This item is only available as the following downloads:
1 FLOATING TREATMENT WETLANDS AS A STORMWATER BEST MANAGEMENT PR ACTICE IN NORTH CENTRAL FLORIDA By NEAL BEERY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENT S FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 Neal Beery
3 To Wade, Pam, Maria, Kevin, and Megan. Each of you propped me up, gave me something to lean o n and pushed me forward
4 ACKNOWLEDGMENTS My thanks go to Mark Clark for lending me his knowledge and his patience. I also thank everyone who helped me out on the ponds: Jason, Patrick, Ben, Matt, Mark, Illea, Laura, Lory, Pam, Maria, Wade and Penelope. Finally, my thanks and a ppreciation go to ACF Environmental for the opportunity to conduct this research as well as Dr. Hochmuth and Dr. Boyer for the time and effort they put into being on my committee.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 14 Stormwater Management ................................ ................................ ........................ 14 Impairment of Natural Water bodies ................................ ................................ .. 14 Management of Stormwater Runoff ................................ ................................ .. 15 Deficiencies in Stormwater Regulation ................................ ............................. 16 Floating Treatment Wetlands ................................ ................................ .................. 20 Naturally Occurring Floating Wetlands ................................ ............................. 20 Constructed Floating Treatment Wet lands (FTWs) ................................ .......... 23 Objectives and Hypotheses ................................ ................................ .................... 25 Objectives ................................ ................................ ................................ ......... 25 Hypot heses ................................ ................................ ................................ ...... 26 2 MATERIALS AND METHODS ................................ ................................ ................ 30 Floating Treatment Wetland Design ................................ ................................ ....... 30 Study Area ................................ ................................ ................................ .............. 31 Site Descriptions ................................ ................................ ................................ ..... 32 Methodology ................................ ................................ ................................ ........... 36 Initial Macrophyte Sampling and Analysis ................................ ........................ 36 Final Harvest and Analysis ................................ ................................ ............... 37 Data Analysis ................................ ................................ ................................ ... 40 3 RESULTS AND DISCUSSION ................................ ................................ ............... 53 General Growth Characteristics ................................ ................................ .............. 53 ................................ ................................ .. 53 ................................ ................................ ............. 54 ................................ ................................ .. 55 Macrophyte Species Dominance and Recruitment ................................ ................. 55 Nutrient and Metals Mass Assimilation ................................ ................................ ... 56 Assimilation Performance by Panel ................................ ................................ ........ 60
6 Assimilation Performance by Species ................................ ................................ ..... 62 Carbon, Nitrogen and Phosphorus Tissue Concentration by Species .............. 64 Metal Tissue Concentration by Species ................................ ........................... 68 Biomass Assimilation by Species ................................ ................................ ..... 69 Carbon Assimilated by Species ................................ ................................ ........ 70 Nitrogen and Phosphorus Mass Assimilation by Species ................................ 71 Metal Mass Assimilation by Species ................................ ................................ 73 Performance by Allocation of Biomass ................................ ................................ ... 74 Carbon, Nitrogen and Phosphorus Concentration by Zone .............................. 75 Metal Concentration by Zone ................................ ................................ ........... 77 Biomass Assimilate d by Zone ................................ ................................ .......... 78 Carbon Mass harvested by Zone ................................ ................................ ..... 79 Nitrogen and Phosphorus Mass Harvest by Zone ................................ ............ 80 Metal Mass Assimilation by Zone ................................ ................................ ..... 82 4 CONC LUSIONS ................................ ................................ ................................ ... 106 Overall Performance of FTWs ................................ ................................ .............. 106 Panels and Edge Effects ................................ ................................ ....................... 109 Macrophyte Species Selection ................................ ................................ .............. 110 Harvesting by Zone ................................ ................................ ............................... 113 Recommendations for Further Research ................................ .............................. 115 APPENDIX REFERENCE LIST ................................ ................................ ................................ ...... 120 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 126
7 LIST OF TABLES Table page 1 1 Treatment efficiencies for wet detention systems based on selected research studies in Florida; from Harper and Baker, 2007, pp. 5 8. ................... 27 1 2 Summary of performance data for total phosphorus (TP), total nitrogen (TN) and biomass removal for a variety of floating treatment wetlands. ..................... 27 2 1 Number of clumps of each species per site. ................................ ....................... 41 2 2 Water quality d ata during the study period. ................................ ........................ 41 2 3 Stormwater runoff characterization for low intensity commercial areas in southern and central Florida; recreated from Harper and Baker (2007). ............ 42 3 1 Self recruited macrophyte species per site. ................................ ........................ 86 3 2 Initial mean constituent mass loading from planted macrophytes. ...................... 87 3 3 Overall mean per square meter mass assimilation. ................................ ............ 87 3 4 Biomass assimilation rate per square meter per day for all parameters st udied. ................................ ................................ ................................ ............... 87 3 5 Comparison of nutrient assimilation rates in this study with other stormwater treatment systems. ................................ ................................ ............................. 88 3 6 Final wet biomass averaged among all three deployment locations. .................. 89 3 7 Overall final wet mass per FTW. ................................ ................................ ......... 89 3 8 Mean net mass accumulation by panel. ................................ ............................. 89 3 9 Mean tissue concentration by panel. ................................ ................................ .. 90 3 10 Carbon biomass concentration (g kg 1 ) by species. ................................ ............ 90 3 11 Nitrogen biomass concentration (g kg 1 ) by species. ................................ .......... 90 3 12 Phosphorus biomass concentration (g kg 1 ) by species. ................................ .... 90 3 13 Site mean zinc (mg kg 1 ) concentrations by species. ................................ .......... 91 3 14 Site mean copper (mg kg 1 ) concentrations by species. ................................ ..... 91
8 3 15 Above mat net dry biomass assimilation (g m 2 ) by species. .............................. 91 3 16 Above mat carbon (g m 2 ) assimilation by species. ................................ ............ 92 3 17 Above mat nitrogen (g m 2 ) assimilation per species. ................................ ......... 92 3 18 Above mat phosphorus (mg m 2 ) assimilation per species. ................................ 92 3 19 Above mat zinc (mg m 2 ) assimilation, per species. ................................ ............ 93 3 20 Above mat copper (mg m 2 ) assimilation, per species. ................................ ....... 93 3 21 Carbon biomass concentration (g kg 1 ) by zone. ................................ ................ 93 3 22 Nitrogen biomass concentration (g kg 1 ) by zone. ................................ .............. 93 3 23 Phosphorus biomass concentration (g kg 1 ) by zone. ................................ ......... 94 3 24 Zinc biomass concentration (mg kg 1 ) by zone. ................................ .................. 94 3 25 Co pper biomass concentration (mg kg 1 ) by zone. ................................ ............. 94 3 26 Cadmium biomass concentration (mg kg 1 ) by zone. ................................ .......... 94 3 27 Chromium biomass concentration (mg kg 1 ) by zone. ................................ ........ 95 3 28 Biomass (g m 2 ) assimilation by zone. ................................ ................................ 95 3 29 Carbon (g m 2 ) mass assimilation by zone. ................................ ......................... 9 5 3 30 Nitrogen (g m 2 ) mass assimilation by zone. ................................ ....................... 95 3 31 Phosphorus harvest (g m 2 ) by zone. ................................ ................................ .. 95 3 32 Zinc (mg m 2 ) mass assimilation by zone. ................................ ........................... 96 3 33 Copper (mg m 2) mass assimilation by zone. ................................ ..................... 96 3 34 Chromium (mg m 2 ) mass assimilation by zone. ................................ ................. 96
9 LIST OF FIGURES Figure page 1 1 Floating wetland design criteria, an excerpt from the draft rule handbook (FDEP, 2010b). ................................ ................................ ................................ ... 28 1 2 Components of a floating wetland; source: Wanielista et al., 2012. .................... 29 1 3 A natural fl oating wetland (foreground) and constructed floating wetland (background) ................................ ................................ ................................ ...... 29 2 1 Unplanted plastic mat consisting of three 1.22 m by 2.44 m panels .................. 42 2 2 Close up view of the FTW mat plastic matrix viewed from below the mat looking up ................................ ................................ ................................ .......... 43 2 3 Floating wetland treatment panel with PVC anchor attachment located in the lower left hand corner ................................ ................................ ......................... 43 2 4 Location of anchor attachment and pan els A, B and C. ................................ ...... 44 2 5 Aerial view of Gainesville, Florida with test locations outlined ............................ 44 2 6 Ae rial view of SS1 (Lake Alice), on the University of Florida campus ................. 45 2 7 Ground level view of SS1 (Lake Alice). ................................ .............................. 45 2 8 Aerial view of SS2, Tumblin Creek stormwater pond, SW 6 th St. to the right ...... 46 2 9 Ground level view of SS 2, Tumblin Creek stormwater p on d .............................. 47 2 10 Aerial view of SS3, RTS stormwater pond, with RTS parking lot to the right ...... 48 2 11 Ground level view of SS3, RTS stormwater pond ................................ ............... 49 2 12 Position of target points for initial plant sampling. ................................ ............... 49 2 13 Separated initial plant samples ................................ ................................ ........... 50 2 14 Locations of depth measurements prior to harvest. ................................ ............ 50 2 15 Above mat, mat, and below mat zones ................................ .............................. 51 2 16 Division of panels into rows and blocks for mat zone stratified random sampling. ................................ ................................ ................................ ........... 52 3 1 Initial (left) and final (r ight) above mat FTW macrophytes ................................ .. 97
10 3 2 Surface area of SS3 and SS1 ................................ ................................ ............. 98 3 3 Mat zone of FTWs. A) Mat at the beginning of the study period ........................ 98 3 4 Examples of extensive root growth on FTWs ................................ ..................... 99 3 5 Aquatic vertebrates and invertebrates found in FTW below mat zone. ............... 99 3 6 SS1 panel growth, panel A, B, and C (left to right) ................................ ........... 100 3 7 SS2 panel growth, panel A, B, and C (left to right) ................................ .......... 100 3 8 SS3 panel growth panel A, B, and C (left to right) ................................ .......... 100 3 9 Example of lack of Pontederia cordata on SS1, panel C ................................ .. 101 3 10 Mean, maximum and minimum of biomass assimilation among species groups on FTWs in this study. ................................ ................................ .......... 101 3 11 Percent of biomass harvested by each zone. ................................ ................... 102 3 12 Percent of carbon assimilated by each zone. ................................ ................... 102 3 13 Percent of nitrogen assimilated by each zone. ................................ ................. 103 3 14 Percent of phosphorus assimilated by each zone. ................................ ........... 103 3 15 Percent of zinc assimilated by each zone. ................................ ........................ 104 3 16 Percent of copper assimilated by each zone. ................................ ................... 104 3 17 Percent of chromium assimilated by each zone. ................................ .............. 105
11 LIST OF ABBREVIATION S B MP Best Management Practice, in the context of water treatment, these are used by resource managers to guide and aid in controlling changes in water quality and quantity DO Dissolved oxygen, the amount of oxygen found in water FTW Floating Treatment Wetlan d, an artificial floating island or mat, installed in stormwater systems, on which plants grow MAPS Managed Aquatic Plant Systems, a best management practice which utilizes growing plants for the use in stormwater treatment TMDL Total Maximum Daily Load, t he allowable upper limit of a constituent found in waters of the State TSS Total Suspended Solids, the amount of solids found in water that can be filtered out using a filter of a determined pore size
12 Abstract of Thesis Presented to the Graduate S chool of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FLOATING TREATMENT WETLANDS AS A STORMWATER BEST MANAGEMENT PR ACTICE IN NORTH CENTRAL FLORIDA By Neal Beery December 2013 Chair: Mark Clark Cochair: George Hocmuth Major: Soil and Water Science Floating treatment wetlands (FTWs) are being considered to reduce pollutant loads associated with non point source s FTWs are artificial floating island s or mats, installed in sto rmwater systems, on which plants grow. The plants utilize nutrients fo und in stormwater and a ssimilate nutrients and other constituents within their tissues. Three FTWs were deployed in stormwater systems in Gainesville, Florida over the course of one growing season. Plant biomass was harvested and analyzed for biomass, carbon, nitrogen, phosphorus, zinc, copper, cadmium and chromium content. The goal of the study was to determine the overall efficacy of FTWs and their various component parts to assim ilate contaminants from within a stormwater basin. The average (m 2 d 1 ) mass assimilation of the FTWs were as follows: biomass 6.86 4.78g, carbon 2.52 1.75g, nitrogen 0.115 0.074g, phosphorus 15.3 8.8mg, metals ranging from a high of zinc (0.311 0.330mg), to a low of cadmium (0.0008 0.0009mg). Mass assimilation (m 2 ) for the above mat zone was highest with biomass of 856.2 649.1g, carbon 322.8 44.1g, nitrogen 9.5 6.2g, phosphorus 0.9 0.42g, metals ranging from a high for zinc (39.2 46.4mg), and cadmium wi th
13 unmeasurable amounts. The planted species with the highest biomass assimilation (m 2 ) were C. flaccida (306.4 392.6g) and self recruited species (296.2 174.4g). It was concluded that FTWs can be an effective stormwater BMP, that C. flaccida and species that that self recruited during the period of deployment were most effective above and below the plastic matrix allowed for the most constituents to be harvested.
14 CHAPTER 1 INTRO DUCTION Stormwater Management Impairment of Natural Waterbodies Surface water quality in many areas around the world is greatly affected by pollutants associated with human activity in both urban and agricultural settings. Pollutants are defined as natur al or artificial substances that occur in concentrations harmful to the environm ent (United Nations, 1997). In the context of water systems, pollutants matter, heavy metals and other chemicals from human manufactured materials (Mitsch et al., 2000). Water quality problems arise when these pollutants are physically transported from upland sources, or are directly released into water bodies through human activities (Lee and Bang, 2000). Degradation of surface water quality through pollution is widespread, including in Florida, where an estimated 550 square miles of estuaries, 2,000 miles of rivers and streams, and more than 375,000 acres of lakes were found to have excessive nutr ient bodies (FDEP, 2010a ) Criteria that set the maximum threshold for pollutants found in waters of the state are part of the Federal Clean Water Act which seeks to pr otect water resources for continued human use and natural function by setting specific maximum values of pollutants that cannot legally be exceeded (FDEP, 2010a ; Chapter 403, Florida Statutes, Section 304.067). The most common pollutants affecting Florid a waters include nitrate, nitrite, ammonia, phosphorus, E. coli petroleum hydrocarbons, mercury, aluminum, lead,
15 cadmium, chromium, copper, zinc, and suspended solids ( Harper and Baker, 1997 ). The quality of water resources is important for the well bein g of human s (drinking water, irrigation, recreation) and natural systems (biodiversity, ecological function) (Wanielista et al., 2012). Protection of these resources through pollution control is required to ensure their continued use (Hubbard 2010; Fulch er, 1994; US EPA, 2009; Chapter 62 40.431, F.A.C.). Management of Stormwater Runoff Increases in human population and their subsequent activities have led to extensive conversion of land to more intensive management and an increase in imp ervious surface s such as roadways, sidewalks, parking lots and conventional roofs. This increase in impervious area, combined with the additional loading of previously mentioned pollutants has led to the detriment of natural water systems. As more surface area becomes imp ervious the volume of precipitation that is able to percolate into the ground during and after a precipitation event decreases, thus exacerbating stormwater runoff issues of both quantity and quality (Staddon, 2011). The United States Environmental Pr otection Agency (2009) has recognized urban stormwater runoff as one of the major contributing factors to the decline in surface water quality in the United States. In urban settings, pollutants in the form of excess fertilizers, organic debris, oil and g rease from vehicles and equipment, or manmade substances (such as residue from the wearing of vehicle brake pads), can be carried by stormwater runoff, water that runs laterally over the landscape (US EPA, 2007 ). Stormwater runoff is often collected in ret ention and detention systems either for temporary or long term storage, whereupon it can be released as surface water into
16 natural water bodies or percolate into soil as groundwater. These systems were originally designed and built mainly for flood contro l and not for removal or co ntrol of pollutants (Wanielista et al 2012). New designs, techniques and technologies are being considered and implemented in order to address the increased pollutant loads associated with non point source pollution. These ef forts have focused on incorporating treatment as a main goal of newly constructed stormwater basins, as well as the retrofitting of older basins to include a variety of best management practices (BMPs) that address pollution. These technologies may includ e entirely human constructed mechanisms such as baffle boxes, or designs that incorporate natural organisms (such as algae and macrophytes) and systems such as constructed wetlands into the treatment process Examples of wetland treatment technologies inc lude shallow littoral zones that allow for the growth of emergent macrophytes in stormwater basins or systems that utilize floating macrophytes positioned in open water areas of a wet detention basin (FDEP, 2010a ; Headley and Tanner, 2006). Deficiencies in Stormwater Regulation Stormwater management and regulation began with the goal of flood control (quantity of stormwater) but this focus is shifting focus to include management for pollutants that enter w ater bodies and groundwater ( quality of stormwater). Florida regulations regarding stormwater have been updated over the past forty years to include goals for water quality as well as quantity. T ( Chapter 62 40 F.A.C.) [stormwater m If the source of water pollutants cannot be reduced through management education or regulation, then in order to comply with Florida regulation, efforts will be necessary to
17 address the removal of pollutants from managed retention systems and possibly natural systems. qualitative basis, such that the stormwater discharge of an applicant being determined ould require efforts to store stormwater runoff (Chapter 17 4.248, F.A.C.). This regulation was updated in 1982 with Chapter 17 25 of the F.A.C. which required stormwater permits fo r all new stormwater discharges and required ret rofits for discharges that saw increases in flow or pollutant loading; this legislation did not include the use of BMPs for treatment of stormwater quality. The current criteria for stormwater treatment in Florida w ere set in 1995 by Chapter 62 40 of the F.A.C This criteria states that a minimum of 80% of the average annual load of pollutants of concern must be removed by stormwater treatment systems; a dditionally, a stormwater system that discharges into an Outstanding Florida Water (water bodies that a re must remove 95% of the average annual pollutant load (FDEP, 2012) rule does not currently include incentives to utilize newer technologies and techniqu es that would bring new stormwater treatment systems closer to meeting th e new criteria (Livingston, 2009 ). Harper and Baker (2007) state in their report to the Florida Department of Environmental Protection (FDEP) that several commonly used stormwater sy stems often do not meet set criteria for stormwater treatment ( Table 1 1 ) Their study compiled mean performance efficiencies of ten stormwater management systems (all were wet detention systems that had close to 14 day resid ence times). Mean
18 performance efficiencies were as follows: total nitrogen reduction was 37%, total phosphorus reduction was 69%, TSS reduction was 77%, and reductions of copper and zinc metals was 69% and 85%, respectively. This data suggests that the l evel of treatment efficiency in these systems is not adequate to meet the current or proposed stormwater treatment design to meet criteria is at least in part contrib uting to the 2009). Proposed New Statewide Stormwater Rule According to a 2009 of Watershed Management, new regula tion on the design and implementation of stormwater treatment systems was proposed and began development in order to treatment failing to meet permit design criteria (Living ston, 2009) As of January 2013, the new regulation was still under development. This new regulation, generally referred to as the Statewide Stormwater Rule, will sometimes be referred to as the Rule in this thesis Within the most recent draft of this have been set for the design and implementation of particular BMPs (FDEP, 2010b ) The new Rule also proposes that, for most discharge permits, the performance of new or retrofitted stormwater systems must meet t he following requirements: either, reduction by 85% of pollutant loading that are attributable to development, or the pollutant loading must be less than or equal to the amount of pollutant loading prior to any development of the land, whichever is less ( F DEP 2010b ).
19 One of the BMPs identified in S design criteria for use of Managed Aquatic Plant Systems (MAPS) which includes the use of floating wetlands. Floating wetlands are described in the Handbook as islands or mats on which plants grow, utilizing nutrients found in stormwater and accumulating nutrients and other constituents of stormwater within their tissues. The Handbook recognizes that removal of biomass from the floating wetlands and transportati on of this biomass away from the stormwater treatment system allows for nutrients and other constituents of concern to be removed from the stormwater system ( FDEP, 2010b ). The Handbook indicates that when floating wetlands are applied as a BMP within a sto rmwater treatment system, the proposed nutrient reduction credit is 20% 40% for both total nitrogen and total phosphorous. This proposed range of nutrient removal is a significant portion of the proposed 85% removal criteria (as well as the existing 80% r emoval criteria), and implies that floating wetlands and other MAPS, when properly utilized, may be an effective tool for permit applicants seeking to boost the performance of their stormwater treatment system. Section 14.4 o f the Draft Applicants Handbook sets out criteria for the use of floating wetlands as BMPs. Figure 1 1 is an excerpt from the Handbook that describes these criteria (emphasis and notes are from the source) ( FDEP, 2010b ) If the Statewide Stormwater Rule is adopted into Florida law with an Applicant s Handbook containing provisions for floating wetlands as MAPS, the amount of treatment credit awarded to floating wetlands could motivate many stormwater managers to install these systems. The environmental pro ducts firm that provided the floating wetland MAPS design for this research wishe d to construct and evaluate performance of their
20 system so that they could determine treatment credit that would be allotted when market ing th eir systems in order to meet prob able demand from stormwater managers seeking to boost stormwater treatment system performance Floating Treatment Wetlands Naturally Occurring Floating Wetlands In naturally occurring floating wetlands, the general components are the mat (substrate made of live and dead organic, mi neral, trapped gases and vegetation roots and rhizomes ), roots extending below the mat, leaf and shoot biomass extending above the mat the free water within and below the wetland, and in some cases a layer of organic matter tha t rests below the wetland which has sloughed off the mat (Sasser et al., 1991; Clark, 2000). Figure 1 2 depicts these components in a constructed floating wetland In North America, floating wetlands are most prevalent in Lou isiana where they occur in both freshwater and brackish coastal marshes of the Mississippi River Delta (Sasser, 1994). They are also found in Arkansas, Georgia and Florida, where they are found inland in freshwater systems (Sasser and Gosselink, 1984; Cla rk, 2000). Floating wetlands can be found in varying forms throughout the world, and are also known as flotant (Sasser, 1994), tussocks, floatons or sudds in the ir respective locales (Mallison et al. 2001). Floating wetlands can exhibit great ecologica l differences from the rooted wetlands in adjacent littoral zones. They create conditions and behave in ways that make them ecologically important and distinct from rooted wetlands (Sasser, 1994). Their flotation allows them to escape the effects of hydr opattern (depth, duration, frequency of flooding) that other rooted wetlands a re subjected to, which in turn
21 influences vegetation, biogeochemical processes, and wildlife. The hydrologic regime of the adjacent rooted wetland can exhibit fluctuations that causes vegetative succession to favor plants that can cope with changes or are best suited for the prominent flooded conditions, while floating wetlands hydrologic regime remain s relatively stable unless their buoyancy is altered or the water body in which they float is no longer as deep as their mat (Sasser, 1994). Floating wetlands in different regions have a variety of plant species that grow upon them. Even in the same region, such as Orange Lake in Florida, floating islands can be dominated by di ffere nt plant species; Mallison et al. (2001) describe floating wetlands dominated by pickerelweed ( Pontederia cordata ) growing in close proximity to floating wetlands dominated by Cuban bulrush ( Scirpus cubensis) and Hydrocotyle sp.. This is of interest not only for the apparent variety in species that colonize the wetlands, but also for the characteristics of the species themselves. Also in Orange Lake, Mallison et a l. (2001) observed floating wetlands with wetland obligate species such as cattail ( Typha la tifolia ), in addition to broom sedge ( Andropogon virginicus ) and dog fennel ( Eupatorium capillifolium ), both of which are facultative species, meaning they are able to live in both moist and non flooded conditions, but not necessarily flooded conditions T he presence of wetland and non wetland plants is important to note because it shows that on a single floating wetland, the hydrologic conditions can be such that a plant which thrives in anaerobic conditions and a plant that cannot tolerate permanently flo oded conditions can be supported (Mitsch and Gosselink, 1997). A wide variety of vegetation types can be found on floating wetlands, mostly wetland plants due to the
22 stress of low oxygen in most of the mat. The plant species may determine the buoyancy or longevity of the floating wetland, as the plant litter can accumulate and provide substrat e for other species to colonize. T his is especially true when the floating floating plants (such as Hydrocotyle sp.) creating a nuc leus and then accumulating enough substrate for plants that require some sort of support (Clark, 2000). The biomass assimilation of floating wetland plants varies as widely as the plant species that are found on the wetlands. Sasser (1994) found Hydrocoty le sp. to have a biomass assimilation rate of 0.1 2 g m 2 yr 1 whereas, in the same region of Louisiana, maidencane ( Panicum hemitomon ) had a rate of 636.17 g m 2 yr 1 Assimilation of floating wetland biomass is linked to drivers similar to that of roote d wetlands, without the important characteristic of flooding depth (Clark, 2000). Water is constantly available to a floating wetland, and thus so long as the plant species on the wetland are supplied with ample nutrient resources for growth they are able to assimilate biomass. This is not true of rooted wetlands where a drop in water level can drive productivity down due to the lack of water. The main driver for Assimilation in fl oating wetland plants is often nutrient availability, since the plants ar e not connected with the geological substrate they do not have the soil as a source of nutrients (Mallison et al. 2001). In this way, hydropattern can have an indirect effect on floating wetlands, such as when an influx of sediment brings new nutrients i nto a waterbody (Clark, 2000). In the same sense, the productivity of floating wetlands can increase in water bodies undergoing eutrophication due to
23 and therefore, o ther than climactic influence ( such as wind, solar irradiance, humidity), the principal driving force for their Assimilation is the physical and chemical constituents of the water they are growing in (Headley and Tanner, 2006). Constructed Floating Treat ment Wetlands (FTWs) Floating treatment wetlands are an emerging technology which can be used for removing nutrients and pollutants from stormwater and agricultural runoff. They behave in much the same way that natural floating wetlands do ( Figure 1 3 ) but are designed, deployed and maintained by humans, instead of occurring naturally. Certain considerations in their design and maintenance must be made since FTWs may be deployed in already functioning stormwater treatment sys tems, natural water bodies, and in close proximity to humans and wildlife (Hubbard, 2010). Constructed FTWs are comprised of the same basic components as that of natural floating wetlands, that is, emergent macrophytes rooted in a floating mat, suspended over free water (Headley and Tanner, 2006 ). The macrophytes can either be rooted in growing media or affixed some other way to the mat (Headley and Tanner, 2011). The mat, which is typically the constructed portion of the FTW, can be made of plastic or organic materials such as coconut fiber or bamboo, and can be rigid or flexible. The mat gives support and buoyancy to the macrophytes on the floating wetland (Headley and Tanner, 2006). In natural ecosystems, floating wetland plants die and senesce whic h allows nutrients and other components to leach into the water column (Sasser et al. 1991). Constructed floating treatment wetlands function in much the same manner as natural floating wetlands, including the return of nutrients to the water column thro ugh senescing vegetation. This makes a harvesting plan for floati ng treatment wetlands
24 necessary if optimizing nutrient removal is the principal objective. Therefore, at least a portion of the vegetation biomass growing on FTWs deployed for pollution cont rol must be harvested seasonally in order to remove the plant bound nutrients or pollutants from the system (Headley and Tanner, 2006). A unique opportunity is available for the utilization of floating treatment wetlands due to their sole source of nutr ients being the water column in which they float. When compared with rooted wetlands, floating wetlands would be a better option for some water quality treatment purposes, such as the removal of nutrients and other pollutants in a basin with fluctuating w ater depth, since constructed floating treatment wetlands behave in the same way as natural floating wetlands with respect to hydropattern. For example, if a stormwater pond receives runoff in seasonally varying amounts, thus causing the water level to ri se or fall, emergent vegetation on the littoral shelf may not be in contact with the pollutant laden water during certain periods, and can even die or be displaced by species that thrive in the drier conditions. Utilizing floating wetlands to provide trea tment of the water may be more reliable since the wetland will rise and fall with water levels, thus providing treatment at all stages of the hydropattern. This approach to water treatment has been studied in Auckland, Australia as well as Germany, Indi a and the United States (Headley and Tanner, 2008). Several studies have been done in the past five years (Chua et al., 2012; Wanielista et al 2012; DeBusk et al., 2004; Stewart et al., 2008), at the laboratory, mesocosm, and pond sized scales, though a s recently as 2006 Headley and Tanner report ed
25 summary of some of the performance data for floating treatment wetlands in stormwater applications can b e found in Table 1 2 Additionally, using a mesocosm scale study, Van de Moortel et al. (2010) found that FTWs have proven to be effective at removal of nitrogen, phosphorus, total organic carbon, and several heavy metals, wh en com pared to a control group with no FTW. Average removal efficiencies of FTWs for TN, and TP were 42% and 22% respectively, compared to 15% and 6% for the control. Objectives and Hypotheses Objectives The possible adoption of a unified R ule for Florida served as impetus for this research. Th e new statewide Rule and its inclusion of Managed Aquatic Plant Systems (such as floatin g treatment wetlands) increases the possibility that these BMPs could be prevalent across the state The ir possible prevalence increases the importance of determining the actual performance of FTWs. As stated earlier, quality of stormwater that enters waters of the state. Allotting more treat ment credi t to floating wetlands than is justified by their actual performance could allow for stormwater treatment systems to be permit ted without meeting the legal 80% (or proposed 85%) nutrient reduction criteria. This failure to meet the reduction cri teria could impact Therefore, it is important to the natural resources of Florida that research is carried out which evaluates the efficacy of floating wetlands used as BMPs. Furthermore, the methods used to evaluate the efficacy of floating wetland MAPS permitted for use in stormwater treatment systems must become standardized and accurate. This is to
26 ensure that new floating wetland designs are tested equally and so floatin g wetlands installed as MAPS are as effective as the treatment credit they are allotted. This research project sought to deploy floating treatment wetlands in actual stormwater retention settings in order to determine the ability of floating wetlands to ta ke up nutrients and heavy metal s as well as to determine possible issues pertaining to the maintenance efforts of FTWs for maximum benefits as stormwater BMPs. Hypotheses H1: Floating treatment wetlands assimilate total carbon ( TC), total nitrogen (TN), t otal phosphorous (TP), cadmium (Cd), chromium (Cr), copper (Cu), and zinc (Zn). H2 : Within floating treatment wetlands there is an edge effect where more TC, TN, TP, Cd, Cr, Cu, and Zn are assimilated by the outer portions of the floating wetland than th e inner portion. H3: Macrophyte species planted on floating treatment wetlands that have been shown in previous field studies to assimilate more TC, TN, TP, Cd, Cr, Cu, and Zn than other species will assimilate more nutrients or metals. H4: Partitionin g of nutrients (TC, TN, TP) and metals (Cd, Cr, Cu and Zn) below the mat (roots) within the mat (rhizomes) and above the mat (petioles and leaves) will be different and follow similar distribution pat terns to previous field studies.
27 Table 1 1. Tre atment e fficiencies for wet detention systems based on selected research studies in Florida; from Harper and Baker, 2007, pp. 5 8. Location, Land Use Reported Removal Efficiency (%) TN TP Zn Cu Brevard County, FL, Commercial -69 --Boca Raton, F L, Residential 12 55 --Maitland, FL, Highway 35 81 92 56 EPCOT, FL, Highway 44 62 88 Orlando, FL, Urban -38 --Orlando, FL, Residential -91 96 90 DeBary, FL, Commercial & Residential 30 70 95 50 Tampa, FL, Light Commercial -65 51 -Ta mpa, FL, Commercial 63 90 87 55 Melbourne, FL 36 65 -92 Mean 37 69 85 69 Table 1 2. Summary of performance data for total phosphorus (TP), total nitrogen (TN) and biomass removal for a variety o f floating treatment wetlands. Study Removal (g m 2 da y 1 ) Biomass TN TP Debusk, Dierberg and Reddy (2001) 42 0.031 0.370 Hubbard, Gascho and Newton (2004) 33.9 0.046 1.096 .006 0.162 Chua et al (2012) 0.002 0.0160 .0002 .0016 Tanner et al (2011) 0.16 0.24 .0023 0.0054 Wen and Reckna gel (2002) 0.043 0.086 White (2009) 0.001 0.002
28 Floating Wetland Design Criteria: (SUBJECT TO CHANGE AS MORE DATA BECOMES AVAILABLE) (a) The area of floating wetland mats shall be at least five percent (5%) of the surface area of the wet detention pond. (What about load reduction if > 5%) (b) The floating wetland island or mats shall use a variety of plants that have been documented to have high nutrient uptake in their plant tissues. Some proven plants include Canna flaccida, Juncus effus es, Spartina spp., Pontederia cordata, ADD TO LIST/EDIT (c) Floating wetland mats or islands shall be installed and maintained in accordance with permitted design (d) Where necessary, exclusion netting sh all be used on floating islands or mats to prevent turtles, grass carp, or other animals from eating the plant roots or plants such that they adversely affect the successful growth of the aquatic plants. The applicant may propose alternative mechanisms to minimize eating of plant roots or plants based on an affirmative demonstration, based on calculations or other information, that the alternative design is appropriate for the specific site conditions an d will meet the above considerations. (e) Within 6 months of installation, the float ing wetland island or mat shall have at l east 90 percent coverage with no more than 10% consisting of exotic or nuisance species. (f) Plants on the mats or islands shall be removed and replaced at a minimum on an annual basis. The harvested plant and potting materials shall be removed and disposed of in such a manner that nutrients will not re enter the stormwater treatment system. Figure 1 1 Floating wetland design criteria, an excerpt from the draft rule h andbook ( FDEP, 2010b )
29 Figure 1 2. Components of a floating wetland; source: Wanielista et al., 2012. Figure 1 3 A natural floating wetland (foreground) an d constructed floating wetland (background) photo courtesy of Neal Beery Constructed FTW Natural floating wetland
30 CHAPTER 2 MATERIALS AND METHOD S Floating Treatment Wetland Design The FTWs for this research were constructed and maintained by following the instruction of ACF Environmental Inc., who provided the materials, design and on site assistance in the construction and deployment of the FTWs. Each FTW consisted of three 1.22 m by 2.44 m positively buoyant plastic panels strung together with nylon rope to create a single mat 2.44 meters wide by 3.66 meters long with an area of 8.93 m 2 ( Figure 2 1 ) Sealed 1 inch PVC pipes, running along the width of the each individual panel were installed within the matrix to add buoyancy Each panel consisted of a composite open cube matrix 5 x 5 cm square with 4 cm diameter circular holes throughout ( Figure 2 2 ). Each 1.22 m by 2.44 m section will be referred to as a Panel and all three attached Panel s installed with macrophytes will be referred to a s a Floating Treatment Wetland (FTW). Three species of wetland plants were used to initially populate each floating treatment wetland: Juncus effusus (common rush), Pontederia cordata (pickerelweed), and Canna flaccida (canna). ACF Environmental was re sponsible for choosing and supplying the plant species used. The plants were supplied as bare root individuals washed clean of soil and substrate. Each FTW was planted by placing the individual plants through the holes in the plastic matrix so the roots extended below the mat and the petioles extended above the surface in the manner of emergent macrophytes. Cable ties were used to secure the individual plants to the plastic matrix when the young plants were too small to remain fixed in place Approximat ely thirty
31 each wetland in a generally random pattern with approximately three clumps per square one hole of the pl astic matrix. This planting pattern was overseen by employees of ACF Environmental. Exact numbers of clump s fo r each Site are found in Table 2 1 An anchor was tied to a PVC pipe that was fixed to one corner of the FTW with t he PVC pipe secured to the platform ( Figure 2 3 ). The anchor was installed in order to keep the FTWs from migrating toward the shore, thus keeping the conditions consistent over time. An anchor scope ranging from 3:1 to 5:1 was used to allow each FTW to pivot within the free water but prevent the FTW from running into the littoral zone. The PVC attachment was fixed to the platform in order to keep it from cropping plant material if the PVC attachment were to swivel around. At each site, the Panel affixed with the anchor attachment was referred to as Panel A, the middle Panel was referred to as Panel B, and the Panel furthest from the anchor attachment was Panel C ( Figure 2 4 ). Each FTW was towed by canoe to the area in which it would remain anchored for the duration of the study. Each area was determined to have adequate depth and clearance from any possible obstructions. Where possible, photographs were taken of the floating islands from eac h end of the three Panel s making up the platforms in order to provide visual documentation of plant growth and condition over the period of the study. Study Area The C ity of Gainesville, located in north central Florida (latitude 29 longitude 82 W), has a humid subtropical climate, and receives an average annual rainfall of 122.8 cm. In the summer months, the minimum and maximum d aily mean
32 temperatures are 18C and 35 C respectively, and the daily mean relative humidity is 67% I n the winter mon ths, minimum and maximum daily mean monthly temperatures are 4C and 22 C respectively and the daily mean relative humidity is 76% (Irmak et al., 2002; NOAA, 2012). The Gainesville a rea experiences its peak evapotranspiration rate in May, with the mean r ate of 146 mm month 1 and its minimum evapotranspi ration rate in December, with the rate of 52 mm month 1 (Irmak et al., 2002) Over the study period, which ran from March 9th 2010 through December 12 th 2010, a weather station (approximately 10 miles north northwest of the study sites) received 75.3 cm of rainfall, and experienced an average daily minimum temperature of 16 C and an average daily maximum temperature of 29 C, an average relative humidity of 79%, an average windspeed of 7.06 km hr 1 and an average evapotranspiration rate of 97 mm month 1 (FAWN, 2012). The climactic data for the study period seemed to fall within normal ranges for the region. Site Descriptions The study locations which will be referred to as SS1, SS2, and SS3 consisted of three individual stormwater catchment basins in Gainesville, Florida ( Figure 2 5 ). All sites were located in areas of disturbed urban soils (UF CALM, 2012; USDA 2012) and were in close proximity to roadways, parking lo ts, conventionally roofed buildings, and other developed surfaces from which they received stormwater runoff. These sites were chosen for their variety of possible characteristics, which were anticipated to be representative conditions of variation found in stormwater ponds. All sites were located in developed urban areas and so were expected to exhibit elevated nitrogen levels (Harper and Baker, 2007). One site, SS3, was also suspected to receive
33 high levels of heavy metals due to its proximity to indus trial runoff from a transit bus parking lot and maintenance area (Bringham et al., 2002). Water quality data was taken at the three sites intermittently during the study period ( Table 2 2 ). For TN and TP, water samples were tak en from the ambient water near, but not under, each FTW by taking grab samples at approximately 35 cm deep. S amples were preserved with concentrated sulfuric acid, stored at 4C until they were submitted for analysis. Analysis was performed by the UF/IFA S ANSERV Advanced Research Laboratory. TSS samples were analyzed according to EPA Method 160.2. A YSI multi parameter probe was used to determine temperature, pH and DO (dissolved oxygen). As a comparison with other stormwater runoff characteristics, Table 2 3 was recreated from a table from Harper and Baker (2007), which compiles data from stormwater runoff characterization studies performed in several parts of Florida. In comparison to the overall means SS1 and SS3 had co mparable TN and higher TP concentrations whileSS2 had lower TN and TP concentrations. It must be noted, however, that the Harper and Baker study was a compilation of data for the water quality of stormwater runoff itself, not for the water column of storm water basins. TSS levels for SS1, SS2 and SS3 were similar to that of another study, performed in Tampa, Florida, which reported TSS levels for three stormwater ponds as ranging between 4 and 10 mg L 1 These levels are considered normal to low for storm water wet retentio n ponds (Rushton et al., 2004). SS1 was located on Lake Alice, a small lake and stormwater catchment basin on the southwestern portion of the Universi ty of Florida campus ( Figure 2 6 Figure 2 7 ). It
34 is a eutrophic lake of approximately 81 acres that receives runoff from much of the University of Florida campus (Mitsch, 1976; UF CALM, 2012). During the period of this study, grass carp, turtles, alligators and a wide variety of water birds were seen at the site. The area of land that drains to Lake Alice covers a surface area of 1,106 acres, of which an estimated 42% is impervious. The calculated pollutant removals originally permitted for the stormwater treatment system were as fol lows: TN removal of 15,874.6 kg yr 1 ,TP removal of 4,538.0 kg yr 1 TSS removal of 507,209.5 kg yr 1 Zn removal of 32.6 kg yr 1 Cu removal of 14.9 kg yr 1 Cr removal of 16.2 kg yr 1 ( SJRWMD 2010). SS2 was located on a stormwater pond on SW 6 th Street at Tumblin Creek Park ( Figure 2 8 Figure 2 9 ). Tumblin Creek Park is a small recreational park located original pe rmit, the stormwater pond covered an area of 1.08 acres, received the first flush of runoff from a developed area of 50.6 acres (approximately 70 percent of which is impervious), provided 3673 m 3 of water quality treatment volume, and discharges treated st ormwater into Tumblin Creek. The calculated pollutant removals for the permitted stormwater treatment system were as follows: TN removal of 128.67 kg yr 1 ,TP removal of 19.10 kg yr 1 TSS removal of 9,457.00 kg yr 1 Zn removal of 15.42 kg yr 1 (SJRWMD, 2011; SJRWMD 2002). The stormwater pond at SS2 was constructed with a littoral shelf, which is a stormwater BMP in which a shallow area with a low slope supports the growth of emergent macrophytes (NCDENR, 2005 ). The photograph in Figure 1 3 which depicts a natural floating wetland in the foreground, was taken at SS2 during the course of this
35 study. The perimeter of this stormwater pond is populated with emergent macrophytes and natural floating wetland mats. During the perio d of study it was noted that the Tumblin Creek Park stormwater retention system is maintained for aesthetic value with the use of a product similar to Aquashade, a blue dye that absorbs solar radiation of the wavelength that is useful to photosynthetic pl ants. These types of dyes are often used to control algae and aquatic plant populations within water bodies, thus promoting water clarity. The shading provided by these blue dyes has been found to decrease the growth of aquatic plants due to shading of l ight (Manker and Martin, 1984). This decreased growth in aquatic plants may have had some effect on increased nutrient availability, and thus the growth of the FTW macrophytes, at this site. SS3 was located on a stormwater pond adjacent to the Gainesville Rapid Transit System ( RTS ) Downtown Station on SE 10 th Avenue ( Figure 2 10 Figure 2 11 ). This approximately 1.5 acre stormwater pond is located in a stormwater park maintained for water storage, treatment and general aesthetic value. This pond was designed to receive runoff from the parking and maintenance lot for busses at the Gainesville RTS Main Street Station. A variety of wetland and upland species were found growing in a densely populated buffer zone and littoral zone surrounding the pond. Early in the study period, Hydrilla verticillata a non native submersed aquatic macrophyte, was found growing thickly throughout the pond. Sometime during the study period the presence of H verticill ata severely declined. It was unknown whether this was due to management efforts (such as with the use of an aquatic herbicide, that may have affected the FTW located at this site) or natural factors. It was anecdotally
36 mentioned by a stormwater manager for the site that treated wastewater effluent may have been discharged into the stormwater pond for a short period at the beginning of the study, but was cut off at some point. This could not be confirmed, but also may have played a role in the nutrient d ynamics affecting the Hydrilla population as well as the FTW. Methodology Initial Macrophyte Sampling and Analysis On March 31, 2010, a fter a 22 day period of acclimation post planting, each FTW site was visited for the purpose of taking initial macrophyt e samples Each species of originally installed macrophyte was sampled to determine the initial tissue concentration for the constituents of interest. Additionally, the initial biomass of each Panel was calculated using these samples by multiplying the m ean mass of the six sampled clumps (two clumps of each species from each of the three FTWs) by the number of clumps installed on each Panel. Two clumps of each species were taken from each FTW for initial sampling. One clump was taken from Panel A and a s econd clump was taken from Panel C near the anchor attachment edge. The clump sampled was the one occurring closest to a target point 45 cm from the edge in the center of each Panel as shown in Figure 2 12 Macrophyte samples were kept on ice and were sorted and pr ocessed for analysis off site. Each clump was rinsed, the number of individual plants per clump was recorded, and each clump was separated with sciss ors into categories of petiole (live and dead), rhizome and root ( Figure 2 13 ). Plant tissue samples were dried at 45 o C for a minimum of ten days. The tissue was then ground using a ball mill. Tissue was ground into a homogenous powder that
37 could pass through a 2 mm sieve and stored in opaq ue plastic scintillation vials until submittal for analysis. Plant tissue analysis was performed by the UF/IFAS ANSERV Advanced Research Laboratory in Gainesville, Florida. S amples were analyzed for concentrations of total nitrogen, total phosphorous, to tal carbon, cadmium, copper, chromium and zinc. Final Harvest and Analysis On December 2, 2010 macrophytes of SS3 and Panels A and B of SS2 were ha rvested. On December 17, 2010 macrophytes of Panel C of SS2 were harvested, and on December 2 2, 2010 macroph ytes of SS1 were harvested. T he protocol for harvesting the macrophytes of the respective FTWs follows. Wh ile still floating in situ, the depth between the water surface and the plastic matrix was measured a t the midpoint of the outer edge of each (see th e red x in Figure 2 14 ) Photograp hs of the macrophytes on each FTW were also taken Each FTW was towed to shore by canoe. Each of the three Panel s of the respective FTWs w ere separated and photograph s of the to p and botto m of each Panel were taken when possible. Prior to harvesting, a s urvey for species richness was taken by visually looking over each Panel and clipping a small sample o f each species found for later identification. After identification, these clippings were reincorporated into their appropriate species designation. Each Panel was harvested separately. The harvest was separated into three vertical biomass zones relative to the plastic matrix that made up each mat. These zones were Above Mat, Mat, and Be low Mat ( Figure 2 15 ). Above Mat biomass was harvested by cutting macrophytes down to where their rhizomes began (live and
38 senesced petioles and leaves were considered Above Mat biomass, but rhizome and root structures, even i f they were physically above the plastic matrix were not). For the Above Mat Zone, biomass was sorted by the three initially installed Species ( P. cordata, C. flaccida and J. effusus ) and a fourth all other species of macro phytes. The Below Mat Zone was harvested by shearing off all biomass (including any rhizome, root or any other plant structure) found below the plastic matrix; special care was taken to get as close to the plastic matrix as possible. No sorting by species was done for the Below Mat Zone, as physically identifying and separating clustered root structures was not possible. For the Above Mat and Below Mat Zones, sub samples were taken when the biomass exceed ed an amount that would be practical for the tissue drying and grinding process. T his estimation was conducted by a visual examination after each FTW was brought ashore; only site SS2 required sub sampling. When sub samples were taken, the wet mass of all harvested biomass was determined using a top loadi ng scale. Then, manageable amounts of biomass were separated into two to four sub samples ranging from 2.25 kg to 5 kg then weighed when wet. After the sub samples were dried, and the ratio of wet to dry mass for the sub samples was found, this wet to dr y ratio was then used to calculate the dry weight of the full harvest. The Mat Zone was harvested by including any biomass that was not harvested for the Above Mat or Below Mat divisions. No sorting by species was done for the Mat Zone, as physically i dentifying and separating clustered roo t structures was not practical.
39 Due to the physical difficulty of removing rhizome and other material from within the plastic matrix, biomass in the Mat Zone was sub sampled using a stratified random subsampling. The construction design of each Panel divided the plastic matrix into five rows of five blocks. Figure 2 16 shows a repre sentation of these divisions of a Panel The rows run along the long edge of each Panel and the blocks run a long the short edge Thus, each Panel was divided into a five row by five block grid. A random number list was generated for numbers 1 through 5, and this list was used to select a single block out of each row in a stratified random manner. The biomass from the five selected blocks (one block for each of the five rows) was harvested completely from the Mat Zone. This biomass was then combined to make the sub sample of the Mat biomass from each Panel As t his biomass represented 1/5 of the biomass for the Mat Zone of each Panel, the dry mass for that Panel was calculated by multiplying the mass for its sub sample by five. All plant tissue samples from the Final Harvest were dried at 45 o C for a minimum of ten days. The tissue was then ground first usi ng a Wiley Mill and then using a ball mill. Tissue was ground into a homogenous powder that could pass through a 2 mm sieve and stored in opaque plastic scintillation vials until submittal for analysis. Plant tissue analysis was performed by the UF/IFAS ANSERV Advanced Research Laboratory in Gainesville, Florida. S amples were analyzed for concentrations of total nitrogen, total phosphorous, total carbon, cadmium, copper, chromium and zinc. After the total biomass for both the initial planting and final harvest was determined, biomass was multiplied by the concentration of TN, TP, TC, Cd, Cu, Cr, Zn to determine the total mass of each constituent for each Site, Panel, Zone and Species.
40 The initial mass for TN, TP, TC, Cd, Cu, Cr, and Zn was subtracted f rom the final mass for each constituent, which resulted in the net mass assimilation (removed from water column) by the FTW per Site, Panel, Zone and Species over the study period. Data Analysis The statistical package JMP 8.0 (SAS) was used to run stat istical analysis on the data for this research. Data was compiled into sets depending on what result was being tested (for example, when testing for results between Zones, the per Site per Panel Zone data was compiled). Tests for normality were then run on each data set using JMP. When data sets were found to be normal, an ANOVA test was run to determine if significant differences existed between independent groups. If significant differences rmine which particular groups significantly differed from the others. When data was found to be non normal, Wilcoxon tests were performed to determine if there were significant differences between the means of two samples (for example, between the initial mean biomass of a Species group and the final mean biomass of that Species group) and the Kruskal Wallis test was used to determine if significant differences were present when the degrees of freedom was greater than 1 (for example, between the mean net b iomass assimilation of all four Species groups). An alpha level ( statistical tests.
41 Table 2 1 Number of clumps of each s pecies per s ite. Site Species Clumps SS1 Juncus effusus 52 Pontederia cordata 41 Canna flaccida 23 SS2 Juncus effusus 48 Pontederia corda ta 45 Canna flaccida 28 SS3 Juncus effusus 37 Pontederia cordata 32 Canna flaccida 28 Table 2 2. Water quality data during the study period. Site Date Parameter TN (mg L 1 ) TP (m g L 1 ) TSS (mg L 1 ) C pH DO (mg L 1 ) SS1 5/25/2010 1.08 0 .69 7/15/2010 0.36 0.49 7/20/2010 0.12 0.50 1.66 9/26/2010 1.97 0.54 10/5/2010 2.14 0.55 6.57 24 9.1 14.8 Mean 1.13 0.56 4.12 24 9.1 14.8 SS2 5/25/2010 0.58 0.05 7/15/2010 0.16 0.05 7/20/2010 0. 03 0.03 1.1 0 9/26/2010 0.64 0.04 10/5/2010 0.72 0.04 1.5 0 24 7 .0 4.8 Mean 0.43 0.04 1.30 24 7.0 4.8 SS3 5/25/2010 1.03 0.06 7/15/2010 0.23 0.13 7/20/2010 0.12 0.10 3.14 9/26/2010 1.2 0 0.10 10/5/20 10 1.33 0.09 5.99 23 7.2 7.5 Mean 0.78 0.10 4.56 23 7.2 7.5
42 Table 2 3 Stormwater runoff characterization for low intensity commercial areas in southern and central Florida ; recreated from Harper and Baker (2007) Location Mean Concentration (mg l 1 ) TN TP Zn Cu Cd Cr Orlando, FL, Areawide Study 0.89 0.16 ----Ft. Lauderdale, FL, Coral Ridge Mall 1.10 0.10 0.128 0.015 -Tampa, FL, Norma Park 1.19 0.15 0.037 ---Orlando, FL, International Market Place 1.53 0.19 0.168 0.031 0 .008 0.013 DeBary, FL 0.761 0.26 0.028 0.01 0.0005 0.003 Bradfordville, FL 2.14 0.16 ----Tallahassee, FL, Cross Creek Mall 0.925 0.15 0.045 0.008 Sarasota County 0.88 0.31 ----Tampa, FL, Florida Aquarium 0.761 0.215 0.09 0.019 0.003 -Overall Mean 1.18 0.179 0.094 0.018 0.006 0.013 Figure 2 1. Unplanted plastic m at consisting of three 1.22 m by 2.44 m panels photo courtesy of Neal Beery.
43 Figure 2 2. Close up view of the FTW mat plastic matrix viewed from below the mat lo oking up photo courtesy of Neal Beery. Figure 2 3. Floating wetland treatment panel with PVC anchor attachment located in the lower left hand corner photo courtesy of Neal Beery.
44 Figure 2 4. Location of anchor attachment and p anels A, B and C. Figure 2 5 Aerial view of Gainesville, Florida with test locations outlined photo courtesy of USGS, 2012.
45 Figure 2 6. Aerial view of SS1 (Lake Alice), on the University of Florida campus photo courtesy of USGS, 2012. Figure 2 7. Ground le vel view of SS1 (Lake Alice).
46 Figure 2 8. Aerial view of SS2, Tumblin Creek stormwater pond, SW 6 th St. to the right photo courtesy of USGS, 2012.
47 Figure 2 9. Ground level view of SS2, Tumblin Creek stormwater pond photo courtesy of Neal Beer y.
48 Figure 2 10. Aerial view of SS3, RTS stormwater pond, with RTS parking lot to the right photo courtesy of USGS, 2012.
49 Figure 2 11. Ground level view of SS3, RTS stormwater pond photo courtesy of Neal Beery. Figure 2 12. Position of ta rget point s for initial plant sampling
50 Figure 2 13. Separated initial plant samples photo courtesy of Neal Beery. Figure 2 14 Locations of depth measurements prior to harvest.
51 Figure 2 15 Above m at, mat, and below mat z ones photos cou rtesy of Neal Beery
52 Figure 2 16 Division of p anel s into rows and blocks for m at z one stratified random sampling
53 CHAPTER 3 RESULTS AND DISCUSSI ON Measurements were made for assimilation of biomass, carbon, nitrogen, phosphorus, zinc, copper, cadmium and chro mium. Measurements of each of these constituents were taken for the Above Mat portions of four species groups ( J. effusus C. flaccida and P. cordata and self taken for all biomass of the Mat and Below Mat Zones with no separation of species. Measurements were taken for these separate groups on each of the three Panels (A, B, and C) of each of the three deployed FTWs (SS1, SS2, and SS3). These measurements were taken for the initially deployed FTWs and for the final growth of the FTWs at the end of the study. General Growth Characteristics Visually, the size and extent of coverage of the FTWs increased from the initial planting over the course of the study ( Figure 3 1 ). At the end of the period of study, the surface area of SS2 was almost entirely covered with macrophyte growth. Some surface area of SS3 was devoid of macrophytes, and large portions of the surface area of SS1, were devoid o f macrophytes ( Figure 3 2 ). The loss of individual macrophytes may have occurred due to grazing or trampling by wildlife. The occurrence of grass carp, Ctenopharyngodon idella was noted during the study period at SS1. Duri ng a high water period following heavy rain, grass carp were seen grazing on flooded grass banks, and it is speculated that they may have been grazing on the Below Mat portion of the SS1 macrophytes throughout the study. Common Moorhens, Gallinula chlorop us were seen on the SS3 FTW, and
54 evidence of their grazing was observed on the Above Mat portion of that FTW. The killing or grazing of macrophytes by turtles may also have occurred, as turtles were seen at SS3 and SS1, resting on top of crushed or bent macrophytes. At the end of the study period, the Mat Zone (approximately 5 cm thick) of the FTWs contained rhizome, root and petiole sections of macrophytes as well as algal growth and organic detritus ( Figure 3 3 ). When the FTWs were deployed into their respective stormwater retention ponds, the plastic matrix that made up the mat section was devoid of any material, except where macrophytes had been installed and passed through the mat. At the en d of the study period, where macrophytes remained or had been recruited, the Mat Zone was filled with live and non living biomass, mostly made up of rhizome sections of macrophytes. When the mat section had little biomass or other materials, light and a ir could easily pass through (Part A of Figure 3 3 ) from above the mat. At the end of study, when the Mat Zone was filled with organic material, light and air would not have been able to pass easily through the mat section, pos sibly stressing algal, bacterial and macrophytic organisms due to lack of light and oxygen. Also, such dense material, especially living rhizomes, proved to be exceedingly difficult to remove from the mat Zone during the harvest of the FTWs. This has imp lications for the management of FTWs as BMPs, as stormwater managers may either design the plastic matrix so that materials could be more easily removed, or leave the materials in the mat Zone and redeploy the FTWs since this material is less prone to senes cence and annual decomposition.
55 Where macrophytes were present, the root growth of FTWs was extensive ( Figure 3 4 ). Roots extended to depths (below the mat portion) of 10cm to 50cm in some cases. In so me cases roots grew in very thickly matted clusters, and contained organic detritus, inorganic minerals and sometimes aquatic vertebrate and invertebrates such as tadpoles, fish and crustaceans (Figures 3 5). FTWs were harvested carefully and manually i n this study, and any animals that were found were not included in measurements. However, if applied as a stormwater BMP, harvesting of FTW roots may be done in a manner that any animal biomass within root clusters would be harvested along with macrophyte biomass. Additionally, root clusters may have been thick enough to support anaerobic conditions that may lead to the occurrence of denitrification processes, a method of nutrient removal that was not measured by this study. Macrophyte Species Dominance a nd Recruitment Each of the three Sites were initially planted with the macropyhte species J. effusus C. flaccida and P. cordata These three species remained on all Panels of all Sites except for SS1, where at the end of the study P. cordata was not fo und on any Panel and J. effusus was not found on Panel A. During the study, FTWs were not managed through selective plant removal to retain only the originally installed species, thus allowing any new species to self recruit. Many species not installed i nitially were found on the sites at the end of the study ( Table 3 1 ). A variety of self recruited species were found at all three sites, and on all Panels at each of the Sites. At least twenty distinct species were identified at each Site, some of which were native to Florida and others were non native species ( IFAS, 2013 ). The
56 mass assimilation (biomass and other c gory rivaled or exceeded the final biomass of the three initially installed spe cies. Quantitative Assimilation portion of this chapter. Two of the self recruited species, Acer rubrum and Salix caroliniana are considered woody species and can become the size o f shrubs or trees. If allowed to grow large enough, their mass and density would likely change the buoyancy characteristics of the mat and eventually cause the FTW to sink. Species such as these would need to be periodically removed in order to preven t p ossible damage to the FTW. Nutrient and Metals Mass Assimilation The principle of stormwater treatment by FTWs relies on healthy initially installed macrophytes and any self recruited macrophytes to remove nutrients from the water column and to add this material to the FTW in amounts greater than what was installed (Kadlec and Wallace, 2009). The amount of initially installed mass is important to know in order to determine the net effect of the FTW once harvested. Table 3 2 lists the initial mean masses for the various constituents over the three study sites, along with standard deviations, minimum and maximum values for the Sites. There were no significant differences found among site s nor were there significant difference s found among Panels for total initial dry biomass (ANOVA, DF=2, The removal of pollutants by FTWs was found by taking the difference of the final mass of the constituent from the initially installed mass of the constituent. The mass removals for this study were normalized on a per square meter basis. T his normalization was done in order to allow managers to estimate the surface area
57 coverage required to reach removal goals, as well as for other researchers to make comparisons to other FTWs, MAPS, or BMPs in general. The mean site dry biomass assimilati on over the course of the study was 1 908 1330 g m 2 the minimum site mean assimilation was 655.8 g m 2 and the maximum site mean was 3749 g m 2 ( Table 3 3 ). This range illustrates the potential of FTWs to remove biomass for a representation of stormwater treatment basins in similar climactic conditions. Although not assimilated from the water column, carbon content of the biomass is also of interested to assess CO 2 assimilation as well as overall biomass characteristics. O ver the course of the study average mean carbon mass was 700.2 488.2 g m 2 The mean nitrogen and phosphorus assimilation s over the course of the study were 32.0 20.5 g m 2 and 4.25 2.44 g m 2 respectively. The average carbon, nitrogen and phosp horus content of the biomass were 34.51.9%, 1.810.04 % and 0.2 810.058 % respectively. Among the metals evaluated, Zinc was assimilated in higher amounts than any of the ot her metals, with a mean mass of 86.5 91.6 mg m 2 as compared to means of 9.43 9 .54 mg m 2 0.23 4 0.263 mg m 2 and 5.11 4.82 mg m 2 for copper, cadmium and chromium respectively. Significant differences were found between final and initial mass for each of the constituents (Wilcoxon Signed Rank, FTWs had significantly more mass, of all measured constituents, per square meter at the end of the study than they did when macrophytes were initially installed. The FTWs were capable of removing some am ount of all constituents over the period of the study,
58 and therefore can be considered an effective method of stormwater treatment when compared to non usage of FTWs. The mass assimilation s were converted to a rate (g m 2 day 1 ) by dividing the mass per sq uare meter results by the 278 day study period ( Table 3 4 ). This rate applies to this FTW design with an installation date of mid March and a harvest date of early to mid December. Table 3 5 shows c omparative mass assimilation rates for biomass, nitrogen and phosphorus across a variety of wetland systems, both natural and constructed. Generally, the results of mass assimilation rates found by this study fell within the widely varying range reported by a variety of plan t based water treatment systems. assimilation of 6.9 g m 2 d 1 are similar to that of Headley and Tanner (2008) who found a range of 2.7 5.9 g m 2 d 1 for FTWs in mesocosm settings, and are low whe n compared to the high of 33.9 g m 2 d 1 found by Hubbard et al. (2004) using FTWs in swine waste lagoons. Hogg and Wein (1988) reported a value of 1939 g dry biomass m 2 for natural floating mats populated by Typha latifolia This value is not a rate li ke the other values, but the mean dry biomass assimilated by FTWs in this study was comparable, with 1908 g dry biomass m 2 This study found a nitrogen assimilation rate of 0.115 g N m 2 day 1 which was comparable to other FTWs in the literature, but low er than emergent treatment wetlands. For nitrogen assimilation a low of .0017 g N m 2 d 1 was reported for FTWs in an urban stormwater setting (Chua et al., 2012), and a high of 2.2 2.6 g N m 2 d 1 was reported for emergent wetlands (Bachand and Horne, 2 000). A high assimilation rate of
59 1.096 g N m 2 d 1 was found by Hubbard et al. (2004) for an FTW system, which was substantially greater than the results of this study, and was most likely due to the extremely high nutrient content of the swine waste lag oon site. While considering these values it is important to note that the referenced studies may have included losses of nitrogen due to transformations and volatilization of nitrogenous compounds. The current study only measured the nitrogen that was ta ken up by macrophytes, which may be reflected in lower values when compared to other studies. Periphyton raceways studied by DeBusk et al. (2004) had low p hosphorus assimilation rates, 0.88 m g P m 2 d 1 when compared to the rate of 15.3 m g P m 2 d 1 f or this study. However, the high rate of 0.162 g P m 2 d 1 reported by Hubbard et al. (2004) for FTWs in swine lagoons was higher than that of this study, again, most likely due to the extremely high nutrient content of the sites at which the Hubbard stud y was conducted. The total mass of FTWs may have implications for the logistics and timing of the periodic assimilation of biomass. Three adults had difficulty manually removing a single 1.22 m by 2.44 m FTW Panel from most of the stormwater ponds. Also, the bulk biomass harvested from the FTW with the greatest growth (SS2) more than filled the bed of a conventional pickup truck. The harvesting plans for FTWs will need to include considerations for the wet weight and volume of biomass being harvested. T he overall management of FTWs will need to take into consideration the transport, storage and disposal of the removed biomass. At the end of the growing period, the FTWs were removed from the stormwater basin sites and the biomass was harvested and proce ssed for evaluation. After having
60 been harvested and prior to drying, the biomass contained the full weight of the water contained within the materials found on the FTW, and was referred to as the wet biomass. The mean final wet biomass of the FTWs was 1 4.7 kg m 2 with close to half of that mass being from the matted, spongy roots in the Below Mat Zone ( Table 3 6 ). Site SS2 had the greatest total wet biomass, with 28.8 kg m 2 and SS1 had the least total wet biomass, with 4.7 4 kg m 2 For FTWs of the dimension used in this study (2.44 m by 3.66 m), the mean final wet biomass per FTW was 131 kg, and the maximum FTW wet biomass was 257 kg ( Table 3 7 ). The FTW design used in this study easily allowed for panels to be removed separately; however, a mass of approximately 260 kg distributed evenly over three panels yielded a mass of around 86 kg per panel, which proved to be a difficult amount of biomass to maneuver and lift by hand. Additionally, the ma ss of the plastic FTW mat would add to the mass of the FTWs as a whole. Assimilation Performance by Panel Comparisons between Panels at each site were made in order to determine if an edge effect (where the middle Panel biomass and tissue characteristic s differed from the Panels along the outer edge ), possibly due to factors related to the FTW. These factors may have included light availability, air movement, temperature, dissolved oxygen levels, and or nutrient concentrations (Sasser, 1994). It was hy pothesized that the outer portions of the FTWs in this study would have greater biomass and pollutant assimilation capacity due to their location on the FTW which may have experienced higher light intensity and duration, higher air velocity, and higher nut rient concentrations because water would not have to pass through other roots before reaching the inner portions.
61 Visual inspection among panels at the end of the study did not app ear to have differences ( Figure 3 6 Figure 3 7 and Figure 3 8 ). No noticeable patterns in petiole height, root length, leaf coloration, or general health were evident between Panels. Visually, growth on SS2 and SS3 seemed evenly distributed in biomass and species diversity across the FTW whereas growth was sparse overall on SS1 and some species were present on some Panels but not on others. This result on SS1 may have been due more to the mortality of specific individuals by trampling or grazin g rather than patterns associated with an edge effect. When quantitatively comparing the biomass and tota l mass of nutrients and metals average mass in Panel B was consistently lower than Panels A or C, however no significant differences were found amo ng Panels ( Table 3 8 ), thus Hypothesis 2 was rejected. It had been hypothesized that a buffering effect would exist, where the area of the FTWs closest to the edge of each panel would receive more nutrients and have a higher di ssolved oxygen level, allowing for greater macrophyte growth and FTW performance. The amounts of dissolved nutrients, dissolved oxygen and nutrient containing suspended particles have been found to decrease with respect to distance away from their source in emergent wetlands (Reddy and DeLaune, 2008). In this study, the source was the free water surrounding the FTWs, and so it was thought that the interior portions of the FTWs would receive less. If this buffering effect did occur to any degree it had no effect on FTW performance for any constituent. This was most likely due to the scale of the FTWs in this study, where there was never more than 1.22 meters between any point on the FTW and open water. However, if larger surface area was covered in conti guous FTW panels, an effect may have been observed. Also, the
62 B Panels may have received the same amount of constituents as A and C Panels from free water below the FTWs. Additionally, no significant difference in overall tissue concentration between Pan els were found for any parameter ( Table 3 9 ) This indicates that nutrient and metal availability in the free water below the FTWs was not as affected by outer Panels or that water exchange with the roots is not unidirectional as originally expected. In addition, the lack of difference in tissue concentration between Panels suggested that growth of the macrophytes was not affected by other factors differing between Panels, or those facto rs did not differ among Panels. Panels A between any Panels, for both mass assimilation and tissue concentration, indicated the absence of an edge effect on FTWs with a surface area of 8.93 m 2 The lack of edge effect at the 2.44 m by 3.66 m dimension implies that stormwater managers implementing FTWs as MAPS need not be concerned with edge effects affecting FTW performance at this scale. However, i ncreasing the surface area of an FTW, or placing multiple FTWs adjacent to one another, may result in performance altering edge effects not seen in this study. Assimilation Performance by Species When designing and installing FTWs for treatment of storm water, the species that are installed must be considered for their performance capabilities. Headley and The choice of species often comes down to selecting locally occurring native species that exhibit vigorous growth within polluted waters under the local climatic conditions.
63 used for this study were recommended by ACF Environmental as part of their FTW he three initially installed species (as well as Spartina spp. ) as part of its Floating Wetland Design Criteria, and describes them as having been documented as having high nutrient tissue concentrations, but does not offer references or values (FDEP, 2010 b ). The species used for this study were J. effusus, C. flaccida, and P. cordata as well as additional self recruited species. The self recruited species were referred to as the Other group and consisted of any species not included in the three initiall y installed species groups. No management of individual macrophyte species was performed over the period of study to select for any particular species, and no pruning or harvesting was conducted prior to the end of the study period. This allowed for c omparisons to be made between the mass assimilation capabilities of the four Species groups at the end of the study. It is important to note that these among Species comparisons were only made for the Above Mat portions of the macrophytes, as the Mat and Below Mat portions were not able to be separated by species in a practical or accurate manner. It is also important to note that upon harvesting the Above Mat biomass was not washed or cleaned of biofilms (such as algae, periphyton, and bacteria) or detri tus; therefore, the per species concentration and mass data included these materials in addition to the tissue of the macrophytes. Macrophytes were installed as seedlings or young plants purchased from a nursery, and contained nutrients within their tissu e at the time of installation. This method of installing seedlings is seen as the most preferable method, as other studies
64 have suggested that it would lead to higher success rates (Headley and Tanner, 2006). As the macrophyte seedlings matured and grew on the FTWs they absorbed nutrients and metals from the free water below the mat and carbon from the atmosphere. As no growth media, such as peat, potting soil or plant fibers, was used for the FTWs in this study, the macrophytes relied entirely on the fr ee wa ter as a source of nutrients. Carbon, Nitrogen and Phosphorus Tissue Concentration by Species It is important to note that there were multiple species groups growing on the FTWs in this study. When comparisons between species groups were made, it w as a comparison of how different groups either thrived or declined on the same FTW Mat, not a comparison made between different FTWs installed with a single species group that did not compete with other species. When reading the following descriptions of n itrogen, phosphorus and carbon concentrations in dry biomass of the three macrophyte species and Other species groups refer to Table 3 10 Table 3 11 and Table 3 12 The concentrations given in these tables are for dried, ground and homogenized macrophyte tissue. The columns of the three sites for each groups for mean concentration (Wilcoxon signed rank test, N=3 that are not connected by the same letter are significantly different. Headley and Tanner (2006) concluded that Juncus species were able to grow well in floating wetland systems in a relatively similar environment (Auckland, New Ze aland), but another study found this species was not successful when used in swine lagoons or grown with one quarter strength Hoaglund solution (Hubbard et al., 2004).
65 The current study found a mean carbon content of 411.5 8.0 g C kg 1 and Ho (1979) re ported a comparable carbon content for dry J. effusus biomass (480 g C kg 1 ). The current study found a J. effusus dry biomass nitrogen concentration ranging from 9.36 to 10.89 g N kg 1 Professional Service Industries, Inc. (PSI) (2010), and McJannet et al. (1995) reported similar ranges of 8.9 to 12.6 g N kg 1 and 9.4 to 10.9 g N kg 1 for J. effusus dry biomass in an FTW application and natural populations, respectively. The current study found a J. effusus phosphorus concentration of 0.85 to 1.47 g P kg 1 which was similar to that of the 2010 PSI study reporting on FTW macrophytes (0.6 to 0.9 g P kg 1 ), but lower than that reported by McJannet et al. (2.1 to 3.5 g P kg 1 ) for emergent macrophytes. This may have been due to the fact that the McJannet et al. study was for emergent J. effusus that had access to phosphate in the soil porewater, while the 2010 PSI study was for floating J. effusus that only had access to the phosphate in the free water of the stormwater b asins. Canna flaccida has also been studied in FTW and emergent wetland applications. This study found a range for dry C. flaccida biomass carbon of 353.8 to 372.1 g C kg 1 ; no similar research was found to make a comparison of carbon content for Canna This study found the tissue nitrogen concentration for dry C. flaccida to range from 9.62 to 15.71 g N kg 1 which is similar to that of the PSI (2010) study (12.1 to 18.4 g N kg 1 ) and the Chang et al. (2012) study (9.6 g N kg 1 co ncentration). This study found the phosphorus concentration to range from 0.72 to 2.89 g P kg 1 which is similar to the 2010 PSI study (0.9 to 1.3 g P kg 1 ) (2010) and the Chang et al. study (2.5 g P kg 1 for dry shoot biomass).
66 This study found a carb on content ranging from 375.5 to 388.2 g C kg 1 for dry P. cordata content ranging from 403.3 to 430.1 g C kg 1 This study found a nitrogen content for dry P. cordata biomass of 9.74 to 14.02 g N kg 1 which was somewhat lower than 1 ) and a study performed by Holt et al. (1999) (39 g N kg 1 ). The Vogel thesis reported values for P. cordata growing on an FTW application in Florida that was more than doub le the value found at any site in the current study. This study found a phosphorus concentration range of 0.75 to 1.10 g P kg 1 which was similar to that of P. cordata in an FTW setting as found in the PSI study (.37 to 0.60 g P kg 1 ) but considerably lo wer than emergent P. cordata as shown by the Holt et al. study (5.1 g P kg 1 ). The high phosphorus content found by Holt et al. was most likely due to the emergent nature for that study, where higher phosphorus was available in the wetland soil at higher levels than that of the free water at SS1, SS2, or SS3. Also, iron plaque tends to accumulate on wetland soils due to oxidation of reduced soluble iron; this plaque often co precipitates phosphorus, which may have led to higher phosphorus concentrations in the root zone for emergent wetlands. The mean dry biomass carbon concentration found for the Other Species group at each site was 381.8 10.2 g C kg 1 Very little literature for the specific set of self recruited species in this study was found, bu t typical carbon content for wetland species has been found to be 38%, or 380 g C kg 1 a very similar value to that found in the current study (Radr, 2001). This study found the Other group to have a nitrogen concentration ranging from 11.6 to 13.0 g N kg 1 In a synthesis study of 41 different wetland species growing in emergent wetland conditions, McJannet et al. (1995) found
67 that the mean nitrogen concentration was similar, with a range from 10 to 14 g N kg 1 dry biomass. This study found a range of phosphorus content from 1.0 to 2.2 g P kg 1 for the Other species group, whereas the p hosphorus content for the 41 species in the McJannet et al. study ranged from 0.29 to 0.35 g P kg 1 dry biomass. The reasons for the differing ranges are not fully unde rstood, but could be due to the variety of species studied by McJannet et al. not being the same variety as studied in the current study. McJannet et al. (1995) found lower P concentrations in facultative wetland species when compared to obligate wetland s pecies. J. effusus is a facultative wetland species, C. flaccida and P. cordata are obligate wetland species, and the Other group contained both facultative and obligate wetlands species. J. effusus had a lower phosphorus concentration than C. flaccida a nd the Other species group, but higher than P. cordata Facultative and obligate wetland species designations were not an accurate predictor of relative phosphorus tissue content in this study. In the context of carbon fixation potential, Hypothesis 3 sta ted that the species group with higher concentrations reported in literature would have the highest concentration, and thus the highest potential to remove carbon. J. effusus was reported to have the highest carbon tissue concentration in the literature, and in the current study J. effusus was the speci es found to be significantly greater than any other species groups in this FTW study, therefore the hypothesis is confirmed for carbon. conce ntration (Kruskal P. cordata and the Other group had no significant difference, but J. effusus and C. flaccida differed from
68 each other and the two previously mentioned species groups, with J. effusus having a higher carbon concentration and C. flaccida having a significantly lower concentration. No significant difference was found among species groups for nitrogen and phosphorus dry biomass concentrations (Kruskal Wallis Rank Sum, N=4, DF=3, of nitrogen assimilation potential, Hypoth esis 3 stated that the species with higher concentrations reported in literature would also have the highest concentration in this study, and thus the highest potential to remove nitrogen. Though P. cordata was r eported to have the highest nitrogen tissue concentration in the literature, no species was found to significantly differ from any other species in this FTW study, therefore the hypothesis is rejected for nitrogen. In the context of phosphorus assimilati on potential, Hypothesis 3 stated that the species with higher concentrations reported in literature would have the highest concentration, and thus the highest potential to remove phosphorus. Though J. effusus was reported to have the highest phosphorus t issue concentration in the literature, no species was found to significantly differ from any other species in this FTW study, therefore the hypothesis is rejected for phosphorus. Metal Tissue Concentration by Species The mean dry biomass zinc concentration ranged from 28.3 15.1 mg kg 1 for the Other group to 39.1 19.9 mg kg 1 for C. flaccida ( Table 3 13 ). The mean dry biomass copper concentration ranged from 2.17 0.83 mg kg 1 for the Other group to 5.87 1.68 mg kg 1 for C. flaccida ( Table 3 14 ). For Above Mat biomass of the species groups at most of the sites, cadmium and chromium tissue concentrations were not found at levels high enough to exceed the minimum analytical detection limits, an d so were not included in the results of this study. For zinc and copper, the results of all
69 species groups in this study fell within the range, albeit the low end, reported in the literature. Fritioff (2005) reported concentrations of heavy metals in pl ant tissue grown in stormwater treatment areas; zinc ranged from 16 to 220 mg kg 1 dry biomass, copper ranged from 3 to 60 mg kg 1 dry biomass, and cadmium ranged from 0.04 to 35 mg kg 1 dry biomass. There was no significant difference between species gr oups for dry biomass zinc concentration (Kruskal concentrations, J. effusus C. flaccida and P. cordata did not differ significantly, but J. effusus and C. flaccida s lower concentration; the Other group and P. cordata had no significant difference (Wilcoxon Signed Rank, DF=3, reach minimum detection levels, and so were not included. In the context of heavy metal assimilation potential, Hypothesis 3 stated that the species with higher concentrations reported in literature would have the highest concentration, and thus the highest potential to remove metals. No species utilized in this st udy was reported in the literature to have specifically higher metals concentrations than others, therefore the hypothesis was neither confirmed nor rejected. Biomass Assimilation by Species Dry b iomass assimilation ( Table 3 15 ) for the Above Mat zone on a per square meter basis varied widely depending on the s pecies, with a minimum of 8.6 g m 2 for P. cordata at SS1 and a maximum of 861.6 g m 2 for C. flaccida at SS2. The mean biomass assimilation ranged less widely among av eraged across all sites ( Figure 3 10 ) with P cordata having the lowest assimilation ( 85.8 g m 2 ) and C. flaccida having the highest assimilation t (306.4 g m 2 ).
70 Significant differences were found between the Other group and J. effusus as well as between the Other group and P. cordata with the Other group being greater than J. effusus and P. cordata (Wil coxon Signed Rank N=3 difference was found between J. effusus C. flaccida and P. cordata althou gh the mean biomass Assimilation for C. flaccida was 306.4 392.6 g m 2 and the mean assimilation s for J. effusus and P. cordata were 88.6 101.9 g m 2 and 85.8 69.3 g m 2 respectively. This lack of significant difference among these Species groups w as likely due to the large standard deviation for C. flaccida possibly because of either the high mortality at SS1 or the dominance of C. flaccida growth at SS2 ( Figure 3 9 ) Carbon Assimilated by Species The site mean carbon assimilation ranged from a low of 32.8 g C m 2 for P. cordata to a high of 114.2 g 68.0 C m 2 for the Other group ( Table 3 16 ). There is little comparable literature for Above Mat carbon assimilation by FTWs. However, a the sis by Vogel (2012) that focused on FTWs in South Florida reported a range of 56 g C m 2 to 318 g C m 2 for all mat zones (which included what corresponds to the Above, Mat, dy. The maximum was that of C. flaccida on SS2 at 309.1 g C m 2 suggesting that this species can be utilized for maximized potential assimilation of biomass carbon in storm water systems in Florida. For all species groups, carbon assimilation was significantly greater than the per square meter initial mass of c arbon (Wilcoxon Signed Rank, N=3 demonstrating that all species groups were effective at assimilating new carbon. Across
71 all species groups for each panel at each site, carbon assimilation had a strong positive linear relationship with dry biomass assimilation as s hown in the following equation: Carbon Mass Assimilation, kg =0.365(Biomass Assimilation kg) + 0.834 R 2 =0.9972 The significant differences between species groups were the same as for dry biomass assimilation but differed from that of the dry biomass tissue carbo n concentration ( Table 3 16 ), indicating that it was the ability of FTWs to incorporate biomass, not increase carbon tissue concentrations, that drove the net assimilation of carbon in this study. When considering Hypothesis 3, C. flaccida and the Other group had approximately three times greater mean mass carbon assimilation than J. effusus and P. cordata The site mean carbon assimilation for C. flaccida did not significantly differ from the means for J. effusus or P. cord ata ; however, the large standard deviation for C. flaccida probably influenced this lack of significant difference. When considering the maximum mass assimilation performance of the four Species groups, C. flaccida and the Other group had the greatest pot ential to assimilate carbon, and so should be considered over the other two species when designing FT W systems to assimilate carbon. Nitrogen and Phosphorus Mass Assimilation by Species The Other species group had the highest site mean Above Mat nitrogen assimilation with 3.70 2.04 g N m 2 and J. effusus had the lowest with 0.86 0.92 g N m 2 ( Table 3 17 ). This range was considerably lower than the nitrogen mass assimilation Vogel reported, 10.0 to 34.7 g N m 2, in an FTW application for species including C. flaccida J. effusus and P. cordata as well as some of the same self recruited species. This may have been due to the longer growing season experienced rent study, across all
72 species groups for each panel at each site, nitrogen assimilation had a strong positive linear relationship with dry biomass assimilation as s hown in the following equation: Nitrogen Mass Assimilation g =10.64(Biomass Assimilation kg) + 0.330 R 2 =0.952 The results of this study for site mean phosphorus mass assimilation ranged from 21.7 41.2 mg P m 2 for P. cordata to 387.8 152.3 mg P m 2 for the Other species group ( Table 3 18 ). These results were comparable to the range of phosphorus assimilation found by White (2009) (47 to 184 mg P m 2 ), but the high and low assimilation values differed for C. flaccida and J. effusus J. effusus was higher, with a mean of 184 mg P m 2 assi milated when compared to 67.6 94.1 mg P m 2 C. flaccida was lower, with a mean of 47 mg P m 2 when compared to 234.6 mg P m 2 for the current study. It was hypothesized that the species with higher assimilati on s reported in the literature would have higher assimilation s in the current study. This was rejected, as species reported to have high assimilation s in the literature had low assimilation s in this study and vice versa. For the current study, across a ll species groups for each panel at each site, phosphorus assimilation had a positive linear relationship with dry biomass assimilation however the correlation was not as strong as for carbon or nitrogen. The relationship is s hown in the following equatio n: Phosphorus Mass Assimilation g =0.955(Biomass Assimilation kg) + 0.062 R 2 =0.664 For both nitrogen and phosphorus mass assimilation species groups had significant differences similar to that found for biomass assimilation (Wilcoxon Signed
73 assimilation characteristics across species, as opposed to varying concentration of nutrients. There were significant differences between the final and initial mean mass assimilation s for all species groups, and therefore all species groups were capable at removing nitrogen and phosphorus, confirming Hypothesis 1 for each species. Metal Mass Assimilation by Species The mass assimilation of zinc per squa re meter was lowest for P. cordata (3.01 4.06 mg Zn m 2 ) and highest for C. flaccida (20.70 28.56 mg Zn m 2 ) ( Table 3 19 ). The species groups significantly differed in the same pattern as for biomass assimilation (Wilcoxon assimilation was driven by biomass growth. The range of zinc assimilation was somewhat lower than that of the result by Fritioff (2005) for emergent macrophytes in a stormwater setting (48.8 mg Zn m 2 ), thou gh the Fritioff study was a measure of emergent macrophyte zinc in a stormwater setting. When considering Hypothesis 3 for metal mass assimilation there were no consistently reported results in the literature so the hypothesis could be neither confirmed nor denied. For the current study, across all Species groups for each Panel at each Site, zinc assimilation had a relatively strong positive linear relationship with dry biomass assimilation as s hown in the following equation: Zinc Mass Assimilation mg = 67.89(Biomass Assimilation kg) 14.10 R 2 =0.8848 The mass assimilation of copper was lowest for P. cordata (0.24 0.36 mg Cu m 2 ) and highest for C. flaccida (2.18 2.86 mg Cu m 2 ) ( Table 3 20 ). The Species groups si gnificantly differed in the same pattern as for biomass assimilation (Wilcoxon Signed assimilation was driven by biomass growth. The range of copper assimilation was comparable to that of the result by Fritioff (2005)
74 for emergent macrophytes in a stormwater setting (1 mg Cu m 2 ), though the Fritioff study was a measure of emergent macrophyte copper in a stormwater setting. It is important to note that for the means of J. effusus and P. cordata there was l zinc or copper mass per square meter and final zinc or copper mass per square meter. This indicates that over the course of the study the FTWs, on a mean basis, were a source of zinc and copper to the sites in which they were installed, with the metals installed in the was the relatively high tissue concentration in the initially installed individuals of J. effusus and P. cordata coupled with the loss of most or al l of the initial biomass (and therefore metals contained in the macrophyte tissue ) over the course of the study. Performance by Allocation of Biomass The design of the FTW used in this study led to the FTW being harvested in three distinct sections, the A bove Mat biomass, the Mat biomass and the Below Mat biomass. The entire FTW was removed from the water in order to facilitate the harvesting of the Below Mat and Mat biomass. The complete removal of the material that had grown within the Mat was difficul t and time consumi ng due to the small openings of the plastic matrix that made up the floating platform. When FTWs are deployed in the field these same issues may be experienced by managers, and therefore a harvesting scheme that minimizes the difficulty and time of harvesting efforts is important. Additionally, while addressing the issues of time and effort of harvesting, a harvesting scheme that efficiently and effectively removes the constituent(s) of interest is required. Examples of this would be a focus on the removal of zinc from a
75 stormwater pond near a manufacturing site, or the removal of nitrogen from stormwater near a heavily fertilized area. Targeted harvesting certain portions of the FTWs (Above Mat, Mat or Below Mat) may be ineffective due to their relative harvest difficulty, as well as their being low mass of the constituent(s) of interest. However, greater amounts of the constituent(s) of interest in a particular portion of the FTW may make harvesting an individual portion an effective means of pollutant removal while maintaining a critical propagule for future plant establishment without having to replant the FTW. This part of the study sought to determine which, if any, portion of the FTW were more or less effective at removing the va rious constituents. Carbon, Nitrogen and Phosphorus Concentration by Zone For th e current study, the Below Mat carbon biomass concentration was found to be similar to that of the Above Mat and Mat Zones with a site mean of 331.5 41.2 for Below Mat, 3 38.0 49.1 for Above Mat and 367.0 7.4 g kg 1 for the Mat ( Table 3 21 ) The mean carbon biomass concentrations did not significantly differ between zones. A study by Xian et al. (2007) conducted in Scotland found that the carbon biomass concentration in the shoots of J. effusus (shoot concentrations are analogous to the Above Mat Zone in the current study, and root concentrations are analogous to Below Mat) were similar to that of the roots across three sites (479 g C kg 1 470 g C kg 1 and 468 g C kg 1 for roots, 479 g C kg 1 477 g C kg 1 and 481 g C kg 1 for shoots, respective of sites). Hypothesis 4 is supported by this data as previous research has found carbon concentrations to be similar for Above Mat and Below M at Zones. For the current study, the Below Mat nitrogen biomass concentration was found to be greater than that of the Above Mat and Mat Zones with a site mean of 24.1 2.5 for Below Mat, 10.5 1.0 g kg 1 for Above Mat and 19.6 0.3 g kg 1 for the Mat Zone
76 ( Table 3 22 ) All three zones were found to be significantly different from each other. At the end of the study conducted by PSI (2010), it was found that the nitrogen biomass concentration in the shoots of P. cordata ( shoot concentrations are analogous to the Above Mat Zone in the current study, and root concentrations are analogous to Below Mat) was higher than that of the roots (5.0 g N kg 1 for roots, 14.0 g N kg 1 for shoots). For two other species, C. flaccida and J. effusus PSI (2010) found that root nitrogen biomass concentration was slightly higher than that of shoot nitrogen concentration ( C. flaccida root was 22.0 g N kg 1 shoot was 18.4 g N kg 1 ; J. effusus root was 13.7 g N kg 1 shoot was 10.7). These ni trogen concentrations were similar to those found in the current study, and the results of generally higher concentrations in the root (Below Mat) was similar as well, thus supporting Hypothesis 4 for nitrogen concentrations. Below Mat phosphorus biomass c oncentration was found to be greater than that of the Above Mat and Mat Zone s, with a site mean of 4.3 4 0.92 g P kg 1 for Below Mat, 1.1 1 0.29 g P kg 1 for Above Mat and 2.98 0.84 g P kg 1 for the Mat Zone ( Table 3 23 ) All three zones were found to be significantly different from each other. At the end of the study conducted by PSI (2010) it was found that the phosphorus biomass concentration in the shoots of P. cordata were similar (0.4 g P kg 1 for roots, 0.37 g P kg 1 for shoots). In the PSI study, for two species, C. flaccida and J. effusus it was found that root phosphorus biomass concentration was similar, and considerably higher than that of shoot phosphorus concentration, respectively ( C. flaccida root was 1.5 g P kg 1 shoot was 1.3 g P kg 1 ; J. effusus root was 1.2 g P kg 1 shoot was 0.6 g P kg 1 ). 1 and a leaf/shoot concentration of around 2.0 g P kg 1 were found in a study by Hadad and Maine (2007). The
7 7 phospho rus concentrations found in the two referenced studies were on the same order of magnitude as those found in the current study, though slightly lower overall and the results of generally higher concentrations ( J. effusus roots in the PSI study were found to have higher phosphorus concentrations than shoots) in the Below Mat for the current study was found to be at differ with the PSI study. Hypothesis 4 is not fully supported by this data for phosphorus concentrations by zone when considering all species reported in the referenced literature. Metal Concentration by Zone Although specific data was not given by Fritioff (2005), the general statement was made that across several emergent wetlands species; Zn, Cu, Cd and Cr were found to accumulate in greater concentration in roots than shoots for rooted wetland plants. This conclusion was explained by the bioavailability of heavy metals in the soils and sediments of the sites. Although this same study reports that free floating wetland species, such as Lemn a minor and Eichornia crassipes have been found to accumulate higher levels of heavy metals than emergent species, no comparisons between roots and shoots was made. The current study found that Below Mat metal concentrations were significantly greater tha n those of Above Mat Zones for the metals Cu, Cd, and Cr, but Zn concentrations did not significantly differ by zone. For Cu, Cd and Cr the Below Mat concentrations were higher than those of the Above Mat, and slightly higher than those of the Mat Zone ( T able 3 24 Table 3 25 Table 3 26 and Table 3 27 ). Studies by Rai (2008) as well as Deng et al. (2004) found mean Zn biomass concentrations across several emergent wetland species to be slightly higher in the root zone when compared to the shoot zone (Rai: 56.4 mg Zn kg 1 for roots, 41.7 mg Zn kg 1 for shoots; Deng et
78 al.: 78 mg Zn kg 1 for roots, 36 mg Zn kg 1 for shoots). For Cu, root biom ass concentrations were found to be much greater than shoot concentration (Rai: 732 mg Zn kg 1 for roots, 73.6 mg Zn kg 1 for shoots; Deng et al.: 871 mg Zn kg 1 for roots, 67 mg Zn kg 1 for shoots). It was hypothesized that the partitioning of constitu ents by zone would follow patterns found in previous field studies. This hypothesis was rejected for Zn, and supported for Cu. Zones did not significantly differ for Zn partitioning, and the Below Mat Zone was greater than (and significantly different fr om) the Above Mat portion, as was found by previous studies. Biomass Assimilate d by Zone On an average sites basis, biomass harvested was the great est in the Above Mat Zone (856 649 g m 2 ), and the Below Mat Zone had the lowest biomass harvested (486 4 00 g m 2 ) ( Table 3 28 ). However there were no significant di fferences found between Zones. The current study found that 44.9% of the biomass harvest was in the Above Mat Zone, and 55.1% was found in the Mat and Below Mat Zon es, combined (29.7% from Mat, and 25.4% from Below Mat) ( Figure 3 11 ). PSI (2010) divided their individual macrophytes into two positions, root and shoot, rather than three (Above Mat, Mat, and Below Mat, with Above Mat and Be low Mat being analogous to shoot and root, respectively), however an approximate comparison between the current study and the PSI study can be made. Though, the exact sampling technique was not described in the PSI paper it is assumed that the rhizome sec tion of the plant was included as the Agrostis alba P. cordata C. flaccida and J. effusus ) that made up the majority of the biomass in their
79 study, PSI (2010) found 27.6% of t difference between shoots and roots, the results of this study were similar in that the greater biomass harvest than the Above Mat Zone. A study by Ladislas et al. (2013), which was a similar FTW study involving floating macrophytes, found that for J. effusus the shoots made up 47.1% of the total biomass, while roots made up 52.9%, and for Carex riparia root biomass made up 47.7% and shoot biomass made up 52.3%. These results were similar to that of the current study. For a study by Hadad and Maine (2007), the results were basically similar to that of the current study, with approximately half of the biomass being found in the biomass harvest, in that the combined Mat and Below Mat Zones had greater biomass harvest; however, it is important to note that there were no statistically significant differences between any of the three groups. Carbon Mass harvested by Zone The harvest of carbon on a mean site basis mirrored the harves t of biomass, with the Above M at Zone removing the most (323 244 g m 2 ) and the Below Ma t Zone removing the least (168 141 g m 2 ) ( Table 3 29 ). As with biomass, there were no significant differences between zones for carbo n assimilation Patterns evident in the harvest of biomass by zone were consistent with patterns for the assimilation of carbon by zone. The percent of total carbon harvest within each zone was found to be very similar to that of the percent of total biom ass harvested by zone, with 46.1% of the total carbon
80 mass assimilation in the Above Mat Zone and 53.9% for the Mat and Below Mat Zone combined ( Figure 3 12 ). As with the comparisons by species, it can be said that the biomass found in each zone was driving the mass of carbon found in each zone. As carbon is typically found to be directly related to the total biomass of wetland macrophytes (Hadad and Maine, 2007), the similarities between carbon and biomass are not surprising, both for the current study and the comparisons to other studies in the literature. Previous studies on partitioning of carbon assimilation by zone were not found on an area basis for floating macrophytes, therefore Hypothesis 4 could be neither supported nor rejected. Nitrogen and Phosphorus Mass Harvest by Zone The Below Mat Zone assimilated the greates t amount of nitrogen (11.4 8.58 g m 2 ), on a mean site basis, whereas the Above Mat Zone assimilated the least (9.5 5 6.17 g m 2 ) ( Table 3 30 ). Again, there were no significant differences found between zones for nitrogen assimilation s. This was expected, as the mean assimilation s per zone were within 2 grams per square meter of one another. This overall even distribution of nitrogen throughout the Zones is important for the management of excess nitrogen in surface waters, as it shows that no one zone can be discounted when considering harvesting efforts or FTW design. This will be discussed furt her in the conclusions sec tion. In a study by Chen et al. (2009) an ornamental species of Canna lily had 86.9% of nitrogen assimilation performed by what is analogous to its Above Mat Zone, 5.4% by its Mat Zone, and 7.5% by its Below Mat Zone. The results of the current study do n ot reflect the results of Chen et al. (2009). For P. cordata in the same Chen et al. study the percent assimilation by position was as follows: 85.7% for Above Mat, 2.1% for Mat,
81 and 12.1% for Below Mat. Again, these results were not similar to what was found in the Chen et al. study. As the Chen et al. study would have predicted that the Above Mat Zone would have assimilated a greater portion of the total nitrogen than the Mat and Below Mat Zones combined, Hypothesis 4 is rejected for nitrogen mass ass imilation This may have been due to the late season harvest of the FTWs, which could have allowed time for the wetland macrophytes to begin shifting their nutrients to their roots and rhizomes for storage. The Below Mat Zone assimilated the greatest amo unt of phosphorus (1.83 1.35 g m 2 ), on a mean site basis, whereas the Above Mat Zone assimilated the least (0.90 0.42 g m 2 ) ( Table 3 31 ). Again, there were no significant differences found between Zones for phosphorus as similation s. However, the maximum phosphorus assimilation for the Below Mat Zone was nearly twice that of the maximum assimilation for the Above Mat Zone. The greater amounts of phosphorus assimilated by the Mat and Below Mat Zones could have implication s for the design and maintenance of FTWs when the focus is on the removal of phosphorus. The current study found that 21.2% of the phosphorus assimilation was in the Above Mat Zone, and 78.9% was found in the Mat and Below Mat Zones, combined (with 35.7% a s Mat, and 43.2% as Below Mat) ( Figure 3 14 ). In a study by Chen et al. (2009) an ornamental species of Canna lily had 93.5% of phosphorus assimilation performed by what is analogous to its Above Mat Zone, 2.4% by its Mat Zone and 4.0% by its Below Mat Zone. The results of the current study do not reflect these results. For P. cordata in the same Chen et al. (2009) study the percent assimilation by Zone was as follows: 91.7% for Above Mat, 1.7% for Mat, and 6.6% for Below M at. Again, these
82 results were not similar to what was found in this study. As this other study would have predicted that the Above Mat Zone would have assimilated a greater portion of the total phosphorus than the Mat and Below Mat Zones combined, Hypoth esis 4 is rejected for phosphorus mass assimilation This result could have been due to adsorption of phosphorus to the detrital organic matter and mineral matter caught within the root mats (Sasser et al., 1991) that made up most of the Below Mat Zone of each FTW. Metal Mass Assimilation by Zone In general, it has been concluded in the available literature that metals are found in greater amounts in the lower portions of wetland plants, namely the roots and rhizomes. This is due to the adsorption and abs orption of the roots that are in close free water, this may not necessarily be the case (Fritioff, 2005). However, the matted root systems of FTWs also provide surface area in which mineral and organic materials are caught and held; these materials provide binding sites for metals, or may contain metals themselves (Headley and Tanner, 2008). The assimilation of Zn on a mean site basis mirrored the assimilation of biomas s, with the Above Mat Zone removing the most (39.2 46.4 mg m 2 ) and the Below Mat portion removing the least (20.2 20.3 mg m 2 ) ( Table 3 23 ). As with biomass, there were no significant differences between zones for zinc as similation It is also interesting that while the site minimum assimilation s for Zn are very close for the Above Mat and Below Mat Zones, the maximum Above Mat Zn assimilation for a single site is more than twice that of the maximum Below Mat Zn assimilat ion for a single site. This indicates that the potential for the Above Mat Zone to remove Zn is possibly greater than the potential for t he Below Mat Zone to remove Zn.
83 The percent of total Zn mass assimilated was found to be similar to that of the perce nt assimilation s by zone for biomass (and carbon). The current study found that 45.3% of the biomass assimilation was in the Above Mat Zone, and 54.7% was found in the combined Mat and Below Mat Zones, combined (with 31.3% as Mat, an d 23.4% as Below Mat) ( Figure 3 15 ). Studies by Deng et al. (2009), as well Fritioff and Greger (2003), measured the mass of various heavy metals within the tissues of a suite of wetland species and concluded that, with few exceptions, heavy metals are mostly found in the roots of these plants. This is partly supported by the study by Ladislas et al. (2013) which found that for J. effusus the shoots made up 58.4% of the total zinc, while roots made up 41.6%, and for Carex riparia shoot biomass made up 35.5% and root biomass made up 64.5%. So, for Carex riparia the Ladislas et al. study agrees with the other studies, but for J. effusus the study does not agree. The current study found that when roots and rhizomes were combined (as Below Mat and Ma t), they assimilated greater amounts of Zn than the shoot portion (Above Mat Zone). Though some literature is unclear as to whether upper or lower portions of wetland plants should be expected to be able to remove Zn in higher amounts, most of the literat ure found seems to agree with the results of the current study, thus Hypothesis 4 is supported for Zn mass assimilation The assimilation of Cu was around one tenth that of the assimilation of Zn, however, the highest and lowest assimilation s by zone were not consistent between the two metals. Although no zone significantly differed from the rest, the Above Mat Zone had the highest assimilation (3.5 3.8 mg m 2 ) and the Mat Zone (not the Below Mat Zone as it was with Zn) had the lowest assimilation of Cu (2.5 2.0 mg m 2 ) ( Table 3
84 33 ). Another way that Zn and Cur assimilation patterns differed between zones is that the minimum assimilation s as well as the maximum assimilation s were similar for the Above Mat and Below Mat Zone s. These similar assimilation patterns, paired with the fact that the Above Mat and Below Mat Zones outperformed the Mat Zone resulted in the combined Above Mat and Below Mat Zones removing 73.2% of the Cu (37.3% for Above Mat, 35.9% for Below Mat), wit h the Mat Zone only removing 26.8 % ( Figure 3 16 ). From a management perspective, the majority of Cu being found in the Above Mat and Below Mat Zones allows for managers concerned with the metal to make an informed decision on their design and maintenance of an FTW. Previous studies on partitioning of heavy metal assimilation by zone found the Below Mat portions to be higher, whereas this study found no significant difference among zones for Cu assimilation there fore Hypothes is 4 was rejected. Cadmium assimilation did not occur at an appreciable level (by concentration or mass), and will not be discussed. The assimilation of Cr was less than one tenth that of the assimilation of Zn, but close to the assimilation of Cu by zone in order of magnitude. However, a distinctly different pattern of assimilation by zone was found for Cr. The Mat and Below Mat portions were greater than that of the Above Mat Zone, and were significantly different (both the Mat and Below Mat Zones were significantly different from the Mat Zone). The Above Mat Zone had the lowest assimilation (0.46 2 0.40 3 mg m 2 ) and the Below Mat Zone had the highest assimilation of Cr (2.63 2.59 mg m 2 ) ( Table 3 34 ). The maximum site assimilation for the Above Mat Zone did not even reach the average assimilation s for Mat and Below Mat.
85 The low performance of the Above Mat Zone for Cr assimilation is pronounced when considering the percentage of Cr assimilation by zone, with the Above Mat Zone only representing 9.0% of the total, and the combined Mat and Below Mat Zones removing 91.0 % (39.6% for the Mat Zone, and 51.4% for the Below Mat Zone) ( Figure 3 17 ). From a management perspective, almost all of th e assimilated Cr being found in the Mat and Below Mat Zones allows for managers concerned with the metal to make an informed decision on their design and maintenance. Leaving the Mat Zone (which is more difficult and possibly costly to harvest) in place o n the FTW after harvest would leave nearly forty percent of the Cr behind. Previous studies on partitioning of heavy metal assimilation by Zone found the Below Mat portions to be higher. This study found the Below Mat Zone to be significantly greater tha n the Above Zone for Cr assimilation, therefore Hypothesis 4 was supported
86 Table 3 1. Self recruited macrophyte species per s ite. Species Common Name SS1 SS2 SS3 Acer rubrum red maple X Alternanthera philoxeroides alligator weed X Aster elliottii Elliott's aster X Aster lateriflorus calico aster X Bidens laevis bur marigold X X Carex unk. sedge X Cyperus odoratus fragrant flatsedge X X X Cyperus polystachyos flat sedge X X Cyperus sp. spikerush X Cyperus surinamensis tropical fl atsedge X Diodia virginiana Virginia buttonweed X Eclipta alba yerba de tajo X X Eleocharis sp spikerush X X Erigeron vernus early whitetop fleabane X Eupatorium capillifolium dog fennel X X X Galium tinctorum stiff marsh bedstraw X Hydroc otyle umbulata pennywort X X X Juncus sp. rush X Limnobium spongia American spongeplant X Ludwigia leptocarpa angelstem primrose willow X X Ludwigia peruviana water primrose X X Ludwigia repens red ludwigia X X X Mikania scandens c limbing hempv ine X Phyla nodiflora matchstick X X Pistia stratiotes bristlegrass X Pluchea odorata sweetscent X X Pluchea rosea rosy camphorweed X X X Poaceae sp. grass, unknown XX XX XX Polygonum hydropiperoides wild water pepper X X Polygonum sp. knotwe ed X Rhynchospora caduca beak rush X Rhynchospora odorata fragrant beak rush X Sacciolepis striata American cupscale grass X X Sali x caroliniana coastal plain willow X X X Salvinia minima water spangle X Setaria geniculata Knotroot foxtail X Setaria viridis green bristlegrass X X Typha latifolia common cattail X Wolffia sp. watermeal X Note: XX indicates multiple species of a genus were present.
87 Table 3 2. Initial mean constituent mass loading from planted macrophytes. Paramet er Unit Mean Std. Dev. Minimum Maximum Dry Biomass (g m 2 ) 253.9 26.9 216.2 277.1 Carbon (g m 2 ) 97.0 10.7 81.9 106 Nitrogen (g m 2 ) 3.77 0.40 3.21 4.14 Phosphorus (g m 2 ) 0.828 0.086 0.710 0.912 Zinc (mg m 2 ) 13.1 1.40 11.2 14.3 Copper ( mg m 2 ) 2.07 0.22 1.76 2.25 Cadmium (mg m 2 ) 0.026 0.002 0.022 0.028 Chromium (mg m 2 ) 0.564 0.057 0.484 0.615 Table 3 3. Overall mean per square meter mass assimilation. Parameter Unit Mean Std. Dev. Min Max W p value Biomass (g m 2 ) 1908 1330 655 .8 3750 6.67 <0 .0001 Carbon (g m 2 ) 700.2 488.2 227.3 1372 6.78 <0 .0001 Nitrogen (g m 2 ) 32.0 20.5 10.8 59.7 5.83 <0 .0001 Phosphorus (g m 2 ) 4.25 2.44 2.25 7.68 5.12 <0 .0001 Zinc (mg m 2 ) 86.5 91.6 18.3 216 4.55 <0 .0001 Copper (mg m 2 ) 9.43 9.54 2.11 22.9 4.35 <0 .0001 Cadmium (mg m 2 ) 0.23 4 0.26 3 0.047 0.606 2.59 0 .0225 Chromium (mg m 2 ) 5.11 4.82 1.64 11.92 3.40 <0 .0001 Means, standard deviations, minimums and maximums are on a mass assimilation per Site basis. The W and p values were f ound using the Wilcoxon S ign ed R ank test for the significant differences between Final and Initial mass. Table 3 4 Biomass assimilation rate per square meter per day for all parameters studied. Parameter Unit Mean Std. Dev. Min Max Dry Biomass (g m 2 d 1 ) 6.86 4.78 2.36 13.5 Carbon (g m 2 d 1 ) 2.52 1.76 0.818 4.94 Nitrogen (g m 2 d 1 ) 0.115 0.074 0.039 0.215 Phosphorus ( m g m 2 d 1 ) 15.3 8.8 8.09 27.6 Zinc ( m g m 2 d 1 ) 0.311 0.330 0.066 0.777 Copper ( m g m 2 d 1 ) 0.034 0.034 0.008 0.0 82 Cadmium ( m g m 2 d 1 ) 0.0008 0.0009 0.0002 0.0022 Chromium ( m g m 2 d 1 ) 0.018 0.017 0.006 0.043
88 Table 3 5. Comparison of nutrient assimilation rates in this study with other s tormwater treatment systems. Reference Description Mass Assimilation Biomass Nitrogen Phosphorus ( g m 2 d 1 ) ( g m 2 d 1 ) (m g m 2 d 1 ) This study stormwater FTW, Florida 6. 86 0.115 15.3 FTW, urban runoff Vetiver 0.0017 0 16 FTW, urban runoff Polygonum 0.0028 0 4 Hubbard et al (2004) FTW, swine waste, Typha latifolia 33.9 1.097 162.4 FTW, swine waste, Juncus effusus 1.3 0.046 6.57 FTW, swine waste, Panicum hem atomon 20.0 0.663 98.56 FTW, 1/4 strength Hoaglund solution Typha latifolia 22.7 0.718 126.89 FTW, 1/4 strength Hoaglund solution Juncus eff. 11.2 0.263 38.6 FTW, 1/4 strength Hoaglund solution Panicum hem. 18.3 0.410 55.64 Tanner et al. (2011) F TW, batch fed mescosm 0.64 0 .76 54 58 FTW, flow through mesocosm, low high loading 0.16 0 .24 2.3 5.4 Tanner and Headley (2008) FTW mesocosm (**rhizome mass not included) 2.7 5.9 ** Wen and Recknagel (2002) FTW, Myriophyllum, P aspalum and Ranunculus 43 86 Jangrel Bratli (2011) s tormwater FTW, unfertilized 0.17 8.5 s tormwater FTW, fertilized 0.16 3220 Hogg and Wein (1988 ) Natural floating Typha mats (*g m 2 found in situ) 1939* Chang et al. (2002) emerg ent mesocosm, Scirpus and Pontederia 0.036 1.5 Durham emergent treatment wetland, swine waste, Typha 1.35 200 Bachande and Horne (2000) emergent treatment wetland 2.2 2.6 Spieles and Mitsch (2000) emergent treatment wetland, Ohio 18 20 US EPA ( 1999 ) emergent treatment wetlands 0.2 0 60 DeBusk at al (2004) periphyton raceways Florida Everglades 0.88
89 Table 3 6. Fina l wet biomass averaged among all three deployment locations. Final Wet Biomass (kg m 2 ) Mean Minimum Maximum Above Mat 4.20 1.72 8.63 Mat 4.10 1.57 6.76 Below Mat 6.39 1.45 13.4 All 14.7 4.74 28.8 Table 3 7. Overall final wet mass per FTW. Fina l Wet Biom ass (kg per FTW) Mean Minimum Maximum Above Mat 37.5 15.3 77.0 Mat 36.6 14.0 60.3 Below Mat 57.1 13.0 120.0 All 131 42.3 257 Note: FTW dimensions were 2.44 m by 3.66 m Table 3 8. Mean net mass accumulation by p anel. Parameter Unit A B C Biomass (g m 2 ) 689 473 590 320 630 549 Carbon (g m 2 ) 252 172 219 121 229 200 Nitrogen (g m 2 ) 11.7 7.9 10 .3 4.8 10.3 8.1 Phosphorus (g m 2 ) 1.63 1.07 1.33 0.49 1.29 0.91 Zinc (mg m 2 ) 25.2 24 23.1 21.7 38.1 45.9 Coppe r (mg m 2 ) 3.76 4.07 2.59 2 .00 3.07 3.48 Cadmium (mg m 2 ) 0.145 0.189 0.0473 0.0391 0.0416 0.0357 Chromium (mg m 2 ) 2.324 2.499 1.32 0.919 1.468 1.398 *Note: There were no statistical differences found between the Kruskal Wallis test, blocked by Site. Degrees of freedom for all tests was 2, and number of samples per Panel was 3.
90 Table 3 9. Mean tissue c oncentration by p anel. Parameter Unit A B C Carbon (g m 2 ) 347 25 360 15 349 26 Nitrogen (g m 2 ) 15.9 1.2 16.8 1.1 17.4 0.7 Phosphorus (g m 2 ) 2.31 0.24 2.41 0.25 2.57 0.49 Zinc (mg m 2 ) 29.7 7.5 43.3 13.8 37.7 10.6 Copper (mg m 2 ) 5 .00 2.53 4.67 2.21 4.4 0 1.6 0 Cadmium (mg m 2 ) 0.123 0.102 0.0775 0.0256 0.0789 0.0365 Chr omium (mg m 2 ) 2.38 1.27 2.36 0.68 2.3 0 0.67 Kruskal Wallis test blocked by Site Degrees of freedom for all tests was 2, and nu mber of samples per Panel was 3 Table 3 10. Carbon biomass concentration (g kg 1 ) by species. Mean St. Dev. Min Max Sig. Diff. Juncus effusus 411.5 8.0 400.3 418.4 A Canna flaccida 361.7 7.7 353.8 372.1 B Pontederia cordata 381.9 6.3 375.5 388.2 C Other 381. 8 10.2 367.6 390.8 C Note: Species with similar letters were found to have no significant difference for their Site means using the Wilcoxon Signed Rank test for significant difference ( = 0.05, DF=3 n=3 ), blocked by Site. Table 3 11. Nitrogen biomass concentration (g kg 1 ) by species. Mean St. Dev. Min Max Sig. Diff. Juncus effusus 10.23 0.64 9.36 10.89 A Canna flaccida 12.13 2.60 9.62 15.71 A Pontederia cordata 11.88 2.14 9.7 4 14.02 A Other 12.51 0.62 11.64 12.98 A Note: Species with similar letters were found to have no significant difference for their Site means using the Wilcoxon Signed Rank test for significant difference ( = 0.05, DF=3 n=3 ), blocked by Site. Table 3 12 Phosphorus biomass concentration (g kg 1 ) by species. Mean St. Dev. Min Max Sig. Diff. Juncus effusus 1.09 0.27 0.85 1.47 A Canna flaccida 1.72 0.89 0.72 2. 89 A Pontederia cordata 0.93 0.17 0.75 1.10 A Other 1.51 0.52 1.06 2.23 A Note: Specie s with similar letters were found to have no significant difference for their Site means using the Wilcoxon Signed Rank test for significant difference ( = 0.05, DF=3 n=3 ), blocked by Site.
91 Table 3 13. Site mean zinc (mg kg 1 ) concentrations by species. Mean St. Dev. Min Max Sig. Diff. Juncus effusus 31.2 8.90 19.2 40.6 A Canna flaccida 39. 1 19.9 19.1 66.3 A Pontederia cordata 34.6 20.9 13.7 55.5 A Other 28.3 15.1 15.6 49.6 A Note: Species with similar letters were found to have no significant difference for their Site means using the Wilcoxon Signed Rank test for significant difference ( = 0.05, DF=3 n=3 ), blocked by Site. Table 3 14. Site me an copper (mg kg 1 ) concentrations by species. Mean St. Dev. Min Max Sig. Diff. Juncus effusus 5.48 1.75 3.36 7.66 A Canna flaccida 5.87 1.68 3.62 7.65 A Pontederia cordata 3.07 2.05 1.03 5.12 A B Other 2.17 0.83 1.14 3.17 B Note: Species with similar letters were found to have no significant difference for their Site means using the Wilcoxon Signed Rank test for significant difference ( = 0.05, DF=3 n=3 ), blocked by Site. Table 3 15 Above m at net dry biomass assimilation (g m 2 ) by species. Mean St. Dev. Min Max Sig. Diff. F > I Juncus effusus 88.6 102 6.21 228 B Canna flaccida 306 393 21.9 862 A B Ponted eria cordata 85.8 69.3 8.63 156 B Other 296 174 105 527 A Note: In Sig. Diff. column, Species not connected by the same letter are significantly DF =3, n=3), blocked by Site. In F>I column, an asterisk (*) indicates there was significantly more Final mass for that Species than Initial mass, and NS indicates there was no significant difference between Final mass and Initial mass (Wilcoxon Signed Rank,
92 Table 3 16 Above mat c arbon (g m 2 ) assimilation by s pecies. Mean St. Dev. Min Max Sig. Diff. F> I Juncus effusus 36.9 42.4 2.42 95.8 B Canna flaccida 110 141 7.73 309 A B Pontederia cordata 32.8 26.2 3.21 58.4 B Other 114 68.0 38.7 204 A Note: In Sig. Diff. column, Species not connected by the same letter are significantly DF=3, n=3), blocked by Site. In F>I co lumn, an asterisk (*) indicates there was significantly more Final mass for that Species than Initial mass, and NS indicates there was no significant difference between Final mass and Initial mass (Wilcoxon Signed Table 3 17 Above m at n itrogen (g m 2 ) assimilation per species. Mean St. Dev. Min Max Sig. Diff. F> I Juncus effusus 0.862 0.923 0.032 2.13 B Canna flaccida 3.03 3.75 0.223 8.33 A B Pontederia cordata 0.940 0.753 0.122 1.55 B Other 3.70 2.04 1.43 6.39 A Note: In Sig. Diff. column, Species not connected by the same letter are significantly DF=3, n=3), blocked by Site. In F>I column, an asterisk (*) indicates there was significantly more Final mass for that Species than Initial mass, and NS indicates there was no significant difference between Final mass and Initial mass (Wilcoxon Signed Table 3 18 Above m at phosphorus (mg m 2 ) assimilation per species. Mean St. Dev. Min Max Sig. Dif f. F> I Juncus effusus 67.6 94.1 14.5 199 B Canna flaccida 235 266 29.4 610 A B Pontederia cordata 21.7 41.2 32.7 66.9 B NS Other 388 152 252 601 A Note: In Sig. Diff. column, Species not connected by the same letter are signific antly DF=3, n=3), blocked by Site. In F>I column, an asterisk (*) indicates there was significantly more Final mass for that Species than Initial mass, and NS indicat es there was no significant difference between Final mass and Initial mass ( Wilcoxon Signed
93 Table 3 19 Above m at zinc (mg m 2 ) assimilation, per s pecies. Mean St. Dev. Min Max Sig. Diff. F> I Juncus effusus 3.24 4.26 0.64 9.17 B NS Canna flaccida 20.7 28.6 0.388 61.1 A B Pontederia cordata 3.01 4.06 0.440 8.71 B NS Other 10.9 10.6 1.83 25.8 A Note: In Sig. Diff. column, Species not connected by the same letter are significantly different. Statistical tests were DF=3, n=3), blocked by Site. In F>I column, an asterisk (*) indicates there was significantly more Final mass for that Species than Initial mass, and NS indicates there was no significant difference b etween Final mass and Initial mass (Wilcoxon Signed Table 3 20 Above m at copper (mg m 2 ) assimilation, per s pecies. Mean St. Dev. Min Max Sig. Diff. F> I Juncus effusus 0.370 0.572 0.0723 1.18 B NS Canna flaccida 2.18 2.86 0.07 71 6.23 A B Pontederia cordata 0.237 0.358 0.0494 0.741 B NS Other 0.492 0.257 0.294 0.854 A Note: In Sig. Diff. column, Species not connected by the same letter are significantly different. Statistical tests were performed using the Wilco DF=3, n=3), blocked by Site. In F>I column, an asterisk (*) indicates there was significantly more Final mass for that Species than Initial mass, and NS indicates there was no significant difference between Final mass and Ini tial mass (Wilcoxon Signed Table 3 21. Carbon biomass concentration (g kg 1 ) by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 338 49.1 275 395 A Mat 367 7.39 359 377 A Below Mat 332 41.2 274 371 A Note: Zones with similar lette rs were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 22. Nitrogen biomass concentration (g kg 1 ) by z one. Mean St. Dev. Min Max Si g. Diff. Above Mat 10.5 1.05 9.02 11.5 A Mat 19.6 0.335 19.1 19.9 B Below Mat 24.1 2.55 22.1 27.7 C Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for signifi cant difference ( = 0.05, DF=2, n=3 ), blocked by Site.
94 Table 3 23. Phosphorus biomass concentration (g kg 1 ) by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 1.11 0.290 0.779 1.48 A Mat 2.98 0.841 2.15 4.14 B Below Mat 4 .34 0.918 3.60 5.63 C Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 24. Zinc biomass concentration (mg kg 1 ) by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 32. 7 8. 29 22.8 43. 1 A Mat 42. 3 18.9 24.3 68.4 A Below Mat 43.2 19.0 18.0 64.0 A Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 25. Copper biomass concentration (mg kg 1 ) by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 3. 76 1.5 5 1.6 4 5.3 1 A Mat 4. 77 1.95 2.0 4 6.49 A B Below Mat 6.8 0 3. 47 2.0 4 10.2 B Note: Zones w ith similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 26. Cadmium biomass concentration (mg kg 1 ) by z one. Mean St. D ev. Min Max Sig. Diff. Above Mat 0.00768 0.00983 0.000592 0.0216 A Mat 0.130 0.109 0.0227 0.28 0 A B Below Mat 0.219 0.0738 0.122 0.300 B Note: Zones with similar letters were found to have no significant difference for their Site means usin g the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site.
95 Table 3 27. Chromium biomass concentration (mg kg 1 ) by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 0.370 0.123 0.274 0.544 A Mat 3.27 1.37 1.49 4.81 B Below Mat 5.29 1.82 2.79 7.10 C Not e: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 28. Biomass (g m 2 ) assimilation by z one. Mean St. D ev. Min Max Sig. Diff.* Above Mat 856 649 349 1770 A Mat 566 300 207 941 A Below Mat 486 400 99.1 1040 A Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 29. Carbon (g m 2 ) mass assimilation by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 323 244 127 667 A Mat 209 111 74.7 345 A Below Mat 168 141 25.4 360 A Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 30. Nitrogen (g m 2 ) mass assimilation by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 9.55 6.17 4.57 18.2 A Mat 11.1 5.96 3.98 18.6 A Below Mat 11.4 8.58 2.23 22.8 A Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.0 5, DF=2, n=3 ), blocked by Site. Table 3 31. Phosphorus harvest (g m 2 ) by z one. Mean St. Dev. Min Max Sig. Diff. Above Mat 0.901 0.417 0.425 1.44 A Mat 1.52 0.733 0.836 2.53 A Below Mat 1.83 1.35 0.575 3.71 A Note: Zones with similar letters we re found to have no sig diff for their Site means using the Kruskal Wallis test for sig difference ( = 0.05, DF=2, n=3 ), blocked by Site.
96 Table 3 32. Zinc (mg m 2 ) mass assimilation by zone. Mean St. Dev. Min Max Sig. Diff. Above Mat 39.2 46.4 5.38 105 A Mat 27.1 25.1 6.38 62.3 A Below Mat 20.2 20.3 5.22 48.9 A Note: Zones with similar l etters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 33. Copper (mg m 2) mass assimilation by zone. Mean St. Dev. Min Max Sig. Diff. Above Mat 3.51 3.88 0.556 9.00 A Mat 2.53 2.01 0.966 5.37 A Below Mat 3.38 3.65 0.590 8.54 A Note: Zones with similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant differe nce ( = 0.05, DF=2, n=3 ), blocked by Site. Table 3 34. Chromium (mg m 2 ) mass assimilation by zone. Mean St. Dev. Min Max Sig. Diff. Above Mat 0.462 0.403 0.142 1.03 A Mat 2.02 1.83 0.702 4.60 B Below Mat 2.63 2.59 0.724 6.29 B Note: Zones w ith similar letters were found to have no significant difference for their Site means using the Kruskal Wallis test for significant difference ( = 0.05, DF=2, n=3 ), blocked by Site.
97 A B C Figure 3 1. Initial (left) and fin al (right) above m at FTW macrophytes. A) SS1 B) SS2 C) SS3 ; photos courtesy of Neal Beery.
98 A B Figure 3 2 Surface area of SS3 and SS1 A) SS3 with oncomplete cover, B) SS1 with incomplete cover of macrophyte growth ; photos courtesy of Neal Beery. A B Figure 3 3 Mat z one of FTWs. A) Mat at th e beginning of the study period, B) Mat during the harvest at the end of the study period ; photos courtesy of Neal Beery.
99 A B Figure 3 4. Examples of extensive root growth on FTWs. A) Surface area coverage by macrophyte roots on a panel from SS3, B) Roots hanging below the m at of SS3 ; photos courtesy of Neal Beery. A B C Figure 3 5. Aquatic vertebrates and invertebrates found in FTW below mat z one. A) Tadpoles (possibly Rana catesbeiana ), B) c atfish (poss ibly of the Ictaluridae family), C) f reshwater shrimp (possibly Palaemonetes paludosus ) ; photos courtesy of Neal Beery.
100 Figure 3 6 SS1 panel growth, p anel A, B, and C (left to right) ; photos courtesy of Neal Beery. Figure 3 7 SS2 panel growth, p anel A, B, and C (left to right) ; photos courtesy of Neal Beery. Figure 3 8 SS3 panel growth, p anel A, B, and C (left to right) ; photos courtesy of Neal Beery.
101 Figure 3 9 Example of lack of Pontederia cordata on SS1, p ane l C ; photo courtesy of Neal Beery. Figure 3 10 Mean, maximum and minimum of biomass assimilation among s pec ies groups on FTWs in this study
102 Figure 3 11. Percen t of biomass harvested by each z one. Figure 3 12. Percent of carbon assimilated b y each z one.
103 Figure 3 13. Percent of nitrogen assimilated by each z one. Figure 3 14. Percent of phosphorus assimilated by each z one.
104 Figure 3 15. Perce nt of zinc assimilated by each z one. Figure 3 16. Percent of copper assimilated by eac h z one.
105 Figure 3 17. Percent o f chromium assimilated by each z one.
106 CHAPTER 4 CONCLUSIONS The goal of this study was to measure the efficacy of FTWs deployed in several stormwater treatment systems representative of wet retention ponds in a humid sub tropical climate. With evolving stormwater regulation in Florida, stormwater managers could receive treatment credit for the use of BMPs including FTWs, making a known performance standard for FTWs important in the proper allocation of these credits. If FTWs are to be used as BMPs the design and management for optimal, consistent, and predictable treatment performance are also important for proper treatment credit allocation. For this study, the installed FTWs were allowed to function throughout a growing season and then were harvested, and their various constituents were analyzed. removing various contaminants from the stormwater at the study sites. This study sought to measure var ious features of FTWs that may have an impact on BMP design or management. Effectiveness was evaluated by categories reflected by the FTW design, including the FTW as a whole, by separate panels that make up the FTW, by each species group originally insta lled (and a conglomerate group of self recruited macrophyte species), and by the three vegetative zones relative to the constructed Mat (Above Mat, Below Mat, and the Mat itself). Overall Performance of FTWs The hypothesis that the FTW design would remove constituents of interest was supported. Over the period of the study the installed macrophytes as well as macrophytes that self recruited added biomass to the FTWs at each of the three sites.
107 st (carbon, nitrogen, phosphorus, zinc, copper, cadmium and chromium) and increased throughout the study, thus increasing the mass of the harvested FTWs to an amount greater (and significantly different) than that of the mass found in the initially install ed FTWs. This conclusion may seem obvious, or not in need of testing. However, determining that this FTW design in this application is actually effective in removing the constituents of interest allows for purposeful further study of FTWs of this design. The performance of the FTWs could be gauged by their visual appearance. The FTWs with greater assimilation of biomass and other constituents were evident through a higher density of plant growth. The originally installed species remained present on th e FTWs that performed well, as opposed to those that performed less well, where they were more absent and replaced by self recruiting species. These visual cues could be useful to managers as they make decisions about when to harvest, when to continue to use FTWs as a BMP, and various other changes in design that may arise. The FTWs were installed at three separate sites for this study, and the biomass Assimilation varied. For example, for biomass Assimilation ranged from 655.8 g m 2 to 3750 g m 2 (mean of 1908 1330 g m 2 ). The maximum mass assimilation of all other constituents occurred at the site which showed the maximum biomass Assimilation Also, the site with the minimum biomass Assimilation showed the lowest mass assimilation for most other con stituents; the exceptions were copper and chromium, whose per site minimum mass assimilation s were found at the site with the second highest biomass assimilation which was likely the influence of higher metal concentrations in the water column.
108 The bio mass that FTW macrophytes produced during the study was made up of material assimilated from the sites and the surrounding atmosphere. Carbon was taken from the surrounding atmosphere, and all other constituents were removed from the free water of the sto rmwater systems. The biomass to carbon assimilation ratio was consistently around 3:1, and the site mean nitrogen assimilation was 700.2 488.2 g m 2 (2.5 2 1.76 g m 2 on a per day basis over the period of the study). The biomass to nitrogen assimilati on ratio was consistently around 60:1, and the site mean nitrogen assimilation was 32.0 20.5 g m 2 (0.115 0.07 4 g m 2 on a per day basis over the period of the study). The biomass to phosphorus assimilation ratio was consistently around 450:1, and the Site mean phosphorus assimilation was 4.25 2.44 g m 2 (15 .3 8 .8 mg m 2 on a per day basis). The metal with the highest assimilation was zinc with a assimilation of 86.5 91.6 mg m 2 (0.311 0.330 mg m 2 d 1 ), and the lowest metal assimilation was c admium with a assimilation of 0.233 0.263 mg m 2 Thus, FTWs are capable of removing metals that are of interest to stormwater managers. Positive relationships were found between the Assimilation of FTW dry biomass and Assimilation of carbon, nitrogen, phosphorus and zinc. The relationship for phosphorus to biomass was the least strong, with an r 2 value of 0.664, and the other relationships had r 2 values of 0.885 (zinc), 0.952 (nitrogen), and 0.997 (carbon). If further research finds similar relations hips for these constituents, a general model can be generated in order to determine the mass of constituents produced by FTWs across sites by only measuring biomass. This would help stormwater managers by providing a performance. Research would need to be done across the state to account for varying growth conditions, such as length of
109 growing season as well as variations in water quality, but would help ensure that such a model is an appropriate tool to help ensure FTWs perform to expectations set out by permitted treatment credit. As can be determined from the high standard deviations on many of the means for FTW performance in this study, there was widely variable performance among SS1, SS2 and SS3. As this study sought to determine FTW effectiveness in general, the disparate minimum and maximums of the three sites show that there is the potential for FTW performance to be high or low, although it is not yet known what factors specifically determined performance, further investigation of these factors would be beneficial to optimize treatment potential and limit overestimation performance of FTWs in some areas. It will be the task of stormwater managers to determine if an FTW design is appropriate at certain sites and to plan and maintain the FTWs in a manner that achieves the desired results. It is important to note that all sites in this study were relatively large and deep wet detention ponds with nutrient levels that seemed adequate. Possible problems may hav e arisen due to wildlife and water chemistry, such a pH levels, but overall, the FTWs in this study were successful in assimilating the constituents of interest in a range of conditions. Panels and Edge Effects There were no issues with edge effects among panels. It was thought that the outer panels may have taken up or physically blocked nutrients, moving air, dissolved oxygen or sunlight before reaching the inner portions of the FTWs, reducing the growth and overall performance of the inner panels. This was not found to be the case. However, the FTWs in this study were single 8.93 m 2 installations that did not take up a
110 study, the hypothesis that an edge effect would exist (H2) was not supported. Had the FTWs been installed in larger contiguous patterns there could have been issues with buffering inner portions or not protecting outer portions, thus affecting performance. This design issue may be important to stormwa ter managers seeking to maximize FTW performance. Also, issues such as the FTWs shading the free water and modifying water column dissolved oxygen levels may affect the performance of the overall stormwater system. Anecdotally, this study had a separate portion that was not reported on in this thesis which found possible issues with the reduction of dissolved oxygen when FTWs were placed in isolated chambers. This reduction of oxygen levels in the water column is hypothesized to be the result of a large percentage of the surface area being covered by the FTW and a high biological oxygen demand associated with the Mat and Below Mat portions of the FTW. More information on these and similar issues will be needed in the future to determine the appropriate m aximum surface area coverage and surface area density of FTWs in stormwater systems. Macrophyte Species Selection The macrophytes installed on an FTW are the active part of the design; they assimilate the constituents of interest and are then harvested. I n order to maximize the performance of FTWs, the species that are installed or allowed to self recruit are important. As it was found that the biomass Assimilation was the principal factor determining the amount of potential pollutants harvested, it is im portant to choose and maintain the species that grow successfully and attain a high level of Assimilation Nitrogen, phosphorus and zinc biomass concentrations had no significant differences among species. For constituents such as carbon, J. effusus sign ificantly differed in
111 concentration from the other species and had the highest concentration of carbon; however, this species did not have the highest mass assimilation of the four species groups. This indicated that it was not the constituent biomass con centration of particular species that made them perform better or worse for mass assimilation, rather, it was the general biomass that was grown by that species that lead to greater performance. C. flaccida and the Other Species groups assimilated the m ost biomass, w ith their site means being 306 390 g m 2 and 296 174 g m 2 respec tively, as opposed to 88.6 102 g m 2 and 85 .8 69.3 g m 2 for J. effusus and P. cordata respectively. The success of these two species groups for the biomass constituen t indicated that utilization of C. flaccida would be appropriate for FTW design, and that space can be left between initially installed individuals in order to allow for self recruitment. Self recruitment may also be useful because these individuals will most likely be currently present in the sites in which FTWs are installed and therefore may be well suited to the conditions at particular sites. If stormwater managers are using FTWs to fix carbon from the atmosphere, C. flaccida and self recruited specie s would be the most effective, according to the results of this stud y. A mean assimilation of 110 141 g m 2 for C. flaccida and 114 68.0 g m 2 for the Other Species group were found, and were nearly three times greater than the other two species. The FTW design that gave greater potential to harvest carbon was found to utilize the same species groups as for the maximization of biomass harvest. The pattern for biomass and carbon harvest maximization applied to the assimilation of nitrogen and phosphoru s as well. C. flaccida and the Other Species
112 group had the greatest mass assimilation s of these two constituents with 3.03 3.75 g m 2 and 3.70 2.04 g m 2 respectiv ely for nitrogen assimilation 235 266 g m 2 and 388 152 g m 2 for phosphorus assimi lation respectively. These mass assimilation s were three to four times greater than the other two species groups. It was found that C. flaccida and self recruited species should be utilized for nutrient removal by stormwater managers working in similar conditions to those of this study. When considering the assimilation of zinc and copper, the pattern for the previously mentioned constituents applied once again. C. flaccida and the Other Species group assimilated the greatest metal mass of the specie s g roups: zinc assimilation was 20.7 28.6 mg m 2 and 10.9 10.6 mg m 2 respectively, and copper assimilation was 2.18 2.86 mg m 2 and 0.49 2 0.257 mg m 2 respectively. For zinc assimilation s, C. flaccida and the Other group assimilated three to seven times more than the other two groups. For copper assimilation s, C. flaccida and the Other group assimilated two to six times more than the other two groups. Also, for both zinc and copper, J. effusus and P. cordata did not have significantly more metal mass at the final harvest than they did when initially installed. It was thus concluded that C. flaccida and the Other group were the only two species groups effective at removing zinc and copper from stormwater systems. As with the other constituents, i t was the amount of biomass that was grown by these two species groups that lead to their efficacy, not the concentrations at which the metals were found in the biomass of particular species groups. Chromium and cadmium were not found in appreciable amoun ts at the end of
113 were not effective in removing those two metals or that concentrations in the water column were not at high enough levels for these plants to accumul ate Harvesting by Zone The hypothesis that the zone found by other studies to have greater potential for assimilating the constituents of interest would have similar potentials in this study was supported in some instances, but not in others. This depen ded mostly on which constituent was being considered. For biomass, 44.9% of the total biomass was harvested from the Above Mat Zone, compared to 29.7% and 25.4% for the Mat and Below Mat Zones, respectively. However, due to the large standard deviation, there was no significant difference among Zones for the mass of biomass harvested. When the easily harvested zones, Above Mat and Below Mat, are combined, 70.3% of the biomass would be assimilated by only harvesting these two zones and leaving the Mat Zo ne as propagule for the next growing season. Carbon harvesting results closely mirrored those for biomass, and the same conclusion was drawn, especially since there was no significant difference among zones for carbon biomass concentration. Nitrogen bio mass concentration significantly differed among all zones, with the Below Mat Zone having the highest concentration (24.1 2.5 g kg 1 ), and the Mat and Above Mat having lower concentrations (19.6 0.3 g kg 1 and 10.5 1.0 g kg 1 respectively). However, the mass of nitrogen assimilation did not significantly differ assimilation ranged from 29.9% to 35.6% of the total, and the combined Above Mat and Below Mat Zones assimilation was 65.5% of the total. When combined, these two zones contained the majority of the nitrogen mass.
114 Phosphorus biomass concentration significantly differed among all zones, with the Below Mat Zone having the highest concentration (4.34 0.92 g kg 1 ), and the Mat and Above Mat having lower conc entrations (2.98 0.84 g kg 1 and 1.11 0.29 g kg 1 respectively). However, the mass of phosphorus assimilation did not significantly differ among zones, due to the high standard deviations associated with the means. The Below Mat Zone was responsible for 43.2% of the phosphorus that was assimilated and the Above Mat Zone only assimilated 21.2% of the total. The Mat Zone was closer to that of the Below Mat Zone with 35.7% of the total. Still, 64.4% of the assimilated phosphorus would be assimilated if only the easily harvested Above and Below Mat Zones were removed. It was hypothesized that metals would be assimilated in greater amounts by the Below Mat Zone. This was found to be true for chromium, but not for zinc, copper (or cadmium, which was not found in appreciable levels in any zone). Zinc had no significant differences among zones for concentrations and mass assimilation s, but the Above Mat zinc mass assimilation was 45.3% of the total (68.7% of the total zinc assimilated was found in the Abo ve Mat and Below Mat Zones combined, leaving 31.3% in the Mat Zone). If managers are focused on zinc removal harvesting only the Above Mat and Below mat Zones would remove more than the majority of the total. Copper had no significant differences among zones for mass assimilation and the percentages of the total assimilated copper for each zone were similar, however, harvesting only the Above Mat and Below Mat Zones assimilated 73.2% of the total. For chromium, the Above Mat Zone was less than and signi ficantly different from the other two zones, and only made up 9.0% of the total amount of chromium assimilated
115 Leaving the Mat Zone on the FTW would have left 39.6% of the total behind; the Below Mat Zone made up 51.4% of the total chromium assimilated The majority of the chromium assimilated could be harvested by removing the Below Mat and Above Mat Zones. It was concluded that managers could harvest only the Above Mat and Below Mat Zones and still remove 60 to 70% of assimilated nutrients, and 60 to 8 0% of assimilated metals. This would allow managers to expend fewer resources reinstalling macrophytes, while leaving the Mat Zone to reestablish biomass on the FTWs to be harvested later. Most of the Mat Zone is made up of biomass that had taken nutrien ts and metals from the stormwater system; only 14.1 g m 2 of biomass was made up of the rhizome portions of the installed macrophytes, whereas 580.4 g m 2 of biomass was found in the Mat Zone at the end of the study. There would be little risk of loading the stormwater system with nutrients or metals from the initial installation after a harvest, as the Mat Zone mass that is left over came from the system, not from outside of the system. In addition, much of the mass associated with the Mat Zone is living tissue that will not senesce at the end of the growing season, as would be the case for Above and Below Mat Zones. Recommendations for Further Research This study sought to determine whether FTWs were an effective method for stormwater management across a number of constituents commonly focused on by stormwater BMPs. It was generally concluded that FTWs can be an effective stormwater BMP, that this size of FTW installed in large enough stormwater systems pose no negative impact on FTW performance, that C. flaccida and self recruited
116 harvested FTW zones allows for most of the assimilated constituents to be harvested. There were other more particular issues and questions that were brought forth by this general study. Some issues arose with the previous maintenance (unrelated to this study) of the stormwater system sites. Possible grazing by grass carp, which were installed at SS1 as aquatic weed control, may have led to d ecreased FTW performance at that site. The presence of grazers of any sort must be considered or mitigated through exclusionary devices. It was not directly confirmed, but herbicides may have been used at SS3 on submerged aquatic vegetation during the st udy; this may not have had an effect on the FTW macrophyte growth, but herbicides and other chemicals, such as the dye used at SS2, need to be considered before installing FTWs. The physical design of stormwater systems may also cause potential limitations for the functioning of both FTWs and the stormwater systems themselves. In addition to the previously described issues with surface area coverage, the depth of a stormwater system could affect FTW performance. The root depth approached a meter below the Mat Zone for FTWs in this study. If the depth of the stormwater system were to be less than the length of the FTW roots, the roots could interact with the bottom of the stormwater system, thus allowing nutrient uptake from sediments and restricting uptak e from the water column. Alternatively, if the depth of the water column were significantly greater than the root length, then the treatment potential facilitated by contact between roots and water column would be limited. Along with issues arising from the design and maintenance of stormwater systems in which FTWs would be installed is the issue of water chemistry at all
117 widely in their levels of available nutrients, pH, oxygen, turbidity, and other factors, and the performance of FTWs may vary due to the effects of these site conditions. Research needs to address possible effects of these conditions on FTW performance in order for stormwater managers to properly utilize FTWs as BMPs. The nutrient levels in the sites in this study were all sufficient for FTW growth, but this may not be the case in every stormwater system. Harvesting schemes were another aspect of FTW design and maintenance that further research should ad dress. This study had one harvest in a calendar year, allowing for FTW macrophytes to grow essentially one entire growing season. As the Mat Zone could be left unharvested and regrow quickly in the middle of a growing season, multiple harvests per growin g season could increase the efficacy of FTWs to remove certain constituents. Research could address this, as well as determining if performance is increased by harvesting only the Above Mat Zone, or both the Above Mat and Below Mat Zones. Additionally, t he amount of effort required to remove the bulk mass may need to be addressed, as some mechanical device may make harvesting FTWs significantly more efficient and therefore less costly. The FTWs in this study were left with 80% (the portion that was not ta ken as part of the sub sampling of this Zone) of the Mat Zone biomass after harvesting and were allowed to grow throughout another season. At the end of a second growing season, the FTWs at SS2 and SS3 were found to have visually similar amounts of growth as compared to the first growing season of this study (SS1 had very little growth). Further study is necessary to determine the efficacy of this approach, and whether maximization
118 of performance requires FTW plants to be entirely reinstalled after a cer tain number of growing seasons. Issues may arise due to the nature of removing what may be unwanted constituents from stormwater systems by utilizing FTWs. In order for FTWs to effectively remove constituents from stormwater systems and ultimately keep t hem from reaching natural surface water bodies, the biomass (and the constituents of interest) must be harvested and taken off site. The disposal of this biomass must be addressed in order for FTWs to be an effective BMP. Possible disposal methods would be conventional landfills, composting of the biomass, and use as fuel in biomass combusting energy Assimilation The constituents may leave the watershed of the stormwater system from which they were removed and placed in another. An example of an issue arising from the disposal of FTW biomass is composting. Composting of the biomass may be a beneficial method, as the end result could be used for soil amendments and cut down on the amount of fertilizer brought in from outside sources, however the biomass may be high in heavy metals or leach nutrients into surface waters or ground water. The nature of the problems being addressed by the utilization of FTWs may dictate the appropriate method of disposal, but this issue will need to be considered by stormwa ter managers who install and maintain FTWs. This study found general values for FTW performance across three representative stormwater systems in north central Florida. If utilized as a stormwater BMP, the credit allotted to FTWs for treatment must corres pond to quantifiable and appropriate indicators of performance. The assimilation of constituents on a mass per square meter per day basis was used in this study, and may prove to be a useful unit of
119 measure for further studies as well as governmental agen cies overseeing the implementation of FTWs. The guidelines for approving, installing, maintaining and assessing FTWs must be consistent in order for this BMP to be utilized in a manner that ensures the performance that it is credited while remaining a via ble option for stormwater managers.
120 LIST OF REFERENCES Bachand, P.A.M., Horne, A.J., 2000. Denitrification in constructed free water surface wetlands: I. Very high nitrate removal rates in a macrocosm study. Ecol. Eng. 14, 9 15. Bingham, R.L., Neal, H. V., El Agroudy, A.A., 2002. Characterization of the potential impact of stormwater runoff from highways on the neighboring water bodies case study: Tamiami trail project. In: Sevent h Biennial Stormwater Research and Watershed Management Conference Chan g, N. B., Islam, M. K., Wanielista, M.P 2012 Floating wetland mesocosm assessment of nutrient removal to reduce ecotoxicity in stormwater ponds. Int. J Environ. Sci. Te. 9 (3), 453 462. Chen, Y., Bracy, R.P., Owings, A.D., Merhaut, D. J. 2009. Nitrogen and phosphorous removal by ornamental and wetland plants in a greenhouse recirculation research system. Hortscience 44 (6), 1704 1711. Chua, H.C.L., Tan, S.B.K., Sim, C.H., Goyal, M.K., 2012. Treatment of baseflow from an urban catchmen t by a floating wetland system. Ecol. Eng. 49, 170 180. Clark, M.W., 2000. Biophysical characterization of floating wetlands (flotant) and vegetative succession of a warm temperate aquatic ecosystem. University of Florida Gainesville, FL (PhD dissertat ion) DeBusk, T.A., Dierber, F.E., Reddy, K. R. 2001. The use of macrophyte based systems for phosphorus removal: An overview of 25 years of research and operational results in florida. Water Sci. Technol. 44 (11 12) 39 46. DeBusk, T.A., Grace, K.A., Die rberg, F.E., Jackson, S.D., Chimney, M.J., Gu, B., 2004. An investigation of the limits of phosphorus removal in wetlands: A mesocosm study of a shallow periphyton dominated treatment system Ecol. Eng. 23, 1 14. Deng, H., Ye, Z.H., Wong, M. H. 2004. A ccumulation of lead, zinc, copper and cadmium by 12 wetland plant species thriving in metal contaminated sites in C hina. Environ P ollut. 132 (1 ) 29 40. Deng, H., Ye, Z.H., Wong, M. H. 2009 Lead, zinc and iron (F e 2+) tolerances in wetland plants an d relation to root anatomy and spatial pattern of ROL. Env iron. Exp. Bot. 65 (2), 353 362. Durham, S., 2006. Floating above lagoon wastewater. Agr Res. 54 (8), 11 11.
121 Florida Department of Environmental Protection ( FDEP ) 2010b Environmental resourc for stormwater treatment systems in Florida : March 2010 draft. http://www.dep. state.fl.us/wat er/wetlands/erp/rules/stormwater/index.htm Florida Department of Environmental Protection (FDEP), 2010a Integrated Water Quality Assessment for Florida: 2010 305(b) Report and 303(d) List Update, 127 129. Florida Department of Environmental Protection (FDEP), 2012. Factsheet about outstanding Florida waters. http://www.dep.state.fl.us/water/wqssp/ofwfs.htm Fritioff, ., 2005. Metal accumulation by plants. Stockholm University, Sweden (PhD thesis). Fritioff, ., Greger, M. 2003. Aquatic and terrestrial plant species with potential to remove heavy metals from stormwater. Int J of Phytoremediat. 5 (3), 211 224. Fulcher, G.A. 1994. Urban stormwater quality from a residential catchme nt. Sci. Total Environ. 146 147, 535 542. Hadad, H.R., Maine, A. M. 2007. Phosphorous amount in floating and rooted macrophytes growi ng in wetlands from the middle Paran river floodplain (A rgentina). Ecol Eng 31 (4), 251 258. Harper, H.H., Baker, D .M., 2007. Evaluation of current stormwater design criteria within the state of Florida. Environmental Research & Design, Inc. http://www.dep. state.fl.us/water/wetlands/erp/ Headley, T.R., Tanner, C.C., 2006. Application of floating wetlands for enhanced stormwater treatment: A review. Auckland Regional Council, Auckland, New Zealand (ARC Technical Publication No. HAM2006 123). Headley, T.R., Tanner, C.C., 2008. Floating Vegetated Islan ds for Stormwater Treatment: Removal of Copper, Zinc, and Fine Particulates. Auckland Regional Council, Auckland, New Zealand (ARC Technical Publication No. TR2008/030). Headley, T.R., Tanner, C.C., 2011. Components of floating emergent macrophyte tre atment wetlands influencing removal of stormwater pollutants Ecol. Eng., 37, 474 486. Ho, Y. B. 1979 Chemical composition studies on some aquatic macrophytes in three scottish lochs. II. potassium, sodium, calcium, magnesium and iron. Hydrobiologia 6 4 (3), 209 213.
122 Hogg, E.H., Wein, R.W., 1988. The contribution of Typha components to floating mat buoyanc y. Ecology. 69 (4), 1025 1031.Mitsch, W.J., 1976. Ecosystem modeling of water hyacinth management in Lake Alice, Florida. Ecol. Model. 2 (1), 69 8 9. Hollander, M., Wolfe, D.A., 1999. Nonparametric Statistical Methods, second ed. John Wiley & Sons, Inc., New York. Holt, T.C., Maynard, B.K., Johnson, W. A. 1999. 447 nutrient removal by five ornamental wetland plant species grown in treatment Assim ilation wetland biofilters. Hortscience 34 (3), 521. Hubbard, R.K., 2010. Floating vegetated mats for improving surface water quality, in: Shah, V (Ed.) Emerging Environmental Technologies, Volume II. Springer, New York, pp. 211 244. Hubbard, R.K., Gascho, G.J., Newton, G.L., 2004. Use of floating vegetation to remove nutrients from swine lagoon wastewater. Am Soc of Agr Eng 47 (6), 1963 1972. Institute of Food and Agriculture Sciences (IFAS), 2013. Center for Aquatic and Invasive Plants. Uni versity of Florida, Gainesville, Florida http://plants.ifas.ufl.edu/ Irmak, S., Haman, D.Z., Jones, J. W. 2002. Evaluation of class A pan coefficients for estimating reference evapotranspi ration in humid location. J I rrig. Drain. E ASCE Engineering 128 (3), 153 159. Jangrell Bratli, A.S., 2011. The effects of artificial floating wetland i slands on water quality in a eutrophic l ake University of South Florida, St. Petersburg, FL (MS th esis). Kadlec, R.H., Wallace, S.D., 2009. Treatment Wetlands, second ed. CRC Press, Boca Raton, Florida. Ladislas, S., Grente, C ., Chazarenc, F., Brisson, J., Andrs, Y. 2013. Performances of two macrophytes species in floating treatment wetlands for cadmium, nickel, and zinc removal from urban stormwater runoff. Water Air Soil Poll. 224 (2), 1 10. Lee, J.H., Bang, K.W., 2000. Characterization of urban stormwater runoff. Water Res. 34 (6), 1773 1780. Livingston, E. H., 2009. Presentation. Histo program; the proposed statewide stormwater rule, how we got there. Presentation to the American Academy of Environmental Engineers on September 22, 2009. http://www.aaees.org/downloadcenter/Presentation Livingston.pdf
123 Mallison, C.T., Stocker, R.K., Cichra, C.E., 2001. Physical and vegetative characteristics of floating islands. J. Aquat. Plant Manage 39, 107 111. Manker, D.C. Martin, D.F., 1984. Investigation of two possible modes of action of the inert dye Aquashade on H ydrilla J. Environ. Sci. Health. A19 (6), 725 733 McJannet, C.L., Keddy, P.A., Pick, F. R. 1995 Nitrogen and phosphorus tissue concentrations in 41 we tland plants: A comparison across habitats and functional groups. Funct. Ecol. 9 (2), 231 238. Mitsch, W.J., 1976. Ecosystem modeling of water hyacinth in Lake Alice, Florida. Ecol. Model. 2, 69 89. Mitsch, W.J., Gosselink, J.G., 1997. Wetlands, third ed. John Wiley and Sons, Inc., Hoboken, New Jersey. p 166. Mitsch, W.J., Horne, A.J., Nairn, R.W., 2000. Nitrogen and phosphorous retention in wetlands ecological approaches to solving nutrient problems. Ecol. Eng. 14, 1 7. National Oceanic and A tmospheric Administration (NOAA), 2012. National Climactic Data Center Climate Information. http://www.ncdc.noaa.gov/climate information North Carolina Department of Environment and Natural Res ources (NCDENR) Division of Water Quality 2005. Updated Draft Manual of Stormwater Best Management http://townhall.townofchapelhill.org/agendas/2006/10/09/9/ Professional Serv ice Industries, Inc. (PSI), 2010. The effectiveness of vegetated floating mats in sequestering nutrients in a structurally controlled waterbody. Lee County Department of Natural Resources, Fort Myers, FL (PSI project No. 0552189). Rad r, R.B., Batzer, D. P., Wissinger S.A., 2001. Bioassessment and Management of North American Freshwater Wetlands John Wiley & Sons, New York. Rai, P. K. 2008. Heavy metal pollution in aquatic ecosystems and its phytoremediation using wetland plants: An ecosustainable app roach. Int J of Phytoremediat. 10(2), 131 160. Reddy, K.R., and DeLaune, R.D., 2008. Biogeochemistry of wetlands: science and applications. Taylor & Francis Group, LLC. Boca Raton, Florida. Rosner, B., 2010. Fundamentals of Biostatistics, seventh e d. Cengage Learning, Boston, Massachusetts.
124 Rushton, B., Huneycutt, D., Teague, K., 2004. Characterization of three stormwater ponds. Southwest Florida Water Management District, Resource Management Department. DEP Contract No. WM 716. ver Water Management District (SJRWMD), 2002. Standard general environmental resource permit technical staff report. Application #: 40 001 82960 1. stormwater management master plan and permit application renewal. Application #: 4 001 15570 3. loads for SW 5th master basin. Application #: 40 001 82960 1. Sasser, C.E., 1994. Vegetation dynamics in relation to nutrients in floating marshes in Lousiana, USA. Utrecht University, Netherlands. Louisiana State University, Baton Rouge, Louisiana (Faculty thesis). Sasser, C.E., Gosselink, J.G., 1984. Vegetation and primary Assimilation in a floating freshwater marsh in Louisiana. Aquat. Bot 20, 245 255. Sasser, C. E., Gosselink, J.G., Shaffer G.P., 1991. Distribution of nitrogen and phosphorus in a louisiana freshwater floating mars h. Aquat. Bot. 41 (4), 317 31. Spieles, D.J., Mitsch, W.J 1999 The effects of season and hydrologic and chemical loading on nitrate retention in constructed wetlands: A comparison of low and high nutrient riverine systems. Ecol. Eng 14 ( 1), 77 91. Staddon, C. 2011. Stormwater management, i n Cohen N., Robbins P. (Eds.), Green cities: An A to Z guide. SAGE Publications Inc., Thousand Oaks, California, pp. 403 405 Stewart, F.M., Mulholland, T., Cunningham, A.B., Kania, B.G., Osterlund, M.T., 2008. Floating islands a s an alternative to construc ted wetlands for treatment of excess nutrients from agricultural an d municipal wastes: Results of laboratory scale tests Land Con. Recl. 16 (1), 25 33. Tanner, C.C., Sukias, J., Park, J., Yates C., Headley, T., 2011. Floating treatment wetlands: A n ew tool for nutrient management in lakes and waterways http://www.massey.ac.nz/~flrc/workshops/11/Manuscripts/Tanner_2011.pdf U.S. Environmental Protection Agency (US EPA), 1999. Free water surface wetlands for wastewater treatment: A technology assessment. EPA 832 S 99 002. Of fice of Water, Washington, DC.
125 U S Environmental Protection Agency (EPA), 2007. Total maximum daily loads with stormwater sources: A summary of 17 TMDLs. EPA 841 R 07 002 http://water.epa.gov/lawsregs/lawsguidance/cwa/tmdl/upload/17_TMDLs_ Stormwater_ Sources.pdf U S Environmental Protection Agency (EPA), 2009. Section 303(d) Water: Total maximum daily loads for Florida http://water.epa.gov/lawsregs/lawsguidance/ cwa/tmdl/factsheet.cfm U.S. Geolo gical Survey (USGS). 2012. Earth Explorer. http://earthexplorer.usgs.gov/ United Nations Statistical Division, 1997. Glossary of environment statistics, studies in methods, Series F, No. 67. United Nations, New York. United States Department of Agriculture (USDA), 2012. Web soil s urvey Soil Survey Staff, Natural R esources Conservation Service. http://websoilsurvey. sc.egov.usda.gov/A pp/HomePage.htm Uni versity of Florida Conservation Area Land Management Plan s (UF CALM), 2012. http://www.facilities.ufl.edu/favicon.ico University of Florida, Florida Automated Weather Network (UF FAWN), 2012. Florida Automated Weather Network Home http://fawn.ifas.ufl.edu/favicon.ico Van de Moortel, A .M. K ., Meers, E., De Pauw, N., Tack, F.M. G. 2010 Effects of vegetation, season and tempe rature on the removal of pollutants in experimental floating t reatment wetlands. Water Air Soil Poll 212 (1), 281 297. Vogel, J.A., 2011. The effects of artificial floating wetland island construction materials on plant biomass. University of South Flo rida, St. Petersburg, FL (MS thesis). Wanielista, M.P., Chang, N., Chopra, M., Xuan, Z., Islam, K., Marimon, Z., 2012. Floating wetland systems for nutrient removal in stormwater ponds. FDOT Report, Project BDK78 985 01. Wen, L., Recknagel, F. 2002. I n situ removal of dissolved phosphorus in irrigation drainage water by planted floats: Preliminary results from growth chamber experiment. Agricult ure, Ecosystems Env. 90 (1), 9 15. White, S., 2009. Floating treatment systems, report. Clemson Universit y, Clemson, South Carolina (whitepaper) http://beemats.com/uploads/2158/Clemson_ University_ 5_28_09.doc Xian, G., Crane, M., Su, J. 2007. An analysis of urban development and its environmental impact on the Tampa Bay watershed. J Env. Manag. 85 (4), 965 976.
126 BIOGRAPHICAL SKETCH Neal Beery was born in Jacksonville, Florida. His playgrounds have been tracts of Florida upland and lowlands, rivers, creeks and coastline. His wish is to work towards the conservation of natural resources, especially water. Neal attended the Paxon School for Advanced Studies in high school, where he took a course in Environmental Science and realized that a career in natural resource conser vation was possible. His time as an undergraduate at the University of Florida narrowed his interests to the science and policy of water and wetlands. Neal completed an internship with the Estuaries Policy Division of the Conservancy of Southwest Florida and gained an appreciation of the efforts that many people around the state put towards protecting the resources we utilize and enjoy every day. A short stint working with the e put Neal in further contact with his Wetlands professor, Dr. Mark Clark. Neal somehow convinced Mark Clark to take him on as a student in the Wetlands Biogeochemistry Laboratory within the UF Soil and Water Science Department, where he conducted and reported on this research. Now, Neal is beginning his career working towards protecting and conserving his former and current playground, the outdoors of Florida.