grouped in to the same statistically similar groups for all of the analysis conducted, a pr inciple components analysis (PCA) was used to further investigate if the ch emical characteristics of the le achates were truly different. The relative importance of each principle co mponent is represented in Figure 2-15. The total variance explained by component 1 and 2 is 76 % of the total variance in observed data. Plotting vectors of analytical data al ong the axis of component 1 and 2 of the PCA results in a two dimensional representati on of relationships among the data vectors
49 (Figure 2-16). Projection of data groups into principl e component dimensional space (Figure 2-17), shows no species groups overlap each other, meaning species separation is distinct. There appears to be a gr ouping of species, however, in which Eleocharis , Typha , and Cladium occupy a similar region of PCA space. These plants are also the most abundant and dominant species in the Everglades at this time. This grouping suggests that the parameters measured may be reflections of plant life strategies, employed to be effective in life in an o ligotrophic wetland system. This is further evidenced by the grouping of Taxodium , Nuphar , and Nymphea , the three species most adapted to deep water habitats. Finally, these results indicate that with the parameters measured, no two species groups occupy the same dimensional space with respect to the principle components 1 and 2. This suggests again that based upon these parameters of characterization, DOM produced from thes e plants is significantly different. Summary and Conclusions The tissue contents of structural material s such as hemicellulose, cellulose, and lignin, and the wide range of compounds found in the soluble fiber fraction of plant tissues suggest that there are possible differe nces in the manner in which a particular species will decompose in wetland systems. Also, the overall abundance of these basic plant structural materials w ould also suggest that upon se nescence, the DOM released from these materials would be related to th e structural compounds present in the plant tissues. The level of measurable essentia l nutrients, such as carbon, nitrogen, and phosphorus, should likewise be reflected in th e DOM pool produced from plant materials upon senescence and exposure to the aquati c environment. However, based upon the parameters used to describe the POM or se nescent plant tissue, there were only a few significant linkages established th at could be used to predic t the characteristics of the
50 DOM derived. These relationships were on a ve ry basic level, describing TC of leachates and nutrient ratios. More detailed tissue an alysis may be necessary to understand and link tissue characteristics to the DOM pr oduced from leaching these materials. This study has found that different speci es of wetland vegetation commonly found in the Everglades ecosystem do produce charac teristically different DOM products based upon major nutrient content (TC, TN, TP), to tal carbohydrate and phenolic content, and molecular size fractionation. Re sults of significant groupings in these analyses, and thus the separation observed among species in the principle component analysis, support this conclusion.
51 10010Kilometers N E W S #RB #HL #WCA-2B MDLS ENP BCNP-S BCNP-N WCA-3A #WCA-3B WCA-1 WCA-2A #MSIR 1:1,200,000Florida #Study Site Figure 2-1. Map of areas where plant sa mples were obtained in 2002 and 2003.Water Conservation Area (WCA-1) is also known as the Loxahatchee National Wildlife Refuge. Big Cypress National Pr eserve (BCNP) is separated into northern and southern halves, BCNPN and BCNP-S respectively. The Miccosukee and Seminole Indian Reserv ation is denoted by MSIR and the Everglades National Park by ENP.
52 0 100 200 300 400 500 600 0510152025 Time (hrs) DOC (mg C L-1) Taxodium Nuphar Figure 2-2. Results of a 24 hour water extraction of ground Taxodium and Nuphar tissues. The leachates were measured at 1,3,6,12, and 24 hours for dissolved organic carbon. The observed decline in concentration of DOC after 12 hours suggests coagulation and precipitation of DOC. y = 0.65x 4.8 R2 = 0.83 0 6 12 18 51015202530 POM N:PDOM N:P Figure 2-3. Regression analys is of DOM and POM N:P mass ratios (n=27 observations).
53 0 10 20 30 40 50 60 EleoTypCladSpar ThalNuphNymPanTax SpeciesSoluble Fiber Content (Percent) Figure 2-4. Analysis of sol uble fiber fraction of senescent plant tissues. Species are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Soluble fiber content is presented as percent of total fiber content in plant tissues by mass. 0 10 20 30 40 50 60 EleoTypCladSpar ThalNuphNymPanTax Species Hemicellulose Content (Percent) Figure 2-5. Analysis of hemicellulose fracti on of senescent plant tissues. Species are abbreviated as follows: : Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Hemicellulose content is presented as pe rcent of total fiber content in plant tissues by mass.
54 0 10 20 30 40 50 60 EleoTypCladSpar ThalNuphNymPanTax Species Cellulose Conten t (Percent) Figure 2-6. Analysis of cellulose fracti on of senescent plant tissues.Species are abbreviated as follows: : Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Cellulose content is presented as percent of total fiber content in plant tissues by mass. 0 10 20 30 40 50 60 EleoTypCladSpar ThalNuphNymPanTax Species Lignin Conten t (Percent) Figure 2-7. Analysis of lignin fraction of senescent plant ti ssues. Species are abbreviated as follows: : Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Lignin content is presente d as percent of to tal fiber content in plant tissues by mass.
55 y = 7.85x 362 R2 = 0.82 0 50 100 150 200 250 505560657075 POM Soluble Fiber + Hemicellulose (percent)DOM Total Carbon mg C g-1 POM Figure 2-8. Regression analys is of DOM total carbon expr essed as mg C per g tissue versus the combined percent value of POM soluble fiber and hemicellulose (n=27 observations). 0 10 20 30 40 50 60 70 EleoTypCladSparThalNuphNymPanTax SpeciesPercent Leachate <1 KDaA C B B D D C B A Figure 2-9. Molecular weight fractionation analysis <1 KDa. Values represent the percent of total carbon in leachates that was found to be < 1 KDa in size. Letters above bars indica te significantly different groups and error bars represent +\std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax).
56 0 10 20 30 40 50 60 70 EleoTypCladSparThalNuphNymPanTax SpeciesPercent Leachate 1 KDa-3KDa A A A E D C C C B Figure 2-10. Molecular weight fractionation analysis 1-3 KDa. Values represent the percent of total carbon in leachates that was found to be >1 KDa and <3 Kda in size. Letters above bars indicate si gnificantly different groups and error bars represent +\std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). 0 10 20 30 40 50 60 70 EleoTypCladSparThalNuphNymPanTax SpeciesPercent Leachate 3-10 KDaA C C E B D C B B Figure 2-11. Molecular weight fractionation analysis 3-10 KD a. Values represent the percent of total carbon in leachates that was found to be >3 KDa and <10 Kda in size. Letters above bars indicate si gnificantly different groups and error bars represent +\std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladiu m (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nym phea (Nym); Panicum (Pan); Taxodium (Tax).
57 0 10 20 30 40 50 60 70 EleoTypCladSparThalNuphNymPanTax SpeciesPercent Leachate >10 KDaA B B B A B B B B Figure 2-12. Molecular weight fractionation analysis >10 KDa. Values represent the percent of total carbon in leachates that was found to be >10 Kda in size. Letters above bars indica te significantly different groups and error bars represent +\std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). 0 10 20 30 40 50 60 EleoTypCladSparThalNuphNymPanTax SpeciesLeachate Carbohydrate Content (percent)A A A C C B C B D Figure 2-13. Analysis of leachate total car bohydrate content. Values indicate percent of total carbon in leachates in the form of carbohydrate. Letters above bars indicate significantly diffe rent groups and error bars represent +\std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax).
58 0 10 20 30 40 50 EleoTypCladSparThalNuphNymPanTaxSpeciesCarbon as Phenolic Compounds (percent) Figure 2-14. Total phenolics content analysis. Values are represented as percent of total organic carbon in leachates in the fo rm of phenolics compounds. . Letters above bars indicate significantly differ ent groups and error bars represent +\std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Figure 2-15. Scree plot show ing the proportion of variance in the original data (all attributes) summarized by each of prin cipal components. PC Axes 1,2 and 3 summarize over 90% of the original varian ce (24 attributes). These can used independently (the PC axes have correlation 0.00 by definition).
59 Comp. 1Comp. 2 -0.3-0.2-0.10.00.10.20.3 -0.3-0.2-0.10.00.10.20.3 1 2 3 7 8 9 13 14 15 19 20 21 25 26 27 31 32 33 37 38 39 43 44 45 49 50 51 -4-2024 -4-2024 Sol..Fib Hemi Cell Lign T.TC T.TN T.TP T.C.N T.C.P T.N.P T.E4.E6 L.Pot.C Carbo Phen L.E4.E6 L..1KDa L.1.3KDa L.3.10KDa L..10KDa L.TC L.TN L.TP L.C.N L.C.P L.N.P L.SRP L.NO3 Figure 2-16. Location of individual samples within PC space (axes 1 and 2 only) and the vectors describing the correlation between the composite axes and the original sample attributes. The longer the vector , the stronger the co rrelation; vectors that are primary left-right are uncorrela ted with the composite axes 2. Axis one appears to be a nutrient content ax is and axis 2 appears to be a carbon quality axis .
60 -4-20246 Comp.1 -5 -3 -1 1 3Comp.2ele ele ele typ typ typ cla cla cla spar spar spar thal thal thal nup nup nup nym nym nym pan pan pan tax tax tax Figure 2-17. Individual samp les projected into principal components space (for axes 1 and 2 only). Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Table 2-1. Nutrient analysis for all speci es bulk tissue samples. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Total carbon, total nitrogen, and total phosphorus are represented by TC, TN, and TP respectively. Ratios of carbon to nitrogen, carbon to phosphorus, and nitrogen to phosphorus are indicated by C:N, C:P, and N:P. Species TC g kg-1 TN g kg-1 TP g kg-1 C:N C:P N:P Eleo 428 10.1 0.382 42.4 1121 26.4 Typ 459 7.7 0.342 59.6 1342 22.5 Clad 442 7.6 0.358 58.2 1236 21.2 Spar 462 8.9 0.736 51.9 628 12.1 Thal 416 9.5 0.703 43.8 592 13.5 Nuph 440 19 1.310 23.2 337 14.6 Nym 438 15.9 1.120 27.6 391 14.2 Pan 437 12.7 1.148 34.4 380 11.0 Tax 478 17.4 1.808 27.5 264 9.6
61 Table 2-2. Nutrient analysis of leachate samples. Species names are abbreviated as follows: : Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Total carbon, total nitrogen, a nd total phosphorus are represented by TC, TN, and TP respectively. Ratio s of carbon to nitrogen, carbon to phosphorus, and nitrogen to phosphorus are indicated by C:N, C:P, and N:P. Soluble reactive phosphorus is deno ted by SRP, ammonium nitrogen by NH4(N), and nitrate nitrogen by NO3(N). Units are in grams of each analyte leached per kg of tissue. Species TC g kg-1 TN g kg-1 TP g kg-1 C:N C:P N:P SRP g kg-1 NH4(N) g kg-1 NO3(N) g kg-1 Eleo 40.2 2.39 0.16 16.8 250 15 0.052 0.29 0.012 Typ 34.2 1.25 0.12 27.3 294 11 0.014 ND 0.030 Clad 21.5 0.89 0.11 24.1 196 8.1 0.086 ND 0.006 Spar 38.3 3.81 0.69 10.1 55 5.5 0.089 1.38 0.011 Thal 87.3 6.19 0.65 14.1 134 9.5 0.143 0.52 0.026 Nuph 179.2 6.55 1.11 27.3 161 5.9 0.771 0.02 0.161 Nym 219.6 5.53 0.88 39.7 250 6.3 0.109 ND 0.065 Pan 48.2 2.88 1.02 16.7 47 2.8 0.772 0.22 0.014 Tax 105.3 2.04 1.20 51.5 88 1.7 0.990 ND 0.032 Table 2-3. Tissue NaOH extraction E4/E6 analys is and leachate E4/E6 analysis. Species names are abbreviated as follows: : Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). The E4/E6 ratio is the ratio of absorbance at the wavelengths 465nm and 665nm. Species POM E4/E6 DOM E4/E6 Eleo 11 8.7 Typ 9.2 11 Clad 6.1 9.1 Spar 5.5 5.2 Thal 4.3 8.6 Nuph 5.1 8.0 Nym 3.2 1.9 Pan 8.7 8.2 Tax 8.1 9.9
62 Table 2-4. Isotopic analysis of nitrogen and carbon in both bulk tissue and leachate samples from all species. Species are abbreviated as follows : Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax).Values are represented in Â‰ difference from zero. Species POM N15 DOM N15 POM C13 DOM C13 Eleo 2.89 0.70 -29.96 -28.1 Typ 0.71 0.05 -27.77 -29.4 Clad 2.02 -1.20 -26.96 -27.4 Spar 0.11 1.20 -15.24 -16.4 Thal 4.00 7.40 -27.68 -29.2 Nuph 1.08 -4.20 -25.44 -25.4 Nym 1.55 -0.70 -27.15 -25.8 Pan 2.46 -0.50 -29.13 -27.3 Tax 0.64 0.10 -25.39 -25.9
63 CHAPTER 3 DECOMPOSITION AND MICROBIAL UTIL IZATION OF DISSOLVED ORGANIC MATTER DERIVED FROM MACROPHYTE S: A COMPARISON OF PLANT SPECIES FROM THE FLORIDA EVERGLADES. Introduction Dissolved organic matter (DOM) is importa nt as a carbon source for biosynthesis and as an energy source for bacterial meta bolism in aquatic ecosystems. Many studies have examined the role of DOM in a quatic ecosystem metabolism and secondary production and found that DOM is often an abundant source of energy to bacterial communities (Wetzel, 1984; Benner et al ., 1986; Moran and Hodson, 1990; Wetzel, 1992). Aquatic bacteria are uni que in that they are the only organisms capable of utilizing this vast pool of carbon for biosynt hesis and metabolism. Through biosynthesis, these bacterial communities can make this of ten large pool of carbon available to higher trophic levels through grazing of bacterial biom ass by higher heterotrophic organisms. To date, most of the investigations have focu sed on bulk utilization of DOM from various sources and focused on the reac tivity of the recalcitrant, hi gh molecular weight aromatic portions of this DOM (McKnight and Aiken, 1998; Pempkowiak et al., 1998; Pullin et al., 2004). However, relatively fewer stud ies have investigated the uptake of low molecular weight (LMW) monomers such as amino acids and free sugars (Kirchman, 2003). Due to their relatively small size, th ese molecules are taken into bacter ial cells easily through transport across cellular memb ranes. Amino acids in the DOM pool are rapidly utilized by bacterial communities because of their ease of uptake. Amino acids, in the form of protein, can represent a la rge portion of the labile DOM fraction and
64 support bacterial growth (Hoch and Kirc hman, 1995; Tranvik and Jorgensen, 1995; Rosenstock and Simon, 2001). An example of that abundance can be found in the planktonic community where on average, 50% of the biomass (dry weight) is protein. In a review of amino acids in aquatic syst ems, Kirchman (2003) found roughly 20% of bacterial production was due to amino acid upt ake. In a study of coastal waters in Delaware, Hoch and Kirchman (1995) obser ved up to 100% of bacterial productivity depended upon amino acids. Protein is thus understandably the most examined constituent of the DOM pool. Mono and polysaccharides, another group of LMW compounds, comprised of sugars such as glucose, are also very important portions of the DOM pool that are degradable by bacteria. These carbohydrates are utilized quic kly due to their ease of uptake and readily degradable stores of ener gy for metabolism. Like amino acids, glucose can account for 20-30 % of bacterial production and up to 97% in extr eme cases (Rich et al., 1997). Glucose and fructo se are the two most abundant carbohydrates in aquatic environments (Kirchman, 2003). Most res earch concerning monosaccharide utilization by bacteria is found in comp arison studies between monosaccharides and amino acids. Dissolved amino acids (DAA) comparison st udies found amino acids more important than glucose for growth and metabolism (Tra nvik and Jorgensen, 1995). This is likely due to the fact that DAA can be used fo r both energy production and biosynthesis, while glucose and fructose are used almost exclus ively for energy producti on. Combinations of sugars and proteins, such as DNA and RNA (pyrimidine bases and pentose sugars) are also found in abundance (up to 20% of the DOM pool) in some aquatic systems dominated by phytoplankton (Kirchman, 2000).
65 Current theory on bacterial utilization of aquatic DOM holds that low molecular weight materials are more readily utilized and that high molecular weight materials are require enzymatic or physical alteration befo re they can be utilized for bacterial metabolism or biosynthesis (Amon and Benner, 1996). Also, the nature of this material with respect to aromaticity (phenolic structure content) can be a dete rmining factor in the extent to which DOM is utilized by microbial communities in that the higher the aromatic content, the less bioavailabl e the substrate is (Almendr os and Dorado, 1999; Moran and Hodson, 1994). Few studies have investigated the labili ty of DOM derived from decomposition of aquatic macrophytes (Findlay et al., 1986; Mann and Wetzel, 1996; Hullar et al., 1996; Scully et al., 2004), although th is group of plants can be a dominant source of organic matter to wetland ecosystems (Wetzel, 2001). Furt her, little attenti on has been given to the role of different species of macrophytes in wetlands and their contributions to the DOM pool. With changes in plant community structure being a common observation in wetlands undergoing eutrophicati on or high levels of distur bance (Newman et al., 1996; Childers et al., 2003; King et al., 2004; Weisner and Miao, 2004), understanding the ecological implications to DOM dynamics and th e related trophic inte ractions of changes in vegetation are of great ecolo gical concern. This research aims to evaluate different types of wetland vegetation commonly found in the Florida Everglades in terms of the lability of DOM derived from these plants se nescent tissues and the potential contribution of these species to the DOM pool and food web dynamics. A second goal of this research is to compare the characteristics of the DOM known to affect bacterial utilization (carbohydrate content, protein content, phenol ic content, and molecular weight) to the
66 rates of DOM degradation and the bacterial growth efficiency (BGE) to determine how the source and chemical composition of DOM aff ect the biodegradability of this material. This research is based upon the hypothesis that these vegetation types, which have previously been shown to produce chemica lly different DOM, will also produce DOM that differs in its bulk decomposition a nd ability to support bacterial growth. Materials and Methods Plant Species and Sample Locations To investigate the relationships between plant derived organic matter and DOM in wetlands, selected species of dominant wetl and vegetation were c hosen and the freshly senescent standing detrital material collected from sites within the Florida Everglades. Specific vegetation types chosen for this st udy were common to the greater Everglades wetland ecosystem and included Typha domingensis Pers., Cladium jamaicense Crantz, Panicum hemitomon Shult., Spartina bakerii Merr., Eleocharis interstincta (Vahl) R.and S . , Thalia geniculata L., Taxodium disticum (L.) Rich., Nymphea odorata Ait., and Nuphar luteum (L.) Sibth. and S. Plant detrital ma terial was collected above the current water level to avoid possible previously leached material. Stems, leaves, and inflorescence were collected and combined. Collection of senescent Nuphar and Nymphea that was not in the water column wa s accomplished by locating sites of water drawdown induced senescence where the plant material was exposed above the saturated soil surface. In the case of Taxodium , the only woody species coll ected, only leaves were used in leachate experiments. Sites of collection of these plants in cluded the Loxahatchee National Wildlife Refuge (WCA-1), Water Conservation Area (WCA) 2A, 2B, 3A, 3B , the Miccosukee and Seminole Indian Reservations, Big Cypre ss National Preserve, and the Everglades
67 National Park (Figure 3-1). Collection of samples occurred during the months of November Â– December 2002 and in 2003 duri ng months of natura l senescence. Plant detrital material collection consisted of harvest of recently senesced above ground biomass. For each species, a minimum of 10 samples were collected, returned to the laboratory and immediately dried at 55 ÂºC. The temperature of 55Âº C was chosen to avoid artificial lignification of the tissues observed with highe r drying temperatures (Roberts and Rowland, 1998). Tissue samples were subsequently ground in a large Wiley mill to 1mm and thoroughly mixed to produce 3 aggregate samples. Dissolved Organic Matter Production and Characterization The production and characterization of DOM for these experiments was described in detail previously (see Chapter 2). Br iefly, to make DOM samples, ground detrital material was leached using di stilled deionized water for th ree hours and then sequentially filtered, ending with 0.2 Âµm. Samples of DOM were then diluted to appropriate concentrations for experiments and analysis. Molecular size fractions of leachates were determined by ultra filtration techniques uti lizing Amicon 8500 continu ously stirred ultra filtration cells. Millipore regenerated cellulose filter membranes (76 mm diameter) of nominal molecular weight limit 1000 (YM1 ), 3000 (YM3), and 10,000 (YM10) Daltons were used and calculations of molecular weight fractions were made according to Tadanier et al. (2000). Total Carbon (TC) of leachates was determined on a Shimadzu TOC-5050. Total Nitrogen (TN) and tota l phosphorus analysis were performed following methods for Total Kjeldahl Nitrogen (TKN) which entailed acid digestion and autoclaving of samples to digest organic ma trix and release nitrogen for colorimetric analysis (EPA method 351.2).
68 To determine total carbohydrat e content in the leachates, a modified version of the Phenol-Sulfuric Acid Method was used (Duboi s et al., 1956; Liu et al., 1973). Total phenolic content, content of both mono a nd poly phenolics compounds, was determined following the colorimetric method of Price and Butler (1977). Protein content was calculated base d upon Total Keldjahl Nitrogen (TKN) measurements of leachates. Estimates of pr otein content were based upon the convention that organic nitrogen is predominantly in th e form of protein, representing 16% of the total mass. Nitrate and ammonium values were subtracted from the TKN value for each leachate and then the remaining nitrogen value multiplied by a factor of 6.24 to determine crude protein content on a mass basis af ter the methods of Ran et al. (2004). Experimental Design To test the degradability of the nine DOM leachates, reactor ve ssels were used in triplicate for each species leachate that enabled decomposition to occur in the dark under controlled aerobic and temperat ure conditions. A set of reac tors of D-glucose solutions served as controls for comparisons of decom position rates. Reactor vessels consisted of 500 ml wide mouth glass Erlenmey er flasks with two holes in the stopper. In one hole a glass tube was inserted with a 20 gauge syringe needle attached at the end. This tube enabled aeration at the bottom of the vessel and the needle helped minimize the bubble size and thus maximize the effective aeration at the set flow rate. A second much shorter glass tube was also inserted into the stopper to serve as an air outle t. By only extending the tube 1 cm into the headspace area of the flask, it enabled air to escape while minimizing losses by evaporation. Air was pumpe d into the reaction ve ssels by adjustable flow electric aquarium aeration pumps connected to a manifold which serviced all of the vessels via surgical tubing airlines. A total of 30 vessels were used in the experiment, 9
69 species in triplicate and one set of triplicat e glucose controls. All DOM leachates were standardized to a con centration of 100 mg C L-1, prior to inoculation. To ensure nutrient limitation would not affect the decomposition ra tes, additions of concentrated nutrient solutions of SRP, NO3 -, and NH4 + were added to the leachates to give concentrations of 3 mg L-1, 5mg L-1 and 5 mg L-1 respectively. This resulted in a CNP mass ratio of 100:10:3. Based upon reported values for bacter ial biomass (Wetzel, 2001), these levels of nutrients would be ample even if all car bon in leachates was to be incorporated into bacterial biomass and no respir ation was to occur. To acc ount for losses of water through evaporation, the control vessels were monitored for loss of water and necessary additions of DDI water mere made to all reactor vessels based upon those observations. All additions were made prior to reactor sampling. To ensure the decomposition conditions would not favor any particular species, an inoculum was chosen from a nearby wetland ar ea that was not dominated by any of the species being investigated. This was done to ensure that the microbial community used to inoculate the leachates was not attuned to utilizing a particular s ubstrate from one of the species in question and ther efore would not overestimate the lability or quality of the DOM. On the first day of the experiment, a 2L water sample was collected from a shaded and undisturbed area of Lake Ali ce on the University of Florida campus and quickly filtered first at 0.45 um and then through 0.2 Âµ m. The last 100ml retained in the filtration unit, which contained a concentrated microbial community, was used as inoculum for the decomposition reactors. On e ml aliquots were introduced into each reactor at time=0. The amount of DOC in the inoculum was measured and accounted for.
70 All reactor vessels were sampled immediat ely prior to inocula tion and then again on the time series of 0.5, 1, 2, 4, 6, and 8 days after inoculation. On each sampling time, 10 ml of sub sample were taken from each reactor. Half of the sub sample volume was filtered via syringe filtration (0.2 Âµm) for DOC analysis and the other 5 ml was used for TOC analysis. Both DOC and TOC we re analyzed on a Shimadzu TOC-5050 (Columbia, Maryland) for organic carbon. Th e TOC measurement was indicative of both bacterial biomass and DOM and the DOC measur ement indicated the amount of leachate carbon left in the reactor vessels. At the end of the 8 day incubation, reactors were taken offline; all water was filtered at 0.2 Âµ m, and stored in the dark at 4Â°C until the next phase of the experiment. At this time, SRP, TKN, NO3 and NH4 + were also measured on filtered water samples to ensure that th e cultures had not been limited by nutrient availability. To further investigate th e degradability past the initial stage of rapid decomposition, all reactor vessel so lutions were filtered at 0.2 Âµ m and diluted with 50% DDI water. This was done to lower the con centration of any substr ates or products of decomposition produced by bacterial communities in the initial experiment that may inhibit further bacterial util ization of the leachate DOM or bacterial growth. A second nutrient addition was made using the same con centrations as were previously used, fresh inoculum was added, and leachates were again allowed to undergo decomposition. Samples were taken as before, but on a l onger time scale. Sampling times were 4, 8, 16, 24, and 32 days from the second inoculation, comb ined with the initial 8 days for a total of 40 days of decomposition.
71 Bacterial Growth Efficiency and Decomposition Rates Bacterial growth efficiency (BGE) was calculated after the methods of Mann and Wetzel (1996) where BGE = (bacterial bi omass carbon incorporation/ carbon removed from substrate pool) x 100. Carbon incorporat ed into bacterial bi omass was calculated by subtracting the carbon value of the 0.22 um f iltrate from that of the bulk DOM sample. The BGE was calculated for the initial 24 hours as this is often the highest BGE rate during the decomposition process (Mann and We tzel, 1996). The necessity of measuring BGE early in the experimental phase is supported by Leff and Meyer (1991). Decomposition rates were calculated by log transformation of the data and plotting against time in days. A linear regression mode l was then fit to the data and the slope (m) of the regression line through linearized data used as the rate of decomposition (K) per unit of time (day). All statis tical analysis (linear regres sion, comparison of means, and multiple regression analysis were perfor med using NCSS software (Number Cruncher Statistical System, East Kaysville, Utah). Results and Discussion Dissolved Organic Matter Characterization The potentially leachable carbon, that amount of carbon lost on leaching in DDI water for three hours, and the le achable nutrients were found to be quite variable for the nine species investigated (Tab le 3-1). These results suggest that these vegetation types could potentially provide very different levels of DOM with different nutrient contents to the surrounding wetland system. The DOM samples were standardized to equal concentrations of dissolved organic ca rbon (DOC) so that comparisons of DOM degradation rates and BGE would be mo re represented. Standardized sample concentrations of DOM (Table 3-2) are the beginning characteristics prior to the
72 inoculation of the reaction ve ssels. These values do not include the nutrient additions that were made to the DOM samples just before the experiment was initiated. Total carbohydrate content (TCC), reveals a range of values (29-49% of total carbon) and indicates that ther e are significant differences in DOM samples with respect to TCC (Table 3-3). These values are highe r on average than the range of values for carbohydrate content in studies of natural DOM (Volk et al., 1997) which was expected due to the increased bioavaila bility of these compounds in natural environments (Opsahl and Benner, 1996). Because the nutritive a nd metabolic value of monosaccharides has been shown in the literature (Kirchman, 2003), it was expected that TCC would greatly affect the decomposition rates and the total carbon lost form the DOM samples during decomposition, but not enhance BGE due to its relatively low value for biomass incorporation. Although the analysis for TCC is not compound or si ze specific, it was thought to be an adequate measure of labile components of the DOM. Therefore, species such as Typha , Eleocharis , Thalia , Nymphea , and Taxodium were expected to have high decomposition rates and relatively accelerat ed total carbon loss after 8 days. Likewise, protein content estimates for th e DOM samples showed that there were significant differences among the species in terms of protein cont ent (Table 3-3). Further, this suggests that there should be some differences in BGE on DOM based upon protein content, as protein has been shown to be critical for bacterial growth and less important as a source of energy for me tabolism (Tranvik and Jorgensen, 1995; Rosenstock and Simon, 2001). Species such as Eleocharis , Spartina , Thalia , and Panicum , which contained the highest amounts of protein, were expected to have higher
73 BGE, which in turn could also increase th e total loss of carbon from these species DOM samples due to the high demand for metabolic substrate. Unlike carbohydrate and protei n content, total phenolic s content (TPC) of the standardized DOM samples prior to the decomposition experiments indicates that inhibition of microbial growth and meta bolism of DOM is possible due the higher concentrations of some samples (Table 3-3). It has been shown that phenolics compounds inhibit bacterial growth and u tilization of DOM in various aquatic environments and therefore the species with hi gher TPC concentrations were expected to have lower rates of decom position, BGE, and total carbon loss over the term of the decomposition experiment (Almendros and Dorado, 1999; Bano et al., 1997, Findlay and Sinsabaugh, 2003). Of special interest were Typha , Nuphar and Nymphea , which had markedly higher percentages of phenolics co mpounds compared to other species, and therefore expected to have decreased microbial utilizati on of carbon reflected in lower BGE and decomposition rates. Results of the molecular weight fractiona tion indicate a large ra nge of values for low molecular weight fracti ons of <1 KDa and 1-3 KDa in the DOM samples from different species, while the highest molecular weight fraction, that which was greater than 10 KDa, was similar among the nine species (T able 3-3). Similar results have been observed in the Everglades by Wang et al. ( 2002) who found close to 50% of DOM to be below 1 KDa. Much attention has been given to the molecular size of DOM constituents and their lability. Currently, most research agrees in that the smaller (<1 KDa) DOM is much more readily available for rapid mi crobial utilization (Amon and Benner, 1996; Hernes and Benner, 2003; Young et al., 2004). This is due to the relative ease at which
74 these materials can be taken up into the bacter ial cells. The larger fractions require the use of extracellular enzymes by bacteria to cleave off manageable sized molecules for metabolism or biosynthesis (Wetzel, 2001). Therefore, it was expected that DOM samples with higher amounts of <1KDa material, such as Cladium , Spartina , and Thalia , would decompose faster and possibly indica te higher BGE values (Amon and Benner, 1996). Decomposition The results of the decom position experiments (Figures 3-2 and 3-3) indicate Typha decomposed slowly compared to others in Phase 1, but continued to lose carbon during Phase 2, suggesting Typha to be a more difficult substrate to use, but provides a longer term source of available carbon to the bacterial community. Eleocharis and Cladium were very similar to each other in rapid loss of carbon and relative little activity in Phase 2, suggesting rapidly utilizable compounds were used quickly in Phase 1. This also suggests that other factors, possibly phenolics, inhibited further decomposition (Phase 2) due to complexity of remaining substrate. Nuphar , decomposed in a similar manner as Typha (Figure 3-4 and 3-5) in that loss in Phase 1 was slow compared to others, but continued in Phase 2. These results again suggest that the substrate complexity in Nuphar may be less readily available on the short term, but does provide a prolonged usable so urce of carbon to bacterial communities. Like Cladium and Eleocharis , Spartina and Thalia decomposed rapidly in Phase 1 but were fairly stable in Phase 2 indicating that the readily available substrate in the DOM produced from these species is rapidly utiliz ed by bacteria and the remainder likely requires modification to become bioavailable.
75 Nymphea , Panicum and Taxodium decomposed in a similar fashion to each other but more slowly compared to Eleocharis , Cladium , in early Phase 1 (Figures 3-6 and 37). In Phase 2, Nymphea and Taxodium exhibited a minor level of further decomposition, suggesting that these two species were so mewhat intermediate between the rapidly utilizable substrates of some species and the slowly utilized, but su stained bioavailability of species such as Typha and Nuphar . Glucose decomposition, used as a comparison in this experiment, showed somewhat slower loss of carbon in Phase 1 of the decomposition. This is likely due to the lack of organic substrates needed for biosynthesis and therefore bacterial populat ions were using glucose for metabolic maintenance. This would in turn slow the use of glucose as the bacterial community was not increasing in size. In Phase 2, more dramatic carbon loss was observed suggesting that time and size of microbial pool were the factors most important to decomposition of glucose. Nitrogen in the form of nitrate and a mmonium, as well as, phosphorus as SRP, were monitored throughout both phases of decomposition. In Phase 1, all samples were dramatically reduced with resp ect to nutrients, but none deplet ed suggesting that nutrient limitation was not a factor in decomposition process. In Phase 2, samples reacted differently as those species that continued to lose carbon also reduced concentrations of available N and P, while those that did not s howed less utilization of N and P available. As in Phase 1, no samples revealed nutrient limitation as a possible reason for lack of decomposition of substrate and therefore it wa s concluded that remaining substrate after 40 days of decomposition was in fact, not bioavailable in its present form.
76 Loss of carbon after Phase 1 (8 days) and phase 2 (40 days) of decomposition experiments represent the ra pidly utilized DOM pool and the total bioavailable DOM pool respectively (Table 3-4). All DOM samp les were greater than the glucose control with respect to carbon loss in th e first eight days. This sugge sts that the glucose control, employed here for comparison of substrate qua lity, was lacking in vital substrate for biosynthesis and therefore the bacterial inoc ulum could not metabolize as much of the carbon as the microbial commun ities in other DOM samples. Further evidence of this can be found in the very low BGE of glucose samples (Table 3-4). The 8 day carbon loss was highest in Spartina (79 %) and lowest in Typha (44%). Nuphar was not significantly different that Typha in 8 day carbon loss, suggesting th at both these species produce less rapidly bioavailable substrates compared to other species DOM. After 40 days Typha and Nuphar were still the lowest va lues of carbon loss indica ting a low percentage of bioavailable DOM. Most of the other species did not change signifi cantly past the Phase 1 and this is reflected in the decomposition rates (K day-1) (Table 3-4). Species that had slower decomposition rates ( Typha and Nuphar ) continued to lose carbon in Phase 2, while those species with high decomposition rates, such as Nymphea , Spartina , and Thalia , lost a majority of their carbon in the fi rst few days of Phase 1. The differences in decomposition rates and total loss of carbon after 40 days for different species DOM suggests that these plants offer differing leve ls of bioavailable DOM to the microbial communities. Further, the time scale on which this material is utilized is variable, suggesting that DOM from different plant so urces could affect microbial communities on different temporal and spat ial scales when hydrology is considered. The calculated bacterial growth efficiencies (BGE) of th e DOM samples (Table 3-4) further suggests
77 that DOM from these different species reacts differently to decomposition. For example, Taxodium and Spartina have BGE of 12% and 13% re spectively, yet there is a 15 % difference in their respective 40 day losses of carbon and a si gnificant difference in their respective decomposition rates. This suggest s that BGE is not de pendant upon the total amount of carbon that is bioavailable, but ra ther upon the differences of substrate types and qualities. Another intere sting example is that of Cladium and Eleocharis , similar decomposition curves, rates and total carbon lo ss after 40 days, but very different in terms of the BGE, further suggesting that s ubstrate quality is va riable and therefore reflected in the ability of microbial communities to use substrates for biosynthesis. Of particular interest is the comparison of Typha and Cladium . A prominent observation of concern in the Northern Everglades is the encroachment of Typha into areas once dominated by Cladium (Newman et al., 1996; Weisner and Miao, 2004). These two plant types differ in their respective values for to tal carbon loss to microbial communities, the time scale on which this mate rial is available, and the level of BGE provided by the DOM. While Typha DOM is relatively slow to decompose, almost half of the bioavailable carbon is availabl e for bacterial biomass production while Cladium DOM decomposes more rapidly and has a gr eater amount of bioavailable carbon, but only about one third is available for bacter ial biomass production. These results suggest that there exists a signifi cant potential for changes to DOM dynamics concomitant with vegetative shifts in plant communities in the Everglades. As stated previously, there are few studi es which examine the DOM derived from wetland and aquatic vegetation. Findlay et al. (1986) measured decomposition of leachates derived from living leaves of Alternanthera philoxeroides and found that 70%
78 of the carbon in those leachates was lost in the first 72 hrs of decomposition and a total of 79% lost in 30 days, similar to the results observed for Spartina . The measured BGE on live Alternanthera DOM was 59% which exceeds any va lues observed in this study, but would be expected as the plant material had not been subjected to natural senescence and likely had high concentrations of utilizable s ubstrates and nutrients. Similar results were observed by Sun et al. (1997) who found BGE values of 59% for leachates of Nuphar leutem and Alternanthera philoxeroides and BGE value of 17.4% for Taxodium . Again, as in the study by Findlay et al. (1986), leachates were made from living tissues, and therefore higher values for BGE are expected when compared to this study. Mann and Wetzel (1996) reported BGE values of 16-34% for Juncus effuses derived DOM and 445% for Typha latifolia . Comparisons of Typha data from this study with that of Wetzel and Mann (1996) suggests that in the case of Typha , species within this genus may contain similar levels and qualities of subs trates utilizable for bacterial biomass production. In contrast to the findings of this experiment, recent studies of natural DOM samples reveal much lower levels of bioavailability and BGE. For example, Bano et al. (1997) report BGE of 22% for natural DOM from the Okeefenokee Swamp, and only 1.8% of the DOM was actually utilized. Si milarly, Moran and Hodson (1994) found that humic rich DOM in coastal waters of Geor gia only lost 24% of the carbon in 7 weeks with no nutrient limitation. A recent study by Cammack et al. (2004) reported BGE for DOM samples from 28 lakes in southern Quebec with an average BGE of 21 %. The results found here are within the ranges of values reported in the literatu re. In most cases, these values are for highly humified DOM or DOM of terrestrial origin. The results of this study suggest that aquatic and wetland vegetation are potentially very important
79 sources of DOM. Further, the amounts a nd rates of carbon loss may exceed those suggested by studies of more terrestrial or recalcitrant DOM. Linkages Stepwise regression was employed to identify the characteristics most important to decomposition (carbon loss) from DOM samp les derived from wetland vegetation. The analysis indicated that 86% of variability was explained by Phenol and 1-3 KDa. Using all parameters in multiple regression models re vealed that co-linearity existed within the parameters due to the relationships of MWF summing to 1. To deal with co-linearity problems, the MWF data was partitioned into ratios and log transformed to create a new set of variables. Upon running the multiple regression models with the new variables, the explained variability did not increase. Theref ore, the two best predictors of lability of DOM for these experiments were input in to the multiple regression model and the resulting r2 value of 0.86 used. Many iterations of the regression model were attempted using different combinations of the variable s measured, but the best explanation was due to phenol and MWF of 1-3KDa, with phenolic s content being negatively correlated with carbon loss and the MWF of 1-3 KDa being positively correlated with carbon loss. However, the press r2 value (0.48), a metric for compar ison of relationships to larger populations, suggests that more data would be necessary to apply this relationship outside of the Everglades. The expectation that car bohydrate content would be a major factor in the amount of total carbon loss to decompositi on is supported in the literature (Brinkman et al., 2004) but was not supported in these experimental result s. There is the possibility that some of the carbohydrate material was associated with the high MWF and therefore was unavailable for microbial utilization. Li kewise, protein content, known to increase BGE and potentially accelerate carbon loss wa s not found to be significant in predicting
80 carbon loss or BGE. Again, it is possible that the majority of protein in the samples was in the form of high MWF and therefore largely unavailable at this time. Recent studies by Hagerman et al. (1998) and Northup et al. (1998) both suggest that phenolic compounds in DOM can bind to proteins and en zymes, effectively taking them out of the pool of bioavailable DOM. With the leve ls of phenolics compounds measured here, protein binding with phenolics has great potential to affect the results. Phenolic content played a major role in affecting the decomposition process but how exactly is unclear. While phenolic compounds are more difficult to break down, many times they can be found in the low MWF, effectively masking the metabolic value of the low MWF. An interesting finding in this study was that the amount of phenolic compounds present seemed to decrease the amount of total carbon lost; however, the amount of carbon lost after the phase 1 deco mposition (8 days) was positively correlated to the amount of phenolic mate rial in the initial DOM (r2=0.75, P<0.01). These findings suggest that there are two possible mechanisms in which phenolics are important. First, phenolic compounds interfere with enzymatic activities of bacter ia, or bind with proteins in the DOM pool, effectively reducing th e amount of carbon the microbial community can access (population control or enzymatic cont rol). Then, as decomposition progresses, the microbial communities respond with more enzyme production and therefore increase decomposition of DOM, but take more time to do so. Another possible explanation and one relatively new to the study of DOM is the possible negative effects of secondary plant metabolites, compounds produced by pl ants as a chemical defense against herbivory. These compounds are often aromatic in nature and have received very little attention in aquatic plants, although the effects of secondary plant metabolites are well
81 known in terrestrial plants and their antim icrobial, antifungal and antiherbivory effects are well documented (Haslam, 1989 and referenc es therein). Recent investigations by Sutton and Portier (1991) and Kubanek et al. ( 2001) suggest that the role of secondary plant metabolites is vastly underestimated a nd relatively unknown. Kubanek et al. (2001) have shown that Saururus cernus contains compounds that deter crayfish feeding and possibly have antimicrobial effects. Fu rther evidence by Newman et al. (1998) and Wilson et al. (1999) suggest these compounds are present in other aquatic vegetation types as well. Both Cladium and Typha have been documented as having secondary metabolites useful in deterring herbivory and plant competition respectively (Weisner and Miao, 2004) and allelopathly has been observed in Eleocharis interstincta (Sutton and Portier 1991). It is th en plausible that these compounds coul d be present in some or all of the species investigated here and possibly be affecting th e microbial utilization of DOM from these plants. Since there is no eviden ce that these compounds persist temporally in natural systems, it is possible that they dela y the utilization of car bon or cause microbial communities to shift to resistant species eff ectively delaying the loss of carbon from the DOM pool. Ecological Significance Incorporation of the findings of this study into a conceptual model is very instructive for the purposes of evaluating the ecological significance of aquatic and wetland vegetation sources of DOM to DOM dynamics (Figure 3-8). This model represents the flow of orga nic carbon (in the form of DOM) from bulk plant tissue through the biotic decomposition and transf ormation process. Organic carbon (as DOM) is leached from senescent plant material and into the DOM pool. It is either bioavailable or recalcitrant to microbial utilization. If r ecalcitrant, it is considered to follow the route
82 of export from the immediate system. Bi oavailable DOM is then acted upon by the microbial communities present and a portion is mineralized through respiration to CO2. The remaining portion is incorporated into ba cterial biomass through biosynthesis. This portion is then available to higher trophic levels within the system or exported as particulate organic matter to go through the system of decompositi on again. Calculated values for each of the steps in this conceptu al model were derived from these experiments (Table 3-5). Comparison of species agains t each other finds a significant difference among the species in their potential level of contribution of organic carbon to the DOM pool per Kg of tissue and that th e submergent aquatic vegetation, Nuphar and Nymphea , contribute extreme amounts of DOM to th e surrounding DOM pool. This suggests that different species can have diffe rent levels of impact in terms of bulk contribution of carbon. Investigation of the potential bioava ilable contribution by each species again reveals that there is a significant differ ence among species of plants, resulting in significantly different levels of export of DOM from the system as well. In kind, the highest producers of DOM are also the highe st exporters. Producti on of bioavailable organic carbon is dominated by the su bmergent vegetation species with Taxodium and Thalia producing about half as much bioa vailable carbon. The remaining species contribute about half as much again as Thalia and Taxodium . Microbial respiration of plant derived organic carbon is expectedly higher than th at portion used in bacterial growth which suggests that species with high amounts of carbon used for respiration could be significant contributors to nutrie nt cycling as well. Microbial biomass production seems to be relatively high in Thalia , Nymphea , and Nuphar , while the other species support very little microbial biomass production. Th is has significant ecological
83 implications in that a large amount of carbon from these three species is actually made available for heterotrophic productio n from microbial biomass. Of special interest here is the comparison of Typha and Cladium . As mentioned previously, Typha has been observed encroaching upon historically Cladium dominated areas of the Northern Everglad es. In this model it appear s that per kg of plant tissue, Typha contributes more potentially leachable carbon, more carbon to export, more bioavailable carbon, and supports the growth of more bacterial biomass than does Cladium . While studies have shown the detrimental ecological effects of Typha invasion, none have determined the effects of vegetation shifts from Cladium to Typha on carbon dynamics (Weisner and Miao, 2004). At this time, the effects of this change in DOM source are not fully realized, but this study s uggests that an almost 2 fold increase in leachable carbon, twice the export, and almost twice the microbial biomass production may negatively affect the DOM dynamics of the Florida Everglades. Summary and Conclusions Investigation of the decom position of DOM derived from the nine common species of wetland vegetation from the Everglades re vealed that the decomposition rates for the nine species tested were not the same. The results suggest that DOM lability or bioavailability is variable based upon the spec ies of origin. Likewise, bacterial growth efficiency measured on DOM samples suggests a species dependant factor as well, with secondary plant metabolites being a possible fact or in the inhibition of bacterial growth. The model presented here for these vegeta tion types suggests loss of carbon from DOM pool is dependant upon the amount of pheno lic compounds present and the amount of carbon in the relatively low molecular weight region of 1-3 KDa, has been shown to account for 86% of the variability in the obser ved decomposition of this material. It has
84 been shown here that these nine species of wetland plants do have significantly different amounts of phenolic compounds and low mo lecular weight compounds in the DOM derived from their tissues. With this in mi nd, it is important to recognize the large range of potential these species have to contribu te to the overall DOM pool and DOM dynamics in wetlands where they are found. Further, in areas such as the water conservation areas in the northern Everglades where large scal e shifts in vegetation community structure have been observed, the potential change in trophic dynamics stemming from altered degradability of DOM inputs to the system c ould be very significant. Because of the energetic value of DOM to wetland and aquatic systems in general with respect to trophic dynamics, a shift from one DOM source to anot her with greater or lesser bioavailability could alter functioning of bacterial communities in downstream systems and likewise alter nutrient dynamics as well. At this tim e the consequences of vegetation community alteration are not fully realized and therefor e it is recommended that the relationships between vegetation community structure a nd DOM cycling be further investigated.
85 10010Kilometers N E W S #RB #HL #WCA-2B MDLS ENP BCNP-S BCNP-N WCA-3A #WCA-3B WCA-1 WCA-2A #MSIR 1:1,200,000Florida #Study Site Figure 3-1. Map of areas where plant sa mples were obtained in 2002 and 2003.Water Conservation Area (WCA-1) is also known as the Loxahatchee National Wildlife Refuge. Big Cypress National Preserve (BCNP) is separated into northern and southern halves, BCNPN and BCNP-S respectively. The Miccosukee and Seminole Indian Reserv ation is denoted by MSIR and the Everglades National Park by ENP.
86 10 30 50 70 90 110 02468 Time (days)DOC mg C L-1 Eleo Typ Clad Figure 3-2. Results of Phase 1 decompos ition as loss of carbon over 8 days for Eleocharis , Typha , and Cladium denoted as (Eleo), (Typ), and (Clad) respectively. Error bars indi cate +/1 standard deviation. 0 10 20 30 40 50 60 70 816243240 Time (days)DOC mg C L-1 Eleo Typ Clad Figure 3-3. Results of Phase 2 decomposition as loss of carbon over 32 more days for Eleocharis , Typha , and Cladium denoted as (Eleo), (Typ), and (Clad) respectively. Error bars indi cate +/1 standard deviation.
87 10 30 50 70 90 110 02468 Time (days)DOC mg C L-1 Spar Thal Nuph Figure 3-4. Results of Phase 1 decompos ition as loss of carbon over 8 days for Spartina , Thalia , and Nuphar denoted as (Spar), (Thal), and (Nuph) respectively. Error bars indicate +/1 standard deviation. 0 10 20 30 40 50 60 70 816243240 Time (days)DOC mg C L-1 Spar Thal Nuph Figure 3-5. Results of Phase 2 decompositi on as loss of carbon over 32 more days for Spartina , Thalia , and Nuphar denoted as (Spar), (Thal), and (Nuph) respectively. Error bars indi cate +/1 standard deviation.
88 10 30 50 70 90 110 02468 Time (days)DOC mg C/L Nym Pan Tax Gluc Figure 3-6. Results of Phase 1 decompos ition as loss of carbon over 8 days for Nymphea , Panicum , Taxodium , and Glucose denoted as (Nym), (Pan), (Tax) and (Gluc) respectively. Error bars indicate +/1 standard deviation. 0 10 20 30 40 50 60 70 816243240 Time (days)DOC mg C L-1 Nym Pan Tax Gluc Figure 3-7. Results of Phase 2 decomposition as loss of carbon over 32 more days for Nymphea , Panicum , Taxodium , and Glucose denoted as (Nym), (Pan), (Tax) and (Gluc) respectively. Error bars indicate +/1 standard deviation.
89 Figure 3-8. Conceptual diagra m of biotic degradation and transformation pathways of DOC derived from senescent plant tissue s through abiotic leaching process. Letters A-E indicate va riable values of DOC moving from one pool to another. D C B A E Particulate Organic Matter 1 Kg Leachable DOC DOC Pool Biosynthesis Export Respiration Bioavailable DOC Microbial Community Microbial Biomass
90 Table 3-1. Nutrient analysis of leachate samples. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Total carbon, total nitrogen, a nd total phosphorus are represented by TC, TN, and TP respectively. Ratio s of carbon to nitrogen, carbon to phosphorus, and nitrogen to phosphorus are indicated by C:N, C:P, and N:P. Soluble reactive phosphorus is deno ted by SRP, ammonium nitrogen by NH4(N), and nitrate nitrogen by NO3(N). Units are in grams of each analyte leached per kg of tissue. Species TC g kg-1 TN g kg-1 TP g kg-1 C:N C:P N:P SRP g kg-1 NH4(N) g kg-1 NO3(N) g kg-1 Eleo 40.2 2.39 0.16 16.8 250 15 0.052 0.29 0.012 Typ 34.2 1.25 0.12 27.3 294 11 0.014 ND 0.030 Clad 21.5 0.89 0.11 24.1 196 8.11 0.086 ND 0.006 Spar 38.3 3.81 0.69 10.1 55 5.53 0.089 1.38 0.011 Thal 87.3 6.19 0.65 14.1 134 9.50 0.143 0.52 0.026 Nuph 179 6.55 1.11 27.3 161 5.90 0.771 0.02 0.161 Nym 219 5.53 0.88 39.7 250 6.33 0.109 ND 0.065 Pan 48.2 2.88 1.02 16.7 47 2.82 0.772 0.22 0.014 Tax 105 2.04 1.20 51.5 88 1.70 0.990 ND 0.032 Table 3-2. Nutrient analysis of standard ized DOM sample prior to experimentation. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Total carbon, total nitrogen, and total phosphorus are represented by TC , TN, and TP respectively. Soluble reactive phosphorus is denoted by SRP, ammonium nitrogen by NH4(N), and nitrate nitrogen by NO3(N). Units are in mg L-1. Species ------------TC ------------TN ------------TP ---mg L-1-SRP ------------NH4-N ------------NO3 Â–N ------------Eleo 100 5.95 0.40 0.130 0.720 0.03 Typ 100 3.66 0.34 0.040 0.000 0.09 Clad 100 1.94 0.51 0.400 0.000 0.03 Spar 100 9.95 2.99 1.780 3.610 0.03 Thal 100 7.09 2.24 1.640 0.590 0.03 Nuph 100 3.66 0.62 0.430 0.010 0.09 Nym 100 2.52 0.51 0.050 0.000 0.03 Pan 100 5.99 2.12 1.600 0.450 0.03 Tax 100 1.94 1.14 0.940 0.000 0.03
91 Table 3-3. Characterizati on analysis of standardized DOM sample prior to experimentation. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Carbohydrate and phenolics content are presented as percen t of total organic carbon in sample. Estimated protein is presented as mg protein L-1. Molecular weight fractions (MWF) are presented as percent of total carbon in sample. Species Carbohydrate (%carbon) Phenolics (%carbon) Estimated Protein MWF <1 KDa MWF 13 KDa MWF 310 KDa MWF >10 KDa Eleo 40.1 4.7 32.4 4.3 15.8 28.2 51.6 Typ 48.9 37.4 22.2 15.4 12.0 12.3 60.2 Clad 31.4 11.3 11.9 20.4 2.9 13.7 62.9 Spar 28.7 7.5 39.3 29.4 2.8 10.0 57.7 Thal 44.4 11.7 40.4 32.2 2.0 2.1 63.6 Nuph 31.3 30.4 22.2 6.5 30.3 12.3 50.8 Nym 37.9 27 15.5 16.1 8.4 18.0 57.3 Pan 35.5 4.7 34.4 15.1 15.0 8.9 60.9 Tax 42.8 7.5 11.9 19.2 13.8 7.4 59.6 Control 100 NA NA 100 NA NA NA Table 3-4. Results of standardized DOM sa mple decomposition experiments. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Eight day (Phase 1) and 40 day (Phase 2) carbon loss is presented as percent of total carbon present at beginning of the experiment. Bacterial growth efficien cy (BGE) is presen ted as percent of microbially utilized carbon in corporated into biomass. Species Phase 1C loss (%) Phase 2 C loss (%) Decomposition Rate (K day-1) BGE (%) Eleo 64.7 68.7 0.048 18 Typ 43.8 57.3 0.032 46 Clad 68.1 70.1 0.053 33 Spar 79.4 79.0 0.073 13 Thal 73.9 75.9 0.070 37 Nuph 44.7 54.6 0.034 21 Nym 64.1 69.8 0.079 25 Pan 61.5 60.9 0.047 14 Tax 58.5 64.2 0.046 12 Control 28.0 75.3 0.030 7
92 Table 3-5. Calculated values of carbon transfer for five pathways (A-E) of decomposition and transformation of dissolved organic carbon (DOC) derived from senescent plant tissue. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). All values are in grams of carbon in the form of DOC per kg of tissue leached. Species Potentially Leachable Carbon (A) Bioavailable DOC (C) Microbial Biomass Production (E) Microbial Respiration (D) Export (B) Eleo 40.2 27.7 5.0 22.7 12.6 Typ 34.3 19.6 8.9 10.7 14.6 Clad 21.5 15.1 5.0 10.1 6.4 Spar 38.4 29.3 3.9 25.5 9.0 Thal 87.4 52.8 19.7 33.1 34.6 Nuph 179 97.9 20.3 77.6 81.6 Nym 220 153 38.3 115 66.3 Pan 48.2 27.2 3.9 23.3 21.0 Tax 105 67.6 8.1 59.6 37.7
93 CHAPTER 4 PHOTOLYTIC MINERALIZATION AND PHOTO-BLEACHING OF DISSOLVED ORGANIC MATTER (DOM) DERIVED FR OM VEGETATION TYPES OF THE FLORIDA EVERGLADES Introduction The scientific community has long been aw are of the effects of solar radiation on aquatic ecosystems. The historical viewpoint was one of the biological effects of light on such processes as photosynthesis, photot axis, and biorhythms by photosynthetically active radiation (PAR 400-750 nm). This unde rstanding included biological effects of ultraviolet radiation (UV 200-400 nm) as it pe rtains to the damaging and inhibitory effects it has on aquatic bi ota (Zafiriou, 2002; Khan and Wetzel, 1999). However, relatively recent discoveries of the abiotic affects of solar ra diation, particularly UV, have greatly broadened our understand ing of the functional role of light in the chemistry and energy dynamics of aquatic ecosystems. No where is this importa nce of light better demonstrated than in the dynamics of DOM (Findlay and Sinsabaugh, 2003). The DOM in aquatic systems can have great and broad reaching effects on heterotrophic production and carbon cycling, metal and nutrient availability, water chemistry, and water column light attenuati on (Wetzel, 1992; Scully and Lean, 1994). Recent investigations have revealed that photolysis, the lysing of carbon to carbon bonds in DOM, and photo-bleaching, the destruc tion of chromophores within DOM, can influence the lability of DOM in aquatic syst ems (Scully et al., 2004; Pullin et al., 2004; Whitehead et al., 2000).
94 While incomplete photolysis is comm on in aquatic ecosystems, complete photolysis (direct mineraliza tion of DOM to CO or CO2) is also possible (Valentine and Zepp, 1993; Zou and Jones, 1997). Th is result is due in part to the fact that most natural DOM in aquatic systems is humic in origi n. Humic materials are often composed of organic functional groups such as phenolics, aromatics, a nd carboxylics that have double C bonds (unsaturated) (Thurman, 1985). Thes e carbonyl structures are chromophoric in nature and most are susceptible to UV photolysis and photobleaching (Zepp, 1988). Because of the highly variable nature of humic material composition, humic portions of DOM can react to photolysis at different levels (Clair and Sayer, 1997), however, aliphatic moieties of DOM are not known to react substantially to UV photodegradation due to their low occurrence of conjugated (double) bonds. Although the full ecological significance of photochemical reactions in natural waters has not been fully realized, much re search has been conducte d to investigate the possible effects of photolysis on aquatic ecos ystems (Mopper et al., 1991; Bushaw et al.; 1996). Because much of DOM in aquatic ecosy stems is humic in nature, it is often recalcitrant to microbial util ization, and therefore, photolysis could potentially provide a mechanism to breakdown the recalcitrant porti ons of natural DOM a nd make it available once again for bacterial utiliz ation. This would allow DOM th at is largely sequestered from aquatic ecosystem energetics to be mobilized. Wetzel et al. (1995) found that exposure to UV light produced numerous smalle r molecular weight or ganics from humic and fulvic acid fractions of DOM and that the small organic fraction stimulated bacterial growth. Similar results have been found in st udies of rivers, wetlands , lakes, and marine systems (Engelhaupt et al., 2003; Reit ner et al., 1997; Gellar, 1986).
95 As well as potentially increasing microbial access to refractory DOM, the nutrients bound within DOM can also be accessed. Wa ng et al. (2000) repor ted the production of up to 1.9 uM ammonium per hour by photolysis of natural DOM from various sources. These findings are in accordance with Bush aw et al. (1996) who reported ammonium production of 86 uM per day from photochemical decomposition of whole DOM samples from Okeefenokee Swamp, GA. Photolytic production of a mmonium has great implications especially in n itrogen poor aquatic systems. Further, the release of organically bound nitrogen and phosphorus can be a significant pathway in oligotrophic systems (Wetzel, 2001). These findings have im portant implications for nutrient cycling in aquatic systems, especially wetland syst ems such the oligotrophic Everglades, which have high DOM loads, and relativ ely tight nutrient cycles. Previous work has shown that DOM deri ved from different species of wetland vegetation in the Everglades to be chemica lly and physically different, and that biotic degradation of this vegetation derived DOM was also somewhat species dependant which suggest that changes in plant communities can alter the DOM cycle (see Chapter 2). The purpose of this study was to evaluate the phot olytic susceptibility of DOM derived from different species of wetland vegetation comm only found in the Everglades wetlands to photo-bleaching and direct minera lization and determine if this too is species dependant and if so, what effects this might have on the DOM cycling. A secondary goal of this research was to link observed photolytic mineralization and photo-bleaching to characteristics of DOM thought to be important to photolytic decomposition.
96 Materials and Methods Plant Species and Sample Locations To investigate the relationships between plant derived organic DOM and photolysis in wetlands, selected species of dominant we tland vegetation were chosen and the freshly senescent standing dead biomass collected fr om sites within the Florida Everglades. Specific vegetation types chosen for this st udy were common to the greater Everglades wetland ecosystem and included Typha domingensis Pers., Cladium jamaicense Crantz, Panicum hemitomon , Spartina bakerii , Eleocharis interstincta , Thalia geniculata , Taxodium disticum , Nymphea odorata , and Nuphar luteum . All plant material was collected above the current wa ter level to avoid possible pr eviously leached material. Stems, leaves, and inflorescence were collect ed and combined. Collection of senescent Nymphea and Nuphar that was not in the water column was accomplished by locating sites of water drawdown induced senescence wh ere the plant material was exposed above the saturated soil surface. In the case of Taxodium , the only woody species collected, only leaves were used in leachate experiments. Sites of collection of these plants in cluded the Loxahatchee National Wildlife Refuge (WCA-1), Water Conservation Area (WCA) 2A, 2B, 3A, 3B , the Miccosukee and Seminole Indian Reservations, Big Cypre ss National Preserve, and the Everglades National Park (Figure 4-1). Collection of samples occurred during the months of November Â– December 2002 and in 2003 duri ng months of natural senescence. Plant tissue collection consisted of harvest of recently senesced above ground biomass. For each species, a minimum of 10 samples were collected, returned to the laboratory and immediately dried at 55 ÂºC. The temperature of 55Âº C was chosen to avoid artificial lignification of the tissues observed with higher drying temper atures (Roberts and
97 Rowland 1998). Tissue samples were subseque ntly ground in a large Wiley mill to 1mm and thoroughly mixed to produce 3 aggregate samples. Dissolved Organic Matter Production and Characterization The production and characterization of DOM for these experiments was described in detail previously (Chapter 2). Briefly, to make DOM samples, ground detrital material was leached using distilled de ionized water for three hours an d then sequentially filtered, ending with 0.2 Âµm. Samples of DOM were then diluted to appropriate concentrations for experiments and analysis. Molecular size fract ions of leachates were determined by ultra filtration techniques utilizing Amicon 8500 cont inuously stirred ultra filtration cells. Millipore regenerated cellulose filter memb ranes (76 mm diameter) of nominal molecular weight limit 1000 (YM1), 3000 (YM3), and 10,000 (YM10) Daltons were used and calculations of molecular weight fractions were made according to Tadanier et al. (2000). Total Carbon (TC) of leachates was determined on a Shimadzu TOC-5050. Total phenolic content, content of both mono a nd poly phenolics compounds, was determined following the colorimetric method of Price and Butler (1977). Experimental Design DOM samples were exposed to UV radiation under normal environmental conditions in natural sunlight (UVA (320-400nm)+UVB(280-320nm)+PAR(400700nm)) to determine the extent of direct photolytic mineralization of DOM and the relative level of susceptibility of bulk DOM to photo-bleaching. Samples were placed in UV exposure tubes made from optically pure quartz glass tubing (10mm diameter 1mm wall thickness) in 15cm segments with a f ood grade inert silicone stopper (size 000) in each end to retain the sample. Every effort was made to minimize the size of the bubble captured in the UV exposure containers when placing the stopper. Samples were bubbled
98 with oxygen prior to inclusion within the quartz tubing cells and sodium azide was added as an antimicrobial agent (1 mg L-1). Samples from each species were divided into 21 tubes (7 sample dates x 3 experimental rep licates x 9 species for a total of 189 sample tubes) and incubated from sunrise to sunset in a shallow plastic bin (painted flat black) with a constant flow of water to keep the te mperature at 17ÂºC (Figure 4-2). Experiments were conducted on an elevated table in an open field located at the University of Georgia Marine Institute on Sapelo Island, Georgi a (31.39 N, 81.28 W) during the month of September 2004. In the experiment, DOM samples were exposed to 9 days of natural sunlight, and each day, 3 sample tubes of each species were removed from the exposure pool. Each sample tube held a pproximately 12 ml of sample at an initial c oncentration of 100 mg C L-1 as DOM. Blanks, made up of DDI water and sodium azide were also removed each day. Dark blanks of each speci es DOM were also incubated in the pool at the same time. These dark samples were kept from UV exposure by wrapping tubes in aluminum foil. Natural Sunlight Exposure Cumulative exposure was measured on site everyday with a Solar Light Co. (Glenside, Pennsylvania) model PMA 2100 dua l channel recording light meter equipped with a model PMA 2107 UVA+UVB light sensor and a model PMA 2131 PAR light sensor programmed to record light m easurements in both the UVA+UVB and PAR ranges every minute. At the end of each day of exposure, cumulative UV and PAR exposure was tabulated and samples were remove d and stored at 4ÂºC until returned to the laboratory for analysis of TOC and loss of chromophoric DOM.
99 Quantification of Photo-Min eralization and Photobleaching Photolysis or photo-mineralization of DOM was quantified by analysis of DOC at each step of the experiment. Loss of carbon as DOC (minus changes in dark controls) was interpreted as direct photo-mineralization to CO2. DOC was plotted against total exposure to determine photo-mineralization re quirements and temporal scales for DOM samples. Photodegradation rates were calculate d by linearizing the data from the first 3-4 days of exposure and taking the slope of th e line fit to the data as the decomposition constant (K) per unit time (day). Photo-bleaching was qualitatively docume nted by measuring light absorbance at the diagnostic wavelength 325 nm before and after the exposure experiment. This wavelength was chosen after repeated scan s of samples from 300 to 400 nm, and an optical activity maxima was found at 325, co rresponding to similar areas of optical activity for chromophoric dissolved organic matter (CDOM) in other studies (Boehme and Coble, 2000; Baker and Spencer, 2004). The specific absorbance of chromophoric dissolved organic matter (aCDOM) was calculated as the absorbance value divided by the concentration of dissolved or ganic carbon (DOC) as follows: aCDOM = a325 /[DOC] mg L-1 Specific absorbances at time 0 and after 9 days exposure were then subtracted from each other to calculate change in aCDOM. This value for change was then compared to the initial measured characteristics of the samples DOM to investigate relationships between DOM characteristics and the amount of obser ved photo-bleaching. A similar method was used to qualitatively examine the loss of ar omatic or hydrophobic structures in the DOM samples. The wavelength employed was 254 nm, commonly called sp ecific ultraviolet absorbance (SUVA), which is derived in th e same manner as that for CDOM at 325 nm
100 (Stewart and Wetzel, 1981; D illing and Kaiser, 2000). Abso rption at this wavelength is known to directly relate to concentration of aromatic and hydr ophobic moieties in the DOM sample (Dilling and Kaiser, 2000). The absorbance is divided by the concentration of carbon in the sample to normalize and make samples comparable. Results and Discussion Characterization of Dissolved Organic Matter Analysis of DOM characteristics yielded si gnificantly different levels of phenolic compounds and molecular weight fractions, especially in the <1 KDa and 1-3 KDa molecular cut offs, of the DOM samples (Tab le 4-1). Phenolic compounds presented as percent of carbon in the DOM pool ranged from 5 to 37 % and are in order with levels of polyphenols found in the DOM from estuarine plant leachates by Scully et al. 2004. Molecular weight fractions were significantly different as well, among species, with the exception of the >10 KDa fraction which only showed two significantly different groupings of species. The specific UV abso rption of DOM at 254 nm (SUVA) (Dilling and Kaiser, 2000), revealed a ra nge of values after normaliza tion to carbon content before exposure to UV light. These values indicat e a wide range of aromatic and hydrophobic contents for the nine vegetation types. Values ranged for 0.0116 ( Taxodium ) to 0.0333 ( Nymphea ) almost a three fold difference. Perc ent of carbon in the form of phenolic compounds explained 81% of the variability in the SUVA measurements prior to UV exposure (r2=0.81 P<0.001). Likewise, percent car bon in phenolic compounds explained 78% of the variability in the absorbance values of CDOM found in the DOM samples, which had a similar range of values as SUVA measurements, 0.0078 ( Taxodium ) to 0.0213 ( Typha ).
101 The relationship of percent carbon as phe nolic compounds to the absorbances at 254 nm and 325 nm indicates that ther e are other hydrophobic, aromatic, and chromophoric organic structures in the DOM sa mples that may affect the results of the photolytic decomposition experiments outside of the compounds quantified in the phenolic compound class. Photolytic Decomposition and Mineralization Experimental exposure of DOM derived from the nine different species of wetland vegetation revealed that photol ytic mineralization can play a significant role on the fate of this material on a short time scale (Figure 4-3). These results suggest that although the exposure was run for nine consecutive days, most of the direct photolysis occurred within a relatively short time span. This is evid enced by the reactions of most species that showed little or no photolysis activity after 3-5 days. The exception was Thalia which continued to lose carbon from direct photolys is until the end of the experimental phase. A similar trend was observed when car bon loss was related to cumulative UV dose instead of days of exposure (Figure 4-4). Th ese results indicate th at the total amount of carbon lost to direct photolysis is variab le based upon species and the molecular characteristics there in. In support of th e hypothesis that DOM de rived from different plant species would react differe ntly to photolytic degradation, these results indicate that there were significant differences among the sp ecies tested with re spect to total carbon mineralized within 9 days of UV exposure. Scully et al. (2004) found similar results when comparing photolytic decomposition of DOM derived from periphyton, sawgrass, mangroves, and sea grasses in the southern Everglades estu arine ecotone. In further support of these findings, Farjalla et al. (2001) reported a significant difference in photolysis between DOM derived from Phragmites and Hydrocaris . While there were no
102 significant differences found between all sp ecies examined, there were significant differences among groupings of species. DOM from Thalia was significantly less affected by UV exposure than any ot her species tested. Likewise, Spartina was affected to a greater extent than any ot her species in terms of direct photolysis and organic carbon mineralization. Cladium and Typha were less affected by photolysis than Nuphar and Nymphea , although not significantly; however, these f our species were significantly more affected than Eleocharis , Panicum and Taxodium (ANOVA, =0.05, P<0.01), DuncanÂ’s Multiple Range post hoc). Significant differe nces among these groupings suggests, at the very least, that DOM derived from diffe rent species of wetland vegetation react differently to UV induced mine ralization of organic carbon. Comparison of total loss of carbon to direct photolysis to phenolic content and molecular weight fractions (MWF) revealed a lack of any significant relationship. It was expected that the phenolic c ontent would correlate well with the amount of carbon lost, however no such relationship was found (r2=0.28, P<0.21). This surprising result suggests that there are other structures w ithin the DOM that are being photolytically mineralized. Similarly, phenolic content did not describe all th e variability in the initial DOM samples with respect to specific abso rbance at 254nm, which is indicative of aromatic and hydrophobic compounds. The re sults suggest other moieties, likely unsaturated, aromatic compounds, or possibl y LMW organic acids (as reported by Pullin et al., 2004) were responsible fo r the resulting direct photoly tic mineralization of organic carbon in these DOM samples. Further, th ese compounds make up a significant portion of carbon compounds found in the DOM derived fr om the species tested as indicated by the percentages of organic carbon lost to dire ct photolysis, suggesti ng that more rigorous
103 characterization of the arom atic portions of DOM would be necessary to predict susceptibility of DOM to direct photolysis. Lack of any significant relationship betw een mineralization of organic carbon and molecular weights was also surprising a nd suggests that the aromatic and hydrophobic materials largely responsible for reactivity to UV light are likely mixed throughout the molecular weight fractions and thus no particul ar fraction is responsible for a majority of photolytic mineralization. Scully et al. (2004) re ported results to the contrary in that most of the photolytic activity occu rred in what would be the 3-10 KDa fraction in this experiment. A possible explan ation could be that, unlike Sc ully et al. (2004) the high concentrations of DOM used in this experiment could have promoted the incorporation of aromatic moieties into other fractions such as proteins (>10 KDa). Also, the possibility of LMW organic acids being somewhat suscep tible to direct photol ysis (as reported by Pullin et al., 2004) could explai n some of the lack of relationship between molecular weight and mineralization of organic carbon. Rates of photolytic decomposition were f ound to be independent of the amount of the total carbon mineralized and the phenolic content in the DOM samples. Photolytic decomposition rates were found to range from a low of 0.0226 day-1 ( Thalia ) to a high of 0.0848 day-1 ( Spartina ) (Table 4-2). This broad range suggests underlying factors are responsible for the different rates of minera lization. The incorporat ion of photosensitive aromatics into other fracti ons of organic carbon may al so serve here to explain differences observed in apparent rates of photolytic decomposition. Phenolic compounds binding with proteins have been previously reported in DOM studies (Hagerman et al., 1998; Northup et al., 1998) and could account for the lack of relationship between
104 phenolics and photolysis in some samples. Further, incorporation of photosensitive compounds into other organic structures, eff ectively protecting them from photolysis is likely not limited to phenolics. This incorpora tion could also act to slow the process of photolysis as evidenced by Boavida and Wet zel (1998), who reported the inactivation of phosphatase enzymes by humic substances a nd the subsequent re activation of these enzymes after exposure to UV light. The expe riment by Boavida and Wetzel (1998) also suggested that extraction of humic comple xes from the DOM pool decreased the amount of UV exposure necessary to reactivate enzy mes. These findings suggest the bulk DOM also plays some role in the shading or protection of photosensitive compounds. Of special interest is the degree to which DOM produced from these different species is susceptible to photolytic mi neralization. In this experiment, Spartina lost the greatest amount of carbon to di rect photolysis (34%) althoug h it contained relatively low amounts of carbon in phenolic compounds. C onversely, the lowest amount of carbon lost to direct photolysis was from Thalia (13%) which corresponded to a relatively greater amount of phenolic compound content than Spartina . These results suggest again that there are other organic structures outside of the polyphenols measured here that were responsible for direct minera lization of organic carbon. This was supported by the change in specific absorbance at 254 nm after the UV exposure (Figure 4-5) with respect to phenolic content. Approximately 25% of th e variability observed in the change in absorbance at 254 nm was unaccounted for by pheno lic content. Hence it is reasonable to conclude that phenolic content, while a relia ble predictor of change in aromatic and hydrophobic content in DOM exposed to UV radi ation, is not comprehensive enough to use as a predictor of direct photolysis.
105 These results suggest that aromatic and hydrophobic content of DOM is only a predictor of photolysis potential. The re lationship between the actual loss of organic carbon to direct mineralization and form of organic carbon in the DOM pool is one of complexity requiring more rigorous char acterization of the aromatic and hydrophobic portions of the DOM pool to understand. Photo-Bleaching Specific absorbance at 325 nm (aCDOM) was found to decrease in all samples from the initial values of DOM sa mples prior to UV light expos ure (Table 4-2) indicating photo decomposition of chromophoric structures in the DOM samples. Because phenolic structures are so often chromophoric in na ture, a comparison of change in specific absorbance at 325 nm was made with the measured amount of phenolic compounds in the initial DOM samples (Figure 4-6). Before exposure, phenolic compound content was found to correlate we ll with initial aCDOM, with a linear regression r2=0.84 (P<0.001). This correlation held through the exposures and resulted in an r2=0.65 (P<0.001) after the exposure experiment. These results sugge st that much of the DOM in phenolic compound form underwent some level of photo-bleaching; however, the amount of chromophoric materials that were completely mineralized is unknown. Likely, some of the CDOM was directly mineralized and some partially decomposed to non-aromatic or chromophoric structures as evidenced by the l ack of any relationship between the percent change in aCDOM and the percent loss of carbon to direct photolytic mineralization. Decrease of aCDOM ranged from a low of 13% ( Thalia ) to a high value of 55% ( Taxodium ). Of interest were the species Taxodium , Typha , and Nuphar which exhibited between 40-55% losses of chro mophoric activity through exposure to UV. These species demonstrated that loss of chromophoric activity, or aCDOM, did not correspond to a loss of
106 organic carbon from photo-mineralization. This suggests that while a large percentage of chromophores were deactivated, they were likely decomposed to smaller, more bioavailable, compounds versus being direc tly mineralized. Many studies have shown that UV exposure enhances bioavailability of DOM (Keiber et al., 1990; Miller and Moran, 1997; Wetzel et al., 1995; Hapeman et al., 1998; Frimmel, 1998; Lehto et al., 2003). This photolytic decomposition is not necessari ly reflected in direct mineralization. Vahatalo and Wetzel (2004) reported that a 70 day exposure to UV radiation decomposed close to 100% of the CDOM but only 41% DOC was lost, suggesting that CDOM deactivation does not necessarily resu lt in loss of carbon from the DOM pool, but rather the structure was altered and became mo re or less bioavailable in the process. Wetzel et al. (1995) repo rted subtle changes in 13C NMR spectra of UV exposed DOM but observed a significant increase in bacteria l activity following exposure. These results suggest again that subtle changes to CDOM structure can have significant effects on DOM dynamics. While photo-mineralization ceased after 3-5 days in all but one species, there was evidence that photo-bleaching could continue on a longer scale. Th e highest amount of chromophore loss was 55%, translating to 45% of the original CDOM s till existed intact. Similar results have been reported by Del Castillo et al. (199 9) who identified chromophores resistant to photo-bleaching. Some regeneration of chromophores has been reported recently, implying that these materials may be recalcitrant to photodecomposition and photo-degradation (Boe hme and Coble, 2000). In this study, nine plant species tested for susceptibili ty to photo-bleaching all exhibited results indicating that there was loss of chromophoric activity after exposure to UV radiation and
107 there were considerable differences in the re activities of different species to UV radiation exposure. The extent of the transformation could not be determined by this experiment, nor could the amount of direct photolysis due to destruction of chromophores. Lack of relationship between phenolics and direct photolysis coupled w ith the correlation between phenolic compounds and photo-bleaching indicates that the two processes are likely affecting different compounds in the DOM samples. As in the direct photolysis experiment, a more rigorous characterization of the aromatic and hydrophobic portions of the DOM pool is required to more accurately predict the effects photo-bleaching will have on a particular kind of DOM. Ecological Significance Globally, wetlands are consta ntly being encroached upon due to development and human usage. Due to this anthropogenic influence, many wetlands are impacted by changes in hydrology and nutrien t inputs. This often result s in changes to plant and animal communities and the ecological function of the wetlands themselves. As a case study, in the Everglades system, major distur bances in nutrient cycling and hydrology have occurred and one of the resulting observa tions has been large areas of destabilized vegetation communities. Changes in plant communities could have significant impacts on the cycling of nutrients, especially DOM. Previous work has es tablished significant differences in the physical and chemical pr operties of DOM derived from the nine species of wetland vegetation tested here (Cha pter 2). Significant differences in the cycling of carbon in the form of DOM de rived from these plants have also been established (Chapter 3). Therefore, the differences among species tested here with respect to their susceptib ility to photolysis, either direct or indirect, suggest that shifts in
108 vegetation communities on a spatial scale coul d alter DOM dynamics and alter the level of importance of UV radiation in DOC cycling. The ecological significance of photolysis a nd photo-bleaching can be viewed at two levels for plant derived DOM in the Everglades . The first being the effects of photolysis on DOM dynamics, especially the microbial l oop, which represents the trophic exchange of energy, a major driving force for any a quatic ecosystem (Covert and Moran, 2001). There are numerous studies reporting increa sed bacterial activity associated with UV exposure of DOM, presumably increasing the amount of DOC removed from the DOM pool (Findlay and Sinsabaugh, 2003). Of si milar importance woul d be the increased availability of nutrients occluded in more recalcitrant portions of DOM that are made bioavailable through photolytic transformation, a result that is also established in the recent literature (Zepp, 2003). In this study microbial decomposition was not coupled with the photolysis experiments, so the net effect on the microbial loop is uncertain. Although negative effects on bi oavailability of DOM due to UV exposure have been documented (Benner and Biddanda, 1998; Tranvi ck and Kokalj, 1998; Anasio et al., 1999; Tranvik and Bertilsson, 2001) the moun ting evidence found in the literature suggests that photolysis will likely enhan ce microbial utilization of DOM due to the nature and source of the materi al being studied here. This study has shown that from 13 to 34% of the carbon in the DOM samples could be lost due to direct photolysis. Of the remaining portion, some level of susceptibil ity to photo-bleaching has been shown to occur, and would likely result in enhanced bioavailability. A second level to evaluate the significance of photo-degr adation of DOM is on the level of indirect effects. How will change s to the dominant UV exposure response of the
109 DOM pool affect the system? Again the litera ture supports the role of CDOM in aquatic systems as one of protection of biota from harmful affects of UV radiation as well as controlling the penetration of light into th e water column (Schnider and Curtis, 1997; Osburn et al., 2001). Changes in either of these functional roles due to changes in dominant DOM source could ha ve wide ranging and possi ble negative effects on the Everglades ecosystem. Because photolysis is much more important in slow moving waters, it likely plays a significant role in the organic carbon dynamics of a system such as the Everglades (Clair and Sayer, 1997). Summary and Conclusions The results of this study suggest UV i nduced photolysis is regulated by the variability in chemical composition of DOM an d its parent material. Direct photolysis, the mineralization of DOM to CO2, accounts for approximately 13-34% of carbon. Large changes in SUVA and aCDOM from 28-71% and 13-55% , respectively, indicate that aromatic, hydrophobic, and chromophoric structur es of DOM can be dramatically altered or decomposed through UV exposure. Further, these changes were found to be related to phenolic compound content, which varies significantly among species. However, results suggest that other photo sens itive compounds, such as non-phenolic based organics, may play a significant role in the photolysis and photo-bleaching processes. Although bacterial response to UV exposure was not meas ured here, it was thought that it will have a stimulatory effect on the microbial utiliza tion of DOM, due to the sources and chemical nature of the material examined here. Thr ough recent research it is clear that photolysis, either direct or indirect, ha s broad reaching implications for the organic matter dynamics and biological processes in aquatic ecosystems. This study has demonstrated that significant differences do exist among species in their reactio ns to UV exposure and it is
110 based upon the characteristics of the plants th emselves, therefore, it is concluded that shifts in plant communities in the Evergl ades, resulting in cha nges in the dominant sources of DOM to the system, could have si gnificant effects to the cycling of DOM in this system, which could also result in alteration to nutrient cycling and biological processes.
111 10010Kilometers N E W S #RB #HL #WCA-2B MDLS ENP BCNP-S BCNP-N WCA-3A #WCA-3B WCA-1 WCA-2A #MSIR 1:1,200,000Florida #Study Site Figure 4-1.Map of areas where plant sa mples were obtained in 2002 and 2003.Water Conservation Area (WCA-1) is also known as the Loxahatchee National Wildlife Refuge. Big Cypress National Pr eserve (BCNP) is separated into northern and southern halves, BCNPN and BCNP-S respectively. The Miccosukee and Seminole Indian Reserv ation is denoted by MSIR and the Everglades National Park by ENP.
112 = quartz glass exposure tubes View from above photolysis experiment container Side view of experimental container Coolant water OUT Coolant water IN = quartz glass exposure tubes = water level inside exposure container = quartz glass exposure tubes View from above photolysis experiment container View from above photolysis experiment container Side view of experimental container Coolant water OUT Coolant water IN = quartz glass exposure tubes = water level inside exposure container Side view of experimental container Coolant water OUT Side view of experimental container Coolant water OUT Coolant water IN = quartz glass exposure tubes = water level inside exposure container Coolant water IN = quartz glass exposure tubes = water level inside exposure container = quartz glass exposure tubes = water level inside exposure container Figure 4-2. Diagram of experiment al photolysis exposure chamber.
113 60 70 80 90 100DOC mg C L-1 Eleo Typ Clad 60 70 80 90 100DOC mg C L-1 Spar Thal Nuph 60 70 80 90 100 0369 UV Exposure (Days)DOCmg C L-1 Nym Pan Tax Figure 4-3 Loss of dissolved organic carbon over 9 days of exposure to naturalsunlight. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax).
114 60 70 80 90 100DOC mg C L-1 Eleo Typ Clad 60 70 80 90 100 0150300450600750 UV Exposure (j cm-2)DOC mg C L-1 Nym Pan Tax 60 70 80 90 100DOC mg C L-1 Spar Thal Nuph Figure 4-4. Loss of dissolved organi c carbon based upon quantified UVA and UVB exposure. Error bars represent +/std. error. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax).
115 y = 0.0002x + 0.0056 R2 = 0.74 0 0.004 0.008 0.012 0.016 010203040 Percent Carbon in Phenolic CompoundsChange in Specific Absorbanc e @ 254 nm Figure 4-5. Linear regression an alysis of change in specif ic absorbance at 254 nm from time =0 to time=9 days of UV exposure. Change in absorbance was calculated by subtracting post exposure absorbance values from initial value before exposure. y = 0.0002x + 0.0018 R2 = 0.65 0 0.002 0.004 0.006 0.008 0.01 010203040 Percent Carbon in Phenolic CompoundsChange in Specific Absorbanc e @ 325 nm Figure 4-6. Linear regression an alysis of change in specif ic absorbance at 325 nm from time =0 to time=9 days of UV exposure. Change in absorbance was calculated by subtracting post exposure absorbance values from initial value before exposure.
116 Table 4-1. Characterizati on analysis of standardized DOM sample prior to experimentation. Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). SUVA and aCDOM absorbance at 254 and 325 nm respectively are presented as absorbance divided by carbon concentration. Phenolics content are presented as percent of total organic carbon in sample. Mole cular weight fractions (MWF) are presented as percent of total carbon in sample. Species SUVA Abs. 254 nm Phenolics (%carbon) CDOM Abs. 325 nm MWF <1 KDa MWF 1-3 KDa MWF 3-10 KDa MWF >10 KDa Eleo 0.0214 4.7 0.0109 4.3 15.8 28.2 51.6 Typ 0.0330 37 0.0213 15.4 12.0 12.3 60.2 Clad 0.0225 11 0.0119 20.4 2.9 13.7 62.9 Spar 0.0195 7.5 0.0103 29.4 2.8 10.0 57.7 Thal 0.0220 12 0.0160 32.2 2.0 2.1 63.6 Nuph 0.0314 30 0.0177 6.5 30.3 12.3 50.8 Nym 0.0333 27 0.0196 16.1 8.4 18.0 57.3 Pan 0.0234 4.7 0.0136 15.1 15.0 8.9 60.9 Tax 0.0116 7.5 0.0078 19.2 13.8 7.4 59.6 Table 4-2. Summary of resu lts of photolysis and photoblea ching experiments. Loss of carbon to direct photolysis is presented in % of initial C present. Loss of absorbance in SUVA and aCDOM is also as % of initial values. Decay rates are presented as K per unit if time (d ay). Species names are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nymphea (Nym); Panicum (Pan); Taxodium (Tax). Species Loss of Carbon (%) Change in SUVA (%) Change in aCDOM (%) Photolytic Mineralization Rate K day-1 Eleo 18 34 27 0.0453 Typ 23 42 40 0.0808 Clad 22 31 20 0.0703 Spar 34 25 19 0.0848 Thal 13 36 13 0.0226 Nuph 26 44 40 0.0822 Nym 25 28 20 0.0727 Pan 17 35 31 0.0441 Tax 16 71 55 0.0504
117 CHAPTER 5 SPATIAL DISTRIBUTION AND VEG ETATION EFFECTS ON DISSOLVED ORGANIC CARBON IN THE GREATER EVERGLADES ECOSYSTEM. Introduction The Everglades ecosystem has received much attention in the last decade due in part to its uniqueness as an oligotrophic ecosy stem and the large scale restoration effort at the Federal and State level. The degradation of the Evergl ades during the early and mid twentieth century has led to th e documentation of widespread degradation of the system. Major changes to hydrology were brought about by the introduction of a series of canals and water control structures stil l used today to manage water flow. Inputs of agricultural and urban run-off have increased phosphorus inputs into a historic ally oligotrophic system resulting in nutrient enrichment of periphyton, vegetation, soil, and water (Newman et al., 1996; Newman et al., 1997; Reddy et al., 1999; DeBusk et al., 2001; Fisher and Reddy, 2001; DeBusk and Reddy, 2003). Collectively these anthropogenic changes have brought about many changes to th e historic state of the Everglades. One particularly disturbing observ ation has been that of wi despread shifts in plant communities and ecotypes throughout the system (Newman et al., 1996; Newman et al., 1998; Davis, 1991). The shift in dominant vege tation types in many areas of the northern Everglades ( Typha encroachment) and loss of ecotypic features (ridge and slough) in the southern Everglades has become a topic of c oncern as the ecological significance is not yet fully understood (King et al., 2004). The Everglades are a macrophyte dominated wetland system and rich in organic carbon in the form of plant biomass, organic soils,
118 and dissolved organic carbon in the wate r column (DeBusk and Reddy, 1998). Because macrophytic vegetation can be a very signi ficant source of DOC, changes in plant communities and associated ecotypes could alter DOC and associated nutrient cycling (Wetzel, 1992). The role of emergent vegetation in wetla nds and aquatic systems has received little attention with respect to itÂ’s contribution to the overall production of dissolved organic carbon (DOC) (Bertilsson and Jones, 2003; Farja lla et al., 2001; Scu lly et al., 2004; Mann and Wetzel, 1996; Findlay et al., 1986), even though they can be a dominant source of DOC within systems (Wetzel, 1992; Findlay et al., 1986). It is well known that DOC in wetland and aquatic ecosystems plays a significa nt role in system trophic interactions, nutrient cycling, and the biogeochemistry of metals and potentially harmful pollutants (Findlay and Sinsabaugh, 2003 and references th erein). Previous research has shown that dominant vegetation types in the Florida Ever glades have significantly different potential contributions to the DOC pool. It has al so been shown that species dependant characteristics can influence the biotic and ab iotic transformations of this DOC and thus have effects on the overall dynamics of DOC in the system. Based on my earlier studies, I hypothesize that various plant communities will produce different leve ls of DOC at the community and ecotype level, thus influenci ng spatial distribution of DOC in the water column. Therefore, the purpose of this study was to investigate the influence of various vegetation communities and associated ecotype s on bulk levels of DOC observed across a large wetland system and determine if different plant communities are in fact drivers of spatial variability in DOC concentration of wetland ecosystems.
119 Methods Study Site The Florida Everglades, a 6880 km2 macrophyte dominated sub-tropical wetland located in the southern portion of the state of Florida, USA, provides an opportunistic location to investigate the role of emergent aquatic vegetati on in the production of DOC within wetland ecosystems. This wetland sy stem contains charac teristic vegetation communities and subsequent ecotypes based on geomorphological features within the system. The main vegetation types include Cladium jamaicense , Eleocharis interstincta , Nymphea odorata , and large areas of Taxodium disticum . There are also many other species present to a lesser degree, including many species of tree and shrubs found on tree islands. Other than tree islands, two ot her unique ecotypes found in the Everglades are the ridge and slough ecotype found in associ ation in central and southern Everglades. These ecotypes are thought to form in areas where geomorphological conditions provide and elevated area (ri dge) dominated by Cladium , in close proximity to a more open water system (slough) characterized by the presence of Nymphea and often periphyton. Other ecotypes do exist such as wet prairie areas characterized by Eleocharis , assorted grasses, such as Panicum , and to a much less extent Cladium and other macrophytes sparsely distributed within an area that has no distin ct channels as seen in the ridge and slough ecotypes. Areas of mangrove prairie are f ound at the ecotone of fr esh and salt water in the southern most reaches of the Everglad es National Park (ENP) and are dominated Rhizaphora mangle . Large expanses of shallow to limestone and exposed limestone called marl prairie are also found in the sout heastern and western portions of the ENP and Big Cypress National Preserve (BCNP) and are dominated by an assortment of grasses and to a large extent benthic periphyton.
120 Sampling The large scale sampling effort (1283 s ites) encompassed the following areas: Water Conservation Area(s) (WCA) 1, 2A, 2B , 3A, 3B, the Holeyl and and Rotenberger Tracts (HLRB), Miccosukee and Seminole Indi an Reservations (MSR), Big Cypress National Preserve (BCNP), the Everglades National Park (ENP), and various small private and public lands south east of ENP collectively termed Model Lands (MDLS) (Figure 5-1). To determine appropriate samp ling stations within this large area, a randomly stratified sampling design was employed. Landscapes were divided by hydrologic units and then subdivided in to eco-regions based on a number of characteristics available from previous resear ch in these areas. Prio r to sampling, all sites were determined by randomly generating Lat/ Long co-ordinates within each given ecoregion (with care to overlap prev ious sampling efforts). This sampling effort effectively covered the current extent of the Everglades wetland ecosystem east of the Fakahatchee Strand (Figure 5-2). All sites were acce ssed via helicopter begi nning with WCA-1 in May 2003 and ending with MDLS in Decembe r 2003. The opportunity arose to take water samples concurrently with the Evergl ades Soil Mapping project designed to map soil nutrients across the gr eater Everglades basin. Water samples were taken at each site in 150 ml Nalgene high density polyethylene sample bottles. Water was collected by subm erging bottle at least 10 cm into the water column taking care to minimize introduction of particulate matter into the sample. In areas where water depth was less than 10cm, water was collected by submerging bottle as far into water column as possible to collect the sample and stored on ice until returned to the laboratory for analysis. Samples we re filtered at 0.45 um and acidified with concentrated sulfuric acid to a pH of approximately 2.0. Water column DOC
121 concentration was derived by high te mperature oxidation of DOM to CO2 coupled with IR detection of CO2 in a Shimadzu TOC-5050 (Columbia, Maryland). Data Analysis Data analysis consisted of comparing mean values for the extreme examples of ecotypes and vegetation comm unities within each hydrologic unit (HU). This method entailed picking sites that were clearly dom inant in ecotype (example 80% or greater ridge) or clearly dominant vegetation (example 90% of vegetation was Cladium , and at least 50% coverage). In some ecotypes, such as sloughs or marl prairies where vegetation is sparse, percent cover of vegetation could have been less than 50%. NCSS (Number Cruncher Statistical System, East Kaysville, Utah) software was used to perform one way ANOVA and DuncanÂ’s Multiple Range post hoc tests for differences in means (alpha =0.05 for all tests). In some cases, data were log transformed to meet requirements of normality. Because this study was conducted as an addendum to another ongoing research objective, site selection was not determin ed with DOC and vegetation communities as the primary concern. Therefore the total number of sites utilized in this portion of data analysis was gr eatly reduced with respect to the total number of samples used in the spatial analysis due to the larg e number of sites with heterogeneous vegetation communities. Results of the spatial analysis we re used to determine areas of spatial autocorrelation which were avoided in choosing sites for vegetati on and ecotype comparisons. Descriptive statistics were also used to de termine mean value (95% confidence level) and range of DOC values in each HU. Spatial analysis of DOC within the syst em was performed using ESRI software (Redlands, California). The Geos tatistical Analyst extension of ArcGIS 8.3 was used to interpolate DOC surfaces with in each HU by several methods (Local Polynomial, Radial
122 Basis Function, and Ordinary Kriging). For each HU, models were compared to determine the Â“bestÂ” interpolation. Mode l comparisons were based on comparison of prediction errors. The best interpolation me thod was identified as Ordinary Kriging in most cases and so to keep consistency in the presentation of data, this method was employed for all HU. Surfaces generated from using this method were merged into one surface to create a DOC model for the entire Ev erglades system. Spatial analysis of each individual HU was done separately due to temporal and hydrologic barriers between HUs. Results and Discussion Hydrologic Units Mean values of DOC for a ll hydrologic units are repr esented in Table 5-1. The highest mean values for DOC occur in WCA-1 and HLRB, as well as the highest range of values for DOC. This suggests a very signi ficant influence of ag ricultural drainage waters. Organic soils in the Everglades Agri cultural Area (EAA) have been drained and are under agricultural production. Drainage wa ters from the EAA have been shown to have relatively high levels of DOC by prev ious studies (Qualls and Richardson, 2003). The high values associated with the northern portions of the Evergl ades also correlate with areas of nutrient impact documented by previous studies (Newman et al. 1997; Reddy et al., 1999; DeBusk et al., 2001; DeBusk and Reddy, 2003). Water Conservation Area-2B presents itself as a model of what would be expected of DOC concentrations (mean value 18.9 mg C L-1) for the northern portions of the Everglades. This area is uni que in the WCAÂ’s as it has no si gnificant inflows or outflows and is fed mainly by groundwater inputs. Becau se of its hydrologic isolation, it serves as a control and is useful for comparison of th e other northern areas as it contains mostly
123 Cladium communities with scattere d open water habitats. By comparison, WCA-1, 2A, and 3A, as well as, the HLRB tracts tend to be higher in DOC and have larger ranges in values. This further supports the sugges tion of significant DOC inputs from canals draining the EAA. The mean values trend downward with d ecreasing latitude sugge sting a shift from dominance of allochthanous DOC to insitu production of DOC (autochthonous sources). Further, the downward trend in concentration wi th decreasing latitude suggests that much of the DOC in the system is degraded as th e water travels southwar d toward Florida Bay, and thus little of the input of DOC in the northern reaches is actually exported. Vegetation communities tend to become less dense in the deeper water habitats of the southern reaches of the WCAÂ’s and in the very shallow soil habitats of marl prairies and wet prairies of the ENP and BCNP. This is evidenced by the lower mean DOC values observed in the southern areas of the Evergl ades such as BCNP, ENP and MDLS. This could allow for more photolytic decompos ition and mineralization of DOC as well. Previous studies (chapters 3 and 4) have shown that macrophyte derived DOC is often highly bioavailable and quickly utilized (days) and that this DOC pool can be abiotically mineralized by UV radiation exposure. Qualls and Richardson (2003) and Scully et al. (2004) have shown that Ever glades DOC is UV sensitive and rates from bulk DOC and macrophyte derived DOC are comparable to t hose found in previous portions of this study (chapter 4). Davis (1991) reported that there was a si gnificant difference in plant productivity in the northern Everglades with respect to the southern portions of the system. Therefore, it was expected that ev en without significant inputs of DOC from EAA drainage waters, there would be slightly higher values observed in the areas. This is
124 supported by the use of WCA-2B as a model of expected DOC values, where by comparison, the southern reaches are significa ntly lower in DOC production, presumably due to less dense vegetation communities and possibly higher UV expos ure (Scully et al., 2004). Spatial Distribution As expected, comparison of prediction e rrors showed the local polynomial models did not fit the data as well as the other models. Visual inspection showed these models to be very smooth. For some HUs Kriging mode ls suffered because it was difficult to fit a semiviariogram to the data, suggesting the sc ale of sampling did not match the scale of spatial autocorrelation. Universal Kriging (UK) performed slightly better than Ordinary Kriging (OK) by prediction error comparison, however visual inspection showed UK to result in Â“roughÂ” looking surfaces sugge sting the data were over-fitted, while OK produced moderately smoother surfaces. In many HUs, the spline models were very smooth, but grossly over or under predicted DOC where samples were sparse. Ordinary Kriging consistently produced the "best" re sults (combination of low prediction error and smoothness of surface) and was thus used to model all HU independently. A spatial distribution of DOC within the Greater Everglades Ecosystem (all mapped HU combined) is presented in Figure 5-3. Despite the differences discussed above, all the models produced very similar visual results when viewed at a regional scale (i.e. 1:500,000). At the 1:1,200,000 scale of each HU (the scale of Figure 5-3), visual differences between most of the spatial models were not apparent, suggesting the sa mpling density was adequate for regional scale interpretations (i.e. those focused on gl obal changes within an HU). Changes that are global to an HU are considered lo cal changes for the entire Everglades.
125 While there were few comparisons made be tween hydrologic units due to temporal differences in sampling, the observed DOC conc entrations suggest a latitudinal trend as mentioned previously in the discussion of bulk DOC measurements for each HU. As the WCAÂ’s extend southward, the southern reache s of these areas often impound water and thus vegetation communities are less dense. Th is can be seen in the satellite images (Figure 5-4) available through the South Fl orida Water Management District website (www.sfwmd.gov). Note the darkened areas in the southern portions of each WCA and the corresponding trend in DOC. Inputs of high DOC water can be identified in WCA-1, HLRB, and to a lesser extent, northern WC A-3A where canal inputs bring EAA water into the areas. Water conservation Area 1 indicated a distin ct gradient form north to south in DOC concentration. This gradient suggests high inputs to the northern portions and possibly higher plant production in the northern portions. Note here that wa ter depth increases with distance south. While canal influen ce on all sides of WCA-1 would lead one to expect more of a concentric ring effect of DOC to the area, this was not observed. Water control structures force canal water inflow into two distinct areas of WCA2A, one at the northern most point of the HU, and multiple points along the northeast border with WCA-1. There was approximately 3 weeks between the sampling of WCA-1 and WCA-2A, but the gradient of DOC seemed to stay intact, suggesting constant input of water along the canal interface in this area . Interior portions of WCA-2A were found to have lower concentrations of DOC and sugge st a lack of drainage water influence. These areas are considered to be more pr istine in terms of water quality and plant community (Reddy et al., 1999; Newman et al., 1996). The dominant flow path of water
126 from the established nutrient transects of WC A-2A seems to carry DOC laden waters due south, suggesting that drainage water or plant communities have a significant affect on DOC concentration as the norther n most points of this area are where the most prolific Typha infestations occur. Qualls and Ri chardson (2003) reported increased DOM production in the Typha impacted areas vs. the more pristine Cladium dominated landscape. Trends in WCA-2B were hard to determine as the number of sites were limited in this area, but observations of low plant dens ity in the north east portion of the HU and increasing plant density in the southweste rn portion suggest the importance of plant density on DOC production. This plant density change is likely due to drawdown effects in the higher elevation areas of the north east portion, resulting in more robust plant communities in the southwestern portions and thus more DOC. The Holeyland and Rotenberger tracts (H LRB) were observed to be a region of high drainage derived DOC inputs. The canal inputs in the north and south of these areas show strong gradients of DOC. These gradient s overlap into WCA-3A and MSIR. There exists control structures at the southern e nd of HLRB, and the resulting effects of this high DOC drainage water can be seen in nor thern WCA-3A and in the northeast portion of MSIR. The Miami Canal is perforated in spots and runs diagonally through the plume of higher DOC with an additional outfall located in the top left corner of the secondary observed area of high DOC. These observations strongly suggest that the influence of drainage waters is significan t throughout the northern Evergl ades. Further, there are no physical boundaries between MSIR and north ern WCA-3A and even though there was a
127 significant time lapse between sampling of thes e areas (1 month), the influence of the drainage water inputs was still evident. The observed ecotype and vegetation hom ogeneity of BCNP, which is mostly a mosaic of Taxodium stands and domes with a sparse under story of gr asses, is evidenced here with the lack of spatial differentiati on of DOC concentrations. However, in the southern most reaches of BCNP, sparse Cladium meadows and wet prairies are the dominant feature. The best example of DOC relationships w ith ecotype and vegetation occurs in the ENP. Satellite imagery of the ENP shows Shark Slough, the dominant drainage feature in the ENP, to be bordered by progressively hi gher elevation and thus different vegetation communities. These wet and marl prairies bound the slough to the east and west. The wet prairies also grade into BCNP and MDLS . The trend of DOC seen in the ecotype and vegetation comparisons is reflected in the spatial map of DOC as the Shark River Slough area of the ENP, where the ridge/slough systems domina te, contains the highest DOC bordered by progressively lower DOC areas as the system grades into wet and marl prairie. The southern edge of the ENP is fringed by extensive mangrove swamps where DOC levels tend to rise again (Scully et al ., 2004). This is also reflected in the comparison of mangroves in both ecotype s and vegetation of ENP and MDLS. Vegetation Previous characterization of DOM deri ved from different species of wetland vegetation in the Everglades (Chapter 2) f ound there to be significant differences in the potential for different plant t ypes to contribute DOC to the sy stem. Of particular interest to this study were the difference in potentia l contribution from sel ected species and the observed effects in the field. Eleocharis was found to potentially contribute 40 g of DOC
128 per kg of senescent biomass, which is signifi cantly higher than the potential contributions of Typha (34 g/kg) and Cladium (22 g/kg). However, Panicum (48 g/kg) had a potentially larger contribution, as did Ta xodium (105 g/kg) and Nymphea (220 g/kg). In the analysis of vegetation types and m ean DOC concentratio ns, periphyton and mangroves were introduced, for which there is no potentially leachable DOC data. Scully et al. (2004) conducted leaching experi ments on these materials, but did not report the values; therefore evaluation of potential co ntributions of this type cannot be made for periphyton and mangroves. It has been reported elsewh ere (Qualls and Richardson, 2003; Davis et al., 2003; Jaffe et al., 2004) and observed in th is study that DOC increases in the mangrove zones and so it is assume d that mangroves contribute a significant amount of DOC. Analysis of dominant vegetation communities is summarized in Table 2. Of particular interest is the trend of Typha dominated areas to be consistently higher producers of DOC over its ecotype (ridge) counter part Cladium . In all areas where Typha had considerable presence, with the exception of WCA-1 where any statistical differentiation was hindered by high alloch thonous inputs of DOC, it had the highest mean concentration of DOC. In WCA-1, results did suggest Typha to be higher producer of DOC than Cladium , but no significant difference was found. Although in some areas, dominant ridges of Typha were not sampled densely enough to make statistical comparisons; the data collected suggest that this trend is consistent throughout the system, as was observed in WCA-2A, WCA3A, and HLRB. This observation is of particular interest as the ecological impact of the increasing invasion of Typha into the Everglades is unknown. Further, in many cases Cladium dominated areas reflect higher
129 DOC concentrations than traditional slough a nd wet prairie species, thus supporting the trend of ridge ecotypes being higher than slough and wet prairie ecotypes in DOC production. It is important to note that in areas where Cladium does not dominate other species (except Typha ) the observed density of Cladium is considerably lower, which allows other vegetation types found in great er densities to produce more DOC per unit area. In BCNP, Cladium was not observed in very dense stands as was the case in other areas, but grasses and sedges of the wet prairi es and depressional marshes often exceeded 70% coverage. There was no significant difference (P>0.35) found between Cladium dominated sites and sites dominated by IN T SS (Integrated slough species = Nymphea , Utricularia , and scattered periphyton, none of which was clearly dominant and therefore grouped together for the analysis). In this case, it is likely that Nymphea, a potentially large and significant contributor of DOC, is too sparse to produce significantly higher DOC levels as suggested by its potential, and so was overshadowed by the more dense areas of Cladium . This was also evident in WCA-3A and WCA-3B where Cladium dominated areas were found to have significantly higher levels of DOC than areas of periphyton and integrated slough species. In this case peri phyton dominated areas we re not significantly different from Cladium areas, suggesting that periphyt on has the potential to produce similar amounts of DOC. Another possible e xplanation would be th at since periphyton mats were floating on the surface in these ar eas during sampling, they effectively shade DOC from other sources from UV radiati on and subsequent mineralization (Wetzel, 2001).
130 In the analysis of vegetation communities in HLRB, Panicum dominated sites were available in to an extent to enable comparisons with Typha and Cladium communities. While Cladium was found to have significantly (P<0.05) lower levels of DOC than Typha dominated areas, as expected ba sed upon potential production of DOC, Panicum dominated areas, which have he potentia l to produce more DOC than either Typha or Cladium were found to be similar to those dominated by Cladium . This is likely the result of density differences discussed previ ously. While Panicum had been shown to have a higher potential production of DOC per kg of senescent biomass, this plant was not observed in densities equivalent to those of Cladium or Typha . However, in MSIR, Panicum was found in densities (visual observations) similar to Cladium and were also found to be associated with significantly high er DOC concentrations. Levels of DOC in Eleocharis dominated areas of MSIR were f ound to be significantly lower than Cladium (similar to results of WCA-1 and ENP) a nd was observed in low enough densities to suggest that in this case, mass of Eleocharis was insufficient to reflect its potential contribution. Analysis of plant communities common in BCNP found that Taxodium (highest potential) was associated with the highest levels of DOC. This sugge sts that the potential contribution of Taxodium was reflected in this area. Integrated grasses had the second highest value, which suggests thes e species ability to produce DOC. Cladium was found to be unexpectedly low; how ever, as mentioned before, Cladium densities in this HU were very low. In the cas e of periphyton, it is importan t to note that periphyton communities were associated with the benthic substratum (limestone outcrop and soil surface) in this HU which suggests that in fact, periphyton may shade DOC, rather than
131 produce significant quantities, as periphyton communities in ENP and MDLS were also benthic in nature and in bot h of those HU, periphyton dominated areas had the lowest DOC values. Comparisons of DOC values associated with plant communities in the ENP suggest that densities of Eleocharis were too low to reflect pot ential of DOC production with respect to Cladium . Mangroves were introduced in the analysis of this HU and while elevated DOC concentrations have been obser ved in this area previously (Qualls and Richardson, 2003; Davis et al., 2003; Jaffe et al., 2004), the sa mpling of this study only penetrated the transition zone (fringe) of fresh water mars h and mangrove systems. In the MDLS, sampling of clearly mangrove domin ated areas reflected the inputs of DOC from this vegetation type more clearly. In this HU Cladium was relatively sparse in comparison to other HU and results suggest simila rity with integrated grasses in terms of DOC production. As mentioned previousl y, periphyton was confined to subsurface benthic structure and likely c ontributed little to water co lumn DOC concentrations. Ecotypes Because ecotypes in the Everglades are ch aracteristically dominated by particular plant communities, a comparison of ecotypes an d their associated DOC concentrations is valuable to further investigate trends in DOC with respect to ve getation. Analysis of mean values of DOC for ecotypes is summar ized in Table 5-3. In most cases, slough ecotypes tend to be significantly lower pr oducers of DOC than ridge systems (WCA-2A, WCA-3A, MSIR). Again this is likely due to the low densities in which these plants exist in slough systems. In areas such as WCA-1 and HLRB, wh ere high inputs of allochthanous DOC occur, the trend is reve rsed. This is possibly due to increased transport of DOC laden waters of EAA draina ge origin in the open, less vegetated slough
132 systems. In the ENP, the distinction betw een ridge and slough ecotypes in terms of DOC is not observed. This is likely result of th e loss of the classical open water slough ecotype from the invasion of slough systems by ridge an d wet prairie species due to persistent low water levels (altered hydrol ogy). It is also noteworthy to mention that shallow water habitats such as wet prairies and marl prai ries had consistently lower levels of DOC except in WCA-3A where wet pr airie ecotypes had very high percent coverage (70-80%) of vegetation in contrast to other areas wh ere vegetation percent cover was often much lower (20-30%), suggesting a better reflection of this ecotype (and Eleocharis , the dominant species) to produce DOC. In BCNP, no significant difference was found between th e wet prairie and depressional marshes, which was expected as there was little difference in vegetation. These ecotypes were based more on hydrologic continuity than on distinct vegetation groupings. In the instances where there was differing plant communities, the densities were quite low. In the ENP and MDLS, ma rl prairies, dominated by benthic periphyton and sparse vegetation, were found to be signi ficantly lower in DOC than other ecotypes, reflecting again, the differences in vege tation type and potential DOC production. In the analysis of ecotypes, the results act to support the findings of the vegetation analysis. This was expected as ecotypes are generally characterized by their plant communities, or lack there of. With th e exception of WCA-1, HLRB, and WCA-2A, trends expected in vegetation were fo und to follow through to ecotype as well. Conclusions In the greater Everglades, DOC concentrati ons were found to be variable across the landscape, as well as, within smaller hydrologi c units (HU). Results suggest a latitudinal gradient of DOC concentration and a signi ficant input of DOC from EAA drainage
133 waters. The existence of this gradient also suggests that as water moves from north to south through the Everglades system, a la rge portion of the DOC is utilized or photolytically mineralized before reaching the mangrove interface of ENP. Within individual HU, both ecotype and subseque nt vegetation type seem to contribute significantly different levels of DOC to the system. Therefore, plant type could be a significant factor in DOC production and contribution to DOC dynamics in the Everglades. Typha has been shown to contribute greater amounts of DOC in areas where it has proliferated and become dominant. This is important in that it is presently unknown how ongoing shifts in ecotype (loss of ridge/sl ough system in ENP) and dominant vegetation (large scale invasion of Typha in WCA-1, 2A, 3A) in the Everglades could alter the system at the level of cycling of DOC or at the ecosystem level. Spatial analysis suggest that in the no rthern borders of WCA-1 and HLRB, inputs of water from the EAA cause the relationships between ecotypes and their contribution of DOC to break down or reverse. The impact s of these high con centration inputs are currently unknown, but warrant further investigation.
134 10010Kilometers N E W S #RB #HL #WCA-2B MDLS ENP BCNP-S BCNP-N WCA-3A #WCA-3B WCA-1 WCA-2A #MSIR 1:1,200,000Florida #Study Site Figure 5-1. Map of area where water sa mples were obtained in 2002 and 2003.Water Conservation Area (WCA-1) is also known as the Loxahatchee National Wildlife Refuge. Big Cypress National Pr eserve (BCNP) is separated into northern and southern halves, BCNPN and BCNP-S respectively. The Miccosukee and Seminole Indian Reserv ation is denoted by MSIR and the Everglades National Park by ENP.
135 # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # ## # # ## # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # ## # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## ## # # ## # # ## # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # ## # # # # # # # ## # # ## # # # # # # ## # ### ## # # ## # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # ## # # # # # # # # # ## # ## # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # ## # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # ## # ## # # # # # # # # # # # # # # # ## # # # ## ## # # # # # ### # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # ## # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # ## # # # # # # # # ## # # # # # ## # # # # # # # # # # # # # # # # # # # # # # # # # # # # 10010Kilometers N E W S Figure 5-2. Locations of all water sampling locations within the greater Everglades basin. Hydrologic units (HU) are abbrev iated as follows: Water Conservation Area (WCA); Holeyland and Rotenber ger tracts (HLRB); Miccosukee and Seminole Indian Reservations (MSI R); Big Cypress National Preserve (BCNP); Everglades National Park (ENP ); and the collective properties south east of the park termed the Model Lands (MDLS). WCA 1 WCA -2A WCA -3A WCA -2B WCA -3B ENP MDLS BCNP MSIR HLRB 1:1,000,000
136 Florida N E W S 10010Kilometers #Study Area 1:1,200,000 #WCA 1 #WCA 2A #WCA 2B #HL/RB #ENP #WCA 3B #WCA 3A #MIC #BCNP-S DOC (mg/L) 5 10 10 15 15 20 20 25 25 30 30 35 35 40 40 45 45 50 50 60 No Data #MDLS #BCNP-N Figure 5-3. Spatial distribution of DOC in th e greater Everglades basin. Hydrologic units (HU) are abbreviated as follows: Water Conservation Area (WCA); Holeyland and Rotenberger tracts (HL RB); Miccosukee and Seminole Indian Reservations (MSIR); Big Cypress National Preserve, north and south (BCNP-N & S); Everglades National Park (ENP); and the collective properties south east of the park termed the Model Lands (MDLS).
137 Figure 5-4. Satellite image of the greater Ev erglades basin study area. Hydrologic units (HU) are outlined with light borders (source: South Florida Water Management District). ENP W C A -1 MSIR BCNP W C A -3B W C A -3 A W C A -2B W C A -2 A MDLS HLRB
138 Table 5-1. Summary of disso lved organic carbon (DOC) measurements throughout the greater Everglades basin. Hydrologic un its (HU) are abbreviated as follows: Water Conservation Area (WCA); Hole yland and Rotenberg er tracts (HLRB); Miccosukee and Seminole Indian Reserv ations (MSIR); Big Cypress National Preserve (BCNP); Everglades Nation al Park (ENP); and the collective properties south east of the park term ed the Model Lands (MDLS). Numbers in () denote +/std. dev. of the mean HU Sample Size (n) Mean DOC mg C L-1 Mean Water Depth (cm) Max DOC mg C L-1 Min DOC mg C L-1 WCA-1 119 38.0 (13) 27 83.2 16.5 WCA-2A 98 25.4 (6.2) 34 53.7 17.5 WCA-2B 20 18.9 (5.3) 48 33.1 13.8 WCA-3A 256 20.7 (6.6) 64 42.6 8.2 WCA-3B 62 16.6 (4.0) 38 23.6 9.1 HLRB 95 32.1 (14) 16 83.9 8.9 MSIR 78 18.4 (4.1) 54 30.5 9.9 BCNP 186 10.9 (3.4) 18 23.6 4.3 ENP 323 13.5 (4.0) 35 24.3 3.6 MDLS 46 12.0 (3.9) 11 20.0 4.4
139 Table 5-2. Summary of plant community dissolved organic carbon (DOC) comparisons throughout the greater Everglades ba sin. Hydrologic units (HU) are abbreviated as follows: Water Cons ervation Area (WCA); Holeyland and Rotenberger tracts (HLRB); Miccosuk ee and Seminole Indian Reservations (MSIR); Big Cypress National Preserve (BCNP); Everglades National Park (ENP); and the collective properties sout h east of the park termed the Model Lands (MDLS). Plant types are abbreviated as follows: Cladium (Clad); Typha (Typ); Eleocharis (Eleo); periphyton (Peri); Panicum (Pan); Integrated slough species (Int SS); Integrated grasses (Int G); Taxodium (Tax); and mangrove (Man). Values presented are mean values (ANOVA Duncans Multiple Range, P<0.05, significant difference indicated by *). The symbol --indicates insufficient data for comparisons. HU Clad Typ Eleo Peri Pan Int SS Int G Tax Man WCA-1 41.4 43.7 45.1 ------------WCA-2A 23.8 28.5* --------------WCA-2B 20.1 --------16.3 ------WCA-3A 21.2 25.9* --20.7 ----------WCA-3B 16.4 ----16.8 --13.3* ------HLRB 26.9 32.1* ----25.0 --------MSIR 17.1 --13.2* --21.8** --------BCNP 10.9 ----8.7* ----12.6** 14.1** --ENP 14.1 --11.6* 8.1** --------11.2* MDLS 12.1 ----8.5* ----12.4 --14.9**
140 Table 5-3. Summary of ecot ype comparisons with respect to DOC concentration (ANOVA Duncans Multiple Range P< 0.05, significant difference indicated by *). . Hydrologic units (HU) ar e abbreviated as follows: Water Conservation Area (WCA); Holeyla nd and Rotenberger tracts (HLRB); Miccosukee and Seminole Indian Reserv ations (MSIR); Big Cypress National Preserve (BCNP); Everglades Nation al Park (ENP); and the collective properties south east of the park term ed the Model Lands (MDLS). Ecotypes are abbreviated as follows: ridge (R), slough (S); wet prairie (WP), depressional marsh (DM), and marl pr airie (MP). The symbol --indicates insufficient data for comparisons HU R S WP DM MP WCA-1 41.3 45.4* 31.5 ----WCA-2A 26.2 22.2* ------WCA-2B 20.1 16.8 ------WCA-3A 22.4 16.2* 26.4 ----WCA-3B 17.9 14.0* ------HLRB 29.4 37.5* ------MSIR 19.3 15.7* ------BCNP ----10.8 10.4 --ENP 15.6 14.3 12.2* --9.7** MDLS 14.8 --11.3* --9.9*
141 CHAPTER 6 SUMMARY AND SYNTHESIS Review of Objectives The central hypothesis of this research wa s that vegetation taxa and its chemical composition are significant factors in form ation, characteristics, and reactivity of dissolved organic matter (DOM) in wetlands. Specific hypotheses test ed in this study were: 1) DOM from different vegetation type s is chemically different, and thus the functional role of this material in wetland systems is variable; 2) DOM from different plant sources will react differently to microbi al degradation; 3) DOM of different plant origin will vary in the degree of suscep tibility to UV photolysis; and 4) spatial distribution of DOM in large wetland systems is dependant on vegetation communities. To test these hypotheses, a series of expe riments with the following objectives were conducted. Characterize both POM and DOM from a va riety of dominant wetland vegetation types to determine if differences of DOM exist and if there are significant links between POM characteristics and resulti ng DOM characteristics (both chemical and physical). Determine the biodegradability (relative amount of microbial utilization) of DOM produced by different dominant wetland ve getation types and relate this to the characteristics (physical and chemical ) of DOM from each vegetation type Determine both the effects of abiotic degradation via UV photolysis and photobleaching on plant derived DOM and the potential amount of DOC mineralization resulting from UV light exposure. Investigate the spatial distribution of DOM and the functional role of vegetation types in various hydrologic units of the Everglades wetland ecosystem.
142 To satisfy objective 1, senescent plant ma terial from several vegetation types, collected in the Everglades wetland system, was subjected to leaching with water (to produce DOM) and chemical analysis of plan t fiber and nutrients. The results of the characterization and exploration of linka ges between POM and DOM revealed that although POM from different plant types was composed of similar structural building blocks, levels of nutrients and soluble orga nic materials were highly variable. The DOM produced from these materials was significantly different both chemically and physically. The characteristics measured in these experiments were by no means exhaustive or completely definitive of chemical character; however, they were sufficient to determine whether or not different species produce physically and chemically different DOM. Significant linkages of chemical and physical characteristics of DOM to the source POM were not as apparent as expected. It is likely th at more rigorous chemical characterization of these materials is n ecessary to elucidate these relationships. To complete objective 2, a set of expe riments was conducted to evaluate the decomposition and bacterial utilization of di ssolved organic matter (DOM) derived from wetland vegetation types commonly found in the Everglades under aerobic conditions and in the presence of unlimited nitrogen and phosphorus supply. Dissolved organic matter was characterized based on prominent ch aracteristics thought to enhance or retard decomposition (carbohydrate content, protein content, phenolic content, and molecular size). Total loss of carbon over a two phase (40 day) decom position trial resulted in a range of carbon loss from 57% ( Typha domingensis ) to 79% (Spartina bakerii ) compared to a glucose control, which lost 75%. Decomposition rates varied from 0.032 day-1 ( Typha domingensis ) to 0.079 day-1 ( Nymphea odorata ) as compared to the glucose
143 control (0.030 day-1). Bacterial growth efficiency (BGE), the percent of carbon utilized that went to biosynthe sis, ranged from 12% ( Taxodium disticum ) to 46% ( Typha domingensis ) as compared to the low BGE of glucose (7%). Glucose, used as a comparative control in these experiments, was likely used almost exclusively for metabolic maintenance due to the lack of ami no acids, or suitable precursosrs, needed for protein synthesis. Results suggest that characteristics meas ured were significa ntly different among species; however, no significant relationshi ps were found between characteristics measured and decomposition rates or bacteria l growth efficiency. Multiple regression analysis of total carbon lost at 8 days with ch aracteristics measured re vealed that phenolic content and percent of carbon in the 1-3 KDa molecular weight fraction predicted 86% of variability in the loss of carbon measurements . Results further suggested some complex interaction of carbohydrate, protein, and phenolics compounds making some of these fractions more or less available depending on the interactions. Also, the amount of these compounds in the relatively unavailable molecular weight fraction >10 KDa was not determined, and therefore could be reason for the lack of correla tion between measured characteristics and decomposition rates and BGE. The possible role of secondary plant metabolites, compounds produced as chemical de fenses to herbivory, in slowing bacterial degradation and incorporation of DOM is proposed as a seco nd possible source of variability in the data. Results of calculati ons of potential for e xport, mineralization, and bacterial incorporation suggest that shifts in plant communities, presently observed in the Everglades, could have significant impact on the DOM dynamics of the system.
144 Relatively recent investigations into the reactivity of DOM in wetlands and other aquatic systems has brought into question the role of abiotic decomposition of DOM. To satisfy objective 3, DOM derived from wetland vegetation types, common to the Florida Everglades, was exposed to natu ral sunlight over the course of nine days. Ultraviolet radiation was measured during the exposures , and samples of DOM were measured for loss of carbon as DOC. Photolytic minera lization was significantly different among groups of species, and pheno lic content was correlated (r2=0.81, P<0.001) with changes in specific absorbance at 254nm, indicating loss of aromatic and hydrophobic structure. Photo-bleaching, measured by loss of absorban ce at 325 nm, was found to be correlated (r2=0.65, P<0.001) with phenolic compound conten t in DOM samples, but results suggest that other compounds are also under going photo-bleaching and contributing to overall direct photolysis losses of carbon. In an effort to better understand the role of ecosystem types and vegetation communities in dissolved organic car bon (DOC) production, surface water samples collected from 1283 sites within the greater Everglades basin were analyzed for DOC (objective 4). Samples were collected in summer and fall of 2003 from Water Conservation Areas 1, 2A, 2B, 3A, 3B, th e Holey Land and Rotenberger Tracts, the Miccosukee and Seminole Indian Reservations , Big Cypress Nationa l Preserve, and the Everglades National Park. Surface water DOC concentrations varied significantly, with values ranging from 4.4 Â– 83.8 mg C L-1. Geostatistical methods and GIS techniques were employed to generate maps of DOC le vels throughout the greater Everglades basin. Results of the mapping effort strongly suggest a significant influence of canal inputs of DOC into the Northern Everglades. Seconda rily, the observed gradient in DOC suggests
145 that most DOC is consumed within the Everglad es and that relatively little is exported. Observed ecosystem types (ridge, slough, wet prairie) and vegetation communities (percent cover per species) at selected sites within each hydrologic unit were documented. Comparisons of mean DOC conc entration were made, and in many cases, vegetation type and ecotype were signif icantly different with respect to DOC concentration suggesting that inputs from these communities and ecotypes influence DOC dynamics in the Everglades. Synthesis The results presented in this study sugge st that DOM dynamics are linked to the source of the DOM. It has been shown he re that different plant species produce characteristically different DOM, with variable levels of reactivity in abiotic and biotic decomposition. Much research into DOM dynami cs started with char acterization of this material and observation of differences in DOM from different sources. Historically, differentiation of DOM from terrestrial and a quatic sources has provi ded insight into the bioavailability of this material and its role in the ecosystem in which it is found. With this set of studies, evidence has been presented to support another level of complexity in what is understood to be the DOM cycle in wetla nds and aquatic ecosystems in general. A greater understanding of how particular species affect DOM dynamics, and organic carbon cycling specifically, helps to determ ine the ecological function and value of different plant species in a given environment. Calculations of potential pr oduction, export, bacterial in corporation, and abiotic / biotic mineralization have significant impli cations for wetland systems (Figure 6-1). Pathways of organic carbon production, minerali zation, and export for the plants studied are presented in Table 6-1. Values in the (A) column represent the amount of organic
146 carbon (g kg-1) produced from 1 kg of plant mate rial, and corresponded to the (A) pathway on Figure 6-1 representing the DOC pool generated from I kg of plant particulate organic matter (POM). Similarly, the pathways of photolysis (B), bioavailable DOC (C), microbial biomass production (D), mi crobial respiration (E), and export (F), calculated from this study are summarized. Differences observed in the potential movement of DOC from pools to degradati on pathways suggest th at changes are the primary sources of this material could ha ve significant impact on the cycling of DOC within a wetland system. The observation of large scale shifts in vegetation communities, seen globally in wetlands systems due to alterations of hydr ology and nutrient inputs derived from human activity, could have broad reaching effects on the DOM cycle. If the Everglades system is used as an example, the shift from historic Cladium dominated marsh to dense stands of Typha could have severe effects on the ecology of the system. Dissolved organic matter is provides a variety of ecologi cal functions to the systems where it is found. In the Everglades, a change form one major sour ce of DOM to another could have cascading effects on the bacterial communities adapted to utilize this material. In turn, movement of this material through higher trophic levels may be altered and have broad reaching effects on the native biota. Alterations in bi ogeochemical cycling of major nutrients such as N and P could potentially exacerbate the eutrophication problem already at the forefront of concern in the Everglades ba sin. Changes in dominant DOM source could also have unknown effects in the cy cling of metals such as mercury.
147 Conclusions The results of this research suggest th at plant types are very important in determining the reactivity and fate of DOM produced autochthonously in the Everglades ecosystem. From these experiments and i nvestigations, I conclude the following: Different species of wetland vegeta tion produce physically and chemically different types of DOM. Certain characteristics of the parent pl ant material, such as fiber fractions and nutrient ratios, can be used to predict potential DOM production and nutrient content. Plant derived DOM reacts differently to abiotic and biotic decomposition processes, depending upon species of origin. Characteristics of DOM such as phe nolic content and molecular weight fractions are significant predictors of reactivity of DOM to abiotic and biotic decomposition. A significant gradient of DOC exists in the Everglades system and the inputs of DOC from the northern mo st areas surrounding canal discharge areas suggest a significant impact of th ese drainage waters with respect to DOC loading to the system. Plant community structure is a signif icant factor in DOM production in the Everglades, and shifts in this struct ure may have significant impacts on the cycling of organic carb on within the system. In conclusion, while the effects of thes e shifts in plant communities are still unknown, the results of this research have shown that plant species are a significant element in the characteristics of DOM derive d from them. Further, these plant sources produce both physically and chemically diffe rent DOM with subsequently different reactivity to degradation processes. Thes e conclusions indicate a need for further research into how changes to plant comm unities will impact the functioning of the Everglades with respect to DOM and associat ed biogeochemistry and system energetics.
148 Future Research Needs The future needs of research in this ar ea of plant derived DOM are many. To better understand the overall ecological co ntribution of differe nt plant species to carbon cycling, especially DOM cycling, more studies must compare different plant types in context of community structure and potenti al contribution of organic ca rbon. This requires stringent characterization of both particulate and dissolv ed organic fractions a nd elucidation of the linkages between them. This goal will also re quire refinement of the characteristics of importance selected to describe this materi al. The role of secondary plant compounds, those used for defense against herbivory or fo r allelopathic effects, must be thoroughly investigated to determine if they exert a ny level of control over DOM bioavailability. Finally, the functional role of species divers ity and plant community structure on organic matter cycling, with special respect to DOM, must be addressed, as wetlands and other aquatic systems continue to undergo shifts in plant communities due to anthropogenic influence. The ecological effects of these shifts in plant communities are still not known and without proper attention, c ould negatively impact aquatic and wetland systems, such as the Everglades, and cause furt her degradation of these systems.
149 Litter Microbial biomass DOC CO2 Leaching UV Import Export CO2 B A C D E F Litter Microbial biomass DOC CO2 CO2 Leaching UV Import Export CO2 CO2 B A C D E F Figure 6-1. Amended conceptual model of dissolved organic carbon cycling in a wetl and ecosystem. Pools of carbon storage and degradation pathways are denoted by bold letters as follows: leachable DOC pool from 1 kg of particulate organic matter (A); DOC lost to direct photo-mineraliza tion (B); bioavailable DOC (C); DOC in corporated into microbial biomass (D); DOC lost to microbial respirati on (E); DOC lost to export (F).
150 Table 6-1. Summary of poten tial carbon storage and losses in the dissolved organic carbon cycle of a wetland ecosystem. Sp ecies are abbreviated as follows: Eleocharis (Eleo); Typha (Typ); Cladium (Clad); Spartina (Spar); Thalia (Thal); Nuphar (Nuph); Nym phea (Nym); Panicum (Pan); Taxodium (Tax). Cloumns represent pathways of DOC and loss from the carbon cycle. All values are in g kg-1.representing production and loss of DOC from 1 kg of plant detrital material. Species Leachable DOC (A) Photomineralized DOC (B) Bioavailable DOC (C) Microbial Growth (D) Microbial Respiration (E) Export (F) Eleo 40.2 7.3 27.7 5.0 22.7 12.6 Typ 34.3 7.8 19.6 8.9 10.7 14.6 Clad 21.5 4.8 15.1 5.0 10.1 6.4 Spar 38.4 13 29.3 3.9 25.5 9.0 Thal 87.4 11 52.8 19.7 33.1 34.6 Nuph 179 46 97.9 20.3 77.6 81.6 Nym 220 55 153 38.3 115 66.3 Pan 48.2 8.4 27.2 3.9 23.3 21.0 Tax 105 16 67.6 8.1 59.6 37.7
151 LIST OF REFERENCES Aiken, G., D. McKnight, R. Wershaw, and P. MacCarthy. 1985. Humic Substances in Soil, Sediment, and Water. 1st ed. Wiley-Interscience, New York. pp 443 Almendros, G., and J. Dorado. 1999. Mol ecular characteristics related to the biodegradability of humic acid preparations. European Journal of Soil Science 50: 227-236 Amon, R.M. and R. Benner. 1996. Bacterial utilization of differe nt size classes of dissolved organic matter. Limnol. Oceanogr. 41: 41-51 Amon, R., H. Fitznar, and R. Benner. 2001. Linkages among bioreactivity, chemical composition, and diagenic state of marine dissolved organic matter. Limnol. Oceanogr. 46: 287-297 Anasio, A., T. Denward, L. Tranvik, and W. Graneli. 1999. Decreased bacterial growth on vascular plant detritus due to photochemical modification. Aquat. Microb. Ecol. 17: 159-165 Bano, N., M. Moran, and R. Hodson. 1997. B acterial utilization of dissolved humic substances from a freshwater swam p. Aquat. Microb. Ecol. 12: 233-238 Benke, A., I. Chaubey, M. Ward, E. Dunn. 2000. Flood pulse dynamics of an unregulated river floodplain in the southeastern U.S. Coastal Plain. Ecology 81: 27-35 Benner, R. and B. Biddanda. 1998. Photochemical transformations of surface and deep marine dissolved organic matter: effect s on bacterial growth. Limnol. Oceanogr. 43: 1373-1378 Benner, R. and R. Hodson. 1985. Microbi al degradation of the leachable and lignocellulosic components of leaves and wood from Rhizophora mangle in a tropical mangrove swamp. Marine Ec ological Progress Series 23: 221-230 Benner,R., A. Maccubbin, and R. Hodson. 1984. Preparation, characterization, and microbial degradation of specifically radi olabled (14C) lignocellu loses from marine and freshwater macrophytes. A ppl. Environ. Micro. 47: 381-389 Benner, R., M. Moran, and R. Hodson. 1986. Biogeochemical cycling of lignocellulosic carbon in marine and freshwater ecosystems : relative contributions of prokaryotes and eucaryotes. Limnol. Oeanogr. 31: 89-100
152 Benner, R., S. Newell, A. Maccubbin, and R. Hodson. 1984. Relativ e contributions of bacteria and fungi to rates of degradation of lignocellulosi c detritus in salt marsh sediments. Appl. Environ. Microbiol. 48: 36-40 Bertilsson, S. and J.B. Jones. 2003. Supply of dissolved organic matter to aquatic ecosystems: autochthonous sources. In: Aquatic Ecosystems: Interactivity of Dissolved Organic Matter. Findlay, S.E. and R.L. Sinsabaugh (Eds). Academic Press, Boston, MS. 512 p. Bertilsson, S. and L. Tranvik. 1998. Phot ochemically produced carboxylic acids as substrates for freshwater bacteriopl ankton. Limnol. Oceanogr. 43: 1862-1870 Bertilsson, S. and L. Tranvik. 2000. Photoche mical transformation of dissolved organic matter in lakes. Limnol. Oceanogr. 45(4): 753-762 Betts, K. 1998. Growing concern about disinfec tion byproducts. Environ. Sci. Technol. 32: 546 Boavida, M. and R.G. Wetzel. 1998. Inhibi tion of phosphatase activity by dissolved humic substances and hydrolytic reactiva tion by natural ultraviolet light. Freshwat. Biol. 40: 285-293 Boehme, J., and P. Coble. 2000. Characteri zation of colored dissolved organic matter using high-energy laser fragmentation. Environ. Sci. Tech. 34(15): 3283-3290 Bunte, C., and M. Simon. 1999. Bacter ioplankton turnover of dissolved free monosaccharides in a mesotrophic la ke. Limnol. Oceanogr. 44: 1862-1870 Burdige, D., and K. Gardner. (1998) Molecu lar weight distribution of dissolved organic carbon in marine sediments pore wate rs. Marine Chemistry 62: 45-64 Bushaw, K. and many others. 1996. Photoche mical release of biologically available nitrogen from aquatic dissolved or ganic matter. Nature 381: 404-407 Chefetz, B., A. Deshmukh, P. Hatcher, and E. Guthrie. 2000. Pyrene sorption by natural organic matter. Environ. Sci. Technol. 34: 2925-2930 Childers, D., R. Doren , R. Jones, G. Noe, M. Rugge, and L. Scinto. 2003. Decadal change in vegetation and soil phosphorus pa tterns across the Everglades landscape. J. Environ. Qual. 32: 344-362 Clair, T. and B. Sayer. 1997. Environmental variability in the reactivity of freshwater dissolved organic carbon to UV-B. Biogeochemistry 36: 89-97 Covert, J., and M. Moran. 2001. Molecular ch aracterization of the estuarine bacterial communities that use high and low molecular weight fractions of dissolved organic carbon. Aquat. Microb. Ecol. 25(2): 127-139
153 Crews, P., J. Rodriguez, and M. Jaspars. 1998. Organic Structure Analysis. Oxford Univ. Press, New York. 549 pp. Dai, K., C. Johnson, and C. Driscoll. 2001. Organic matter chemistry and dynamics in clear-cut and unmanaged hardwood forest ecosystems. Biogeochem. 54: 51-83 Dai, X., C. Ping, and G. Michaelson. 2002. Characterizing soil organic matter in artic tundra soils by different anal ytical approaches. Organic Geochemistry 33: 407419 Davis, C. and A. Van der Valk. 1983. Litter decomposition in prairie glacial marshes. In R. Good, D. Whigham, and R. Simpson (E ds.): Freshwater Wetlands: Ecological Processes and Management Potential. Academic Press, New York NY Davis, S. 1991. Growth, decomposition, and nutrient retention of Cladium jamaicense Crantz and Typha domingensis Pers. In the Florida Everglades. Aquatic Botany 40: 203-224 Davis, S., D. Childers, J. Day, D. Rudnick, F. Sklar. 2003. Factors affecting the concentration and flux of materials in tw o southern Everglades mangrove wetlands. Mar. Ecol. Prog. Ser. 253: 85-96 Davis, W., C. Erickson, C. Johnson, J.J. De lfino, and J. Porter. 1999. Quantitative Fourier transform infrared spectroscopi c investigation of humic substance functional group composition. Chemosphere 38: 2913-2928 Dawson, T., S. Maubelli, A. Plamboeck, T. Temp ler, and K. Tu. 2002. Stable isotopes in plant ecology. Ann. Rev. Ecol. Syst. 33: 507-529 DeBusk, W.F., S. Newman, and K.R. Re ddy. 2001. Spatio-temporal patterns of soil phosphorous enrichment in everglades wate r conservation area 2A. J. Environ. Qual. 30(4): 1438-1446 DeBusk, W.F. and K.R. Reddy. 1998. Turnover of detrital organic carbon in a nutrient impacted Everglades marsh. Soil Sci. Soc. Amer. J. 62(5): 1460-1468 DeBusk, W.F. and K.R. Reddy. 2003. Nutrient and hydrology effects on soil respiration in a northern Everglades marsh. J. Environ. Qual. 32(2): 702-710 DeBusk, W.F., J. White, and K.R. Reddy. 2001. Carbon and nitrogen dynamics in wetland soils. In: Modeling Carbon and N itrogen Dynamics for Soil Management. M. Shaffer et al. Eds. Lewis Pub. Boca Raton.pp 27-54 Del Castillo, C., P. Coble, J. Morell, J. Lopez, and J. Corredor. 1999. Analysis of the optical properties of the Orinoco Ri ver plume by absorption and fluorescence spectroscopy. Marine Chemistry 66: 35-51
154 Del Giorgio, P. and J. Cole. 1998. Bacterial grow th efficiency in natural aquatic systems. Annu. Rev. Ecol. Syst. 29: 503-541 Dilling, J. and K. Kaiser. 2002. Estimati on of the hydrophobic fraction of dissolved organic matter in water samples usi ng UV photometry. Wat. Res. 36 p 5037-5044 Driebe, E. and T. Whitman. 2000. Cottonwood hybridization affect tannin and nitrogen content of leaf litter and alters decomposition. Oecologia 123: 99-107 Dubois, M., K. Giles, J. Hamilton, P. Rebers, and F. Smith. 1956. Colorimetric method for the determination of sugars and related substances. Analytical Chemistry 28(3) p. 350-356 Engelhaupt, E., T. Bianchi, R. Wetzel, and M. Tarr. 2003. Photochemical transformations and bacteria l utilization of high molecula r weight dissolved organic carbon in a southern Louisiana tidal stream (Bayou Trepagnier). Biogeochemistry 62: 39-58 Ervin, G. N. and R. G. Wetzel. 2003. An ecological perspective of allelochemical interference in land Â–water interface communities. Plant and Soil 256 (1) 13-28 Espeland, E. and R. Wetzel. 2001. Comple xation, stabilization, and UV photolysis of extracellular and surface bound glucosidase and alkaline phosphatase: Implications for biofilm biota. Microb. Ecol. 42(4): 572-585 Farjalla, V., A. Anesio, S. Be rtilsson, and W. Graneli. 2001. Photochemical reactivity of aquatic macrophyte leachates: abiotic transformations and bacterial response. Aquat. Microb. Ecol. 24: 187-195 Fievre, A, T. Solouki, A. Minshall, and W. Cooper. 1997. High resolution Fourier transform ion cyclotron resonance mass spectrome try of humic and fulvic acids by laser desorption/ionization and electrospray ionization. Energy and Fuels 11: 554-560 Findlay, S., L. Carlough, M. Crocker, H. Gill, J. Meyer, and P. Smith. 1986. Bacterial growth on macrophyte leachate and fate of bacterial productio n. Limnol. Oceanogr. 31(6): 1335-1341 Findlay, S. and R. Sinsibaugh.(eds.) 2003. Aquatic Ecosystems: Interactivity of Dissolved Organic Matter. Academic Press, Boston. 512 pp. Fisher, S. and G. Likens. 1973. Energy fl ow in Bear Brook, New Hampshire: an integrative approach to stream ecosystem metabolism. Ecol. Monogr. 43:421-439 Fisher, M. and K.R. Reddy. 2001. Phosphorous flux from wetland soils affected by long term nutrient loading. J. Environ. Qual. 30(1): 261-271 Freeman, C., N. Ostle, and H. Kang. 2001. An en zymatic latch on the global carbon store. Nature 409: 149
155 Frimmel, F. H. 1998. Impact of light on the pr operties of aquatic natu ral organic matter. Environ. Inter. 24: 559-571 Fukushima, T., T. Ishibashi, and A. Imai. 2001. Chemical charac terization of DOM in Hiroshima Bay, Japan. Estuar. Coast. Shelf Sci. 53: 51-62 Fustec, E., G. Gilbert, D. Lacaze, and M. Escautier. 1988. Degradation of soft wood 14C lignocelluoses and its relation to the formation of humic substances in river and pond environments. Hydrobiologia 159: 169-176 Geller, A. 1986. Comparison of mechanisms enhancing biodegradability of refractory lake water constituents. Limnol. Oceanogr. 31(4) 755-764 Gjessing, E.T. and many others. 1998. Multi -method characterization of natural organic matter isolated from water: characterizati on of reverse osmosis-isolates from water of two semi identical dystrophic lake ba sins in Norway. Wat. Res. 32(10): 31083124 Griffiths, H. 1998. Stable Isotopes: In tegration of Biologi cal, Ecological, and Geochemical Processes. Bios Scientific Publishers, NY. 325 pp. Hagerman, A., M. Rice, and N. Ritchard. 1998. Mechanisms of protein precipitation for two tannins, pentagalloyl glucose and epicat echin (16) catechin (procyanidin). J. Agric. Food. Chem. 46: 2590-2595 Hansell, D.A. and C.A. Carlson (Eds.) 2002. Biogeochemistry of Marine Dissolved Organic Matter. Academic Press, Amsterdam. 774 pp. Hapeman, C., S. Bilboulin, B. Anderson, and A. Torrents. 1998. Structural influences of low molecular weight DOC mimics on photoly tic fate of atrizine. Env. Tox. Chem. 17(6): 975-981 Haslam, E. 1989. Plant Polyphenols. Camb ridge University Press, Cambridge.358 pp. Hayes, M.B. 1998. Humic substances: Prog ress towards more realistic concepts of structures. In: Humic Substances: Structur es, Properties, and Uses. G. Davies and E. Ghabbour (Eds.) Royal Society of Chemistry, MPG Books, UK. 296 pp. Herbert, B. and P. Bertsch. 1995. Characteri zation of dissolved colloidal organic matter in soil solution: a review. In Carbon Fo rms and Functions in Forest Soils. W. McFee and J.M. Kelly (Eds). Soil Sci. Soc. Am. Wisconsin. 213 pp. Hernes, P. and R. Benner. 2003. Photochemical and microbial degradation of dissolved lignin phenols: Implications for the fate of terrigenous dissolv ed organic matter in marine environments. J. Geophys. Res. Oceans 108 (C9): Art. No. 3291
156 Hessen, D, and L. Tranvik. 1998. Aquatic Humic Substances : Ecology and Biogeochemistry. (Ecological Studies Vol. 133). Springer-Verlog, New York pp. 386 Hoch, M. and D. Kirchman. 1995. Ammonium uptake by heterotrophic bacteria in the Delaware estuary and adjacent coastal waters. Limnol. Oceanogr. 40: 886-897 Hoffman, S., M. Shafer, C. Babiarz, and D. Armstrong. 2000. A cr itical evaluation of tangential flow ultra filtration for trace metal studies in freshwater systems. Environ. Sci. Technol. 34: 3420-3427 Hongave, D. 1994. Sunlight degradation of aquatic humic substances. Acta Hydrochimica et Hydrobio logica 22(3): 117-120 Hoppe, H. 1983. Significance of exoenzyma tic activities in the ecology of brackish water: Measurements by means of methyl umbelliferyl-substrates. Mar. Ecol. Prog. Ser. 11: 299-308 Hullar, M., B. Fry, B. Peterson. and R. Wright . 1996. Microbial util ization of estuarine dissolved organic carbon: a stable isotope tracer approach tested by mass balance. Appl. Environ. Microb. 62(7): 2489-2493 Imai, A., T. Fukushima, K. Matsushige , and Y. Kim. 2001. Fractionation and characterization of DOM in a shallow eutroph ic lake, its inflowing rivers, and other organic matter sources. Wat. Res. 35 (17): 4019-4028 Ivanoff, D.B., K.R. Reddy, and S. Robinson. 1998. Chemical fractionation of organic phosphorus in selected histosol s. Soil Sci. 163: 36-45 Jaffe, R., J. Boyer, X. Lu, N. Maie, C. Yang, N. Scully, and S. Mock. 2004. Source characterization of dissolved organic ma tter in a subtropical mangrove-dominated estuary by fluorescence analysis. Mar. Chem. 84: 195-210 Johnson, M., W. Huang, and W. Weber. 2001. A distributed reactivity model for sorption by soils and sediments 13. Simula ted diagenesis of natural sediment organic matter and its impact on sorpti on/desorption equilibria. Environ. Sci. Technol. 35: 1680-1687 Jones, R.I. 1992. The influence of humic subs tances on lacustrine food chains. Hydrobiol. 229: 73-91 Jorgensen, N., and R. Jensen. 1994. Microb ial fluxes of free monosaccharides and total carbohydrates in freshwater determined by PAD-HPLC. FEMS Microbiology Ecology 14: 79-93
157 Kaiser, K., G. Guggenberger, L. Haumaier, and W. Zech. 2001. Seasonal Variations In the chemical composition of DOM in organi c forest floor laye r leachates of old growth Scots Pines ( Pinus sylvestris ) amd European Beech ( Fagus sylvatica ) stands in northeastern Bavaria, Germany. Biogeochem. 55:103-143 Kalbitz, K., S. Geyer, and W. Geyer. 2000. A comparative characterization of dissolved organic matter by means of original aqueous samples and isolated humic substances. Chemosphere 40: 1305-1312 Kaplan,L., R. Larson, and T. Bott. 1980. Patterns of Dissolved organic carbon in transport. Limnol. Oceanogr. 25(6): 1034-1043 Khan, W. and R. G. Wetzel. 1999. Effects of microscale water level fluctuations and altered ultra violet radia tion of periphytic microbiota. Microb. Ecol. 38: 253-263 Kieber, R., X. Zhou, and K. Mopper. 1990. Formation of carbonyl compounds from UV induced photodegradation of humic substa nces in natural wa ters: fate of riverine carbon to the sea. Limnol. Oceanogr. 35: 1503-1515 King, S., C. Richardson, D. Urban, and E. Romanowicz. 2004. Spatial dependency of vegetation-environment linkages in an anthropogenically influenced wetland ecosystem. Ecosystems 7(1): 75-97 Kirchman, D. 2000. Microbial Ecology of the Oceans. p. 261-288. Wiley-Liss. New York. Kirchman, D. 2003. The contribution of m onomers and other low molecular weight compounds to the flux of dissolved organic material in aquatic ecosystems. In: Aquatic Ecosystems Interactivity of Dissolv ed Organic Matter. Findlay, S., and R. Sinsabaugh Eds. Academic Press, New York. 512 pp. Klavins, M. 1997. Genisis, sources, and sinks of aquatic humic substances. In:Aquatic Humic Substance: Characterizati on, Structure, and Genisis. p.105-123 Knicker, H. and P. Hatcher. 1997. Survival of protein in an organo-rich sediment: Possible encapsulation in organic matte r. Naturwissenschaften 84: 231-234 Kracht, O. and G. Gleixner. 2000. Isotope analysis of pyrolysis products from Sphagnum peat and dissol ved organic matter from bog water. Organic Geochemistry 31:645-654 Kroer, R. 1993. Bacterial growth efficiency on natural dissolved or ganic matter. Limnol. Oceanogr. 38(6): 1282-1290 Kronberg, L. 1999. Characterization of aqua tic humic substances. In Limnology of Humic Waters. J. Keskitalo and P. El oranta (Eds.) Backhuys Pub.Leiden.
158 Kubanek, J., M. Hay, P. Brown, N. Lindquist, and W. Fencial. 2001. Lignoid chemical defenses in the freshwater macrophyte Saururus cernuus . Chemoecology 11:1-8 Leff, L. and J. Meyer. 1991. Biological availability of dissolved organic carbon along the Ogeechee River. Limnol. Oceanogr. 36(2): 315-323 Lehto, K., J. Puhakka, and H. Lemmetyinen. 2003. Photodegradation products of PAHs in water and their amenability to biodegra dation. Polycyclic Aromatic Compounds 23: 401-416 Lewis, W. 2002. Yeild of n itrogen from minimally disturbe d watersheds of the United States. Biogeochem. 57: 375-385 Liu, D., P. Wong, and B. Dutka. 1973. Determ ination of carbohydrate in lake Sediment by a modified phenol-sulfuric acid method. Wat. Res. 7p. 741-746 Ljungdahl, L. and K. Eriksson. 1985. Ecology of microbial cellulose degradation. In: Advances in Microbial Ec ology. K. Marshall Ed. p. 237-299 Lobartini, J., K. Tan, L. Asmussen, R. Leona rd, D. Himmelsbach, and A. Gringle. 1991. Chemical and spectral differences in humi c matter from swamps, streams, and soils in the southeastern united states. Geoderma 49: 241-254 Mann, C. and R. G. Wetzel. 1996. Loading an d utilization of dissolved organic carbon from emergent macrophytes. Aquat. Bot. 53: 61-72 Mannino, A. and H. Harvey. 2000. Biochemical composition of particles and dissolved organic material along an estuarine gradie nt: Sources and implications for DOM reactivity. Limnol. Oceanogr. 45: 775-788 McDowell, W. and G.E. Likens. 1988. Origin, composition and flux of dissolved organic carbon in the Hubbard Brook Valley. Ecol. Monogr. 58: 177-195 McKnight, D.M. and G.R. Aiken. 1998. Sour ces and age of aquatic humus. In Aquatic Humic Substances. D. Hassen and L.J. Tranvik Eds. Ecological Studies 133. Springer-Verlog Germany. McKnight, D., E. Thurman, and R. Wershaw. 1985. Biogeochemistry of ThoreauÂ’s Bog, Concord, Massachusetts. Ecology 66: 1339-1352 Miller, W. and M. Moran. 1997. Interacti on of photochemical and microbial processes in the degradation of refractory dissolve d organic matter from a coastal marine environment. Limnol. Oceanogr. 42: 1317-1324 Mopper, K., X. Zhou, R. Keiber, D. Keib er, R. Sikorski, and R. Jones. 1991. Photochemical degradation of dissolved organic carbon and its impact on the oceanic carbon cycle. Nature 353:60-62
159 Moran, M. and R. Hodson. 1989. Formation and b acterial utilization of dissolved organic carbon derived from detrital lignocellulose. Limnol. Oceanogr. 34: 1034-1047 Moran, M. and R. Hodson. 1990. Bacter ial production on humic and non-humic components of dissolved organic car bon. Limnol. Oceanogr. 35: 1744-1756 Moran, M. and R. Hodson. 1994. Dissolved hum ic substances of vascular plant Origin in a coastal marine environment. Limnol. Oceanogr. 39: 762-771. Morris, D. and B. Hargraves. 1997. The role of photochemical degradation of dissolved organic carbon in regulating the UV transp arency of three lakes on the Poconos Plateau. Limnol. Oceanogr. 42(2) 239-249 Mulholland, P.J. 1981. Organic carbon flow in a stream swamp ecosystem. Ecol.Mono. 51: 307-322 Newman, S., J. Grace, and J. Koebel. 1996. Effects of nutrients and hydroperiod on Typha , Cladium , and Eleocharis : Implications for Evergl ades restoration. Ecol. Appl. 6(3): 774-783 Newman, S., K. Reddy, W. DeBusk, Y. Wa ng, G Shih, and M. Fisher. 1997. Spatial distribution of soil nutrients in a northern Everglades marsh: Water Conservation Area 1. Soil Sci. Soc. Am. J. 61(4): 1275-1283 Newman, S., J. Schuette, J. Grace, K. Rutche y, T. Fontaine, K. Reddy, and M. Pietrucha. 1998. Factors influencing cattail abundan ce in the northern Everglades. Aquat. Bot. 60(3): 265-280 Nguyen, L.M. 2000. Organic matter compos ition, microbial biomass and microbial activity in a gravel-bed constructed wetla nd treating farm dairy wastewaters. Ecol. Engin. 16: 199-221 Northup, R., R. Dahlgreen, and J. McColl. 1998. Polyphenols as regulators of plant Â– litter-soil interaction in northern CaliforniaÂ’s pygmy forest: A positive feedback? Biogeochemistry 42: 189-220 Obernosterer, I., P. Ruardij, and G. Herndl . 2001. Spatial and diurnal dynamics Of DOM fluorescence and peroxide and the photochemical oxygen demand of surface water DOM across the subtropical Atlan tic Ocean. Limnol. Oceanogr. 46(3) 632643 Opsahl, S. and R. Benner. 1999. Charac terization of carbohydrates during early diagenisis of five vascular plan t tissues. Organ. Geochem. 30: 83-94 Osburn, C., D. Morris, K. Thorn, and R. Moelle r. 2001. Chemical and optical changes in freshwater DOM exposed to solar radiation. Biogeochem. 54: 251-278
160 Parish, C.C. 1999. Determination of total lipi d, lipid classes, and fatty acids in aquatic samples. In: Lipids In Freshwater Ecosys tems. M.T. Arts and B. Wainman (Eds.) Springer-Verlag New York. Pempkowiak, J., M. Piotrowska-Syzpryt, a nd J. Kozuch. 1998. Rates of diagenic changes of humic substances in Baltic surface sediments. Environment International. 24: 589-594 Preston, C.M. 1996. The carbon chemistry of peat: Current knowledge and research needs for climate change issues. Proceeding of: Long-Term Dynamics and Contemporary Carbon Budget of Northern Peatlands. Natural Resources Canada. Preston, C.M., 2001. Carbon 13 solid-state NMR of soil organic matterusing the technique effectively. Can. J. Soil Sci. 81: 255-270 Price, M. L. and L. Butler. 1977. Rapi d visual estimation and spectrophotometric determination of tannin content of so rghum grain. J. Agriculture and Food Chemistry 25: 1268-1273 Pullin, M., S. Bertilsson, J. Goldstone, and B. Voelkner. 2004. Effects of sunlight and hydroxyl radical formation on dissolved organi c matter: Bacterial growth efficiency and production of carboxylic acids and other substances. Limnol. Oceanogr. 49(6) 2011-2022 Puupponen-Pimia, R., L. Nohynek, C. Meier, M. Kahkonen, M. Heinonen, A. Hopia, and k. Oksman-Caldentey. 2001. Antimicrobi al properties of phenolic compounds from berries. J. Appl. Microbiol. 90: 494-507 Qualls, R. and B.L. Haines. 1991. Geochemist ry of dissolved organic nutrients in water percolating through a forest ecosystem . Soil Sci. Soc. Am J. 55:1112-1123 Qualls, R., B. Haines, W. Swank, and S. Ty ler. 2002. Retention of dissolved Organic nutrients in a forested ecosy stem. Biogeochem. 61: 115-133 Qualls, R. and C. Richardson. 2003. Factors co ntrolling the concentration, export, and decomposition of dissolved organic nutrien ts in the Everglades of Florida. Biogeochem. 62: 197-229 Ran N, Agami M, Oron G. 2004. A pilot st udy of constructed wetlands using duckweed (Lemna gibba L.) for treatment of domestic primary effluent in Israel. Water Research 38 (9): 2241-2248 Reddy, K.R., T. Feijtel, and W.H. Patric k. 1984. Effect of soil redox conditions on microbial oxidation of organic matter. In : The Role of Organic Matter in Modern Agriculture. Y.Chen and Y. Avnimelec h, Eds. Martinus Nijhoff Publishers, Dordrecht. Pp 117-156
161 Reddy, K.R., and D.A. Graetz. 1988. Carbon and n itrogen Dynamics in wetland soils. In: The Ecology and Management of Wetlands, Vo l. I. DD Hooks et al. Eds. Timber Press, Portland. Pp. 307-318 Reddy, K.R., G.A. OÂ’Connor, and C. Schelske . 1999. Phosphorus Biogeochemistry in Subtropical Ecisystems. Lewis Publishers, Boca Raton. p. 707 Reitner, B., G. Herndl, and A. Herzig. 1997. Role of UV-B radiation on photochemical and microbial oxygen consumption in a humic-rich lake. Limnol. Oceanogr. 42(5) 950-960 Rich, J., M. Gosselin, E. Sherr, B. She rr and D. Kirchman. 1997. High bacterial production, uptake and concentrations of di ssolved organic matter in the central Arctic Ocean. Deep S ea Research II 44: 1645-1663 Richardson, C., G. Ferrell, P. Vaithiyanathan. 1999. Nutrient effects on stand structure, resorption efficiency, and secondary com pounds in Everglades sawgrass. Ecology 80(7) 2182-2192 Richardson, S., J. Simmons, G. Rice. 2002. Disinfection byp roducts: the next generation. Environ. Sci. Technol. 36: 1198-1207 Roberts, J., and A. Rowland. 1998. Cellulo se fractionation in decomposition studies using detergent fiber pretreatment met hods. Commun. Soil Sci. Plant Anal. 25: 269-277 Rosenstock, B., and M. Simon. 2001. Sources and sinks of dissolved free amino acids and protein in a large and deep meso trophic lake. Limnol . Oceanogr. 46: 644-654 Rousseau 1999. Refractory organic matter in sediments from the northwest African upwelling system: Abundance, chemical structure, and origin. Organic Geochemistry 30: 83-99 Ryan, D., C. Shih, and D. OÂ’Connor. 1996. Fluorescence spectroscopic studies of Alfulvic acid complexation in acidic soluti ons. In Humic and Fulvic Acids: Isolation, Structure, and Environmental Ro le. J. Gafney, N. Marley, and S. Clark (Eds.) ACS Symposium Series 651 Sachse, A., D. Babenzein, G. Ginzel, J. Gelbrecht, and C. Steinberg. 2001. Characterization of dissolved organi c carbon (DOC) in a dystrophic lake and adjacent fen. Biogeochem. 54: 279-296 Salisbury, F. and C. Ross. 1992. Plant Physiology 4th Ed. Wadsworth Inc., Belmont, CA. pp. 682 Santos, E and A. Duarte. 1998. The influe nce of pulp mill effluents on the composition of the humic fraction of aquatic organi c matter. Wat. Res. 32(3): 597-608
162 Schiff, S., R. Avarena, S. trumbore, M. Hinton, and P. Dillon. 1997. Export of DOC from forested catchments on the Precambria n Sheild of Central Ontario: clues from 13C and 14C. Biogeochemistry 36:43-67 Schindler, D. and P. Curtis. 1997. The role of DOC in protecting freshwaters subjected to climatic warming and acidification from UV exposure. Biogeochem. 36: 1-8 Schindler, D., P. Curtis, S. Bayley, B. Pa rker, K. Beaty, M. Stainton. 1997. Climate induced changes in the DOC budgets of bor eal lakes. Biogeochemistry 36: 9-28 Schwarzenbach, R., P. Gschwend, and D. Imboden. 2003. Environmental Organic Chemistry, 2nd ed. Wiley-Interscience, New Jersey. p. 1313 Scully,N, and D. Lean. 1994. The attenuation of ultraviolet radiation in temperate lakes. Arch. Hydrobiol. Be ih. 43: 135-144 Scully, N., N. Maie, S. Dailey, J. Boyer, R. Jones, and R. Jaffe. 2004. Early diagenesis of plant derived dissolved organic matter along a wetland mangrove, estuary ecotone. Limnol. Oceanogr. 49(5): 1667-1678 Sharpless, C., D. Seibold, and K. Linden. 2003. Nitrate photosensiti zed degradation of atrizine during UV water treatment. Aquatic Sciences 65(4) 359-366 Stevenson, F.J. 1994. Humus Chemistry: Ge nesis, Composition, and Reactions. Wiley, New York. Stewart, A. and R. Wetzel. 1981. Asymme trical relationships between absorbance, fluorescence, and dissolved organic carbon. Limnol. Oceanogr. 26(3): 590-597 Strobel, B., H. Hansen, O. Borggaard, M. Andersen, and K. Raulund Rasmussen. 2001. Composition and reactivity of DOC in forest floor soil solutions in relation to tree species and soil type. Biogeochem. 56: 1-26 Strober, Q., R. Jones, and D. Schmidt. 1995. Ultra trace level mercury in the Everglades ecosystem: a multi-media canal pilot study. Water Air Soil Pollut. 80: 991-1001 Sun, L., E. Perdue, J.L. Meyer, and J. We is. 1997. Use of elemental composition to predict bioavailability of dissolved orga nic matter in a Georgia river. Limnol. Oceanogr. 42: 714-721 Sutton, D. L. and K. Portier. 1991. Influence of spikerush plants on growth and nutrient content of Hydrilla . J. Aquat. Plant Manag. 29: 6-11 Tadanier, C., D. Berry, and W. Knocke. 2000. Dissolved organic matter apparent molecular weight distribution and number average apparent molecular weight by batch ultrafiltration. Envi ron. Sci. Tech. 34(11): 2348-2353
163 Tan, K. 2003. Humic matter in soil and the envi ronment: principles and controversies. Marcel Dekker, New York. pp. 386 Thomsen, M., P. Lassen, S. Dobel, P. Hansen, L. Carlsen, B. Mogensen. 2002. Characterization of humic materials of different origin: a multivariate approach for quantifying the latent properties of DOM. Chemosphere 49: 1327-1337 Thurman, E.M. 1985. Organic Geochemistry of Natural Waters. Dr. W. Junk, Dordrecht, Netherlands. Tranvik, L. 1992. Allochthonous dissolved organic matter as an energy source for pelagic bacteria and the concept of th e microbial loop. Hydrobiol. 229: 107-114 Tranvik, L., and S. Bertilsson. 2001. Cont rasting effects of solar UV radiation on dissolved organic matter sources for bact erial growth. Ecol. Lett. 4: 458-463 Tranvik, L. and N. Jorgensen. 1995. Colloidal and dissolved organic matter in lake watercarbohydrate and amino acid compos ition, and ability to support bacterial growth. Biogeochem. 30: 77-97 Tranvik, L., and S. Kokalj. 1998. Decrease d biodegradability of algal DOC due to interactive effects of UV ra diation and humic matter. Aquat. Microb. Ecol. 14: 301-307 Twilley, R., M. Brinson, and G. Davis. 1977. Phosphorus adsorption, translocation, and secretion in Nuphar luteum . Limnol. Oceanogr. 22: 1022-1032 United States Environmental Protection Agency (1993) Manual of Methods for Chemical Analysis of Water and Wastes. EPA Publication 600/R-93/100 Washington D.C. Vahatalo, A, and R. Wetzel. 2004. Photoc hemical and microbial decomposition of chromophoric dissolved organic matter during long (months-years) exposures. Marine Chemistry 89: 313-326 Valentine, R., and R. Zepp. 1993. Fo rmation of carbon monoxide from the photodegradation of terrestria l dissolved organic carbon in natural waters. Environ. Sci. Technol. 27: 409-412 Van Soest, J.P. 1970. Forage and Fiber Analysis Handbook No. 379. USDA. Washington, D.C. Villar, C., L. de Cabo, P. Vaithiyanathan, a nd C. Bonetto. 2001. Litter decomposition of emergent macrophytes in a floodplain marsh of the lower Parana River. Aquat. Bot. 70: 105-116 Volk, C., C. Volk, and L. Kaplan. 1997. Chemical composition of biodegradable dissolved organic matter in stream water. Limnol. Oceanogr. 42 (1) 39-44
164 Waiser, M. and R.D. Robarts. 2000. Ch anges in composition and reactivity of allochthanous DOM in a prairie saline lake. Limnol. Oceanogr. 45(4): 763-774 Wang, W., M. Tarr, T. Bianchi, and E. Engelhaupt. 2000. Ammonium production from aquatic humic and colloidal matte r. Aquatic Geochemistry 6: 275-292 Wang,Y., Y. Hseih, W. Landing, Y. Choi, V. Salters, and D. Campbell. 2002. Chemical and carbon isotopic evidence for the source a nd fate of dissolved organic matter in the northern Everglades. Biogeochemistry 61: 269-289 Waterman, P. and S. Mole. 1994. Analysis of Phenolic Plant Metabolites. Blackwell Scientific Publications, Oxford. 238 p. Weinberg, H. 1999. Disinfecti on byproducts in drinking water: the analytical challenge. Anal. Chem. 71: 801-809 Weisner, S. and S. Miao. 2004. Use of morphological variability in Cladium jamaicense and Typha domingensis to understand vegetation changes in an Everglades marsh. Aquatic Botany 78: 319-335 Wershaw, R. L., 2004. Evaluation of concep tual models of natural organic matter (humus) from consideration of the ch emical and biological processes of humification. Scientific Investigations Report 2004-5121. U.S. Department of Interior Wetzel, R. G. 1984. Detrital dissolved a nd particulate organic carbon functions in aquatic ecosystems. Bull. Mar. Sci. 35(3): 503-509 Wetzel, R.G. 1992. Gradient dominated ecosyst ems: sources and regulatory functions of dissolved organic matter in freshwater ecosystems. Hydrobiol. 229: 181-198 Wetzel, R.G. 1993. Humic compounds from we tlands: Complexation, inactivation, and reactivation of surface bound a nd exocellular enzymes. Verh. Internat. Verein. Limnol. 25: 122-128 Wetzel, R.G. 2001. Limnology: Lake and River Ecosystems. 3rd Ed. Academic Press. San Diego. 1005 pp. Wetzel, R.G., P. Hatcher, and T. Bianch i. 1995. Natural photo lysis by ultraviolet irradiance of recalcitrant dissolved organic matter to simple substrates for rapid bacterial metabolism. Limnol. Oceanogr. 40: 1369-1380 Wetzel, R.G. and B. Manny. 1977. Seasonal cha nges in particulate and dissolved organic carbon and nitrogen in a hardwater stream. Arch. Hydrobiol. 80:20-39
165 Whitehead, R.F., S. Mora, S. Demers, M. Go sselin, P. Monfort. and B. Mostajir. 2000. Interactions of ultravio let-B radiation, mixing, and biological activity on photobleaching of natural chromophoric di ssolved organic matter: A mesocosm study. Limnol. Oceanogr. 45(2): 278-291 Williams, P. 2000. Heterotrophic bacteria and the dynamics of dissolved organic material. In: Microbial Ecology of the O ceans. D. Kirchman Ed. Wiley Liss, New York. pp. 153-200 Wu, W., P. Chadik, W. Davis, J. Delfino, and D. Powell. 2000. The Effect of structural characteristics of humic substances on disinfection by-product formation in chlorination. In: Natural Organi c Matter and Disinfection By-Products Characterization and Control in Drinking Water. S. Barrett Ed. American Chemical Society. Young K., P. Maurice, K. Docherty, S. Br idgham. 2004. Bacterial degradation of dissolved organic matter from two northe rn Michigan streams. Geomicrobiology Journal 21 (8): 521-528 Zafiriou, O.C. 2002. Sunburnt organic matter: Biogeochemistry of light altered substrates. Limnology and Ocean ography Bulletin. 11(4) 69-74 Zegouagh, Y., S. Derenne, C. Largeau, P. Bertrand, M. Sicre, A. Saliot, and B. Zepp, R.G. 1999. Environmental photo processes involving natural organic matter. In: Frimmeland Christman (Eds) Humic substanc es and their Role in the environment. (pp 193-214) J. Wiley and Sons. New York, NY. Zepp, R.G. 1988. Environmental photoprocesses involving natural organic matter. In: Frimmeland Christman (Eds) Humic substanc es and their Role in the environment. (pp 193-214) J. Wiley and Sons. New York, NY Zepp, R.G., T. Callaghan, and D. Erickson. 2003. Interactive effects of ozone depletion and climate change on biogeochemical cycles . Photochem. Photobiol. Sci. 2(1): 51-61 Zou, Y. and R. Jones. 1997. Photochemistry of natural dissolved organic matter in lake and wetland waters: production of carbon monoxide. Wat. Res.31(4) 850-858
166 BIOGRAPHICAL SKETCH Todd Zachary Osborne was born October 1, 1973, at Andrews Air Force Base, Maryland, to Zack and Edna Osborne. Todd spen t the next 31 years of his life in, on, or under the water. Whether surfing, fishi ng, snorkeling, boating, or just running amok, water is the focus of his recreation, passi on, and academic pursuits. Growing up in a military family, he had the opportunity to se e the world at a young age. They moved often, but retreated to the family farm in th e Blue Ridge Mountains of Virginia when time allowed, and it was on this farm that Todd l earned to appreciate the natural world. With the tireless support of his family and friends, and the inspiration of a few assorted teachers and charac ters, Todd has Â“tested the wate rsÂ” on four continents and a handful of islands in the Ca ribbean and south Pacific. With passion for the protection of aquatic ecosystems and the environment in general, Todd has managed to make a career of what he loves to do; and fortunately, he looks forward to going to the Â“officeÂ”.