|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 INDICATORS OF REDUCTION IN SOIL (IRIS) IN VARIOUS CONDITIONS OF SATURATION IN SOUTH FLORIDA MARL BY CHRISTINE A. COFFIN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012
2 2012 Christine A. Coffin
3 To my husband for his support patience and love
4 ACKNOWLEDGMENTS University of Florida Advisory Committee members, Drs. Yuncong Li (advis or), Katie Migliaccio and Ed Hanlon, provided the guidance and focus to produce some meaningful results to further the application of soil science. I am grateful to my advisor, who saw the potential in this project and provided the resources necessary duri ng fiscally challenging economic times. Education Center, as well as Pam Moon for her generosity, and several graduate students who took the time to share their knowledge. I would like to recognize U nited S tates D epartment of A griculture N atural R esources C onservation S ervice staff and associates, Cynthia Stiles, Patty Jones, Edwin Dunkinson and Steve Monteith for providing assistance with Indicators of Reductio n in Soil ( IRIS ) tubes used in this research, and hope these results can assist in evaluating and improving IRIS technology. Lastly, I deeply appreciate friends and family who have provided encouragement and tolerated my absence in their life for many year s.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 7 LIST OF FIGURES ................................ ................................ ................................ ......................... 8 LIST OF ABBREVIATIONS ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 11 CHAPTER 1 INTRODUCTION AND LITERATURE RE VIEW ................................ .............................. 13 Rationale ................................ ................................ ................................ ................................ 13 Research Goals ................................ ................................ ................................ ....................... 14 Application ................................ ................................ ................................ ............................. 15 Marl Soil Formation and Geographic Extent ................................ ................................ .......... 15 South Florida Hydrologic Alterations ................................ ................................ .................... 17 Hydric Soils Classification ................................ ................................ ................................ ..... 18 Criteria and Field Indicators of Hydric Soils ................................ ................................ .......... 19 Evaluating the Hydric Status of Marl Soils ................................ ................................ ............ 21 Morphologic Indicators of Soil Saturation ................................ ................................ ............. 22 Reduction in Saturated Soil Conditions ................................ ................................ .................. 23 Measurement of Oxidation Reduction Potentials in Soils ................................ ...................... 23 Iron Indicators of Reduction in Soil (FIRIS) ................................ ................................ .......... 24 Manganese Indicators of Reduction in Soil (MIRIS) ................................ ............................. 25 Carbonates in South Florida Marl Soils ................................ ................................ .................. 27 Research Objectives ................................ ................................ ................................ ................ 28 2 ANAEROBIC SOIL DETER MINATIONS ................................ ................................ ........... 35 Introduction ................................ ................................ ................................ ............................. 35 Materials and Methods ................................ ................................ ................................ ........... 36 Selection and Extraction of Soil Mesocosms ................................ ................................ .. 36 Installation of Equipment for Monitoring Reduction ................................ ...................... 37 Platinum electrodes ................................ ................................ ................................ .. 37 Salt bridges ................................ ................................ ................................ ............... 38 Indicators of Reduction in Soil (IRIS) ................................ ................................ ..... 38 Treatments ................................ ................................ ................................ ....................... 39 Soil Properties and Sampling ................................ ................................ .......................... 40 Organic matter me asurement by Loss of Ignition (LOI) ................................ .......... 41 Carbonate composition ................................ ................................ ............................. 41 Soil bulk density (Db) ................................ ................................ .............................. 41
6 Soil particle size ................................ ................................ ................................ ....... 42 Iron, manganese and calcium content of soils ................................ .......................... 42 Cation exchange capacity (CEC) ................................ ................................ ............. 43 Temperature, pH, and electrical conductivity ................................ .......................... 43 Methods of Analyzing Oxide Removal from IRIS Tubes ................................ ............... 44 Visual method ................................ ................................ ................................ .......... 45 Trace method ................................ ................................ ................................ ............ 45 Scan method ................................ ................................ ................................ ............. 46 Results ................................ ................................ ................................ ................................ ..... 46 Properties of Selected Marl Soils ................................ ................................ .................... 46 Water Table Treatment Effects on Electro chemical Soil P roperties ............................. 48 Anaerobic Conditions as Determined by Redox Potential ................................ .............. 49 Determination of Anaerobic Conditions with FIRIS ................................ ....................... 53 Determination of Anaerobic Conditions with MIRIS ................................ ..................... 54 Comparison of IRIS Analysis Methods ................................ ................................ ........... 56 Discussion ................................ ................................ ................................ ............................... 57 3 SUMMARY AND CONCLUSI ONS ................................ ................................ ..................... 83 Introduction ................................ ................................ ................................ ............................. 83 Objective 1 ................................ ................................ ................................ .............................. 83 Objective 2 ................................ ................................ ................................ .............................. 84 Objective 3 ................................ ................................ ................................ .............................. 85 APPENDIX A HYDRIC SOILS TECHNIC AL STANDARD ................................ ................................ ...... 87 Hydric Soil Testing Methods ................................ ................................ ................................ .. 87 Collection of Oxidation Reduction Readings with Pt El ectrodes ................................ .......... 88 B CONSTRUCTION OF EQUI PMENT ................................ ................................ ................... 90 Introduction ................................ ................................ ................................ ............................. 9 0 Plati num Electrodes ................................ ................................ ................................ ................ 90 Salt Bridges ................................ ................................ ................................ ............................. 91 Soil Columns ................................ ................................ ................................ .......................... 92 Manganese Oxid e Paint ................................ ................................ ................................ .......... 93 IRIS Rotation Device for Digital Scanning ................................ ................................ ............ 96 C DETAILED SOIL DESCRI PTIONS ................................ ................................ ..................... 98 LIST OF REFERENCES ................................ ................................ ................................ ............. 100 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 104
7 LIST OF TABLES Table page 1 1 Extent of marl soils in Florida ................................ ................................ ........................... 29 1 2 Criteria for hydric soils ................................ ................................ ................................ ...... 30 1 3 Oxidants in soil redox processes, listed in order of preference ................................ ......... 30 1 4 Characteristics of soil carbonates ................................ ................................ ....................... 31 2 1 Characteristics of three marl soils selected for this s tudy ................................ .................. 60 2 2 Correlation coefficients among water table, time, insects and soil properties ................... 61 2 3 Anaerobic conditions determined by platinum electrodes, iron Indicators of Reduction in Soil (FIRIS) and manganese Indicators of Reduction in Soil (MIRIS) ....... 61 2 4 Analysis of variance ................................ ................................ ................................ ........... 62 2 5 Correlation coefficients among various methods of determining redox potential ............. 63 2 6 Comparison of results using different techniques to estimate oxide removal f rom FIRIS and MIRIS tubes ................................ ................................ ................................ ..... 64
8 LIST OF FIGURES Figure page 1 1 Areas mapped as marl in south Florida ................................ ................................ .............. 32 1 2 Major Land Resource Region U in Florida ................................ ................................ ........ 33 1 3 Standard redox (Eh)/pH line for determining anaerobic conditions ................................ .. 34 2 1 Map of soil collection sites in southeastern Miami Dade County, Florida. ...................... 65 2 2 Schematic of redox equipment as installed in soil mesocosms. ................................ ........ 66 2 3 Wireworm excavation in Soil B mesocosm ................................ ................................ ....... 67 2 4 Changes in electrical conductivity associated with water table treatments. ...................... 68 2 5 Average pH response to water table treatments in three marl soils. ................................ .. 68 2 6 Corrected Eh measured by platinum and reference electrodes ................................ .......... 69 2 7 Average Eh for each marl soil in varying water table treatment depths ............................ 70 2 8 Distribution of Eh values above and below the standard redox line. ................................ 71 2 9 Temporal relationship between Eh and degree of saturation in each soil type ................. 72 2 10 Iron oxide removal from Indicators of R eduction in Soil (IRIS) ................................ ....... 73 2 11 The measured response of redox tools to water table treatments with time ...................... 74 2 12 Manganese oxide removal from IRIS installed in three water table regimes .................... 75 2 13 IRIS performance in marl soils with flooded conditions ................................ ................... 76 2 14 Comparison of digital image techniques used to analyze an IRIS tube ............................. 77 2 15 Iron oxide removal on IRIS tubes treated with the 30 cm water table .............................. 78 2 16 Manganese oxide removal from IRIS tubes as estimated by three methods ..................... 79 2 17 Iron oxide removal from IRIS tubes as estimated by three methods ................................ 79 2 18 Correlation of three methods used to determine manganese oxide removal ..................... 80 2 19 Correlation of the three methods used to determine iron oxide rem oval ........................... 81 2 20 Iron and manganese coated IRIS tubes after receiving the same treatments. .................... 82
9 LIST OF ABBREVIATION S ASTM American Society for Testing and Materials C Degrees Celsius CC Cubic centimeter CEC Cation exchange capacity CM Centimeter D B Bulk density the weight per volume, typically expressed in grams per cubic centimeter for soils DI Deionized water DDI Double deionized water DPI Dots per inch, indicatin g visual quality of a digital image EC Electrical c onductivity EDTA Ethylenediaminetetraacetic acid E H Reduction o xidation p otential FIRIS Iron oxide coated Indicators of Reduction in Soil G Gram IC Inorganic carbon ICP I nductively coupled plasma atomic emission spectrometry IFAS Institute of Food and Agricultural Science IRIS Indicators of Reduction in Soil includes FIRIS and MIRIS L Liter M Moles per liter a concentration express ed in units of mol /L MEQ M illiequivalent MG Milligram MIRIS Manganese oxid e coated Indicators of Reduction in Soil ML Milliliter
10 MM Mil l imeter MWCO Molecular weight cut off NRCS Natural Resources Conservation Service a Federal Agency NSSL National Soil Survey Laboratory NTCHS National Technical Committee for Hydric Soils NWI Na tional Wetlands Inventory OM Organic matter P P T Platinum PVC Polyvinyl chloride pipe RPM Revolutions per minute SHE Standard h ydrogen e lectrode S I L Silt loam soil texture based on particle size analysis TREC Tropical Rese arch and Educat ion Center UF University of Florida U S M icro siemens USDA United States Department of Agriculture USFWS United States Fish and Wildlife Service
11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INDICATOR S OF REDUCTION IN SOIL (IRIS) IN VARIOUS CONDITIONS OF SATURATION IN SOUTH FLORIDA MARL By Christine A. Coffin August 2012 Chair: Yuncong Li Major: Soil and Water Science Marl soils of south Florida lack redoximorphic features traditionally used to determine if anaerobic conditions exist. Soils with anaerobic conditions rely on alternative oxidants, such as manganese oxide (MnO 4 ) and iron oxide (FeO 3 ) to facilitate microbial respiration processes. For field investigations of anaerobic soils the platinum (Pt) electrode is a standard tool to measure redox potentials, but is not cost effective. Indicators of Reduction in Soils (IRIS) were developed as an alternative to Pt electrodes for use in field investigations of hydric soils. In this study, three marl soil types were extracted to compare anaerobic determination results of Pt electrodes, iron oxide coated IRIS (FIRIS), and manganese oxide coated IRIS (MIRIS). IRIS w ere installed in the upper 45 cm of the soils and treated with 0 cm, 15 cm and 30 cm water table depths. IRIS tubes were extracted at intervals of 14, 28, and 56 days to quantify metal oxide removal using visual, trace and scan methods. Results of the stud of the soil during the 56 days when inundated to the soil surface. Findings suggest that in high calcareous soils FIRIS need 56 days of installation time to reduce enough iron oxide to ref lect anaerobic conditions. Metal oxide removal in saturated conditions from MIRIS averaged just 29% of the amount removed from FIRIS, and MIRIS failed to reflect anaerobic soil conditions
12 over the 56 day installation. Pt electrode readings indicated anaero bic reduction within seven days, allowing for anaerobic conditions to be determined in as little as 14 days in calcareous marl soils
13 CHAPTER 1 INTRODUCTION AND LIT ERATURE REVIEW Rationale Soils with marl deposits at the surface are found in the extreme southern portions of Florida. The marls are taxonomically classified as carbonatic Typi c fluvaquents and represent seven soil series located in Miami Dade, Monroe, Broward and Collier c ounties in Florida. An inventory of marl soils in south Florida is app roximately 500 ,000 acres including federally owned lands within two national parks and a preserve (Davis et al ., 2005 ; USDA NRCS, 1998 ). In protected, natural areas such as Everglades National Park, Biscayne National Park, t he Big Cypress Preserve, and oth er protected coastal wetlands, marl soils exist in functioning wetland systems of inland freshwater prairies and along coastal fringes where seasonally flowing, slow shallow water environ s still exist. Many protected species use the marl lands for habitat, most notably is the Cape Sable Seaside Sparrow ( Ammodramus maritimus mirabilis ) the American Alligator ( Alligator mississippiensis ) and a vast array of native and migratory bird species. Although such designated areas are protected, the natural hydroper iod has been altered by water control structures installed during the mid 1900s to drain the Everglades for flood protection. nited S tates Fish and Wildlife Service (USFWS) in the 1979 publica tion Classification of Wetlands and Deep Water Habitats of the United States to describe soils found in wetland systems. The term has been widely adopted and officially recognized as one identifying characteristic of a wetland along with wetland hydrology and h ydrophytic vegetation. T o consistently identify, delineate, and regulate wetlands, a formal system of hydric soil s classification has been developed and continues to evolve as knowledge expands. Anaerobic c onditions and saturation must both exist to s upport the determination that a soil is hydric.
14 Once a soil is classified as hydric it has regulatory connotations relative to land use and management A h ydric soil is defined as having formed in conditions of saturation, flooding or ponding for long eno ugh period s during the growing season to develop anaerobic conditions in th e upper part (USDA NRCS, 2006). During the development of the first national list of hydric soils t he National Technical Committee for Hydric Soils (NTCHS) was formed to oversee th e process S oils are evaluated according to the criteria, indicators and test methods approved by the NTCHS. Anaerobic conditions can be documented by either r edox p otential (Eh) readings, iron reduction on buried iron oxide coated Indicators of Reduction in Soils (FIRIS ) tubes, the presence of reduced iron (Fe 2+ ) in the soil, or positive reaction to alpha alpha d ipyridyl solution. The latter two options can only be used on soils with sufficient iron present, which is not characteristic of marl soils in so uth Florida. FIRIS and manganese coated ( MIRIS ) tubes were developed for use by field soil scientists in problem soils that lack readily assessable morphologic indicators of redu ction and anaerobic conditions. Research Goal s Th is study compare s the result s obtained using platinum ( Pt ) electrodes, FIRIS, and MIRIS in determining anaerobic soil conditions in calcareous marl soils An evaluation of the potential of these tools for use by field soil scientists conduc ting hydric soil determinations would be hel pful. C hemical and physical properties of the three m arl soils used in the research can help identify characteristics that may affect anaerobic redox processes As part of the evaluation of FIRIS and MIRIS examine how hydric soil determinations might vary by analysis methods used to determine the percent meta l oxide removal from Indicators of Reduction in Soil ( IRIS ) tubes
15 Application IRIS technology has the potential to provide a cost effective alternative to scientists conducting hydric soil investigati ons in problem soils by all ow ing rapid in situ assessment s with minimal expense for equipment and labor. Hydric soil determinations are a key component in developing interpretive land use limitations and are a requirement of regu latory wetland determinati ons IRIS could have potential application in monitoring the progress of wetland restoration creation, and enhancement projects associated with the Comprehensive Everglades Restoration P lan in south Florida. Marl Soil Formation and Geographic Extent To de termine where marl soils exist, it is important to understand what marl is, and the differences used to describe marl In the National Soil Survey Handbook marl is described as a limnic layer that is light colored and reacts with hydrochloric acid (USDA NR CS, 1997) The two characteristics used to verify marl soils in the field is a moist soil color value of 5 or more and reaction with dilute (1M) HCl to evolve carbon dioxide CO 2 as shown in Equation 1 1. CaCO 3 + 2HCl = H 2 O + CO 2 (g) +Ca 2+ + 2Cl ( 1 1 ) Limnic materials are used as modifiers of texture to describe either the origin or the material itself These materials were deposited in water by precipitation or through the action of aquatic organisms or derived from plants and organisms (USDA NRCS, 1 997). Marl is primarily calcium carbonate, typically found below the surface of organic soils where the marl formed in a preexisting open water environment (USDA NRCS, 1999). Marl soil forms in areas inundated with calm, shallow waters over a limestone sub strate In the 1996 Dade County Soil Survey (USDA NRCS, 1996) the depth to limestone varies in each of the marl soil series: Biscayne marl = limestone bedrock at 1 cm to 51 cm below the soil surface
16 Perrine marl = limestone bedrock at 51 cm to 1 m below th e soil surface Pennsuco marl = limestone bedrock at 1m to 2 m below the soil Equation 1 2 explains what occurs when periphytic algal mats dominated by cyanobacteria species cause an increase in water pH through respiration, leading to saturation and preci pitation of calcium carbonate from the water column. The powder fine precipitate accumulates on the inundated substrate to form these waterborne soils. The rate at which calcitic marls accumulate is directly related to their spatial extent and productivity (Davis et al., 2005). Ca 2 + + 2HCO 3 CaCO 3 + CO 2 + H 2 O ( 1 2 ) Algal and bacterial mats are able to survive the fluctuating hydroperiods in marl prairies and sloughs of the Everglades ecosystem. Wetlands with these abbreviated hydroperiods are dry long enough each year to prohibit the accumulation of organic material through oxidation, preventing the formation of organic soil layers (Davis et al., 2005). With the annual return of the rainy season, around June in south Florida, these periphyton mats are rehydrated in shallow waters that overlay limestone bedrock. Marl soils are located across the globe where these conditions are or were present at various geologic periods with favorable water levels and alkaline chemistry The term marl is widely used in English language geology, while the terms m ergel and s eekreide are used in European references (Strasser et al., 2008). It is curious that marl soils are not mentioned in areas of the Far East, Russia, Central/South America, or Australia. Soils in parts of the world have not been mapped, or may not identify marl by the same terminology or concepts that are used in the United States. A search of the Natural R esources C n ational soils database revealed that in the United States, marl at the soil surface is not common or expansive in spatial distribution Outside the Everglades, there are lacustrine marl soils of the Great Lakes and in
17 coastal areas of South Carolina. Figure 1 1 shows the distribution of marls in Florida, which are primari ly limited to Miami Dade, Monroe, Broward and Collier counties associated with the southeastern areas of the greater Everglades ecosystem. In this subtropical environment marls form under shallow surface waters that are seasonally present from 2 to 9 month s in duration each year (Davis et al., 2005). An inventory of the acres of marl soils in south Florida are listed in Table 1 1. The actual acres are significantly higher since federally owned lands within Everglades National Park, Biscayne National Park an d Big Cypress National Preserve are not included in the most recent published soil surveys. A t least 190,000 h ectares of marl prairies surround the Shark River Slough in the heart of Everglades National Park (Davis et al., 2005). South Florida Hydrologic A lterations In protected, natural areas of south Florida, such as Everglades National Park, Biscayne National Park, the Big Cypress Preserve and protected coastal wetlands, marl soils often exist in functioning wetland systems In areas where seasonally, sl ow flowing, shallow water is present, inland freshwater prairies sloughs, and coastal fringe ecosystems are sustained Although many of these designated areas are protected, their natural hydroperiods were either abbreviated or removed by water control st ructures installed to provide flood protection for developing areas. If the wetland hydrology at a site has been removed, then the wetland functions and values are also altered. An over simplified relationship of the interdependence of the site characteris tics is represented by the formula in Equation 1 3. WETLANDS = w etland h ydrology + h ydric s oils + h ydrophytic v egetation ( 1 3 ) Everglades hydrology was significantly altered, beginning at the turn of the 20 th Century and peaked during the early 1960s to f acilitate the objective of reclaiming swamp land for land uses considered more productive at that time Drainage projects not only provided flood
18 protection but also enabled year round crop production on marl soils that were drained through a vast network of canals and levees Before the 1960s, the east glades area south of Miami and east of Homestead was used during the dry season (December April) to produce seasonal truck crops Post harvest, these areas were left idle until the following dry season ret urned and the water table fell below the shallow root systems of these crops Before initiation of systematic drainage in south Florida marl soils had hydroperiods lasting from 2 to 9 months in duration After drainage canals and pumps were installed, the eastern glades stayed dry for 9 months of the year (Davis et al., 2005). At present, drained marl soils are used for the year round production of food crops (okra, malanga, bananas, squash, beans, corn), field grown ornamental plants (woody landscape plan ts and large palms), or container grown ornamental plants (bedding plants, annuals, herbs, orchids and bromeliads). Hydric Soils Classification At the Federal Government level wetland regulations primarily derive from the 1972 Clean Water Act and/or the S wampbuster provisions of the 1985 Food Security Act In addition, to Federal Legislation there may be s tate and c ounty level regulations for wetland use and permitting, but varies across the United States from state to state as well as between counties Hy dric soils are an integral component in jurisdictional wetland determinations that affect land use planning and management by individuals and bodies of government. To assist the USFWS in delineating wetlands for the National Wetland Inventory (NWI) in the late 1970s, the U nited S tates D epartment of Agriculture Natural Resources Conservation Service (NRCS) was asked to develop a national h ydric s oils list that would enable the use of remote sensing and existing soil surveys for the production of NWI maps The culmination of that collaborative effort resulted in the first version of the N TCHS published in 1985 (Richardson and Vepraskas, 2001). During the development of the first n ational l ist of h ydric
19 s oils in the early 1980s, there was disagreement among NRCS personnel throughout the country on what criteria and which soils should be included The N T C H S was formed to oversee the process to be used as well as develop a scientifically sound list that could withstand scrutiny (Tiner, 1999) Marl soils in this research were evaluated according to the criteria, indicators and test methods approved by the NTCHS. NRCS maintains the official list of hydric soils covering all soil map units delineated throughout the United States and its territories During the 20th Century, most soils in the United States were mapped on a countywide basis and some have been updated since the original publications were completed Each map unit in a soil survey is either classified as hydric or non h ydric according to the current crit eria of a hydric soil The list provides a quick indication of whether a specific site may contain hydric soils and/or wetlands. A soil can be saturated and not anaerobic if the water has a significant amount of dissolved oxygen present Likewise, fine tex tured soils can show anaerobic conditions when not saturated if the water holding capacity of the soil matrix excludes O 2 while large macro pores in the soil structure contain air Capillary fringe is not always considered to be part of the saturated zone as its influence above the water table is not easily measured and can be a dynamic property affecting soil water state (Vepraskas and Sprecher, 1997). Migliaccio et al (2008) found that capillary rise in Biscayne marl soils had a significant influence o n the soil water content in the soil layers above the water table. Criteria and Field Indicators of Hydric Soils For a soil series or map unit to be included in the hydric soils list, at least one of the hydric soil criteria in Table 1 2 must be met Crite ria 1, 3 and 4 are also indicators of hydric soils that can be determined in the field Criteria 2 are used for the purpose of searching the National Soil Information System database in preparation of hydric soil lists and are not field indicators
20 ( USDA NR CS, 2002 ) One limitation placed on the criteria is soil temperature, which is not a concern in the Everglades region where soils are warm enough to facilitate year round biologic activity An on site investigation is necessary to determine the soil proper ties and document any evidence of saturation in the upper 30 cm of the soil. Correct and consistent determinations are necessary if they are to be credible and defensible in a legal context. Today, a n official list of 48 indicators is used to identify and delineate hydric soils in the United States The indicators are applicable to one or more of the 26 geographically defined Land Resource Regions (LRR) in the United States, which are alphabetically labeled A Z Land Resource Regions serve to group region s with similar soils and land uses for broad management of soil survey activities and data management by NRCS. Figure 1 2 shows that most of Florida is in LRR U Florida Subtropical Fruit, Truck Crop, and Range Region, extending from Key West north to abou t Gainesville, and encompassing the area of interest in this research Indicators are alphanumerically coded, with a letter preceding the number to indicate the conditions that each indicator can be applied; A = All soil types, S = Sandy soils, F = Fine te xtured soils (loamy and clayey), T = Test Indicators Each LRR has an official list of hydric soil indicators that can be used to identify and delineate hydric soils. Indicator A4, hydrogen sulfide odor, may be present in coastal areas of marl accumulation receiving sulfur rich marine waters Soils collected in this study do not include marls receiving tidal waters. A soil can have more than one indicator present, which provides additional confidence in a hydric determination. The primary indicator relative to marl soils in LRR U is F10, as follows (USDA NRCS, 2006): All mineral layers above a marl layer must have a color chroma of 2 or less, OR F10 is only for use in Land Resource Region U The marl layer must have a color value of 5 or more The marl layer m ust start within 10 cm of the soil surface
21 Evaluating the Hydric Status of Marl Soils Labeling a soil as marl indicates formation in water from the biogenic process of CaCO 3 precipitation in response to algal respiration Marls in south Florida are taxono mically classified as carbonatic Typic fluvaquents and represented by 7 soil series in the 4 counties listed in Table 1 1 Some silt loam soils existing in Broward and Collier are not positively identified as mar l soils, as it is thought they may have deri ved from non biogenic processes of erosion and deposition ( H. Yamataki personal communication 2010). In the 1998 Collier County Soil Survey of Florida, the Pennsuco soil series is not labeled as Pennsuco marl, but rather as Pennsuco silt loam Understand ing that it is a marl soil or marl variant requires the reader to review the Pennsuco series description as it occurs in that survey area, which states the soil formed in marl and is associated with wet prairies (USDA NRCS, 1998) The concern is not simply the marl label itself, but the hydric soil criteria that are applicable once a soil is labeled marl. In October of 1999 some soils in Collier County, Florida were investigated for hydric soil identification problems Pennsuco marl was one soil sampled and found to contain 30% CaCO 3, 7% organic matter, and the balance was quartz sand Another observation during that investigation i color values of 3 and 4 were not uncommon Following t hat investigation, W ade Hurt with the NTCHS proposed the following considerations to address the problems associated with hydric soil identification in south Florida: Marl should be defined to contain a minimum of 60% CaCO 3 with no secondary carbonates Ma rly mineral materials be allowed u sing indicator F10, as long as the CaCO 3 content is at least 15%, with no secondary carbonates
22 Include colors with a value and chroma of 4/1 in indicator F10 The 2010 update to the Hydric Soil Field Indicators, had none of the proposed changes, or expansion of criteria incorporated into the F10 indicator (USDA NRCS, 2010). Morphologic I ndicators of Soil Saturation The most common features of a soil, used in situ, to evaluate if a soil is saturated for a significant period i s the organic matter content, a gleyed matrix, and/or redoximorphic features (Tiner, 1999) Iron reduction can be visibly determined with the change from reddish brown ferric iron into a grayish ferrous iron; the resulting gray soil matrix is then referred (USDA NRCS, 2010) Gleying is not used as an indicator in marls since the gray color is not derived from the presence of Fe 2 + bu t rather the biogenic formation of calcium carbonates. When marl occurs beneath organic soil layers, this posit ion indicates a hydroperiod of sufficient duration to allow the accumulation of organic matter at the surface A review of typical pedon descriptions of marl soils in the 1996 Dade County Soil Survey reveals marls occur at the soil surface and are not over lain by an organic horizon. The color of marl does not change when dried because it contains too little organic matter to coat the carbonate particles. Most of the samples of marl from the United States have organic matter content s between 4 % and 20 % (USDA NRCS, 1999). In lieu of sufficient organic matter at the soil surface, a gleyed matrix can be used to indicate soil saturation O xygen (O 2 ) is the preferred electron receptor for soil microbes, but in anaerobic conditions microbes will reduce ferric iron (Fe 3 +) to ferrous iron (Fe 2 + ) In 1999, Chen et al explored background values of iron and other elements in marl soils from undisturbed sites in Everglades National Park (ENP) and compared these element concentrations with disturbed soils throughout Flori da Iron content in the upper 20 cm of marl soils ranged from 1295 mg/kg
23 to 100,385 mg/kg Iron and the other elements were determined to exist in higher concentrations in the ENP marl soils when compared with other soils throughout Florida (Chen et al., 2 000 ). While investigating the arsenic content of surface soils in Florida, Chen et al (2002) reported an average CaCO 3 content of 56% in south Florida marls. Since these soils are dominated by calcium carbonates (calcite), they offer little to no morpholo gical evidence of soil processes regarding saturation and anaerobic conditions Redoximorphic features such as redox concentrations and redox depletions are also visible soil characteristics expressed when a soil contai ns sufficient iron or manganese, but are not produced in marl soils for the in situ assessment of anaerobic and saturated soil condition s Reduction in Saturated Soil Conditions Reduction is a measurable process that can be utilized to document anaerobic soil condition, a prerequisite of hydr ic soils A lack of dissolved O 2 an organic food source, and the presence of anaerobic microbes to degrade the organic matter are all necessary for reduction to occur. Microbial respiration then requires an electron acceptor once dissolved In natural soi l environments, electrons are accepted by the oxidants shown in Table 1 3 listed in order of preference In the absence of dissolved O 2 NO 3 would first be utilized and CO 2 would be the last desired acceptor since significant energy must be expended on t he part of the microbe to transfer the electron (Mitsch and Gosselink, 2000). Measurement of Oxidation Reduction Potentials in Soils The ability to quantitatively measure energy associated with the electron transfer was discovered during the nineteenth cen tury while studying elements using electricity The quantity of electrical charge that passes through an electrochemical cell is proportional to the substance produced at the electrode when a redox reaction occurs Anaerobic conditions can be documented us ing Pt and reference electrodes to measure the voltage difference and o btain a
24 redox potential reading In the presence of free O 2 in the soil, potentials ranging from + 400 to + 700 mv should be expected Once saturated, O 2 can be depleted within several ho urs to several days with values dropping in the range of + 400 to 400 mv (Mitsch and Gosselink, 2000) Measuring the reduction and oxidation potential of a soil by collect ing conduct ance readings between Pt and reference electrodes is an established metho d (Appendix A ). Eh is pH dependent and the standard used to evaluate readings is + 175 m V at a pH of 7 .0 For each whole pH unit greater than 7 a correction factor of 60 mV is added to the Eh reading; + 60 mV is added for each pH unit less than 7 The value of 60 is derived from the Equation 1 4 where a line with a 60 slope is derived when plotting iron redox potentials against pH. Eh volts = Eh o 0.059 log reduced species ( 1 4 ) N (oxidized species) ( H) m Figure 1 3 demonstrates that in anaerobic conditions iron reduction potential is achieved at greater reduction potentials in non alkaline soils A query of the NRCS soil survey data base ( online at http://soildatamart.nrcs.usda.gov/ ) shows the pH in m arl samples at the 25 cm depth in Broward, Collier, Miami Dade and Monroe counties ranges from 7.0 to 8.3 Saturated Ag/AgCl electrodes with 4 M KCl electrolyte solution will vary in conductance from temperature fluctuations of just 0.13 mv for each + 1 C ( Ansuini and Dimond 1994) A small temperature correction factor is recommended when using saturated Ag/AgCl electrodes at temperatures that vary from 25 C Under climate c ontrolled lab conditions with an ambient air temperature set at 25 C, correctio n factors for temperature were not expected to be significant or necessary but were recorded. Iron Indicators of Reduction in Soil (FIRIS) Outside of a research lab the Pt electrode method i s not practical for routine use by field soil scientists due to th e installation of expensive equipment, labor requirements and expos ure to
25 hazards from weather conditions and wildlife, as well as vandalism or theft. In 2006 Jenkinson and Franzmeier introduced the use of iron coated IRIS tubes as a method to determine reduction in soil s (Castenson and Ra b enhorst, 2006) FIRIS tubes have been accepted by the NTCHS as an alternative to ol for documenting anaerobic soil conditions associated with wetlands and hydric soils. For soils low in iron, FIRIS can serve as a surroga te representing Fe 3 + as if naturally present Most Florida soils, including marls, do not contain sufficient amounts of iron to produce morphologic indicators that allow for rapid identification of an aerobic conditions in the field. In contrast to electr odes, IRIS tubes have distinct advantages in that they are relatively easy to install and extract, inexpensive, and lightweight enough to carry on foot to remote locations not accessible by vehicles Castenson and Rabenhorst (2006) described the constructi on and methodology used to evaluate F IRIS tubes for investigating soil reducing conditions The iron coated IRIS tubes are now referred to as FIRIS to differentiate them from the newly developed manganese coated IRIS, or MIRIS. FIRIS tubes for this project were provided courtesy of the U nited States Department of Agriculture Natural Resources Conservation Service, National Soil Survey Lab in Lincoln, Nebraska. Iron transformations in soils supporting rice production found that water soluble/exchangeable i ron was dependent more on pH than either iron or organic matter content (Mandal and Mitra, 1982) Soils derived from carbonates can maintain high saturations of siderite (FeCO 3 ) that do not dissolute due to the extreme buffering capacity of marl soils (Sti les et al., 2010) In high pH, alkaline soils, reduction of Fe 3 + to Fe 2 + is suppress ed and expected to limit FIRIS usefulness (Stiles et al., 2010) Manganese Indicators of Reduction in Soil (MIRIS) Since 2009 MIRIS tubes have been investigated as an alte rnative to FIRIS to accurately reflect saturated, reduced soil conditions in significantly less time than FIRIS reacts in alkaline
26 soils Dr. Cynthia Stiles, while at the U nited States Department of Agriculture (U SDA ) National Soil Survey Center perform ed bench testing of the MIRIS before an initial field test was conducted in Alaska during the summer of 2009. In the Brooks Range region of north central Alaska, FIRIS and MIRIS where installed in a calcareous fen wetland utilizing two sets of 5 MIRIS and one set of 5 FIRIS tubes The tubes were oriented along a wetland transect with high pH/Eh conditions To observe manganese oxide transformation in non hydric soils, one set of MIRIS tubes was installed in a well drained mid slope position Within four day s, the wetland MIRIS tubes displayed oxide paint removal The MIRIS positioned in the well drained slope did not show significant manganese oxide removal after 7 days, confirming manganese oxide removal in saturated conditions ( Stiles et al., 2010 ). After 42 days the five FIRIS tubes that were installed adjacent to the MIRIS had <10% oxide removal on four of the five tubes and a concentration of iron oxide on the upper 10 cm of the tube R apid removal of manganese oxide in high pH/Eh conditions support s th e idea that MIRIS could be effective in identifying the presence of reducing conditions that are unfavorable for iron reduction ( Stiles et al., 2010 ) Observing IRIS in calcareous, marl soils of south Florida provides a measure of performance to build upon the 2009 investigation by Stiles et al in Alaska. MIRIS manufacturing methods at this early stage of development is similar to th e methods used for FIRIS manufacture except manganese oxide paint production and application to the polyvinyl chloride (PVC) tubes are more labor intensive and time consuming In Appendix B there is a description of the entire process used to produce manganese oxide paint and to fabricate the MIRIS tubes. This research project provide d an additional evaluation of the new
27 MIRIS on calcareous marl soils that are problematic in their inability to reflect changes in soil hydrologic regimes. Carbonates in South Florida Marl Soils In h as been suggested that marl soils, as a positive indicator of their biogenic origin and classificatio n as marl, should contain no secondary carbonates. The formation of marl in south Florida involves in the precipitation of CaCO 3 in shallow water environments fed by water from surficial limestone aquifers. Calcite and dolomite account for 90% of natural c arbonates, are known to occur in soils as primary or secondary carbonates (Table 1 4). In soils, the most common carbonates are calcite, dolomite and magnesite (Ming, 2007). Rabenhorst et al. (1984) were able to use stable C isotopes to differentiate prima ry and secondary carbonates in limestone soils of Texas The methodology was based on the assumption that soil CO 2 is incorporated into secondary carbonates, which have more negative C 13 values than marine limestone (West et al., 1988) Marl soils of south Florida do not fit this model since Florida marls are formed in response to respiration by periphytic algae in shallow freshwater as well as saltwater environs Morphologic characteristics of pedogenic carbonates often used in arid regions are not present in the humid climate of the Everglades ecosystem with an annual rainfall of a pproximately 152 centimeters. Segregation of primary from secondary carbonates in soils derived from highly calcareous materials is challenging as current analytical methods cann ot identify the Mg as deriving from MgCO 3 or CaMg(CO 3 ) 2 (Goh and Mermut, 2008) Soil samples collected from this research will have total soil carbonates determined from the volume of CO 2 gas generated upon dissolution of solid carbonates with HCl acid, us ing a Scheibler apparatus (USDA, 1993).
28 Research Objectives Iron and Manganese are not expected to be present in appreciable amounts in natural systems of the greater Everglades ecosystem to reflect reducing conditions, but can be provided to the soils via FIRIS and MIRIS tubes to simulate redox potentials of soils with naturally sufficient reducible iron and manganese oxides The newest MIRIS tubes were fabricated and tested against the original FIRIS to see if MIRIS tubes can reflect reduction at a faster and more reliable rate than with the FIRIS, which is known reduce iron oxides at a slow rate in high pH conditions. When reduced, both metals are soluble and move with soil water, so that there is visible evidence of translocation by absence of metal oxid es from the tube surfaces. The first objective of this research was to evaluate the performance of the recently developed F IRIS and MIRIS tubes against the established, conventional Pt and reference electrodes A comparison of the quantity of metal oxides removed in a time interval allows a comparison of anaerobic soil determinations by each tool. Results can provide guidance on the most appropriate tool for use in high pH, calcareous soils in south Florida. The second objective was to determine if general soil characteristics of marl might have some influence on redox processes since FIRIS and MIRIS were both dev eloped for use in problem soils. South Florida marls selected for this research are integral components of wetland ecosystems that are being restor ed as part of the Comprehensive Everglades Restoration Project Marl soil properties may help in identifying the limits for use of IRIS tools or improve IRIS technology for use in alkaline problem soils. The final objective w as to analyze results obtained from three methods used to determine the percent metal oxide removal from the surface of IRIS tubes In keeping with the concept of using trained soil scientists to provide a visual estimate of the percent of oxide removal, the visual estimate would be co mpared against a digital scan of the tubes Digital analysis by hand
29 tracing as well as direct scanning with a bar/hand scanner will include both techniques currently used. Table 1 1 Extent of m arl soils in Florida Soil s eries Acres in Miami Dade Co unt y Acres in Monroe Co unty Acres in Collier Co unty Acres in Broward Co unty Total acres by soil series Biscayne 67,364 ---67,364 Cudjoe -3,410 --3,410 Key West -450 --450 Lignumvitae -1,360 --1,360 Pennsuco 14,490 -7,697 1,886 2 4,073 Perrine 50,362 --276 50,638 Saddlebunch -1,140 --1,140 Total acres 132,216 6,360 7,697 2,162 148,435 -Indicates soil series not present in the county. Data compiled from the U S D ept. of A griculture database o n official soil series M ore info at https://soilseries.sc.egov.usda/ osdquery.aspx (22 Jul y 2012).
30 Table 1 2 Criteria for h ydric s oils Table 1 3 Oxidants in soil redox processes, listed in order of preference by reducing bacteria Oxidized form Reduced form Redox potential (mV) NO 3 N 2 O, N 2 NH 4 + +250 Mn 4 + Mn 2 + +225 Fe 3 + Fe 2 + +100 to 100 SO 4 2 S 2 100 to 200 CO 2 CH 4 < 200 Data adapted from: Mitsch W.J. and J.G. Gosselink, 2000. Wetla nds, 3rd Edition. John Wiley and Sons, Inc., Canada. Table 6 2, p. 169.
31 Table 1 4 Characteristics of s oil c arbonates
32 Figure 1 1 Areas m apped as m arl in s outh Florida. Source: Martin Figu eroa, Senior Soil Scientist with the U nited S tates D ept. of A griculture N atural R esources C onservation S ervice in the M ajor L and R esource A rea 8 office in Ft. Myers, Florida (October 2009).
33 Figure 1 2 Major Land Resource Region U in Florida. Map so urce: United States Department of Agriculture, Natural Resources Conservation Service. 2006. Land Resource Regions and Major Land Resource Areas of the United States, the Caribbean, and the Pacific Basin. U S Department of Agriculture Handbook 296. Online at : http://www.mo15.nrcs.usda.gov/technical/mlra_fl.html (May 2012)
34 Figure 1 3 Standard redox (Eh) /pH line for determining anaerobic conditions. Adapted from The Hydric Soil Techni cal Standard 2007, Technical Note 11, p 9. Available at : http://soils.usda.gov/use/hydric/ntchs/tech_notes/index.html
35 CHAPTER 2 ANAEROBIC SOIL DETER MINATIONS Introduction Anaer obic soils incapable of displaying morphologic indicators to allow rapid, in situ assessments by field scientists in soil mapping, site use limitations, and wetland determinations or delineations are referred to as problem soils (Vepraskas and Sprecher 19 97) Iron Indicators of Reduction in Soil ( FIRIS ) and m anganese Indicators of Reduction in Soil ( MIRIS ) tubes were recently developed as alternatives for assessing anaerobic soils that lack readily assessable morphologic indicators of reduction and anaerob ic conditions FIRIS tubes have been in use since 2006 and accepted by the National Technical Committee for Hydric Soils ( NT C HS ) as an alternative to electrodes for determining hydric status of a soil through the verification of sustained anaerobic conditi ons (NTCHS, 2007) The performance of FIRIS or MIRIS has not been evaluated in south Florida where alkaline soils and wetland s dominate the landscape Marl soils of south Florida are considered problem soils. Whether marl soils are oven dry or saturated th ey maintain a light grey color, which typically indicates a reduced state in most mineral soils with sufficient iron content. This study provides an additional evaluation of the performance of FIRIS and MIRIS in high pH soils of a different ecological sett ing than Alaska to see how these recently developed redox tools may be applied. In this research, simultaneous investigations of both the FIRIS and MIRIS tubes abilities to reflect reduction in high pH, saturated calcareous marls will be conducted As part of this study a visual method, scan method and trace method were used on s elect Indicators of Reduction in Soil ( IRIS ) tubes to compare the percentage of metal oxides that would be calculated by each method.
36 Materials and Methods Selection and Extraction of Soil Mesocosms Three marl soil map units in Miami Dade County, Florida were selected to represent varying land uses and hydrologic alterations common on undeveloped marl soil areas of south Florida. Depth to limestone contact was 51centimeters to 1 met er in the Perrine soil and 1 to 2 meters in the Pennsuco soils (USDA NRCS, 1996). Additional considerations of site selection included land use and drainage for the possible variances in nutrient content and redox species. To best represent the in situ so il matrix, horizons/layers were extracted with minimal disturbance to the soil from each of the three sites shown in Figure 2 1. Pennsuco marl (Soil A) c ollection site had no compaction and the marl had a loose consistency similar to a powdery snow 25.4 c entimeter diameter polyvinyl chloride ( PVC ) pipes, 61 centimeters in length and beveled on one end were manually inserted to the 45 centimeter target soil A 25.4 centimeter diameter pipe gripper was then inserted into the top of the submerged column and e xpanded to achieve an air tight seal, preventing the soil core being vacuumed from the bottom of PVC pipe when extracted to the surface A manually powered lever and fulcrum was constructed of 15 cm x 15 cm beams and heavy duty chain to extract the soil me socosm to the surface Immediately after each extraction, a sand filter, 25.4 centimeter PVC cap, and 1.6 centimeter clear T ygon tube were installed onto the PVC pipe before carrying to the vehicle and placing the rack for transport to the lab Perrine mar l drained (Soil B) and the Pennsuco marl drained (Soil C) mesocosms were collected using a backhoe during the first week of August, 2011. After transporting all twenty seven mesocosms to the T ropical R esearch and E ducation C enter (TREC) lab, they were le ft in the rack, under cover, to drain overnight The following day they were move d into a climate controlled lab where each PVC column was placed into a custom built wooden cradle to
37 facilitate installation of plumbing appurtenances and provide structural stability during data and sample collection The twenty seven soil columns, containing 9 mesocosms from each of the three marl soils, were randomly located throughout the lab space to minimize the impacts of micro climate variations. Once mesocosms were i n the lab, water table levels were maintained approximately 30 centimeters below the surface in each column throughout the next 21 day s until treatment water tables we re established. Maintaining soil moisture minimize d development of surface cracks in the soil from the air condition ing which could affect redox potentials by allowing oxygen penetration deeper into the soil than would naturally occur Installation of Equipment for Monitoring Reduction P latinum e lectrodes A Campbell Scientific Data Logger (mod el CR1000) was utilized to test the platinum ( Pt ) electrodes before installation into the soil mesocosms Analog inputs 1 and 2 were wired for differential measurements that would be manually connected to each of the 27 Pt electrodes and the Ag/AgCl refere nce electrode before to each reading A laptop was used to interface with the data logger, using Loggernet (version 4.0) software to receive and record the data of each reading which represented a 10 second average of measured redox potential. During the last week of August 2011, the water that had been maintained in the bottom of the mesocosms to prevent drying and cracking, was drained for installation of the Pt electrodes and the other redox equipment to be installed The electrodes were labeled and ins talled in the center of the soil column by boring to 23 cm depth with a 2.54 centimeter diameter auger The platinum tip was then inserted into undisturbed soil to the target 25 cm depth, as prescribed for collecting redox readings in fine textured soils ( NTCHS 2007) From the extracted soil, thick slurry was made with deionized water ( DI ) water, and then poured back into the auger hole to vertically anchor each Pt electrode. The 0 cm and 15 cm water table depths guarantee the Pt
38 electrode would be submerg ed and reduction was expected A water table at 30 cm provide d an opportunity to see if the capillary movement in marls is sufficient enough for reducing conditions to occur at the 25cm redox reading depth. Salt b ridges A KCl salt bridge has shown to be ef fective in providing consistent electrical conductivity while not influencing redox processes (Veneman and Pickering 1983) To install the salt bridges, a 2.54 centimeter diameter soil boring was made in the same manner described for installing the Pt ele ctrodes. A soil slurry was made from the extracted soil with DI water and poured around the bridge in each soil me s ocosm to secure the bridge position and ensure electrical conductivity. After installation, each salt bridge was labeled and a removable PVC cap was kept on the top of each bridge to keep the exposed agar gel from drying out and shielding the surface from any foreign particles that might alter reference electrode conductance I ndicators of Reduction in Soil (IRIS) The FIRIS were stored in clima te controlled lab until ready to be installed in order to prevent exposure to moisture Installation procedures for IRIS are described below and are the same, except that the diameter of the FIRIS tubes was 1.27 centimeters while the MIRIS were 1.91 centim eters The diameter of the soil boring is important in providing direct contact between the oxide paints on the IRIS tubes and soil, since using soil slurry would alter the soil structure and not best represent the soil in field site conditions. A soil bor ing with a diameter smaller than the IRIS tube diameter is also a concern, as there is the potential to remove oxide paints during installation from excessive abrasion. Three 1.27 centimeter and three 1.91 centimeter borings were made to the 45 centim eter depth in each soil column. The extracted soil was collected from three layers from each column, labeled and allowed to air dry for analyses of soil properties. To protect the oxide paint during
39 insertion into the soil borings, each IRIS tube was envel oped in a clear plastic sleeve before inserting to the 45 centimeter depth. After installed, the protective sleeve was vertically removed to expose the tube surface to the s urround ing soil. Due to the size limitations of 25.4 centimeter diameter soil colum ns, only 3 each of the FIRIS and MIRIS tubes could be placed in each mesocosm to spatially accommodate all the equipment as depicted in Figure 2 2. Each IRIS tube was labeled to identify the soil mesocosm and treatment variables to which each tube would be subject to. Treatments Three marl soil series selected from the sites shown in Figure 2 1 had different land uses that could in fluence the biogeochemistry of each marl soil : Pennsuco marl, undrained from a natural wetland area Pennsuco marl, drained fro m an ornamental field nursery Perrine marl, drained from an annually tilled, row crop field Three water table levels were selected to evaluate performance of IRIS in various degrees of soil saturation. The water table treatments at 15 cm and 0 cm provide a simulation of the soils in conditions of saturation and inundation, respectively 30 cm below the soil surface 15 cm below the soil surface 0 cm at the soil surface Three extraction intervals were selected to determine how rapidly the FIRIS vs. MIRI S would reflect the reducing conditions in the soils 14 day s after installation 28 days after installation 56 day s after installation At each time interval, one FIRIS and one MIRIS w ere extracted from each of the 27 soil mesocosms so that the performance of both types of IRIS could be evaluated within the same
40 water table and sampling date. Redox readings, ambient air temperature and soil water pH were collected during th e late afternoon/early evening for 56 days from September 1 through October 27, 2011. During the first week, readings were taken 1 time each day; during the 2nd week, reading were taken 1 time every two days; during the 3rd week, readings were taken 1 time every three days; during days 28 through 56 readings were taken every 7 days The f requency of the redox readings meets the minimum requirement of no more than 7 days between readings as specified by the Hydric Soils Technical Standard If a predominance of reduction continues for at least 14 consecutive day s, then anaerobic conditions can be documented by this method (NTCHS 2007). With the extraction of 9 soil profiles for each of the three marl soils, three replica tes were used to evaluate t he IRIS tubes Cumulatively, the replicates would provide 162 IRIS ( 3 soils x 3 water table tr eatments x 3 extraction intervals x 6 IRIS per mesocosm = 81 FIRIS and 81 MIRIS) to be evaluated from the 27 soil mesocosms. Soil P roperties and S ampling Soil samples were collected from each of the 27 soil mesocosms during installation of the IRIS tubes t o evaluate soil properties that are known to effect reduction and nutrient availability Marls soils collected for this research are taxonomically classified as Entisols, are relatively young in their development with only A and C horizons described in the top 45 centimeter of the soil profile (USDA NRCS 1996) The Perrine marl, drained soil at the second collection site is annually disk plowed to a 15.2 centimeter depth and the other two sites were not tilled and had little to no soil horizon development Soil samples were collected from three layers in each of the 27 soil mesocosms; Layer A = 0 to 15 cm, Layer B = 15 to 30 cm, Layer C = 30 to 45 cm The 81 samples were placed in soil sampling bags, labeled, air dried for 42 days and passed through a 2mm mesh screen in preparation of the soil analyses described below Soil and soil water
41 samples were processed at the TREC lab from November 2011 through January 2012, with 3 replicates f rom each soil series analyzed. Organic matter measurement by Loss of Ign ition (LOI) The evaluation of organic matter ( OM ) content by the Loss on Ignition (LOI) American Society for Testing and Materials ( ASTM ) standard D2974 07 method C by dry combustion was chosen for the ease of use (ASTM, 2000) Ten grams of air dried, sie ved soils was processed from each of the 81 soil samples (3 layers x 27 mesocosms ) for determination of organic matter using Equation 2 1. Organic carbon was then estimated at 50% of OM content. % OM = oven dry soil weight soil weight after ignition x 1 00 ( 2 1 ) oven dry soil weight Carbonate c omposition Eighty one soil samples were collected from each of the three soil layers (0 to 15 cm, 15 to 30 cm, and 30 to 45 cm depths). Determination of total carbonate content was determined by using a Scheible r apparatus to measure the volume of CO 2 gas generated when the soil carbonates react with 6 M HCl (Dreimanis, 1962). The percent of inorganic carbon was calculated as 12% of the CaCO 3 as prescribed by the procedure used at TREC for determining inorganic c arbon in soils. Soil b ulk d ensity (Db) When adding DI water to replace the water lost from evaporation in the lab, there was a concern that Perrine marl water table might equalize too slowly to maintain the desired water table level and affect redox. A bul k density test was conducted by the ring method to see if the compacted layer below the plowed surface of the Perrine marl might have some influence on soil wat er movement in the mesocosms. A standard 7.6 centimeter ring could not be used to collect the sa mples due to the installation of 8 pieces of the redox equipment in each me s ocosm A
42 smaller ring with an inside diameter of 5.1 centimeters was fabricated from galvanized steel pipe and used successfully. Samples were pulled directly from the soil mesocos ms in November 2011, after the 56 days of redox readings were collected and soil disturbance would no longer be an issue. Three mesocosms of each marl soil series had samples collected from both the surface (0 to 2.5 cm) and subsurface layers (15 to 17.5 cm). A total of 18 samples were weighed for moist weight, oven dried for 24 hours at 95 C, weighed again for oven dry weight, and bulk density calculated in g/c c Soil p article s ize Particle size was determined using a Bouyoucos hydrometer (Gavlak et al ., 2003). The h y drometer method removes CaCO 3 with acid before drying the soil samples for particle size analysis. Since marl soils are predominantly composed of CaCO 3 this step was omitted from the process to avoid significant alteration of the marl soils. Three mesocosms from each of the three marl soil series had 500 grams of soil col lected from all three layers (0 to 15 cm, 15 to 30 cm and 30 to 45 cm). The 27 soil samples were oven dried for 24 hours at 95 C and passed through a 2 observed in the sieved samples, but without acid dissolution, could not be separated from the similar sized sand particles. Iron, mangane se and calcium content of soils Total iron, manganese, and calcium of the marl soils were measured by inductively coupled plasma atomic emission spectrometry (ICP) at the Analytical Research Laboratory on the U niversity of F lorida main campus in Gainesvill e, Florida using E nvironmental P rotection Agency standard m ethod 200.7. M etals were ac id digested and extracted from 81 soil samples in the Soil and Water Lab at TREC Twenty seven of the extractions were selected for shipment to
43 Gainesville. The samples p rovided three replicates for each of the three layers in each of the three soil types (3x3x3=27). Cation e xchange c apacity (CEC) A method was first described by Polemio and Rhoades in 1977, where Na and Cl were determined by ICP without prior removal of C aCO 3 was used for CEC. The method has two step s using NaOAc NaCl + 60% ethanol as the saturating solution then extracted by application of Mg(NO 3 ) 2 (Wang et al ., 2005). The r esults from ICP were used in Equation 2 2 to calculate CEC in cmol/kg. CEC = (N a t Na sol ) = Na t [Cl t (Na/Cl) sat.sol ] ( 2 2 ) Na t = total sodium Na sol = soluble sodium Cl t = total chloride Na/Cl sat.sol = ratio of sodium to chloride in the saturating solution Temperature, pH, and e lectrical c onductivity Digital and mercury thermometers were used to check and record ambient temperatures when each redox reading was collected Temperatures in the climate controlled laboratory remained stable at 22.8 to 23.9 C, except on two occasions when the air conditioning unit was not working properly and temperatures rose to 26.7 C and 28.9 C for less than 24 hours Those temperature increases represent a maximum variance in the redox reading of less than 0.8 mv, ha ving inconsequential effects. United States Department of Agriculture (USDA) published soil survey data for Miami Dade County shows pH values form marl soils at the 25 cm depth ranges from 7.4 to 8.4 (USDA NRCS 1996 ) The 25 cm depth corresponds to the de pth platinum electrodes were installed to collect redox readings. To document soil pH, 20 ml soil water samples were
44 collected from the valve in the bottom of each soil mesocosm before each reading was taken with the data logger. Water levels in the mesoco sms were checked the same time each day to maintain the water table treatment levels at 0 cm, 15 cm or 30 cm depths If water had been lost to evaporation, then DI water was added after the soil water samples and the redox readings had been taken The pH o f the DI water being added to maintain the water table levels was monitored over the 56 days to identify any possible influence the DI water chemistry may have on redox processes. Electrical conductivity of the water samples was also checked as an indicato r of salinity The Pennsuco marl, undrained soil collection site was about 2 kilometers west of Biscayne Bay and subject to salt water intrusion as well as windblown salts associated with tropical storms or hurricanes approaching from the eastern Atlantic A total of 513 soil water samples were collected, filtered and analyzed at the Soil and Water Lab at TREC in October and November 2011 Electrical c onductivity and pH were measured using an Accumet Excel SL50 Dual Channel conductivity meter and an Accumet AR60 Dual Channel pH meter. Methods of A nalyzing O xide R emoval from IRIS T ubes Following removal of IRIS tubes, metal oxide removal can be estimated visually or it can be digitally scanned and analyzed using image analysis software. This latter method of digital analysis can be accomplished with two different techniques. The first technique is to scan the actual IRIS tube, which may require multiple images be cut and pasted together to attain a single seamless image of a round tube. A second technique is t o wrap the tube in clear acetate and hand trace the reduced areas, then the acetate sheet can laid flat and scanned as a single image for import into the software. Each of these three techniques was utilized on select IRIS tubes extracted for this study an d will be referred to as either the visual method, scan method or trace method when discussed.
45 After each extraction the IRIS tubes were rinsed under a gentle flow of tap water to remove loose soil particles, and then placed in a test tube rack to dry. Onc e in a completely aerobic environment, metal oxide reduction ceases preserving the removal pattern on the PVC tubes. After completely dry, each labeled tube was stored into a 6 mil plastic sleeve to prevent further alteration/removal of any remaining oxide paint until analysis could be performed. The initial scope of the research was to only use the visual estimate following the guidelines provided by in the Hydric Soils Technical Standard (NTCHS, 2007). T he successful construction of a rotation device that allowed the adaptation of a hand/bar scanner for recording the surface of the IRIS tubes as a single image prompted the desire to compare the results attained from each of three techniques when applied to the same set of IRIS tubes. Visual method The firs t technique used was the visual estimate which required determining if there was any oxide removal within the upper 15 cm of the tube, at what depth removal began, and then what percent age of oxides were removed from a 15 cm long section. Tubes that did no t have oxide removal beginning within the upper 15 cm of the soil, were recorded as having 0% removal since deeper depths are not considered in anaerobic determinations (NTCHS, 2007). Over one day, all 162 tubes were visually estimated for oxide removal an d 96 were estimated to have 0% removal and 10 showed 30% or more removal of oxides in the upper 30 cm of the tubes The sets of 3 that each of th e 10 IRIS tubes were grouped with in a single me s ocosm was selected for further digital analysis unless they we re observed to have no oxide removal. Trace method The trace method was executed by wrapping a sheet of clear acetate around each IRIS tube, securing with scotch tape, and marking the points of overlap. Each area of oxide removal was outlined with a perman ent black magic marker and filled in. For tubes with significant oxide
46 removal of 30% or more, this can be time consuming. A benefit of this technique is if allows exclusion of any areas of oxide removal that occurred during installation and extraction of the IRIS tubes, rather than by electro chemical reduction processes. Linear scrapes did occur on a few of the IRIS tubes used in this project and showed up on more MIRIS than FIRIS tubes and could be a consequence of the variability in the two oxide paints regarding adhesion to the PVC tube during manufacture. Scrapes may have occurred at installation from the polycarbonate protective sleeve or by shell fragments in the soil at time of extraction. Scan method Direct scanning of the PVC tubes was done at the University of Maryland (Rabenhorst el al., 2009) using a modified flatbed scanner, which required cutting and pasting of multiple images from different sides of the tube. To cut down on the time constraints associated with direct scanning, a rotation devi ce was designed and built to allow the use of a hand/bar scanner to be employed in scanning all 360 of the PVC tube as one continuous image. Belt adjustments between the turning rods allowed the device to scan both the 1.27 cm diameter FIRIS tubes as well as the 1.9 cm diameter MIRIS tubes. The hand scanner used is this study, a VuPoint Magic Wand, cap tured the images in color at 300 dpi reso lution and downloaded as jp e g image files. Two images of each tube were taken to provide an additional option if one image was distorted or blurry. Results Properties of Selected Marl Soils Table 2 1 provides a list of all the measured and observed soil properties for the three marl soils sampled to aid in identifying relationships between soil redox processes and inherent soil properties The three soils are very similar in most properties, except for % silt, bulk d ensity, OM electrical conductivity ( EC ) structure and live macro fauna Soil A has less silt and more
47 sand than the other soils, which may be due to the noticeable presence of mollusk shell fragments that did settle out as sand sized particles in soil te xture determinations by the hydrometer method Dissolution of calcium carbonates before determining soil texture would have dissolved shell fragments, but was not done to prevent the dissolution of the calcitic marl soil itself. The presence of a plow laye r in Soil B prompted the decision to test bulk density ( Db ) in the top two layers of each soil, with three replicates The compacted layer (15 to 30 cm) in Soil B had a Db of 1.1 g/cc, compared to 0.65 and 0.89 g/cc in Soils A and C, respectively. All thre e soils were below the 1.3 g/cc bulk density of a mineral silt loam soils and Soil A aligned more closely with D b values of 0.5 g/cc measured in organic muck soils used for farming in the northern Everglades (Wright and Inglett, 2009). Table 2 1 lists So il A as having 8.3% organic matter, twice the amount measured in Soils B or C which was unexpected since Soil A had color values of 7and 6 in the upper 30 cm and sparse vegetation adjacent to the soil sampling site Soils B and C were noticeably darker in appearance with color values of 5 in the upper 30 cm One possible explanation of the light color in Soil A is that precipitating CaCO 3 from active marl formation results in organic particles being coated or masked In the Bahama Banks, 24 0 kilometers sou theast of the study area, researchers used a scanning electron microscope on actively forming calcitic soils to identify 0 to 5% organic materials oriented in thin concentric films with carbonate precipitates in shallow coastal waters (Davies et al., 1978) E C in Soil A revealed this site to be in the upper end of th e slightly saline range of 700 to 2000 uS /cm (R hoades et al. 1992), which is 2 to 3 times higher than in Soils B and C, respectively Increased salinity in Soil A is most likely due to its loca tion in a coastal wetland
48 area subject to saltwater intrusion and windblown salts during tropical storms and hurricane events The shores of Florida Bay are less than 2 kilometers east of where Soil A was collected, and well within the boundaries of the sa ltwater intrusion line, which extended 6 kilometers westward from the coast at the time it was delineated (Shonenshine, 1995). Deionized water used to maintain the water table treatments showed no changes in electrical conductivity or pH with time and did not contribute to any changes within the mesocosms (data not shown). Examination of the soil from each site revealed that Soil C possessed subangular soil structure, where Soils A and B had no discernible structure (massive) Soil structure provides aera tion and allows water to infiltrate and permeate at a faster rate than massive soils lacking structure Figure 2 3 shows significant burrowing by wireworms in Soil B that excavated tunnels to the soil surface shortly after water table treatments were estab lished at the 15 and 30 cm depths in the mesocosms Wireworms are controlled in agricultural fields by continuous flooding for 42 or more days ( Hall and Cherry 1993) and were not apparent in mesocosms with water tables maintained at the soil surface A si ngle mole cricket surfaced in mesocosm B6 4 after establishing the water table at 15 cm and was promptly removed, but had been burrowing below the soil surface since extraction more than 21 days earlier. Water T able T reatment E ffects on E lectro chemical So il Properties Electrical conductivity in soils A, B and C increased by 25.6%, 10.9% and 15.1%, respectively as water table depth s decreased from 30 cm to 0 cm as show n in Figure 2 4 Soil A ha d the largest increase of ~600 uS /cm between the 30 cm and 0 cm water table depths. Higher salinity in Soil A is likely attributed to its close proximity to the saline coastal waters of Biscayne Bay. In contrast, Soil B and C receive in land freshwater sources derived from an annual rainfall amount of 145cm and irriga tion water pumped from the Biscayne aquifer (USDA NRCS, 1996).
49 Water table treatment depths had little impact on pH levels (Figure 2 5 ) wi th pH variations of +/ 0.1, and within the standard error ranges for all but Soil B Previously drained, alkaline so ils tend to decrease in pH with time due to the build up of CO 2 gases that produce carbonic acid as a consequence of iron reduction (Mitsch and Gosselink, 2000) Marl soils possess such extreme buffering capacity (Zhang et al., 2002) that small amounts of carbonic acid would have little ability to alter pH in these soils High pH resulted in requirements for lower reduction potentials to achieve the reduction level set in the Hydric Soils Technical Standard (NTCHS, 2007). Over 56 days 18 pH measurements w ere taken from each of the mesocosms to coincide with the collection of redox readings for comparison against the standard redox potential A s trong correlation (0.21 P<0.01 ) was identified between pH and soil with time (Table 2 2 ), but changes in pH w ere small (between 7.7 and 8.1) during the 56 day time span, regardless of soil type or water table depth (Figure 2 5 ). The primary effect of high pH was the requirement of standard redox values less than + 175 mV (at pH of 7.0) to verify reduced conditions. A naerobic Conditions as Determined by Redox Potential In this study, initial redox readings were taken from the Pt electrodes on the same day water table treatments were established in the soil mesocosms Even though all soils had been drained of gravitatio nal water 7 days before the experiment began, high variability of the initial redox potential was apparent in all soil mesocosms regardless of soil type The highest redox potential ( Eh ) reading on day 0 had a high of + 635 mV in Soil C with a 30 cm water t able depth to a low of 54 mV in Soil A with water table at the soil surface, revealing a wide range of soil moisture contents before establishing water table treatment depths Initial readings were taken 6 hours after water tables were established so that other soil properties such as saturated hydraulic
50 conductivity and capillary water movement (Migliaccio et al., 2008) could have contributed to the wide range of the initial Eh readings. Eh values dropped ~400 mV in the first 7 days in all soils except S oil C with the 30 cm water table treatment (Figure 2 6) During the first 24 hours Eh in Soil C drops by just less than 200 mV, but by day 2 the Eh rebounds upward slightly until leveling off in a highly aerated zone of + 400 mV by day 14 Redox potentials in Soils A and B also dropped quickly once water was applied to the mesocosms to establish a 30 cm water table depth, but unlike Soil C Soils A and B did not rapidly rebound into the aerobic zone Following initial establishment of water table depths, sma ll daily additions of DI water were necessary to maintain constant 30cm water table levels Soils A and B did not come back above the standard redox line to an aerobic state until after 51 and 58 days, respectively T he presence of subangular soil structu re in Soil C is a characteristics that s eparates Soil C from the other two soils and likely to impact hydraulic conductivity (Table 2 1 ) Soil C is not in continuous production of field grown ornamental shrubs and trees with a persistent ground cover of gr ass and herbaceous weeds; all conditions favoring the development and persistence of soil structure Soils A and B both lacked structure and the ability to drain quickly, but from different conditions Soil A is still actively forming with ongoing CaCO 3 pr ecipitation and lacking the time necessary for a soil to develop structure Soil B is annually tilled with heavy equipment resulting in a compacted layer that could explain the lower Eh readings in Figure 2 7 than for Soil A with the same 15 and 30 cm wate r table treatment depths. Compacted soils or plow pans are known to result in significant reductions in hydraulic conductivity, permeability and air diffusion (Horn et al., 1995).
51 Figure 2 6 shows all three soils had average redox potentials in the aerobi c zone on the day that water tables were established at a 15 cm depth ; and all more than 200 mV high er than anaerobic standard lines Soils A and B both had rapid drops in average Eh and moved into the anaerobic zone with + 81 mV readings by day 3 for Soil A, and by day 5 for Soil B Soil C was relatively slow to reach a reduced state with an average redox reading of + 43 mV after 11 days, then continuing on a slow, steady decline so that by day 44 all three soils coincide around a median Eh of 210 mV. The p latinum electrode collects redox potential readings from a fixed depth of 25 cm, 10 cm below the 15 cm water table treatment depth, where anaerobic conditions were expected to develop Review of the three 15 cm water table repetitions for Soil C, rather th an the average, showed delayed responses of 9, 11 and 22 days before reaching the anaerobic zone As with the 30 cm water table treatment, soil structure for Soil C provides a likely explanation for the difference in response time to soil saturation Space s between soil aggregates in the subangular structure that contained O 2 may have been entrapped by the downward moving DI water, diffusing slowly into the soil water, then utilized as preferential oxidants by soil bacteria. Figure 2 6 shows a similar respo nse time to saturated conditions for each of the three soils observed in the 0 and 15 cm water table treatment In flooded conditions, Soil A started out in a reduced state since that soil was kept moist right up until drained 7 days before the water table s treatments depths were applied Marl soils in field conditions will develop surface cracks from shrinkage as they dry out Soil B started out at a higher Eh value than Soil C then rapidly dropped 250 mV to reach a reduced state after only 2 days while So il C took 11 days to drop 280 mV and arrive at a reduced state A slow, steady reduction occurred in the three soils so they
52 merged on day 37 at an average Eh of 228 mV and remained there for the final 21 day s of monitoring. At the 0 and 15 cm depth water table treatment depths, all soils were expected to reach a reduced redox value shortly after water tables were established based solely on the corresponding 25 cm depth of platinum electrode tips Whether the soils would have a reduced redox potential due to the 30 cm water table treatment was questionable since the water table would be maintained just 5 cm below the tip of the Pt electrode, providing an opportunity for O 2 to occupy soil voids adjacent the Pt electrode Figure 2 8 shows that Soil C with al l three water table treatments had the highest redox values of the three soils and maintained aerobic conditions with the 30 cm water table treatment throughout the entire 56 day sampling period The two soils (A and B) lacking any discernible soil structu re displayed their water holding capacities were sufficient to develop saturated soil conditions within 5 days Once Soils A and B were reduced, the se soils maintained that reduced state for more than the 14 consecutive day s resulting in positive confirmat ion of anaerobic soil conditions according to the Hydric Soil Technical Standard (NTCHS, 2007). Table 2 2 shows strong correlation between redox readings (Eh Electrode) and soil, water table and time (days) Figure 2 9 displays the relation between Eh va lues and water table depths with time for each soil Soil C shows initial fluctuations in Eh, but then follows a constant, downward movement with time The 30 cm water table Soil C maintained a high, steady aerobic state during the 56 days, compared to Soi ls A and B with fluctuations in the reduction pattern during the first 21 days Reactions of Soils A and B were comparable to each other in terms of initial and final Eh values and the general pattern of Eh fluctuations throughout the 56 days.
53 Determinatio n of Anaerobic Conditions with FIRIS FIRIS tubes are currently accepted for use by the N TCHS as an alternative to Pt electrodes for determining anaerobic conditions, a prerequisite to classifying soils as hydric At present FIRIS tubes are recommended to b e installed for 28 days and require that at least 30% of the iron oxides be removed from a 15 cm long section that must start within 15 cm of the soil surface (NTCHS, 2007) In this study, FIRIS were compared to Pt electrode performance as well as MIRIS, w hich resulted in the selection of three extraction intervals of 14 days 28 days and 56 days to provide a comparison of each tool at those specific intervals Unlike Pt electrodes that can continue to collect data for long periods, IRIS tubes can only refl ect the soil conditions that exist between initial installation and the day of extraction since they are not reinstalled upon exposure to an aerated environment. The first e xtraction of FIRIS tubes after 14 days showed an average of less than 1% iron oxide removal in all soils for both the 15 cm and 30 cm water table treatments (Figure 2 1 0 ). FIRIS reaction in flooded conditions with the water table at 0 cm depth during that same time showed measureable oxide removal. The results varied by soil type with So il A showing 0% removal while Soil C displayed 8% oxide removal on that initial extraction This gradation of performance between Soils A, B and C continue with the next two extractions at 28 days and 56 days Electrical conductivity in Table 2 1 is the on ly measured soil property that follows the oxide removal pattern across the three soils. Soils A and B are within the slightly saline range of 700 to 2000 uS /cm, while Soil C is < 700 uS /cm and considered non saline (Rhoades et al. 1992 ). Table 2 2 shows no significant correlation between FIRIS performance and EC readings. Studies into energy harvested from underwater reduction do show that bacterial species do vary across the salinity of classes of fresh, brackish or saline water bodies (Holmes et al., 20 04).
54 Research into the direct effects of salinity on iron oxide reducing soil bacteria species or their performance was not part of this study nor could research this specific be located, indicating this could be an indirect factor in iron oxide reduction and removal. The primary difference between FIRIS and Pt electrode performance was that the electrodes captured Eh readings showing favorable conditions for Fe 3 + reduction allowing the document ation of anaerobic conditions in as little as 14 days in floode d conditions. In Table 2 3 FIRIS in flooded conditions did not meet the minimum 30% oxide removal until sometime between the extraction on day 2 8 with 24% removal and the final extraction on day 56 with 48.2% oxide removal. Figure 3 1 1 disp lays the longer of reaction time of both IRIS tubes as compared with the Pt electrodes that reflect ed anaerobic Eh values within 7 days Redox measurements are able to capture reduction potentials immediately but oxide reduction and removal from the PVC tubes takes longe r. F IRIS technology d id reflect the same redox trend as Pt electrodes, but at a slower rate in the marl soils utilized for this study Determination of Anaerobic Conditions with MIRIS MIRIS were developed building upon FIRIS manufacture and research to date This study represents the first comprehensive redox investigation to include MIRIS since its development and field test in Alaska during the summer of 2010 (Stiles et al., 2010) At this time the same Hydric Soils Testing Standards listed for FIRIS perfo rmance (NTCHS, 2007) will be applied to MIRIS In Table 1 3 Mn 4 + reduction is measured to occur before Fe 3 + reduction at corresponding higher Eh values, indicating a conservative approach in the application of the iron oxide removal standards If MIRIS te chnology is expanded and studied further, a set of evaluation standards separate from FIRIS may be prescribed. MIRIS installed in the 15 cm and 30 cm water table treatments showed the same lack of oxide removal (Figure 2 1 2 ) as displayed in the FIRIS insta lled with the same set of variables
55 Based on visual estimates, averages of 0.7% and 0.9% of manganese oxides were removed from the 54 MIRIS installed at the 15 cm and 30 cm water table depths, respectively Under the flooded soil conditions displayed in F igure 2 1 3 there is a steady increase in oxide removal with time for each soil type, but only MIRIS installed in Soil C achieved >30% manganese oxide removal by the final extraction on day 56 Table 2 4 affirms the significant relationship MIRIS has to wa ter table treatment, soil type and the length of reaction time The response pattern of MIRIS and FIRIS to water table levels and time are almost identical, with a comparative removal ratio of 0.29 manganese oxides for every to 1.0 iron oxide removed, as a veraged from the full 56 day period in the 0 cm water table treatment From the comparative removal ratio, it could be projected that 71% more time is required for manganese oxide removal than iron oxide removal in south Florida marl soils in flooded condi tions which would be close to 90 day s. Comparing MIRIS with the standard Pt electrode in Table 2 5 there is low or no significance between these two tools, depending on the analysis method used to calculate the percent oxide removal from the tubes Figure 2 1 1 shows manganese oxide removal increases gradually and continually through final extraction on day 56 For that same period, we can see Eh readings dr o p by an average of 400 mV during the first 14 days, continue to drop slightly, and then level off by day 29 This pattern show that Pt electrodes are quick to confirm saturated conditions, and when sustained can confirm anaerobic conditions with in 14 days Looking at Table 2 3 we can see that within the data collection period of 56 days the Pt electrodes verified anaerobic soil conditions across all three water table treatments by maintaining reduced conditions (< + 175 mV) at the 25 cm depth for least 14 consecutive day s The exception was Soil C with the 30 cm water table treatment, which accounts for the aerobic
56 Eh readings (> + 175 mV) in Table 2 3 The IRIS were also able to confirm anaerobic soil conditions within the 56 days, but only for the 0 cm water table treatment with 33% of FIRIS and 4% of MIRIS having > 30% oxide removal. IRIS installed in soils with 15 and 30 cm water tables did not reflect the anaerobic conditions shown in the Eh readings during the same 56 days The progressive pattern of reduction conveyed in Figure 2 1 1 demonstrates that both FIRIS and MIRIS react to reduced conditions in so uth Florida marl soils, but at very different rates FIRIS with the same treatments variables of soil, water table and time are able to reduced iron oxides at an average of 3.4 times faster than occurs for manganese oxides reduction and removal from MIRIS tubes. Comparison of IRIS Analysis M ethods Once tracing and scanning wa s accomplished, the image wa s imported into Digimizer 4.0 by MedCalc software for binary image analysis (Figure 2 14 ) T he per centages of areas with data (dark areas) and without (white areas) were calculated to determine the percent oxides removed from 18 IRIS tubes The areas without data represent areas where oxides were removed from the tube. The three tubes in Figure 2 15 were recorded as having 0% oxide removal, although iron oxide was removed in a zone of inundation below the 30 cm depth which cannot be included in the anaerobic soil determination Table 2 6 shows the 3 MIRIS and 18 FIRIS that met the selection criteria for further analysis by trace and scan digital methods. All I RIS tubes in the T able 2 6 were treated with a 0 cm water table depth, which shows strong correlation in Table 2 2 for all methods, but not as significant for MIRIS tubes estimated by either of the digital analysis techniques. Figure 2 16 shows the lack of manganese oxide removal in all soils but Soil C. A small data set of only three MIRIS tubes were digitally analyzed, so standard error values are more pronounced for the tra ce
57 and scan methods. In Figure 2 17 we can see that iron oxide removal is also mos t abundant for FIRIS installed in Soil C, and to a lesser extent in Soils A and B. The small data set available for MIRIS tube analy sis by digital methods result ed in almost perfect correlation in Figure 2 1 8 where the three data points come from one Soil C mesocosm with pronounced removal of manganese oxides in flooded conditions In Figure 2 1 9 there are 18 data points from seven soil mesocosms representing the three soil types treated with the 0 cm water table The larger and more diverse sample set for the FIRIS, provides additional confidence in the correlation data shown in Table 2 5 Visual methods have the lowest correlations but the differences were comparatively small and shows that either visual, trace or scan methods can be successfully utilized to provide reliable estimates of oxide removal from IRIS tubes. Discussion General characteristics of the marl soils selected for this study were similar in texture, pH, and nutrient content. Land use and ecological setting of each marl soil resulted in m oderate variations in physical and chemical properties. Soil salinity of Soil A, adjacent to a coastal wetland system, had an average EC of value of 1777 u S /cm, that places it in the upper end of the slightly saline catego ry with a range of 700 to 2000 uS / cm (Rhoades et al. 1992). Soil B had 53% lower EC readings than in Soil A, but still in the lower end of the slightly saline category. Soil C was classified as freshwater with 71% lower EC value as found in Soil A. There was significant correlation betwee n EC and redox potentials identified from data analyzed by SAS. Another characteristic of Soil C that set it apart was a subangular soil structure that developed under the production of ornamental trees and shrubs without tillage of the soil. Structure is impacts aeration, hydraulic conductivity, and drainage that influence saturation, and redox potentials. The correlation of soil structure and Eh were found to be just as significant as
58 identified between EC and Eh. In general, Eh increases with salinity an d aeration meaning that higher Eh readings in Soils A and C are expected as displayed in Figure 2 7. Vertical capillary water movement and dissolved oxygen content of marl soils are two properties that could explain the redox patterns observed in this stud y. Pt electrodes, FIRIS, and MIRIS performance were compared in their ability to reflect redox state of the soil through Eh readings, iron oxide removal or manganese oxide removal. The Pt electrodes performed consistently during the sampled 56 days and dis played the distinct advantage of providing measured redox potentials in real time, allowing for rapid assessment of reducing and anaerobic conditions. In flooded conditions, Eh readings were able to provide confirmation of anaerobic conditions after just 1 4 days, which met the Hydric Soils Testing Standards (NTCHS, 2007). A major disadvantage of Pt electrodes is the high cost to purchase or manufacture with platinum. Field installation of electrodes requires significant equipment set up and protection measu res, both limit the number of data collection points that can be analyzed at any one time. FIRIS performed as expected with a delay in iron oxide reduction associated with high pH, calcareous soils. With flooded soil conditions, iron oxides are readily red uced and could be confirmed with the last of three extractions (Figure 2 20) 56 days after installation. When water tables were at the 15 and 30 cm depths, the required 30% removal of oxides from the upper part of the FIRIS did not occur by the end of the 56 days. For use in alkaline soils with pH values of 7.5 to 8.5 the FIRIS require additional time to reduce iron oxides, and 56 days instead of the current 28 day installation time is recommended (NTCHS, 2007). In 2009, MIRIS tubes were developed for use in high pH alkaline soils where FIRIS tubes are known to perform slowly. A rapid field test in the summer of 20 09 in alkaline soils of Alaska
59 had >30% manganese oxide reduction and removal from MIRIS within one week and comparable results were expected in this study Of the 81 MIRIS tubes installed in this study, only one had > 30% manganese oxide removal > 30% (54%), with 27 of the 81 tubes installed in soils that were flooded (0 cm water table) throughout the 56 days sampling period. The poor performance of MIRIS in this study is not fully understood from the data collected. During the manufacture of the MIRIS it was obvious that the manganese oxide paint is difficult to apply uniformly and does not adhere to the PVC tube as well as in the FIRIS tubes, but those issues alone do not explain the poor reduction results. Additional microbial studies into the manganese reducing Geobacter species could help to determine if manganese reducing bacteria are present in sufficient populations to reduce manganese oxide s, as manganese is not present in the natural systems of south Florida. Once testing of IRIS concluded, the amount of oxide removal was determined using three different techniques. The visual, trace and scan methods were applied to select FIRIS and MIRIS t ubes and found to correlate well with similar estimates in oxide removal across all three soils indicating that any of the techniques could be applied to provide reliable estimates of oxide removal. Deciding which IRIS analysis method(s) to use will likely depend on the end use of the data, the number of tubes to be processed and an assessment of oxide removal extent upon extraction. If processing time or ease of use is a primary concern, then visual estimates should be used When results are likely to be c hallenged or to fall close to the 30% removal standard (NTCHS, 2007), then verifying results by more than one method may be desired Digital analysis can provide quantitative documentation directly from a scan of the IRIS tube surface. Direct scanning can be fast when image editing is not required before analysis Tubes with linear
60 scratches or other features that should be edited would benefit from the trace method by selectively excluding unrepresentative features to improve accuracy before binary image a nalysis proceeds Table 2 1 Characteristics of three marl soils selected for this study Characteristic Soil A Soil B Soil C Taxonomic class Typic fluvaquents Typic fluvaquents Typic fluvaquents Map unit name Pennsuco marl Perrine marl, drained Pennsuc o marl, drained % Sand 40.30 34.16 34.45 %Silt 28.11 42.51 41.41 % Clay 31.59 23.34 24.14 Bulk density (g/cm3) 0.54 0.98 0.82 Structure Massive Massive Subangular blocky % Organic matter 8.30 3.86 4.60 CEC (meq/100 g) 21.28 14.58 15.47 Total Na (mg /L) 1237.16 850.96 901.13 Total Cl (mg/L) 209.28 146.70 154.12 pH 7.91 7.83 8.12 EC ( uS /cm) 1777.00 843.00 517.00 % CaCO 3 71.51 81.64 82.88 Total Ca (mg/L) 1452.99 1586.79 1555.00 Total N % 0.30 0.13 0.20 Total C% 14.30 13.14 13.46 Total Mn (mg/L) 0.03 0.86 0.17 Total Fe (mg/L) 5.61 7.06 3.53 Soil insects observed None Wireworms None
61 Table 2 2 Correlation coefficients among water table, time, insects and soil properties Soil type Water table Days Soil insects pH Electrical conductivity Structure Eh electrode 0.35** 0.44** 0.31** 0.24** 0.21** 0.33** 0.41** FIRIS visual 0.22* 0.52** 0.29** NS NS NS 0.23* FIRIS trace NS 0.57** 0.30** NS NS NS NS FIRIS scan NS 0.59** 0.28* NS NS NS NS MIRIS visual 0.25* 0.39** 0.36** NS NS N S NS MIRIS trace NS 0.31* NS NS NS NS 0.30* MIRIS scan NS 0.31* NS NS NS NS 0.30* P < 0.05, ** P < 0.01, NS = no significant correlation; Statistical Analysis Software Table 2 3 Anaerobic conditions determined by platinum electrodes iron Indi cators of Reduction in Soil (FIRIS) and manganese Indicators of Reduction in Soil (MIRIS) in 3 water table treatments
62 Table 2 4 Analysis of variance
63 Table 2 5 Correlation coefficients among various methods of determining redox potential
64 Table 2 6 Comparison of results using different techniques to estimate oxide removal from FIRIS and MIRIS tubes IRIS tube label % Removal visual method % Removal trace method % Removal scan method Median (%) Average (%) C0 1_Mn1 1.0 1.0 0.3 1.0 0.8 C0 1_Mn2 16.0 21.0 21.0 21.0 19.3 C0 1_Mn3 54.0 60.0 60.0 60.0 58.0 A0 1_Fe2 16.0 21.0 20.0 20.0 19.0 A0 1_Fe3 30.0 41.0 40.0 40.0 37.0 A0 2_Fe2 10.0 15.0 34.0 15.0 19.7 A0 2_Fe3 40.0 78.0 77.0 77.0 65.0 B0 1_Fe1 5.0 3.0 9. 0 5.0 5.7 B0 1_Fe2 10.0 13.0 12.0 12.0 11.7 B0 1_Fe3 60.0 60.0 59.0 60.0 59.7 B0 3_Fe2 5.0 7.0 5.0 5.0 5.7 B0 3_Fe3 33.0 47.0 39.0 39.0 39.7 C0 1_Fe1 15.0 27.0 34.0 27.0 25.3 C0 1_Fe2 80.0 87.0 84.0 84.0 83.7 C0 1_Fe3 85.0 99.0 100.0 99.0 66.7 C0 2 _Fe1 4.0 4.0 5.0 4.0 4.3 C0 2_Fe2 15.0 23.0 24.0 23.0 20.7 C0 2_Fe3 50.0 52.0 49.0 50.0 50.3 C0 3_Fe1 5.0 15.0 21.0 15.0 13.7 C0 3_Fe2 60.0 66.0 63.0 63.0 63.0 C0 3_Fe3 99.0 81.0 78.0 81.0 86.0 Alphanumeric labels for IRIS tubes explained : 1 st charac ter of the label reveals the soil (A, B, or C ); 2 nd character is the water table treatment depth (0 15 or 30 cm); 3 rd character is the mesocosm number for the order they were collected in the field (1 9); 4 th character indicates the type of IRIS (Fe = iro n, or Mn = manganese); the last character indicates the time the tubes were installed in the soil before extracted for analysis (1 = 14 days, 2 = 28 days, and 3 = 56 days)
65 Figure 2 1. Map of soil collection sites in southeastern Miami Dade County, Flo rida
66 Figure 2 2 Schematic of redox equipment as installed in soil mesocosms
67 Figure 2 3 Wireworm excavation in Soil B mesocosm
68 Figure 2 4 Changes in electrical conductivity associated with water table treatments in each marl soil Figure 2 5 Average pH response to water table treatments in three marl soils
69 Figure 2 6 Corrected Eh measured by platinum and reference electrodes. Comparison is by water table treatment depths o f 0 cm 1 5 cm and 30 cm over 56 consecutive day s. St andard redox lines covering the pH range show aerobic conditions above the line or anaerobic when below
70 Figure 2 7 Av erage Eh for each marl soil in varying water table treatment depths
71 Figure 2 8 Distribution of Eh values above and below the sta ndard redox line (NTCHS, 2007). Three soils were treated with water tables maintained at 0 cm, 15 cm or 30 cm depths for 56 consecutive day s.
72 Figure 2 9 Temporal relationship between Eh and degree of saturation in each soil type Soil A is Pennsuc o marl ; Soil B is Perrine marl, drained ; and Soil C is Pennsuco marl drained.
73 Figure 2 1 0 Iron oxide removal from I ndicators of R eduction in Soi l (IRIS) installed in three water table regimes Estimates by three different methods are notated by the pr e fix letter on FIRIS label s in the legend: v isual estimate (V ), trace method (T ) and scan method (S ) visual estimate
74 Figure 2 1 1 The measured response of redox tools to water table treatments with time A verages were generated from the three soils and the three water table treatment depths Oxide removal estimate calculated using the visual estimate and notated by the pre fix letter on IRIS label s in the legend: v isual estimate (V ).
75 Figu re 2 1 2 M anganese oxide removal from IRIS installed in three water table regimes Estimates by three different methods are notated by the pre fix letter on MIRIS label s in the legend: v isual estimate (V ), trace method (T ) and scan method (S ) visual es timate
76 Figure 2 1 3 IRIS performance in marl soils with flooded conditions. A) FIRIS tubes, B) MIRIS tubes with m etal o xide removal calculated from the averages for the three 0 cm water table treatment repetitions
77 Figure 2 14 Comparison of di gital image techniques used to analyze an IRIS tube. A) T raced image using acetate and permanent marker. B) Digitally s canned tube C) B inarization of the scanned image in preparation for digital analysis. 0 15 cm indicates the zone of removal, which began at the soil surface.
78 Figure 2 15 Iron oxide removal on IRIS tube s treated with the 30 cm water table Depth of the water table is indicated by the dashed line and shows no removal above that line.
79 Figure 2 16 M anganes e oxide removal from IRIS tubes as estimated by three methods. Method is notated by the pre fix letter on the M IRIS label in the legend: visual estimate (V ), trace method (T ) and scan method (S ). Figure 2 17 Iron oxide rem oval from IRIS tubes as estimated by three methods The m ethod is notated by the pre fix letter on FIRIS label in the legend: v isual estimate (V ), trace method (T ) and scan method (S ).
80 Figure 2 18 Correlation of three methods used to determine mang anese oxide removal. MIRIS extracted from a Soil C mesocosm treated with 0 cm water tables maintained for 56 consecutive days are compared Linear regression line and R 2 value s provided. Note: Only one set of manganese tubes had oxide removal in a quantity significant enough to warrant digital analysis by the trace and scan methods.
81 Figure 2 1 9 Correlation of the three methods used to determine iron oxide removal FIRIS from the nine soil mesocosms treated with a 0 cm water table and maintained for 56 consecutive day s are compared. L inear regression line and R 2 value s provided.
82 Figure 2 2 0 Iron and manganese coated IRIS tubes after receiving the same treatment s Each set of 3 w as installed in a Soil A mesocosm with a 0 cm water table depth
83 CHAPTE R 3 SUMMARY AND CONCLUSI ONS Introduction Expectations were that iron coated Indicators of Reduction in Soil ( FIRIS ) would not reduce within the recommended installation time of 28 days and the manganese coated Indicators of Reduction in Soil ( MIRIS ) would reduce within 14 days. In this study, the hypothesis was not proven with the iron oxides reducing 3.4 times faster than the manganese oxides. Soil structure and salinity were the two soil properties that had positive correlation to redox potential increase s in the data, with the latter potentially impacting anaerobic microbe reduction rates. Results indicate an extended installation time for FIRIS, additional testing of MIRIS, as well as confirmation of the reliability of plati num ( Pt ) electrodes. Objective 1 The first objective of this research was to evaluate the performance of the recently developed Indicators of Reduction in Soils (IRIS) tubes against the established, conventional method of measuring the soil redox potential using Pt and reference electr odes Once flooded conditions were applied (0 cm water table) the Pt electrodes averaged 5.2 days to reflect anaerobic reduction potentials, by dropping below the pH dependent standard reduction line ( + 175 mV at pH 7, 60 slope) Real time redox readings ( Eh ) from the Pt electrode were able to verify anaerobic conditions 22.8 days after flood ing With the same treatments, average iron oxide removed from FIRIS was estimated at 5.2% after 14 days 23.8% after 28 days and 48.2% after 56 day s of flooding Manga nese oxide removal from MIRIS was lower than expected at 1.2% after 14 days 6.4% after 28 days and 17.4% after 56 day s of f looding. Applying the H ydric S oil T echnical S tandard iron oxide removal to both types of IRIS,
84 anaerobic condit ions could be verified on FIRIS from the extraction at 56 days ; the MIRIS failed to reflect anaerobic conditions during the 56 days with flooding The results support the use of FIRIS for inundated conditions with an installation interval of 56 days which limit s the applicability since most hydric soil investigations are on sites saturated below the soil surface. In contrast, < 1% of oxides were removed from the upper part of either FIRIS or MIRIS in the 15 and 30 cm water table treatments conclu ding that in high pH, calcareous soils of south Florida, the most reliable method is the Pt electrode although it is not preferred for field investigations. Additional research into MIRIS paint formulation, similar to the work by Dr. M. Rabenhorst et al. on iron ox ide paint formulation, may improve reduction rates as well as the adhesion of the manganese oxide paint to polyvinyl chloride tubes during fabrication. Evaluation of MIRIS in other high pH, problem soils would be beneficial to identify use limitations and improve the technology at this early development stage of MIRIS tubes. Objective 2 D etermin ing what general soil characteristics of marl might influence redox processes was the second objective. Soil structure, electrical conductivity and pH were found to have significant (P < 0.01) correlation with Pt electrode readings. Electrical conductivity in this situation was a consequence of the proximity to coastal waters of Biscayne Bay Higher salinity in Soil A is the one characteristic that can explain the hig her Eh readings than found in Soil B with the same water table treatments. The aeration provided by soil structure expla ins why Soil C quickly rebound ed from water applications when the soil water tables were established for each soil mesocosm Soil C had an average Eh + 212 mV higher than in Soil A, and + 260 mV higher than in Soil B as an average from all water table treatments. The resistance to reduction provided by good soil structure could be promoted through reduced tillage practices and permanent cov er crops for low lying areas prone to subterranean flooding or sea level rise.
85 Research into anaerobic microbial species involved in metal oxide reduction would be beneficial in recognizing how impacts of land use or ecological communities may affect popul ations of microbes and their respiration rates. Soil salinity impacts to reduction rates in this study showed positive correlation with Eh resulting in less reduced conditions. The impacts to microbes in brackish and saltwater environments would be helpfu l to understanding IRIS use limitations in tidally influenced wetland systems. Objective 3 The final objec tive was to compare the results obtained from the visual, trace and scan methods of estimating the percent oxide removal from IRIS tubes. Visual estim ates of oxide removal were 17% and 20% l ess that when compared t o trace (R 2 = 0.88) and scan methods (R 2 = 0.84) respectively. The variance between the visual, scan and trace method results, indicate that oxide removal within +/ 13% of the soil determinations, depending on the method selected for IRIS analysis. The trace and scan digital methods corresponded most closely (R 2 = 0.97) with the trace method estimates 4% lower than the scan me thod If processing time and ease of use are primary concerns, then visual estimates can be used. If oxide removal r esults are close to the 30% removal criteria then evaluation by more than one method may be desired Digital analysis provides quantitative documentation directly from a scan of the IRI S tube surface. Direct scanning requires less time than the trace method when image editing is not necessary before analysis. The rotation device constructed for this project could be refined to allow portable h and /bar scanners to be readily adapted to scanning round IRIS tubes of variable diameters. The t race method pair s professional judgment with technology for digital analysis. Tubes with linear scratches or other features that require editing would benefit f rom the trace method by selective exclu sion of unrepresentative features
86 Deciding which IRIS analysis method(s) to use should be determined by the end use of the data, the number of tubes to be processed and the extent of oxide removal observed upon extr action. Additional study with IRIS tubes from other problem soils and from a large group of individuals performing visual estimates would be helpful to determine if findings in this study are applicable in other areas where IRIS used. Development of recomm endations for oxide removal estimate methods and the various software applications used for image analysis would further the standardization of IRIS technology for research purposes
87 APPENDIX A HYDRIC SOILS TECHNIC AL STANDARD Hydric Soil Testing Methods The Hydric Soil Technical Standard ha s been specified and approved for use by the National Technical Committee for Hydric Soils (NTCHS) in investigating the surface hydrology of soils with regard to anaerobic conditions and saturation Although other metho ds may exist, the methods discussed here are accepted by the NTCHS for the purposes of evaluating: wetland functions, hydric soil status, or field indicator criteria (NTCHS, 2007) Saturation and anaerobic conditions must both exist to prove a soil is hydr ic. Redox potential (Eh) measurements of < + 175 m V at pH 7 must exist to document anaerobic conditions and are adjusted for pH on a line with a slope of negative 60 ( Figure 1 3). Positive reaction to alpha alpha d ipyridyl through a pink or red color chang e reflects the presence of reduced iron in soils with sufficient iron present, but low iron content in marl soils does not provide the indicator reaction necessary to utilize this solution. Soils low in iron, such as marl, can be evaluated for redox potent ial using iron coated Indicators of Reduction in Soil ( FIRIS ) tubes Confirmation of anaerobic conditions in the field by Indicators of Reduction in Soil ( IRIS ) tubes requires that at least 3 of the 5 buried IRIS tubes have iron removed from at least 30% o f a 15 cm long area along the tube, which must start within 15 cm of the soil surface (NTCHS, 2007). Saturated conditions in situ can only be proven with piezometer data, which should include site specific water table data from an open well and site specif ic precipitation records Piezometers must be installed at 25 cm and 100 cm depths and read at least every 7 days The period of readings should cover a minimum cycle of dry wet dry (April through November in
88 s outh F lorida ) or up to one year If water occu rs within 25 cm of the soil surface for at least 14 consecutive days then the saturation criteria for a hydric soil has been met (NTCHS, 2007). When evaluating hydric soils status, b oth saturat ed and anaerobic conditions must occur more than 50% of the tim e, this timeframe might be >1 in 2 years, >5 in 10 years, etc The required precipitation data help to determine if the wet period of interest is occurring during a year with low, normal, or high precipitation rates when compared with historical data for a location (Sprecher and Warne, 2000) Historical precipitation data from 1971 2000 are available online from the U nited States Department of Agriculture N atural R esources C onservation S ervice, National Water and Climate Center In this research, water tabl es w ere managed in the laboratory. Collection of Oxidation Reduction Readings with Pt Electrodes Measuring the reduction and oxidation potential of a soil by taking conductance readings between platinum (Pt) and reference electrodes is an established metho d used in laboratory and field research When documenting soil saturation to meet hydric soil criteria, in addition to the electrodes a salt bridge(s) and a piezometer(s) are needed to document the water table (NTCHS 2007) Outside of a research lab this method is not practical for routine use by field soil scientists due to the installation of expensive equipment, labor requirement s, and exposure of equipment to hazards from weather conditions and wildlife, as well as vandalism or theft. When the use of Pt electrodes is employed, redox readings need to be collected at least every 7 days, since Eh fluctuate s with time. The voltage should be read at a depth of 25 cm from the mineral soil surface in loamy and clayey soils, which categorizes the marl soils in this research If a predominance of reduction continues for a minimum of 14 consecutive day s, then anaerobic conditions can be documented at the location of the electrode Since Eh varies
89 spa tially as well as with time, multiple electrodes are installed throughout the study area. Soil pH should be collected at the same time redox readings are taken to determine the correct ed Eh Figure 1 3 demonstrates that in anaerobic conditions iron reduction potential is achieved at high er Eh values than in non alkali ne soils. Ambient soil temperature to correct the reference electrode based on the type of electrode and the electrolyte used should be recorded at the time redox readings are taken. Interpretations of the readings are based on values obtained from using a standard hydrogen reference electrode (SHE), which is not suitable for field use. More durable electrodes can be used, but must include a correction factor to represent the expected value if a SHE had been used. Calomel or Ag/AgCl reference electrodes are the most common types used with differing correction factors of + 250 mV and + 200 mV Equations A 1 and A 2 explain the redox reactions for each of these electrode types Calomel electrodes contain mercury, which requires careful handling and disposal, so they were not considered for used in this study. Hg 2 Cl 2s + 2e 2Hg s + 2Cl aq ( A 1 ) AgCl s + e Ag s + Cl aq ( A 2 ) A small temperature correction factor is recommended when using saturated Ag/AgCl electrodes used at temperatures that vary f rom 25 C Increase in temperatures above 25 C are corrected by 0.13 mV/C, while temperatures below 25 C are adjusted using +0.13 mV/C ( Ansuini and Dimond, 1994) Under climate controlled lab conditions with an ambient air temperature set at 25 C, c orrection factors for temperature were not expected to be significant or necessary but were recorded.
90 APPENDIX B CONSTRUCTION OF EQUI PMENT Introduction Execution of research required 10 months of construction and testing of the equipment and components b ecause similar instrument s were not available on the market. D etail ed information for constructing component s is described in this chapter to assist future research and development of the Indicato rs of Reduction in Soil (IRIS). T he quality of the component s used in this research was integra l to the quality of the data collected and analyzed. Platinum Electrode s Thirty 45 centimeter platinum electrodes (Pt) were fabricated for the project due to lack of availability and to control costs. Electrodes were c onstructed with guidance from the methods and materials recommended by Vepraskas and Cox at North Carolina State University. The platinum wire and b rass rod were connected using electrical solder as they could not be fused with a propane torch since their melting points are too far apart at 1772 C and 882 C, respectively. The quality of the constructed electrodes was tested using an Ag/AgCl reference electrode filled with a 4.0 M saturated KCl solution. A ferrous Solution) of known conductance was the standard buffer solution with a target reading of + 475 mV, and a variance of +/ 20 mV was determined acceptable (Light 1972). D eionized water (D I ) water was also used to check for variance in a non buffered solution and ident ify any major variations. Electrodes were tested 5 times throughout 7 months, between initial construction and final installation. Considering the quality of the redox readings, the meter used to measure conductivity between the electrodes should have an i nput resistance of at least 20 giga ohms to minimize electrical feedback to reduce fluctuations in the readings (Rabenhorst et al., 2009). An
91 inexpensive handheld volt meter (Extech Auto Ranging Digital Multimeter Model: MN36) was used to complete initial bench tests on the electrodes. The day before installation into mesocosms the electrodes were tested using a Campbell Scientific Data Logger ( m odel CR1000), with a resistance rating of 20 giga ohms Salt Bridge s Good conductance between the platinum and re ference electrodes is necessary to obtain representative redox readings One of the three water level treatments to be used in this research will result in a 30 cm deep water table which is 5 cm below the Pt electrode tip at a 25 cm depth The other two wa ter level treatments will inundate the 25 cm deep platinum tip Installation of salt bridges in the mesocosms with the 30 cm water table was desired since the degree of soil saturation above the water table was not known, and variability in conductance cou ld affect the quality of the redox readings. Installation of salt bridges in all the soil mesocosms, instead of just those with the 30 cm deep water level treatment improve d the efficiency and reduced costs of collecting redox readings from the 27 soil mes ocosms Salt bridges eliminated the need to clean the reference electrode between each reading and facilitated the use of a single Ag/AgCl reference electrode for redox readings in all 27 mesocosms, which cut equipment costs by $2 600.00 An additional ben efit to using salt bridges in all the soil mesocosms was that the reference electrode would stay clean, as the need to clean soil from the reference electrode tip between readings was eliminated. Between July and August of 2011, 30 salt bridges were constr ucted from 61 centimeter lengths of 1.91 centimeter diameter polyvinyl chloride ( PVC ) pipe The bridges were constructed in the lab by dissolving a saturated KCl solution into an a gar powder that had been diluted in boiling, deionized water was allowed to cool slightly to form a pourable gel that sets up in the PVC tubes when cooled To minimize air pockets in the gel, liquid agar was pour ed at a 45
92 degree angle in a steady stream to allow the hot, liquid gel to displace air as each tube was filled To prev ent the agar gel from desiccation and cracking, the bridges were stored in a saturated KCl solution until ready for installation The day before installation the salt bridges were rinsed with DI water and dried to remove external salt deposits from the PVC tubes, and a permanent PVC cap was glued on the bottom of each Three 0.3 centimeter diameter holes were drilled around the circumference of the 1.9 centimeter PVC salt bridges at the 25 cm depth to expose the KCl agar gel at the same depth as the Pt el ectrode tip. Soil C olumns Soil columns were designed and constructed from PVC irrigation pipe materials for durability, light weight and non reactive nature of the material. Size of the columns was relatively large for soil mesocosms used in lab analyses, with a length of 61 centimeters and a diameter of 25.4 centimeters Using a bulk density of 1.41 g/cc for a silt loam soil (marl is typically silt loam ) and considering soil water weight each soil column weigh a pproximately 100 pounds. The large size and weight of the soil columns required additional planning of the sampling and transporting from the field to the lab. A primary concern during field collection would be to extract soil cores to avoid disturbing the soil layers and structure as much as possib le. The 25.4 centimeter diameter of the PVC pipe was necessary to accommodate the installation of 8 pieces of equipment into each soil mesocosm ; which included 3 iron coated Indicators of Reduction in Soil (F IRIS ) 3 manganese coated Indicators of Reductio n in Soil (M IRIS ) 1 Pt electrode and 1 salt bridge. The 61 centimeter le ngth of the PVC columns allowed for a 45 centimeter soil sample to be collected with enough space left accommodating water above the soil surface, which allowed the addition of water to establish and maintain the specified water table depths. The bottom rim of each PVC
93 pipe was beveled to reduce soil resistance and allow for more efficient penetration during soil sampling in the field. A 25.4 centimeter PVC cap was fitted with a PVC dr ain pipe and ball valve to control water levels in each column and to facilitate collecting water samples from the bottom of the column. A wooden cradle was built for each column to accommodate the plumbing appurtenances coming from the bottom center of ea ch column and to provide access for repairs if any leaks were to develop during the research. To monitor and adjust water levels, a 0.95 centimeter clear poly tube was installed vertically into the side of each column with a sponge filter in the elbow fitt ing to minimize movement of soil particles between the tube and the PVC pipe. To prevent soil particles from exiting the drain pipe during water sampling, 23 centimeter diameter by 2.54 centimeter thick sand filters were constructed of silica sand that was soaked in 1 M HCl for two hours, rinsed with DI water and air dried. The acid washed sand was placed into 30 fine mesh polyester bags that were fabricated to fit flush inside each PVC cap between the soil sample in the column and the drain pipe in the bott om center of the PVC cap. Manganese Oxide Paint With instruction from Patty Jones and Steve Monteith at the N ational S oil S urvey L ab (NSSL) the manganese oxide paint was manufactured at the T ropical R esearch and E ducation C enter facility in Homestead from February through May of 2011. Potassium permanganate was first treated with high levels of potassium to form acid birnessite for the final manganese substance (Stiles et al ., 2010) Birnessite is the most common form of mineralized manganese found in soils It is a poorly crystalline, tetravalent, oxide of manganese MnO 2 ) and readily synthesized in the lab (McKenzie 1971) Since the process to manufacture the manganese methods below describe the processes Three batches w ere produced to ensure enough of the
94 product would be available since it requires two months after manufactured before the paint is ready for application. Dialysis tubing was prepared for the leaching of oxides from the manganese paint. Five 61 centimete r sections of 12,000 to 14,000 molecular weight cut off ( MWC O ) dialysis tubing was heated to 80 C for 30 minutes in a solution of 0.01 mol/L of NaHCO 3 and 0.01 mol/L of sodium e thylenediaminetetraacetic acid ( NaEDTA ) with enough DI water to equal 1 L Once cooled, the tubing was transferred into an air tight glass container with a 50% ethanol soluti on for storage in the refrigerator until ready to be filled with the acid birnessite, produced in the next step. Acid birnessite was then prepared by dissolving 4.0 mol es of potassium permanganate ( KMnO 4 ) into enough DI water to equal 2.5 L of solution Th e solution was brought to a boil in a 3.5 L glass flask using a stirrer bar on a hot plate under a fume hood 163.5 ml of 12 M HCl was added to the boiling permanganate solution, 1 drop at time to prevent boiling. Equation B 1 shows the reduction of potass ium permanganate with the addition of the acid ( + H) and reduction of the chloride ion Acid solution was cooled overnight, allowing the acid birnessite precipitate to settle The decanted precipitate was transferred into four 250 ml centrifuge bottles, fil led with DI water and centrifuged at 2000 rpm for 30 minutes, decanted; repeated again After the second and final decanting, DI water was added to the centrifuge bottle and shook for 10 minutes. KMnO 4 + H + + Cl Mn 2 + + Cl 2 + H 2 O + K + ( B 1 ) Each of t he prepared dialysis tube was filled with the contents of the four centrifuge bottles; tubing was secured on each end with plastic clips and immersed in 2 to 3 gallons of DI water DI water immediately turned a bright violet/purple color on initial immersi on in the bath, but this dissipated after the initial soaking The DI water bath was changed every hour the first
95 day, and then twice a day until the salts were determined to have completed leaching by use of a silver nitrate indicator solution. Silver nit rate (AgNO 3 ) indicator solution was prepared by diluting 9g of AgNO 3 in 500 ml DI water This was applied with a disposable pipette adjacent to the submerged dialysis tubing to check for completion of salt leaching If salts are still leaching the indicato r solution turns cloudy white Staff at the NSSL indicated it should take about 7 days to complete leaching, but a negative AgNO 3 test was achieved after the 2 nd day for all three batches This could have been related to the larger quantity of DI water use d since the NSSL stab were using only 2 L of DI water, and 2 to 3 gallons of DI water were used each time the water baths were changed for this project due to the large size ( 49 liter s ) plastic tubs used for submersion of the dialysis tubes containing the birnessite. After consulting with Patty Jones at the NSSL on the premature AgNO 3 test I was advised to use observable changes to the dialysis tubing and the precipitate it contained Swelling of the dialysis tubing, followed by a squeaky balloon like text ure would develop and finally the precipitate would change from a purple black color to a dark brown color Over days 3 to 5 day, these indicators were all observed Daily testing with the AgNO 3 indicator solution continued to produce no reaction and confi rmed the leaching process had completed Each of the 3 batches was extracted from the dialysis tubes after 6 days in the DI water baths. The solid adduct remaining in the dialysis tubing was placed into 50 ml centrifuge tubes and processed at 2000 rpm for 5 minutes and decanted The adduct was rinsed two more times by refilling tubes with DI water, agitating on a shaker for 5 minutes, centrifuged at 2000 rpm for 5 minutes, and decanted On the final decanting, an equal volume of the liquid to solids was lef t in the tubes Each 50 ml centrifuged tube produced between 7.5 to 15 ml of adduct; 81 tubes were
96 processed The tubes were then labeled and stored in the refrigerator for 42 days (mid April to the end of May 2011) to stabilize the manganese paint to ensu re the proper conversion of hydrous oxides before fabrication of the actual MIRIS tubes (personal communication with Cynthia Stiles at the N S SL in Lincoln, Nebraska April 2011). IRIS Rotation Device for Digital Scanning After extracting MIRIS and FIRIS t ubes the percentage of oxide paint that has been removed from the PVC pipe must be measured. This can be done by two or more sets of human eyes, or the tubes can be scanned and digitally analyzed by an image software application. Flatbed scanners have been used for this purpose, but require taking multiple images from different sides of the tube, which then need to be cropped and merged to produce a single image for analysis (Rabenhorst et al ., 2008). Relatively inexpensive, portable, handheld/bar scanners are now widely available and had the potential to be used to analyze the extracted IRIS tubes. The idea of using a hand scanner for this research was suggested by MIRIS researcher, Ed Dunkinson, in July of 2011, but a device to adapt the scanner was needed The model available (Magic Wand by VuPoint Solutions; cost $100.00) for this research had a scanning window 0.5 cm wide that is 1.5 cm from the roller bar. The distance between the scanning window and the roller bar that activates the device to record th e image is to too far to produce a clear, undistorted image of either the 1.27 centimeter FIRIS or the 1.91 centimeter MIRIS tubes. To overcome this obstacle, a device with similar mechanics to a lathe was fabricated from wood, screws, a metal rod, and rub ber rollers off of a fax machine. An IRIS tubes is mounted onto a horizontal dowel that is perpendicular to second roller bar that rotates simultaneously with the dowel, allowing the IRIS tube to be centered in the scanning window while a 360 image of the entire circumference of the tube is recorded. The 25 cm by 3 cm bar scanner is place in a stationary cradle, centered above the IRIS tube so that the controls of the scanner are easily
97 accessible. The device could serve as a prototype for a more refined p iece of equipment to use with the handheld/bar scanners
98 APPENDIX C DETAILED SOIL DESCRI PTIONS Pennsuco marl (Soil A) with persistent shallow water above the surface, this site actively produces marl as part of a larger marl prairie ecological community Vegetation consisted of periphytic algae floating above the soil surface and sparse saw grass inundated with 30 cm of standing water, at the time of collection during the last week of July 2011 Here is a description of the upper 45 centimeters of Pennsuc o marl soil profile collected at L at itude 25.4347, Long itude 80.3555: 0 to 10 cm 10YR 7/2 clay loam; very small distinct organic bodies ranging in color from 10YR 4/2 to 10YR 2/1 in 15% to 20% of the matrix; common, fine roots; massive structure. 10 to 35 cm 10YR 6/2 clay loam; organic staining around few and very fine root channels; massive structure. 35 to 45 cm 10YR 4/2 loam; small shell fragments (< 1 mm) throughout matrix; few medium and fine roots; massive structure. Perrine marl, dr ained (Soil B) The perimeter of the site is drained by field ditches to allow annual production of tilled, irrigated, vegetable crops between December and April each year Sweet corn had been harvested from the field several months prior to sample collect ion, which was evident by the remaining corn stubble on the soil surface Limestone bedrock at the sample site was between 51 centimeters and 56 centimeters below the soil s urface and the water table was measured at 71 centimeters below the soil surface on the day of sample collection Here is a description of the upper 45 centimeters of the Perrine marl profile collected at L at itude 25.4644, Long itude 80.4616: 0 to 12 cm 10 YR 5/2 loam; massive structure 12 to 28 cm 10 YR 5/2 loam; massive structure; dens e plow layer from 12 to 22 cm 28 to 35 cm 10 YR 6/1 loam; CaCO 3 precipitation as small distinct concretions in 20% of matrix; massive structure
99 35 to 45 cm 10 YR 5/2 loam; CaCO 3 precipitation as small distinct concretions in 5% of matrix; massive s tructure Pennsuco marl, drained (Soil C). Historically, the site was part of a long, narrow drainage slough (a lso referred to as a marl finger) at a naturally lower elevation in the landscape. A review of aerial photos showed the during the 1960s the canal was excavated in the center of the drainage slough adjacent to this site, as part of a larger regional flood control project throughout central and south Florida. A 3 to 5% slope away from the canal bank indicates the site received marl spoil when the can al was excavated. Marl is more than 1.8 meters deep at this site, providing ideal conditions for the existing field nursery operation. Vegetation consists of a variety of field grown ornamental plants and palms with a permanent ground cover of volunteer gr asses and herbs between the plants and rows. Here is a description of the upper 45 centimeters of the Perrine marl profile collected at Latitude 25.4644, Long itude 80.4616: 0 to 10 cm 10YR 5/2; loam; weak, medium subangular blocky structure 10 to 38 cm 10 YR 5/2; loam; weak, very coarse subangular blocky structure 38 to 45 cm 10 YR 4/1 ; clay loam; moderate, medium subangular blocky structure
100 LIST OF REFERENCES Ansuini F.J., and J.R. Dimond. 1994. Factors a ffecting the a ccuracy of r eference e lectrodes. M ater. Perform. 33 ( 11 ) 14 17. ASTM 2000. Standard test methods for moisture, ash, and organic matter of peat and other organic soils. Method D 2974 00. American Society for Testing and Materials. West Conshohocken, PA. Castenson K.L., and M.C. Rabenhor st. 2006. Indicator of Reduction in Soil (IRIS): e valuation of a n ew a pproach for a ssessing r educed c onditions in s oil Soil Sci. Soc. Am. J. 70:1222 1226. doi:10.2136/sssaj2005.0130 Chen, M., L. Q. Ma, and W. G. Harris. 2002. Arsenic concentrations in Florida surface soils: Influence of soil type and properties. Soil Sci. Soc. Am. J. 66:640 646. doi:10.2136/sssaj2002.6320 Chen, M., L.Q. Ma and Y.C. Li. 2000. Concentrations of P, K, Al, Fe, Mn, Zn, Cu and As in marl soils from south Florida Soil an d Crop Sciences Society of Fl orida Proceedings 59:124 129. Chen M., L. Q. Ma, and W. G. Harris. 2002. Arsenic c oncentrations in Florida s urface s oils: Influence of soil t ype and properties Soil Sci. Soc. Am. J. 66:640 646 doi:10.2136/sssaj2002.6320 Davies P.J., B. Bubela, and J. F er guson 1978. The formation of ooids. Sedimentology. 25:703 730. Davis S, E. Gaiser, W. Loftus, and A. Huffman 2005. Southern m arl p rairies co nceptual e cological m odel Wetlands. 25:821 831. Dreimanis A 1962. Qu antitative gasometric determination of calcite and dolomite by using Chit tick Apparatus. J. Sediment. Petrol. 32 ( 3 ): 520 529. Gavlak R.G., D.A. Horneck, and R.O. Miller 2003 Soil, p lant and w ater r eference m ethods for the w estern r egion, 2 nd e dition W estern r egional e xtension publication no. 125. Oregon State University, Corvallis, OR http://isnap.oregonstate.edu/WERA_103/ / Methods/WCC 103 M anual 2003 Soil%20Sand Silt Clay.pdf (accessed 11 Dec. 2011). p. 129 31 Goh T.B, and A.R. Mermut. 2008. Carbonates. In: M. Carter a nd E. G. Gregorich, e ditors, Sampling and Methods of Analysis, 2nd Edition. CRC Press, Boca Raton, FL. p. 215 223. Ha ll, D.G., and R.H. Cherry. 1993. Effect of temperature and flooding to control the wireworm Melanotus communis (Coleoptera: Elateridae). Fla. Entomol. 76(1): 155 160.
101 Holmes D.E., 2004. Microbial c ommunities a ssociated with e lectrodes h arvesting e lectricity from a v ariety of a quatic s ediments. Microb. Ecol. 48 : 178 190. Horn R., H. Domialb, A. Slowihka Jurkiewiczb, and C. van Ouwerkerk 1995. Soil compaction processes and their effect s on the structure of arable soils and the environment. Soil Tillage Res. 35:23 36. Jenkinson B.J., and D.P. Franzmeier. 2006. Development and evaluation of i ron c oated t ubes that i ndicate r eduction in s oils. Soil Sci. Soc. Am. J. 70 : 183 91. doi:10.2 136/sssaj2004.0323 Light T.S. 1972. Standard s olution for r edox p otential m easurements Analytical Chemistry 44 ( 6 ): 1038 1039. doi :10.1021/ac60314a021 Loeppert R.H., C.T. Hallmark, and M.M. Koshy 1984. Routine p rocedure for r apid d etermination of s oi l c arbonates Soil Sci. Soc. Am. J. 48 (5) : 1030 1 033. doi:10.2136/sssaj1984.03615995004800050016x Mandal L.N., and R.R. Mitra.1982. Transformation of iron and manganese in rice soils under different m oisture regimes and organic matter applications. Pla nt Soil. 69 : 45 5 6. McKenzie R.M. 1971. The s ynthesis of b irnessite, c ryptomelane, and s ome o ther o xides and h ydroxides of m anganese Mineral. Mag 38 : 493 502. Migliaccio, K.B ., B. Schaffer, J. Crane and R. Muoz Carpena. 2008. Assessing capillary rise in a field nursery considering irrigation management Paper No: 083323 presented at Annual Meeting of the American Society of Agricultural Engineers Providence, RI. June 29 July 2, 2008 Ming, D. 2007. Carbonates. In : R. Lal editor, Encyclopedia of S oil Science, 2nd Edition. Francis and Taylor, New York NY p. 202 2 05. Mitsc h W.J. and J.G. Gosselink. 2000. Wetlands 3rd e dition John Wiley and Sons, Inc., Canada p. 155 204. NOAA. 1996. Florida Keys National Marine Sanctuary Final Management Pl an /Environmental Impact Statement, Vol. II of III. U S D ept of Commerce National Oceanic and Atmospheric Administration National Ocean Service Office of Ocean and Coastal Resource Management San ctuaries and Reserves Division. U.S. Government Prin ting Office Washington DC. p 12. NTCHS. 2007. The h ydric soil technical s tandard Technical n ote 11. National Technical Committee for Hydric Soils U S D ep artment of A gric ulture N at ural R esources C onserv ation S erv ice N at ional S oil S urvey C ent er Li ncoln, N E ftp://ftp fc.sc.egov .usda.gov/NSSC/Hydric_Soils/note11.pdf ( accessed 6 Jun e 2012) p 3 28.
102 Rabenhorst M.C., R. Bourgault, and B. R. James. 2008. Iron o xyhydroxide r educt ion in s imulated w etland s oils: E ffects of m ineralogical c omposition of IRIS p aints Soil Sci. Soc. Am. J. 72 : 1838 1842. doi:10.2136/sssaj2007.0368 Rabenhors t, M.C., W.D. Hively, and B.R. James. 2009. Measurements of s oil redox potential. Soil Sci. Soc Am. J. 73 : 668 674. doi:10.2136/sssaj2007.0443 Rhoades, J.D., A. Kandiah, and A.M. Mashal i. 1992. The use of saline waters for crop production: guidelines on water, soil and crop management. Chapter 2. Saline waters as resources. Food and Agriculture Organization of United Nations, Rome 1992. Table 1 http://www.fao.org/docrep/T0667E/t0667e05.htm ( accessed 17 Jun e 2012). Richardson J.L., and M.J. Vepraskas 2001. Wetland s oils g enesis, h y drology, l andscapes, and c lassification Lewis Publishers of CRC Press Boca Raton, FL. p. 19 28. Shonenshine, R.S. 1995. Delineation of saltwater intrusion in the Biscayne Aquifer, eastern Dade County, Florida. Water Resources Investigations Report 96 4 285. U.S. Geologic Survey Library Reston, VA. http://fl.water.usgs.gov/Miami/online_reports/wri964285/ index.html (accessed 6 June 2012). Sprecher S.W, and A.G. Warne. 2000. Accessing and u sing m eteorological d ata to e valuate w etland h ydrology Army Engineer Research and Development Center, Vicksburg, MS. http://el.erdc.usace.army.mil/elpubs/ pdf/wrap00 1/contents.pdf ( accessed 7 Jul y 2012). Stiles C.A., E.T. Dunkinson, C. Ping, and J. Kidd 2010. Initial f ield i nstallation of m anganese i ndicators of r eduction in soils, Brooks Range, Alaska. Soil Survey Horizons 51: 102 107. Strasser M. C. Schindler, and F. S. Anselmetti. 2008. Late p leistocene earthquake triggered moraine dam failure and outburst of Lake Zurich, Switzerland J. Geophys. Res. 113:1 16. Tiner R.W. 1999. Wetland i ndicators, a guide to w etland i dentification, d elineati on, c lassification, and m apping. CRC Press of Boca Raton, FL. p 148 149 ; 170 USDA NRCS. 1996. Soil s urvey of Dade County a rea, Florida. U.S. Dept. of Agriculture Natural Resources Conservation Service Na tional Cooperative Soil Survey division U S Governm ent Printing Office, Washington, DC. USDA NRCS 1997. The n ational s oil s urvey h andbook title 4 30 VI U.S. Dept. of Agriculture Natural Resources Conservation Service. Amend. 22 part 618 .67 (h) ( xi ). http://soils.usda.gov/technical/handbook/contents/part618.html ( accessed 7 Jul y 2012). USDA NRCS. 1998. Soil s urvey of Collier County a rea, Florida. Published by the National Cooperative Soil Survey. U S Government Pr inting Office, Washington, DC.
103 USDA NRCS. 1999. Soil t axonomy, 2nd e dition. Agriculture H andbook No. 436 U S. Government Printing Office, Washington, DC. p. 89. USDA NRCS. 2002. Changes in h ydric s oils of the United States U.S. Dept. of Agriculture N atural Resource s C onservation Servi c e. U.S. Gov Printing Office, Washington, DC. FR Doc. 02 23683 http://www.gpo.gov/fdsys/pkg/FR 2002 09 18/html/02 23683.htm ( accessed 23 Jul y 2012). USDA NRCS. 2010 Field i ndicators of h ydric s oils in the United States. Version 7.0. L.M. Vasi las, G.W. Hurt, and C.V. Noble, editors U nited S tates Dep ar t ment of Agriculture Natural Resources Conservation Service in cooperat ion with the Natio nal Technical Committee for Hydric Soils Veneman P.L., and E.W. Pickering. 1983. Salt bridge for field redox potential measurements Commun. Soil Sci. Plant Anal. 14 : 669 677. Vepraskas M.J., and S.W. Sprecher.1997. Aquic conditions and hydric soil s: the p roblem s oils. Special p ublication no. 50. Soil Science Socie ty of America Inc., Madison, WI. p. 7 11. Wang Q., Y. Li and W. Klassen 2005. Determination of c ation e xchange c apacity on l ow to h ighly c alcareous s oils. Commun. Soil Sci. Plant Anal 36:1479 1498. West L.T., L.R. Drees, L.P. Wilding, and M.C. Rabenhorst 1988. Differentiation of p edogenic and l ithogenic c arbonate f orms in Texas Geoderma. 43:271 287. Wright A.L., P.W. Inglett. 2009. Soil o rganic c arbon and n itrogen and d istribut ion of c arbon 13 and n itrogen 15 in a ggregates of E verglades h istosols Soil Sci. Soc. Am. J. 73: 427 433. Zhang M., Y. Li, P.J. Storrella and M. McLaughlin. 2002. Replacing c alcareous s oils with a cid s oils to g row r ainforest s s pecies at Fairchild Tro pical Garden, Miami. Proceedings of the Florida State Horticultural Society 115:146 149.
104 BIOGRAPHICAL SKETCH Christine Coffin began her interest in soils as an undergraduate at the University of Maryland with an elected course in soils which sparked a deep interest and ultimately resulted in switching to a major in soil science In 1991, she served as a Summer Intern with the U nited S tates D epartment of A griculture N atural R esources C onservation S ervice (USDA NRCS) in Sussex County, Delaware and part icipated in the University of Maryland Soil Judging Team. After earning a Bachelor of Science degree in 1992, she planned to mov e to south Florida and work on the conservation of the Everglades, but was offered a career position with USDA NRCS in Delaware as a f ield Soil Scientist. In that position she conducted soil salinity studies, soil survey updates in the coastal plain and piedmont regions of the mid Atlantic. Extensive survey of both coastal and inland wetlands w as the focus of her work in that posi tion, as well as digital cartography and survey m a n u script development. An opportunity to work in the Everglades region became available in 1995, at which time she relocated to Davie, Florida as a District Conservationist (DC) with USDA NRCS. In 1996, she accepted a DC position in Homestead Florida where she continues her career in natural resources conservation In 2005 she decided to work towards a Master of Science d egree through the University of Flor ida while continuing to work full time. Soil resea rch in this study was conducted at the U niversity of Florida I nstitute of Food and Agricultural Science Tropical Research and Education Center in Homestead, Florida.