|UFDC Home||myUFDC Home | Help|
PHOSPHORUS RELEASE AND STORAGE BY TWO ISOLATED WETLANDS IN
THE NORTHERN LAKE OKEECHOBEE DRAINAGE BASIN
CHRISTY L. SACKFIELD
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING
UNIVERSITY OF FLORIDA
CHRISTY L. SACKFIELD
I would like to thank my committee chair, Dr. Kenneth L. Campbell, for providing
me the opportunities to study, learn, and grow academically under his guidance. His
support has played a tremendous role in the success of this project. I would also like to
thank committee member Dr. Mark Clark for his enthusiasm for wetlands and his
dedication to this project, and thanks to Dr. Sanjay Shukla for his time and support.
Several people in the Agricultural and Biological Engineering department
contributed to the success of this project. I am sincerely grateful for Billy Duckworth's
help with fieldwork. Ralph Hoffman, Bob Tonkinson, and James Rummell helped with
materials and construction. Thanks to Ms. Yu Wang and the ladies in the lab in the Soil
and Water Science Department for their guidance in the laboratory. Thanks to Dr. Portier
for his assistance with statistical analysis.
I appreciate all the encouragement from my fellow graduate students, professors,
and friends. So many people made this research an experience of a lifetime. This project
would not have been possible without funding from the South Florida Water
Management District, the Florida Department of Environmental Protection, and the
Florida Department of Agriculture and Consumer Services.
And finally, I would especially like to thank my family and Davis whose love and
support have pushed me above and beyond what I thought possible. I could not have
done this without them.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iii
LIST OF TABLES ............... ............. ......... ...................... vi
LIST OF FIGURE S ......... ..................................... ........... vii
ABSTRACT .............. .................. .......... .............. ix
1 IN TR OD U CTION ............................................... .. ......................... ..
1.1 Lake Okeechobee and W ater Quality Issues ........................................ ...............1
1.2 Beef Cattle Ranches and Water Quality Issues ............................................... 1
1.3 Hydrologic Conditions Relative to Florida ......... ......................5
1.4 Phosphorus in W wetlands ................................................ ............................. 6
1.5 Soil/W ater Phosphorus Interactions (P-Cycle) .................................. .............. .8
1.6 M easuring Phosphorus Uptake/Release .............................. ..............................11
1.7 Using Mesocosms for In-Situ Analysis of Phosphorus.............. ...... ........ 12
1.8 G oals and Objectives ............. ........................ .... .... .. ............ 14
2 M E TH O D O L O G Y ...................................................................................... 16
2 .1 Site D description ......................................... ...................16
2.2 Trial Run at Stormwater Ecological Enhancement Project..............................20
2.3 In-Situ W etland M esocosm Study .............................. .................................. 22
H hypothesis ........... ...... ... ...................................... .................. 23
2 .4 W ater Q u ality A naly sis.............................................................. .....................29
2 .5 S o il A n aly sis ................................................. ................ 3 1
2 .6 V egetation A naly sis.......... ............................................................. .... .... .... .. 34
2 .7 Statistical M ethods........... ........................................ ................ .. .... ...... 35
3 RESULTS AND DISCU SSION ........................................... .......................... 36
3.1 W after Level Results.............................. ............ ............. 37
3.2 Water Quality Data and Statistical Results................................................ 38
3.3 Phosphorus Results (Uptake/Release) ...................................... ............... 42
3.4 Soil Results .............................................. ........................ .... 49
3.5 Vegetation Results ......................................... ..................... ........ 50
3.6 B rom ide R esults................................................. 51
S ite 1 C control ................................................................ ............ ...... 54
Site 1 D ilute ................................................................... ......55
Site 1 Spike ............................................................................................... 56
Site 4 Control ............................................................... ..... ............. 57
Site 4 D ilute ........................................................................................... ... ..... 58
Site 4 Spike .................................. .............................. ....... 59
3.7 D term nation of EPCw Sm ax, DPS x .................................................................60
3.8 Langmuir-Hinshelwood Model of Phosphorus .................. .............64
4 C O N C L U S IO N S ................................................................................................... 6 8
5 SUGGESTIONS FOR FUTURE RESEARCH ......................................................72
APPENDIX: TRIAL RUN RESULTS ................................. ...............74
LIST OF REFERENCES ........................ .................. 77
BIOGRAPHICAL SKETCH ................................. ...........................................81
LIST OF TABLES
1-1. Mean soluble reactive P (SRP) and total P (TP) concentrations in dairy and beef
cattle m anure (G raetz and N air 1995) .............................................. .................. 4
2-1. Polycarbonate tube dimensions for trial run at Stormwater Ecological
Enhancement (SEEP) site on University of Florida campus...............................20
2-2. Treatments applied at SEEP trial run. ............................................ ............... 21
2-3. Treatments applied to mesocosms in wetlands at Pelaez & Sons Ranch ................23
2-4. Treatments applied at Pelaez Ranch based on water volumes in mesocosms...........28
3-1. Statistical analysis of phosphorus data, Pelaez & Sons Ranch...............................40
3-2. Statistical analysis of nitrogen data, Pelaez & Sons Ranch.................. ............41
3-3. Soil results (0-10 cm depth) for Site 1 and Site 4, Pelaez & Sons Ranch ...............49
3-4. Vegetation results for Site 1 and Site 4, Pelaez & Sons Ranch.............................51
3-5. Illustration of possible leakage underneath mesocosm walls in Site 4 control
m esocosm s. .......................................... ............................ 58
A-1. Trial run results from the Stormwater Ecological Enhancement Project (SEEP)
site, U university of Florida cam pus. ........................................ ....... ............... 74
A-2. Water quality results from trial run at Pelaez & Sons Ranch, Okeechobee County,
F lo rid a ........................................................... ................ 7 5
LIST OF FIGURES
2-1. Map ofPelaez & Sons Ranch, Okeechobee County, Florida, with wetland and
m esocosm locations........... ............................................................ ...... ........ 17
3-1. W after depths at Site 1.......................................... .......................... 38
3-2. W after depths at Site 4 .......................... ...................... ... .. ......... ........... 38
3-3. Mean SRP and standard deviations over time for Site 1. ........................................44
3-4. Mean SRP and standard deviations over time for Site 4..................... ..............45
3-5. M ass versus water depth over time for Site 4.................................. ............... 46
3-6. M ass versus w ater depth over tim e for Site 1....................................... ...................48
3-7. Illustration of poor mixing and ET effects using bromide data ..............................53
3-8. Effects of ET and groundwater loss in control and dilute mesocosms at Site 1.........55
3-9. Effects of ET and groundwater loss in spike mesocosms at Site 1. .........................57
3-10. Effects of ET and groundwater loss for control mesocosms at Site 4....................58
3-11. Effects of ET and groundwater loss for dilute mesocosms at Site 4....................59
3-12. Effects of ET and groundwater loss on spike mesocosms at Site 4.....................60
3-13. Schematic of graph used to determine phosphorus retention and release for
iso lated w etlan d .................................................... ................ 6 1
3-14. EPCw value for Site 1, Pelaez & Sons Ranch ...................................................62
3-15. EPCw graph for Site 4, Pelaez & Sons Ranch. ...................................................... 63
3-16. Predicted vs. measured SRP conc.using Langmuir-Hinshelwood model. ..............67
A-1. Water level changes in Site 1 and Site 4 during trial run .......................................75
A-2. Trial run, Site 1, soluble reactive phosphorus conc. vs. time. ..................................76
A-3. Trial run, Site 4, soluble reactive phosphorus cone. vs. time.................................76
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering
PHOSPHORUS RELEASE AND STORAGE BY TWO ISOLATED WETLANDS IN
THE NORTHERN LAKE OKEECHOBEE DRAINAGE BASIN
Christy L. Sackfield
Chair: Kenneth L. Campbell
Major Department: Agricultural and Biological Engineering
A mesocosm study was conducted to evaluate phosphorus removal potential of two
isolated wetlands located on the Pelaez & Sons beef cattle ranch in Okeechobee County,
Florida. Research has shown that wetlands function as sinks or as sources of nutrients
depending on various properties of the sediment, vegetation, and water column. Best
management practices for beef cattle ranching are currently being implemented in the
Okeechobee basin to reduce P loads to Lake Okeechobee and to help achieve total
maximum daily loads set by the Florida Department of Environmental Protection.
Phosphorus retention and storage by isolated wetlands in the Okeechobee drainage basin
is one of several BMPs that are currently being evaluated on the Pelaez & Sons beef
cattle ranch. The results from the in-situ mesocosms indicated phosphorus retention
under nutrient loading conditions in a native vegetation wetland on the ranch. Little to no
phosphorus retention occurred at an improved, floralta planted wetland. Studying
wetlands in-situ as opposed to conducting sediment core studies in a laboratory can
potentially provide more realistic estimates of P release or retention under varying water
column P concentrations; however, results from this study were limited due to
desiccation of the wetlands.
1.1 Lake Okeechobee and Water Quality Issues
Excessive phosphorus loads to Lake Okeechobee over the past several decades
have led to algal blooms and fish kills, discouraged recreation, and caused taste and odor
problems in the drinking water from the lake (Nordlie 2001). The Florida Department of
Environmental Protection has established a total maximum daily load (TMDL) of 140
metric tons of total phosphorus (TP) per year to help restore the health of the lake. This
translates into a maximum in-lake phosphorus concentration of 40 ppb. Despite land
acquisition and other direct management efforts, the lake is still suffering from high
nutrient inputs from the surrounding watershed indicating that further research is needed
on best management practices (BMPs) and their implementation to improve water
1.2 Beef Cattle Ranches and Water Quality Issues
Prior to implementation of BMPs in the northern Lake Okeechobee watershed, a
study was conducted from 1985 to 1989 to assess average annual P budgets for each of
25 tributary basins according to their land uses (Fonyo et al. 1991). A follow-up study
was conducted from 1997-2001 to update P budgets based on land management changes
that occurred since the previous study (Hiscock et al. 2003). These two studies were
compared to evaluate effectiveness of the implementation of recent BMPs and
management practices on reducing P loads to Lake Okeechobee. Results indicated that
net P imports to the lake decreased by 28% from the previous study (2380 to 1717 t P),
primarily as a result of changes to four main land uses, dairy (1170 to 458 t P), improved
pasture (1010 to 559 t P), row crops (72 to 545 t P), and sod farms (-70 to -239 t P)
(Hiscock et al. 2003).
Changes in net P imports from dairies were primarily caused by the implementation
of two important programs, the 1987 Florida Department of Environmental Protection
Dairy Rule and the Dairy Buy-Out Program. These programs led to a decrease in the
number of dairies in the northern Lake Okeechobee watershed as well as improved
management practices. Dairy farms only account for 2% of the land use area in the Lake
Okeechobee watershed while improved pasture accounts for 36% (South Florida Water
Management District 2004). Decreases in net phosphorus imports from improved pasture
are the result of lower fertilizer application and improved management practices (Hiscock
et al. 2003).
The significance of dairies in regards to net P imports to Lake Okeechobee are
decreasing. This leaves beef cattle ranches and improved pastures as the current leading
source of net P imports to the Lake. The animal densities and runoff phosphorus
concentrations associated with beef cattle are low compared to dairy cattle; however, the
vast acreage (approximately 470,000 acres) that beef cattle ranches encompass makes
them a major contributor of phosphorus to Lake Okeechobee (SFWMD).
There are 25 drainage basins in the Lake Okeechobee watershed, and of these, four
were identified as priority basins, contributing a disproportionate amount ofP loads to the
lake. These four basins are the lower Kissimmee River Basins (S-65D, S-65E, and S-
154) and the Taylor Creek-Nubbin Slough basin (S-191) (South Florida Water
Management District 2004). The Lake Okeechobee drainage basin consists of
approximately 4,400 square miles, and the four priority basins cover about 450 square
miles of the watershed. These priority regions contribute approximately 35 percent of the
potentially controllable phosphorus load, and account for most of the over-target load to
the lake (SFWMD 2004). Agricultural activity is a major contributor to water quality
problems in these regions. The average annual phosphorus discharge concentration for
improved pastures in the S-154 basin, according to Rules of the SFWMD: Works of the
District Basins Chapter 40E-61, F.A.C is 0.35 mg P/L. Any discharge found to be in
excess of this limit must be reduced to avoid penalty.
According to the U.S. Environmental Protection Agency, runoff from agricultural
activities is considered a nonpoint source of pollution meaning that pollutants are carried
by precipitation over and through the ground as a diffuse source to nearby water bodies.
This process is often expedited on improved pastures due to intensive drainage canals and
high water tables that are common in the central and south Florida region. The most
obvious form of pollution from beef cattle ranches comes from cattle manure, of which a
significant portion is phosphorus attached to manure particles (Reddy et al. 1996). The
amount of manure production varies according to cattle size, age, and productivity of the
animal. The impact of runoff from beef ranches and dairies on water quality can vary
according to cattle manure composition.
Mean soluble reactive phosphorus (SRP) and total phosphorus (TP) concentrations
were found to be higher in dairy manure compared to beef manure (2760 vs. 2100 mg P
kg-1 for SRP and 6500 vs. 5300 mg P kg-1 for TP, respectively (Table 1-1) (Mathews
1992, Graetz and Nair 1995). This is likely due to differences in the diets of beef cattle
and dairy cattle. Beef cattle consume grass with little mineral P supplement while dairy
cattle receive a significant P supplement (Graetz and Nair 1995). However, if you
consider that the amount of land used for improved pasture (36%) now greatly exceeds
the amount of land used for dairy (2%), it becomes evident that the pollution load from
beef cattle ranches can become a serious threat to water quality.
Table 1-1. Mean soluble reactive P (SRP) and total P (TP) concentrations in dairy and
beef cattle manure (Graetz and Nair 1995)
Manure Types SRP (mg P kg-1) TP (mg P kg-1)
Dairy 2760 (1330)a 6500 (2480)
Beef 2100 (880) 5300 (980)
aNumbers in parenthesis are standard deviation values.
Over-fertilization of pastures and direct deposition of cattle manure in waterways
are common problems that can also lead to water quality violations in receiving water
bodies. Results from one study conducted at the MacArthur Agro-ecology Research
Center (MAERC) in 1998 indicated that P loads from improved pastures are often linked
to historical phosphorus fertilizer application, which was applied regularly for 15-20
years prior to 1987 (Bohlen 2003). One case study, conducted in 1997, identified that
bahiagrass roots were capable of penetrating into the subsurface hard pan (Bh horizon).
This suggests that this type of deep-rooted forage may utilize phosphorus located in the
hard pan; therefore, the hard pan can serve as a valuable source of naturally occurring
phosphorus for bahiagrass. Ranchers are able to further reduce their fertilizer
applications if bahiagrass forage is utilized on their ranches and if a substantial pool of
phosphorus is stored in the subsurface hard pan (Bh horizon) from historical P loading.
This information, as well as information from other studies, allowed ranchers to reduce
the amount of phosphorus fertilizer applied to their fields, which saves them money, and
reduces their phosphorus loads in runoff, improving water quality. This study as well as
several other studies provided useful recommendations for the development of a BMP
manual for beef cattle ranches.
In June of 1999, the Florida Cattlemen's Association (1999) published a manual
describing Water Quality Best Management Practices for Cow/Calf Operations in order
to protect water bodies of Florida and to maintain compliance with state water quality
standards. The effectiveness of these BMPs is being evaluated under the Lake
Okeechobee cow-calf BMP research program, which was started by the University of
Florida, Institute of Food and Agricultural Sciences (UF-IFAS). The BMPs
recommended in the cow/calf operations manual are specific to the unique soil and water
conditions of Florida and were developed to help cattlemen meet water quality standards
as well as maintain production and income.
Some common BMPs for cow-calf operations include fencing off cattle from
waterbodies, supplying alternate sources of water, providing ample shade for cattle away
from waterbodies, and installing structures that provide transit of livestock across
waterways without causing sedimentation and/or erosion.
1.3 Hydrologic Conditions Relative to Florida
Soils in the northern Lake Okeechobee watershed consist primarily of Spodosols,
which are characterized by three general horizons, the A, E, and Bh horizons (Graetz and
Nair 1995). Spodosols account for approximately 64% of soils in Okeechobee County
(Nair and Graetz 2002). These soils can often be identified by their slightly acidic pH,
sandy composition, and shallow water tables. In general, the soils in the Okeechobee
drainage basin have <4% clay content and slopes that range from 0-2%, so there is little
to no erosion to contribute to water quality issues (Haan 1995). The A horizon is the
uppermost soil layer which contains small amounts of organic matter and is usually 10 to
50 cm thick. The E horizon contains white sand and varies in thickness. The subsurface
hardpan Bh horizon, or spodic layer, has a low permeability and is identified by a buildup
of organic matter and metals; principally, iron (Fe) and aluminum (Al). For much of the
year, the water table exists between the spodic horizon and the soil surface. The A and E
horizons have a low phosphorus retention capacity; however, the spodic horizon is
capable of retaining high levels of P, if water is allowed to infiltrate into this layer.
According to one study (Campbell et al. 1995), a significant improvement in water
quality occurred in soils with a well-defined spodic horizon. Deep, coarse sand soils with
a barely detectable spodic horizon did not show significant improvements in water
quality in this same study (Campbell et al. 1995). Deep percolation of soluble P into the
iron (Fe) and aluminum (Al) rich lower spodic (Bh) horizons is restricted due to low
permeability; therefore, high soluble P concentrations exist in surface flow or lateral
subsurface flow above the spodic layer (Graetz and Nair 1995).
The topography of south Florida is essentially flat with an abundance of streams,
ponds, marshes, and wetlands. Wetlands currently account for approximately 18% of the
land cover in the Lower Kissimmee River region, where this research takes place (Reddy
et al. 1996). Historically, most improved pastures were once swamplands that were
developed using intensive agricultural drainage techniques. Some pastures maintained
low-lying depressions that were more difficult to completely drain, forming small ponds
or wetlands across the landscape. These wetlands have been identified as having a major
role in reducing the amount of nutrients in pasture runoff
1.4 Phosphorus in Wetlands
Phosphorus is an essential nutrient for plant and animal growth; however, excessive
accumulation of nutrients in wetlands or any other waterbody can cause undesirable
environmental and economic consequences. Phosphorus can be found in the soils,
minerals, living organisms, and water columns of lakes and wetlands. Wetlands may act
as sinks, sources, or transformers of nutrients within a watershed. Although wetlands
generally serve as nutrient sinks, changing hydrologic conditions, decaying vegetation,
and animal use can alter the nutrient cycle leading to an export of nutrients from the
wetland (Mitsch and Gosselink 1993).
The biotic and abiotic factors that influence the P cycle in wetlands are dependent
on water levels, amount of available P, soil P adsorption, and seasonal changes in uptake
by microorganisms (Daniel et al. 1998). The phosphorus cycle does not have a
significant gaseous loss mechanism as does the nitrogen cycle, and it occurs primarily in
the sedimentary cycle. The primary storage mechanism for phosphorus in wetlands is
sediment accumulation, accounting for more than 95% of the P in natural wetlands
(Faulkner and Richardson 1989). Vegetation may account for large amounts of
phosphorus storage during the active growing season, but vegetation serves as a relatively
short-term storage mechanism overall because once the plants senesce, some of the
bound nutrients are released back into the soil or water column (Reddy et al. 1996).
Microorganisms and sediments sorptive characteristics control the initial uptake of P in
Although some wetlands have a short-term ability to remove nutrients, there exists
the possibility that they may become nutrient saturated leading them to release nutrients
to downstream waterbodies. The transport, storage and biological availability of P in a
watershed may be greatly influenced by the cycling of P in wetlands (DeBusk 1999);
therefore, it is necessary to closely examine nutrient dynamics in these systems in relation
to in-situ experimentation to obtain realistic P uptake rates.
1.5 Soil/Water Phosphorus Interactions (P-Cycle)
It is important to understand the phosphorus cycle in wetlands in order to
understand how wetlands may serve to improve water quality for improved pastures, as
identified in the cow-calf BMP manual. Wetlands perform many important
biogeochemical functions in watersheds, which include trapping sediments, storing
nutrients, and transforming inorganic nutrients to organic forms. Some of the major
processes that influence the phosphorus cycle in wetlands are listed as follows: (1)
diffusion, (2) plant uptake, (3) litterfall, (4) sedimentation, (5) decomposition, (6)
sorption, and (7) burial and peat accretion (Debusk 1999). Phosphorus sorption in
wetlands occurs primarily at the oxidized sediment-water interface by diffusive processes
(Reddy et al. 1996).
Diffusion involves concentration gradients that exist between the soil pore water
and the water column. Dissolved forms of P are transferred from regions of high
concentration to regions of lower concentration. For instance, if the P concentration of
the soil porewater is greater than the P concentration of the water column, then
phosphorus will diffuse out of the soil and into the overlying water until equilibrium is
reached and the gradient no longer exists. However, equilibrium is constantly changing
due to several factors such as runoff events carrying nutrients into the wetland or wind
affects re-suspending P-rich sediments into the water column. Diffusion is one of the
primary mechanisms by which phosphorus is transferred between soil and the water
column within a wetland. Phosphorus sorption by wetlands soils can only occur when
inorganic P is in direct contact with sorptive sites. This requires diffusion of water
column P into the underlying soils before phosphorus can be retained (Reddy et al. 1995).
Plants are capable of extracting needed nutrients by their roots either from the soil
or from the water column if roots are located in the water column. However,
macrophytic uptake of nutrients occurs primarily through roots in the soil on a daily to
weekly time scale. Algae grow attached to plant stems, the sediment and litter, or within
the water column. Algal response to nutrient availability can be more rapid, on the order
of hours to days. The form of P taken up by plants and algae is inorganic and is primarily
orthophosphate (HPO42- and H2P04-). If an equilibrium P concentration has been reached
between the soil and water column, and if no new nutrients are added, then the plants will
generally take up phosphorus from the soil, causing a shift in equilibrium concentration.
Plants play a major role in the phosphorus cycle because they can assimilate large
amounts of nutrients into their biomass. However, once the growing season has ended
and the plants begin to senesce, nutrients will continue to be pumped by the plant, but the
nutrients will not necessarily be utilized by the plant and will therefore be quickly cycled
back into the system.
Litterfall contributes to the phosphorus cycle in that once plants die or senesce, the
dead plant tissue will fall to the soil surface releasing a portion of the phosphorus that
was once incorporated in its biomass. Depending on the hydrologic condition, litter
quality and plant productivity, an accumulation of dead biomass will occur. This layer of
detritus will provide nutrients for other living plants and organisms in the wetland once it
has been broken down into a useable form. More readily available forms of nutrients in
the wetland will be consumed first. Therefore, much of the P storage by plants is
transient, returning much of their nutrients back into the nutrient cycle.
Sedimentation often occurs in wetlands because particulate matter that was once
suspended in the water column settles out due to lower velocities. These lower velocities
as well as shallow water depths and emergent vegetation, serve to filter sediments out of
the water column. Erosion is a major contributor of nutrient pollution in many parts of
the United States; however, its role in sandy soils with no clay coatings or soils where
metal oxide associates with parent material is not as profound due to low phosphorus
retention capacities. Low relief topography reduces impacts from erosion in Florida.
Wetlands can "fill-in" over time due to heavy sediment loading from the surrounding
Decomposition is the breakdown of organic matter, such as dead plant tissue, by a
variety of microorganisms that utilize organic carbon as an energy source. These organic
compounds must be broken down into smaller components and ultimately into
orthophosphate, which can be utilized as a nutrient or diffused back into the soil or water
column. Temperature, pH, nutrient availability, carbon quality, redox potential, and
availability of electron acceptors including oxygen, nitrate, sulfate, manganese, and iron
influence decomposition rates.
Sorption involves two processes that include the following: 1) adsorption of
orthophosphate by clays and iron or aluminum oxides in the soil, and 2) precipitation of
phosphate with either iron and aluminum oxides or dissolved calcium, to form solid
compounds in the soil or water column. Phosphate minerals are often very stable in the
soil and contribute to the long-term storage of P in wetlands. Soils with high
concentrations of soil-extractable aluminum and soils that are predominantly mineral
soils have shown high phosphate removal rates, due to their sorptive capacities
(Richardson 1985). The phosphorus retention capacity of wetland soils is directly
correlated with the oxalate extractable iron and aluminum content and organic matter
content according to various studies (Reddy and D'Angelo 1994, Reddy et al. 1996).
Burial and peat accretion takes place when organic matter is gradually buried in the
soil profile, making it more resistant to decomposition. Peat accretion occurs when
deposition of organic matter is more rapid than the decomposition rate. Peat accretion is
a very slow process and often does not produce conditions preferable for reducing high P
loading rates to levels that are acceptable for water quality targets.
If a wetland is being used as a BMP to help reduce phosphorus loads in runoff to
downstream waterbodies, then it is important to know exactly how much P is entering the
wetland as compared to how much P is leaving the wetland. If the wetland is not
successful at reducing P loads to acceptable levels, then it is possible that the wetland has
reached its maximum potential for P storage and other means are necessary to further
reduce P loads from that particular site. If high levels of P are released from a wetland,
then the water quality of nearby streams and waterbodies will ultimately suffer. Excess
phosphorus loads can lead to eutrophication and subsequent destruction of ecological
1.6 Measuring Phosphorus Uptake/Release
In order to assess the potential for a wetland to store phosphorus and remove P
from runoff, it is important to determine historical P loading to the wetland and P loading
capacity of the wetland. Analyzing the amount of P in wetland soils can provide great
insight into these types of questions because storage of P in soils is often the primary
means of P storage in wetlands. Common measurement techniques involve taking
several soil cores from the wetland site, storing them on ice, and returning them to the
laboratory for testing. Chemical fractionation schemes are then used to identify forms of
inorganic and organic P (labile and non-labile pools). One such method involves
subsequent extractions of soils with KC1, NaOH, and HC1 (Reddy et al. 1995). The major
inorganic P pools include the following: loosely adsorbed P, Fe- and Al- bound P, and
Ca- and Mg- bound P. The following describes a laboratory procedure performed in the
Soil and Water Science department at the University of Florida:
1. Extraction with KC1 represents the labile pool of inorganic phosphorus,
which is important for plant growth and for controlling P concentrations of
the overlying water. This pool of phosphorus typically ranges from 0.1 to
1.1% of the total P in wetlands.
2. Extraction with NaOH represents the P associated with amorphous
oxyhydroxide surfaces and crystalline Fe and Al oxides. Typical ranges for
this pool are 17-43% of total P.
3. Extractions with HC1 represent the P associated with Ca and Mg minerals.
This pool is essentially unavailable to biological assimilation and is not often
found in the Okeechobee watershed because of low pH.
*Residual P not extracted by KC1, NaOH, and HC1 is considered to be highly
resistant and biologically unavailable. Fulvic acid bound P (NaOH-P) is
biologically reactive and can be 6-57% of total P.
This method is commonly used to develop Phosphorus Sorption Isotherms for
various wetland soils.
1.7 Using Mesocosms for In-Situ Analysis of Phosphorus
The influence of vegetation, organisms, grazing, seasons, etc. are not taken into
account during the phosphorus sorption isotherm method. By removing the soil cores
from their natural environment and testing them in a controlled laboratory environment,
the results are indicative of phosphorus uptake/release by soil only, which may or may
not be useful for real world applications. Phosphorus sorption isotherms are conducted at
a certain controlled temperature and agitation of soil within the solution maximizes
uptake/release of phosphorus by the soil. In the field however, especially in Florida,
environmental conditions are highly variable and extraction results from laboratory
isotherms may not provide the best possible information about phosphorus uptake and
release. The calculation of maximum retention capacity based on isotherms usually
underestimates potential P sorption by soils under field conditions (Reddy et al. 1995)
because in the field there are numerous factors contributing to phosphorus uptake/release
from the water column and soil porewater, for example, vegetation and algae. The
overall P sorption in the field would be higher because there would be more available
sinks for P compared to isotherm studies.
Another approach to measuring P uptake/release in soils is using mesocosm type
structures for an in-situ experiment as opposed to a laboratory type method. Mesocosms
are middle-sized (typical range of m2 to ha), constructed ecosystems that serve as useful
tools for ecological research. They represent extensions of the microcosm method
because their relatively large size allows the incorporation and use of much more
ecological complexity than is possible in conventional microcosms (Kangas and Adey
1996). Another definition of mesocosms is that they are outdoor semi-controlled
ecosystems whose physical dimensions and basic water chemistry are known and
controlled (Boyle and Fairchild 1997). The initial physical, chemical, and biological
conditions within mesocosms will ultimately shape the final results of experiments. In
Florida, mesocosm studies are used to provide the unique soil-type constraints, water
flows, evaporation, rainfall, wind effects, and water chemistry. One study on the "role of
mesocosm studies in ecological risk analysis" recommended that researchers develop
closer working relationships with mathematical modelers in linking computer models to
the outcomes of mesocosm studies (Boyle and Fairchild 1997). By coordinating the
development of models with the formation of tests and validations, the realism and
accuracy of both activities can be improved (Boyle and Fairchild 1997). The results of
this type of collaborative effort could help provide a foundation for the development of
ecological indicators and improve strategies for environmental monitoring and
assessment on the regional level (Boyle and Fairchild 1997).
Unlike microcosm studies, mesocosm studies can be designed to include critical
ecological variables such as site-specific water-quality conditions, native flora and fauna,
and various strengths and complexities of ecological interactions (Boyle and Fairchild
1997). Factors that influence the fate of nutrients, such as phosphorus, within a system
can be broken down into three main components, physical, chemical, and biological.
Some physical differences include solar radiation, temperature, and the transparency of
the water. Chemical factors that might alter the behavior of nutrients in a system include
pH, redox potential, and level of critical inorganic nutrients. Biological factors may
include type of crops, standing versus floating, levels of organic carbon, and rates of
metabolic activity. Differences in these ecological conditions occur regionally, and are
expected to produce differences in the results of mesocosm experiments as well as model
outputs (Boyle and Fairchild 1997).
1.8 Goals and Objectives
1. Evaluate the influence of nutrient loading on water column nutrient concentrations
in two isolated wetlands by measuring changes in water quality over time with an
in-situ mesocosm approach.
2. Compare nutrient uptake/release in floralta planted wetland to native
3. Address seasonal/water level effects on nutrient uptake/release in wetlands.
4. Demonstrate the effectiveness this in-situ mesocosm approach for
determining equilibrium phosphorus concentration (EPC) values for the wetlands.
A mesocosm study was conducted on two isolated wetlands located on Pelaez &
Sons beef cattle ranch in Okeechobee County, Florida (latitude: 270 9' 25.2" N,
longitude: 80 12' 57.6" W). A circular area with a diameter of approximately 68 cm was
enclosed within each mesocosm structure. The area contained the natural soil, water, and
organisms in order to produce in-situ results. Some mesocosms were spiked with a
treatment that represented a typical runoff effect, which would be contaminated with N
and P from cattle manure or fertilizer. Treatments of 0.5 ambient concentration, ambient
concentration, and two times ambient concentration were used to represent a range of
runoff events. The range of treatments was limited by the size of the mesocosms and
2.1 Site Description
This research was conducted on Pelaez & Sons beef cattle ranch located on 128th
Street in Okeechobee County, Florida (Priority Basin S-154). The ranch is approximately
617 hectares and was historically and is currently managed as a cow/calf operation. In
the late 50's and early 60's, however, row crops such as tomatoes were produced on
portions of the property. The property can produce enough forage to support an
estimated 580 animal units; however, the system currently supports an average of 533
cows, with an average weight of 454 kg and 35 bulls with an average weight of 612 kg.
The native cover used for grazing is primarily 'Pensacola' bahiagrass (Paspalum
notatum) with some fields planted with Floralta limpograss (Hemarthria altissima). The
ranch is divided into approximately fifteen separate pastures and is maintained
hydrologically through a series of canals and culverts. Figure 2-1 illustrates locations of
the two wetlands on the ranch as well as approximate locations of the mesocosms within
Figure 2-1. Map ofPelaez & Sons Ranch, Okeechobee County, Florida, with wetland and
The study site is characterized by subtropical climate conditions with distinct wet
and dry seasons. Average annual rainfall is approximately 1160 mm, with most rainfall
occurring during the wet season months of May through October. Four storms directly
affected Florida around the time of this experiment in 2004, Hurricanes Charley, Frances,
Ivan, and Jeanne. Soils are generally Spodosols that are poorly drained fine sands with
neutral to slightly acidic pH. A spodic horizon or hardpan exists at a depth of 50 to 76
cm below the soil surface and a typical soil profile consists of the A, E, Bh, and Bw
horizons. The sandy pasture soils and lateral groundwater movement in the uplands
cause much of the P to be transported to the wetlands or channels that cut through the
ranch rather than being held efficiently on site (Graetz and Nair 1995).
These soil conditions are indicative of high water tables and saturated soil
conditions during the wet season; therefore, there are several depressional wetlands
located on the ranch, covering approximately 7% of the ranch area. Two specific
wetlands on the ranch were chosen for this study. Site 1 has a drainage area of
approximately 58 ha and is characterized by Myakka Fine Sand and Immokalee Fine
Sand. The average elevation at this site is 9.3 meters above mean sea level. The wetland
drains to a flume that connects to a channel that transports water off of the ranch and
eventually towards Lake Okeechobee. Drainage from the watershed eventually ends at
gated control structures leading to Lake Okeechobee (Reddy et al. 1996). Site 4 has a
drainage area of approximately 22 ha and is characterized by Myakka Fine Sand. The
topsoil layer in this wetland is believed to be below the hardpan meaning that the soil is
primarily characterized by coarse, white sand. The average elevation of this wetland is
approximately 9.9 meters above mean sea level and this wetland also drains to a flume
that is connected to a channel. Myakka and Immokalee soils are characterized by their
poor phosphorus retention capacity and high water permeability. These wetlands are
seasonal with periods of standing water in the wet season (May-Oct) and drying out in
the dry season (Nov-Apr).
The wetlands on this ranch provide a variety of ecological functions as well as
hydrologic functions. The wetlands vary in size, shape, and vegetative cover, but they
are similar in that they are both isolated, receiving inputs from surface runoff or
subsurface flow and precipitation only. The wetland at Site 1 is covered with native
vegetation including Panicum hemitomon, Juncus effuses, Pontederia cordata, and
Polygonum hydropiperiodes. The vegetation at Site 4 wetland is dominantly Floralta
limpograss (Hemarthria altissima) with some bahiagrass (Paspalum notatum) and some
Polygonum hydropiperiodes. Research has been conducted on the usefulness of Floralta
for grazing purposes, and the average daily gain (ADG) of cows grazing on Floralta
compared to Bahiagrass was found to be somewhat similar (Sollenberger et al. 1988).
Surface water in the wetlands and ditches on the beef cattle ranch are colored,
indicating humic and fulvic acids within the water column. Organic phosphorus
associated with these acids accounts for greater than 40% of total phosphorus in the water
column. Humic and fulvic acids give the water a tea color and are relatively resistant to
biological degradation (Reddy and D'Angelo 1994).
The wetlands for this study were selected on the basis that they play some role in
improving water quality on the ranch. Restoration of hydrology of isolated wetlands is
expected to reduce phosphorus loads to Lake Okeechobee by storing surface runoff and
assimilating excessive nutrients before they are transported downstream. The
implementation of landowner and typical cost-share Best Management Practices for
addressing nonpoint sources are expected to achieve a significant reduction in TP loading
to the lake. The Best Management Practices that will be implemented at this site include
raising the water level (flashboard risers) to hold water in the wetlands and slowly
releasing it after rains, and restoring a longer hydroperiod in the isolated wetlands.
Reddy et al. (1996) suggested that increasing the hydraulic retention times of a wetland
would serve to improve the overall phosphorus retention capacity of the wetland. Also,
alternative water supplies will be installed on the ranch to help keep cattle out of the
wetlands and canals to prevent direct contamination by cattle manure. This study was
conducted in the pre-BMP phase of the project meaning that water was allowed to flow
from the wetlands to the flumes naturally without implementation of any water control
2.2 Trial Run at Stormwater Ecological Enhancement Project
A trial run experiment was conducted at the Stormwater Ecological Enhancement
Project (SEEP) site on the University of Florida campus using smaller mesocosm
structures to assess the time frame needed to conduct a complete experiment on the
wetlands at Pelaez & Sons Ranch. Table 2-1 describes the characteristics of the
polycarbonate tubes used for the trial run experiment.
Table 2-1. Polycarbonate tube dimensions for trial run at Stormwater Ecological
Enhancement (SEEP) site on University of Florida campus.
Total Length of Tube = 50.8 cm
Diameter (Outside) =15.2 cm
Wall Thickness = 0.1 cm
Diameter (Inside) 15 cm
Cross-Sectional Area= 176.7 cm2
Total Volume = 5442.8 cm3 (5.4 Liters)
Before the polycarbonate tubes were installed, a composite water sample was taken
from three locations within SEEP. This was conducted before serious disturbance of the
site occurred. The sample was filtered, acidified and placed on ice to be taken back to the
laboratory for analysis of soluble reactive phosphorus (SRP). Comparing ambient water
quality in the SEEP to the water quality inside the small mesocosms provides useful
information on the effect of the mesocosm structure on water quality (ambient vs.
Nine polycarbonate tubes were installed at SEEP on July 15, 2004. Native
vegetation was carefully retained in each tube, and then each tube was pushed
approximately 10 cm into the soil. Care was taken to ensure a proper seal was formed
between the bottom of the mesocosm and the surrounding soil to avoid leakage.
An assumption was made that the ambient soluble reactive phosphorus level in the
SEEP was approximately 1 mg/L. Using this assumption, treatments were applied with
the goal of raising the SRP concentration in the spiked treatment mesocosms to 3 mg/L
and reducing the SRP concentration in the diluted treatment mesocosms to 0.5 mg/L.
Table 2-2 shows the treatments applied to the mesocosms, with triplicates of each
Table 2-2. Treatments applied at SEEP trial run.
Control (Cl, C2, C3) No treatment applied
Spiked (S1, S2, S3) Added -150 mL of 100 mg/L P solution
Diluted (D1, D2, D3) 50% dilution with deionized water
Treatments were carefully applied; ensuring that water level inside the mesocosms
remained the same as water level outside the mesocosms (1/2" tube used as a siphon).
This was done to prevent porewater exchange with the water column due to a head
pressure differential. After treatments were applied, water samples were taken from each
tube with a syringe, filtered, acidified, and placed on ice for transport to the lab for
analysis of SRP. Water samples were taken again on July 16, 2004 (composite ambient
and samples from each mesocosm). Two more water sampling dates had been set for the
18th and 20th; however, on the 18th, all of the mesocosms had been flooded due to a storm
event raising water levels. Further water sampling would not have provided meaningful
information. Samples taken were refrigerated until analysis, which took place within 28
days of sampling. SRP results from the trial run experiment are listed in Table A-1.
It was noted during the trial run experiment that the water level inside the
mesocosms did not consistently coincide with water levels outside the mesocosms. The
clay content of the soil allowed a strong seal to form between the bottom of the tube and
the soil. Because SEEP site serves as a detention basin, it receives high runoff loads from
the surrounding area during storm events. As the water level at SEEP rose, it did not rise
as quickly inside the tube because the tube only received water input from rainfall and
not from surface runoff The same occurred as the water level dropped after the rainfall
event. The water level dropped much more quickly in SEEP as compared to the water
level inside the mesocosms. This led to the conclusion that future mesocosm designs
may need to be modified to include some type of bladder that allows water level to
equilibrate between the inside and outside of the structure without head differentials
potentially resulting in mass flow of porewater into or out of the mesocosms. In soils
with lower clay content, the differences in water level would possibly not have been as
profound because the levels would have equilibrated naturally through the soil.
2.3 In-Situ Wetland Mesocosm Study
This experiment was conducted to evaluate various phosphorus interactions within
the wetlands. An in-situ experimental design was used so that all factors influencing
phosphorus storage and release in wetlands could be evaluated. The direction of
phosphorus flux is primarily regulated by the concentration gradients across the
soil/water interface; therefore, in-situ mesocosms were subjected to high, low, and
ambient phosphorus levels to compare the flux of phosphorus between the soil and
overlying water column. Changes in water quality concentrations of the water column
were regularly monitored over time in hopes of obtaining phosphorus uptake and release
* The spike and dilute treatments will cause a shift in equilibrium between the
phosphorus concentration in the overlying water column and the soil porewater,
leading to a diffusive flux
* SRP concentrations in all mesocosms will converge at an equilibrium phosphorus
concentration where uptake equals release
* The SRP concentration in the water column will decrease over time in the spiked
mesocosms due to uptake by soil, vegetation, algae, etc.
* The SRP concentration in the water column will increase over time in the dilute
mesocosms due to release of phosphorus from the soil, vegetation, algae, etc
The treatments for this experiment are shown in Table 2-3.
Table 2-3. Treatments applied to mesocosms in wetlands at Pelaez & Sons Ranch.
Mesocosm Wetland Phosphorus Treatments
1-3 Floralta Wetland (Site 4) 0.5A
4-6 Floralta Wetland (Site 4) A (control)
7-9 Floralta Wetland (Site 4) 2A
10-12 Native Vegetation (Site 1) 0.5A
13-15 Native Vegetation (Site 1) A (control)
16-18 Native Vegetation (Site 1) 2A
*Each mesocosm will be replicated three times for comparison. A=Ambient.
For the actual experiment, a total of eighteen mesocosms were constructed, nine
mesocosms for Site 1 and nine for Site 4. The mesocosms were constructed out of
Fiberglass Kalwall material (Sun-Lite HP), which was ordered from Solar Components
Corporation. This material was chosen for being lightweight, shatterproof, highly
transmissive of sunlight (85-90%), impact resistant, and easily manageable. The
mesocosms had a wall thickness of 0.1016 cm, were 61-91 cm in height, and
approximately 68 cm in diameter when completed. The circular mesocosms were
constructed from a rectangular sheet; therefore, the edges had to be overlapped and sealed
using a fiberglass epoxy and aluminum tape along the outside wall for additional support.
The original kalwall sheet height was 152 cm and was cut to produce mesocosms at 60
cm and 90 cm heights, to reduce costs. The mesocosms did not have inflows or outflows;
they simply isolated the areas within the mesocosms from the rest of the wetland to
assess effects of various influential factors on phosphorus retention in the soils and
floodwaters. Bladders were not incorporated into the design of these mesocosms, as
suggested from the trial run, to avoid compromising the structural integrity of the
mesocosms. Water level differences between the inside and outside of the mesocosms
during the experiment were not expected to be as significant at Pelaez & Sons due to the
sandy soil conditions. Also, water level fluctuations in the two wetlands were not
expected to be large. The mesocosms were open at the tops and bottoms so that they
could be pushed into the wetland soil as well as exposed to the atmosphere at the top to
maintain field conditions as much as possible. It was assumed that light and temperature
remained similar inside and outside of the mesocosm because of the material chosen for
the mesocosm design.
Once the mesocosms were constructed, solutions for treatments of the mesocosms
were prepared. Water quality results from a trial run experiment at Pelaez & Sons Ranch
were used to design and prepare the treatments for the actual experiment. Results from
this trial run are presented in the appendix, Table A-2, and Figures A-i, A-2, and A-3.
Based on the trial run, ambient soluble reactive phosphorus (SRP) concentration at Site 1
was 0.23 mg/L and the ambient ammonium (NH4+) concentration for this site was 0.20
mg/L. For Site 4, ambient SRP concentration was 1.6 mg/L and ambient NH4
concentration was 0.2 mg/L. The treatments used for the SEEP experiment were
approximately the same treatments used in this experiment, three control mesocosms (no
treatment), three 50% diluted with de-ionized water, and three nutrient enriched (spiked)
mesocosms. The desired concentrations after applying the treatments for the spiked
mesocosms were 3.0 mg/L SRP and 0.40 mg/L NH4+ for Site 1 and 3.2 mg/L SRP and
0.40 mg/L NH4+ for Site 4.
A 500 mg P/L solution was prepared using phosphorus fertilizer (KH2PO4) and a
100 mg N/L solution was prepared using nitrogen fertilizer (NH4C1). These solutions
were prepared in separate six-gallon containers to be carried into the field for treatment
application. A 2000-ppm bromide tracer solution was also prepared to add to each
mesocosm. Bromide was used successfully as a subsurface water-movement tracer in an
experiment conducted on flat, sandy, high water table soils in the Kissimmee River Basin
of Florida, which drains into Lake Okeechobee in a former study (Campbell et al. 1995).
Other tracers such as chloride and rhodamine tracers did not prove useful in the region of
interest due to high background concentrations and sorption of the tracer to various
elements, making detection and recovery difficult. For these reasons, bromide, a
conservative tracer, was used to raise the bromide concentration in each mesocosm to 5
mg/L to provide some insight into the hydrologic conditions occurring inside the
mesocosms. Evapotranspiration and groundwater losses were estimated using bromide
Once the equipment was prepared and carried into the field, locations for the
mesocosms within the wetlands were identified. General locations for the mesocosms
had been identified during the trial run experiment. An area was chosen within the
wetland that provided similar vegetative cover for all the mesocosms to be placed in the
wetland as well as similar water levels. It was difficult to find locations for all eighteen
mesocosms in the wetlands that provided similar conditions. The area in which the
mesocosms could be placed was limited by water depth.
On October 20, 2004, installation of the mesocosms within the wetlands
commenced. Each mesocosm was carefully placed over the vegetation and lowered
down to the soil. Once soil was encountered, pressure was gently applied by hand to
minimize movement of the mesocosm while a knife was used to cut through vegetative
roots and the soil surrounding the tube. A board was placed on top of the mesocosm so
even pressure could be applied while pushing the structure into the soil. Two people
were required to successfully install the mesocosms in the wetlands. The mesocosms
were pushed approximately 10 cm into the soil. Care was taken to ensure that the
mesocosms were deep enough in the soil to prevent leakage under the bottom walls of the
mesocosm. To test for leaks, water level inside a test mesocosm was lowered below the
outside water level, and a stopwatch was used to determine how long it took the water
level to fill back up (water level would not change or would change very slowly inside
the mesocosm if properly constructed). Installing the structures while attempting to
cause as little disturbance as possible was an important aspect of the experiment. Once
the mesocosms were installed, they were allowed to settle for 24 hours before treatment
After a 24-hour settling period, ambient water samples were taken from each
wetland to establish initial water quality conditions. Water samples were taken from each
mesocosm for the same purpose. Comparing ambient water samples from the wetlands to
initial condition samples from each mesocosm provides an estimate of the degree of
disturbance from installation.
Treatments were then applied on October 21, 2004. The control mesocosms did
not receive any treatment except for the bromide solution. All eighteen tubes received a
bromide addition based on water levels that was expected to raise the concentration to 5
mg Br/L inside each tube. Each Spike mesocosm received additions from the phosphorus
and nitrogen solutions according to their water depths to raise the N and P concentrations
to a predetermined level. Because the amount of solution added to the spike tubes was
minimal compared to the total water volume in the tube, water levels inside and outside
the tube remained the same. For the dilutions, 50% of the water volume within the tube
had to be removed and then replaced with de-ionized water. To do this, a 1 1/2" siphon
tube was used to connect the inside water level to the outside water level while a clear
plastic trash bag placed inside the mesocosm was slowly filled with the appropriate
amount of de-ionized water. The siphon tube allowed water levels to constantly
equilibrate, pushing 50% of the water volume out of the mesocosm and replacing it with
the de-ionized water volume inside the plastic bag. Once water levels were the same
inside and outside the mesocosm, the siphon tube was carefully removed and de-ionized
water inside the trash bag was released into the mesocosm, thus completing the 50%
dilution treatment. The de-ionized water had to be carefully released from the plastic bag
to avoid disturbance. The bromide was added after the dilution so that bromide would
not be lost through the siphon tube. Each tube was very carefully stirred for complete
mixing of the water column following treatment application. Treatments were applied at
random to the mesocosms installed in each wetland, as indicated in Table 2-4.
Table 2-4. Treatments applied at Pelaez Ranch based on water volumes in mesocosms.
Circumference (cm) 208.8 214.9 219.5 214.9 208.8 211.8 199.6 210.3 208.8
Radius (cm) 33.2 34.2 34.9 34.2 33.2 33.7 31.8 33.5 33.2
Bromide Addition (mL) 107.0 156.0 108.0 138.0 171.0 112.0 117.0 142.0 95.0
Ammonium Chloride Addition (mL) 95.0 114.0 76.0
Phosphorus Addition (mL) 261.0 313.0 210.0
Water Addition (L) 27.6 34.1 22.5
Control Dilute Spike
Site 4 1 2 3 1 2 3 1 2 3
Circumference (cm) 210.8 210.3 216.4 210.3 210.3 216.4 213.4 210.3 213.4
Radius (cm) 31.5 31.4 32.4 31.4 31.4 32.4 31.9 31.4 31.9
Bromide Addition (mL) 140.0 120.0 137.0 106.0 116.0 163.0 129.0 134.0 132.0
Ammonium Chloride Addition (mL) 103.0 107.0 106.0
Phosphorus Fertilizer Addition (mL) 165.0 172.0 170.0
Water Addition (L) 21.2 23.2 32.6 ---
Water quality sampling was again conducted within each mesocosm immediately
following treatment application to assess effectiveness of the treatments. Water quality
results taken immediately after treatment application and until the end of the experiment
were used to evaluate concentration changes over time. Water levels were also recorded
each time water samples were taken. Water sampling was conducted on day 1, 2, 4, and
8 following treatment application. Details on water sampling techniques and water
quality parameters are discussed later.
Vegetation and soil samples were collected at the end of the experiment.
Vegetation above the water surface was clipped, collected in individual Ziplock bags for
each mesocosm, and then stored on ice for transport back to the lab. Vegetation below
the water surface and above the soil surface was also clipped and stored in a separate
ziplock bag for analysis. Soil cores were taken from within each mesocosm using a
special wetland-coring device. Cores were stored in ziplock bags and placed on ice for
transport to the lab.
2.4 Water Quality Analysis
Water samples were collected and analyzed for soluble reactive phosphorus (SRP),
total phosphorus (TP), nitrate (NO3), ammonium (NH4+), total kjeldahl nitrogen (TKN),
and bromide (Br). The next several paragraphs describe methods used for water quality
Water total Kjeldahl nitrogen (TKN) was assessed using a Kjeldahl digestion
method in compliance with EPA Method #351.2. Reagents used include concentrated
sulfuric acid (H2S04) and Kjeldahl salt catalyst (K2SO4 + CuSO4). Samples were taken
in the field with a syringe, acidified with concentrated H2S04, and then stored on ice for
transport to the lab for analysis.
Water total phosphorus was assessed using an autoclave method. Reagents
included 11 N H2SO4 and potassium persulfate. Samples were digested in the autoclave
for 60 minutes at 1210C and 15 psi. Water TP in the sample digests was determined
using a TechniconTM Autoanalyzer, EPA Method 365.1 (EPA 1993). Samples were
taken in the field with a syringe, acidified with concentrated H2SO4, and then stored on
ice for transport to the lab for analysis.
Soluble reactive phosphorus, nitrate, and ammonium concentrations were
determined using a TechniconTM Autoanalyzer. Samples were taken in the field with a
syringe, filtered through 0.45-rm filter paper, acidified with concentrated H2SO4, and
then stored on ice for transport to the lab for analysis.
Bromide was analyzed using a Perkin Elmer Series 200 Auto-sampler and a
RFICTM Ion Pac AS4A-SC Analytical 4x250 mm Ion Exchange Column. A wavelength
of 205 nm and a flow rate of 1 mL/min were used on the water samples to determine
bromide concentrations. The eluent used was 20 mmol sodium borate powder
The procedures used for collecting water samples in the field were as follows: One
composite sample was taken outside of the mesocosms, and one sample was taken inside
each mesocosm (sampling before treatment application and then Day 1, Day 2, Day 4,
and Day 8 following treatments).
The soluble versus particulate forms of P are useful in quantifying physical
behavior during transport. Transport processes affect the availability of P for
assimilation by biota and retention by soils and sediments (Reddy et al. 2003).
Groundwater monitoring wells were installed around the perimeter of the wetlands
to determine gradients between the surface water in the wetland and groundwater;
however, these data were not available during the time frame of this experiment. The
wells were instrumented with water level recorders to track water level changes. Water
levels were to be measured both above and below the spodic horizon in order to estimate
the flow through this layer.
2.5 Soil Analysis
Soil cores were taken at the end of the experiment with a special coring device that
was developed at the University of Florida Soil and Water Science Department. Soils
were analyzed for bulk density, water content, pH, total phosphorus from ashing, water-
soluble phosphorus, and Fe/Al oxalate bound phosphorus.
To calculate bulk density and water content, the whole wet soil sample was
weighed in the original collection bag, noting the number of soil cores per bag. The
volume of the soil sample was calculated based on the size of the corer used to take the
sample and the number of cores. The corer had a cross sectional area of 38.5 cm2 and a
length of 10 cm. The sample was then homogenized and a 0.28 kg sub-sample was
taken, picking out roots from the sample. A portion of the sub-sample was placed in an
aluminum tin, weighed, and then dried at 700C for a minimum of 72 hours. After drying,
the soil sample and tin were weighed again. This procedure allowed for the calculation
of water content and bulk density of the soil samples. Once the soil sample was dried, it
was hand ground and placed in a scintillation vial for eventual analysis of total
phosphorus from ashing. A portion of the wet, homogenized soil sample was stored for
eventual water extraction to assess water-soluble phosphorus in the soil.
Soil pH was determined using an ORION SA 720, Fisher AR50 pH meter. The
meter was calibrated using a 2-point calibration with buffer solutions pH 7.00 then pH
4.00 to obtain a standard curve. Twenty mL of deionized water were added to 10 mL of
wetland soil sample then stirred and allowed to equilibrate before taking the pH
measurement (Hanlon 1994, Thomas 1996).
Total phosphorus was determined using the ignition or ash method in compliance
with EPA Method 365.1. Soils were dried (at 700C for 72 hours), ground in a ball mill,
and passed through a #40 mesh sieve. Then, 0.2-0.5 g prepared soil as well as an external
QC standard (peach leaves) were placed in a muffle furnace for 30 min. at 2500C then 3-4
hr. at 550C. Weights were determined before and after being placed in the furnace for
loss of ignition (LOI) calculations. The ash was moistened with distilled, de-ionized
water and 20 mL of 6.0 M hydrochloric acid (HC1). Samples were then placed on a hot
plate at 100-1200C until dry, and then an additional 30 minutes on high heat. Samples
were re-moistened with DDI water and 2.25 mL of 6.0 M HC1, and then placed back on
the hot plate to bring to a boil. The samples were filtered with Whatman #41 filter paper,
and then transferred to 20-mL HPDE scintillation vials for TP analysis. The TP
concentrations of digested solutions were determined analytically on a TechniconTM
Autoanalyzer. Calculations were as follows (Anderson 1976, Jackson 1958) (EPA
Total Phosphorus (pg g-')= TP conc. (pg mL-1) 50 (mL) / soil wt. (g)
Loss of Ignition (%) (LOI) = 100-[Ash weight (g) / soil wt. (g) 100]
Water-soluble phosphorus was determined using 2.00 g of air-dried soil sieved to
pass a 2-mm sieve into a 50-mL centrifuge tube. 20 mL of de-ionized water was added,
and then the tubes were placed on their sides in a reciprocating shaker at low speed for
one hour. The samples were centrifuged at 6000 rpm for 10 min. A Millipore filtration
apparatus and 0.45-im filters were used to filter the supernatant from the tubes into 20-
mL scintillation vials. The soluble reactive phosphorus concentration in the samples was
determined using the automated ascorbic method (Luscombe et al. 1979).
Soil analysis for oxalate extractable iron and aluminum bound phosphorus was
conducted by preparing a soil:solution ratio of 1:40 (0.5 g dry ground soil and 20 mL
oxalate reagent) (McKeague et al. 1966, Sheldrick 1984). The oxalate reagent was
prepared using 350 mL of ammonium oxalate and 267.5 mL of oxalic acid. Samples
were placed in tubes and shaken in the dark for 4 hours. Then the samples were
centrifuged at 6000 rpm for 10 min, and filtered with 0.45-[im filter paper with the lights
off. The pH of the solution was checked (pH should be 3.0). Filtered samples were sent
to the UF/IFAS Analytical Research Laboratory (ARL) for determination of oxalate
extractable Fe/Al bound phosphorus (EPA Method 200.7). The oxalate extraction was
used to determine the Fe and Al bound P, which is the reactive fraction of amorphous Fe-
An in-depth soil characterization was not conducted for this experiment due to the
fact that the experiment was conducted in-situ and there were several factors affecting
phosphorus assimilation in the soils. Other research studies involved with phosphorus
assimilation in soils have conducted chemical fractionation schemes as mentioned earlier.
Phosphorus sorption isotherms are then developed based on the results from this
procedure. These isotherms can be used to determine the maximum phosphorus retention
capacity and uptake rates of soils, but again, the calculation of maximum retention
capacity based on isotherms usually underestimates potential P sorption by soils under
field conditions (Reddy et al. 1995). Phosphorus release rates from soils are probably
slower under field conditions compared to rates observed under laboratory conditions
(Reddy et al. 1996). An attempt was made to estimate maximum phosphorus retention
capacity and EPCw values for the wetland soils in this research based on oxalate
extractable Fe/Al results and previous results from soil cores taken within the Lake
Okeechobee drainage basin.
In one experiment, batch incubation studies were conducted on selected wetlands
soils in the Okeechobee Basin to demonstrate their capacity to retain or release P (Reddy
et al. 1995). The average EPCw (threshold P concentration in water column where P
retention = P release) for wetlands soils in the Okeechobee Basin was found to be 0.42
mg P L-1 (Reddy et al. 1995). Soils with strong binding capacities have low EPC values
(Reddy et al. 1996). Chemical fractionation was used in the study to identify forms of
inorganic and organic P (labile and non-labile pools). Other reported results for S-154
Basin (Pelaez Ranch included in this basin) from Reddy et al. (1995) include the
following: pH = 6.21, TOC (mg g-) = 110, Bulk density (g cm-3) = 0.46.
2.6 Vegetation Analysis
Vegetation type and biomass were identified for each wetland site. Above water
vegetation volume and below water, above soil vegetation volume were determined in
At the end of each experiment, the vegetation from inside the mesocosms was
clipped at the water surface and then again at the ground surface and brought to the lab
for biomass analysis. The division of vegetative growth above the water surface as
compared to below the water surface is often important for data interpretation, especially
if the data can be normalized according to vegetative biomass within each mesocosm.
2.7 Statistical Methods
Water quality data were analyzed using Statistical Analysis Software (SAS, 2003)
version 9.1.0. An analysis of variance (ANOVA) was performed on the data using the
GLM procedure, which uses the method of least squares to fit general linear models. The
Waller-Duncan k-ratio t test was performed on all the main effect means. Significant
differences were determined at the p<0.01 level.
RESULTS AND DISCUSSION
The native vegetation wetland (Site 1) exhibits lower ambient nutrient
concentrations compared to the improved, floralta planted wetland (Site 4), as a result of
differences in cattle stocking densities in the surrounding pastures, soil characteristics,
and vegetative cover. The wetland at Site 4 receives higher nutrient loading on an annual
basis from higher cattle stocking densities on the surrounding pasture compared to Site 1,
and has a limited phosphorus storage capacity due to low Fe/Al content of the soil within
the wetland. The soil at Site 4 is coarse, white sand, characterized by a low phosphorus
storage capacity and high infiltration rates. At Site 1, lower cattle stocking densities are
evident and the storage capacity of the wetland soil is improved by having higher Fe/Al
content. Organic matter buildup at Site 1 is higher than at Site 4 due to vegetative
densities within the wetlands. The maximum phosphorus storage capacity of the soil was
found to be higher at Site 1 compared to Site 4.
Rapidly decreasing water levels during the time frame of this experiment produced
unfavorable conditions for data collection and analysis. In general, water samples were
taken at the midpoint of the water column. As water levels decreased, the location of
sampling moved closer to the sediment-water interface. Phosphorus gradients within the
water column have been identified in lakes, and it is believed that such gradients also
exist in wetlands. Phosphorus concentrations are generally higher near the sediment-
water interface. It is possible that as the water level declined, changes in water column
phosphorus concentrations were compounded due to a naturally existing gradient within
the water column.
3.1 Water Level Results
Initial water depths were approximately 16 cm above the soil surface and declined
rapidly during the course of the experiment. These low water levels were not ideal for
taking water quality samples. Sediments play a more dominant role in the metabolism of
the water column at shallow water depths. Water levels declined slightly more rapidly at
Site 4 compared to Site 1 (Figs. 3-1 and 3-2). The higher organic matter content of the
soil at Site 1 possibly serves as a "sponge", helping to maintain saturated soil conditions
for longer periods of time under decreasing water level conditions. Infiltration through a
thick organic layer is usually limited.
Three to five water depth measurements were taken at each mesocosm. An average
daily value was determined for each wetland to produce Figs. 3-1 and Fig. 3-2. Water
depths inside the mesocosms were influenced by proximity of the mesocosms to the
canals that run through the wetlands. Slightly higher water depths were measured in
mesocosms located closest to the canals. Average daily water depths were also
determined per treatment for some analyses.
Site 1 Native Vegetation Wetland
20.0 Average Water Depth in Mesocosms
^ 16.0 -- v, y = -1.8611x + 71263
R2 = 0.9659
10/21/2004 10/23/2004 10/25/2004 10/27/2004 10/29/2004 10/31/2004
Figure 3-1. Water depths at Site 1.
Site 4 Improved Wetland
Average Water Depth in Mesocosms
120 y = -2.0305x + 77743
R2 = 0.9437
10/20/2004 10/22/2004 10/24/2004 10/26/2004 10/28/2004 10/30/2004
Figure 3-2. Water depths at Site 4.
3.2 Water Quality Data and Statistical Results
Water quality results and statistical analyses are presented in Table 3-1 and Table
3-2. The time, 0.5 days, indicates samples that were taken before treatments were
applied. Results from day 1 indicate samples that were taken immediately following
treatment application. The trends over time, which will be discussed in future sections,
were examined from day 1 to the end of the experiment. The wetland at Site 4 was
completely dry by October 29, 2004; therefore, no water quality samples were taken.
Samples were taken on October 29 at Site 1 wetland, but water levels were near zero.
There was no water inside three of the nine mesocosms located in Site 1 wetland
by day 9 (one control, one dilute, and one spike were dry). Results for SRP, TP, NH4+,
NO3-, and TKN on day 9 were determined from water quality samples taken from the
other six mesocosms that were not dry (two replicates of each treatment analyzed).
One control, two dilutes, and one spike mesocosm were dry by October 25 in Site
Mean concentrations in the same column with the same letter (a, b, or c) in front
indicate values that are not statistically different. For instance, on the first day, before
any treatments were applied (time 0.5), SRP concentrations for all three treatments were
statistically the same. On the second day, after treatment application, the control and
dilute were statistically similar to each other, but different from the spike treatment at
Site 1. At Site 4, SRP concentrations for all three treatments were statistically different
immediately following treatment application. The same occurred for total phosphorus
At both Site 1 and Site 4, ammonium concentrations were only statistically
different immediately following treatment application. Nitrate concentrations remained
statistically similar throughout the time frame of this experiment. No relationships were
determined from the TKN results.
Table 3-1. Statistical analysis of phosphorus data, Pelaez & Sons Ranch.
Soluble Reactive Phosphorus (SRP)
Tie (day) 0.5
1 2 3 5 9
0.5 1 2
Root MSE 0.029 0.59 0.284 0.321 0.287 0.086 0.133 0.303 0.102 0.178 0.219
TRTp-value 0.4293 0.0002 0.0002 0.0008 0.0008 0.001 0.31 0.0001 <0001 <0001 0.015
0.017 0.344 0.164 0.186 0.165 0.061 0.077 0.175 0.059 0.103 0.155
stdev 0.013 1.032 0.49 0.555 0.494 0.014 0.087 0.517 0.102 0.258 0.204
nman a-0.047 a-4.098 a-2.122 a-1.788 a-1.660 a-1.462 a-1.729 a-3.197 a-3.400 a-3.564 a-3.744
stdev 0.045 0.027 0.026 0.021 0.047 0.126 0.119 0.09 0.128 0.146 0.234
nman a-0.068 b-0.082 b-0.095 b-0.068 b-0.120 b-0.188 a-1.868 b-1.849 b-1.822 b-1.912 b-2.090
stdev 0.018 0.006 0.024 0.037 0.017 0.078 0.177 0.033 0.069 0.085 novalue
nman a-0.035 b-0.022 b-0.036 b-0.048 b-0.083 b-0.140 a-1.903 c-0.570 c-0.795 c-0.826 c-0.795
Total Phosphorus (TP)
Site 1 Site 4
Tme(day) 0.5 1 2 3 5 9 0.5 1 2 3 5
RootMSE 0.163 0.403 0.501 0.652 0.425 0.148 0.151 0.228 0.172 0.177 0.346
TRTp-value 0.1634 <0001 0.0006 0.0064 0.0068 0.0024 0.288 <0001 <0001 <0001 0.033
0.094 0.233 0.289 0.377 0.245 0.104 0.087 0.132 0.099 0.102 0.244
stdev 0.142 0.65 0.796 1.084 0.705 0.085 0.144 0.388 0.256 0.199 0.288
nman a-0.298 a-5.149 a-3.463 a-2.768 a-2.227 a-2.266 a-1.922 a-3.540 a-3.877 a-3.792 a-4.041
stdev 0.158 0.243 0.233 0.161 0.079 0.23 0.066 0.061 0.104 0.192 0.288
nman a-0.546 b-0.478 b-0.831 b-0.435 b-0.836 b-0.718 a-2.060 b-2.019 b-2.038 b-2.088 b-2.237
stdev 0.186 0.067 0.253 0.274 0.194 0.071 0.209 0.042 0.11 0.131 novalue
nman a-0.280 b-0.132 b-0.467 b-0.364 b-0.605 b-0.542 a-2.136 c-0.642 c-0.921 c-0.944 b-0.951
*LSMEANs are the same for each treatment.
*All C centrations measured in mg/L
*STDEVaries for each TRT and time step.
*Site 1 Native Vegetatin Wetland, Site 4 Impoved Wetland
Lake Okeechobee has been impacted from phosphorus loading from its
surrounding watershed. Phosphorus trends over time were analyzed in greater detail
because phosphorus is the major nutrient of concern. This in-situ mesocosm experiment
was primarily beneficial in determining phosphorus uptake/release dynamics within the
wetlands. Changes in nitrogen concentrations were not as important in this research.
Table 3-2. Statistical analysis of nitrogen data, Pelaez & Sons Ranch.
Time (day) 0.5
1 2 3 5 9
0.5 1 2
Root MSE 0.016 0.105 0.025 0.092 0.079 0.423 0.049 0.038 0.065 0.066 0.036
TRT p-value 0.381 0.0086 0.28 0.592 0.399 0.718 0.968 <.0001 0.201 0.554 0.582
error) 0.009 0.061 0.014 0.053 0.046 0.299 0.028 0.022 0.037 0.038 0.025
stdev 0.025 0.171 0.021 0.144 0.02 0.217 0.064 0.065 0.104 0.024 0.017
mean a-0.046 a-0.392 a-0.037 a-0.141 a-0.043 a-0.399 a-0.091 a-0.570 a-0.175 a-0.098 a-0.059
stdev 0.005 0.061 0.032 0.059 0.134 0.665 0.053 0.011 0.013 0.041 0.047
mean a-0.026 b-0.058 a-0.057 a-0.112 a-0.126 a-0.585 a-0.082 b-0.041 a-0.122 a-0.160 a-0.092
stdev 0.01 0.012 0.019 0.038 0.022 0.218 0.018 0.008 0.039 0.105 no value
mean a-0.034 b-0.014 a-0.021 a-0.061 a-0.045 a-0.221 a-0.083 b-0.029 a-0.067 a-0.136 a-0.045
Site 1 Site 4
Time (day) 0.5 1 2 3 5 9 0.5 1 2 3 5
Root MSE 0.003 0.018 0.005 0.002 0.002 0.015 0.016 0.019 0.005 0.005 0
TRT p-value 0.037 0.961 0.152 0.079 0.422 0.318 0.682 0.246 0.63 0.593 no value
error) 0.001 0.01 0.003 0.001 0.001 0.01 0.009 0.011 0.003 0.003 0
stdev 0.003 0.019 0.003 0 0 0.023 0.012 0.033 0.006 0.003 0
mean b-0.007 a-0.022 a-0.009 a-0.011 a-0.011 a-0.027 a-0.018 a-0.038 a-0.015 a-0.009 a-0.011
stdev 0 0.019 0.005 0.003 0.003 0.011 0.023 0.003 0.006 0.005 0
mean a-0.011 a-0.022 a-0.016 a-0.007 a-0.009 a-0.019 a-0.027 a-0.013 a-0.015 a-0.006 a-0.011
stdev 0.003 0.016 0.006 0 0 0.004 0.005 0 0 0.005 no value
mean a-0.015 a-0.025 a-0.007 a-0.011 a-0.011 a-0.046 a-0.016 a-0.011 a-0.011 a-0.006 a-0.011
Total Kjeldahl Nitrogen (TKN)
Site 1 Site 4
Time (day) 0.5 1 2 3 5 9 0.5 1 2 3 5
Root MSE 2.928 2.439 3.92 4.079 7.736 14.851 0.723 0.218 8.037 0.426 4.872
TRT p-value 0.862 0.02 0.174 0.347 0.832 0.994 0.394 <.0001 0.349 0.0006 0.546
error) 1.691 1.408 2.263 2.355 4.466 10.501 0.417 0.126 4.64 0.246 no value
4.47 3.66 5.657 6.942 3.567 3.642 0.349 0.347 13.868 0.476 6.867
a-6.969 a-9.92 a-11.729 a-10.396 a-15.918 a-30.56 a-5.217 a-5.826 a-13.404 b-5.008 a-10.568
1.783 1.761 2.68 1.015 8.494 2.221 0.151 0.066 1.204 0.446 0.565
a-6.417 a-7.235 a-8.378 a-6.740 a-17.651 a-29.56 a-5.427 b-5.389 a-6.378 a-5.884 a-5.883
1.602 1.159 2.63 0.828 9.73 25.367 1.193 0.132 0.144 0.344 no value
a-5.655 b-2.057 a-4.760 a-5.236 a-13.767 a-31.19 a-6.055 c-2.133 a-3.218 c-3.161 a-3.542
*LSMEANs are the same for each treatment.
*All Concentrations measured in mg/L
*STDEV varies for each TRT and time step.
*Site 1 Native Vegetation Wetland, Site 4 Improved Wetland
3.3 Phosphorus Results (Uptake/Release)
Assuming that soluble reactive phosphorus concentrations in the soil porewater and
water column were already at equilibrium prior to the experiment, it was expected that
the system would show a net decrease in water column SRP concentration in the spiked
mesocosms over time as a new equilibrium was established in response to the nutrient
load. The SRP concentration in the spiked water column was expected to be greater than
the SRP concentration in the soil porewater, causing a net flux of SRP from the water
column into the sediment through diffusive processes. Results indicated a decrease in
SRP concentration over time at Site 1 as expected, but not at Site 4. Site 1 wetland
showed a typical two-part response to the nutrient loading, a quick, initial SRP uptake
followed by a slower, more drawn out uptake rate (see Fig. 3-3). Site 4 wetland showed
an increase in SRP concentration over time in the spiked mesocosms, possibly due to
evapotranspiration loss and lack of adsorptive sites in the soil in this wetland. As water
levels dropped at Site 4, the concentration of SRP in the mesocosms increased (see
Figure 3-4). Obtaining an equilibrium phosphorus concentration value was difficult for
the time frame of the experiment due to water levels decreasing to zero; however, SRP
concentrations in the spiked mesocosms at Site 1 appear to be decreasing towards a
minimum value or equilibrium.
The dilute treatment was expected to result in a net flux of phosphorus from the soil
into the overlying water column. Site 1 showed an increasing trend in SRP concentration
in the water column over time. Dilute mesocosms at Site 4 showed increasing SRP
concentrations from Oct 21 through Oct 23 followed by leveling off of the SRP
concentration from Oct 23 through Oct 25. The dissolved phosphorus concentration in
the soil porewater tends to be naturally higher than the phosphorus concentration in the
overlying water column in many wetlands, causing a flux from the soil to the water
column (Reddy and D'Angelo 1994). This could explain some of the increases in the
SRP concentration in the water column in the control mesocosms over time.
Changes in soluble reactive phosphorus concentrations were primarily linear for
most treatments, increasing over time. The graphs in Figs. 3-3 and 3-4 demonstrate the
mean daily SRP concentrations as well as standard deviations for each treatment over
Because changes in concentrations over time are affected by the changing water
level, it is useful to look at water column phosphorus mass over time. An increase in the
mass of phosphorus in the water column within a mesocosm would potentially reflect a
release of phosphorus from the soil. The factors influencing water depth within the
mesocosms include groundwater loss and evapotranspiration. As evapotranspiration
leaves the mesocosm, nutrients such as phosphorus remain in the water column causing
an increase in phosphorus concentration; however, the mass of phosphorus within the
water column would remain the same. If only groundwater loss takes place (no
evapotranspiration), the phosphorus concentration of the water column within the
mesocosm would not change because phosphorus and groundwater would be leaving at
the same rate; however the mass of P would decrease. Looking at the mass of P within
the water column can help determine the dominant process for water loss from the
mesocosms and help identify when uptake/release of phosphorus from the soil takes
Site 1 Native Vegetation Wetland Control Mesocosms
(with Standard Deviations)
10/21/04 10/23/04 10/25/04 10/27/04 10/29/04 10/31/04
Site 1 Native Vegetation Wetland Dilute Mesocosms
(with standard deviations)
10/23/04 10/25/04 10/27/04 10/29/04
Site 1 Native Vegetation Wetland Spike Mesocosms
(with standard deviations)
10/23/04 10/25/04 10/27/04 10/29/04 10/31/04
Figure 3-3. Mean SRP and standard deviations over time for Site 1.
y = 0.0141x 539.72
R = 0.8951
y = 0.015x 573.77
R2 = 0.9964
Site 4 Floralta Wetland Control Mesocosms
(with standard deviations)
y = 0.0569x 2177.3
R = 0.8067
10/22/04 10/23/04 10/24/04 10/25/04 10/26/04
Site 4 Floralta Wetland Dilute Mesocosms
(with standard deviations)
10/22/04 10/23/04 10/24/04 10/25/04 10/26/04
Site 4 Floralta Wetland Spike Mesocosms
(with standant deviations)
Figure 3-4. Mean SRP and standard deviations over time for Site 4.
y = 0.1344x 5142
R2 = 0.9596
10/22/04 10/23/04 10/24/04 10/25/04
(,"ith standard deviations)
The control treatments at Site 4 showed a linearly decreasing mass of P with
decreasing water depth, as expected (Fig. 3-5). Significant uptake/release was not
expected in the control treatments because it was assumed that the water column and soil
porewater were already at equilibrium. Phosphorus moved out of the mesocosm with
groundwater loss over time and was possibly adsorbed to soil particles. The dilute
treatments at Site 4 showed an initial release of phosphorus due to diffusion of P from a
higher concentration within the soil porewater moving to a lower P concentration within
the water column (Fig. 3-5). After this initial release, groundwater loss was the dominant
factor influencing the mass of P within the mesocosms. The spike treatments at Site 4
demonstrate some release of P from the sediment may have occurred because the mass
did not decrease as expected from 10/22 to 10/23 but rather stayed relatively constant
(Fig. 3-5). As P mass was lost through groundwater, it was initially replaced by P release
from the sediment. This release of P from the sediment in the spike treatment was not
expected, but is supported by soil data indicating low P storage capacity of this improved,
non-native wetland. Phosphorus mass decreased with time due to groundwater loss.
Site 4 Floralta Wetland Control Mesocosms
80 y = -12.491x + 478274 14 "
S2 = 0.9732 12 "
Sy = -2.6162x + 100170 10 g.
40 R 0663 8
10/21/04 10/22/04 10/23/04 10/24/04 10/25/04 10/26/04
Mass P m Water Depth
Linear (Mass P) Linear (Water Depth)
Figure 3-5. Mass versus water depth over time for Site 4.
Site 4 Floralta Wetland Dilute Mesocosms
y =-3.0087x + 11519^^^^ ^ 8.0 Q
R2 = 0.8935 6.0
21/04 10/22/04 10 23 4 10 24 4 1 25 114 10/26/04
+ Mass P m Water Depth Linear (Water Depth)
e tiS 4 Floralta Wetland Spike Mesocosms
S t 16.0
y = -2.6541x + 101619 8.0
R2 = 0.9538 6.0
10/22/04 10/23/04 1024 04 1025 04 10/26/04
Mass P Water Depth Linear (Water Depth)
Figure 3-5. Continued.
The control treatment at Site 1 shows an initial rapid decrease in P mass followed
by a slightly increasing P mass within the mesocosms (Fig. 3-6). The initial rapid drop
could be due to either P uptake possibly by the soil, algae, or macrophytes or could be the
result of settling following disturbance from installation of the mesocosm. Higher levels
of Fe/Al content of the soil support adsorption on sediment at this site. ET contributed to
some increase in P concentration in the control mesocosms. Convective circulation could
have caused mixing of sediments within the water column leading to an increase in P
mass within the mesocosms. The dilute treatments showed a release of P into the water
column at Site 1 (Fig. 3-6). Adsorption of P occurred in the spike treatments at Site 1 due
to diffusion of P from the higher concentration in the water column to lower
concentration in the soil porewater.
Site 1 Native Vegetation Wetland Control Mesocosms
y = -1.8389x + 70414
R2 = 0.9036
21/04 10/23/04 10/25/04 10/27/04 10/29/04 10/31
Mass P m Water Depth Linear (Water Depth)
Site 1 Native Vegetation Wetland Dilute Mesocosms
y = -2.0275x + 77636
R2 = 0.9935
10/23/04 10/25/04 10/27/04 10/29/04
-- Mass P m Water Depth -- Linear (Water Depth)
Figure 3-6. Mass versus water depth over time for Site 1.
Site 1 Native Vegetation Wetland Spike Mesocosms
1 y = -1.637x + 62685 12
S80 = 0.9133 0
10/21/04 10/23/04 10/25/04 10/27/04 10/29/04 10/31/04
Mass P Water Depth Linear (Water Depth)
Figure 3-6. Continued.
3.4 Soil Results
Soil cores were taken within each mesocosm at the end of the experiment at
depths of 0 to 10 cm below the ground surface. The coring device allowed surface water
to be removed as well as coarse litter on the soil surface. The soil results in Table 3-3
were determined from a composite soil sample from each wetland because soil results
were not expected to be significantly different among treatments.
Table 3-3. Soil results (0-10 cm depth) for Site 1 and Site 4, Pelaez & Sons Ranch.
Water Water Soil Bulk Oxalate Extractions
Content Extractable TP Density P Fe Al
Soil Data pH bywt.(%) P (mg/kg) (mg/kg) (g/cm") (mg/kg) (mg/kg) (mg/kg)
Site 1 Wetland (NV) 5.07 78 1.690 644.72 0.258 149.68 4,296 2,772
Site 4 Wetland (IMP) 6.20 24 6.030 37.44 1.401 8.316 36.96 47.04
NV = Native Vegetation Wetland, IMP = Improved Wetland
The improved wetland (Site 4) showed higher concentrations of water extractable P
compared to the native vegetation wetland at Site 1. Water-soluble P has been used to
indicate the potential for P to be released from manure-impacted soils (Nair et al. 1999).
This would indicate that there is a higher potential for P release from Site 4.
Concentrations of water soluble phosphorus and oxalate extractable P have been found to
decrease with depth in the soil profile (Nair et al. 2004). Soil total phosphorus
concentrations were higher at the native vegetation wetland compared to the improved
wetland. Soil total phosphorus concentrations in the A horizon of manure-impacted
spodosols in Nair and Graetz (2002) were in the range of 25 mg P/kg to 2300 mg P/kg.
The study by Nair and Graetz (2002) was conducted on dairy farms; however, total
phosphorus concentrations in the soils presented here fall within that range.
3.5 Vegetation Results
This experiment took place late in the growing season; therefore, vegetation did
not play a significant role in nutrient uptake within the mesocosms. A limited amount of
vegetation was observed inside each mesocom (Table 3-4) due to limited vegetative
growth in desired water depths. Nutrient results were not normalized according to
vegetative biomass within each mesocosm for this reason.
Emergent macrophytes such as Polygonum sp. and Panicum sp. exhibited lower
storage of phosphorus than other macrophytes such as Hydrocotyle sp. in one study by
Reddy et al. 1996.
There was higher vegetative growth in the spike mesocosms at Site 1 compared to
Site 4. Epiphytic algal growth would have contributed to rapid phosphorus uptake within
the mesocosms. The initial rapid SRP uptake in spiked treatment mesocosms at Site 1 is
potentially explained by higher algae growth within the mesocosms due to higher
Table 3-4. Vegetation results for Site 1 and Site 4, Pelaez & Sons Ranch.
(dbow sil/bowwate) Wter/Abo Ab )eW ter
Site 1 VegAtatimType (nL) Soil BHam s (g) mass (g)
Ctoil- 1 Pdygiznh J #dtippeq i s 40 4.899 1.449
Ctoil -2 Pdjygzmunhtdtnpetis 35 4.500 1.253
ratod-3 Pcnactmh~ n toma 8 1.169 0.289
Iihte- 1 PdygiJunth#dppets 50 6.894 2123
Ilkte- 2 PdjygznzmJth d petds 45 5.918 0.454
Ilkte- 3 PdjygiJMn tdppeks 100 14.936 2245
Sike- 1 Dilr#nii-Rttitkin 55 11.589 3.059
Spke-2 Pcnarcmhitc n 40 5.273 0.73
Spike-3 Pcnacnmh~ntomam 50 8.910 32539
3.6 Bromide Results
Bromide data were used to estimate the amount of ET occurring within the
mesocosms. According to data collected at the Florida ONA weather station, the
average ET during the time frame of this experiment (Oct 21-29, 2004) was
approximately 0.25 cm/day. ET was calculated using the Penman calculation method.
Data were also collected at Buck Island Ranch in Highlands County, FL in 2003. These
data indicated an ET rate of 0.31 cm/day. An average daily ET rate was calculated for
each wetland on Pelaez & Sons beef cattle ranch using bromide tracer data. The average
daily ET obtained for Site 1 was 0.31 cm/day, and the average ET for Site 4 was 0.36
cm/day. Differences in ET rates for the two wetlands were possibly due to vegetative
cover, root-zone water levels, and soil properties (Hunt and Hunt 2001).
Bromide results from all three treatments at Site 4 and the dilute treatment at Site 1
indicated that there was poor mixing of the bromide solution within the water column in
the first sample taken for analysis. Bromide levels were above the desired target level on
the first sampling date and dropped to the target level of approximately 5-6 mg/L by the
second sampling date. From the second sampling date onward, the bromide
concentration, in general, showed an increasing trend due to ET loss (Fig. 3-7).
Increases in SRP concentrations in all three treatments at Site 4 from the second
sampling date until the end of the experiment were potentially explained by ET loss.
Minor uptake/release from the sediment occurred; however, the changes in SRP
concentration were predominantly explained by the bromide results and ET loss at this
The diagram in Figure 3-7 illustrates initial poor mixing of the bromide solution
within the water column followed by increasing bromide concentration due to ET loss.
_11.0 Bromide Concentrations at Site 1 Wetland
SPoor Mixing of Bromide Solution
Within water column
SIncreasing Br Cone. Due to ET Loss
Target Bromide Concentration
10/20/2004 10/22/2004 10/24/2004 10/26/2004 10/28/2004 10/30/2004
--Control *- Dilute -- Spike
Figure 3-7. Illustration of poor mixing and ET effects using bromide data.
Changes in SRP mass over time within the mesocosms were predicted using a
simple water balance approach. The water volume inside the mesocosms decreased
linearly each day as a result of evapotranspiration and groundwater loss. No rainfall
occurred during the time frame of this experiment. Surface flow was restricted by the
mesocosms, and groundwater flow into the mesocosms was not likely, as water levels
were dropping rapidly. The initial water depth was determined from daily water depth
measurements taken in the field. A decreasing, linear relationship was determined for
water depth over time in both wetlands. An average initial water volume inside the
mesocosms was then determined by multiplying the initial water depth by the average
surface area. An average initial SRP concentration was determined for each treatment
from water quality results and was then used to estimate the mass of SRP within the
mesocosms. For calculation purposes, water loss through ET was calculated first,
followed by water loss through groundwater. As ET occurred within the mesocosms,
SRP concentrations increased, and the mass of SRP remained constant. SRP is removed
at the same rate as groundwater; therefore, ET is the only factor affecting the increase in
Differences between the predicted SRP mass over time and the measured SRP
mass over time should indicate uptake or release that may be occurring within the system.
Site 1 Control
The bromide concentrations within the control mesocosms increased linearly from
day 1 to day 9 as a result of ET loss. ET rates were determined as the amount of water
that had to be extracted from the mesocosm in order to produce the measured increase in
bromide concentration. Comparing the measured SRP mass to the predicted SRP mass
due to ET and groundwater loss proved useful in determining the actual rates of retention
and release of SRP within the mesocosms.
Figure 3-8 shows that the mass of SRP in the control mesocosms decreased as
predicted due to ET and groundwater losses. SRP concentrations increased linearly with
time, as discussed previously, due to ET loss. The dominant factor influencing the loss of
SRP mass over time was groundwater loss and almost no uptake/release of SRP occurred
within the control treatment mesocosms at Site 1. Minor release of SRP possibly
occurred due to diffusive gradients between the soil porewater and the overlying water
Site 1 Native Wetland Control Treatment
Predicted vs. Measured SRP Mass
10/20/2004 10/22/2004 10/24/2004 10/26/2004 10/28/2004 10/30/2004
-*- Predicted from ET and G Loss Measured Value
Site 1 Native Wetland Dilute Treatment
Predicted vs Measured SRP Mass
10/20/2004 10/22/2004 10/24/2004 10/26/2004 10/28/2004 10/30/2004
Predicted fromET and G Loss --Measured Value
Figure 3-8. Effects of ET and groundwater loss in control and dilute mesocosms at Site 1.
Site 1 Dilute
Bromide concentrations increased linearly in the dilute mesocosms, similar to the
control mesocosms, as a result of ET loss. The rate of increase in SRP concentrations
was slightly higher for the dilute treatments compared to the control treatments because
of SRP release from the sediment, whereas the control treatments were mainly affected
by ET loss.
The measured SRP masses in the dilute treatment mesocosms at Site 1 were not
predicted by ET and groundwater loss calculations (Fig. 3-8). The mass of SRP
increased from day 1 to day 5 as a result of release of SRP from a source such as the
sediment. The mass of SRP in the water column should have decreased due to
groundwater loss; however, as release of SRP occurred, the mass in the water column
increased. A total average release of 0.084 mg SRP/L occurred for the dilute treatment
mesocosms. This release was expected because the SRP concentration of the water
column was lowered by 50% on day 1 causing a shift in equilibrium between the soil
porewater and the overlying water column. Release slowed or stopped by day 5 and
groundwater loss controlled the loss of SRP mass.
Site 1 Spike
Bromide concentrations in the spike treatment mesocosms at Site 1 increased,
similar to the control and dilute treatment mesocosms. The SRP concentration dropped
rapidly, as discussed previously, and was followed by a slower decrease with time. The
mass of SRP was not predicted by ET and groundwater calculations (Fig. 3-9). The
measured mass of P was lower than the predicted mass, indicating uptake of SRP in the
The decrease in measured SRP mass in the mesocosm is not believed to be the
result of dilution (leakage) with outside water (which had a lower SRP concentration)
because an increase was seen in the bromide concentration. The bromide concentration
would have decreased as well as the SRP concentration had dilution been the major factor
affecting concentration changes in the mesocosms. Total uptake in the spike mesocosms
averaged 0.735 mg SRP/ L.
Site 1 Native Wetland- Spike Treatment
Predicted vs. Measured SRP Mass
10/20/2004 10/22/2004 10/24/2004 10/26/2004 10/28/2004 10/30/2004
--- Predicted from ET and G Loss Measured Value
Figure 3-9. Effects of ET and groundwater loss in spike mesocosms at Site 1.
The desired SRP concentration for the spike mesocosms was 3.0 mg/L. However,
measured concentrations indicated an initial SRP concentration of 4.1 mg/L (See Fig. 3-
3). This is possibly the result of poor mixing of the spike solution within the water
column, which was also evident in the bromide results.
The rate of uptake is higher than rate of release at Site 1. The rate of adsorption is
generally faster than the rate of desorption according to numerous soil phosphorus
models such as the Langmuir-Hinshelwood model (Overman et al. 1999).
Site 4 Control
Bromide concentrations in the control treatment mesocosms increased non-linearly
at Site 4 as a result of ET loss and SRP concentrations increased linearly with time.
Changes in SRP mass were predicted very well by ET and groundwater loss calculations
indicating that no uptake or release of SRP occurred in the control mesocosms. Figure 3-
10 illustrates the effect of ET and groundwater loss on SRP mass in the control
treatments at Site 4.
Site 4 Improved Wetland Control Treatment
Predicted vs. Measured SRP Mass
10/20/2004 10/21/2004 10/22/2004 10/23/2004 10/24/2004 10/25/2004 10/26/2004
-*- Predicted from ET and G Loss -- Measured Value
Figure 3-10. Effects of ET and groundwater loss for control mesocosms at Site 4.
A very minor dilution of 0.142 L possibly occurred from 10/21 to 10/22 in the
control mesocosms causing both the bromide concentration and the SRP concentration to
decrease slightly (Table 3-5).
Table 3-5. Illustration of possible leakage underneath mesocosm walls in Site 4 control
SRP Conc. (mg/L) Bromide Conc. (mg/L) Water Volume (L)
10/21/04 1.868 6.858 49.00
10/22/04 1.849 6.840 46.88
Difference 0.019 0.018 2.12
Site 4 Dilute
Bromide concentrations in the dilute treatment mesocosms at Site 4 increased from
day 2 to day 5 due to ET loss. There was some indication of poor mixing of the bromide
solution within the water column in the dilute treatments as indicated by a dramatic drop
in bromide concentration from 10/21 to 10/22. SRP concentrations increased from day 1
to approximately day 3 and then leveled off as a result of ET loss and SRP release from
the sediment. The measured mass of SRP was slightly above the predicted mass due to
release of SRP from the sediment, as expected in the dilute treatments (Fig. 3-11) due to
Site 4 Improved Wetland Dilute Treatment
Predicted vs. Measured SRP Mass
10/20/2004 1(I 21 211114 1(I 22 21114 1(I 23 21114 10/24/2004 1(I 25 2'1114 10/26/2004
-- Predicted from ET and G Loss -- Measured Value
Figure 3-11. Effects of ET and groundwater loss for dilute mesocosms at Site 4.
SRP release slowed or stopped by day 2 of the experiment and began to follow the
trend predicted by groundwater and ET loss from day 2 to day 5. The soil at Site 4 has a
poor phosphorus retention capacity and the release of SRP from the sediment was
possibly caused by naturally occurring gradients between the soil porewater and the
overlying water column in wetlands.
Site 4 Spike
Bromide concentrations in the spike treatment mesocosms increased non-linearly
from day 1 to day 5 due to ET loss. SRP concentrations increased linearly, as discussed
previously, as a result of ET loss. The increases in SRP concentration in the spiked
mesocosms also resulted from the low phosphorus sorption capacity of the soil in this
wetland. Phosphorus can be released from the sediment at this site even at very low
concentrations. The rate of loss of SRP mass from the spiked mesocosms was predicted
accurately by groundwater and ET loss calculations indicating little to no SRP uptake
(Fig. 3-12), compared to Site 1 spike mesocosms which did show SRP uptake.
Site 4 Improved Wetland Spike Treatment
Predicted vs. Measured SRP Mass
S 150 -----------------------
10/20/2004 10/21/2004 10/22/2004 10/23/2004 10/24/2004 10/25/2004 10/26/2004
Predicted fromET and G Loss --Measured Value
Figure 3-12. Effects of ET and groundwater loss on spike mesocosms at Site 4.
Slight fluctuations of uptake/release of SRP between the sediment and water
column occurred in these mesocosms as a result of natural processes in which
equilibrium will eventually become established.
Total average release for the spike mesocosms at Site 4 was 0.247 mg SRP/L.
Total average uptake for the spike mesocosms at Site 4 was 0.306 mg SRP/L. The
resulting difference is an uptake of 0.059 mg SRP/L for this improved wetland site.
The desired SRP concentration for the spike mesocosms was 3.2 mg/L. This
concentration was achieved successfully within the spike treatment mesocosms at Site 4.
3.7 Determination of EPCw, Smax, DPSox
A typical graph used to determine phosphorus (P) retention and release for soil
core studies conducted in isolated wetlands is shown in Figure 3-13.
Water Column P concentration
Figure 3-13. Schematic of graph used to determine phosphorus retention and release for
Water column P concentration (mg/L), cumulative P flux (release/retention) = mg
m-2, (Pk) represents initial desorbed P per unit surface area of soils (mg P m-2), slope (Ka)
is the P assimilation or soil buffering capacity coefficient (L m-2), and equilibrium P
concentration (EPCw) is the water column concentration where retention equals release
An attempt was made to determine EPCw values for the two isolated wetlands on
Pelaez & Sons ranch using this approach. An EPCw value of approximately 1.6 was
determined for the wetland at Site 1, as shown in Fig. 3-14, using only the results from
the spike treatment mesocosms. This is supported by SRP concentrations in the spiked
mesocosms at Site 1 decreasing to a value of around 1.6 mg/L. However, the data do not
indicate a release of phosphorus from the soil at concentrations below the EPCw value,
possibly due to phosphorus being tightly bound by the soil due to high iron and
aluminum content of the soil. Using data from all three treatments gives a much lower
EPCw value (approximately 0.35); however, if the EPCw value were this low, then SRP
concentrations in the spiked mesocosms would have continued to decrease to 0.35 rather
than 1.6 mg/L.
Site 1 Native Vegetation Wetland All Treatments
350EPCw = X Intercept = 1.6 (0.35 all data)
300 y= 133.88x 219.31
3 250 R2 = 0.9984
a y = 64.253x 22.355
SR2 = 0.7549
4.5 4 3.5 3 2.5 2 1.5 1 0.5 0
Water Column P Concentration
Figure 3-14. EPCw value for Site 1, Pelaez & Sons Ranch.
Data from Site 4 were inconclusive and did not provide useful information for
determining an EPCw value for this wetland, as shown in Fig. 3-15. Possible errors in
the data could have been caused by poor mixing of treatments within the water column
within the mesocosms. The low iron/aluminum content of the soil at this site indicates
that release could occur at very low phosphorus concentrations because there are no
strong binding sites for phosphorus in the soil matrix.
Site 4 Improved Wetland
EPC cannot be determined
0 .0 -,,,,,,
s y = 1.6669x 9.8696
1 R2 = 0.0167
4 3.5 3 2.5 2 1.5 1 0.5 0
Water Column P Concentration
Figure 3-15. EPCw graph for Site 4, Pelaez & Sons Ranch.
Reddy et al. (1995) used the following relationship to determine the maximum
sorption capacity of soil in the Lake Okeechobee drainage basin: Smax = 1.74 + 0.172
[Fe + Al], r2=0.78, n=285. Smax was calculated for the two isolated wetlands on Pelaez
& Sons Ranch using this relationship and the Fe/Al soil data collected from the wetlands.
Smax was calculated as 2.15 mmoles P/kg at Site 4 and 32.64 mmoles P/kg at Site 1.
The oxalate extractable Fe/Al bound P was found to be 0.27 and 4.83 mmoles P/kg at Site
4 and Site 1 respectively. This method has been primarily applied to uplands in past
experiments and is usually dependent on agitation (shaking) of a soil sample within a
water column. These relationships were used here for illustrative purposes.
Nair and Graetz (2002) suggested that management recommendations for reducing
P loss through subsurface drainage should consider the degree of phosphorus saturation
(DPS) and water-soluble P. The DPS relates a measure of P already adsorbed by soil to
its P adsorption capacity. It has been indicated as a good measure of a soil's P release
capacity (Nair et al. 2004). DPS values can be used to predict P loss from a soil
irrespective of the depth of the soil within a profile. The following relationship was
described in Nair et al. (2004): DPSox= [(Ox-P)/a(Ox-Fe + Ox-Al)] *100, where alpha is
an empirical factor that compares different soils with respect to P saturation and
concentrations are in mmoles/kg. A value of 0.55 was developed for spodosols in the
Lake Okeechobee drainage basin in Nair and Graetz (2002). This relationship was used
to calculate the DPS for the wetlands in this research.
DPS (Site 1) = [(4.83)/(0.55*(76.92+102.74))] *100 = 4.89%
DPS (Site 4) = [(0.27)/(0.55*(0.66+1.74))]*100 = 20.3%
Low DPS values do not necessarily indicate that a soil is suitable for further P
loading (Nair and Graetz 2002). Consideration of the capacity of a soil to retain P should
be made for management purposes. The sorption capacity of a soil is influenced by soil
properties such as clay content and Al and Fe concentrations (Nair and Graetz 2002). A
change point or value above which there is a rapid increase in water-soluble P and
therefore a likelihood of a negative impact of P in the soil on water quality was identified
in Nair et al. (2004). The change point of DPSox is approximately 20%. This would
mean that the wetland at Site 4 has a higher potential to negatively impact water quality
compared to Site 1. This is also supported by the higher concentrations of water-soluble
P at Site 4 compared to Site 1.
3.8 Langmuir-Hinshelwood Model of Phosphorus
A mathematical model is needed to describe the dynamics of phosphorus uptake in
isolated wetland environments. Water quality results from this in-situ wetland mesocosm
approach indicated a rapid initial uptake of water column P concentration (for nutrient
enriched treatments) followed by a gradual decline at the native vegetation wetland at
Site 1. Similar trends for phosphorus were described in a batch reactor study using the
Langmuir-Hinshelwood Model (Overman and Scholtz 1999). The Langmuir-
Hinshelwood model was used in batch reactor studies to relate soil phosphorus chemistry
to the soil/solution ratio and to initial P in the reactor. Although phosphorus uptake in
wetlands is much more complicated than phosphorus uptake in a controlled batch reactor
study, the trend for phosphorus removal from the water column in-situ could be described
using the Langmuir-Hinshelwood model. The Langmuir-Hinshelwood model for
phosphorus can be described by the following scheme (Overman and Scholtz 1999):
P + S -- A F + S
where P = concentration of solution P, mg/L; S = concentration of available sites
for P adsorption, mg/L, A = concentration of adsorbed P, mg/L; F = concentration of
fixed P, mg/L; ka = rate coefficient for adsorption, L mg-1 h-1; kd = rate coefficient for
desorption, h-1; and kr = rate coefficient for reaction, h-1.
This model is separated into two components, the reversible adsorption of P from
solution to the colloidal surface (Langmuir component) and the irreversible reaction of
the surface species (Hinshelwood component). The rate of adsorption of P to soil
particles is dependent on the number of sites available at which adsorption can take place,
the initial concentration of P, and the equilibrium process of adsorption.
Differences between batch reactor studies and the study conducted in this research
include the following:
1. Various environmental factors were allowed to influence the rate of P adsorption and
uptake in the wetland opposed to only soil adsorption occurring in batch reactor
2. Mixing of water column P concentration and soil porewater occurred naturally,
through groundwater loss, diffusion, or convective circulation rather than applied
Using the Langmuir-Hinshelwood model to describe results from this in-situ
mesocosm approach requires making the following assumptions:
1. Phosphorus uptake is primarily influenced by soil uptake/release in the two wetland
systems addressed in this research rather than other available uptake sites such as
plants, algae, micro invertebrates, etc.
2. Mixing between solution P concentration and the soil porewater occurred within the
mesocosms due to convectional circulation, diffusion or rapid groundwater loss (as
opposed to direct agitation as in the batch reactor study).
The Langmuir-Hinshelwood model adequately describes results from a batch
reactor study because the process of P adsorption is driven by kinetics and complete
mixing of the system; however, in low flow velocities, as seen in wetland environments,
transfer of P from solution to particle surface is more often driven by diffusion. The
interaction between solution P concentration and available sites for adsorption is limited
in wetland environments. Although the Langmuir-Hinshelwood model describes the
kinetics of phosphorus, it fails to identify the mechanisms of soil P chemistry.
Figure 3-16 shows measured SRP values from the in-situ mesocosm approach (Site
1 Spiked Mesocosms) compared to predicted SRP concentrations using the Langmuir-
Hinshelwood model. Initial SRP concentration used in the model was taken from
measured data from Site 1 wetland. The value for Smax was determined to be 1,300
mg/L for Myakka soil under water saturated conditions in Overman and Scholtz (1999).
Values for ka, kd, and kr were determined using a transient analysis and visual
Measured vs Predicted SRP
(Langmuir Hinshelwood Model)
0 24 48 72 96
ka=.0004,kd=.55,k1.01 --- Measured SRP
Figure 3-16. Predicted vs. measured SRP conc.using Langmuir-Hinshelwood model.
In the batch reactor study analyzed by Overman and Scholtz (1999), values for ka,
kd, and kr were determined to be 0.05 L mg-1 h-1, 0.25 h-1, and 0.005 h-1 respectively with
an Smax value of 47 mg/L for the Bh horizon of Myakka soil (unsaturated). The rate
coefficient, ka, is lower for the wetlands compared to the batch reactor study because in
the batch reactor study, the soil is completely mixed within a phosphorus solution;
therefore, there is a greater chance that all of the adsorption sites in the soil will come in
contact with phosphorus in the solution. In the wetland soil, only the top layer of the soil
will be in direct contact with the phosphorus in the water column above, and diffusion
must occur before phosphorus can be adsorbed to sorption sites lower in the soil profile.
The objective of this research was not to identify the individual processes affecting
phosphorus chemistry in wetlands but rather to obtain general information about
uptake/release rates for each wetland system as a whole. The in-situ mesocosm approach
has the potential to provide valuable information for assessing the capability of wetlands
to improve water quality in agricultural settings. This method also shows potential for
improving modeling efforts of wetlands.
The spike and dilute treatments did cause a shift in equilibrium phosphorus
concentrations between the overlying water column and the soil porewater; however,
SRP concentrations did not converge at an equilibrium concentration during the time
frame of this study. Phosphorus uptake occurred as expected in the spike mesocosms at
Site 1, but not at Site 4. Phosphorus release from soil porewater into the overlying water
column occurred at Site 1 as expected, but it was difficult to determine if release occurred
at Site 4 because the data were explained almost entirely by ET effects.
Soluble reactive phosphorus was analyzed in greater detail than other water quality
parameters because phosphorus is the nutrient of concern in the Lake Okeechobee
drainage basin. This parameter is believed to provide the most insight into uptake/release
capabilities of the wetlands on the ranch. Phosphorus is often the primary target element
of nutrient removal systems because of its critical role in limitation of algal growth in
lakes (Elder et al. 1997). Changes in ammonium and nitrate within the mesocosms over
time were not significant. No significant differences in TKN concentrations were
observed between sampling dates.
Isolated wetlands do not receive hydrologic inputs from streams or other direct
supplies; therefore, they rely on rainwater, surface runoff, or groundwater to supply
hydrologic conditions conducive to a wetland environment. The wetlands studied in this
experiment are constantly exposed to highly variable hydrologic conditions. The
wetlands were more often exposed as compared to flooded; however, wetland vegetation
was prominent in the areas, especially at Site 1. Because the experiment had to be
conducted while the wetlands were flooded rather than exposed, the experiment was
limited due to the short flooding duration of the wetlands. The experiment took place in
the middle of a very active 2004 hurricane season; however, water control techniques had
not been implemented on the ranch at the time of the experiment. This research was
conducted in the pre-BMP phase of the overall project, and water control techniques were
not to be implemented until 2005. Flood control practices were implemented near Lake
Okeechobee to prevent severe flood damage to farms in the drainage basin from the
hurricanes. The floodgates near the lake were opened to allow rapid drainage of the
watershed. This, along with no rainfall in the week of the experiment, possibly
contributed to the rapid water level decline in the wetlands during the course of the
experiment, which took place in October 2004.
The mesocosm structure itself did not significantly affect the results of the
experiment. The Kalwall fiberglass material used was effective for the purposes of this
Because this experiment was conducted near the end of the growing season,
vegetation did not contribute significantly to nutrient uptake within the wetlands.
Minimal amounts of vegetation were incorporated inside each mesocosm. This was due
in part to the fact that areas with deep enough water for the experiment did not contain
much vegetation, possibly due to cattle or higher soil elevations around root systems.
Therefore, vegetation data were not used to normalize nutrient concentrations within the
When considering these isolated wetlands as a means to improve water quality in
the Okeechobee drainage basin, it is important to keep in mind that some studies have
indicated a net release of nutrients from sediments that had been exposed to air and were
subsequently rewetted (Baldwin 1996). This is an important consideration for these
wetlands because they are so often exposed, instead of flooded. After a long period of
exposure, these wetlands have the potential to contribute high phosphorus loads upon
initial rewetting, especially at Site 4 where phosphorus is not held tightly in the soil
Results from this study indicate that a lower residence time may be required to
improve water quality leaving the ranch. Treatment in the spike mesocosms at Site 1
occurred within the first 48 hours. Treatment continued after 48 hours but at a much
The two wetlands on Pelaez & Sons beef cattle ranch are quite different. The
wetland at Site 4 has undergone numerous improvement techniques and is less natural
than the wetland at Site 1. Site 4 wetland is drained more often, has a higher pH, is
planted with Floralta limpograss rather than native vegetative cover, and generally
supports more cattle on an annual basis than Site 1 wetland. Differences in initial
ambient conditions could be attributed to these factors. Differences in oxalate extractable
Fe/Al bound P play a significant role on phosphorus adsorption by the soils in each
wetland. Site 1 exhibited a higher phosphorus adsorption capacity compared to Site 4,
and is most likely the result of Fe/Al content of the soils. Site 4 contained coarse, sandy,
white soils compared to darker, mineral soils at Site 1. The higher level of water
extractable P found in the soil at Site 4 compared to Site 1 was likely the result of higher
nutrient loads in runoff from the surrounding pastures.
When taking water quality samples to assess the nutrient removal potential of these
wetlands, it is important to keep in mind the overall results, on an annual basis, rather
than a single measurement.
SUGGESTIONS FOR FUTURE RESEARCH
Installing the mesocosms by hand within the wetlands proved to be an arduous,
time-consuming task. Most of the time was spent cutting through vegetative roots that
would have prevented an adequate seal from forming between the bottom of the
mesocosm wall and the sediment. The mesocosms were constructed of a fairly flexible
Kalwall fiberglass material. Pressure had to be applied gently to the top of the mesocosm
as the roots were cut to avoid buckling of the mesocosm walls. To improve the strength
of the material during installation, a metal collar could have been designed to fit around
the bottom of the mesocosm with a sharpened edge that would aid in cutting through the
roots. Potentially, the collar could be designed so that it could be easily detached and re-
attached between experiments. This would reduce installation time between experiments
if multiple tests were to be performed, reduce disturbance to the wetland, and allow
recovery of the system between experiments once the mesocosm was removed from the
Another issue involved the amount of time required for taking water samples.
Water samples were taken individually by hand, meaning that only one sample could be
taken from each mesocosm per day. Multiple people would have been required to obtain
more than one sample per day from all of the mesocosms. Setting up an automatic
sampler with multiple sampling lines could have reduced water sampling labor and
allowed more time to collect information such as dissolved oxygen levels, temperature,
and pH. A platform would need to be constructed to house the sampling device, and the
samples could be collected from the sampler at regularly scheduled times. An automatic
sampler would increase the accuracy of the sampling technique and potentially reduce
error. Samples are more easily filtered and acidified within a laboratory environment as
compared to in the field. Walking in the wetland from mesocosm to mesocosm to take
water samples each day caused degradation of the wetland environment around the
mesocosms. This would have had an effect on future experiments conducted in the same
In order to avoid water level problems, a minimum water depth of 15 cm should be
maintained during the course of the experiment when possible. Using this minimum
water depth should hopefully prevent sediment control over the water column and
desiccation problems. Installing water level recorders in the wetlands with real-time data
applications would have been useful in planning times to conduct the experiment to avoid
water level issues.
Tracking temperature, dissolved oxygen levels, and pH in the water column
throughout the experiment would have been useful in characterizing the system and
potentially explaining some of the results. Numerous models dealing with phosphorus
sorption require a maximum sorption value, which is often dependent on dissolved
oxygen levels. Also, dissolved oxygen, temperature, and pH have direct effects on
nutrient dynamics within wetlands.
TRIAL RUN RESULTS
Table A-i. Trial run results from the Stormwater Ecological Enhancement Project
(SEEP) site, University of Florida campus.
FIRST SAMPLE DATE SECOND SAMPLE DATE THIRD SAMPLE DATE
7/15/2004 7/16/2004 7/18/2004
*Water Depth was 8.25" *Water Depth was 24"
SRP mg/1 SRP mg/1
Ambient AB1 0.0583 Ambient AB11 0.0383
C1 AB2 0.0615 C1 AB12 0.1284
C2 AB3 0.0324 C2 AB13 0.0469 MESOCOSMS FLOODED
C3 AB4 0.0673 C3 AB14 0.0698
NO DATA AVAILABLE
D1 AB5 0.0364 D1 AB15 0.0720
D2 AB6 0.0475 D2 AB16 0.0641
D3 AB7 0.0201 D3 AB17 0.0264
S1 AB8 3.1040 S1 AB18 1.7334
S2 AB9 7.2234 S2 AB19 2.3195
S3 AB10 1.0287 S3 AB20 1.6091
*C = Control, D = 50% diluted with H20, S = Spiked with 100mg/1 P solution.
Table A-2. Water quality results from trial run at Pelaez & Sons Ranch, Okeechobee
Water TKN NH4 SRP N03
Sample Depth Cone Cone Conc TP Cone
Date Site # Type (m) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L)
09/18/04 1 Ambient 0.28 3.910 0.144 0.143 0.304 0.017
09/19/04 1 Ambient 0.21 6.254 0.160 0.124 0.492 0.017
09/18/04 1 BS1T1 0.28 4.730 0.299 0.421 0.534 0.115
09/18/04 1 S1T1 0.28 5.726 0.216 3.803 4.142 3.901
09/19/04 1 S1T2 0.21 4.847 0.420 3.256 3.387 1.646
09/19/04 1 S1T3 0.21 4.965 0.339 3.196 3.389 1.537
09/19/04 1 S1T4 0.21 4.378 0.153 2.951 3.304 1.265
09/18/04 4 Ambient 0.27 5.902 0.186 1.594 1.907 0.017
09/19/04 4 Ambient 0.22 5.551 0.237 1.611 1.745 0.072
09/18/04 4 C4T1 0.25 5.316 0.161 1.656 1.921 0.028
09/19/04 4 C4T2 0.20 5.726 0.137 1.668 1.872 0.017
09/19/04 4 C4T3 0.20 5.316 0.092 1.617 1.930 0.017
09/19/04 4 C4T4 0.20 5.316 0.121 1.668 1.918 0.012
09/18/04 4 BS4T1 0.28 5.668 0.191 1.820 1.867 0.213
09/18/04 4 S4T1 0.28 6.195 0.108 3.679 4.090 2.399
09/19/04 4 S4T2 0.26 6.019 0.256 4.211 4.480 2.454
09/19/04 4 S4T3 0.26 5.844 0.289 4.339 4.498 2.535
09/19/04 4 S4T4 0.26 5.785 0.205 4.354 4.523 2.535
*BS=before spike, S=Spike, C=Control, T=time
*No replicates for Trial Run Experiment
Site 1 and Site 4 Water Depths
09/17/04 09/18/04 09/19/04 09/20/04
-- Site 1 -A- Site 4 Linear (Site 1) Linear (Site 4)
Figure A-1. Water level changes in Site 1 and Site 4 during trial run
Site 1 Trial Run Experiment
A- Before Spike
Figure A-2. Trial run, Site 1, soluble reactive phosphorus cone. vs. time.
Site 4 Trial Run Experiment
0 1 2 3 4 5
-- Ambient Spike -A- Before Spike --- Control
Figure A-3. Trial run, Site 4, soluble reactive phosphorus conc. vs. time.
LIST OF REFERENCES
Anderson, J.M. 1976. An ignition method for determination of total phosphorus in lake
sediments. Water Research 10:329-331.
Baldwin, D.S. 1996. Effects of exposure to air and subsequent drying on the phosphate
sorption characteristics of sediments from a eutrophic reservoir. Limnology and
Oceanography 41: 1725-1732.
Bohlen, P. 2003. Nutrient Loads in Surface Runoff from Cattle Ranches in South
Florida. Myakka: A Soil and Water Science Department Publication, University
of Florida, IFAS. Volume 3, Number 3.
Boyle, T.P., and J.F. Fairchild. 1997. The role ofmesocosm studies in ecological risk
analysis. Ecological Applications 7:1099-1102.
Campbell, K.L., J.C. Capece, and T.K. Tremwel. 1995. Surface/subsurface hydrology
and phosphorus transport in the Kissimmee River Basin, Florida. Ecological
Daniel, T.C., A.N. Sharpley, and J.L. Lemunyon. 1998. Agricultural phosphorus and
eutrophication: A symposium overview. J. Environ. Qual. 27:251-257.
DeBusk, W.F. 1999. Phosphorus Cycling in Wetlands. SL170, Fact Sheet of the Soil
and Water Science Department, Florida Cooperative Extension Service, Institute
of Food and Agricultural Sciences, University of Florida.
Elder, J.F., B.J. Manion, and G.L. Goddard. 1997. Mesocosm experiments to assess
factors affecting phosphorus retention and release in an extended Wisconsin
wetland; prepared in cooperation with the Wisconsin Department of Natural
Resources. Middleton, Wis.: U.S. Dept. of the Interior, U.S. Geological Survey;
Denver, CO: Branch of Information Services.
EPA. 1993. Methods for the determination of inorganic substances in environmental
EPA. 1991. Methods for the determination of metals in environmental samples. 200.7.
Faulkner, S.P., and C.J. Richardson. 1989. Physical and chemical characteristics of
freshwater wetlands soils. In: D.A. Hammer (ed.), Constructed Wetlands for
Wastewater Treatment. Lewis Publishers, Chelsea, MI. pp 41-72.
Florida Cattlemen's Association. 1999. Water Quality BMPs for Cow/Calf Operations
Fonyo, C., R. Fluck, R., W. Boggess, H. Dinkler, and L. Stanislawski. 1991.
Biogeochemical behavior and transport of P in the Lake Okeechobee Basin. Area
3 Final Report. Departments of Agricultural Engineering and Food and Resource
Economics, IFAS, University of Florida.
Gachter, R., J.S. Meyer, and A. Mares. 1988. Contribution of bacteria to release and
fixation of phosphorus in lake sediments. Limnol. Oceanogr. 33:1542-1558.
Graetz, D.A., and V.D. Nair. 1995. Fate of phosphorus in Florida Spodosols
contaminated with cattle manure. Ecological Engineering 5:163-181.
Graetz, D.A., V.D. Nair, K.M. Portier, and R.L. Voss. 1999. Phosphorus accumulation
in manure-impacted Spodosols of Florida. Agric. Ecosyst. Environ. 75: 31-40.
Haan, C.T. 1995. Fate and transport of phosphorus in the Lake Okeechobee Basin,
Florida. Ecological Engineering 5:331-339.
Hanlon, E.A. 1994. IFAS Extension Soil Testing Laboratory Chemical Procedure and
Training Manual, June 1994.
Hiscock, J.G., C.S. Thourot, and J. Zhang. 2003. Phosphorus budget land use
relationships for the northern Lake Okeechobee watershed, Florida. Ecological
Engineering 5: 63-74.
Hunt, L., and R.J. Hunt. 2001. Estimating evapotranspiration in natural and constructed
wetlands. Wetlands 21:614-628.
Jackson, M.L. 1958. Soil Chemical Analysis. Prentice Hall, Inc., Englewood Cliffs, NJ.
Kangas, P., and W. Adey. 1996. Guest Editorial: Mesocosms and ecological
engineering. Ecological Engineering 6:1-5.
Kuipers, J.S. 2002. Prescribed grazing plan prepared for Pelaez & Sons, Okeechobee
County, Florida, in cooperation with Okeechobee Soil & Water Conservation
District and United States Department of Agriculture Natural Resources
Luscombe, P.C., J.K. Syers, and P.E.H. Gregg. 1979. Water extraction as a soil testing
procedure for phosphorus. Commun. Soil Sci. Plant Anal. 10:1361-1369.
Mathews, C.B. 1992. Cattle manure characterization and decomposition and its potential
impact on water quality. M.S. Thesis, University of Florida, 149 p.
McKeague, J.A., and J.H. Day. 1966. Dithionite- and oxalate-extractable Fe and Al as
aids in differentiating various classes of soils. Canadian Journal of Soil Science
Mitsch, W.J. and J.G. Gosselink. 1993. Wetlands, 2nd edition. Van Norstrand Reinhold
Nair, V.D. and D.A. Graetz. 2002. Phosphorus saturation in Spodosols impacted by
manure. Journal of Environmental Quality 31:1279-85.
Nair, V.D., K.M. Portier, D.A. Graetz, and M.L. Walker. 2004. An environmental
threshold for degree of phosphorus saturation in sandy soils. Journal of
Environmental Quality 33:107-113.
Nordlie, T. 2001. "UF Scientists say protect Lake Okeechobee before it's too late." UF
Overman, A.R., and R.V. Scholtz. 1999. Langmuir-Hinshelwood model of soil
phosphorus kinetics. Commun. Soil. Sci. Plant Anal. 30:109-119.
Reddy, K.R., M. Clark, J. Jawitz, T. Debusk, M. Annable, W. Wise, and S. Grunwald.
2003. Phosphorus Retention and Storage by Isolated and Constructed Wetlands
in the Okeechobee Basin. Annual Report submitted to the Florida Department of
Agriculture and Consumer Services Office of Water Policy, Tallahassee, Florida.
Reddy, K.R., E.G. Flaig, and D.A. Graetz. 1996. Phosphorus storage capacity of
uplands, wetlands and streams of the Lake Okeechobee Watershed, Florida.
Agriculture, Ecosystems and Environment 59:203-216.
Reddy, K.R., O.A. Diaz, L.J. Scinto, and M. Agami. 1995. Phosphorus dynamics in
selected wetlands and streams of the lake Okeechobee Basin. Ecological
Reddy, K.R., and E.M. D'Angelo. 1994. Soil processes regulating water quality in
wetlands. Global Wetlands: Old World and New. Edited by W.J. Mitsch.
Richardson, C.J. 1985. Mechanisms controlling phosphorus retention capacity in
freshwater wetlands. Science 228:1424-1428.
Sheldrick, B.H. 1984. Analytical method manual (no 84-011: Acid ammonium oxalate
extractable Fe and Al). Land Resource Research Insititute, Ottawa, ON, Canada.
South Florida Water Management District. 2004. Lake Okeechobee: Watershed
http://www.sfwmd.gov/org/wrp/wrp_okee/2 wrp_okeeh20shed/2 wrp_okeeh2
South Florida. 1997. Rules of the South Florida Water Management District: Works of
the District Basins Chapter 40E-61, F.A.C. Amended June 30, 1997.
Statistical Analysis Software (SAS) 9.1.0. 2003. SAS Institute, Cary, NC.
Thomas, G.W. 1996. Soil pH and Soil Acidity. In Methods of Soil Analysis. Part 3.
Chemical Methods. J.M. Bigham (Ed.), Soil Science Society of America, ASA,
Madison, Wisconsin. p475.
US Environmental Protection Agency, Region 4. 2003. Draft TMDL for Lake
Okeechobee tributaries S-133, S-135, S-154, S-191.
Christy L. Sackfield was born in Reidsville, North Carolina, in 1981. After
graduating high school in 1999, she moved to Raleigh, NC, to attend North Carolina
State University. She graduated with a degree in biological and agricultural engineering
in May 2003. After graduation she entered the master's program in water resources with
the Agricultural and Biological Engineering department in August 2003 to work on water
quality in the Lake Okeechobee drainage basin.