|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
AGRONOMIC AND ENVIRONMENTAL CHARACTERIZATION OF PHOSPHORUS IN
BIOSOLIDS PRODUCED AND/OR MARKETED IN FLORIDA
SARAH L. CHINAULT
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
O 2007 Sarah L. Chinault
To Mom, Dad, Amanda, Chris, and Richelle.
I would first like to thank Dr. O'Connor, my supervisory committee chair. While there
were times of great frustration (for both of us!) I am much better prepared for a professional
career thanks to his teachings. Thanks also go to Drs. Sartain and Elliott for serving on my
committee. I must thank Scott Brinton. Without his help, this proj ect would have been beyond
my capability. Scott has the ability to handle endless interruptions and stupid questions without
ever losing patience, for which I am grateful. Scott tended to daily greenhouse responsibilities,
and provided maj or help during harvests and teachings. Thanks go to Sampson Agyin-
Birikorang, who helped with the statistical manipulations of data. Sampson also has endless
patience and always made time to assist, even when his own duties were great. When I was
swamped with laboratory analyses, Matt Miller was a great help with dishes and other "lab
monkey" duties, which helped me maintain my sanity. Thanks go to Maria Silveria, who helped
me learn laboratory procedures when I started my working on my degree. Thanks go to the
College of Agriculture and Life Science (CALS) for providing a matching assistantship to pay
for my tuition. Additional stipend and research support was provided by the Florida Water
Environment Association (FWEA)-Utilities Council and Milorganite, Inc. Lastly, thanks go to
my family and friends, who supported me emotionally and financially (Dad!) through all my
years at UF.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ........._... ...........4........... .....
LIST OF TABLES ............... ............7... ......... ....
LIST OF FIGURES ............... ...........8.............. ....
AB STRAC T ......__.............. ............12.. ....
1 INTRODUCTION ............... ...........14...................
Hypotheses and Research Obj ectives ............... ............................17
Study Approach ............... ...........18...................
2 MATERIALS AND METHODS ............... ...........19..................
M ateri als ........... .. ... ........ ............. ..........1
Year 1 Glasshouse and Static Incubation Studies ............... ....................19
Laboratory Characterization ............... ............20.. ...............
Dynamic Laboratory Incubation ............... ..............................23
Glasshouse Study ............... ............26.. ...............
Statistical Analysis ............... ...........29...................
3 RESULTS AND DISCUSSION ............... ...........31..................
Laboratory Characterization ............... ...........31...................
Dynamic Laboratory Incubation ............... ...............................35
Glasshouse Study ............... ............42.. ...............
Y ields.............. .... ............. ...........4
Phosphorus Uptake and Relative Phytoavailability ............... ...............45
Phosphorus Leaching ............... ...........49...................
4 CONCLUSIONS ............... ............70.. ...............
Dynamic Laboratory Incubation ............... ..............................70
Glasshouse Study ................... .. ........... ............72 .......
Land Applying BPR or BPR-like Residuals ............... .......... ............74
APPENDIX GLAS HOUSE ANOVA AND CUMULATIVE DATA. ................. ............76
LIST OF REFERENCES ...........__ ...........82.......... .....
BIOGRAPHICAL SKETCH ............... ............85.. ...............
LIST OF TABLES
2-1 Selected properties oflImmokalee fine sand ............... ......... .............30
3-1 Materials received for laboratory characterization ............... ............ ......51
3-2 Selected general chemical properties of materials ............... ........... ........52
3-3 Selected phosphorus properties of materials ............... .......... ............53
3-4 Comparison of TP concentrations determined by 3 digestion methods ................... ...54
3-5 Dynamic laboratory incubation: cumulative SRP released and SRP released
(% of applied P). ........._ .. ...........55......__. ....
3-6 Cumulative Bahiagrass tissue P concentrations after 4 harvests........... ..._.........65
3-7 Relative phosphorus phytoavailability of biosolids to TSP. ............... ..............66
3-8 Various RPP estimate model fit statistics ....._____ ......___ ............_67
3-9 Point estimates of RPP for Milorganite and Lakeland NS
biosolids ............... ............69.. ...............
A-1 Cumulative Bahiagrass P uptake after 4 harvests ANOVA. ............... ...............76
A-2 Cumulative Bahiagrass yield after 4 harvests ANOVA. ............... ............... 77
A-3 Cumulative P leached in glasshouse experiment ANOVA ............... .............78
A-4 Cumulative Bahiagrass yields after 4 harvests ............... ............... ...._79
A-5 Cumulative P uptake for Bahiagrass after 4 harvests ........._._. .........___.....80
A-6 Cumulative P leached after 4 leachings ............... ..........................81
LIST OF FIGURES
3-la P rate: 56 kg P ha- Cumulative SRP released (% of applied P) as a function of
P source. ........._.._ ............56....___ .....
3-1b P rate: 224 kg P ha- Cumulative SRP released (% of applied P) as a function of
P source. ........._.._ ...........56...___ .....
3-2a P rate: 56 kg P ha- Dynamic laboratory incubation: cumulative P released (% of applied
P) as a function of biosolids-PWEP. ............... ............ ......... .....57
3 -2b P rate: 224 kg P ha- Dynamic laboratory incubation: cumulative P released (% of
applied P) as a function of biosolids-PWEP ............... ......... .............57
3-3a P rate: 56 kg P ha- Dynamic laboratory incubation: cumulative P released as a function
of bi osoli ds-P SI ............ ............58...__ __ ....
3-3b P rate: 224 kg P ha- Dynamic laboratory incubation: cumulative P released as a function
of bi osoli ds-P SI ............ ...........58....__ ....
3-4a P rate: 56 kg P ha- Dynamic laboratory incubation: cumulative P released (% of applied
P) per released event. ............... ...........59..................
3 -4b P-Rate: 224 kg P ha- Dynamic laboratory incubation: cumulative P released (% of
applied P) per released event. ............... ......... ......... .........6
3-5a P rate: 56 kg P ha- Cumulative Bahiagrass yields after 4 harvests .............. .... .......61
3-5b P-rate: 112 kg P ha- Cumulative Bahiagrass yields after 4 harvests ........... ....... .....61
3-5c P rate: 224 kg P ha- Cumulative Bahiagrass yields after 4 harvests ............. ...... ......62
3-6 Cumulative Bahiagrass yield-weighted tissue P concentrations after 4 harvests ...........62
3-7 P rate: 224 kg P ha- Bahiagrass N tissue concentrations in each harvest. .................. 63
3-8 P rate: 224 kg P ha- Bahiagrass yield-weighted N concentrations after 4 harvests.......63
3-9a P rate: 56 kg P ha- Cumulative Bahiagrass P uptake after 4 harvests. ................... ....64
3-9b P-rate: 112 kg P ha- Cumulative Bahiagrass P uptake after 4 harvests ................... ....64
3-9c P-rate: 224 kg P ha- Cumulative Bahiagrass P uptake after 4 harvests. ................... ..64
3-1 0 Cumulative Bahiagrass P uptake as a function of P applied ............... ..............66
3-11 Plateau model: P uptake from Milorganite treatments as a function of P applied........... 68
3-12 Relative phosphorus phytoavailability (RPP) as a function of biosolids-PWEP. ...........69
LIST OF ABBREVIATIONS
AN ammonium nitrate
ANOVA analysis of variance
BAP biologically available phosphorus
BPR biological phosphorus removal
DDI distilled, de-ionized water
EC electrical conductivity
FDEP Florida Department of Environmental Protection
ICP inductively coupled plasma spectrophotometry
LOI loss on ignition
PSC phosphorus source coefficient
PSI phosphorus saturation index
PV pore volume
PWEP percent water-extractable phosphorus
RPA relative phosphorus adsorption
RPP relative phosphorus phytoavailability
SRP soluble reactive phosphorus
TC total carbon
TN total nitrogen
TP total phosphorus
TSP triple super phosphate
UJF University of Florida
USEPA United States Environmental Protection Agency
WEP water-extractable phosphorus
WTR water treatment residual
WWTP wastewater treatment plant
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
AGRONOMIC AND ENVIRONMENTAL CHARACTERIZATION OF PHOSPHORUS IN
BIOSOLIDS PRODUCED AND/OR MARKETED IN FLORIDA
Sarah L. Chinault
Chair: George O'Connor
Major Department: Soil and Water Science
Land application of biosolids can result in an accumulation of soil phosphorus (P). While
excess soil P is typically not harmful to crops, P can migrate offsite and can lead to surface and
groundwater impairment. Increased concern over accelerated eutrophication of water bodies has
led to heightened scrutiny and regulation of biosolids land application. Regulation could include
limiting application rates to match crop P requirements (P-based rates). If P-based regulations are
imposed, it will be critical to understand biosolids P phytoavailability and the potential for
surface and groundwater impairment from biosolids P. We conducted laboratory and glasshouse
studies to provide understanding. A laboratory incubation was conducted using small soil
columns to assess P release potential: 11 biosolids and triple super phosphate (TSP) were
individually mixed with 400 g of a typical low-P sorbing Florida soil (Immokalee fine sand) at
two rates: 56 and 224 kg P ha- .
Columns were leached bi-weekly for 14 weeks to attain 60 mL of drainage in each
leaching. Soluble reactive P (SRP) was determined and summed over the 8 teachings.
Cumulative P release (as a percentage of P applied) was greatest from biological P removal
(BPR) and BPR-like biosolids. Phosphorus release from Milorganite, a thermally dried biosolids
high in Fe and Al oxides (44 g kg- ), released ~39% of applied P at the 56 kg P ha-l rate,
indicating that P release was limited by association of P with Fe and Al and by heat drying. In
contrast to the moderate P release from Milorganite, P release from the Lakeland NS biosolids, a
BPR-like product with low percent solids (3%), and low Fe and Al content (~12 g kg- ), was
~90% of P applied at 56 kg P ha- .
A glasshouse study utilizing large soil columns was conducted: 7 biosolids and TSP were
individually mixed with 4 kg samples of Immokalee soil and Bahiagrass (Paspalum notatum
Flugge) was grown. Bahiagrass was harvested monthly for 4 months and tissue total P (TP) was
measured. Four of the biosolids chosen for the glasshouse study had high PWEP values (>15%),
and their P was expected be highly plant available. Indeed, the OCUD S, Lakeland NS, GRU,
and Boca Raton biosolids fit into the "high" category of relative P phytoavailability (RPP; >75%
of TSP). Two biosolids with low P-solubility (PWEP I 1.1%) fit into the moderate RPP category
(25-75% of TSP): Milorganite, and GreenEdge. One biosolids (Disney) with a moderate PWEP
(8.4%) also fit into the moderate RPP category. Results of the glasshouse study indicate that no
change in biosolids application rate is needed (or justified) for BPR and BPR-like biosolids with
RPP values in the high RPP category. However, application rates for Disney and GreenEdge
could be approximately doubled to meet crop P needs. Phosphorus in Milorganite is only ~1/3 as
phytoavailable as P in TSP, indicating that Milorganite may be applied at P rates as great as 224
kg P ha-l (based strictly on phytoavailability). Given the wide range of RPP values and leaching
risks of various biosolids, land application of biosolids should not be regulated en masse. PWEP,
and PSI are easily determined and excellent gauges of both biosolids agronomic value and P
leaching hazard. PWEP is a better indication of how a biosolids might affect the environment
than biosolids-TP or soil test P.
Throughout history, human and animal wastes have been recycled to agricultural land to
supply nutrients for crops. However, the development of centralized wastewater treatment
systems, and an ever-increasing human population has created biosolids disposal challenges. The
U.S. produced an estimated 6.5 X 106 Mg of biosolids in 2000 and approximately 50-60% of
biosolids were disposed of via land-based recycling (USEPA, 1999). Biosolids production is
expected to increase to 7.5 X 106 Mg yrl by 2010. The state of Florida produces about 2.7 X 105
dry Mg of biosolids per year; 66% of the biosolids are land applied, 17% are land filled, and 17%
are marketed and distributed to the public (FDEP, 2005). In addition to biosolids produced
within the state of Florida, 9. 1 X 104 Mg of class AA pelletized biosolids are imported each year.
Given the large quantities of biosolids produced, land-based recycling of biosolids becomes a
critical disposal route for municipalities. Biosolids land application can also be an
environmentally sound and beneficial practice.
Biosolids are a source of essential plant elements such as N, P, sulfur, and micronutrients.
Thus, land-based recycling of biosolids to agricultural land can supply farmers with an economic
alternative to chemical fertilizers. Biosolids are also high in organic matter, which can be
important in Florida, where soils are typically sandy and low in organic matter. While the land
application of biosolids is beneficial, long-term biosolids application can result in accumulation
of soil P (O'Connor et al., 2005). When biosolids are applied at an N-based rate to meet crop N
requirements, P is typically oversupplied due to differences in biosolids N:P than crop needs.
The excessive P accumulation it is not harmful to crops (Person et al., 1994), but leads to
accumulation of P in amended soils and potential environmental problems. Migration of P off-
site to ground and surface waters is a cause for concern as P is generally the limiting nutrient in
fresh water ecosystems for accelerated eutrophication (Carpenter et al., 1998). P can migrate off-
site either in dissolved form, as biosolids particles, or attached to soil particles (Elliott et al.,
2005). Most soils in the U.S. have sufficient P-sorbing capacity (due to Fe and Al oxides) to
prevent P leaching. Elliott et al. (2002) showed that even sandy Florida soils with moderate P-
sorbing capacity can prevent P leaching. In soils with adequate P-sorbing capacity, P loss occurs
mostly through erosion and runoff, and erosion control measures can help control P loss in these
soils. Atlantic Coastal Plain soils (including most Florida soils), however, are naturally low in Fe
and Al oxides. In such soils, P leaching and loss of P dissolved in subsurface flow become the
dominant P loss mechanism (He et al., 1999).
Excess P in water bodies can lead to accelerated eutrophication. Increased concern over
water quality has led to amplified scrutiny of biosolids land application programs and
recommendations that biosolids application be limited to meet crop P needs (P-based). Limiting
biosolids land application to P-based rates would significantly reduce biosolids application rates.
When application rates are decreased, more land area is required for disposal, increasing disposal
costs. Lower biosolids application rates also mean additional N fertilizer is required, increasing
costs to farmers, and making land-based recycling of biosolids for agronomic benefits less
Recent research suggests that limiting biosolids application to P-based rates is probably
unnecessary for conventionally produced biosolids (Brandt et al., 2004). Biosolids-P is not
necessarily as labile as P in mineral fertilizers or manures, and liability can be greatly influenced
by the wastewater treatment process (Maguire et al., 2001). Treatment processes such as thermal
drying can significantly reduce P liability of biosolids compared to conventionally produced cake
(Smith et al., 2002). A possible exception may be biological P removal (BPR) biosolids, in
which the wastewater treatment process is engineered to promote luxury uptake of P by bacteria.
While the BPR process reduces P concentrations in wastewater effluent, the resulting biosolids
have a higher concentration of P than conventionally produced residuals, and labile P in BPR
biosolids may be higher than for conventional or heat-dried biosolids (Elliott et al., 2002; Brandt
et al., 2004; O'Connor et al., 2004).
Total P (TP) concentration of the biosolids is generally a poor indicator of P liability and
phytoavailability. Only a small fraction of P from most conventionally produced biosolids is
soluble, making most biosolids less likely to negatively impact the environment compared to
soluble P sources (mineral fertilizer, manures). The water-extractable P (WEP) content of
biosolids has been highly correlated to P liability (Brandt et al., 2004). For typical biosolids,
percent water-extractable P (PWEP = WEP/TP*100) is < 5% (Brandt et al., 2004), and
phytoavailability (plant-available) P is only 40-50% of TSP (USEPA, 1995). Biosolids produced
using iron (Fe) and aluminum (Al) salts may have even lower P liability and PWEP values
(<0.5%) (Corey, 1992; Brandt et al., 2004). BPR biosolids may have much greater soluble P
concentrations than conventionally produced biosolids, and greater PWEP (~ 14%) (Brandt et
The phosphorus saturation index (PSI) has also been correlated to environmental P
leaching risk (Elliott et al., 2002). PSI is calculated as the molar ratio of oxalate-extractable P to
Fe and Al ([Pox]/[Alox + Feox]). When biosolids have PSI values less than 1.1, little P leaching is
expected to occur, even in sandy, low P-sorbing soils (Elliott et al., 2002). Both PWEP and PSI
can serve as a priori estimates of biosolids-P liability. Both measures can better estimate how P
in biosolids will affect the environment when land applied than biosolids-TP or soil test P
measures (Maguire et al., 2001; Elliott et al., 2002; Brandt et al., 2004; O'Connor et al., 2004).
Given the nutrient concerns that accompany biosolids land application, each state's
Natural Resources Conservation Service (NRCS) is required by the Unified Strategy for Animal
Feeding Operations (USDA/USEPA, 1999) to develop P management strategies. Under code
590, states have three options to manage P: agronomic soil test P recommendations,
environmental soil P thresholds, or a P site index to evaluate vulnerability to potential P loss
(Elliott and O'Connor, 2007). Most states have chosen a P-index approach as both P source and
transport factors are taken into account. The P-index acknowledges that for negative
environmental impacts to occur, both a soluble P source and a transport mechanism are required
(Sharpley et al., 2003). The state of Florida is in the midst of developing a P-index to manage P
Hypotheses and Research Objectives
A good understanding of biosolids P phytoavailability relative to fertilizer-P is critical to
making adjustments to biosolids application rates should P-based application restrictions be
imposed. All biosolids do not have the same potential to negatively affect the environment,
hence, it is important to know which biosolids-P chemical characteristics can be used to judge a
priori the environmental impact a biosolids will have once land applied.
* Hypothesis 1: Relative P phytoavailability is greater for BPR biosolids than conventionally
* Hypothesis 2: P leaching is significantly greater from (sandy, low P-sorbing) soil amended
with BPR residuals than the same soil amended with conventionally treated biosolids.
* Hypothesis 3: P liability of all biosolids is less than fertilizer-P (TSP).
* Objective 1: Characterize the agronomic value of various biosolids important in Florida so
adjustments can be made to biosolids application rates if biosolids P-based regulations are
* Objective 2: Develop sufficient data to demonstrate that most biosolids pose less of an
environmental hazard (with the possible exception of BPR biosolids) than readily soluble P
* Objective 3: Expand the database of knowledge to include a wider variety of biosolids
production schemes, specifically materials produced or marketed in Florida. Research was
focused on materials not previously studied in detail, specifically BPR materials.
A full laboratory characterization was conducted on 21 biosolids produced or marketed in
Florida to identify materials warranting further study in glasshouse and laboratory experiments.
Seven residuals were chosen for a glasshouse study to judge agronomic P availability and
environmental liability. Eleven biosolids were chosen for a dynamic laboratory incubation, the
purpose of which was to quantify water soluble P, examine the kinetics of P release, and
determine the leaching hazard of various biosolids materials.
MATERIALS AND METHODS
Wastewater treatment plants (WWTPs) throughout Florida were invited to participate in
the research effort by completing a survey (developed by Robert Morrell of PBS&J consulting)
of plant operations and materials characteristics. Completed surveys were submitted by 24
WWTPs, representing 22 residuals products, and results were forwarded to UF personnel. Based
on survey results, 20 Florida products were requested from WWTPs for analysis. A few samples
of biosolids produced in other states, but marketed in Florida, were also requested. Milorganite, a
commercially available biosolids product from Wisconsin, was chosen to be included in this
study. In total, 21 materials were subjected to laboratory characterization.
Year 1 Glasshouse and Static Incubation Studies
In 2005, glasshouse and static laboratory incubation studies were conducted with 9
biosolids: Milorganite, GreenEdge, Jacksonville Cake, Tampa, Disney Compost, Pinellas Cake,
Boca Raton, and Orange County Utilities Division (OCUD) east dry and cake materials. The
residuals were individually amended to a typical Florida Spodosol (Immokalee fine sand) at 2 P-
application rates: 56 and 224 kg P hal to represent P-based and N-based biosolids application
rates, respectively. Most of the soil was used in the glasshouse study, but some was used in a
static laboratory incubation. The glasshouse study was intended to provide measures of P
leaching risk and relative P phytoavailability to pasture grass of various biosolids. Amended soils
used in the static laboratory incubation were sampled at various times for WEP and TP
determination, and the data were intended to yield information about the kinetics of biosolids-P
release. Both the glasshouse study and the static laboratory incubation experiments were
terminated in December 2005 because of apparent un-equal P loadings in treatments. We have
no explanation for the unequal P loads, but without equal P loadings, conclusions on relative P
phytoavailability and P leaching were not possible. Similar studies were begun anew in 2006.
Biosolids percent solids was determined by drying (105o C) "as is" materials to constant
weight. Biosolids pH and electrical conductivity (EC) were determined using a 1:10 ratio of
biosolids (dry weight equivalent) to distilled, de-ionized water (DDI), and equilibrated by static
incubation at room temperature for 2 hours. After the 2 hr equilibration, the biosolids-DDI water
mixture was stirred and allowed to settle, and pH and EC were determined on the solution. Total
carbon/ total nitrogen (TC/TN) was determined on dried, ball-milled biosolids by combustion at
10100 C using a Carlo Erba (Milan, Italy) NA-1500 CNS analyzer.
Loss on ignition (LOI), a measure of biosolids organic matter content, was determined by
measuring 0.2 g of oven-dried and ball-milled biosolids into pre-weighed 50 mL beakers. The
materials were ashed at 2500C for 30 minutes, followed by ashing at 5500C for an additional 4
hours. The samples were allowed to cool, and the beakers were weighed again. Mass lost during
the ash process is a measure of organic matter (Sparks, 1996).
To analyze for total phosphorus (TP), samples of biosolids were first oven-dried and
ground in a ball mill to a fine powder. The biosolids (0.2 g) samples were then ashed (2500C for
30 minutes and 5500C for 4 hours) in a muffle furnace to destroy organic matter. The ashed
samples were digested with ~2 mL of DDI water and 20 mL of 6 M HC1. The beakers were
placed on a hotplate at 110o C until the HCI evaporated, and the residual was dry. The hotplate
temperature was then increased to the hottest setting (~3000C) for 1 hour. The beakers were
allowed to cool, and ~2 mL of DDI and 2.25 mL of 6 M HCI was added to the residual material
and the beakers were placed on a hot plate (highest setting) until small bubbles started to form
(solution began to boil). The solution was removed from the hotplate and allowed to cool. The
solution was transferred quantitatively to a funnel containing a Whatman #42 fi1ter, and allowed
to drain into 50 mL volumetric flasks. The fi1ter paper was rinsed 3 times with DDI and allowed
to drain completely between each rinse (Andersen, 1976). The solutions were brought to volume
and analyzed for P using the molybdenum blue method (Murphey and Riley, 1962). Digests from
the Andersen (1976) method were also analyzed for total Fe, Al, and Ca via inductively coupled
plasma spectroscopy (ICP).
We also determined biosolids-TP using EPA method 3050A (USEPA, 1995) to verify TP
values determined by the Andersen (1976) method were accurate. Method EPA3050A required 2
g (dry weight equivalent) of biosolids. Samples were pre-digested with 10 mL of 1:1 HNO3,
covered with reflux caps, and heated at 110oC for 10-15 minutes. After the pre-digestion, an
additional 5 mL of HNO3 WAS added and the samples were refluxed for 30O minutes. This second
step was repeated until brown fumes were no longer visible. The reflux caps were removed and
the digestion solution was allowed to evaporate (2-3 h) to a volume of 5 mL. The flasks were
removed from the hotplate and allowed to cool. Then, 30 mL of 30% hydrogen peroxide and 2
mL of DDI was added and new reflux caps were place on top of the flasks. The samples were
returned to the hotplate set at 110oC until effervescence subsided. Additional 1 mL aliquots of
30% hydrogen peroxide were added (not exceeding 10 mL) until effervescence ceased. After the
peroxide reaction, 5 mL of concentrated HNO3 WAS added and the samples were refluxed for 30
minutes. The samples were then cooled, and filtered through a Whatman #41 fi1ter into 50 mL
volumetric flasks. The fi1ters were rinsed 3 times with DDI and allowed to drain completely
between each rinse. Filtered digests were brought to volume and analyzed for P via ICP.
Biosolids-TP was also determined by an outside laboratory via perchloric acid digestion
(Association of Official Analytical Chemists, 1990; The Fertilizer Institute, 1982) to further
confirm TP values. One gram of dried and powdered biosolids were boiled with 30 mL of
concentrated HNO3 for ~20 minutes. The samples were cooled and 20 mL of perchloric acid was
added until the solutions were colorless and started to fume. Once samples were colorless, and
white fumes were noted, the digestion was continued for an additional 20 minutes. Following the
second 20-minute digestion, 30 mL of DDI was added and the samples were boiled for 5
minutes. Samples were then cooled and filtered through an Ahlstrom Grade 54, 12.5 cm filter
paper into a 100 mL volumetric flask. Samples were brought to volume and analyzed for P via
Water-extractable P (WEP) was determined using a 1:200 biosolids (dry weight
equivalent) to DDI water (Sharpley and Moyer, 2000). The residuals/water mixture was shaken
on an orbital shaker at 200 strokes per minute for 1 hour. A sub-sample of the mixture was
centrifuged for 10 minutes at 4000 rpm, and the supernatant was vacuum-filtered through a 0.45
Cpm- Eilter. The solution was analyzed for P using the molybdenum blue method (Murphey and
Riley, 1962). Percent WEP (PWEP) was calculated by dividing WEP by TP and multiplying by
100 (PWEP = WEP/TP*100). Brandt et al. (2004) concluded that PWEP was a good measure of
the environmentally relevant portion of total P in biosolids and manures.
Oxalate-extractable P, Fe, and Al was determined by shaking 0.5 (dry weight equivalent)
of biosolids with 30 mL of 0. 175 M ammonium oxalate and 0. 1 M oxalic acid for 4 hours on an
orbital shaker at 200 strokes per minute (Loeppert and Inskeep, 1997). The samples were shaken
in a sealed box to avoid exposure to light. After shaking, the samples were centrifuged for 10
minutes at 4000 rpm, and vacuum filtered through a Whatman #42 fi1ter. The supernatant was
analyzed for P, Fe, and Al via ICP. Moles of oxalate-extractable P, Fe, and Al were used to
calculate PSI, an indication of labile P in residuals [PSI = moles P/ (moles Fe + mole Al)] (Elliott
et al., 2002).
Fe-oxide extractable P (Fe-strip P) in biosolids was determined using Fe-oxide
impregnated strips (Chardon, 1996). Strips were prepared by soaking Whatman #42 filter paper
in 0.65 M FeCl3-6H20 + 0.6 M HCI solution overnight. The strips were removed and allowed to
air dry. The strips were immersed for 30 seconds in 2.7 M NH40H, then rinsed twice in DDI
water, left to sit in clean DDI water for 1 hour, and then air dried before use. Fe-strip P was
determined by shaking the prepared strips with 1 g (dry weight equivalent) of biosolids and 60
mL of 0.01 M CaCl2 in a 120 mL bottle for 16 hours on an orbital shaker at 125
oscillations/minute. The original method was changed from 40 mL CaCl2 to 60 mL CaCl2 to
ensure the biosolids were completely covered with solution. During the CaCl2 WASh, P is
removed from the biosolids and is retained on the Fe-strip. After shaking, the strips were
removed from the glass bottles and rinsed thoroughly with DDI to remove biosolids particles.
The rinsed strips were placed in 125 mL Erlenmeyer flasks and shaken with 40 mL of 0. 1 M
H2SO4 On an orbital shaker for 1 hour at 125 oscillations/minute. The 0.1 M H2SO4 extracts P
from the Fe-strip, and the solutions were then analyzed for P using the molybdenum blue method
(Murphey and Riley, 1962). Fe-strip P is measure of biologically available P in soil (Sharpley,
1993a and b).
Dynamic Laboratory Incubation
Eleven biosolids and TSP were chosen for a dynamic laboratory incubation. Biosolids
were mixed at 2 rates, equivalent to 56 and 224 kg P ha l, with 400 g samples of Immokalee A
horizon soil (Immokalee fine sand, sandy, siliceous, hyperthermic Arenic Alaquods; Table 2-1)
and enough water (40 mL) to reach field capacity. The biosolids-amended soil was incubated for
2 weeks in zip-lock plastic bags. The bags of soil were mixed and opened daily during the initial
incubation to avoid anaerobic conditions. As in the glasshouse study (below), ammonium nitrate
was added to the soil to supply additional N. Originally, we intended to equalize varying N levels
among the materials at both P rates to 300 kg N ha-l plant available N (PAN). We initially
applied 75 kg N hal to the biosolids-amended soil at the start of the 2-week equilibration, with
the intention of split applying the remaining N after leachings. However, the columns of soil
were near field capacity for the duration of the experiment, and applying AN in solution would
have induced unintentional leaching; applying AN in crystalline form would not allow for N
distribution throughout the soil column. Thus, the remaining 225 kg N ha-l was not applied.
Incubation columns were constructed of 17 cm x 5 cm sections of polyvinyl chloride
(PVC) tubing with screening in the bottom to prevent soil loss. At the end of the initial 2-week
bag incubation, 400 g biosolids-amended soils were packed into the incubation columns to a
depth of 13 cm and a bulk density of 1.51 g cm-3. A total of 75 columns was used (12 materials x
2 rates x 3 reps + 3 controls = 75). The columns were positioned vertically in wooden racks for
Preliminary experiments were conducted to determine a representative column pore
volume (PV). Two methods were used to determine PV, gradual wetting and flood wetting. Four
incubation columns were filled with 400 g of dry soil and weighed. Tap water was gradually
added to two of the columns until 1 or 2 drops of water exited the bottom (gradual wetting). The
retained water volume (~118 mL) was assumed to be 1 PV. The other two columns were flooded
with 250 mL of tap water and allowed to drain overnight. When drainage was complete, the
columns were re-weighed and the difference in weight (~120 g) was assumed to be 1 PV. One
PV was taken to be 120 mL.
Every 2 weeks, sufficient tap water (adjusted to pH 5.0) was added to the columns to
result in 60 mL (1/2 PV) of drainage. We adjusted the pH of the water to 5.0 to simulate the pH
of rainfall in south Florida. Laboratory tap water was analyzed for soluble reactive (inorganic) P
(below detection limit; 0.001 ppm). For the first 2 leachings, P analysis of the leachates included
total phosphorus (TP), total dissolved phosphorus (TDP) and soluble reactive phosphorus (SRP);
pH and EC was also determined. Total P in the leachate was determined by digesting 5 mL of
unfiltered leachate with 0.3 5 g of potassium persulfate and 1 mL of 5.5 M H2SO4 On a digestion
block at 1250 C until only 5 mL of solution remained (USEPA, 1993). The temperature was then
increased to 150o C until 0.5 mL of solution remained. Reflux caps were placed on top, and the
temperature was further increased to 3800 C and digestion continued until the solution was clear.
TDP was determined in the same way, however the leachates were vacuum filtered through a
0.45 Cpm- filter to remove particulates before digestion. The digested leachate was analyzed for P
via the molybdenum blue method (Murphey and Riley, 1962). SRP was determined by vacuum
filtering leachate through a 0.45 Cpm- filter. The filtered leachate (undigested) was analyzed for P
using the molybdenum blue method. Organic P was determined by subtracting SRP from TDP.
Results from the first 2 leachings, revealed that the SRP represented the maj ority (>80%) of the
total P leached from the amended soils, and that organic P was minimal (<10% of TP). Thus in
subsequent leaching events, TP and TDP were measured only for select (highly colored or
cloudy) samples: Lakeland NS, OCUD S, OCUD E cake and OCUD E dry. TDP and TP values
were determined on these samples to ensure organic P and particulate P was minimal. Leachate
pH and EC was measured directly on unfiltered/undigested leachate.
Seven biosolids and TSP were chosen for the glasshouse experiment for a total of 8
materials. Two biosolids were selected to represent (expected) low-P availability biosolids:
Milorganite, and GreenEdge (PWEP I 1.1%). Four biosolids were selected as high P-availability
biosolids (Boca Raton, GRU, Lakeland NS, and OCUD S), which are BPR or BPR-like residuals
(PWEP > 15%). One biosolids selected (Disney) had a moderate PWEP (8.4%) and biosolids-P
was expected to be moderately plant available. Biosolids were also selected based on PSI and
PWEP values. BPR and BPR-like biosolids typically have PSI and PWEP values greater than
traditionally produced residuals. We chose biosolids with high PSI and PWEP values to expand
the existing relative P phytoavailability database. Much work has been done on the
phytoavailability and liability of conventionally produced biosolids, including some Florida
materials, and some BPR materials (O'Connor et al., 2004; Elliott et al., 2002), but the database
is limited. Materials were mixed with 4 kg of A horizon Immokalee soil (Table 2-1) at 3 rates:
56, 112, and 224 kg P ha- The 56 and 224 kg P ha-l rates represent P-based and N-based
application rates, respectively. The 112 kg P ha-l rate was added to better define the relative P
phytoavailability response curve of the materials. Immokalee fine sand was chosen to represent a
typical, P-deficient, low-P sorbing Florida soil. A P-deficient and low P-sorbing soil is necessary
to represent a "worst-case" scenario of biosolids land application, and to maximize plant
response to P additions. The Immokalee soil does not retain P and can allow P movement to
ground and surface waters.
The biosolids-amended soils were equilibrated (at field capacity) in zip-lock plastic bags
for 2 weeks in the laboratory prior to use in the glasshouse. Ammonium nitrate (AN) was added
to the amended soils to equalize varying amounts of N supplied by the biosolids. N was applied
at an equivalent rate of 300 kg ha-l plant available nitrogen (PAN). The GreenEdge material
supplied the most PAN (240 kg ha l) at the high P rate. We opted to supply an additional 60 kg
hal of N to ensure N limitations would not affect P release or uptake. This rate is excessive for
Bahiagrass (179 kg N ha-l recommended; Kidder et al., 1998), but could not be avoided due to
the N:P ratio of the materials and the high targeted P rate of 224 kg ha- N mineralization was
assumed to be 40% of the total N in all of the biosolids, based on previous experience (O'Connor
and Sarkar, 1999). The total supplemental AN needed was split-applied, with '/ applied prior to
incubation, and the remaining split applications added throughout the growing season (following
the first 3 leaching events). AN is fully soluble in water and therefore is immediately available
for plant uptake. Thus, AN was split-applied to ensure sufficient N was available for uptake
throughout the growing season. A blend of potassium-magnesium sulfate ("sul-po-mag"; 22% S,
18% K, 11% Mg) was added (0.91 g, equivalent to ~444 kg ha l) to supply adequate and uniform
S, K, and Mg.
Containers for the glasshouse study were constructed of 15 cm diameter X 45 cm long
sections of polyvinyl chloride (PVC) tubing. A screen was fitted to the bottom to prevent soil
loss. A PVC cap fitted with plastic tubing was attached to the bottom of the column to allow
leachate collection. A total of 100 columns was used (8 materials X 4 replicates X 3 rates + 4
controls = 100). The columns were arranged in a randomized complete block design in wooden
racks holding 6 columns each. The experiment was blocked to minimize glasshouse positioning
effects. The columns were rotated one position within each block weekly to further minimize
glasshouse effects on grass growth. O'Connor et al. (2004) used the same container and
experimental design to quantify P uptake and leaching for 12 biosolids products in the same soil.
Thirty cm of sand (~8.5 kg) was packed into the columns as a support layer (TP = 12 g
kg- WEP = 0.12 mg kg- pH = 5.1, RPA = 8.6%). The sand was included to provide additional
rooting depth for the grass. The sand was flooded and allowed to freely drain to wet the layer to
field capacity and to remove soluble constituents. Following the 2-week laboratory incubation, 4
kg of biosolids-amended soil was placed on top of the sand layer in each column, and 5 g of
Bahiagrass seed was planted. Five grams of Bahiagrass seed per pot is a seeding rate equivalent
to~-28.6 Mg ha- the recommended seeding rate is ~10.9 Mg ha-l (Chambliss et al., 2001). We
increased the seeding rate to ensure sufficient and rapid soil surface coverage. The seeds were
covered with a layer of sand (~0.5 cm) and misted every 3-4 hours until germination. The grass
was allowed to grow for approximately 7 weeks before the first harvest. Subsequent harvests
occurred at approximately 4-week intervals.
The Bahiagrass was harvested to a height of 3.8 cm using scissors. The wet clippings
were placed in pre-weighed paper bags and dried at 680 C to constant weight to represent grass
yields. The dried tissue was ground to pass a #20 sieve with a Wiley mill, digested (Andersen,
1976) and analyzed for P via the molybdenum blue method (Murphey and Riley, 1962).
Phosphorus uptake was calculated as yield times tissue P concentration. Individual
harvest P uptake masses were summed to give cumulative P uptake mass. Tissue N content of
plants from the 224 kg P ha-l rate for all treatments was also measured for all 4 harvests. Tissue
N was determined by grinding the tissue to a fine powder, and analyzing for N by combustion at
10100 C using a Carlo Erba (Milan, Italy) NA-1500 CNS analyzer.
Following each harvest, sufficient tap water (adjusted to pH 5.0) was applied to each
column to yield ~500 mL (~0.25 pore volume) of leachate. Soluble reactive (inorganic) P (SRP)
of glasshouse tap water was measured (below detection limit; 0.001 ppm) via the molybdenum
blue method (Murphey and Riley, 1962). Leachate was analyzed for SRP using the molybdenum
blue method. Leachate volume times leachate P concentration yielded mass of P leached.
Leachates from the Lakeland NS material (high P rate) were highly colored. TP was determined
on the leachates to ensure organic P loss from the material was not significant.
Cumulative P leached from the laboratory incubation was subj ected to a time-series
analysis (SAS Institute, 1989). Data were tested for normality using PROC UNIVARIATE. To
normalize yield and P uptake data, logarithmic transformations were needed. P leaching data
were normalized with a square transformation. Transformed cumulative yield, P uptake, and P
leached from the glasshouse study were statistically analyzed using PROC GLM. Means
separation of treatment differences was by Tukey test (p < 0.05) on transformed data. Relative
phytoavailability was calculated by fitting a linear regression to P uptake as a function of P
applied data using a slope- ratio approach. All P sources were regressed through a common
intercept of 8.48 mg, the average value of P uptake for the control, resulting in a response
proportional to the rate of P application.
Table 2-1. Selected properties of Immokalee fine sand.
Parameter, units Value
Sand, % 95a
Organic matter, g kg-l 7.0 a
Oxalate-extractable P, mg kg-l 13.1 a
Oxalate-extractable Fe, mg kgl 85.6 a
Oxalate-extractable Al, mg kgl 40.1 a
PSI" 0.14 a
RPAd, % 2.0
Total P, mg kg-l 15.5
Mehlich 1-extractable P, mg kgl 1.47
KCL-extractable P, mg kgl 1.9 a
NaOH-extractable P, mg kg-l 3.3 a
HCl-extractable P, mg kg-l 0.9 a
Sequenced sum, mg kg-l 6.1 a
aData from O'Connor et al. (2004). cPhosphorus Saturation Index. dRelative phosphorus
adsorption [fraction of 400 mg P kg-l soil load sorbed (Harris et al., 1996)].
RESULTS AND DISCUSSION
The laboratory characterization was key to determining which materials warranted further
study in the dynamic laboratory incubation and glasshouse studies. Twenty-one biosolids were
received and analyzed (Table 3-1). As detailed below, both basic chemical properties and P
characteristics were important for material selection. Recall that a goal of both the glasshouse
and dynamic laboratory incubation experiments was to include materials not previously studied,
specifically BPR and BPR-like products.
Total nitrogen (TN) concentrations ranged from 16 to 70 g kgl (1.6 to 7%). Biosolids
C:N ratios were generally low (6-15), with the exception of the West Palm Beach Compost
material (C:N of 26) (Table 3-2). The TN and C:N ratios are typical of U. S. produced biosolids
(USEPA, 1995). C:N ratios are important for predicting N mineralization from biosolids. In
general, when the C:N ratio of an organic material is less than 20: 1, N will be released into the
soil. If the C:N ratio exceeds 30: 1, N can be immobilized by soil microbes, and unavailable for
plant uptake. Thus, we selected materials with C:N ratios below 20: 1. The Disney material was
chosen as the representative compost residual to avoid possible complications with N
immobilization from high C:N ratio of the West Palm Beach biosolids.
Biosolids pH varied with source and product form (Table 3-2). Cakes were circum-
neutral to alkaline, and thermally dried products were slightly acidic. Lime stabilized products
had the highest pH values, as lime stabilization increases the pH of the final product (pH > 12).
We excluded materials with pH values > 9 from the glasshouse study as Bahiagrass prefers an
acidic environment (pH~-5.0; Chambliss and Adjei, 2006). O'Connor et al. (2004) reported
severely limited growth of Bahiagrass when biosolids with high pH (>12) values were applied to
the Immokalee soil at N-based rates.
TP concentrations of biosolids (Table 3-3) ranged from 7.9 to 33 g kg-l (0.79 to 3.3%),
typical of TP concentrations of biosolids produced conventionally nationwide (~20 to 40 g kg- ),
and > 40 g kg-l for BPR materials (USEPA, 1995). The difference in TP values results from the
differences in wastewater characteristics and treatment practices (Brandt et al., 2004). TP
concentrations were not used to include/exclude materials, but TP concentration was an
important measurement, as the laboratory determined TP values were used to calculate the mass
of biosolids needed to attain selected P application rates.
TP values for the materials were determined using 3 methods: Andersen (Andersen,
1976), EPA 3050A (USEPA, 1995) and perchloric acid (Association of Official Analytical
Chemists, 1990; The Fertilizer Institute, 1982) (Table 3-4). The Andersen and EPA 3050A
analyses were performed at UF. The perchloric acid analysis was conducted at an independent
laboratory for additional TP verification. The ability to recover P from the biosolids differs
among the techniques, which is reflected in the variability of the reported TP concentrations. UF
researchers were not provided information on how the individual WWTPs analyzed materials for
TP, which could explain the variability between determined and producer-supplied P values. A
standard reference biosolids material (National Institute of Standards and Technology, Standard
2781: Domestic Sludge) was used to gauge the ability of digestion methods to recover P. All
three methods had excellent P recovery percentages (88-96%).
Researchers chose to use the TP values determined by the Andersen (1976) method for
this study, primarily because of its ease of use. The EPA 3050A method is tedious and time
consuming, making the method impractical for the large number of samples involved in this
study. The perchloric acid analysis requires special facilities to handle the explosive nature of
perchloric acid, which were not available in the UF laboratory. In contrast, the Andersen (1976)
and EPA 3050A methods use hydrochloric and nitric acid, respectively, which are much safer to
handle. In a preliminary experiment, the Andersen-TP concentrations were used to calculate the
quantity of biosolids necessary to reach a target P rate of 224 kg P ha- A small quantity (200g)
of Immokalee soil was amended with the calculated masses of biosolids. When the biosolids-
amended soils were analyzed for TP using the Andersen (1976) method, the target P rate was
reached, indicating that the TP values obtained from the Andersen (1976) method were adequate
measures of the total P in the biosolids.
Total Fe, Al, and Ca concentrations were measured on the Andersen-TP extracts via ICP
(Table 3-2). Total metal concentrations were representative of biosolids produced nationally and
reflected the individual biosolids treatment processes. Total Fe ranged from 2.2 to 60 g kg-l (0.2
to 6%), Al ranged from 1.7 to 24 g kg-l (0.17 to 2.4%) and Ca ranged from 13 to 310 g kg-l (1.3
to 31%). The Lakeland Glendale material is lime stabilized, which is reflected in the high total
Ca (310 g kg- ). The greater a residual's Fe and Al content, the better the material is able to
retain P, leading to lower phytoavailability and lower leaching risks (Elliott et al., 2002). As
discussed below, Milorganite (41 g kg-l Fe, 2.7 g kg-l Al) has relatively high amounts of Fe and
Al, and resulted in decreased P release and P phytoavailability.
Water-extractable P (WEP) ranged from 0.04 to 14 g kg-l (Table 3-3). Brandt et al.
(2004) demonstrated that WEP is a good measure of the environmentally relevant portion of P in
biosolids. WEP is used to calculate percent water-extractable P (PWEP = WEP/ TP*100). Most
of the non-BPR materials analyzed had PWEP values < 5%, which is typical of non-BPR
biosolids produced in the U.S. (Brandt et al., 2004). BPR materials, (Boca Raton, OCUD E cake
and dry) had PWEP values > 11%. Brandt et al. (2004) determined the average PWEP for
various BPR biosolids sampled nationwide to be > 14%. PWEP was high for both Lakeland NS
(47%) and GRU (26%). These materials are not produced via BPR processes, but personal
communication with plant operators suggests that P removal is likely occurring in both systems.
We have categorized these materials as "BPR-like".
Both the Lakeland NS and GRU materials are low percent solids (3 and 5%, respectively,
Table 3-2), and we decided to analyze the liquid and solid phases of both materials to determine
the concentration of TP. Phosphorus in the liquid phase would be expected to be highly labile,
and we suspected that the high PWEP values were due in part to high concentrations of P in the
liquid phase. To determine the TP concentrations in the separated liquid and solid phases of each
residual, we centrifuged sub-samples at 12,000 rpm for 15 minutes to separate the liquid phase
from the solid phase. The separated phases were then analyzed for TP, the solid phase via the
Andersen (1976) method, and the liquid phase via persulfate digestion (USEPA, 1993). Analysis
showed that 50% and 20% of TP was in the liquid phase for the Lakeland NS and GRU
materials, respectively. The smaller concentration of P in the liquid phase for GRU than
Lakeland NS is reflected in the smaller PWEP value for GRU (26%) versus Lakeland NS (47%).
Disney is a BPR biosolids, however the PWEP (8.4%) is below average for a BPR
residual. The Disney material is a composted mixture of biosolids, food, and yard waste. Brandt
et al. (2004) reported that composting an anaerobically digested cake decreased WEP by 10-fold.
The yard waste itself could decrease PWEP; the Disney compost was sieved to pass through a 2
mm sieve prior to analysis, however small pieces of wood were obvious. The yard waste
composted with the Disney biosolids decreases the mass of biosolids per kg of land applied
finished product (composted biosolids + yard waste, even after sieving), thus reducing labile P.
Oxalate-extractable Fe, Al, and P concentrations (Table 3-3) were used to calculate PSI
values for the materials. PSI can be used a priori to gauge the liability of P in many biosolids in
sandy soils (Elliott et al., 2002). PSI relates the moles of oxalate-extractable P to the moles of
extractable Fe and Al (PSI = moles P / [moles Fe + moles Al]). Biosolids with high
concentrations of Fe and Al tend to have less labile P. PSI has no meaning for biosolids whose P
chemistry is controlled by Ca (lime stabilized materials). Elliott et al. (2002) suggested a critical
PSI value of 1.1 for non-lime stabilized materials. That is, if the PSI of a material exceeds the
critical value, appreciable P leaching may occur from amended, sandy soils with limited P
retention capacity. Biosolids with PSI values I 1.1 resulted in minimal P leaching in the Elliott et
al. (2002) study. Most of the samples analyzed in the current study were BPR or BPR-like
materials, and the PSI values exceed the critical value proposed by Elliott et al. (2002). We
would expect significant leaching when these materials were land applied to Immokalee soil that
retains P poorly.
Dynamic Laboratory Incubation
Eleven biosolids and TSP were individually mixed with 13 cm (400 g) samples of
Immokalee soil for the dynamic incubation study. Soil columns were leached a total of 8 times
(~4 total PV) over 5.5 months. The first 7 leachings were conducted bi-weekly to attain ~60 mL
(1/2 PV) of drainage each time. Significant P release (leaching) ceased after 5 leachings (2.5
months). A time-series analysis (SAS Institute, 1989) showed no difference in cumulative P
released for leachings 5-7. We waited 2 additional months to conduct the final leaching to allow
for P dissolution and distribution through the individual columns of soil and to confirm P release
from all materials was maximized. Indeed, P loss from all materials during leaching 8 was
minimal (< Img). Data from the final leaching was added to the time-series analysis and, again,
showed no change in cumulative P released between leachings 5-8. Three months passed
between leachings 5 and 8. Because cumulative P release did not change over the 3-month time
frame, we assumed that P release from the materials had ceased, and the experiment was
terminated. SRP in leachate was taken to represent total P released, as SRP constituted the
maj ority (>85% of TP) of P lost to leaching. To calculate P released as a percentage of applied P,
the mean mass of P released from the control treatment was subtracted from the mass of P
released for all other treatments, and the difference divided by the appropriate mass of P
One obj ective of the study was to quantify water soluble P (mass of P released) from
various biosolids representing a range of chemical characteristics. Biosolids with lower
quantities of water soluble P are less likely to negatively affect the environment. The quantity of
water soluble P is governed by several factors, including biosolids treatment processes
[especially heat drying (Smith et al., 2002; Maguire et al., 2001)], and biosolids chemical
composition (especially the quantity of Fe and Al oxides in the material). Residuals high in Fe
and Al oxides retain P, and result in less P leaching (Brandt et al., 2004). Thus, we expected
significantly more P release from BPR biosolids low in Fe and Al oxides than thermally dried
residuals high in Fe and Al.
ANOVA showed significant source (P-source), rate (P-rate) and source by rate
interactions. As a result, comparisons of cumulative P released and cumulative P leached as a
percent of applied P are discussed within each P application rate and not across P application
rates. Cumulative P release was greatest from BPR and BPR-like materials (Lakeland NS, GRU,
Boca Raton, and OCUD E dry and cake) (Table 3-5). Figures 3-la and b show cumulative P
release (percentage of applied P) in bar graph form for the 56 and 224 kg P hal rates,
respectively. P release from Lakeland NS, GRU, Boca Raton, and OCUD E cake biosolids was
equal to TSP at both the 56 and 224 kg P ha-l application rates (Figures 3-la and b). The OCUD
E dry material was produced by thermally drying the OCUD E cake material. The OCUD E cake
is undigested biosolids and is not lawfully land applied, but was included for scientific interest.
Heat drying the OCUD E cake created a class AA material that may be land applied. Heat drying
typically decreases labile P (Smith et al., 2002), and P release was less for the OCUD E dry
biosolids than for the OCUD E cake biosolids. However, P release from the OCUD E dry
biosolids at the 56 kg P ha-l application rate was equal to that from TSP, indicating the OCUD E
dry biosolids still had high quantities of labile P.
P release from the Disney biosolids was less than TSP and most BPR and BPR-like
biosolids at both the 56 and 224 kg P ha-l rates. At the 56 kg P ha-l rate, P release from Disney
was equal to GRU. As previously mentioned, Disney is a composted mix of biosolids, yard, and
food waste, which results in a PWEP (8.4%) less than the other BPR and BPR-like biosolids (>
15%) in this study. The PSI (0.45) of Disney reflects relatively large quantities of Fe and Al,
which likely reduced P release.
At 56 kg P ha- P release from Milorganite and GreenEdge was significantly lower than
from TSP. Both Milorganite and GreenEdge are thermally dried. Thermal drying typically
lowers quantities of labile P. Milorganite also has relatively high total Fe + Al concentrations (44
g kg- ) compared to typical biosolids total Fe + Al concentrations (20 g kg- ), which we expected
to further reduce P liability. P release from Milorganite and GreenEdge was not significantly
different, suggesting that P release from Milorganite and GreenEdge could be controlled by
physical properties (pellet dissolution). When the laboratory incubation was dismantled, pellets
of Milorganite and GreenEdge were still apparent, suggesting that even though the columns were
near field capacity for the duration of the experiment, the length of time was not sufficient for
complete pellet dissolution. Milorganite Greens Grade, which is the same product as
Milorganite, but passed through a smaller sieve, has a slightly greater PWEP value (0.85%) than
Milorganite (0.58%). We delayed the final leaching for 2 months following leaching 7 to allow
dissolved P to diffuse through soil pores (a time limited reaction), and the columns of soil were
maintained near field capacity during the 2-month break. Despite the elapsed time and sufficient
moisture, P release from all materials was < 1 mg in leaching 8. Time series analysis also
indicated that P release was minimal over the last 3-month period (leachings 5-8), suggesting
that P release from the P sources was maximized.
We expected GreenEdge to release a greater quantity of P than Milorganite based on the
biosolids PSI values, total Fe and Al concentrations, and PWEP values. GreenEdge has greater
PSI (1.0) and PWEP (1.1%) values than Milorganite, (PWEP: 0.58%, PSI: 0.55) and Milorganite
has greater Fe and Al (44 g kg- ) concentrations than GreenEdge (Fe + Al = 23 g kg- ). Based on
the P chemistry, we expected a greater quantity of P release from GreenEdge, however the data
did not support our a priori assumption. We hypothesized that calcium was influencing P
liability, however the Ca concentration of GreenEdge is not great enough to suggest that Ca is
controlling P solubility. We then ground a sample of GreenEdge with a mortar and pestle and
determined PWEP on the finely ground biosolids. The PWEP of GreenEdge did not increase
after grinding, (1.3%, compared to 1.1% on un-ground biosolids) indicating that pellet size was
not influencing P solubility. We currently have no explanation for the equal quantities ofP
released from GreenEdge and Milorganite. Butkes et al. (1998) studied P retention of water
treatment residuals (WTRs) and suggested that cationic polymers added during the water
treatment processes can retain P. During the dewatering of GreenEdge, a polymer (Ciba Zetag
8849FS) is added, and it is possible that the dewatering polymer has the capacity to sorb P, thus
reducing P liability.
Cumulative mass of P released from the OCUD S material was equal to TSP at the 56 kg
P ha-l application rate and greater than TSP at the 224 kg P ha-l rate (Figures 3-la and b). The
OCUD S material is a conventionally produced biosolids; however, the PWEP (21%) and the
PSI (2.9) values of the OCUD S material are high, and consistent with the high quantity of P
leached. Given the treatment process of the OCUD S biosolids anaerobicallyy digested) we
would not expect PWEP and PSI values to be so high. However the total Fe + Al concentrations
of the OCUD S material are low (11Ig kg- ), resulting in the high PSI value and quantities of P
released equal to TSP.
As previously mentioned, biosolids-TP is not a good indication of P liability. However,
PWEP can be used to gauge the environmental impact a residual will have once land-applied. A
logarithmic relationship exists between cumulative P released (% of applied P) and biosolids-
PWEP at both rates (Figures 3-2a and b). The correlation was similar for the 56 kg P ha-l rate (r2
= 0.65) and the 224 kg P ha-l application rate (r2 = 0.69). The correlation between cumulative P
released (% of applied P) and biosolids-PWEP was not strong enough to indicate that PWEP can
be used to predict the amount of P release that will occur when a biosolids is land applied.
However, Brandt et al. (2004) demonstrated that PWEP is a superior measure of the
environmentally relevant portion of P in biosolids and manures than biosolids-TP or soil test P.
Biosolids-PWEP can be used a priori to gauge the potential of a residual to negatively affect the
environment. PWEP is a measure of the water-soluble P in biosolids, thus materials with PWEP
values > 14% (vertical line in Figures 3-2a and b) should be assumed to have a larger negative
environmental impact than biosolids with PWEP values < 14%. Recall that Brandt et al. (2004)
reported an average PWEP value for BPR biosolids sampled nationwide to be > 14%. Given the
ease with which WEP/PWEP can be determined, PWEP can be used as a quick and efficient
gauge of a residual's potential to negatively impact the environment once land applied. Elliott et
al. (2006) used source WEP to improve P source coefficient assignments (PSC = 0. 102 x
WEP0.99) for state P-indices, and runoff dissolved P was well correlated (r2 = 0.80) with source-
Figures 3-3a and b show P released (% of applied P) as a function of PSI for the 56 and
224 kg P ha-l rates, respectively. Elliott et al. (2002) demonstrated that for biosolids with PSI
values < 1.1, no appreciable leaching occurred from another sample of the Immokalee soil
amended with biosolids. Several materials used in this study had PSI values above the suggested
critical point of 1.1, which would portend significant P losses measured from the BPR and BPR-
like materials. Data from the current study indicate that a critical PSI value of ~2.0 better
separates biosolids where leaching was greatest, suggesting that the critical PSI value be raised
from 1.1 to ~2.0. However, the Elliott et al. (2002) study differed from the current study in 2
ways: 1) the Elliott et al. (2002) study utilized much larger soil columns, where '/ of a pore
volume was 500 mL, and 4 monthly leachings were necessary to reach 1 pore volume of
drainage; and 2) Bahiagrass was grown in the Elliott et al. (2002) soil columns, reducing the
quantity of P available for leaching. Data published separately (O'Connor et al., 2004) showed
that Bahiagrass took up 29-57% of applied P at the 56 kg P ha-l application rate. In the dynamic
laboratory incubation, ~4 pore volumes of drainage were collected versus 1 pore volume of
leachate collected in the Elliott et al. (2002) study. The small columns in the current study
included no plants and were flushed with more pore volumes of water than the large glasshouse
columns, which would encourage P release and movement downward (and out) of the small
columns. Given the differences in experimental design between the current study and Elliott et
al. (2002) study, raising the critical value from 1.1 to ~2 is probably not justified. While there is
no clear mathematical relationship between cumulative P release and PSI, it is evident that P
release increases when PSI > 2.0. P release appears to be maximized (~80% of applied P) at PSI
values > 2.0
The second maj or obj ective of dynamic laboratory study was to examine the kinetics of P
release. Figures 3-4a and b show cumulative P released as a percentage of P applied per leaching
event for the 56 and 224 kg P ha-l rates, respectively. While the ultimate quantity of P released is
important, the rate at which P is released is also important. When large amounts of P are released
from a material quickly, more P is in the soil solution at any given time and is subj ect to leaching
in periodic rain events, increasing the risk of ground and surface water impairment. Slow rates of
P release provide plants more time to take up biosolids-P, decreasing the amount of P subject to
loss from the soil profile.
Materials with high PWEP values (OCUD E cake, Lakeland NS, GRU, OCUD S)
released large quantities ofP quickly (within the first 3 leachings). At the 224 kg P ha-l
application rate (Figure 3-4b), the thermally dried materials (Milorganite, GreenEdge, and
Tallahassee) and the conventionally produced Broward cake released < 7% of applied P during
the first leaching event and the remaining materials released >15% of applied P. The quantity of
P released from Milorganite and GreenEdge was ~55% less than P released by the highly-P
soluble materials during the first leaching event. Phosphorus release from the cake and slurry
materials began to decrease (slow) in leaching 4, but P release from Milorganite and GreenEdge
continued to increase through leaching 5. As discussed above, high rates of P release from BPR
and BPR-like materials increase the quantity of P in the soil solution at any given point in time.
The increase in soil solution P suggests that BPR and BPR-like biosolids could be excellent
fertilizers, but also makes BPR and BPR-like residuals more likely to negatively affect the
environment, when soil solution P is washed through the soil profie into ground and surface
We can roughly convert total PVs leached to residence time in a Hield setting. One PV of
leachate represents ~5.64 cm drainage. Cumulative leaching (~480 mL) represented 22.6 cm of
drainage. Yearly rainfall in south Florida averages ~140 cm yrl (Obeysekera et al., 2004) and
evapotranspiration is ~70% of rainfall (Nachabe et al., 2005). Subtracting evapotranspiration
from rainfall (140 cm yr- 100 cm yr- ) yields 40 cm yr- drainage. Assuming 40 cm drainage
yr- we can calculate the number of PVs leached per year (40 cm yr- /5.64 cm PV-1 = 7.09 PV yr
1). Thus, the 4 PVs leached in this study equate to ~7 months in the Hield. This study represents
an extreme case of biosolids land application where the soil used had minimal P sorbing capacity
and no plants were grown to take up applied P. This experiment was key to understanding
biosolids P leaching characteristics in a representative "worst-case" scenario.
A glasshouse study was run concurrently with the laboratory leaching study. Seven
biosolids were used and TSP was included as a reference. Biosolids were mixed with 13 cm (4
kg) samples of Immokalee soil and placed on top of 30 cm of base sand. Two biosolids evaluated
had low PWEP (<1.1%) values: Milorganite, and GreenEdge, 1 biosolids had a moderate PWEP
(Disney, 8.4%), and 4 biosolids had high PWEP (215%) values: Lakeland NS, OCUD S, GRU,
Disney, and Boca Raton. Bahiagrass was grown and harvested monthly for 4 months. After each
harvest, columns were leached to attain ~500 mL (1/2 pore volume) drainage. Statistical analysis
(ANOVA) showed significant treatment (biosolids), rate (P-rate), and rate X treatment effects on
yield, P uptake, and P leaching (SAS Institute, 1999; Tables A-1, 2, and 3, respectively). Thus,
results are discussed within each P application rate and not across P application rates.
We provided excessive N and other macronutrients (S, K, Mg) to isolate P as the only
nutrient variable. The lowest rate of P application (56 kg P ha- ) provides more than adequate P
for Bahiagrass. As a result, yields (Appendix Table A-4) were expected to be equal across all
materials and application rates. Equal yields are critical since harvest yields are used to calculate
P uptake. At the 56 kg P ha-l application rate, there were no significant differences in cumulative
yield between biosolids source treatments, and all treatment yields were significantly greater
than the control (Figure 3-5a). At the 112 kg P ha-l application rate, all treatment yields were
greater than the control, but the cumulative yield for OCUD S biosolids was greater than yields
for TSP, Disney, and GreenEdge (Figure 3-5b). At the 224 kg P ha-l application rate (Figure 3-
Sc), all the yields for all materials except for Lakeland NS and Disney were different from the
control. The cumulative yield for the Lakeland NS material was smaller than yields for all other
materials with the exception of the Disney compost. Yields at the 224 kg P ha-l application rate
for Bahiagrass grown in soil amended with Lakeland NS were highly variable, and Bahiagrass
growth was inexplicably reduced in 2 of the 4 soil columns. At the high P rate, the cumulative
yield of the OCUD S biosolids was not different from the yields of the other materials (Figure 3-
Sc). Because ANOVA showed rate and treatment interactions affecting yield, we cannot
statistically compare yields for treatments across rates. Figure 3-6 shows that most cumulative
yields approached the average yields across treatments (except controls) (~28 mg; horizontal line
on Figure 3-6).
To validate that Bahiagrass N uptake was sufficient, we analyzed tissue from all four
harvests at the highest rate of P application (Figure 3-7). We analyzed only tissue from the 224
kg P ha-l rate because most of the N at this rate was assumed to come from N mineralization
from the biosolids. If we misjudged the N mineralization rate to be 40%, we would anticipate N
deficiencies to be more apparent at the 224 kg P ha-l rate. Recall that N was equalized across P
application rates and treatments using ammonium nitrate. More ammonium nitrate was needed at
the 56 and 112 kg P ha-l rates than at the 224 kg P ha-l rate. The ammonium nitrate would be
immediately available to the Bahiagrass, thus we would not expect N deficiencies due to slower
(or smaller) than expected N mineralization from the biosolids. Tissue N concentration was used
to calculate yield-weighted N concentrations for each treatment over the entire growing season
using eqution 3-1.
trt = treatment
H1 = harvest
H12 = harvest 2
trt yieldH1 x NH1 + trt yieldH2 x NH2 + trt yieldH2 x NH2 trt yieldH2 x NH2 (3-1)
cumulative trt yield
Measured N concentrations were sufficient for grazing beef cattle (minimum 1.12 g N g-
1; NRC, 1996). Based on the N content per harvest (Figure 3-7), it appears we underestimated the
quantity of N that would mineralize from the biosolids initially (first harvest), and overestimated
N mineralization throughout the rest of the growing season. Figure 3-8 shows that yield-
weighted N concentrations for all biosolids treatments were above the minimum N
concentrations required for beef cattle. The yield-weighted tissue N concentrations were
sufficient for us to believe N was not affecting yields. Micronutrients were also considered as the
source of yield variation. Micronutrients were not supplied during the growing season, however
research has shown micronutrient deficiencies in Bahiagrass are rare and do not affect yields
(Chambliss and Adjei, 2006).
P Uptake and Relative Phytoavailability
The 224 kg P ha-l rate of P application resulted in the greatest cumulative P uptake
(Figures 3-9a c; Appendix Table A-2). P uptake ranged from 23-52 % of P applied at the low
application rate (Figure 3-9a). In a similar study by O'Connor et al. (2004), P uptake ranged
from 29-57% of applied P at the same low P application rate. At the 224 kg P ha-l rate, P uptake
in the current study ranged from 7-28% (Figure 3-9c), whereas O'Connor et al. (2004) reported P
uptake ranging from 1 1-29% of applied P at the 224 kg P ha-l rate. Recall that yields for
Bahiagrass grown in two replicates of soil amended with the Lakeland NS biosolids were
reduced at the highest rate of P application; therefore, average P uptake at this rate was lowered.
Cumulative yield-weighted tissue P concentrations ranged from 0.5 g kg-l to 4.5 g kg-l (Table 3-
7). Bahiagrass grown in soil with no added P (control) accumulated the least P, and soil amended
with TSP at 224 kg P ha-l resulted in the highest yield-weighted P concentration. Adj ei and
Rechcigl (2002) suggested that a sufficient tissue P concentration for Bahiagrass is 2.0 g kg-l
However, the critical concentration (P concentration necessary for survival) for Bahiagrass is
likely lower, and may even be closer to 1.0 g kgl (personal communication, Dr. Jerry Sartain,
2007). Milorganite applied at the 56 and 112 kg P ha-l resulted in tissue P concentrations slightly
below the hypothesized critical value of 1.0 g kg l. However, cumulative yields of Milorganite
and GreenEdge treatments were not different from yields in the fertilizer-P (TSP) treatment
(Figures 3-5a-c). Thus, the low tissue P concentrations in Milorganite and GreenEdge treatments
did not limit above ground growth. Bahiagrass grown in control columns (no added P) yielded
less than all other treatments. Over the course of the 4-month growing season, no above ground P
deficiencies were noted for any treatment. Plant roots were not examined for P-defieiency as the
glasshouse study is ongoing.
Relative phytoavailability (RPP) was estimated using a slope-ratio approach. Cumulative
P uptake was plotted as a function of P applied for each biosolids (Figure 3-10). Linear
regressions were fit to the data for each P source with a common intercept of 8.48 mg (mean
cumulative P uptake for the control columns). To calculate RPP, the slope of the regression line
for each material was divided by the slope of the regression line for TSP (Table 3-7). O'Connor
et al. (2004) used the same approach to estimate RPP for 12 biosolids on the Immokalee soil.
Table 3-7 also lists RPP values for biosolids produced or marketed in Florida determined by
O'Connor et al. (2004). We attempted to fit both linear and 2nd degree polynomials to the data
from the current study. The polynomials resulted in slightly greater r2 ValUeS, but greater
coefficients of variability (CV) (Table 3-8). Comparison of polynomials by comparing slopes is
also difficult, leading us to choose linear regressions to determine RPP. While a linear regression
through only the 56 and 112 kg P hal rates had slightly higher r2 and lower CV values than
linear regressions through all 3 rates, r2 and CV values for regressions including all 3 rates were
acceptable and significant at p < 0.05 (Table 3-8). Recall the purpose of including 3 rates of P
application was to better define the RPP curve; therefore, we opted to include all rates for
O'Connor et al. (2004) proposed 3 groupings of relative phytoavailability: high (>75% of
TSP), moderate (25-75% of TSP), and low (<25% of TSP). With the exceptions of Disney, all of
the BPR and BPR-like materials fit into the high RPP category. The Disney biosolids is a
composted mixture of BPR biosolids and yard waste, with PWEP (8.4%) and PSI (0.43) values
less than the PWEP and PSI of the other BPR biosolids included in the glasshouse study. The
PSI value for the Disney biosolids indicates a greater quantity of Fe and Al than the BPR or
BPR-like biosolids. Stratful et al. (1999) reported that BPR biosolids contained more
phytoavailable P compared to conventionally produced residuals.
The OCUD S material was 110% as phytoavailable as TSP. Interestingly, the OCUD S
material is not a BPR material, but has a high PWEP (21%), exceeding to the average PWEP
value (14%), for BPR materials (Brandt et al., 2004). To calculate the regression slope for
Lakeland NS biosolids, 2 P uptake replicates were excluded at the 224 kg P ha-l rate. Recall that
Bahiagrass yields were inexplicably reduced in 2 replicates of the Lakeland NS amended soil at
the high P rate, reducing P uptake. Excluding data points reduces power of statistical
comparison; however when cumulative yields from all 4 replicates were used for regression, the
linear correlation was very poor (r2 = 0.09). Thus, the regression coefficient and r2 ValUeS foT
Lakeland NS listed in Table 3-7 reflect only 2 replications.
The linear regression model described the data for Milorganite poorly (r2 = 0.51). We,
therefore, estimated RPP values for Milorganite via point estimates (Table 3-9) calculated using
Point Estimate RPP = (P uptakesou~rc P uptakecontroi) / P appliedsource (3 -2)
(P uptakeTrSP- P uptakecontrol) / P appliedTrSP
At the 56, 112, and 224 kg P ha-l rates, P in Milorganite was 38%, 31%, and 23% as
phytoavailable as TSP, respectively. Thus, on average, Milorganite is ~31% as phytoavailable as
TSP, placing it in the lower end of the moderate category of RPP values proposed by O'Connor
et al., (2004). Milorganite would be assumed a priori to be in the low category of RPP values,
based on PWEP (0.58%) and PSI (0.55) values, however the data support a moderate RPP for
Milorganite. We also fit the P uptake data for Milorganite to a plateau model (Equation 3-2,
a = concentration of P in control (no added P)
b = P concentration plateau increment above the background (mg)
c = rate constant for change in uptake for a given change in the cumulative amount of P applied
P applied = rate of P application
Uptake (mg) = a + b(1-e-c*Papplied) (3-2; Barbarick et al., 1995)
The good fit of the data to the plateau model confirms the unique P characteristics of
Milorganite. We can calculate a similar RPP for Milorganite by dividing the mean cumulative P
uptake for Milorganite (30.9 mg) by the mean cumulative P uptake for TSP (106 mg) at the 224
kg P ha-l application rate: 30.9 mg/ 106 mg *100 = 29% RPP. Even at the high P loads (224 kg P
ha- ) associated with N-based application rates, Milorganite is only ~29% as phytoavailable as
TSP (consistent with point estimate).
The regression equation for the Lakeland NS uptake data resulted in a weak r2 ValUe
(0.54), so we confirmed the RPP calculated via the slope-ratio approach using point estimates
(Table 3-9). Averaging the point estimates for 56, 112, and 224 kg P ha-l resulted in a RPP of
90% for the Lakeland NS biosolids (consistent with slope-ratio approach).
We attempted to correlate RPP with biosolids-PWEP, but the correlation did not
adequately explain the relationship between RPP and biosolids-PWEP (r = 0.64). However,
when RPP is plotted as a function of biosolids-PWEP, (Figure 3-12) a logarithmic relationship
exists (r2 = 0.79). The logarithmic correlation is not strong enough to predict RPP using PWEP.
However, it appears that biosolids-PWEP > 14% segregates biosolids with greater RPP values
from biosolids with lesser RPP values. Recall the same trend was noted with the P release data
from the dynamic laboratory incubation; biosolids with PWEP values > 14% could be assumed
to result in a larger negative environmental impact than biosolids with PWEP values < 14%. In
the glasshouse study, biosolids with PWEP values < 14% had RPP values below 60% (moderate
RPP category), compared to biosolids with PWEP values > 14%, which had RPP values > 90%
(high RPP category).
The glasshouse study conducted here, and earlier work by O'Connor et al. (2004),
demonstrates the wide range of RPP values possible for various biosolids. Knowledge of
biosolids-RPP is key to adjusting application rates, should P-based limitations on biosolids land
application be imposed. Results of the glasshouse study indicate that no change in biosolids
application rate is needed (or justified) for BPR and BPR-like biosolids with RPP values in the
high RPP category. However, application rates for Disney and GreenEdge could be
approximately doubled that necessary to meet crop P needs. Based on the plateau model,
Milorganite could be applied at rates > 224 kg P ha l, and will only be about 1/3 as available as
TSP. The low RPP of Milorganite does not necessarily imply that Milorganite is a poor P-
fertilizer; Bahiagrass grown in Milorganite-amended soil showed no signs of above ground P
deficiency during the course of the experiment and Bahiagrass yields were equivalent to TSP
treatment yields for each application rate.
A similar glasshouse study by Elliott et al. (2002) used a base sand with minimal P-
sorbing capacity. The base sand used in this study had moderate P-sorbing capacity (RPA =
8.6%). Elliott et al. (2002) showed that a Florida sand with even moderate P-sorbing capacity
(RPA=15.3%) can limit P leaching. Minimal P leaching occurred in the current glasshouse. Only
TSP and BPR materials applied at the highest rate resulted in significant P leaching (Appendix
Table A-3). The base sand is likely retarding the advance of P through the soil column. This
study is on going, and it is possible that P retained in the column through the first growing
season will eventually leach during subsequent croppings. Given the deep rooting of Bahiagrass,
the P sorbed to the base sand is also accessible for plant uptake.
OC C 0 0
"C~ E C0O 0C
r=l o L
00\01A N IC 0\CC40 4
10 A s30
O er O O vO v
O\ iN v,3 j I Cb
NM Ne NNr e c i c
0\ O0ccr ccrOOMc
iD ero 0 bab er0 OOMe
OW0M dM C~333j C3 0
N ++ CMMM333 C)
CCImo 0o 0\0 C-I)O \ I~ \0
0\W ~ o m Nv~ i v~i ~j C C
O O O N ON C
d o Oo i i O mo a n
ccrcrb c cc
CCI 3 C~ 3
O E o
ICCIme-i a CC er0 a
MM MM N mw M IAm N ~O~OI 0i \30 0j D~0 C
co coo 0
coovemoo\co o a r
vc~ v, bb O\ v, 3
3 d 0\
O O O
O O O o o
O b @ O m m cc, Oc c
NNW MNN NY N
-co O\eem R -
NN N N N N N ICN ~ \C)3j C C C
a C,, &&& o
Q, on = = a
seco o 3 5 %& # o d
$$mA QUU OO QEE Ha i
Table 3-4. Comparison of biosolids-TP concentrations determined
by 3 digestion methods
OCUD E Cake
aProducer-supplied data. bND: not determined.
Table 3-5. Dynamic laboratory incubation: cumulative net
SRP released and SRP released as a percentage
of applied P. Means & 1 standard error, n = 3.
% of applied
80.9 & 3.08
74.0 & 2.69
38.5 A 10.8
19.0 & 3.73
40.0 & 4.29
27.4 & 2.04
42.0 & 1.34
39.9 & 1.80
67.0 & 2.32
68.5 & 2.63
87.7 & 4.90
79.3 A 1.84
90.0 & 2.21
62.4 & 1.87
92.8 & 2.15
79.9 & 1.14
69.5 & 2.05
62.4 & 1.05
97.3 & 3.37
80.0 + 0.667
61.4 & 7.31
31.5 & 2.08
52.8 & 1.34
40.3 A 1.80
0.67 & 0.062
8.33 & 0.318
30.5 A 1.11
3.96 & 1.11
7.84 & 1.54
4.12 & 0.442
11.3 & 0.842
5.66 & 0.252
21.6 & 2.42
6.90 + 0.239
28.2 & 1.08
9.03 & 0.505
32.7 & 0.756
9.27 & 0.228
25.7 & 0.770
9.56 & 0.347
32.9 & 0.275
7.16 & 0.221
25.7 & 0.469
10.0 + 0.212
33.0 + 0.434
6.33 & 0.753
13.0 + 0.856
5.43 & 0.138
16.6 & 0.742
OCUD E Cake
OCUD E Cake
OCUD E Dry
OCUD E Dry
kg P ha '
d'' .i": "
si"qB r' rArg, oC~
Figure 3-1. Cumulative SRP released (% of applied P) as a function of P source. Means capped
with the same letter are not different (Tukey's Test, p<0.05). A) P-rate: 56 kg P ha-1.
B) P-Rate: 224 kg P ha-1.
30 -1 P
y = 10.4 Ln(x) + 45.4
CV = 18.7%
0 10 20 30 40 50 60 70 80 90
y = 11.1 Ln(x) + 30.7
CV = 21.5%
0 10 20 30 40 50
60 70 80 90
Figure 3-2. Dynamic laboratory incubation: cumulative P released (% of applied P) as a function
of biosolids-PWEP. Error bars represent 1 standard error. Vertical line at biosolids-
PWEP 14% indicates when increased negative environmental impact may occur.
A) P-Rate: 56 kg P ha- B) P-Rate: 224 kg P ha-l
1.0 1.5 2.0 2.5 3.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5
Figure 3-3. Dynamic laboratory incubation: cumulative P released as a function of PSI. Dashed
vertical line at 1.1 PSI represents change point proposed by Elliott et al. (2002). Solid
vertical line at 2.0 PSI represents increased critical PSI value suggested by data in the
current study. A) P-Rate: 56 kg P ha- B) P-Rate: 224 kg P ha- .
* I ~
I 1 III~
0 00 00 0 ~D v \
0 8 OS
' l l I I~~
00 0 0 0
0 a mm
fi i *=<0
0 0 0
0 0 0
Figure 3-5. Cumulative Bahiagrass yields after 4 harvests. P-rate: 112 kg P ha l. Means capped
with the same letter are not different (Tukey Test, p<0.05). A) P-Rate: 56 kg P ha-l
B) P-Rate: 112 kg P ha- C) P-Rate: 224 kg P ha-l
Figure 3-5. Continued
30 T ... T "-
~a II II 1 I I II II-I-F[ IO Boca Raton
O ~Lakeland NS
2 OCUD S
56 112 224
P Applied (kg ha l)
Figure 3-6. Cumulative Bahiagrass yields after 4 harvests for all P application rates. Horizontal
line at ~28 g dry matter yield represents average dry matter yield across rates and
treatments. Error bars represent 1 standard error.
O Harvest 1
El Harvest 2
5 Harvest 3
5 Harvest 4
Figure 3-7. Bahiagrass N tissue concentration per harvest. P-rate: 224 kg P ha l. Error bars
represent 1 standard error.
OCUD S Greenedge
GRU IVilorganite Lakeland
TSP Boca Raton Disney
Figure 3-8. Bahiagrass yield-weighted N concentration after 4 harvests. P-rate: 224 kg P ha .
Error bars represent 1 standard error.
Milorganite Disney Green Edge Boca Raton
T SP OCUD S Lakeland NS
GRU Boca Raton
Milorganite Disney GreenEdge Lakeland NS Boca Raton
TSP OCUD S
112 kg P ha- Means
p<0.05). A) P-Rate: 56 kg P
Figure 3-9. Cumulative Bahiagrass P uptake after 4 harvests. P-rate =
capped with the same letter are not different (Tukey Test,
ha- B) P-Rate: 112 kg P ha- C) P-Rate: 224 kg P ha- .
Table 3-6. Cumulative Bahiagrass yield-weighted tissue P concentrations after four
0.531 & 0.06
1.86 & 0.09
2.71 & 0.21
4.20 + 0.40
0.91 & 0.06
0.97 & 0.10
1.18 & 0.08
1.07 & 0.17
1.39 & 0.7
1.90 + 0.03
1.43 & 0.11
1.90 + 0.05
2.50 + 0.10
1.65 & 0.13
2.25 & 0.09
3.22 & 0.26
1.59 & 0.16
2.35 & 0.24
3.16 & 0.49
2.45 & 0.04
3.50 + 0.21
1.71 & 0.09
2.41 & 0.14
3.42 & 0.16
Table 3-7. R phosphorus phytoavailability of biosolids to TSP
P Source r2 COefflcient RPP (%) Category Reference Study
TSP 0.89 0.472 100 high Current
OCUD S 0.90 0.519 110 high Current
GRU 0.91 0.439 93 high Current
Lakeland NS 0.54 0.435 92 high Current
Boca Raton 0.86 0.429 91 high Current
GreenEdge 0.88 0.222 47 moderate Current
Disney 0.70 0.220 47 moderate Current
Regression equation: y = x + 8.48 Current
Milorganite 31* moderate Current
TSP 0.400 100 O'Connor et al. (2004)
Largo Cake 0.297 74 moderate O'Connor et al. (2004)
Largo Pellets 0.193 48 moderate O'Connor et al. (2004)
Baltimore Cake 0.136 34 moderate O'Connor et al. (2004)
Tarpon Springs Cake 0.124 31 moderate O'Connor et al. (2004)
Regression equation: y x + 20.544;
R2 0.91 O'Connor et al. (2004)
aDetermined via point estimates.
A Lakeland NS
X OCUD S
m Boca Raton
P Applied (kig P ha )
Figure 3-10. Cumulative Bahiagrass P uptake as a function of P applied
~O mmoo ~ acrvccN~
oom e moom01A
3CN Nb Nb W
C~O\~~~ C ~ a
N~3C)3 C~ C NC
00 001 iDcr
WN~o O cOONDY
Table 3-9. Point estimates of RPP for Milorganite and Lakeland NS biosolids
P Source Rate RPP % Average
TSP 56 100 100
TSP 112 100
TSP 224 100
Milorganite 56 38 31
Milorganite 112 31
Milorganite 224 23
Lakeland NS 56 108 90
Lakeland NS 112 106
Lakeland NS 224 55
""""""""""" """-'' 30
Uptake = 8.55 + 22.2(1-e-0 0203*P apphied) -5
P Rate (kg P ha l)
Figure 3-11i. Plateau model: P uptake of Milorganite as a function of P applied
y = 14.6 Ln(x) + 42.5
r2 = 0.793
0 10 20 30 40 50 60 70 80 90
Figure 3-12. Relative phosphorus phytoavailability (RPP) as a function of biosolids-PWEP.
Vertical line at 14% PWEP indicates when increased negative environmental impact may occur.
Dynamic Laboratory Incubation
The obj ectives of the dynamic laboratory incubation were to quantify soluble P release,
study kinetics of P release, and evaluate the leaching hazard of various biosolids sources. The
experiment was designed to mimic an extreme situation of biosolids land application; the soil
used had minimal P-sorbing capacity and no plants were grown to utilize supplied P. We accept
our second hypothesis that P leaching would be significantly greater from BPR and BPR-like
products than conventionally treated materials. P release from BPR and BPR-like biosolids was
equal to TSP, thus we rej ect our third hypothesis that P liability from all biosolids would be less
than TSP. This experiment is key to understanding the environmental hazard specific residuals
Biosolids-PWEP is an excellent indication of how a biosolids will impact the
environment when land applied to sandy, low P-sorbing soils. Biosolids with high PWEP values,
including BPR and BPR-like materials yielded the greatest cumulative P leached. Biosolids with
PWEP values > 14% should be assumed to have a larger potential negative environmental impact
than biosolids with PWEP values <14%. PSI can also be used to gauge the environmental impact
of biosolids land applied to sandy soils with minimal P-sorbing capacity. Given the observed
trends in PWEP and PSI values based on biosolids treatment process, (i.e. PWEP and PSI
increase for BPR and BPR-like biosolids) environmental hazard can be roughly gauged by a
biosolids treatment process. While exceptions exist, much research has shown that BPR and
BPR-like materials have a greater risk of P loss compared to conventionally produced and
pelletized biosolids. In this study, the differences in cumulative P mass leached between the
dried and BPR or BPR-like materials appear to reflect both physical and chemical controls on P
solubility. The pellets of thermally dried materials did not completely dissolve over the course of
the experiment, leading to smaller quantities of P released. Relatively high concentrations of Fe
and Al in the Milorganite material also apparently decreased P release.
Thermally dried and conventionally produced residuals have a slower rate of P release
than BPR and BPR-like products. Knowledge of the kinetics of P release is important to
understanding the effects a residual will have on the environment. A slower rate of P release is
desirable, because opportunity for plant uptake is increased and there is less P in the soil solution
at any given moment. The less P in the soil solution at any given time means less P is available
for leaching through and out of the soil profile, risking impairment of water bodies.
The dynamic laboratory incubation demonstrates that biosolids land application should
not be regulated by assuming all biosolids have equal amounts of labile P. Measurements such as
PWEP and PSI, should be considered in regulating biosolids land application. Assuming all
biosolids to have equal amounts of labile P, and requiring P-based application rates without
considering a residual's individual environmental hazard would unfairly burden municipalities
facing disposal problems or unfairly advantage WWTP producing BPR products. Blanket
regulation of biosolids land application is also unwise given the benefits biosolids can have to
soil and crops when land applied. Biosolids land application can decrease chemical fertilizer
inputs. Chemical fertilizers designed to be water-soluble and provide instant plant nutrition are a
greater environmental hazard than conventionally produced or thermally dried residuals. While
BPR and BPR-like biosolids treatment processes are environmentally beneficial to reduce P in
wastewater effluent, these materials likely pose a greater environmental hazard due to more P
and greater P liability when land applied.
If biosolids are to be applied to agricultural land under P-based restrictions, the quantity
of P that will be available to the crop becomes critical. If a residual has low phytoavailability,
and is applied at a P-based rate, the crop will be N and P deficient, requiring additional mineral
fertilizer input. For BPR or BRP-like materials, P-based application to crops would provide
sufficient P, but require supplemental N fertilizer to meet crop needs. Three BPR or BPR-like
materials, and 1 conventionally produced biosolids (OCUD S) examined herein, fit into the high
category (>75% of TSP) proposed by O'Connor et al. (2004). One BPR (Disney) and 2 thermally
dried materials (Milorganite and GreenEdge) were in the moderate RPP category (25-75% of
TSP). Milorganite fits into the moderate category proposed by O'Connor et al. (2004), despite
expectation that it would be a low RPP material. Based on determined RPP values for BPR and
BPR-like biosolids, we accept our first hypothesis that RPP would be greater from BPR and
BPR-like biosolids than conventional biosolids [with the exception of the conventionally
digested OCUD S biosolids (RPP = 110% of TSP)]. Materials in the high category for RPP also
had the greater cumulative P leached in the laboratory incubation. Materials with high water-
soluble P have more P available for plant uptake. Materials with high RPP make excellent
fertilizers, but can also pose a greater environmental risk for P loss.
The 1995 U.S. EPA design manual (USEPA, 1995) suggests the average "relative
effectiveness" for biosolids-P to be 50% of mineral fertilizer. While the relative effectiveness
factor admits not all P in biosolids is phytoavailable, this study, as well as a similar study by
O'Connor et al. (2004), shows the wide range in the relative phytoavailability of biosolids.
Again, P phytoavailability and leaching hazard are linked to biosolids treatment processes. The
Milorganite biosolids, which is relatively high in Fe and Al, showed the lowest relative
phytoavailability (3 1%). The moderate RPP of the Disney compost is likely due to fact that the
material is a mixture of composted yard waste and biosolids, lowering the quantity of
soluble/available P. Materials with the greatest PWEP values showed the greatest relative
phytoavailability. If and when biosolids P-based regulation of biosolids land application is
imposed, it would be wise to consider individual residual's P characteristics (PWEP and PSI).
Given the range of biosolids P characteristics, regulating all biosolids as if residuals were all the
same could unnecessarily limit beneficial biosolids land application. Most states use a P-index
approach to predicting P loss from a material, and most states do not differentiate biosolids by
treatment process or characteristics such as PWEP. Currently, Florida uses the same source
coefficient for all biosolids (0.015). Elliott et al. (2006) suggested calculating PSC values by
multiplying by the WEP (PSC = 0. 102 x WEP0.99), based on runoff P studies.
Research not detailed in this thesis used 4 biosolids (Milorganite, OCUD S, Lakeland NS,
and Disney) and TSP in rainfall simulations and measured TP, TDP, and biologically (algae)
available P (BAP) in runoff and leachate from Immokalee soil. Flow-weighted TP, TDP, and
BAP were all highly correlated to biosolids-PWEP values. With the exception of Lakeland NS,
BAP losses from all biosolids and TSP was predominantly via leaching. The low solids content
(3%), of the Lakeland NS biosolids, however, resulted in extensive soil surface coverage with a
layer of fine material that was particularly susceptible to rain drop impact and runoff loss.
Cumulative BAP in runoff and leachate for all biosolids treatments was significantly less than
from TSP treatments. Thus, in the short term, biosolids-P (even with very high PWEP values) is
much less of an environmental threat than fertilizer-P. The results of the rainfall simulation
confirm that leaching is the predominant P loss mechanism in typical Florida sands.
Land-Applying BPR or BPR-Like Residuals
Municipalities using BPR biosolids treatment processes have several options to reduce
the impact the residuals will have on the environment; however, each option has advantages and
disadvantages. If supported by the P-Index, biosolids can be land applied at P-based rates, which
will reduce the labile P in the soil solution at any given time, decreasing the chance of negative
environmental impact. P-based application rates of BPR and BPR-like biosolids will provide
sufficient crop P, but will require additional mineral fertilizer input of N. While applying
residuals at a P-based rate is environmentally sound, more land area is required for disposal and
disposal costs will increase. Farmers will also face increased costs from purchasing additional N
fertilizer, which may discourage them from utilizing biosolids as fertilizer. Soil incorporation can
also reduce P loss, especially in soils with sufficient P sorbing capacity.
Biosolids can also be co-applied with WTRs. O'Connor and Elliott (2000) suggested that
co-applying biosolids with water treatment residuals (WTR) increases soil P retention capacity
and reduce P mobility. Agyin-Birikorang et al. (2007) demonstrated that P sorbed to WTRs is
The dynamic laboratory incubation experiment demonstrated the benefits of high Fe and
Al concentrations in limiting P loss, and WWTPs could add Fe and Al salts to BPR and BPR-like
products to reduce environmental risk, however this would significantly increase biosolids mass,
leading to transportation and disposal problems.
The simplest approach is to apply biosolids to soils with sufficient P-sorption capacity
(easily determined). Elliott et al. (2002) demonstrated that even soils with moderate P-sorbing
capacity could prevent significant P loss. Applying biosolids to land with sufficient P-sorbing
capacity may require transporting biosolids longer distances, increasing transportation costs.
Lastly, thermally drying biosolids can reduce labile P and significantly reduce biosolids
mass, decreasing the negative environmental impact. The trend for thermal drying to reduce P
liability was not noted when the OCUD E cake material was dried, however the OCUD E cake
material is un-digested, and therefore may not be land applied, which may produce unique P
GLASSHOUSE ANOVA AND CUMULATIVE DATA
Table A-1. Cumulative Bahiagrass P uptake after 4 harvests.
Source DF Squares Mean Square
Model 27 30.6 1.13
Error 72 1.56 0.022
Corrected Total 99 32.2
R-Square Coeff Var Root MSE
0.952 3.75 0.147
Source DF Type I SS Mean Square
Treatment (source) 7 24.3 3.04
Block 3 0.024 0.008
P-Rate 2 4.81 2.40
Treatment*Rate 14 1.45 0.104
Source DF Type III SS Mean Square
Treatment (source) 7 10.8 1.55
Block 3 0.024 0.008
P- Rate 2 4.81 2.40
Treatment*Rate 14 1.45 0.104
Pr > F
F Value Pr >
F Value Pr >
Table A-2. ANOVA for cumulative Bahiagrass vield after 4 harvests.
'ar Root MSE
DF Type I SS
2 1 246
DF Type III SS
R-Square Coeff V
F Value Pr > F
F Value Pr > F
F Value Pr > F
Table A-3. ANOVA for cumulative P leached in glasshouse experiment.
Type I SS
Type III SS
Corrected Total 99
R-Square Coeff Var
Mean Square F Value
Mean Square F Value
Mean Square F Value
Pr > F
Pr > F
Pr > F
Table A-4. Glasshouse study: cumulative Bahiagrass yields after 4 harvests. Means & 1 standard
error, n = 4.
kg P ha '
15.9 & 1.3
26.6 & 1.0
26.2 & 0.49
25.1 & 0.84
26.6 & 1.1
28.4 & 0.70
26.3 & 2.0
28.6 & 1.0
26.7 & 0.23
28.5 & 0.63
25.7 & 0.85
26.4 & 0.88
23.4 & 0.62
27.6 & 0.55
29.9 & 1.0
30.8 & 1.0
28.1 & 0.79
29.6 & 0.21
30.1 & 0.48
29.5 A 1.3
30.6 & 0.63
16.8 & 4.0
30.9 & 1.2
33.5 A 1.6
Table A-5. G
lasshouse study: cumulative P uptake for Bahiagrass after 4 harvests. Means & 1
andard error, n = 4.
Cumulative P Uptake
8.48 & 1.0
49.4 & 1.1
71.0 & 3.6
105 & 6.6
24.0 & 1.1
27.6 & 1.8
30.9 & 2.0
30.5 A 1.5
37.1 & 0.72
54.0 + 88
36.8 & 2.3
50. 1 1.2
58.5 A 1.7
45.4 & 2.0
67.1 & 2.5
99.0 & 4.7
44.6 & 2.9
69.8 & 3.4
94.9 & 7.0
53.2 & 1.7
75.1 & 6.4
58.3 A 14
52.8 & 1.3
79.5 A 1.6
115 & 6.0
Table A-6. Glasshouse experiment: cumulative P leached after 4 leachings. Means & 1 standard
error, n = 4.
0.044 & 0.006
0.000 + 0.005
0.895 & 0.373
33.6 & 4.88
% of P Applied
0.042 & 0.005
0.46 & 0.181
0.053 & 0.011
0.025 & 0.003
0.012 & 0.002
0.046 & 0.005
0.034 & 0.005
0.021 & 0.002
0.004 & 0.002
0.001 & 0.000
0.001 & 0.002
0.069 & 0.009
0.226 & 0.051
3.30 + 0.906
0.051 & 0.004
0.147 & 0.096
1.25 & 0.301
0.070 + 0.005
0.254 & 0.057
1.55 & 0.518
0.076 & 0.011
0.290 + 0. 119
1.27 & 0. 131
kg P ha
0.028 & 0.009
0.423 & 0.106
13.6 & 3.73
0.009 & 0.004
0.261 & 0.199
0.030 + 0.005
0.480 + 0. 118
6.33 & 2.13
0.035 & 0.012
0.554 & 0.246
5.20 + 0.538
LIST OF REFERENCES
Agyin-Birikorang, S., G.A. O'Connor, L.W. Jacobs, K.C. Makris, and S.R. Brinton. 2007. Long-
term phosphorus immobilization by a drinking water treatment residual. J. Environ. Qual.
Andersen, J.M. 1976. An ignition method for determination of total phosphorus in lake
sediments. Water Research 10: 329-331.
Association of Official Analytical Chemists. 1990. Official methods of analysis of the
Association of Official Analytical Chemists. 15th ed. Arlington, VA.
Brandt, R.C. and H.A. Elliott 2005. Agricultural recycling of BiosetTIL. COnsulting report
prepared for Biosolids Distribution Services, Inc.
Brandt, R.C., H.A. Elliott, and G.A. O'Connor. 2004. Water-extractable phosphorus in biosolids:
implications for land-based recycling. Water Environ. Research, 76:121-129.
Butkes, M.A., D. Grasso, C.P. Schuthess, and H. Wijnja. 1998. Surface Complexation Modeling
of Phosphate Adsorption by Water Treatment Residual. J. Environ. Qual. 27: 1055-1065.
Carpenter, S.R., N.F. Caraco, D.L., R.W. Howarth, A.N. Sharpley, and V.H. Smith. 1998. Non-
point pollution of surface water with phosphorus and nitrogen. Ecol. Applic., 8:559-568.
Chambliss, C.G. and M.B. Adjei. Bahiagrass. 2006. Bahiagrass. UF/IFAS Publication SS-AGR-
36. Univ. of FL, Gainesville.
Chambliss, C., P. Miller, and E. Lord. 2001. Florida Cow-Calf Management, 2nd Ed.- Forages.
UF/IFAS Publication AN118. Univ. FL, Gainesville.
Chardon, W.J., R.G. Menon, and S.H. Chien. 1996. Iron-oxide impregnated filter paper (Pi test):
I. A review of its development and methodological research. Nutr. Cycling Agroecosyst.
Corey, R.B. 1992. Phosphorus regulations: Impact of sludge regulations. Crop Soils. 20:5-10.
Elliott, H.A. and G.A. O'Connor. 2007. Phosphorus management for sustainable biosolids
recycling in the United States. Soil Biol. and Biochem. Dec. 26, 2006 [in press].
Elliott, H.A., R.C. Brandt, P.J.A. Kleinman, A.N. Sharpley, and D.B. Beegle. 2006. Estimating
source coefficients for phosphorus site indices. J. Environ. Qual. 35:2195-2201.
Elliott, H.A., R.C. Brandt, and G.A. O'Connor. 2005. Runoff phosphorus losses from surface-
applied biosolids. J. Environ. Qual. 34: 1632-1639.
Elliott, H.A., G.A. O'Connor, and S. Brinton. 2002. Phosphorus leaching of biosolids-amended
sandy soils. J. Environ. Qual. 31:681-689.
[FDEP] Florida Department of Environmental Protection. 2005. Summary of Class AA residuals.
www. dep.state.fl. us/water/wastewater/dom/docs/2004AA. pdf. Florida Department of
Environmental Protection. Tallahassee, Florida. March 2007.
Harris, W.G., R.D. Rhue, G. Kidder, R.B. Brown, and R. Little. 1996. Phosphorus retention as
related to morphology and sandy coastal plain soil materials. Florida Coop. Ext. Serv.
Circ. 817. Univ. of Florida, Gainesville.
He, Z.L., A.K. Alva, and Y.C. Li. 1999. Sorption-desorption and solution concentration of
phosphorus in fertilized sandy soil. J. Environ. Qual. 28: 1804-1810.
Kidder, G., E.A. Hanlon, C.G. Chambliss. UF/IFAS Standardized Fertilizer Recommendations
for Agronomic Crops. Univ. FL., Coop. Ext. Serv., IFAS, SL-129. 1998.
Loeppert, R.H., W.P. Inskeep. 1997. Iron (Chapter 23). In: Sparks et al. (eds). Methods of Soil
Analysis. Part 3: Chemical Methods. SSSA, Madison, WI. P. 649-650.
Maguire, R.O., J.T. Sims, S.K. Dentel, F.J. Coale, and J.T. Mah. 2001. Relationship between
biosolids treatment process and soil phosphorus availability. J. Environ. Qual. 30:1023-
Murphey, J. and J.P. Riley. 1962. Phosphorus analysis procedure. Methods of soil analysis. Part
2. A.L. Page (Ed.) 413-426. ASA and SSSA. Madison, WI.
Nachabe, M., N. Shah, M. Ross, and J. Vomacka. 2005. Evapotranspiration of two vegetation
covers in a shallow water table environment. Soil Sci. Soc. Am. J. 69:492-499.
NRC. 1996. Nutrients requirements of beef cattle, 6th Ed; National Academy Press: Washington,
Obeysekera, J., J. Browder, L. Hornung, and M.A. Harwell. 2004. The natural south Florida
system I: Climate, geology, and hydrology. Urban Ecosystems. 3:223-244.
O'Connor, G.A. and D. Sarkar. 1999. Fate of land applied residuals-bound phosphorus. DEP
WM 661. Florida Environ. Protection Agency, Tallahassee.
O'Connor, G.A., S. Brinton, and M.L. Silveira. 2005. Evaluation and selection of soil
amendments for field testing to reduce P losses. Soil Crop Sci. Soc. Florida Proc. 64:22-
O'Connor, G.A. and H.A. Elliott. 2000. Co-application of biosolids and water treatment
residuals. Final Report. Florida Department of Environ. Protection, Tallahassee, Florida.
O'Connor, G.A., D. Sarkar, S.R. Brinton, H.A. Elliott, and F.G. Martin. 2004. Phytoavailability
of biosolids phosphorus. J. Environ. Qual. 33:703-712.
Person, A.E., P.E. Speth, R.B. Corey, T.W. Wright, P.L. Schlecht. 1994. Effect of twelve years
of liquid sludge application on soil phosphorus level. In Sewage Sludge: Land Utilization
and the Environment; Clapp, C.E., Ed.; Soil Science Society of America: Madison, WI.
[SAS] Statistical Analysis System. 1989. SAS/STAT user' s guide. Version 6. 4th ed. SAS Inst.,
Sharpley, A.N. 1993a. An innovative approach to estimate biologically available phosphorus in
agricultural runoff using iron oxide-impregnated paper. J. Environ. Qual. 22:597-601.
Sharpley, A.N. 1993b. Assessing phosphorus bioavailability in agricultural soils and runoff. Fert.
Sharpley, A.N. and B. Moyer. 2000. Phosphorus forms in manure and compost and their release
during simulated rainfall. J. Environ. Qual. 29:1462-1469.
Sharpley, A.N., Weld, J.L., Beegle, D.B., Kleinman, P.J.A., Gburek, W.J., Moore, Jr., P.A.,
Mullins, G., 2003. Development of phosphorus indices for nutrient management planning
strategies in the United States. Soil and Water Conservation 58: 137-152.
Smith, S.R., D.M. Bellet-Travers, R. Morris, and J.N.B. Bell. 2002. Fertilizer value of enhanced-
treated and conventional biosolids products. Paper 10. In Proc. CIWEM Conf.-Biosolids:
The Risks and Benefits, an Update on the Latest Res., London. 9 Jan. 2002. Chartered
Inst. Of Water and Environ. London.
Sparks, D.L. 1996. Methods of soil analysis. Part 3. SSSA Book Ser. 5. SSSA, Madison, WI.
Stratful, I., S. Brett, M.B. Scrimshaw, and J.N. Lester. 1999. Biological phosphorus removal: Its
role in phosphorus recycling. Environ. Technol. 20:681-695.
The Fertilizer Institute. 1982. Fertilizer Sampling and Analytical Methods, 4th Ed. The Fertilizer
Institute. Washington D.C.
USDA/USEPA, 1999. Unified national strategy for animal feeding operations. March 9, 1999. U.S.
Government Printing Office, Washington, DC.
USEPA. 1995. Process design manual: land application of sewage sludge and domestic septage.
EPA/625/R-95/001, Office of research & development, Cincinnati, OH.
USEPA. 1993. Methods for determination of inorganic substances in environmental samples.
Revision 2.0 365.1. Washington, D.C.
USEPA. 1999. Biosolids generation, use and disposal in the United States.
www.epa.gov/epaoswer/non-hw/compost/biosoldpf USEPA. Washington, D.C. March
Sarah Chinault was born in Lakeland, Florida, on November 24, 1981 to Edward and
Linda Chinault. Sarah has an older brother (Chris, 28) and a younger sister (Amanda, 23). Sarah
obtained her B.S. degree from the University of Florida in 2004, maj oring in environmental
science. Sarah's career goals include working for an environmental advocacy or animal rights
group. Always an animal lover, Sarah has 5 small dogs that she considers children: Mocha (10),
Mini-Me (7), Maynard (4), Hank (11), and Bauer (1). Mitch (11) recently passed from cancer
and will be missed.