<%BANNER%>

Influences of Nutrient Loading, Vegetative Habitats and Simulated Drought on Microbial Enzyme Activities in the Everglades


PAGE 1

INFLUENCES OF NUTRIENT LOADING, VEGETATIVE HABITATS AND SIMULATED DROU GHT ON MICROBIAL ENZYME ACTIVITITES IN THE EVERGLADES By CHRISTOPHER RYAN PENTON A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

ii ACKNOWLEDGMENTS This research was supported by funds from the South Florida Water Management District. I would like to thank Dr. Ramesh Reddy for his support and for offering this unorthodox opportunity. I am indebted to Dr. Susan Newman for her patience, support, and guidance through the learning curve. She taught me the rigors of scholarly pursuit, pushed me to excel, and was always there to debate fine details with red ink. A large portion of thanks goes to my wife Stephanie, who always supported me through my late night sessions even when she was devoting mo st of her time to medical school. Her love and support guided me through the rough times. Fi nally, I would also lik e to give thanks to my family whose love, devotion and encouragement throughout my life have always motivated me to strive for knowledge a nd excel in whatever path I choose.

PAGE 3

iii TABLE OF CONTENTS page ACKNOWLEDGMENTS…………………………………………………………….......ii LIST OF TABLES………………………………………………………………………...v LIST OF FIGURES……………………………………………………………………...vii ABSTRACT…………………………………………………………………………….viii CHAPTERS 1 INTRODUCTION………………………………………………………...1 Background………………………………………………………..1 Statement of the Problem…………………………………….........4 Objectives and Scope of Research………………………………...5 2 EXPERIMENTAL CONSIDERATIONS IN DETERMINING OPTIMUM ENZYME ASSAY CONDITIONS IN WETLAND SYSTEMS…………………………………………...7 Introduction………………………………………………………..7 Materials and Methods…………………………………………...10 Results……………………………………………………………13 Discussion………………………………………………………..19 Conclusions………………………………………………………22 3 EFFECTS OF HABITAT DIFFERENTIATION ON MICROBIAL ENZYME ACTIVITIES IN THE EVERGLADES……………...23 Introduction………………………………………………………23 Materials and Methods…………………………………………...28 Results……………………………………………………………32 Discussion………………………………………………………..45 Conclusions………………………………………………………50

PAGE 4

iv 4 NUTRIENT LOADING EFFECTS ON BIOGEOCHEMICAL AND MICROBIAL ENZYME DYNAMICS IN THE EVERGLADES………………………………………………….52 Introduction……………………………………………………....52 Materials and Methods…………………………………………...55 Results……………………………………………………………64 Discussion………………………………………………………..81 Conclusions………………………………………………………91 5 MICROBIAL ENZYME RESPO NSES TO DECREASED WATER LEVELS IN EVERGLADES PEAT AND MARL SEDIMENTS……………………………………...94 Introduction………………………………………………………94 Materials and Methods…………………………………………...96 Results…………………………………………………………..102 Discussion………………………………………………………114 Conclusions……………………………………………………..117 6 SUMMARY AND CONCLUSIONS…………………………………..120 REFERENCES…………………………………………………………………………125 BIOGRAPHICAL SKETCH…………………………………………………………...136

PAGE 5

v LIST OF TABLES Table page 3-1 Mean chemical properties of the soil and benthic layers………………………..35 3-2 Mean enzyme activities of the soil and benthic layers……………………...…...37 3-3 Correlation coefficients for be nthic enzyme and nutrient data………………….37 3-4 Correlation coefficients for soil enzyme and nutrient data……………………...39 3-5 Overall benthic and soil norma lized total enzy me activities…………………….39 3-6 Enzyme ratios……………………………………………………………………42 3-7 Correlation coefficients for soil and benthic layer enzyme ratio components and nutrient data…………………………………………………………………42 4-1 Benthic nutrient, enzyme, and ratio para meters in P enriched and reference sites in LNWR and WCA-2A…………………………………………………………66 4-2 Benthic nutrient, enzyme, and ratio para meters in P enriched and reference sites in WCA-3A and ENP-TS………………………………………………………..67 4-3 Benthic layer nutrient a nd enzyme correlation coefficients……………………..69 4-4 Soil layer nutrient, enzyme, and ratio para meters in P enriched and reference sites in LNWR and WCA-2A…………………………………………………………73 4-5 Soil layer nutrient, enzyme, and ratio para meters in P enriched and reference sites in WCA-3A and ENP-TS………………………………………………………..74 4-6 Soil layer nutrient a nd enzyme correlation coefficients…………………………76 4-7 Average benthic and soil CRR values…………………………………………...78 4-8 Impact index………………………………………………………….…………..80 5-1 Mean enzyme activities………………………………………………………...103

PAGE 6

vi 5-2 Mean enzyme activities…………………………………………………….…..104 5-3 Mean Enzyme Index of Carbon Quality (EICQ)…………...………………….107 5-4 Cumulative enzyme activity slopes (CES)……………………………….…….108

PAGE 7

vii LIST OF FIGURES Figure page 2-1 Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments and (c) varying MUFcellobioside substrat e concentrations..14 2-2 Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments and (c) varying MUFphosphatase substrate concentrations..15 2-3 Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments and (c) varying MUFglucoside substrat e concentrations..18 2-4 Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments and (c) varying L-Le ucine amidomethylcoumarin substrate concentrations....20 3-1 Radar plots....40 3-2 Benthic microbial N and P resource allo cation in terms of C mineralization...44 3-3 Soil microbial N and P resource alloca tion in terms of C mineralization.....44 3-4 Bubble graph comparison of (a) be nthic and (b) soil enzyme ratio components Ecell/En, Ecell/Ep and Ecell/Eox......47 4-1 Graphs showing spikes in enzyme activ ities at the transitional sites 4-2 Benthic layer bubble plot of Ece ll/Ep vs. Ecell/En with Ecell/Eox.......83 4-3 Benthic plot of Log Ecell/En vs Log Ecell/Ep... 5-1 Benthic cumulative enzyme activity (CEA)110 5-2 0 to -10 cm cumulative enzyme activity (CEA)..111 5-3 Mean lignin, cellulose, and nutrient contents of ENP-TS and WCA-3A cores.3

PAGE 8

viii Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INFLUENCES OF NUTRIENT LOADING, VEGETATIVE HABITATS AND SIMULATED DROUGHT ON MICROBIAL ENZYME ACTIVITI ES IN THE EVERGLADES By Christopher Ryan Penton August, 2004 Chair: Dr. K.R. Reddy Cochair: Dr. Susan Newman Major Department: Soil and Water Science Wetlands frequently function as a long term carbon (C) sink as organic matter accumulates in response to reduced decompositi on rates. In general, enzyme catalyzed reactions are considered the rate-limiting step in organic matter degradation. Nutrient status, vegetative community and hydrological regimes significantly influence microbial enzyme activities. This thesis exams the e ffects of nutrients, vegetative communities and water level drawdown on enzymes involved in C, nitrogen (N), phosphorus (P) and lignin mineralization within four hydrologically di stinct areas of the Florida Everglades. Examining enzyme activities along distin ct phosphorus gradients, this study demonstrates that P enriched sites exhi bit lower microbially perceived N and P limitations on C mineralization in addition to enhanced cellulose decomposition rates. Phosphorus loading resulted in a decreased microbial mobilization of resources for P mineralization which resulted in a greater en ergetic allocation for C mineralization. Nitrogen appears to become less limiting to C mineralization in the enriched sites within

PAGE 9

ix Everglades National Park. A simple two component model, incorporating total phosphorus and the relationship between the en zymes involved in C and P mineralization accounted for 62% of the variability in cellulose decomposition rates. Vegetative composition was found to si gnificantly alter th e allocation of enzymatic resources due to varying substrate complexities. Carbon mineralization in the open water was significantly less influences by lignin than the sawgrass habitats. An index relating hydrolytic and oxidative enzy mes, the Enzyme Index of Carbon Quality, was significantly greater in the open water ha bitats. This enzymatic index suggests more favorable C mineralization conditions with in the open water communities and provides insight into the development of the sloughsawgrass ridge topography of the Everglades. Laboratory water level drawdown was found to effect the microbial allocation of enzymatic resources in intact Everglades soil cores. Peat dominated sediments exhibited elevated enzyme activities and a larger response to lower water levels over time. Enzymatic N mineralization incr eased significantly in the surficial soil horizons as a response to water level drawdown. The resu lts indicate that extended periods of low water may result in enhanced C and N mi neralization of Everglades sediments. The results of these studies indicate that microbial enzyme activities are responsive to changes in nutrient conditi ons, vegetation and water levels. The extrapolation of these enzyme activities to potential decompositi on based on perceived C qualities connects the microbial condition to eco system level processes. Changes to the Everglades ecosystem as a result of mana gement practices can thus be initially recognized through alterations in the highl y responsive microbial component that functions in nutrient cycling and serves as the basis for the trophic food chain.

PAGE 10

1 CHAPTER 1 INTRODUCTION Background Wetlands often have decreased O2 availability as a result of decreased O2 diffusion rates through water, resulting in decreas ed overall supply (P onnanmperuma, 1972). Additionally, increased O2 demand, due to generally high carbon (C) availability, results in overall decreases in decom position rates in wetlands (Red dy and D’Angelo, 1994). An important and significant role of wetlands in the global ecosystem is C sequestration. The accumulation of organic matter and thus the role in peat development and global C sequestration may be affected by changes in ambient nutri ent conditions (Wetzel, 1991; Newman et al., 2003), vegetative composition (DeBusk and Reddy, 1998; Berg, 2000; Fioretto et al., 2000; Kourtev et al., 2002a ), and differing hydrological regimes (Volk, 1973; Elliott et al., 1984; DeBusk, 1996). Th e mineralization of organic matter by the resident heterotrophic microbial communities w ithin a system plays a crucial role in the cycling of C and exerts an influence on the overall energy flow within a system (Elliot et al., 1984). The activity of the he terotrophic bacteria and fungi whose activity is related to the biochemical structure, nutri ent content, and quantity of av ailable substrates produces enzymes that are often considered the rate limiting steps in decomposition (Chr st and Rai, 1993). The synthesis and activities of the enzymes involved in organic nutrient mineralization are most regulated by the induction of macrophytic and macromolecular substrates within soils (Burns, 1986; Nausch et al., 1998) as well as the intrinsic factors

PAGE 11

2 of plant litter (Linkins et al., 1990a; Carreiro et al., 2000; K ourtev et al., 2002a). Lignin content has been specifically correlated with decomposition rates (Meentemeyer, 1978; Mellilo et al., 1989), especially when nitrogen (N) is readil y available (Mellilo et al., 1982; Taylor et al., 1989). Nitrogen content, presented as C:N or lignin:N are also often used as predictors of litter decomposition rates (Mellilo et al., 1982; Taylor et al., 1989; Sinsabaugh et al., 1993; DeBusk and Re ddy, 1998; Carreiro et al., 2000) and may influence the decomposition of large molecula r weight organic matter (Sinsabaugh et al., 1993; Carreiro et al., 2000). Phosphorus (P) avai lability has also been shown to control decomposition rates in generally P-lim ited systems (Newman et al., 2001). Decomposition of organic matter is a co mmunity level process that involves specific interactions within a microbial co nsortium (Sinsabaugh et al., 1991). The advantage of enzyme assays is that they pr esent specific information on one process in a complex community. The microbial degradatio n of particulate organic matter, such as plant litter, has been shown to be most influenced by the enzymes involved in lignocellulose degradat ion, P cycling, and N cycling (S insabaugh et al., 1991; Sinsabaugh and Moorhead, 1996), which are often considered the rate limiting steps in degradation (Chr st and Rai, 1993). Strong relationshi ps have been established between lignocellulose-degrading enzymes and litter mass loss rates among vary ing litter qualities (Sinsabaugh et al, 1992a). Enzyme activities have the potential to affect all major wetland functions where decomposition is low. Peat accumulation is dependent upon a lower rate enzymatic activity resulting in C storage and inorganic nutrients rema in sequestered within the poorly degraded peat matrix when decomposition is low. This impairment of nutrient

PAGE 12

3 cycling by the microbial consortia causes i norganic nutrients to accumulate and to be retained within the wetland (Freeman et al ., 1996). However, the interpretation and relation of individual enzyme activities to higher trophic proc esses and observations is often a difficult and tedious process that freque ntly results in vague relationships to soil and microbial parameters in the system (Marsden and Gray, 1986). A current strategy involves a resource a llocation rationale in interpreting enzyme activities and the MARCIE (M icrobial Allocation of Resources Among Community Indicator Enzymes) model for exposi ng linkages between individual enzymes (Sinsabaugh and Moorhead, 1994; Sinsaba ugh and Findlay, 1995; Sinsabaugh et al., 1996, 2002). The model indicates that the expres sion of enzymes is tied to environmental nutrient availabilities and that the distribution of enzyme activ ities can be interpreted as a resource allocation strategy (Sinsabaugh et al., 2002). Underlying th is concept is the relationship between ligno cellulose degradation a nd environmental N and P concentrations. Components of this model have been linked to bacterial productivity, microbial biomass, and particulate orga nic carbon turnover times (Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997). Specifically, the Enzyme Index of Carbon Quality (EICQ), which relates the activities of hydrolytic enzymes to those involved in lignin degradation, has been shown to decrease as decomposition proceeds, which is reflective of lower substrate quality (Kourtev et al., 2002b). This index has also been correlated with microbial biomass, productivity and negatively correla ted with particulate organic carbon (POC) turnover time (Sinsa baugh and Findlay, 1995). Furthermore, enzymes are divided into functional units base d on C, N and P mineralization (Ecell, En, Ep) and lignin degradation (Eox). Enzyme ratios using these compartmentalized units

PAGE 13

4 such as Ecell/Ep and Ecell/En, which reflect apparent P and N control on C mineralization, respectively, have been correlated with bact erial productivity (Sinsabaugh and Findlay, 1995; Sinsabaugh and Moorhead, 1994). Statement of the Problem The Florida Everglades was once a vast peat wetland that encompassed 10,000 km2 and stretched from Lake Okeechobee southward to the shores of Florida Bay. In the late 1800’s and early 1900’s large areas of the origin al landscape was altered for agricultural and urban development. Water Conservation Areas (WCAs) were cr eated as a result of compartmentalization by dikes, levees, a nd water control structures. The remnant Everglades consists mainly of four manage d, hydrologically distinct areas: Arthur R. Marshall Loxahatchee National Wildlif e Refuge (LNWR), WCA-2, WCA-3, and Everglades National Park (ENP). Intensive agricultural development in the northern periphery has resulted in the establishment of P and, to a le sser extent, N gradients within these four hydrologic component s of the oligotrophic Everglades. As a consequence, shifts in macrophyte species compositi on (Davis, 1943, 1991; Vaithiyanaithan and Richardson, 1999), increases in net primary production (NPP) (Davis, 1991) and peat accumulation (Reddy et al., 1993), loss or taxonomic shifts of native periphyton assemblages (McCormick and O’Dell, 1996; McCormick et al., 1996), increases in microbial activity and biomass (White a nd Reddy, 2000), and other ecological changes have been observed in Everglades areas re ceiving nutrient inputs from canal water with total P (TP) concentrations as much as 10 to 30 fold higher than background conditions (McCormick et al., 1996). Altered surface flow as a result of compartmentalization and hydrological management also affected the C cycling with in the Everglades. The net area covered by

PAGE 14

5 the deeper sloughs or open water habitats ha s diminished, being replaced by shallower, monotypic sawgrass stands. This affected the natural landscape of the Everglades which was dominated by a slough, slough-ridge topogr aphy. The development and maintenance of this topographical relief is not well unders tood and has been attr ibuted to particulate transport and differential decomposition. Water level fluctuations as a conseque nce of natural seas onal variations or hydrologic management have the potential to dr astically affect the ability of a wetland system, such as the Everglades, to sequester C. Previous studies have shown increases in CO2 flux with lowered water table depths (V olk, 1973) and the highest total C flux under drained conditions in Everglad es peat sediments (DeBusk, 1996). The mineralization of soil organic N sources has been found to be gr eater in aerobic than anaerobic conditions (Reddy and Patrick, 1984; McLatchey and Re ddy, 1998), to increase after the drying of wet soils (Cabrera 1993, Bridgham et al., 1998), and specifically, to be 2-3 times greater (White, 1999; Venterlink et al., 2002) in drai ned wetland soils. The impact of water fluctuations thus has a potentially large role in the nutrient cycli ng and retention of the Everglades. Objectives and Scope of Research The overall goal of this study was to de termine the influences of different vegetative habitats, nutrient loading and water table draw down on microbial enzyme activities in the Everglades. Chapters 2 th rough 5 address more sp ecific objectives: 1.) Develop an appropriate enzyme experi mental method for performing assays in wetland systems. 2.) Determine the effects of vegetativ e habitats on microbial enzyme activities in Water Conservation Area 3A of the Everglades.

PAGE 15

6 2.) Determine the relationships betwee n nutrient conditions and microbial enzyme activities among all four hy drologic units of the Everglades. 3.) Determine the validity of different enzyme models in predicting potential decomposition among different litter qua lities and in varying nutrient conditions. 4.) Determine the effects of a simulated drought on enzyme activities in WCA3A and ENP-Taylor Slough (ENP-TS) cores. The overall hypotheses for this study were: (i) Deeper water slough habitats exhibit accelerated potential decomposition due to subs trate quality differenc es; (ii) Nutrient loading generally enhances microbial enzyme activities; and (iii) Decreased water levels result in an enhancement of enzyme activit ies, suggesting greater potential decomposition based on perceived soil quality.

PAGE 16

7 CHAPTER 2 EXPERIMENTAL CONSIDERATIONS IN DETERMINING OPTIMUM ENZYME ASSAY CONDITIONS IN WETLAND SYSTEMS Introduction Fluorogenic substrates have been widely adopted in investigations of enzyme activities in a variety of systems such as lakes, grasslands, wetlands, streams, groundwater and oceans (Chrst, 1989; Freem an et al., 1995; Miettinen et al., 1996; Sinsabaugh and Foreman, 2001; Debosz et al., 1999; Mayr et al., 1999, Shackle et al., 2000; Wittmann et al., 2000; Burns and Ryder, 2001; Newman et al., 2003; Saiya-Cork et al., 2002; Wittmann et al., 2004). These com pounds have become popular due to their production of fluorescent compounds, which e xhibit less interference by highly colored phenolic compounds than colorimetric substrates (Sinsabaugh et al., 1991; Freeman et al., 1995). Methylumbelliferyl (MUF) and amidomethylcoumarin (AMC) substrates are among the most widely adopted fluorogenic substrates. Fundamentally different methodological ap proaches have been adopted in the literature. The most recent approaches include the use of microtiter plates (Marx et al., 2001). Standardized approaches use s ubstrate saturating conditions, standard temperature, and an assay pH that maximi zes the fluorescence of the fluorochrome. Enzymes have differing optimum pH levels wh ich have influenced the use of different pH conditions between enzymes (Parham a nd Deng, 2000). The environmental approach utilizes substrate concentrations similar to the local environment w ith assay temperature and pH approximating field conditions. A comp romise between these approaches is to

PAGE 17

8 use substrate saturating conditions with a ssay temperature and pH approximating field conditions. However, temp eratures as high as 45 C have been used (Sinsabaugh et al., 1991). Sample alkalinization prior to fluores cence measurement is al so often performed to maximize the fluorescence intens ity (Freeman et al., 1995). Enzyme activity is generally more reso lvable with substrate saturation and the resultant generally linear function of enzyme activity allows for shorter incubation times and less error. Strong correlations in the l iterature between poten tial enzyme activities and nutrient conditions or corr elated processes allow enzyme activities to reflect the magnitude or process relationships among samp les or between systems (Cembella et al., 1984; Jansson et al., 1988; Foreman et al., 1998). The methodological heterogeneity of enzyme assays is a reflection of ecological diversity and localized microbial community structures among studies. For example, substrate incubation times in the literature ra nge from 10 minutes to 3 days (Miettinen et al., 1996; Foreman et al., 1998; Debosz et al ., 1999; Burns and Ryder, 2001; Kourtev et al., 2002b). Assay incubation times are generall y based on the ability to measure within the linear portion of the time curve while limiting the opportunity for microbial growth (Freeman et al., 1995, 1996). Reported MUF a nd AMC substrate concentrations are also highly variable between studies and ranges from 1 to 500 M in soil and water column enzyme studies (Chrst, 1989; Sinsabaugh et al., 1997; Debosz et al., 1999; Burns and Ryder, 2001; Shackle et al., 2000; Sinsaba ugh and Foreman, 2001; Saiya-Cork et al., 2002; Wittmann et al., 2004). An emphasis of setting substrate concentrations at saturating conditions has been documented in several studies (Chrst, 1989; Sinsabaugh and Findlay, 1995; Burns and Ryder, 2001).

PAGE 18

9 The majority of enzyme studies are perfor med on soil, water, and litter samples. However, the use of enzymes in wetland systems is comparatively rare. The combination of soil and water matrices introduces additi onal issues in the development of enzyme assays. Fluorescence interference, due to th e rather large accumulation of organic matter in various stages of decay in wetland syst ems as well as the accumulation of phenolic compounds can play a major role in enzyme activity determination. Therefore it is essential to calculate a quench adjustment for each sample by placing standards in sample solutions and relate the fluoresences to th e standard curve so that samples can be accurately compared (Freeman et al., 1996; Ma rx et al., 2001). However, this does not eliminate other issues that can introduce pr oblems between samples. The presence of inhibitors, alternative reaction paths and competing substrates can also affect the results of an assay (Sinsabaugh et al., 1991). Enzyme activity is usually expressed as the net change in fluorescence from an initial to a final measurement over time to determine the substrate conversion rate. Laboratory replicate errors can be quite large in soil enzyme assays due to difference in particle distribution as well as sample heterogeneity on a minute scale, even in homogenized samples. What is not clear from the literature is the problem associated with relying on these endpoint values. The use of net en zyme activity changes without observing changes in fluorescence over smaller increments of the total incubation time can result in larger replicate errors. The us e of automated spectrofluorimeters allows the investigator to observe the en zyme kinetics over time and thus identify any replicates from analysis that produce both outlying net acti vity and erratic kinetics over time, thus reducing replicate error a nd increasing precision.

PAGE 19

10 The goal of this study was to develop an enzyme assay protocol for Everglades soils to be utilized in further investigations of enzyme activities and their relation to nutrient regimes. The objectives of this st udy were to address the effects of (1) varying substrate concentrations; (2) incubation times; and (3) the use of multiple time point measurements in determining optimum enzyme assay conditions. Materials and Methods Soil cores were collected in Wa ter Conservation Area 2A, a 447 km2 impounded wetland in the northern Everglades that has r eceived agricultural drainage for over 30 years (Davis, 1991). Three soil cores were collected at nutrient enriched (F1) and reference (U3) transect sites along a distinct P gradient. Soil sampling was accomplished using a 5 cm diameter piston type corer in the spring of 2001. The benthic layer or flocculent detrital layer was separated from the remaining soil to be utilized in this experiment. This layer, composed of ac tive algal remains and plant components in various stages of decay, was chosen due to greater microbial ac tivity and variability between sites. The samples were stored on ice until transfer to the laboratory. Large roots and rocks were removed. The samples were homogenized for 5 min using a handheld Biospec Biohomogenizer™ in a 500 mL beaker. Approximately 10 g wet weight was transferred to a second 25 0 mL beaker and 100 mL deionized H2O was added. The suspension was homogenized for an additional 5 min and 1 mL was transferred to another 250 mL b eaker to which 99 mL of DI H2O was added. The suspension was mixed and 50 mL of the su spension was transferred to a 50 mL Eppendorf centrifuge tube. This suspension was used for the assays and refrigerated until use. Previous studies have shown that freezing generally caused increased enzyme activities due to the disruption of enzyme complexes and cell lysis as compared to

PAGE 20

11 refrigeration and was thus avoided (Sinsa baugh and Linkins, 1989; Sinsabaugh et al., 1991). Biostatic agents were not used in this study as their functions differ among samples and have been found to repress the ac tivity of some enzymes (Sinsabaugh et al., 1991). Four enzymes were investigated ut ilizing the following substrates: MUF-D glucoside (Sigma M3633), MUF-cellobios ide (Sigma M6018), MUF-phosphate (Sigma M8168) and L-leucine amidomethylcoumarin (S igma L2145). These substrates were used for the determination of -glucosidase, cellobiohydrolase, phosphatase, and leucine aminopeptidase activities, respectively. Enzyme substrate solutions were prepared in 10 mM Tris-HCl, pH 8.5 in varying concentration ranges. Buffered conditions st abilize the pH depende nt intensity of the fluorescent product (Chr st and Krambeck, 1986). Enzyme assays were performed using a Cytofluor 600™ (Perseptive Biosyste ms, Inc. Framingham, MA) automated spectrofluorometer at with Kineticalc™ software at 360 nm excitation and 460 nm emission. Samples were analyzed in Corni ng 48-well microplates. 360 L Tris-HCl pH 8.5 was added to each sample well followed by 400 L soil suspension and 40 L of the enzyme substrate. Triplicate replicates of each suspension were used. The spectrofluorimeter was set to 25 C, to shake for 3 sec before each reading, and for end-point analysis. Sensitivity was set at 95, which refers to the sensitivity of the fluorimeter with a maximum value of 100. The plate was inserted a nd the initial (Ti) fluorescence was recorded. A reading was pe rformed every 10 minutes for the duration of the incubation. A similar detailed spect rofluorimeter method used 1 minute intervals

PAGE 21

12 between readings (Marx et al., 2001). A fina l fluorescence (Tf) was recorded by setting the fluorimeter to end point anal ysis at the end of the run. Various amounts of substrates were adde d to the samples to establish enzyme saturation (Chr st, 1991). Incubations were car ried out at 5, 10, 30, 60, 125, 250, and 500 M for MUF-cellobioside; 5, 10, 20, 30, 50, 125, 250, and 500 M for MUFphosphate; 5, 10, 20, 30, 50, 70, 100, 250, 500 M for MUF-glucoside and 5, 10, 20, 50, 100, 300, 400, 800, and 1600 M for L-Leucine amidomethylcoumarin. These substrate concentrations reflect in-well measurements as a consequence of dilution and are similar to the range of substrate con centrations used in another methodological study (Marx et al, 2001). In most cases fluorescence readings were taken over 2 hours. After evaluating the kinetic curves, the initial fluorescence (Ti) was subtracted from the final fluorescence (Tf) across incubation time to determine the potential enzyme activity. The largest difference in fluorescence (Tf-Ti) should yield the best resolution of activity for a given site. Alkalinization of the samples pr ior to fluorescence measurement was found unnecessary as all sample fluoresences were resolvable due to the high sensitivity of the spectrofluorimeter. Further conversi on of potential enzyme activity to mols MUF or AMC released g-1 AFDM h-1 was not necessary to meet the goals of this study. All figures represent averaged data with standard errors. Data from selected substrate concentrations are presented in figures to represent the extremes associated with each enzyme.

PAGE 22

13 Results Cellobiohydrolase The effect of substrate concentration a nd incubation time on the potential activity of cellobiohydrolase was investig ated at the enriched and reference sites. Tf-Ti values are simply referred to as “activity ” for the purposes of this study. The response of the highest MUF-cell obioside concentra tion (500M) over incubation time varied between the enriched (F 1) and reference (U3) sites (Figures 2-1a & 2-1b). Mean activities over the incubation period were 30 2 and -240 (unitless) with standard errors 179% and 48% of the mean fo r the F1 and U3 sites, respectively. Both sites exhibited erratic readings with initial reductions in activ ity that was reflected in the large replicate errors. F1 responded differen tly than U3 with net increases in activity after 120 mins. Positive activity changes were exhibited at the lowest MUF-cellobioside concentration of 5 M at both sites with mean activities of 223 and 295 and standard errors 20% and 17% of the mean at the F1 and U3 sites, respectively. These activities were considered low and reflect an insuffici ent quantity of substrate to obtain near saturating conditions. This low resolution in the difference between Tf and Ti values occurred in the range of 5 to 30 M. The behavior of cellobiohydrolase was f ound to be most linear at a substrate concentration of 60 M. This concentration reflected the gr eatest activity among the sites before a large decrease in activity at 125 M (Figure 2-1c). Mean activities of 673 and 503 with standard errors 9% and 15% of the mean occurred for the F1 and U3 sites, respectively. Due to linear increases in activity, extended incubation times were not

PAGE 23

14 a.) 3100 4100 5100 6100 7100 8100 9100 10100 020406080100120140 b.) 3100 4100 5100 6100 7100 8100 9100 10100 020406080100120140 c.) 0 100 200 300 400 500 600 700 800 0100200300400500600 Figure 2-1. Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments a nd (c) varying MUF-cellobioside substrate concentrations on net activities. Activit y ( Tf-Ti ) Activit y ( Tf-Ti ) Activit y ( Tf-Ti ) Time(min) Time(min) LEGEND 500 M 60 M 5 M F1 U3 Activit y ( Tf-Ti ) Substrate Concentration (M)

PAGE 24

15 a.) 0 2000 4000 6000 8000 10000 12000 020406080100120140 b.) 0 2000 4000 6000 8000 10000 12000 020406080100120140 c.) 0 2000 4000 6000 8000 10000 12000 14000 16000 0100200300400500 Figure 2-2. Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments and (c) varying MUF-phosphate substrate concentrations on net activities. Activit y ( Tf-Ti ) Activit y ( Tf-Ti ) Time(min) Time(min) F1 U3 Activit y ( Tf-Ti ) Substrate Concentration (M) LEGEND 500 M 50 M 5 M

PAGE 25

16 considered important at 60 M as substrat e saturating conditions were maintained throughout the duration of the run. However, at higher substrate concentrations, such as 500 M, incubation time does appear to influenc e the determination of activity (Fig 2-1a & 2-1b). At 500 M, incubation times exceeding approximately 80 minutes for F1 and 120 minutes for U3 would be necessary in order to resolve positive activities at both sites. Phosphatase Mean activities for F1 and U3 sites we re 3717 and 5433 (unitless) with standard errors 2.5% and 5.0% of the mean at 500 M respectively (Figures 2-2a & 2-2b). Relative differences between sites substantiall y decreased at concentrations at or above 250 M. In contrast to the MUF-cellobioside assay, high concentratio ns did not result in erratic behavior over the inc ubation period but rather remain ed linear (Figures 2-2a & 22b). The lowest concentration of 5 M exhibited very linear relationships over time in all replicates in both nutrient conditions. M ean activity values were 287 and 4050 with standard errors 6.5% and 1.2% of the mean fo r the F1 and U3 site s, respectively. The low activity at the F1 site reflected the need for a higher substrate concentration in order to counteract potential interferences in future experi mental runs where microbial phosphatase production may be lower. The most appropriate concentration fo r the purpose of future studies was determined to be in the 50 M range (Fi gure 2-2c), before a substantial decrease in activity occurred in the U3 sample. Linear responses occurred in both the F1 and U3 samples over the length of the incubation pe riod with mean activity values of 2770 and 9067 with standard errors 2.5% and 2.9% of th e mean, respectively. These values reflect the highest activity value while still minimizi ng laboratory error. Incubation time at 50 M does not appear to be an important c onsideration since linear relationships were

PAGE 26

17 exhibited in both samples. This concentra tion was 8 times lower than another wetland study that examined phosphatase linearity in order to establish optimum substrate concentrations (Kang and Freeman, 1999). -glucosidase The highest MUF-glucoside concentra tion of 500 M yiel ded Tf-Ti activity differences with average valu es of 565 and 989 in F1 and U3 samples respectively, with laboratory replicate standard errors of 1.8% a nd 9.3% of the mean (Figures 2-3a & 2-3b). Quasi-linear graphs were produced with the highest activity values occurring at the U3 site. 5 M average activity values were 267 and 88 with standard errors representing 7.6% and 50.8% of the mean for the F1 and U3 sites, respectively. The relationship of enzyme activity between sites shifted from the general tr end at approximately 200 M and continued until approximately 300 M (Fig ure 2-3c). This il lustrates the profound effect that substrate concentrations can have on formulating conclusions. The most consistent substrate concentrati on in terms of net activity, linearity and error was identified at 100 M. These runs yielded mean activities of 416 and 248 with standard errors of 8.1% and 4.8% of the mean for the F1 and U3 sites, respectively. There was a linear response of enzyme activ ity over time with no apparent effect of increasing incubation time on mean activ ities or differences among the sites. Leucine Aminopeptidase Leucine aminopeptidase exhibited typical act ivity responses to increasing substrate concentrations (Figure 2-4c). Peak activity was around 800 M and decreased in response to a higher or lower concentration. As substrate concentration was decreased from 800 M, the differences between the tw o sites and the replic ate errors also

PAGE 27

18 a.) 2500 3000 3500 4000 4500 5000 020406080100120140 b.) 7400 7600 7800 8000 8200 8400 8600 8800 9000 020406080100120140 c.) 0 200 400 600 800 1000 1200 0100200300400500600 Figure 2-3. Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments and (c) varying MUF-glucoside substrate concentrations on net activities. Activit y ( Tf-Ti ) Activit y ( Tf-Ti ) Time(min) Time (min) LEGEND 500 M 100 M 5 M U3 F1 Activit y ( Tf-Ti ) Substrate Concentration (M)

PAGE 28

19 decreased. At 1600 M the mean activitie s were 11167 and 18833 with standard errors of 16.5% and 13.4% of the mean for the F1 and U3 sites, respectively. The 300 M runs exhibited mean activities of 6197 and 11866 with standard errors of 1.0% and 1.0% of the mean for the F1 and U3 sites, respectively. Linear graphs at 300 M were the most consistent, showing no apparent influence of incubation time on site differences or individual activities (Figs 2-4a & 2-4b). Mean activity values for the 10 M runs were 417 and 710 with standard errors representing 5.9% and 7.8% of the mean. This concentration represented the lowest activity difference between samples among the range of substrate concentrations. Discussion These results demonstrate the need for the investigation of incubation time and substrate concentration effects to ensure the validity of data in future experiments. Enzyme activity is extremely sensitive to cha nges in environmental conditions, thus it is necessary to construct the assa ys such that enzyme effici ency is maintained through a range of conditions that may be encountered. The enzymes in this study were responsive to changes in nutrient concen trations. These responses were evident in different relationships to changes in substrate concen tration and incubation time. In some cases, different contrasts between the enriched a nd reference samples were observed as the substrate concentration was altered. This may lead the investigator to assume inverse relationships between samples, which are not necessarily reflected in the majority of other substrate concentrations Therefore, the most contrasting conditions must be assayed in order to develop the most consis tent substrate concentration and incubation time in relation to changes in sample properties.

PAGE 29

20 a.) 0 10000 20000 30000 40000 50000 60000 0102030405060 b.) 0 10000 20000 30000 40000 50000 60000 0102030405060 c.) 0 5000 10000 15000 20000 25000 30000 0500100015002000 Figure 2-4. Effects of (a) time on net activities in F1 sediments, (b) time on net activities in U3 sediments a nd (c) varying L-Leucine amidomethylcoumarin substrate concentrations on net activities. Activit y ( Tf-Ti ) Activit y ( Tf-Ti ) Time(min) Time(min) LEGEND 1600 M 300 M 10 M F1 U3 Activit y ( Tf-Ti )

PAGE 30

21 The use of the curves produced over the total incubation peri od is necessary to observe enzyme behavior over time. While the endpoint based activity (Tf-Ti) produced as a result of incubation may indicate a net ch ange in substrate conve rted, there is often erratic conversion behavior over time. Without the use of these curves, the data would be considered valid since there would be no reas on to discard any values that appear to reflect net activity. The resultant experimental protocol for future studies discards any non-linear runs from further analysis when the activity reflects outlying values. In determining optimum substrate concentrations, it is generally evident that very high and low concentrations tended to produce erratic changes in ac tivity over time and lower differences between sites. The high c oncentration response can be attributed to intermolecular quenching in which fluorescen ce is masked by the presence of other fluorescent and non-fluorescent molecules. Ther efore, there can be a false decrease or a lesser increase in substrat e conversion as measured by the fluorimeter. At low concentrations, saturation kinetics did not a ppear to be satisfied, thereby resulting in a lower perceived reaction rates. This is due to the lack of adequate s ubstrate that fails to retain a constant reaction velocity. The appropriate MUF and AMC substrate concentrations that were determined in th is study fall within the range of 40 to 200 M documented in other studies of water column and soil enzymes (Burns and Ryder, 2001; Shackle et al., 2000; Sinsabaugh and Forema n, 2001; Saiya-Cork et al., 2002). Incubation time was shown to influen ce the resolution between the two study sites. In most cases, incubation times le ss than 30 minutes re sult in a decrease in resolution. In addition, substrat e conversion is generally errati c with large errors between replicates. An incubation time of 45 minutes is deemed suitable fo r capturing activity as

PAGE 31

22 well as enabling a rapid turnover of samples in future studies. This lies within the ranges observed in other studies (K ourtev et al., 2002b; Foreman et al, 1998; Burns and Ryder, 2001). Short incubation times minimize any ch anges that may occur in the microbial community (Hoppe, 1993). It should be noted that longer incubation times are acceptable if enzyme dynamics or environmental fact ors warrant. Such incubations may be performed in order to resolve very lo w enzyme activities or overcome phenolic quenching in heavily lignified or tannic sample s. However, in order to investigate the temporal function of the enzymes over the in cubation period, the samples must remain in the automated spectrofluorimeter, thus dela ying the processing of other samples. Therefore, the most expeditious 45-minute in cubation is preferred to decrease sample storage time and increase efficiency. Conclusions These results demonstrate the need for the investigation of incubation time and substrate concentration effects to ensure the validity of data in future experiments. Further, it is prudent that enzyme kinetic s be incorporated in to the experimental methodology due to the differences among subs trate concentration responses over time that were observed in this study. This process can be done expeditiously using a microplate approach to enzyme assays and analyzing with an automated spectrofluorimeter, compared to older cuvette based measurements. It is also important to analyze a range of concentrations in the most widely contrasting environmental conditions within the study area in order to determine the most appropriate substrate concentration. However, the final determin ation of the warranted incubation conditions will be dependent on the specific goals of the study, sample matrix, number of assays, the desire for precision versus accuracy, and available analytical equipment.

PAGE 32

23 CHAPTER 3 EFFECTS OF HABITAT DIFFERENTIATION ON MICROBIAL ENZYME ACTIVITIES IN THE EVERGLADES Introduction Organic matter accumulation within wetla nds is a consequence of the balance between net primary production (NPP) and microbial heterotrophic metabolism. Microbial decomposers play a crucial role in carbon (C) cycling and are responsible for driving the C energy flow up the detrital food chain. The minera lization of organic nutrients by the microbial community exerts an appreciable influence on energy flow by regulating nutrient availability for furthe r decomposition and primary production (Elliot et al., 1984). Mineralization of plant matter is governed by the chemical and physical properties of available substrates such as lignin, nitrogen (N), and phosphorus (P) (DeBusk and Reddy, 1998; Berg, 2000; Fioretto et al., 2000; Kourte v et al., 2002a) as well as other environmental and physiochemical influences. Therefore, changes in these parameters associated with different litter types and nutrient conditi ons have the potential to alter peat accumulation rates a nd potentially topography over time. The microbial degradation of particulate or ganic matter (POM), su ch as plant litter, has been shown to be most influenced by the enzymes involved in lignocellulose degradation, P cycling and N cycling (S insabaugh et al., 1991; Sinsabaugh and Moorhead, 1994), which are often considered the rate limiting steps in degradation (Chr st and Rai, 1993). Due to the sometimes complex macrophytic structure, a large quantity of diverse enzymes ma y be necessary to complete degradation (Eriksson and

PAGE 33

24 Wood, 1985; Ljungdahl and Eriksson, 1985). The synthesis and activity of enzymes may be most regulated through the induction by macrophytic and macromolecular substrates present in the soils (Burns, 1986; Nausch et al., 1998), with strong relationships established between lignoce llulose-degrading enzymes a nd litter mass loss rates among various litter qualities (Sinsaba ugh et al., 1992a). Consequent ly, soils beneath different plant species have been shown to support microbial communities that differ in both structure and function (Degens and Harris, 1997; Grayston et al., 2001; Kourtev et al., 2002a; Kourtev et al., 2003) a nd can affect sediment nutri ents through pl ant uptake and rhizosphere characteristics (Templer et al ., 1998). Litter breakdown rates and enzyme activities have been shown to vary among speci es and have been attributed to intrinsic factors of the leaves (Linkins et al., 1990a & 1990b; Carreiro et al., 2000; Kourtev et al., 2002b) such as specific chemical compositi on (Linkins et al., 1990a & 1990b). For example, litter N content, presented as C:N or lignin:N are often used as predictors of decomposition rates (Melillo and Aber., 1982; Sinsabaugh et al., 1993; DeBusk and Reddy, 1998; Carreiro et al., 2000). Microbial degradation by-products and e xogeneous nutrients may also influence microbial activity. For example, polyphenol s, the by-products of lignin degradation, have the potential to inhibit enzyme co mplexes. These compounds can, in certain appropriate environmental conditions, beco me the dominant regulatory mechanisms involved in microbial respir ation (McClaughtery and Linki ns, 1990; Wetzel, 1993). The availability of nitrogen and ot her nutrients has been shown to control the early phases of decay, when mass loss is less th an 30% (Taylor et al., 1989). Conversely, the addition of N has been shown to retard the decomposition of large mole cular weight organic matter

PAGE 34

25 by the repression of phenol oxidase (Sinsaba ugh et al., 1993; Carreir o et al., 2000). As decomposition proceeds, nutrient availability becomes progressively less important and lignin content controls litte r decay rates (Taylor et al., 1989; Berg, 2000). As a consequence of these complex interactions the interpretation of individual enzyme activities as simple responses to environmental substrate concentrations may not serve to adequately predict the actual microbial community dynamics. A current strategy for predicting microbial degradation rates i nvolves the use of a resource allocation rationale and the MA RCIE (Microbial Allocation of Resources Among Community Indicator Enzymes) model for exposing linkages between individual enzymes (Sinsabaugh and Moorhead, 1994; Sins abaugh and Findlay, 1995; Sinsabaugh et al., 1997, 2002). The model is based on the premise that enzyme mediated decomposition of complex molecules are the ra te-limiting step in C mineralization. It relates bacterioplankton producti on or litterbag mass loss through a first order model that includes C, N, and P allocation. It further indi cates that the expression of enzymes is tied to environmental nutrient avai labilities and that the distri bution of enzyme activities can be interpreted as a reso urce allocation strategy (Sinsabaugh et al., 2002). Related to the MARCIE model, the Enzyme Index of Carbon Quality (EICQ) is a relative index of the normalized activities of the hydrolytic en zymes to oxidative or lignin degrading enzymes. EICQ has been shown to be correlated with microbial biomass (r=0.71), productivity (r=.80) and negatively corre lated with particulate organic carbon (POC) turnover time (r=-0.99) (Sinsabaugh a nd Findlay, 1995). EICQ values and other enzyme ratios reflect microbial community responses to the pe rceived abundance of nutrients as well as lignin a nd thus respond to changes in substrate and environmental

PAGE 35

26 nutrient conditions in terms of microbial resource allocation dyna mics. MARCIE model components, such as Ecell/Ep and Ecell/E n, which reflect apparent phosphorus and nitrogen control on C mineralization, have also been correlated with bacterial productivity (Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997) An underlying concept is that lignocellulo se degradation by extracellular enzymes is tied to environmental N and P concentrations. Sin ce only a certain amount of metabolic energy can be utilized in the production of enzyme s, an abundant expression of one enzyme resulting from a lack of direc tly utilizable substrate will re sult in less energy available for the production of other enzymes. In organic matter dominated wetlands, such as the Everglades, the microbial mineralization of organic matter is a key pr ocess involved in the initial development, accumulation, and maintenance of the peat profile. Differential decomposition processes over time may result in the development of topographic features related to local biogeochemical characteristics which are, in turn, influenced by larger scale ecosystem processes. Specifically, the Everglades we re historically dominated by a slough-ridge landscape interspersed with tree islands. This landscape consisted of dense sawgrass ridges with soil surfaces 2 to 3 feet higher than the adjacent deeper sloughs (Baldwin and Hawker, 1915). However, after over 50 y ears of compartmentalization and altered surface flow, the area covered by the deeper sl oughs is diminishing, being replaced by shallower, monotypic sawgrass (Cladium jamaicense crantz ) stands. Changes in overall productivity have accompanied shifts in vegetative communities, which have the potential to alter the nutrient storag e capability of the system (Davis, 1991).

PAGE 36

27 Differences between the habitats are prim arily observed in plant litter chemistry, characteristics, and localized nutrient condi tions. For example, approximately 13-fold greater root P accumulation rates in catta il over sawgrass have been documented (Lorenzen et al., 2001) as well as greater P accumulation in cattail in high P conditions (Davis, 1991; Koch and Reddy, 1992; Chiang et al., 2000). Conversely, sawgrass root phosphatase activities were also found to be significantly greater than cattail in low P conditions (Kuhn et al., 2002). In a 4 year P dosing study in the Everglades, tissue P accounted for approximately 5% of the adde d P in sawgrass while less than 2% was recovered from plant tissue at the slough site du e to the disappearance of the native community (Chiang et al., 2000). Localized dissolved organic nitrogen (DON) and dissolved organic carbon (DOC) bioavailability to a microbial community has also been shown to differ between cattail a nd spikerush (Eleocharis spp.) stands in a coastal humic lake (Stepanauskas et al., 2000). These di fferences in nutrient conditions and plant characteristics ultimately influence the e fficiency and composition of the resident heterotrophic microbial decomposers, which sh ould be exhibited as changes in enzyme activities. The main objective of this study was to determine the effects of vegetative habitat types and nutrient loading on apparent enzyme activity and apply the results to varying enzyme models to account for differences in potential decom position among the three dominant habitat types (sawgr ass, cattail, and open water). By investigating these relationships, one possible fact or in the development of th e topographical relief of the Everglades may be more fully understood.

PAGE 37

28 Materials and Methods Study Sites The field study sites for this project we re within Water Conservation Area 3A (WCA-3A) located in Broward County, FL WCA-3A is a freshwater marsh incompletely impounded by a combination of le vees and canals and covers an area of 2,339 km2. The area is a mosaic of sawgrass stands and open water areas or sloughs interspersed with tree islands. The result of over 40 years of agricultural runoff into the Everglades has been the establishment of a P gradient originating in the northern regions of the Everglades at points downstream of discharges. Cattail (Typha spp.) has invaded into areas previously dominated by sawgrass (Davis, 1991; DeBusk et al., 1994; DeBusk et al., 2001). Six sites were selected for sampling, re presenting P enriched and slightly P enriched (designated reference) cattail, sawg rass, and open water sites. Enriched sites were designated ECS, ESS, EOS, ECB, ESB, and EOB where the first letter is designated E=enriched, the second; C=cattail, S=sawgra ss, O=open water, and the third; S=soil layer, and B=benthic layer. Reference P sites followed the same nomenclature with RCS, RSS, ROS, RCB, RSB, and ROB where R=re ference site. GPS coordinates were 26 04.900’ 80 49.520’ for the enriched site and 26 02.410’ 80 48.870’ for the reference site. Sites were located near established transects utilized by the South Florida Water Management District for water quality monito ring. Dense cattail, sparse sawgrass, and open water areas were present at the P enri ched sites, located approximately 1.22 km from the canal inflow. Sparse cattail, stands of sawgrass, and open water or slough communities consisting of water lily (Nymphaea) spikerush and periphyton constituted

PAGE 38

29 the reference (slightly impacted) site, loca ted approximately 5.83 km from the canal inflow. Sampling Soil cores were obtained utilizing a 5 cm stainless steel piston type corer with butyrate inserts on September 9th, 2001. Cores were collected in triplicate at each habitat-site combination. The coring procedure involve d pushing the coring apparatus through the soil layer to a de pth of approximately 30 cm. A serrated metal knife was used to cut around the perimeter of the tube to sever large roots and other plant matter. The core was extruded and the benthic matter, defined as the uncons olidated or pourable core fraction, was separated from the soil laye r. The soil section wa s extruded to a depth of 10 cm and both the benthic and soil layers were stored in pl astic bags on ice for transport to the laboratory. Soil Preparation Soil sample analysis began within 24 hour s after field collection. Each layer, corresponding to the 0 to –10 cm soil or bent hic, was prepared separately. 10 g subsamples were placed in pre-weighed aluminum pans for dry mass (DM) and ash free dry mass (AFDM) determination. Dry mass was determined by incubating the pans in a drying oven for 36 hours at 105 C. Ash weight determination involved ashing the samples at 500 C for 2 hours and AFDM was calcula ted by subtracting the ash weight from the dry weight measurements. Samples were transferred to a 500 mL beaker and large objects, such as rocks, were discarded. The samples were homogen ized for 10 minutes with a Biospec Biohomogenizer™, resulting in a so il slurry. 10 g of the slu rry was added to 100 mL DI H2O and homogenized for an additional 5 minut es. 10 mL of the suspension was added

PAGE 39

30 to 90 mL of DI H2O and subsequently transferred to a 100 mL centrifuge tube. The suspensions were refrigerated un til use (Sinsabaugh et al., 1991). Enzyme Analysis Hydrolytic enzyme activity was determin ed using methylumbelliferyl (MUF) and aminomethylcoumarin (AMC) substrates. Subs trate concentrations were optimized at saturating conditions. The activities of -glucosidase (BGL), cellobiohydrolase (CBH), phosphatase (PHO), leucine aminopeptidase (LEU), phenol oxidase (PHE) and peroxidase (PER) were assayed using MUF-D-glucoside (Sigma M3633), MUFcellobioside (Sigma M6018), MUFphosphate (Sigma M8168), L-Leucine amidomethylcoumarin (Sigma L2145) L-3,4-dihydroxyphenylalanine (DOPA), and DOPA + H2O2 as substrates, respectively. MUF and AMC substrate conversion wa s measured using a Cytofluor 600™ automated spectrofluorimeter (PerSeptive Biosystems, Inc., Framingham, MA) with Kineticalc™ software at 360 nm excitation a nd 460 nm emission at 20 C. Assays were performed using Corning 48-well culture plates in which 400 L of sample, 360 L of 10 mM Tris-HCl pH 8.5, and 40 L of substrate were added. Stock substrate concentrations were 2000 M for MUF-D glucoside, 1000 M for MUF-phosphate, 1200 M for MUF-cellobioside, and 6000 M for L-Leucine aminomethylcoumarin resulting in well concentr ations of 100, 50, 60 and 300 M, respectively. Each sample analysis was performed in quadruplicate. In itial and final fluorescence measurements, as well as measurements every five minutes, were taken during the 1 hour incubation. Graphs produced from the readings taken every five minutes were analyzed to ensure that linear kinetics were being obser ved. Final apparent enzyme ac tivities were calculated as moles substrate released g-1 AFDM h-1.

PAGE 40

31 Phenol oxidase activity was determin ed by adding 2.0 mL soil suspension to 2.0 mL 10 mM L-DOPA (L-dihydroxphenylalanine) in 10 mM Tris-HCL pH 8.5 in 10 mL Eppendorf™ centrifuge tubes. Peroxidase ac tivity was determined by adding 2.0 mL soil suspension to 2.0 mL 10 mM L-DOPA with 200 L 0.3% H2O2. The solutions were vortexed for 30 seconds and placed on a shaker plate in a light-proof box for 45 minutes. The solutions were then centrifuged at 3000 rpm for 30 seconds and 500 L supernatant was extracted and quadruplicates were placed in a Corning™ 48-well culture plate. Controls consisting of 250 L DI H2O and 250 L 10 mM L-DOPA solution were added to the remaining wells. Absorbance was r ead at 360/460 nm excitation/emission on the spectrofluorimeter. Quenching was performe d using the soil suspensions with a total volume of 500 L in each well. Sample nutrient analysis was perfor med by DB Labs, Rockledge, FL. Total phosphorus (TP) (EPA 365.2), total nitrogen (TN) (MVP), total organic carbon (TOC) (MVP), calcium (Ca)(SW7140), magnesium (Mg)(SW7450), and lignin (AOAC 973.18) analysis was performed using standa rd methods on homogenized samples. Models Extracellular enzymes can be grouped into four categories: Ecell (BGL and CBH), En (LEU), Ep (PHO), and Eox (PHE and PER); allowing the enzymes to be separated into those involved in C, N, and P mineralization as well as lignin degradation, respectively. While there are many enzymes in each category, it is assumed that the activity of one, or more preferably a few enzymes in each group are sufficiently correlated such that the activ ity of a few can act as indi cators for the entire group.

PAGE 41

32 Enzyme activities are averag ed if there is more than one enzyme in the group (Sinsabaugh et al., 1997). Potential decomposition models were constructed from enzymic data using normalized enzyme activity across both the be nthic and soil layers. This was done in order to compare different enzymes at the sa me scale so that the more active enzymes do not heavily weight the calcul ations of model components. The Enzymic Index of Carbon Quality (EICQ) compares the activities of oxidative to hydrolytic enzymes (Sinsabaugh and Findlay, 1995). In addition, several other indices were formulated from the data. Ecell/En reflects apparent N control over C mi neralization, Ecell/Ep is a relative measure indicating P control over C mine ralization, and Ecell/Eox reflects apparent lignin control over C mineralization. To improve normality and heteroscedescity, data were log transformed before statistical analysis using SAS version 8 statistical software (SAS, 1999). A split-plot model for the data was adopted for subse quent ANOVA analysis using PROC GLM in SAS. The whole plot corresponded to the sites in reference to the nut rient gradient while the subplots referred to the specific vegeta tive communities within the whole plots. Multiple comparisons were made between habitats, sites, and habitat-site combinations. Other analyses were performed using the Sta tistical Package for the Social Sciences (™ SPSS) version 11.0.0. All regressions and signifi cant differences are significant at the p<.05 level unless otherwise noted. Results Nutrient Composition Average nutrient concentrations for the bent hic and soil layers with standard errors are presented in Table 3-1. Benthic TP valu es were significantly different among sites

PAGE 42

33 and all habitats in the reference plots while soil TP varied among individual sites. Benthic TP concentrations were gene rally higher than soil TP values. Total nitrogen values were significantly different among individual sites while only the sawgrass was affected by location on the grad ient in the benthic layer. Soil TN values were significantly different among individua l sites and between the enriched and reference open water and cattail habitats. TN concentrations were higher in the enriched habitats with the exception of the sawgrass so il and benthic samples. TN was correlated with TP in the benthic layer (Table 3-3). Benthic TOC values were significantly di fferent among sites, although the effect of the nutrient gradient was onl y significant at the open water site. Soil TOC was significantly different among sites with the ca ttail (p<0.0001) and ope n water affected by location on the gradient. There was no clear relationship between TOC concentrations in the benthic and soil layers. Benthic TOC wa s significantly correlated with TP and TN, while soil TOC was correlated w ith TP (Tables 3-3 and 3-4). Carbon to nitrogen (C:N) ratios exhibited differences based on habitat type in both the soil and benthic layers, range from 11.4 to 16.2, and were significantly correlated with PHO and LEU in the soil layer. The open water habitat exhi bited the lowest C:N ratios among habitats with the largest diffe rence occurring between the open water and sawgrass habitats. There was no clear re lationship between depth and C:N ratios. Lignin data were only available for the soil layer with a small sample size (n=6). Lignin content increased from the reference to the enriched sites and was the highest in the sawgrass habitat. Lignin ranged from 6% at the RCS site to 49% at the ESS site.

PAGE 43

34 Enzyme Activities All of the enzyme assays, with the excepti on of PER, yielded detectable activities. PER activity was not detectable in the majority of samples. Standard errors of hydrolytic enzyme activity were generally lower than the oxidative enzymes, ranging from 4% to 46% of the mean. Laboratory replicate e rror was generally less than field replicate variability. Benthic -glucosidase (BGL) activity did not vary significantly between habitats or along the gradient (Table 3-2). Though not significant, the ca ttail and open water communities generally exhibited higher activitie s in the enriched sites. Benthic BGL was significantly correlated with PHO, CBH, and LEU, while soil layer BGL was only weakly correlated with LEU (Tables 3-3 & 3-4) BGL activities were consistently higher in the benthic layer, ranging from approximately 2 to 40 times the activity in the soil layer, however, benthic and soil layer BGL ac tivities were not signi ficantly correlated to one another. Soil layer BGL was significantl y correlated with TN but did not exhibit significant differences between hab itats or along the gradient. Phosphatase activities were significantly lower than those reported in Everglades periphyton mats (Newman et al., 2001). While soil PHO significantly (p<0.0001) increased with distance from the inflow, there were no significant differences among habitats. This increase in PHO with distan ce from the inflow has been documented in other Everglades areas (Wright and Reddy, 2001b) as well as decrea ses in PHO with P loading in periphyton (Newman et al., 2001). Benthic PHO was signi ficantly correlated with BGL, CBH, and LEU while soil layer P HO was negatively correlated with LEU and PHE. Benthic PHO activities were greater th an in the soil, ranging from approximately 2 to 100 times the activities of the soil layer wi th the only significan t change occurring

PAGE 44

35 Table 3-1. Mean chemical properties of the soil and benthic layers. Units are expressed as g kg-1, lignin content expressed as % by mass with corresponding standard errors. TN= total nitr ogen, TP=total phosphorus, TOC=total organic carbon, C:N=ratio of TOC to TN. Be n=Benthic layer, Soil=Soil 0 to -10 cm layer. N/A represents lack of replication due to insufficient sample. ND=not determined. Site TN TP TOC C:N Lignin Cat 36.9 0.5 0.9 0.0 445.5 3.5 12.1 0.3 ND Ben Enr Saw 28.7 1.1 0.7 0.1 439.7 16.1 15.3 0.2 ND Open 37.0 0.3 1.7 0.1 435.3 2.6 11.8 0.0 ND Cat 6.0 0.7 0.1 0.0 702.3 5.5 11.8 0.5 ND Ben Ref Saw 28.9 1.1 0.5 0.1 403.0 3.0 14.5 0.2 ND Open 22.7 0.2 0.3 0.0 262.3 15.1 11.5 0.6 ND Cat 32.4 0.2 1.3 0.5 408.0 NA 12.6 0.8 41.7 NA Soil Enr Saw 28.1 1.0 0.9 0.1 454.7 21.9 16.2 0.5 49.4 NA Open 38.0 0.4 1.0 0.1 438.5 2.5 11.5 0.1 39.2 NA Cat 28.4 1.0 1.1 0.1 377.3 18.2 13.3 0.2 5.6 NA Soil Ref Saw 29.0 5.1 0.5 0.1 394.0 12.5 13.9 1.5 32.6 NA Open 30.2 1.9 0.3 0.0 343.0 28.0 11.4 0.2 29.5 NA

PAGE 45

36 between the enriched and reference cattail habitats. This relationship has been documented in other aquatic systems in Florida (Newman and Reddy, 1992; Wright and Reddy, 2001b). Benthic PHO was correlated to the nutrient parameters TN, TOC, and TP, while soil PHO was negatively correlated with TOC and TP. Neither benthic or soil CBH activity were significantly different between habitats or along the gradient. As with BGL, there were no trends evident among the habitats or any consistent relationship with the gradient Benthic CBH was significantly correlated with BGL, PHO, and LEU. There were no si gnificant correlations among enzymes in the soil layer (Table 3-3). Benthic CBH was corre lated to the nutrient parameters TN and TP. No significant correlations between CBH and nutrient parameters were observed in the soil layer. Cattail and open water benthic LEU was signi ficantly higher at the enriched sites. Open water and sawgrass soil layer LEU was si gnificantly higher at the enriched sites. Sawgrass LEU was lower (p<0.05) than the other habitats at the enriched sites. Benthic LEU was significantly correlated with BGL, PH O, and CBH (Table 3-3), while soil layer LEU was correlated with PHO and PHE (Table 3-4). Benthic LEU was higher than soil layer activities in the enriched sites, however this relationship was mixed at the reference sites. Benthic LEU was negatively correlate d to the nutrient parameters TN, TOC, and TP while the soil LEU was negatively correlated with TOC and TP. Phenol oxidase did not vary significantly between habitats or along the gradient. Standard errors were greater in the PHE assays than the hydrolytic assays, consistent with higher variability reported by Sinsabaugh (personal communi cation). Activities were

PAGE 46

37 Table 3-2. Mean enzyme activities of the soil and benthic layers. Data reflects means with corr esponding standard errors. ND=not determined. Units for -glucosidase (BGL), cellobi ohydrolase (CBH), phosphatase (PHO), and leucine aminopeptidase (LEU) are moles substrate released g-1 AFDM h-1. Units for PHE and PER are moles DICQ released g-1 AFDM h-1. Site BGL CBH PHO LEU PHE PER ECB 0.0800.031 0.0720.0227.343.77 4.541.75 124.552.3 ND ESB 0.0230.002 0.0300.0042.560.35 1.690.44 45.96.3 ND EOB 0.080 0.023 0.0540.0136.222.20 5.061.48 37.210.0 ND RCB 0.0440.018 0.0190.0091.871.10 0.870.47 75.911.8 ND RSB 0.0440.013 0.0300.0084.280.88 1.960.50 60.235.5 ND ROB 0.0250.010 0.0340.0193.492.22 2.420.97 14.914.9 ND ECS 0.0150.001 0.0380.0020.130.017 1.170.16 55.57.1 0.120.02 ESS 0.0100.001 0.0250.0050.0620.0090.770.16 35.18.4 0.160.03 EOS 0.0160.004 0.0260.0050.110.017 1.110.19 18.49.7 0.070.04 RCS 0.0010.001 0.0320.0040.810.12 1.030.10 59.37.3 ND RSS 0.0180.008 0.0340.0110.76 0.14 2.090.37 98.14.5 ND ROS 0.0120.002 0.021 0.0041.540.13 2.250.01 78.814.2 ND Table 3-3. Correlation coefficien ts for benthic enzyme and nutr ient data. Correlations significant at p<0.05 unless otherwise noted. *=p< 0.0001, NS=not significant. BGL= -glucosidase, PHO= phosphatase CBH= cellobiohydrolase, LEU= leucine aminopeptidase, PHE= phenol oxidase, TN=tot al nitrogen and TOC=total organic carbon. BGL PHO CBH LEU PHE TN TOC PHO 0.73 CBH 0.83* 0.89* LEU 0.71 0.94* 0.92* PHE NS NS NS NS TN NS 0.60 0.45 -0.65 NS TOC NS 0.54 NS -0.64 NS -0.99* TP NS 0.56 0.46 -0.63 NS 0.89* 0.87*

PAGE 47

38 generally higher at the referen ce habitats in the soil layer and were within the ranges of reported activities in the Everglades (Wright and Reddy, 2001b) and other systems (Sinsabaugh and Linkins, 1990; Sinsabaugh et al., 1992b). Another study found no significant changes in phenol oxidase activ ity along the gradient in WCA2A in the Everglades (Wright and Reddy, 2001). The pos itive correlation of soil PHE with LEU suggests that a greater proporti on of N may be in recalcitra nt form. Soil PHE was also negatively correlated with TOC. There was not a significant trend of decreasing phenol oxidase activity with depth. Overall normalized hydrolytic (EThyd) and hydrolytic plus oxidative (EThyd+ox) enzyme activities in the benthic layer exhibite d mixed relationships w ith the gradient and among habitats (Table 3-5). EThyd values were within the ra nge reported in a previous study of a riverine system (S insabaugh et al., 1997). EThyd and EThyd+ox in both the cattail and open water habitats were significantly greate r at the enriched sites. However, there were no consistently significant differences between habitats. The differences between the EThyd and EThyd+ox values demonstrate the contributi on of apparent lignin degradation to the overall enzyme activity. The refere nce open water benthic layer (ROB) exhibited the smallest contribution of oxidases to th e overall enzyme activity (7%) while the reference sawgrass habitat soil laye r (RSS) showed the greatest (37%). The contribution of normalized enzyme activities are illustra ted in radar plots for the benthic and soil layers (Figure 3-1) for visu alizing the relationships pres ented for the calculated MARCIE model components. It is interesting to note that the shapes of the radar plots for each layer are similar in both the enriched and re ference sites. However, energy allocation does appear to be different in the soil versus benthic layers.

PAGE 48

39 Table 3-4. Correlation coefficients for soil en zyme and nutrient data. Regressions significant at p< 0.05 unless otherwise noted. *= p<0.0001, NS=not significant. BGL= -glucosidase, PHO= phosphatase CBH= cellobiohydrolase, LEU= leucine aminopeptidase, PHE= phenol oxidase, TN=tot al nitrogen and TOC=total organic carbon. BGL PHO CBH LEU PHE TN TOC PHO NS CBH NS NS LEU NS -0.71 NS PHE NS 0.71 NS 0.64 TN 0.60 NS NS NS NS TOC NS -0.79 NS -0.50 -0.65 NS TP NS -0.54 NS -0.65 NS NS 0.57 Table 3-5. Overall benthic and soil normalized total enzyme activities. Total activities are expressed as the sum of the hydrolytic (BGL, CBH, PHO, LEU) and oxidative (PHE) enzymes (EThyd+ox) as well as the hydrolytic enzymes only (EThyd) with standard errors. Enr=enriched site, Ref= reference site. Values are unitless. ET (hyd) ET (hyd+ox) Cat 2.27 0.88 2.82 1.11 Benthic EnrSaw 0.81 0.11 1.01 0.13 Open 2.10 0.60 2.27 0.56 Cat 0.71 0.34 1.05 0.30 Benthic RefSaw 1.11 0.28 1.38 0.29 Open 0.94 0.35 1.01 0.39 Cat 0.59 0.02 0.84 0.06 Soil EnrSaw 0.38 0.06 0.54 0.03 Open 0.49 0.09 0.57 0.08 Cat 0.51 0.05 0.77 0.07 Soil RefSaw 0.75 0.20 1.18 0.22 Open 0.65 0.05 1.00 0.10

PAGE 49

40 a.) enriched b.) reference 0.0 0.2 0.4 BGL PHO CBH LEU PHE ECB ESB EOB 0.0 0.2 0.4 BGL PHO CBH LEU PHE RCB RSB ROB c.) enriched d.) reference 0.0 0.2 0.4 BGL PHO CBH LEU PHE ECS ESS EOS 0.0 0.2 0.4 BGL PHO CBH LEU PHE RCS RSS ROSFigure 3-1. Radar Plots of be nthic (a & b) and soil (c & d) layer average normalized enzyme activities for the enriched and reference ha bitats showing the basis for MARCIE model components. MARCIE Model Components Ecell/Ep values reflect the apparent P control on C mi neralization based on resource allocation (Table 3-6). Higher Ecel l/Ep values indicate de creased production of PHO in relation to the hydrolytic enzymes involved in C mineralization. Therefore, higher Ecell/Ep values suggest decreased control of P on C mineralization. Benthic and soil Ecell/Ep were significantly higher at all the enriched ha bitats. The enriched benthic layer did not show any significant differences among the habitats. However, all of the habitats exhibited significantly different bent hic Ecell/Ep values at the reference sites

PAGE 50

41 with the open water habitats having the lowest values. Significantly greater values were found in the soil, compared to the benthic, among the enriched habitats. Benthic Ecell/Ep was negatively correlated with TN and TOC wh ile soil Ecell/Ep was positively correlated with TOC and TP (Table 3-7). Ecell/En values reflect apparent N cont rol on C mineralization. Higher Ecell/En values, like Ecell/Ep, indicat e decreased LEU production in relation to C mineralization, which in turn suggests lesser control of N on C mineralization. Ecell/En values were within the high range of values reported in a study of a riverine system (Sinsabaugh et al., 1997). The enriched site benthic and soil layers exhibited no si gnificant differences among the habitats. However, at the reference site, in both laye rs, all of the habitats were significantly different than one another with the lowest Ecell/ En associated with the open water habitat. Benthic Ecell/En was negativ ely correlated with TN, TOC, and TP while soil values were positively co rrelated with TOC and TP. The relative degrees of apparent N and P influences on C mineralization can be compared simultaneously in order to further di fferentiate the sites (F igures 3-2 & 3-3). The reference benthic layer sites generally li e along a line from the open water habitats (low N, low P) to the cattail habitats (high N, high P). The grouping of these habitats is more distinct than that of the enriched ha bitats, which are generally in the two high P quadrants, with varying relative N levels. In comparison, the soil layer relationships are more distinctive in relation to the gradient, although not as clear as the benthic layer in terms of habitat grouping. Enriched sites are found in the upper por tion of the high N, high P cate gory while the majority of

PAGE 51

42 Table 3-6. Enzyme ratios calculated with normalized apparent enzyme activity means and associated standard errors. Values expressed are unitless. Ecell/Eox is the appa rent lignin influence on C mine ralization, Ecell/Ep and Ecell/En are th e apparent phosphorus and ni trogen influences on C mineralization and EICQ is the Enzyme Index of Carbon Quality. ND reflects a lack of repli cation due to zero oxidative enzyme activities in the denominator, thus refl ecting conservatively low estimates. Site Ecell/ Eox Ecell/Ep Ecell/En EICQ ECB 1.18 0.111.43 0.26 1.11 0.20 5.35 0.54 ESB 1.06 0.061.26 0.10 1.16 0.28 4.99 0.41 EOB 4.47 2.481.33 0.13 0.86 0.13 19.02 10.48 RCB 0.81 0.412.38 0.41 2.52 0.30 3.42 1.32 RSB 0.58 0.150.97 0.19 1.16 0.09 3.25 0.33 ROB 2.24 ND 0.57 0.03 0.71 0.10 7.98 ND ECS 0.91 0.0926.19 2.691.57 0.25 3.44 0.20 ESS 1.10 0.4633.81 2.021.55 0.26 4.00 1.24 EOS 1.24 0.1122.60 2.061.25 0.04 4.50 0.27 RCS 0.63 0.083.03 0.12 1.29 0.16 2.95 0.19 RSS 0.48 0.154.04 0.62 0.79 0.13 2.68 0.39 ROS 0.39 0.061.28 0.22 0.47 0.07 2.92 0.23 Table 3-7. Correlation coe fficients for soil and benthic layer enzyme ratio components and nutri ent data. Regressions significant at P<0.05 unless otherwise noted. NS=not significant. Ecell/Ep and E cell/En are the apparent phosphorus and nitrogen influenc es on C mineralization and EICQ is the Enzyme Index of Carbon Quality TN= total nitrogen, TOC=total organic carbon and TP=total phosphorus. Benthic Soil TN TOC TP TN TOC TP Ecell/En -0.77 -0.70 -0.60 Ecell/En NS 0.63 0.87 Ecell/Ep -0.56 -0.51 NS Ecell/Ep NS 0.81 0.62 Ecell/Eox NS NS 0.57 Ecell/EoxNS 0.68 0.56 EICQ 0.47 NS 0.62 EICQ 0.51 NS NS

PAGE 52

43 reference sites are found in the low N, high P quadrant in the soil layer. The reference habitat soil layers exhibit similarities to the corresponding benthic la yers, with apparent grouping, especially in terms of the open water habitat. The reference open water habitat is again found in the lowest section of the lo w N, high P quadrant with the cattail habitats generally grouped in a higher N area than the other habitats. The differences between the benthic and soil layers are probably due to lo wer microbial activities in the soil layer, which results in a lower demand for N and P as well as shifts in community composition due to decreased carbon quality with depth. Ecell/Eox values reflect the apparent lignin control on C mi neralization. Higher Ecell/Eox values indicate d ecreased PHE production, which is a reflection of apparent lignin content, suggesting lesser lignin cont rol on C mineralization. The reference open water habitat benthic Ecell/Eox was significantly higher than the othe r habitats while the enriched open water was highe r (p<0.05) than the sawgrass. These differences were not found in the soil layers. The enriched sites generally exhibited higher values than the reference, with only 3 of 6 comparisons in both layers significantly different. Ecell/Eox was correlated with TP in the benthic layer in addition to TOC and TP in the soil layer. By combining the average Ecell/Ep, Ecel l/En, and Ecell/Eox values for each habitat, the general trends for each habitat concerning P, N, and lignin influences on C mineralization can be constructed (Figure 34). Comparing the enriched to reference benthic sites, the largest appare nt shifts in N and P dynamics occurs in the cattail habitats while the smallest shift occurs in the sawgra ss sites. The open water habitats exhibited

PAGE 53

44 ECB ECB ESB ESB EOB EOB RCB RSB RSB ROB ROB RSB ECB ESB EOB RCB RCB-0.4 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 -0.3-0.2-0.100.10.20.30.40.50.6 Ecell/EnEcell/EpLow N, High P High N, Low P Low N, Low P High N, High P Figure 3-2. Benthic microbial N and P resource allocation in terms of C mineralization. The Y-axis repres ents apparent P limitation while the X-axis represents apparent N limita tion on C mineralization. Valu es are log-transformed for comparison. N and P conditions reflect microbially perceived concentrations. ECS ESS RCS RCS RSS ROS ROS ECS ECS ESS ESS EOS EOS EOS RCS RSS RSS ROS-0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 -0.6-0.5-0.4-0.3-0.2-0.100.10.20.30.4 Ecell/EnEcell/EpLow N, High PHigh N, High P High N, Low P Low N, Low P Figure 3-3. Soil microbial N a nd P resource allocation in term s of C mineralization. The Y-axis represents apparent P lim itation while the X-axis represents apparent N limitation on C mineralization. Values are log-transformed for comparison. N and P conditions reflect microbially perceived concentrations.

PAGE 54

45 the least amount of lignin influence in both the enriched and reference sites in the benthic layer, which are accompanied by the greatest amount of apparent N limitation with the exception of the ROS site. The reference open water site also exhibited the greatest N and P limitation in relation to C mineralizati on in both layers. Th e distinct grouping in the soil layer reflects primarily changes in P among the enriched habitats and changes in N among the reference habitats (Figure 3-4). Ho wever, this relationship is not evident in the benthic layer. Enzyme Index of Carbon Quality (EICQ) values did not vary in a significantly consistent fashion along the gr adient. However, open water habitat EICQ values were significantly higher in the benthic layers at both the enriched and reference sites, indicating greater perceived C quality. EI CQ was only weakly correlated with the benthic nutrient parameters TN and TP as well as soil TN. Discussion The use of enzyme comparisons, such as Ecell/Ep and Ecell/En, are based on the premise that the use of energy to produce cer tain extracellular enzymes reduces the net energy available for the expression of other en zymes. The relationships between certain enzyme groups are unique and serve to convey in formation that is specifically related to perceived environmental conditions. Th e MARCIE model components, originally developed as a relationship to mass loss rates (Sinsabaugh and Moorhead, 1994 & 1996), may be used as a tool for determining mi crobial productivity a nd perceived nutrient limitations (Jackson et al., 1995; Sinsabaugh a nd Findlay, 1995; Sinsab augh et al., 1997). The use of these models as predictive compone nts in this paper is due to the lack of corresponding decomposition data provided by li tter bags or other means. However,

PAGE 55

46 relationships with published microbial comm unity characteristics and activities are referenced to provide a strong ba sis for the use of these models. The model components Ecell/En and Ecell/ Ep generally predicted lower N and P limitation on C mineralization at the enriched ha bitats. This is concurrent with decreases in nutrient limitations that ha ve been documented in other enriched Everglades regions (McCormick et al., 1996). Ecell/Ep values sp anned an almost 20-fold greater range than Ecell/En. This relationship is consistent with 10-fold greater ranges observed in an aquatic study (Sinsabaugh et al., 1997) and is due larg ely to greater phosphatase variability along the primarily P gradient Additionally, Ecell/Ep was as much as 4-fold lower than values reported in an aquatic study (Sinsabaugh et al., 1997), reflecting the low P nature of this system. Positive relationships between Ecell/En and Ecell/Ep and microbial productivity have also been obser ved (Sinsabaugh et al., 1997). The greater potential productivity at the enriched habitats is further supported by generally higher Ecell/Eox values at these sites, reflecting lower apparent lignin influence on C mineralization. The shift in the microbial response alon g the gradient may reflect changes in microbial community composition that have been obs erved in enriched sites in the Everglades (DeBusk and Reddy, 1998), with subsequent increas ed C mineralization rates. Changes in the apparent quality of the li tter with nutrient concentrations such as decreased C:N and C:P ratios and (Shave r and Melillo, 1984; Craft and Richardson, 1995; Bridgham et al., 1996; DeBusk and Re ddy, 1998) have the potential to influence microbial activity which may be reflected by sh ifts in Ecell/Eox, Ecell/En, and Ecell/Ep along the gradient. These changes has been obs erved in shifts of sawgrass C:N:P ratios

PAGE 56

47 a.) Enr Cattail 1.18 Enr Sawgrass 1.06 Enr Open 4.47 Ref Cattail 0.81 Ref Sawgrass 0.58 Ref Open 2.24 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.00.51.01.52.02.53.0 Ecell/EnEcell/Ep b.) Enr Cattail 0.91 Enr Sawgrass 1.10 Enr Open 1.24 Ref Cattail 0.62 Ref Saw 0.49 Ref Open 0.39 0 5 10 15 20 25 30 35 40 0.00.20.40.60.81.01.21.41.61.82.0 Ecell/EnEcell/Ep Figure 3-4. Bubble graph comparison of (a ) benthic and (b) soil enzyme ratio components Ecell/En, Ece ll/Ep, and Ecell/Eox, representing apparent nitrogen, phosphorus, and lignin control on carbon mineralization, respectively. Bubble size repr esents Ecell/Eox. Ref=reference site, Enr= enriched site, open=open water ha bitats. Note the change in scale between the two layers.

PAGE 57

48 between enriched and reference sites in WC A-2A in the Everglades with values of 2520:63:1 and 360:9:1 in the enriched and re ference sites, respectively, which were correlated with higher C turnover rate s (Koch and Reddy, 1992). Additionally, the general increases in TP, TN, and TOC of both the benthic and soil layers at the enriched habitats appears to suppor t greater the higher Ecell/Ep and Ecell/En values. The supposition of greater productivity at the enriched habitats is also supported by larger EICQ values, which are generally limited to the more active benthic layer. Greater EICQ values in the benthic layer are s ynonymous with early stages of decomposition, which have been shown to have a higher carbon quality (Sinsabaugh and Linkins, 1990; Sinsabaugh and Moorhead, 1994; Sinsab augh and Findlay, 1995). The positive relationships between EICQ and productivity (r=0.80), mi crobial biomass (r=0.71) as well as the negative relati onship with POC turnover time (r=-0.99) (Sinsabaugh and Findlay, 1995) therefore also suggest greater C mineralizati on at the enriched sites. Additionally, this rela tionship between EICQ and microbi al biomass is supported in the Everglades by significantly hi gher microbial biomass C (M BC) and microbial biomass P (MBP) observed within the benthic layer al ong a P gradient in WCA-2A (Wright and Reddy, 2001b), greater soil respiration in highe r P conditions (DeBusk, 1996), as well as faster decomposition in a P dosing study in WCA-1A (Newman et al., 2001). High levels of productivity (Sinsabaugh et al ., 1997) have also been asso ciated with more eutrophic conditions. This relationship is depende nt on a sufficient carbon flow that includes saccharides and amino acids (Sinsabaugh et al., 1997). Increases of TOC, general increases in primary production, and lower a pparent N limitation on C mineralization at the enriched habitats further supports these re lationships. Therefore, the consequence of

PAGE 58

49 these elevated model parameters is predic ted to be increased decomposition at the enriched habitats. However, increases in net primary pr oduction, resulting in greater C flow, may exceed any increase in decompos ition, resulting in a net accumulation of organic matter (Davis, 1991). Ecell/En and Ecell/Ep values did not pred ict more favorable microbial conditions in the open water habitat. Rather, this habitat was generally the most limiting in terms of N and P on C mineralization. This may be a consequence of algal competition for N and P resources due to the more prolific periphyton community in the open water habitats. This contradiction in prediction between the N and P models with Ecell/ Eox and EICQ may be due to varying shifts in the type of mi crobial community responses to different environmental changes. The nutrient gradient is most related to ch anges in N and P and thus it would be expected that these nutrien ts would be the main driving force behind microbial changes when habitat types re main constant. Conversely, changes in vegetative types would be expect ed to influence the actual de gradability of the C source by the microbial community, due to varying shielding effects on nut rients by refractory compounds within the plant structures. In f act, the mineralization of organic C was most affected by C quality with total P concentr ation and the lignocellulose index (LCI) accounting for 91% of the variab ility in aerobic C mineraliza tion of plant litter along the WCA-2A gradient (DeBusk and Reddy, 1998). Thus, different enzyme components would be involved in regulati ng the productivity between these two types of changes. These contrasts in enzyme components mo st related to productivity have been documented with the BGL and PHO correlatio n with bacterial pr oduction in nutrient enriched and unenriched me socosms, respectively (Chr st and Rai, 1993). This suggests

PAGE 59

50 that the coupling and de-coupling of en zymes to production may be dependent on localized biogeochemical properties. The components Ecell/Eox and EICQ predic ted that the open water habitats were likely to have higher potential decomposition ra tes than the other two vegetative habitats, especially in the benthic layer. EICQ ha s been shown to be most affected by litter quality, independent of specific site characte ristics in a litter bag study (Kourtev, 2002b). The lower apparent lignin influence on C mi neralization at these habitats is most certainly related to changes in substrate st ructure and composition. Generally lower C:N ratios also predict a more favorable substrat e quality within the open water habitats. This increase in potential decomposition over time would be expected to develop a lower elevation in relation to the sawgrass habitat, in particular. Additionally, this increase is coupled with a lower C input due to lower pr imary productivity in the form of vegetative turnover. Conversely, the lower potential degr adation at the sawgrass habitats, coupled to a much greater C input, would be expected to exceed the metabolic capability of the resident microbial community, resulting the accumulation of organic matter. The combination of these factors support field observations of lower elevation open water habitats and higher elevation sawgrass stands. Conclusions Lower apparent lignin influence on C mine ralization, lower C:N values, and higher EICQ values all predict that the open water habitat benthic layers are likely to have greater substrate qualities than cattail a nd sawgrass habitats with higher potential decomposition. When coupled to a lower C input, the results support field observations of lower elevations within the open water habitats. Therefore, this study

PAGE 60

51 points to the potentially si gnificant role that different ial decomposition has on the development of the Everglades landscape. The limited range of enzymes used in this study, while allowing for the expeditious processing of samples, may not accurately reflect the full arra y of mineralization pathways. The use of a greater range, po ssibly utilizing exoand endocelullases, chitinase, and other enzymes will allow for a greater enzymatic resolution. Additionally, the use of litter decomposition bags in this type of study would allow the effects of substrate concentration on enzyme activities to be extrapolated to actual decomposition rates.

PAGE 61

52 52 CHAPTER 4 NUTRIENT LOADING EFFECTS ON BI OGEOCHEMICAL AND MICROBIAL ENZYME DYNAMICS IN THE EVERGLADES Introduction The accumulation of organic carbon within wetlands is the result of the balance between net primary production (NPP) and microbial heterotrophic metabolism. Microbial decomposers play a crucial role in carbon (C) cycling and are responsible for driving the C energy flow up the detrital food chain. The minera lization of organic nutrients by the microbial community exerts an appreciable influence on energy flow by regulating nutrient availability for furthe r decomposition and prim ary productivity (Elliot et al., 1984). Although most of the organic C in aquatic and marsh systems is processed and recycled entirely by the microbial comm unity without entering the higher order food webs (Wetzel, 1984), alterations in microbial decomposition associated with nutrient loading have the potential to drastically aff ect the higher trophic levels of the system. The relationships between the heterotrophic microbial community and other effects of nutrient loading may serve to further f acilitate the understanding of ecosystem component linkages within wetlands. Due to th eir contribution to the ability of the soil to degrade organic matter, enzyme activities ma y be especially significant in determining changes in environmental conditions and thr ough the subsequent effects on the resident microbial community (Frankenberger and Dick, 1983; Dick, 1997). The decomposition of plant organic matter is largely regulated by the synthesis and activities of extracellular enzymes produced by microbial communities, which operate at

PAGE 62

53 the biochemical level (Sinsabaugh and Moor head, 1994), and are induced by the presence of macromolecular and macrophytic substrates within soils (Nausch et al., 1998). Due to the complex structure of macrophytic tissue, a large quantity of diverse enzymes may be necessary to complete degradation (Eri ksson and Wood, 1985; Lj ungdahl and Eriksson, 1985). The mineralization of plant matter is governed by the chemical and physical properties of available substrates such as lignin, nitrogen (N), and phosphorus (P) availability (DeBusk and Reddy, 1998; Berg, 200 0; Fioretto et al., 2000; Kourtev et al., 2002a & b) as well as other environmen tal and physicochemical influences. The nature of the many interactions that occur between extracellular enzymes and inducing and repressing environmental com ponents are not well understood and may play a significant role in regulating enzyme activit y. Some examples of potentially significant interactions include the polyphenolic inhibiti on of enzyme complexes as a consequence of lignin degradation. These compounds ca n, in certain appropriate environmental conditions, become the dominant regulat ory mechanisms involved in microbial respiration (McClaughtery and Linkins, 1990; We tzel, 1993). The addition of N has been shown to retard the decom position of large molecular weight organic matter by the repression of phenol oxidase (Eriksson et al ., 1990; Blanchette 1991; Sinsabaugh et al., 1993; Hammel, 1997; Carreiro et al., 2000) and UV photolys is has been shown to enhance enzyme activities by relieving the inhibition by humic compounds (Wetzel et al., 1995; Boavida and Wetzel, 1998; Wetzel, 2000). As a consequence of these complex interactions, the interpretati on of individual enzyme activiti es as simple responses to environmental substrate concentrations may not serve to adequately predict the actual microbial community response.

PAGE 63

54 Enzyme activities have the potential to re flect changes in nutrient cycling as a result of changes in water and soil quality (Wetzel, 1991). For example, in a P-limited system the activities of phosphatases may be a dominant regulatory mechanism controlling microbial productiv ity. Phosphatase regenera tes inorganic P through the hydrolysis of organic P to i norganic P (Wetzel, 1991). It is repressed by P enrichment and dissolved reactive phosphorus (DRP) (Jansson et al., 1988; Chr st, 1991; Newman et al., 2003), and has been recommended as a parameter to assess P impact in the Everglades (Newman et al., 2003). Phosphatase is one example of a suite of enzymes involved in organic matter de gradation and nutrient cycling that can be affected by nutrient loading (Wetzel, 1991; Newman and Reddy, 1993; Marx et al., 2001). Interpreting and relating i ndividual enzyme activities to higher order trophic processes and observations is often a diffi cult and tedious process and are often not linked to soil and microbial parameters in the field (Marsden and Gray, 1986). A current strategy involves the use of a resource allo cation rationale and the MARCIE (Microbial Allocation of Resources Among Community Indicator Enzymes) model for exposing linkages between individual enzymes (Sinsa baugh and Moorhead, 1994; Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997, 2002). The model is based on the enzyme mediated decomposition of complex mol ecules being the rate-limiting step in decomposition. It further indicates that the expression of enzymes is tied to environmental nutrient availabili ties and that the distribution of enzyme activities can be interpreted as a resource allocation strate gy (Sinsabaugh et al., 2002). An underlying concept is that lignocellulo se degradation by extracellular enzymes is tied to environmental N and P concentrations. For example, model components such as

PAGE 64

55 Ecell/Ep and Ecell/En, which reflect appa rent P and N control on C mineralization, respectively, have been correlated with b acterial productivity (S insabaugh and Findlay, 1995; Sinsabaugh et al., 1997). Since a cer tain amount of metabolic energy can be utilized in the production of enzymes, an abundant expression of one enzyme resulting from a lack of directly utilizable substrat e will result in less energy available for the production of other enzymes. The objectives of this experiment were to (1) determine the effects of nutrient loading on microbial enzyme activities in the four hydrologic units of the Everglades, (2) investigate differences in microbial responses to nutrient impacts among the areas, and (3) establish a potential decomposition model relating enzyme activities to cotton strip decomposition rates (CRR). In addition, determ inations of possible interactions between site specific biogeochemical parameters and microbial enzyme activities were investigated. Materials and Methods Site Description The Florida Everglades is an oligotrophic system with referen ce surface water total phosphorus (TP) levels averaging less than 10 g L-1 throughout the inte rior of the marsh (McCormick and O’Dell, 1996; McCormick et al., 2000). The result of decades of nutrient loading from the Everglades Agricult ural Area (EAA) has been the establishment of P and, to a lesser extent, N gradients within the four hydrologic compartments of the remnant Everglades. Consequently, shifts in macrophyte species composition (Davis, 1943, 1991; Vaithiyanaithan and Richardson, 19 99), increases in NPP (Davis, 1991) and peat accumulation (Reddy et al., 1993), lo ss or taxonomic shifts of native periphyton assemblages (McCormick and O’Dell, 1996), in creases in microbial activity and biomass

PAGE 65

56 (White and Reddy, 2000), and other ecologi cal changes have been observed in Everglades areas receiving nutri ent inputs from canal waters with TP concentrations as much as 10-30 fold higher than background (McCormick et al., 1996). Field study sites were locate d in transects situ ated along nutrient gr adients in four distinct hydrologic un its of the Everglades: Arthur R. Marshall Loxahatchee National Wildlife Refuge (LNWR), Water Conservati on Area 2A (WCA-2A), Water Conservation Area 3A (WCA-3A), and Taylor Slough within Everglades National Park (ENP-TS). Loxahatchee National Wildlife Refuge (LNWR) is the northernmost of the hydrologic areas in the Everglades, comple tely impounded by levees and canals, and encompasses 566 km2. The phosphorus gradient in this ar ea exhibits a steep decline with surface water and soil TP levels decreasing to reference levels with in 2.2 km of the L-7 canal (SFWMD, 2003). LNWR hydrology is pr imarily rainfall driven (54%) and is, unlike the rest of the Everglades, an acidic soft water system. Sites were chosen on a transect located along a nutrien t gradient extending from th e L-7 canal and represented enriched (X1), transitional (X 2), and reference (X4) P condi tions which are designated as the ENR, TRANS, and REF sites, respect ively. The average surface water TP concentrations were between 7.2 and 12.3 g L-1 between 1996 and 2001 for the reference site (SFWMD, 2003). Water Conservation Area 2A is a 442 km2 area located in the northern Everglades, completely enclosed by canals a nd levees, and is the most studied among the regions. The primary source of water and nutrient loading to the area are the S-10 structures that transfer water from agricultural areas and LNWR via the Hillsboro Canal. Nitrogen and phosphorus gradients exist in the water column and periphyton tissue

PAGE 66

57 (McCormick and O’Dell, 1996). Soil P concen trations have ranged from approximately 400 to 1600 mg kg-1 in the reference and enriched ar eas, respectively (R eddy et al., 1993; DeBusk et al., 1994). Three sites were samp led, designated F1, F4, and U3. These sites are referred to as ENR, TRANS, and REF fo r the enriched, transitional and reference nutrient sites, respectiv ely. The enriched site is charac terized by robust stands of Typha domingensis Pers. (cattail) and a floating surface layer of Lemnaceae (duckweed). The surface waters contain a large quantity of par ticulate organic matter that is composed of plant matter in various stages of decay. The reference site consists of Eleocharis spp. (spike rush), Nymphae (water lily), Utricularia spp. (blatterwort) and benthic and floating periphyton mats surrounded by stands of Cladium jamaicense Crantz (sawgrass), with little suspended organic matter. Water depth was, on average, shallower at the reference site with clear surface water and particulate pe at intermixed in some areas with a thin carbonate layer underlying the surficial peri phyton mats. Water Conservation Area 3A (WCA-3A) encompasses 2,012 km2 and is predominantly a vast sawgrass marsh interspe rsed with sloughs, tr ee islands, and wet prairies. It is the only area not completely enclosed by leve es. The highest annual mean surface water TP levels in the inflow and interior marsh were 67.3 and 20.3 g L-1 for 1978-2000 (Newman et al., 2002). The extent of dow nstream enrichment is similar to that of LNWR and extends approximately 3 km from the inflow (Newman et al., 2002). Sampling sites were located al ong a SFWMD water quality mon itoring transect located at 0.5E, 1.5W, and Nmeso for the enriched, transiti onal, and reference sites, respectively. The reference site consisted mostly of water lilies, spikerush, and periphyton.

PAGE 67

58 Everglades National Park (ENP) is a 5,569 km2 wetland consisting primarily of marl forming wet prairies, sawgrass stands, fr eshwater sloughs, and mangrove stands at the southern periphery. Marl forming prai ries are characterized by the formation of calcitic mud, especially in the southern regi ons. Surface water flow into Taylor Slough originates at the S-332 structur e at the southeastern side of the park. The overlying water column at the enriched site contains less su spended organic matter, resulting in a clearer profile when compared to the other enriched sites. Lower net primary productivity (NPP) at the enriched site is manifested in decr eased litterfall, as compared to the other hydrologic units. Vegetative change s relating to nutrient input occur in a relatively short distance from the canal inflow and is genera lly limited to the deeper slough region, less than 50 meters in width. The reference site is characterize d by dense periphyton and epiphyton accompanied by spikerush. Sa mpling sites were located along a SFWMD water quality monitoring transe ct designated as 0.5W, 1.5E, and Smeso for the enriched, transitional, and reference sites, respectively (Newman et al., 2002). Sampling Soil cores were obtained using a 10 cm thin-walled stainless steel corer on 12/14/2001, 5/18/2002, and 10/14/2002. Cores were co llected in triplicate at each site. The coring procedure involved pushing the coring apparatus through the soil layer to a depth of approximately 30 cm. During inser tion, a serrated metal knife was used to cut around the perimeter of the corer to sever large roots and other plant matter. The top of the core was sealed with a plastic cap allo wing the core to be excavated from the soil whole. The core was extruded and the benthi c matter, defined as the unconsolidated or pourable core fraction, was separated from the soil layer. The so il section was extruded

PAGE 68

59 to a depth of 10 cm and both the benthic matte r and soil sections were stored in plastic bags on ice for transport to the laboratory. Soil Preparation Soil sample analysis began within 24 hour s of field collection. Each layer, corresponding to the 0 to –10 cm soil and benthi c layers, were prepared separately. 10 g sub-samples were placed in pre-weighed and pre-ashed aluminum pans for dry mass (DM) and ash free dry mass (AFDM) dete rmination. Dry mass was determined by incubating the pans in a drying oven for 36 hours at 105 C. Ash weight determination involved ashing the samples at 500 C for 2 hours. Samples were transferred to a 500 mL beaker and large objects, su ch as snail shells and rocks, were discarded. The samples were homogenized for 10 minutes with a Biospec Biohomogenizer™, resulting in a soil or benthic slurry. 10 g of the slurry was diluted to a concentration of 10-3 with deionized water and ho mogenized for an additional 5 minutes. The resulting suspension was tr ansferred to a 100 mL centrifuge tube and refrigerated until use (Sinsabaugh et al., 1991). Enzyme analysis began within 6 hours of sample preparation. Enzyme Analysis Hydrolytic enzyme activity was determin ed using methylumbelliferyl (MUF) and amidomethylcoumarin (AMC) substrates. Subs trate concentrations were optimized at saturating conditions. The activities of -glucosidase (BGL), phosphatase (PHO), leucine aminopeptidase (LEU), phenol oxidase (PHE) and peroxidase (PER) were assayed using MUF-D-glucoside (Sigma M3633), MUF-phosphate (Sigma M8168), L-Leucine amidomethylcoumarin (Sigma L2145) L-3,4-dihydroxyphenylalanine (DOPA), and DOPA + H2O2 as substrates, respectively.

PAGE 69

60 MUF and AMC substrate enzymatic analys is was measured using a Cytofluor 600™ (PerSeptive Biosystems, Inc., Framingham, MA) automated spectrofluorimeter with Kineticalc™ software at 360 nm excita tion and 460 nm emission at 20 C. Assays were performed using Corning 48 -well culture plat es in which 400 L of sample, 360 L of 10 mM Tris-HCl pH 8.5, and 40 L of substrate was a dded. Stock substrate concentrations were 2000 M for MUF-D glucoside, 1000 M for MUF-phosphate, and 6000 M for L-Leucine aminomethylcoumarin resulting in well concentrations of 100 M, 50 M, and 300 M, respectively. Each sample run was performed in quadruplicate. Initial and final fluorescence measurements as well as measurements every five minutes were taken during the 1 hour incubation. Changes in fluorescence were determined by subtracting the initial from the final fl uorescence. Graphs produced from the readings taken every five minutes were analyzed to ensure that linear kinetics was being observed. Concentrations of MU F and AMC released were calculated by the application of standard curves to the in itial and final fluoresences. The absolute difference in concentrations yielded the substr ate released during the incubation period. The effects of quenching on MUF and AMC subs trates were determined in order to account for fluorescence blocki ng or absorption effects cau sed by coloring, particle suspension, humic matter, self-quenching, or ot her inhibitions in the suspensions. MUF and AMC standards were placed in each sample suspension to determine the quench percentage of each matrix. The final and initial fluoresences were then converted using the appropriate quench percentage. Phenol oxidase and peroxidase anal ysis was performed by adding 2.0 mL soil suspension to 2.0 mL 10 mM L-dihydroxyphenyl alanine (L-DOPA) dissolved in 10 mM

PAGE 70

61 Tris-HCl pH 8.5 in 10 mL Eppendorf™ centrif uge tubes. The solutions were vortexed for 30 seconds and placed on a shaker plate in a light-proof box for 45 minutes. The solutions were then centrifuged at 3000 rpm for 30 seconds, 500 L supernatant was then extracted and placed in quadruplicate in a Co rning™ 48-well culture plate. Controls consisting of 250 L DI H2O and 250 L 10 mM L-DOPA solution were added to the remaining wells. Sample nutrient analyses was perfor med by DB Labs, Rockledge, FL. Total phosphorus (TP) (EPA 365.2), total nitrogen (TN) (MVP), total organic carbon (TOC) (MVP), calcium (Ca) (SW7140), and magne sium (Mg) (SW7450), and lignin (AOAC 973.18) analysis was performed using sta ndard methods on homogenized samples. Potential enzyme activities are expressed in moles MUF released g-1 AFDM h-1 for glucosidase (BGL) and phosphatase (PHO), moles AMC released g-1 AFDM h-1 for leucine aminopeptidase (LEU), and moles DICQ released g-1 AFDM h-1 for phenol oxidase (PHE) and peroxidase (PER). Cotton Strips Cotton strips (Shirley Institute, Manchester, England) were prepared by cutting 12 cm wide strips and securing them in duplicate to stainless steel wire frames (6mm thick) with staples and waterp roof tape (Newman et al., 2001). The assemblies were deployed within 10 meters of the soil coring sites at a de pth of at least 15 cm below the soil surface for a period of 2 weeks at all sampling sites, corresponding within a week to soil coring events. Retrieval involved measuring the dist ance from the top of the frame to the soil layer using visual methods to determine the su rface of the soil. The depth of the benthic layer was also recorded. Control strips on wire frames were quickly deployed in duplicate and immediately retrieved. The retrie ved strips were then washed in ambient

PAGE 71

62 water to remove excess detrital matter. The cotton strips were removed from the wire frame and cut into 2 cm wide strips 10 cm be low the soil surface mark. 2 cm wide strips were also cut in the corres ponding benthic layer. The stri ps were washed in DI H2O and allowed to soak for 10 minutes to ensure sa turation prior to stre tching. A tensiometer (Chatillon TCD-200) with a digital force gauge (DFIS 200, Chati llon, Greensboro, North Carolina, USA) was used to determine the rema ining tensile strength of the cotton strips. This value was subtracted from the contro l strip reading from the corresponding area. The loss of tensile strength ove r time is expressed as a rate constant. Since the loss of tensile strength is analogous to first-order decay, the calculation of the rate constant requires linearization of the curve. Cotton Rottenness Rate (CRR) was calculated as (Hill et al., 1985): CRR= {[((yo-y)/y)1/3)] / # of days deployed}* 3 65, where yo is the tensile strength of the control strips and y is the tensile strength of the test strips at each 2 cm increment. Models Extracellular enzymes were grouped into four categories: Ecell (BGL), En (LEU), Ep (PHO), and Eox (PHE and PER). This grouping allows the enzymes to be grouped into those involved in C, N, and P mine ralization as well as lignin degradation, respectively. Enzyme activities are norma lized on a scale of 0-1 to eliminate the weighting effects of the more active enzymes. Enzyme ratios were formulated to reflect the premise of resource allocation and were based on assumptions derived from the MARCIE (Microbial Allocation of Resour ces Among Community Indicator Enzymes) model (Sinsabaugh and Moorhead, 1 994, 1996; Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997, 2002). The model is based on the premise that the enzyme mediated decomposition of complex molecu les is the rate limiting step in C

PAGE 72

63 mineralization, indicating that the expression of enzymes are tied to environmental nutrient availabilities and that the distribution of enzyme activ ities can be interpreted as a resource allocation strategy (Sinsabaugh et al., 2002). MARCIE model components, such as Ecell/Ep and Ecell/En, which reflect apparent P and N control on C mineralization, respectively, have been correlated with bact erial productivity (Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997). These ratios are based on the underlying concept that lignocellulose de gradation by extracellu lar enzymes is tied to environmental N and P concentrations. Since only a certain amount of energy is available for enzyme production, an abundant expressi on of one enzyme resulting from the lack of directly utilizable substrate will result in less energy availabl e for the production of other enzymes. Ecell/En is the ratio calculat ed by: BGL/LEU, which reflects apparent N control over potential cellulos e decomposition. Ecell/Ep are relative measures indicating P control over potential cellu lose decomposition and is calculated by: BGL/PHO. Ecell/Eox reflects apparent lignin control ov er potential cellulose decomposition and is calculated by: BGL/(average(PER+PHE)). The Enzyme Index of Carbon Quality (EICQ) is a relative inde x of the normalized activities of the hydrolytic enzymes to the oxi dative or lignin degrading enzymes. EICQ has been correlated with microbial bi omass (r=0.71), productivity (r=0.80), and negatively correlated with particulate or ganic carbon (POC) turnover time (r=-0.99) (Sinsabaugh and Findlay, 1995). Statistics Data was statistically analyzed with SAS v.8 statistical software (SAS, 1999) using mixed model repeated measures to determine significant differences (p<0.05) between sampling periods, soil depths, hydrologic units, and sites. The benthic and soil

PAGE 73

64 layers were analyzed independently. Due to significant AREA*SITE*SAMPLING PERIOD interactions, contrast statements were used to differentiate among specific sites. Data was log-transformed to improve normality and heteroscedescity. Regressions were performed using SYSTAT 10.2 (SYSTAT, 20 02) for individual time periods. Mean values of all sampling dates were combined fo r tables and charts. Significant differences and correlation coefficients are significant at the p<0.05 le vel, unless otherwise noted. Results Mean values were calculated from the th ree sampling periods in order to analyze the effects of the canal inflows in LN WR, WCA-2A, WCA-3A, and ENP-TS. The results of mixed model repeated measures an alysis on differences between enriched and reference sites are presented in Tables 4-1 a nd 4-2 for the benthic data and Tables 4-4 and 4-5 for the soil data. Due to the complexity of responses at the tr ansitional sites, these results are presented separately. Benthic Data Nutrient data (Tables 4-1 & 4-2) for the be nthic layers reflected significantly higher total phosphorus (TP) at the enri ched sites in all four areas. The greatest P concentrations were present in the enriched LNWR and WCA-3A sites. TP concentrations were within the ranges reported in othe r studies (Koch and Reddy, 1992; Reddy et al., 1993; DeBusk et al., 1994; Wright et al., 2001b). Among the reference sites, the lowest average TP concentrations were within ENP-TS at 0.1 g kg-1. These low values were also reported in an earlier study (Wright et al., 2001b). Generally, there were not consistently si gnificant differences between the enriched and reference TN and TOC values in the four areas (Tables 4-1 & 4-2). TN did vary significantly along the gradient in LNWR. TN and TOC concentrations were similar to

PAGE 74

65 those reported in a previous study (Wright et al., 2001b). LN WR, WCA-2A, and WCA3A exhibited a similar range of TN and TOC va lues. ENP-TS TN concentrations were as much as 64% lower at the ENR site and 70% lower at the REF site. Correspondingly, ENP-TS TOC concentrations were as much as 61% and 64% lower at the ENR and REF sites, respectively. TN was most strongly co rrelated with TOC and lignin and negatively correlated with calcium (Table 4-3). TOC was most strongly correlated with TP, TN, lignin, and negatively correlated wi th calcium (Table 4-3). Calcium concentrations were the highest in ENP-TS, with values as great as 400 to 1800% higher in the enriched and reference sites, respectively. The only significant change along the gradient occurred in LNWR with a higher concentra tion at the enriched site. Increases in Ca have been shown to increase with P loading due to release from organic matter (Newman et al., 2001). Mg concentrations were between 150% and 338% greater in LNWR and WCA2A than the other two areas sampled and did not vary significantly along the gradient. Benthic C:P values were signi ficantly higher at the refere nce sites in all four study areas. The largest difference in C:P values within an area occurred in ENP-TS with values of 236 and 1803 at the ENR and REF si tes, respectively. Benthic C:N ratios ranged from 11.5 to 14.4 and were significantly lower at the reference sites (6 of 12 comparisons), with the exception of ENP-TS. The largest range of N:P values were 18.2 to 123.6 in ENP-TS, reflecting the minimum and maximum values among the areas. N:P values increased significantly with increasi ng distance from the canal in all areas.

PAGE 75

66Table 4-1. Benthic nutrient, enzyme, and ratio parameters in P enriched and referenc e sites in LNWR and WCA-2A. Values presented are means averaged over the three sampling periods with standa rd errors. Significant differences are given between the enriched and background sites in the respective areas. Each asterisk re flects differences present at the p<0.05 level per sampling period. Analysis was divided into the three sampling periods (1, 2, 3). Glucosidase, Leucine aminopeptidase, phosphatase units are moles AMC or MUF released g-1 AFDM h-1. Phenol oxidase and peroxidase units are moles DICQ released g-1 AFDM h-1. C:P= Total organic carbon to to tal phosphorus, C:N=Total organic carbon to total nitrogen, N:P=To tal nitrogen to total phosphorus LNWR WCA-2A Parameter Units Enriched Reference (P<0.05) Enriched Reference (P<0.05) (n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3 Total Phosphorus g kg-1 1.65 0.41 0.59 0.02 * 1.36 0.17 0.43 0.04 * Total Nitrogen g kg-1 34.99 3.59 40.14 1.40 * 30.99 1.38 31.67 0.61 Total Organic Carbon g kg-1 440.44 12.37 467.33 20.34 447.00 26.53 410.94 22.86 Calcium g kg-1 44.00 10.50 13.78 0.78 * 41.72 10.72 42.39 10.52 Magnesium g kg-1 3.96 0.28 2.37 0.11 3.77 0.49 3.68 0.12 C:P 312.6 62.1 796.2 9.3 * 347.7 70.6 978.3 140.2 * C:N 12.8 1.0 11.7 0.2 * 14.4 0.3 13.0 1.0 * N:P 23.6 3.2 68.2 1.0 * 24.0 4.4 74.3 5.4 * Lignin % 40.5 1.8 32.4 1.3 42.5 1.2 33.4 7.5 Cellulose % 13.5 0.9 12.8 0.0 17.4 1.6 18.1 3.9 Glucosidase 0.27 0.01 0.27 0.03 0.23 0.01 0.11 0.01 Leucine aminopeptidase 3.49 0.17 4.92 0.40 3.00 0.17 1.51 0.08 Phosphatase 2.77 0.29 16.21 0.39 * 3.18 0.64 11.20 2.56 Phenol oxidase 59.45 7.93 77.46 5.02 25.05 2.48 18.54 1.21 Peroxidase 21.94 4.45 15.41 8.30 6.83 1.10 4.43 2.25 Ecell/Ep 31.44 14.11 0.97 0.24 * 38.98 20.59 0.39 7.47 * Ecell/En 0.57 0.14 0.37 0.06 0.63 0.15 0.20 0.10 Ecell/Eox 4.51 2.47 4.30 2.51 7.14 2.28 17.58 2.09

PAGE 76

67Table 4-2. Benthic nutrient, enzyme, and ratio parameters in P enriched and referenc e sites in WCA-3A and ENP-TS. Values presented are means aver aged over the three sampling peri ods with standard errors. Si gnificant differences are giv en between the enriched and background sites in the respectiv e areas. Each asterisk refl ects differences present at th e p<0.05 level per sampling period. Analysis was di vided into the three sampling peri ods (1, 2, 3). Glucosidase, Leu cine aminopeptidase and Phosphatase units are moles AMC-MUF released g-1 AFDM h-1. Phenol oxidase and Peroxidase units are moles DICQ released g-1 AFDM h-1. WCA-3A ENP-TS Parameter Impacted Reference (P<0.05) Impacted Reference (P<0.05) (n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3 Total phosphorus g kg-1 1.66 0.25 0.48 0.05 * 0.74 0.20 0.10 0.01 * Total nitrogen g kg-1 33.16 1.12 37.53 1.95 13.2 4.77 11.98 0.90 Total organic carbon g kg-1 438.11 9.99 427.67 16.88 173.63 46.82 168.00 8.19 * Calcium g kg-1 64.72 39.35 25.78 6.34 166.44 58.77 191.11 28.04 Magnesium g kg-1 1.17 0.10 1.13 0.16 1.43 0.05 2.44 0.16 C:P 283.7 48.8 840.8 5.4 * 236.3 21.6 1803.2 137.9 * C:N 13.2 0.3 11.5 0.3 13.1 1.5 14.4 0.1 N:P 21.6 4.3 75.2 0.4 * 18.2 1.1 123.6 8.3 * Lignin % 21.6 4.0 26.5 2.7 20.1 11.1 7.2 4.5 Cellulose % 15.4 4.7 11.5 0.4 20.9 4.8 8.4 5.0 Glucosidase 1.07 0.29 0.26 0.02 0.67 0.01 0.15 0.01 Leucine aminopeptidase 4.39 0.65 3.23 0.10 8.13 0.79 3.07 0.05 * Phosphatase 2.32 0.56 18.09 3.23 * 11.19 3.18 61.33 0.48 * Phenol oxidase 186.6 46.0 105.2 4.1 287.3 89.1 94.9 2.3 Peroxidase 10.8 4.3 17.3 4.1 10.8 2.2 10.4 2.2 Ecell/Ep 53.20 22.73 1.51 0.61 * 3.47 0.80 0.11 0.02 * Ecell/En 1.33 0.14 0.57 0.09 * 1.59 0.41 0.44 0.09 * Ecell/Eox 4.22 1.01 1.44 0.46 2.22 0.43 0.96 0.20 *

PAGE 77

68 Lignin and cellulose data is based up on data from the 2nd and 3rd sampling periods. Lignin values ranged from 7% to 43% Decreases in lignin content occurred in LNWR, WCA-2A, and ENP-TS with increasi ng distance from the canal. The lowest lignin value was associated with the REF EN P-TS site as was the lowest cellulose content. Lignin was correlated with TP, TN, TOC, and negatively correlated with Ca. Cellulose content ranged from 8% to 21% w ith mixed relationships along the gradient. Various hydrolytic enzyme activities were asso ciated with the P gradients in each area. Potential benthic BGL activities ranged from 0.11 to 1.07 (Table 41 & Table 4-2), and were higher in the ENR sites. However, thes e differences were only significant in 3 of 12 cases. There was no change in BGL along a gr adient observed in an earlier study in WCA-2A (Wright and Reddy, 2001b). WCA-2A generally exhibited the lowest BGL activities. BGL was correlated with TP, LEU, and PHE. Simila rly, LEU significantly decreased between the ENR and REF sites in on ly 1/3 of the cases, with the exception of LNWR, and ranged from 1.51 to 8.13 moles g-1 AFDM h-1. The lowest LEU values were also generally found in WCA-2A and were only significantly correlated with PHE. PHO was significantly higher in the reference sites in 75% of the comparisons and ranged from 2.32 to 61.33 moles MUF released g-1 AFDM h-1. The decrease in PHO with increased proximity to the canal has been obs erved in other studies in the Everglades (Wright and Reddy, 2001b) as well as with decreased phosphata se activities in periphyton associated P loading. Significant correlations were found with TP, TN, TOC, and lignin. The highest activitie s were associated with the ENP-TS sites, reflecting the lowest benthic TP values.

PAGE 78

69 Table 4-3. Benthic layer nutri ent and enzyme correlation coefficients. All values are significant at the p<0.05 level. TN=t otal nitrogen, TP =total phosphorus, TOC=total organic carbon, C:P=rati o of TOC to TP, Ca=calcium, Mg=magnesium, GLU=glucosidase, E cell/En= apparent N control on carbon mineralization, Ecell/Ep=apparent P c ontrol on carbon mineralization LEU= leucine aminopeptidase, PHO=phosphatase, PHE=phenol oxidase, CRR=cotton rottenness rate. TP TN TOC Lig Cell Ca Mg C:P Ecell/ En Ecell/ En BGLLEUPHO TN .69 TOC .68 .98 Lignin .56 .66 .69 Cellulose .40 .41 Ca -.45 -.80 -.79 -.81 Mg C:P -.78 -.41 Ecell/En -.36-.36 Ecell/Ep .76 .41 .44 .39 -.63.46 GLU .39 -.54.57 .42 LEU .60 PHO -.66 -.48 -.48 -.50 .10 .44 -.91 PHE -.23 .37.51 CRR .70 .65 -.69 .62 -.57 Benthic phenol oxidase activit ies decreased between the ENR and REF sites in all areas but LNWR. However, this decreas e was only significant in 25% of the comparisons. A lack of significant changes in PHE was previously documented within WCA-2A (Wright and Reddy, 2001b). The lowest PHE activities were present in WCA2A with the largest range of activities in ENP-TS. Benthic PHE was correlated with BGL and PHO. Peroxidase activities were not significantly diffe rent between the ENR and REF sites and were not correlated with other enzymes or nutrients. Average benthic Ecell/Ep values decreas ed with distance from the canal inflow (Tables 4-1 & 4-2) and were si gnificantly lower in the refere nce sites in 11 of 12 cases. Values were as much as much as 20 times hi gher than those reported in an aquatic study

PAGE 79

70 (Sinsabaugh et al., 1997). The lowest Ecel l/Ep values were within ENP-TS, reflecting higher PHO activity in relation to BGL as well as the low TP values. Ecell/Ep was correlated with the nutrient parameters TP, TN TOC, and lignin. The correlations with the enzymes inclusive in the ratio formul ation, BGL and PHO, indicate the greater variability of phosphorus in this study. Benthic Ecell/En significantly increased with P loading in all four areas in 5 of 12 comparisons. The largest range occurred within EN P-TS. There were no correlations to components that were not related to the rati o. The most significant changes in benthic Ecell/Eox values occurred within ENP-TS, w ith higher values at the enriched site. Conversely, Ecell/Eox was signifi cantly higher in one time peri od at the reference site in WCA-2A. Ecell/Eox was not significantly correlated with any other parameters. Enzyme Index of Carbon Quality (EICQ) valu es generally did not predict increased C quality at the enriched site s, with the exception of WC A-2A and WCA-3A, though these differences were not significant. Soil Data Average 0 to –10 cm soil TP values ranged from 0.2 to 1.0 g kg-1 with significantly lower concentrations in 7 of 12 cases associated with the REF sites (Tables 4-4 & 4-5). The lowest TP va lues were associated with the ENP-TS sites. Soil TP was correlated with TN, TOC, and lignin, and wa s generally lower than the benthic layer (Table 4-6). The lowest TP and TN values we re both associated with the ENP-TS sites. TN mean concentrations ranged from 8 to 37 g kg-1 and were significantly different in 5 of 12 cases with 2 of 3 instances reported in ENP-TS. TN was strongly correlated with TOC, lignin, and negatively correlated with Ca. There was no significant change in TN content with depth. TOC ranged from 86 to 473 g kg-1 with significantly lower values

PAGE 80

71 once again associated with the ENP-TS site s in 1/3 of the comparisons. Soil TOC was correlated with lignin and ne gatively correlated with Ca. There was no consistent relationship between TOC content and depth. Calcium concentrations ranged between 13 and 220 g kg-1. The ENP-TS Ca concentrations were as much as 17 times gr eater than other areas. In relation to the gradient, Mg and Ca were not significantly lo wer at the reference sites. Mg values ranged from 0.9 to 3.6 g kg-1. Ca was also negatively corr elated with lignin. Neither Ca nor Mg demonstrated consiste nt relationships with depth. Soil mass-mass C:P ratios were significantly higher (6 of 6 cases) at the WCA-3A and ENP-TS reference sites, reflecting the in fluence of the P gradient. C:P was weakly correlated with both TN and lignin, and was highe r in the soil layer, with the exception of ENP-TS. Conversely, soil C:N values were si gnificantly lower at 50% of the reference sites and ranged from 11 to 16. C:N ratios did not vary significantly within ENP-TS along the gradient. Like C:P, the soil C:N va lues were the lowest at the ENP-TS sites and exhibited higher values in the soil layer, with the excep tion of ENP-TS. N:P values were significantly higher at the reference si tes in 11 of 12 comparisons and were the lowest in the ENP-TS sites. Additionally, soil N:P values were greater than the benthic layers, with the exception of the ENP-TS enri ched site. These increases with soil depth are analogous to increases in N:P observed with increasing mass loss (Sinsabaugh et al., 1993). C:P and N:P values were comparable to soil data gathered in WCA-2A while C:N values were on the lower range (Reddy et al ., 1993). Lignin data ranged from 11% to 52% in the soil layer and generally increas ed with depth. Cellulose content also

PAGE 81

72 increased with depth, ranging from 13% to 27% and was generally hi gher at the reference sites, with the ex ception of ENP-TS. Compared to the benthic layer, hydrolytic enzyme activities in the soil layer had fewer significant differences al ong the P gradient (Tables 4-4 & 4-5). Average soil layer BGL activity was consistently lower than the benthic activity. Soil BGL ranged from 0.05 to 0.39 moles g-1 AFDM h-1 and decreased from the ENR to REF site in all areas with significant differences in 2 of 3 sa mpling periods in WCA-3A and ENP-TS. The largest contrast between the ENR and REF sites occurred in ENP-TS. Soil BGL was negatively correlated with Mg. Soil leucine aminopeptidase (LEU) activity was lower in 10 of 12 sites in comparison to the benthic layer and ranged from 1.92 to 3.58 moles g-1 AFDM h-1. LEU was generally not significan tly different between the enriched and reference sites. LEU was negatively correlated with TN, TOC, and lignin. There was not a consistent relationship between LEU activity and the P gradient. Soil PHO also did not significantly vary between the REF and ENR sites. Values were consistently lower in the soil layer, reflecting activities as great as 10 times lower those in the corresponding benthic layer. This depth relationship has been previously documented in other areas (Newman and Reddy, 1993; Wright and Re ddy, 2001b). Soil PHO was negatively correlated with TP, TN, TOC, lignin, and positiv ely correlated with Ca (Table 4-6). Soil PHE and PER activities increased with dept h. PHE was weakly negatively correlated with TN, TOC, and lignin, while PER was not correlated with any parameters. The increase of oxidative activity with depth is in disagreement with earlier studies within the

PAGE 82

73Table 4-4. Soil layer nutrient, enzyme, and ratio parameters in P enriched and reference sites in LNWR and WCA-2A. Values presented are means averaged over the three sampling periods with standard errors. Significant differences are given between the enriched and back ground sites in the respective areas. Each aste risk reflects differences present at the p <0.05 level. Analysis was divide d into the three sampling periods (1, 2, 3). Gl ucosidase, Leucine aminopeptidase, phosphatas e units are moles AMC or MUF released g-1 AFDM h-1. Phenol oxidase and peroxidase units are moles DICQ released g-1 AFDM h-1. C:P= Total organic carbon to total phosphorus, C:N= Total organic carbon to total nitrogen, N:P=Total nitrogen to total phosphorus. LNWR WCA-2A Parameter Units Impacted Reference (P<0.05) Impacted Reference (P<0.05) (n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3 Total phosphorus g kg-1 0.91 0.14 0.37 0.09 0.72 0.13 0.33 0.04 * Total nitrogen g kg-1 31.98 1.74 35.58 3.15 29.12 2.77 33.61 1.21 Total organic carbon g kg-1 471.11 3.74 462.11 28.15 443.00 17.14 464.67 8.17 Calcium g kg-1 30.33 3.10 12.61 0.20 34.83 6.77 28.6 2.69 Magnesium g kg-1 3.41 0.28 1.96 0.08 3.64 0.09 3.22 0.21 C:P 711.1 122.3 1271.9 58.7 743.9 200.3 1478.4 174.7 C:N 14.9 0.9 13.2 0.6 * 16.4 1.3 13.9 0.7 N:P 47.1 5.8 97.7 8.1 * 46.0 10.3 105.7 7.6 * Lignin % 52.1 3.1 45.7 2.1 50.1 2.8 46.5 3.6 Cellulose % 17.3 0.2 18.8 0.1 18.2 1.7 19.7 0.7 Glucosidase 0.24 0.06 0.14 0.06 0.08 0.00 0.05 0.01 Leucine aminopeptidase 1.92 0.13 2.69 0.05 1.95 0.03 2.16 0.07 Phosphatase 1.89 0.34 2.46 0.07 1.35 0.05 1.14 0.06 Phenol oxidase 110.5 8.33 223.0 45.9 78.6 4.0 142.8 23.4 Peroxidase 24.3 0.8 36.1 5.3 44.5 9.1 41.1 2.2 Ecell/Ep 18.47 8.40 8.52 3.25 13.08 5.47 0.84 0.99 * Ecell/En 0.60 0.20 0.32 0.15 0.27 0.04 0.10 0.03 Ecell/Eox 0.88 0.32 1.44 1.11 0.24 0.06 2.29 0.06

PAGE 83

74 Table 4-5. Soil nutrient, enzyme, and ratio parameters in P enriched and reference sites in WCA-3A and ENP-TS. Values presented are means averaged over the three sampling periods with standard errors. Significant differe nces are given between the enriched and backgr ound sites in the respective areas. Each aste risk reflects differences present at the p<0.05 level per sampling period. Analysis was divide d into the three sampling periods (1, 2, 3). Glucosidase, Leucin e aminopeptidase and Phosphatase units are moles AMC-MUF released g-1 AFDM h-1. Phenol oxidase and Peroxidase units are moles DICQ released g-1 AFDM h-1. WCA-3A ENP-TS Parameter Impacted Reference (P<0.05) Impacted Reference (P<0.05) (n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3 TP g kg-1 0.95 0.42 0.31 0.04 * 0.37 0.11 0.23 0.08 TN g kg-1 25.43 9.13 37.21 1.09 7.65 1.63 12.74 2.05 * TOC g kg-1 343.13 129.19 473.39 5.00 86.36 25.78 153.48 24.17 * Ca g kg-1 87.86 72.95 19.06 0.86 183.44 78.51 219.89 30.89 Mg g kg-1 1.01 0.20 0.88 0.03 1.56 0.33 2.92 0.61 C:P 489.3 142.7 1601.9 220.8 * 239.3 6.4 775.9 60.0 * C:N 13.1 0.6 12.5 0.3 * 10.9 1.1 12.0 0.1 N:P 38.7 13.2 127.4 14.7 * 20.3 0.4 62.5 6.7 * Lignin % 20.5 11.5 45.1 1.8 19.0 12.2 10.9 1.8 Cellulose % 18.0 1.6 18.7 1.6 27.4 8.4 13.3 0.1 BGL 0.39 0.02 0.21 0.02 * 0.28 0.07 0.06 0.01 * LEU 2.87 0.08 2.31 0.02 3.48 0.04 2.72 0.17 PHO 0.80 0.14 2.18 0.23 3.87 0.54 10.84 0.20 PHE 253.0 32.0 222.1 34.0 577.9 76.0 252.4 18.0 PER 19.2 3.6 32.6 13.6 62.0 14.0 24.9 13.7 Ecell/Ep 90.19 39.15 9.37 3.13 * 2.74 0.71 0.32 0.10 * Ecell/En 1.28 0.27 0.57 0.11 0.63 0.23 0.21 0.07 * Ecell/Eox 4.97 3.19 1.29 0.47 1.86 1.20 0.34 0.16

PAGE 84

75 Everglades (McLatchey and Reddy, 1998). So il PHE was highest in ENP-TS. Neither enzyme was observed to have a signif icant response across the gradient. Soil Ecell/Ep values were lower in th e reference sites in each area, however statistical significance was generally limite d to WCA-3A and ENP-TS. The lowest Ecell/Ep values occurred in ENP-TS. Soil E cell/Ep was higher than the benthic layer at the reference sites and lower at the enrich ed sites, with the exception of WCA-3A. Ecell/Ep was negatively correlated with PHE. As in the benthic layer, PHO was more linearly related to the Ecell/Ep ratio (r2 =-0.65) than BGL (r2 =0.32) (Table 4-6). Soil Ecell/En was also lower in the reference sites, although significant differences were confined to ENP-TS. The highe st Ecell/En and Ecell/Eox values occurred within WCA3A. Ecell/Eox was not significantly different along the gradient. Additionally, Ecell/Eox values were generally lower in the soil layer, possibly reflecting lowe r litter quality based on perceived lignin content. EICQ values pr edicted greater carbon qual ity at the enriched sites in WCA-2A, WCA-3A, and ENP-TS, although these differences were not significant. CRR Cotton Rottenness Rate (CRR) was averaged over the three sa mpling dates (Table 4-7). Benthic and soil CRR decreased with increasing distance from the canals in all areas. The greatest CRR in both the benthic and soil layers occurred at the WCA-2A ENR site. However, the most dynamic change in CRR was associated with the ENP-TS benthic layers between the enriched and refe rence sites. There were not consistent differences between the benthic and soil la yers and values were within the ranges reported in the mesocosm P loading expe riment in LNWR (Newman et al., 2001).

PAGE 85

76 Table 4-6. Soil layer nutrient and enzyme correlation coefficients. Values in bold are significant at th e p<0.0001 level, all other va lues are significant at the p<0.05 level. Li g=lignin, Cell=cellulose. TN=t otal nitrogen, TP=total phosphorus, TOC=total organic carbon, C:P= ratio of TOC to TP, Ca=calcium, Mg=magnesium, GLU=glucosidase, E cell/En= apparent N control on carbon mineralization, Ecell/Ep=apparent P c ontrol on carbon mineralization LEU= leucine aminopeptidase, PHO=phosphatase, PHE=phenol oxidase, CRR=cotton rottenness rate. TP TN TOC Lig Ca Mg Ecell/ En Ecell/ Ep PHO TP TN .63 TOC .63 .99 Lig .48 .76 .79 Cell Ca -.59 -.62 -.78 Mg C:P .48 .51 .41 EC/EN .37 -.49 EC/EP .53 .61 .37 .62 BGL -.44 .93.57 LEU -.69 -.54 -.39 PHO -.49 -.61 -.62-.51.57 -.81 PHE -.37 -.40 -.39 -.48 .57 CRR A model was created using enzyme data to test the validity of prediction against actual CRR values. Ecell/Ep and TP were found to be the most powerful predictors of CRR, accounting for 62% of the variability in the benthic layer. The resulting benthic model: Predicted CRR= (0.04(LnEc:Ep)+1.4) + (0.35(LnTP)+1), was strongly correlated in sampling period 1 (r2 =0.92) and more weakly corre lated in sampling period 3 (r2=0.46). This benthic model predicte d significantly greater cellulose decomposition at the enriched sites in all four areas at all time periods (p< 0.0001). A separate significant model could not be created for the soil laye r. However, the benthic model predicted significantly greater decomposition in the en riched sites in 75% of the contrasts.

PAGE 86

77 Impact Index An impact index was constructed to compare the varying parameters on the same scale and to allow the comparisons between th e enriched and reference sites within each area (Table 4-8). In general, impact index values above 0.5 reflect severe changes in a specific parameter between the enriched and reference sites. Positive values reflect increased values at the enriched site while negative values reflect higher values at the reference site. The equation for each parameter is: Impact Index = log ((Enriched P site) / (Reference P site)) As expected, the most responsive nutrient parameters to the P gradient were the Prelated biogeochemical indicators TP, C:P, N: P, PHO, and Ecell/Ep in both the benthic and soil layers. TOC showed relatively little change in relation to the P gradient while TN demonstrated a small response in the soil layer. Differences were also observed with BGL in both layers. The majority of large imp act indices in the benthic layer were within ENP-TS, even though there is a less pronounced P gradient in this area. WCA-2A, a heavily impacted area, did not exhibit the la rge differences that were expected. There were greater index values associated with the benthic layer, indicati ng that this layer is more responsive to changes in biogeochemical conditions. Transitional Sites Large increases in several of the indi vidual enzyme activities occurred at the transitional sites in the four hydrologic units (Figure 4-1a & 4-1b). The greatest increases in benthic activity occurred at the WCA-2A TRANS site with BGL and LEU with 10 of 12 comparisons significant. Spikes in LEU were also observed within the transitional sites in LNWR, WCA-2A, and EN P-TS. PHO also increased in the soil layer WCA-2A,

PAGE 87

78 Table 4-7. Average benthic and soil layer CRR values over the three sampling periods with standard errors. Soil layer CRR values represent 2 of 3 sampling periods due to insufficient de pth deployment of the cotton st rip racks. ENR=Enriched, Trans=Transitional, REF=reference sites. Benthic layer Soil layer LNWR ENR 23.2 0.5 24.7 0.5 LNWR TRANS 22.9 0.8 21.0 0.7 LNWR REF 20.3 0.8 19.4 0.3 WCA-2A ENR 25.7 0.1 25.4 0.2 WCA-2A TRANS 20.7 0.7 18.1 1.0 WCA-2A REF 18.5 1.3 18.7 1.1 WCA-3A ENR 24.7 0.1 25.1 0.1 WCA-3A TRANS 23.5 0.6 22.6 0.4 WCA-3A REF 17.2 1.1 20.7 1.0 ENP-TS ENR 24.6 0.5 22.2 0.2 ENP-TS TRANS 17.8 0.8 20.9 0.6 ENP-TS REF 16.3 1.3 19.6 0.6

PAGE 88

79 a.) Benthic layer 0.0 0.1 0.2 0.3 0.4 0.5 1enr1trans1ref2enr2trans2ref3enr3trans3ref4enr4trans4ref Ec En 1TRANS4TRANS 3TRANS 2TRANS b.) Soil layer 0.00 0.05 0.10 0.15 0.20 1enr1trans1ref2enr2trans2ref3enr3trans3ref4enr4trans4ref En Ep 1TRANS4TRANS 3TRANS 2TRANS Figure 4-1. Graphs showing spikes in enzyme ac tivities at the transitional sites. Mean enzyme activities have been normali zed for visual comparison. Ec and En, normalized glucosid ase and leucine aminopeptidas e activities exhibited the largest spikes in the benthic layer while En and Ep (phosphatase) exhibited the larg est in the soil layer. Bars represent standard errors. 1=LNWR, 2=WC A-2A, 3=WCA-3A, 4=ENP-TS. Enr=enriched, trans= transitional, ref=reference sites.

PAGE 89

80 Table 4-8. Impact index (log(impacted/refer ence)) for nutrient and enzymic data in th e soil and benthic laye rs for LNWR, WCA-2 A, WCA-3A, and ENP-TS for averages cal culated over the three sampling periods Bolded items reflect parameters that are equal to or greater than 0.5, indi cating large changes over the gradient. Negative values indicate larger p arameters at the reference sites while positi ve values reflect larger valu es at the enriched sites. Benthic Soil LNWR WCA-2A WCA-3A ENP-TS LNWR WCA-2A WCA-3A ENP-TS Parameter TP 0.42 0.49 0.53 0.83 0.38 0.34 0.29 0.22 TN -0.06 -0.01 -0.05 0.04 -0.04 -0.07 -0.38 -0.23 TOC -0.03 0.04 0.01 -0.02 0.01 -0.02 -0.25 -0.28 Ca 0.48 -0.01 0.27 -0.14 0.38 0.07 0.22 -0.24 Mg 0.22 0.00 0.02 -0.23 0.24 0.06 0.05 -0.28 C:P -0.42 -0.46 -0.47 -0.88 -0.27 -0.36 -0.54 -0.49 C:N 0.04 0.05 0.06 -0.05 0.05 0.05 0.02 0.00 N:P -0.5 -0.5 -0.53 -0.83 -0.32 -0.40 -0.56 -0.49 Lignin 0.10 0.12 -0.09 0.47 0.06 0.03 -0.47 0.07 Cellulose 0.02 -0.01 0.10 0.48 -0.04 -0.04 -0.02 0.29 BGL -0.04 0.45 0.56 0.66 0.15 0.27 0.26 0.56 LAP -0.19 0.27 0.13 0.24 -0.16 -0.06 0.08 0.13 PHO -1.06 -0.49 -0.95 -0.81 -0.03 -0.16 -0.44 -0.44 PHE -0.29 -0.04 0.21 0.51 0.02 0.20 0.10 0.20 PER 0.21 0.03 0.13 -0.03 -0.13 -0.01 -0.11 0.39 Ecell:Ep 1.03 1.04 1.54 1.42 0.19 0.42 0.85 0.96 Ecell:En 0.13 0.24 0.41 0.44 0.28 0.32 0.29 0.47 Ecell:Eox -0.09 0.28 0.50 0.40 0.28 0.19 0.38 0.48

PAGE 90

81 WCA-3A, and ENP-TS sites. However, these spikes were not correlated with increased CRR. The significant benthic BGL and LEU increas e at the WCA-2A transitional site is accompanied by a generally much lower TN (31 g kg-1), lignin (21%), cellulose (9%), the lowest TOC (352 g kg-1) of any site not in ENP-TS, as well as elevated Ca (81 g kg-1) and Mg (81 mg kg-1 ). This site is also dominated by Chara spp ., which is a characteristic algae in transitional P sites and is a level II indicator of P enrichment (U.S. EPA, 2002). The spike in the benthic LNWR transitional site is also accompanied by a lower TN (31.23 g kg-1) and TOC (438.1 g kg-1). Increases also occurred in the soil LEU and PHO at the transitional sites in WCA2A and ENP-TS. The lowest TP (0.2 g kg-1) and TN (6 g kg-1) of any site was recorded at the transitional site in ENP-TS which acc ounts for the greatest PHO and LEU activity of any soil layer. This relationship in the soil layer is not as clear at the WCA-2A transitional site. Discussion The following discussion has been groupe d into several sections that first compare the general effects of the P gradient between the enriched and reference sites. The discussion generally pertains to the benthi c layer results, as this layer is the most biologically active (White and Reddy, 2000) a nd is more responsive to changes in environmental conditions. Secondly, specific hydrologic units are compared in relation to their response to the P gradient. Lastly, a brief discussion of the observations made at the transitional sites is included.

PAGE 91

82 Phosphorus Gradients The nutrient and enzyme parameters of the benthic layer varied in differing degrees along the nutrient gradient. The correlati ons between TOC-TP and TOC–TN reflect increased vegetative input associated with increased TP values in both the benthic and soil layers. Aboveground production in cat tail stands ranges from 3035 to 1077 g m2 y-1 in the enriched and reference areas in WCA-2A, respectively, while sawgrass ranges from 1943 to 986 g m2 y-1 (Davis, 1991). These changes in productivity and thus nutrient input from vegetative sources are reflected in soil nutrient contents. Decreases in P with increasing distance from the canals are reflected in the larger C:P values at the reference sites in both layers. Increases in C:P rati os in WCA-2A have been correlated with decreased C turnover rates (Koch and Reddy, 1992), while lower C:P ratios have been correlated with enhanced microbial respiration (Ama dor and Jones, 1993). While individual enzyme activities can provide some insight into microbial nutrient perceptions, relative indices may provide a more productive look at community responses to environmental ch ange. Ecell/En, Ecell/Ep, a nd Ecell/Eox relationships in the different areas are presented in Figure 4-2. The higher Ecell/Ep values at the enriched sites reflect a decrease in apparent P c ontrol on C mineralizati on. The greatest shift within WCA-3A suggests that this area ha s the largest P imp act on the microbial community. The increase in Ecell/Ep that occurs in WCA-3A ma y be due to a the weighting effect of a higher C mineralizati on rate. This may be due to decreased inhibition from polyphenols, a ge nerally lower water level du ring the sampling events, or the lower lignin content at the WCA-3A enrich ed site. The lowest shift in Ecell/Ep occurs in ENP-TS (Figure 4-2), suggesti ng that the microbial communities are less affected across the gradient by changes in P conditions. The impact index does not

PAGE 92

83 Enr LNWR Ref LNWR Enr WCA-2A Ref WCA-2A Enr WCA-3A Ref WCA-3A Enr ENP-TS Ref ENP-TS-10 0 10 20 30 40 50 60 70 00.20.40.60.811.21.41.61.82 Ecell/EnEcell/Ep Figure 4-2. Benthic layer bubble plot of Ecell/Ep vs. Ecell/En with Ecell/Eox represented by th e bubble size. Arrows denote shifts from the enriched to the corresponding reference sites. support this by suggesting that the highest TP impact (0.83) occurs along the ENP-TS gradient. However, the TP values are lower at both the enriched and reference sites, so the increases may not be enough to induce reso urce allocation shifts along the gradient. In this case it does not appear that the imp act index accurately reflects the severity of impacts in relation to significant changes that may be occurring within the communities. The greatest shift in P influence on C minera lization therefore app ears to occur within WCA-3A with the smallest shift within ENP-TS. Ecell/En values were higher at all the enriched sites, once again to differing degrees. The largest shift occurred within ENP-TS (Figure 4-2), which also had the highest Ecell/En impact index. The higher Ecell/En and Ecell/Ep values at all the enriched sites are concurrent with decreases in nutrient limitations that have been documented (McCormick et al., 1996), as well as lower P mineralization rates in enriched areas of the Everglades (Newman et al., 2003; White and Reddy, 2000).

PAGE 93

84 Lower Ecell/En values at the reference sites may be attributed to shifts from primarily macrophyte to algal i nputs at the reference sites (F igure 4-2). This would be manifested in higher LEU activity, in relati on to BGL, since algae generally contain a higher protein content than macrophytes (B oschkler and Cappenberg, 1998). This is especially important since LEU may also function in C mineralization from protein sources. One driving force behind higher Ecell/E n values at the enriched sites may be the generally elevated BGL activities, which ha s been correlated with microbial production in nutrient enriched mesocosms (Chr st and Rai, 1993). Overall, the increases in Ecell/En at the enriched sites mimic a southerly pattern: LNWR
PAGE 94

85 The increase in Ecell/Eox values at the en riched sites, with th e notable exception of WCA-2A, indicate decreased apparent lignin control on C mineralization. This would indicate a greater C mineralization potential. However, since enzyme activities in the view of a resource allocation model are interdependent, othe r factors may be influencing these trends. The higher Ecell/En values at the enriched sites in the four areas are a response to increased perceived N content. This is not apparent in the TN content of the soils which suggests that this parameter is not well suited for judging N availability, since there are varied forms of N within the syst em. A large portion of endogenous N may also be shielded by refractory compounds within the litter (Sinsabaugh and Moorhead, 1994) while exogenous N loading has been shown to repress lignin degradation (Eriksson et al., 1990; Blanchette, 1991; Sinsabaugh et al., 1993 ; Hammel, 1997; Carre iro et al., 2000) by the repression of phenol oxidase. Therefor e a connection may be established between N and lignin influence on C mineralization wh ereas individual enzyme activities do not necessarily point to this conclu sion. Therefore it appears that P loading has resulted in a generally lower level of ligni n influence on C mineralizati on, although a higher level of microbially-perceived exogenous N may be repressing oxidative enzyme activity. The combination of factors resulting from nutrient loading to the four areas of the Everglades appears to have resulted in a de crease in N, P, and lignin influence on the microbial community. This decrease of nutrient limitation on the community should result in greater mineralization rates of plan t matter, which is supported by consistently higher CRR rates at the enriched sites as well as a greater microbial biomass in another study in WCA-2A (Wright and Reddy, 2001a). The CRR decomposition model, utilizing both TP and Ecell/Ep as parameters, indicate s the relatively strong influences that the P

PAGE 95

86 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 -1.2-1-0.8-0.6-0.4-0.200.20.40.6 Ecell/EnEcell/Ep ENR REF TRANS High P, Low N Low P, High N Low P, Low N High P, High N Figure 4-3. Benthic plot of Log Ecell/En vs. Log Ecell/Ep. Each point represents the mean from one sampling period. The X axis represents apparent N influence on C mineralization, the Y axis represent apparent P influence on C mineralization. Values greate r than 1 on the Y axis indicates high perceived P availability. gradient have on cellulose decomposition rates. The coupling of Ecell/Ep and TP in the CRR model indicates the relativ e strength of the enzyme parameters BGL and PHO in predicting cellulose decomposition. This is comparable to the correlation between BGL and PHO with bacterial production in nutrien t enriched and un-enriched mesocosms, respectively (Chr st and Rai, 1993). However, as th e interactions of other components within the system indicate, CRR may not be ade quate to significantly predict actual litter decomposition rates, especially in re lation to differing litter types. The lack of consistent predictive power in EICQ values sugg ests that the dynamic changes in nutrient conditions relating to th e canal inflows may influence the significance of the model. The more complex interactions that may be affecting the communities such

PAGE 96

87 as microbial community shifts, vegetative ch anges, and differences in interactions between the heterotrophic and algal commun ities may outweigh the singular dynamics of the individual enzyme activity components of the EICQ model, espe cially in terms of lignin degradation. Therefore it does not a ppear that the EICQ model is a powerful predictor of microbial responses in this system among the different hydrologic units. Additionally, the use of only 5 en zymes in this study may be restrictive on the validity of the EICQ model. ENP-TS Special consideration is attributed to th e ENP-TS sites due to the large shift in apparent N dynamics across the gradient, as co mpared to the relatively large shifts in P dynamics in the other areas. As previously discussed, ENP-TS had the greatest shift in apparent N influence and the smallest shift in apparent P influence on C mineralization. The highest Ecell/En and lowest Ecell/Ep valu es at the ENP-TS en riched sites suggests that P is the most limiting of th e enriched sites, which is expected, and that N is the least limiting of any site, which is not expected. Although this site ha s a higher LEU activity, the relationship with the amount of C mineralization that appears to be occurring indicates that, energetically, this community is expending a smaller proportion of its resources on the acquisition of N from organic sources. The lowest overall Ecell/Eox at the ENP-TS enriched site points to increased lignin control on C mineralization. This indicates th at perceived lignin c ontent is playing an important part in the C mineralization pro cess, with increased combined oxidative enzyme activities, although % ligni n content is lowest at this site. Interactions among environmental and microbial parameters may pl ay an important part in the elevation of apparent lignin control. Elevated PHE activ ity may be partially due to the lack of

PAGE 97

88 microbially available N, reflected in elev ated LEU activity, which has been shown to possibly repress PHE activity. Overall enzyme activities within ENP-TS were not drastically lower than the other enriched sites; in fact, the majority of activities were higher. One contributing factor to this elevated activity may be due to the UV photolysis of inhibitory polyphenols and other dissolved C compounds (Wetzel et al., 1995; Boavida and Wetzel, 1998; Wetzel, 2000). It has been reported that 90% of the solar radiation is absorbed at a depth of 2.5 cm in the northern Everglades while in the surface wate r of the southern Everglades the depth was approximately 10 cm (Qualls and Richardson, 2003). Greater UV penetration at the ENP-TS enriched site may serve to alleviate the inhibition by polyphenols, resulting in more effi cient community energetics. At the ENP-TS reference site, appa rent P limitation on C mineralization is increased to the highest level of any site wh ile TP, TN, TOC, lignin, and cellulose are all the lowest. It should be noted that phosphatase activities re flect the summation of algal and bacterial expressed enzymes (Jansson et al., 1988). The production of phosphatase is most probably the primary motive force, based on the resource allocation strategy, leading to the decreased production of the C and N acquiring enzymes at this site. However, BGL activity may also be tied to the resident periphyton community. Photosynthetically produced extracellular organic carbon (EOC) from the periphyton community may supply greater amounts of labile carbon that is sufficiently degraded for direct microbial uptake (E speland et al., 2001). This may result in the markedly decreased production of BGL observe d at these sites if the major ity of DOC released is of sufficiently low molecular mass. When this readily utilizable carbon is available, there

PAGE 98

89 may be no need for microorganisms to acquire it through enzymatic action (Chr st and Rai, 1993). The supplementation with EOC, resulting in lower BGL activities, may also be exhibited among the suite of cellulase enzy mes, which would coincide with the lower CRR at the ENP-TS reference site. In comparison with the other areas, ENPTS appears to be a unique area of the Everglades that is relatively un-impacted fr om canal inflows. The relative lack of differences in apparent P limitation suggest s very little P impact s to the resident communities. The large shifts in apparent N limitation cannot easily be explained through these techniques and suggests that so me unique N anomaly is occurring near the inflow, in terms of the microbial resource allocation view. The fact that Ecell/En increases at the enriched sites with southe rly canal flow suggests that this may be a consequence of agricultural loading, runoff from the nor therly areas or processes occurring within the canals. Transitional Sites Transitional sites exhibit characteristics of both enriched and reference P regions, which may provide optimal conditions for microbial activity. The increases in enzyme activities among the transitional sites in the four areas suggest that different biogeochemical conditions support the microbial communities. Spec ifically, the lower nutrients present at the WCA-2A transitional site are even lower than the reference sites. The elevated Ca, as much as 2 times the concentration at the enriched and reference sites, may serve to relieve the inhibition by humic aci ds, which is analogous to the situation at the enriched site in ENP-TS. The lower TO C and cellulose suggest s the possibility of relative C limitations, which are not necessari ly supported by higher BGL activities, but once again must be viewed as a relationship to N and P mineraliza tion. The unique algal

PAGE 99

90 composition, with an abundant Chara spp. mat composing the majority of the benthic layer, has been shown to replace the Utricularia spp. and periphyton communities after one year of P loading (Craft and Richardson, 1995). Elevated Ca levels in the benthic and TP in the soil may be attributed to the CaCO3 and CaHPO4 bands of the cell walls of Chara spp. (Kiyosawa, 2001). The higher TP in the soil is supported by elevated Ecell/Ep, suggesting a lower P limitation on C mi neralization than the enriched site in WCA-2A. Therefore, some of the enzyme activities reflected at this site may be attributed to the metabolic a nd structural attributes of th e resident algal community. Several reasons may account for the nutrien t dynamics at the transitional sites. Upstream porewater dynamics may be exerting pressures on the transitional site. The increase in TP in the soil layer, which is gr eater than that at the enriched site, suggests that P is possibly being released from the soil at sites closer to the inflow. This may be due to the reduction of P loading from the canal s in recent years to a concentration lower than that of the soil. The deposition of P downstream appears to result in the expansion of the P front and may continue until a relativ ely static equilibrium is reached. Secondly, and more specifically to the WCA-2A transitio nal site, the elevated TP may be due to P removal capabilities of the Chara spp. communities and eventual burial into the soil layer. The most consistent increase at the tran sitional sites in both the soil and benthic layers was in relation to LEU. The decrease of average TN at the transitional sites in 6 of 8 cases supports these elevated activities. However, WCA-2A appears to be unique in this regard. This is the only area, in both la yers, to show an increase in Ecell/En at the transitional site, indicating that N is less lim iting to the microbial community as related to C mineralization.

PAGE 100

91 Conclusions Interpreting individual enzy me activities is often difficult due to the wide range of factors that control the induction and repressi on of activities. Add itional factors include inhibitory and enhancing effects of organi c and inorganic compounds as well as other environmental influences. The use of enzyme ratios that express activities in terms of microbially perceived C, N, a nd P relationships in conjunction with nutrient data appears to assist in beginning to understand so me of the complex relationships. The results of this study indi cate that great care is requi red in any interpretations of raw enzyme activities to higher order processe s such as decomposition processes. Local and regional influences must be accounted for before any conclusions are drawn. The comparison of enzyme activities to decompositi on rates, either using litter bags, cotton strips, or other methods can a ssist in linking microbial activi ties to some of these higher order processes. The relative impacts of P enrichment on microbial enzyme activities differs among the hydrologic units of th e Everglades but generally re sults in decreased apparent N and P control (higher Ecell/Ep and Ecell/ En) and greater cellulo se decomposition rates (CRR). Site characteristics appear to contribute sharply to these differences. Plant matter induction, algal composition, UV phot olysis of humic substances, and photosynthetically-produced extracellular organi c carbon are some of the factors that are hyothesized to contribute to these differenc es. However, the increases in primary production, resulting from a gr eater C flow, may exceed any increases in decomposition, resulting in a net accumulation of organic matter (Davis, 1991; Reddy et al., 1993; Craft and Richardson, 1993).

PAGE 101

92 It appears that the enzyme dynamics at the ENP-TS enriched site are mainly driven by the effects of a smaller P input relative to the other hydrologic units of the Everglades. The smaller P input has resulted in a smalle r increase in NPP than the other enriched sites, leading to a relative C limitation in a ddition to maintaining a relative P limitation. This lower C input leads to a decreased input of detritus in the wate r and a more pristine water column which may increase UV inactiv ation of inhibitory polyphenols. Nitrogen needs are still being met as the microbial community does not appear to be limited by N in relation to C mineralizati on. Hypothetically, this c ondition would continue until increases in net primary production (NPP) e qual or exceed microbial C mineralization. The transitional sites frequently exhib it unique characteristics, manifested in elevated enzyme activities that appear to re flect changes that may be occurring in terms of structure and function of the algal, heterotrophic, and macrophytic communities. Variations in nutrient contents and enzyme ac tivities at these sites from what would be expected from a quasi-linear reduction in nutri ent contents from the enriched to reference sites suggests that there are more complex interactions occurring that may be early indicators of eutrophication. The exponential model, using CRR as the defining decomposition criteria, utilized Ecell/Ep and TP as the major derivatives. Ho wever, the use of cott on strips restricts the model to account for only cellulose degradati on. Litter bags or similar methods may be more suitable to develop a more robust enzyme model. Additionally, the use of a greater range of enzymes and more robust nutrient analyses may increase the resolution of the model as well as provide a more thorough unde rstanding of the relatio nships within the system. Lastly, the fractionation of the alga l, plant, and microbial decomposer enzyme

PAGE 102

93 activities would allow for a much greater reso lution into the sources and consequences of resource allocation dynamics.

PAGE 103

94 CHAPTER 5 MICROBIAL ENZYME RESPONSES TO DECREASED WATER LEVELS IN EVERGLADES PEAT AND MARL SEDIMENTS Introduction Generally anaerobic conditions in wetlands are due to the decreased diffusion of O2 in water as well as increased demand as a consequence of high carbon (C) availability. This results in an overall reduction in decomposition (Reddy and D’Angelo, 1994) which accounts for the role that inundated wetlands play in C sequestration. Organic matter accumulation in wetlands may be significantly affected by changes in water levels as a result of the characteristic hydroperiod which encompasses frequency, duration and depth. Organic mineralization rates are contro lled by the microbial community that exert an appreciable influence on the energy flow of a system (Elliot et al., 1984) and have the potential to be affected by water level fluctuations. The microbial degradation of organic matter been shown to be most influenced by the enzymes involved in li gnocellulose degr adation, P cycling, and N cycling (Sinsabaugh et al., 1991; Sinsabaugh and M oorhead, 1996), which ar e often considered the rate limiting step s in degradation (Chr st and Rai, 1993). Due to their functional role in the degradation of organic matter, enzyme activities may be especially significant in determining changes in environmental conditi ons through their subsequent effects on the resident microbial community (Frankenberg er and Dick, 1983; Dick, 1997). Enzyme activities have the potential to affect all major wetland functions where decomposition is low: Peat accumulation is dependent upon a lowe r rate of enzymatic ac tivity resulting in

PAGE 104

95 C storage and inorganic nutrients remain se questered within the poorly degraded peat matrix when decomposition is low. This im pairment of nutrient cycling causes inorganic nutrients to accumulate and to be retained within the wetland (Freeman et al., 1996). Phenol oxidase activity, which controls lignin degradation and is dependent on O2 availability, has demonstrated varying resu lts to water level decreases and has been referred to as an “enzymic latch” control ling C mineralization (Freeman et al., 2001). Phenol oxidase (PHE) has been observed to increase with depth (Lhdesmki and Piispanen, 1988), decrease w ith decreased dissolved O2 availability (Pind et al., 1994), exhibit no discernible variability with depth (Duxbury and Tate, 1981), increase in drought conditions (Freeman et al., 2001) and not respond predictably to drought conditions (Williams et al., 2000b). The mech anisms behind PHE variability may also include induction by the pres ence of certain phenolic materials among other inducible and repressible controls. Other studies have investigated the e ffects of a water le vel decrease on flooded sediment microbial respirati on. Linear increases in CO2 flux were found with lowered water table depth down to 50 cm in Evergl ades sawgrass peat sediments (Volk, 1973). Increased microbial biomass a nd C flux in drought conditions has also been observed in peatland cores (Blodau et al., 2004). The highe st total C flux as a function of combined CO2 and CH4 evolution was found under controlled drainage conditions 15 cm below an Everglades peat surface and was thought to be time dependent due to the water holding capacity of the peat matrix (DeBusk, 1996). The mineralization of soil organic nitr ogen (N) sources has been found to be greater in aerobic than an aerobic conditions (Reddy and Patrick, 1984; McLatchey and

PAGE 105

96 Reddy, 1998) and to increase afte r the drying of wet soils (Cabrera 1993; Bridgham et al., 1998). N ammonification rates were specificall y found to be 2-3 times greater in drained wetland soils (White and Reddy, 2001; Venterlin k et al., 2002). Be nthic ammonification rates have also been shown to be approximate ly 2 times greater in Everglades sediment cores than the corresponding deep er layers (White and Reddy, 2001). The objectives of this study were to determine; (1) the response of enzyme activities to a controlled water level decrea se in soil cores extracted from a marl dominated wetland (Taylor Slough in Everglad es National Park (ENP-TS)) and a peat dominated wetland (Water Cons ervation Area 3A (WCA-3A)); (2) the differences in response to drained and flooded conditions between the two sites; and (3) any effects that an extended laboratory incubati on has on enzyme activities. Materials and Methods Site Description The Florida Everglades is an oligotrophic system with referen ce surface water total phosphorus (TP) levels averaging less than 10 g L-1 throughout the inte rior of the marsh (McCormick et al., 2003). Field study sites we re located within the interior of WCA-3A and ENP-TS. WCA-3A encompasses 2,012 km2 and is predominantly a vast peat sawgrass marsh interspersed with sloughs tree islands, and wet prairies. It is the only area not completely enclosed by levees. The highest annual mean surface water TP levels in the inflow and interior marsh were 67.3 and 20.3 g L-1, respectively for 1978-2000 (Newman et al., 2002). The vegetation at the sampling site is generally composed of water lilies, spikerush, and periphyton assemblages.

PAGE 106

97 Everglades National Park is a 5,569 km2 wetland consisting primarily of marl forming wet prairies, sawgrass stands, fres hwater sloughs, and mangrove stands at the southern periphery. Marl forming prairies are characterized by the formation of calcitic mud, especially in the southern regions. The sampling site is characterized by dense periphyton assemblages, epiphyton and spikerush. Sampling Soil cores were obtained using a 12 cm th in-walled, serrated edge stainless steel corer with butyrate sleeves on March 3rd, 2003. Th irty cores were collect ed at each site. The coring procedure involved pushing the coring apparatus through the soil layer to a depth of approximately 50 cm. During inser tion a serrated metal knife was used to cut around the perimeter of the corer to sever larg e roots and other plan t matter. The core was removed intact with overlying water and tr ansported to the labor atory for subsequent analysis. Incubation Setup Soil cores were prepared for an extended 12 week incubation to investigate the changes associated with decreased water le vels. Aluminum foil was wrapped around the cores to the level of the soil to exclude light from penetrating into the soil profile. The cores were left unsealed at the top and holes were drilled in the cores at 20 cm below the soil surface for the dry treatments and 20 cm above the soil surface for the wet treatments, allowing any excess added water to be drained for water level maintenance. The dry treatment soil surface was generally within 5 cm of the top of the core to prevent any shading from the core walls. The core s were placed in a random arrangement in a Percival Scientific™ Growth Chamber Model e-36HID light chamber at a light intensity of 1500 micromole/m1/sec set on a diurnal light cycle. Water was transported from the

PAGE 107

98 field study sites approximately once every tw o weeks to replenish water lost through evaporation. Soil Preparation Soil sample analysis for the initial time pe riod was initiated with in 24 hours of field collection. Each layer, corre sponding to the -10 to -20 cm soil, 0 to –10 cm soil, and benthic layers, was prepared separately. 10 g sub-samples were placed in pre-weighed and pre-ashed aluminum pans for dry mass (DM) and ash free dry mass (AFDM) determination. Dry and ash weights were de termined by incubating the pans in a drying oven for 36 hours at 105 C and ashing the samples at 500 C for 2 hours, respectively. Fresh samples used for enzyme analysis were transferred to a 500 mL beaker and large objects, such as snail shells and rocks, were discarded. The samples were homogenized for 10 minutes with a Biospec Biohomogenizer™, result ing in a soil or benthic slurry. 10 g of the slurry was diluted to a concentration of 10-3 with deionized water and homogenized for an additional 5 minutes. The resulting suspension was transferred to a 100 mL centrifuge tube and refrigerated. En zyme analysis began within 6 hours of sample preparation. Enzyme Analysis Hydrolytic enzyme activity was determin ed using methylumbelliferyl (MUF) and aminomethylcoumarin (AMC) substrates. Subs trate concentrations were optimized at saturating conditions. The activities of -glucosidase (BGL), phosphatase (PHO), leucine aminopeptidase (LEU), phenol oxidase (PHE) and peroxidase (PER) were assayed using MUF-D-glucoside (Sigma M3633), MUF-phosphate (Sigma M8168), L-Leucine amidomethylcoumarin (Sigma L2145) L-3,4-dihydroxyphenylalanine (DOPA), and DOPA + H2O2 as substrates, respectively.

PAGE 108

99 MUF and AMC substrate enzymatic analys is was measured using a Cytofluor 600™ automated spectrofluorimeter (PerSept ive Biosystems, Inc., Framingham, MA) with Kineticalc™ software at 360 nm excita tion and 460 nm emission at 20 C. Assays were performed using Corning 48 -well culture plat es in which 400 L of sample, 360 L of 10 mM Tris-HCl pH 8.5, and 40 L of substrate were a dded. Stock substrate concentrations were 2000 M for MUF-D glucoside, 1000 M for MUF-phosphate, and 6000 M for L-Leucine amidomethylcoumarin resulting in well concentrations of 100 M, 50 M, and 300 M, respectively. Each sample run was performed in quadruplicate. Initial and final fluorescence measurements as well as measurements every five minutes were take n during the 1 hour incubation. Graphs produced from the readings taken every five minut es were analyzed to ensure that linear kinetics was being observed. Concentrations of MUF and AMC re leased were calculated by the application of standard curves to the initial and final fluoresences. The difference in concentrations yielded the substrate released during the incubation period. The effects of quenching on MUF and AMC subs trates were determined in order to account for fluorescence blocki ng or absorption effects cau sed by coloring, particle suspension, humic matter, self-quenching, or ot her inhibitions in the suspensions. MUF and AMC standards were placed in each sample suspension to determine the quench percentage of each matrix. The final and initial fluoresences were then converted using the appropriate quench percentage. Phenol oxidase and peroxidase analyses were performed per Sinsabaugh (personal communication). 2.0 mL soil suspensi on was added to 2.0 mL 10 mM L-DOPA dissolved in 10 mM Tris-HCL pH 8.5 in 10 mL Eppendorf™ centrifuge tubes. The

PAGE 109

100 solutions were vortexed for 30 seconds and plac ed on a shaker plate in a light-proof box for 45 minutes. The solutions were then centrifuged at 3000 rpm for 30 seconds, 500 L supernatant was then extracted and placed in quadruplicate in a Corning™ 48-well culture plate. Controls consisting of 250 uL DI H2O and 250 L 10 mM L-DOPA solution were added to the remaining wells. Sample nutrient analysis was perfor med by DB Labs, Rockledge, FL. Total phosphorus (TP) (EPA 365.2), total nitrogen (TN) (MVP), total organic carbon (TOC) (MVP), calcium (Ca) (SW7140), and ligni n (AOAC 973.18) analyses were performed using standard methods on homogenized samples. Potential enzyme activities are expressed in moles MUF released g-1 AFDM h-1 for BGL and PHO, moles AMC released g-1 AFDM h-1 for LEU, and moles DICQ released g-1 AFDM h-1 for PHE and PER. Models Extracellular enzymes were grouped into four categories: Ecell (BGL), En (LEU), Ep (PHO), and Eox (PHE and PER). This allowed the enzymes to be grouped to indicate C, N, and P mineralization as well as lignin degradation, respectivel y. Enzyme activities were normalized on a scale of 0-1 to eliminat e the weighting effects of the more active enzymes. Enzyme ratios were formulated to examine resource allocation and were based on assumptions derived from the MARCIE (M icrobial Allocation of Resources Among Community Indicator Enzyme s) model (Sinsabaugh and Moorhead, 1994; Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997; Sinsabaugh et al., 2002). The model is based on the premise that the enzyme mediated deco mposition of complex molecules is the rate limiting step in C mineralizati on, indicating that the expressi on of enzymes are tied to environmental nutrient availabili ties and that the distribution of enzyme activities can be

PAGE 110

101 interpreted as a resource allocation strategy (Sinsabaugh et al., 2002). Ecell:Eox values were formulated by dividing the normali zed BGL activity by the average of the normalized oxidative enzyme activities (PHE and PER). An Enzyme Index of Carbon Quality (EICQ) measure was calculated to assess the apparent microbial perceived carbon quality. EICQ is a rela tive index of the normalized activities of the hydrolytic enzymes to the oxidative or lignin degr ading enzymes (Sinsabaugh and Linkins, 1990; Sinsabaugh et al., 1992b; Sinsabaugh, 1994; Si nsabaugh and Findlay, 1995). EICQ has been correlated with microbial biomass (r=0.71), productivity (r=0.80), and negatively correlated with particulate organic car bon (POC) turnover time (r=-0.99) (Sinsabaugh and Findlay, 1995). Cumulative activities were ca lculated by integrating th e area under a plot of time vs. enzyme activity (Sinsaba ugh et al., 1993). The cumulative area corresponding to each time period was divided by the length of inc ubation. Regressions were performed on the cumulative activities for each treatment. Statistics Data was statistically analyzed w ith SAS™ v.8 (SAS, 1999) using repeated measures mixed models analysis to determ ine significant differences (p<0.05) between sampling periods, treatments, and sites. The different laye rs were analyzed independently and contrast statements were us ed to differentiate individual sites where interaction terms were significant. Data was log-transformed to improve normality and heteroscedescity. Regressions were perf ormed using SYSTAT 10.2 (SYSTAT, 2002). Mean values of the sampling pe riods were combined for tables and charts. Differences designated statistically significant we re performed at the p<0.05 level.

PAGE 111

102 Results Enzyme Activities By 12 weeks the moisture contents of the ENP-TS dry cores were 42 and 68%, compared to 89 and 76% for the wet cores in the benthic and 0 to -10 cm layers, respectively. The moisture contents at 12 weeks of the WCA-3A dry cores were 36 and 78%, compared to 93 and 90% for the wet core s in the benthic and 0 to -10 cm layers, respectively. BGL average va lues ranged from 0.005 to 0.252 moles MUF g-1 AFDM h1 (Table 5-1). The highest enzyme activities were generally associated with the benthic layer in both WCA-3A and ENP-TS. Wet trea tments generally exhi bited higher activities across all time periods in ENP-TS and 4 of 9 comparisons in WCA-3A, although the difference was only significant in one cas e. Among the areas, significantly higher glucosidase activities were consistently associ ated with the WCA-3A cores in the benthic and 0 to -10 cm layers (p<0.05). Additionall y, there were significant changes in enzyme activity over the incubation in both treatments wi thin the benthic and 0 to -10 cm layers. PHO exhibited a broad range of activities from 0 to 22.28 moles MUF g-1 AFDM h-1 (Table 5-1). Activities significantly decreased with depth in the 0, 2, and 4 week collections. However, by 12 weeks the greatest activity generally moved from the benthic to the 0-10 cm soil layer. Benthi c wet treatments in both ENP-TS and WCA-3A were significantly higher at 2 and 4 weeks. Ho wever, at 12 weeks the trend was reversed with the dry treatments exhibiting significan tly higher activities in both areas within the benthic layer. There was no clear di fference between the two sites. LEU activities ranged from 0 to 9.76 moles AMC g-1 AFDM h-1 and generally decreased with depth (Table 5-2). Benthic layer LEU did not vary significantly between treatments in ENP-TS. However, WCA-3A LE U activities were significantly higher in

PAGE 112

103Table 5-1. Mean enzyme activities expr essed as moles substrate released g-1 AFDM h-1 with standard errors. Column heading numbers denote the sampling period in w eeks (0, 2, 4, and 12) and enzyme measur ed. Note that time 0 values are replicated in the wet and dry rows as these serve as controls for both trea tments. TS=ENP-TS. 3A=WCA-3A. BGL activities are expressed as 102. Zero enzyme activities reflect a lack of measurable activity. BGL PHO 0 2 4 12 0 2 4 12 Benthic 7.82.5 5.01.9 5.20.2 9.96.1 22.283.76 17.0912.26 13.730.900.010.01 WET 0 -10 2.90.1 2.50.9 3.20.5 0 1.350.94 1.180.03 1.760.21 1.081.08 TS 10-20 5.52.4 1.41.1 4.61.3 0.70.7 0.780.26 0.480.06 0.830.36 0.080.08 Benthic 7.82.5 1.10.1 3.20.5 10.84.7 22.283.76 3.631.52 5.461.19 4.062.32 DRY 0-10 2.90.1 1.50.9 1.10.2 3.72.4 1.350.94 0.610.07 0.870.20 5.021.84 10-20 5.52.4 1.00.4 4.31.2 0.50.5 0.780.26 0.260.04 0.620.06 2.840.25 Benthic 9.91.2 8.90.8 23.82.4 20.70.5 8.561.43 15.401.96 26.524.350 WET 0-10 3.00.6 2.60.3 6.41.1 4.91.1 0.730.13 0.950.08 1.440.40 2.260.61 3A 10-20 1.00.1 0.50.1 3.10.1 0.50.5 0.430.04 0.360.06 0.770.19 1.580.07 Benthic 9.91.2 12.63.1 15.02.2 25.24.8 8.561.43 7.922.36 4.931.30 5.660.73 DRY 0-10 3.00.6 2.60.4 3.80.7 6.40.5 0.730.13 0.810.12 0.770.14 2.950.38 10-20 1.00.1 0.50.1 1.30.1 0.90.1 0.430.04 0.330.03 0.430.02 1.620.15

PAGE 113

104Table 5-2. Mean enzyme activities expr essed as moles substrate released g-1 AFDM h-1 with standard errors. Column heading numbers denote the sampling period in w eeks (0, 2, 4, and 12) and enzyme measur ed. TS=ENP-TS. 3A=WCA-3A. Note that time 0 values are replicated in the wet and dry rows as these serve as controls for both treatments. Zero enzyme activities reflect a lack of measurable activity. LEU PHE 0 2 4 12 0 2 4 12 Benthic 2.390.68 3.572.684.420.28 9.556.79 476.32171.5299.9025.43 108.7623.6895.7153.91 WET 0 -10 0.680.30 0.400.280.370.8 0.080.08 256.46119.6055.3014.76 69.226.36 26.6126.61 TS 10-20 0.800.57 0.560.260.930.42 0 256.31119.7550.0516.50 86.1716.68 58.2558.24 Benthic 2.390.68 1.490.531.550.20 8.066.07 476.32171.5262.074.58 81.5210.43 71.586.30 DRY 0-10 0.680.30 0.550.180.400.17 1.130.17 256.46119.6048.186.48 42.073.17 74.049.67 10-20 0.800.57 0.430.111.130.54 0.550.55 256.31119.7562.848.36 91.7713.44 31.4531.45 Benthic 1.850.36 7.221.030 0 113.7512.70 15.964.13 97.237.75 52.5215.80 WET 0-10 0.210.07 0.350.080 0 21.499.12 9.780.61 22.6922.36 26.442.76 3A 10-20 0.100.01 0.240.020 0 21.6010.12 10.930.94 18.2518.25 13.1913.19 Benthic 1.850.36 5.911.335.901.00 9.763.06 113.7512.70 34.5714.36 31.062.75 21.430.24 DRY 0-10 0.210.07 1.200.032.590.48 7.462.46 21.499.12 6.954.03 23.925.32 15.671.64 10-20 0.100.01 0.370.100.890.14 2.530.44 21.6010.12 6.274.27 23.660.65 17.164.66

PAGE 114

105 the dry treatments at 4 and 12 weeks. Dry tr eatment activities in the 0 to -10 cm layer were significantly higher in WCA-3A but were not significantly higher in ENP-TS until 12 weeks. There were significant time eff ects in the benthic but not the 0 to -10 cm layer. Comparing areas, there were genera lly no significant differe nces in the benthic layer. However, the WCA-3A dry treatments were significantly higher than the ENP-TS dry treatments in the 0 to -10 cm layer. Phenol oxidase activities were wide ly variable and ranged from 6.95 to 476.32 moles DICQ g-1 AFDM h-1 (Table 5-2). The majority of benthic and 0 to -10 cm layer phenol oxidase activities were significantly higher in the ENP-TS cores. Significant differences between treatments were confin ed to the 4 and 12 week WCA-3A samples, with greater PHE activity in the wet core s. There was not a consistently clear relationship with depth. Peroxidase activities ranged from 0 to 59.31 moles DICQ g-1 AFDM h-1 with the majority of enzyme activities unresolvable. There were no significant differences between treatments, time periods, or depth. Th is can be attributed to the very large replicate errors. Ecell/Eox values reflect the apparent lignin influence on C mineralization. Significantly higher values were associated with the WCA-3A cores in all depths and treatments. Values generally decreased with depth, reflecting a gr eater lignin influence on C mineralization. WCA-3A dry treatments were significantly higher than the wet treatments at 4 and 12 weeks in the benthic la yer and only at 2 weeks in the 0 to -10 cm layer. Conversely, the ENP-TS benthic and 0 to -10 cm layers generally exhibited higher values in the wet treatment w ith only 1 significant contrast.

PAGE 115

106 EICQ values reflect the total apparent perceived carbon quality as a function of hydrolytic and oxidative enzyme activities (Tab le 5-3). WCA-3A benthic dry treatments were significantly greater than all the ENPTS benthic dry treatments as were the wet treatments, although only in 50% of the cases. The effect of drying on the sediments was most apparent in the WCA-3A cores in both layers with significantly greater EICQ values in the dry treatments in 5 of 6 cases, compared to the wet. However, EICQ values were not generally significantly different between the treatments in the ENP-TS cores until 12 weeks. Cumulative Activities Cumulative enzyme activities were interpre ted in two fashions. The slopes of the log regressions of each period’s cumulative en zyme activity (CES) provide insight into the rate of total change occurring in e ach treatment over the total incubation period (Table 5-4). Cumulative enzyme activity was calculated according to the trapezoidal rule for integrating the cumulative activity under a curve. Sec ondly, net cumulative enzyme activity-day (CEA) was calculated by the cumulative integrated enzyme activity calculated using the trapezoidal rule for each time period divided by the number of days (Figure 5-1). While CES measures the net rate of changes of a particular treatment over the time period, CEA measures the average accumulation of enzyme activity per day. BGL CES values exhibited the greatest changes over time in the WCA-3A cores in both the benthic and 0 to -10 cm layers (Table 5-4). There was not a consistent trend in rates of change among the treatments. BGL sl opes were generally lower than that of the other enzymes, reflecting sma ller net changes over time. CEA exhibited greater values associated with the WCA-3A cores and wet tr eatments within each site in the benthic

PAGE 116

107Table 5-3. Mean Enzyme Index of Carbon Qu ality (EICQ) values calculated as the ratio of normalized hydrolytic enzymes to normalized oxidative enzymes with standard errors. Ecell/Eox va lues calculated as the ratio of the mean normalized BGL to the mean normalized oxidative enzyme activities (PHE and PER) with standard errors. Va lues are unitless. B=benthic layer, 10=0 to -10 cm layer, 20=-10 to -20 cm layer. TS =ENP-TS, 3A=WCA-3A. Top row indicates the time period in weeks. X indicates a lack of standard error du e to only one replicate in cluded in the analysis. Ecell/Eox EICQ 0 2 4 12 0 2 4 12 Benthic 2.060.21 0.640.17 0.620.110.880.47 2.500.780.890.511.260.250.690.43 WET 0 -10 0.960.34 0.510.22 0.520.110 0.600.230.270.040.320.040.140.14 TS 10-20 1.770.65 0.310.25 0.620.150.14X 0.870.360.200.130.370.100.07X Benthic 2.060.21 0.180.04 0.490.121.010.39 2.500.780.300.050.630.051.180.16 DRY 0-10 0.960.34 0.200.11 0.300.060.630.36 0.600.230.140.040.290.050.880.40 10-20 1.770.65 0.090.04 0.570.150.13X 0.870.360.060.020.380.121.040.38 Benthic 10.511.845.784.36 2.900.064.370.88 6.570.646.424.701.880.101.460.29 WET 0-10 4.043.48 0.870.12 1.180.782.300.73 3.161.680.430.320.470.322.040.02 3A 10-20 2.861.43 0.130.04 0.99X 0.480.09 1.390.690.100.030.38X 3.972.93 Benthic 10.511.846.471.21 5.730.7311.011.356.570.645.091.174.930.528.771.08 DRY 0-10 4.043.48 3.950.35 1.910.163.540.78 3.161.682.820.082.210.085.060.87 10-20 2.861.43 1.330.89 0.680.080.47X 1.390.690.920.400.790.111.850.44

PAGE 117

108 Table 5-4. Cumulative enzyme activity sl opes (CES) based on non-normalized enzyme activities. Values are unitless. Li near regressions were performed to determine the relationships over time. GLU PHO LEU PHE PER Benthic 1.3 (r2=0.93) 195.8 (r2=0.99) 126.6 (r2=0.98) 936.5 (r2=0.45) 236.9 (r2=0.97) WET 0 -10 0.2 (r2=0.98) 22.4 (r2=0.98) 2.1 (r2=0.78) 344.5 (r2=0.32) 333.1 (r2=0.98) TS 10-20 0.4 (r2=0.94) 6.0 (r2=0.97) 6.0 (r2=0.98) 843.5 (r2=0.70) 2.1 (r2=0.60) Benthic 1.25 (r2=0.93) 45.8 (r2=0.48) 89.6 (r2=0.96) 528.7 (r2=0.21) 158.8 (r2=0.97) DRY 0-10 0.4 (r2=0.92) 56.0 (r2=0.96) 12.8 (r2=0.94) 594.7 (r2=0.49) 132.8 (r2=0.94) 10-20 0.3 (r2=0.93) 33.0 (r2=0.96) 13.8 (r2=0.98) 587.7 (r2=0.59) 79.8 (r2=0.99) Benthic 4.0 (r2=0.99) 47.5 (r2=0.73) -21.8 (r2=1) 1220.8 (r2=0.95) 2.35 (r2=0.96) WET 0-10 1.0 (r2=0.99) 33.1 (r2=0.98) -1.3 (r2=0.97) 423.8 (r2=0.97) 584.8 (r2=0.99) 3A 10-20 0.3 (r2=0.99) 21.9 (r2=0.98) -0.8 (r2=0.99) 242.7 (r2=0.95) -5.6 (r2=0.90) Benthic 3.5 (r2=0.98) 69.7 (r2=0.91) 137.7 (r2=0.99) 229.4 (r2=0.49) 17.2 (r2=0.96) DRY 0-10 0.9 (r2=0.97) 34.2 (r2=0.97) 97.7 (r2=0.98) 332.0 (r2=0.97) 9.2 (r2=0.15) 10-20 0.2 (r2=0.98) 19.2 (r2=0.96) 33.3 (r2=0.98) 346.5 (r2=0.97) 10.5 (r2=0.21)

PAGE 118

109 layer (Figure 5-1). This trend was also pr esent in the 0 to -10 cm layer, with the exception of similar values among both the wet and dry ENP-TS treatments (Figure 5-2). PHO CES values did not exhibit a clear trend between the TS and WCA-3A cores (Table 5-4) or between treatme nts, indicating that the rate of change was variable among cores among all layers. However, the benthi c CEA showed that the wet treatments in both ENP-TS and WCA-3A resulted in a gr eater net accumulation of enzyme activity over time in the benthic layer (Figure 5-1). Dry treatments in both areas were similar while the WCA-3A cores were higher in the wet treatments. The relationship between ENP-TS CEA values treatments differed be tween the benthic and 0 to -10 cm layers (Figure 5-2). Benthic PHO CE A was higher in the wet treatment but the 0 to -10 cm PHO CEA was greater in the dry treatment. CE S values decreased with depth in 3 of the four treatments. LEU CES values were variable in the be nthic layer (Table 5-4) with a negative slope in the WCA-3A wet cores as a result of non-resolvable LEU activity at 4 and 12 weeks. The 0 to -10 cm and -10 to -20 cm la yers exhibited larger CES values in the in the dry treatments in both sites. Benthi c LEU CEA mirrored the CES values with the reduced accumulation in the WCA-3A wet treatment once again due to the nonresolvable activity at 4 and 12 weeks (Figure 51) as did the 0 to -10 cm layer (Figure 52). CEA and CES values generally decrea sed with depth, indicating reduced enzyme production and activity. PHE CES values were lower in the dry trea tments in the benthic layer (Table 5-4). However, in the 0 to -10 cm and -10 to -20 cm layers greater CES values were associated

PAGE 119

110 A.) 1.4 1.7 4.2 4.60 1 1 2 2 3 3 4 4 5 5 TS B DryTS B Wet3A B Dry3A B Wet GLU B.) 146.2 250.3 143.4 343.9 90.7 139.7 164.4 32.60 50 100 150 200 250 300 350 400 TS B DryTS B Wet3A B Dry3A B Wet PHO LEU C.) 2588.8 3205.6 847.9 1683.7 13.4 41.4 150.6 194.40 500 1000 1500 2000 2500 3000 3500 TS B DryTS B Wet3A B Dry3A B Wet PHE PER Figure 5-1. Benthic cumulative enzyme activity (CEA) expressed as the integrated area under the data points plotted against time based on mean values. A.) BGL, B.) PHO and LEU, and C.) PHE and PER. Units are cumulative enzyme activity expressed as total mols MUF, AMC, or DICQ accumulated g-1 AFDM over 84 days.

PAGE 120

111 A.) 0.5 0.5 1.1 1.20.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 TS B DryTS B Wet3A B Dry3A B Wet GLU B.) 54.0 33.7 35.9 37.7 16.6 7.3 90.8 1.80 10 20 30 40 50 60 70 80 90 100 TS B DryTS B Wet3A B Dry3A B Wet PHO LEU C.) 1718.7 1639.2 435.3 520.4 36.5 609.2 155.3 285.40 200 400 600 800 1000 1200 1400 1600 1800 2000 TS B DryTS B Wet3A B Dry3A B Wet PHE PER Figure 5-2. 0 to -10 cm cumulative enzyme ac tivity (CEA) expressed as the integrated area under the data points plot ted against time based on mean values. A.) BGL, B.) PHO and LEU, and C.) PHE a nd PER. Units are cumulative enzyme activity expressed as total mols MUF, AMC, or DICQ accumulated g-1 AFDM over 84 days.

PAGE 121

112 with the dry treatments. CEA values were greater in the ENP-TS cores in both the benthic and 0 to -10 cm layers (Figures 51 & 5-2) and mimicked the CES relationships between treatments with the exception of the WCA-3A cores in the 0 to -10 cm layer. There was not a consistent trend with de pth in either the CES or CEA values. The rate of increase of PER CES in the benthic layer was higher in the ENP-TS cores. ENP-TS wet treatments exhibited great er CES values in both the benthic and 0 to -10 cm layers (Table 5-4) which was also re flected in the CEA values (Figures 5-1 & 52). PER CEA values mirrored the relationshi ps between treatments in a similar fashion to BGL, PHO, and PHE in the benthic laye r. PER CES values decreased between the benthic and -10 to -20 cm laye rs, reflecting decreases in th e rate of PER production with depth. Nutrients Site core variability generally exceeded any differences between wet/dry treatments as well as time effects. Therefore, the most consistent difference in nutrient concentrations as well as li gnin and cellulose occurred in comparisons between WCA-3A and ENP-TS soil cores (Figure 5-3). Averages were calculated from a total of 15 cores per layer-site combination from 0, 2, and 12 w eek incubations. Average cellulose content ranged from 6.2 to 41% and was significantly greater in the WCA-3A cores, with an inverse relationship in the -10 to -20 cm layer. Cellulose content increased dramatically in the ENP-TS cores in the -10 to -20 cm layer, otherwise there was not a significant relationship with increasing depth. Average lignin content ranged from 7.7 to 47.7% with significantly higher concentrations found within the WCA-3A cores. Lignin content

PAGE 122

113 A.) B.) 40.4 6.2 12.8 6.3 12.2 41.0 15.5 9.2 30.6 16.0 7.7 47.7 0 10 20 30 40 50 60 TS B 3A BTS -103A -10TS -203A -20Sample% Abundance Cellulose % Lignin % C.) 13 38 15 41 11 36 158 432 183 474 497 137 0 100 200 300 400 500 600 TS B 3A BTS -103A -10TS -203A -20Sampleg/kg TN TOC 0.46 0.25 0.17 0.34 0.21 0.16 0.0 0.1 0.2 0.3 0.4 0.5 TS B 3A BTS -103A -10TS -203A -20Sampleg/kg TP Figure 5-3. Mean lignin, cellulose, and nut rient contents of ENP-TS and WCA-3A cores benthic, 0 to 10 cm soil, and -10 to -20 soil layers. (A) cellulose and lignin content. (B) total nitrogen and tota l organic carbon (C) total phosphorus. Bars represent standard errors.

PAGE 123

114 increased with depth in the WCA-3A cores but did not exhibit a consistent trend in the ENP-TS cores. Average total nitroge n (TN) and total organic carbon (TOC) concentrations exhibited the same type of relationship with ranges of 11 to 41 g kg-1 and 137 to 497 g kg-1, respectively. WCA-3A cores had si gnificantly greater TP with an overall range of 0.16 to 0.46 g kg-1, and a significant decrease with depth. Discussion The surface layers are the most biological ly active and important in terms of degradation in inundated systems. For this reason the majority of the discussion will pertain to findings related to the benthic a nd 0 to -10 cm layer since the water level decrease in the dry treatments involved a grea ter exposure of these layers to increased O2 availability. The discussion is grouped accordin g to the effects that lowered water levels and sediment types have on oxidative enzyme activity, N mineralization and microbiallyperceived carbon quality (EICQ). Oxidative Enzymes The dependence of the oxidative enzymes on O2 availability predicts that lowered water levels would result in increased activity (Pind et al., 1994). The results from this study demonstrate the apparent variability of PHE and PER activity in the field that has been previously documented as a response to depth and O2 availability (Duxbury and Tate, 1981; Lhdesmaki and Piispanen, 1988; Pind et al., 1994; Freeman et al., 1991 & 2001; Williams et al., 2000b) with the dry cores exhibiting lower cumulative PHE and PER activities over the incubation period. This contradicts a recent study that showed a 7 fold increase in PHE activity in aerated condi tions (Freeman et al., 2001) but agrees with another study (Williams et al., 2000b). The l ack of a predictable increase within the

PAGE 124

115 study time frame suggests that enzyme inhib ition may be occurring or that an extended incubation time may be required. It has been suggested that persiste nt drought conditions may be required before PHE activity increa ses (Williams et al., 2000b), which appears to be the situation in this study. While the oxidative enzymes do not app ear to respond predictably to lowered water levels, the apparent lignin influence on C mineralization (Ecell/Eox), a function of the oxidative activities, does exhibit a treatment effect. After initial decreases at 2 weeks, the Ecell/Eox values increase at a more rapid rate in the dry treatments, especially in WCA-3A. This suggests that C mineralization in the dry treatments is less negatively influenced by lignin, which would result in more favorable potential decomposition. This is less significantly pron ounced within ENP-TS, while the WCA-3A benthic layer exhibits much less lignin influence by 12 weeks. The combination of greater C mineraliza tion and lower oxidative enzyme activity in the WCA-3A cores results in a lower a pparent lignin control on C mineralization in both the benthic and 0 to -10 cm layers. T hus, compared to the ENP-TS cores, the WCA3A sediments appear to be more favorable for potential decomposition in terms of lignin influence and have a greater Ecell:Eox respons e with lower water levels. Much of this difference between areas is most probably due to differences in substrate composition (Williams et al., 2000a), phenolic concentr ations (Freeman et al., 2001) and C availability. N Mineralization Based on other studies, N mineralization was expected to increase in simulated drought conditions (Reddy and Patrick, 1984; Cabrera 1993; Bridgham et al., 1998;

PAGE 125

116 McLatchey and Reddy, 1998; White, 1999; Vent erlink et al., 2002). Enhanced LEU activity was observed in the WCA-3A 0 to -10 cm layer dry treatments over the incubation period in the CES, CEA, and individual enzyme activities. However, this difference was only exhibited by the ENP-TS co res at 12 weeks. The relegation of this increase largely to the 0 to -10 cm laye r is analogous to larger increases in N mineralization rates in this layer in aer obic versus methanogenic conditions (White, 1999). Therefore it appears that O2 availability is limiting N mineralization in the WCA3A cores with the ENP-TS cores responding slow er to a decrease in water levels. The difference between the two sites may be due to the greater abundance of organic matter for C mineralization in the WCA-3A cores wh ich would energetically lead to greater N mineralization. EICQ The Enzyme Index of Carbon Quality (EIC Q) was expected to increase during a simulated drought as conditions became more favorable for microbial production. The largest predicted effect on the EICQ was e xhibited by both layers in the WCA-3A cores where values were significantly higher in th e dry treatments across the incubation period. However, the ENP-TS dry treatment in the 0 to -10 layer was only significantly higher at the 12 week collection. Thus drought duration ap pears to play a significant role in the cumulative effects on potential decompos ition. Higher EICQ values in the dry treatments, which have been correlated to microbial biomass and productivity (Sinsabaugh and Findlay, 1995), are supported by greater C flux a nd soil respiration (DeBusk, 1996) as a result of drainage as we ll as increased microbial biomass and C flux (Blodau et al., 2004).

PAGE 126

117 The time effect may be due to the slow de ssication of the 0 to -10 cm layer due to the capillary draw of water from the saturated zone. Additionally, much of this layer may still contain a substantial por tion of anaerobic microsites (K ettunen et al., 1998). Larger differences between the two treatments ma y therefore be expected over an extended incubation time. This is supported by a previous study in which increases of soil respiration rates due to soil drainage was seen to be time dependent (DeBusk, 1996). Lastly, the higher EICQ in the WCA-3A core s may be attributed to the soil composition differences between the two site s. WCA-3A is primarily an organic peat matrix while the ENP-TS profile is uniquely dominated by a r obust periphyton mat in various stages of decay. Therefore, the nutr itional basis for the resident heterotrophic microbial communities is distinct between the two sites. It should be noted that the majority of enzyme activities showed a large reduction in activity between the initial collections following the field sa mpling event. This initial decrease in activity suggests that a core effect is occurr ing, possibly due to the lessening of nutrient and microbial wast e exchange between the porewat er and surface water due to the lack of a constantly replenished source. Additionally, the accu mulation of DOC from mineralization processes may be occurring with in the porewater (Blodau et al., 2004). This also suggests that a large effect of drought may occur with in the initial time frame of hours or days following a treatment. The use of cumulative enzyme activities, either in the form of slopes or net activity reduces this effect in interpreting the data. Conclusions The ecological implications of these findi ngs suggest that a lengthened hydroperiod with a significant drought period will result in decreased net organic matter accumulation

PAGE 127

118 as a result of higher perceived C quality. Th ese findings are based primarily on the use of the EICQ model in predicting potential decom position rates as well as the apparent lignin influence on carbon mineralization. Increase s in C quality with decreased water level appear to be especially appare nt in the 0 to -10 cm layer in the organic peat dominated WCA-3A, where microbial activity is usua lly relatively limited due to anaerobic conditions. The increase of N mineralization in the dry treatments of the 0 to -10 cm layer in both areas points to decreased O2 limitation. This points to overall increased microbial productivity, especially since the enzyme used in this study (LEU) to assay for N mineralization also has the cap acity to function in the mineralization of C. Lastly, the higher organic matter WCA-3A soils res pond greater to drought conditions with significantly higher N mineraliza tion and EICQ in the dry trea tments which was expected due to a higher organic matter content. The results from enzyme studies are usually presented solely in terms of activity per period or treatment. The use of cumula tive enzyme activities, expressed either as slopes representing the rate of change over ti me or as net cumulative activities conveys information regarding trends in terms of time series studies. These cumulative activities are usually expressed in rela tion to litter mass loss rate s in decomposition studies. However, this study demonstrates the validity of using these comparisons, especially when significant differences between treatm ents are relegated to later time periods. Due to the constraints of this study it is impossible to accurately model the effects of a prolonged dry-down. However, many of the differences between wet and dry treatments were becoming increasingly contras ting with time. The current trend suggests that a longer incubation period would result in larger EICQ and N mineralization

PAGE 128

119 differences between the treatments as satu rated pores empty. Conversely, the large differences between initial c onditions and the 2 week co res suggest that significant changes occur within the cores as a result of containment. Therefore a more stringent series of assays on a time frame of hours or days are needed in order to further understand these microbial shifts as a functi on of laboratory containment. This has potential implications on the validity of numerous soil microcosm experiments where prolonged incubation times are involved.

PAGE 129

120 CHAPTER 6 SUMMARY AND CONCLUSIONS Field and laboratory studies were conducted to examine the influences of nutrient loading, vegetative habitat, and simula ted drought conditions on microbial enzyme activities in the Everglades. A summary of resu lts as they relate to the study objectives is presented below. 1) Develop an appropriate experime ntal method for performing enzyme assays in wetland systems. The investigation of the e ffects of varying substrate concentrations on measured enzyme activity was deemed important in developing assays that are accurate in estimating in situ activities. The inclusi on of kinetic activity r eadings over time was found to be significant in determining the va lidity of measured enzyme activities and provided a method of discarding outlying valu es. The importance of incubation time was found to vary among substrate concentrations with a 45 minute incubation time deemed adequate in this study. Optimum substrate concentrations were determined to be 60 M, 50 M, 100 M, and 300M for cellobiohydrolase, phosphatase, glucosidase, and leucine aminopeptidase, respectively. These subs trate concentrations exhibited a linear relationship over time in both enriched and re ference P sediments and were utilized for further experimentation.

PAGE 130

121 2) Determine the effects of vegetative habitats on microbial enzyme activities in Water Conservation Area 3A of the Everglades. The apparent lignin influence on organic C mineralization (Ecell/Eox) appeared to be the driving force behind changes in microbi al enzyme activities in the benthic layer among different vegetative habitats. The si gnificantly higher Ecell/Eox values in the open water habitats, coupled with higher Enzyme Index of Carbon Quality (EICQ) measures indicate that this habitat is more suited to potential decomposition. Lower C:N ratios of the benthic matter associated with the open water habitats also predict greater potential decomposition. When coupled to th e lower C input in the open water habitats, it is suggested that the increases in poten tial decomposition would result in a lower elevation over time that supports the current Everglades topography. 3) Determine the relationships between nutrient conditions and microbial enzyme activities among four hy drologic units of the Everglades. Phosphorus was less limiting to C mineraliza tion, expressed as Ecell/Ep values, at the enriched sites with the smallest apparent limitation within Water Conservation Area 3A. Conversely, the largest apparent P limitati on occurred within Everglades National Park (ENP-TS). Shifts in apparent N limitation on C mineralization, expressed as Ecell/En, occurred between the enriched a nd reference sites in Loxahatchee National Wildlife Refuge (LNWR), Water Conservati on Area 2A, Water Cons ervation Area 3A, and Taylor Slough within Everglades National Park. A trend of increasing Ecell/En values at the enriched sites in a southerly direction suggests that a decreasing N limitation

PAGE 131

122 is occurring that is not reflect ed in soil nutrient conditions at these sites. A significantly larger Ecell/En difference between the enri ched and reference sites within ENP-TS coupled with the lowest Ecell/Ep shift sugge sts that this area is unique in regards to microbial enzyme activities. The combination of decreased apparent N and P limitation on C mineralization results in greater potenti al decomposition at the enriched sites. An enzyme decomposition model was cons tructed using tensile strength loss of cotton strips as a basis for cellulose d ecomposition. An exponential model utilizing Ecell/Ep and total P (TP) accounted for between 46% and 92% of the variability in cotton rottenness rates (CRR), reflecting the large in fluence of P availability on the microbial community within the Ever glades ecosystem. 4) Determine the validity of different enzyme models in predicting potential decomposition among different litter qualities and in varying nutrient conditions. The use of the EICQ, a relative meas ure of hydrolytic to oxidative enzyme activities, appears to most valid when comp aring changes in vegetative types or changes in O2 availability. This is due to the weight that the EICQ model places on the oxidative enzymes involved in lignin degradation. However, this model was found to be inconsistent when predicting potential d ecomposition along nutrient gradients in the Everglades. Rather, the model parameters Ecell/Ep and Ecell/E n, which relate the apparent P and N influence on C minera lization respectively, predict cellulose decomposition rates, accounting for between 46 % and 92% of the variability. The most appropriate model, which varies among th e different comparisons in these studies,

PAGE 132

123 suggests the need for the inve stigation of the underlying conc epts and comparisons used for model development. Additi onally, variations in the valid ity of the EICQ model were exhibited when comparing sites according to habitat or nutrient differences. These studies suggest that the use of this model appears to be restricted to comparisons between differing vegetative types where substrate st ructure and lignin content influence oxidative enzyme activities. 5.) Determine the effect of a simulate d drought on enzyme activities in WCA3A and ENP-TS cores. The results from this study suggest that prolonged drought periods will result in increased C mineralization in the otherwise inun dated surficial soil layers, especially in peat sediments. Elevated EICQ values, in creases in individual enzyme activities, and decreased oxidative activities all suggest a greater microbi al perceived carbon quality under dry conditions. Most of these changes be tween treatments occu rred in the 0 to -10 cm layer. The primarily peat sediments of WCA-3A, compared to the marl sediments of ENP-TS, exhibited larger cha nges in the dry treatments. Relatively larger differences with time between the wet and dry treatments suggest that a longer incubation period may result in even greater cha nges in potential decomposition. The microbial component of the ecosystem serves as the basis for the trophic food chain and is temporally the most responsive to alterations in envi ronmental conditions. Changes in the microbial component are mani fested in higher order trophic processes. As such, this study provides valuable data concerning the effects of vegetative habitats, nutrient loading and water table drawdown on microbial enzyme activities that are

PAGE 133

124 responsible for nitrogen and phosphorus cycl ing, carbon mineralizati on and are the rate limiting steps in organic matte r decomposition. The results from this study may be utilized to better understand the dynamics of the Everglades ecosystem, especially in terms of the Comprehensive Evergl ades Restoration Plan (CERP).

PAGE 134

125 LIST OF REFERENCES Amador, J.A., Jones, R.D., 1993. Nutrient lim itations on microbial respiration in peat soils with different total phosphorus cont ent. Soil Biol. Biochem. 25, 793-801. Baldwin, M., Hawker H.W., 1915. Soil survey of the Fort Lauderdale area, Florida. In: Field Operations of the Bureau of Soils, 1915. Washington, D.C: U.S. Department of Agriculture, pp. 751-798. Berg, B., 2000. Litter decomposition and organic matter turnover in northern forest soils. For. Ecol. Manage. 133, 13-22. Blanchette, R.A., 1991. Delignicati on by wood-decay fungi. Annual Rev. Phytopathology 29, 381-398. Blodau, C., Basiliko, N., Moore, T.R., 2004. Carbon turnover in peatland mesocosms exposed to different water tables. Biogeochem. 67, 331-351. Boavida, M. –J., Wetzel, R.G., 1998. Inhi bition of phosphatase activity by dissolved humic substance and hydrolytic reactiv ation by natural UV. Freshwater Biology 40, 285-293. Boschker, H.T.S., Cappenberg, T.E., 1998. Patterns of extracellular enzyme activities in littoral sediments of Lake Gooimer, The Netherlands. FEMS Microbial Ecology 25, 79-86. Bridgham, S. D., Pastor, J., Jannsens, J.A ., Chapin, C., Malterer, T.J., 1996. Mulitple limiting gradients in peatlands: A call for a new paradigm. Wetlands 16, 45-65. Bridgham, S. D., Updegraff, K., Past or, J., 1998. Carbon, nitrogen, and phosphorus mineralization in northern wetlands. Ecology 79, 1545-1561. Burns R. G., 1986. Interaction of enzymes with soil mineral and organic colloids. In: Huang P.M., Schnitzer M. (Eds), Interactions of Soil Minerals with Natural Organics and Microbes. Madis on, WI: SSSA Spec. 17. SSSA, pp. 429-451. Burns, A., Ryder, D.S., 2001. Response of bacterial extracellular enzymes to inundation of floodplain sedi ments. Freshwater Biology 46, 1299-1307. Cabrera, M. L., 1993. Modelling the flush of nitrogen mineralization caused by drying and rewetting soils. Soil Sci. Soc. Am. J. 57, 63-66.

PAGE 135

126 Carreiro, M.M., Sinsabaugh, R.L., Repert, D.A., Parkhurst, D.F., 2000. Microbial Enzyme shifts explain litter decay res ponses to simulated nitrogen deposition. Ecology, 81, 2359-2365. Cembella, A.D., Antia, N.J., Harrissoon, P.J., 1984. The utilization of inorganic and organic phosphorus compounds as nutrients by eukaryotic microalgae: a multidisciplinary pers pective. Part 1. CRC Crit. Rev. Microbiol. 10, 317-391. Chiang, C., Craft, C.B., Rogers, D.W., Rich ardson., C.J., 2000. Effects of 4 years of nitrogen and phosphorus additions on Ev erglades plant communities. Aquatic Botany 68, 61-78. Chrst, R.J., 1989. Characteriza tion and significance of B-gluc osidase activity in lake water. Limnol. Oceanogr. 34, 660-672. Chrst, R.J., 1991. Environmental control of the synthesis and activity of aquatic microbial ectoenzymes. In: Chr st, R.J. (Ed.). Microbial Enzymes in Aquatic Environments. New York: Springer, pp. 29-59. Chrst, R.J., Krambeck, H.J., 1986. Fluores cence correction for measurements of enzyme activity in natural waters using methylumbelliferyl-substrates. Archives of Hydrobiology 106, 79-90. Chrst, R.J., Rai, H., 1993. Ectoenzyme activ ity to bacterial secondary production in nutrient impoverished and nutrient enriched freshwater mesocosms. Microb. Ecol. 25, 131-150. Craft, C.B., Richardson, C.J., 1993. P eat accretion and phosphorus accumulation along a eutrophication gradient in th e northern Everglades. Biogeochem. 22, 133156. Craft, C. B., Richardson, C.J., 1995. Respons e of Everglades plant communities to nitrogen and phosphorus a dditions. Wetlands 15, 258-271. Davis, J.H., Jr., 1943. The natural features of southern Florida. Florida Geological Survey Bull. 25. Davis, S.M., 1991. Growth, decom position, and nutrient retention of Cladium jamaicense Crantz and Typha domingensis Pers. in the Florida Everglades. Aquat. Bot. 40, 203-224. Debosz, K., Rasmussen, P.H., Pederson, A. R., 1999. Temporal variations in microbial biomass C and cellulolytic enzyme activity in arable soils: Effects of organic matter input. Applied Soil Ecol. 13, 209-218.

PAGE 136

127 DeBusk, W.F., 1996. Organic matter turnove r along a nutrient gradient in the Everglades. Ph.D. Dissertation, Un iv. of Florida, Gainesville, FL. DeBusk, W.F., Reddy, K.R., 1998. Turnover of detrital organic carbon in a nutrientimpacted Everglades marsh. Soil Sci. Soc. Am. J. 62, 1460-1468. DeBusk, W. F., Newman, S., Reddy, K.R., 2001 Spatio-temporal patterns of soil phosphorus enrichment in Everglades WCA-2A. Environ. Qual. 30, 1438-1446. DeBusk, W.F., Reddy, K.R., Koch, M.S., Wang, Y., 1994. Spatial distribution of soil nutrients in a northern Everglades mars h: Water Conservation Area 2A. Soil Sci. Soc. Am. J. 58, 543-552. Degens, B. P., Harris, J.A., 1997. Devel opment of a physiological approach to measuring the catabolic diversity of soil microbial communities. Soil Biol. Biochem. 29, 1309-1320. Dick, R. P., 1997. Soil enzyme activities as in tegrative indicators of soil health. In: Pankhurst, C. E., Doube, B. M., Gupta, V. V. S. R. (Eds.). Biological Indicators of Soil Health. Wallingford, UK: CAB International, pp. 121-156. Duxbury, J. M., Tate III, R.L., 1981. The effect of soil depth and crop cover on enzymatic activities in Pahokee muck. Soil Sci. Soc. Am. J. 45, 322-328. Elliott, E. T., Coleman, D.C., Ingham, R.E., Trofymow, J.A., 1984. Carbon and energy flow through microflora and micr ofauna in the soil subsystem of terrestrial ecosystems. In: Klug, M.J., Reddy, C.A. (Eds.). Current Perspectives in Microbial Ecology. Wa shington, D.C.: American Society for Microbiology, pp. 424-433. Eriksson, K.–E., Blanchette, A., Ander, P., 1990. Microbial and Enzymatic Degradation of W ood and Wood Components New York: Springer-Verlag. Eriksson, K.-E., Wood, T.M., 1985. Biodegradati on of cellulose. In: T. Higuchi (Ed.) Biosynthesis and Biodegrad ation of Wood Components. New York: Academic Press, pp. 469-503. Espeland, E.M., Francoeur, S.N., Wetzel, R.G ., 2001. Influence of Algal Photosynthesis on biofilm bacterial produc tion and associated gluc osidase and xylosidase activities. Microb. Ecol. 42, 524-530. Fioretto, A., Pap, S., Gurcio, E., Sorrentino, G., Fuggi, A., 2000. Enzyme dynamics on decomposing leaf litter of Cistus incanus and Myrtus communis in a Meditteranean ecosystem. Soil Biol. Biochem. 32, 1847-1855.

PAGE 137

128 Foreman, C.M., Franchini, P., Sinsabaugh, R. L., 1998. The trophic dynamics of riverine bacterioplankton: Relationships amo ng substrate availability, ectoenzyme kinetics, and growth. Limnol. Oceanogr. 44, 1344-1352. Frankenberger, W.T., Dick, W. A., 1983. Relationship betwee n enzyme activities and microbial growth and activity indices in soil. Soil Sci. Soc. Am. J. 47, 945-951. Freeman, C., Liska, G., Ostle, N.J., Jones, S.E., Lock, M.A., 1995. The use of fluorogenic substrates for measuring enzyme activity in peatlands. Plant and Soil 175, 147-152. Freeman, C., Liska, G., Ostle, N.J., Lo ck, M.A., Reynolds, B., Hudson, J., 1996. Microbial activity and enzymic decompositi on processes following peatland water table drawdown. Plant and Soil 180, 121-127. Freeman, C., Ostle, N., Kang, H., 2001. Pea tland phenol oxidase: An enzymic latch on a global carbon store. Nature 409, 149. Grayston, S. J., Griffith, G.S., Mawdsley, J.L., Campbell, C.D., Bardgett, R.D., 2001. Accounting for variability in soil micr obial communities of temperate upland grassland ecosystems. Soil Biol. Biochem. 33, 533-551. Hammel, K. E., 1997. Fungal degradation of lignin. In: Cadisch, G., Giller, K. E. (Eds.). Driven by Nature. Walli ngford, UK: CAB International. Hill, M.O., Latter, P.M., Bancroft, G ., 1985. A standard curve for inter-site comparison of cellulose degradation us ing the cotton strip method. Can. J. Soil Sci. 65, 89-95. Hoppe, H.G. 1993. Use of fluorogenic mode l substrates for extracellular enzyme activity (EAA) measurement of bacteria. In: Kemp, K., Sherr, B., Cole, J.J. (Eds.). Handbook of Methods in Aquatic Microbi al Ecology. Chelsea, MI: Lewis Publishers, 777 pp. Jackson, C.R., Foreman, C.M., Sinsabaugh, R.L., 1995. Microbial enzyme activities as indicators of organic matter processi ng rates in a Lake Erie coastal wetland. Freshwater Biology 34, 329-342. Jansson, M., Olsson, H., Pettersson, K., 1988. Phosphatases; orig in, characteristics and function in lakes. Hydrobiologia 170, 157-175. Kettunen, A., Kaitala, V., Lehtinen, A., Lohila, A., Alm, J., Silvola, J., Martikainen, P., 1997. Methane production and oxidation potent ials in relation to water table fluctuations in two boreal mire s. Soil Biol. Biochem. 31, 1741-1749.

PAGE 138

129 Kiyosawa, K., 2001. Ca2+ and phosphate releases from calcified Chara cell walls in concentrated KCl solution. J. Exp. Bot. 52, 223-229. Koch, M.S., Reddy, K.R., 1992. Distributi on of soil and plant nutrients along a trophic gradient in the Florida Evergl ades. Soil Sci. Soc. Am. J. 56, 1492-1499. Kourtev, P.S., Ehrenfeld, J.G., Haggblom, M., 2002a. Exotic plant species alter the microbial community structure and f unction in the soil. Ecology 83, 31523166. Kourtev, P. S., Ehrenfeld, J.G., Haggblom, M., 2003. Experimental analysis of the exotic and native plant species on the structure and function of soil microbial communities. Soil Biol. Biochem. 35, 895-905. Kourtev, P.S., Ehrenfield, J.G., Huang., W.Z., 2002b. Enzyme activities during litter decomposition of two exotic and two native plant species in hardwood forests of New Jersey. Soil Bio. Biochem. 34, 1207-1218. Kuhn, N. L., Mendelssohn, I.A., McKee, K. L., Lorenzen, B., Brix, H., Miao, S., 2002. Root phosphatase activity in Cladium ja maicense and Typha domingensis grown in Everglades soil at ambient and el evated phosphorus le vels. Wetlands 22, 794-800. Lahdesmaki, P., Piispanen., R., 1988. Degradation products and the hydrolytic enzyme activities in the soil hum ification processes. Soil Biol. Biochem. 20, 287-292. Linkins, A.E., Sinsabaugh, R.L., McCl augherty, C.A., Mellilo, J.M., 1990a. Comparison of cellulase ac tivity on decomposing leaves in a hardwood forest and woodland stream. Soil Biol Biochem. 22, 423-425. Linkins, A.E., Sinsabaugh, R.L., McClaughe rty, C.A., Melillo, J.M., 1990b. Cellulase activity on decomposing l eaf litter in microcosms. Plant and Soil 123, 17-25. Ljundahl, L.G., Eriksson, K.E., 1985. Ecology of microbial cellulose degradation. Advances in Microb. Ecol. 8, 237-299. Lorenzen, B., Brix, H., Mendelssohn, I.A., McKee, K.L., Miao, S., 2001. Growth, biomass allocation and nutr ient use efficiency in Cladium jamaicense and Typha domingensis as affected by phosphorus an d oxygen availability. Aquatic Botany 70, 117-133. Marsden, W. L., Gray, P.P., 1986. Enzymatic hydrolysis of cellulose in lignocellulose materials. CRC Critical Re views in Biotechnology. 3, 235-276.

PAGE 139

130 Marx, M.C., Wood, M., Jarvis, S.C., 2001. A microplate fluorimetric assay for the study of enzyme diversity in soil s. Soil Biol. Biochem. 33, 1633-1640. Mayr, C., Miller, M., Insa m, H., 1999. Elevated CO2 alters community-level physiological profiles and enzyme activi ties in alpine grassland. J. Microb. Methods 36, 35-43. McClaughterty, C.A., Linkins, A.E., 1990. Te mperature responses of enzymes in two forest soils. Soil Biol. Biochem. 22, 29-33. McCormick, P., O’Dell, M., 1996. Quantifying periphyton responses to phosphorus in the Florida Everglades: a synoptic-expe rimental approach. J. N. Am. Benthol. Soc. 15, 450-468. McCormick, P.V., Newman, S., Miao, S.L., Gawlik, D.E., Marley, D., Reddy, K.R., Fontaine, T.D., 2001. Effects of anthropogenic phosphorus inputs on the Everglades, In: J.W. Porter and K.G. Porter, Eds. The Everglades, Florida Bay, and Coral Reefs of the Florida Ke ys: An Ecosystem Sourcebook. Boca Raton, FL: CRC Press, p. 83-126. McCormick, P.V., Newman, S., Payne, G ., Miao, S., Fontaine, T., 2000. Ecological effects of P enrichment. Chapter 3 in 2001 Everglades Consolidated Report, South Florida Water Management District, West Palm Beach, FL. McCormick, P.V., Rawlick, P.S., Lurding, K., Smith, E.P., Sklar, F.H., 1996. Periphyton–water quality relationships along a nutrient gradient in the Florida Everglades. J. N. Am. Benthol. Soc. 15, 433–449. McLatchey, G. P., Reddy, K.R., 1998. Regul ation of organic ma tter decomposition and nutrient release in a wetla nd soil. J. Environ. Qual. 27, 1268-1274. Meentemeyer, V., 1978. Macroclimate and ligni n control of litter decomposition rates. Ecology 59, 465-472. Melillo, J.M., Aber, J.D., 1982. Nitrogen a nd lignin control of hardwood leaf litter decomposition dynamics. Ecology 63, 621-626. Melillo, J.M., Aber, J.D., Linkins, A.E., Ri cca, A., Fry, B., Nadelhoffer, K.J., 1989. Carbon and nitrogen dynamics along the decay continuum: plant litter to soil organic matter. Plant and Soil 115, 189-198. Miettinen, I.R., Vartiainen, T., Martik ainen, P.J., 1996. Bacterial enzyme activities in ground water during bank filtr ation of lake water. Water Res. 30, 2495-2501.

PAGE 140

131 Nausch, M., Pollehne, F., Kerstan, E., 1998. Ex tracellular activities in relation to hydrodynamics in the Pomerian Bight (Sout hern Baltic Sea). Microb. Ecol. 36, 251-258. Newman, S., Reddy, K.R., 1993. Alkaline phospha tase activity in the sediment-water column of a hypereutrophic lake. J. Environ. Qual. 22, 832-838. Newman, S., Kumpf, H., Laing, J.A., Kennedy, W.C., 2001. Decomposition responses to phophorus enrichment in an Everglades (USA) slough. Biogeochem. 54, 229-250. Newman, S., McCormick, P.V., Backus, J.G ., 2003. Phosphatase activity as an early Warning indicator of wetland eutrophicati on: problems and prospects. J. Applied Phycology 15, 45-59. Newman, S., McCormick, P.V., Trexler, J.C ., Reddy, K.R., Miao, S.L., Shuford, R.B.E., Baker, S., Pagen, X., Wright, A.L, 2 002. Effects of changes in phosphorus levels on the central and southern Evergl ades. South Florida Water Management District. West Palm Beach, FL. Parham, J.A., Deng, S.P., 2000. Detection, qu antification and char acterization of -glucosaminidase activity in soil Soil Biol. Biochem. 32, 1183-1190. Pind, A., Freeman, C., Lock, M.A., 1994. Enzy mic degradation of phenolic materials in peatlands-Measurement of phenol oxidase activity. Plant Soil 159, 227-231. Ponnanmperuma, F.N., 1972. The chemistry of submerged soils. Advances in Agronomy 24, 29-96. Qualls, R. G., Richardson, C.J., 2003. Factor s controlling concentr ation, export, and decomposition of dissolved organic nutrien ts in the Everglades of Florida. Biogeochem. 62, 197-229. Reddy, K.R., D’Angelo, E.M., 1994. Soil pr ocesses regulating water quality in wetlands. In: Mitsch, W.J. (Ed.) Global Wetlands: Old World and New. Amsterdam: Elsevier, pp. 309-324. Reddy, K.R., DeLaune, R.D., DeBusk, W.F., Koch, M.S., 1993. Long term nutrient accumulation rates in the Everglades. Soil. Sci. Am. J. 57, 1147-1155. Reddy, K R., Patrick, W.H., 1984. Nitrogen transformations and loss in flooded soil and sediments. CRC Critical Reviews in Environ. Control. 13, 273-309. Saiya-Cork, K. R., Sinsabaugh, R.L., Zak, D.R., 2002. Effects of long-term nitrogen deposition on extracellular enzyme activity in an Acer saccharum forest soil. Soil Biol Biochem. 34, 1309-1315.

PAGE 141

132 SAS v.8. Cary, NC: SAS Institute Inc. 1999. SFWMD. 2003. Everglades Consolidated Report. South Florida Water Management District, West Palm Beach, Florida. Shackle, V.J., Freeman, C., Reynolds, B ., 2000. Carbon supply and the regulation of enzyme activity in constructed wetlands. So il Biol. Biochem. 32, 1935-1940. Shaver, G. R., Mellilo, J.M., 1984. Nutrient budgets marsh plants: Efficiency concepts and relation to availability. Ecology 65, 1491-1510. Sinsabaugh, R.L., 1994. Enzymic analysis of mi crobial pattern and process. Biology and Fertility of Soils 17, 69-74. Sinsabaugh, R.L., Findlay, S., 1995. Micr obial production, enzyme activity, and carbon turnover in surface se diments of the Hudson River Estuary. Microb. Ecol. 30, 127-141. Sinsabaugh, R. L., Foreman, C.M., 2001. Activity profiles of bacterioplankton in a eutrophic river. Freshwater Biology 46, 1-12. Sinsabaugh, R.L., Linkins, A.E., 1990. Enzymic and chemical analysis of particulate organic matter from a boreal ri ver. Freshwater Biology 23, 301-309. Sinsabaugh, R.L., Moorhead, D.L., 1994. Re source allocation to extracellular enzyme production: A model of phosphorus and nitrogen c ontrol of litter decomposition. Soil Biol. Biochem. 26, 1305-1311. Sinsabaugh, R. L., Moorhead, D.L., 1996. S ynthesis of litter qua lity and enzymic approaches to decomposition modeling. In: Cadish, G., Giller, K. E. (Eds.). Driven by Nature: Plant Litter Qual ity and Decomposition. Wallingford, UK: CAB International, pp. 363-375. Sinsabaugh, R.L., Antibus, R.K., Linkins, A. E., 1991. An enzymic approach to the analysis of microbial activity during plant litter decomposition. Agriculture, Ecosystems and Environment 34, 43-54. Sinsabaugh, R.L., Antibus, R.K., Linkins A.E., McClaugherty, C.A., Rayburn, L., Repert, D., Weiland, T., 1993. Wood Decomposition: Nitrogen and phosphorus dynamics in relation to extr acellular enzyme activity. Ecology 74, 1586-1593. Sinsabaugh, R. L., Antibus, R.K., Linkins, A.E., Rayburn, L., Repert, D., Weiland, T., 1992a. Wood decomposition over a first-order watershed: mass loss as a function of exoenzyme activity. Soil Biol. Biochem. 24, 743-749.

PAGE 142

133 Sinsabaugh, R.L., Carreiro, M., Repert, D. A., 2002. Allocation of extracellular enzymatic activity in relation to litter composition, N deposition, and mass loss. Biogeochemistry 60, 1-24. Sinsabaugh, R.L., Findlay, S., Franchini, P., Fischer, D., 1997. Enzymatic analysis of riverine bacterioplankton pr oduction. Limnol. Oceanogr. 42, 29-38. Sinsabaugh, R.L., Liska, G., Ostle, N.J ., Lock, M.A., Reynolds, B., Hudson, J., 1996. Microbial activity and enzymic decompos ition processes following peatland water table drawdown. Plant and Soil 180, 121-127. Sinsabaugh, R.L., Weiland, T., Linkins, A. 1992b. Enzymic and molecular analysis of microbial communities associated with lotic particulate organic matter. Freshwater Biology 28, 393-404. SPSS for Windows. Version 11.0.0. Chicago (IL): SPSS Inc, 2001. Stepanauskas, R., Farjalla, V., Tranvik, L., Sv ensson, J., Esteves, F., Graneli, W., 2000. Bioavailability and sources of DOC a nd DON in macrophyte stands of a tropical coastal lake. Hydrobiologia 436, 241-248. SYSTAT Version 10.2. Richmond, CA : Systat Software Inc, 2002. Taylor, B., Parkinson, D., Parsons, W.F.J ., 1989. Nitrogen and lignin content as predictors of litter decay rates: A microcosm test. Ecology 70, 97-104. Templer, P., Findlay, S., Wigand, C., 1998. Sediment chemistry associated with native and non-native emergent macrophytes of a Hudson River marsh ecosystem. Wetlands 18, 70-78. U.S. EPA. 2002. Methods for Evaluati ng Wetland Condition: Vegetation-Based Indicators of Wetland Nutrient Enrichment Washington, D.C: Office of Water, U.S. Environmental Protection Agency, EPA-822-R-02-024. Vaithiyanathan, P., Richardson, C.J., 1999. Macrophyte species changes in the Everglades: Examination along a eutr ophication gradient. J. Environ. Qual. 28, 1347-1358. Venterlink, H. O., Davidsson, T.E., Kiehl, K., Leonardson, L., 2002. Impact of drying and re-wetting on N, P, and K dynamics in a wetland soil. Plant and Soil 243, 119-130. Volk, B. G., 1973. Everglades histosol subsidence: 1. CO2 evolution as affected by soil type, temperature and moisture. Soil and Crop Science Society of Florida Proc. 32, 132-135.

PAGE 143

134 Wetzel, R.G., 1991. Extracellular enzymatic in teractions: Storage, redistribution, and interspecific communication. In : Chr st, R.J. (Ed.) Microbial Enzymes in Aquatic Environments. New York: Springer-Verlag, pp. 6-28. Wetzel, R.G., 1993. Humic compounds from wetlands: Complexati on, inactivation, and reactivation of surface-bound and extracellular enzymes. Verh. Int. Verien. Limnol. 25, 122-128. Wetzel, R. G., 2000. Natural photodegradation by UV-B of dissolved organic matter of difference decomposing plant sources to readily degradable fatty acids. Verhand. Internat. Verein. Limnol. 27, 2036-2043. Wetzel, R.G., Hatcher, P.G., Bianchi, T. S., 1995. Natural photoly sis by ultraviolet irradiance of recalcitra nt dissolved organic matter to simple substrates for rapid bacterial metabolism. Limnology and Oceanography 40, 1369-1380. White, J. R., 1999. Influence of phosphorus loading on biogeochemical cycling of nitrogen in northern Everglades soils Ph. D. Dissertation. Univ. of Florida, Gainesville, FL. White, J. R., Reddy, K.R., 2000. The effect of phosphorus on potential N mineralization rates of northern Ever glades soils. Soil Sci. Soc. Am. J. 64, 1525-1534. White, J.R., Reddy, K.R. 2001. Influence of se lected inorganic el ectron acceptors on organic nitrogen mineralization in Everglades soils. Soil Sci. Soc. Am. J. 65, 941-948. Williams, C. J., Shingara, E. A., Yavitt, J. B. 2000b. Phenol oxidase activity in peatlands in New York State: Response to summer drought and peat type. Wetlands 20, 416-421. Williams, M. A.., Rice, C.W., Owensby, C.E ., 2000a. Carbon dynamics and microbial activity in tallgrass prairi e exposed to elevated CO2 for 8 years. Plant Soil 227, 127-137. Wittmann, C., M., Khknen, A., Ilvesniemi, H ., Kurola, J., Salkinoja-Salonen, M.S., 2004. Areal activities an st ratification of hydrolytic enzymes involved in the biogeochemical cycles of carbon, nitr ogen, sulfur and phosphorus in podsolized boreal forest soils. Soil Biol. Biochem. 36, 425-433. Wittmann, C., Suominen, K.P., SalkinojaSalonen, M.S., 2000. Evaluation of ecological disturbance and intrinsic bioremediation potential of pulp-mill contaminated lake sediment using ke y enzymes as probes. Environmental Pollution, 107, 255-261.

PAGE 144

135 Wright, A. L., Reddy, K.R., 2001a. Heterotr ophic microbial activity in northern Everglades wetland soils. Soil Sci. Soc. Am. J. 65, 1856-1864. Wright, A.L., Reddy, K.R., 2001b. Phosphorus loading effects on extracellular enzyme activity in Everglades wetland soils. Soil Sci. Am. J. 65, 588-595.

PAGE 145

136 BIOGRAPHICAL SKETCH Christopher Ryan Penton was born in the deep fried southern town of Tuscaloosa, AL. After moving to Texas, Georgia and Sout h Carolina he finally came to rest in the quaint, one redlight town of Lecanto, FL. Hi s inquisitive mind, always taking toys apart, blowing up everything with his chemistry set, and asking his parents endless questions eventually propelled him into a B.S. degree in microbiology. He then went to work as a high school chemistry teacher where he continue d with the fascinati on of watching things explode (much to the delight of his students) a nd then went on to work in the Everglades with the South Florida Water Management Di strict. He has one brother Garrett, who attends the University of Florida on a track scholarship and two sisters, an Australian shepherd and a black lab. His wife Ste phanie, who ardently supported him through the past several years, has also recently completed her educa tion, culminating in an M.D. degree. He and his wife curre ntly reside in Michigan, wher e the snow is as abundant as the Florida sun.


Permanent Link: http://ufdc.ufl.edu/UFE0005181/00001

Material Information

Title: Influences of Nutrient Loading, Vegetative Habitats and Simulated Drought on Microbial Enzyme Activities in the Everglades
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0005181:00001

Permanent Link: http://ufdc.ufl.edu/UFE0005181/00001

Material Information

Title: Influences of Nutrient Loading, Vegetative Habitats and Simulated Drought on Microbial Enzyme Activities in the Everglades
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0005181:00001


This item has the following downloads:


Full Text












INFLUENCES OF NUTRIENT LOADING, VEGETATIVE
HABITATS AND SIMULATED DROUGHT ON MICROBIAL ENZYME
ACTIVITIES IN THE EVERGLADES















By

CHRISTOPHER RYAN PENTON


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2004














ACKNOWLEDGMENTS

This research was supported by funds from the South Florida Water Management

District. I would like to thank Dr. Ramesh Reddy for his support and for offering this

unorthodox opportunity. I am indebted to Dr. Susan Newman for her patience, support,

and guidance through the learning curve. She taught me the rigors of scholarly pursuit,

pushed me to excel, and was always there to debate fine details with red ink. A large

portion of thanks goes to my wife Stephanie, who always supported me through my late

night sessions even when she was devoting most of her time to medical school. Her love

and support guided me through the rough times. Finally, I would also like to give thanks

to my family whose love, devotion and encouragement throughout my life have always

motivated me to strive for knowledge and excel in whatever path I choose.
















TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ................................................................... ii

LIST OF TABLES................... ................... ................... ..... ........v

LIST OF FIGURES................... .......................................... ........vii

ABSTRACT.................... ...... .................... ......... viii

CHAPTERS

1 INTRODUCTION............... ..... ............... ......... ...1

Background .................. ............... .. .......... .............
Statement of the Problem ........... ............ .... .........4
Objectives and Scope of Research..................................... 5

2 EXPERIMENTAL CONSIDERATIONS IN DETERMINING
OPTIMUM ENZYME ASSAY CONDITIONS IN
WETLAND SYSTEM S ... ...7... ......... .... ............7

Introduction............. .. .............................7
Materials and Methods........... ......... ............10
Results.................. ................... ................... ......... 13
Discussion ............. .............. ......... .......19
Conclusions........... ..... .................22

3 EFFECTS OF HABITAT DIFFERENTIATION ON MICROBIAL
ENZYME ACTIVITIES IN THE EVERGLADES..................23

Introduction................ ... .... ............ ............ 23
Materials and Methods ................... ..........................28
Results.................. ................... ................... ......... 32
Discussion ............... .................. ....... ......45
Conclusions........... ..... .................50














4 NUTRIENT LOADING EFFECTS ON BIOGEOCHEMICAL AND
MICROBIAL ENZYME DYNAMICS IN THE
EVERGLADES........... ........................ ................. 52

Introduction............ .......................................52
M materials and M ethods................... .......... ............. ......55
Results.............. ..... .................... 64
Discussion........... ................ .......... 81
Conclusions........... ........ ................. 91

5 MICROBIAL ENZYME RESPONSES TO DECREASED
WATER LEVELS IN EVERGLADES PEAT
AND M ARL SEDIM ENTS ............... ............... ...........94

Introduction......... .......... ..................94
Materials and Methods.......... ............ ..........96
Results............ ............. ............. .. .......... 102
Discussion........... ....... ............. ......... 114
Conclusions................................... 117

6 SUMMARY AND CONCLUSIONS...................................120

REFEREN CES ...................................... .......................... .. ......... 125

BIOGRAPHICAL SKETCH................................ ..................136
















LIST OF TABLES


Table page

3-1 Mean chemical properties of the soil and benthic layers.............. ..........35

3-2 Mean enzyme activities of the soil and benthic layers............ ............37

3-3 Correlation coefficients for benthic enzyme and nutrient data.....................37

3-4 Correlation coefficients for soil enzyme and nutrient data................... ....39

3-5 Overall benthic and soil normalized total enzyme activities......................39

3-6 Enzyme ratios................... ........................................... ....... 42

3-7 Correlation coefficients for soil and benthic layer enzyme ratio components
and nutrient data................... .............. ..................... .......... 42

4-1 Benthic nutrient, enzyme, and ratio parameters in P enriched and reference sites
in LNWR and WCA-2A. ........... ..............................66

4-2 Benthic nutrient, enzyme, and ratio parameters in P enriched and reference sites
in W CA -3A and EN P-TS............ ................. .......... ......... ....67

4-3 Benthic layer nutrient and enzyme correlation coefficients.......................69

4-4 Soil layer nutrient, enzyme, and ratio parameters in P enriched and reference sites
in LNWR and WCA-2A. ........... ..............................73

4-5 Soil layer nutrient, enzyme, and ratio parameters in P enriched and reference sites
in W CA -3A and EN P-TS............ ........................... ......... ....74

4-6 Soil layer nutrient and enzyme correlation coefficients............ ..........76

4-7 Average benthic and soil CRR values............. ......................78

4-8 Impact index........................................... ........ .........80

5-1 M ean enzym e activities............ .................. ............ ........... 103









5-2 M ean enzym e activities........... .................. ........................ 104

5-3 Mean Enzyme Index of Carbon Quality (EICQ)....................................107

5-4 Cumulative enzyme activity slopes (CES)......................................108
















LIST OF FIGURES


Figure page

2-1 Effects of (a) time on net activities in Fl sediments, (b) time on net activities
in U3 sediments and (c) varying MUF-cellobioside substrate concentrations.....14

2-2 Effects of (a) time on net activities in F sediments, (b) time on net activities
in U3 sediments and (c) varying MUF-phosphatase substrate concentrations.....15

2-3 Effects of (a) time on net activities in F sediments, (b) time on net activities
in U3 sediments and (c) varying MUF-glucoside substrate concentrations........18

2-4 Effects of (a) time on net activities in F sediments, (b) time on net activities
in U3 sediments and (c) varying L-Leucine amidomethylcoumarin substrate
concentrations................... ................... ......................... .......20

3-1 Radar plots................... ................... ..................... .......40

3-2 Benthic microbial N and P resource allocation in terms of C mineralization......44

3-3 Soil microbial N and P resource allocation in terms of C mineralization.........44

3-4 Bubble graph comparison of (a) benthic and (b) soil enzyme ratio
components Ecell/En, Ecell/Ep and Ecell/Eox.................... .............47

4-1 Graphs showing spikes in enzyme activities at the transitional sites.............79

4-2 Benthic layer bubble plot of Ecell/Ep vs. Ecell/En with Ecell/Eox.................83

4-3 Benthic plot of Log Ecell/En vs Log Ecell/Ep....................................86

5-1 Benthic cumulative enzyme activity (CEA).........................................110

5-2 0 to -10 cm cumulative enzyme activity (CEA)..................................... 111

5-3 Mean lignin, cellulose, and nutrient contents of ENP-TS and WCA-3A
cores ......... ......... ......... ............ ......... 113














Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

INFLUENCES OF NUTRIENT LOADING, VEGETATIVE
HABITATS AND SIMULATED DROUGHT ON
MICROBIAL ENZYME ACTIVITIES IN THE EVERGLADES

By

Christopher Ryan Penton

August, 2004

Chair: Dr. K.R. Reddy
Cochair: Dr. Susan Newman
Major Department: Soil and Water Science

Wetlands frequently function as a long term carbon (C) sink as organic matter

accumulates in response to reduced decomposition rates. In general, enzyme catalyzed

reactions are considered the rate-limiting step in organic matter degradation. Nutrient

status, vegetative community and hydrological regimes significantly influence microbial

enzyme activities. This thesis exams the effects of nutrients, vegetative communities and

water level drawdown on enzymes involved in C, nitrogen (N), phosphorus (P) and lignin

mineralization within four hydrologically distinct areas of the Florida Everglades.

Examining enzyme activities along distinct phosphorus gradients, this study

demonstrates that P enriched sites exhibit lower microbially perceived N and P

limitations on C mineralization in addition to enhanced cellulose decomposition rates.

Phosphorus loading resulted in a decreased microbial mobilization of resources for P

mineralization which resulted in a greater energetic allocation for C mineralization.

Nitrogen appears to become less limiting to C mineralization in the enriched sites within









Everglades National Park. A simple two component model, incorporating total

phosphorus and the relationship between the enzymes involved in C and P mineralization

accounted for 62% of the variability in cellulose decomposition rates.

Vegetative composition was found to significantly alter the allocation of

enzymatic resources due to varying substrate complexities. Carbon mineralization in the

open water was significantly less influences by lignin than the sawgrass habitats. An

index relating hydrolytic and oxidative enzymes, the Enzyme Index of Carbon Quality,

was significantly greater in the open water habitats. This enzymatic index suggests more

favorable C mineralization conditions within the open water communities and provides

insight into the development of the slough-sawgrass ridge topography of the Everglades.

Laboratory water level drawdown was found to effect the microbial allocation of

enzymatic resources in intact Everglades soil cores. Peat dominated sediments exhibited

elevated enzyme activities and a larger response to lower water levels over time.

Enzymatic N mineralization increased significantly in the surficial soil horizons as a

response to water level drawdown. The results indicate that extended periods of low

water may result in enhanced C and N mineralization of Everglades sediments.

The results of these studies indicate that microbial enzyme activities are

responsive to changes in nutrient conditions, vegetation and water levels. The

extrapolation of these enzyme activities to potential decomposition based on perceived C

qualities connects the microbial condition to ecosystem level processes. Changes to the

Everglades ecosystem as a result of management practices can thus be initially

recognized through alterations in the highly responsive microbial component that

functions in nutrient cycling and serves as the basis for the trophic food chain.














CHAPTER 1
INTRODUCTION

Background

Wetlands often have decreased 02 availability as a result of decreased 02 diffusion

rates through water, resulting in decreased overall supply (Ponnanmperuma, 1972).

Additionally, increased 02 demand, due to generally high carbon (C) availability, results

in overall decreases in decomposition rates in wetlands (Reddy and D'Angelo, 1994). An

important and significant role of wetlands in the global ecosystem is C sequestration. The

accumulation of organic matter and thus the role in peat development and global C

sequestration may be affected by changes in ambient nutrient conditions (Wetzel, 1991;

Newman et al., 2003), vegetative composition (DeBusk and Reddy, 1998; Berg, 2000;

Fioretto et al., 2000; Kourtev et al., 2002a), and differing hydrological regimes (Volk,

1973; Elliott et al., 1984; DeBusk, 1996). The mineralization of organic matter by the

resident heterotrophic microbial communities within a system plays a crucial role in the

cycling of C and exerts an influence on the overall energy flow within a system (Elliot et

al., 1984). The activity of the heterotrophic bacteria and fungi whose activity is related to

the biochemical structure, nutrient content, and quantity of available substrates produces

enzymes that are often considered the rate limiting steps in decomposition (Chr6st and

Rai, 1993).

The synthesis and activities of the enzymes involved in organic nutrient

mineralization are most regulated by the induction of macrophytic and macromolecular

substrates within soils (Burns, 1986; Nausch et al., 1998) as well as the intrinsic factors









of plant litter (Linkins et al., 1990a; Carreiro et al., 2000; Kourtev et al., 2002a). Lignin

content has been specifically correlated with decomposition rates (Meentemeyer, 1978;

Mellilo et al., 1989), especially when nitrogen (N) is readily available (Mellilo et al.,

1982; Taylor et al., 1989). Nitrogen content, presented as C:N or lignin:N are also often

used as predictors of litter decomposition rates (Mellilo et al., 1982; Taylor et al., 1989;

Sinsabaugh et al., 1993; DeBusk and Reddy, 1998; Carreiro et al., 2000) and may

influence the decomposition of large molecular weight organic matter (Sinsabaugh et al.,

1993; Carreiro et al., 2000). Phosphorus (P) availability has also been shown to control

decomposition rates in generally P-limited systems (Newman et al., 2001).

Decomposition of organic matter is a community level process that involves

specific interactions within a microbial consortium (Sinsabaugh et al., 1991). The

advantage of enzyme assays is that they present specific information on one process in a

complex community. The microbial degradation of particulate organic matter, such as

plant litter, has been shown to be most influenced by the enzymes involved in

lignocellulose degradation, P cycling, and N cycling (Sinsabaugh et al., 1991; Sinsabaugh

and Moorhead, 1996), which are often considered the rate limiting steps in degradation

(Chr6st and Rai, 1993). Strong relationships have been established between

lignocellulose-degrading enzymes and litter mass loss rates among varying litter qualities

(Sinsabaugh et al, 1992a).

Enzyme activities have the potential to affect all major wetland functions where

decomposition is low. Peat accumulation is dependent upon a lower rate enzymatic

activity resulting in C storage and inorganic nutrients remain sequestered within the

poorly degraded peat matrix when decomposition is low. This impairment of nutrient









cycling by the microbial consortia causes inorganic nutrients to accumulate and to be

retained within the wetland (Freeman et al., 1996). However, the interpretation and

relation of individual enzyme activities to higher trophic processes and observations is

often a difficult and tedious process that frequently results in vague relationships to soil

and microbial parameters in the system (Marsden and Gray, 1986).

A current strategy involves a resource allocation rationale in interpreting enzyme

activities and the MARCIE (Microbial Allocation of Resources Among Community

Indicator Enzymes) model for exposing linkages between individual enzymes

(Sinsabaugh and Moorhead, 1994; Sinsabaugh and Findlay, 1995; Sinsabaugh et al.,

1996, 2002). The model indicates that the expression of enzymes is tied to environmental

nutrient availabilities and that the distribution of enzyme activities can be interpreted as a

resource allocation strategy (Sinsabaugh et al., 2002). Underlying this concept is the

relationship between lignocellulose degradation and environmental N and P

concentrations. Components of this model have been linked to bacterial productivity,

microbial biomass, and particulate organic carbon turnover times (Sinsabaugh and

Findlay, 1995; Sinsabaugh et al., 1997). Specifically, the Enzyme Index of Carbon

Quality (EICQ), which relates the activities of hydrolytic enzymes to those involved in

lignin degradation, has been shown to decrease as decomposition proceeds, which is

reflective of lower substrate quality (Kourtev et al., 2002b). This index has also been

correlated with microbial biomass, productivity and negatively correlated with particulate

organic carbon (POC) turnover time (Sinsabaugh and Findlay, 1995). Furthermore,

enzymes are divided into functional units based on C, N and P mineralization (Ecell, En,

Ep) and lignin degradation (Eox). Enzyme ratios using these compartmentalized units









such as Ecell/Ep and Ecell/En, which reflect apparent P and N control on C

mineralization, respectively, have been correlated with bacterial productivity (Sinsabaugh

and Findlay, 1995; Sinsabaugh and Moorhead, 1994).

Statement of the Problem

The Florida Everglades was once a vast peat wetland that encompassed 10,000 km2

and stretched from Lake Okeechobee southward to the shores of Florida Bay. In the late

1800's and early 1900's large areas of the original landscape was altered for agricultural

and urban development. Water Conservation Areas (WCAs) were created as a result of

compartmentalization by dikes, levees, and water control structures. The remnant

Everglades consists mainly of four managed, hydrologically distinct areas: Arthur R.

Marshall Loxahatchee National Wildlife Refuge (LNWR), WCA-2, WCA-3, and

Everglades National Park (ENP). Intensive agricultural development in the northern

periphery has resulted in the establishment of P and, to a lesser extent, N gradients within

these four hydrologic components of the oligotrophic Everglades. As a consequence,

shifts in macrophyte species composition (Davis, 1943, 1991; Vaithiyanaithan and

Richardson, 1999), increases in net primary production (NPP) (Davis, 1991) and peat

accumulation (Reddy et al., 1993), loss or taxonomic shifts of native periphyton

assemblages (McCormick and O'Dell, 1996; McCormick et al., 1996), increases in

microbial activity and biomass (White and Reddy, 2000), and other ecological changes

have been observed in Everglades areas receiving nutrient inputs from canal water with

total P (TP) concentrations as much as 10 to 30 fold higher than background conditions

(McCormick et al., 1996).

Altered surface flow as a result of compartmentalization and hydrological

management also affected the C cycling within the Everglades. The net area covered by









the deeper sloughs or open water habitats has diminished, being replaced by shallower,

monotypic sawgrass stands. This affected the natural landscape of the Everglades which

was dominated by a slough, slough-ridge topography. The development and maintenance

of this topographical relief is not well understood and has been attributed to particulate

transport and differential decomposition.

Water level fluctuations as a consequence of natural seasonal variations or

hydrologic management have the potential to drastically affect the ability of a wetland

system, such as the Everglades, to sequester C. Previous studies have shown increases in

CO2 flux with lowered water table depths (Volk, 1973) and the highest total C flux under

drained conditions in Everglades peat sediments (DeBusk, 1996). The mineralization of

soil organic N sources has been found to be greater in aerobic than anaerobic conditions

(Reddy and Patrick, 1984; McLatchey and Reddy, 1998), to increase after the drying of

wet soils (Cabrera 1993, Bridgham et al., 1998), and specifically, to be 2-3 times greater

(White, 1999; Venterlink et al., 2002) in drained wetland soils. The impact of water

fluctuations thus has a potentially large role in the nutrient cycling and retention of the

Everglades.

Objectives and Scope of Research

The overall goal of this study was to determine the influences of different

vegetative habitats, nutrient loading and water table drawdown on microbial enzyme

activities in the Everglades. Chapters 2 through 5 address more specific objectives:

1.) Develop an appropriate enzyme experimental method for performing assays

in wetland systems.

2.) Determine the effects of vegetative habitats on microbial enzyme activities in

Water Conservation Area 3A of the Everglades.









2.) Determine the relationships between nutrient conditions and microbial

enzyme activities among all four hydrologic units of the Everglades.

3.) Determine the validity of different enzyme models in predicting potential

decomposition among different litter qualities and in varying nutrient

conditions.

4.) Determine the effects of a simulated drought on enzyme activities in WCA-

3A and ENP-Taylor Slough (ENP-TS) cores.

The overall hypotheses for this study were: (i) Deeper water slough habitats exhibit

accelerated potential decomposition due to substrate quality differences; (ii) Nutrient

loading generally enhances microbial enzyme activities; and (iii) Decreased water levels

result in an enhancement of enzyme activities, suggesting greater potential decomposition

based on perceived soil quality.














CHAPTER 2
EXPERIMENTAL CONSIDERATIONS IN DETERMINING OPTIMUM ENZYME
ASSAY CONDITIONS IN WETLAND SYSTEMS

Introduction

Fluorogenic substrates have been widely adopted in investigations of enzyme

activities in a variety of systems such as lakes, grasslands, wetlands, streams,

groundwater and oceans (Chr6st, 1989; Freeman et al., 1995; Miettinen et al., 1996;

Sinsabaugh and Foreman, 2001; Debosz et al., 1999; Mayr et al., 1999, Shackle et al.,

2000; Wittmann et al., 2000; Burns and Ryder, 2001; Newman et al., 2003; Saiya-Cork et

al., 2002; Wittmann et al., 2004). These compounds have become popular due to their

production of fluorescent compounds, which exhibit less interference by highly colored

phenolic compounds than colorimetric substrates (Sinsabaugh et al., 1991; Freeman et al.,

1995). Methylumbelliferyl (MUF) and amidomethylcoumarin (AMC) substrates are

among the most widely adopted fluorogenic substrates.

Fundamentally different methodological approaches have been adopted in the

literature. The most recent approaches include the use of microtiter plates (Marx et al.,

2001). Standardized approaches use substrate saturating conditions, standard

temperature, and an assay pH that maximizes the fluorescence of the fluorochrome.

Enzymes have differing optimum pH levels which have influenced the use of different

pH conditions between enzymes (Parham and Deng, 2000). The environmental approach

utilizes substrate concentrations similar to the local environment with assay temperature

and pH approximating field conditions. A compromise between these approaches is to









use substrate saturating conditions with assay temperature and pH approximating field

conditions. However, temperatures as high as 45 'C have been used (Sinsabaugh et al.,

1991). Sample alkalinization prior to fluorescence measurement is also often performed

to maximize the fluorescence intensity (Freeman et al., 1995).

Enzyme activity is generally more resolvable with substrate saturation and the

resultant generally linear function of enzyme activity allows for shorter incubation times

and less error. Strong correlations in the literature between potential enzyme activities

and nutrient conditions or correlated processes allow enzyme activities to reflect the

magnitude or process relationships among samples or between systems (Cembella et al.,

1984; Jansson et al., 1988; Foreman et al., 1998).

The methodological heterogeneity of enzyme assays is a reflection of ecological

diversity and localized microbial community structures among studies. For example,

substrate incubation times in the literature range from 10 minutes to 3 days (Miettinen et

al., 1996; Foreman et al., 1998; Debosz et al., 1999; Burns and Ryder, 2001; Kourtev et

al., 2002b). Assay incubation times are generally based on the ability to measure within

the linear portion of the time curve while limiting the opportunity for microbial growth

(Freeman et al., 1995, 1996). Reported MUF and AMC substrate concentrations are also

highly variable between studies and ranges from 1 to 500 [tM in soil and water column

enzyme studies (Chr6st, 1989; Sinsabaugh et al., 1997; Debosz et al., 1999; Bums and

Ryder, 2001; Shackle et al., 2000; Sinsabaugh and Foreman, 2001; Saiya-Cork et al.,

2002; Wittmann et al., 2004). An emphasis of setting substrate concentrations at

saturating conditions has been documented in several studies (Chr6st, 1989; Sinsabaugh

and Findlay, 1995; Burns and Ryder, 2001).









The majority of enzyme studies are performed on soil, water, and litter samples.

However, the use of enzymes in wetland systems is comparatively rare. The combination

of soil and water matrices introduces additional issues in the development of enzyme

assays. Fluorescence interference, due to the rather large accumulation of organic matter

in various stages of decay in wetland systems as well as the accumulation of phenolic

compounds can play a major role in enzyme activity determination. Therefore it is

essential to calculate a quench adjustment for each sample by placing standards in sample

solutions and relate the fluoresences to the standard curve so that samples can be

accurately compared (Freeman et al., 1996; Marx et al., 2001). However, this does not

eliminate other issues that can introduce problems between samples. The presence of

inhibitors, alternative reaction paths and competing substrates can also affect the results

of an assay (Sinsabaugh et al., 1991).

Enzyme activity is usually expressed as the net change in fluorescence from an

initial to a final measurement over time to determine the substrate conversion rate.

Laboratory replicate errors can be quite large in soil enzyme assays due to difference in

particle distribution as well as sample heterogeneity on a minute scale, even in

homogenized samples. What is not clear from the literature is the problem associated

with relying on these endpoint values. The use of net enzyme activity changes without

observing changes in fluorescence over smaller increments of the total incubation time

can result in larger replicate errors. The use of automated spectrofluorimeters allows the

investigator to observe the enzyme kinetics over time and thus identify any replicates

from analysis that produce both outlying net activity and erratic kinetics over time, thus

reducing replicate error and increasing precision.









The goal of this study was to develop an enzyme assay protocol for Everglades

soils to be utilized in further investigations of enzyme activities and their relation to

nutrient regimes. The objectives of this study were to address the effects of (1) varying

substrate concentrations; (2) incubation times; and (3) the use of multiple time point

measurements in determining optimum enzyme assay conditions.

Materials and Methods

Soil cores were collected in Water Conservation Area 2A, a 447 km2 impounded

wetland in the northern Everglades that has received agricultural drainage for over 30

years (Davis, 1991). Three soil cores were collected at nutrient enriched (Fl) and

reference (U3) transect sites along a distinct P gradient. Soil sampling was accomplished

using a 5 cm diameter piston type corer in the spring of 2001. The benthic layer or

flocculent detrital layer was separated from the remaining soil to be utilized in this

experiment. This layer, composed of active algal remains and plant components in

various stages of decay, was chosen due to greater microbial activity and variability

between sites. The samples were stored on ice until transfer to the laboratory. Large

roots and rocks were removed. The samples were homogenized for 5 min using a

handheld Biospec BiohomogenizerTM in a 500 mL beaker. Approximately 10 g wet

weight was transferred to a second 250 mL beaker and 100 mL deionized H20 was

added. The suspension was homogenized for an additional 5 min and 1 mL was

transferred to another 250 mL beaker to which 99 mL of DI H20 was added. The

suspension was mixed and 50 mL of the suspension was transferred to a 50 mL

Eppendorf centrifuge tube. This suspension was used for the assays and refrigerated until

use. Previous studies have shown that freezing generally caused increased enzyme

activities due to the disruption of enzyme complexes and cell lysis as compared to









refrigeration and was thus avoided (Sinsabaugh and Linkins, 1989; Sinsabaugh et al.,

1991). Biostatic agents were not used in this study as their functions differ among

samples and have been found to repress the activity of some enzymes (Sinsabaugh et al.,

1991).

Four enzymes were investigated utilizing the following substrates: MUF-P-D

glucoside (Sigma M3633), MUF-cellobioside (Sigma M6018), MUF-phosphate (Sigma

M8168) and L-leucine amidomethylcoumarin (Sigma L2145). These substrates were

used for the determination of P-glucosidase, cellobiohydrolase, phosphatase, and leucine

aminopeptidase activities, respectively.

Enzyme substrate solutions were prepared in 10 mM Tris-HCl, pH 8.5 in varying

concentration ranges. Buffered conditions stabilize the pH dependent intensity of the

fluorescent product (Chr6st and Krambeck, 1986). Enzyme assays were performed using

a Cytofluor 600TM (Perseptive Biosystems, Inc. Framingham, MA) automated

spectrofluorometer at with KineticalcTM software at 360 nm excitation and 460 nm

emission. Samples were analyzed in Corning 48-well microplates. 360 pL Tris-HCI

pH 8.5 was added to each sample well followed by 400 [tL soil suspension and 40 [tL of

the enzyme substrate. Triplicate replicates of each suspension were used.

The spectrofluorimeter was set to 250 C, to shake for 3 sec before each reading,

and for end-point analysis. Sensitivity was set at 95, which refers to the sensitivity of the

fluorimeter with a maximum value of 100. The plate was inserted and the initial (Ti)

fluorescence was recorded. A reading was performed every 10 minutes for the duration

of the incubation. A similar detailed spectrofluorimeter method used 1 minute intervals









between readings (Marx et al., 2001). A final fluorescence (Tf) was recorded by setting

the fluorimeter to end point analysis at the end of the run.

Various amounts of substrates were added to the samples to establish enzyme

saturation (Chr6st, 1991). Incubations were carried out at 5, 10, 30, 60, 125, 250, and

500 pM for MUF-cellobioside; 5, 10, 20, 30, 50, 125, 250, and 500 [tM for MUF-

phosphate; 5, 10, 20, 30, 50, 70, 100, 250, 500 pM for MUF-glucoside and 5, 10, 20, 50,

100, 300, 400, 800, and 1600 pM for L-Leucine amidomethylcoumarin. These substrate

concentrations reflect in-well measurements as a consequence of dilution and are similar

to the range of substrate concentrations used in another methodological study (Marx et al,

2001). In most cases fluorescence readings were taken over 2 hours. After evaluating

the kinetic curves, the initial fluorescence (Ti) was subtracted from the final fluorescence

(Tf) across incubation time to determine the potential enzyme activity. The largest

difference in fluorescence (Tf-Ti) should yield the best resolution of activity for a given

site. Alkalinization of the samples prior to fluorescence measurement was found

unnecessary as all sample fluoresences were resolvable due to the high sensitivity of the

spectrofluorimeter. Further conversion of potential enzyme activity to pmols MUF or

AMC released g-1 AFDM h-1 was not necessary to meet the goals of this study.

All figures represent averaged data with standard errors. Data from selected

substrate concentrations are presented in figures to represent the extremes associated with

each enzyme.









Results

Cellobiohydrolase

The effect of substrate concentration and incubation time on the potential activity

of cellobiohydrolase was investigated at the enriched and reference sites. Tf-Ti values

are simply referred to as "activity" for the purposes of this study.

The response of the highest MUF-cellobioside concentration (500PM) over

incubation time varied between the enriched (Fl) and reference (U3) sites (Figures 2-la

& 2-1b). Mean activities over the incubation period were 302 and -240 (unitless) with

standard errors 179% and 48% of the mean for the Fl and U3 sites, respectively. Both

sites exhibited erratic readings with initial reductions in activity that was reflected in the

large replicate errors. Fl responded differently than U3 with net increases in activity

after 120 mins.

Positive activity changes were exhibited at the lowest MUF-cellobioside

concentration of 5 [tM at both sites with mean activities of 223 and 295 and standard

errors 20% and 17% of the mean at the Fl and U3 sites, respectively. These activities

were considered low and reflect an insufficient quantity of substrate to obtain near

saturating conditions. This low resolution in the difference between Tf and Ti values

occurred in the range of 5 to 30 [aM.

The behavior of cellobiohydrolase was found to be most linear at a substrate

concentration of 60 [aM. This concentration reflected the greatest activity among the sites

before a large decrease in activity at 125 [aM (Figure 2-1c). Mean activities of 673 and

503 with standard errors 9% and 15% of the mean occurred for the Fl and U3 sites,

respectively. Due to linear increases in activity, extended incubation times were not

















10100 i--
oi *- ~-- -- -i -- fr --i-- --A

9100-

8100

7100

6100

5100

4100


0 20 40 60 80 100 120 14

Time (min)

100

1100 : i t -- k ----------------

i100

'100

.100

i100

100

10 -----------------------------------------
.100
0 20 40 60 80 100 120 140
Time (min)


LEGEND


A 500 gM
* 60 gM
* 5 M


0 100 200 300 400 500 600
Substrate Concentration ([M)



Figure 2-1. Effects of (a) time on net activities in Fl sediments, (b) time on net activities
in U3 sediments and (c) varying MUF-cellobioside substrate concentrations
on net activities.



























2000


0 20 40 60 80 100 120 140

Time (min)


LEGEND


A 500 [M
* 50 [iM
+ 5 [iM


-A
--A'-


-..*....*- -r----+-~'


0 20 40 60 80 100
Time (min)


2000


120 140


100 200 300 400

Substrate Concentration ([tM)


Figure 2-2. Effects of (a) time on net activities in Fl sediments, (b) time on net activities
in U3 sediments and (c) varying MUF-phosphate substrate concentrations
on net activities.









considered important at 60 [tM as substrate saturating conditions were maintained

throughout the duration of the run. However, at higher substrate concentrations, such as

500 lM, incubation time does appear to influence the determination of activity (Fig 2-la

& 2-1b). At 500 [aM, incubation times exceeding approximately 80 minutes for Fl and

120 minutes for U3 would be necessary in order to resolve positive activities at both sites.

Phosphatase

Mean activities for Fl and U3 sites were 3717 and 5433 (unitless) with standard

errors 2.5% and 5.0% of the mean at 500 [aM, respectively (Figures 2-2a & 2-2b).

Relative differences between sites substantially decreased at concentrations at or above

250 [aM. In contrast to the MUF-cellobioside assay, high concentrations did not result in

erratic behavior over the incubation period but rather remained linear (Figures 2-2a & 2-

2b). The lowest concentration of 5 [aM exhibited very linear relationships over time in all

replicates in both nutrient conditions. Mean activity values were 287 and 4050 with

standard errors 6.5% and 1.2% of the mean for the Fl and U3 sites, respectively. The

low activity at the Fl site reflected the need for a higher substrate concentration in order

to counteract potential interference in future experimental runs where microbial

phosphatase production may be lower.

The most appropriate concentration for the purpose of future studies was

determined to be in the 50 [aM range (Figure 2-2c), before a substantial decrease in

activity occurred in the U3 sample. Linear responses occurred in both the Fl and U3

samples over the length of the incubation period with mean activity values of 2770 and

9067 with standard errors 2.5% and 2.9% of the mean, respectively. These values reflect

the highest activity value while still minimizing laboratory error. Incubation time at 50

[aM does not appear to be an important consideration since linear relationships were









exhibited in both samples. This concentration was 8 times lower than another wetland

study that examined phosphatase linearity in order to establish optimum substrate

concentrations (Kang and Freeman, 1999).

P-glucosidase

The highest MUF-glucoside concentration of 500 pM yielded Tf-Ti activity

differences with average values of 565 and 989 in Fl and U3 samples respectively, with

laboratory replicate standard errors of 1.8% and 9.3% of the mean (Figures 2-3a & 2-3b).

Quasi-linear graphs were produced with the highest activity values occurring at the U3

site. 5 pM average activity values were 267 and 88 with standard errors representing

7.6% and 50.8% of the mean for the Fl and U3 sites, respectively. The relationship of

enzyme activity between sites shifted from the general trend at approximately 200 [tM

and continued until approximately 300 [tM (Figure 2-3c). This illustrates the profound

effect that substrate concentrations can have on formulating conclusions.

The most consistent substrate concentration in terms of net activity, linearity and

error was identified at 100 aM. These runs yielded mean activities of 416 and 248 with

standard errors of 8.1% and 4.8% of the mean for the Fl and U3 sites, respectively.

There was a linear response of enzyme activity over time with no apparent effect of

increasing incubation time on mean activities or differences among the sites.

Leucine Aminopeptidase

Leucine aminopeptidase exhibited typical activity responses to increasing substrate

concentrations (Figure 2-4c). Peak activity was around 800 [aM and decreased in

response to a higher or lower concentration. As substrate concentration was decreased

from 800 [aM, the differences between the two sites and the replicate errors also

























3000


2500


20 40 60 80 100 120 140

Time (min)


9000

8800

8600

8400


7400
0 20 40 60 80
Time (min)


LEGEND


A 500 [M
* 100 [M
* 5 [M


iT T- -


100 120 140


0 100 200 300 400 500 600
Substrate Concentration ([M)


Figure 2-3. Effects of (a) time on net activities in F sediments, (b) time on net activities
in U3 sediments and (c) varying MUF-glucoside substrate concentrations
on net activities.


, 1 1,









decreased. At 1600 [tM the mean activities were 11167 and 18833 with standard errors

of 16.5% and 13.4% of the mean for the Fl and U3 sites, respectively.

The 300 utM runs exhibited mean activities of 6197 and 11866 with standard

errors of 1.0% and 1.0% of the mean for the F and U3 sites, respectively. Linear graphs

at 300 utM were the most consistent, showing no apparent influence of incubation time on

site differences or individual activities (Figs 2-4a & 2-4b). Mean activity values for the

10 [tM runs were 417 and 710 with standard errors representing 5.9% and 7.8% of the

mean. This concentration represented the lowest activity difference between samples

among the range of substrate concentrations.

Discussion

These results demonstrate the need for the investigation of incubation time and

substrate concentration effects to ensure the validity of data in future experiments.

Enzyme activity is extremely sensitive to changes in environmental conditions, thus it is

necessary to construct the assays such that enzyme efficiency is maintained through a

range of conditions that may be encountered. The enzymes in this study were responsive

to changes in nutrient concentrations. These responses were evident in different

relationships to changes in substrate concentration and incubation time. In some cases,

different contrasts between the enriched and reference samples were observed as the

substrate concentration was altered. This may lead the investigator to assume inverse

relationships between samples, which are not necessarily reflected in the majority of

other substrate concentrations. Therefore, the most contrasting conditions must be

assayed in order to develop the most consistent substrate concentration and incubation

time in relation to changes in sample properties.














60000

50000

40000

30000

20000

10000


60'

50'

40'

30l


--4


--A-








-------*----------- -*- --------------*

0 10 20 30 40 50 60
Time (min)

000

000

000


20000


10000


LEGEND


A 1600 [M
* 300 [M
* 10 M


-- .--------------- ----- ----4- 4- -*


0 10


20 30 40 50 60

Time (min)


500 1000 1500


2000


Figure 2-4. Effects of (a) time on net activities in Fl sediments, (b) time on net activities
in U3 sediments and (c) varying L-Leucine amidomethylcoumarin substrate
concentrations on net activities.









The use of the curves produced over the total incubation period is necessary to

observe enzyme behavior over time. While the endpoint based activity (Tf-Ti) produced

as a result of incubation may indicate a net change in substrate converted, there is often

erratic conversion behavior over time. Without the use of these curves, the data would be

considered valid since there would be no reason to discard any values that appear to

reflect net activity. The resultant experimental protocol for future studies discards any

non-linear runs from further analysis when the activity reflects outlying values.

In determining optimum substrate concentrations, it is generally evident that very

high and low concentrations tended to produce erratic changes in activity over time and

lower differences between sites. The high concentration response can be attributed to

intermolecular quenching in which fluorescence is masked by the presence of other

fluorescent and non-fluorescent molecules. Therefore, there can be a false decrease or a

lesser increase in substrate conversion as measured by the fluorimeter. At low

concentrations, saturation kinetics did not appear to be satisfied, thereby resulting in a

lower perceived reaction rates. This is due to the lack of adequate substrate that fails to

retain a constant reaction velocity. The appropriate MUF and AMC substrate

concentrations that were determined in this study fall within the range of 40 to 200 [tM

documented in other studies of water column and soil enzymes (Bums and Ryder, 2001;

Shackle et al., 2000; Sinsabaugh and Foreman, 2001; Saiya-Cork et al., 2002).

Incubation time was shown to influence the resolution between the two study

sites. In most cases, incubation times less than 30 minutes result in a decrease in

resolution. In addition, substrate conversion is generally erratic with large errors between

replicates. An incubation time of 45 minutes is deemed suitable for capturing activity as









well as enabling a rapid turnover of samples in future studies. This lies within the ranges

observed in other studies (Kourtev et al., 2002b; Foreman et al, 1998; Burns and Ryder,

2001). Short incubation times minimize any changes that may occur in the microbial

community (Hoppe, 1993). It should be noted that longer incubation times are acceptable

if enzyme dynamics or environmental factors warrant. Such incubations may be

performed in order to resolve very low enzyme activities or overcome phenolic

quenching in heavily lignified or tannic samples. However, in order to investigate the

temporal function of the enzymes over the incubation period, the samples must remain in

the automated spectrofluorimeter, thus delaying the processing of other samples.

Therefore, the most expeditious 45-minute incubation is preferred to decrease sample

storage time and increase efficiency.

Conclusions

These results demonstrate the need for the investigation of incubation time and

substrate concentration effects to ensure the validity of data in future experiments.

Further, it is prudent that enzyme kinetics be incorporated into the experimental

methodology due to the differences among substrate concentration responses over time

that were observed in this study. This process can be done expeditiously using a

microplate approach to enzyme assays and analyzing with an automated

spectrofluorimeter, compared to older cuvette based measurements. It is also important

to analyze a range of concentrations in the most widely contrasting environmental

conditions within the study area in order to determine the most appropriate substrate

concentration. However, the final determination of the warranted incubation conditions

will be dependent on the specific goals of the study, sample matrix, number of assays, the

desire for precision versus accuracy, and available analytical equipment.














CHAPTER 3
EFFECTS OF HABITAT DIFFERENTIATION ON MICROBIAL ENZYME
ACTIVITIES IN THE EVERGLADES

Introduction


Organic matter accumulation within wetlands is a consequence of the balance

between net primary production (NPP) and microbial heterotrophic metabolism.

Microbial decomposers play a crucial role in carbon (C) cycling and are responsible for

driving the C energy flow up the detrital food chain. The mineralization of organic

nutrients by the microbial community exerts an appreciable influence on energy flow by

regulating nutrient availability for further decomposition and primary production (Elliot

et al., 1984). Mineralization of plant matter is governed by the chemical and physical

properties of available substrates such as lignin, nitrogen (N), and phosphorus (P)

(DeBusk and Reddy, 1998; Berg, 2000; Fioretto et al., 2000; Kourtev et al., 2002a) as

well as other environmental and physiochemical influences. Therefore, changes in these

parameters associated with different litter types and nutrient conditions have the potential

to alter peat accumulation rates and potentially topography over time.

The microbial degradation of particulate organic matter (POM), such as plant litter,

has been shown to be most influenced by the enzymes involved in lignocellulose

degradation, P cycling and N cycling (Sinsabaugh et al., 1991; Sinsabaugh and

Moorhead, 1994), which are often considered the rate limiting steps in degradation

(Chr6st and Rai, 1993). Due to the sometimes complex macrophytic structure, a large

quantity of diverse enzymes may be necessary to complete degradation (Eriksson and









Wood, 1985; Ljungdahl and Eriksson, 1985). The synthesis and activity of enzymes may

be most regulated through the induction by macrophytic and macromolecular substrates

present in the soils (Burns, 1986; Nausch et al., 1998), with strong relationships

established between lignocellulose-degrading enzymes and litter mass loss rates among

various litter qualities (Sinsabaugh et al., 1992a). Consequently, soils beneath different

plant species have been shown to support microbial communities that differ in both

structure and function (Degens and Harris, 1997; Grayston et al., 2001; Kourtev et al.,

2002a; Kourtev et al., 2003) and can affect sediment nutrients through plant uptake and

rhizosphere characteristics (Templer et al., 1998). Litter breakdown rates and enzyme

activities have been shown to vary among species and have been attributed to intrinsic

factors of the leaves (Linkins et al., 1990a & 1990b; Carreiro et al., 2000; Kourtev et al.,

2002b) such as specific chemical composition (Linkins et al., 1990a & 1990b). For

example, litter N content, presented as C:N or lignin:N are often used as predictors of

decomposition rates (Melillo and Aber., 1982; Sinsabaugh et al., 1993; DeBusk and

Reddy, 1998; Carreiro et al., 2000).

Microbial degradation by-products and exogeneous nutrients may also influence

microbial activity. For example, polyphenols, the by-products of lignin degradation,

have the potential to inhibit enzyme complexes. These compounds can, in certain

appropriate environmental conditions, become the dominant regulatory mechanisms

involved in microbial respiration (McClaughtery and Linkins, 1990; Wetzel, 1993). The

availability of nitrogen and other nutrients has been shown to control the early phases of

decay, when mass loss is less than 30% (Taylor et al., 1989). Conversely, the addition of

N has been shown to retard the decomposition of large molecular weight organic matter









by the repression of phenol oxidase (Sinsabaugh et al., 1993; Carreiro et al., 2000). As

decomposition proceeds, nutrient availability becomes progressively less important and

lignin content controls litter decay rates (Taylor et al., 1989; Berg, 2000). As a

consequence of these complex interactions, the interpretation of individual enzyme

activities as simple responses to environmental substrate concentrations may not serve to

adequately predict the actual microbial community dynamics.

A current strategy for predicting microbial degradation rates involves the use of a

resource allocation rationale and the MARCIE (Microbial Allocation of Resources

Among Community Indicator Enzymes) model for exposing linkages between individual

enzymes (Sinsabaugh and Moorhead, 1994; Sinsabaugh and Findlay, 1995; Sinsabaugh et

al., 1997, 2002). The model is based on the premise that enzyme mediated

decomposition of complex molecules are the rate-limiting step in C mineralization. It

relates bacterioplankton production or litterbag mass loss through a first order model that

includes C, N, and P allocation. It further indicates that the expression of enzymes is tied

to environmental nutrient availabilities and that the distribution of enzyme activities can

be interpreted as a resource allocation strategy (Sinsabaugh et al., 2002).

Related to the MARCIE model, the Enzyme Index of Carbon Quality (EICQ) is a

relative index of the normalized activities of the hydrolytic enzymes to oxidative or lignin

degrading enzymes. EICQ has been shown to be correlated with microbial biomass

(r-0.71), productivity (r=.80) and negatively correlated with particulate organic carbon

(POC) turnover time (r=-0.99) (Sinsabaugh and Findlay, 1995). EICQ values and other

enzyme ratios reflect microbial community responses to the perceived abundance of

nutrients as well as lignin and thus respond to changes in substrate and environmental









nutrient conditions in terms of microbial resource allocation dynamics. MARCIE model

components, such as Ecell/Ep and Ecell/En, which reflect apparent phosphorus and

nitrogen control on C mineralization, have also been correlated with bacterial

productivity (Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997). An underlying

concept is that lignocellulose degradation by extracellular enzymes is tied to

environmental N and P concentrations. Since only a certain amount of metabolic energy

can be utilized in the production of enzymes, an abundant expression of one enzyme

resulting from a lack of directly utilizable substrate will result in less energy available for

the production of other enzymes.

In organic matter dominated wetlands, such as the Everglades, the microbial

mineralization of organic matter is a key process involved in the initial development,

accumulation, and maintenance of the peat profile. Differential decomposition processes

over time may result in the development of topographic features related to local

biogeochemical characteristics which are, in turn, influenced by larger scale ecosystem

processes. Specifically, the Everglades were historically dominated by a slough-ridge

landscape interspersed with tree islands. This landscape consisted of dense sawgrass

ridges with soil surfaces 2 to 3 feet higher than the adjacent, deeper sloughs (Baldwin and

Hawker, 1915). However, after over 50 years of compartmentalization and altered

surface flow, the area covered by the deeper sloughs is diminishing, being replaced by

shallower, monotypic sawgrass (Cladium jamaicense crantz) stands. Changes in overall

productivity have accompanied shifts in vegetative communities, which have the

potential to alter the nutrient storage capability of the system (Davis, 1991).









Differences between the habitats are primarily observed in plant litter chemistry,

characteristics, and localized nutrient conditions. For example, approximately 13-fold

greater root P accumulation rates in cattail over sawgrass have been documented

(Lorenzen et al., 2001) as well as greater P accumulation in cattail in high P conditions

(Davis, 1991; Koch and Reddy, 1992; Chiang et al., 2000). Conversely, sawgrass root

phosphatase activities were also found to be significantly greater than cattail in low P

conditions (Kuhn et al., 2002). In a 4 year P dosing study in the Everglades, tissue P

accounted for approximately 5% of the added P in sawgrass while less than 2% was

recovered from plant tissue at the slough site due to the disappearance of the native

community (Chiang et al., 2000). Localized dissolved organic nitrogen (DON) and

dissolved organic carbon (DOC) bioavailability to a microbial community has also been

shown to differ between cattail and spikerush (Eleocharis spp.) stands in a coastal humic

lake (Stepanauskas et al., 2000). These differences in nutrient conditions and plant

characteristics ultimately influence the efficiency and composition of the resident

heterotrophic microbial decomposers, which should be exhibited as changes in enzyme

activities.

The main objective of this study was to determine the effects of vegetative habitat

types and nutrient loading on apparent enzyme activity and apply the results to varying

enzyme models to account for differences in potential decomposition among the three

dominant habitat types (sawgrass, cattail, and open water). By investigating these

relationships, one possible factor in the development of the topographical relief of the

Everglades may be more fully understood.









Materials and Methods

Study Sites

The field study sites for this project were within Water Conservation Area 3A

(WCA-3A) located in Broward County, FL. WCA-3A is a freshwater marsh

incompletely impounded by a combination of levees and canals and covers an area of

2,339 km2. The area is a mosaic of sawgrass stands and open water areas or sloughs

interspersed with tree islands. The result of over 40 years of agricultural runoff into the

Everglades has been the establishment of a P gradient originating in the northern regions

of the Everglades at points downstream of discharges. Cattail (Typha spp.) has invaded

into areas previously dominated by sawgrass (Davis, 1991; DeBusk et al., 1994; DeBusk

et al., 2001).

Six sites were selected for sampling, representing P enriched and slightly P

enriched (designated reference) cattail, sawgrass, and open water sites. Enriched sites

were designated ECS, ESS, EOS, ECB, ESB, and EOB where the first letter is designated

E=enriched, the second; C=cattail, S=sawgrass, O=open water, and the third; S=soil

layer, and B=benthic layer. Reference P sites followed the same nomenclature with RCS,

RSS, ROS, RCB, RSB, and ROB where R=reference site. GPS coordinates were

26'04.900' 80o49.520' for the enriched site and 26'02.410' 80o48.870' for the reference

site. Sites were located near established transects utilized by the South Florida Water

Management District for water quality monitoring. Dense cattail, sparse sawgrass, and

open water areas were present at the P enriched sites, located approximately 1.22 km

from the canal inflow. Sparse cattail, stands of sawgrass, and open water or slough

communities consisting of water lily (Nymphaea), spikerush and periphyton constituted









the reference (slightly impacted) site, located approximately 5.83 km from the canal

inflow.

Sampling

Soil cores were obtained utilizing a 5 cm stainless steel piston type corer with

butyrate inserts on September 9 2001. Cores were collected in triplicate at each

habitat-site combination. The coring procedure involved pushing the coring apparatus

through the soil layer to a depth of approximately 30 cm. A serrated metal knife was

used to cut around the perimeter of the tube to sever large roots and other plant matter.

The core was extruded and the benthic matter, defined as the unconsolidated or pourable

core fraction, was separated from the soil layer. The soil section was extruded to a depth

of 10 cm and both the benthic and soil layers were stored in plastic bags on ice for

transport to the laboratory.

Soil Preparation

Soil sample analysis began within 24 hours after field collection. Each layer,

corresponding to the 0 to -10 cm soil or benthic, was prepared separately. 10 g sub-

samples were placed in pre-weighed aluminum pans for dry mass (DM) and ash free dry

mass (AFDM) determination. Dry mass was determined by incubating the pans in a

drying oven for 36 hours at 1050 C. Ash weight determination involved ashing the

samples at 5000 C for 2 hours and AFDM was calculated by subtracting the ash weight

from the dry weight measurements.

Samples were transferred to a 500 mL beaker and large objects, such as rocks, were

discarded. The samples were homogenized for 10 minutes with a Biospec

BiohomogenizerTM, resulting in a soil slurry. 10 g of the slurry was added to 100 mL DI

H20 and homogenized for an additional 5 minutes. 10 mL of the suspension was added









to 90 mL of DI H20 and subsequently transferred to a 100 mL centrifuge tube. The

suspensions were refrigerated until use (Sinsabaugh et al., 1991).

Enzyme Analysis

Hydrolytic enzyme activity was determined using methylumbelliferyl (MUF) and

aminomethylcoumarin (AMC) substrates. Substrate concentrations were optimized at

saturating conditions. The activities of P-glucosidase (BGL), cellobiohydrolase (CBH),

phosphatase (PHO), leucine aminopeptidase (LEU), phenol oxidase (PHE) and

peroxidase (PER) were assayed using MUF-P-D-glucoside (Sigma M3633), MUF-

cellobioside (Sigma M6018), MUF-phosphate (Sigma M8168), L-Leucine

amidomethylcoumarin (Sigma L2145), L-3,4-dihydroxyphenylalanine dopaA), and

DOPA + H202 as substrates, respectively.

MUF and AMC substrate conversion was measured using a Cytofluor 600TM

automated spectrofluorimeter (PerSeptive Biosystems, Inc., Framingham, MA) with

KineticalcTM software at 360 nm excitation and 460 nm emission at 200 C. Assays were

performed using Corning 48-well culture plates in which 400 [iL of sample, 360 [iL of

10 mM Tris-HCI pH 8.5, and 40 [iL of substrate were added. Stock substrate

concentrations were 2000 [iM for MUF-P-D glucoside, 1000 [iM for MUF-phosphate,

1200 [iM for MUF-cellobioside, and 6000 [iM for L-Leucine aminomethylcoumarin

resulting in well concentrations of 100, 50, 60 and 300 [iM, respectively. Each sample

analysis was performed in quadruplicate. Initial and final fluorescence measurements, as

well as measurements every five minutes, were taken during the 1 hour incubation.

Graphs produced from the readings taken every five minutes were analyzed to ensure that

linear kinetics were being observed. Final apparent enzyme activities were calculated as

moless substrate released g-1 AFDM h-1.









Phenol oxidase activity was determined by adding 2.0 mL soil suspension to 2.0

mL 10 mM L-DOPA (L-dihydroxphenylalanine) in 10 mM Tris-HCL pH 8.5 in 10 mL

EppendorfFM centrifuge tubes. Peroxidase activity was determined by adding 2.0 mL soil

suspension to 2.0 mL 10 mM L-DOPA with 200 [iL 0.3% H202. The solutions were

vortexed for 30 seconds and placed on a shaker plate in a light-proof box for 45 minutes.

The solutions were then centrifuged at 3000 rpm for 30 seconds and 500 [iL supernatant

was extracted and quadruplicates were placed in a CorningTM 48-well culture plate.

Controls consisting of 250 [tL DI H20 and 250 [tL 10 mM L-DOPA solution were added

to the remaining wells. Absorbance was read at 360/460 nm excitation/emission on the

spectrofluorimeter. Quenching was performed using the soil suspensions with a total

volume of 500 [iL in each well.

Sample nutrient analysis was performed by DB Labs, Rockledge, FL. Total

phosphorus (TP) (EPA 365.2), total nitrogen (TN) (MVP), total organic carbon (TOC)

(MVP), calcium (Ca)(SW7140), magnesium (Mg)(SW7450), and lignin (AOAC 973.18)

analysis was performed using standard methods on homogenized samples.

Models

Extracellular enzymes can be grouped into four categories: Ecell (BGL and

CBH), En (LEU), Ep (PHO), and Eox (PHE and PER); allowing the enzymes to be

separated into those involved in C, N, and P mineralization as well as lignin degradation,

respectively. While there are many enzymes in each category, it is assumed that the

activity of one, or more preferably a few enzymes in each group are sufficiently

correlated such that the activity of a few can act as indicators for the entire group.









Enzyme activities are averaged if there is more than one enzyme in the group

(Sinsabaugh et al., 1997).

Potential decomposition models were constructed from enzymic data using

normalized enzyme activity across both the benthic and soil layers. This was done in

order to compare different enzymes at the same scale so that the more active enzymes do

not heavily weight the calculations of model components. The Enzymic Index of Carbon

Quality (EICQ) compares the activities of oxidative to hydrolytic enzymes (Sinsabaugh

and Findlay, 1995). In addition, several other indices were formulated from the data.

Ecell/En reflects apparent N control over C mineralization, Ecell/Ep is a relative measure

indicating P control over C mineralization, and Ecell/Eox reflects apparent lignin control

over C mineralization.

To improve normality and heteroscedescity, data were log transformed before

statistical analysis using SAS version 8 statistical software (SAS, 1999). A split-plot

model for the data was adopted for subsequent ANOVA analysis using PROC GLM in

SAS. The whole plot corresponded to the sites in reference to the nutrient gradient while

the subplots referred to the specific vegetative communities within the whole plots.

Multiple comparisons were made between habitats, sites, and habitat-site combinations.

Other analyses were performed using the Statistical Package for the Social Sciences (TM

SPSS) version 11.0.0. All regressions and significant differences are significant at the

p<.05 level unless otherwise noted.

Results

Nutrient Composition

Average nutrient concentrations for the benthic and soil layers with standard errors

are presented in Table 3-1. Benthic TP values were significantly different among sites









and all habitats in the reference plots while soil TP varied among individual sites.

Benthic TP concentrations were generally higher than soil TP values.

Total nitrogen values were significantly different among individual sites while only

the sawgrass was affected by location on the gradient in the benthic layer. Soil TN values

were significantly different among individual sites and between the enriched and

reference open water and cattail habitats. TN concentrations were higher in the enriched

habitats with the exception of the sawgrass soil and benthic samples. TN was correlated

with TP in the benthic layer (Table 3-3).

Benthic TOC values were significantly different among sites, although the effect of

the nutrient gradient was only significant at the open water site. Soil TOC was

significantly different among sites with the cattail (p<0.0001) and open water affected by

location on the gradient. There was no clear relationship between TOC concentrations in

the benthic and soil layers. Benthic TOC was significantly correlated with TP and TN,

while soil TOC was correlated with TP (Tables 3-3 and 3-4).

Carbon to nitrogen (C:N) ratios exhibited differences based on habitat type in both

the soil and benthic layers, range from 11.4 to 16.2, and were significantly correlated

with PHO and LEU in the soil layer. The open water habitat exhibited the lowest C:N

ratios among habitats with the largest difference occurring between the open water and

sawgrass habitats. There was no clear relationship between depth and C:N ratios.

Lignin data were only available for the soil layer with a small sample size (n=6).

Lignin content increased from the reference to the enriched sites and was the highest in

the sawgrass habitat. Lignin ranged from 6% at the RCS site to 49% at the ESS site.









Enzyme Activities

All of the enzyme assays, with the exception of PER, yielded detectable activities.

PER activity was not detectable in the majority of samples. Standard errors of hydrolytic

enzyme activity were generally lower than the oxidative enzymes, ranging from 4% to

46% of the mean. Laboratory replicate error was generally less than field replicate

variability.

Benthic 1-glucosidase (BGL) activity did not vary significantly between habitats or

along the gradient (Table 3-2). Though not significant, the cattail and open water

communities generally exhibited higher activities in the enriched sites. Benthic BGL was

significantly correlated with PHO, CBH, and LEU, while soil layer BGL was only

weakly correlated with LEU (Tables 3-3 & 3-4). BGL activities were consistently higher

in the benthic layer, ranging from approximately 2 to 40 times the activity in the soil

layer, however, benthic and soil layer BGL activities were not significantly correlated to

one another. Soil layer BGL was significantly correlated with TN but did not exhibit

significant differences between habitats or along the gradient.

Phosphatase activities were significantly lower than those reported in Everglades

periphyton mats (Newman et al., 2001). While soil PHO significantly (p<0.0001)

increased with distance from the inflow, there were no significant differences among

habitats. This increase in PHO with distance from the inflow has been documented in

other Everglades areas (Wright and Reddy, 2001b) as well as decreases in PHO with P

loading in periphyton (Newman et al., 2001). Benthic PHO was significantly correlated

with BGL, CBH, and LEU while soil layer PHO was negatively correlated with LEU and

PHE. Benthic PHO activities were greater than in the soil, ranging from approximately 2

to 100 times the activities of the soil layer with the only significant change occurring









Table 3-1. Mean chemical properties of the soil and benthic layers. Units are
expressed as g kg1, lignin content expressed as % by mass with corresponding
standard errors. TN= total nitrogen, TP=total phosphorus, TOC=total organic
carbon, C:N=ratio of TOC to TN. Ben=Benthic layer, Soil=Soil 0 to -10 cm
layer. N/A represents lack of replication due to insufficient sample. ND=not
determined.

Site TN TP TOC C:N Lignin
Cat 36.9 0.5 0.9 0.0 445.5 3.5 12.1 0.3 ND
Ben Enr Saw 28.7 1.1 0.7 0.1 439.7 16.1 15.3 0.2 ND
Open 37.0 0.3 1.7 0.1 435.3 2.6 11.8 0.0 ND
Cat 6.0 + 0.7 0.1 0.0 702.3 + 5.5 11.8 0.5 ND
Ben Ref Saw 28.9 1.1 0.5 0.1 403.0 3.0 14.5 0.2 ND
Open 22.7 0.2 0.3 0.0 262.3 15.1 11.5 0.6 ND
Cat 32.4 0.2 1.3 0.5 408.0 NA 12.6 0.8 41.7 NA
Soil Enr Saw 28.1 + 1.0 0.9 0.1 454.7 21.9 16.2 0.5 49.4 NA
Open 38.0 0.4 1.0 0.1 438.5 2.5 11.5 0.1 39.2 NA
-- - - ----------------------------- ------- --. .Y --- ^- --
Cat 28.4 + 1.0 1.1 + 0.1 377.3 + 18.2 13.3 + 0.2 5.6 + NA
Soil Ref Saw 29.0 5.1 0.5 0.1 394.0 12.5 13.9 1.5 32.6 NA
Open 30.2 1.9 0.3 0.0 343.0 28.0 11.4 0.2 29.5 NA









between the enriched and reference cattail habitats. This relationship has been

documented in other aquatic systems in Florida (Newman and Reddy, 1992; Wright and

Reddy, 2001b). Benthic PHO was correlated to the nutrient parameters TN, TOC, and

TP, while soil PHO was negatively correlated with TOC and TP.

Neither benthic or soil CBH activity were significantly different between habitats

or along the gradient. As with BGL, there were no trends evident among the habitats or

any consistent relationship with the gradient. Benthic CBH was significantly correlated

with BGL, PHO, and LEU. There were no significant correlations among enzymes in the

soil layer (Table 3-3). Benthic CBH was correlated to the nutrient parameters TN and

TP. No significant correlations between CBH and nutrient parameters were observed in

the soil layer.

Cattail and open water benthic LEU was significantly higher at the enriched sites.

Open water and sawgrass soil layer LEU was significantly higher at the enriched sites.

Sawgrass LEU was lower (p<0.05) than the other habitats at the enriched sites. Benthic

LEU was significantly correlated with BGL, PHO, and CBH (Table 3-3), while soil layer

LEU was correlated with PHO and PHE (Table 3-4). Benthic LEU was higher than soil

layer activities in the enriched sites, however this relationship was mixed at the reference

sites. Benthic LEU was negatively correlated to the nutrient parameters TN, TOC, and

TP while the soil LEU was negatively correlated with TOC and TP.

Phenol oxidase did not vary significantly between habitats or along the gradient.

Standard errors were greater in the PHE assays than the hydrolytic assays, consistent with

higher variability reported by Sinsabaugh (personal communication). Activities were










Table 3-2. Mean enzyme activities of the soil and benthic layers. Data reflects
means with corresponding standard errors. ND=not determined. Units for
3-glucosidase (BGL), cellobiohydrolase (CBH), phosphatase (PHO), and
leucine aminopeptidase (LEU) are [moles substrate released g-1 AFDM h-1.
Units for PHE and PER are moles DICQ released g-1 AFDM h-1.


BGL
0.0800.031
0.0230.002
0.080 0.023
0.0440.018
0.0440.013
0.0250.010

0.0150.001
0.0100.001
0.0160.004
0.001+0.001
0.0180.008
0.0120.002


CBH
0.0720.022
0.0300.004
0.0540.013
0.0190.009
0.0300.008
0.0340.019

0.0380.002
0.0250.005
0.0260.005
0.0320.004
0.0340.011
0.021 0.004


PHO
7.343.77
2.560.35
6.222.20
1.87+1.10
4.280.88
3.492.22

0.130.017
0.0620.009
0.110.017
0.81+0.12
0.76 0.14
1.540.13


Table 3-3. Correlation coefficients for benthic enzyme and nutrient data. Correlations
significant at p<0.05 unless otherwise noted. *=p<0.0001, NS=not significant.
BGL= P-glucosidase, PHO= phosphatase, CBH= cellobiohydrolase, LEU=
leucine aminopeptidase, PHE=phenol oxidase, TN=total nitrogen and
TOC=total organic carbon.


BGL PHO CBH LEU PHE TN TOC
PHO 0.73
CBH 0.83* 0.89*
LEU 0.71 0.94* 0.92*
PHE NS NS NS NS
TN NS 0.60 0.45 -0.65 NS
TOC NS 0.54 NS -0.64 NS -0.99*
TP NS 0.56 0.46 -0.63 NS 0.89* 0.87*


Site
ECB
ESB
EOB
RCB
RSB
ROB

ECS
ESS
EOS
RCS
RSS
ROS


LEU
4.541.75
1.690.44
5.061.48
0.870.47
1.960.50
2.420.97

1.170.16
0.770.16
1.110.19
1.030.10
2.090.37
2.250.01


PHE
124.552.3
45.96.3
37.210.0
75.911.8
60.235.5
14.914.9

55.57.1
35.18.4
18.49.7
59.37.3
98.14.5
78.814.2


PER
ND
ND
ND
ND
ND
ND

0.120.02
0.160.03
0.070.04
ND
ND
ND









generally higher at the reference habitats in the soil layer and were within the ranges of

reported activities in the Everglades (Wright and Reddy, 2001b) and other systems

(Sinsabaugh and Linkins, 1990; Sinsabaugh et al., 1992b). Another study found no

significant changes in phenol oxidase activity along the gradient in WCA2A in the

Everglades (Wright and Reddy, 2001). The positive correlation of soil PHE with LEU

suggests that a greater proportion of N may be in recalcitrant form. Soil PHE was also

negatively correlated with TOC. There was not a significant trend of decreasing phenol

oxidase activity with depth.

Overall normalized hydrolytic (EThyd) and hydrolytic plus oxidative (EThyd+ox)

enzyme activities in the benthic layer exhibited mixed relationships with the gradient and

among habitats (Table 3-5). EThyd values were within the range reported in a previous

study of a riverine system (Sinsabaugh et al., 1997). EThyd and EThyd+ox in both the cattail

and open water habitats were significantly greater at the enriched sites. However, there

were no consistently significant differences between habitats. The differences between

the EThyd and EThyd+ox values demonstrate the contribution of apparent lignin degradation

to the overall enzyme activity. The reference open water benthic layer (ROB) exhibited

the smallest contribution of oxidases to the overall enzyme activity (7%) while the

reference sawgrass habitat soil layer (RSS) showed the greatest (37%). The contribution

of normalized enzyme activities are illustrated in radar plots for the benthic and soil

layers (Figure 3-1) for visualizing the relationships presented for the calculated MARCIE

model components. It is interesting to note that the shapes of the radar plots for each

layer are similar in both the enriched and reference sites. However, energy allocation

does appear to be different in the soil versus benthic layers.









Table 3-4. Correlation coefficients for soil enzyme and nutrient data. Regressions
significant at p<0.05 unless otherwise noted. *=p<0.0001, NS=not significant.
BGL= P-glucosidase, PHO= phosphatase, CBH= cellobiohydrolase, LEU=
leucine aminopeptidase, PHE=phenol oxidase, TN=total nitrogen and
TOC=total organic carbon.


PHO
CBH
LEU
PHE
TN
TOC
TP


BGL PHO CBH LEU PHE
NS
NS NS
NS -0.71 NS
NS 0.71 NS 0.64
0.60 NS NS NS NS
NS -0.79 NS -0.50 -0.65
NS -0.54 NS -0.65 NS


TOC







0.57


Table 3-5. Overall benthic and soil normalized total enzyme activities. Total activities
are expressed as the sum of the hydrolytic (BGL, CBH, PHO, LEU) and
oxidative (PHE) enzymes (EThyd+ox) as well as the hydrolytic enzymes only
(EThyd) with standard errors. Enr=enriched site, Ref= reference site. Values
are unitless.


ET (hyd) ET (hyd+ox)
Cat 2.27 0.88 2.82 1.11
Benthic Enr Saw 0.81 0.11 1.01 0.13
Open 2.10 0.60 2.27 + 0.56
Cat 0.71 + 0.34 1.05 0.30
Benthic Ref Saw 1.11 0.28 1.38 0.29
Open 0.94 0.35 1.01 0.39
Cat 0.59 0.02 0.84 0.06
Soil Enr Saw 0.38 0.06 0.54 0.03
Open 0.49 0.09 0.57 0.08
Cat 0.51 + 0.05 0.77 + 0.07
Soil Ref Saw 0.75 0.20 1.18 0.22
Open 0.65 0.05 1.00 + 0.10













b.) reference


c.) enriched


d.) reference


Figure 3-1. Radar Plots of benthic (a & b) and soil (c & d) layer average normalized
enzyme activities for the enriched and reference habitats showing the basis
for MARCIE model components.


MARCIE Model Components

Ecell/Ep values reflect the apparent P control on C mineralization based on

resource allocation (Table 3-6). Higher Ecell/Ep values indicate decreased production of

PHO in relation to the hydrolytic enzymes involved in C mineralization. Therefore,

higher Ecell/Ep values suggest decreased control of P on C mineralization. Benthic and

soil Ecell/Ep were significantly higher at all the enriched habitats. The enriched benthic

layer did not show any significant differences among the habitats. However, all of the

habitats exhibited significantly different benthic Ecell/Ep values at the reference sites


a.) enriched









with the open water habitats having the lowest values. Significantly greater values were

found in the soil, compared to the benthic, among the enriched habitats. Benthic Ecell/Ep

was negatively correlated with TN and TOC while soil Ecell/Ep was positively correlated

with TOC and TP (Table 3-7).

Ecell/En values reflect apparent N control on C mineralization. Higher Ecell/En

values, like Ecell/Ep, indicate decreased LEU production in relation to C mineralization,

which in turn suggests lesser control of N on C mineralization. Ecell/En values were

within the high range of values reported in a study of a riverine system (Sinsabaugh et al.,

1997). The enriched site benthic and soil layers exhibited no significant differences

among the habitats. However, at the reference site, in both layers, all of the habitats were

significantly different than one another with the lowest Ecell/En associated with the open

water habitat. Benthic Ecell/En was negatively correlated with TN, TOC, and TP while

soil values were positively correlated with TOC and TP.

The relative degrees of apparent N and P influences on C mineralization can be

compared simultaneously in order to further differentiate the sites (Figures 3-2 & 3-3).

The reference benthic layer sites generally lie along a line from the open water habitats

(low N, low P) to the cattail habitats (high N, high P). The grouping of these habitats is

more distinct than that of the enriched habitats, which are generally in the two high P

quadrants, with varying relative N levels.

In comparison, the soil layer relationships are more distinctive in relation to the

gradient, although not as clear as the benthic layer in terms of habitat grouping. Enriched

sites are found in the upper portion of the high N, high P category while the majority of









Table 3-6. Enzyme ratios calculated with normalized apparent enzyme activity
means and associated standard errors. Values expressed are unitless.
Ecell/Eox is the apparent lignin influence on C mineralization, Ecell/Ep and
Ecell/En are the apparent phosphorus and nitrogen influences on C
mineralization and EICQ is the Enzyme Index of Carbon Quality. ND reflects
a lack of replication due to zero oxidative enzyme activities in the
denominator, thus reflecting conservatively low estimates.


Site Ecell/ Eox


ECB
ESB
EOB
RCB
RSB
ROB
ECS
ESS
EOS
RCS
RSS
ROS


Ecell/Ep Ecell/En


EICQ


1.18 0.11 1.43 0.26 1.11 0.20 5.35 0.54
1.06 0.06 1.26 0.10 1.16 0.28 4.99 0.41
4.47 + 2.48 1.33 0.13 0.86 0.13 19.02 10.48
0.81 0.41 2.38 0.41 2.52 0.30 3.42 1.32
0.58 0.15 0.97 0.19 1.16 0.09 3.25 0.33
2.24 ND 0.57 0.03 0.71 0.10 7.98 ND
0.91 0.09 26.19 2.69 1.57 0.25 3.44 0.20
1.10 + 0.46 33.81 2.02 1.55 0.26 4.00 1.24
1.24 0.11 22.60 2.06 1.25 0.04 4.50 + 0.27
0.63 0.08 3.03 0.12 1.29 0.16 2.95 0.19
0.48 0.15 4.04 0.62 0.79 0.13 2.68 0.39
0.39 0.06 1.28 0.22 0.47 0.07 2.92 0.23


Table 3-7. Correlation coefficients for soil and benthic layer enzyme ratio
components and nutrient data. Regressions significant at P<0.05 unless
otherwise noted. NS=not significant. Ecell/Ep and Ecell/En are the apparent
phosphorus and nitrogen influences on C mineralization and EICQ is the
Enzyme Index of Carbon Quality. TN= total nitrogen, TOC=total organic
carbon and TP=total phosphorus.


Benthic Soil
TN TOC TP TN TOC TP
Ecell/En -0.77 -0.70 -0.60 Ecell/En NS 0.63 0.87
Ecell/Ep -0.56 -0.51 NS Ecell/Ep NS 0.81 0.62
Ecell/Eox NS NS 0.57 Ecell/Eox NS 0.68 0.56
EICQ 0.47 NS 0.62 EICQ 0.51 NS NS









reference sites are found in the low N, high P quadrant in the soil layer. The reference

habitat soil layers exhibit similarities to the corresponding benthic layers, with apparent

grouping, especially in terms of the open water habitat. The reference open water habitat

is again found in the lowest section of the low N, high P quadrant with the cattail habitats

generally grouped in a higher N area than the other habitats. The differences between the

benthic and soil layers are probably due to lower microbial activities in the soil layer,

which results in a lower demand for N and P as well as shifts in community composition

due to decreased carbon quality with depth.

Ecell/Eox values reflect the apparent lignin control on C mineralization. Higher

Ecell/Eox values indicate decreased PHE production, which is a reflection of apparent

lignin content, suggesting lesser lignin control on C mineralization. The reference open

water habitat benthic Ecell/Eox was significantly higher than the other habitats while the

enriched open water was higher (p<0.05) than the sawgrass. These differences were not

found in the soil layers. The enriched sites generally exhibited higher values than the

reference, with only 3 of 6 comparisons in both layers significantly different. Ecell/Eox

was correlated with TP in the benthic layer in addition to TOC and TP in the soil layer.

By combining the average Ecell/Ep, Ecell/En, and Ecell/Eox values for each

habitat, the general trends for each habitat concerning P, N, and lignin influences on C

mineralization can be constructed (Figure 3-4). Comparing the enriched to reference

benthic sites, the largest apparent shifts in N and P dynamics occurs in the cattail habitats

while the smallest shift occurs in the sawgrass sites. The open water habitats exhibited













0.6

0.5

0.4

0.3

0.2

0.1



-0.1
0 -



-0.2

-0.3

-0.4
-0.3


0.3 0.4 0.5 0.6


Figure 3-2. Benthic microbial N and P resource allocation in terms of C mineralization.
The Y-axis represents apparent P limitation while the X-axis represents
apparent N limitation on C mineralization. Values are log-transformed for
comparison. N and P conditions reflect microbially perceived concentrations.

1.8
Low N, High P High N, High P
1.6 M ESS ESS
ECS ESS A ECS
1.4 EOSI
EOS A ECS
1.2 EOS

1

0.8
0.8RSS
0.6
0.6 RSS
U M R RSS A RCS A RCS
0.4 RCS

0.2 0 ROS
*ROS
0 RROS

-0.2


-0.4 L
-0.6


-0.5 -0.4 -0.3 -0.2 -0.1
Ecell/En


0 0.1 0.2 0.3 0.4


Soil microbial N and P resource allocation in terms of C mineralization. The
Y-axis represents apparent P limitation while the X-axis represents apparent
N limitation on C mineralization. Values are log-transformed for
comparison. N and P conditions reflect microbially perceived concentrations.


-0.2 -0.1 0 0.1 0.2
Ecell/En


Figure 3-3.









the least amount of lignin influence in both the enriched and reference sites in the benthic

layer, which are accompanied by the greatest amount of apparent N limitation with the

exception of the ROS site. The reference open water site also exhibited the greatest N

and P limitation in relation to C mineralization in both layers. The distinct grouping in

the soil layer reflects primarily changes in P among the enriched habitats and changes in

N among the reference habitats (Figure 3-4). However, this relationship is not evident in

the benthic layer.

Enzyme Index of Carbon Quality (EICQ) values did not vary in a significantly

consistent fashion along the gradient. However, open water habitat EICQ values were

significantly higher in the benthic layers at both the enriched and reference sites,

indicating greater perceived C quality. EICQ was only weakly correlated with the

benthic nutrient parameters TN and TP as well as soil TN.

Discussion


The use of enzyme comparisons, such as Ecell/Ep and Ecell/En, are based on the

premise that the use of energy to produce certain extracellular enzymes reduces the net

energy available for the expression of other enzymes. The relationships between certain

enzyme groups are unique and serve to convey information that is specifically related to

perceived environmental conditions. The MARCIE model components, originally

developed as a relationship to mass loss rates (Sinsabaugh and Moorhead, 1994 & 1996),

may be used as a tool for determining microbial productivity and perceived nutrient

limitations (Jackson et al., 1995; Sinsabaugh and Findlay, 1995; Sinsabaugh et al., 1997).

The use of these models as predictive components in this paper is due to the lack of

corresponding decomposition data provided by litter bags or other means. However,









relationships with published microbial community characteristics and activities are

referenced to provide a strong basis for the use of these models.

The model components Ecell/En and Ecell/Ep generally predicted lower N and P

limitation on C mineralization at the enriched habitats. This is concurrent with decreases

in nutrient limitations that have been documented in other enriched Everglades regions

(McCormick et al., 1996). Ecell/Ep values spanned an almost 20-fold greater range than

Ecell/En. This relationship is consistent with 10-fold greater ranges observed in an

aquatic study (Sinsabaugh et al., 1997) and is due largely to greater phosphatase

variability along the primarily P gradient. Additionally, Ecell/Ep was as much

as 4-fold lower than values reported in an aquatic study (Sinsabaugh et al., 1997),

reflecting the low P nature of this system. Positive relationships between Ecell/En and

Ecell/Ep and microbial productivity have also been observed (Sinsabaugh et al., 1997).

The greater potential productivity at the enriched habitats is further supported by

generally higher Ecell/Eox values at these sites, reflecting lower apparent lignin influence

on C mineralization. The shift in the microbial response along the gradient may reflect

changes in microbial community composition that have been observed in enriched sites in

the Everglades (DeBusk and Reddy, 1998), with subsequent increased C mineralization

rates.

Changes in the apparent quality of the litter with nutrient concentrations such as

decreased C:N and C:P ratios and (Shaver and Melillo, 1984; Craft and Richardson,

1995; Bridgham et al., 1996; DeBusk and Reddy, 1998) have the potential to influence

microbial activity which may be reflected by shifts in Ecell/Eox, Ecell/En, and Ecell/Ep

along the gradient. These changes has been observed in shifts of sawgrass C:N:P ratios
























Enr Open 4 47
15


10


05


RefCattaill081 Q9


Enr Cattail 1 18
SEnr Sawgrass 1 06

Ref Sawgrass 0 58


00
00 05 10 15
Ecell/En


20 25 30


00 02 04 06


Enr Cattail 091


EnrOpen 1 24


Ref Saw 049
ag=


08 1 0
Ecell/En


Ref Cattail 0 62

12 14 16 18 20


Bubble graph comparison of (a) benthic and (b) soil enzyme ratio
components Ecell/En, Ecell/Ep, and Ecell/Eox, representing apparent
nitrogen, phosphorus, and lignin control on carbon mineralization,
respectively. Bubble size represents Ecell/Eox. Ref=reference site, Enr=
enriched site, open=open water habitats. Note the change in scale between
the two layers.


Figure 3-4.


Ref Open 0 39&









between enriched and reference sites in WCA-2A in the Everglades with values of

2520:63:1 and 360:9:1 in the enriched and reference sites, respectively, which were

correlated with higher C turnover rates (Koch and Reddy, 1992). Additionally, the

general increases in TP, TN, and TOC of both the benthic and soil layers at the enriched

habitats appears to support greater the higher Ecell/Ep and Ecell/En values.

The supposition of greater productivity at the enriched habitats is also supported by

larger EICQ values, which are generally limited to the more active benthic layer. Greater

EICQ values in the benthic layer are synonymous with early stages of decomposition,

which have been shown to have a higher carbon quality (Sinsabaugh and Linkins, 1990;

Sinsabaugh and Moorhead, 1994; Sinsabaugh and Findlay, 1995). The positive

relationships between EICQ and productivity (r=0.80), microbial biomass (r=0.71) as

well as the negative relationship with POC turnover time (r=-0.99) (Sinsabaugh and

Findlay, 1995) therefore also suggest greater C mineralization at the enriched sites.

Additionally, this relationship between EICQ and microbial biomass is supported in the

Everglades by significantly higher microbial biomass C (MBC) and microbial biomass P

(MBP) observed within the benthic layer along a P gradient in WCA-2A (Wright and

Reddy, 2001b), greater soil respiration in higher P conditions (DeBusk, 1996), as well as

faster decomposition in a P dosing study in WCA-1A (Newman et al., 2001). High levels

of productivity (Sinsabaugh et al., 1997) have also been associated with more eutrophic

conditions. This relationship is dependent on a sufficient carbon flow that includes

saccharides and amino acids (Sinsabaugh et al., 1997). Increases of TOC, general

increases in primary production, and lower apparent N limitation on C mineralization at

the enriched habitats further supports these relationships. Therefore, the consequence of









these elevated model parameters is predicted to be increased decomposition at the

enriched habitats. However, increases in net primary production, resulting in greater C

flow, may exceed any increase in decomposition, resulting in a net accumulation of

organic matter (Davis, 1991).

Ecell/En and Ecell/Ep values did not predict more favorable microbial conditions in

the open water habitat. Rather, this habitat was generally the most limiting in terms of N

and P on C mineralization. This may be a consequence of algal competition for N and P

resources due to the more prolific periphyton community in the open water habitats. This

contradiction in prediction between the N and P models with Ecell/Eox and EICQ may be

due to varying shifts in the type of microbial community responses to different

environmental changes. The nutrient gradient is most related to changes in N and P and

thus it would be expected that these nutrients would be the main driving force behind

microbial changes when habitat types remain constant. Conversely, changes in

vegetative types would be expected to influence the actual degradability of the C source

by the microbial community, due to varying shielding effects on nutrients by refractory

compounds within the plant structures. In fact, the mineralization of organic C was most

affected by C quality with total P concentration and the lignocellulose index (LCI)

accounting for 91% of the variability in aerobic C mineralization of plant litter along the

WCA-2A gradient (DeBusk and Reddy, 1998). Thus, different enzyme components

would be involved in regulating the productivity between these two types of changes.

These contrasts in enzyme components most related to productivity have been

documented with the BGL and PHO correlation with bacterial production in nutrient

enriched and unenriched mesocosms, respectively (Chr6st and Rai, 1993). This suggests









that the coupling and de-coupling of enzymes to production may be dependent on

localized biogeochemical properties.

The components Ecell/Eox and EICQ predicted that the open water habitats were

likely to have higher potential decomposition rates than the other two vegetative habitats,

especially in the benthic layer. EICQ has been shown to be most affected by litter

quality, independent of specific site characteristics in a litter bag study (Kourtev, 2002b).

The lower apparent lignin influence on C mineralization at these habitats is most

certainly related to changes in substrate structure and composition. Generally lower C:N

ratios also predict a more favorable substrate quality within the open water habitats. This

increase in potential decomposition over time would be expected to develop a lower

elevation in relation to the sawgrass habitat, in particular. Additionally, this increase is

coupled with a lower C input due to lower primary productivity in the form of vegetative

turnover. Conversely, the lower potential degradation at the sawgrass habitats, coupled

to a much greater C input, would be expected to exceed the metabolic capability of the

resident microbial community, resulting the accumulation of organic matter. The

combination of these factors support field observations of lower elevation open water

habitats and higher elevation sawgrass stands.

Conclusions

Lower apparent lignin influence on C mineralization, lower C:N values, and higher

EICQ values all predict that the open water habitat benthic layers are likely to have

greater substrate qualities than cattail and sawgrass habitats with higher potential

decomposition. When coupled to a lower C input, the results support field

observations of lower elevations within the open water habitats. Therefore, this study









points to the potentially significant role that differential decomposition has on the

development of the Everglades landscape.

The limited range of enzymes used in this study, while allowing for the expeditious

processing of samples, may not accurately reflect the full array of mineralization

pathways. The use of a greater range, possibly utilizing exo- and endocelullases,

chitinase, and other enzymes will allow for a greater enzymatic resolution. Additionally,

the use of litter decomposition bags in this type of study would allow the effects of

substrate concentration on enzyme activities to be extrapolated to actual decomposition

rates.














CHAPTER 4
NUTRIENT LOADING EFFECTS ON BIOGEOCHEMICAL AND MICROBIAL
ENZYME DYNAMICS IN THE EVERGLADES

Introduction

The accumulation of organic carbon within wetlands is the result of the balance

between net primary production (NPP) and microbial heterotrophic metabolism.

Microbial decomposers play a crucial role in carbon (C) cycling and are responsible for

driving the C energy flow up the detrital food chain. The mineralization of organic

nutrients by the microbial community exerts an appreciable influence on energy flow by

regulating nutrient availability for further decomposition and primary productivity (Elliot

et al., 1984). Although most of the organic C in aquatic and marsh systems is processed

and recycled entirely by the microbial community without entering the higher order food

webs (Wetzel, 1984), alterations in microbial decomposition associated with nutrient

loading have the potential to drastically affect the higher trophic levels of the system.

The relationships between the heterotrophic microbial community and other effects of

nutrient loading may serve to further facilitate the understanding of ecosystem

component linkages within wetlands. Due to their contribution to the ability of the soil to

degrade organic matter, enzyme activities may be especially significant in determining

changes in environmental conditions and through the subsequent effects on the resident

microbial community (Frankenberger and Dick, 1983; Dick, 1997).

The decomposition of plant organic matter is largely regulated by the synthesis and

activities of extracellular enzymes produced by microbial communities, which operate at









the biochemical level (Sinsabaugh and Moorhead, 1994), and are induced by the presence

of macromolecular and macrophytic substrates within soils (Nausch et al., 1998). Due to

the complex structure of macrophytic tissue, a large quantity of diverse enzymes may be

necessary to complete degradation (Eriksson and Wood, 1985; Ljungdahl and Eriksson,

1985). The mineralization of plant matter is governed by the chemical and physical

properties of available substrates such as lignin, nitrogen (N), and phosphorus (P)

availability (DeBusk and Reddy, 1998; Berg, 2000; Fioretto et al., 2000; Kourtev et al.,

2002a & b) as well as other environmental and physicochemical influences.

The nature of the many interactions that occur between extracellular enzymes and

inducing and repressing environmental components are not well understood and may play

a significant role in regulating enzyme activity. Some examples of potentially significant

interactions include the polyphenolic inhibition of enzyme complexes as a consequence

of lignin degradation. These compounds can, in certain appropriate environmental

conditions, become the dominant regulatory mechanisms involved in microbial

respiration (McClaughtery and Linkins, 1990; Wetzel, 1993). The addition of N has been

shown to retard the decomposition of large molecular weight organic matter by the

repression of phenol oxidase (Eriksson et al., 1990; Blanchette 1991; Sinsabaugh et al.,

1993; Hammel, 1997; Carreiro et al., 2000) and UV photolysis has been shown to

enhance enzyme activities by relieving the inhibition by humic compounds (Wetzel et al.,

1995; Boavida and Wetzel, 1998; Wetzel, 2000). As a consequence of these complex

interactions, the interpretation of individual enzyme activities as simple responses to

environmental substrate concentrations may not serve to adequately predict the actual

microbial community response.









Enzyme activities have the potential to reflect changes in nutrient cycling as a

result of changes in water and soil quality (Wetzel, 1991). For example, in a P-limited

system the activities of phosphatases may be a dominant regulatory mechanism

controlling microbial productivity. Phosphatase regenerates inorganic P through the

hydrolysis of organic P to inorganic P (Wetzel, 1991). It is repressed by P enrichment

and dissolved reactive phosphorus (DRP) (Jansson et al., 1988; Chr6st, 1991; Newman et

al., 2003), and has been recommended as a parameter to assess P impact in the

Everglades (Newman et al., 2003). Phosphatase is one example of a suite of enzymes

involved in organic matter degradation and nutrient cycling that can be affected by

nutrient loading (Wetzel, 1991; Newman and Reddy, 1993; Marx et al., 2001).

Interpreting and relating individual enzyme activities to higher order trophic

processes and observations is often a difficult and tedious process and are often not

linked to soil and microbial parameters in the field (Marsden and Gray, 1986). A current

strategy involves the use of a resource allocation rationale and the MARCIE (Microbial

Allocation of Resources Among Community Indicator Enzymes) model for exposing

linkages between individual enzymes (Sinsabaugh and Moorhead, 1994; Sinsabaugh and

Findlay, 1995; Sinsabaugh et al., 1997, 2002). The model is based on the enzyme

mediated decomposition of complex molecules being the rate-limiting step in

decomposition. It further indicates that the expression of enzymes is tied to

environmental nutrient availabilities and that the distribution of enzyme activities can be

interpreted as a resource allocation strategy (Sinsabaugh et al., 2002). An underlying

concept is that lignocellulose degradation by extracellular enzymes is tied to

environmental N and P concentrations. For example, model components such as









Ecell/Ep and Ecell/En, which reflect apparent P and N control on C mineralization,

respectively, have been correlated with bacterial productivity (Sinsabaugh and Findlay,

1995; Sinsabaugh et al., 1997). Since a certain amount of metabolic energy can be

utilized in the production of enzymes, an abundant expression of one enzyme resulting

from a lack of directly utilizable substrate will result in less energy available for the

production of other enzymes.

The objectives of this experiment were to (1) determine the effects of nutrient

loading on microbial enzyme activities in the four hydrologic units of the Everglades, (2)

investigate differences in microbial responses to nutrient impacts among the areas, and

(3) establish a potential decomposition model relating enzyme activities to cotton strip

decomposition rates (CRR). In addition, determinations of possible interactions between

site specific biogeochemical parameters and microbial enzyme activities were

investigated.

Materials and Methods

Site Description

The Florida Everglades is an oligotrophic system with reference surface water total

phosphorus (TP) levels averaging less than 10 pg L-1 throughout the interior of the marsh

(McCormick and O'Dell, 1996; McCormick et al., 2000). The result of decades of

nutrient loading from the Everglades Agricultural Area (EAA) has been the establishment

of P and, to a lesser extent, N gradients within the four hydrologic compartments of the

remnant Everglades. Consequently, shifts in macrophyte species composition (Davis,

1943, 1991; Vaithiyanaithan and Richardson, 1999), increases in NPP (Davis, 1991) and

peat accumulation (Reddy et al., 1993), loss or taxonomic shifts of native periphyton

assemblages (McCormick and O'Dell, 1996), increases in microbial activity and biomass









(White and Reddy, 2000), and other ecological changes have been observed in

Everglades areas receiving nutrient inputs from canal waters with TP concentrations as

much as 10-30 fold higher than background (McCormick et al., 1996).

Field study sites were located in transects situated along nutrient gradients in four

distinct hydrologic units of the Everglades: Arthur R. Marshall Loxahatchee National

Wildlife Refuge (LNWR), Water Conservation Area 2A (WCA-2A), Water Conservation

Area 3A (WCA-3A), and Taylor Slough within Everglades National Park (ENP-TS).

Loxahatchee National Wildlife Refuge (LNWR) is the northernmost of the

hydrologic areas in the Everglades, completely impounded by levees and canals, and

encompasses 566 km2. The phosphorus gradient in this area exhibits a steep decline with

surface water and soil TP levels decreasing to reference levels within 2.2 km of the L-7

canal (SFWMD, 2003). LNWR hydrology is primarily rainfall driven (54%) and is,

unlike the rest of the Everglades, an acidic soft water system. Sites were chosen on a

transect located along a nutrient gradient extending from the L-7 canal and represented

enriched (XI), transitional (X2), and reference (X4) P conditions which are designated as

the ENR, TRANS, and REF sites, respectively. The average surface water TP

concentrations were between 7.2 and 12.3 [ig L-1 between 1996 and 2001 for the

reference site (SFWMD, 2003).

Water Conservation Area 2A is a 442 km2 area located in the northern

Everglades, completely enclosed by canals and levees, and is the most studied among the

regions. The primary source of water and nutrient loading to the area are the S-10

structures that transfer water from agricultural areas and LNWR via the Hillsboro Canal.

Nitrogen and phosphorus gradients exist in the water column and periphyton tissue









(McCormick and O'Dell, 1996). Soil P concentrations have ranged from approximately

400 to 1600 mg kg-1 in the reference and enriched areas, respectively (Reddy et al., 1993;

DeBusk et al., 1994). Three sites were sampled, designated Fl, F4, and U3. These sites

are referred to as ENR, TRANS, and REF for the enriched, transitional and reference

nutrient sites, respectively. The enriched site is characterized by robust stands of Typha

domingensis Pers. (cattail) and a floating surface layer of Lemnaceae (duckweed). The

surface waters contain a large quantity of particulate organic matter that is composed of

plant matter in various stages of decay. The reference site consists of Eleocharis spp.

(spike rush), Nymphae (water lily), Utricularia spp. (blatterwort) and benthic and floating

periphyton mats surrounded by stands of Cladium jamaicense Crantz (sawgrass), with

little suspended organic matter. Water depth was, on average, shallower at the reference

site with clear surface water and particulate peat intermixed in some areas with a thin

carbonate layer underlying the surficial periphyton mats.

Water Conservation Area 3A (WCA-3A) encompasses 2,012 km2 and is

predominantly a vast sawgrass marsh interspersed with sloughs, tree islands, and wet

prairies. It is the only area not completely enclosed by levees. The highest annual mean

surface water TP levels in the inflow and interior marsh were 67.3 and 20.3 .ig L-1 for

1978-2000 (Newman et al., 2002). The extent of downstream enrichment is similar to that

of LNWR and extends approximately 3 km from the inflow (Newman et al., 2002).

Sampling sites were located along a SFWMD water quality monitoring transect located at

0.5E, 1.5W, and Nmeso for the enriched, transitional, and reference sites, respectively.

The reference site consisted mostly of water lilies, spikerush, and periphyton.









Everglades National Park (ENP) is a 5,569 km2 wetland consisting primarily of

marl forming wet prairies, sawgrass stands, freshwater sloughs, and mangrove stands at

the southern periphery. Marl forming prairies are characterized by the formation of

calcitic mud, especially in the southern regions. Surface water flow into Taylor Slough

originates at the S-332 structure at the southeastern side of the park. The overlying water

column at the enriched site contains less suspended organic matter, resulting in a clearer

profile when compared to the other enriched sites. Lower net primary productivity (NPP)

at the enriched site is manifested in decreased litterfall, as compared to the other

hydrologic units. Vegetative changes relating to nutrient input occur in a relatively short

distance from the canal inflow and is generally limited to the deeper slough region, less

than 50 meters in width. The reference site is characterized by dense periphyton and

epiphyton accompanied by spikerush. Sampling sites were located along a SFWMD

water quality monitoring transect designated as 0.5W, 1.5E, and Smeso for the enriched,

transitional, and reference sites, respectively (Newman et al., 2002).

Sampling

Soil cores were obtained using a 10 cm thin-walled stainless steel corer on

12/14/2001, 5/18/2002, and 10/14/2002. Cores were collected in triplicate at each site.

The coring procedure involved pushing the coring apparatus through the soil layer to a

depth of approximately 30 cm. During insertion, a serrated metal knife was used to cut

around the perimeter of the corer to sever large roots and other plant matter. The top of

the core was sealed with a plastic cap allowing the core to be excavated from the soil

whole. The core was extruded and the benthic matter, defined as the unconsolidated or

pourable core fraction, was separated from the soil layer. The soil section was extruded









to a depth of 10 cm and both the benthic matter and soil sections were stored in plastic

bags on ice for transport to the laboratory.

Soil Preparation

Soil sample analysis began within 24 hours of field collection. Each layer,

corresponding to the 0 to -10 cm soil and benthic layers, were prepared separately. 10 g

sub-samples were placed in pre-weighed and pre-ashed aluminum pans for dry mass

(DM) and ash free dry mass (AFDM) determination. Dry mass was determined by

incubating the pans in a drying oven for 36 hours at 1050 C. Ash weight determination

involved ashing the samples at 5000 C for 2 hours.

Samples were transferred to a 500 mL beaker and large objects, such as snail shells

and rocks, were discarded. The samples were homogenized for 10 minutes with a

Biospec BiohomogenizerTM, resulting in a soil or benthic slurry. 10 g of the slurry was

diluted to a concentration of 10-3 with deionized water and homogenized for an additional

5 minutes. The resulting suspension was transferred to a 100 mL centrifuge tube and

refrigerated until use (Sinsabaugh et al., 1991). Enzyme analysis began within 6 hours of

sample preparation.

Enzyme Analysis

Hydrolytic enzyme activity was determined using methylumbelliferyl (MUF) and

amidomethylcoumarin (AMC) substrates. Substrate concentrations were optimized at

saturating conditions. The activities of P-glucosidase (BGL), phosphatase (PHO), leucine

aminopeptidase (LEU), phenol oxidase (PHE) and peroxidase (PER) were assayed using

MUF-P-D-glucoside (Sigma M3633), MUF-phosphate (Sigma M8168), L-Leucine

amidomethylcoumarin (Sigma L2145), L-3,4-dihydroxyphenylalanine dopaA), and

DOPA + H202 as substrates, respectively.









MUF and AMC substrate enzymatic analysis was measured using a Cytofluor

600TM (PerSeptive Biosystems, Inc., Framingham, MA) automated spectrofluorimeter

with KineticalcTM software at 360 nm excitation and 460 nm emission at 200 C. Assays

were performed using Coming 48-well culture plates in which 400 [iL of sample, 360

[IL of 10 mM Tris-HCI pH 8.5, and 40 [iL of substrate was added. Stock substrate

concentrations were 2000 [iM for MUF-P-D glucoside, 1000 [iM for MUF-phosphate,

and 6000 [iM for L-Leucine aminomethylcoumarin resulting in well concentrations of

100 [iM, 50 [iM, and 300 [iM, respectively. Each sample run was performed in

quadruplicate. Initial and final fluorescence measurements as well as measurements

every five minutes were taken during the 1 hour incubation. Changes in fluorescence

were determined by subtracting the initial from the final fluorescence. Graphs produced

from the readings taken every five minutes were analyzed to ensure that linear kinetics

was being observed. Concentrations of MUF and AMC released were calculated by the

application of standard curves to the initial and final fluoresences. The absolute

difference in concentrations yielded the substrate released during the incubation period.

The effects of quenching on MUF and AMC substrates were determined in order to

account for fluorescence blocking or absorption effects caused by coloring, particle

suspension, humic matter, self-quenching, or other inhibitions in the suspensions. MUF

and AMC standards were placed in each sample suspension to determine the quench

percentage of each matrix. The final and initial fluoresences were then converted using

the appropriate quench percentage.

Phenol oxidase and peroxidase analysis was performed by adding 2.0 mL soil

suspension to 2.0 mL 10 mM L-dihydroxyphenylalanine (L-DOPA) dissolved in 10 mM









Tris-HCI pH 8.5 in 10 mL EppendorfFM centrifuge tubes. The solutions were vortexed

for 30 seconds and placed on a shaker plate in a light-proof box for 45 minutes. The

solutions were then centrifuged at 3000 rpm for 30 seconds, 500 [iL supernatant was then

extracted and placed in quadruplicate in a CorningTM 48-well culture plate. Controls

consisting of 250 [IL DI H20 and 250 [iL 10 mM L-DOPA solution were added to the

remaining wells.

Sample nutrient analyses was performed by DB Labs, Rockledge, FL. Total

phosphorus (TP) (EPA 365.2), total nitrogen (TN) (MVP), total organic carbon (TOC)

(MVP), calcium (Ca) (SW7140), and magnesium (Mg) (SW7450), and lignin (AOAC

973.18) analysis was performed using standard methods on homogenized samples.

Potential enzyme activities are expressed in moles IMUF released g-1 AFDM h-1

for glucosidase (BGL) and phosphatase (PHO), moles AMC released g-1 AFDM h-1 for

leucine aminopeptidase (LEU), and moles DICQ released g-1 AFDM h-1 for phenol

oxidase (PHE) and peroxidase (PER).

Cotton Strips

Cotton strips (Shirley Institute, Manchester, England) were prepared by cutting 12

cm wide strips and securing them in duplicate to stainless steel wire frames (6mm thick)

with staples and waterproof tape (Newman et al., 2001). The assemblies were deployed

within 10 meters of the soil coring sites at a depth of at least 15 cm below the soil surface

for a period of 2 weeks at all sampling sites, corresponding within a week to soil coring

events. Retrieval involved measuring the distance from the top of the frame to the soil

layer using visual methods to determine the surface of the soil. The depth of the benthic

layer was also recorded. Control strips on wire frames were quickly deployed in

duplicate and immediately retrieved. The retrieved strips were then washed in ambient









water to remove excess detrital matter. The cotton strips were removed from the wire

frame and cut into 2 cm wide strips 10 cm below the soil surface mark. 2 cm wide strips

were also cut in the corresponding benthic layer. The strips were washed in DI H20 and

allowed to soak for 10 minutes to ensure saturation prior to stretching. A tensiometer

(Chatillon TCD-200) with a digital force gauge (DFIS 200, Chatillon, Greensboro, North

Carolina, USA) was used to determine the remaining tensile strength of the cotton strips.

This value was subtracted from the control strip reading from the corresponding area.

The loss of tensile strength over time is expressed as a rate constant. Since the loss of

tensile strength is analogous to first-order decay, the calculation of the rate constant

requires linearization of the curve. Cotton Rottenness Rate (CRR) was calculated as (Hill

et al., 1985): CRR= { [((yo-y)/y)1/3)] / # of days deployed}* 365, where yo is the tensile

strength of the control strips and y is the tensile strength of the test strips at each 2 cm

increment.

Models

Extracellular enzymes were grouped into four categories: Ecell (BGL), En (LEU),

Ep (PHO), and Eox (PHE and PER). This grouping allows the enzymes to be grouped

into those involved in C, N, and P mineralization as well as lignin degradation,

respectively. Enzyme activities are normalized on a scale of 0-1 to eliminate the

weighting effects of the more active enzymes. Enzyme ratios were formulated to reflect

the premise of resource allocation and were based on assumptions derived from the

MARCIE (Microbial Allocation of Resources Among Community Indicator Enzymes)

model (Sinsabaugh and Moorhead, 1994, 1996; Sinsabaugh and Findlay, 1995;

Sinsabaugh et al., 1997, 2002). The model is based on the premise that the enzyme

mediated decomposition of complex molecules is the rate limiting step in C









mineralization, indicating that the expression of enzymes are tied to environmental

nutrient availabilities and that the distribution of enzyme activities can be interpreted as a

resource allocation strategy (Sinsabaugh et al., 2002). MARCIE model components,

such as Ecell/Ep and Ecell/En, which reflect apparent P and N control on C

mineralization, respectively, have been correlated with bacterial productivity (Sinsabaugh

and Findlay, 1995; Sinsabaugh et al., 1997). These ratios are based on the underlying

concept that lignocellulose degradation by extracellular enzymes is tied to environmental

N and P concentrations. Since only a certain amount of energy is available for enzyme

production, an abundant expression of one enzyme resulting from the lack of directly

utilizable substrate will result in less energy available for the production of other

enzymes. Ecell/En is the ratio calculated by: BGL/LEU, which reflects apparent N

control over potential cellulose decomposition. Ecell/Ep are relative measures indicating

P control over potential cellulose decomposition and is calculated by: BGL/PHO.

Ecell/Eox reflects apparent lignin control over potential cellulose decomposition and is

calculated by: BGL/(average(PER+PHE)).

The Enzyme Index of Carbon Quality (EICQ) is a relative index of the normalized

activities of the hydrolytic enzymes to the oxidative or lignin degrading enzymes. EICQ

has been correlated with microbial biomass (r=0.71), productivity (r=0.80), and

negatively correlated with particulate organic carbon (POC) turnover time (r=-0.99)

(Sinsabaugh and Findlay, 1995).

Statistics

Data was statistically analyzed with SAS v.8 statistical software (SAS, 1999)

using mixed model repeated measures to determine significant differences (p<0.05)

between sampling periods, soil depths, hydrologic units, and sites. The benthic and soil









layers were analyzed independently. Due to significant AREA*SITE*SAMPLING

PERIOD interactions, contrast statements were used to differentiate among specific sites.

Data was log-transformed to improve normality and heteroscedescity. Regressions were

performed using SYSTAT 10.2 (SYSTAT, 2002) for individual time periods. Mean

values of all sampling dates were combined for tables and charts. Significant differences

and correlation coefficients are significant at the p<0.05 level, unless otherwise noted.

Results

Mean values were calculated from the three sampling periods in order to analyze

the effects of the canal inflows in LNWR, WCA-2A, WCA-3A, and ENP-TS. The

results of mixed model repeated measures analysis on differences between enriched and

reference sites are presented in Tables 4-1 and 4-2 for the benthic data and Tables 4-4 and

4-5 for the soil data. Due to the complexity of responses at the transitional sites, these

results are presented separately.

Benthic Data

Nutrient data (Tables 4-1 & 4-2) for the benthic layers reflected significantly higher

total phosphorus (TP) at the enriched sites in all four areas. The greatest P concentrations

were present in the enriched LNWR and WCA-3A sites. TP concentrations were within

the ranges reported in other studies (Koch and Reddy, 1992; Reddy et al., 1993; DeBusk

et al., 1994; Wright et al., 2001b). Among the reference sites, the lowest average TP

concentrations were within ENP-TS at 0.1 g kg-1. These low values were also reported in

an earlier study (Wright et al., 2001b).

Generally, there were not consistently significant differences between the enriched

and reference TN and TOC values in the four areas (Tables 4-1 & 4-2). TN did vary

significantly along the gradient in LNWR. TN and TOC concentrations were similar to









those reported in a previous study (Wright et al., 2001b). LNWR, WCA-2A, and WCA-

3A exhibited a similar range of TN and TOC values. ENP-TS TN concentrations were as

much as 64% lower at the ENR site and 70% lower at the REF site. Correspondingly,

ENP-TS TOC concentrations were as much as 61% and 64% lower at the ENR and REF

sites, respectively. TN was most strongly correlated with TOC and lignin and negatively

correlated with calcium (Table 4-3). TOC was most strongly correlated with TP, TN,

lignin, and negatively correlated with calcium (Table 4-3).

Calcium concentrations were the highest in ENP-TS, with values as great as 400 to

1800% higher in the enriched and reference sites, respectively. The only significant

change along the gradient occurred in LNWR with a higher concentration at the enriched

site. Increases in Ca have been shown to increase with P loading due to release from

organic matter (Newman et al., 2001). Mg concentrations were between 150% and

338% greater in LNWR and WCA-2A than the other two areas sampled and did not vary

significantly along the gradient.

Benthic C:P values were significantly higher at the reference sites in all four study

areas. The largest difference in C:P values within an area occurred in ENP-TS with

values of 236 and 1803 at the ENR and REF sites, respectively. Benthic C:N ratios

ranged from 11.5 to 14.4 and were significantly lower at the reference sites (6 of 12

comparisons), with the exception of ENP-TS. The largest range of N:P values were 18.2

to 123.6 in ENP-TS, reflecting the minimum and maximum values among the areas. N:P

values increased significantly with increasing distance from the canal in all areas.












Table 4-1.


Benthic nutrient, enzyme, and ratio parameters in P enriched and reference sites in LNWR and WCA-2A. Values
presented are means averaged over the three sampling periods with standard errors. Significant differences are given
between the enriched and background sites in the respective areas. Each asterisk reflects differences present at the p<0.05
level per sampling period. Analysis was divided into the three sampling periods (1, 2, 3). Glucosidase, Leucine
aminopeptidase, phosphatase units are moles AMC or MUF released g-1 AFDM h-1. Phenol oxidase and peroxidase
units are moles DICQ released g-1 AFDM h-1. C:P= Total organic carbon to total phosphorus, C:N=Total organic


carbon to total nitrogen, N:P=Total nitrogen to total phosphorus

LNWR WCA-2A

Parameter Units Enriched Reference (P<0.05) Enriched Reference (P<0.05)
(n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3


Total Phosphorus g kg-1
Total Nitrogen g kg-1
Total Organic Carbon g kg-1
Calcium g kg-1
Magnesium g kg-1
C:P
C:N
N:P
Lignin %
Cellulose %
Glucosidase
Leucine aminopeptidase
Phosphatase
Phenol oxidase
Peroxidase
Ecell/Ep
Ecell/En
Ecell/Eox


1.65 0.41
34.99 3.59
440.44 12.37
44.00 10.50
3.96 0.28
312.6 62.1
12.8 1.0
23.6 3.2
40.5 1.8
13.5 0.9
0.27 0.01
3.49 0.17
2.77 0.29
59.45 7.93
21.94 4.45
31.44 14.11
0.57 0.14
4.51 2.47


0.59 + 0.02
40.14 1.40
467.33 20.34
13.78 0.78
2.37 0.11
796.2 9.3
11.7 0.2
68.2 1.0
32.4 1.3
12.8 0.0
0.27 0.03
4.92 0.40
16.21 0.39
77.46 5.02
15.41 8.30
0.97 0.24
0.37 0.06
4.30 2.51


* 1.36 0.17
* 30.99 1.38
447.00 26.53
41.72 10.72
3.77 0.49
347.7 70.6
14.4 0.3
24.0 4.4
42.5 1.2
17.4 1.6
0.23 0.01
3.00 + 0.17
* 3.18 0.64
25.05 2.48
6.83 1.10
* 38.98 20.59
0.63 0.15
7.14 2.28


0.43 0.04
31.67 0.61
410.94 22.86
42.39 10.52
3.68 0.12
978.3 140.2
13.0 1.0
74.3 5.4
33.4 7.5
18.1 3.9
0.11 0.01
1.51 0.08
11.20 2.56
18.54 1.21
4.43 2.25
0.39 7.47
0.20 + 0.10
17.58 2.09













Table 4-2.


Benthic nutrient, enzyme, and ratio parameters in P enriched and reference sites in WCA-3A and ENP-TS. Values
presented are means averaged over the three sampling periods with standard errors. Significant differences are given
between the enriched and background sites in the respective areas. Each asterisk reflects differences present at the
p<0.05 level per sampling period. Analysis was divided into the three sampling periods (1, 2, 3). Glucosidase, Leucine
aminopeptidase and Phosphatase units are moles AMC-MUF released g-1 AFDM h-1. Phenol oxidase and Peroxidase
units are moles DICQ released g-1 AFDM h-1.


WCA-3A ENP-TS

Parameter Impacted Reference (P<0.05) Impacted Reference (P<0.05)
(n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3


Total phosphorus
Total nitrogen
Total organic carbon
Calcium
Magnesium
C:P
C:N
N:P
Lignin
Cellulose
Glucosidase
Leucine aminopeptidase
Phosphatase
Phenol oxidase
Peroxidase
Ecell/Ep
Ecell/En
Ecell/Eox


g kg-1
g kg-1
g kg-1
g kg-1
g kg-1



%


1.66 0.25
33.16 1.12
438.11 9.99
64.72 39.35
1.17 0.10
283.7 48.8
13.2 0.3
21.6 4.3
21.6 4.0
15.4 4.7
1.07 0.29
4.39 0.65
2.32 0.56
186.6 46.0
10.8 4.3
53.20 22.73
1.33 0.14
4.22 1.01


0.48 0.05
37.53 1.95
427.67 16.88
25.78 6.34
1.13 0.16
840.8 5.4
11.5 0.3
75.2 0.4
26.5 2.7
11.5 0.4
0.26 0.02
3.23 0.10
18.09 3.23
105.2 4.1
17.3 4.1
1.51 0.61
0.57 0.09
1.44 0.46


* 0.74 0.20
13.2 4.77
173.63 46.82
166.44 58.77
1.43 0.05
* 236.3 21.6
* 13.1 + 1.5
* 18.2 1.1
20.1 + 11.1
20.9 4.8
0.67 0.01
8.13 0.79
* 11.19 3.18
287.3 89.1
10.8 2.2
* 3.47 0.80
1.59 0.41
2.22 0.43


0.10 + 0.01
11.98 0.90
168.00 + 8.19
191.11 28.04
2.44 0.16
1803.2 137.9
14.4 0.1
123.6 8.3
7.2 4.5
8.4 5.0
0.15 0.01
3.07 0.05
61.33 0.48
94.9 2.3
10.4 2.2
0.11 0.02
0.44 0.09
0.96 0.20


*1 *









Lignin and cellulose data is based upon data from the 2nd and 3rd sampling

periods. Lignin values ranged from 7% to 43%. Decreases in lignin content occurred in

LNWR, WCA-2A, and ENP-TS with increasing distance from the canal. The lowest

lignin value was associated with the REF ENP-TS site as was the lowest cellulose

content. Lignin was correlated with TP, TN, TOC, and negatively correlated with Ca.

Cellulose content ranged from 8% to 21% with mixed relationships along the gradient.

Various hydrolytic enzyme activities were associated with the P gradients in each area.

Potential benthic BGL activities ranged from 0.11 to 1.07 (Table 4-1 & Table 4-2), and

were higher in the ENR sites. However, these differences were only significant in 3 of 12

cases. There was no change in BGL along a gradient observed in an earlier study in

WCA-2A (Wright and Reddy, 2001b). WCA-2A generally exhibited the lowest BGL

activities. BGL was correlated with TP, LEU, and PHE. Similarly, LEU significantly

decreased between the ENR and REF sites in only 1/3 of the cases, with the exception of

LNWR, and ranged from 1.51 to 8.13 moless g-1 AFDM h-1. The lowest LEU values

were also generally found in WCA-2A and were only significantly correlated with PHE.

PHO was significantly higher in the reference sites in 75% of the comparisons and ranged

from 2.32 to 61.33 moles MUF released g-1 AFDM h-1. The decrease in PHO with

increased proximity to the canal has been observed in other studies in the Everglades

(Wright and Reddy, 2001b) as well as with decreased phosphatase activities in

periphyton associated P loading. Significant correlations were found with TP, TN, TOC,

and lignin. The highest activities were associated with the ENP-TS sites, reflecting the

lowest benthic TP values.











Table 4-3. Benthic layer nutrient and enzyme correlation coefficients. All values are
significant at the p<0.05 level. TN=total nitrogen, TP=total phosphorus,
TOC=total organic carbon, C:P=ratio of TOC to TP, Ca=calcium,
Mg=magnesium, GLU=glucosidase, Ecell/En= apparent N control on carbon
mineralization, Ecell/Ep=apparent P control on carbon mineralization LEU=
leucine aminopeptidase, PHO=phosphatase, PHE=phenol oxidase,
CRR=cotton rottenness rate.


Ecell/ Ecell/
TP TN TOC Lig Cell Ca Mg C:P En En BGL LEU PHO
TN .69
TOC .68 .98
Lignin .56 .66 .69
Cellulose .40 .41
Ca -.45 -.80 -.79 -.81
Mg
C:P -.78 .41
Ecell/En -.36 -.36
Ecell/Ep .76 .41 .44 .39 -.63 .46
GLU .39 -.54 .57 .42
LEU .60
PHO -.66 -.48 -.48 -.50 .10 .44 -.91
PHE -.23 .37 .51
CRR .70 .65 -.69 .62 -.57


Benthic phenol oxidase activities decreased between the ENR and REF sites in all

areas but LNWR. However, this decrease was only significant in 25% of the

comparisons. A lack of significant changes in PHE was previously documented within

WCA-2A (Wright and Reddy, 2001b). The lowest PHE activities were present in WCA-

2A with the largest range of activities in ENP-TS. Benthic PHE was correlated with

BGL and PHO. Peroxidase activities were not significantly different between the ENR

and REF sites and were not correlated with other enzymes or nutrients.

Average benthic Ecell/Ep values decreased with distance from the canal inflow

(Tables 4-1 & 4-2) and were significantly lower in the reference sites in 11 of 12 cases.

Values were as much as much as 20 times higher than those reported in an aquatic study









(Sinsabaugh et al., 1997). The lowest Ecell/Ep values were within ENP-TS, reflecting

higher PHO activity in relation to BGL as well as the low TP values. Ecell/Ep was

correlated with the nutrient parameters TP, TN, TOC, and lignin. The correlations with

the enzymes inclusive in the ratio formulation, BGL and PHO, indicate the greater

variability of phosphorus in this study.

Benthic Ecell/En significantly increased with P loading in all four areas in 5 of 12

comparisons. The largest range occurred within ENP-TS. There were no correlations to

components that were not related to the ratio. The most significant changes in benthic

Ecell/Eox values occurred within ENP-TS, with higher values at the enriched site.

Conversely, Ecell/Eox was significantly higher in one time period at the reference site in

WCA-2A. Ecell/Eox was not significantly correlated with any other parameters.

Enzyme Index of Carbon Quality (EICQ) values generally did not predict increased C

quality at the enriched sites, with the exception of WCA-2A and WCA-3A, though these

differences were not significant.

Soil Data

Average 0 to -10 cm soil TP values ranged from 0.2 to 1.0 g kg-1 with

significantly lower concentrations in 7 of 12 cases associated with the REF sites (Tables

4-4 & 4-5). The lowest TP values were associated with the ENP-TS sites. Soil TP was

correlated with TN, TOC, and lignin, and was generally lower than the benthic layer

(Table 4-6). The lowest TP and TN values were both associated with the ENP-TS sites.

TN mean concentrations ranged from 8 to 37 g kg-1 and were significantly different in 5

of 12 cases with 2 of 3 instances reported in ENP-TS. TN was strongly correlated with

TOC, lignin, and negatively correlated with Ca. There was no significant change in TN

content with depth. TOC ranged from 86 to 473 g kg-1 with significantly lower values









once again associated with the ENP-TS sites in 1/3 of the comparisons. Soil TOC was

correlated with lignin and negatively correlated with Ca. There was no consistent

relationship between TOC content and depth.

Calcium concentrations ranged between 13 and 220 g kg-'. The ENP-TS Ca

concentrations were as much as 17 times greater than other areas. In relation to the

gradient, Mg and Ca were not significantly lower at the reference sites. Mg values

ranged from 0.9 to 3.6 g kg-1. Ca was also negatively correlated with lignin. Neither Ca

nor Mg demonstrated consistent relationships with depth.

Soil mass-mass C:P ratios were significantly higher (6 of 6 cases) at the WCA-3A

and ENP-TS reference sites, reflecting the influence of the P gradient. C:P was weakly

correlated with both TN and lignin, and was higher in the soil layer, with the exception of

ENP-TS. Conversely, soil C:N values were significantly lower at 50% of the reference

sites and ranged from 11 to 16. C:N ratios did not vary significantly within ENP-TS

along the gradient. Like C:P, the soil C:N values were the lowest at the ENP-TS sites

and exhibited higher values in the soil layer, with the exception of ENP-TS. N:P values

were significantly higher at the reference sites in 11 of 12 comparisons and were the

lowest in the ENP-TS sites. Additionally, soil N:P values were greater than the benthic

layers, with the exception of the ENP-TS enriched site. These increases with soil depth

are analogous to increases in N:P observed with increasing mass loss (Sinsabaugh et al.,

1993). C:P and N:P values were comparable to soil data gathered in WCA-2A while C:N

values were on the lower range (Reddy et al., 1993). Lignin data ranged from 11% to

52% in the soil layer and generally increased with depth. Cellulose content also









increased with depth, ranging from 13% to 27% and was generally higher at the reference

sites, with the exception of ENP-TS.

Compared to the benthic layer, hydrolytic enzyme activities in the soil layer had

fewer significant differences along the P gradient (Tables 4-4 & 4-5). Average soil layer

BGL activity was consistently lower than the benthic activity. Soil BGL ranged from

0.05 to 0.39 .moles g-1 AFDM h-1 and decreased from the ENR to REF site in all areas

with significant differences in 2 of 3 sampling periods in WCA-3A and ENP-TS. The

largest contrast between the ENR and REF sites occurred in ENP-TS. Soil BGL was

negatively correlated with Mg.

Soil leucine aminopeptidase (LEU) activity was lower in 10 of 12 sites in

comparison to the benthic layer and ranged from 1.92 to 3.58 [moles g-1 AFDM h-1.

LEU was generally not significantly different between the enriched and reference sites.

LEU was negatively correlated with TN, TOC, and lignin. There was not a consistent

relationship between LEU activity and the P gradient. Soil PHO also did not

significantly vary between the REF and ENR sites. Values were consistently lower in the

soil layer, reflecting activities as great as 10 times lower those in the corresponding

benthic layer. This depth relationship has been previously documented in other areas

(Newman and Reddy, 1993; Wright and Reddy, 2001b). Soil PHO was negatively

correlated with TP, TN, TOC, lignin, and positively correlated with Ca (Table 4-6). Soil

PHE and PER activities increased with depth. PHE was weakly negatively correlated

with TN, TOC, and lignin, while PER was not correlated with any parameters. The

increase of oxidative activity with depth is in disagreement with earlier studies within the












Table 4-4.


Soil layer nutrient, enzyme, and ratio parameters in P enriched and reference sites in LNWR and WCA-2A. Values
presented are means averaged over the three sampling periods with standard errors. Significant differences are given
between the enriched and background sites in the respective areas. Each asterisk reflects differences present at the p<0.05
level. Analysis was divided into the three sampling periods (1, 2, 3). Glucosidase, Leucine aminopeptidase, phosphatase
units are moles AMC or MUF released g-1 AFDM h-1. Phenol oxidase and peroxidase units are moles DICQ released
g-1 AFDM h-1. C:P= Total organic carbon to total phosphorus, C:N=Total organic carbon to total nitrogen, N:P=Total


nitrogen to total phosphorus.

LNWR WCA-2A

Parameter Units Impacted Reference (P<0.05) Impacted Reference (P<0.05)
(n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3


Total phosphorus
Total nitrogen
Total organic carbon
Calcium
Magnesium
C:P
C:N
N:P
Lignin
Cellulose
Glucosidase
Leucine aminopeptidase
Phosphatase
Phenol oxidase
Peroxidase
Ecell/Ep
Ecell/En
Ecell/Eox


g kg-1
g kg-1
g kg-1
g kg-1
g kg1-




%


0.91 0.14
31.98 1.74
471.11 3.74
30.33 3.10
3.41 0.28
711.1 + 122.3
14.9 0.9
47.1 5.8
52.1 3.1
17.3 0.2
0.24 0.06
1.92 0.13
1.89 0.34
110.5 8.33
24.3 0.8
18.47 8.40
0.60 + 0.20
0.88 0.32


0.37 0.09
35.58 3.15
462.11 28.15 *
12.61 0.20
1.96 0.08
1271.9 58.7
13.2 0.6 *
97.7 8.1
45.7 2.1
18.8 0.1
0.14 0.06
2.69 0.05
2.46 0.07
223.0 45.9
36.1 5.3
8.52 3.25
0.32 0.15
1.44 1.11


* 0.72 0.13
29.12 2.77
443.00 17.14
34.83 6.77
3.64 0.09
* 743.9 200.3
16.4 1.3
* 46.0 10.3
50.1 2.8
18.2 1.7
* 0.08 0.00
1.95 0.03
1.35 0.05
78.6 4.0
44.5 9.1
13.08 5.47
* 0.27 0.04
0.24 0.06


0.33 0.04
33.61 + 1.21
464.67 8.17
28.6 2.69
3.22 0.21
1478.4 174.7
13.9 0.7
105.7 7.6
46.5 3.6
19.7 0.7
0.05 0.01
2.16 0.07
1.14 0.06
142.8 23.4
41.1 2.2
0.84 0.99
0.10 + 0.03
2.29 0.06














Table 4-5. Soil nutrient, enzyme, and ratio parameters in P enriched and reference sites in WCA-3A and ENP-TS. Values
presented are means averaged over the three sampling periods with standard errors. Significant differences are given
between the enriched and background sites in the respective areas. Each asterisk reflects differences present at the
p<0.05 level per sampling period. Analysis was divided into the three sampling periods (1, 2, 3). Glucosidase, Leucine
aminopeptidase and Phosphatase units are moles AMC-MUF released g1 AFDM h-1. Phenol oxidase and Peroxidase
units are moles DICQ released g-1 AFDM h-1.

WCA-3A ENP-TS

Parameter Impacted Reference (P<0.05) Impacted Reference (P<0.05)
(n=9) (n=9) 1 2 3 (n=9) (n=9) 1 2 3


0.95 0.42
25.43 9.13
343.13 129.19
87.86 72.95
1.01 0.20
489.3 142.7
13.1 0.6
38.7 13.2
20.5 11.5
18.0 1.6
0.39 0.02
2.87 0.08
0.80 + 0.14
253.0 32.0
19.2 3.6
90.19 39.15
1.28 0.27
4.97 3.19


0.31 0.04
37.21 1.09
473.39 5.00
19.06 + 0.86
0.88 0.03
1601.9 220.8
12.5 0.3
127.4 14.7
45.1 + 1.8
18.7 1.6
0.21 0.02
2.31 0.02
2.18 0.23
222.1 34.0
32.6 13.6
9.37 3.13
0.57 0.11
1.29 0.47


* 0.37 0.11
7.65 1.63
86.36 25.78
183.44 78.51
1.56 0.33
* 239.3 6.4
* 10.9 1.1
* 20.3 0.4
19.0 12.2
27.4 8.4
* 0.28 0.07
3.48 0.04
3.87 0.54
577.9 76.0
62.0 14.0
* 2.74 0.71
0.63 0.23
1.86 1.20


0.23 0.08
12.74 2.05
153.48 24.17
219.89 30.89
2.92 0.61
775.9 60.0
12.0 + 0.1
62.5 6.7
10.9 1.8
13.3 0.1
0.06 0.01
2.72 0.17
10.84 0.20
252.4 18.0
24.9 13.7
0.32 0.10
0.21 0.07
0.34 0.16


g kg-1
g kg-1
g kg-1
g kg-1
g kg-1


TP
TN
TOC
Ca
Mg
C:P
C:N
N:P
Lignin
Cellulose
BGL
LEU
PHO
PHE
PER
Ecell/Ep
Ecell/En
Ecell/Eox









Everglades (McLatchey and Reddy, 1998). Soil PHE was highest in ENP-TS. Neither

enzyme was observed to have a significant response across the gradient.

Soil Ecell/Ep values were lower in the reference sites in each area, however

statistical significance was generally limited to WCA-3A and ENP-TS. The lowest

Ecell/Ep values occurred in ENP-TS. Soil Ecell/Ep was higher than the benthic layer at

the reference sites and lower at the enriched sites, with the exception of WCA-3A.

Ecell/Ep was negatively correlated with PHE. As in the benthic layer, PHO was more

linearly related to the Ecell/Ep ratio (r2 =-0.65) than BGL (r2 =0.32) (Table 4-6). Soil

Ecell/En was also lower in the reference sites, although significant differences were

confined to ENP-TS. The highest Ecell/En and Ecell/Eox values occurred within WCA-

3A. Ecell/Eox was not significantly different along the gradient. Additionally, Ecell/Eox

values were generally lower in the soil layer, possibly reflecting lower litter quality based

on perceived lignin content. EICQ values predicted greater carbon quality at the enriched

sites in WCA-2A, WCA-3A, and ENP-TS, although these differences were not

significant.

CRR

Cotton Rottenness Rate (CRR) was averaged over the three sampling dates (Table

4-7). Benthic and soil CRR decreased with increasing distance from the canals in all

areas. The greatest CRR in both the benthic and soil layers occurred at the WCA-2A

ENR site. However, the most dynamic change in CRR was associated with the ENP-TS

benthic layers between the enriched and reference sites. There were not consistent

differences between the benthic and soil layers and values were within the ranges

reported in the mesocosm P loading experiment in LNWR (Newman et al., 2001).










Table 4-6. Soil layer nutrient and enzyme correlation coefficients. Values in bold
are significant at the p<0.0001 level, all other values are significant at the
p<0.05 level. Lig=lignin, Cell=cellulose. TN=total nitrogen, TP=total
phosphorus, TOC=total organic carbon, C:P=ratio of TOC to TP, Ca=calcium,
Mg=magnesium, GLU=glucosidase, Ecell/En= apparent N control on carbon
mineralization, Ecell/Ep=apparent P control on carbon mineralization LEU=
leucine aminopeptidase, PHO=phosphatase, PHE=phenol oxidase,
CRR=cotton rottenness rate.


Ecell/ Ecell/
TP TN TOC Lig Ca Mg En Ep PHO
TP
TN .63
TOC .63 .99
Lig .48 .76 .79
Cell
Ca -.59 -.62 -.78
Mg
C:P .48 .51 .41
EC/EN .37 -.49
EC/EP .53 .61 .37 .62
BGL -.44 .93 .57
LEU -.69 -.54 -.39
PHO -.49 -.61 -.62 -.51 .57 -.81
PHE -.37 -.40 -.39 -.48 .57
CRR


A model was created using enzyme data to test the validity of prediction against

actual CRR values. Ecell/Ep and TP were found to be the most powerful predictors of

CRR, accounting for 62% of the variability in the benthic layer. The resulting benthic

model: Predicted CRR= (0.04(LnEc:Ep)+1.4) + (0.35(LnTP)+1), was strongly correlated in

sampling period 1 (r2 =0.92) and more weakly correlated in sampling period 3 (r2=0.46).

This benthic model predicted significantly greater cellulose decomposition at the

enriched sites in all four areas at all time periods (p<0.0001). A separate significant

model could not be created for the soil layer. However, the benthic model predicted

significantly greater decomposition in the enriched sites in 75% of the contrasts.









Impact Index

An impact index was constructed to compare the varying parameters on the same

scale and to allow the comparisons between the enriched and reference sites within each

area (Table 4-8). In general, impact index values above 0.5 reflect severe changes in a

specific parameter between the enriched and reference sites. Positive values reflect

increased values at the enriched site while negative values reflect higher values at the

reference site. The equation for each parameter is:

Impact Index = log ((Enriched P site) / (Reference P site))

As expected, the most responsive nutrient parameters to the P gradient were the P-

related biogeochemical indicators TP, C:P, N:P, PHO, and Ecell/Ep in both the benthic

and soil layers. TOC showed relatively little change in relation to the P gradient while

TN demonstrated a small response in the soil layer. Differences were also observed with

BGL in both layers. The majority of large impact indices in the benthic layer were within

ENP-TS, even though there is a less pronounced P gradient in this area. WCA-2A, a

heavily impacted area, did not exhibit the large differences that were expected. There

were greater index values associated with the benthic layer, indicating that this layer is

more responsive to changes in biogeochemical conditions.

Transitional Sites

Large increases in several of the individual enzyme activities occurred at the

transitional sites in the four hydrologic units (Figure 4-la & 4-1b). The greatest increases

in benthic activity occurred at the WCA-2A TRANS site with BGL and LEU with 10 of

12 comparisons significant. Spikes in LEU were also observed within the transitional

sites in LNWR, WCA-2A, and ENP-TS. PHO also increased in the soil layer WCA-2A,









Table 4-7. Average benthic and soil layer CRR values over the three sampling periods
with standard errors. Soil layer CRR values represent 2 of 3 sampling periods
due to insufficient depth deployment of the cotton strip racks. ENR=Enriched,
Trans=Transitional, REF=reference sites.


LNWR ENR
LNWR TRANS
LNWR REF

WCA-2A ENR
WCA-2A TRANS
WCA-2A REF

WCA-3A ENR
WCA-3A TRANS
WCA-3A REF

ENP-TS ENR
ENP-TS TRANS
ENP-TS REF


Benthic layer
23.2 0.5
22.9 0.8
20.3 0.8

25.7 0.1
20.7 0.7
18.5 1.3

24.7 0.1
23.5 0.6
17.2 1.1

24.6 0.5
17.8 0.8
16.3 1.3


Soil layer
24.7 0.5
21.0 0.7
19.4 0.3

25.4 0.2
18.1 1.0
18.7 1.1

25.1 0.1
22.6 0.4
20.7 1.0

22.2 0.2
20.9 0.6
19.6 0.6










a.) Benthic layer


lenr Itrans 1ref 2enr 2trans 2ref 3enr 3trans 3ref 4enr 4trans 4ref


b.) Soil layer


lenr Itrans 1ref 2enr 2trans 2ref 3enr 3trans 3ref 4enr 4trans 4ref


Figure 4-1. Graphs showing spikes in enzyme activities at the transitional sites. Mean
enzyme activities have been normalized for visual comparison. Ec and En,
normalized glucosidase and leucine aminopeptidase activities exhibited the
largest spikes in the benthic layer while En and Ep phosphatasee)
exhibited the largest in the soil layer. Bars represent standard errors.
1=LNWR, 2=WCA-2A, 3=WCA-3A, 4=ENP-TS. Enr=enriched, trans=
transitional, ref=reference sites.















Table 4-8. Impact index (log(impacted/reference)) for nutrient and enzymic data in the soil and benthic layers for LNWR, WCA-2A,
WCA-3A, and ENP-TS for averages calculated over the three sampling periods. Bolded items reflect parameters that are
equal to or greater than 0.5, indicating large changes over the gradient. Negative values indicate larger parameters at the
reference sites while positive values reflect larger values at the enriched sites.

Benthic Soil

LNWR WCA-2A WCA-3A ENP-TS LNWR WCA-2A WCA-3A ENP-TS
Parameter

TP 0.42 0.49 0.53 0.83 0.38 0.34 0.29 0.22
TN -0.06 -0.01 -0.05 0.04 -0.04 -0.07 -0.38 -0.23
TOC -0.03 0.04 0.01 -0.02 0.01 -0.02 -0.25 -0.28
Ca 0.48 -0.01 0.27 -0.14 0.38 0.07 0.22 -0.24
Mg 0.22 0.00 0.02 -0.23 0.24 0.06 0.05 -0.28
C:P -0.42 -0.46 -0.47 -0.88 -0.27 -0.36 -0.54 -0.49
C:N 0.04 0.05 0.06 -0.05 0.05 0.05 0.02 0.00
N:P -0.5 -0.5 -0.53 -0.83 -0.32 -0.40 -0.56 -0.49
Lignin 0.10 0.12 -0.09 0.47 0.06 0.03 -0.47 0.07
Cellulose 0.02 -0.01 0.10 0.48 -0.04 -0.04 -0.02 0.29
BGL -0.04 0.45 0.56 0.66 0.15 0.27 0.26 0.56
LAP -0.19 0.27 0.13 0.24 -0.16 -0.06 0.08 0.13
PHO -1.06 -0.49 -0.95 -0.81 -0.03 -0.16 -0.44 -0.44
PHE -0.29 -0.04 0.21 0.51 0.02 0.20 0.10 0.20
PER 0.21 0.03 0.13 -0.03 -0.13 -0.01 -0.11 0.39
Ecell:Ep 1.03 1.04 1.54 1.42 0.19 0.42 0.85 0.96
Ecell:En 0.13 0.24 0.41 0.44 0.28 0.32 0.29 0.47
Ecell:Eox -0.09 0.28 0.50 0.40 0.28 0.19 0.38 0.48









WCA-3A, and ENP-TS sites. However, these spikes were not correlated with increased

CRR.

The significant benthic BGL and LEU increase at the WCA-2A transitional site is

accompanied by a generally much lower TN (31 g kg-1), lignin (21%), cellulose (9%), the

lowest TOC (352 g kg-1) of any site not in ENP-TS, as well as elevated Ca (81 g kg-1) and

Mg (81 mg kg-1 ). This site is also dominated by Chara spp., which is a characteristic

algae in transitional P sites and is a level II indicator of P enrichment (U.S. EPA, 2002).

The spike in the benthic LNWR transitional site is also accompanied by a lower TN

(31.23 g kg-1) and TOC (438.1 g kg-1).

Increases also occurred in the soil LEU and PHO at the transitional sites in WCA-

2A and ENP-TS. The lowest TP (0.2 g kg-1) and TN (6 g kg-1) of any site was recorded

at the transitional site in ENP-TS which accounts for the greatest PHO and LEU activity

of any soil layer. This relationship in the soil layer is not as clear at the WCA-2A

transitional site.

Discussion

The following discussion has been grouped into several sections that first

compare the general effects of the P gradient between the enriched and reference sites.

The discussion generally pertains to the benthic layer results, as this layer is the most

biologically active (White and Reddy, 2000) and is more responsive to changes in

environmental conditions. Secondly, specific hydrologic units are compared in relation

to their response to the P gradient. Lastly, a brief discussion of the observations made at

the transitional sites is included.









Phosphorus Gradients

The nutrient and enzyme parameters of the benthic layer varied in differing degrees

along the nutrient gradient. The correlations between TOC-TP and TOC-TN reflect

increased vegetative input associated with increased TP values in both the benthic and

soil layers. Aboveground production in cattail stands ranges from 3035 to 1077 g m2 y1

in the enriched and reference areas in WCA-2A, respectively, while sawgrass ranges

from 1943 to 986 g m2 y1 (Davis, 1991). These changes in productivity and thus nutrient

input from vegetative sources are reflected in soil nutrient contents. Decreases in P with

increasing distance from the canals are reflected in the larger C:P values at the reference

sites in both layers. Increases in C:P ratios in WCA-2A have been correlated with

decreased C turnover rates (Koch and Reddy, 1992), while lower C:P ratios have been

correlated with enhanced microbial respiration (Amador and Jones, 1993).

While individual enzyme activities can provide some insight into microbial

nutrient perceptions, relative indices may provide a more productive look at community

responses to environmental change. Ecell/En, Ecell/Ep, and Ecell/Eox relationships in

the different areas are presented in Figure 4-2. The higher Ecell/Ep values at the enriched

sites reflect a decrease in apparent P control on C mineralization. The greatest shift

within WCA-3A suggests that this area has the largest P impact on the microbial

community. The increase in Ecell/Ep that occurs in WCA-3A may be due to a the

weighting effect of a higher C mineralization rate. This may be due to decreased

inhibition from polyphenols, a generally lower water level during the sampling events, or

the lower lignin content at the WCA-3A enriched site. The lowest shift in Ecell/Ep

occurs in ENP-TS (Figure 4-2), suggesting that the microbial communities are less

affected across the gradient by changes in P conditions. The impact index does not


















0.

LU


83



70

60 -
Enr WCA-3A
50-

40 nr

30


Ref WCA-2A
10


Ref ENP-TS
10
0 02 04 06 08 1 12 14 16 18 2
Ecell/En


Figure 4-2. Benthic layer bubble plot of Ecell/Ep vs. Ecell/En with Ecell/Eox
represented by the bubble size. Arrows denote shifts from the enriched
to the corresponding reference sites.



support this by suggesting that the highest TP impact (0.83) occurs along the ENP-TS


gradient. However, the TP values are lower at both the enriched and reference sites, so

the increases may not be enough to induce resource allocation shifts along the gradient.


In this case it does not appear that the impact index accurately reflects the severity of


impacts in relation to significant changes that may be occurring within the communities.

The greatest shift in P influence on C mineralization therefore appears to occur within


WCA-3A with the smallest shift within ENP-TS.


Ecell/En values were higher at all the enriched sites, once again to differing


degrees. The largest shift occurred within ENP-TS (Figure 4-2), which also had the


highest Ecell/En impact index. The higher Ecell/En and Ecell/Ep values at all the


enriched sites are concurrent with decreases in nutrient limitations that have been


documented (McCormick et al., 1996), as well as lower P mineralization rates in enriched


areas of the Everglades (Newman et al., 2003; White and Reddy, 2000).









Lower Ecell/En values at the reference sites may be attributed to shifts from

primarily macrophyte to algal inputs at the reference sites (Figure 4-2). This would be

manifested in higher LEU activity, in relation to BGL, since algae generally contain

a higher protein content than macrophytes (Boschkler and Cappenberg, 1998). This is

especially important since LEU may also function in C mineralization from protein

sources. One driving force behind higher Ecell/En values at the enriched sites may be the

generally elevated BGL activities, which has been correlated with microbial production

in nutrient enriched mesocosms (Chr6st and Rai, 1993). Overall, the increases in

Ecell/En at the enriched sites mimic a southerly pattern: LNWR
3A
less important in the southern regions. Compared to the northern Everglades,

significantly lower C/N ratios of the DOC/DON fraction of surface and shallow pore

water were found within the southern Everglades (Qualls and Richardson, 2003), which

mirrors the decreased apparent N control on C mineralization (Ecell/En). Several

hypotheses concerning the southerly trend of decreasing DOC/DON ratios are presented

by Qualls and Richardson (2003) and include increased rates of biodegradation, greater

plant production of soluble organic matter, and greater production of soluble organic

matter by-products where peat is being decomposed faster.

Plots of log Ecell/En vs. log Ecell/Ep indicate higher perceived P, and to a lesser

extent, N in the enriched sites (Figure 4-3). Reference sites generally reside in the low N,

low P quadrant. The general lack of sites residing in low P, high N conditions indicates

that N loading may not be a substantial issue in the Everglades, which is consistent with

the view that the system is P, not N limited.









The increase in Ecell/Eox values at the enriched sites, with the notable exception of

WCA-2A, indicate decreased apparent lignin control on C mineralization. This would

indicate a greater C mineralization potential. However, since enzyme activities in the

view of a resource allocation model are interdependent, other factors may be influencing

these trends. The higher Ecell/En values at the enriched sites in the four areas are a

response to increased perceived N content. This is not apparent in the TN content of the

soils which suggests that this parameter is not well suited for judging N availability, since

there are varied forms of N within the system. A large portion of endogenous N may also

be shielded by refractory compounds within the litter (Sinsabaugh and Moorhead, 1994)

while exogenous N loading has been shown to repress lignin degradation (Eriksson et al.,

1990; Blanchette, 1991; Sinsabaugh et al., 1993; Hammel, 1997; Carreiro et al., 2000) by

the repression of phenol oxidase. Therefore a connection may be established between N

and lignin influence on C mineralization whereas individual enzyme activities do not

necessarily point to this conclusion. Therefore it appears that P loading has resulted in a

generally lower level of lignin influence on C mineralization, although a higher level of

microbially-perceived exogenous N may be repressing oxidative enzyme activity.

The combination of factors resulting from nutrient loading to the four areas of the

Everglades appears to have resulted in a decrease in N, P, and lignin influence on the

microbial community. This decrease of nutrient limitation on the community should

result in greater mineralization rates of plant matter, which is supported by consistently

higher CRR rates at the enriched sites as well as a greater microbial biomass in another

study in WCA-2A (Wright and Reddy, 2001 la). The CRR decomposition model, utilizing

both TP and Ecell/Ep as parameters, indicates the relatively strong influences that the P












2.5

2

1.5

1

0.5

0

-0.5

-1

-1 C5


High P, Low N High P, High N


A *


---------------------------- A-6 ----------------------------------





A A A
A



LA

Low P, Low N Low P, High N


-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6
Ecell/En



Figure 4-3. Benthic plot of Log Ecell/En vs. Log Ecell/Ep. Each point represents the
mean from one sampling period. The X axis represents apparent N influence
on C mineralization, the Y axis represent apparent P influence on C
mineralization. Values greater than 1 on the Y axis indicates high perceived
P availability.



gradient have on cellulose decomposition rates. The coupling of Ecell/Ep and TP in the


CRR model indicates the relative strength of the enzyme parameters BGL and PHO in


predicting cellulose decomposition. This is comparable to the correlation between BGL


and PHO with bacterial production in nutrient enriched and un-enriched mesocosms,


respectively (Chr6st and Rai, 1993). However, as the interactions of other components


within the system indicate, CRR may not be adequate to significantly predict actual litter


decomposition rates, especially in relation to differing litter types.


The lack of consistent predictive power in EICQ values suggests that the dynamic


changes in nutrient conditions relating to the canal inflows may influence the significance


of the model. The more complex interactions that may be affecting the communities such


* ENR
*REF
ATRANS









as microbial community shifts, vegetative changes, and differences in interactions

between the heterotrophic and algal communities may outweigh the singular dynamics of

the individual enzyme activity components of the EICQ model, especially in terms of

lignin degradation. Therefore it does not appear that the EICQ model is a powerful

predictor of microbial responses in this system among the different hydrologic units.

Additionally, the use of only 5 enzymes in this study may be restrictive on the validity of

the EICQ model.

ENP-TS

Special consideration is attributed to the ENP-TS sites due to the large shift in

apparent N dynamics across the gradient, as compared to the relatively large shifts in P

dynamics in the other areas. As previously discussed, ENP-TS had the greatest shift in

apparent N influence and the smallest shift in apparent P influence on C mineralization.

The highest Ecell/En and lowest Ecell/Ep values at the ENP-TS enriched sites suggests

that P is the most limiting of the enriched sites, which is expected, and that N is the least

limiting of any site, which is not expected. Although this site has a higher LEU activity,

the relationship with the amount of C mineralization that appears to be occurring

indicates that, energetically, this community is expending a smaller proportion of its

resources on the acquisition of N from organic sources.

The lowest overall Ecell/Eox at the ENP-TS enriched site points to increased lignin

control on C mineralization. This indicates that perceived lignin content is playing an

important part in the C mineralization process, with increased combined oxidative

enzyme activities, although % lignin content is lowest at this site. Interactions among

environmental and microbial parameters may play an important part in the elevation of

apparent lignin control. Elevated PHE activity may be partially due to the lack of









microbially available N, reflected in elevated LEU activity, which has been shown to

possibly repress PHE activity.

Overall enzyme activities within ENP-TS were not drastically lower than the

other enriched sites; in fact, the majority of activities were higher. One contributing

factor to this elevated activity may be due to the UV photolysis of inhibitory polyphenols

and other dissolved C compounds (Wetzel et al., 1995; Boavida and Wetzel, 1998;

Wetzel, 2000). It has been reported that 90% of the solar radiation is absorbed at a depth

of 2.5 cm in the northern Everglades while in the surface water of the southern

Everglades the depth was approximately 10 cm (Qualls and Richardson, 2003). Greater

UV penetration at the ENP-TS enriched site may serve to alleviate the inhibition by

polyphenols, resulting in more efficient community energetic.

At the ENP-TS reference site, apparent P limitation on C mineralization is

increased to the highest level of any site while TP, TN, TOC, lignin, and cellulose are all

the lowest. It should be noted that phosphatase activities reflect the summation of algal

and bacterial expressed enzymes (Jansson et al., 1988). The production of phosphatase is

most probably the primary motive force, based on the resource allocation strategy,

leading to the decreased production of the C and N acquiring enzymes at this site.

However, BGL activity may also be tied to the resident periphyton community.

Photosynthetically produced extracellular organic carbon (EOC) from the periphyton

community may supply greater amounts of labile carbon that is sufficiently degraded for

direct microbial uptake (Espeland et al., 2001). This may result in the markedly

decreased production of BGL observed at these sites if the majority of DOC released is of

sufficiently low molecular mass. When this readily utilizable carbon is available, there









may be no need for microorganisms to acquire it through enzymatic action (Chr6st and

Rai, 1993). The supplementation with EOC, resulting in lower BGL activities, may also

be exhibited among the suite of cellulase enzymes, which would coincide with the lower

CRR at the ENP-TS reference site.

In comparison with the other areas, ENP-TS appears to be a unique area of the

Everglades that is relatively un-impacted from canal inflows. The relative lack of

differences in apparent P limitation suggests very little P impacts to the resident

communities. The large shifts in apparent N limitation cannot easily be explained

through these techniques and suggests that some unique N anomaly is occurring near the

inflow, in terms of the microbial resource allocation view. The fact that Ecell/En

increases at the enriched sites with southerly canal flow suggests that this may be a

consequence of agricultural loading, runoff from the northerly areas or processes

occurring within the canals.

Transitional Sites

Transitional sites exhibit characteristics of both enriched and reference P regions,

which may provide optimal conditions for microbial activity. The increases in enzyme

activities among the transitional sites in the four areas suggest that different

biogeochemical conditions support the microbial communities. Specifically, the lower

nutrients present at the WCA-2A transitional site are even lower than the reference sites.

The elevated Ca, as much as 2 times the concentration at the enriched and reference sites,

may serve to relieve the inhibition by humic acids, which is analogous to the situation at

the enriched site in ENP-TS. The lower TOC and cellulose suggests the possibility of

relative C limitations, which are not necessarily supported by higher BGL activities, but

once again must be viewed as a relationship to N and P mineralization. The unique algal









composition, with an abundant Chara spp. mat composing the majority of the benthic

layer, has been shown to replace the Utricularia spp. and periphyton communities after

one year of P loading (Craft and Richardson, 1995). Elevated Ca levels in the benthic

and TP in the soil may be attributed to the CaCO3 and CaHP04 bands of the cell walls of

Chara spp. (Kiyosawa, 2001). The higher TP in the soil is supported by elevated

Ecell/Ep, suggesting a lower P limitation on C mineralization than the enriched site in

WCA-2A. Therefore, some of the enzyme activities reflected at this site may be

attributed to the metabolic and structural attributes of the resident algal community.

Several reasons may account for the nutrient dynamics at the transitional sites.

Upstream porewater dynamics may be exerting pressures on the transitional site. The

increase in TP in the soil layer, which is greater than that at the enriched site, suggests

that P is possibly being released from the soil at sites closer to the inflow. This may be

due to the reduction of P loading from the canals in recent years to a concentration lower

than that of the soil. The deposition of P downstream appears to result in the expansion

of the P front and may continue until a relatively static equilibrium is reached. Secondly,

and more specifically to the WCA-2A transitional site, the elevated TP may be due to P

removal capabilities of the Chara spp. communities and eventual burial into the soil layer.

The most consistent increase at the transitional sites in both the soil and benthic

layers was in relation to LEU. The decrease of average TN at the transitional sites in 6 of

8 cases supports these elevated activities. However, WCA-2A appears to be unique in

this regard. This is the only area, in both layers, to show an increase in Ecell/En at the

transitional site, indicating that N is less limiting to the microbial community as related to

C mineralization.









Conclusions

Interpreting individual enzyme activities is often difficult due to the wide range of

factors that control the induction and repression of activities. Additional factors include

inhibitory and enhancing effects of organic and inorganic compounds as well as other

environmental influences. The use of enzyme ratios that express activities in terms of

microbially perceived C, N, and P relationships in conjunction with nutrient data appears

to assist in beginning to understand some of the complex relationships.

The results of this study indicate that great care is required in any interpretations of

raw enzyme activities to higher order processes such as decomposition processes. Local

and regional influences must be accounted for before any conclusions are drawn. The

comparison of enzyme activities to decomposition rates, either using litter bags, cotton

strips, or other methods can assist in linking microbial activities to some of these higher

order processes.

The relative impacts of P enrichment on microbial enzyme activities differs

among the hydrologic units of the Everglades but generally results in decreased apparent

N and P control (higher Ecell/Ep and Ecell/En) and greater cellulose decomposition rates

(CRR). Site characteristics appear to contribute sharply to these differences. Plant

matter induction, algal composition, UV photolysis of humic substances, and

photosynthetically-produced extracellular organic carbon are some of the factors that are

hyothesized to contribute to these differences. However, the increases in primary

production, resulting from a greater C flow, may exceed any increases in decomposition,

resulting in a net accumulation of organic matter (Davis, 1991; Reddy et al., 1993; Craft

and Richardson, 1993).