Citation
Organic matter turnover along a nutrient gradient in the Everglades

Material Information

Title:
Organic matter turnover along a nutrient gradient in the Everglades
Creator:
DeBusk, William F., 1953-
Publication Date:
Language:
English
Physical Description:
vi, 176 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Dissertations, Academic -- Soil and Water Science -- UF
Soil and Water Science thesis, Ph. D
The Everglades ( local )
Peat ( jstor )
Nutrients ( jstor )
Microcosms ( jstor )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 167-175).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by William F. DeBusk.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
023805759 ( ALEPH )
35771602 ( OCLC )

Downloads

This item has the following downloads:

organicmattertur00debu.pdf

EJAARXOVF_ZF9VCH.xml

organicmattertur00debu_0171.txt

organicmattertur00debu_0145.txt

organicmattertur00debu_0010.txt

organicmattertur00debu_0088.txt

organicmattertur00debu_0078.txt

organicmattertur00debu_0087.txt

organicmattertur00debu_0017.txt

organicmattertur00debu_0006.txt

organicmattertur00debu_0060.txt

organicmattertur00debu_0033.txt

organicmattertur00debu_0082.txt

organicmattertur00debu_0147.txt

organicmattertur00debu_0161.txt

organicmattertur00debu_0142.txt

organicmattertur00debu_0001.txt

organicmattertur00debu_0165.txt

organicmattertur00debu_0016.txt

organicmattertur00debu_0056.txt

organicmattertur00debu_0103.txt

organicmattertur00debu_0115.txt

organicmattertur00debu_0061.txt

organicmattertur00debu_0148.txt

organicmattertur00debu_0155.txt

organicmattertur00debu_0108.txt

organicmattertur00debu_0098.txt

organicmattertur00debu_0118.txt

organicmattertur00debu_0096.txt

organicmattertur00debu_0093.txt

organicmattertur00debu_0140.txt

organicmattertur00debu_0112.txt

organicmattertur00debu_0150.txt

organicmattertur00debu_0135.txt

organicmattertur00debu_0000.txt

organicmattertur00debu_0110.txt

organicmattertur00debu_0077.txt

organicmattertur00debu_0089.txt

organicmattertur00debu_0034.txt

organicmattertur00debu_0090.txt

organicmattertur00debu_0178.txt

organicmattertur00debu_0081.txt

organicmattertur00debu_0076.txt

organicmattertur00debu_0095.txt

organicmattertur00debu_0121.txt

organicmattertur00debu_pdf.txt

organicmattertur00debu_0007.txt

organicmattertur00debu_0097.txt

organicmattertur00debu_0074.txt

organicmattertur00debu_0124.txt

organicmattertur00debu_0164.txt

organicmattertur00debu_0137.txt

organicmattertur00debu_0141.txt

organicmattertur00debu_0018.txt

organicmattertur00debu_0036.txt

organicmattertur00debu_0094.txt

organicmattertur00debu_0063.txt

organicmattertur00debu_0024.txt

organicmattertur00debu_0035.txt

organicmattertur00debu_0102.txt

organicmattertur00debu_0011.txt

organicmattertur00debu_0004.txt

organicmattertur00debu_0084.txt

organicmattertur00debu_0139.txt

organicmattertur00debu_0045.txt

organicmattertur00debu_0157.txt

organicmattertur00debu_0176.txt

organicmattertur00debu_0180.txt

organicmattertur00debu_0071.txt

organicmattertur00debu_0175.txt

organicmattertur00debu_0126.txt

organicmattertur00debu_0092.txt

organicmattertur00debu_0072.txt

organicmattertur00debu_0169.txt

organicmattertur00debu_0038.txt

organicmattertur00debu_0021.txt

organicmattertur00debu_0111.txt

organicmattertur00debu_0113.txt

organicmattertur00debu_0172.txt

organicmattertur00debu_0003.txt

organicmattertur00debu_0046.txt

organicmattertur00debu_0130.txt

organicmattertur00debu_0144.txt

organicmattertur00debu_0013.txt

organicmattertur00debu_0177.txt

organicmattertur00debu_0181.txt

organicmattertur00debu_0151.txt

organicmattertur00debu_0179.txt

organicmattertur00debu_0109.txt

organicmattertur00debu_0032.txt

organicmattertur00debu_0005.txt

organicmattertur00debu_0019.txt

organicmattertur00debu_0028.txt

organicmattertur00debu_0100.txt

organicmattertur00debu_0043.txt

organicmattertur00debu_0122.txt

organicmattertur00debu_0070.txt

organicmattertur00debu_0012.txt

organicmattertur00debu_0014.txt

organicmattertur00debu_0047.txt

organicmattertur00debu_0062.txt

organicmattertur00debu_0163.txt

organicmattertur00debu_0129.txt

organicmattertur00debu_0020.txt

organicmattertur00debu_0039.txt

organicmattertur00debu_0083.txt

organicmattertur00debu_0015.txt

organicmattertur00debu_0167.txt

organicmattertur00debu_0120.txt

organicmattertur00debu_0106.txt

organicmattertur00debu_0182.txt

organicmattertur00debu_0160.txt

organicmattertur00debu_0162.txt

organicmattertur00debu_0009.txt

organicmattertur00debu_0008.txt

organicmattertur00debu_0105.txt

organicmattertur00debu_0158.txt

organicmattertur00debu_0031.txt

organicmattertur00debu_0037.txt

organicmattertur00debu_0065.txt

organicmattertur00debu_0154.txt

organicmattertur00debu_0041.txt

organicmattertur00debu_0153.txt

organicmattertur00debu_0040.txt

organicmattertur00debu_0134.txt

organicmattertur00debu_0166.txt

organicmattertur00debu_0069.txt

organicmattertur00debu_0044.txt

organicmattertur00debu_0146.txt

organicmattertur00debu_0050.txt

organicmattertur00debu_0049.txt

organicmattertur00debu_0168.txt

organicmattertur00debu_0053.txt

organicmattertur00debu_0026.txt

organicmattertur00debu_0002.txt

EJAARXOVF_ZF9VCH_xml.txt

organicmattertur00debu_0030.txt

organicmattertur00debu_0107.txt

organicmattertur00debu_0123.txt

organicmattertur00debu_0149.txt

organicmattertur00debu_0104.txt

organicmattertur00debu_0125.txt

organicmattertur00debu_0058.txt

organicmattertur00debu_0067.txt

organicmattertur00debu_0143.txt

organicmattertur00debu_0064.txt

organicmattertur00debu_0042.txt

organicmattertur00debu_0052.txt

organicmattertur00debu_0117.txt

organicmattertur00debu_0119.txt

organicmattertur00debu_0085.txt

organicmattertur00debu_0054.txt

organicmattertur00debu_0136.txt

organicmattertur00debu_0073.txt

organicmattertur00debu_0022.txt

organicmattertur00debu_0174.txt

organicmattertur00debu_0116.txt

organicmattertur00debu_0079.txt

organicmattertur00debu_0025.txt

organicmattertur00debu_0091.txt

organicmattertur00debu_0131.txt

organicmattertur00debu_0075.txt

organicmattertur00debu_0156.txt

organicmattertur00debu_0101.txt

organicmattertur00debu_0059.txt

organicmattertur00debu_0132.txt

organicmattertur00debu_0048.txt

organicmattertur00debu_0099.txt

organicmattertur00debu_0128.txt

organicmattertur00debu_0029.txt

organicmattertur00debu_0133.txt

organicmattertur00debu_0173.txt

organicmattertur00debu_0086.txt

organicmattertur00debu_0152.txt

organicmattertur00debu_0068.txt

organicmattertur00debu_0023.txt

organicmattertur00debu_0127.txt

organicmattertur00debu_0138.txt

organicmattertur00debu_0051.txt

organicmattertur00debu_0080.txt

organicmattertur00debu_0170.txt

organicmattertur00debu_0159.txt

organicmattertur00debu_0114.txt

organicmattertur00debu_0027.txt

organicmattertur00debu_0057.txt

organicmattertur00debu_0066.txt

organicmattertur00debu_0055.txt


Full Text













ORGANIC MATTER TURNOVER ALONG A NUTRIENT GRADIENT
IN THE EVERGLADES












By


WILLIAM F. DEBUSK












A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA 1996



UNtVE'SITY OF iFLO,"'A LAS













ACKNOWLEDGEMENTS


This research was supported in part by the U. S. Department of Agriculture

National Research Initiative Competitive Grants program (Grant No. 92-37102-7542). The grant was awarded to Louisiana State University and University of Florida. Graduate Fellowships were funded in part by the USDA National Needs Fellowship Program for studies in the field of water science. I thank those who selected me as a recipient of this award. Thanks also go to all who donated their assistance, support and expertise, especially Dr. K. R. Reddy and Ms. Yu Wang. Most importantly, thanks go to my wife Patty for her moral (and, of course, financial) support.






























ii














TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS............................. ............... 11

ABSTRACT............... ................................ .. v

CHAPTERS

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

Background..................... ........ ....... .............. 1
W etland Carbon Cycle ................................................. .............. 2
M icrobial Ecology ................................................................ 8
Factors Affecting Decomposition Rate............................................... 9
Modeling Decomposition of Heterogeneous Substrates............................ 13
Ecosystem Models..................... ...... .................. 17
Everglades Study Site.......................... ................19
Objectives and Scope of Research .................................... ................ 21

2 ORGANIC C MINERALIZATION AS A FUNCTION OF
NUTRIENT ENRICHMENT AND HYDROLOGY....................... 23

Introduction ........................................................................ 23
M aterials and M ethods .................................. .................... ............ 26
Results ........................ ............... ................. 35
Discussion ................................ ....... ....... 51
Summary and Conclusions.............................. ....................61

3 TURNOVER OF ORGANIC CARBON POOLS ALONG
THE WCA-2A NUTRIENT GRADIENT ..................... ............. 62

Introduction ...................... ... ... ..... .. .............. 62
Materials and Methods ................. ....... ................... 67
Results................................ .. .. ...... ...................74
D iscussion ........................................................................ 88
Summary and Conclusions ............................................. .... ......98

4 REGULATORS OF ORGANIC MATTER DECOMPOSITION
ALONG THE WCA-2A NUTRIENT GRADIENT...................... 101

Introduction........................ .......... ................ 101
Materials and Methods ......................................... 105
Results ........................... ........ .................. 112
Discussion .............................................................. .............. 134
Summary and Conclusions......................................................... 141



Ili








5 DETRITAL CARBON MODEL ...................................................... 143

Introduction ......................................................... ........... ....143
Materials and Methods ................................................................ 143
Results and Discussion .............................................................. 155
Conclusions .............................................................................. 161

6 SUMMARY AND CONCLUSIONS ............................................ 163

LIST OF REFERENCES ....................................................................... 167

BIOGRAPHICAL SKETCH ..................................................... ........ 176













































iv













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

ORGANIC MATTER TURNOVER ALONG A NUTRIENT GRADIENT IN THE EVERGLADES

By
William F. DeBusk

August, 1996



Chairman: Dr. K. R. Reddy
Co-chairman: Dr. J. W. Jones
Major Department: Soil and Water Science

Organic matter accumulation in wetlands represents a potential long-term sink and source for organic carbon (C) and associated nutrients and contaminants. Turnover of organic C was measured in a nutrient-impacted sawgrass and cattail marsh in Everglades Water Conservation Area 2A (WCA-2A). Controlled laboratory incubations and a microcosm study were conducted to determine potential rates of C mineralization in plant litter and peat along a gradient of phosphorus (P) enrichment. Field incubations at 10 sites along the nutrient gradient measured in situ organic matter decomposition rate throughout the floodwater, litter and peat profile.

Organic C mineralization in wetland microcosms was significantly enhanced by. interactive effects of increased P availability and decreasing water table. Approximately 90% of the variability in potential organic C mineralization in peat and plant litter, measured under aerobic and anaerobic conditions, was explained by total P and lignocellulose content of the organic substrate. Anaerobic mineralization rates were 32% of the rates measured under aerobic conditions. In situ organic matter decomposition rate was higher in nutrientv








enriched areas of WCA-2A than in the low-nutrient interior marsh. Decomposition rate typically was at a maximum in the floodwater and litter layer and decreased with depth in the peat profile. Field studies provided evidence that microbial decomposers obtain nutrients, especially P, from the surrounding floodwater and soil porewater as well as from the organic substrate.

Results of laboratory and field studies indicate that organic C turnover in WCA-2A is strongly affected by P availability, although 02 availability is the major controlling factor. Availability of C (substrate quality) and nitrogen (N) may limit turnover rate under Penriched conditions. Experimental findings from these studies provide insight into the effects of accelerated nutrient loading on C cycling and net accumulation of organic matter and nutrients in wetlands.

































vi














CHAPTER 1
INTRODUCTION


Background


Organic carbon (C) accumulation in wetlands is the mass balance between net

primary production (C fixation) and heterotrophic metabolism (C mineralization). Organic C in plant litter, peat or soil organic matter (SOM) serves as the source of energy to drive the detrital food chain in wetlands. Most of the organic matter produced in wetlands is deposited directly in the detrital pool (Moran et al., 1989; Wetzel, 1992), thus microbial decomposers play the major role in C cycling and energy flow in wetlands.

Burial of organic matter as peat provides a means for long-term storage of elements associated with organic C, such as nutrients and heavy metals (Clymo, 1983). Allochthonous compounds may be incorporated into peat and soil organic matter through plant uptake and senescence, immobilization within the soil organic matrix by physical/chemical processes such as adsorption, occlusion and precipitation, or through uptake by microbial decomposers, with storage either within living cells or metabolic byproducts. On a much broader scale, storage of organic C in wetland soil is an important component of the global carbon cycle and thus may impact large-scale processes such as global warming and ozone depletion (Happell and Chanton, 1993; Whiting, 1994).

Under favorable conditions for organic matter decomposition, stored nutrients or contaminants may be released through mineralization and then recycled in the ecosystem or exported from the system (Ponnamperuma, 1972; Reddy and D'Angelo, 1994). The rate of net organic matter accumulation is a critical determinant of how a wetland functions as an ecological unit within the landscape. The storage function is equally important for natural



1








wetlands, especially those which represent an ecotone between terrestrial and aquatic ecosystems, and created wetlands, which may be used for treatment of wastewater or runoff (Howard-Williams, 1985).


Wetland Carbon Cycle


From a conceptual standpoint the continuum of organic C transformations and

flows which constitute the wetland C cycle may be represented as a collection of discrete storage units, or compartments, with simultaneous transfer of mass among the compartments (Figure 1-1). The vegetation component (including macrophytic and algal species) represents transformers of inorganic C (CO2) to organic C (primary production) through the process of photosynthesis. The heterotrophic microfauna represent transformers of organic C back to inorganic C through cellular respiration. Organic C is stored in the system in living (vegetation and microbial biomass) and non-living (dead plant tissue, plant litter, peat or SOM) components. Non-living storage of organic C is proportionally large in wetlands in relation to other ecosystems; this storage provides a substantial energy reserve to the ecosystem which is slowly released through the detrital food web (Wetzel, 1992).


Peat


Peat is the result of biological, chemical and physical changes imposed on plant remains over an extended time period. Extent of decomposition, or humification, is qualitatively assessed by the extent to which plant structure is preserved (Given and Dickinson, 1975; Clymo, 1983). Numerous organic constituents have been isolated from peat, and many have been used in assessing the degree of decomposition, although these are generally classes and subclasses of organic compounds rather than discrete compounds. For example, peat material soluble in non-polar solvents is often termed "wax." but includes numerous compounds other than waxes (esters of fatty acids with alcohols other





3








C02


cO2
Plant tissue CH4
Leaching
Decomposition

Decomposition/ileaching F F Mineralization Lte,- Microbial Litter biomass +-- DOC HCO3~

Burial
Microbial HCO3 PEAT +- om-+sDOC Leaching biomass CH4
Decomposition Decomposition/eaching Mineralization



Figure 1-1. Conceptual diagram of the organic carbon cycle in a wetland ecosystem.





4


than glycerol) (Clymo. 1983). Although a small portion of total peat mass, this fraction is of interest when determining origins of peat, since fatty acids of lipids can be traced directly or indirectly to the original plant type (Given and Dickinson, 1975; Borga et al., 1994). Various types of acid or alkali hydrolysis procedures have been used to isolate fractions roughly equivalent to cellulose, hemicellulose and lignin. Proportions of cellulose and hemicellulose in plant litter and peat tend to decrease with age (i.e. decomposition or "humification"), while lignin content increases with age; these three structural groups have all been used to characterize the degree of peat decomposition (Clymo, 1983; Brown et al., 1988; Bohlin et al., 1989). Analysis of peat has also revealed a large variety of phenolic compounds, many of which may be extracted in the "lignin" fraction. Concentration of cellulose and lignin is much higher in the fibrous fraction (remnant plant parts) of peat, while humic acids are much more prevalent in the humus fraction, although fulvic acids were found in somewhat greater amounts in the fibrous fraction (Given and Dickinson, 1975). Humic acid content of temperate peats showed a tendency to increase with age and depth, but peats in tropical and subtropical regions, including the Everglades, did not reflect this trend. Interpretation of chemical analyisis of peat may be clouded by uncertainty about the extent to which present characteristics represent historical differences in vegetation versus variability of decomposition processes.

The degree of decomposition generally increases with depth of the peat, therefore, the organic matter becomes increasingly humified at greater depth (Clymo, 1983). Assuming an historically constant quality and quantity of substrate addition to the soil, the rate of decomposition is greatest in the upper regions of the profile, decreasing with depth. This is due not only to antecedent decomposition of substrate in the soil profile, but to the vertical gradient of soil environmental parameters. The latter category may include dissolved 0, concentration (in saturated soil), soil moisture (unsaturated soil), and nutrient availability. All of these variables, particularly the first two, are functions of hydrologic conditions. Dissolved O, concentration is affected both by the rate of diffusion from the





5


soil surface, which is greatly reduced in flooded soils, and the 0, demand created by organic matter along the diffusion path (Howeler and Bouldin, 1971). Dissolved Organic C


The ecological significance of dissolved organic C (DOC) in wetlands has not been clearly defined. Even in terrestrial and aquatic systems, for which a greater depth of knowledge exists for DOC dynamics, the role of the dissolved fraction of organic C has not been well established. Among the reasons for this is the fact that DOC represents a broad spectrum of organic compounds of varying environmental recalcitrance (Wetzel, 1984), thus it may not be appropriate to treat DOC as a homogeneous category.

Cook and Allan (1992a) measured several DOC fractions in old field soils

representing various stages of succession (12-62 years), and observed changes in DOC composition during incubation period of 210 days. They found that the total amount and chemical composition of DOC in soil solution does not correspond to potential biodegradability. The Leenheer fractionation scheme failed to measure changes in DOC substrate quality, although the total mass of DOC decreased during the incubation period, and the total N concentration increased. The authors suggested that analysis of soil water DOC fails to account for the bulk of the microbial activity, which is intimately associated with surfaces of detritus and particulate soil organic matter. Cook and Allan (1992b) measured soil DOC concentration in conjunction with instantaneous rate of C mineralization to determine whether DOC represents the primary source of energy and nutrient release for soil metabolism. The size of the soil DOC pool was weakly correlated with instantaneous C mineralization rate.

The DOC pool in aquatic ecosystems is a relatively stable component, both in terms of the size and quality of the pool (Wetzel, 1984). Decomposition of organic substrates in aquatic systems involves both labile and complex (recalcitrant) organic matter, the latter comprising the bulk of the DOC. Turnover of highly labile, energy-rich substrates may





6


approach a rate of 5-10 times per day, thus actual concentration (storage) of labile organic C is generally extremely low. Despite the fact that recalcitrant DOC and POC are slow to mineralize, these pools represent a major portion of the organic C processed by the heterotrophic community due to the relatively massive size of these pools.

During transport, selective removal of organic compounds occurs due to microbial utilization and chemical adsorption or precipitation, thus recalcitrance of DOC increases downgradient. Decomposition of dissolved organic compounds (resulting from partial decomposition of particulate organic matter) occurs primarily at surfaces in the litter layer, soil/sediment and among epiphytic microflora. Microbial Biomass


Most of the organic C fixed in aquatic and marsh systems (by both phytoplankton and macrophytes) is processed and recycled entirely by bacteria, without entering the food web, i.e. higher animals (Wetzel, 1984). Decomposition of organic matter is the primary ecological role of the heterotrophic microflora in soils, as it provides for mineralization of growth-limiting nutrients and formation of recalcitrant organic compounds (e.g. humus) which contribute to the chemical stability of the sytem (Swift, 1982). Microbial biomass comprises only a small fraction of the non-living organic matter, yet most of the net ecosystem production passes through the microbial component at least once and typically several times (Elliott et al., 1984; Heal and Ineson, 1984; Van Veen et al., 1984). Microbial decomposers derive their energy and C for growth from detrital organic C and facilitate recycling of energy and C within and external to the wetland ecosystem (Wetzel, 1984; 1992). Soil microbes may exert a significant influence on ecosystem energy flow in the form of feedback, since mineralization of organically-bound nutrients is a regulator of nutrient availability for both primary production and decomposition (Elliott, et al., 1984).

The soil biomass constitutes a major C sink, in that it represents a significant

portion of the "active" organic C (Paul and van Veen, 1978). Nutrients may be held tightly





7


within the microbial biomass component of a low-nutrient ecosystem reflecting efficient recycling of remineralized organic compounds (Melillo et al., 1984). Microbes and plants often compete for nutrients in ecosystems with limited nutrient input and tight nutrient cycles (Lodge et al., 1994). Transient environmental conditions may stress the microbial community and result in fluctuations in biomass. This temporary increase in microbial mortality may result in significant remineralization of nutrients and induce a pulse of available nutrients for the plant community (Lodge et al., 1994).

The wetland environment is characterized by widespread anoxia, thus the

importance of anaerobic metabolism in organic matter turnover is greatly increased in wetlands versus terrestrial ecosystems (Ponnamperuma, 1972; Reddy and D'Angelo, 1994). The predominant type of anaerobic respiration in wetlands depends on the relative availability of alternate electron acceptors. Depending on geographic location and local anthropogenic influences, anaerobic metabolism in freshwater wetlands may be regulated by inputs of NO-, Mn4 Fe'" or SO42. Utilization of these electron acceptors by heterotrophic microflora follows a thermodynamically predictable sequence, but also depends on availability of the compounds (Westermann, 1993; Reddy and D'Angelo, 1994). Microbial respiration in freshwater wetlands is frequently limited by electron acceptor availability, rather than C availability as in terrestrial ecosystems. In the complete absence of electron acceptors, methanogenesis is the major metabolic pathway in anaerobic wetland soils (Westermann, 1993). Microbial respiration in estuarine ecosystems such as salt marshes and mangrove swamps is dominated by sulfate reducing bacteria, due to the high concentration of sulfate in seawater (Howarth, 1993). Sulfate reducers and methanogens utilize most of the same metabolic by-products (short-chain fatty acids) of fermenting bacteria, which do not require an external source of electron acceptors (Oremland, 1988; Howarth. 1993). Sulfate reducing bacteria have the ability to outcompete methanogenic bacteria for these substrates in the presence of sulfate. Where the availability of suitable C substrate is sufficiently high, methanogenesis and sulfate





8


reduction both may contribute significantly to soil respiration in high-sulfate soils (Oremland, 1988).


Microbial Ecology


Degradation of complex substrates such as plant detritus represents the actions of a diverse and heterogeneous assemblage of microorganisms which exhibits successional trends similar to those observed at the macro-scale (Swift, 1976). Saito et al. (1990) documented succession of cellulolytic fungi and bacteria to a more diversified bacterial assemblage during decomposition of pure cellulose in a waterlogged soil. The concept of r and K strategists in population biology has been applied to microbial populations in natural systems (Heal and Ineson, 1984; Swift, 1984; Atlas, 1986). An r strategist microorganism is often characterized by rapid growth rate and dominance in environments where resources are abundant, especially under fluctuating environmental conditions. In a marsh ecosystem, such conditions might result from cyclic drying and reflooding or pulsed loading of nutrients. The r strategists are exploitative, opportunistic and density independent, and would be expected to dominate during early successional stages (Insam and Haselwandter, 1989). The microbial K strategists would include slow growing, oligotrophic, humusdegrading organisms which are adapted to resource-limited environments (high stress). Growth of K strategists is generally damped, density-dependent and controlled by interspecific competition (Heal and Ineson, 1984). The r and K strategy concept generally has been related to microbial response to environmental stress as a selective pressure. Recently, the idea of disturbance as an additional selective pressure has been put forth. This has led to additional classifications of organisms based on their response to various combinations of stress and disturbance (Heal and Ineson, 1984; Swift, 1984; Atlas, 1986).

Indices based on microbial activity and organic C have been proposed to provide an operationally-defined means for describing the response of soil microbial populations to resource quality and environmental conditions. The ratio of microbial biomass C to soil





9


organic C (C,/Cor) has been related to soil C availability and the tendency for a soil to accumulate organic matter (Anderson and Domsch, 1989: Sparling, 1992). Another, more widely investigated, microbial index is the metabolic quotient or specific respiration rate (qCO,), the ratio of the basal respiration rate (as C02-C) per unit microbial biomass C (C,c) (Insam and Haselwandter, 1989; Anderson and Domsch, 1990,1993; Wardle, 1993; Ohtonen, 1994). The qCO, has been used as a response variable to effects of temperature, soil management. ecosystem succession and heavy metal stress, and is apparently a significantly more robust parameter than C,JCorg (Anderson and Domsch, 1993). Relationships have been shown between qCO, and r and K strategies of stress-induced selection, and also with the general theory of ecosystem development proposed by Odum (1969). Increased qCO, is associated with r-selected microbial populations, resulting from high resource availability and simple substrate-decomposer relationships in early successional ecosystems (Insam and Haselwandter, 1989: Wardle, 1993). Lower qCO, values may be indicative of K-selected populations dominating mature systems, especially low-nutrient systems with closed cycling of resources. In addition, increased qCO, levels are apparently related to ecosystem disturbance, such as pollution (Ohtonen, 1994).


Factors Affecting Decomposition Rate


The decomposition/mineralization process in wetlands differs from that in upland ecosystems in a number of ways (Reddy and D'Angelo, 1994). Thepredominance of aerobic conditions in upland soils generally results in rapid decomposition of organic matter such as plant and animal debris. Net retention of organic matter is minimal in this case, and consists of accumulation of highly resistant compounds which are relatively stable even under favorable conditions for decomposition (Jenkinson and Rayner, 1977; Paul, 1984). The decomposition process occurs at a significantly lower rate in wetland soils, due to frequent-to-occasional anaerobic conditions throughout the soil profile resulting from





10


flooding. Because of this, significant accumulation of moderately decomposable organic matter occurs, in addition to lignin and other recalcitrant fractions (Clymo, 1983).

There is evidence that growth strategies of the heterotrophic community reflect

those of the plant community through their direct response to variation in resource quality, which is a function of plant growth strategies, and their similar response to common environmental conditions (Heal and Ineson, 1984). Carbon and nutrient utilization by the microbial decomposer community responds to three main groups of factors: (1) substrate quality, (2) physicochemical environment and (3) other organisms. Substrate quality is a general term which refers to the combination of physical and chemical characteristics which determine its potential for microbial growth. Substrate quality is not determined by any particular factor; however, nitrogen and lignin content have been suggested as indicators of biodegradability (Heal et al., 1981; Minderman, 1968; Andr6n and Paustian, 1987; Melillo et al., 1989). Physicochemical factors include temperature, pH, exogenous nutrient supply, moisture content (for non-flooded conditions) and oxygen or alternate electron acceptor availability (Swift et al., 1979; Heal et al., 1981; Reddy and D'Angelo, 1994).

Nutrient availability affects decomposition rate by limiting microbial growth.

Growth-limiting nutrients may be obtained by the microbial decomposers from the organic substrate or from dissolved compounds in the water and porewater (Godshalk and Wetzel, 1978). Microbial decomposers colonizing nutrient-depleted substrates (e.g. high C:N or lignin:N ratio) tend to scavenge significant amounts of nutrients from the surrounding media, resulting in net immobilization of growth-limiting nutrients in the system (Melillo et al., 1984). Although nutrient loading is typically greater in wetlands than in uplands due to location within the landscape, nutrient availability may be low relative to the pool of available organic C in wetlands (Reddy and D'Angelo, 1994). Nitrogen (N) and phosphorus (P) both have been identified as microbial growth-limiting nutrients in wetlands (Westermann, 1993). Nitrogen, unlike P, may be lost from wetlands through








microbial metabolism via denitrification, as well as through ammonia volatilization (Reddy and D'Angelo, 1994).

Substrate availability to microbial decomposers is determined by molecular size

distribution, availability of nutrients in the substrate as well as in the environmental matrix, and the amount of exposed surface area (Heal et al., 1981). Preferential utilization of low molecular weight compounds by microbial decomposers alters the chemical character of the organic resource such that the proportion of complex, slowly degradable compounds increases over time. Thus, the substrate quality of resources over a wide range of initial composition tends to converge to a more uniform composition, and hence, degradability (Melillo et al., 1989).

The basic cell wall construction in vascular plants includes a framework component of a-cellulose, a matric component of linear polysaccharides (hemicellulose), and an encrusting component composed of lignin (Zeikus, 1981). Lignin occurs in cells of conductive and supportive tissue, and thus is not found in algae and mosses. The presence of lignin is the ultimate limiting factor in decomposition of vascular plant tissue (Zeikus, 1981).

The heterogeneous group of organisms known as white-rot fungi, found primarily in terrestrial habitats, are the most active and complete decomposers of lignin (Eriksson and Johnsrud, 1982). Partial chemical modification of lignin has been documented for other eucaryotic microbes, most notably the brown-rot and soft-rot fungi, as well as for certain species procaryotes (Zeikus, 1981). Certain species of the genera Streptomyces, Norcardia, Bacillus, Azotobacter and Pseudomonas are included in the latter category. Numerous species of fungi capable of cellulose and lignin degradation, including Hyphomycetes, have been identified in oxidized zones of wetland soils (Westermann, 1993).

Lignin degradation by white-rot fungi is highly oxidative, and may involve singlet oxygen and hydroxyl radicals in chemical oxidation of aromatic ring structures or





12


intermonomeric linkages (Benner et al., 1984). However, recent research has cast doubts on the unconditional requirement of molecular 0, for lignin degradation. Wetland plant decomposition studies have demonstrated decomposition of lignin in anoxic salt marsh, freshwater marsh and mangrove sediments using "4C labelling techniques (Benner et al., 1984a). Bacterial degradation of lignin from Spartina alterniflora predominated over fungal degradation in studies of decomposition in salt marsh sediment (Benner et al., 1984b).

The capacity for cellulose depolymerization is shared by several species of bacteria, actinomycetes and microfungi (Sagar, 1988b). Cellulose degradation readily occurs under anaerobic conditions, although at a reduced rate, mediated primarily by bacteria of the genus Clostridium (Swift et al., 1979). The ratio of cellulose to lignin degradation rate has been shown to be similar under both aerobic and anaerobic conditions (Benner et al., 1984a).

Aside from its effects on lignin degradation and other extracellular

depolymerization, 02 depletion forces a major shift in microbial metabolism of monomeric C compounds (e.g. glucose, acetate), from aerobic to anaerobic pathways (Westermann, 1993). Catabolic energy yields for bacteria utilizing alternate electron acceptors (NO, Mn4, Fe'3, SO4=, CO2) are lower than for 02, thus microbial growth rates are generally lower in anaerobic environments (Westermann, 1993; Reddy and D'Angelo, 1994). In addition, sulfate reducing and methanogenic bacteria must depend on fermenting bacteria (e.g. Clostridium spp.) to produce substrate in the form of short chain C compounds, such as volatile fatty acids, from the breakdown of mono- and polysaccharides (Howarth, 1993). Thus, although C metabolism occurs in the absence of 02, and even in the complete absence of electron acceptors, the decomposition process for plant litter and soil organic matter is often significantly curtailed.





13


Modeling Decomposition of Heterogeneous Substrates


Conceptual models developed for describing decomposition of heterogeneous substrates such as plant residues and soil organic matter generally fall into one of four major groups: single homogeneous compartment, two-compartment, multi-compartment and non-compartmental heterogeneous models (Jenkinson, 1990). Compartmental models utilize separation of the organic matter (or specifically organic C) into discrete compartments with significantly different turnover times. One approach attempts to quantitatively separate certain chemical fractions of organic matter which are presumed to control decomposition kinetics of the heterogeneous substrate, or can be described as functions of some readily measured chemical parameter or environmental variable (Minderman, 1968). Many researchers have used a more operationally defined approach to group different fractions of organic matter. A common conceptual scheme for organic matter fractions involves grouping according to empirically derived measures of biodegradability. Resulting categories may simply be termed "labile", "resistant", "stable", etc. This approach has been the most commonly used for ecosystem C and N models (Van Veen and Paul, 1981; Parton et al., 1987; Jenkinson, 1990; Grant et al., 1993a,b). Another approach to dealing with heterogeneous substrates treats the substrate as a single component of variable quality, or overall biodegradability (Godshalk and Wetzel, 1978; Moran et al., 1989). This is more often seen as a theoretical approach to description of decomposition, and has not been widely implemented in soil C and N models.

Minderman (1968) evaluated decomposition data for plant litter from four forest types. It was shown that total mass loss over time could not be accurately described by a simple first-order decay model. However, when litter was separated into six individual components, viz phenols, waxes, lignin, cellulose, hemicellulose and sugars, mass loss for each component followed first-order decay kinetics. Mass loss of the sum of the 6 components was similar to total mass loss for forest litter. Thus, it was concluded that a





14


multiple, or composite, exponential decay equation was a more appropriate model for litter decomposition than a single compartment model. It was also concluded that long-term (e.g. >10 years) decomposition kinetics is regulated by the resistant fractions of litter. This conceptualization of litter decomposition was based on only one year of decomposition data, and was not rigorously evaluated with the available data nor validated with additional data sets.

First-order rate constants (k) for decomposition of hardwood leaf litter were shown to be highly correlated with initial tissue lignin:N ratio (Melillo et al., 1982). It was also shown that, for each type of litter, N content of the residual material was linearly related to mass loss (inverse-linear relationship) during the first year of decomposition. Further research on plant litter decomposition during a 77-month period established a strong relationship between substrate biodegradability and the ratio of lignin to total lignocellulose (lignin+cellulose) (Melillo et al., 1989). This lignocellulose index (LCI), or the proportion of remaining structural material occurring as lignin, varies widely among types of plant litter, e.g. from about 0.2 in "high quality" litter to about 0.6 in "low quality" litter. After an initial phase of degradation in which most of the simple compounds, including sugars and amino acids, are rapidly lost, the LCI for plant residues slowly converges on the 0.70.8 range found in soil organic matter. The implication of this was that, after initial stages of decomposition, the rate of decomposition is regulated only by environmental conditions, and is no longer a function of initial substrate composition.

Aber et al. (1990) contended that the first-order rate constant (k) and the slope of

the inverse-linear relationship of mass remaining vs. N content can be predicted from initial C fraction analysis (extractables, cellulose and lignin) and initial N content. They also accurately predicted decomposition of heterogeneous material using exponential decay models for the individual C fractions, although this was validated for only a small variety of litter types.





15


Andr6n and Paustian (1987) compared zero-order, single first-order, parallel (twocompartment composite) first order, consecutive (two-compartment) first-order and a fourcompartment model. Overall, the simple first-order model incorporating environmental effects provided the best description of the data, although the temperature-corrected 2compartment models were also satisfactory, and gave a good estimate of the initial size of the labile (water soluble) pool. However, this pool could also be determined using the single-compartment model with Y-intercept determined by curve-fitting (therefore Yintercept is equivalent to the resistant fraction).

A single-compartment exponential decay model using an exponentially decreasing rate coefficient (k) was used by Godshalk and Wetzel (1978) to describe decomposition of five species of aquatic macrophytes. The form of the model was dW/dt = -k*W, where k= a*exp(-b*t) and W is percent of initial weight remaining, t is time in days, k is the firstorder rate coefficient and a and b are rate parameters. The exponentially decreasing rate coefficient reflected the increasing overall recalcitrance of the substrate as the composition shifted to a greater proportion of highly resistant compounds. Decay rates were negatively correlated to the total fiber content of the substrate, but not well correlated to individual components such as cellulose, hemicellulose and lignin.

A study by Moran et al. (1989) provided a critical evaluation of single- and multicompartment exponential decay models for plant litter, based on chemical components of the litter. The authors measured in situ and laboratory decomposition rates of whole litter and the lignocellulose components of litter for the emergent macrophytes Spartina alterniflora and Carex walteriana. Following an initial rapid (ca. two weeks) phase of mass loss from leaching and decomposition of non-lignocellulosic components, losses of C due to decomposition were attributable mainly to the lignocellulose fraction. By tracking the mass loss over time for individual chemical components of substrate, they were able to mechanistically construct a composite exponential decay model of whole-litter decomposition:





16


N, = Soexp(-kt)+Hoexp(-k2t)+Coexp(-k3t)+Loexp(-kt) [1-1] where So, Ho, CO and L0 represent initial amounts of soluble material, hemicellulose, cellulose and lignin in the substrate. Initial values for the four components shown in equation 1-1 were experimentally determined quantities. First-order rate constants (k) were estimated for each compartment using a curve-fitting routine. However, evaluation of alternative decomposition models revealed that the whole-litter decomposition process (at least for a one-half year period) was more accurately described by the decaying-coefficient model (Godshalk and Wetzel, 1978). The main shortcoming of the composite exponential model was the relatively poor fit (underestimation) during later stages of the experimental decomposition period, i.e. after about four months. This was attributed for the most part to an apparent decrease in specific decomposition rate (k) over time for individual components of the substrate. The authors suggested that selective degradation of less resistant fractions within each component (e.g. various phenolic subunits of the lignin component), along with the increasing importance over time of physical protection and humification, were probable causes for the observed decrease in component decay rates. It was concluded that the enormous number of biochemically distinct components, each with a characteristic biodegradability (and distinct rate constant) presents an inherent limitation on the use of the composite exponential model for accurate mechanistic descriptions of decomposition.

A single-compartment decomposition model was proposed by Bosatta and Agren (1985), in which heterogeneity was described by a continuously varying quality variable q. A continuity equation for substrate C density was the basis for describing the flow of C along a continuum of time and quality,

=r(q,t)/8t = -S(q,t) + OF(q,t)/aq [1-2] where r = density of substrate C, S = loss or sink of C due to microbial degradation, and F = flow of C into a defined region of q space (quality). Under this concept, mass of substrate C over the entire span of time and quality can be viewed as a surface plot of p(q,t), where t and q are represented by the x and y axes, with p as the height (z axis). The





17


total mass of C remaining at time t may be calculated by integrating over q0 to q, (range of quality) with time held constant. This model is quasi-mechanistic in that it represents a substrate as a composite of compounds with varying biodegradability while avoiding the pitfalls of artificially-imposed boundaries between biochemical components. It may be viewed as a two-dimensional version of the decaying coefficient model described previously. However, validation of this type of model using independent data sets has not yet been reported in the literature, therefore, the validity of the model for description of short- or long-term C dynamics is undetermined.


Ecosystem Models


Numerous ecosystem-scale models of organic C and N cycling and soil organic

matter accumulation have been developed for grasslands and agricultural systems. Many of these employ the concept of discrete compartments for organic C or N pools with different turnover times. A long-term organic C model was constructed for evaluating turnover of soil organic matter in virgin and cultivated grasslands (Paul and Voroney, 1980; Van Veen and Paul, 1981). Organic matter was separated into decomposable, ligniferous and recalcitrant fractions for above- and below-ground compartments. In addition, the soil profile was divided into three depth increments. Microbial biomass was considered separately as the "processing" compartment for soil organic matter. Microbial transformation and turnover of organic C and N were the basis for a soil organic matter model for predicting N immobilization in cultivated soils (Van Veen et al., 1984). A more detailed treatment of microbial kinetics was used in a soil organic matter model designed to evaluate short-term C and N dynamics (Grant et al., 1993a,b). The model was based on literature-derived equations for microbial activity, then validated for prediction of temporal trends in mineralization and immobilization of C and N in added substrates ranging from glucose to crop residues.





18


Two well-documented models are particularly broad in scope, one in a temporal

sense and the other from a spatial perspective. The Rothamsted turnover model (Jenkinson and Rayner, 1977: Jenkinson, 1990) was developed for prediction of long-term soil organic matter dynamics in croplands. Organic matter entering the model system is tranferred stepwise through five compartments representing decomposable plant material, resistant plant material, humified organic matter, microbial biomass and CO, lost from the system. Development of the Rothamsted model has benefitted from the availability of field data for a period of well over 100 years. The Century model of soil organic matter (C and N) in Great Plains grasslands (Parton et al., 1987) has been used to simulate effects of climatic gradients and grazing on soil organic matter levels and plant productivity over a wide geographic region. This model assigns soil organic C and N to three separate compartments based on turnover time: active (including microbial biomass), slow and passive soil organic matter. In addition, plant residue is divided into structural (slow) and metabolic (rapid) pools. The Century model has also been used to simulate short- and longterm effects of fire on N cycling in soil and plants (Ojima et al., 1994).

A simulation model of organic C mineralization was developed for management of cultivated Histosols in the Everglades (Browder and Volk, 1978). Organic C was partitioned into a nonliving pool, according to degradability or chemical structure, and an active living pool containing microbial biomass. Water table (which controlled soil moisture and 02 availabilty), temperature and soil organic C content were used as effects to predict rates of soil subsidence and release of N compounds and organic acids to surface and groundwater. An ecosystem model of the Everglades sawgrass marsh was developed for predicting effects of various anthropogenic impacts on ecosystem structure and function (Bayley and Odum, 1976). Simulations included manipulation of hydroperiod and water depth, which influenced plant growth and fire occurrance. Phosphorus concentration in surface inflows was varied to simulate the effects of nutrient loading from anthropogenic sources, which impacted plant growth rates. Peat accumulation was simulated as a function





19


of plant growth and fire. The simulation model suggested that the relatively simple ecosystem is highly unpredictable and sensitive to inflow of high P water and hydroperiod. However, calibration and validation of the model were hindered by a lack of reliable data for many system processes.


Everglades Study Site


The Everglades of south Florida is a mosaic of wetland ecosystems extending from near Lake Okeechobee southward to Florida Bay (Figure 1-2). The various ecosystems making up the pre-drainage Everglades, which encompassed an area of about 10 000 km2, were highly adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred primarily through rainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959).

Recent development of the Everglades and surrounding watershed has created changes in nutrient loading and hydrology (SFWMD, 1992). Most significantly, a large area of the northern Everglades was drained and converted to agricultural production during the first half of this century. This area of sugar cane, vegetable and sod farming is referred to as the Everglades Agricultural Area (EAA). The remainder of the northern Everglades was divided into three Water Conservation Areas (Figure 1-2) in the 1960s, for water storage and flood control. Water level within the WCAs is controlled by a system of levees, pumps and floodgates. Currently, the Everglades consists of the WCAs to the north and Everglades National Park to the south.

Drainage of the EAA has resulted in widespread oxidation of the organic soil and concomitant mineralization and leaching of organically-bound nutrients. As a result,





20








r------ ------I Lake
Oklloseuho

WS Palm
Bech MAP VIEW




I \CiFan



mnW, CaSa-domsed
































Figure 1-2. Everglades WCA-2A study area, showing sampling
originating from the S- surface inflow.rd

S-1OD

100






WCA-2A










0 5 10 Kilmeters







Figure 1-2. Everglades WCA-2A study area, showing sampling
transect across the nutrient enrichment gradient,
originating from the S-10C surface inflow.





21


nutrients from organic soil mineralization, along with additional nutrients from fertilizers, have been transported via drainage canals toward the WCAs for approximately 30 years. Loading of agricultural drainage water into the WCAs has resulted in nutrient enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod, in the encroachment of cattail (Typha domingensis Pers.) and other rapidly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward and Ornes, 1983; Toth, 1987, 1988).

Accelerated nutrient loading in northern WCA-2A (Figure 1-2) during the past three decades has created a distinct nutrient (especially P) gradient in water, soils and plant tissue (Davis, 1991; Koch and Reddy, 1992; DeBusk et al, 1994). Changes in species composition of periphyton and macrophyte communities, along with an overall increase in net primary productivity have been documented along this gradient (Davis, 1991; SFWMD, 1992). Soil dating by analysis of 137Cs peaks has indicated that peat accumulation rate has increased in nutrient-enriched areas of WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993).


Objectives and Scope of Research


The overall objective of this study was to determine turnover time of organic C

pools in plant litter and peat along the gradient of nutrient enrichment in Everglades WCA2A. It is hypothesized that turnover of organic C changes along vertical and lateral profiles within the study area. Chapters 2 through 4 address more specific objectives:

* Determine the effects of nutrient enrichment and flooding on mineralization of organic

C in the peat-litter profile. It is hypothesized that increased nutrient loading and

decreasing water table interactively increase C mineralization rate in the litter and peat.

* Determine the effect of nutrient enrichment on turnover of organic C pools along the

WCA-2A nutrient gradient, and to examine the relationships between size of the





22


microbial biomass C pools and turnover time of associated organic C pools. It is hypothesized that turnover time for major C pools increases (decomposition rate

decreases) downgradient from the inflow of nutrient-laden water. It is also

hypothesized that turnover time increases in successively older organic C pools.

* Determine the effect of nutrient enrichment on in situ decomposition rate along a vertical

profile in the water column and peat, specifically the significance of various

environmental and substrate-related factors on decomposition rate.

* Develop a mass balance for organic C using experimental data, and construct a

conceptual model to describe turnover of organic C pools and net C accumulation.

Experimental data will be generated from a combination of field, greenhouse and laboratory studies, presented in the following chapters. Chapter 2 describes a greenhouse study using Everglades microcosms to determine effects of water table and nutrient enrichment on whole-profile C mineralization. Laboratory incubations of Everglades soil and plant litter to determine turnover of organic C pools are presented in Chapter 3. Studies involving in situ decomposition of plant material and an artificial substrate are described in Chapter 4. Development of a conceptual model and synthesis of experimental results are presented in Chapters 5 and 6.













CHAPTER 2
ORGANIC C MINERALIZATION AS A FUNCTION OF
NUTRIENT ENRICHMENT AND HYDROLOGY Introduction


Decomposition of organic matter is governed by the chemical composition of the

substrate and external, or environmental, factors. Among the more important environmental factors are temperature, moisture, nutrients and electron acceptors (Swift et al.. 1979; Heal et al., 1981; Reddy and D'Angelo, 1994). The most important electron acceptor in terms of organic matter turnover is O,. The presence of floodwater severely limits availability of 0, in wetlands, therefore decomposition proceeds at a highly reduced rate. In the absence of 02, breakdown of organic carbon (C) by microbial decomposers is accomplished using alternate electron acceptors, such as NO3, Mn', Fe 3, and SO-, which results in lower energy yield for the organisms (Reddy and D'Angelo, 1994). The hydroperiod of a wetland, which encompasses frequency, duration and depth of flooding, is thus a major determinant in the accumulation of organic matter.

Substrate quality, a general term referring to the degradability of an organic

substrate by microbial decomposition, also has a major impact on rate of decomposition (Swift et al, 1979; Heal et al., 1981; Heal and Ineson, 1984). Generally included in the concept of substrate quality are the availability of organic C (which is a function of lignocellulose content and other factors) and nutrients for cell maintenance and growth. The required nutrients for microbial decomposition of organic C may be aquired from the substrate itself, e.g. as part of the original plant material, or from the floodwater or porewater (Swift et al, 1979). In a nutrient-enriched, or eutrophic, wetland, plant litter (substrate), floodwater and porewater may serve as nutrient sources for decomposers. In


23





24


nutrient-poor, or oligotrophic wetlands, the organic substrate may be the main source of nutrients for microbial decomposers, if nutrient concentration in the water and porewater is extremely low.

Nutrient availability affects decomposition rate through its effects on microbial growth. Although nutrient loading is typically greater in wetlands than in uplands due to location within the landscape, nutrient availability may be low relative to the pool of available organic C in wetlands (Reddy and D'Angelo, 1994). Nitrogen (N) and phosphorus (P) both have been identified as microbial growth-limiting nutrients in wetlands (Westermann, 1993). Nitrogen, unlike P, may be lost from wetlands through microbial metabolism via denitrification, as well as through ammonia volatilization (Reddy and D'Angelo, 1994).

Th Everglades encompasses a variety of wetland ecosystems which were historically adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred primarily through rainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974) (Figure 2-1). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959).

Loading of agricultural drainage water into the WCAs has resulted in nutrient

enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus (P) enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod, in the encroachment of cattail (Typha domingensis Pers.) and other rapidly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward and Ornes, 1983; Toth, 1987, 1988).






25













Okeechabee.

West Pan
Beach MAP VIEW \ Fort



usrni Cadail-dominaled Mked
NGtional Park / sawgra8-etaJJ 10C
10A





WCA-2A









0 5 10 KilometersWCA-2B








Figure 2-1. Site map for WCA-2A study area, showing locations of sampling
sites. Coordinates for sampling sites are listed in Table 2-1.





26


The main objective of this study was to determine the effects of nutrient enrichment and flooding on turnover, or mineralization, of soil organic C along the WCA-2A nutrient gradient. It is hypothesized that increased nutrient loading and decreasing water table interactively increase C mineralization rate in the litter and peat.


Materials and Methods


Site Description


Field study sites were located in WCA-2A, a 447 km2 region of the northern Everglades (Figure 2-1). Surface water flows into WCA-2A from the Hillsboro Canal through the four S-10 water control structures and from the North New River Canal through the S-7 pump station. Most of the hydraulic loading is through the S-10C and S10D structures into the northern portion of WCA-2A. The general direction of flow is from north to south. Water depth is usually less than one meter, and varies considerably, both seasonally and year-to-year, with occasional dry periods (SFWMD, 1992; personal observations). The bulk of the surface outflow is through three control structures at the south end of WCA-2A, into WCA-3 (Figure 2-1).

Soil in WCA-2A consists of Everglades and Loxahatchee peats (Gleason et al., 1974). Everglades peat, the most common soil in the Everglades, is associated with the sawgrass marsh community. It is dark brown, finely fibrous to granular, with circumneutral pH, relatively high N content and low SiO,, Fe and Al content. Peat depth in WCA-2A ranges from about 1 to 2 m, and age of basal peats is estimated to be 2000 to 4800 yr. Beneath the peat lies a bedrock of Pleistocene limestone, with intermediate layers of calcitic mud, sandy clay and sand in several areas (Gleason et al., 1974).

The primary sources of nutrient loading to WCA-2A are the S-10 structures which convey water from the Hillsboro Canal and WCA-1 (Figure 2-1). A distinct gradient of N and, most significantly, P enrichment in water, plants and soil has formed between the





27


high-nutrient region adjacent to the inflows and the low-nutrient interior marsh of WCA-2A (Koch and Reddy, 1992; SFWMD, 1992; DeBusk et al., 1994). A vegetation gradient coincides with the nutrient gradient; most notable is the gradient from sawgrass marsh with scattered aquatic slough in the interior to cattail and mixed emergents near the inflows. The vegetation gradient was divided into three discrete categories for the purposes of the current study: cattail-dominated, sawgrass-dominated and mixed cattail and sawgrass (Figure 2-1).

A north-south transect approximately 10 km long was established for soil, water and vegetation sampling along the nutrient gradient in WCA-2A. Sampling sites were located along the transect at 7 locations, starting near the S-10C inflow structure at the Hillsboro Canal and ending in the interior marsh region (Figure 2-1; Table 2-1). These locations were chosen as representative of a wide range of soil and water P concentration and vegetation type. A total of 10 sampling sites were established among the 7 locations on the transect, as follows. One site each was situated at distances of 0.75 and 2.2 km from the inflow (sites T 1 and T2), within the highly nutrient-impacted cattail dominated area (Figure 2-1). Two sites each were located at distances of 3.1, 4.0 and 5.0 km from the inflow, in the moderately impacted transitional vegetation zone characterized by patches of cattail and sawgrass and mixed stands of cattail and sawgrass. At each of these three distances, one sampling site was located within a cattail stand (sites T3, T4 and T5 respectively) and one within a sawgrass stand (sites C1, C2 and C3 respectively). The minimally impacted interior region was represented by sites within the sawgrass marsh at distances of 6.9 and 10.2 km from the inflow (sites C4 and C5). This area is characterized by a significant coverage of aquatic slough habitat; however, sampling was limited to the marsh community to maintain continuity among all sites.


Sampling Methodology


Duplicate soil cores were taken on July 11, 1994 at each of the 10 sampling sites in WCA-2A. An additional (triplicate) core was obtained for cattail and sawgrass sites at each





28









Table 2-1. Locations of WCA-2A sampling sites for
microcosm study and approximate downstream distance from the S-10OC surface water inflow.
Site Latitude N Longitude W Distance
deg mein g min km
T1 26 21.8 80 21.1 0.75 T2 26 21.1 80 21.2 2.2 T3,C1 26 20.5 80 21.3 3.1 T4,C2 26 20.0 80 21.4 4.0 T5,C3 26 19.5 80 21.4 5.0 C4 26 18.5 80 21.5 6.9 C5 26 16.8 80 21.5 10.2





29


end of the respective ranges along the transect, i.e. at sites Ti, T5, Cl and C5. This scheme provided triplication at the sites of highest and lowest impact and at 2 transitional sites. Replicate cores within each site were located approximately 2 m apart. In all, 24 soil cores were obtained.

Soil cores were collected intact in 50 cm-length sections of rigid, clear acrylic

tubing with inside diameter of 14.6 cm (nominal dimensions: 6-inch diameter with 1/8-inch wall thickness). The coring procedure was optimized during previous field trips, such that disruption of litter and peat stratigraphy was avoided and compaction of the core was minimized (less than 5%). Penetration of the loosely packed litter layer and mats of fine roots at the easily compressed peat surface may result in significant compaction using traditional methods of core driving using a hammer or pile driver. Coring for this study was accomplished by pushing the acrylic tubes downward by hand while cutting around the perimeter of the core tubes with a serrated knife (i.e. breadknife) to sever pieces of plant litter and roots. The coring tubes were pushed into the soil until the peat surface was aligned with marks placed 30 cm from the bottom of the tubes. Intact cores were excavated using a shovel, and the bottom openings were plugged with a polypropylene disks (2.54 cm thick) machined to a diameter slightly smaller than the core I.D. and fitted with dual rubber O-rings for a water-tight seal. Core tubes were capped for transport from the field, such that air headspace was reduced or eliminated, to prevent turbulent mixing and possible disruption of the core stratigraphy.


Soil-Water Microcosms


The microcosm study was set up in an open-sided greenhouse on the University of Florida campus. Intact soil cores (soil-water microcosms) were placed upright in aluminum racks situated inside polyethylene livestock troughs. The troughs were filled with water to a level coinciding with the top of the peat in the core tubes, creating near-ambient temperature water baths to buffer temperature in the microcosms. Water temperature was continuously





30


monitored during the study with thermocouples connected to a data logger (Model CR10, Campbell Scientific, Logan UT).

Partial shading was provided (to approximate field light conditions) by draping shade cloth over the tubs containing the microcosms. Measured levels of PAR (photosynthetically active radiation) beneath the shade cloth were less than 25% of full sunlight at midday. Spot measurement of PAR beneath the canopy of a dense cattail stand in a nearby planted wetland yielded values of approximately 120-150 gE m-2 s-', or about half the rate measured for the microcosms. However, the canopy density of cattail and sawgrass at the WCA-2A sampling sites was estimated to be significantly lower than in the experimental wetland.

Floodwater depth was adjusted in each microcosm to 10 cm by siphoning off

excess water. Therefore, the soil-water microcosms represented the top 30 cm (meanS.D. = 30.51.5 cm) of the WCA-2A peat profile, plus the overlying layer of plant litter and 10 cm of surface water. During the experimental period, distilled water was added as needed to microcosms to compensate for evaporation of surface water. Microcosms were allowed to stabilize for about 2 weeks, before initiation of experiments. The following studies were performed during the months of August and September, a period of minimal day-to-day temperature variation.


Dissolved 0z and pH Profiles


Vertical profiles of dissolved 02 and pH were measured in each microcosm during mid-day, over a period of one week. A needle-type microelectrode (Model 757; Diamond General, Ann Arbor, MI) was mounted on a motorized screw-drive apparatus, developed in-house for redox profiling of lake sediment cores. The microelectrode was slowly (typically 3 mm mind) driven vertically through the water column and litter layer to the peat surface. These measurements were timed to coincide with maximum algal photosynthetic activity, thus approximating maximum expected values of dissolved O2 and pH in the water





31

column. In addition, diel measurements of dissolved 02 were taken within the litter layer of microcosms representing high- and low-impact areas (sites Ti and C5). Profiles of dissolved 02 were recorded, as described above, and repeated every 6 hours for a 24-hour period.


Soil Respiration


In preparation for soil respiration measurements, the plastic tubs containing the soil-water microcosms were covered with sheets of plywood to excude sunlight and terminate algal activity. Respiration of the heterotrophic microflora could be better estimated without interference from algal photosynthesis and respiration. Microcosm core tubes were capped with polyethylene plugs, which were securely attached with silicone glue to provide an airtight seal. The caps contained inlet and outlet ports for gas exchange from the core headspace. A gas manifold was set up to feed compressed air to each microcosm through plastic tubing. Air was bubbled through the surface water at a depth of about 5 cm, at a rate of approximately 30 mL min-. The air was passed through a 2N NaOH solution prior to delivery to the microcosms to scrub CO2 from the inflow gas.

Soil respiration was estimated by measuring CO2 and CH4 evolved from the microcosms, using the following procedure. Outflow air from each microcosm was sampled through a short outflow tube (Pharmed brand rubber tubing) in the top with a I mL syringe fitted with a 25-gauge hypodermic needle. The syringe containing gas sample was immediately inserted into a butyl rubber stopper to retard leakage through the needle. A second syringe was used to take a duplicate sample in the same manner. This procedure was repeated for all 24 microcosms. Triplicate samples of the inflow air stream were also taken, for calculation of mass inflow rates for CO2 and CH4 (if detectable). At each sampling event, air outflow rate was measured for each microcosm, using a Manostat Calcuflow flowmeter (Manostat, New York, NY). The syringes containing outflow air samples were immediately taken to the lab for analysis of CO, and CH4. Respiration





32


measurements were made under aerated (air bubbled into the water column) and non-aerated (air delivered to the headspace above the water surface) conditions in the water column, to simulate both aerobic and anaerobic flooded conditions in the field.

Mass of CO, and CH4 in the 1 mL air samples was determined by direct sample

injection into a dual-detector gas chromatograph (Hewlitt-Packard 5840A, Avondale, PA). Separate determinations of CO, and CH4 were made from the duplicate samples taken from each microcosm. Sample gases were analyzed for CO, and CH, analysis using thermal conductivity (TCD) and flame ionization detectors (FID), respectively. For CO, analysis, a Poropak N (Supelco, Bellefonte, PA) column was used, with He as a carrier gas. Oven, injector and detector temperatures were set to 60, 140 and 200 "C. For CH4 analysis, a Carboxen 1000 (Supelco, Bellefonte, PA) column was used, with a N2 carrier gas. Oven, injector and detector temperatures were 120, 120 and 200 'C. Mass flux of CO, and CH4 from each microcosm was calculated as:

(outflow conc. inflow conc.) x (air flow rate).

Following measurement of CO, and CH4 flux under flooded conditions, one

replicate microcosm from each of the 4 triplicated site (T1, T5, Cl and C5) was extruded from the core tube and sectioned into 2-cm "slices", with the litter layer collected separately. The discrete depth intervals thus obtained were placed in air-tight polyethylene containers and refrigerated at 4 "C. Samples were later homogenized by mixing with a spatula; large pieces of plant material were chopped into smaller (ca. 1 cm) pieces to increase the homogeneity of the sample. Subsamples of the peat and litter samples were used for determination of moisture content, bulk density and analysis for total C, N and P. Additional subsamples were used in an aerobic soil respiration study under controlled conditions (described below). Total C and N analyses were performed on oven-dried (60 C), ground (< 0.2 mm) subsamples using a Carlo-Erba NA-1500 CNS Analyzer (HaakBuchler Instruments, Saddlebrook, NJ). Separate subsamples were analyzed for total P following combustion (ashing) at 550 C for 4 h in a muffle furnace and dissolution of the





33


ash in 6 M HCI (Anderson, 1976). The digestate was analyzed for P using the automated ascorbic acid method (Method 365.4, USEPA, 1983).

Soil respiration (CO, and CH4 flux) measurements were repeated under drained

conditions. A 0.5 cm drain hole was drilled through the side wall of each core tube, about 1 cm from the bottom plug. A polypropylene fitting packed with glass wool was used to connect a short length (approx. 40 cm) of 0.5 cm diameter Tygon tubing to the drain hole. Water was allowed to drain from the microcosms to the desired level. The Tygon tubing was then taped to the outside of the core tube in a vertical orientation, to serve as an indicator of water level in the microcosm (water table as opposed to saturated capillary fringe). Fine adjustment of water level was performed by extracting water with a 50 mL syringe fitted with a long, large-bore needle placed along the inner wall of the microcosms.

Experimental determination of soil respiration was repeated under conditions of transient water level. Measurement of CO, and CH4 flux were made according to the protocol described above, as water level in the microcosms was sequentially lowered to the peat surface and 5, 10 and 15 cm below the peat surface. Water level was maintained at each depth for approximately 5 days, with flux measurements made on the last day. Additional flux measurements were made at 7 and 19 days, while the microcosms were drained to -15 cm.


Aerobic Soil Incubation: Potential Soil Respiration


Subsamples of the litter and 2-cm vertical sections of the "sacrificed" replicate microcosms from sites Tl, T5, C1 and C5 (0.75, 3.1, 5.0 and 10.3 km from inflows) were used for a laboratory study of soil respiration potential. Short-term aerobic assays of CO2 production were set up in triplicate 160-mL glass serum bottles containing 10 g wet weight (approximately 1 g dry weight equivalent) of sample, either litter or peat from each 2-cm section of the 30 cm profiles. Bottles were covered with clear plastic (PVC) film, to retard moisture loss while allowing gas exchange, and placed in a temperature-controlled





34


incubator equipped with a reciprocal shaker table. The sample bottles were incubated in the dark at 27 'C, with constant agitation to ensure uniform exchange of gases between soil and air over the short time intervals of measurement. Samples were pre-incubated for 48 hours to allow stabilization of microbial populations in the soil and litter.

Soil respiration was determined by measuring short-term increase in CO, in the

headspace of closed serum bottles containing litter or peat samples. A static system, rather than a more conventional flow-through (respirometer) method. To begin the respiration assay, sample bottles, which had been provided continuous exchange of air, were stoppered with sleeve-type rubber serum stoppers. One mL of headspace was sampled from each serum bottle with a gas-tight syringe, for the initial (time = 0) sample. The sample bottles were immediately returned to the incubator/shaker. Headspace samples were taken again at 10, 20 and 30 minutes. Sample bottles were shaken continuously except for the brief period of time required for sampling. Erroneous results have been reported for static incubations of calcareous soils, associated with solubility and retention of CO2 as HCO3 (Martens, 1987). However, incubation time was shortened sufficiently to minimize the CO, gradient between headspace and soil solution, and maintain a linear increase in CO, headspace concentration over the incubation period.

Headspace samples were analyzed for CO2 by GC-TCD, as described above. The value thus obtained was then extrapolated from injection volume to the sample headspace volume. The time rate of change of CO2 in the serum bottle was used for calculation of CO2 production rate. For the short (30 min.) interval used in the assay, accumulation of CO2 in the headspace was a linear function of time, i.e. production rate was constant. Aerobic soil respiration was determined by relating the rate of CO2 production to the sample mass.





35


Results


Soil Chemical Characteristics


Soil total C and N content was relatively invariant among the four cores analyzed, and within each profile (Table 2-2). Total C content of the peat in the four selected cores was about 44% (w/w), nearly all of which was organic C (based on an organic matter content of 88-90%). Storage of N relative to total C content, expressed as C:N ratio, was consistently high (low C:N ratio) in soil and litter of the four microcosms (Table 2-2). Magnitude of C:N ratio varied only slightly around a mean of roughly 15, with slightly greater values found for site C5.

Analysis of total P revealed a distinct trend in soil and litter P enrichment among sites (Figure 2-2). In the case of microcosm Tl, substantial P enrichment (compared with the interior marsh site C5) was found throughout most of the top 30 cm of peat and litter. Microcosm C5, in contrast, showed slightly elevated total P concentration in the litter and the upper 8-10 cm of peat, decreasing below about 20 cm. Significant P enrichment also occurred in the litter and upper portion of the peat profile of microcosms Cl and T5 (Figure 2-2). The overall extent of P enrichment at these transitional sites was less than for site T1; P concentration approached "background" levels for sawgrass peat below about 15 cm in T5 and about 28 cm in C1. The C:P ratio of litter and the upper 80% of the soil profile in microcosm Tl was approximately 300 (Figure 2-3). In contrast, C:P ratio in microcosm C5 ranged from about 800-900 in the upper profile to 2000 in the deeper peat.


Dissolved 0, and pH Profiles


Dissolved 02 profiles (Figure 2-4) revealed a distinct spatial pattern of intense,

localized 02 production and rapid utilization in the litter layer and at the soil surface. These local 02 sources and sinks resulted in highly variable concentration and numerous microgradients within the profile. Profiles of pH showed a high degree of similarity to





36




Table 2-2. Chemical analysis of the peat-litter profile of wetland microcosms
from sites T-1, T-5, C-1 and C-5 in Everglades WCA-2A.
Ash
Site Depth content Total C Total N Total P C:N C:P
cm % ---- g kg '---- mg kg'

T-1 Litter 10.4 446 27.3 1563 16.3 285
0-2 10.3 445 30.6 1600 14.5 278 2-4 11.1 434 29.5 1604 14.7 271 4-6 11.7 447 31.5 1635 14.2 273 6-8 13.7 443 33.2 1558 13.3 284 8-10 13.0 441 30.6 1592 14.4 277 10-12 12.5 437 29.4 1545 14.9 283 12-14 11.4 445 29.8 1491 14.9 299 14-16 11.6 460 28.6 1436 16.1 320 16-18 12.8 448 28.6 1475 15.7 304 18-20 13.3 426 27.3 1421 15.6 300 20-22 11.2 439 30.0 1342 14.6 327 22-24 11.8 438 28.8 1324 15.2 331 24-26 10.9 435 28.2 1044 15.4 417 26-28 15.4 421 28.7 703 14.7 599 28-30 16.2 411 31.0 452 13.3 911
T-5 Litter 10.5 437 30.0 1141 14.5 383
0-2 10.5 421 30.5 1184 13.8 356 2-4 11.2 440 33.6 1312 13.1 335 4-6 9.3 436 34.5 1355 12.6 321 6-8 9.9 433 34.6 1216 12.5 356 8-10 10.6 436 34.9 897 12.5 486 10-12 12.1 442 35.7 448 12.4 986 12-14 10.1 458 36.4 291 12.6 1574 14-16 9.0 464 37.8 252 12.3 1839 16-18 9.2 456 38.2 239 11.9 1903 18-20 9.3 465 37.6 257 12.4 1808 20-22 10.4 456 37.8 232 12.1 1964 22-24 11.5 427 35.9 213 11.9 2004 24-26 14.7 444 36.2 199 12.3 2224





37







Table 2-2--continuued.
Ash
Site Depth content Total C Total N Total P C:N C:P cm % ---- gkg'---- mg kg"

C-1 Litter 12.8 435 29.9 1328 14.6 328
0-2 11.5 501 34.4 1232 14.6 406 2-4 11.7 440 30.0 1181 14.7 372 4-6 12.0 482 31.3 981 15.4 491 6-8 11.5 442 29.1 1059 15.2 418 8-10 11.1 439 28.0 905 15.7 485 10-12 11.1 399 26.7 793 15.0 503 12-14 12.2 423 28.6 730 14.8 579 14-16 11.1 382 25.8 677 14.8 565 16-18 10.7 440 29.0 531 15.2 830 18-20 11.2 433 27.1 471 16.0 919 20-22 9.9 383 28.0 408 13.7 938 22-24 10.3 408 28.9 294 14.1 1389 24-26 9.8 372 25.9 314 14.4 1181 26-28 8.5 518 39.0 278 13.3 1863 28-30 8.5 417 33.6 246 12.4 1698
C-5 Litter 16.6 476 37.5 534 12.7 893
0-2 13.7 451 35.6 586 12.7 770 2-4 13.5 491 38.6 599 12.7 820 4-6 12.0 451 32.8 546 13.8 825 6-8 11.4 469 27.7 493 16.9 952 8-10 11.0 494 28.0 357 17.7 1386 10-12 10.4 473 27.2 336 17.4 1409 12-14 9.8 429 22.3 307 19.3 1398
14-16 9.5 467 26.1 291 17.9 1606 16-18 9.3 442 25.7 275 17.2 1605 18-20 9.4 458 27.7 252 16.5 1816 20-22 9.6 423 26.6 243 15.9 1739 22-24 9.3 414 26.3 236 15.7 1753 24-26 9.4 425 27.1 215 15.6 1976 26-28 9.5 463 28.8 211 16.1 2192





38



TOTAL P (mg kg1 )

0 500 1000 1500 2000 0 500 1000 1500 2000
LITTER
0-2 2-4 4-6 6-8
8-10 10-12 12-14
14-16 16-18 18-20
20-22 22-24
E 24-26 Site T1i Site T5
26-28 0.75 km 5km S 28-30


LITTER
0-2 C 2-4
4-6 6-8
8-10
10-12 12-14 14-16
16-18 18-20
20-22 22-24
24-26 I Site Cl Site C5
26-28_ 3 km 10 km
28-30



Figure 2-2. Soil total P profiles in selected microcosms from WCA-2A.





39


C:P RATIO

0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500
LITTER
0-2 2-4 4-6 6-8
8-10
10-12
12-14 14-16 16-18 18-20
20-22 22-24
S22-2 Site T1 Site T5 E 24-26
S 26-28 0.75 km 5km S 28-30

O LITTER
0-2 O 2-4
4-6 6-8
8-10
10-12 12-14 14-16 16-18 18-20
20-22 22-24
24-26- Site C1 Site CS
26-28- 3 km 10 km
28-30


Figure 2-3. C:P ratio in profiles of selected microcosms from WCA-2A.





40


those for dissolved 02, resulting from localized changes in CO, concentration associated with algal photosynthesis.

Zones of high, and occasionally supersaturated, levels of dissolved 02 occurred within the macro-litter layer (primarily large pieces of cattail leaves) throughout the water column of microcosms T 1, T2 and T3 (nutrient-enriched), associated with filamentous algal mats embedded within the litter. The range of pH values in these profiles was approximately 8.2 to 8.6. A similar scenario was observed for microcosms C1 and C2, where the litter layer was somewhat less developed. At site C3, an extensive loosely-arranged macro-litter layer (primarily sawgrass) was covered with attached periphyton, which produced a 2-3 cm deep zone of 02 supersaturation. This resulted in pH levels of about 9, compared with about 8.3 in the overlying water column. The traces of dissolved 02 indicate diffusion from the photosynthetic source to the overlying water and underlying litter and soil. The solid horizontal lines at depth=0 in Figure 2-4 represent the approximate location of the peat surface. A layer of fine organic sediment, about 1 to 1.5 cm deep, covered the peat in all microcosms. Thus, the "sediment-water interface" actually occurred above the horizontal line at depth=0. Depletion of O, with distance was more rapid in the downward direction, due to greater demand and slower diffusion in the litter, fine sediment and peat than in open water. While the prodigious amount of 02 generated locally within the water column was sufficient to oxygenate much or the litter layer, the peat remained anoxic, due to total depletion of 02 near or above the peat surface.

Dissolved 02 profiles in microcosms T4, T5, C4 and C5, representing the

transitional and low-nutrient areas in the WCA-2A marsh, reflected the reduced size and depth of litter and the presence of a benthic periphyton layer above the fine sediment and peat (Figure 2-4). Distinct peaks of dissolved O2 occurred near the surface of the periphyton layer, representing supersaturated conditions within the zone of maximum photosynthesis. Steep gradients of dissoved 0, resulted from diffusion from the narrow photosynthetic zone to the overlying water column, and especially into the sediment, which






41











E




--~~ -o



o 4






-- -'- -



0
rE: o






a...-- ------ -o




.L,
( ,.
0 Mc o c o
I a o n I
o o
T-------~ - \ r











(tuo) HLda





42


represented a strong 02 sink. The production of 02 in the periphyton and depletion in the sediment layer was accompanied by parallel increases and decreases in pH.

Diel measurements of dissolved O, in two selected microcosms, representing

high-nutrient (T1) and low-nutrient (C5) sites, revealed considerably different responses depending on time of day (Figure 2-5). The midday dissolved O0 profile in microcosm T1 was similar to the profile in Figure 2-4 for the same site, although the diel measurement was performed on a separate replicate core. Beween noon and 6 PM, production of O, in the litter algal mat had decreased considerably, due to diminished light availability. However, as a result of several hours of 02 production and vertical migration through diffusion, a redistribution of 02 had occurred, and the lower litter layer and fine sediment were partially oxidized, as well as the upper litter layer. The large 02 demand, created by the high availability of C and nutrients for microbial decomposers, in the litter and sediment caused rapid depletion of dissolved O2 between 6 PM and midnight. During the night (by 6 AM the next day) the litter layer, which encompassed the entire water column, had become essentially anaerobic.

As shown previously in Figure 2-4, significant oxygenation of the periphyton and litter layer occurred during the middle part of the day in the C5 microcosm (Figure 2-5). As in the T1 microcosm, a vertical redistribution of 02 had occurred by 6 PM, although substantially less symmetrical in the C5 microcosm. The absence of litter above the periphyton layer permitted greater diffusive flux of 02 into the water column of C5, while the significantly more consolidated sediment and peat restricted diffusion of 02, such that the demand exceeded the supply. While O2 demand was higher in the sediment-upper peat layers of T5, the unconsolidated nature of these layers allowed much greater rate of 02 diffusion than in C5. The decrease in dissolved O, during the night was attenuated in the C5 microcosm relative to T1, probably due to the quantity and quality (degradability) of the litter, and availability of nutrients for microbial decomposition. The result of this was an





43














I0









L0






ZZ 0
-i o



















u. .
0 .J co










(w ) -.......d
Ca ,





z0 eO O












1--d





44


oxidized water column (about 20% saturation) during the night, and a partially oxidized litter layer.


Column COz and CH, Flux


Mean temperature in the microcosms for the duration of the study was 27 *C. Temperatures recorded during gas flux measurements ranged from 24-30 "C, thus deviating no more than 30 from the mean temperature. A temperature correction was applied to the data, based on experimental results of Volk (1973), who measured soil respiration over a range of temperature and soil moisture in intact cores of Monteverde muck (sawgrass peat) from the northern Everglades. Data from that study were used to calculate a Q10 value of 1.82 for microbial respiration within the range of temperatures encountered during the present study. Based on this relationship, all CO2 and CH4 flux data were normalized to the mean temperature of 27 "C.

Flux data for each microcosm (i.e. individual field reps) are shown in Figure 2-6 for flooded and drained conditions, plotted as a function of distance from the inflow. The general trend for gaseous C flux (CO2 + CH4) from the microcosms was that of increased flux with decreasing water table and decreasing distance from the inflow, that is, there was a positive response to the nutrient gradient. However, these tendencies did not always apply when CH4 flux was considered separately, or when flux data were divided into subgroups by water table, vegetation type or sample site.

Flux of CO2 was relatively high under flooded conditions, even when compared to flux measured under saturated (no floodwater) and partially drained conditions (Figure 26). As previously noted, aeration was provided to the water column to approximate the oxygenation brought about by algal photosynthesis. The mean flooded CO, flux, for all sites combined, of 2.6 gg C cmi2 h-' (horizontal dotted line) was significantly greater (a =

0.05) than the mean values of 0.8 and 1.9 gg C cm-2 h-' for saturated (water table at soil surface) and drained (-5 cm) conditions, respectively. However, mean CO2 flux for





45


flooded microcosms was significantly lower than in more extensively drained (-10 and -15 cm) microcosms. Values for mean CO, flux in the latter were 3.2 and 5.0 gLg C cm2 h-2, respectively.

Flux of CO2 under flooded conditions decreased significantly with increasing

sampling distance from the WCA-2A inflow, according to linear regression anaylsis (a =

0.05). This trend was significant primarily by virtue of the difference in means between cattail and sawgrass sites (significant according to ANOVA, ac-0.05). Within vegetation types, a significant trend in flux vs. distance from inflow was observed for cattail (T1-T5) sites, but not for sawgrass (C1-C5) sites. These trends are also evident by visual inspection of data points relative to the overall mean (broken line) in Figure 2-6.

The lowest CO, flux resulted from lowering the water level to the peat surface. Under these conditions, the litter layer was compressed at the soil surface, and remained saturated. Thus, aeration of the litter layer was restricted to diffusion from the surface. Given the high 02 demand of the litter, depletion of 02 occurred within a few millimeters of the surface of the litter. This was confirmed with spot measurements of dissolved O2 with a micro-electrode, as described previously. Unlike the case of flooded microcosms, there was no significant trend in CO2 flux with distance from the inflow and no difference among cattail vs. sawgrass dominated sites.

Mean CO2 flux increased significantly with decreasing water table from 0 to -15 cm (Figure 2-6). Multiple linear regression analysis indicated a significant (a = 0.05) linear response of CO2 flux to water table, with no significant response to distance from inflow. However, there was significant (a = 0.05) interaction between water table and distance from inflow. The linear model using both water table and distance as factors (independent variables) explained 85% of the variability in CO2 flux. With water table as the only factor, the model explained 81% of the variability in flux. The positive response to water table decrease was also significant (a = 0.05) for each sampling site considered individually. Evidence of the interaction between water table and distance from inflow was the





46


8
FLOODED A Cladium O FLOODED 41
6 ~- 0---e------o Typha 0 o 10
4- O00 -102

a10-4 2 10-3

10

SATURATED O SATURATED
6- 10"1
D O

O A
2- 10.3

I 0 ... I. . 1 10-4 I N 8- 1 N E DRAINED (-5 cm) E O 6 0 10-1 0 O - 0 -or O O 4 0 10-2 aP
Sx
X 2- -- A 10-3 X DRAINED (-5 cm)
.L 0 . -r .... .. .... .... .... 10 .L

6 DRAINED (-10 cm) O
6 10-2


2 A 003
4- O10-2

2-O10

0.0
DRAINED (-10 cm)



0 00
6 02 0 10.1 1 88- -- -----10-2
A A

2- 10
DRAINED (-15 cm) DRAINED (-15 cm)
0 .... I ...'lI. .I. ''I . .... I. .'h' I'..g'I 'II 10
0 2 4 6 8 10 0 2 4 6 8 10

DISTANCE FROM INFLOW (km)



Figure 2-6. Wetland microcosm CO, (left) and CH4 (right) flux under flooded and drained
conditions. Data points represent means of replicate flux measurements for each field rep (one data point for each microcosm). Symbols differentiate
between Cladium and Typha dominated sampling sites.





47


development of a visible trend in flux along the distance axis with increasing depth of water table (Figure 2-6). For a water table of -15 cm this trend was significant (a--0.05), which was also the case under flooded conditions. Additionally, at -15 cm water table depth the mean CO2 flux in microcosms from cattail sites was significantly greater than for sawgrass sites.

The mean CH4 flux (all microcosms combined) was much greater under flooded conditions (0.148 gg C cm-2 h-1) than when the microcosms were drained, though differences among means were not significant for a = 0.05. In general, CH4 flux was approximately one to two orders of magnitude smaller than corresponding CO2 flux in the microcosms. Data presented in Figure 2-6 suggest that CH4 flux at cattail-dominated sites was greater than at sawgrass-dominated sites, but this comparison was not statistically significant. A significant trend in CH4 flux along the nutrient gradient existed only for drained conditions with water table at -15 cm. In this case, unlike the trend for CO2 flux, CH4 flux increased with distance from the inflow. When microcosms in this group were analyzed by vegetation type (cattail vs. sawgrass) the trend was significant for sawgrassdominated sites (C1-C5) only.

Response of CO2 and CH, flux to water table and distance from inflow is

summarized in Figure 2-7 by a series of curves representing total C (CO2-C + CH4-C) flux vs. distance for each water level, with vegetation types combined. Flux of CO2 represented roughly 90-99% of the total C flux, thus the trends in CO2 flux discussed earlier also apply to flux of total C. In particular, total C flux increased significantly with lowering of the water table from the peat surface (saturated conditions) to 15 cm below the peat surface. In addition, the decrease in total C flux with increasing distance from water inflows was significant only for flooded conditions and a water table of -15 cm.





48








7


It:6-1



SDrained (-15 cm)
4- -""" """ '----J
%1b Drained (-10 cm)




-1 Saturated (no floodwater) 0 2 4 6 8 10
DISTANCE FROM INFLOW (km)


Figure 2-7. Total C flux (CO,+CH4) from wetland microcosms as a function of water
table and distance of sample site from surface water inflow.





49


Potential Soil Respiration


Aerobic incubation of plant litter and peat subsamples from 2-cm depth intervals

provided an index of potential soil respiration, or organic C mineralization rate, in the field. An useful method of comparing mineralization rate among sites is to express soil respiration as specific C loss (mg CO2-C g' soil C d') (Figure 2-8). This form of expression normalizes soil respiration rate to the C content the substrate. Specific loss rate is equivalent to the first order rate constant (k) for the exponential decay model

C(t) = Coe-kt, [2-1] where C is concentration or mass as a function of time and CO is the initial quantity of C. Data points in Figure 2-8 can be expressed as k = d-' by multiplying the values by 10' (i.e. g'1 C g-' C d1 = d-l).
Mean specific CO2 loss rate for the T 1 (highly nutrient impacted site) soil profile and litter (0.89 mg C g-' C d') was approximately double the values for the C1 (0.50), C5 (0.39) and T5 (0.38) profiles. The highest rates in all profiles occurred in the litter layer and near the peat surface (Figure 2-8), indicating higher microbial activity in those strata. In the T5 and C5 soils, representing transitional and unimpacted sites, rates in the lower 20 cm of the profile were well below 0.5 mg C g-1' C d', with minimum rates in those profiles of 0.07 and 0.05 mg C g-' C d', respectively. However, the T5 profile was characterized by higher rates near the surface than the C5 profile. In the Cl profile, change in specific CO, loss rate from top to bottom was not as pronounced, that is, rate in the upper portion of the profile was not elevated to a great extent, however the sharp decrease in rate which occurred in T5 and C5 was observed only below about 22 cm in C1. In comparison, specific CO, loss rate in the T 1 profile was highly elevated near the surface, and although the rate dropped substantially below about 5 cm, remained above 0.2 mg C g' C d'l,through the entire profile.





50


SPECIFIC C LOSS (mg C02-C g-1 soil C d-1)

0 0.5 1.0 1.5 2.0 2.5 3.0 0 0.5 1.0 1.5 2.0 2.5 3.0

LITTER
0-2 2-4 4-6 6-8
8-10
10-12 12-14
14-16 16-18 18-20 20-22 22-24
E 24-26 Site Ti Site T5 %- 26-28 0.75 km 5 km I 28-30


o LITTER
0-2 O 2-4
4-6 6-8
8-10
10-12 12-14 14-16 16-18 18-20
20-22 22-24
24-26 Site C1 Site C5
26-28 3 km 10 km
28-30


Figure 2-8. Potential soil respiration (aerobic incubation) of litter and 2-cm depth
increments of selected WCA-2A microcosms, expressed as C-specific CO,
loss (i.e. soil organic C basis). Values represent means and standard error of
3 replicate incubations.





51


The range of values for specific CO, loss rate in the Ti profile (highest impact) is equivalent to a range of 0.0002 to 0.003 d-' for the first-order decay constant k, under aerobic conditions. Similarly, the corresponding values for k in the C5 profile ranged from

0.00005 to 0.001 d-'. Thus, the litter and surficial peat are substantially more degradable than the older peat lower in the profile. Based on these results, turnover time (for aerobic conditions) ranged from about 1 2.7 years for litter and 14 55 years for peat in the lower part of the 30-cm profiles.

For comparison to microcosm C flux measurements, the aerobic CO, loss rate in the laboratory incubation experiment was expressed as aereal loss rate, determined individually for litter and 2-cm depth increments (Figure 2-9). The resulting trends in C loss rate were similar to those observed for specific C loss. However, the differences in rate among the four sites were not as pronounced. Lower bulk density at sites closer to the inflow (e.g. site Ti) resulted in attenuated values when expressed on an aereal basis. Integration of the curves in Figure 2-9 yielded the sum total of potential, or aerobic, respiration in the soil profiles. The resulting values for the T1, T5, Cl and C5 profiles were 22.8, 11.2, 21.8 and 14.1 ptg C cm-2 1, respectively. These values may be considered "potential" or maximum rates under completely aerobic conditions, compared with total gaseous C (CO, + CH4) flux measured on flooded and partially drained microcosms. In general, potential (aerobic) respiration rates were about 3 times greater than those measured under partially-drained (water table at -15 cm) conditions.


Discussion


The purpose of this study was to evaluate the response of heterotrophic microbial activity in the litter and peat along a nutrient gradient, under flooded and drained conditions. A single set of soil microcosms, with sequential levels of treatment (water table) imposed on the soil cores, was used to determine the influence of transient changes in water table depth on soil respiration. A benefit of this experimental design was the





52


AREAL C LOSS (gg CO2-C cm2 h-1)
0 1 2 3 4 0 1 2 3 4
LITTER
0-2 2-4
4-6 6-8
8-10 10-12 12-14 14-16
16-18 18-20 20-22 22-24
24-26 Site T1 Site T5 S 26-28 0.75 km 5 km
28-30
I

W LITTER
0-2 2-4 Co 4-6
6-8
8-10
10-12 12-14 14-16 16-18
18-20 20-22 22-24 24-26
26-28 Site Cl Site C5
28-30- 3km 10 km


Figure 2-9. Potential soil respiration (aerobic incubation) of litter and 2-cm depth
increments of selected WCA-2A microcosms, expressed as CO2 loss on an
areal basis. Values represent means and standard error of 3 replicate
incubations.





53


elimination of spatial variability from statistical comparison of flux at various levels of water table. This required the assumption that changes observed in the response (flux) were a function of changes in water table, and not due to an unrelated, time-dependent factor. Measurement of flux during the "equilibration" period (flooded conditions), which was of greater duration than the subsequent period of draining, showed that microbial respiration activity had stabilized, that is, initial changes due to core transport or other disturbance were no longer occurring.

Vertical profiles of dissolved O, in the water column and litter layer of the

microcosms (Figures 2-6 and 2-7) were evidence of the potential for significant, though highly localized, oxidized zones in the litter layer. Light availability in the water column, and penetration into the litter layer, under field conditions is dependent on density of the plant canopy (including floating vegetation) and of plant litter within the water column. The oligotrophic sawgrass marsh, as well as the aquatic sloughs, in the interior portions of WCA-2A generally support a lower standing crop of live macrophytes and plant litter (Toth, 1987; Davis, 1991). Light availability in the water column is quite high, giving rise to extensive periphyton growth (SFWMD, 1992). It should be expected that the water column, and at least the upper portion of plant litter, would be substantially oxygenated during much of the day. On the other hand, assumptions that the water column, and especially the litter layer, in eutrophic areas of the marsh are often completely reduced (Belanger et al., 1989), should be qualified, according to results of the present study. Although the native periphyton has been displaced in the nutrient-enriched areas of WCA-2A, various types of non-native algal mats have been observed within the macrolitter (referring to size as well as origin), depending on light availability. In spite of the high 0, demand created by the relative abundance of C and nutrients in the plant litter, the O0 supplied by algal photosynthesis may be significant in terms of supporting aerobic microbial heterotrophs. The highly localized nature of oxygenation in the litter, created by steep gradients between pockets of O2 supply and demand makes quantification of the





54


extent of oxygenation difficult. Furthermore, spot measurements of dissolved O in the field, without consideration of spatial or temporal variability, could give very misleading information on 02 availability in the water column and litter layer.

The artificial aeration system utilized in the microcosm water columns created a relatively uniform level of oxygenation, both spatially and temporally, tending to average out regions of high and low concentration. Aeration was provided to the water column of the flooded microcosms to compensate for the lack of photosynthetic activity and prevent a condition of complete anoxia in the water column. As discussed earlier, photosynthesis in the microcosms would have presented numerous problems associated with accurate determination of CO2 production. Flux of CO2 from the calcareous peat could have been drastically affected by shifting of the carbonate equilibrium as a response to depletion of CO2 by photosynthetic activity.

Presumably, constant aeration of the water column greatly enhanced decomposition of the extensive, and relatively labile, plant litter and algal detritus. The potential contribution of this component is considerable, as demonstrated during the controlled incubations in the laboratory. As a result, CO, flux from flooded microcosms probably was more representative of maximum daytime levels under field conditions. Taking into account diurnal variation in dissolved 02 concentration in the water column and litter, the mean daily rate of respiration in the field was probably considerably lower than indicated by microcosm measurements. Therefore, flux measurements made during aeration of the microcosm water column should be considered maximum, or potential, rates sustainable in the field for short periods of time and are useful primarily for comparison among sites along the nutrient gradient and for scaling rate parameters in an ecosystem model.

Methane flux data exhibited much greater variability than CO2 flux measurements. Because of the absence of emergent macrophytes, an important conduit for CH, emissions from wetlands (Schutz et al., 1991), and the low solubility of CH4 in water, ebullition was the primary mechanism for efflux of CH4 from the microcosms. Since ebullition is an





55


intermittent process, CH4, measurements taken over short time intervals would presumably yield highly variable results. A secondary, albeit important, source of uncertainty inherent in measuring CH, flux across an anaerobic-aerobic interface is the loss of CH, through oxidation to CO,. This occurs in aerobic regions of soil and floodwater through the action of methane oxidizing bacteria which utilize CH4 as an energy source and 02 as the terminal electron acceptor. Thus, the amount of CH4 collected above the microcosm soil or floodwater surface may substantially underestimate the actual CH4 production in the soil. Happell and Chanton (1993) estimated that an average of 46% of CH4 produced in a north Florida swamp forest soil was oxidized to CO, near the sediment surface.

Conversely, it would be incorrect to assume that all of the CO, released from the microcosms is attributable to the activity of aerobic heterotrophs. Oxidation of methane to CO2 by methanotrophic bacteria is potentially a major sink for methane in wetlands (Kelly and Chynoweth, 1979; Schipper and Reddy, 1996). Furthermore, CO, is also a secondary product of methanogenic bacteria, and is evolved in varying proportion to CH4 during methanogenesis (Oremland, 1988). An additional source of CO, production in anaerobic soils or microsites is sulfate reduction. Normally a more significant process in salt marshes due to elevated SO42- levels (Howarth, 1993), it may be a major pathway of anaerobic decomposition in WCA-2A, given the high concentrations of SO42- measured in the floodwater (Schipper and Reddy, 1994).

Measurements of CH4 flux in wetlands have generally indicated lower rates under drained or partially drained conditions, because of the decreased number anaerobic sites and increased consumption (oxidation) of CH4 by aerobic soil microorganisms (Happell and Chanton, 1993; Moore and Dalva, 1993). This trend was not supported by results from the present study. Mean flux of CH4 did not change significantly under drained conditions, between water table depths of 0 and 15 cm (Figure 2-6). It is possible that production of CH4 decreased during that time, but the decrease was offset by increased emission of trapped CH4 bubbles from the lower, saturated portion of the profile, as the





56


pressure head decreased with drainage. This type of response to changing water table might be expected to occur under transient conditions of drainage.

Soil respiration in the microcosms increased significantly along the nutrient gradient in WCA-2A. from areas of low to high impact (oligotrophic to eutrophic). Analysis of total N and P on four selected soil profiles indicated greater P enrichment toward the S-10C inflow, while total soil N was relatively unchanged. Previous studies in WCA-2A have shown a well-defined soil and water P gradient south (downflow) of the S-10 inflows (Koch and Reddy, 1992 ; DeBusk et al., 1994). Increased accumulation of both N and P has occurred in the impacted regions of WCA-2A, through accelerated accretion of organic matter (Craft and Richardson, 1993; Reddy et al., 1993). However, enrichment of the peat with N (higher concentration on a mass basis) has not occurred to a great extent, while considerable P enrichment has occurred (DeBusk et al., 1994). Thus, it would appear that response of C flux (soil respiration) to distance from inflow would be the result of P enrichment near the inflow. Data from the laboratory incubation study support this argument.

A significant relationship (a = 0.05) was found between total P concentration and aerobic soil respiration, but respiration and total N content were not significantly correlated. However, a linear model of respiration as a function of both total P and total N showed a significant (ar = 0.05) interaction between the two factors. It is possible that N becomes a limiting nutrient for microbial activity in regions of excessive P enrichment. The multiple linear regression model, however, explained only 46% of the variability in soil respiration, in part because the relationship between total P and respiration was not linear. Respiration was better described as a power function of total P (total P raised to a power), such that a log-log plot of respiration vs. total P yielded a straight line. This model accounted for 67% of the variability in respiration. The improved fit using this type of model probably results from the log-normal distribution of both sets of data. However, visual inspection of the data suggests that respiration may actually be asymptotic at higher levels of total P. This





57


also may suggest a co-limiting factor for microbial activity, possibly N. A portion of the variability in respiration not explained by variation in total P may also be due to differences in C availability, since the data set from the laboratory study include plant litter and peat from a wide range of depth and therefore age and biodegradability.

A major effect on soil respiration in the microcosms was water table. Increase in respiration as the water table was lowered (disregarding flooded conditions, which are not directly comparable) was highly significant. Increased soil respiration with drainage (which results in increased 02 availability) has been reported in other studies. For example, a linear increase in CO, flux with decreasing water table to a depth of 50 cm was reported for undisturbed cores of Montverde muck (sawgrass peat) from the Everglades (Volk, 1973). Similarly, CO2 flux increased linearly with decreasing water table to a 40 cm depth in intact peat cores from Canadian wetlands (Moore and Dalva, 1993).

Of particular interest in this study, however, is the interactive effect of water table (02 availability) and distance from inflow (P enrichment) on respiration (Figure 2-7). A positive response of respiration to decreasing water table would result in an upward shift in the plot of C flux vs. distance, with no change in slope. However, in this case, both a shift along the Y-axis and a significant increase in slope were observed, clearly indicating an interactive effect. Examination of the plots in Figures 2-8 and 2-10 provides more insight into this interaction. As discussed previously, the aerobic incubation of litter and peat subsamples from selected microcosms was considered as a measure of potential respiration (not limited by 02 availability). Specific C loss (Figure 2-8) provided a means for direct comparison of relative biodegradability, since these values were based only on the organic component of the peat or litter, and variation in respiration due to availability of 0, had been removed. The trend of low biodegradability in the older peat (lower depth) and increasing near the surface was common to all profiles, but more exaggerated in profiles from nutrient-enriched areas. This is well-defined in Figure 2-10, which shows cumulative effect of this trend as one moves down the profile. It also relates potential respiration rate to





58


POTENTIAL RESPIRATION (gg C cm-2 h-1)
0 5 10 15 20 25 0 5 10 15 20 25
LITTER
0-2 Site T1 Site T5
2-4 0.75 km 5 km
4-6 6-8
8-10
10-12 12-14 14-16 16-18
18-20
20-22 22"24 E 24-26
26-28"
28-30


o LITTER
0-2
S 2-4- Site C1 Site C5 O 2-4 S 4-6 3 km 10 km

6-8

10-12 12-14

16-18 18-20
20-22 22-24
24-26
26-28
28-30 :


Figure 2-10. Potential soil respiration (aerobic incubation) of litter and 2-cm depth
increments of selected WCA-2A microcosms, expressed as cumulative, or
depth-integrated CO, loss. The bottom value represents potential respiration
for the entire litter and soil profile.





59


surface area, to account for differences in bulk density at these sites. It is apparent that relatively labile, or decomposable, organic matter has accumulated to a greater depth at the more highly nutrient-impacted sites. Thus, not only has there apparently been a Penrichment effect, but also a P-accumulation effect, that is, a larger accumulation of the more recent, enriched organic matter associated with nutrient loading during the past 3 decades. In support of this concept are estimates of accretion based on 137Cs dating of intact cores which show significantly increased rates of peat accretion closer to the inflows in WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993).

Potential soil respiration rate (determined from aerobic incubations) expressed on a cumulative soil area or volume basis, greatly exceeded actual measured rates in corresponding intact microcosms, even under partially drained conditions. Whole-column gaseous C flux from microcosms T1, T5, Cl and C5 were 5.8, 5.5, 4.9 and 3.8 jgg C cm2 h', respectively. In comparison, potential rates for combined depths, integrated over the entire 30 cm profile, were 22.8, 11.2, 21.8 and 14.1 gLg C cm2 h-', for the same sites. If potential respiration only in the upper 15 cm of the profile (including litter) is considered, extrapolated values are 14.8, 8.5, 16.5 and 12.1, still substantially higher than actual rates. This suggests that 02 availability was limited in the soil above the water table. Observations made during the experiment support this possibility, since it was apparent that the peat was saturated above the water table. This saturated zone, or capillary fringe, was approximately 2-3 cm thick, by visual inspection, thus 02 availability might be expected to be close to zero in this portion of the profile. Undoubtably, a substantial number of microsites in the soil above the capillary fringe were also water-filled, due to the hysteresis effect commonly observed during soil drainage. The water content of the soil above the capillary fringe would be expected to decrease with time. Thus, the response of soil respiration rate to soil drainage is probably time-dependent. Time was not included as a variable in the present study, although the significance of time as a factor in respiration response to drainage was considered beforehand. Instead, the intent of the microcosm study was to approximate





60


transient hydrologic conditions for measuring the response of soil respiration to drainage or drought. This was also a consideration for imposing sequential levels of treatment on the same set of microcosms, as discussed earlier.

In comparison to C flux measured during the present study, the average CO, flux from intact (80 cm deep) peat columns from 3 Canadian wetlands ranged from 1.9 to 11.7 jtg C cm-2 h- under saturated conditions and 9.0-15.8 .g C cmn2 h' for a water table depth of 40 cm (Moore and Dalva, 1993). Total C flux from undisturbed cores of from the Everglades (Montverde muck, a sawgrass peat) averaged 8.4 jtg C cm2 h-' for a water table depth of 15 cm (Volk, 1973). Potential respiration (aerobic incubation) integrated over the upper 30 cm of peat from short pocosin, tall pocosin and gum swamp sites in North Carolina averaged 6.5, 12.5 and 11.5 tg CO,-C cm-2 h-' (Bridgham and Richardson, 1992). The sampling sites ranged from nutrient-deficient to moderately nutrient-rich, respectively. Using in situ measurement of CO, and CH4 flux from north Florida swamp forests, Happell and Chanton (1993) estimated the average flux of CO, under flooded conditions to be approximately 3-5 times less than under drained conditions. After adjusting for estimated root respiration, the average rate of C mineralization under flooded conditions, based on combined CO2 and CH, fluxes, was 11.9 mol C m-2 y', equivalant to

1.6 glg C cm-2 h. In situ measurement of CH, flux from a WCA-2A slough community, using a static chamber method, yielded an average value of 0.09 g m2 d', equivalent to

0.28 jg C cm2 h' (Schipper and Reddy, 1994)

Living macrophytes were not included as part of the microcosms, in order to differentiate respiration of microbial decomposers from plant respiration. It must be realized, when extrapolating results of this study to field conditions, that macrophytes interact closely with soil microorganisms, therefore microbial processes in the microcosms were certainly affected to some degree. For example. wetland macrophytes provide various amounts of organic C and 0, to the rhizosphere, both of which affect the metabolism of heterotrophic microbes, at least near the rhizoplane (root surfaces) (Reddy and D'Angelo,





61


1994). On the other hand, living roots which were severed during coring might have become a source of labile C, though not a source of 02, for microbial decomposers. Periphyton presented an additional problem, since neither algal photosynthesis nor respiration were desirable during C flux determinations. However, algae was considered to be physically inseparable from the litter layer, especially fine particulate organic detritus, therefore no attempt was made to remove it. Therefore, dead algal biomass resulting from total shading of the microcosms during flux measurements was another potential source of labile C for decomposers. However, the most labile pool of organic C originating from recently dead plant biomass was probably mineralized during the period of about 2 weeks between the onset of shading and measurement of C flux.

Summary and Conclusions


Based on results of this study, the following conclusions were made:

Photosynthetic activity of the native periphyton in unimpacted areas and filamentous algae in the extensive litter layer of impacted areas may provide significant oxidation of the litter layer during the daytime, potentially enhancing decomposition.

Soil respiration increased with lowering of the water table, in direct proportion

(linear response) to depth of the water table; however, magnitude of the response increased with the degree of nutrient enrichment. Response of soil respiration to nutrient enrichment was significant under flooded and highly drained conditions.

There was a statistically significant interaction between water table depth and

nutrient enrichment as factors affecting soil respiration rate, however respiration rate was more dependent on depth of the water table.

Potential soil respiration was found to be a good indicator of actual respiration, and was significantly correlated with soil total P concentration.













CHAPTER 3
TURNOVER OF ORGANIC CARBON POOLS ALONG THE WCA-2A NUTRIENT GRADIENT


Introduction


A primary characteristic of wetlands which distinguishes them from upland ecosystems is their propensity for accumulating organic carbon (C). Organic C accumulation in wetlands is the net result of primary production (C fixation) and decomposition (C mineralization). In wetlands with extended hydroperiod, decomposition of dead plant material proceeds at a reduced rate, leading to substantial accumulation of organic C (as organic matter), which may include extensive peat deposits.

Decomposition of organic matter is governed by the chemical composition of the

substrate and external, or environmental, factors. Among the more important environmental factors are temperature, moisture, nutrients and electron acceptors (Swift et al., 1979; Heal et al., 1981; Reddy and D'Angelo, 1994). Unlike terrestrial ecosystems, decomposition in wetlands is generally not moisture-limited and is frequently electron acceptor-limited. In the latter case, the primary controller of decomposition is, of course, O,. In many freshwater wetlands, however, alternate electron acceptors for anaerobic microbial respiration, such as NO-, Mn*, Fe 3 and SO4=, are often in short supply relative to available organic C, leaving methanogenesis as the principal mode of respiration (Westermann, 1993).

Nutrient availability affects decomposition rate through its effects on microbial growth. Although nutrient loading is typically greater in wetlands than in uplands due to location within the landscape, nutrient availability may be low relative to the pool of available organic C in wetlands (Reddy and D'Angelo, 1994). Nitrogen (N) and phosphorus (P) both have been identified as microbial growth-limiting nutrients in


62





63


wetlands (Westermann, 1993). Nitrogen, unlike P, may be lost from wetlands through microbial metabolism via denitrification, as well as through ammonia volatilization (Reddy and D'Angelo, 1994).

In addition to environmental conditions, decomposition rate is significantly affected by chemical and physical composition of the organic substrate (Swift et al, 1979; Heal et al, 1981). The term "substrate quality" generally refers to the availability of C compounds and associated nutrients for microbial utilization (Heal et al., 1981; Heal and Ineson, 1984). Lignin and cellulose fractions are generally considered to be key components of the "C quality" of an organic substrate (Colberg, 1988; Moran et al., 1989). Lignin is more resistant to breakdown than cellulose, therefore substrate cellulose content decreases more rapidly during decomposition (Colberg, 1988; Melillo et al., 1989). Degradation of lignin and cellulose is slower under anaerobic conditions, although the relative decay rates for each remains approximately the same (Benner et al., 1984). Initial substrate composition has been used as a predictor of decomposition rate (Swift et al., 1979). In several cases, initial lignin content and lignin:N ratio of plant tissue were shown to be highly correlated with decomposition rate (Godshalk and Wetzel, 1978; Berg and Staaf, 1981; Melillo et al., 1982).

The ligno-cellulose index (LCI) was proposed as a measure of substrate C quality along the "decay continuum" of plant leaves to soil organic matter (Melillo et al, 1989). Increased LCI during decomposition reflects the increasing concentration of lignin relative to the total ligno-cellulose content of the organic substrate, thus the substrate becomes increasingly resistant to decomposition. In addition to the relative increase in plant lignin, an accumulation of lignin derivatives generated as by-products of microbial metabolism (i.e. humus) occurs during decomposition (Zeikus, 1981).

Microbial decomposers, primarily bacteria and to a certain extent fungi and

protozoa, play the lead role in C cycling and energy flow in wetlands (Benner et al., 1984; Wetzel, 1984; Westermann, 1993). Although they represent only a small fraction of the





64


total C and organic matter in soils, microbial decomposers are responsible for processing nearly all of the organic C produced in the ecosystem (Jenkinson and Ladd, 1981; Van Veen et al., 1984). Research in terrestrial ecosystems has shown that the microbial biomass also constitutes a major C sink, in that it represents a significant portion of the "active" organic C (Paul and van Veen, 1978). Repeated cycling of organic C through the microbial biomass results in loss of organic C from the detrital pool via aerobic (CO, loss) or anaerobic (CO, and CH4 loss) microbial respiration. This same process also results in changes in substrate composition as original plant material is lost and by-products of microbial metabolism accumulate (Swift, 1982; Heal and Ineson, 1984).

The current study examines the influence of nutrient loading on selected microbial processes regulating turnover of organic C pools in a northern Everglades marsh. The Everglades encompasses a variety of wetland ecosystems which were historically adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred pri ly throughrainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974) (Figure 3-1). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959).

Recent development of the Everglades and surrounding watershed has created changes in nutrient loading and hydrology (SFWMD, 1992). Most significantly, a large area of the northern Everglades was drained and converted to agricultural production during the first half of this century. This area of sugar cane, vegetable and sod farming is referred to as the Everglades Agricultural Area (EAA). The remainder of the northern Everglades was divided into three Water Conservation Areas (Figure 3-1) in the 1960s, for water storage and flood control. Water level within the WCAs is controlled by a system of levees,





65










Lake

West Palm
Beach MAP VIEW




\ Fort Luderdale Miami I /






8-10C





WCA-2A 0100







0 5 10 Kilometers WCA-2B








Figure 3-1. Site map for WCA-2A study area, showing locations of sampling sites.
Coordinates for sampling sites are listed in Table 3-1.





66


pumps and floodgates. Currently, the Everglades consists of the WCAs to the north and Everglades National Park to the south.

Drainage of the EAA has resulted in widespread oxidation of the organic soil and concomitant mineralization and leaching of organically-bound nutrients. As a result, nutrients from organic soil min lization, a ith additional nutrients from fertilizers, have been transported via drainage canals toward the WCAs for approximately 30 years. Loading of agricultural drainage water into the WCAs has resulted in nutrient enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus (P) enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod, in the encroachment of cattail (Typha domingensis Pers.) and other rapidly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward and Ornes, 1983; Toth, 1987, 1988).

Accelerated nutrient loading in northern WCA-2A (Figure 3-1) during the past three

decades has created a distinct nutrient (especially P) gradient in water, soils and plant tissue L (Davis, 1991; Koch and Reddy, 1992; DeBusk et al, 1994). Changes in species composition of periphyton and macrophyte communities, along with an overall increase in net primary productivity have been documented along this gradient (Davis, 1991; SFWMD, 1992). Soil dating by analysis of '37Cs peaks has indicated that peat accumulation rate has increased in nutrient-enriched areas of WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993).

The main objective of this research was to determine the effect of nutrient

enrichment on turnover of organic C pools along the WCA-2A nutrient gradient. A further objective was to examine the relationships between size of the microbial biomass C pools and turnover time of associated organic C pools. It is hypothesized that turnover time for major C pools increases (decomposition rate decreases) downgradient from the inflow of nutrient-laden water. It is also hypothesized that turnover time increases in successively older organic C pools.





67


Materials and Methods


Site Description


Field study sites were located in WCA-2A, a 447 km2 region of the northern Everglades (Figure 3-1). Surface water flows into WCA-2A from the Hillsboro Canal through the four S-10 water control structures and from the North New River Canal through the S-7 pump station. Most of the hydraulic loading is through the S-1OC and S10D structures into the northern portion of WCA-2A. The general direction of flow is from north to south. Water depth is usually less than one meter, and varies considerably, both seasonally and year-to-year, with occasional dry periods (SFWMD, 1992; personal observations). The bulk of the surface outflow is through through three control structures at the south end of WCA-2A, into WCA-3 (Figure 3-1).

Soil in WCA-2A consists of Everglades and Loxahatchee peats (Gleason et al., 1974). Everglades peat, the most common soil in the Everglades, is associated with the sawgrass marsh community. It is dark brown, finely fibrous to granular, with circumneutral pH, relatively high N content and low SiO,, Fe and Al content. Peat depth in WCA-2A ranges from about I to 2 m, and age of basal peats is estimated to be 2000 to 4800 yr. Beneath the peat lies a bedrock of Pleistocene limestone, with intermediate layers of calcitic mud, sandy clay and sand in several areas (Gleason et al., 1974).

The primary sources of nutrient loading to WCA-2A are the S-10 structures which convey water from the Hillsboro Canal and WCA-1 (Figure 3-1). A distinct gradient of N and, most significantly, P enrichment in water, plants and soil has formed between the high-nutrient region adjacent to the inflows and the low-nutrient interior marsh of WCA-2A (Koch and Reddy, 1992; SFWMD, 1992; DeBusk et al., 1994). A vegetation gradient coincides with the nutrient gradient; most notable is the gradient from sawgrass marsh with scattered aquatic slough in the interior to cattail and mixed emergents near the inflows. The





68


vegetation gradient was divided into three discrete categories for the purposes of the current study: cattail-dominated, sawgrass-dominated and mixed cattail and sawgrass (Figure 3-1).

Ten field sampling sites were established along the nutrient and vegetation gradient on a 10 km transect extending from the S-10C inflow into the interior marsh. The sites, numbered 1 through 10, were located at distances of 0.07, 0.3, 0.6, 1.2, 2.0, 2.9, 3.9,

4.8, 6.6 and 9.8 km downstream from the inflow (Table 3-1; Figure 3-1). Plant and Soil Sampling


Live and standing dead (attached to plant) cattail and sawgrass leaves were collected at sites along the sampling transect in March and June 1995. Live cattail leaves were collected at sites 1, 6 and 8 and live sawgrass leaves at sites 6, 8 and 10 on March 8. In addition, dead cattail leaves were collected at sites 1 through 8 and dead sawgrass leaves at sites 6 through 10. On June 6, live and dead plant leaves were collected from cattails at sites 1 through 6 and from sawgrass at sites 5 through 10. Five entire leaves were removed per plant, and repeated for three different plants at each site. Leaf samples were dried in a forced-draft drying room at 60 "C, then all leaves in each sample were cut into pieces of about 2 cm length and mixed thoroughly to produce a homogeneous composite sample. All live and standing dead plant samples were analyzed for total C, N and P content. Lignin and cellulose content were determined for selected samples from the March 1995 sampling event. These included live cattail from sites 1 and 6, live sawgrass from sites 6 and 10, dead cattail from sites 1, 4 and 6 and dead sawgrass from sites 6, 9 and 10.

Soil cores were also obtained during the June 1995 field sampling event. The plant litter layer on the soil surface was sampled first, by placing a short section of 15 cm diameter PVC pipe over the sample area and manually transferring the litter contained within the pipe to a zippered plastic bag. A serrated knife was used to cut through the litter, around the inside perimeter of the pipe, to precisely delineate the sample area. Next, a simple coring apparatus consisting of 7.6 cm diameter aluminum pipe was used to obtain





69









Table 3-1. Locations of sampling sites in WCA-2A, and
distance from S-10OC inflow.
Distance
Site Latitude N Longitude W from inflow
deg min deg min km

1 26 22.09 80 21.07 0.1 2 26 21.98 80 21.09 0.3 3 26 21.81 80 21.12 0.6 4 26 21.53 80 21.20 1.2 5 26 21.05 80 21.21 2.0 6 26 20.53 80 21.27 2.9 7 26 20.02 80 21.37 3.9 8 26 19.52 80 21.39 4.8 9 26 18.51 80 21.46 6.6 10 26 16.81 80 21.48 9.8





70


intact samples of the top 30 cm of the soil profile. The coring tube was pushed into the soil within the area from which litter had been removed. A serrated knife was used to cut through the fine root mat at the top of the soil profile, facilitating penetration of the coring tube without compaction of the soil. When additional force was required, the coring tube was hammered into the soil after being fitted with a solid aluminum plug (with an air vent) to distribute the force of impact from the mallet. The coring tube was then excavated, and the intact soil core extruded through the top using a plunger apparatus. The upper 10 cm of the peat profile was separated as it was extruded, and placed in a zippered plastic bag. Similarly, the 10-30 cm layer of peat was extruded and placed in a separate bag. The 0-10 and 10-30 cm layers of the peat profile were designated as "surface peat" and "buried peat".

The above procedure was repeated three times at each sampling site, within a radius of approximately 5 m. Samples contained in the sealed plastic bags were immediately placed on ice and transported to the lab within 24 hours. After removal of live roots from the samples, wet weights were recorded for each, then replicate samples were thoroughly mixed to create a single composite sample. Subsamples of litter and peat were dried to constant weight in a forced-draft oven at 60 'C, for determinations of moisture content and dry bulk density. The composited samples were stored in leak-proof polypropylene jars in a refrigerator at 4 'C.


Plant and Soil Analysis


Total C and N analysis was performed on dried, finely ground (< 0.2 mm) samples using a Carlo-Erba NA-1500 CNS Analyzer (Haak-Buchler Instruments, Saddlebrook, NJ). Total P analysis was performed on separate subsamples following combustion (ashing) at 550 "C for 4 h in a muffle furnace and dissolution of the ash in 6 M HCl (Anderson, 1976). The digestate was analyzed for P using the automated ascorbic acid method (Method 365.4, USEPA, 1983). Lignin and cellulose content were determined by a standard procedure using acid-detergent and H2SO, extractions (AOAC, 1990).





71


Litter and peat samples were analyzed for C in the microbial biomass using the

chloroform fumigation-extraction (CFE) technique (Horwath and Paul, 1994), with a slight modification. Field-moist samples (ca. 0.5 g dry mass) were fumigated in a vacuum dessicator with ethanol-free chloroform, which was contained in a beaker next the samples. Immediately prior to fumigation, 0.5 mL of chloroform was added directly to each sample, to enhance distribution of chloroform within the wet samples (Ocio and Brookes, 1990). The remainder of the fumigation and extraction process was performed according to Horwath and Paul (1994). Triplicate fumigated and non-fumigated samples were extracted with 25 mL of 0.5 M K2SO4, centrifuged and the supernatant filtered though glass fiber filters (Gelman A/E, Gelman Sciences, Ann Arbor MI) using a vacuum filtration system. The ratio of extractant to dry soil was increased over the recommended 5:1 (w/w) because of the extremely high organic content of the soil and litter. The filtered extracts were analyzed for total organic C (soluble organic C) on a Dohrmann DC- 190 TOC analyzer (Rosemount Analytical Inc., Santa Clara, CA). Microbial biomass C was calculated from the difference in IKSO4 extractable C between fumigated and non-fumigated samples. A correction factor (kEc), which accounts for the efficiency of the fumigation process, is generally used to obtain a direct estimate of microbial C from the flush of extractable C following fumigation (Horwath and Paul, 1994). A value of kEc = 0.37 was used in this case, based on extensive calibration previously carried out for organic soils (Sparling et al., 1990).


Microbial Respiration


Microbial respiration associated with decomposing standing dead material, plant litter and peat was measured for estimation of C mineralization rate. Aerobic respiration was measured for standing dead leaves using bottle incubations and measurement of accumulated headspace CO2. Triplicate 1 g (dry weight) subsamples of standing dead cattail and sawgrass leaves from all sites and collection dates were placed in separate 160 mL





72


serum bottles. Deionized water (5 to 7 mL as required) was added to the dried plant material to re-wet the substrate for microbial activity. Each bottle was stoppered with a sleeve-type rubber septum and incubated in the dark at 25 C. After a 24-hour preincubation period, headspace gas was sampled in the bottles. Sampling was repeated after 12 and 24 hours, and subsequently every 24 hours for a total of four days of sampling. Headspace gas was sampled (1 mL) using zero dead volume 1 mL insulin syringes (Becton Dickinson, Lincoln Park, NJ). Syringe needles were inserted into a rubber stopper for short-term storage of samples prior to analysis on a gas chromatograph

(GC) for CO2. Increase in CO, in the bottle headspace was linear over the four-day period. Rate of CO2 evolution was calculated from the slope of the best-fit linear regression line.

Soil (litter and peat) respiration was measured using an air flow-through system (respirometer) which traps evolved CO2 on a continuous basis (Zibilske, 1994). Use of a continuous flow apparatus for respiration measurement is recommended for calcareous soils, such Everglades peat, to avoid problems associated with retention of microbiallyevolved CO2 as bicarbonate (Martens, 1987). The respirometer consisted of an air supply (compressed air cylinders), CO, scrubber (2N NaOH trap), incubation vessels and a separate CO2 trap (0.05 to 0.1 N NaOH) for each incubation vessel. These components were connected by plastic (PVC) tubing and a gas manifold to create a continuous flow of CO2-free air through each incubation vessel and into the individual CO, traps. Incubation vessels were fashioned from 250 mL Mason jars with air-tight lids. The jar lids were modified to accomodate plastic fittings with valves for controlling gas inflow and outflow. Air flow rate through each incubation vessel was maintained at 25 mL min'.

The NaOH traps provided continuous collection of CO, evoloved during microbial respiration. NaOH in the traps was changed periodically, as determined by the rate of CO, accumulation. Free (remaining) NaOH in the traps was titrated with standardized HCI, following addition of BaC1, to precipitate the Na2CO3 in solution as insoluble BaCO3, to





73


determine the amount of NaOH which had reacted with CO,. Molar quantity of CO, evolved during the incubation period was determined stoichiometrically (Zibilske, 1994).

Subsamples (10 g wet weight) of litter, surface peat (0-10 cm depth) and buried peat (10-30 cm depth) were placed in 50 mm diameter plastic petri dishes. A thick glass fiber prefilter was placed beneath each sample to increase aeration of the sample by maintaining drained, yet moist, conditions in the sample. The petri dishes containing the samples were placed inside the Mason jars and incubated in the dark at 25 "C. A 48 hour pre-incubation period was found to be sufficient time to achieve stabilization of respiration rate. Samples were incubated for one week following pre-incubation, then removed from the incubation chambers, dried at 60 "C in a forced-draft oven and weighed to determine sample dry mass. The incubation was performed in triplicate through successive incubation of the entire sample set.

Anaerobic respiration was measured in the same fashion, with the following

modifications. The source air was replaced by N2 gas (prepurified grade) and sample size was increased to 50 g wet weight. Incubation period was shortened to three days, following a 48 hour pre-incubation. A preliminary study suggested that a longer incubation period would result in depletion of alternate electron acceptors, and would therefore not reflect conditions in the field at the time of sampling. At the end of the three day incubation period the inlet and outlet ports of each incubation vessel were closed, and the NaOH traps were removed and titrated as described above. Headspace gas in each vessel was sampled immediately following cessation of gas flow through a rubber septum in the lid, using 1 mL insulin syringes. Syringe needles were inserted into a rubber stopper for short-term storage of samples prior to analysis on a GC for methane (CH,). Sampling was repeated after 6 hours, and gas samples were analyzed for CH4. Preliminary studies showed that accumulation of methane in the incubation vessels was linear over time.

Sample gases were analyzed on a Hewlett-Packard 5840A GC (Hewlett Packard, Avondale, PA), using thermal conductivity (TCD) and flame ionization detectors (FID) for





74


CO2 and CH4 analysis, respectively. For CO, analysis, a Poropak N (Supelco, Bellefonte, PA) column was used, with He as a carrier gas. Oven, injector and detector temperatures were set to 60, 140 and 200 'C. For CH4 analysis, a Carboxen 1000 (Supelco, Bellefonte, PA) column was used, with a N, carrier gas. Oven, injector and detector temperatures were 120, 120 and 200 "C.


Statistical Analyses


Simple and multiple linear regression procedures were used for evaluating

continuous variables (continuous Y vs. continuous X), and analysis of variance (ANOVA) procedures were used for categorical data (continuous Y vs. nominal or ordinal X). All statistical procedures were performed using JMP software (SAS Institute, Cary, NC).


Results


Chemical Analysis of Substrate


Total C content was similar in plant tissue, standing dead, litter, surface peat and buried peat (Tables 3-2 and 3-3). Average values for the ten sampling sites were slightly but significantly higher (a = 0.05) in standing dead than in peat. Total N content increased significantly from standing dead to litter to peat. Total P content was significantly lower in standing dead than live plant tissue, litter and surface (0-10 cm) peat (Tables 3-2 and 3-3; Figure 3-2). Lignin content increased significantly (a = 0.05) from the plant tissue to peat along the decay sequence, while cellulose content decreased.

Average total C content of live sawgrass leaves was significantly higher (a = 0.05) than cattail. Conversely, total N and P concentrations were significantly higher in live tissue of cattail than in sawgrass. Differences between means of C, N and P content of standing dead material in cattail and sawgrass were not significant. Mean total N content of






75


Table 3-2. Chemical analysis of live and standing dead plant tissue collected from 10
sites along the WCA-2A nutrient gradient. Each sample was a composite of
leaf samples from five plants.
Plant Sample Sample
Site type type date Total C Total N Total P Lignin Cellulose
- Rmk- kj ---- % ---1 Cattail Live Mar-95 402 11.4 918 5.0 36.5 Jun-95 409 15.7 1491 Dead Mar-95 466 8.0 339 15.3 38.8 Jun-95 456 4.1 266
2 Cattail Live Jun-95 438 10.0 1067 Dead Mar-95 460 7.3 343 Jun-95 452 3.6 262 3 Cattail Live Jun-95 451 8.8 982 Dead Mar-95 463 5.7 375 Jun-95 465 4.7 275 4 Cattail Live Jun-95 436 10.3 1166
Dead Mar-95 457 5.3 277 10.0 42.0 Jun-95 454 3.0 258 5 Cattail Live Jun-95 422 9.5 1202 Dead Mar-95 457 6.4 373 Jun-95 444 4.1 305 Sawgrass Live Jun-95 475 5.9 527 Dead Jun-95 457 3.9 225
6 Cattail Live Mar-95 432 8.0 844 6.2 34.7 Jun-95 460 8.2 662 Dead Mar-95 463 6.0 310 11.4 42.9 Jun-95 451 3.4 249 Sawgrass Live Mar-95 458 7.8 487 7.4 33.4 Jun-95 464 7.3 673 Dead Mar-95 466 4.3 487 12.9 34.3 Jun-95 470 5.4 370 7 Cattail Dead Mar-95 468 4.7 315
Sawgrass Live Jun-95 459 7.8 844 Dead Mar-95 457 5.3 478 Jun-95 463 3.4 227 8 Cattail Live Mar-95 422 8.1 684 Dead Mar-95 468 4.0 205 Sawgrass Live Mar-95 463 6.6 478 Jun-95 461 8.2 958 Dead Mar-95 456 5.7 241 Jun-95 462 3.2 159 9 Sawrass Live Jun-95 475 5.2 322
Dead Mar-95 454 4.0 140 16.1 35.5 Jun-95 467 3.2 106 10 Sawgrass Live Mar-95 456 6.0 242 7.0 33.3 Jun-95 479 4.4 157 Dead Mar-95 447 4.6 61 14.2 38.1 Jun-95 459 3.6 66





76


Table 3-3. Physical and chemical analysis of soil cores collected June 6, 1995
from 10 sites along the WCA-2A nutrient gradient. Each value
represents a single composite sample of three cores from each site.
The soil profile was sampled in three depth increments: litter layer,
surface peat (0-10 cm) and buried peat (10-30 cm).
Depth Bulk Total Total Total
Site increment density N C P Lignin Cellulose
g cm - g kg' - mg kg' %

1 Litter 26.6 433 1582 33.7 17.2
0-10 cm 0.053 28.2 376 1497 39.0 12.7 10-30 cm 0.117 25.4 358 1195 47.7 7.9
2 Litter 26.7 463 1291
0-10 cm 0.043 28.6 440 1337 10-30 cm 0.063 30.3 459 1172
3 Litter 25.7 454 1461
0-10 cm 0.044 30.7 433 1622 10-30 cm 0.065 24.4 388 880
4 Litter 25.0 454 1833 30.2 22.9
0-10 cm 0.034 24.8 439 1416 37.1 18.5 10-30 cm 0.092 31.5 451 982 51.7 13.9
5 Litter 27.8 453 1743
0-10 cm 0.041 28.3 443 1353 10-30 cm 0.094 32.2 453 611
6 Litter 20.1 467 1084 37.8 22.4
0-10 cm 0.057 29.1 451 1060 52.2 14.9 10-30 cm 0.086 35.8 457 284 57.0 12.2
7 Litter 17.5 452 928
0-10 cm 0.056 27.0 439 1146 10-30 cm 0.067 31.3 453 466
8 Litter 19.6 443 1038
0-10 cm 0.056 26.1 440 905 10-30 cm 0.106 26.3 422 279
9 Litter 12.2 452 345 31.1 27.3
0-10 cm 0.064 28.8 448 696 44.9 18.8 10-30 cm 0.088 29.0 453 269 56.9 14.2
10 Litter 16.4 410 275 32.8 21.5
0-10 cm 0.057 28.9 431 475 43.9 16.1 10-30 cm 0.077 24.7 471 236 58.8 14.5





77


standing dead was significantly higher in samples collected in March than in June (a =

0.05). Similarly, total P content of standing dead was higher in March, but the difference was not significant at a = 0.05.

Total N content of live plant tissue decreased significantly with increasing distance from the S-10C inflow, according to linear regression analyisis (a = 0.05). When cattail and sawgrass were considered separately, a similar decreasing trend was observed, but was not significant. Thus, plant type was probably the main factor affecting plant tissue N along the transect, since the N content of live cattails was found to be higher than sawgrass, even at the same distance from the inflow (Table 3-2). Total N content of the dead organic matter compartments did not vary significantly along the sampling transect.

Total P content of live and dead plant material, plant litter and both peat layers

decreased significantly with increasing distance from the inflow (Figure 3-2). As with total N content, the concentration of total P in live cattail leaves was higher than in sawgrass at comparable distances from the inflow (Table 3-2). For cattail and sawgrass plants considered separately, a decreasing trend in total P content of live plants along the transect (increasing distance from the inflow) was apparent, although not significant at a = 0.05. In addition, standing dead total P content of sawgrass decreased significantly along the transect. However, total P concentration in standing dead plant tissue did not differ significantly between cattail and sawgrass plants.

Mean concentrations of total N and total P in standing dead (sampling sites and

plant types combined) were higher for the March sampling event than in June (Table 3-2), although only the difference for total N was significant. Since the spring and early summer period is characterized by new growth following winter senescense, standing dead material sampled in June was presumed repesent the same stock that was sampled in March. Thus, the material sampled in June would be expected to be older and therefore more decomposed.





78




2000

Live plant 1500- .- Standing dead



1000

soo





l.D -*C-" Litter
NE 500



S- - --m- 0
O 0









1000 .
CL f catail 0-10 cm
1500







500



0- I I
0 1 2 3 4 5 6 7 8 9 10 DISTANCE (km)


Figure 3-2. Total P concentration in organic C pools, representing the decay continuum
from plants to peat, as a function of distance from the S-10C inflow. Values for live plant and standing dead plant material represent composite (averaged)
data for cattails and sawgrass.





79


Lignin content was inversely proportional to cellulose content among all samples analyzed (Tables 3-2 and 3-3). Preferential utilization of cellulose over lignin as a C source for microbial decomposers resulted in depletion of cellulose relative to lignin. The shift from cellulose-dominated to lignin-dominated substrate along the decay continuum has been previously described using the ligno-cellulose index (LCI) (Melillo et al., 1989). The LCI is the proportion of lignin in the ligno-cellulose complex, or

LCI = lignin [3-1]
lignin + cellulose

so that the LCI increases with decomposition, with a maximum possible value of one. The LCI did not vary significantly (a = 0.05) with distance from the inflow for any live or dead organic matter component (Figure 3-3). There was also no significant difference (a = 0.05) in LCI between cattail and sawgrass plants (Table 3-2). However, LCI increased significantly (a = 0.05) along the decay sequence, including live plant tissue (Figure 3-3). Furthermore, the LCI for each component along the sequence was significantly (a = 0.05) higher than the previous component.

Microbial Respiration


Mean CO2 production rate during aerobic incubation did not differ significantly (a = 0.05) between cattail and sawgrass standing dead (Table 3-4). However, CO2 production was significantly higher (a = 0.05) for standing dead (plant type and sample site combined) collected in March than in June. There was also a significant trend (a = 0.05) of decreasing CO2 production in standing dead with increasing distance from the inflow. A similar trend was observed for the litter and peat components (Table 3-5); the trends were significant (a = 0.05) for litter and surface (0-10 cm depth) peat.

Specific C loss was calculated from gaseous C loss rate and substrate total C content as:





80






0.9

X 0.5 z 0.7

S0.6 N *30 . ---.i t
Z 0.5

Z 0

LU 0.4

0.1
0 1 2 3 4 5 6 7 8 9 10 DISTANCE (kin)

0 Live plant - 0-10 cm S--0- Standing dead .- 10-30 cm

--0-- Litter



Figure 3-3. Ligno-cellulose index (LCI) of organic C pools, representing the decay
continuum from plants to peat, as a function of distance from the S-10C
inflow. Values for live plant and standing dead plant material represent
composite (averaged) data for cattails and sawgrass.





81









Table 3-4. Potential C mineralization of dead cattail and sawgrass leaves
(standing dead), measured as CO, production rate during aerobic incubation at 25C. Values represent means with standard error in
parentheses (n=3).
Plant CO, production rate
Site type March 1995 June 1995
-------- mg Cg-' d' -------1 Cattail 1.36 (0.09) 0.85 (0.04) 2 Cattail 1.34 (0.07) 0.87 (0.03) 3 Cattail 1.62 (0.03) 1.04 (0.02) 4 Cattail 1.21 (0.01) 0.73 (0.01) 5 Cattail 1.50 (0.09) 1.23 (0.09)
Sawgrass 1.16 (0.15)
6 Cattail 1.24 (0.03) 0.82 (0.02)
Sawgrass 1.26 (0.08) 1.17 (0.11)
7 Cattail 1.08 (0.08)
Sawgrass 1.31 (0.03) 0.84 (0.05)
8 Cattail 0.78 (0.02)
Sawgrass 1.12 (0.05) 0.85 (0.07)
9 Sawgrass 1.01 (0.04) 0.85 (0.01) 10 Sawgrass 0.74 (0.02) 0.48 (0.03)






82


Table 3-5. Potential C mineralization rate and microbial
biomass C content of soil litter and peat (0-10 and
10-30 cm) layers. Potential mineralization was
measured as CO, production rate during aerobic incubation at 25"C. Values represent means with
standard error in parentheses (n=3).
Depth
Site increment CO, production Microbial biomass mg C g, d' mg C g-'

I Litter 1.763 (0.224) 25.50 (2.48)
0-10cm 0.575 (0.003) 8.19 (4.02) 10-30 cm 0.195 (0.023) 2.83 (0.83)
2 Litter 1.756 (0.206) 20.84 (2.61)
0-10cm 0.644 (0.057) 5.76 (2.94) 10-30cm 0.241 (0.016) 3.81 (0.69)
3 Litter 1.659 (0.105) 35.13 (1.54)
0-10cm 0.934 (0.071) 14.60 (2.20) 10-30cm 0.210 (0.020) 4.46 (2.68)
4 Litter 1.934 (0.144) 31.86 (1.28)
0-10cm 1.390 (0.105) 18.14 (1.17) 10-30 cm 0.193 (0.019) 6.65 (2.44)
5 Litter 1.652 (0.236) 37.44 (5.63)
0-10cm 0.649 (0.073) 9.64 (0.89) 10-30cm 0.130 (0.010) 4.61 (2.40)
6 Litter 0.964 (0.052) 15.77 (1.71)
0-10 cm 0.261 (0.034) 3.80 (0.29) 10-30cm 0.100 (0.009) 3.39 (1.45)
7 Litter 0.682 (0.035) 13.84 (2.11)
0-10cm 0.345 (0.018) 6.46 (0.45) 10-30cm 0.173 (0.012) 4.17 (2.28)
8 Litter 1.127 (0.066) 20.49 (2.46)
0-10 cm 0.356 (0.017) 6.21 (1.58) 10-30 cm 0.077 (0.008) 3.00 (1.68)
9 Litter 0.695 (0.044) 7.95 (1.50)
0-10 cm 0.256 (0.014) 5.48 (1.45) 10-30cm 0.090 (0.010) 2.24 (2.48)
10 Litter 0.410 (0.025) 9.27 (0.94)
0-10cm 0.261 (0.019) 3.42 (0.86) 10-30cm 0.104 (0.013) 2.94 (1.61)





83


mg substrate C lost g substrate day [3-2]
g substrate day mg substrate C

This quantity is equivalent to the first-order rate constant k, used in a simple exponential decay model. Thus it carries implications for biodegradability of the substrate. Specific C loss under aerobic conditions decreased with increasing distance from the inflow (significant (a = 0.05) for litter and buried peat) and along the decay sequence from standing dead to peat (Figure 3-4). Sharp increases in specific C loss occurred in samples from site 4 for both litter and surface peat. Aerobic specific C loss decreased sequentially from litter to surface peat to buried peat; however, values for standing dead were not significantly different (a = 0.05) from the litter layer.

Trends in CO2 and CH, production during anaerobic incubation of litter and peat were similar to those for aerobic CO, production (Table 3-6). Specific anaerobic C loss in litter samples decreased significantly (a = 0.05) with increasing distance from the inflow (Figure 3-5). Specific loss was somewhat higher for peat samples near the inflow, but there was no significant trend (a = 0.05) with distance. A sharp rise in specific anaerobic C loss occurred at sites 3 and 4 in the surface peat, similar to the peak associated with specific aerobic C loss. For all sites combined, specific anaerobic C loss decreased significantly (a = 0.05) between the litter, surface peat and buried peat layers.

Production of CH4 accounted for approximately 1 to 25% of the total C evolved during anaerobic respiration (Figure 3-6). The relative proportion of CH4, production was highest in the litter layer at sites 1-9, and in the surface peat at sites 2-5. For all other samples, CH4 production accounted for less than 2% of total gaseous C loss under anaerobic conditions.

Microbial biomass C was significantly higher (a = 0.05) in the litter layer than in surface and buried peat layers (Table 3-5). Biomass C decreased significantly (a = 0.05) with increasing distance from the inflow, although for individual components the decrease was significant only in the litter layer. The same trends were observed for microbial





84







0.005- 0- Standing dead

0.004- 0- Litter
-0- 0-10 cm


0 V
,.O

I 0.002- . /


U) 0.001 *
-\ .g -

0.000
0 1 2 3 4 5 6 7 8 9 10 DISTANCE (km)


Figure 3-4. Specific C loss during aerobic incubation of detrital organic C pools as a
function of distance from the S- IOC inflow. Values for standing dead plant material represent composite (averaged) data for cattails and sawgrass. Data
are mean values from triplicate incubations.





85







0.0015

- Litter
-0 0-10 cm
I'
SoohA- 10-30 cm S0.0010- \ O I .1 I \



0.
)



0.0000
0 1 2 3 4 5 6 7 8 9 10 DISTANCE (km)


Figure 3-5. Specific C loss during anaerobic incubation of detrital organic C pools as a
function of distance from the S-10OC inflow. Values for standing dead plant
material represent composite (averaged) data for cattails and sawgrass. Data are
mean values from triplicate incubations.





86







30

0Q Litter
Q -0- 0-10 cm Z 2- 10-30 cm S 20
- \Q


O I\ / I





0 1 2 3 4 5 6 7 8 9 10






Figure 3-6. Methane production in litter, surface peat (0-10 cn) and buried peat (10-30
%

o. r r .
0 1 2 3 4 5 6 7 8 9 10 DISTANCE (kin)


Figure 3-6. Methane production in litter, surface peat (0-10 cm) and buried peat (10-30
cm), expressed as percent of total C (CO2 + CH4) loss during anaerobic
incubations.






87


Table 3-6. Potential anaerobic decomposition rate for soil litter and peat
(0-10 and 10-30 cm) layers. Potential decomposition was
measured as CO2 and CH, production rate during anaerobic incubation at 25C. Values represent means, with standard
error in parentheses (n=3).
Depth
Site increment CO, production CH, production
-------- g C g' d -------1 Litter 483.8 (6.6) 44.60 (0.58)
0-10 cm 192.5 (19.6) 0.95 (0.17) 10-30 cm 91.2 (5.8) 0.41 (0.07)
2 Litter 403.0 (4.3) 73.26 (4.49)
0-10 cm 241.6 (12.8) 26.77 (0.50) 10-30 cm 103.4 (4.5) 0.72 (0.11)
3 Litter 500.8 (22.5) 107.16 (4.40)
0-10 cm 367.2 (7.5) 80.78 (8.15) 10-30 cm 69.6 (3.2) 0.77 (0.12)
4 Litter 474.3 (12.3) 100.06 (3.89)
0-10cm 441.8 (21.6) 129.07 (8.67) 10-30 cm 80.5 (3.0) 0.64 (0.12)
5 Litter 524.2 (41.9) 107.77 (5.81)
0-10cm 231.3 (16.1) 31.12 (8.27) 10-30 cm 57.0 (2.0) 0.52 (0.10)
6 Litter 233.0 (6.8) 76.40 (5.69)
0-10 cm 100.9 (9.4) 0.93 (0.37) 10-30 cm 25.4 (2.4) 0.60 (0.14)
7 Litter 196.1 (3.6) 19.56 (2.23)
0-10 cm 107.4 (12.5) 0.94 (0.19) 10-30 cm 46.4 (2.5) 0.69 (0.14)
8 Litter 350.3 (16.9) 78.40 (1.07)
0-10cm 132.8 (15.1) 0.82 (0.22) 10-30 cm 30.5 (2.1) 0.47 (0.08)
9 Litter 219.2 (13.5) 27.35 (8.42)
0-10 cm 93.0 (12.8) 0.90 (0.15) 10-30cm 61.6 (10.0) 0.54 (0.10)
10 Litter 163.0 (14.2) 2.48 (0.13)
0-10 cm 122.9 (3.8) 0.82 (0.23) 10-30 cm 39.5 (4.9) 0.59 (0.14)





88


biomass expressed as a percentage of total substrate C (Figure 3-7). Sharp peaks in biomass C were observed for litter at sites 3-5 and for surface peat at sites 3 and 4.


Discussion


The continuum of organic C transformations and flows which constitute the

wetland C cycle may be represented conceptually as a collection of discrete storage units, or compartments, with simultaneous transfer of mass among the compartments (Figure 3-8). This type of representation is the basis for numerous conceptual models of C cycling in terrestrial soils and ecosystems, in which soil organic C is classified according to turnover time (Jenkinson, 1990). Peat is represented by an active surface component (0-10 cm depth) and a more stable buried layer (10-30 cm), based on previously determined chemical and physical characteristics (Koch and Reddy, 1992; Reddy et al., 1993; DeBusk et al, 1994). The top 30 cm of peat roughly incorporates the zone of plant root-soil interaction in the Everglades marsh.

The vegetation component represents transformers of inorganic C (CO,) to organic C (primary production) through photosynthesis. The heterotrophic microflora represent transformers of organic C back to inorganic C through cellular respiration. The decay continuum from plant standing dead to peat is represented as a sequence of four compartments. Organic C is stored in the system in living (vegetation and microbial biomass) and non-living (standing dead plant tissue, plant litter and peat) components. Below-ground C pools (roots and rhizomes) were not sampled during this study, thus only living and dead plant leaves were considered. Previous research in the Everglades has shown that belowground biomass accounted for approximately 12% of total sawgrass production (Toth, 1987).

The concept of a decay continuum to describe the conversion of plant material to soil organic matter is based on selective loss of relatively labile constituents and resulting changes in the chemical and biochemical characteristics of the substrate (Melillo et al.,





89








9

8- --- Litter

7- -H 0-10 cm 0O -A. 10-30cm
- 6

M 5



.23
E .
2 I




0 1 2 3 4 5 6 7 8 9 10 DISTANCE (km)


Figure 3-7. Microbial biomass C pools in litter, surface peat (0-10 cm) and buried peat
(10-30 cm) in WCA-2A, expressed as percent of total organic C. Data are
mean values from triplicate analyses.




90




Live CH
plantC04



S Plant
I standing
I dead


El
I Litter g layer

CH

SSurface Microblal
peat decom(0-10 cm)m posers



Buried
peat
(10-30 cm)



Figure 3-8. Conceptual diagram of the organic C cycle and decay continuum in
Everglades WCA-2A.





91


1989). The broadly defined lignocellulose component of the substrate has been considered by many researchers to be the primary indicator of "substrate quality" (Colberg, 1988; Moran et al., 1989 ). Loss of non-lignocellulosic (labile) components of plant detritus occurs rapidly during the intial stages of decomposition, thus lignin and cellulose become the primary C components of the substrate (Moran et al., 1989).

The LCI can be correlated with age of decomposing plant material and has been proposed as an indicator of the state of decomposition (Melillo et al., 1989). The continuous increase in LCI during the decomposition process directly relates to the quality, or availability, of the C substrate. This is a result of "enrichment" of the substrate with the more recalcitrant lignin compounds as well as lignin derivatives occurring as microbial byproducts (Swift, 1982; Heal and Ineson, 1984). In the present study, LCI decreased significantly (a = 0.05) along the decay continuum, or more appropriately, the "decay sequence". This indicated that substrate C quality, or availability, decreased significantly (a = 0.05) from standing dead plant tissue to soil litter layer, surface peat (0-10 cm depth) and buried peat (10-30 cm).

Among other potential factors regulating decomposition rate (Heal et al., 1981) nutrient and O, availability were considered in the present study, as functions of hydroperiod and nutrient loading in WCA-2A. Total N and P content of live plant tissue (above-ground) reflected the historically-monitored gradient of nutrient-enrichment in the surface water (SFWMD, 1992). However, the P enrichment gradient in living and dead plant tissue, including peat, was the most pronounced, in accordance with results of previous comprehensive monitoring of soil chemistry in WCA-2A (DeBusk et al, 1994).

Total P analysis of soil layers (litter, surface peat and buried peat) included both inorganic and organic forms. Previous studies have shown that organic P accounted for approximately 70-80% of total soil P in WCA-2A (DeBusk et al, 1994; Quails and Richardson, 1995). Most of the organic P in the litter and peat is contained in the particulate organic matter, and can be considered as an integral part of the substrate. Substrate-bound





92


organic P was assumed to account for nearly all of the total P analyzed for standing dead material. This organically-bound P may be considered as a component of overall substrate quality, along with substrate C quality (as indicated by LCI values). Dissolved P in the water and soil porewater, on the other hand, could be considered externally available P. The latter pool of P has been measured on several occasions in WCA-2A, and has been shown to be highly variable, both spatially and temporally (SFWMD, 1992).

Both LCI and total P content were significantly related to decomposition rate in the four dead organic matter pools. A multiple linear regression model of aerobic CO2 production rate as a function of total P and LCI explained 91% of the variability in the response variable. Total P and LCI were both highly significant (a = 0.05) effects on CO, production. In addition, there was significant (a = 0.05) interaction between total P and LCI. This is an indication that P content and C availability may be co-limiting factors in decomposition of plant litter and peat in WCA-2A. Previous study of soils in Everglades National Park showed a significant effect of P on mineralization of added C (Amador and Jones, 1993, 1995).

A relatively high degree of variability surrounded the general trend of decreasing substrate CO2 production rate (substrate biodegradability) along the nutrient gradient, especially near the inflow (Table 3-5, Figure 3-4). This local variability was explained by substrate total P content and LCI, which covaried with the mineralization rate (Figures 3-2 and 3-3). The apparent underlying cause of this phenomenon is differential P loading among neighboring sites. There is evidence of hydraulic short-circuiting or channelized flow in WCA-2A, especially during periods of low water, caused by clumping of dense vegetation in the highly nutrient-enriched area and by airboat trails throughout WCA-2A.

Measurement of CO, production during aerobic incubation of organic matter

provided a direct measurement of organic C mineralization, which represents completion of the decomposition process. For short-term studies and recalcitrant substrates, measurement of CO, is an sensitive method for estimating decomposition rate. However, calculation of





93


the first-order decay constant using CO, production techniques over a short time period, e.g. one or two weeks, requires certain assumptions regarding the type of model to be used. These assumptions must be based on prior knowledge of the decay characteristics of the substrate, since curve fitting is best suited for several widely separated points in time. Controlled studies of mass loss in leaves from various plant types indicate that simple carbohydrates and other labile compounds are substantially lost during the first few weeks of decomposition (Moran et al., 1989). Standing dead portions of cattail and sawgrass may become highly leached and substantially decomposed while attached to the plant, due to the residence time on the plant of up to several months. Thus, a multi-phase kinetic model of long-term decomposition of standing dead material, litter and peat was judged to be inappropriate. The use of a simple (single compartment) exponential decay model was supported by data from a previous study of in situ decomposition of cattail and sawgrass standing dead in WCA-2A (Davis, 1991).

Carbon mineralization during aerobic incubation of litter and peat was considered a measure of potential decomposition rate, since 02 is limited or absent under flooded conditions. Availability of 02 in the water column and litter layer is dependent on rates of 02 diffusion across the water-air interface, 02 production in the water column and litter by algal and macrophytic photosynthesis and O2 demand exerted by the substrate. Even under drained conditions, saturation of microsites can occur due to the high capillarity of peat, therefore a significant portion of the soil may remain anaerobic. Potential respiration in peat from three wetland sites in North Carolina increased with overall nutrient availability (Bridgham and Richardson, 1992).

Measurement of aerobic, or potential, decomposition is an indicator of substrate quality, which encompasses availability of C and growth-limiting nutrients. In the context of the present study, potential decomposition rate incorporates the combined effects of LCI and total P. First-order decay constants calculated in this study, from "instantaneous" specific loss rate, varied over an order of magnitude (ca. 10' to 103 d-') among the four





94


organic C compartments (plant standing dead, litter layer, surface peat and buried peat). Corresponding turnover times (l/k) ranged from about 0.7 to 2.7 years for standing dead and litter, and 5 to 10 years for buried peat. As previously stated, these were considered potential values which may be observed in the field under optimal 02 availability.

Decomposition of above-ground standing dead material under field conditions is predominantly aerobic. The exception would apply to those portions of the plant situated below the water surface. Water depth in WCA-2A typically remains below about 50 cm, and frequently is much lower (SFWMD, 1992). However, standing dead decomposition rate may be limited by nutrient availability and, periodically, moisture availability. The latter is an important factor governing decomposition in terrestrial ecosystems (Heal et al., 1981). Decomposition of above-ground dead material may be severely restricted during extended dry periods. Standing dead material was wetted before aerobic incubation, therefore moisture content was assumed to be non-limiting for this study. Nutrient availability was most likely the limiting factor in decomposition of standing dead. Concentrations of N and P in dead plant tissue were significantly lower (a = 0.05) than in the litter layer and peat (Tables 3-2 and 3-3). Standing dead plant material (above-water) is isolated from plant and surface water nutrient sources. As soluble constituents, including newly-mineralized N and P, are leached out the substrate may become excessively nutrientdepleted. Based on analysis of N and P, decomposition activity in standing dead plant tissue was limited by both N and P availability. Average molar N:P ratio was approximately 40 in both standing dead and litter, yet total N and P content of litter was, on the average, about 5 times greater than for standing dead material. This was viewed as evidence of substantial microbial immobilization of both N and P following deposition of dead plant material into the litter layer. Immobilization potential has been linked to initially low N and lignin content of the organic substrate (Melillo et al., 1984). Studies in a cypress swamp in north Florida receiving municipal wastewater showed that, three weeks after




Full Text

PAGE 1

ORGANIC MATTER TURNOVER ALONG A NUTRIENT GRADIENT IN THE EVERGLADES By WILLIAM F. DEBUSK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1996 UNIVERSITY OF FLORIDA LIBRARIES

PAGE 2

ACKNOWLEDGEMENTS This research was supported in part by the U. S. Department of Agriculture National Research Initiative Competitive Grants program (Grant No. 92-37102-7542). The grant was awarded to Louisiana State University and University of Florida. Graduate Fellowships were funded in part by the USDA National Needs Fellowship Program for studies in the field of water science. I thank those who selected me as a recipient of this award. Thanks also go to all who donated their assistance, support and expertise, especially Dr. K. R. Reddy and Ms. Yu Wang. Most importantly, thanks go to my wife Patty for her moral (and, of course, financial) support. ii

PAGE 3

TABLE OF CONTENTS page ACKNOWLEDGEMENTS ii ABSTRACT v CHAPTERS 1 INTRODUCTION 1 Background 1 Wetland Carbon Cycle 2 Microbial Ecology 8 Factors Affecting Decomposition Rate 9 Modeling Decomposition of Heterogeneous Substrates 13 Ecosystem Models 17 Everglades Study Site 19 Objectives and Scope of Research 21 2 ORGANIC C MINERALIZATION AS A FUNCTION OF NUTRIENT ENRICHMENT AND HYDROLOGY 23 Introduction 23 Materials and Methods 26 Results 35 Discussion 51 Summary and Conclusions 61 3 TURNOVER OF ORGANIC CARBON POOLS ALONG THE WCA-2A NUTRIENT GRADIENT 62 Introduction 62 Materials and Methods 67 Results 74 Discussion 88 Summary and Conclusions 98 4 REGULATORS OF ORGANIC MATTER DECOMPOSITION ALONG THE WCA-2A NUTRIENT GRADIENT 101 Introduction 101 Materials and Methods 105 Results 112 Discussion 134 Summary and Conclusions 141 iii

PAGE 4

5 DETRTTAL CARBON MODEL 143 Introduction 143 Materials and Methods 143 Results and Discussion 155 Conclusions 161 6 SUMMARY AND CONCLUSIONS 163 LIST OF REFERENCES 167 BIOGRAPHICAL SKETCH 176 iv

PAGE 5

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ORGANIC MATTER TURNOVER ALONG A NUTRIENT GRADIENT IN THE EVERGLADES By William F. DeBusk August, 1996 Chairman: Dr. K. R. Reddy Co-chairman: Dr. J. W. Jones Major Department: Soil and Water Science Organic matter accumulation in wetlands represents a potential long-term sink and source for organic carbon (C) and associated nutrients and contaminants. Turnover of organic C was measured in a nutrient-impacted sawgrass and cattail marsh in Everglades Water Conservation Area 2 A (WCA-2A). Controlled laboratory incubations and a microcosm study were conducted to determine potential rates of C mineralization in plant litter and peat along a gradient of phosphorus (P) enrichment. Field incubations at 10 sites along the nutrient gradient measured in situ organic matter decomposition rate throughout the floodwater, litter and peat profile. Organic C mineralization in wetland microcosms was significantly enhanced by interactive effects of increased P availability and decreasing water table. Approximately 90% of the variability in potential organic C mineralization in peat and plant litter, measured under aerobic and anaerobic conditions, was explained by total P and lignocellulose content of the organic substrate. Anaerobic mineralization rates were 32% of the rates measured under aerobic conditions. In situ organic matter decomposition rate was higher in nutrient-

PAGE 6

enriched areas of WCA-2A than in the low-nutrient interior marsh. Decomposition rate typically was at a maximum in the floodwater and litter layer and decreased with depth in the peat profile. Field studies provided evidence that microbial decomposers obtain nutrients, especially P, from the surrounding floodwater and soil porewater as well as from the organic substrate. Results of laboratory and field studies indicate that organic C turnover in WCA-2A is strongly affected by P availability, although 0 2 availability is the major controlling factor. Availability of C (substrate quality) and nitrogen (N) may limit turnover rate under Penriched conditions. Experimental findings from these studies provide insight into the effects of accelerated nutrient loading on C cycling and net accumulation of organic matter and nutrients in wetlands. vi

PAGE 7

CHAPTER 1 INTRODUCTION Background Organic carbon (C) accumulation in wetlands is the mass balance between net primary production (C fixation) and heterotrophic metabolism (C mineralization). Organic C in plant litter, peat or soil organic matter (SOM) serves as the source of energy to drive the detrital food chain in wetlands. Most of the organic matter produced in wetlands is deposited directly in the detrital pool (Moran et al., 1989; Wetzel, 1992), thus microbial decomposers play the major role in C cycling and energy flow in wetlands. Burial of organic matter as peat provides a means for long-term storage of elements associated with organic C, such as nutrients and heavy metals (Clymo, 1983). Allochthonous compounds may be incorporated into peat and soil organic matter through plant uptake and senescence, immobilization within the soil organic matrix by physical/chemical processes such as adsorption, occlusion and precipitation, or through uptake by microbial decomposers, with storage either within living cells or metabolic byproducts. On a much broader scale, storage of organic C in wetland soil is an important component of the global carbon cycle and thus may impact large-scale processes such as global warming and ozone depletion (Happell and Chanton, 1993; Whiting, 1994). Under favorable conditions for organic matter decomposition, stored nutrients or contaminants may be released through mineralization and then recycled in the ecosystem or exported from the system (Ponnamperuma, 1972; Reddy and D'Angelo, 1994). The rate of net organic matter accumulation is a critical determinant of how a wetland functions as an ecological unit within the landscape. The storage function is equally important for natural 1

PAGE 8

2 wetlands, especially those which represent an ecotone between terrestrial and aquatic ecosystems, and created wetlands, which may be used for treatment of wastewater or runoff (Howard-Williams, 1985). Wetland Carbon Cycle From a conceptual standpoint the continuum of organic C transformations and flows which constitute the wetland C cycle may be represented as a collection of discrete storage units, or compartments, with simultaneous transfer of mass among the compartments (Figure 1-1). The vegetation component (including macrophytic and algal species) represents transformers of inorganic C (C0 2 ) to organic C (primary production) through the process of photosynthesis. The heterotrophic microfauna represent transformers of organic C back to inorganic C through cellular respiration. Organic C is stored in the system in living (vegetation and microbial biomass) and non-living (dead plant tissue, plant litter, peat or SOM) components. Non-living storage of organic C is proportionally large in wetlands in relation to other ecosystems; this storage provides a substantial energy reserve to the ecosystem which is slowly released through the detrital food web (Wetzel, 1992). Peat Peat is the result of biological, chemical and physical changes imposed on plant remains over an extended time period. Extent of decomposition, or humification, is qualitatively assessed by the extent to which plant structure is preserved (Given and Dickinson, 1975; Clymo, 1983). Numerous organic constituents have been isolated from peat, and many have been used in assessing the degree of decomposition, although these are generally classes and subclasses of organic compounds rather than discrete compounds. For example, peat material soluble in non-polar solvents is often termed "wax," but includes numerous compounds other than waxes (esters of fatty acids with alcohols other

PAGE 9

3 Figure 1-1. Conceptual diagram of the organic carbon cycle in a wetland ecosystem.

PAGE 10

4 than glycerol) (Clymo. 1983). Although a small portion of total peat mass, this fraction is of interest when determining origins of peat, since fatty acids of lipids can be traced directly or indirectly to the original plant type (Given and Dickinson, 1975; Borga et al., 1994). Various types of acid or alkali hydrolysis procedures have been used to isolate fractions roughly equivalent to cellulose, hemicellulose and lignin. Proportions of cellulose and hemicellulose in plant litter and peat tend to decrease with age (i.e. decomposition or "humification"), while lignin content increases with age; these three structural groups have all been used to characterize the degree of peat decomposition (Clymo, 1983; Brown et al., 1988; Bohlin et al., 1989). Analysis of peat has also revealed a large variety of phenolic compounds, many of which may be extracted in the "lignin" fraction. Concentration of cellulose and lignin is much higher in the fibrous fraction (remnant plant parts) of peat, while humic acids are much more prevalent in the humus fraction, although fulvic acids were found in somewhat greater amounts in the fibrous fraction (Given and Dickinson, 1975). Humic acid content of temperate peats showed a tendency to increase with age and depth, but peats in tropical and subtropical regions, including the Everglades, did not reflect this trend. Interpretation of chemical analyisis of peat may be clouded by uncertainty about the extent to which present characteristics represent historical differences in vegetation versus variability of decomposition processes. The degree of decomposition generally increases with depth of the peat, therefore, the organic matter becomes increasingly humified at greater depth (Clymo, 1983). Assuming an historically constant quality and quantity of substrate addition to the soil, the rate of decomposition is greatest in the upper regions of the profile, decreasing with depth. This is due not only to antecedent decomposition of substrate in the soil profile, but to the vertical gradient of soil environmental parameters. The latter category may include dissolved concentration (in saturated soil), soil moisture (unsaturated soil), and nutrient availability. All of these variables, particularly the first two, are functions of hydrologic conditions. Dissolved 0 2 concentration is affected both by the rate of diffusion from the

PAGE 11

5 soil surface, which is greatly reduced in flooded soils, and the 0 : demand created by organic matter along the diffusion path (Howeler and Bouldin, 1971). Dissolved Organic C The ecological significance of dissolved organic C (DOC) in wetlands has not been clearly defined. Even in terrestrial and aquatic systems, for which a greater depth of knowledge exists for DOC dynamics, the role of the dissolved fraction of organic C has not been well established. Among the reasons for this is the fact that DOC represents a broad spectrum of organic compounds of varying environmental recalcitrance (Wetzel, 1984), thus it may not be appropriate to treat DOC as a homogeneous category. Cook and Allan (1992a) measured several DOC fractions in old field soils representing various stages of succession (12-62 years), and observed changes in DOC composition during incubation period of 210 days. They found that the total amount and chemical composition of DOC in soil solution does not correspond to potential biodegradability. The Leenheer fractionation scheme failed to measure changes in DOC substrate quality, although the total mass of DOC decreased during the incubation period, and the total N concentration increased. The authors suggested that analysis of soil water DOC fails to account for the bulk of the microbial activity, which is intimately associated with surfaces of detritus and particulate soil organic matter. Cook and Allan (1992b) measured soil DOC concentration in conjunction with instantaneous rate of C mineralization to determine whether DOC represents the primary source of energy and nutrient release for soil metabolism. The size of the soil DOC pool was weakly correlated with instantaneous C mineralization rate. The DOC pool in aquatic ecosystems is a relatively stable component, both in terms of the size and quality of the pool (Wetzel, 1984). Decomposition of organic substrates in aquatic systems involves both labile and complex (recalcitrant) organic matter, the latter comprising the bulk of the DOC. Turnover of highly labile, energy-rich substrates may

PAGE 12

6 approach a rate of 5-10 times per day, thus actual concentration (storage) of labile organic C is generally extremely low. Despite the fact that recalcitrant DOC and POC are slow to mineralize, these pools represent a major portion of the organic C processed by the heterotrophic community due to the relatively massive size of these pools. During transport, selective removal of organic compounds occurs due to microbial utilization and chemical adsorption or precipitation, thus recalcitrance of DOC increases downgradient. Decomposition of dissolved organic compounds (resulting from partial decomposition of particulate organic matter) occurs primarily at surfaces in the Utter layer, soil/sediment and among epiphytic microflora. Microbial Biomass Most of the organic C fixed in aquatic and marsh systems (by both phytoplankton and macrophytes) is processed and recycled entirely by bacteria, without entering the food web, i.e. higher animals (Wetzel, 1984). Decomposition of organic matter is the primary ecological role of the heterotrophic microflora in soils, as it provides for mineralization of growth-limiting nutrients and formation of recalcitrant organic compounds (e.g. humus) which contribute to the chemical stability of the sytem (Swift, 1982). Microbial biomass comprises only a small fraction of the non-living organic matter, yet most of the net ecosystem production passes through the microbial component at least once and typically several times (Elliott et al., 1984; Heal and Ineson, 1984; Van Veen et al., 1984). Microbial decomposers derive their energy and C for growth from detrital organic C and facilitate recycling of energy and C within and external to the wetland ecosystem (Wetzel, 1984; 1992). Soil microbes may exert a significant influence on ecosystem energy flow in the form of feedback, since mineralization of organically-bound nutrients is a regulator of nutrient availability for both primary production and decomposition (Elliott, et al., 1984). The soil biomass constitutes a major C sink, in that it represents a significant portion of the "active" organic C (Paul and van Veen, 1978). Nutrients may be held tightly

PAGE 13

7 within the microbial biomass component of a low-nutrient ecosystem reflecting efficient recycling of remineralized organic compounds (Melillo et al., 1984). Microbes and plants often compete for nutrients in ecosystems with limited nutrient input and tight nutrient cycles (Lodge et al., 1994). Transient environmental conditions may stress the microbial community and result in fluctuations in biomass. This temporary increase in microbial mortality may result in significant remineralization of nutrients and induce a pulse of available nutrients for the plant community (Lodge et al., 1994). The wetland environment is characterized by widespread anoxia, thus the importance of anaerobic metabolism in organic matter turnover is greatly increased in wetlands versus terrestrial ecosystems (Ponnamperuma, 1972; Reddy and D'Angelo, 1994). The predominant type of anaerobic respiration in wetlands depends on the relative availability of alternate electron acceptors. Depending on geographic location and local anthropogenic influences, anaerobic metabolism in freshwater wetlands may be regulated by inputs of NO ? Mn^ Fe 3+ or S0 4 2 Utilization of these electron acceptors by heterotrophic microflora follows a thermodynamically predictable sequence, but also depends on availability of the compounds (Westermann, 1993; Reddy and D'Angelo, 1994). Microbial respiration in freshwater wetlands is frequently limited by electron acceptor availability, rather than C availability as in terrestrial ecosystems. In the complete absence of electron acceptors, methanogenesis is the major metabolic pathway in anaerobic wetland soils (Westermann, 1993). Microbial respiration in estuarine ecosystems such as salt marshes and mangrove swamps is dominated by sulfate reducing bacteria, due to the high concentration of sulfate in seawater (Howarth, 1993). Sulfate reducers and methanogens utilize most of the same metabolic by-products (short-chain fatty acids) of fermenting bacteria, which do not require an external source of electron acceptors (Oremland, 1988; Howarth, 1993). Sulfate reducing bacteria have the ability to outcompete methanogenic bacteria for these substrates in the presence of sulfate. Where the availability of suitable C substrate is sufficiently high, methanogenesis and sulfate

PAGE 14

8 reduction both may contribute significantly to soil respiration in high-sulfate soils (Oremland, 1988). Microbial Ecology Degradation of complex substrates such as plant detritus represents the actions of a diverse and heterogeneous assemblage of microorganisms which exhibits successional trends similar to those observed at the macro-scale (Swift, 1976). Saito et al. (1990) documented succession of cellulolytic fungi and bacteria to a more diversified bacterial assemblage during decomposition of pure cellulose in a waterlogged soil. The concept of r and K strategists in population biology has been applied to microbial populations in natural systems (Heal and Ineson, 1984; Swift, 1984; Atlas, 1986). An r strategist microorganism is often characterized by rapid growth rate and dominance in environments where resources are abundant, especially under fluctuating environmental conditions. In a marsh ecosystem, such conditions might result from cyclic drying and reflooding or pulsed loading of nutrients. The r strategists are exploitative, opportunistic and density independent, and would be expected to dominate during early successional stages (Insam and Haselwandter, 1989). The microbial K strategists would include slow growing, oligotrophic, humusdegrading organisms which are adapted to resource-limited environments (high stress). Growth of K strategists is generally damped, density-dependent and controlled by interspecific competition (Heal and Ineson, 1984). The r and K strategy concept generally has been related to microbial response to environmental stress as a selective pressure. Recently, the idea of disturbance as an additional selective pressure has been put forth. This has led to additional classifications of organisms based on their response to various combinations of stress and disturbance (Heal and Ineson, 1984; Swift, 1984; Atlas, 1986). Indices based on microbial activity and organic C have been proposed to provide an operationally-defined means for describing the response of soil microbial populations to resource quality and environmental conditions. The ratio of microbial biomass C to soil

PAGE 15

9 organic C (C^JC ) has been related to soil C availability and the tendency for a soil to accumulate organic matter (Anderson and Domsch, 1989; Sparling, 1992). Another, more widely investigated, microbial index is the metabolic quotient or specific respiration rate (^C0 2 ), the ratio of the basal respiration rate (as CO,-C) per unit microbial biomass C (C^) (Insam and Haselwandter, 1989; Anderson and Domsch, 1990,1993; Wardle, 1993; Ohtonen, 1994). The qC0 2 has been used as a response variable to effects of temperature, soil management, ecosystem succession and heavy metal stress, and is apparently a significantly more robust parameter than C mc /C (Anderson and Domsch, 1993). Relationships have been shown between qC0 2 and r and K strategies of stress-induced selection, and also with the general theory of ecosystem development proposed by Odum (1969). Increased qC0 2 is associated with r-selected microbial populations, resulting from high resource availability and simple substrate-decomposer relationships in early successional ecosystems (Insam and Haselwandter, 1989; Wardle, 1993). Lower qC0 2 values may be indicative of AT-selected populations dominating mature systems, especially low-nutrient systems with closed cycling of resources. In addition, increased qC0 2 levels are apparently related to ecosystem disturbance, such as pollution (Ohtonen, 1994). Factors Affecting Decomposition Rate The decomposition/mineralization process in wetlands differs from that in upland ecosystems in a number of ways (Reddy and D'Angelo, 1994). The predominance of aerobic conditions in upland soils generally results in rapid decomposition of organic matter such as plant and animal debris. Net retention of organic matter is minimal in this case, and consists of accumulation of highly resistant compounds which are relatively stable even under favorable conditions for decomposition (Jenkinson and Rayner, 1977; Paul, 1984). The decomposition process occurs at a significantly lower rate in wetland soils, due to frequent-to-occasional anaerobic conditions throughout the soil profile resulting from

PAGE 16

10 flooding. Because of this, significant accumulation of moderately decomposable organic matter occurs, in addition to lignin and other recalcitrant fractions (Clymo, 1983). There is evidence that growth strategies of the heterotrophic community reflect those of the plant community through their direct response to variation in resource quality, which is a function of plant growth strategies, and their similar response to common environmental conditions (Heal and Ineson, 1984). Carbon and nutrient utilization by the microbial decomposer community responds to three main groups of factors: (1) substrate quality, (2) physicochemical environment and (3) other organisms. Substrate quality is a general term which refers to the combination of physical and chemical characteristics which determine its potential for microbial growth. Substrate quality is not determined by any particular factor; however, nitrogen and lignin content have been suggested as indicators of biodegradability (Heal et al., 1981; Minderman, 1968; Andren and Paustian, 1987; Melillo et al., 1989). Physicochemical factors include temperature, pH, exogenous nutrient supply, moisture content (for non-flooded conditions) and oxygen or alternate electron acceptor availability (Swift et al., 1979; Heal et al., 1981; Reddy and D'Angelo, 1994). Nutrient availability affects decomposition rate by limiting microbial growth. Growth-limiting nutrients may be obtained by the microbial decomposers from the organic substrate or from dissolved compounds in the water and porewater (Godshalk and Wetzel, 1978). Microbial decomposers colonizing nutrient-depleted substrates (e.g. high C:N or lignin:N ratio) tend to scavenge significant amounts of nutrients from the surrounding media, resulting in net immobilization of growth-limiting nutrients in the system (Melillo et al., 1984). Although nutrient loading is typically greater in wetlands than in uplands due to location within the landscape, nutrient availability may be low relative to the pool of available organic C in wetlands (Reddy and D'Angelo, 1994). Nitrogen (N) and phosphorus (P) both have been identified as microbial growth-limiting nutrients in wetlands (Westermann, 1993). Nitrogen, unlike P, may be lost from wetlands through

PAGE 17

11 microbial metabolism via denitrification, as well as through ammonia volatilization (Reddy andD'Angelo, 1994). Substrate availability to microbial decomposers is determined by molecular size distribution, availability of nutrients in the substrate as well as in the environmental matrix, and the amount of exposed surface area (Heal et al., 198 1). Preferential utilization of low molecular weight compounds by microbial decomposers alters the chemical character of the organic resource such that the proportion of complex, slowly degradable compounds increases over time. Thus, the substrate quality of resources over a wide range of initial composition tends to converge to a more uniform composition, and hence, degradability (Melilloet al., 1989). The basic cell wall construction in vascular plants includes a framework component of a-cellulose, a matric component of linear polysaccharides (hemicellulose), and an encrusting component composed of lignin (Zeikus, 1981). Lignin occurs in cells of conductive and supportive tissue, and thus is not found in algae and mosses. The presence of lignin is the ultimate limiting factor in decomposition of vascular plant tissue (Zeikus, 1981). The heterogeneous group of organisms known as white-rot fungi, found primarily in terrestrial habitats, are the most active and complete decomposers of lignin (Eriksson and Johnsrud, 1982). Partial chemical modification of lignin has been documented for other eucaryotic microbes, most notably the brown-rot and soft-rot fungi, as well as for certain species procaryotes (Zeikus, 1981). Certain species of the genera Streptomyces, Norcardia, Bacillus, Azotobacter and Pseudomonas are included in the latter category. Numerous species of fungi capable of cellulose and lignin degradation, including Hyphomycetes, have been identified in oxidized zones of wetland soils (Westermann, 1993). Lignin degradation by white-rot fungi is highly oxidative, and may involve singlet oxygen and hydroxyl radicals in chemical oxidation of aromatic ring structures or

PAGE 18

12 intermonomeric linkages (Benner et al., 1984). However, recent research has cast doubts on the unconditional requirement of molecular 0 : for lignin degradation. Wetland plant decomposition studies have demonstrated decomposition of lignin in anoxic salt marsh, freshwater marsh and mangrove sediments using 14 C labelling techniques (Benner et al., 1984a). Bacterial degradation of lignin from Spartina alterniflora predominated over fungal degradation in studies of decomposition in salt marsh sediment (Benner et al., 1984b). The capacity for cellulose depolymerization is shared by several species of bacteria, actinomycetes and microfungi (Sagar, 1988b). Cellulose degradation readily occurs under anaerobic conditions, although at a reduced rate, mediated primarily by bacteria of the genus Clostridium (Swift et al., 1979). The ratio of cellulose to lignin degradation rate has been shown to be similar under both aerobic and anaerobic conditions (Benner et al., 1984a). Aside from its effects on lignin degradation and other extracellular depolymerization, 0 2 depletion forces a major shift in microbial metabolism of monomeric C compounds (e.g. glucose, acetate), from aerobic to anaerobic pathways (Westermann, 1993). Catabolic energy yields for bacteria utilizing alternate electron acceptors (NO,", Mn 4 *, Fe 3+ S0 4 = C0 2 ) are lower than for 0 2 thus microbial growth rates are generally lower in anaerobic environments (Westermann, 1993; Reddy and D'Angelo, 1994). In addition, sulfate reducing and methanogenic bacteria must depend on fermenting bacteria (e.g. Clostridium spp.) to produce substrate in the form of short chain C compounds, such as volatile fatty acids, from the breakdown of monoand polysaccharides (Howarth, 1993). Thus, although C metabolism occurs in the absence of 0 2 and even in the complete absence of electron acceptors, the decomposition process for plant litter and soil organic matter is often significantly curtailed.

PAGE 19

13 Modeling Decomposition of Heterogeneous Substrates Conceptual models developed for describing decomposition of heterogeneous substrates such as plant residues and soil organic matter generally fall into one of four major groups: single homogeneous compartment, two-compartment, multi-compartment and non-compartmental heterogeneous models (Jenkinson, 1990). Compartmental models utilize separation of the organic matter (or specifically organic C) into discrete compartments with significantly different turnover times. One approach attempts to quantitatively separate certain chemical fractions of organic matter which are presumed to control decomposition kinetics of the heterogeneous substrate, or can be described as functions of some readily measured chemical parameter or environmental variable (Minderman, 1968). Many researchers have used a more operationally defined approach to group different fractions of organic matter. A common conceptual scheme for organic matter fractions involves grouping according to empirically derived measures of biodegradability. Resulting categories may simply be termed "labile", "resistant", "stable", etc. This approach has been the most commonly used for ecosystem C and N models (Van Veen and Paul, 1981; Parton et al., 1987; Jenkinson, 1990; Grant et al., 1993a,b). Another approach to dealing with heterogeneous substrates treats the substrate as a single component of variable quality, or overall biodegradability (Godshalk and Wetzel, 1978; Moran et al., 1989). This is more often seen as a theoretical approach to description of decomposition, and has not been widely implemented in soil C and N models. Minderman (1968) evaluated decomposition data for plant litter from four forest types. It was shown that total mass loss over time could not be accurately described by a simple first-order decay model. However, when litter was separated into six individual components, viz phenols, waxes, lignin, cellulose, hemicellulose and sugars, mass loss for each component followed first-order decay kinetics. Mass loss of the sum of the 6 components was similar to total mass loss for forest litter. Thus, it was concluded that a

PAGE 20

14 multiple, or composite, exponential decay equation was a more appropriate model for litter decomposition than a single compartment model. It was also concluded that long-term (e.g. >10 years) decomposition kinetics is regulated by the resistant fractions of litter. This conceptualization of litter decomposition was based on only one year of decomposition data, and was not rigorously evaluated with the available data nor validated with additional data sets. First-order rate constants (k) for decomposition of hardwood leaf litter were shown to be highly correlated with initial tissue lignin:N ratio (Melillo et al., 1982). It was also shown that, for each type of litter, N content of the residual material was linearly related to mass loss (inverse-linear relationship) during the first year of decomposition. Further research on plant litter decomposition during a 77-month period established a strong relationship between substrate biodegradability and the ratio of lignin to total lignocellulose (lignin+cellulose) (Melillo et al., 1989). This lignocellulose index (LCI), or the proportion of remaining structural material occurring as lignin, varies widely among types of plant litter, e.g. from about 0.2 in "high quality" litter to about 0.6 in "low quality" litter. After an initial phase of degradation in which most of the simple compounds, including sugars and amino acids, are rapidly lost, the LCI for plant residues slowly converges on the 0.70.8 range found in soil organic matter. The implication of this was that, after initial stages of decomposition, the rate of decomposition is regulated only by environmental conditions, and is no longer a function of initial substrate composition. Aber et al. (1990) contended that the first-order rate constant (k) and the slope of the inverse-linear relationship of mass remaining vs. N content can be predicted from initial C fraction analysis (extractables, cellulose and lignin) and initial N content. They also accurately predicted decomposition of heterogeneous material using exponential decay models for the individual C fractions, although this was validated for only a small variety of litter types.

PAGE 21

15 Andren and Paustian ( 1987) compared zero-order, single first-order, parallel (twocompartment composite) first order, consecutive (two-compartment) first-order and a fourcompartment model. Overall, the simple first-order model incorporating environmental effects provided the best description of the data, although the temperature-corrected 2compartment models were also satisfactory, and gave a good estimate of the initial size of the labile (water soluble) pool. However, this pool could also be determined using the single-compartment model with Y-intercept determined by curve-fitting (therefore Yintercept is equivalent to the resistant fraction). A single-compartment exponential decay model using an exponentially decreasing rate coefficient (k) was used by Godshalk and Wetzel (1978) to describe decomposition of five species of aquatic macrophytes. The form of the model was dW/dt = -k*W, where k = a*exp(-b*t) and W is percent of initial weight remaining, t is time in days, k is the firstorder rate coefficient and a and b are rate parameters. The exponentially decreasing rate coefficient reflected the increasing overall recalcitrance of the substrate as the composition shifted to a greater proportion of highly resistant compounds. Decay rates were negatively correlated to the total fiber content of the substrate, but not well correlated to individual components such as cellulose, hemicellulose and lignin. A study by Moran et al. (1989) provided a critical evaluation of singleand multicompartment exponential decay models for plant litter, based on chemical components of the litter. The authors measured in situ and laboratory decomposition rates of whole litter and the lignocellulose components of litter for the emergent macrophytes Spartina alterniflora and Carex walteriana. Following an initial rapid (ca. two weeks) phase of mass loss from leaching and decomposition of non-lignocellulosic components, losses of C due to decomposition were attributable mainly to the lignocellulose fraction. By tracking the mass loss over time for individual chemical components of substrate, they were able to mechanistically construct a composite exponential decay model of whole-litter decomposition:

PAGE 22

16 N, = S 0 exp(-k,t)+H 0 exp(-k 2 t)+C 0 exp(-k ; ,t)+L 0 exp(-k 4 t) [ 1 1 ] where S 0 C 0 and L 0 represent initial amounts of soluble material, hemicellulose, cellulose and lignin in the substrate. Initial values for the four components shown in equation 1-1 were experimentally determined quantities. First-order rate constants (k) were estimated for each compartment using a curve-fitting routine. However, evaluation of alternative decomposition models revealed that the whole-litter decomposition process (at least for a one-half year period) was more accurately described by the decaying-coefficient model (Godshalk and Wetzel, 1978). The main shortcoming of the composite exponential model was the relatively poor fit (underestimation) during later stages of the experimental decomposition period, i.e. after about four months. This was attributed for the most part to an apparent decrease in specific decomposition rate (k) over time for individual components of the substrate. The authors suggested that selective degradation of less resistant fractions within each component (e.g. various phenolic subunits of the lignin component), along with the increasing importance over time of physical protection and humification, were probable causes for the observed decrease in component decay rates. It was concluded that the enormous number of biochemically distinct components, each with a characteristic biodegradability (and distinct rate constant) presents an inherent limitation on the use of the composite exponential model for accurate mechanistic descriptions of decomposition. A single-compartment decomposition model was proposed by Bosatta and Agren ( 1985), in which heterogeneity was described by a continuously varying quality variable q. A continuity equation for substrate C density was the basis for describing the flow of C along a continuum of time and quality, dr(q,t)ldt = -S(q,t) + dF(q,t)/dq [1-2] where r = density of substrate C, S = loss or sink of C due to microbial degradation, and F = flow of C into a defined region of q space (quality). Under this concept, mass of substrate C over the entire span of time and quality can be viewed as a surface plot of p(q,t), where t and q are represented by the x and y axes, with p as the height (z axis). The

PAGE 23

17 total mass of C remaining at time t may be calculated by integrating over q 0 to q x (range of quality) with time held constant. This model is quasi-mechanistic in that it represents a substrate as a composite of compounds with varying biodegradability while avoiding the pitfalls of artificially-imposed boundaries between biochemical components. It may be viewed as a two-dimensional version of the decaying coefficient model described previously. However, validation of this type of model using independent data sets has not yet been reported in the literature, therefore, the validity of the model for description of shortor long-term C dynamics is undetermined. Ecosystem Models Numerous ecosystem-scale models of organic C and N cycling and soil organic matter accumulation have been developed for grasslands and agricultural systems. Many of these employ the concept of discrete compartments for organic C or N pools with different turnover times. A long-term organic C model was constructed for evaluating turnover of soil organic matter in virgin and cultivated grasslands (Paul and Voroney, 1980; Van Veen and Paul, 1981). Organic matter was separated into decomposable, ligniferous and recalcitrant fractions for aboveand below-ground compartments. In addition, the soil profile was divided into three depth increments. Microbial biomass was considered separately as the "processing" compartment for soil organic matter. Microbial transformation and turnover of organic C and N were the basis for a soil organic matter model for predicting N immobilization in cultivated soils (Van Veen et al., 1984). A more detailed treatment of microbial kinetics was used in a soil organic matter model designed to evaluate short-term C and N dynamics (Grant et al., 1993a,b). The model was based on literature-derived equations for microbial activity, then validated for prediction of temporal trends in mineralization and immobilization of C and N in added substrates ranging from glucose to crop residues.

PAGE 24

18 Two well-documented models are particularly broad in scope, one in a temporal sense and the other from a spatial perspective. The Rothamsted turnover model (Jenkinson and Rayner, 1977; Jenkinson, 1990) was developed for prediction of long-term soil organic matter dynamics in croplands. Organic matter entering the model system is tranferred stepwise through five compartments representing decomposable plant material, resistant plant material, humified organic matter, microbial biomass and CO, lost from the system. Development of the Rothamsted model has benefitted from the availability of field data for a period of well over 100 years. The Century model of soil organic matter (C and N) in Great Plains grasslands (Parton et al., 1987) has been used to simulate effects of climatic gradients and grazing on soil organic matter levels and plant productivity over a wide geographic region. This model assigns soil organic C and N to three separate compartments based on turnover time: active (including microbial biomass), slow and passive soil organic matter. In addition, plant residue is divided into structural (slow) and metabolic (rapid) pools. The Century model has also been used to simulate shortand longterm effects of fire on N cycling in soil and plants (Ojima et al., 1994). A simulation model of organic C mineralization was developed for management of cultivated Histosols in the Everglades (Browder and Volk, 1978). Organic C was partitioned into a nonliving pool, according to degradability or chemical structure, and an active living pool containing microbial biomass. Water table (which controlled soil moisture and 0 2 availabilty), temperature and soil organic C content were used as effects to predict rates of soil subsidence and release of N compounds and organic acids to surface and groundwater. An ecosystem model of the Everglades sawgrass marsh was developed for predicting effects of various anthropogenic impacts on ecosystem structure and function (Bayley and Odum, 1976). Simulations included manipulation of hydroperiod and water depth, which influenced plant growth and fire occurrance. Phosphorus concentration in surface inflows was varied to simulate the effects of nutrient loading from anthropogenic sources, which impacted plant growth rates. Peat accumulation was simulated as a function

PAGE 25

19 of plant growth and fire. The simulation model suggested that the relatively simple ecosystem is highly unpredictable and sensitive to inflow of high P water and hydroperiod. However, calibration and validation of the model were hindered by a lack of reliable data for many system processes. Everglades Study Site The Everglades of south Florida is a mosaic of wetland ecosystems extending from near Lake Okeechobee southward to Florida Bay (Figure 1-2). The various ecosystems making up the pre-drainage Everglades, which encompassed an area of about 10 000 km 2 were highly adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred primarily through rainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959). Recent development of the Everglades and surrounding watershed has created changes in nutrient loading and hydrology (SFWMD, 1992). Most significantly, a large area of the northern Everglades was drained and converted to agricultural production during the first half of this century. This area of sugar cane, vegetable and sod farming is referred to as the Everglades Agricultural Area (EAA). The remainder of the northern Everglades was divided into three Water Conservation Areas (Figure 1-2) in the 1960s, for water storage and flood control. Water level within the WCAs is controlled by a system of levees, pumps and floodgates. Currently, the Everglades consists of the WCAs to the north and Everglades National Park to the south. Drainage of the EAA has resulted in widespread oxidation of the organic soil and concomitant mineralization and leaching of organically-bound nutrients. As a result,

PAGE 26

20 Figure 1-2. Everglades WCA-2A study area, showing sampling transect across the nutrient enrichment gradient, originating from the S-10C surface inflow.

PAGE 27

21 nutrients from organic soil mineralization, along with additional nutrients from fertilizers, have been transported via drainage canals toward the WCAs for approximately 30 years. Loading of agricultural drainage water into the WCAs has resulted in nutrient enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod, in the encroachment of cattail (Typha domingensis Pers.) and other rapidly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward and Omes, 1983; Toth, 1987, 1988). Accelerated nutrient loading in northern WCA-2A (Figure 1-2) during the past three decades has created a distinct nutrient (especially P) gradient in water, soils and plant tissue (Davis, 1991; Koch and Reddy, 1992; DeBusk et al, 1994). Changes in species composition of periphyton and macrophyte communities, along with an overall increase in net primary productivity have been documented along this gradient (Davis, 1991; SFWMD, 1992). Soil dating by analysis of l37 Cs peaks has indicated that peat accumulation rate has increased in nutrient-enriched areas of WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993). Objectives and Scope of Research The overall objective of this study was to determine turnover time of organic C pools in plant litter and peat along the gradient of nutrient enrichment in Everglades WCA2A. It is hypothesized that turnover of organic C changes along vertical and lateral profiles within the study area. Chapters 2 through 4 address more specific objectives: • Determine the effects of nutrient enrichment and flooding on mineralization of organic C in the peat-litter profile. It is hypothesized that increased nutrient loading and decreasing water table interactively increase C mineralization rate in the litter and peat. • Determine the effect of nutrient enrichment on turnover of organic C pools along the WCA-2A nutrient gradient, and to examine the relationships between size of the

PAGE 28

22 microbial biomass C pools and turnover time of associated organic C pools. It is hypothesized that turnover time for major C pools increases (decomposition rate decreases) downgradient from the inflow of nutrient-laden water. It is also hypothesized that turnover time increases in successively older organic C pools. • Determine the effect of nutrient enrichment on in situ decomposition rate along a vertical profile in the water column and peat, specifically the significance of various environmental and substrate-related factors on decomposition rate. • Develop a mass balance for organic C using experimental data, and construct a conceptual model to describe turnover of organic C pools and net C accumulation. Experimental data will be generated from a combination of field, greenhouse and laboratory studies, presented in the following chapters. Chapter 2 describes a greenhouse study using Everglades microcosms to determine effects of water table and nutrient enrichment on whole-profile C mineralization. Laboratory incubations of Everglades soil and plant litter to determine turnover of organic C pools are presented in Chapter 3. Studies involving in situ decomposition of plant material and an artificial substrate are described in Chapter 4. Development of a conceptual model and synthesis of experimental results are presented in Chapters 5 and 6.

PAGE 29

CHAPTER 2 ORGANIC C MINERALIZATION AS A FUNCTION OF NUTRIENT ENRICHMENT AND HYDROLOGY Introduction <£1 Decomposition of organic matt er is govern ed by the chemical composition of the substrate and external, or environmental, factors. Among the more important environmental factors are temperature, moisture, nutrients and electron acceptors (Swift et al.. 1979; Heal et al., 1981; Reddy and D'Angelo, 1994). The most important electron acceptor in terms of organic matter turnover is 0 : The presence of floodwater severely limits availability of 0 2 in wetlands, therefore decomposition proceeds at a highly reduced rate. In the absence of 0 2 breakdown of organic carbon (C) by microbial decomposers is accomplished using alternate electron acceptors, such as NO,", Mn 4 *, Fe 3+ and S0 4 = which results in lower energy yield for the organisms (Reddy and D'Angelo, 1994). The hydroperiod of a wetland, which encompasses frequency, duration and depth of flooding, is thus a major determinant in the accumulation of organic matter. Substrate quality, a general term referring to the degradability of an organic substrate by microbial decomposition, also has a major impact on rate of decomposition (Swift et al, 1979; Heal et al., 1981; Heal and Ineson, 1984). Generally included in the concept of substrate quality are the availability of organic C (which is a function of lignocellulose content and other factors) and nutrients for cell maintenance and growth. The required nutrients for microbial decomposition of organic C may be aquired from the substrate itself, e.g. as part of the original plant material, or from the floodwater or porewater (Swift et al, 1979). In a nutrient-enriched, or eutrophic, wetland, plant litter (substrate), floodwater and porewater may serve as nutrient sources for decomposers. In 23

PAGE 30

24 nutrient-poor, or oligotrophia wetlands, the organic substrate may be the main source of nutrients for microbial decomposers, if nutrient concentration in the water and porewater is extremely low. Nutrient availability affects decomposition rate through its effects on microbial growth. Although nutrient loading is typically greater in wetlands than in uplands due to location within the landscape, nutrient availability may be low relative to the pool of available organic C in wetlands (Reddy and D'Angelo, 1994). Nitrogen (N) and phosphorus (P) both have been identified as microbial growth-limiting nutrients in wetlands (Westermann, 1993). Nitrogen, unlike P, may be lost from wetlands through microbial metabolism via denitrification, as well as through ammonia volatilization (Reddy and D'Angelo, 1994). The Everglades encompasses a variety of wetland ecosystems which were historically adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred primarily through rainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974) (Figure 2-1). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959). Loading of agricultural drainage water into the WCAs has resulted in nutrient enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus (P) enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod, in the encroachment of cattail (Typha domingensis Pers.) and other rapidly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward and Ornes, 1983; Toth, 1987, 1988).

PAGE 31

Figure 2-1. Site map for WCA-2A study area, showing locations of sampling sites. Coordinates for sampling sites are listed in Table 2-1.

PAGE 32

26 The main objective of this study was to determine the effects of nutrient enrichment and flooding on turnover, or mineralization, of soil organic C along the WCA-2A nutrient gradient. It is hypothesized that increased nutrient loading and decreasing water table interactively increase C mineralization rate in the litter and peat. Materials and Methods Site Description Field study sites were located in WCA-2A, a 447 km 2 region of the northern Everglades (Figure 2-1). Surface water flows into WCA-2A from the Hillsboro Canal through the four S-10 water control structures and from the North New River Canal through the S-7 pump station. Most of the hydraulic loading is through the S-10C and S10D structures into the northern portion of WCA-2A. The general direction of flow is from north to south. Water depth is usually less than one meter, and varies considerably, both seasonally and year-to-year, with occasional dry periods (SFWMD, 1992; personal observations). The bulk of the surface outflow is through three control structures at the south end of WCA-2A, into WCA-3 (Figure 2-1). Soil in WCA-2A consists of Everglades and Loxahatchee peats (Gleason et al., 1974). Everglades peat, the most common soil in the Everglades, is associated with the saw grass marsh community. It is dark brown, finely fibrous to granular, with circumneutral pH, relatively high N content and low Si0 2 Fe and Al content. Peat depth in WCA-2A ranges from about 1 to 2 m, and age of basal peats is estimated to be 2000 to 4800 yr. Beneath the peat lies a bedrock of Pleistocene limestone, with intermediate layers of calcitic mud, sandy clay and sand in several areas (Gleason et al., 1974). The primary sources of nutrient loading to WCA-2A are the S-10 structures which convey water from the Hillsboro Canal and WCA-1 (Figure 2-1). A distinct gradient of N and, most significantly, P enrichment in water, plants and soil has formed between the

PAGE 33

27 high-nutrient region adjacent to the inflows and the low-nutrient interior marsh of WCA-2A (Koch and Reddy, 1992; SFWMD, 1992; DeBusk et aL, 1994). A vegetation gradient coincides with the nutrient gradient; most notable is the gradient from sawgrass marsh with scattered aquatic slough in the interior to cattail and mixed emergents near the inflows. The vegetation gradient was divided into three discrete categories for the purposes of the current study: cattail-dominated, sawgrass-dominated and mixed cattail and sawgrass (Figure 2-1). A north-south transect approximately 10 km long was established for soil, water and vegetation sampling along the nutrient gradient in WCA-2A. Sampling sites were located along the transect at 7 locations, starting near the S-10C inflow structure at the Hillsboro Canal and ending in the interior marsh region (Figure 2-1; Table 2-1). These locations were chosen as representative of a wide range of soil and water P concentration and vegetation type. A total of 10 sampling sites were established among the 7 locations on the transect, as follows. One site each was situated at distances of 0.75 and 2.2 km from the inflow (sites Tl and T2), within the highly nutrient-impacted cattail dominated area (Figure 2-1). Two sites each were located at distances of 3.1, 4.0 and 5.0 km from the inflow, in the moderately impacted transitional vegetation zone characterized by patches of cattail and sawgrass and mixed stands of cattail and sawgrass. At each of these three distances, one sampling site was located within a cattail stand (sites T3, T4 and T5 respectively) and one within a sawgrass stand (sites CI, C2 and C3 respectively). The minimally impacted interior region was represented by sites within the sawgrass marsh at distances of 6.9 and 10.2 km from the inflow (sites C4 and C5). This area is characterized by a significant coverage of aquatic slough habitat; however, sampling was limited to the marsh community to maintain continuity among all sites. Sampling Methodology Duplicate soil cores were taken on July 1 1, 1994 at each of the 10 sampling sites in WCA-2A. An additional (triplicate) core was obtained for cattail and sawgrass sites at each

PAGE 34

28 Table 2-1. Locations of WCA-2A sampling sites for microcosm study and approximate downstream distance from the S-10C surface water inflow. Site Latitude N Longitude W Distance deg min deg min km Tl 26 21.8 80 21.1 0.75 T2 26 21.1 80 21.2 2.2 T3,C1 26 20.5 80 21.3 3.1 T4,C2 26 20.0 80 21.4 4.0 T5,C3 26 19.5 80 21.4 5.0 C4 26 18.5 80 21.5 6.9 C5 26 16.8 80 21.5 10.2

PAGE 35

29 end of the respective ranges along the transect, i.e. at sites Tl, T5, CI and C5. This scheme provided triplication at the sites of highest and lowest impact and at 2 transitional sites. Replicate cores within each site were located approximately 2 m apart. In all, 24 soil cores were obtained. Soil cores were collected intact in 50 cm-length sections of rigid, clear acrylic tubing with inside diameter of 14.6 cm (nominal dimensions: 6-inch diameter with 1/8-inch wall thickness). The coring procedure was optimized during previous field trips, such that disruption of litter and peat stratigraphy was avoided and compaction of the core was minimized (less than 5%). Penetration of the loosely packed litter layer and mats of fine roots at the easily compressed peat surface may result in significant compaction using traditional methods of core driving using a hammer or pile driver. Coring for this study was accomplished by pushing the acrylic tubes downward by hand while cutting around the perimeter of the core tubes with a serrated knife (i.e. breadknife) to sever pieces of plant litter and roots. The coring tubes were pushed into the soil until the peat surface was aligned with marks placed 30 cm from the bottom of the tubes. Intact cores were excavated using a shovel, and the bottom openings were plugged with a polypropylene disks (2.54 cm thick) machined to a diameter slightly smaller than the core I.D. and fitted with dual rubber O-rings for a water-tight seal. Core tubes were capped for transport from the field, such that air headspace was reduced or eliminated, to prevent turbulent mixing and possible disruption of the core stratigraphy. Soil-Water Microcosms The microcosm study was set up in an open-sided greenhouse on the University of Florida campus. Intact soil cores (soil-water microcosms) were placed upright in aluminum racks situated inside polyethylene livestock troughs. The troughs were filled with water to a level coinciding with the top of the peat in the core tubes, creating near-ambient temperature water baths to buffer temperature in the microcosms. Water temperature was continuously

PAGE 36

30 monitored during the study with thermocouples connected to a data logger (Model CR10. Campbell Scientific, Logan UT). Partial shading was provided (to approximate field light conditions) by draping shade cloth over the tubs containing the microcosms. Measured levels of PAR (photosynthetically active radiation) beneath the shade cloth were less than 25% of full sunlight at midday. Spot measurement of PAR beneath the canopy of a dense cattail stand in a nearby planted wetland yielded values of approximately 120-150 |iE m" 2 s" 1 or about half the rate measured for the microcosms. However, the canopy density of cattail and sawgrass at the WCA-2A sampling sites was estimated to be significantly lower than in the experimental wetland. Floodwater depth was adjusted in each microcosm to 10 cm by siphoning off excess water. Therefore, the soil-water microcosms represented the top 30 cm (meantS.D. = 30.51.5 cm) of the WCA-2A peat profile, plus the overlying layer of plant litter and 10 cm of surface water. During the experimental period, distilled water was added as needed to microcosms to compensate for evaporation of surface water. Microcosms were allowed to stabilize for about 2 weeks, before initiation of experiments. The following studies were performed during the months of August and September, a period of minimal day-to-day temperature variation. Dissolved (X and pH Profiles Vertical profiles of dissolved 0 2 and pH were measured in each microcosm during mid-day, over a period of one week. A needle-type microelectrode (Model 757; Diamond General, Ann Arbor, MI) was mounted on a motorized screw-drive apparatus, developed in-house for redox profiling of lake sediment cores. The microelectrode was slowly (typically 3 mm min" 1 ) driven vertically through the water column and litter layer to the peat surface. These measurements were timed to coincide with maximum algal photosynthetic activity, thus approximating maximum expected values of dissolved 0 2 and pH in the water

PAGE 37

31 column. In addition, diel measurements of dissolved 0 2 were taken within the litter layer of microcosms representing highand low-impact areas (sites Tl and C5). Profiles of dissolved 0 2 were recorded, as described above, and repeated every 6 hours for a 24-hour period. Soil Respiration In preparation for soil respiration measurements, the plastic tubs containing the soil-water microcosms were covered with sheets of plywood to excude sunlight and terminate algal activity. Respiration of the heterotrophic microflora could be better estimated without interference from algal photosynthesis and respiration. Microcosm core tubes were capped with polyethylene plugs, which were securely attached with silicone glue to provide an airtight seal. The caps contained inlet and outlet ports for gas exchange from the core headspace. A gas manifold was set up to feed compressed air to each microcosm through plastic tubing. Air was bubbled through the surface water at a depth of about 5 cm, at a rate of approximately 30 mL min"'. The air was passed through a 2N NaOH solution prior to delivery to the microcosms to scrub C0 2 from the inflow gas. Soil respiration was estimated by measuring C0 2 and CH 4 evolved from the microcosms, using the following procedure. Outflow air from each microcosm was sampled through a short outflow tube (Pharmed brand rubber tubing) in the top with a 1 mL syringe fitted with a 25-gauge hypodermic needle. The syringe containing gas sample was immediately inserted into a butyl rubber stopper to retard leakage through the needle. A second syringe was used to take a duplicate sample in the same manner. This procedure was repeated for all 24 microcosms. Triplicate samples of the inflow air stream were also taken, for calculation of mass inflow rates for C0 2 and CH 4 (if detectable). At each sampling event, air outflow rate was measured for each microcosm, using a Manostat Calcuflow flowmeter (Manostat, New York, NY). The syringes containing outflow air samples were immediately taken to the lab for analysis of C0 2 and CH 4 Respiration

PAGE 38

32 measurements were made under aerated (air bubbled into the water column) and non-aerated (air delivered to the headspace above the water surface) conditions in the water column, to simulate both aerobic and anaerobic flooded conditions in the field. Mass of CO, and CH 4 in the 1 mL air samples was determined by direct sample injection into a dual-detector gas chromatograph (Hewlitt-Packard 5840A, Avondale, PA). Separate determinations of C0 2 and CH 4 were made from the duplicate samples taken from each microcosm. Sample gases were analyzed for CO, and CH 4 analysis using thermal conductivity (TCD) and flame ionization detectors (FID), respectively. For C0 2 analysis, a Poropak N (Supelco, Bellefonte, PA) column was used, with He as a carrier gas. Oven, injector and detector temperatures were set to 60, 140 and 200 C. For CH 4 analysis, a Carboxen 1000 (Supelco, Bellefonte, PA) column was used, with a N 2 carrier gas. Oven, injector and detector temperatures were 120, 120 and 200 C. Mass flux of C0 2 and CH 4 from each microcosm was calculated as: (outflow cone. inflow cone.) x (air flow rate). Following measurement of C0 2 and CH 4 flux under flooded conditions, one replicate microcosm from each of the 4 triplicated site (Tl, T5, CI and C5) was extruded from the core tube and sectioned into 2-cm "slices", with the litter layer collected separately. The discrete depth intervals thus obtained were placed in air-tight polyethylene containers and refrigerated at 4 C. Samples were later homogenized by mixing with a spatula; large pieces of plant material were chopped into smaller (ca. 1 cm) pieces to increase the homogeneity of the sample. Subsamples of the peat and litter samples were used for determination of moisture content, bulk density and analysis for total C, N and P. Additional subsamples were used in an aerobic soil respiration study under controlled conditions (described below). Total C and N analyses were performed on oven-dried (60 C), ground (< 0.2 mm) subsamples using a Carlo-Erba NA-1500 CNS Analyzer (HaakBuchler Instruments, Saddlebrook, NJ). Separate subsamples were analyzed for total P following combustion (ashing) at 550 C for 4 h in a muffle furnace and dissolution of the

PAGE 39

33 ash in 6 M HC1 (Anderson, 1976). The digestate was analyzed for P using the automated ascorbic acid method (Method 365.4, USEPA, 1983). Soil respiration (CO, and CH 4 flux) measurements were repeated under drained conditions. A 0.5 cm drain hole was drilled through the side wall of each core tube, about 1 cm from the bottom plug. A polypropylene fitting packed with glass wool was used to connect a short length (approx. 40 cm) of 0.5 cm diameter Tygon tubing to the drain hole. Water was allowed to drain from the microcosms to the desired level. The Tygon tubing was then taped to the outside of the core tube in a vertical orientation, to serve as an indicator of water level in the microcosm (water table as opposed to saturated capillary fringe). Fine adjustment of water level was performed by extracting water with a 50 mL syringe fitted with a long, large-bore needle placed along the inner wall of the microcosms. Experimental determination of soil respiration was repeated under conditions of transient water level. Measurement of C0 2 and CH 4 flux were made according to the protocol described above, as water level in the microcosms was sequentially lowered to the peat surface and 5, 10 and 15 cm below the peat surface. Water level was maintained at each depth for approximately 5 days, with flux measurements made on the last day. Additional flux measurements were made at 7 and 19 days, while the microcosms were drained to -15 cm. Aerobic Soil Incubation: Potential Soil Respiration Subsamples of the litter and 2-cm vertical sections of the "sacrificed" replicate microcosms from sites Tl, T5, CI and C5 (0.75, 3.1, 5.0 and 10.3 km from inflows) were used for a laboratory study of soil respiration potential. Short-term aerobic assays of C0 2 production were set up in triplicate 160-mL glass serum bottles containing 10 g wet weight (approximately 1 g dry weight equivalent) of sample, either litter or peat from each 2-cm section of the 30 cm profiles. Bottles were covered with clear plastic (PVC) film, to retard moisture loss while allowing gas exchange, and placed in a temperature-controlled

PAGE 40

34 incubator equipped with a reciprocal shaker table. The sample bottles were incubated in the dark at 27 C, with constant agitation to ensure uniform exchange of gases between soil and air over the short time intervals of measurement. Samples were pre-incubated for 48 hours to allow stabilization of microbial populations in the soil and Utter. Soil respiration was determined by measuring short-term increase in C0 2 in the headspace of closed serum bottles containing litter or peat samples. A static system, rather than a more conventional flow-through (respirometer) method. To begin the respiration assay, sample bottles, which had been provided continuous exchange of air, were stoppered with sleeve-type rubber serum stoppers. One mL of headspace was sampled from each serum bottle with a gas-tight syringe, for the initial (time = 0) sample. The sample bottles were immediately returned to the incubator/shaker. Headspace samples were taken again at 10, 20 and 30 minutes. Sample bottles were shaken continuously except for the brief period of time required for sampling. Erroneous results have been reported for static incubations of calcareous soils, associated with solubility and retention of C0 2 as HC0 3 (Martens, 1987). However, incubation time was shortened sufficiently to minimize the C0 2 gradient between headspace and soil solution, and maintain a linear increase in C0 2 headspace concentration over the incubation period. Headspace samples were analyzed for C0 2 by GC-TCD, as described above. The value thus obtained was then extrapolated from injection volume to the sample headspace volume. The time rate of change of C0 2 in the serum bottle was used for calculation of C0 2 production rate. For the short (30 min.) interval used in the assay, accumulation of C0 2 in the headspace was a linear function of time, i.e. production rate was constant. Aerobic soil respiration was determined by relating the rate of C0 2 production to the sample mass.

PAGE 41

35 Results Soil Chemical Characteristics Soil total C and N content was relatively invariant among the four cores analyzed, and within each profile (Table 2-2). Total C content of the peat in the four selected cores was about 44% (w/w), nearly all of which was organic C (based on an organic matter content of 88-90%). Storage of N relative to total C content, expressed as C:N ratio, was consistently high (low C:N ratio) in soil and litter of the four microcosms (Table 2-2). Magnitude of C:N ratio varied only slightly around a mean of roughly 15, with slightly greater values found for site C5. Analysis of total P revealed a distinct trend in soil and litter P enrichment among sites (Figure 2-2). In the case of microcosm Tl, substantial P enrichment (compared with the interior marsh site C5) was found throughout most of the top 30 cm of peat and litter. Microcosm C5, in contrast, showed slightly elevated total P concentration in the litter and the upper 8-10 cm of peat, decreasing below about 20 cm. Significant P enrichment also occurred in the litter and upper portion of the peat profile of microcosms C 1 and T5 (Figure 2-2). The overall extent of P enrichment at these transitional sites was less than for site Tl ; P concentration approached "background" levels for sawgrass peat below about 15 cm in T5 and about 28 cm in CI. The CP ratio of litter and the upper 80% of the soil profile in microcosm Tl was approximately 300 (Figure 2-3). In contrast, C:P ratio in microcosm C5 ranged from about 800-900 in the upper profile to 2000 in the deeper peat. Dissolved Q 2 and pH Profiles Dissolved 0 2 profiles (Figure 2-4) revealed a distinct spatial pattern of intense, localized 0 2 production and rapid utilization in the litter layer and at the soil surface. These local 0 2 sources and sinks resulted in highly variable concentration and numerous microgradients within the profile. Profiles of pH showed a high degree of similarity to

PAGE 42

36 Table 2-2. Chemical analysis of the peat-litter profile of wetland microcosms from sites T-l, T-5, C-l and C-5 in Everglades WCA-2A. Ash Site Depth content Total C Total N Total P C:N C:P % a kg •; mo kg T ittfr 10.4 446 27.3 1563 16.3 285 f)-2 If) 3 445 30.6 1600 14.5 278 %4 1 1 1 434 29.5 1604 14.7 271 4-fi 1 1.7 447 31.5 1635 14.2 273 (S-8 1 3 7 1 J. ( 443 33.2 1558 13.3 284 8-in O 1 \J 1 3 0 1 J.U 441 30.6 1592 14.4 277 l \j i — I — u 437 29.4 1545 14.9 283 1 9-14 1 1 4 445 29.8 1491 14.9 299 14-16 1 1.6 460 28.6 1436 16.1 320 16-18 12.8 448 28.6 1475 15.7 304 18-20 13.3 426 27.3 1421 15.6 300 20-22 11.2 439 30.0 1342 14.6 327 22-24 1 1.8 438 28.8 1324 15.2 331 24-26 10.9 435 28.2 1044 15.4 417 26-28 15.4 421 28.7 703 14.7 599 28-30 16.2 41 1 31.0 452 13.3 911 T-5 Litter 10.5 437 30.0 1 141 14.5 383 0-2 If) 5 421 30.5 1 184 13.8 356 2-4 1 1.2 440 33.6 1312 13.1 335 4-6 9 3 436 34.5 1355 12.6 321 68 U O 9 9 7 7 433 34.6 1216 12.5 356 8-10 1 0 f\ I \J.\J 34 9 897 0 7/ 12.5 486 10-12 12.1 442 35.7 448 12.4 986 12-14 10.1 458 36.4 291 12.6 1574 14-16 9.0 464 37.8 252 12.3 1839 16-18 9.2 456 38.2 239 11.9 1903 18-20 9.3 465 37.6 257 12.4 1808 20-22 10.4 456 37.8 232 12.1 1964 22-24 11.5 427 35.9 213 11.9 2004 24-26 14.7 444 36.2 199 12.3 2224

PAGE 43

37 Table 2-2— continuued. Ash Site Depth content Total C Total N Total P C:N C:P CTtl % . o Icq -; me ke 1 C-l Litter 12.8 435 29.9 1328 14.6 328 0-2 11.5 501 34.4 1232 14.6 406 2-4 11.7 440 30.0 1181 14.7 372 4-6 12.0 482 31.3 981 15.4 491 6-8 1 1.5 442 29.1 1059 15.2 418 8-10 11.1 439 28.0 905 15.7 485 10-12 11.1 399 26.7 793 15.0 503 12-14 12.2 423 28.6 730 14.8 579 14-16 11.1 382 25.8 677 14.8 565 16-18 10.7 440 29.0 531 15.2 830 18-20 11.2 433 27.1 471 16.0 919 20-22 9.9 383 28.0 408 13.7 938 22-24 10.3 408 28.9 294 14.1 1389 24-26 9.8 372 25.9 314 14.4 1181 26-28 8.5 518 39.0 278 13.3 1863 28-30 8.5 417 33.6 246 12.4 1698 C-5 Litter 16.6 476 37.5 534 12.7 893 0-2 13.7 451 35.6 586 12.7 770 2-4 13.5 491 38.6 599 12.7 820 4-6 12.0 451 32.8 546 13.8 825 6-8 1 1.4 469 27.7 493 16.9 952 8-10 1 1.0 494 28.0 357 17.7 1386 10-12 10.4 473 27.2 336 17.4 1409 12-14 9.8 429 22.3 307 19.3 1398 14-16 9.5 467 26.1 291 17.9 1606 16-18 9.3 442 25.7 275 17.2 1605 18-20 9.4 458 27.7 252 16.5 1816 20-22 9.6 423 26.6 243 15.9 1739 22-24 9.3 414 26.3 236 15.7 1753 24-26 9.4 425 27.1 215 15.6 1976 26-28 9.5 463 28.8 211 16.1 2192

PAGE 44

38 2000 Figure 2-2. Soil total P profiles in selected microcosms from WCA-2A.

PAGE 45

39 C:P RATIO LITTER 0-2" 2-4" 4-6' 6-8" 8-10' 10-12" 12-14" 14-16' 16-18" 18-20" 20-22' 22-24" 24-26" 26-28" 28-30' LITTER' 0-2" 2-4" 4-6" 6-8" 8-10" 10-12" 12-14" 14-16" 16-18' 18-20' 20-22' 22-24' 24-26' 26-28 28-30 500 1000 1500 2000 2500 0 I I 1 t i 1 I 1 1 i t 1 1 1 1 1 I 1 1 1 1 Site T1 0.75 km 1 1 1. 1 1 1 1 1 Site C1 3 km 500 1000 1500 2000 2500 1 1 1 1 Site C5 \ 10 km Figure 2-3. C:P ratio in profiles of selected microcosms from WCA-2A.

PAGE 46

40 those for dissolved 0 2 resulting from localized changes in CO, concentration associated with algal photosynthesis. Zones of high, and occasionally supersaturated, levels of dissolved O, occurred within the macro-litter layer (primarily large pieces of cattail leaves) throughout the water column of microcosms Tl, T2 and T3 (nutrient-enriched), associated with filamentous algal mats embedded within the litter. The range of pH values in these profiles was approximately 8.2 to 8.6. A similar scenario was observed for microcosms CI and C2, where the litter layer was somewhat less developed. At site C3, an extensive loosely-arranged macro-litter layer (primarily sawgrass) was covered with attached periphyton, which produced a 2-3 cm deep zone of 0 2 supersaturation. This resulted in pH levels of about 9, compared with about 8.3 in the overlying water column. The traces of dissolved 0 2 indicate diffusion from the photosynthetic source to the overlying water and underlying litter and soil. The solid horizontal lines at depth=0 in Figure 2-4 represent the approximate location of the peat surface. A layer of fine organic sediment, about 1 to 1.5 cm deep, covered the peat in all microcosms. Thus, the "sediment-water interface" actually occurred above the horizontal line at depth=0. Depletion of 0 2 with distance was more rapid in the downward direction, due to greater demand and slower diffusion in the litter, fine sediment and peat than in open water. While the prodigious amount of 0 2 generated locally within the water column was sufficient to oxygenate much or the litter layer, the peat remained anoxic, due to total depletion of 0 2 near or above the peat surface. Dissolved 0 2 profiles in microcosms T4, T5, C4 and C5, representing the transitional and low-nutrient areas in the WCA-2A marsh, reflected the reduced size and depth of litter and the presence of a benthic periphyton layer above the fine sediment and peat (Figure 2-4). Distinct peaks of dissolved 0 2 occurred near the surface of the periphyton layer, representing supersaturated conditions within the zone of maximum photosynthesis. Steep gradients of dissoved 0 2 resulted from diffusion from the narrow photosynthetic zone to the overlying water column, and especially into the sediment, which

PAGE 47

41

PAGE 48

42 represented a strong O, sink. The production of 0 2 in the periphyton and depletion in the sediment layer was accompanied by parallel increases and decreases in pH. Diel measurements of dissolved 0 2 in two selected microcosms, representing high-nutrient (Tl) and low-nutrient (C5) sites, revealed considerably different responses depending on time of day (Figure 2-5). The midday dissolved O, profile in microcosm Tl was similar to the profile in Figure 2-4 for the same site, although the diel measurement was performed on a separate replicate core. Beween noon and 6 PM, production of 0 2 in the litter algal mat had decreased considerably, due to diminished light availability. However, as a result of several hours of 0 2 production and vertical migration through diffusion, a redistribution of 0 2 had occurred, and the lower litter layer and fine sediment were partially oxidized, as well as the upper litter layer. The large 0 2 demand, created by the high availability of C and nutrients for microbial decomposers, in the Utter and sediment caused rapid depletion of dissolved 0 2 between 6 PM and midnight. During the night (by 6 AM the next day) the litter layer, which encompassed the entire water column, had become essentially anaerobic. As shown previously in Figure 2-4, significant oxygenation of the periphyton and litter layer occurred during the middle part of the day in the C5 microcosm (Figure 2-5). As in the Tl microcosm, a vertical redistribution of 0 2 had occurred by 6 PM, although substantially less symmetrical in the C5 microcosm. The absence of litter above the periphyton layer permitted greater diffusive flux of 0 2 into the water column of C5, while the significantly more consolidated sediment and peat restricted diffusion of 0 2 such that the demand exceeded the supply. While 0 2 demand was higher in the sediment-upper peat layers of T5, the unconsolidated nature of these layers allowed much greater rate of 0 2 diffusion than in C5. The decrease in dissolved 0 2 during the night was attenuated in the C5 microcosm relative to Tl, probably due to the quantity and quality (degradability) of the litter, and availability of nutrients for microbial decomposition. The result of this was an

PAGE 49

43

PAGE 50

44 oxidized water column (about 20% saturation) during the night, and a partially oxidized litter layer. Column CO ., and CH Flux Mean temperature in the microcosms for the duration of the study was 27 C. Temperatures recorded during gas flux measurements ranged from 24-30 C, thus deviating no more than 3 from the mean temperature. A temperature correction was applied to the data, based on experimental results of Volk (1973), who measured soil respiration over a range of temperature and soil moisture in intact cores of Monteverde muck (sawgrass peat) from the northern Everglades. Data from that study were used to calculate a Q 10 value of 1.82 for microbial respiration within the range of temperatures encountered during the present study. Based on this relationship, all C0 2 and CH 4 flux data were normalized to the mean temperature of 27 C. Flux data for each microcosm (i.e. individual field reps) are shown in Figure 2-6 for flooded and drained conditions, plotted as a function of distance from the inflow. The general trend for gaseous C flux (CO, + CH 4 ) from the microcosms was that of increased flux with decreasing water table and decreasing distance from the inflow, that is, there was a positive response to the nutrient gradient. However, these tendencies did not always apply when CH 4 flux was considered separately, or when flux data were divided into subgroups by water table, vegetation type or sample site. Flux of C0 2 was relatively high under flooded conditions, even when compared to flux measured under saturated (no floodwater) and partially drained conditions (Figure 26). As previously noted, aeration was provided to the water column to approximate the oxygenation brought about by algal photosynthesis. The mean flooded C0 2 flux, for all sites combined, of 2.6 U£ C cm' 2 h" 1 (horizontal dotted line) was significantly greater (a = 0.05) than the mean values of 0.8 and 1.9 ng C cm 2 h 1 for saturated (water table at soil surface) and drained (-5 cm) conditions, respectively. However, mean C0 2 flux for

PAGE 51

45 flooded microcosms was significantly lower than in more extensively drained (-10 and -15 cm) microcosms. Values for mean C0 2 flux in the latter were 3.2 and 5.0 |ig C cm" 2 h'\ respectively. Flux of C0 2 under flooded conditions decreased significantly with increasing sampling distance from the WCA-2A inflow, according to linear regression anaylsis (a = 0.05). This trend was significant primarily by virtue of the difference in means between cattail and sawgrass sites (significant according to ANOVA, a=0.05). Within vegetation types, a significant trend in flux vs. distance from inflow was observed for cattail (T1-T5) sites, but not for sawgrass (C1-C5) sites. These trends are also evident by visual inspection of data points relative to the overall mean (broken line) in Figure 2-6. The lowest C0 2 flux resulted from lowering the water level to the peat surface. Under these conditions, the litter layer was compressed at the soil surface, and remained saturated. Thus, aeration of the litter layer was restricted to diffusion from the surface. Given the high 0 2 demand of the litter, depletion of 0 2 occurred within a few millimeters of the surface of the litter. This was confirmed with spot measurements of dissolved 0 2 with a micro-electrode, as described previously. Unlike the case of flooded microcosms, there was no significant trend in C0 2 flux with distance from the inflow and no difference among cattail vs. sawgrass dominated sites. Mean C0 2 flux increased significantly with decreasing water table from 0 to -15 cm (Figure 2-6). Multiple linear regression analysis indicated a significant (a = 0.05) linear response of CO, flux to water table, with no significant response to distance from inflow. However, there was significant (a = 0.05) interaction between water table and distance from inflow. The linear model using both water table and distance as factors (independent variables) explained 85% of the variability in C0 2 flux. With water table as the only factor, the model explained 81% of the variability in flux. The positive response to water table decrease was also significant (a = 0.05) for each sampling site considered individually. Evidence of the interaction between water table and distance from inflow was the

PAGE 52

46 FLOODED A Cladium o Typha S 0 ^ Q -fr A 620864-j 2642I" O o SATURATED o-o-& a k 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 DRAINED (-5 cm) O O ( 8" 1 1 i r .. 4 1 1 1 1 1 1 1 1 1 1 1 1 1 1 DRAINED (-10 cm) O H o 0 Q A A i | i i i i | i i i i | r 620 O o o o o A A DRAINED (-15 cm) — cr e o FLOODED O 0 A & & A 10" 10"' 10" J 1111 I I I I 1 I I I I I I I I I I I I I I I I SATURATED D O Q O A A A i i i i I i i i i i i i i 1 I 10' 10" 10"' 0 -ct e -a 0 A A 8 **o A A 10" 10"' •10"' DRAINED (-5 cm) i i i i I i i i i I i i i i I i i i i I i i i i I i 1 10" 1 O" 9 ? i--A A ^ A A O •10 -1 •10 -2 •10 ,-3 i i i i | DRAINED (-10 cm) I | I I I I | I I I I | I I I I | !* 10" 1 8'8'aS •10" 1 •10"' 10" J DRAINED (-15 cm) i | i i i i | i i i i | i i i i | i i i i | r 2 4 6 8 10 0 i i i I i i i i I i i i i I i 2 4 6 I ' I 8 10 10" DISTANCE FROM INFLOW (km) CM E o o Ui 3 x O Figure 2-6. Wetland microcosm CO, (left) and CH 4 (right) flux under flooded and drained conditions. Data points represent means of replicate flux measurements for each field rep (one data point for each microcosm). Symbols differentiate between Cladium and Typha dominated sampling sites.

PAGE 53

47 development of a visible trend in flux along the distance axis with increasing depth of water table (Figure 2-6). For a water table of -15 cm this trend was significant (a=0.05), which was also the case under flooded conditions. Additionally, at -15 cm water table depth the mean C0 2 flux in microcosms from cattail sites was significantly greater than for sawgrass sites. The mean CH 4 flux (all microcosms combined) was much greater under flooded conditions (0.148 |ig C cm' 2 h" 1 ) than when the microcosms were drained, though differences among means were not significant for a = 0.05. In general, CH 4 flux was approximately one to two orders of magnitude smaller than corresponding C0 2 flux in the microcosms. Data presented in Figure 2-6 suggest that CH 4 flux at cattail-dominated sites was greater than at sawgrass-dominated sites, but this comparison was not statistically significant. A significant trend in CH 4 flux along the nutrient gradient existed only for drained conditions with water table at -15 cm. In this case, unlike the trend for CO, flux, CH 4 flux increased with distance from the inflow. When microcosms in this group were analyzed by vegetation type (cattail vs. sawgrass) the trend was significant for sawgrassdominated sites (C1-C5) only. Response of C0 2 and CH 4 flux to water table and distance from inflow is summarized in Figure 2-7 by a series of curves representing total C (C0 2 -C + CH 4 -C) flux vs. distance for each water level, with vegetation types combined. Flux of C0 2 represented roughly 90-99% of the total C flux, thus the trends in C0 2 flux discussed earlier also apply to flux of total C. In particular, total C flux increased significantly with lowering of the water table from the peat surface (saturated conditions) to 15 cm below the peat surface. In addition, the decrease in total C flux with increasing distance from water inflows was significant only for flooded conditions and a water table of -15 cm.

PAGE 54

48 7 I0 1 — i — i — | — i — i — i — | — i — i — i — | — i — i — i — | — i — i — i i 0 2 4 6 8 10 DISTANCE FROM INFLOW (km) Figure 2-7. Total C flux (C0 2 +CH 4 ) from wetland microcosms as a function of water table and distance of sample site from surface water inflow.

PAGE 55

49 Potential Soil Respiration Aerobic incubation of plant litter and peat subsamples from 2-cm depth intervals provided an index of potential soil respiration, or organic C mineralization rate, in the field. An useful method of comparing mineralization rate among sites is to express soil respiration as specific C loss (mg C0 2 -C g 1 soil C d 1 ) (Figure 2-8). This form of expression normalizes soil respiration rate to the C content the substrate. Specific loss rate is equivalent to the first order rate constant (k) for the exponential decay model C(t) = C 0 ek \ [2-1] where C is concentration or mass as a function of time and C 0 is the initial quantity of C. Data points in Figure 2-8 can be expressed as k = d' 1 by multiplying the values by 10' 3 (i.e. g 1 C g 1 C d" 1 = d" 1 ). Mean specific CO, loss rate for the Tl (highly nutrient impacted site) soil profile and litter (0.89 mg C g"' C d"') was approximately double the values for the CI (0.50), C5 (0.39) and T5 (0.38) profiles. The highest rates in all profiles occurred in the litter layer and near the peat surface (Figure 2-8), indicating higher microbial activity in those strata. In the T5 and C5 soils, representing transitional and unimpacted sites, rates in the lower 20 cm of the profile were well below 0.5 mg C g" 1 C d"', with minimum rates in those profiles of 0.07 and 0.05 mg C g" 1 C d"\ respectively. However, the T5 profile was characterized by higher rates near the surface than the C5 profile. In the CI profile, change in specific CO, loss rate from top to bottom was not as pronounced, that is, rate in the upper portion of the profile was not elevated to a great extent, however the sharp decrease in rate which occurred in T5 and C5 was observed only below about 22 cm in CI. In comparison, specific CO, loss rate in the Tl profile was highly elevated near the surface, and although the rate dropped substantially below about 5 cm, remained above 0.2 mg C g" 1 C d" 1 through the entire profile.

PAGE 56

50 SPECIFIC C LOSS (mg C0 9 -C g" 1 soil C d" 1 ) 0.5 1.0 1.5 2.0 2.5 3.0 t I i I I | | I i i !„ 0 0.5 1.0 1.5 2.0 2.5 3.0 i i i i 1 1 1 Site T5 5 km i i t 1 1 1 1 1 1 1 1 1 1 1 Site C5 10 km Figure 2-8. Potential soil respiration (aerobic incubation) of litter and 2-cm depth increments of selected WCA-2A microcosms, expressed as C-specific C0 2 loss (i.e. soil organic C basis). Values represent means and standard error of 3 replicate incubations.

PAGE 57

51 The range of values for specific C0 2 loss rate in the Tl profile (highest impact) is equivalent to a range of 0.0002 to 0.003 d 1 for the first-order decay constant k, under aerobic conditions. Similarly, the corresponding values for k in the C5 profile ranged from 0.00005 to 0.001 d" 1 Thus, the litter and surficial peat are substantially more degradable than the older peat lower in the profile. Based on these results, turnover time (for aerobic conditions) ranged from about 1 2.7 years for litter and 14 55 years for peat in the lower part of the 30-cm profiles. For comparison to microcosm C flux measurements, the aerobic C0 2 loss rate in the laboratory incubation experiment was expressed as aereal loss rate, determined individually for litter and 2-cm depth increments (Figure 2-9). The resulting trends in C loss rate were similar to those observed for specific C loss. However, the differences in rate among the four sites were not as pronounced. Lower bulk density at sites closer to the inflow (e.g. site Tl) resulted in attenuated values when expressed on an aereal basis. Integration of the curves in Figure 2-9 yielded the sum total of potential, or aerobic, respiration in the soil profiles. The resulting values for the Tl, T5, CI and C5 profiles were 22.8, 11.2, 21.8 and 14.1 |ig C cm" 2 h"', respectively. These values may be considered "potential" or maximum rates under completely aerobic conditions, compared with total gaseous C (C0 2 + CH 4 ) flux measured on flooded and partially drained microcosms. In general, potential (aerobic) respiration rates were about 3 times greater than those measured under partially-drained (water table at -15 cm) conditions. Discussion The purpose of this study was to evaluate the response of heterotrophic microbial activity in the Utter and peat along a nutrient gradient, under flooded and drained conditions. A single set of soil microcosms, with sequential levels of treatment (water table) imposed on the soil cores, was used to determine the influence of transient changes in water table depth on soil respiration. A benefit of this experimental design was the

PAGE 58

52 AREAL C LOSS (ng CO--C cm" 2 h" 1 ) 4 0 1 2 3 4 i i 1 1 Site T5 5 km Site C1 3 km 1 1 1 1 1 1 1. i i i i i i i i i : < Site C5 10 km ii Figure 2-9. Potential soil respiration (aerobic incubation) of litter and 2-cm depth increments of selected WCA-2A microcosms, expressed as CO, loss on an areal basis. Values represent means and standard error of 3 replicate incubations.

PAGE 59

53 elimination of spatial variability from statistical comparison of flux at various levels of water table. This required the assumption that changes observed in the response (flux) were a function of changes in water table, and not due to an unrelated, time-dependent factor. Measurement of flux during the "equilibration" period (flooded conditions), which was of greater duration than the subsequent period of draining, showed that microbial respiration activity had stabilized, that is, initial changes due to core transport or other disturbance were no longer occurring. Vertical profiles of dissolved 0 2 in the water column and litter layer of the microcosms (Figures 2-6 and 2-7) were evidence of the potential for significant, though highly localized, oxidized zones in the litter layer. Light availability in the water column, and penetration into the litter layer, under field conditions is dependent on density of the plant canopy (including floating vegetation) and of plant litter within the water column. The oligotrophic sawgrass marsh, as well as the aquatic sloughs, in the interior portions of WCA-2A generally support a lower standing crop of live macrophytes and plant litter (Toth, 1987; Davis, 1991). Light availability in the water column is quite high, giving rise to extensive periphyton growth (SFWMD, 1992). It should be expected that the water column, and at least the upper portion of plant litter, would be substantially oxygenated during much of the day. On the other hand, assumptions that the water column, and especially the litter layer, in eutrophic areas of the marsh are often completely reduced (Belanger et al., 1989), should be qualified, according to results of the present study. Although the native periphyton has been displaced in the nutrient-enriched areas of WCA-2A, various types of non-native algal mats have been observed within the macrolitter (referring to size as well as origin), depending on light availability. In spite of the high 0 2 demand created by the relative abundance of C and nutrients in the plant litter, the 0 2 supplied by algal photosynthesis may be significant in terms of supporting aerobic microbial heterotrophs. The highly localized nature of oxygenation in the litter, created by steep gradients between pockets of 0 2 supply and demand makes quantification of the

PAGE 60

54 extent of oxygenation difficult. Furthermore, spot measurements of dissolved 0 2 in the field, without consideration of spatial or temporal variability, could give very misleading information on 0 2 availability in the water column and litter layer. The artificial aeration system utilized in the microcosm water columns created a relatively uniform level of oxygenation, both spatially and temporally, tending to average out regions of high and low concentration. Aeration was provided to the water column of the flooded microcosms to compensate for the lack of photosynfhetic activity and prevent a condition of complete anoxia in the water column. As discussed earlier, photosynthesis in the microcosms would have presented numerous problems associated with accurate determination of C0 2 production. Flux of C0 2 from the calcareous peat could have been drastically affected by shifting of the carbonate equilibrium as a response to depletion of C0 2 by photosynthetic activity. Presumably, constant aeration of the water column greatly enhanced decomposition of the extensive, and relatively labile, plant litter and algal detritus. The potential contribution of this component is considerable, as demonstrated during the controlled incubations in the laboratory. As a result, C0 2 flux from flooded microcosms probably was more representative of maximum daytime levels under field conditions. Taking into account diurnal variation in dissolved 0 2 concentration in the water column and litter, the mean daily rate of respiration in the field was probably considerably lower than indicated by microcosm measurements. Therefore, flux measurements made during aeration of the microcosm water column should be considered maximum, or potential, rates sustainable in the field for short periods of time and are useful primarily for comparison among sites along the nutrient gradient and for scaling rate parameters in an ecosystem model. Methane flux data exhibited much greater variability than C0 2 flux measurements. Because of the absence of emergent macrophytes, an important conduit for CH 4 emissions from wetlands (Schutz et al., 1991), and the low solubility of CH 4 in water, ebullition was the primary mechanism for efflux of CH 4 from the microcosms. Since ebullition is an

PAGE 61

55 intermittent process, CH 4 measurements taken over short time intervals would presumably yield highly variable results. A secondary, albeit important, source of uncertainty inherent in measuring CH 4 flux across an anaerobic-aerobic interface is the loss of CH 4 through oxidation to CO,. This occurs in aerobic regions of soil and floodwater through the action of methane oxidizing bacteria which utilize CH 4 as an energy source and O, as the terminal electron acceptor. Thus, the amount of CH 4 collected above the microcosm soil or floodwater surface may substantially underestimate the actual CH 4 production in the soil. Happell and Chanton (1993) estimated that an average of 46% of CH 4 produced in a north Florida swamp forest soil was oxidized to C0 2 near the sediment surface. Conversely, it would be incorrect to assume that all of the C0 2 released from the microcosms is attributable to the activity of aerobic heterotrophs. Oxidation of methane to C0 2 by methanotrophic bacteria is potentially a major sink for methane in wedands (Kelly and Chynoweth, 1979; Schipper and Reddy, 1996). Furthermore, C0 2 is also a secondary product of methanogenic bacteria, and is evolved in varying proportion to CH 4 during methanogenesis (Oremland, 1988). An additional source of CO, production in anaerobic soils or microsites is sulfate reduction. Normally a more significant process in salt marshes due to elevated S0 4 2 levels (Howarth, 1993), it may be a major pathway of anaerobic decomposition in WCA-2A, given the high concentrations of S0 4 2 measured in the floodwater (Schipper and Reddy, 1994). Measurements of CH 4 flux in wetlands have generally indicated lower rates under drained or partially drained conditions, because of the decreased number anaerobic sites and increased consumption (oxidation) of CH 4 by aerobic soil microorganisms (Happell and Chanton, 1993; Moore and Dalva, 1993). This trend was not supported by results from the present study. Mean flux of CH 4 did not change significantly under drained conditions, between water table depths of 0 and 15 cm (Figure 2-6). It is possible that production of CH 4 decreased during that time, but the decrease was offset by increased emission of trapped CH 4 bubbles from the lower, saturated portion of the profile, as the

PAGE 62

56 pressure head decreased with drainage. This type of response to changing water table might be expected to occur under transient conditions of drainage. Soil respiration in the microcosms increased significandy along the nutrient gradient in WCA-2A, from areas of low to high impact (oligotrophic to eutrophic). Analysis of total N and P on four selected soil profiles indicated greater P enrichment toward the S-10C inflow, while total soil N was relatively unchanged. Previous studies in WCA-2A have shown a well-defined soil and water P gradient south (downflow) of the S-10 inflows (Koch and Reddy, 1992 ; DeBusk et al., 1994). Increased accumulation of both N and P has occurred in the impacted regions of WCA-2A, through accelerated accretion of organic matter (Craft and Richardson, 1993; Reddy et al., 1993). However, enrichment of the peat with N (higher concentration on a mass basis) has not occurred to a great extent, while considerable P enrichment has occurred (DeBusk et al., 1994). Thus, it would appear that response of C flux (soil respiration) to distance from inflow would be the result of P enrichment near the inflow. Data from the laboratory incubation study support this argument. A significant relationship (a = 0.05) was found between total P concentration and aerobic soil respiration, but respiration and total N content were not significantly correlated. However, a linear model of respiration as a function of both total P and total N showed a significant (a = 0.05) interaction between the two factors. It is possible that N becomes a limiting nutrient for microbial activity in regions of excessive P enrichment. The multiple linear regression model, however, explained only 46% of the variability in soil respiration, in part because the relationship between total P and respiration was not linear. Respiration was better described as a power function of total P (total P raised to a power), such that a log-log plot of respiration vs. total P yielded a straight line. This model accounted for 67% of the variability in respiration. The improved fit using this type of model probably results from the log-normal distribution of both sets of data. However, visual inspection of the data suggests that respiration may actually be asymptotic at higher levels of total P. This

PAGE 63

57 also may suggest a co-limiting factor for microbial activity, possibly N. A portion of the variability in respiration not explained by variation in total P may also be due to differences in C availability, since the data set from the laboratory study include plant litter and peat from a wide range of depth and therefore age and biodegradability. A major effect on soil respiration in the microcosms was water table. Increase in respiration as the water table was lowered (disregarding flooded conditions, which are not directly comparable) was highly significant. Increased soil respiration with drainage (which results in increased 0 2 availability) has been reported in other studies. For example, a linear increase in C0 2 flux with decreasing water table to a depth of 50 cm was reported for undisturbed cores of Montverde muck (sawgrass peat) from the Everglades (Volk, 1973). Similarly, C0 2 flux increased linearly with decreasing water table to a 40 cm depth in intact peat cores from Canadian wetlands (Moore and Dalva, 1993). Of particular interest in this study, however, is the interactive effect of water table (0 2 availability) and distance from inflow (P enrichment) on respiration (Figure 2-7). A positive response of respiration to decreasing water table would result in an upward shift in the plot of C flux vs. distance, with no change in slope. However, in this case, both a shift along the Y-axis and a significant increase in slope were observed, clearly indicating an interactive effect. Examination of the plots in Figures 2-8 and 2-10 provides more insight into this interaction. As discussed previously, the aerobic incubation of litter and peat subsamples from selected microcosms was considered as a measure of potential respiration (not limited by 0 2 availability). Specific C loss (Figure 2-8) provided a means for direct comparison of relative biodegradability, since these values were based only on the organic component of the peat or litter, and variation in respiration due to availability of 0 2 had been removed. The trend of low biodegradability in the older peat (lower depth) and increasing near the surface was common to all profiles, but more exaggerated in profiles from nutrient-enriched areas. This is well-defined in Figure 2-10, which shows cumulative effect of this trend as one moves down the profile. It also relates potential respiration rate to

PAGE 64

58 2 -1, POTENTIAL RESPIRATION (^ig Ccm'h') LITTER 0-2" 2-4 4-6" 6-8' 8-1 0" 10-12' 12-14' 14-16" 16-18' 18-20" 20-22" 22-24' 24-26' 26-28 28-30' LITTER' 0-2" 2-4" 4-6" 6-8" 8-10" 10-12" 12-14" 14-16" 16-18" 18-20" 20-22' 22-24" 24-26" 26-28' 28-30" 0 5 10 15 20 25 0 j i 1 1 1 Site T1 0.75 km 5 10 15 20 25 1 1 1 1 1 1 1 1 Site T5 5 km 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Site C1 3 km l„J I I I I I I I I I I I I I I I I I I I I I I Site C5 10 km Figure 2-10. Potential soil respiration (aerobic incubation) of litter and 2-cm depth increments of selected WCA-2A microcosms, expressed as cumulative, or depthintegrated CO, loss. The bottom value represents potential respiration for the entire litter and soil profile.

PAGE 65

59 surface area, to account for differences in bulk density at these sites. It is apparent that relatively labile, or decomposable, organic matter has accumulated to a greater depth at the more highly nutrient-impacted sites. Thus, not only has there apparently been a Penrichment effect, but also a P-accumulation effect, that is, a larger accumulation of the more recent, enriched organic matter associated with nutrient loading during the past 3 decades. In support of this concept are estimates of accretion based on ,37 Cs dating of intact cores which show significantly increased rates of peat accretion closer to the inflows in WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993). Potential soil respiration rate (determined from aerobic incubations) expressed on a cumulative soil area or volume basis, greatly exceeded actual measured rates in corresponding intact microcosms, even under partially drained conditions. Whole-column gaseous C flux from microcosms Tl, T5, CI and C5 were 5.8, 5.5, 4.9 and 3.8 ug C cm" 2 h' 1 respectively. In comparison, potential rates for combined depths, integrated over the entire 30 cm profile, were 22.8, 11.2, 21.8 and 14.1 ug C cm" 2 h"\ for the same sites. If potential respiration only in the upper 15 cm of the profile (including litter) is considered, extrapolated values are 14.8, 8.5, 16.5 and 12.1, still substantially higher than actual rates. This suggests that 0 2 availability was limited in the soil above the water table. Observations made during the experiment support this possibility, since it was apparent that the peat was saturated above the water table. This saturated zone, or capillary fringe, was approximately 2-3 cm thick, by visual inspection, thus 0 2 availability might be expected to be close to zero in this portion of the profile. Undoubtably, a substantial number of microsites in the soil above the capillary fringe were also water-filled, due to the hysteresis effect commonly observed during soil drainage. The water content of the soil above the capillary fringe would be expected to decrease with time. Thus, the response of soil respiration rate to soil drainage is probably time-dependent. Time was not included as a variable in the present study, although the significance of time as a factor in respiration response to drainage was considered beforehand. Instead, the intent of the microcosm study was to approximate

PAGE 66

60 transient hydrologic conditions for measuring the response of soil respiration to drainage or drought. This was also a consideration for imposing sequential levels of treatment on the same set of microcosms, as discussed earlier. In comparison to C flux measured during the present study, the average C0 2 flux from intact (80 cm deep) peat columns from 3 Canadian wetlands ranged from 1.9 to 1 1.7 |ig C cm' 2 h' 1 under saturated conditions and 9.0-15.8 (ig C cm" 2 h" 1 for a water table depth of 40 cm (Moore and Dalva, 1993). Total C flux from undisturbed cores of from the Everglades (Montverde muck, a sawgrass peat) averaged 8.4 ^g C cm' 2 h" 1 for a water table depth of 15 cm (Volk, 1973). Potential respiration (aerobic incubation) integrated over the upper 30 cm of peat from short pocosin, tall pocosin and gum swamp sites in North Carolina averaged 6.5, 12.5 and 1 1.5 |ig C0 2 -C cm' 2 h"' (Bridgham and Richardson, 1992). The sampling sites ranged from nutrient-deficient to moderately nutrient-rich, respectively. Using in situ measurement of CO, and CH 4 flux from north Florida swamp forests, Happell and Chanton ( 1993) estimated the average flux of CO, under flooded conditions to be approximately 3-5 times less than under drained conditions. After adjusting for estimated root respiration, the average rate of C mineralization under flooded conditions, based on combined C0 2 and CH 4 fluxes, was 1 1.9 mol C m 2 y "', equivalant to 1.6 |ig C cm 2 h' 1 In situ measurement of CH 4 flux from a WCA-2A slough community, using a static chamber method, yielded an average value of 0.09 g m 2 d"\ equivalent to 0.28 ug C cm" 2 h" 1 (Schipper and Reddy, 1994) Living macrophytes were not included as part of the microcosms, in order to differentiate respiration of microbial decomposers from plant respiration. It must be realized, when extrapolating results of this study to field conditions, that macrophytes interact closely with soil microorganisms, therefore microbial processes in the microcosms were certainly affected to some degree. For example, wetland macrophytes provide various amounts of organic C and 0 2 to the rhizosphere, both of which affect the metabolism of heterotrophic microbes, at least near the rhizoplane (root surfaces) (Reddy and D'Angelo,

PAGE 67

61 1994). On the other hand, living roots which were severed during coring might have become a source of labile C, though not a source of 0 2 for microbial decomposers. Periphyton presented an additional problem, since neither algal photosynthesis nor respiration were desirable during C flux determinations. However, algae was considered to be physically inseparable from the litter layer, especially fine particulate organic detritus, therefore no attempt was made to remove it. Therefore, dead algal biomass resulting from total shading of the microcosms during flux measurements was another potential source of labile C for decomposers. However, the most labile pool of organic C originating from recently dead plant biomass was probably mineralized during the period of about 2 weeks between the onset of shading and measurement of C flux. Summary and Conclusions Based on results of this study, the following conclusions were made: Photosynthetic activity of the native periphyton in unimpacted areas and filamentous algae in the extensive litter layer of impacted areas may provide significant oxidation of the litter layer during the daytime, potentially enhancing decomposition. Soil respiration increased with lowering of the water table, in direct proportion (linear response) to depth of the water table; however, magnitude of the response increased with the degree of nutrient enrichment. Response of soil respiration to nutrient enrichment was significant under flooded and highly drained conditions. There was a statistically significant interaction between water table depth and nutrient enrichment as factors affecting soil respiration rate, however respiration rate was more dependent on depth of the water table. Potential soil respiration was found to be a good indicator of actual respiration, and was significantly correlated with soil total P concentration.

PAGE 68

CHAPTER 3 TURNOVER OF ORGANIC CARBON POOLS ALONG THE WCA-2A NUTRIENT GRADIENT Introduction A primary characteristic of wetlands which distinguishes them from upland ecosystems is their propensity for accumulating organic carbon (C). Organic C accumulation in wetlands is the net result of primary production (C fixation) and decomposition (C mineralization). In wetlands with extended hydroperiod, decomposition of dead plant material proceeds at a reduced rate, leading to substantial accumulation of organic C (as organic matter), which may include extensive peat deposits. Decomposition of organic matter is governed by the chemical composition of the substrate and external, or environmental, factors. Among the more important environmental factors are temperature, moisture, nutrients and electron acceptors (Swift et al., 1979; Heal et al., 1981; Reddy and D'Angelo, 1994). Unlike terrestrial ecosystems, decomposition in wetlands is generally not moisture-limited and is frequently electron acceptor-limited. In the latter case, the primary controller of decomposition is, of course, 0 : In many freshwater wetlands, however, alternate electron acceptors for anaerobic microbial respiration, such as NO, Mn 4 *, Fe 3+ and S0 4 = are often in short supply relative to available organic C, leaving methanogenesis as the principal mode of respiration (Westermann, 1993). Nutrient availability affects decomposition rate through its effects on microbial growth. Although nutrient loading is typically greater in wetlands than in uplands due to location within the landscape, nutrient availability may be low relative to the pool of available organic C in wetlands (Reddy and D'Angelo, 1994). Nitrogen (N) and phosphorus (P) both have been identified as microbial growth-limiting nutrients in 62

PAGE 69

63 wetlands (Westermann, 1993). Nitrogen, unlike P, may be lost from wetlands through microbial metabolism via denitrification, as well as through ammonia volatilization (Reddy and D'Angelo, 1994). In addition to environmental conditions, decomposition rate is significantly affected by chemical and physical composition of the organic substrate (Swift et al, 1979; Heal et al, 1981) The term "substrate quality" generally refers to the availability of C compounds and associated nutrients for microbial utilization (Heal et al., 1981; Heal and Ineson, 1984). Lignin and cellulose fractions are generally considered to be key components of the "C quality" of an organic substrate (Colberg, 1988; Moran et al., 1989). Lignin is more resistant to breakdown than cellulose, therefore substrate cellulose content decreases more rapidly during decomposition (Colberg, 1988; Melillo et al., 1989). Degradation of lignin and cellulose is slower under anaerobic conditions, although the relative decay rates for each remains approximately the same (Benner et al., 1984). Initial substrate composition has been used as a predictor of decomposition rate (Swift et al., 1979). In several cases, initial lignin content and lignin:N ratio of plant tissue were shown to be highly correlated with decomposition rate (Godshalk and Wetzel, 1978; Berg and Staaf, 1981; Melillo et al., 1982) The ligno-cellulose index (LCI) was proposed as a measure of substrate C quality along the "decay continuum" of plant leaves to soil organic matter (Melillo et al, 1989). increased LCI during decomposition reflects the increasing concentration of lignin relative to the total ligno-cellulose content of the organic substrate, thus the substrate becomes increasingly resistant to decomposition. In addition to the relative increase in plant lignin, an accumulation of lignin derivatives generated as by-products of microbial metabolism (i.e. humus) occurs during decomposition (Zeikus, 1981). Microbial decomposers, primarily bacteria and to a certain extent fungi and protozoa, play the lead role in C cycling and energy flow in wetlands (Benner et al., 1984; Wetzel, 1984; Westermann, 1993). Although they represent only a small fraction of the

PAGE 70

64 total C and organic matter in soils, microbial decomposers are responsible for processing nearly all of the organic C produced in the ecosystem (Jenkinson and Ladd, 1981; Van Veen et al., 1984). Research in terrestrial ecosystems has shown that the microbial biomass also constitutes a major C sink, in that it represents a significant portion of the "active" organic C (Paul and van Veen, 1978). Repeated cycling of organic C through the microbial biomass results in loss of organic C from the detrital pool via aerobic (C0 2 loss) or anaerobic (C0 2 and CH 4 loss) microbial respiration. This same process also results in changes in substrate composition as original plant material is lost and by-products of microbial metabolism accumulate (Swift, 1982; Heal and Ineson, 1984). The current study examines the influence of nutrient loading on selected microbial processes regulating turnover of organic C pools in a northern Everglades marsh. The Everglades encompasses a variety of wetland ecosystems which were historically adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred primarily through rainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974) (Figure 3-1). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959). Recent development of the Everglades and surrounding watershed has created changes in nutrient loading and hydrology (SFWMD, 1992). Most significantly, a large area of the northern Everglades was drained and converted to agricultural production during the first half of this century. This area of sugar cane, vegetable and sod farming is referred to as the Everglades Agricultural Area (EAA). The remainder of the northern Everglades was divided into three Water Conservation Areas (Figure 3-1) in the 1960s, for water storage and flood control. Water level within the WCAs is controlled by a system of levees,

PAGE 71

Figure 31 Site map for WCA-2A study area, showing locations of sampling sites. Coordinates for sampling sites are listed in Table 3-1.

PAGE 72

66 pumps and floodgates. Currently, the Everglades consists of the WCAs to the north and Everglades National Park to the south. Drainage of the EAA has resulted in widespread oxidation of the organic soil and concomitant mineralization and leaching of organically-bound nutrients. As a result, nutrients from organic soil mineralizat ion, along with additional nutrients from fertilizers have been transported via draina ge canals toward the WCAs jOT^^oxum^lx_30j^s. Loading of agricultural drainage water into the WCAs has resulted in nutrient enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus (P) enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod. in the encroachment of cattail (Typha domingensis Pers.) and other rapidly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward andOrnes, 1983; Toth, 1987, 1988). Accelerated nutrient loading in northern WCA-2A (Figure 3-1) during the past three decades has created a distinct nutrient (especially P) gradient in water, soils and plant tissue (Davis, 1991; Koch and Reddy, 1992; DeBusk et al, 1994). Changes in species composition of periphyton and macrophyte communities, along with an overall increase in net primary productivity have been documented along this gradient (Davis, 1991; SFWMD, 1992). Soil dating by analysis of i37 Cs peaks has indicated that peat accumulation rate has increased in nutrient-enriched areas of WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993). The main objective of this research was to determine the effect of nutrient enrichment on turnover of organic C pools along the WCA-2A nutrient gradient. A further objective was to examine the relationships between size of the microbial biomass C pools and turnover time of associated organic C pools. It is hypothesized that turnover time for major C pools increases (decomposition rate decreases) downgradient from the inflow of nutrient-laden water. It is also hypothesized that turnover time increases in successively older organic C pools.

PAGE 73

67 Materials and Methods Site Description Field study sites were located in WCA-2A, a 447 km 2 region of the northern Everglades (Figure 3-1). Surface water flows into WCA-2A from the Hillsboro Canal through the four S10 water control structures and from the North New River Canal through the S-7 pump station. Most of the hydraulic loading is through the S10C and S10D structures into the northern portion of WCA-2A. The general direction of flow is from north to south. Water depth is usually less than one meter, and varies considerably, both seasonally and year-to-year, with occasional dry periods (SFWMD, 1992: personal observations). The bulk of the surface outflow is through through three control structures at the south end of WCA-2A, into WCA-3 (Figure 3-1). Soil in WCA-2A consists of Everglades and Loxahatchee peats (Gleason et al., 1974). Everglades peat, the most common soil in the Everglades, is associated with the sawgrass marsh community. It is dark brown, finely fibrous to granular, with circumneutral pH, relatively high N content and low SiO.,, Fe and Al content. Peat depth in WCA-2 A ranges from about 1 to 2 m, and age of basal peats is estimated to be 2000 to 4800 yr. Beneath the peat lies a bedrock of Pleistocene limestone, with intermediate layers of calcitic mud, sandy clay and sand in several areas (Gleason et al., 1974). The primary sources of nutrient loading to WCA-2A are the S-10 structures which convey water from the Hillsboro Canal and WCA-1 (Figure 3-1). A distinct gradient of N and, most significantly, P enrichment in water, plants and soil has formed between the high-nutrient region adjacent to the inflows and the low-nutrient interior marsh of WCA-2 A (Koch and Reddy, 1992; SFWMD, 1992; DeBusk et al., 1994). A vegetation gradient coincides with the nutrient gradient; most notable is the gradient from sawgrass marsh with scattered aquatic slough in the interior to cattail and mixed emergents near the inflows. The

PAGE 74

68 vegetation gradient was divided into three discrete categories for the purposes of the current study: cattail-dominated, sawgrass-dominated and mixed cattail and sawgrass (Figure 3-1). Ten field sampling sites were established along the nutrient and vegetation gradient on a 10 km transect extending from the S-10C inflow into the interior marsh. The sites, numbered 1 through 10, were located at distances of 0.07, 0.3, 0.6, 1.2, 2.0, 2.9, 3.9, 4.8, 6.6 and 9.8 km downstream from the inflow (Table 3-1; Figure 3-1). Plant and Soil Sampling Live and standing dead (attached to plant) cattail and sawgrass leaves were collected at sites along the sampling transect in March and June 1995. Live cattail leaves were collected at sites 1, 6 and 8 and live sawgrass leaves at sites 6, 8 and 10 on March 8. In addition, dead cattail leaves were collected at sites 1 through 8 and dead sawgrass leaves at sites 6 through 10. On June 6, live and dead plant leaves were collected from cattails at sites 1 through 6 and from sawgrass at sites 5 through 10. Five entire leaves were removed per plant, and repeated for three different plants at each site. Leaf samples were dried in a forced-draft drying room at 60 C, then all leaves in each sample were cut into pieces of about 2 cm length and mixed thoroughly to produce a homogeneous composite sample. All live and standing dead plant samples were analyzed for total C, N and P content. Lignin and cellulose content were determined for selected samples from the March 1995 sampling event. These included live cattail from sites 1 and 6, live sawgrass from sites 6 and 10, dead cattail from sites 1, 4 and 6 and dead sawgrass from sites 6, 9 and 10. Soil cores were also obtained during the June 1995 field sampling event. The plant litter layer on the soil surface was sampled first, by placing a short section of 15 cm diameter PVC pipe over the sample area and manually transferring the litter contained within the pipe to a zippered plastic bag. A serrated knife was used to cut through the litter, around the inside perimeter of the pipe, to precisely delineate the sample area. Next, a simple coring apparatus consisting of 7.6 cm diameter aluminum pipe was used to obtain

PAGE 75

69 Table 3-1. Locations of sampling sites in WCA-2A, and distance from S-10C inflow. Distance Site Latitude N Longitude W from inflow deg min deg min km 1 26 22.09 80 21.07 0.1 2 26 21.98 80 21.09 0.3 3 26 21.81 80 21.12 0.6 4 26 21.53 80 21.20 1.2 5 26 21.05 80 21.21 2.0 6 26 20.53 80 21.27 2.9 7 26 20.02 80 21.37 3.9 8 26 19.52 80 21.39 4.8 9 26 18.51 80 21.46 6.6 10 26 16.81 80 21.48 9.8

PAGE 76

70 intact samples of the top 30 cm of the soil profile. The coring tube was pushed into the soil within the area from which litter had been removed. A serrated knife was used to cut through the fine root mat at the top of the soil profile, facilitating penetration of the coring tube without compaction of the soil. When additional force was required, the coring tube was hammered into the soil after being fitted with a solid aluminum plug (with an air vent) to distribute the force of impact from the mallet. The coring tube was then excavated, and the intact soil core extruded through the top using a plunger apparatus. The upper 10 cm of the peat profile was separated as it was extruded, and placed in a zippered plastic bag. Similarly, the 10-30 cm layer of peat was extruded and placed in a separate bag. The 0-10 and 10-30 cm layers of the peat profile were designated as "surface peat" and "buried peat". The above procedure was repeated three times at each sampling site, within a radius of approximately 5 m. Samples contained in the sealed plastic bags were immediately placed on ice and transported to the lab within 24 hours. After removal of live roots from the samples, wet weights were recorded for each, then replicate samples were thoroughly mixed to create a single composite sample. Subsamples of litter and peat were dried to constant weight in a forced-draft oven at 60 C, for determinations of moisture content and dry bulk density. The composited samples were stored in leak-proof polypropylene jars in a refrigerator at 4 C. Plant and Soil Analysis Total C and N analysis was performed on dried, finely ground (< 0.2 mm) samples using a Carlo-Erba NA-1500 CNS Analyzer (Haak-Buchler Instruments, Saddlebrook, NJ). Total P analysis was performed on separate subsamples following combustion (ashing) at 550 C for 4 h in a muffle furnace and dissolution of the ash in 6 M HC1 (Anderson, 1976). The digestate was analyzed for P using the automated ascorbic acid method (Method 365.4, USEPA, 1983). Lignin and cellulose content were determined by a standard procedure using acid-detergent and H 2 S0 4 extractions (AOAC, 1990).

PAGE 77

71 Litter and peat samples were analyzed for C in the microbial biomass using the chloroform fumigation-extraction (CFE) technique (Horwath and Paul, 1994), with a slight modification. Field-moist samples (ca. 0.5 g dry mass) were fumigated in a vacuum dessicator with ethanol-free chloroform, which was contained in a beaker next the samples. Immediately prior to fumigation, 0.5 mL of chloroform was added directly to each sample, to enhance distribution of chloroform within the wet samples (Ocio and Brookes, 1990). The remainder of the fumigation and extraction process was performed according to Horwath and Paul (1994). Triplicate fumigated and non-fumigated samples were extracted with 25 mL of 0.5 M IC,S0 4 centrifuged and the supernatant filtered though glass fiber filters (Gelman A/E, Gelman Sciences, Ann Arbor MI) using a vacuum filtration system. The ratio of extractant to dry soil was increased over the recommended 5:1 (w/w) because of the extremely high organic content of the soil and litter. The filtered extracts were analyzed for total organic C (soluble organic C) on a Dohrmann DC190 TOC analyzer (Rosemount Analytical Inc., Santa Clara, CA). Microbial biomass C was calculated from the difference in IC,S0 4 extractable C between fumigated and non-fumigated samples. A correction factor (k BC ), which accounts for the efficiency of the fumigation process, is generally used to obtain a direct estimate of microbial C from the flush of extractable C following fumigation (Horwath and Paul, 1994). A value of k EC = 0.37 was used in this case, based on extensive calibration previously carried out for organic soils (Sparling et al., 1990). Microbial Respiration Microbial respiration associated with decomposing standing dead material, plant litter and peat was measured for estimation of C mineralization rate. Aerobic respiration was measured for standing dead leaves using bottle incubations and measurement of accumulated headspace C0 2 Triplicate 1 g (dry weight) subsamples of standing dead cattail and sawgrass leaves from all sites and collection dates were placed in separate 160 mL

PAGE 78

72 serum bottles. Deionized water (5 to 7 mL as required) was added to the dried plant material to re-wet the substrate for microbial activity. Each bottle was stoppered with a sleeve-type rubber septum and incubated in the dark at 25 C. After a 24-hour preincubation period, headspace gas was sampled in the bottles. Sampling was repeated after 12 and 24 hours, and subsequently every 24 hours for a total of four days of sampling. Headspace gas was sampled (1 mL) using zero dead volume 1 mL insulin syringes (Becton Dickinson, Lincoln Park, NJ). Syringe needles were inserted into a rubber stopper for short-term storage of samples prior to analysis on a gas chromatograph (GC) for C0 2 Increase in C0 2 in the bottle headspace was linear over the four-day period. Rate of C0 2 evolution was calculated from the slope of the best-fit linear regression line. Soil (litter and peat) respiration was measured using an air flow-through system (respirometer) which traps evolved C0 2 on a continuous basis (Zibilske, 1994). Use of a continuous flow apparatus for respiration measurement is recommended for calcareous soils, such Everglades peat, to avoid problems associated with retention of microbiallyevolved C0 2 as bicarbonate (Martens, 1987). The respirometer consisted of an air supply (compressed air cylinders), C0 2 scrubber (2N NaOH trap), incubation vessels and a separate C0 2 trap (0.05 to 0. 1 N NaOH) for each incubation vessel. These components were connected by plastic (PVC) tubing and a gas manifold to create a continuous flow of C0 2 -free air through each incubation vessel and into the individual C0 2 traps. Incubation vessels were fashioned from 250 mL Mason jars with air-tight lids. The jar lids were modified to accomodate plastic fittings with valves for controlling gas inflow and outflow. Air flow rate through each incubation vessel was maintained at 25 mL min '. The NaOH traps provided continuous collection of CO, evoloved during microbial respiration. NaOH in the traps was changed periodically, as determined by the rate of C0 2 accumulation. Free (remaining) NaOH in the traps was titrated with standardized HC1, following addition of BaCl 2 to precipitate the Na 2 C0 3 in solution as insoluble BaC0 3 to

PAGE 79

73 determine the amount of NaOH which had reacted with CO-,. Molar quantity of C0 2 evolved during the incubation period was determined stoichiometrically (Zibilske, 1994). Subsamples ( 10 g wet weight) of litter, surface peat (0-10 cm depth) and buried peat ( 10-30 cm depth) were placed in 50 mm diameter plastic petri dishes. A thick glass fiber prefilter was placed beneath each sample to increase aeration of the sample by maintaining drained, yet moist, conditions in the sample. The petri dishes containing the samples were placed inside the Mason jars and incubated in the dark at 25 C. A 48 hour pre-incubation period was found to be sufficient time to achieve stabilization of respiration rate. Samples were incubated for one week following pre-incubation, then removed from the incubation chambers, dried at 60 C in a forced-draft oven and weighed to determine sample dry mass. The incubation was performed in triplicate through successive incubation of the entire sample set. Anaerobic respiration was measured in the same fashion, with the following modifications. The source air was replaced by N 2 gas (prepurified grade) and sample size was increased to 50 g wet weight. Incubation period was shortened to three days, following a 48 hour pre-incubation. A preliminary study suggested that a longer incubation period would result in depletion of alternate electron acceptors, and would therefore not reflect conditions in the field at the time of sampling. At the end of the three day incubation period the inlet and outlet ports of each incubation vessel were closed, and the NaOH traps were removed and titrated as described above. Headspace gas in each vessel was sampled immediately following cessation of gas flow through a rubber septum in the lid, using 1 mL insulin syringes. Syringe needles were inserted into a rubber stopper for short-term storage of samples prior to analysis on a GC for methane (CH 4 ). Sampling was repeated after 6 hours, and gas samples were analyzed for CH 4 Preliminary studies showed that accumulation of methane in the incubation vessels was linear over time. Sample gases were analyzed on a Hewlett-Packard 5840A GC (Hewlett Packard, Avondale, PA), using thermal conductivity (TCD) and flame ionization detectors (FTD) for

PAGE 80

74 C0 2 and CH 4 analysis, respectively. For C0 2 analysis, a Poropak N (Supelco, Bellefonte, PA) column was used, with He as a carrier gas. Oven, injector and detector temperatures were set to 60, 140 and 200 C. For CH 4 analysis, a Carboxen 1000 (Supelco, Bellefonte, PA) column was used, with a N, carrier gas. Oven, injector and detector temperatures were 120, 120 and 200 C. Statistical Analyses Simple and multiple linear regression procedures were used for evaluating continuous variables (continuous Y vs. continuous X), and analysis of variance (ANOVA) procedures were used for categorical data (continuous Y vs. nominal or ordinal X). All statistical procedures were performed using JMP software (SAS Institute, Cary, NC). Results Chemical Analysis of Substrate Total C content was similar in plant tissue, standing dead, litter, surface peat and buried peat (Tables 3-2 and 3-3). Average values for the ten sampling sites were slighdy but significantly higher (a = 0.05) in standing dead than in peat. Total N content increased significantly from standing dead to litter to peat. Total P content was significantly lower in standing dead than live plant tissue, litter and surface (0-10 cm) peat (Tables 3-2 and 3-3; Figure 3-2). Lignin content increased significandy (a = 0.05) from the plant tissue to peat along the decay sequence, while cellulose content decreased. Average total C content of live sawgrass leaves was significantly higher (a = 0.05) than cattail. Conversely, total N and P concentrations were significantly higher in live tissue of cattail than in sawgrass. Differences between means of C, N and P content of standing dead material in cattail and sawgrass were not significant. Mean total N content of

PAGE 81

75 Table 3-2. Chemical analysis of live and standing dead plant tissue collected from 10 sites along the WCA-2A nutrient gradient. Each sample was a composite of leaf samples from five plants. Plant Sample Sample Cellulose Site type type date Total C Total N Total P Lignin --•gkg mg kg' % 1 Cattail Live Mar-95 402 11.4 918 5.0 36. D Jun-95 409 15.7 1491 Dead Mar-95 466 8.0 339 15.3 3a. 0 Jun-95 456 4.1 266 2 Cattail Live Jun-95 438 10.0 1067 Dead Mar-95 460 7.3 343 Jun-95 452 3.6 262 3 Cattail Live Jun-95 451 8.8 982 Dead Mar-95 463 5.7 375 Jun-95 465 4.7 275 4 Cattail Live Jun-95 436 10.3 1166 Dead Mar-95 457 5.3 277 10.0 42.0 Jun-95 454 3.0 258 5 Cattail Live Jun-95 422 9.5 1202 Dead Mar-95 457 6.4 373 Jun-95 444 4.1 305 Sawgrass Live Jun-95 475 5.9 527 Dead Jun-95 457 3.9 225 6 Cattail Live Mar-95 432 8.0 844 6.2 34.7 Jun-95 460 8.2 662 Dead Mar-95 463 6.0 310 11.4 42.9 Jun-95 451 3.4 249 Sawgrass Live Mar-95 458 7.8 487 7.4 33.4 Jun-95 464 7.3 673 Dead Mar-95 466 4.3 487 12.9 34.3 Jun-95 470 5.4 370 7 Cattail Dead Mar-95 468 4.7 315 Sawgrass Live Jun-95 459 7.8 844 Dead Mar-95 457 5.3 478 Jun-95 463 3.4 227 8 Cattail Live Mar-95 422 8.1 684 Dead Mar-95 468 4.0 205 Sawgrass Live Mar-95 463 6.6 478 Jun-95 461 8.2 958 Dead Mar-95 456 5.7 241 Jun-95 462 3.2 159 9 Sawgrass Live Jun-95 475 5.2 322 Dead Mar-95 454 4.0 140 16.1 35.5 Jun-95 467 3.2 106 10 Sawgrass Live Mar-95 456 6.0 242 7.0 33.3 Jun-95 479 4.4 157 Dead Mar-95 447 4.6 61 14.2 38.1 Jun-95 459 3.6 66

PAGE 82

76 Table 3-3. Physical and chemical analysis of soil cores collected June 6, 1995 from 10 sites along the WCA-2A nutrient gradient. Each value represents a single composite sample of three cores from each site. The soil profile was sampled in three depth increments: litter layer, surface peat (0-10 cm) and buried peat (10-30 cm). Depth Bulk Total Total Total Site increment density N C P Lignin Cellulose g c™ 3 g kg mg kg 1 % 1 Litter 26.6 433 1582 33.7 17.2 0-10 cm 0.053 28.2 376 1497 39.0 12.7 10-30 cm 0.117 25.4 358 1195 47.7 7.9 2 Litter 26.7 463 1291 0-10 cm 0.043 28.6 440 1337 10-30 cm 0.063 30.3 459 1172 3 Litter 25.7 454 1461 0-10 cm 0.044 30.7 433 1622 10-30 cm 0.065 24.4 388 880 4 Litter 25.0 454 1833 30.2 22.9 0-10 cm 0.034 24.8 439 1416 37.1 18.5 10-30 cm 0.092 31.5 451 982 51.7 13.9 5 Litter 27.8 453 1743 0-10 cm 0.041 28.3 443 1353 10-30 cm 0.094 32.2 453 611 6 Litter 20.1 467 1084 37.8 22.4 0-10 cm 0.057 29.1 451 1060 52.2 14.9 10-30 cm 0.086 35.8 457 284 57.0 12.2 7 Litter 17.5 452 928 0-10 cm 0.056 27.0 439 1146 10-30 cm 0.067 31.3 453 466 8 Litter 19.6 443 1038 0-10 cm 0.056 26.1 440 905 10-30 cm 0.106 26.3 422 279 9 Litter 12.2 452 345 31.1 27.3 0-10 cm 0.064 28.8 448 696 44.9 18.8 10-30 cm 0.088 29.0 453 269 56.9 14.2 10 Litter 16.4 410 275 32.8 21.5 0-10 cm 0.057 28.9 431 475 43.9 16.1 10-30 cm 0.077 24.7 471 236 58.8 14.5

PAGE 83

77 standing dead was significantly higher in samples collected in March than in June (a = 0.05). Similarly, total P content of standing dead was higher in March, but the difference was not significant at a = 0.05. Total N content of live plant tissue decreased significantly with increasing distance from the S-10C inflow, according to linear regression analyisis (a = 0.05). When cattail and sawgrass were considered separately, a similar decreasing trend was observed, but was not significant. Thus, plant type was probably the main factor affecting plant tissue N along the transect, since the N content of live cattails was found to be higher than sawgrass, even at the same distance from the inflow (Table 3-2). Total N content of the dead organic matter compartments did not vary significantly along the sampling transect. Total P content of live and dead plant material, plant litter and both peat layers decreased significantly with increasing distance from the inflow (Figure 3-2). As with total N content, the concentration of total P in live cattail leaves was higher than in sawgrass at comparable distances from the inflow (Table 3-2). For cattail and sawgrass plants considered separately, a decreasing trend in total P content of live plants along the transect (increasing distance from the inflow) was apparent, although not significant at a = 0.05. In addition, standing dead total P content of sawgrass decreased significantly along the transect. However, total P concentration in standing dead plant tissue did not differ significantly between cattail and sawgrass plants. Mean concentrations of total N and total P in standing dead (sampling sites and plant types combined) were higher for the March sampling event than in June (Table 3-2), although only the difference for total N was significant. Since the spring and early summer period is characterized by new growth following winter senescense, standing dead material sampled in June was presumed repesent the same stock that was sampled in March. Thus, the material sampled in June would be expected to be older and therefore more decomposed.

PAGE 84

78 20001500 1000m Live plant o I a. c/> O x 0. -I < Io I5002000/ q 3 */ \ 1500 f^--\ — Litter — 0-10 cm — A' — 10-30 cm 1000500\ 9 0 -r 2 T 3 6 4 5 DISTANCE (km) 7 i 8 10 Figure 3-2. Total P concentration in organic C pools, representing the decay continuum from plants to peat, as a function of distance from the S10C inflow. Values for live plant and standing dead plant material represent composite (averaged) data for cattails and sawgrass.

PAGE 85

79 Lignin content was inversely proportional to cellulose content among all samples analyzed (Tables 3-2 and 3-3). Preferential utilization of cellulose over lignin as a C source for microbial decomposers resulted in depletion of cellulose relative to lignin. The shift from cellulose-dominated to lignin-dominated substrate along the decay continuum has been previously described using the ligno-cellulose index (LCI) (Melillo et al., 1989). The LCI is the proportion of lignin in the ligno-cellulose complex, or La = ^ [3-1] lignin + cellulose so that the LCI increases with decomposition, with a maximum possible value of one. The LCI did not vary significantly (a = 0.05) with distance from the inflow for any live or dead organic matter component (Figure 3-3). There was also no significant difference (a = 0.05) in LCI between cattail and sawgrass plants (Table 3-2). However, LCI increased significantly (a = 0.05) along the decay sequence, including live plant tissue (Figure 3-3). Furthermore, the LCI for each component along the sequence was significantly (a = 0.05) higher than the previous component. Microbial Respiration Mean C0 2 production rate during aerobic incubation did not differ significantly (a = 0.05) between cattail and sawgrass standing dead (Table 3-4). However, C0 2 production was significantly higher (a = 0.05) for standing dead (plant type and sample site combined) collected in March than in June. There was also a significant trend (a = 0.05) of decreasing C0 2 production in standing dead with increasing distance from the inflow. A similar trend was observed for the litter and peat components (Table 3-5); the trends were significant (a = 0.05) for litter and surface (0-10 cm depth) peat. Specific C loss was calculated from gaseous C loss rate and substrate total C content as:

PAGE 86

80 DISTANCE (km) — • — Live plant — 0-10 cm o Standing dead — &• — 10-30 cm — -— Litter Figure 3-3. Ligno-cellulose index (LCI) of organic C pools, representing the decay continuum from plants to peat, as a function of distance from the S10C inflow. Values for live plant and standing dead plant material represent composite (averaged) data for cattails and sawgrass.

PAGE 87

81 Table 3-4. Potential C mineralization of dead cattail and sawgrass leaves (standing dead), measured as C0 2 production rate during aerobic incubation at 25C. Values represent means with standard error in parentheses (n=3). Plant C0 2 production rate Site type March 1995 June 1995 mgCg'd' 1 Cattail 1.36 (0.09) 0.85 (0.04) 2 Cattail 1.34 (0.07) 0.87 (0.03) 3 Cattail 1.62 (0.03) 1.04 (0.02) 4 Cattail 1.21 (0.01) 0.73 (0.01) 5 Cattail 1.50 (0.09) 1.23 (0.09) Sawgrass 1.16 (0.15) 6 Cattail 1.24 (0.03) 0.82 (0.02) Sawgrass 1.26 (0.08) 1.17 (0.11) 7 Cattail 1.08 (0.08) Sawgrass 1.31 (0.03) 0.84 (0.05) 8 Cattail 0.78 (0.02) Sawgrass 1.12 (0.05) 0.85 (0.07) 9 Sawgrass 1.01 (0.04) 0.85 (0.01) 10 Sawgrass 0.74 (0.02) 0.48 (0.03)

PAGE 88

82 Table 3-5. Potential C mineralization rate and microbial biomass C content of soil litter and peat (0-10 and 10-30 cm) layers. Potential mineralization was measured as CO, production rate during aerobic incubation at 25C. Values represent means with Site Depth increment CO, production Microbial biomass mgCg' d' mo C 8 1 Litter 1.763 (0.224) (Z.48) 0-10 cm 0.575 (0.003) Q 1 O (A fY)\ (4.Ui) 10-30 cm 0.195 (0.023) Z.oi (U.8J) 2 Litter 1.756 (0.206) O A 20.84 (2.61) 0-10 cm 0.644 (0.057) 5. /6 /"> QA \ (2.94) 10-30 cm 0.241 (0.016) 3.8 1 (U.DV) 3 Litter 1.659 (0.105) 35.13 / 1 C A \ (1.54) 0-10 cm 0.934 (0.071) 14.60 (2.20) 10-30 cm 0.210 (0.020) ( A tC 4.46 (2.68) 4 Litter 1.934 (0.144) 3 1 .86 (128) 0-10 cm 1.390 (0.105) 1 O 1/1 18.14 i i i ~t\ (1.17) 10-30 cm 0.193 (0.019) 6.65 / 1 A A \ (2.44) 5 Litter 1.652 (0.236) n a a 37.44 (5.63) 0-10 cm 0.649 (0.073) C\ C A 9.64 (0.89) 10-30 cm 0.130 (0.010) 4.61 (2.40) 6 Litter 0.964 (0.052) 15.77 (1.71) 0-10 cm 0.261 (0.034) 3.80 /A O0\ (0.29) 10-30 cm 0.100 (0.009) l \ A^\ (1.45) 7 Litter 0.682 (0.035) 13.84 (2.11) 0-10 cm 0.345 (0.018) 6.46 (0.45) 10-30 cm 0.173 (0.012) 4.17 (2.28) S Litter 1.127 (0.066) 20.49 (2.46) 0-10 cm 0.356 (0.017) 6.21 (1.58) 10-30 cm 0.077 (0.008) 3.00 (1.68) 9 Litter 0.695 (0.044) 7.95 (1.50) 0-10 cm 0.256 (0.014) 5.48 (1.45) 10-30 cm 0.090 (0.010) 2.24 (2.48) 10 Litter 0.410 (0.025) 9.27 (0.94) 0-10 cm 0.261 (0.019) 3.42 (0.86) 10-30 cm 0.104 (0.013) 2.94 (1.61)

PAGE 89

83 mg substrate C lost g substrate _i r — x — = day [3-2] g substrate • day mg substrate C This quantity is equivalent to the first-order rate constant k, used in a simple exponential decay model. Thus it carries implications for biodegradability of the substrate. Specific C loss under aerobic conditions decreased with increasing distance from the inflow (significant (a = 0.05) for litter and buried peat) and along the decay sequence from standing dead to peat (Figure 3-4). Sharp increases in specific C loss occurred in samples from site 4 for both litter and surface peat. Aerobic specific C loss decreased sequentially from litter to surface peat to buried peat; however, values for standing dead were not significantly different (a = 0.05) from the litter layer. Trends in C0 2 and CH 4 production during anaerobic incubation of litter and peat were similar to those for aerobic CO, production (Table 3-6). Specific anaerobic C loss in litter samples decreased significantly (a = 0.05) with increasing distance from the inflow (Figure 3-5). Specific loss was somewhat higher for peat samples near the inflow, but there was no significant trend (a = 0.05) with distance. A sharp rise in specific anaerobic C loss occurred at sites 3 and 4 in the surface peat, similar to the peak associated with specific aerobic C loss. For all sites combined, specific anaerobic C loss decreased significantly (a = 0.05) between the litter, surface peat and buried peat layers. Production of CH 4 accounted for approximately 1 to 25% of the total C evolved during anaerobic respiration (Figure 3-6). The relative proportion of CH 4 production was highest in the litter layer at sites 1-9, and in the surface peat at sites 2-5. For all other samples, CH 4 production accounted for less than 2% of total gaseous C loss under anaerobic conditions. Microbial biomass C was significantly higher (a = 0.05) in the litter layer than in surface and buried peat layers (Table 3-5). Biomass C decreased significantly (a = 0.05) with increasing distance from the inflow, although for individual components the decrease was significant only in the litter layer. The same trends were observed for microbial

PAGE 90

84 Figure 3-4. Specific C loss during aerobic incubation of detrital organic C pools as a function of distance from the S-10C inflow. Values for standing dead plant material represent composite (averaged) data for cattails and sawgrass. Data are mean values from triplicate incubations.

PAGE 91

85 0.0015 DISTANCE (km) Figure 3-5. Specific C loss during anaerobic incubation of detrital organic C pools as a function of distance from the S-10C inflow. Values for standing dead plant material represent composite (averaged) data for cattails and sawgrass. Data mean values from triplicate incubations.

PAGE 92

86 30 DISTANCE (km) Figure 3-6. Methane production in litter, surface peat (0-10 cm) and buried peat (10-30 cm), expressed as percent of total C (C0 2 + CH 4 ) loss during anaerobic incubations.

PAGE 93

87 Table 3-6. Potential anaerobic decomposition rate for soil litter and peat (0-10 and 10-30 cm) layers. Potential decomposition was measured as C0 2 and CH 4 production rate during anaerobic incubation at 25C. Values represent means, with standard error in parentheses (n=3). Depth Site increment CO ; production CH 4 production --HgCg-' d< 1 l Litter AQ1 8 /a a\ (o.o) AA AO (U.JO ) (J1U cm 1 ni ^ (ly.o) en it\ IU-30 cm ni 0 y 1 .2 (D.8) n 41 U.4 1 l(\ rn\ \\j.\J 1 ) I Litter acw n (A 7\ (4.-5) 7"? 7A f\ in u1 u cm 0/1 1 A 24 1 .0 7A 77 zo. / / (ft SOI in on m 1U-3U cm 1 f\1 A 1UJ.4 fA ^\ (4.3) n 77 U. /Z (f\ 1 1 ~\ (U. 11) 5 Litter JUU.O /77 ^\ (22.3) 1 m 1 A 1U / 10 VH.HU i u-iu cm Jo /.2 (./•->) an 7a oU. / 8 (8. 1 J) 1U-3U cm AO A oy.o /"I 7\ (3.2) n 77 u. / / 1 H7 77 1U /. / / ( J. 8 1 ) n i n rtw. U1 U cm L5 1 .J 1 1 A 11 (lo.l) 11 17 J 1 1 2 rfi 771 (8.Z / ) 1 n in 1U-3U cm <"7 n n\ (2.0) M ^7 U. 32 tc\ in\ (U. 1U) 0 Litter 771 n fA fil (0.8) 7A 4fi /0.4U AO"! n i n u iu Lin 1 no q 1 UVJ. 7 U.7J i ) 10-30 cm 25.4 (2.4) 0.60 (0.14) 7 Litter 196.1 (3.6) 19.56 (2.23) 0-10 cm 107.4 (12.5) 0.94 (0.19) 10-30 cm 46.4 (2.5) 0.69 (0.14) 8 Litter 350.3 (16.9) 78.40 (1.07) 0-10 cm 132.8 (15.1) 0.82 (0.22) 10-30 cm 30.5 (2.1) 0.47 (0.08) 9 Litter 219.2 (13.5) 27.35 (8.42) 0-10 cm 93.0 (12.8) 0.90 (0.15) 10-30 cm 61.6 (10.0) 0.54 (0.10) 10 Litter 163.0 (14.2) 2.48 (0.13) 0-10 cm 122.9 (3.8) 0.82 (0.23) 10-30 cm 39.5 (4.9) 0.59 (0.14)

PAGE 94

88 biomass expressed as a percentage of total substrate C (Figure 3-7). Sharp peaks in biomass C were observed for litter at sites 3-5 and for surface peat at sites 3 and 4. Discussion The continuum of organic C transformations and flows which constitute the wetland C cycle may be represented conceptually as a collection of discrete storage units, or compartments, with simultaneous transfer of mass among the compartments (Figure 3-8). This type of representation is the basis for numerous conceptual models of C cycling in terrestrial soils and ecosystems, in which soil organic C is classified according to turnover time (Jenkinson, 1990). Peat is represented by an active surface component (0-10 cm depth) and a more stable buried layer (10-30 cm), based on previously determined chemical and physical characteristics (Koch and Reddy, 1992; Reddy et al., 1993; DeBusk et al, 1994). The top 30 cm of peat roughly incorporates the zone of plant root-soil interaction in the Everglades marsh. The vegetation component represents transformers of inorganic C (C0 2 ) to organic C (primary production) through photosynthesis. The heterotrophic microflora represent transformers of organic C back to inorganic C through cellular respiration. The decay continuum from plant standing dead to peat is represented as a sequence of four compartments. Organic C is stored in the system in living (vegetation and microbial biomass) and non-living (standing dead plant tissue, plant litter and peat) components. Below-ground C pools (roots and rhizomes) were not sampled during this study, thus only living and dead plant leaves were considered. Previous research in the Everglades has shown that belowground biomass accounted for approximately 12% of total sawgrass production (Toth, 1987). The concept of a decay continuum to describe the conversion of plant material to soil organic matter is based on selective loss of relatively labile constituents and resulting changes in the chemical and biochemical characteristics of the substrate (Melillo et al.,

PAGE 95

89 Figure 3-7. Microbial biomass C pools in litter, surface peat (0-10 cm) and buried peat (10-30 cm) in WCA-2A, expressed as percent of total organic C. Data are mean values from triplicate analyses.

PAGE 96

90 Live plant I Plant standing dead E 3 3 C c o O > (0 o V Q I Litter layer I Surface peat (0-10 cm) I T Buried peat (10-30 cm) Microbial decomposers Figure 3-8. Conceptual diagram of the organic C cycle and decay continuum in Everglades WCA-2A.

PAGE 97

91 1989). The broadly defined lignocellulose component of the substrate has been considered by many researchers to be the primary indicator of "substrate quality" (Colberg, 1988; Moran et al., 1989 ). Loss of non-lignocellulosic (labile) components of plant detritus occurs rapidly during the intial stages of decomposition, thus lignin and cellulose become the primary C components of the substrate (Moran et al., 1989). The LCI can be correlated with age of decomposing plant material and has been proposed as an indicator of the state of decomposition (Melillo et al., 1989). The continuous increase in LCI during the decomposition process directly relates to the quality, or availability, of the C substrate. This is a result of "enrichment" of the substrate with the more recalcitrant lignin compounds as well as lignin derivatives occurring as microbial byproducts (Swift, 1982; Heal and Ineson, 1984). In the present study, LCI decreased significantly (a = 0.05) along the decay continuum, or more appropriately, the "decay sequence". This indicated that substrate C quality, or availability, decreased significantly (a = 0.05) from standing dead plant tissue to soil lifter layer, surface peat (0-10 cm depth) and buried peat ( 10-30 cm). Among other potential factors regulating decomposition rate (Heal et al., 1981) nutrient and 0 2 availability were considered in the present study, as functions of hydroperiod and nutrient loading in WCA-2A. Total N and P content of live plant tissue (above-ground) reflected the historically-monitored gradient of nutrient-enrichment in the surface water (SFWMD, 1992). However, the P enrichment gradient in living and dead plant tissue, including peat, was the most pronounced, in accordance with results of previous comprehensive monitoring of soil chemistry in WCA-2A (DeBusk et al, 1994). Total P analysis of soil layers (litter, surface peat and buried peat) included both inorganic and organic forms. Previous studies have shown that organic P accounted for approximately 70-80% of total soil P in WCA-2A (DeBusk et al, 1994; Quails and Richardson, 1995). Most of the organic P in the litter and peat is contained in the particulate organic matter, and can be considered as an integral part of the substrate. Substrate-bound

PAGE 98

92 organic P was assumed to account for nearly all of the total P analyzed for standing dead material. This organically-bound P may be considered as a component of overall substrate quality, along with substrate C quality (as indicated by LCI values). Dissolved P in the water and soil porewater, on the other hand, could be considered externally available P. The latter pool of P has been measured on several occasions in WCA-2A, and has been shown to be highly variable, both spatially and temporally (SFWMD, 1992). Both LCI and total P content were significandy related to decomposition rate in the four dead organic matter pools. A multiple linear regression model of aerobic C0 2 production rate as a function of total P and LCI explained 91% of the variability in the response variable. Total P and LCI were both highly significant (a = 0.05) effects on CO, production. In addition, there was significant (a = 0.05) interaction between total P and LCI. This is an indication that P content and C availability may be co-limiting factors in decomposition of plant litter and peat in WCA-2A. Previous study of soils in Everglades National Park showed a significant effect of P on mineralization of added C (Amador and Jones, 1993, 1995). A relatively high degree of variability surrounded the general trend of decreasing substrate C0 2 production rate (substrate biodegradability) along the nutrient gradient, especially near the inflow (Table 3-5, Figure 3-4). This local variability was explained by substrate total P content and LCI, which covaried with the mineralization rate (Figures 3-2 and 3-3). The apparent underlying cause of this phenomenon is differential P loading among neighboring sites. There is evidence of hydraulic short-circuiting or channelized flow in WCA-2A, especially during periods of low water, caused by clumping of dense vegetation in the highly nutrient-enriched area and by airboat trails throughout WCA-2A. Measurement of C0 2 production during aerobic incubation of organic matter provided a direct measurement of organic C mineralization, which represents completion of the decomposition process. For short-term studies and recalcitrant substrates, measurement of C0 2 is an sensitive method for estimating decomposition rate. However, calculation of

PAGE 99

93 the first-order decay constant using C0 2 production techniques over a short time period, e.g. one or two weeks, requires certain assumptions regarding the type of model to be used. These assumptions must be based on prior knowledge of the decay characteristics of the substrate, since curve fitting is best suited for several widely separated points in time. Controlled studies of mass loss in leaves from various plant types indicate that simple carbohydrates and other labile compounds are substantially lost during the first few weeks of decomposition (Moran et al., 1989). Standing dead portions of cattail and sawgrass may become highly leached and substantially decomposed while attached to the plant, due to the residence time on the plant of up to several months. Thus, a multi-phase kinetic model of long-term decomposition of standing dead material, litter and peat was judged to be inappropriate. The use of a simple (single compartment) exponential decay model was supported by data from a previous study of in situ decomposition of cattail and sawgrass standing dead in WCA-2A (Davis, 1991). Carbon mineralization during aerobic incubation of litter and peat was considered a measure of potential decomposition rate, since 0 2 is limited or absent under flooded conditions. Availability of 0 2 in the water column and litter layer is dependent on rates of 0 2 diffusion across the water-air interface, 0 2 production in the water column and litter by algal and macrophytic photosynthesis and 0 2 demand exerted by the substrate. Even under drained conditions, saturation of microsites can occur due to the high capillarity of peat, therefore a significant portion of the soil may remain anaerobic. Potential respiration in peat from three wetland sites in North Carolina increased with overall nutrient availability (Bridgham and Richardson, 1992). Measurement of aerobic, or potential, decomposition is an indicator of substrate quality, which encompasses availability of C and growth-limiting nutrients. In the context of the present study, potential decomposition rate incorporates the combined effects of LCI and total P. First-order decay constants calculated in this study, from "instantaneous" specific loss rate, varied over an order of magnitude (ca. 10" 4 to 10" 3 d' 1 ) among the four

PAGE 100

94 organic C compartments (plant standing dead, litter layer, surface peat and buried peat). Corresponding turnover times ( 1/k) ranged from about 0.7 to 2.7 years for standing dead and litter, and 5 to 10 years for buried peat. As previously stated, these were considered potential values which may be observed in the field under optimal 0 2 availability. Decomposition of above-ground standing dead material under field conditions is predominantly aerobic. The exception would apply to those portions of the plant situated below the water surface. Water depth in WCA-2A typically remains below about 50 cm, and frequently is much lower (SFWMD, 1992). However, standing dead decomposition rate may be limited by nutrient availability and, periodically, moisture availability. The latter is an important factor governing decomposition in terrestrial ecosystems (Heal et al., 1981). Decomposition of above-ground dead material may be severely restricted during extended dry periods. Standing dead material was wetted before aerobic incubation, therefore moisture content was assumed to be non-limiting for this study. Nutrient availability was most likely the limiting factor in decomposition of standing dead. Concentrations of N and P in dead plant tissue were significantly lower (a = 0.05) than in the litter layer and peat (Tables 3-2 and 3-3). Standing dead plant material (above-water) is isolated from plant and surface water nutrient sources. As soluble constituents, including newly-mineralized N and P, are leached out the substrate may become excessively nutrientdepleted. Based on analysis of N and P, decomposition activity in standing dead plant tissue was limited by both N and P availability. Average molar N:P ratio was approximately 40 in both standing dead and litter, yet total N and P content of litter was, on the average, about 5 times greater than for standing dead material. This was viewed as evidence of substantial microbial immobilization of both N and P following deposition of dead plant material into the Utter layer. Immobilization potential has been linked to initially low N and lignin content of the organic substrate (Melillo et al., 1984). Studies in a cypress swamp in north Florida receiving municipal wastewater showed that, three weeks after

PAGE 101

95 addition of l5 N-labelled wastewater, nearly all of the N remaining in the peat-floodwater profile was immobilized in the litter layer (DeBusk and Reddy, 1987). Decomposition rates derived experimentally during anaerobic laboratory incubations were analogous to those occurring in the field in 0 : -depleted regions of the peat and litter. Total O, depletion generally occurs only under flooded conditions. Typically, however, only the subsurface (buried) peat is constantly anaerobic. Litter on the peat surface and suspended in the water column may remain aerobic or undergo diurnal aerobic-anaerobic cycles driven by photosynthetic activity in the water (Chapter 2). Under conditions of low light availability and high 0 2 demand, the entire water and Utter layer may remain anaerobic for the duration of flooding. Under anaerobic conditions, alternate electron acceptors must be utilized by microbial decomposers. These include, in order of preference, NO,", Mn 4 *, Fe 3+ S0 4 = and CO, (Westermann, 1993; Reddy and D'Angelo, 1994). Relatively high concentrations of S0 4 = have been measured in surface water and soil porewater in WCA-2A (Chapter 4; Schipper and Reddy, 1994). This may be a result of discharge of brackish groundwater (from saltwater intrusion) into the canals which supply water to WCA-2A. In addition, production of H ; S(g) was detected (by odor) during anaerobic incubations. Thus, circumstantial evidence strongly suggests that anaerobic respiration in WCA-2A soil is dominated by sulfate reduction, especially near the S-10 inflow. Although it has been the general belief that sulfate reduction is the dominant pathway of anaerobic respiration in the presence of high levels of S0 4 = methanogenesis has been shown to occur to a significant extent in high-sulfate soils where labile C was abundant (Howarth, 1993). The proportion of CH 4 produced (relative to total gaseous C production) during anaerobic incubations varied considerably, ranging from less than 1% to about 25% (Figure 3-6). The highest proportion of CH 4 production occurred in the litter layer, followed by the surface peat layer, with extremely low production in the buried peat layer. This trend corresponded to the relative availabilty of C, as determined by specific aerobic C

PAGE 102

96 and LCI. The distribution of relative CH 4 production along the sampling transect roughly corresponded to substrate total P content. Therefore, it appeared that the overall substrate quality (hence, availability) was the primary factor controlling the proportion of CH 4 produced under anaerobic conditions. Mineralization of substrate C under anaerobic conditions represents the "lower potential" for decomposition in field. Specific anaerobic C loss rate (combined loss of CO, and CH 4 ) was, on average, 32% of the aerobic rate. The anaerobic: aerobic respiration ratio was not significantly different (a = 0.05) among litter, surface peat and buried peat layers. In comparison, anaerobic respiration rate of peat from three North Carolina peatlands was 34 to 63% of the aerobic rate (Bridgham and Richardson, 1992). Lignocellulose extracted from Carex walteriana and incubated in anaerobic peat in the Okeefenokee Swamp was decomposed at 37% of the aerobic decomposition rate (Benner et al., 1984). Anaerobic and aerobic mineralization rates were highly correlated (r = 0.97). The primary factors governing aerobic decomposition rate, total P content and LCI, were also significant factors influencing the rate of anaerobic decomposition. The constant difference between aerobic and anaerobic rates suggested that anaerobic metabolic processes were similar throughout the study area. Based on short-term anaerobic incubations, turnover times for litter, surface peat and buried peat under anaerobic conditions were approximately 2-6, 2-14 and 1 1-60 years, respectively. Actual turnover of the standing dead, litter and peat compartments cannot be predicted by potential aerobic and anaerobic decomposition rates alone. There are also factors which govern the transfer of material from one compartment to the next. Litterfall, the physical deposition of dead plant material into the water column and litter layer, is a function of external physical forces as well as the stage of decomposition of the material. The relative movement of organic matter through the vertical profile (burial) is a result of mass loss, compaction and accumulation of newer organic matter. Position in the vertical profile may have a significant effect on decomposition rate, affecting O, and nutrient

PAGE 103

97 availability. The rate-controlling factors discussed earlier interact to control the movement of organic matter through time and vertical space in the wetland system. It was assumed that microbial biomass calculations were representative of the biomass present in the field at the time of collection. Most likely, a diverse assemblage of active and inactive aerobic, facultative anaerobic and obligate anaerobic microorganisms were present in all litter and peat samples. The activity of the different groups would depend on the presence of 0 2 and alternate electron acceptors. This premise was supported by the observed rapid stabilization of microbial activity under both aerobic and anaerobic conditions imposed by controlled incubations. Microbial biomass C was highly correlated with aerobic (r = 0.98) and anaerobic (r = 0.94) C mineralization rates in the soil. This suggests that the efficiency of C utilization was relatively constant across sites and depths within aerobic and anaerobic environments. Decreased substrate availability, whether due to C limitation with depth or P limitation with distance from the inflow, resulted in a smaller population of microbial decomposers operating at a comparable level of efficiency. The ratio microbial C to total organic C (C mc :C loai ) has been used as a qualitative indicator of soil organic C quality or availability (Anderson and Domsch, 1989; Sparling, 1992). The C mc :C Kal reported for WCA-2A litter and peat spanned the range of values reported for agricultural, pasture and forest soils (Anderson and Domsch, 1989). Values for surface and buried peat along the nutrient gradient corresponded to the range of "typical" values for forest and pasture soils (Sparling, 1992). In comparison, C^C,^ values for litter in the high-nutrient cattail marsh were comparable to or higher than previously reported values. The metabolic quotient (^C0 2 ), or ratio of microbial respiration (as C0 2 loss) to biomass C, has been widely used in studies of ecosystem response to environmental stress or change in terrestrial ecosystems (Anderson and Domsch, 1986, 1990, 1993; Insam and Haselwandter, 1989). Metabolic quotient is a reflection of microbial efficiency, as determined by the relative magnitude of maintenance respiration in the population. The
PAGE 104

98 ecosystem succession and heavy metal stress, and is apparently a significantly more robust parameter than C m JC loal (Anderson and Domsch, 1993). Increased qC0 2 (temporally or spatially) has been associated with high resource availability and simple substratedecomposer relationships in early successional ecosystems, while decreasing qC0 2 values may be indicative of maturing systems, especially low-nutrient systems with closed cycling of resources (Insam and Haselwandter, 1989; Wardle, 1993). In addition, increased qC0 2 levels are apparently related to ecosystem disturbance, such as industrial pollution (Ohtonen, 1994). However, qC0 2 values along the WCA-2A transect, calculated with either aerobic or anaerobic respiration rates, did not vary significantly (a = 0.05) with depth or distance from the inflow. This may suggest that the nutrient gradient in the northern half of WCA-2A has reached a quasisteady state, so that high and low nutrient ecosystems are relatively mature, or stable. It is also a possibility that the physical and chemical stratification of the soil environment in the marsh precludes the use of this physiological parameter, which was developed in a terrestrial setting. Summary and Conclusions Total P content and relative lignin enrichment (expressed as LCI) accounted for 91% of the variability associated with substrate biodegradability (potential respiration rate) across the decay continuum. Anaerobic conditions caused a proportional decrease in decomposition rate (increased turnover time) in the litter, surface peat (0-10 cm) and buried peat (10-30 cm) layers of the soil. Anaerobic decomposition rate was approximately onethird the rate of aerobic decomposition. The plant standing dead component and soil litter layer are potentially the most active (rapid turnover) organic C pools, as reflected by their low to moderate values of LCI. High cellulose content and low nutrient (N and P) concentations indicate that C availability does not limit decomposition of standing dead. As the dead leaf tissue ages on the plant, nutrient availability becomes a major limiting factor for microbial activity.

PAGE 105

99 Another probable limitation may be moisture availability, during drought conditions. Carbon availability in the soil litter layer appears to be significantly greater than in peat, probably due to the presence of available cellulose. Within this layer, the major determinants of decomposition rate are nutrient availability (P in this case) and 0 2 availability. Total P concentration exerted a strong influence on decomposition of litter. It is likely that C availability only becomes a limiting factor under conditions of high nutrient availability. Total P content exerted less influence on decomposition rate as the depth and age of the substrate increased. It is evident that C availability becomes increasingly limiting to microbial growth through the peat profile. High LCI values in the buried peat indicate that the more readily decomposable cellulose component is highly depleted, the remainder being physically protected from microbial activity by the lignified matrix of the substrate. The influence of nutrient availability on decomposition diminishes as the proportion of refractory compounds in the substrate increases (with age and depth). The standing dead pool has a high potential for rapid turnover of nutrients, due to its high decomposition rate and leaching of mineralized nutrients. The soil litter layer is potentially very important in short-term nutrient cycling. Due to the relatively high C availability and resulting high population of microbial decomposers, the litter layer may serve as either a source or sink for nutrients. Released nutrients may recycled through plant uptake, micobial re-immobilization or chemical adsorption or precipitation. Alternatively, they may be exported, along with dissolved organic C, to a downstream area. Turnover of peat, especially peat buried deep in the soil profile, is highly restricted under anaerobic conditions, therefore this pool of organic matter represents a potentially permanent sink for nutrients and contaminants. Microbial biomass C was highly correlated with both aerobic and anaerobic decomposition of organic C throughout the decay continuum. In particular, the ratio of microbial C to total soil C explained 88% of the variability in potential (aerobic)

PAGE 106

100 decomposition. Results of this study show microbial biomass to be a promising indicator of organic matter turnover in wetland soil. Substrate composition, related to nutrient and C availability, were also shown to be accurate predictors of organic C turnover. However, the interactions among these factors, along with environmental effects such as 0 2 availability, must be determined beforehand. Further study is needed to develop reliable biological and chemical indicators of organic C turnover in wetlands.

PAGE 107

CHAPTER 4 REGULATORS OF ORGANIC MATTER DECOMPOSITION ALONG THE WCA-2A NUTRIENT GRADIENT Introduction Decomposition of organic matter of plant origin has been widely studied, primarily in terrestrial ecosystems. Numerous factors associated with composition of the decomposing organic matter as well as external, or environmental, factors have been identified as important effects which control decomposition rate. Among the more important of the environmental factors are temperature, moisture, nutrients and electron acceptors (Swift et al., 1979; Heal et al., 1981; Webster and Benfield, 1986; Reddy and D'Angelo, 1994). Effect of flooding and water table on decomposition of litter and peat in wetlands has been widely documented (Lahde, 1966; Heal and French, 1974; Tate, 1979; Moore and Dalva, 1993). Depletion of 0 2 frequently occurs in wetlands, due to high 0 2 demand created by large stores of organic carbon (C) and greatly reduced 0 2 diffusion rate in water. In the absence of 0 2 oxidized compounds such as NO, Mn 4 *, Fe 3+ and S0 4 = are utilized by microbial decomposers as terminal electron acceptors for catabolic processes. However, these alternate electron acceptors frequently become depleted when demand exceeds supply. Because of the decreased energy yield associated with anaerobic catabolism, decomposition proceeds at a significantly lower rate under anaerobic conditions (Reddy and D'Angelo, 1994). An additional consequence of anaerobiosis is the significant reduction, and possibly inhibition, of lignin depolymerization in the absence of 0 2 (Benner et al., 1984; Colberg, 1988). 101

PAGE 108

102 Nutrient availability may also be a significant factor regulating decomposition rate in wetlands. The supply of microbial growth-limiting nutrients from the surrounding water and soil, often in inorganic form, may be used to supplement the nutrient pool contained within the organic substrate, generally in organic form (Heal et al., 1981; Melillo et al., 1984). Decomposition of organic matter is also regulated by chemical composition of the substrate itself, frequently referred to as "substrate quality" (Swift et al, 1979; Heal et al., 198 1). This refers to substrate nutrient content as well as the susceptability of organic C to microbial breakdown. Lignin and cellulose fractions are generally considered to be key components of the "C quality" of an organic substrate (Swift et al, 1979). Lignin is a more recalcitrant group of organic C compounds than is cellulose, therefore substrate cellulose content decreases more rapidly than does lignin content. Furthermore, lignin breakdown is either incomplete or nonexistant in the absence of 0 ; while cellulose is readily broken down, albeit at a significantly reduced rate (Colberg, 1988). Initial lignin and N content of the organic substrate (e.g. plant tissue) have been correlated with decomposition rate (Godshalk and Wetzel, 1978; Melillo et al., 1982; Aber et al., 1990). During the decomposition of plant material, the amount of lignin relative to cellulose increases, thus decreasing the C quality of the substrate. In addition, an accumulation of lignin derivatives generated as by-products of microbial metabolism (e.g. humic compounds) results in further decrease in substrate quality (Swift et al., 1979). Mineralization of organically-bound nutrients and subsequent loss of a portion of the released inorganic nutrients results in further decrease in overall substrate quality. Terrestrial decomposition studies have indicated that substrate quality and exogenous nutrient supply are both important regulators of decomposition during the early stages (Swift et al., 1979; Heal et al., 1981; Melillo et al., 1989). However, there is evidence that, during later stages of decomposition, the chemical quality of substrates of different initial compositions is reduced to a "least common denominator", so that variability in decay rate is a function of environmental factors alone (Melillo et al., 1989).

PAGE 109

103 The current study evaluates decomposition of plant litter in the water column and peat of a nutrient-impacted Everglades marsh. The Everglades encompasses a variety of wetland ecosystems which were historically adapted to low nutrient availability and periodic droughts. Nutrient and hydraulic loading to the Everglades occurred primarily through rainfall, with occasional pulses of water and nutrients from Lake Okeechobee overflow (Davis, 1943; Parker, 1974) (Figure 4-1). Major vegetational communities included sawgrass (Cladium jamaicense Crantz) marsh, wet prairies, sloughs and tree islands (also known as bayheads). The sawgrass marsh remains the dominant plant community in terms of total area, accounting for nearly two-thirds of the vegetative cover in the Everglades (Davis, 1943; Loveless, 1959). Loading of agricultural drainage water into the WCAs has resulted in nutrient enrichment of soil and vegetation in many areas (SFWMD, 1992; DeBusk et al., 1994). Phosphorus (P) enrichment has been a major concern in the Everglades, having been implicated, along with altered hydroperiod, in the encroachment of cattail (Typha domingensis Pers. ) and other rapid ly-growing vegetation into the native sawgrass marsh (Davis, 1943, 1991; Steward and Ornes, 1983; Toth, 1987, 1988). Accelerated nutrient loading in northern WCA-2A (Figure 4-1) during the past three decades has created a distinct nutrient (especially P) gradient in water, soils and plant tissue (Davis, 1991; Koch and Reddy, 1992; DeBusk et al, 1994). Changes in species composition of periphyton and macrophyte communities, along with an overall increase in net primary productivity have been documented along this gradient (Swift and Nicholas, 1987; Davis, 1991; SFWMD, 1992). Soil dating by analysis of 137 Cs peaks has indicated that peat accumulation rate has increased in nutrient-enriched areas of WCA-2A (Craft and Richardson, 1993; Reddy et al., 1993). The main objective of this research was to determine the effect of nutrient enrichment on organic matter decomposition rate along a vertical profile in the water column and peat. The scope of research included in situ determination of decomposition

PAGE 110

Figure 41 Site map for WCA-2A study area, showing locations of sampling sites. Coordinates for sampling sites are listed in Table 4-1.

PAGE 111

105 rates for native plant litter and for a uniform substrate, in the form of pure cotton strips. Supplemental water and soil chemistry data was evaluated with decomposition data to determine the significance of various environmental and substrate-related factors on decomposition rate. Materials and Methods Site Description Field study sites were located in Water Conservation Area 2 A (WCA-2A), a 447 km 2 region of the northern Everglades (Figure 4-1). Surface water flows into WCA-2A from the Hillsboro Canal through the four S-10 water control structures and from the North New River Canal through the S-7 pump station. Most of the hydraulic loading is through the S-10C and S-10D structures into the northern portion of WCA-2A. The general direction of flow is from north to south. Water depth is usually less than one meter, and varies considerably, both seasonally and year-to-year, with occasional dry periods (SFWMD, 1992; personal observations). The bulk of the surface outflow is through through three control structures at the south end of WCA-2A, into WCA-3 (Figure 4-1). Soil in WCA-2A consists of Everglades and Loxahatchee peats (Gleason et al., 1974). Everglades peat, the most common soil in the Everglades, is associated with the sawgrass marsh community. It is dark brown, finely fibrous to granular, with circumneutral pH, relatively high N content and low Si0 2 Fe and Al content. Peat depth in WCA-2A ranges from about 1 to 2 m, and age of basal peats is estimated to be 2000 to 4800 yr. Beneath the peat lies a bedrock of Pleistocene limestone, with intermediate layers of calcitic mud, sandy clay and sand in several areas (Gleason et al., 1974). The primary sources of nutrient loading to WCA-2A are the S-10 structures which convey water from the Hillsboro Canal and WCA-1 (Figure 4-1). A distinct gradient of N and, most significantly, P enrichment in water, plants and soil has formed between the high-nutrient region adjacent to the inflows and the low-nutrient interior marsh of WCA-2A

PAGE 112

106 (Koch and Reddy, 1992; SFWMD, 1992; DeBusk et al., 1994). A vegetation gradient coincides with the nutrient gradient; most notable is the gradient from sawgrass marsh with scattered aquatic slough in the interior to cattail and mixed emergents near the inflows. The vegetation gradient was divided into three discrete categories for the purposes of the current study: cattail-dominated, sawgrass-dominated and mixed cattail and sawgrass (Figure 4-1). Ten field sampling sites were established along the nutrient and vegetation gradient on a 10 km transect extending from the S-10C inflow into the interior marsh. The sites, numbered 1 through 10, were located at distances of 0.07, 0.3, 0.6, 1.2, 2.0, 2.9, 3.9, 4.8, 6.6 and 9.8 km downstream from the inflow (Table 4-1; Figure 4-1). Water-Porewater Chemistry Dissolved nutrients and pH in the floodwater-soil profile of each of the 10 sampling sites were determined during March and October, 1995 using porewater samplers (Hesslein, 1976). Porewater samplers were constructed of Plexiglas sheets (60 x 7 x 2 cm) with lateral (perpendicular to the long axis) cells milled into one side of the sheets. The cells were 1-cm on center, with volume of approximately 8 mL each. Prior to deployment in the field, cells were filled with deoxygenated deionized water, then covered with a sheet of 0.2-|im pore size membrane filter (Supor-200, Gelman Sciences, Ann Arbor, MI) and a protective fiberglass screen. A Plexiglas faceplate (0.3 cm thick) with slots matching the cell openings was used to hold the filter and screen firmly in place. Porewater samplers were installed vertically into the soil, so that approximately 30 cells (30 cm) were situated within the peat. One sampler was installed at each of the 10 sites on March 8 and October 11, 1995. Two additional samplers were installed at sites 4 and 10 in March, to provide triplicate analyses at those sites. Porewater samplers were retrieved after a 13-day equilibration period in the field. Water was removed from each cell, with filter and screening in place, using 10-mL disposable plastic syringes. The syringes were prepared for transport by embedding the needles into silicone rubber beads on Plexiglas trays, and

PAGE 113

107 Table 41 Locations of sampling sites in WC A-2A, and distance from S-10C inflow. Site Latitude N Longitude W Distance rrom inflow deg min deg min km 1 26 22.09 80 21.07 0.1 2 26 21.98 80 21.09 0.3 3 26 21.81 80 21.12 0.6 4 26 21.53 80 21.20 1.2 5 26 21.05 80 21.21 2.0 6 26 20.53 80 21.27 2.9 7 26 20.02 80 21.37 3.9 8 26 19.52 80 21.39 4.8 9 26 18.51 80 21.46 6.6 10 26 16.81 80 21.48 9.8

PAGE 114

108 placing the sample trays in an ice chest. This provided airtight containers devoid of headspace for the samples. Water-porewater samples were analyzed for pH, ammonium-N (NH 4 -N), dissolved reactive P (DRP) and sulfate (SO/). Sample pH was measured within 24 hours of sampling, using a combination pH electrode. Ammonium-N and DRP were determined by autoanalytical methods (EPA Methods 351.2 and 365.1. USEPA, 1983). Sulfate was measured on a Dionex Series 4500i Ion Chromatograph (Dionex, Sunnyvale, CA). Field Decomposition Study Decomposition of plant litter was measured in situ in a vertical profile at each of the 10 sampling sites using a modification of the porewater sampler as a multi-celled decomposition chamber (Schipper and Reddy, 1995). Decomposition chambers were constructed from sheets of ultra-high molecular weight (UHMW) polyethylene (60 x 10 x 2.5 cm), based on the design of the porewater sampler. Sample cells consisted of slots machined through the entire thickness of the sheets, thus the cells were open on each side of the apparatus. In addition, cell height was increased so that cell spacing was 2 cm on center. A faceplate with slots matching the cells was fastened to each side of the apparatus. Strips of plastic foam (open cell) air conditioning filter, held in place by the faceplates, were used to retain dead plant material within the cells while allowing mass transfer of dissolved constituents. Standing dead leaf material (still attached to the plant) from each of the 10 sampling sites was used as the organic substrate for evaluating in situ decomposition at the respective sites. Standing dead material from cattail plants was sampled at sites 1 through 8, and standing dead from sawgrass plants was sampled at sites 6 through 10. In each case, a composite sample of dead leaves from five plants was collected, dried in a forced-air drying room and chopped into 2-cm pieces. From each thoroughly-mixed composite sample, subsamples were obtained for initial chemical analysis. Total C and N analysis was

PAGE 115

109 performed on dried, finely ground (< 0.2 mm) samples using a Carlo-Erba NA-1500 CNS Analyzer (Haak-Buchler Instruments, Saddlebrook, NJ). Total P analysis was performed on separate subsamples following combustion (ashing) at 550 C for 4 h in a muffle furnace and dissolution of the ash in 6 M HC1 (Anderson, 1976). The digestate was analyzed for P using the automated ascorbic acid method (Method 365.4, USEPA, 1983). Ash weights were recorded for calculating sample ash content. Additional subsamples (0.5 g dry weight) were placed in 23 consecutive cells of the decomposition chambers. Decomposition chambers, containing dead plant native to each site, were installed vertically in the peat, with approximately 13 cells (26 cm) situated above the peat surface and 10 (20 cm) within the peat. Decomposition chambers containing cattail standing dead material were installed at sites 1 and 8 in triplicate, while one chamber each was placed at sites 2 through 7. For sawgrass, triplicate chambers were placed at sites 6 and 10, and single chambers at sites 7 through 9. The chambers were removed after approximately 6 months of field incubation on October 1 1 Contaminating soil organic matter (readily distinguishable from the plant material) was removed from the cells by gende washing with deionized water. Plant material was then removed from each cell and processed separately. Samples were dried in a forced-air drying oven at 60C, then weighed to determine mass loss. Samples from selected sites (1, 4, 6, 9 and 10) were analyzed for total N and P by micro-Kjeldahl digestion (Bremner and Mulvaney, 1982) followed by autoanalysis of digestate for NH 4 -N and P (EPA Methods 35 1.2 and 365.1, USEPA, 1983). Sample decomposition rates were calculated from mass loss (ash-free dry weight) and expressed as a first-order decay constant. Decomposition of the standing dead cattail and sawgrass leaves was assumed to follow simple first-order kinetics, based on results of previous decomposition studies in WCA-2A (Davis, 1991). The rate constant for each sample was calculated as

PAGE 116

110 k= ln[C 0 /C(t)] t [4 u t where k is the first order rate constant (d 1 ), C(t) is ash-free dry mass as a function of time, and C 0 is the initial mass of the sample. Thus, k for the entire incubation period was calculated from initial and final (t = 204 d) ash-free dry mass of the sample. Cotton Strip Assay A cotton strip assay (Latter and Howson, 1977; Howson, 1991), which measures short-term loss in tensile strength of buried cotton strips, was used to evaluate organic matter decomposition potential along the nutrient gradient. One strip (60 by 10 cm) of a standard cotton test fabric (Sagar, 1988) was affixed in a lengthwise manner to the back of each of the 10 porewater samplers (described above) for the October sampling event. The strips were attached at the top and bottom of the samplers using duct tape. Porewater samplers with attached cotton strips were installed at the 10 sampling sites, as previously described, on October 11, 1995. Cotton strips were thus oriented vertically in the water column-litter-peat profile, on the back side of the porewater samplers, opposite the open cells. The strips were retrieved along with the porewater samplers on October 24. At the time of retrieval, at each site a "control" cotton strip was inserted into the peat with a sharpshooter spade, then removed immediately for processing with the experimental strips (Harrison et al., 1988). Cotton strips retrieved from the field sites were placed in zippered plastic bags and stored on ice for transport to the laboratory. Cotton strips were prepared and tested according to methods presented in Harrison et al. (1988), with modifications noted in the following summary. The strips were washed with tap water, to remove soil and debris, and air-dried at room temperature. Each strip was cut horizontally at 4-cm intervals, and each sub-strip was labelled according to depth of incubation. Sub-strips were frayed along the previously cut edges to yield strips of exactly 3-cm width (depth increment). Sub-strips

PAGE 117

Ill were water-saturated prior to measurement of tensile strength. A tensometer was used to measure tensile strength of each sub-strip. Tensile strength loss over the 13-day period was determined from average initial (control) and final tensile strengths. Loss of tensile strength over time was expressed as a rate constant with units of time"', analogous to a first-order decay rate constant. Like first-order decay, tensile strength loss is not linear over time, therefore determination of a rate constant from initial and final values requires linearization of the decay curve. The following model was used to determine the "cotton rotting rate" (CRR), expressed as d"' (Hill et al., 1985): where y 0 and y are initial and final tensile strength, and t is the incubation time (days). The CRR was calculated for each sub-strip, so that direct comparison of rate could be made over the vertical profile and among sites. Temperature and Water Depth Air and soil temperature was monitored continuously at site 7, the approximate midpoint along the sampling transect, during the decomposition study period from June 7 through October 1 L. Thermocouples connected to a data logger were placed in the peat at depths of 0 (peat surface), 5, 10 and 20 cm. Hourly readings were averaged for each 24hour period, and stored as the daily mean temperature for each depth. Prior to June 7, air temperature data from the nearby Everglades Nutrient Removal (ENR) experimental field site (SFWMD, 1996) was substituted. Soil temperatures prior to June 7 were estimated from air temperatures, based on regressions of air versus soil temperatures during the June 7 through October 1 1 time period. Mean water depth at sites 1 through 10 during the study period was calculated from daily mean stage data for nearby WCA-2A field site 2-17 (SFWMD, 1996). Water depth was calibrated using depth data from individual sites during two different sampling events. [4-2]

PAGE 118

112 Results Water Chemistry Floodwater-soil porewater pH was generally within the 6.6 to 7.6 range at the beginning and end of the decomposition study (Figures 4-2 and 4-3). The ubiquity of CaC0 3 in the Everglades, a consequence of its geological origins, serves to buffer pH levels in the soil and water. Nevertheless, daytime pH was slightly elevated in the floodwater at most sites during both sampling events, due to the algal (periphyton) photosynthesis, and resulting depletion of dissolved C0 2 in the water. Site 2 represented a departure from the typical marsh water and peat profile (Figure 4-2). During the March to October study period the upper layer of peat (approximately 30 cm thick) over a wide area became increasingly separated from the basal peat. By the end of the study period, the upper peat had become a floating mat. This created an additional soil-water interface at the bottom of the mat. The effect on pH during the October sampling event was to raise pH near the bottom of the mat (Figure 4-3). Concentration of NH 4 -N was higher in the soil porewater than in the floodwater during both sampling events, indicating a net upward flux of NH 4 + from the peat (Figures 4-4 and 4-5). Since flux is a function of the concentration gradient, it may be a net upward or downward flux, depending on various factors governing soil and water chemistry. Surface water chemistry varies temporally near the S-10 inflows, depending on antecedent hydrologic conditions (SFWMD, 1992). Concentration of NH 4 -N in porewater is strongly influenced by mineralization rate of soil organic N to NH 4 + Accumulation of NH 4 -N in the soil or litter depends also on redox conditions. Under anaerobic conditions, nitrification of NH 4 + to NO," is not possible, thus NH 4 -N accumulates and also diffuses toward a region of lower concentration (e.g. the floodwater). Maximum accumulation of NH 4 -N was in the lower portion of the 0-30 cm peat profile at most sites during both sampling events. However, maximum NH 4 -N levels were found near the soil surface at site 5 in March and

PAGE 119

113 S o QLU Q 10 50-5-10 -15 -20 -25 -30 IDS' o-5-10-15-20 -25 -30 pH 8 6 8 6 8 6 f Site 1 — t w • • • • / • • • • • • • • • • • • • • Site 2 Site 3 • • Site 6 Site 7 Site 8 Site 9 o 6> a ^Site 10 Figure 4-2. Profiles of pH in surface water and soil porewater along the WCA-2A transect, sampled March 8-21, 1995.

PAGE 120

114 6 10-50•5-10-15-20? -25O w -30z t 0-5-10-15-20-25-30f J 8 6 Site 1 f Site 6 8 6 Site 2 • 7 Site 7 PH 7 I 8 6 Site 3 ( 7 Site 8 7 7 8 6 Site 4 Site 9 Figure 4-3. Profiles of pH in surface water and soil porewater along the WCA-2A transect, sampled October 1 1-24, 1995.

PAGE 121

115 10-k• NH 4 -N (mg L" 1 ) 8 0 6 8 0 6 8 0 Site 1 Site 2 Site 3 8 o Site 5 / 1 F— • I I .... I I 1 I 1 Site 6 Site 7 Site 8 Site 9 o Site 10 Figure 4-4.Profiles of NH 4 -N in surface water and soil porewater along the WCA2 A transect, sampled March 8-21, 1995.

PAGE 122

116 NH-N (mgL 1 ) 02468024680 2 4 680 246802 468 I -10-15-20? -25-] O — -30 \ Site 1 Q. UJ Q 10, 50-5-10-15-20-25-30} Site 2 Site 3* • • Site 6 I n I,, ,| • Site 7 Site 8 • Site 4 1 Site 9 i .... i .... i • Site 5 Site 10 Figure 4-5. Profiles of NH 4 -N in surface water and soil porewater along the WCA2A transect, sampled October 1 1-24, 1995.

PAGE 123

117 at sites 4, 5, 6, 8 and 9 in October. Concentrations were higher in the water column and litter layer (above the peat surface) at sites 3 through 6 during October than in March. Separation of the uppermost 30 cm of peat from the underlying peat at site 2 resulted in depletion of porewater NH 4 -N. Concentration of NH 4 -N decreased slightly with increasing distance from the inflow. Dissolved reactive P (DRP) profiles at nutrient-enriched and transitional sites closely resembled NH 4 -N profiles for both March and October sampling events (Figures 46 and 4-7). However, DRP concentration dropped sharply beyond site 6 (> 3 km from inflow) to "background" level at sites 9 and 10. This is consistent with the overall pattern of P enrichment south of the S10 inflows in WCA-2A, associated with loading of agricultural drainage. As with NH 4 -N, DRP concentration was higher in porewater than in the surface water, with an apparent upward flux from the peat resulting. Depletion of DRP at site 2 occurred within the floating peat mat, to a greater extent during October sampling. Depth of maximum accumulation of DRP was variable between sampling dates and among sites. Sulfate in WCA-2A water and porewater probably originates from groundwater discharge into the drainage canals which intersect the porous underlying limestone. Porewater and surface water S0 4 = concentration varied from site to site and between sampling events, but there were consistent trends (Figures 4-8 and 4-9). Concentrations were higher in surface water than in porewater, indicative of downward flux into the peat, and also generally higher near the inflow. Sulfate may be rapidly utilized in the peat as an electron acceptor, in the absence of 0 2 by sulfate reducing bacteria involved in organic matter decomposition. Enrichment of S0 4 = in the peat was more extensive near the inflow, due to the higher loading rate at those sites. Enrichment of S0 4 = in the peat was also greater in March than October, presumably associated with antecedent hydrologic conditions (water depth was greater in March).

PAGE 124

118 DISSOLVED REACTIVE P (mg L" 1 ) 0 0.5 1.0 1.5 2.00 0.5 1 0 1.5 2.00 0.5 1.0 1.5 2.00 0.5 1.0 1.5 2.00 0.5 1.0 1.5 2.0 10 5- 0-5 -10 -15^ -20| -25 -30 i .... i .... i r. Site 1 Q. 10 > a 5 0 •5-3 -10 -15 -20^ -25 -30( J i ^ Site 2 1 1 1 1 1 1 1 1 1 1 1 1 •v Site 3 V. Site 5 Site 6 .... i • If Site 7 ( Site 8 i .... i ... i Site 9 Site 10 Figure 4-6. Profiles of dissolved reactive P in surface water and soil porewater along the WCA-2A transect, sampled March 8-21, 1995.

PAGE 125

119 Figure 4-7. Profiles of dissolved reactive P in surface water and soil porewater along the WCA-2A transect, sampled October 1 1-24, 1995.

PAGE 126

120 Figure 4-8. Profiles of sulfate in surface water and soil porewater along the WCA2A transect, sampled March 8-21, 1995.

PAGE 127

121 S0 4 = (mg L' 1 ) 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 Figure 4-9. Profiles of sulfate in surface water and soil porewater along the WCA2 A transect, sampled October 1 1-24, 1995.

PAGE 128

122 Temperature and Water Depth Mean water depth for the sampling transect over the March through October study period is shown in Figure 4-10. Depth at individual sites typically varied less than 10 cm from the mean depth. Water depth was less than 1 m throughout the study period, decreasing to less than 20 cm during the period May 18 through June 22. Field observations made at the time of minimum water depth ( 10 cm mean depth) verified that surface water was present at all sites along the transect. Daily mean air and soil temperatures (measured at site 7) varied within a range of approximately 10 degrees during the field decomposition period (Figure 4-11). Significant relationships between air and soil temperatures were established using linear regression, for the period June 6 through October 1 1 These relationships were used to calculate mean daily temperatures for the four soil depths over the entire study period, using the mean daily air temperature of 27.2 C. Differences in mean temperature from the litter layer to 20 cm peat depth were less than one degree. Mean temperatures for the study period were 27.0, 27.0, 26.8 and 26.4 C at depths of 0 (litter), 5, 10 and 20 cm. Furthermore, during the period June 6 through October 1 1 the difference in daily mean temperature among soil depths averaged 0.8 degrees (standard deviation = 0.7 degrees), and never exceeded 2.5 degrees. Because temperature variation with depth was minor, decomposition rates were not temperature-corrected. Field Decomposition Study Initial N and P content of the standing dead leaf material used in the decomposition study varied significantly along the WCA-2A transect (Table 4-2). Linear regression analysis verified that initial tissue N and P content decreased significantly (a = 0.05) with increasing distance from the surface water inflow (S-10C). However, N and P content did not differ by plant species, i.e. there was not significant difference between cattail and

PAGE 129

123 100 "I 1 1 1 1 1 1 1 80 110 140 170 200 230 260 290 JULIAN DAY Figure 410. Mean surface water depth for the 10 WCA-2A sampling sites during the field decomposition study period (March 22 through October 1 1, 1995). Data were derived from daily stage readings at WCA-2A station 2-17 (SFWMD, 1996) using measured water depth at each site at the beginning and end of the study period.

PAGE 130

124 32 JULIAN DAY Figure 4-11. Daily mean temperature of air and soil at WC A-2A site 7 during the field decomposition study period (March 22 through October 11, 1995). Soil temperature was measured at four depths relative to the peat surface. Air and soil temperature data were recorded at site 7 during the period June 7 through October 1 1 1995. Air temperature data for the March 22 to June 7 period were derived from a nearby Everglades field research site (SFWMD, 1996), using a constant correction factor of + 1C.

PAGE 131

125 Table 4-2. Initial chemical analysis of standing dead leaf material used for field decomposition study in WCA-2A. Site Plant type Total C Total N Total P Ash Lignin Cellulose ----gkg' mg kg' % of dry mass 1 Cattail 466 8.0 339 2.3 15.3 38.8 2 Cattail 460 7.3 343 2.8 3 Cattail 463 5.7 375 2.0 4 Cattail 457 5.3 277 2.5 10.0 42.0 5 Cattail 457 6.4 373 2.3 6 Cattail 463 6.0 310 2.2 11.4 42.9 Sawgrass 466 4.3 487 3.1 12.9 34.3 7 Cattail 468 4.7 315 2.4 Sawgrass 457 5.3 478 4.2 8 Cattail 468 4.0 205 2.8 Sawgrass 456 5.7 241 4.5 9 Sawgrass 454 4.0 140 4.8 16.1 35.5 10 Sawgrass 447 4.6 61 4.2 14.2 38.1

PAGE 132

126 sawgrass. Ash content of the standing dead material increased significantly with distance from the inflow, primarily due to the significant difference between cattail and sawgrass. Neither lignin nor cellulose content differed significantly with distance from the inflow or among plant type. Decay rate (as k) decreased significantly with depth (linear regression analysis, a = 0.05), from top to bottom of the water-soil profile, at all locations except site 8 (Figure 412). Depthaveraged (each site) decay rate decreased significantly with increasing distance from the inflow, according to linear regression analysis. Analyisis of variance (ANOVA) showed that mean (depth averaged) decay rate was significantly higher at sites 1, 2, 4 and 5 than the remaining sites, and significantly lower at sites 9 and 10. When decay rates of cattail (sites 1 through 8) and sawgrass (sites 6 through 10) were evaluated separately, the same significant trend of decreasing rate away from the inflow was observed. At sites 6 through 8, where both cattail and sawgrass samples were incubated, the depth-averaged decay rate of cattail was significandy higher than sawgrass at all three sites. Total N enrichment of the decomposing plant tissue occurred during the 6-month incubation period (Figure 4-13). The increase in N content, presumably due to uptake of N by microbial decomposers associated with the organic substrate, was more pronounced at sites closer to the inflow. Final N concentration also showed a slight decrease from top to bottom of the water-soil profile, similar to the pattern shown by decay rate at the same sites. In addition, increase in substrate N concentration was greater for cattails than for sawgrass at the same site (site 6). Increase in total mass of substrate N indicated that net immobilization of N associated with decomposition occurred along the entire nutrient gradient (Figure 4-14). The initial mass of N for each set of samples is indicated by the vertical dashed lines in Figure 4-14. The difference between the initial and final values represents net immobilization of N. The extent of N immobilization was relatively uniform among the sites sampled, with the exception of site 10, for which there was only limited immobilization of N.

PAGE 133

127 25 2015 1050-5 -10•15 -20 -25 25 20 15^ £ 10 S 5 I 0^ o. -54 UJ Q -10DECAY RATE CONSTANT k (x1 • A O a 5 £ Sites • Cattail Site 2 • Cattail 2 3 4 5601 2345601 23456 Site 3 Cattail V Site 4 Cattail Site 5 Cattail Q o p ox • o a • m o* • o A> : a s • AO K OA Site 6 Sawgrass Site 7 Cattail Site 7 Sawgrass Sites Sawgrass Site 9 Sawgrass a AO AO a§ S JO a so % Site 10 Sawgrass Figure 4-12. Ash-free dry mass loss rate, (expressed as first order decay constant) for cattail and sawgrass standing dead tissue contained in field decomposition chambers. Standing dead plant material was collected at the site of field incubation. Triplicate chambers were placed at sites 1 and 8 for cattail leaves and at sites 6 and 10 for sawgrass leaves.

PAGE 134

128 TISSUE N CONCENTRATION (g kg" 1 ) 2520151050-5 -10-15-20 -25 25 20 -2510 20 30 i i Site 1 Cattail 1 1 1 • • i i i i i i i i • • • • • • • • -J Site 6 i Sawgrass 10 20 30 0 10 20 30 i | | i | l_i i_i 1 1 1 1 1 1 Site 4 Cattail i i i i Sited Sawgrass • Site 6 Cattail L^| ' Siteta Sawgrass Figure. 4-13. Final total N concentration in detrital plant tissue (cattail and sawgrass) at selected sites along the WCA-2 A transect after 6-month field decomposition study. Dashed lines represents initial values for the respective sites.

PAGE 135

129 TISSUE N MASS (mg) 01 2345601 234560 1 23456 E -20 Z -25 F & i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2520151050-5-1015H Sitel Cattail : i 1 ) * i Site 4 Cattail I 2520151050 -5-10-15-20-25 1 1 1 1 1 i.l 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 • Site 6 •^Sawgrass i i i i 1 i i i y • • • • ^ 1 '.' 1 1 1 Site 6 ; Cattail % ; m. Site 9 • Sawgrass j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i Site 10 !• Sawgrass Figure 4-14. Total N mass remaining in detrital plant tissue (cattail and sawgrass) at selected sites along the WCA-2A transect after 6-month field decomposition study. Dashed lines represents initial values for the respective sites.

PAGE 136

130 Total P concentration in the decomposing substrate increased variably during the decomposition period (Figure 4-15). Final substrate P concentration was somewhat higher in the water column and litter layer than in the soil. Increase in P concentration was confined for the most part to the more nutrient-enriched sites (sites 1 through 6). Substrate P concentration was higher in cattail leaves than in sawgrass at the same site (site 6). Maximum P enrichment was found near the soil-water interface. This was especially pronounced at site 6. Little or no P enrichment of substrate occurred at the low-nutrient sites (sites 9 and 10). Net immobilization of P occurred primarily in association with cattail leaves at sites 1 through 6 (Figure 4-16). Net immobilization took place throughout the water and soil profile at the nutrient-enriched sites, but was restricted to the vicinity of the soil-water interface at the remaining sites. Essentially no immobilization of P occurred at site 10, the most nutrient-poor site. For sawgrass leaves at site 6, P was immobilized near the soil surface, but net mineralization and loss of P occurred in the water column and lower peat. The molar N:P ratio of the organic substrate changed little during the decomposition period at the nutrient-enriched sites 1 and 4, but was more variable at the remaining sites (Figure 4-17). Low initial N:P ratios at sites 1 through 6 were a product of the gradient of P enrichment in the soil and vegetation in WCA-2A. The slight decrease in N:P ratio (relative to initial N:P ratio) of cattail leaves near the soil-water interface at these sites were an indication of relative P enrichment over time. This phenomenon also occurred in sawgrass leaves at sites 9 and 10, where initial N:P ratio was much higher. Increased N:P ratio in the water column and in deeper peat layers indicated a relative enrichment of N. This occurred to the greatest extent at sites 6, 9 and 10. Cotton Strip Assay Cotton rotting rate (CRR) along the vertical watersoil profile typically showed a distinct peak near the soil-water interface, especially at the "transitional" sites 5 through 8

PAGE 137

131 TISSUE P CONCENTRATION (mg kg" 1 ) 0 500 1 000 1 500 2000 2500 0 500 1 000 1 500 2000 2500 0 500 1 000 1 500 2000 2500 25i i i i I i i i I t I i i 1 i t i I 20151050-5-10E 15 -20X -25254 UJ 20Q 151050-5-10-15-20-25-Sitel Cattail : I I I I I I I I I I I I I I I I I i Site 6 Sawgrass i i i 1 1 V I Site 4 Cattail 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 : Site 9 Sawgrass • : : 1 1 •• • i i • ii m • Site 6 Cattail 0' 1 1 1 1 1 1 1 1 1 1 1 Site 10 Sawgrass Figure 4-15. Final total P concentration in detrital plant tissue (cattail and sawgrass) at selected sites along the WCA-2 A transect after 6-month field decomposition study. Dashed lines represents initial values for the respective sites.

PAGE 138

132 E o J— Q. Ill Q TISSUE P MASS (mg) 0.2 0.4 0.6 0 2520151050-5-10-15-20-252520151050-5-10-15-20-25— — — J — — — : > m •• 1 • • • • • • • i i : Sitel Cattail ::• I. I Site 6 Sawgrass 0.2 i l i -r 0.4 i I i 0.6 0 0.2 : V Site 4 Cattail V Sited Sawgrass 0.4 i I i i 0.6 Sitee Cattail Site 10 Sawgrass Figure 4-16. Total P mass remaining in detrital plant tissue (cattail and sawgrass) at selected sites along the WCA-2A transect after 6-month field decomposition study. Dashed lines represents initial values for the respective sites.

PAGE 139

133 TISSUE N:P RATIO 100 200 300 0 100 200 300 0 2520151050-5-10-15-20-252520ISMS' 0-5-10-15-20-25-I L— I I I 1—1 1 t I i i i i_ • < i •i. 4 Site 1 Cattail — 1 : : • • • • • • i i v Site 6 Sawgrass -1 I I I I I I I L_J I I : 1 • 3 i 1 1 'i < • 1 5 Site 4 Cattail Sited Sawgrass 100 200 300 : V Site 6 Cattail 1 1 1 1 1 1 1 • 1 Site 10 \ Sawgrass 1 Figure 4-17. Molar N:P ratio of detrital plant tissue (cattail and sawgrass) at selected sites along the WCA-2A transect after 6-month field decomposition study. Dashed lines represents initial values for the respective sites.

PAGE 140

134 (Figure 4-18). A less distinct vertical trend was observed for sites 1 and 10, although the mean CRR was substantially higher at site 1 Moreover, the depth-averaged CRR exhibited a significant decrease along the nutrient gradient (downgradient). Minimum CRR in the vertical profile was typically found in the lower regions of the sampling depth (30 to 40 cm). The exception was at site 2, where the region below the 30 cm soil depth was actually water beneath the floating mat of peat. Discussion Rate of decomposition of organic matter, such as senesced leaf tissue, is governed by numerous factors related to substrate composition and environmental conditions. Substrate quality refers to availability of C and growth-limiting nutrients to microbial decomposers. Initial composition of dead plant material used in the WCA-2A decomposition study varied according to site location. A significant trend of decreasing tissue N and P content with increasing distance from the inflow was found (Table 4-2), although differences between means for cattail and sawgrass were not significant. Mean values of lignin and cellulose content were not significantly different among plant type or with distance from the inflow. Ligno-cellulose index (LCI) (Melillo et al., 1989), the relative proportion of lignin in the ligno-cellulose component of the substrate, ranged from 0.2 to 0.3 among the samples analyzed. These LCI values were intermediate between those measured for live plant tissue and plant litter at the soil surface at corresponding field sites (Chapter 3), although more similar to values for live plants. Soil oganic matter and peat, representing the result of extensive decomposition, typically have LCI values of about 0.8 (Chapter 3; Melillo et al., 1989). Despite the fact that decomposition rate decreased down-gradient from the inflow, the mean rates for each site (depth-averaged, cattail and sawgrass treated separately) were not highly correlated with initial substrate N or P content (r = 0.56 and 0.25, respectively). However, when above-ground cells of the decomposition chambers were considered

PAGE 141

135 CRR (d 1 ) 0.1 0.2 0 0.1 0.2 0 0.1 0.2 0 0.1 0.2 0 0.1 0.2 30E o 2010o-i -10-20-30 -i -40j — i — i — i — 1_ • Site 1 • • • • • • • • • • • • • • • • • • • Site 2 • • Site 3 • I I I I I I I I I I Site 4 Site 5 Q. LU Q 30 20-i 10H 0 -10-3 -20 -30 -40 Site 6 • • • • • • • • • • • • • • • • • • • • • • • .* Site 7 •* Site 8 • • J — I — I — I — L .1— i— L ,1. J-. \ Srte 9 Figure 4-18. Daily cotton rotting rate (CRR) for cotton strips placed vertically in the water and soil profile along the WCA-2A sampling transect.

PAGE 142

136 separately (corresponding to the litter layer and water column), the relationship between initial substrate nutrient content and decomposition rate was somewhat improved, especially for initial N content (r = 0.74 and 0.32 for initial N and P. respectively). The correlation between decomposition rate and initial substrate N was greater still (r = 0.83) in the upper portion of the water column (i.e. > 10 cm above the peat surface) This suggests that N availability may have been a limiting factor for decomposition in the floodwater. Vertical profiles of NH 4 -N (Figures 4-4 and 4-5) were evidence of the depletion of NH 4 -N in the floodwater relative to porewater. At sites 6, 7 and 8, where cattail and sawgrass material were incubated in adjacent decomposition chambers, mass loss was greater in cattail than sawgrass. Nevertheless, at these same sites, total P content was higher in sawgrass than in cattails, and total N content was higher in sawgrass at sites 7 and 8. However, LCI was higher in sawgrass than cattail ( 0.27 and 0.21, respectively) at site 6 (the only site of the three for which lignin and cellulose data were available). It is possible that C availability was a secondary factor controlling decomposition rate, so that, given the same conditions of nutrient availability, the less-lignified substrate decomposed more rapidly. In addition, the higher ash content of sawgrass may be related to availability of C and organically-bound nutrients due to physical protection of the organic matter by mineral deposits, for example silica. Although experimental results suggested that decomposition rate of standing dead material was highly influenced by N and P availability, there was only a weak correlation between decomposition rate and dissolved nutrients (NH 4 -N and DRP). This included March and October water chemistry as well as combined data for the two sampling dates. Furthermore, no correlation was found between decomposition rate and sulfate concentration, for any depth interval. Thus from available data, it is not clear whether the supply of this alternate electron acceptor affects decomposition in the absence of 0 2 The lack of correlation between water/porewater chemistry and decomposition rate may be due to the high spatial and temporal variability of dissolved constituents.

PAGE 143

137 In contrast, final substrate N and P concentrations were highly correlated with decomposition rate (r = 0.91 and 0.71 for N and P, respectively). Final substrate P was more closely correlated with decomposition rate within the peat layer (r = 0.87) and in the water column (> 10 cm above the peat surface) (r = 0.88). This provided indirect evidence t hat decomposition was controlled p rimaril y by e xternal nutrient availability A two-year in situ decomposition study was previously conducted at three sites in WCA-2A, representing nutrient-enriched, transitional and non-enriched areas (Davis, 1991). Cattail and sawgrass standing dead material placed in the litter layer at the three sites decomposed more rapidly at the nutrient-enriched site, and least rapidly at the nonenriched site. In addition, first-order decay rates calculated from Davis' data were higher for cattail (18.1, 9.2 and 9.0 x 10" 4 d" 1 ) than for sawgrass (10.8, 6.6 and 6.2 x 10" 4 d" 1 ), at each site. It is noteworthy that decomposition rate was more highly correlated with total P concentration in the litter layer and the 0-10 and 10-30 cm depth increments of peat (Chapter 3) than with porewater N or P concentration. Similar relationships were found between CRR of cotton strips and soil and water nutrients. The CRR was not correlated with dissolved N or P (or with S0 4 = ) measured simultaneously with the cotton strip incubation. When porewater data from both sampling dates were combined, DRP concentration explained 62 and 65% of the CRR variability within the litter layer (0-10 cm above the peat surface) and in the uppermost 10 cm of peat. On the other hand, soil and litter total P content explained 57% of the variability in CRR, averaged for each of three depth intervals (litter layer, 0-10 cm of peat and 10-30 cm of peat). Results of this study suggest that single measurements of floodwater and porewater nutrient concentrations are not suitable indicators of long-term nutrient availability. Total nutrient concentrations in the litter and peat are probably more useful indicators of nutrient availability, by virtue of integrating rapidly fluctuating processes over time and space.

PAGE 144

138 Increases in substrate nutrient concentration during the 6-month decomposition study (Figures 4-13 and 4-15) showed that decomposition of standing dead plant tissue was initially limited by nutrient availability, rather than C availability. After senescence and before deposition onto the litter layer overlying the peat, the nutrient supply for microbial decomposers on cattail and sawgrass leaves is essentially terminated, except for rainfall and dry precipitation. Nutrients incorporated in the original plant material may be lost by mineralization and leaching. Furthermore, during this period of up to several months, dessication may become a primary factor limiting decomposition. In a nutrient-enriched area such as the region proximal to surface inflows in WCA-2A, standing dead material deposited in the litter layer may decomposed much more rapidly than in nutrient-poor areas. The relatively high-quality C source (as reflected by LCI values) is readily available to microbial decomposers where nutrient supply is sufficient. Enrichment of the substrate with N or P was indicated by increased final concentration over initial concentration (Figures 4-13 and 4-15). The relative amount of N and P enrichment decreased with increasing distance from the inflow. Minimal N enrichment and no P enrichment occurred at site 10, about 10 km from the inflow. At site 6, a transitional area, P enrichment was minimal or negative (P depletion) in the upper water column and lower peat, but relatively high near the soil surface. This may have resulted from release of mineralized P in surrounding litter, or from faunal activities. Since the C quality of substrates from all sites was comparable, based on lignin and cellulose content, the concentrations of N and P at the end of the field incubations could be construed as indicators of N and P availability. This is a refinement of the previously stated concept of assuming total N and P in native litter and peat to be indicators of availability. In the latter case, though the total amount of N and P may be proportional to available N and P, the proportionality constant may vary greatly depending on C availability for microbial decomposers, as well as on the presence of physico-chemical sinks, particularly for P. Similarly, a change in N:P ratio during decomposition (Figure 4-17) may indicate relative

PAGE 145

139 deficits of N and P in the original substrate. It might follow that, between N and P, it is P which limits microbial growth in the non-impacted interior marsh (i.e. site 10) and N in the highly impacted region near the inflow. Microbial respiration studies in peat from Everglades National Park have indicated that P is the primary microbial growth-limiting nutrient in pristine areas, but that N may be limiting in high-P areas (Amador and Jones, 1993; 1995). Closely related to initial and final substrate nutrient concentrations, but carrying somewhat different implications, is the net change in N and P mass during the decomposition period (Figures 4-14 and 4-16). Increased mass of N or P over the initial mass represented net immobilization, which was a function of microbial nutrient demand and nutrient supply, with nutrient demand based on substrate quality. Immobilization of N and P after 6 months was greater near the inflow and minimal at more distant sites. Net release of P occurred in the deeper peat layers at site 6. The high degree of immobilization occurring at nutrient-rich sites was likely due to the low initial nutrient content of the substrate and relatively high proportion of cellulose in the lignocellulose matrix. Immobilization potential has been linked to initially low N and lignin content of the organic substrate (Melillo et al., 1984). Studies in a cypress swamp in north Florida receiving municipal wastewater showed that, three weeks after addition of 15 N-labelled wastewater, nearly all of the N remaining in the peat-floodwater profile was immobilized in the litter layer (DeBusk and Reddy, 1987). Mass immobilization is a more quantitative parameter than enrichment, as expressed by concentration increase, and is thus suited for mass balances of N and P. However, N and P pools such as in newly decomposing litter are timevariant, and data such as presented in Figures 4-14 and 4-16 are essentially a snapshot of a dynamic process. That is, net immobilization may result from nutrient enrichment of the substrate in the short term due to microbial requirements, but subsequent mineralization of substrate organic C results in remineralization and release, and thus mass loss, of nutrients. This sequence of

PAGE 146

140 immobilization and subsequent release of nutrients in decomposing litter is commonly observed in mass loss studies in various ecosystems (Melillo et al., 1984; Reddy and DeBusk, 1991). Although the CRR of cotton strips did not correlate well with nutrient concentrations, it is evident that nutrient availability was the primary factor regulating cotton degradation. The cotton strips provided a readily available cellulose substrate (Sagar, 1988) for a variety of aerobic and anaerobic microbial decomposers. Loss of tensile strength was not indicative of complete decomposition (C mineralization), but of cellulase activity. However, microbial enzyme activity has been shown to be a good indicator of microbial respiration in soils (Linkins et al., 1990; Sinsabaugh et al., 1991). Spatial trends of CRR were similar to trends observed for the plant decomposition study. That is, nutrient availability, and not necessarily measured nutrient concentrations, accounted for the variability in CRR along the gradient. Use of a uniform substrate, initially devoid of nutrients, eliminated variability associated with substrate composition. Increased CRR in the litter layer and upper region of peat was observed at highnutrient and transitional sites. This was probably associated with increased nutrient availability relative to the open water column and increased 0 2 availability relative to the subsurface peat. A similar trend was reported for a study in Shark Slough in the Everglades National Park, in nutrient-enriched flow-through channels (Maltby, 1988). Greater loss of tensile strength occurred in the surface peat and plant detrital (litter) layer than in the water column. Addition of N and P significantly enhanced tensile strength loss, with P addition having the greatest effect in the peat and N+P addition in the water column. A comparison among wetlands in the southeastern U.S. showed that CRR was higher in disturbed wetlands than in undisturbed wetlands (Bridgham et al., 1991). Drainage was the dominant factor affecting CRR, while nutrient availability and pH also exerted strong effects on CRR. Cotton rotting rate, or loss of tensile strength, is not a quantitative measure of decomposition, and at best provides an indicator of spatial or temporal trends in

PAGE 147

141 decomposition potential among related sites (Howson, 1991). Therefore, qualitative comparisons among unrelated sites, especially based on limited data sets, would be inadvisable. Summary and Conclusions Decomposition of site-native plant material increased closer to the inflow, indicating that in situ decomposition in the litter layer increased with nutrient enrichment in WCA-2A. Results also suggested that nutrient enrichment directly or indirectly (via substrate quality or environmental factors) affected decomposition potential in the peat profile. Decomposition rate along the vertical profile, for which substrate composition was uniform, showed a general decrease from floodwater/litter layer to peat, reflecting the influence of environmental factors, including nutrient and 0 2 availability. Due to the short time span (6 months), the study represented a relatively early stage of the decay continuum from plant tissue to peat, and initial substrate quality exerted an effect on decomposition rate. Decomposition rate of the dead cattail and sawgrass leaves, representing newly-deposited plant material to the litter layer, was therefore governed by the interactive effects of initial chemical composition of the dead material and environmental factors. Variability in cotton strip decay reflected only environmental factors, thus the cottona strip assay may have been a more sensitive indicator of the "decomposition potential" at a given location or depth in the floodwater-peat profile. Although CRR was not closely correlated with dissolved nutrient concentration, there was strong evidence, based on previous total N and P analysis of litter and peat, that both N and P supply from the surrounding matrix were controlling factors for decomposition. Substantial immobilization of N and P occurred in the decomposing plant material, especially within the litter layer and near the inflow. The high immobilization potential of the standing dead material was due to a combination of extremely low nutrient content and relatively high C quality. This shows that the newly-deposited litter in nutrient-enriched

PAGE 148

142 areas can serve as a strong sink for nutrients, with relatively rapid turnover. This carries significant implications regarding the importance of the litter layer in short-term nutrient cycling in wetlands. The effects of nutrient availability on decomposition in wetlands has received far less attention than in terrestrial ecosystems. Nevertheless, understanding the impact of nutrient loading on decomposition has major implications with regard to ecosystem stability in natural wetlands and long-term efficiency of wastewater-treatment wetlands.

PAGE 149

CHAPTER 5 DETRITAL CARBON MODEL Introduction Response of C mineralization rates in plant litter and peat to various environmental and substrate-related effects was evaluated and discussed in earlier chapters. A mathematical model with simulation, based on previously discussed concepts and experimental results, will be presented in this chapter. The purpose of the model is to (1) summarize and demonstrate the relationships among turnover of organic C pools along the decay continuum and previously identified rate-limiting factors, (2) test the relative significance of each factor on model output, (3) compare calculated C mineralization rates based on regressions developed in Chapter 3, with observed rates and (4) evaluate dynamic and steady-state responses of the state variables to lowand high-nutrient conditions as predicted by experimentally-determined relationships. Simulation of environmental conditions and substrate composition along the WCA-2A nutrient gradient will provide information on the validity of the conceptual model and accuracy of the mathematical relationships previously developed for C turnover. Finally, the model is intended to provide insight into the dynamics of C turnover in the WCA-2A wetland and to identify areas for further research. Materials and Methods Model Description A compartment model of organic C flows and mineralization in the detrital subsystem of Everglades WCA-2A was developed using the graphical modeling program 143

PAGE 150

144 Stella II for the Macintosh (High Performance Systems, Inc.). The icon-based Stella language contains built-in components and functions which are automatically linked to graphical elements, thus allowing simultaneous construction of the mathematical model and the model diagram. Difference equations are solved numerically, with options for Euler or Runge-Kutta methods, with user-defined time step (dt). Graphical elements of the Stella system represent state variables or storages, flows, parameters, auxiliary variables and input variables (Figure 5-1). Difference equations are automatically formulated for state variables, such that for each time step an appropriate amount of matter, as defined by the connecting flows, is added or subtracted from the previous amount. A flow may connect two state variables or a state variable and the external environment (outside the system boundary). In the latter case, the "open" end of the flow is represented by a cloud symbol. An open inflow thus represents a forcing function, or inflow of matter from outside the system. An open outflow represents loss of matter from the system, and may be considered a sink. Parameters and variables are represented by the same graphical element, and are connected to flows, state variables and other parameters and variables with arrows (Figure 5-1). This indicates a flow of information, rather than matter, in the direction of the arrow. Stella automatically identifies donor (at the origin of arrows) components as inputs for mathematical expressions for processes or auxiliary variables. The Everglades detrital C model consists of four state variables (Figure 5-2; Table 5-1). Standing dead (attached dead) plant material is designated by the variable stdDead, the soil Utter layer by the variable litter, the surface or "active" layer of peat by activePeat and buried or stabilized peat by stablePeat. Input of organic C to the system occurs through senescence of plant tissue. Only above-ground biomass is considered, as in the previous chapters. Mass flow of organic C among compartments occurs as litterfall (deposition of standing dead) and burial of litter and active peat. Loss of C from the model system occurs from each pool via mineralization of organic C by microbial decomposers. Simultaneous

PAGE 151

145 Figure 5-1. Basic components of the Stella II graphical modeling language.

PAGE 152

146 < i

PAGE 153

147 Table 51 Components of the organic C simulation model for the Everglades marsh detrital subsystem. Refer to model diagram in Figure 5-2. State Variables (C storage in g m 2 ): stdDead(t) = stdDead(t dt) + (senesce minSD litterfall) dt litter(t) = litter(t dt) + (litterfall minL burial) dt activePeat(t) = activePeat(t dt) + (burial minAP burialAP) dt stablePeat(t) = stablePeat(t dt) + (burialAP minSP) dt Processes (C flows in g m 2 y '): senesce = (flow from outside the system) detrital production initiated by plant leaf senescence minSD = stdDead kminSD mineralization of standing dead plant material, representing loss of C from the system litterfall = stdDead kLitter transfer of organic Cfrom standing dead to the soil litter layer via litterfall minL = litter kminL mineralization of litter, representing loss of Cfrom the system burial = litter kBurial transfer of organic Cfrom the soil litter layer to active peat minAP = activePeat kminAP mineralization of active peat, representing loss of C from the system burialAP = activePeat kBurialAP transfer of organic C from active peat to stable peat minSP = stablePeat kminSP mineralization of stable peat, representing loss of C from the system Auxiliary Variables: kLitter = 1/restimeSD first-order rate constant for litterfall, based on specified remSD kBurial = 1/restimeL first-order rate constant for burial of litter, based on specified remUT kBurialAP = 1/restimeAP first-order rate constant for burial of active peat, based on specified remAP

PAGE 154

148 Table 5-1 -continued. kminSD = (5.2E-06 sdTP 3.0E-03 sdLCI 5.0E-06 sdTP sdLCI + 0.0022) 365 moisture first-order decay rate constant for mineralization of standing dead, determined from substrate LCI and TP content (mg kg' 1 ) and moisture availability factor kminL = (5.2E-06 litTP 3.0E-03 litLCI 5.0E06 litTP litLCI + 0.0022) 365 floodL first-order decay rate constant for mineralization of litter, determined from substrate LCI and TP content (mg kg') and flooding (0 2 availability) factor kminAP = (5.2E-06 actptTP 3.0E-03 actptLCI 5.0E-06 actptTP actptLCI + 0.0022) 365 floodAP first-order decay rate constant for mineralization of active peat, determined from substrate LCI and TP content (mg kg') and flooding (0 2 availability) factor kminSP = (5.2E-06 staptTP 3.0E-03 staptLCI 5.0E-06 staptTP staptLCI + 0.0022) 365 floodSP first-order decay rate constant for mineralization of stable peat, determined from substrate LCI and TP content (mg kg') and flooding ( 0 2 availability) factor restimeSD = -LN(remSD)/kminSD residence time for C in standing dead compartment based on static system, computed as time until fraction remSD of original mass remains restimeL = -LN(remLIT)/kminL residence time for C in litter compartment based on static system, computed as time until fraction remLIT of original mass remains restimeAP = -LN(remAP)/kminAP residence time for C in active peat compartment based on static system, computed as time until fraction remAP of original mass remains Parameters: sdLCI lignocellulose index (LCI) for standing dead compartment litLCI lignocellulose index (LCI) for litter compartment actptLCI lignocellulose index ( LCI) for active peat compartment staptLCI lignocellulose index (LCI) for stable peat compartment

PAGE 155

149 Table 5-1 -continued. remSD remaining C mass at standing dead outflow, as a fraction of incoming C remLIT remaining C mass at litter outflow, as a fraction of incoming C remAP remaining C mass at active peat outflow, as a fraction of incoming C Input Variables: sdTP average total P content (mg kg ') of standing dead material litTP average total P content (mg kg' 1 ) of soil litter layer actptTP average total P content (mg kg'') of active peat layer staptTP average total P content (mg kg ') of stable peat layer moisture scaling factor (0-1) for standing dead mineralization rate constant based on moisture availability floodL scaling factor ( 0-1) for litter mineralization rate constant based on flooding and 0 1 availability floodAP scaling factor (0-1) for active peat mineralization rate constant based on flooding and 0, availability floodSP scaling factor ( 0-1) for stable peat mineralization rate constant based on flooding and O, availability

PAGE 156

150 decomposition and burial in the model is analogous to the decay continuum discussed in Chapter 3. The continuum is represented in the model system as discrete pools characterized by sequential stages of decomposition. Mass of C remaining in the four pools represents total accumulation of organic C in the detrital subsystem (litter and peat). Flows, with the exception of senescence, are modeled as first-order processes, therefore the magnitude of flow is determined by current storage in the state variables and first-order decay rate constants (Table 5-1). Values for mineralization rate constants (kminSD, kminL, kminAP and bninSP) are calculated in the model based on a regression equation developed from experimental data presented in Chapter 3 (Tables 3-2 and 3-3): k = (5.2 x 10" 6 )7P (3.0 x 10" 3 )lC7 (5.0 x 10" 6 )7P • LCI + 0.0022 [5-1] where k is the first-order aerobic C mineralization rate constant (d 1 ), TP is substrate total P concentration (mg kg" 1 ) and LCI is the lignocellulose index. The term TP*LCI is a representation of the interactive effect of TP and LCI on mineralization rate. The R 2 for equation 5-1 is 0.899; the significance values (P-values) of the terms TP, LCI and TP'LCI are <0.0001, 0.0013 and 0.0034, respectively. Input variables and parameters associated with total P and LCI are listed in Table 5-1. Values for LCI were previously shown to be relatively invariant within compartments, therefore these are considered parameters instead of input variables. Each mineralization constant is attenuated by a scaling factor, ranging in value from 0 to 1 related to moisture and 0 2 availability. The variable moisture represents the estimated occurrence of wetted conditions on the standing dead tissue, from rainfall or dew. The variables floodL, floodAP andfloodSP allow scaling of mineralization rate constants according to flooded and drained conditions, or alternatively, anaerobic and aerobic conditions. Values represent average conditions for at least one time unit, i.e. one year. The practical lower limit for these input variables is 0.32 (constantly anaerobic conditions), based on the experimentally-derived relationship between aerobic and anaerobic

PAGE 157

151 mineralization rate (Chapter 3), ANAEROBIC RATE = 0.32 x AEROBIC RATE. [5-2] Rate constants for transfer of C from one compartment to the next are calculated in the model from the parameters remSD, remUT and remAP (Table 5-1). Compartments are subjectively defined by the degree of decomposition of the organic substrate, expressed as the proportion of mass export to import (excluding mineralization) for each compartment. This is based on a static first-order model of decomposition: C(t) = C(0)e" kmint [5-3] where is the previously determined mineralization rate constant and C(t)/C(0) is the fractional mass of C remaining at time t. Solving equation 5-3 for t yields the turnover time for mass transfer, termed the residence time. The latter term is applied to avoid confusion with total compartment turnover time, which is determined by total mass loss due to transfer and mineralization. Rate constants for mass transfer (kLitter, kBurial and kBurialAP) are subsequently calculated as the reciprocal of the residence time. Sensitivity Analysis Simulations of low-nutrient, transitional and high-nutrient conditions were performed using input and parameter values (Table 5-2) experimentally determined from laboratory and field studies described in previous chapters. Characteristics of the standing dead, litter, active peat and stable peat compartments in the model system were based on standing dead, litter, surface (0-10 cm) and buried (10-30 cm) peat characteristics along the nutrient gradient in WCA-2A (Chapter 3). The ranges of values thus assigned to parameters, variables and inflow are listed in Table 5-2. Total P concentrations for the four compartments under high-nutrient, transitional and low-nutrient conditions (Table 5-2) were determined by averaging data (reported in Chapter 3) for sites 1-5, 6-8 and 9-10, respectively. The exception was the stablePeat compartment, for which average total P levels for all nutrient conditions were determined

PAGE 158

152 Table 5-2. Values of input variables, parameters and inflow for model simulation of lownutrient, transitional and high-nutrient conditions. Refer to Table 5-1 for complete definitions of variables listed in first column. Variable Low-nutrient Transitional High-nutrient sdTP T 101 339 341 litTF 310 1016 1582 actptTP T 585 1037 1445 staptTP 7 252 252 252 moisture 0.4 0.4 0.4 floodL 1.0 0.64 0.32 floodAP 0.32 0.32 0.32 floodSP 0.32 0.32 0.32 sdLCI 0.257 0.257 0.257 litLCI 0.599 0.599 0.599 actpLCI 0.727 0.727 0.727 staptLCI 0.814 0.814 0.814 remSD 0.80 0.80 0.80 remLIT 0.50 0.50 0.50 remAP 0.25 0.25 0.25 senesce* 411 762 1112 jmgkg1 g C m' 2 y 1

PAGE 159

153 from sites 9 and 10 only. The reasons for this are detailed below. Values for LCI were determined from means of experimentally-determined values for each compartment, and were constant for all levels of nutrient enrichment. The variables moisture, floodL, floodAP and floodSP were initially set to typical values based on experimental results in previous chapters and field observations in WCA-2A. Values of 0.32 were used under all nutrient conditions for active and stable peat compartments, which assumed flooded conditions throughout. The variable floo dL was set to 1 under low-nutrient conditions, representing constant aerobic conditions, 0.64 for transitional conditions and 0.32 for highnutrient conditions. The latter setting assumed constant anaerobic conditions, a conservative interpretation of field results in WCA-2A. Input of organic C via plant senescence was based on data for detritus production in high-nutrient cattail and lownutrient sawgrass stands in WCA-2A (Davis, 1991). The fraction of remaining mass for each compartment was chosen based on first-order decay in a static system, as discussed previously. Values for remSD, remLIT and remAP referred to initial mass in each compartment (Table 5-2). Thus 80, 40 and 10% of the original mass of dead plant tissue entering the system was transferred from the standing dead, litter and active peat compartments, respectively. Selection of values for these parameters was roughly based on relative content of cellulose in the lignocellulose component of standing dead, litter and surface (active) peat (Chapter 3). Sensitivity analysis was performed on the parameters, variables and inflow to determine the relative significance of each to model output (Table 5-3). Variables sdTP, litTP and actptTP, and inflow (senesce) were varied within the ranges shown in Table 5-2. Variable staptTP was varied within the experimentally measured range of 252-968 mg kg" 1 The parameter moisture was varied over the range 0.2-0.6 and floodL, floodAP and floodSP were varied from 0.32-1.0. The variables sdLCI and litLCI were varied 20% and actpLCI and staptLCl were varied to -20%. The parameters remSD, remLIT and remAP

PAGE 160

154 Table 5-3. Sensitivity analysis of model parameters, variables and inflow listed in Table 5-2. Values represent steady-state response of the four state variables relative to change in parameter values. Steady-state response (Output % change/% change in parameter) Parameter stdDead litter activePeat stablePeat sdTP -0.74 0.00 0.00 0.00 litTP 0.00 -2.21 0.00 0.00 actptTP 0.00 0.00 -2.42 0.00 staptTP 0.00 0.00 0.00 -25.58 moisture -3.00 0.00 0.00 0.00 floodL 0.00 -3.01 0.00 0.00 floodAP 0.00 0.00 -3.00 0.00 floodSP 0.00 0.00 0.00 -3.00 sdLCI 0.48 0.00 0.00 0.00 litLCI 0.00 1.61 0.00 0.00 actpLCI 0.00 0.00 2.10 0.00 staptLCI 0.00 0.00 0.00 4.70 remSD -2.40 0.69 0.69 0.69 remLIT 0.00 -2.40 0.69 0.69 remAP 0.00 0.00 -2.40 0.69 senesce 1.00 1.00 1.00 1.00

PAGE 161

155 were varied from 0.25-0.75. Model response was calculated as percent change in steadystate values of the four state variables divided by percent change in parameter values. Results and Discussion Varying settings for parameters and variables used in calculating the mineralization rate constant for a specific compartment (i.e. those associated with total P content, flooding or moisture, and LCI) affected only the steady-state value for the specific state variable, while steady-state values of the remaining state variables remained unchanged (Table 5-3). Increases in moistureand flood-related parameters caused a decrease of constant proportion in steady-state mass of all state variables. These parameters were essentially constant multipliers to attenuate mineralization rate constants. In contrast, changes in the variables sdTP, litTP, actptTP, staptTP, sdLCI, litLCI, actptLCI and staptLCI produced variable response. An increased response (either positive or negative) was observed along the sequence from stdDead to stablePeat, probable a function of increased residence time along the sequence. In particular, model output for stablePeat was relatively sensitive to changes in staptTP. Increasing values of the parameters remSD, remLIT and remAP resulted in decrease in steady-state mass for the corresponding state variable, but an increase of smaller proportion in downstream compartments. This is intuitive, since exporting C at an earlier stage of decomposition results in increased mass transferred to subsequent compartments. The response of the model to changes in mass retention was of the same proportion for all compartments. Changes in inflow (senesce) of C to the system resulted in directly proportional (1:1) changes in steady-state mass for all state variables. The high sensitivity of the model to changes in total P content of the stablePeat compartment revealed an area of uncertainty related to appropriate settings for the variable staptTP. Analytical results, presented in Chapter 3, demonstrated a wide range of total P values for peat in the 10-30 cm depth interval. However, this layer of peat was probably not representative of the deeper, older peat. Field measurements along the WCA-2A

PAGE 162

156 transect indicated that total peat depth was generally between 1 and 2 m, thus the oldest, "stable", peat was not sampled during the study. Furthermore, there is no precedent for estimating total P concentrations in old (e.g. 1000 y) Everglades peat formed under highnutrient conditions, since nutrient enrichment is a very recent phenomenon. It is likely that nutrient availability in older peat would be low regardless of the original composition, because of gradual incorporation of P into refractory compounds formed as microbial byproducts. Settings for total P content (staptTP) were therefore conservative, that is, initially set at the lower end of the range observed in the field. Model simulations were used to evaluate the accuracy of the fundamental equations used by the model, i.e. Equations 5-1 and 5-2. Calculated values of mineralization rate constants (k), were compared with experimentally determined values, as reported in Chapter 3 (Figure 5-3). Settings for total P content and LCI corresponded to actual values measured for sites 1, 4, 6, 9 and 10, for which both types of data are available. The parameters related to flooding were set to either 1 .0 or 0.32, corresponding to aerobic or anaerobic laboratory incubations of plant litter and peat substrate. An approximate 1 : 1 correspondence was shown between calculated (predicted) and observed k values (Figure 5-3). However, there was considerable variability unaccounted for by the model, even though each of the regressions represented in Equations 5-1 and 5-2 explained most of the variability in the dependent variable (rate). Thus, the simulation model reflected the cumulative variability contained in the two regression equations. The detrital C model was also used to simulate system response to a transient state of nutrient enrichment (Figure 5-4). Nutrient enrichment of a low-nutrient system was simulated by grading parameter, variable and input values from low-nutrient to highnutrient settings, as shown in Table 5-2. Steady-state values of the state variables under low-nutrient conditions were used as initial conditions for the 120-year simulation. Parameter, variable and input values were increased at a constant rate during the time span of 30-60 years, from low-nutrient to high-nutrient values. Increased mineralization rates

PAGE 163

157 0 0.001 0.002 0.003 0.004 0.005 OBSERVED k(d" 1 ) SD A L-anaer AP-anaer A SP-anaer • L-aer AP-aer 0 SP-aer

PAGE 164

158 120010002000Active peat t Standing dead Stable peat i i I | I T I 20 40 — I — i — i — i — |— 60 80 1 r 100 •30000 ^ 25000 -20000 15000 120 E U) CO < o I< LU Q. LU -J CD < TIME (y)

PAGE 165

159 associated with P supply were offset by the accelerated input of organic C in the form of live and detrital plant tissue production; as a result, mass of C for all state variables increased. A short time delay occurred in the response between sequential compartments. Although the effects (parameters, variables, input) were increased linearly, all state variable responses were curvilinear, due to interactions among effects. Interactions also resulted in minor fluctuations in mass during the transitional period. The state variables stdDead, litter and activePeat reached a new steady-state soon after the model inputs were stabilized. The stablePeat compartment increased out of proportion to the remaining compartments due to the constant value used for staptTP. Thus, the response of staptTP may be considered an artifact of the choice of parameter settings. The litter compartment also displayed a substantial increase, relative to both stdDead and activePeat. In this case, the response was probably associated with the shift from aerobic to anaerobic conditions (floodL), whereas the aeration and moisture status of standing dead material and peat remained constant over time, as observed under field conditions. In addition to dynamic simulation, sequential simulations were run under low nutrient, transitional and high nutrient conditions (Table 5-2) until steady-state was attained. At that point the rate of change over time was zero in all compartments, and the inflow of detrital C was offset by loss of C from the system via mineralization. Steady-state was reached in about 1000 years under low-nutrient, transitional and high-nutrient conditions. Model output for steady-state conditions is summarized in Table 5-4. Rate of mineralization (k^) increased in standing dead and active peat compartments with increased total P loading. Mineralization rate of litter was greater under high-nutrient than low-nutrient conditions, but less than for transitional conditions, where O, availability was greater. Residence time of the organic substrate, representing throughput among sequential compartments due to litterfall and burial, decreased with increasing P loading. This was a result of increased mineralization rate, thus the specified fraction of original mass was attained in a shorter time period.

PAGE 166

160 Table 5-4. Model output at steady state for simulation of low-nutrient, transitional and high-nutrient conditons in the Everglades WCA-2A marsh. Compartment Residence time Total C mass Net accretion y 1 y gC m z cm T ow-nntricnt conditions J i\J VV 1 1 LI LI Iv 11 L ^UllUlllVllil StanHinp Hpad 0.266 0 84 282 Litter 0.395 1.76 349 Active peat 0.109 12.70 1057 4.6 Stable peat 0.005 16662 42.1 Transitional conditions ^tanHincr HpaH 0 402 0 55 345 Litter 0.616 1.13 414 Active peat 0.192 7.23 1114 4.8 Stable peat 0.005 30863 78.0 High-nutrient conditions Standing dead 0.404 0.55 503 Litter 0.453 1.53 821 Active peat 0.267 5.20 1171 5.1 Stable peat 0.005 45046 113.9

PAGE 167

161 Despite increased mineralization rates, mass accumulation of C was greater with increased P loading. This resulted from the increased inflow of detrital C associated with the overall increase in primary production which accompanies an increase in growthlimiting nutrients. Increase in C mass accumulation was most significant in the stable peat compartment, primarily because P concentration was not increased in that compartment. The equivalent vertical accretion of active and stable peat, assuming bulk density values of 0.05 and 0.086 g cm" 3 and a C content of 46% for peat, was over 1 m under high-nutrient conditions (Table 5-4). Most of the accretion was in the form of stable peat. Conclusions The detrital C model provided a summary of the interactions among substrate composition, environmental factors and organic C turnover. It also provided a means for evaluating the mathematical relationship between the response variable mineralization rate and the factors total P content, lignocellulose index and availability of Q,, (represented by flooding). The relative importance of each factor, previously identified from results of laboratory and field studies, to model output was determined through sensitivity analysis. It was shown that total accumulation of organic C in the four compartments was most strongly affected by total P concentration. This matches the conclusions drawn from the experimental phase of the study, discussed in preceding chapters. The model predicted an increase in net accumulation of organic C with increased P enrichment, based on input values from experimental results in this study and literature values for WCA-2A. However, the model is not suitable, in its present form, as a predictive model for turnover or mass accumulation of C in the Everglades. This is primarily due to lack of understanding of processes in key areas. The greatest degree of uncertainty is associated with the aeration status (floodL) of the litter compartment and the total P content (staptTP) of stablePeat. As shown in Table 5-3, the model is sensitive to both parameters, especially staptTP. Therefore, further elucidation of the actual status of

PAGE 168

162 these parameters is necessary to increase the usefulness of the model for predictive purposes. Nevertheless, the general trend of increased organic C accumulation with increased P enrichment, despite higher net primary productivity, is supported by field estimates of peat accretion in WCA-2A.

PAGE 169

CHAPTER 6 SUMMARY AND CONCLUSIONS Laboratory and field studies were conducted to evaluate the influence of nutrient enrichment on turnover of organic C in standing dead plant material, peat and the overlying litter layer in the Everglades WCA-2A marsh. Specific objectives were presented in Chapter 1 and addressed by studies described in Chapters 2 through 4. A summary of experimental results as they relate to previously stated research objectives is presented below. 1) Determine the effects of nutrient enrichment and flooding on organic C mineralization in the soil (peat and litter layer) profile. Mineralization of organic C in wetland microcosms increased in direct proportion to depth of the water table. The magnitude of the response of C mineralization to water table depth decreased along the nutrient enrichment gradient; i.e., decreased with increasing distance from the surface water inflow. A statistically significant interaction was observed between water table depth and nutrient enrichment as factors affecting C mineralization rate; however, the rate was more significantly affected by water table depth. Potential C mineralization rate for litter and peat, measured during aerobic incubation of substrate samples, was 1.5 to 3 times greater than rates measured in drained microcosms, probably due to an abundance of anaerobic microsites in the peat. Potential C mineralization rate was significantly correlated with substrate total P concentration. 2) Determine the effect of nutrient enrichment on turnover of organic C pools along the WCA-2A nutrient gradient, and examine the relationships 163

PAGE 170

164 between size of the microbial biomass C pools and turnover time of associated organic C pools. Total P content and relative lignin enrichment (expressed as LCI) accounted for 91% of the variability associated with substrate biodegradability (potential respiration rate) across the decay continuum. Anaerobic conditions caused a proportional decrease in decomposition rate (increased turnover time) in the litter, surface peat (0-10 cm) and buried peat (10-30 cm) layers of the soil. Anaerobic decomposition rate was approximately onethird the rate of aerobic decomposition. The plant standing dead component and soil litter layer are potentially the most active (rapid turnover) organic C pools, as reflected by their low to moderate values of LCI. High cellulose content and low nutrient (N and P) concentrations indicate that C availability does not limit decomposition of standing dead. As the dead leaf tissue ages on the plant, nutrient availability becomes a major limiting factor for microbial activity. Another probable limitation may be moisture availability, during drought conditions. Carbon availability in the soil litter layer appears to be significantly greater than in peat, probably due to the presence of available cellulose. Within this layer, the major determinants of decomposition rate are nutrient availability (P in this case) and 0 ; availability. Total P concentration exerted a strong influence on decomposition of litter. It is likely that C availability only becomes a limiting factor under conditions of high nutrient availability. Total P content exerted less influence on decomposition rate as the depth and age of the substrate increased. It is evident that C availability becomes increasingly limiting to microbial growth through the peat profile. High LCI values in the buried peat suggest that the more readily decomposable cellulose component is highly depleted, the remainder being physically protected from microbial activity by the lignified matrix of the substrate. The influence of nutrient availability on decomposition diminishes as the proportion of refractory compounds in the substrate increases (with age and depth).

PAGE 171

165 The standing dead pool has a high potential for rapid turnover of nutrients, due to its high decomposition rate and leaching of mineralized nutrients. The soil litter layer is potentially very important in short-term nutrient cycling. Due to the relatively high C availability and resulting high population of microbial decomposers, the litter layer may serve as either a source or sink for nutrients. Released nutrients may recycled through plant uptake, microbial re-immobilization or chemical adsorption or precipitation. Alternatively, they may be exported, along with dissolved organic C, to a downstream area. Turnover of peat, especially peat buried deep in the soil profile, is highly restricted under anaerobic conditions, therefore this pool of organic matter represents a potentially "permanent" sink for nutrients and contaminants. Microbial biomass C was highly correlated with both aerobic and anaerobic decomposition of organic C throughout the decay continuum. In particular, the ratio of microbial C to total soil C explained 88% of the variability in potential (aerobic) decomposition. Results of this study show microbial biomass to be a promising indicator of organic matter turnover in wetland soil. Substrate composition, related to nutrient and C availability, were also shown to be accurate predictors of organic C turnover. However, the interactions among these factors, along with environmental effects such as 0 2 availability, must be determined beforehand. Further study is needed to develop reliable biological and chemical indicators of organic C turnover in wetlands. 3) Determine the effect of nutrient enrichment on in situ decomposition rate along a vertical profile in the water column and peat, specifically the significance of various environmental and substrate-related factors on decomposition rate. Decomposition of site-native plant material increased closer to the inflow, indicating that in situ decomposition in the litter layer increased with nutrient enrichment in WCA-2A. Results also suggested that nutrient enrichment directly or indirectly (via substrate quality

PAGE 172

166 or environmental factors) affected decomposition potential in the peat profile. Decomposition of uniform substrate along the vertical profile exhibited a general decrease from floodwater/litter layer to peat, reflecting the influence of environmental factors, including nutrient and 0 2 availability. The in situ decomposition study represented the early stages of the decay continuum from plant tissue to peat, and initial substrate quality exerted an effect on decomposition rate. Decomposition rate of the dead cattail and sawgrass leaves, representing newlydeposited plant material to the litter layer, was therefore governed by the interactive effects of initial chemical composition of the dead material and environmental factors. Variability in cotton strip decay reflected only environmental factors, thus the cotton strip assay may have been a more sensitive indicator of the "decomposition potential" at a given location or depth in the floodwater-peat profile. Although CRR was not closely correlated with dissolved nutrient concentration, there was strong evidence, based on previous total N and P analysis of litter and peat, that both N and P supply from the surrounding matrix were controlling factors for decomposition. Substantial immobilization of N and P occurred in the decomposing plant material, especially within the litter layer and near the inflow. The high immobilization potential of the standing dead material was due to a combination of extremely low nutrient content and relatively high C quality. This shows that the newly-deposited litter in nutrient-enriched areas can serve as a strong sink for nutrients, with relatively rapid turnover. This carries significant implications regarding the importance of the litter layer in short-term nutrient cycling in wetlands. The effects of nutrient availability on decomposition in wetlands has received far less attention than in terrestrial ecosystems. Nevertheless, understanding the impact of nutrient loading on decomposition has major implications with regard to ecosystem stability in natural wetlands and long-term efficiency of wastewater-treatment wetlands.

PAGE 173

LIST OF REFERENCES Aber, J. D., J. M. Melillo, and McClaugherty. 1990. Predicting long-term patterns of mass loss, nitrogen dynamics, and soil organic matter formation from initial fine litter chemistry in temperate forest ecosystems. Can. J. Bot. 68:2201-2208. Amador, J. A., and R. D. Jones. 1993. Nutrient limitations on microbial respiration in peat soils with different total phosphorus content. Soil Biol. Biochem. 25:793-801. Amador, J. A., and R. D. Jones. 1995. Carbon mineralization in the pristine and phosphorus-enriched peat soils of the Florida Everglades. Soil Sci. 159:129-141. American Association of Analytical Chemists (AOAC). 1990. Official methods of analysis. W. Horwitz (ed.). American Association of Analytical Chemists, Washington, DC. Anderson, J. M. 1976. An ignition method for determination of total phosphorus in lake sediments. Water Res. 10:329-331. Anderson, T.-H., and K. H. Domsch. 1986. Carbon assimilation and microbial activity in soil. Z. Pflanzenernaehr. Bodenk. 149:457-468. Anderson, T. H., and K. H. Domsch. 1989. Ratios of microbial biomass carbon to total organic carbon in arable soils. Soil Biol. Biochem. 21:471-479. Anderson, T. H., and K. H. Domsch. 1990. Application of eco-physiological quotients (<3C0 2 and qD) on microbial biomasses from soils of different cropping histories. Soil Biol. Biochem. 22:251-255. Anderson, T.-H., and K. H. Domsch. 1993. The metabolic quotient for CO2 (4CO2) as a specific activity parameter to assess the effects of environmental conditions, such as pH, on the microbial biomass of forest soils. Soil Biol. Biochem. 25:393-395. Andren, O., and K. Paustian. 1987. Barley straw decomposition in the field: a comparison of models. Ecology 68:1 190-1200. Atlas, R. M. 1986. Applicability of general ecological principles to microbial ecology, p. 339-370. In Poindexter, J. S., and E. R. Leadbetter (ed.) Bacteria in nature, vol. 2. Plenum, New York. Bayley, S., and H. T. Odum. 1976. Simulation of interrelations of the Everglade's marsh peat, fire, water and phosphorus. Ecol. Modelling 2: 169-188. Belanger, T. V., D. J. Scheldt, and J. R. Platko, II. 1989. Effects of nutrient enrichment on the Florida Everglades. Lake and Reservoir Management 5: 101-1 1 1. 167

PAGE 174

168 Benner, R., A. E. Maccubbin, and R. E. Hodson. 1984a. Anaerobic biodegradation of the lignin and polysaccharide components of lignocellulose and synthetic lignin by sediment microflora. Appl. Environ. Microbiol. 47:998-1004. Benner, R., S. Y. Newell, A. E. Maccubbin, and R. E. Hodson. 1984b. Relative contributions of bacteria and fungi to rates of degradation of lignocellulosic detritus in salt-marsh sediments. Appl. Environ. Microbiol. 48:36-40. Berg, B., and H. Staaf. 1981. Leaching, accumulation and release of nitrogen in decomposing forest litter, p. 163-178. In Clark, F. E., and T. Rosswall (ed.) Ecol. Bull. Vol. 33. Bohlin, E., M. Hamalainen, and T. Sunden. 1989. Botanical and chemical characterization of peat using mutivariate methods. Soil Sci. 147:252-263. Borga, P., M. Nilsson, and A. Tunlid. 1994. Bacterial communities in peat in relation to botanical composition as revealed by phospholipid fatty acid analysis. Soil Biol. Biochem. 26:841-848. Bosatta, E., and G. I. Agren. 1985. Theoretical analysis of decomposition of heterogeneous substrates. Soil Biol. Biochem. 17:601-610. Bremner, J. M. and C. S. Mulvaney, 1982. Nitrogen-total, p. 595-624. In A. L. Page (ed.) Methods of soil analysis. Part 2. SSSA, ASA, Madison, WI. Bridgham, S. D., C. J. Richardson, E. Maltby, and S. P. Faulkner. 1991. Cellulose decay in natural and disturbed peatlands in North Carolina. J. Environ. Qual. 20:695-701. Bridgham, S. D., and C. J. Richardson. 1992. Mechanisms controlling soil respiration (CO2 and CH4) in southern peatlands. Soil Biol. Biochem. 24:1089-1099. Browder, J. A., and B. G. Volk. 1978. Systems model of carbon transformations in soil subsidence. Ecol. Modelling 5:269-292. Brown, A., S. P. Mathur, T. Kauri, and D. J. Kushner. 1988. Measurement and significance of cellulose in peat soils. Can. J. Soil Sci. 68:681-685. Clymo, R. S. 1983. Peat. p. 159-224. In Gore, A. J. P. (ed.) Mires: swamp, bog, fen and moor. Elsevier, Amsterdam. Colberg, P. J. 1988. Anaerobic microbial degradation of cellulose, lignin, oligolignols, and monoaromatic lignin derivatives, p. In Zehnder, A. J. B. (ed.) Biology of anaerobic microorganisms. John Wiley & Sons, New York. Cook, B. D., and D. L. Allan. 1992a. Dissolved organic carbon in old field soils: compositional changes during the biodegradation of soil organic matter Soil Biol Biochem. 24:595-600. Cook, B. D., and D. L. Allan. 1992b. Dissolved organic carbon in old field soils: total amounts as a measure of available resources for soil mineralization Soil Biol Biochem. 24:585-594.

PAGE 175

169 Craft, C. B., and C. J. Richardson. 1993. Peat accretion and phosphorus accumulation along a eutrophication gradient in the northern Everglades. Biogeochemistry 22:133-156. Davis, J. H., Jr. 1943. The natural features of southern Florida. Florida Geological Survey Bull. 25 Davis, S. M. 1991. Growth, decomposition, and nutrient retention of Cladium jamaicense Crantz and Typha domingensis Pers. in the Florida Everglades. Aquat. Bot. 40:203-224. DeBusk, W. F., and K. R. Reddy. 1987. Removal of floodwater nitrogen in a cypress swamp receiving primary wastewater effluent. Hydrobiologia 153:79-86. DeBusk, W. F., K. R. Reddy, M. S. Koch, and Y. Wang. 1994. Spatial distribution of soil nutrients in a northern Everglades marsh: Water Conservation Area 2A. Soil Sci. Soc. Am. J. 58:543-552. Elliott, E. T., D. C. Coleman, R. E. Ingham, and J. A. Trofymow. 1984. Carbon and energy flow through microflora and microfauna in the soil subsystem of terrestrial ecosystems, p. 424-433. In Klug, M. J., and C. A. Reddy (ed.) Current perspectives in microbial ecology. American Society for Microbiology, Washington. Eriksson, K. E., and S. C. Johnsrud. 1982. Mineralisation of carbon, p. 134-153. In Burns, R. G., and J. H. Slater (ed.) Experimental microbial ecology. Blackwell Scientific, Oxford. Given, P. H., and C. H. Dickinson. 1975. Biochemistry and microbiology of peats, p. 123-212. In Paul, E. A., and A. D. McLaren (ed.) Soil Biochemistry Vol. 3. Marcel Dekker, New York. Gleason, P. J., A. D. Cohen, H. K. Brooks, P. Stone, R. Goodrick, W. G. Smith, and W. Spackman, Jr. 1974. The environmental significance of Holocene sediments from the Everglades and saline tidal plain, p. 287-341. In Gleason, P. J. (ed.) Environments of south Florida: present and past. Miami Geological Society, Miami, FL. Godshalk, G. L., and R. G. Wetzel. 1978. Decomposition of aquatic angiosperms. II. Particulate components. Aquat. Bot. 5:301-327. Grant, R. F., N. G. Juma, and W. B. McGill. 1993a. Simulation of carbon and nitrogen transformations in soil: mineralization. Soil Biol. Biochem. 25:1317-1329. Grant, R. F., N. G. Juma, and W. B. McGill. 1993b. Simulation of carbon and nitrogen transformations in soil: microbial biomass and metabolic products. Soil Biol. Biochem 25:1331-1338. Happell, J. D., and J. P. Chanton. 1993. Carbon remineralization in a north Florida swamp forest: effects of water level on the pathways and rates of soil organic matter decomposition. Global Biogeochemical Cycles 7:475-490. Harrison, A. F., P. M. Latter, and D. W. H. Walton. 1988. Cotton strip assay: an index of decomposition in soils. ITE Symposium no. 24. Institute of Terrestrial Ecology. Grange-Over-Sands, UK.

PAGE 176

170 Heal, O. W., and D. D. French. 1974. Decomposition of organic matter in tundra, p. 279-309. In Holding, A. J., O. W. Heal, S. F. MacLean, Jr., and P. W. Flanagan (ed.) Soil organisms and decomposition in tundra. Tundra Biome Steering Committee, Stockholm. Heal, O. W., P. W. Flanagan, D. D. French, and S. F. MacLean, Jr. 1981. Decomposition and accumulation of organic matter, p. 587-633. In Bliss, L. C, O. W. Heal, and J. J. Moore (ed.) Tundra ecosystems: a comparative analysis. Cambridge University Press, Cambridge. Heal, O. W., and P. Ineson. 1984. Carbon and energy flow in terrestrial ecosystems: relevance to microflora, p. 394-404. In Klug, M. J., and C. A. Reddy (ed.) Current perspectives in microbial ecology. American Society for Microbiology, Washington. Hesslein, R. H. 1976. An in situ sampler for close interval pore water studies. Limnol. Oceanogr. 21:912-914. Hill, M. O., P. M. Latter, and G. Bancroft. 1985. A standard curve for inter-site comparison of cellulose degradation using the cotton strip method. Can. J. Soil Sci. 65:609-619. Horwath, W. R., and E. A. Paul. 1994. Microbial biomass. p. 753-773. In R. W. Weaver et al. (ed.) Methods of soil analysis, Part 2. Microbiological and biochemical properties. Soil Science Society of America, Madison, WI. HowardWilliams, C. 1985. Cycling and retention of nitrogen and phosphorus in wetlands: a theoretical and applied perspective. Freshwater Biology 15:391-431. Howarth, R. W. 1993. Microbial processes in salt-marsh sediments, p. 239-259. In Ford, T. E. (ed.) Aquatic microbiology. Blackwell Scientific Publications, Howeler, R. H., and D. R. Bouldin. 1971. The diffusion and consumption of oxygen in submerged soils. Soil Sci. Soc. Am. Proc. 35:202-208. Howson, G. 1991. The cotton strip assay: field applications and global comparisons, p. 217-228. In Wilson, W. S. (ed.) Advances in soil organic matter research: the impact on agriculture and the environment. Royal Society of Chemistry, Cambridge, UK. Insam, H., and K. Haselwandter. 1989. Metabolic quotient of the soil microflora in relation to plant succession. Oecologia 79: 174-178. Jenkinson, D. S. 1990. The turnover of organic carbon and nitrogen in soil. Phil. Trans. R. Soc. Lond. B 329:361-368. Jenkinson, D. S., and J. H. Rayner. 1977. The turnover of soil organic matter in some of the Rothamsted classical experiments. Soil Sci. 123:298-305. Jenkinson, D. S., and J. N. Ladd. 1981. Microbial biomass in soil: measurement and turnover, p. 415-471. In Paul, E. A., and J. N. Ladd (ed.) Soil biochemistry, Vol. 5. Marcel Dekker, New York.

PAGE 177

171 Kelly, C. A., and D. P. Chynoweth. 1979. Methanogenesis: a measure of chemoorganotrophic (heterotrophic) activity in anaerobic lake sediments, p. 164-179. In Costerton, J. W., and R. R. Colwell (ed.) Native aquatic bacteria: enumeration, activity and ecology, ASTM STP 695. American Society for Testing and Materials. Koch, M. S., and K. R. Reddy. 1992. Distribution of soil and plant nutrients along a trophic gradient in the Florida Everglades. Soil Sci. Soc. Am. J. 56:1492-1499. Lahde, E. 1966. Vertical distribution of biological activity in peat of some virgin and drained swamp types. Acta Forest. Fenn. 81:1-15. Latter, P. M., and G. Howson. 1977. The use of cotton strips to indicate cellulose decomposition in the field. Pedobiologia 17:145-155. Linkins, A. E., R. L. Sinsabaugh, C. A. McClaugherty, and J. M. Melills. 1990. Cellulase activity on decomposing leaf litter in microcosms. Plant and Soil 123: 17-25. Lodge, D. J., W. H. McDowell, and C. P. McSwiney. 1994. The importance of nutrient pulses in tropical forests. Tree 9:384-387. Loveless, C. M. 1959. A study of the vegetation in the Florida Everglades. Ecology 40:1-9. Maltby, E. 1988. Use of cotton strip assay in wetland and upland environments an international perspective, p. 140-154. In Harrison, A. F., P. M. Latter, and D. W. H. Walton (ed.) Cotton strip assay: an index of decomposition in soils. ITE Symposium no. 24. Institute of Terrestrial Ecology. Grange-Over-Sands, UK. Martens, R. 1987. Estimation of microbial biomass in soil by the respiration method: importance of soil pH and flushing methods for the measurement of respired CO2. Soil Biol. Biochem. 19:77-81. Melillo, J. M., J. D. Aber, A. E. Linkins, A. Ricca, B. Fry, and K. J. Nadelhoffer. 1989. Carbon and nitrogen dynamics along the decay continuum: plant litter to soil organic matter. Plant and Soil 115:189-198. Melillo, J. M., J. D. Aber, and J. F. Muratore. 1982. Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology 63:621-626. Melillo, J. M., R. J. Naiman, J. D. Aber, and A. E. Linkins. 1984. Factors controlling mass loss and nitrogen dynamics of plant litter decaying in northern streams. Bull. Mar. Sci. 35:341-356. Minderman, G. 1968. Addition, decomposition and accumulation of organic matter in forests. J. Ecol. 56:355-362. Moore, T. R., and M. Dalva. 1993. The influence of temperature and water table position on carbon dioxide and methane emissions from laboratory columns of peatland soils J Soil Sci. 44:651-654. Moran, M. A., R. Benner, and R. E. Hodson. 1989. Kinetics of microbial degradation of vascular plant material in two wetland ecosystems. Oecologia 79:158-167.

PAGE 178

172 Ocio, J. A., and P. C. Brookes. 1990. Soil microbial biomass measurements in sieved and unsieved soil. Soil Biol. Biochem. 22:999-1000. Odum, E. P. 1969. The strategy of ecosystem development. Science 164:262-270. Ohtonen, R. 1994. Accumulation of organic matter along a pollution gradient: application of Odum's theory of ecosystem energetics. Microb. Ecol. 27:43-55. Ojima, D. S., D. S. Schimel, W. J. Parton, and C. E. Owensby. 1994. Longand short-term effects of fire on nitrogen cycling in tallgrass prairie. Biogeochemistry 24:67-84. Oremland, R. S. 1988. Biogeochemistry of methanogenic bacteria, p. 641-705. In Zehnder, A. J. B. (ed.) Biology of anaerobic microorganisms. John Wiley & Sons, New York. Parker, G. G. 1974. Hydrology of the pre-drainage system of the Everglades in southern Florida, p. 18-27. In Gleason, P. J. (ed.) Environments of south Florida: present and past. Miami Geological Society, Miami, FL. Parton, W. J., D. S. Schimel, C. V. Cole, and D. S. Ojima. 1987. Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Sci. Soc. Am. J. 51:1173-1179. Paul, E. A. 1984. Dynamics of organic matter in soils. Plant and Soil 76:275-285. Paul, E. A., and J. A. van Veen. 1978. The use of tracers to determine the dynamic nature of organic matter, p. 61-102. In (ed.) 1 1th Congress International Society of Soil Science, Edmonton. Paul, E. A., and R. P. Voroney. 1980. Nutrient and energy flows through soil microbial biomass. p. 215-237. In Ellwood, D. C, J. N. Hedger, M. J. Latham, J. M. Lynch, and J. H. Slater (ed.) Contemporary microbial ecology. Academic Press, London. Ponnamperuma, F. N. 1972. The chemistry of submerged soils. Advances in Agronomv Vol. 24:29-96. y Quails, R. G., and C. J. Richardson. 1995. Forms of soil phosphorus along a nutrient enrichment gradient in the northern Everglades. Soil Sci. 160:183-198. Reddy, K. R., and W. F. DeBusk. 1991. Decomposition of water hyacinth detritus in eutrophic lake water. Hydrobiol. 21 1:101-109. Reddy, K. R., R. D. DeLaune, W. F. DeBusk, and M. S. Koch. 1993. Long-term nutrient accumulation rates in the Everglades. Soil Sci. Soc. Am. J. 57: 1 147-1 155. Reddy, K. R., and E. M. DAngelo. 1994. Soil processes regulating water quality in wetlands, p. 309-324. In Mitsch, W. J. (ed.) Global wetlands: old world and new. Elsevier Science, Amsterdam. Sagar, B. F. 1988a. The Shirley Soil Burial Test Fabric and tensile testing as a measure of biological breakdown of textiles, p. 1 1-16. In Harrison, A. F., P. M. Latter, and D. W. H. Walton (ed.) Cotton strip assay: an index of decomposition in soils. ITE Symposium no. 24. Institute of Terrestrial Ecology. Grange-Over-Sands, UK

PAGE 179

173 Sagar, B. F. 1988b. Microbial cellulases and their action on cotton fibres, p. 17-20. In Harrison, A. F., P. M. Latter, and D. W. H. Walton (ed.) Cotton strip assay: an index of decomposition in soils. ITE Symposium no. 24. Institute of Terrestrial Ecology. Grange-Over-Sands, UK. Saito, M., H. Wada, and Y. Takai. 1990. Development of a microbial community on cellulose buried in waterlogged soil. Biol. Fert. Soils 9:301-305. Schipper, L. A., and K. R. Reddy. 1994. Methane production and emissions from four reclaimed and pristine wetlands of the southeastern United States. Soil Sci. Soc. Am. J. 58:1270-1275. Schipper, L. A., and K. R. Reddy. 1995. In situ determination of detrital breakdown in wetland soil-floodwater profile. Soil Sci. Soc. Am. J. 59:565-568. Schipper, L. A., and K. R. Reddy. 1996. Determination of methane oxidation in the rhizosphere of Sagittaria lancifolia using methyl fluoride. Soil Sci. Soc. Am. J. 60:611-616. Schutz, H., P. Schroder, and H. Rennenberg. 1991. Role of plants in regulating the methane flux to the atmosphere, p. 29-63. In Sharkey, T. D., E. A. Holland, and H. A. Mooney (ed.) Trace gas emissions by plants. Academic, San Diego. Sinsabaugh, R. L., R. K. Antibus, and A. E. Linkins. 1991. An enzymic approach to the analysis of microbial activity during plant litter decomposition. Agric. Ecosystems Environ. 34:43-54. South Florida Water Management District (SFWMD). 1992. Surface water improvement and management plan for the Everglades: Supporting information document. SFWMD, West Palm Beach. South Florida Water Management District (SFWMD). 1996. Hydrometeorological database (DBHYDRO). SFWMD, West Palm Beach. Sparling, G. P. 1992. Ratio of microbial biomass carbon to soil organic carbon as a sensitive indicator of changes in soil organic matter. Aust. J. Soil Res. 30: 195-207. Sparling, G. P., C. W. Feltham, J. Reynolds, A. W. West, and P. Singleton. 1990. Estimation of soil microbial C by a fumigation-extraction method: use on soils of high organic matter content, and a reassessment of the kg^-factor. Soil Biol. Biochem. 22:301-307. Steward, K. K., and W. H. Ornes. 1983. Mineral nutrition of sawgrass {Cladium jamaicense Crantz) in relation to nutrient supply. Aquat. Bot. 16:349-359. Swift and Nicholas, 1987. Periphyton and water quality relationships in the Everglades Water Conservation Areas. Technical Publication 87-2. South Florida Water Management District. West Palm Beach. Swift, M. J. 1976. Species diversity and the structure of microbial communities in terrestrial habitats, p. 185-222. In Anderson, J. M., and A. Macfadyen (ed.) The role of terrestrial and aquatic organisms in decomposition processes. Blackwell Scientific Publications, Oxford.

PAGE 180

174 Swift, M. J. 1982. Microbial succession during the decomposition of organic matter, p. 164-177. In Burns, R. G., and J. H. Slater (ed.) Experimental Microbial Ecology. Blackwell Scientific, Oxford. Swift, M. J. 1984. Microbial diversity and decomposer niches, p. 8-16. In Klug, M. J., and C. A. Reddy (ed.) Current perspectives in microbial ecology. Amer. Soc. for Microbiology, Washington, DC. Swift, M. J., O. W. Heal, and J. M. Anderson. 1979. Decomposition in terrestrial ecosystems. Univ. of California Press, Berkeley. Tate, HI, R. L. 1979. Effect of flooding on microbial activities in organic soils: carbon metabolism. Soil Sci. 128:267-273. Toth, L. A. 1987. Effects of hydrologic regimes on lifetime production and nutrient dynamics of sawgrass. Technical Publication #87-6. South Florida Water Management District, West Palm Beach, FL. Toth, L. A. 1988. Cattail nutrient dynamics. Technical Publication #88-6. South Florida Water Management District, West Palm Beach, FL. U. S. Environmental Protection Agency (USEPA). 1983. Methods for chemical analysis of water and wastes. Environ. Monit. Support Lab., Cincinnati, OH. Van Veen, J. A., and E. A. Paul. 1981. Organic carbon dynamics in grassland soils. I. Background information and computer simulation. Can. J. Soil Sci. 61:185-201. Van Veen, J. A., J. N. Ladd, and M. J. Frissel. 1984. Modelling C and N turnover through the microbial biomass in soil. Plant and Soil 76:257-274. Volk, B. G. 1973. Everglades histosol subsidence: 1. C02 evolution as affected by soil type, temperature and moisture. Soil and Crop Science Society of Florida Proc. 32:132-135. Wardle, D. A. 1993. Changes in the microbial biomass and metabolic quotient during leaf litter succession in some New Zealand forest and scrubland ecosystems. Functional Ecology 7:346-355. Webster, J. R., and E. F. Benfield. 1986. Vascular plant breakdown in freshwater ecosystems. Ann. Rev. Ecol. Syst. 17:567-594. Westermann, P. 1993. Wetland and swamp microbiology, p. 215-238. In Ford, T. E. (ed.) Aquatic microbiology. Blackwell Scientific Publications, Wetzel, R. G. 1984. Detrital dissolved and particulate organic carbon functions in aquatic ecosystems. Bulletin of Marine Science 35:503-509. Wetzel, R. G. 1992. Gradient-dominated ecosystems: sources and regulatory functions of dissolved organic matter in freshwater ecosystems. Hydrobiol. 229:181-198. Whiting, G. J. 1994. CO : exchange in the Hudson Bay lowlands: community characteristics and multispectral reflectance properties. J. Geophvs. Res 99:1519-1528.

PAGE 181

175 Zeikus, J. G. 1981. Lignin metabolism and the carbon cycle, p. 21 1-243. In Alexander, M. (ed.) Advances in microbial ecology, Vol. 5. Plenum Press, New York. Zibilske, L. M. 1994. Carbon mineralization, p. 835-863. In Weaver. R. W. et al. (ed.) Methods of soil analysis, part 2. Microbiological and biochemical properties. Soil Science Society of America, Madison, WI.

PAGE 182

BIOGRAPHICAL SKETCH Bill DeBusk was born in Pensacola, Florida, where he lived until the age of 13. He moved with his family to Gainesville, which has remained his home base, though interrupted by stays in Melbourne Beach, Homestead and the Orlando area. Bill received BS and MS degrees from the University of Florida in botany and environmental science, respectively. His wife Patty, who has kindly supported him for the past four years, hails from Piano, Illinois, home of the famous Piano tackle box. 176

PAGE 183

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. R. Reddy, Chair Graduate Research Professor of Soil and Water Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosopt xef, Cochair ^of Agricultural and Biological Engineering I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, asa_ dissertation for the degree of Doctor of Philosophy. G. Ronnie Best Branch Chief, Southern Science Center National Biological Service I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. \aa. v*> s >, @ B. L. McNeal Professor of Soil and Water Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. P.l/S. C. Rao Graduate Research Professor of Soil and Water Science This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. /) August, 1996 y Dean, College of Agriculture Dean, Graduate School


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EJAARXOVF_ZF9VCH INGEST_TIME 2014-12-11T22:20:04Z PACKAGE AA00026487_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES