TECHNICAL REPORTS: WASTE MANAGEMENT
17β-Estradiol and Testosterone Sorption in Soil with and without Poultry Litter
M. Bera, D. E. Radcliffe,* M. L. Cabrera, W. K. Vencill, A. Thompson, and S. Hassan
17β-estradiol and testosterone are naturally occurring
steroids that co-occur in poultry litter. The effects of litter
on sorption of these hormones to soil are not known.
Sorption isotherms were developed for 14C-labeled
testosterone and 3H-labeled estradiol in a Cecil sandy clay
loam with and without poultry litter addition. The effect
of applying the hormones alone (single-sorbate) or together
(multisorbate) was also investigated. 14C-testosterone
sorption in soil increased from 2 to 48 h and remained
relatively constant thereafter. 3H-estradiol sorption in soil
was relatively constant from 2 to 24 h and then decreased
to 72 h. These differences may reflect transformation of the
parent hormones to products with different solid-phase
affinity. The maximum sorption coefficient (Kd) in soil
for 14C-testosterone (20.2 mL g−1) was similar to that for
3
H-estradiol (19.6 mL g−1) in single-sorbate experiments.
When hormones were applied together, sorption of both
hormones in soil decreased, but the 14C-testosterone Kd
(12.5 mL g−1) was nearly twice as large as the 3H-estradiol
Kd (7.4 mL g−1). We propose this resulted from competition
between the hormones and their transformation products
for sorption sites, with 14C-testosterone and its expected
transformation
product
(androstenedione)
being
better competitors than 3H-estradiol and its expected
transformation product (estrone). When poultry litter was
mixed with soil, sorption increased for 3H-estradiol but
decreased for 14C-testosterone. This may have been because
poultry litter slowed the transformation of parent hormones.
Our results show that poultry litter could have important
effects on the mobility of estradiol and testosterone.
Copyright © 2011 by the American Society of Agronomy, Crop Science
Society of America, and Soil Science Society of America. All rights
reserved. No part of this periodical may be reproduced or transmitted
in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher.
J. Environ. Qual. 40:1983–1990 (2011)
doi:10.2134/jeq2011.0208
Posted online 5 Oct. 2011.
Received 13 June 2011.
*Corresponding author ([email protected]).
© ASA, CSSA, SSSA
5585 Guilford Rd., Madison, WI 53711 USA
T
he presence of estrogens (estradiol, estrone, and estriol)
and androgens (testosterone) in land-applied manures and
domestic and industrial waste, even in very low concentrations,
has become a growing concern due to their adverse effect on
endocrine systems of humans and wildlife (Ying et al., 2002).
Although estradiol and testosterone help to regulate secondary
sex characteristics under normal conditions, chronic exposure to
these steroids has been associated with abnormal physiological
processes and reproductive abnormalities in birds (Moccia et al.,
1986), turtles (Bishop et al., 1991), and mammals (Martineau et
al., 1988). Land-applied manures and wastes have the potential
to reach surface and subsurface waters, and a number of studies
have reported endocrine disruption in aquatic species (Jafari et al.,
2009; Routledge et al., 1998).
According to the USDA National Agricultural Statistics
Service, the United States produces more chickens (Gallus gallus
domesticus) than any other country (USDA–NASS, 2011). The
production of broilers (chickens raised specifically for meat consumption) in the United States increased from 4.1 billion kg in
1968 to 22.3 billion kg in 2010. Georgia produced 1.3 billion
broilers in 2010, more than any other state in the nation. This
rapid growth of broiler production has increased the production
of poultry litter (a combination of bedding material and manure),
which is applied to agricultural land as a source of nutrients and
a means of disposal.
In addition to nitrogen, phosphorus, and potassium, poultry litter contains 17β-estradiol and testosterone. Lorenzen et al.
(2004) estimated the contents of estradiol and testosterone to be
55 and 30 ng g−1 in broiler litter and 70 and 25 ng g−1 in breeder
(chickens raised for egg production) litter, respectively. In an
Arkansas rainfall simulation study, Nichols et al. (1997) sampled
runoff immediately after application of poultry litter at different
rates. Estradiol concentrations increased linearly with application
rate from about 300 to 1300 ng L−1. In a Georgia natural rainfall
study, Finlay-Moore et al. (2000) measured hormone concentrations in runoff from 0.8-ha paddocks fertilized with poultry litter.
Flow-weighted concentrations ranged from 120 to 820 ng L−1 for
estradiol and from 50 to 920 ng L−1 for testosterone. Hormones
pose a potential hazard to humans and wildlife, even in concentrations at the nanogram per liter level (Routledge et al., 1998).
The degree of surface and subsurface water contamination
depends in part on the sorption of estrogen and testosterone to
M. Bera, 396 Ano Nuevo Ave., Apt. #112, Sunnyvale, CA; D.E. Radcliffe, M.L. Cabrera,
W.K. Vencill, A. Thompson, and S. Hassan, Univ. of Georgia, Athens, GA. Assigned to
Associate Editor Dongqiang Zhu.
Abbreviations: CEC, cation exchange capacity; DOM, dissolved organic matter; HPLC,
high-pressure liquid chromatography.
1983
soil. Most studies have shown that estrogen and testosterone
sorption isotherms are linear or nearly linear (Casey et al.,
2003, 2005; Fan et al., 2008; Hildebrand et al., 2006; Kozarek
et al., 2008; Lee et al., 2003; Sangsupan et al., 2006; Sarmah et
al., 2008). However, two studies found that estradiol isotherms
were highly nonlinear (Bonin and Simpson, 2007; Yu et al.,
2004). Sorption coefficients for estradiol generally increase
with organic C content of soil or sediment, but the reported
coefficients vary widely. Some of this variability can be reduced
by expressing the sorption coefficient in terms of the soil
organic carbon content (Koc), but the log values still range from
2.27 to 4.64 for estradiol. According to Lai et al. (2000), the
estimated value for log Koc based on the octanol-water partition
coefficient (Kow) is 3.5, which is within this range. However,
it is clear that other factors in addition to organic C content
affect sorption. Lai et al. (2000) showed that estrogen sorption occurred on iron oxides in the absence of organic matter.
Other studies found more sorption in soils or river bed sediments with high specific surface and cation exchange capacity
(Bonin and Simpson, 2007; Holthaus et al., 2002).
Other factors that can affect sorption include the method
of measuring the estradiol. Some studies have used 14C-labeled
estradiol and were consequently unable to distinguish between
17β-estradiol and its metabolites (Casey et al., 2003; Fan et
al., 2008; Sangsupan et al., 2006). The method of calculating the sorption coefficients has varied, as well. Most studies
estimate sorbed concentrations by calculating the difference
between the initial solution concentration and the solution
concentration after shaking (Casey et al., 2003; Kozarek et al.,
2008; Lee et al., 2003; Sangsupan et al., 2006; Sarmah et al.,
2008). In this case, the sorbed phase includes any fraction that
is irreversibly sorbed. Several studies have calculated sorption
coefficients indirectly, using a model that includes transformation to an irreversibly sorbed form (Casey et al., 2005; Fan et
al., 2008). With this method, the coefficient is for reversible
sorption only.
Fewer studies have reported sorption coefficients for testosterone, but the results are similar to estradiol, with log Koc in a
slightly higher range, 2.97 to 5.06 (Casey et al., 2004; Fan et
al., 2007; Lee et al., 2003). Only two studies have examined
desorption of estradiol from soil, and no studies have looked
at desorption of testosterone from soil. Hildebrand et al.
(2006) found that 82 to 97% of different forms of estradiol
were irreversibly sorbed in four soils. At low initial concentrations (10 μg L−1), however, 0% of estrone, 20% of estradiol,
and 80% of ethynilestradiol did not desorb. Loffredo and
Senesi (2002) also found that ethynilestradiol did not desorb
from surface and deep horizons of two soils.
No studies have examined the effect of poultry litter on
the soil sorption of 17β-estradiol and testosterone. Organic
matter in litter could retard the movement of the hormones
due to strong sorption, or it could enhance movement due
to cotransport with fine particulate matter. Also, because
both hormones are present in poultry litter, they have the
potential to compete for sorption sites. Bonin and Simpson
(2007) showed the competitive effect of different forms of
estrogen on sorption. No studies have examined the competitive effects of estrogen and testosterone.
1984
The objectives of our research were (i) to evaluate the effect
of poultry litter on the sorption of 17β-estradiol and testosterone in soil and (ii) to evaluate the effect of applying the
hormones alone or together. This work was performed in conjunction with a transport study using soil columns and breakthrough curves to be presented in a later paper.
Materials and Methods
Soil and Poultry Litter Samples
Bulk soil samples for sorption experiments were collected from
the 0- to 30-cm soil depth in paddocks located at the Central
Georgia Research and Education Center near Eatonton, GA.
These paddocks had received broiler litter for approximately 12
yr at the time of sample collection. The soil at the site was a Cecil
sandy clay loam (fine, kaolinitic, thermic Typic Kanhapludult),
which is a common soil in the Southern Piedmont physiographic region. Soil samples were air-dried, ground, passed
through a 2-mm sieve, and stored at 20°C. Poultry litter was
obtained from a poultry farm located in northeast Georgia.
Litter was air-dried, ground, and stored at 20°C.
Soil and Poultry Litter Analyses
Soil samples were analyzed for pH (1:5 soil:water ratio), particle size distribution (Gee and Or, 2002), total carbon (C)
by dry combustion (Nelson and Sommers, 1982), and cation
exchange capacity (CEC) and base saturation. To determine CEC, the soil was extracted with Mehlich-1 solution
(Mehlich, 1953), the bases (Ca, Mg, K, and Na where applicable) were determined by inductively coupled plasma mass
spectrometry and then added to the amount of exchangeable
hydrogen. Base saturation was calculated by dividing bases
by the CEC and expressing the result as percentage. Poultry
litter samples were analyzed for total C by dry combustion
(Nelson and Sommers, 1982).
To extract hormones from the litter, 1 g of litter was placed
in narrow mesh Nytex bags and extracted with 15-mL of
dichloromethane (CH2Cl2: high-pressure liquid chromatography [HPLC] grade) in 40-mL amber-colored glass tubes.
The tubes were sonicated for 1 h before being centrifuged
(2000 rpm for 20 min) and the supernatants decanted into
50-mL evaporation tubes. The bagged litter was reextracted
two times in 10 mL of dichloromethane for 30 min, centrifuged, and the resulting supernatants combined. The CH2Cl2
extracts were evaporated (Turbovap II, LV; Zymark Corp.,
Hopkinton, MA) to near dryness under N2. A final drying
step using two successive 1-mL aliquots of acetonitrile under
N2 removed the remaining CH2Cl2. Temperature in the
Turbovap never exceeded 70°C. The dry residues containing
the hormones were derivatized to the trimethysilyl derivatives
and made up to suitable volume with acetonitrile before gas
chromatography analysis.
The concentrations of 17β-estradiol and testosterone
were measured in the poultry litter using a gas chromatograph tandem mass spectrometer (GC–MS–MS; Agilent
Technologies, Palo Alto, CA) operated in selected ion storage
mode following derivatization of the hormones to trimethylsilyl forms with N,O-bis(trimethylsilyl)-trifluoroacetamide containing 1% (w/w) of trimethylchlorosilane. Chormotographic
Journal of Environmental Quality • Volume 40 • November–December 2011
separation was conducted on a J&W capillary column
(30 m by 0.25 mm i.d. with 0.25-mm film thickness; Agilent
Technologies, Wilmington DE) following splitless injection
using 1 mL min−1 He carrier gas. The column temperature
profile was increased from 90 to 250°C (12°C min−1 ramp) followed by a second increase to 270°C (20°C min−1 ramp) and
held for 5 min. Injection port and transfer lines were held at
260 and 70°C, respectively, while the ion trap and source were
held at 210 and 270°C, respectively. Quality assessment and
control procedures included running one blank and one spiked
sample every 10 samples and a duplicate sample every 20 samples. Limit of quantification was 0.7 ng g−1 for 17β-estradiol,
1.7 ng g−1 for estrone, and 0.50 ng g−1 for testosterone. Average
recovery was 86% for 17β-estradiol, 83% for estrone, and 95%
for testosterone.
The concentrations of hormones in the soil samples in our
study were not measured. However, monthly measurements
were made in the paddocks from which our soil samples were
collected. The procedure for extracting and measuring hormones in soil were the same as described above for the poultry
litter samples, except mesh bags were not used and 5 g of soil
was used instead of 1 g of litter.
Hormones
Unlabeled estradiol and testosterone (both >98% purity) were
obtained from Sigma-Aldrich (St. Louis, MO). Solutions were
stored in a refrigerator at 4°C. Radio-labeled 6,7-3H-estradiol
(specific activity 50 Ci mmol−1) and 4-14C-testosterone (specific activity 55 mCi mmol−1) were obtained from American
Radiolabeled Chemicals (St. Louis, MO). In this study, we
refer to the 3H label as “3H-estradiol” and to the 14C label as
“14C-testosterone” but recognize that the label may actually be
in a metabolite of these hormones. Where appropriate, we discuss the transformation products.
Equilibrium Sorption Isotherms
Batch equilibrium sorption experiments were performed
for soil, soil mixed with litter (soil+litter), and litter alone.
Sorption experiments were conducted for 3H-estradiol and
14
C-testosterone separately as well as with both hormones
together for soil to determine the interaction of these hormones. Unlabeled hormones were diluted to 100 μg mL−1 with
HPLC-grade methanol to avoid precipitation due to the inorganic solvent 0.01 M CaCl2. Then they were diluted to 0.001,
0.01, 0.1, 0.5, and 1 μg mL−1 with 0.01 M CaCl2 solution for
sorption isotherms following the procedure used by Sangsupan
et al. (2006). Ten milliliters of unlabeled estradiol or testosterone of each concentration were added to 1 g air-dried soil
or litter in a 15-mL glass centrifuge tube. For soil mixed with
litter, 0.07 g of litter was added to each g of soil. Test tubes
were sealed by Teflon-lined caps. In each test tube, 5 μL of
labeled testosterone and estradiol (0.059 KBq 6,7-3H-estradiol
and 0.015 KBq 4-14C-testosterone) were added. The test tubes
were shaken in a reciprocating shaker at 300 oscillations per
minute for 2, 24, and 48 h at 20°C.
The experiment was repeated for soil (not for soil+litter and
litter alone) with the hormones applied together. One gram
of air-dried soil was shaken with 10 mL of unlabeled solution
(5 mL unlabeled estradiol and 5 mL unlabeled testosterone
Bera et al.: Effect of Poultry Litter on Estradiol and Testosterone Sorption
of corresponding concentrations). Five microliters of labeled
3
H-estradiol and14C- testosterone were added to each test tube
and shaken for 2, 24, 48, 72, 120, and 144 h at 20°C.
Samples were centrifuged for 15 min at 3500 rpm at each
time interval. The centrifugal force was 4400 times the force
of gravity. One milliliter of supernatant was then taken for
analysis of radioactivity by a Beckman LS6500 liquid scintillation counter (Beckman Coulter, Inc., Brea, CA). The
efficiency of the counter was routinely monitored. Counting
efficiency for 3H was 88% and for 14C was 98%. The instrument was regularly calibrated using standard 3H and 14C
quench standards provided by Beckman Coulter, Inc. (Brea,
CA). An external quench monitor (cesium-137) was used to
compensate for quench. The instrument software converted
raw counts per minute data to disintegrations per minute
based on quench correction.
Three replications of each experimental unit were performed. The difference in initial and final radioactivity was
considered as that adsorbed by 1 g of soil, litter, or the mixture.
Sorption experiments were performed in sterile (0.12 mg mL−1
HgCl2) and nonsterile medium. Mercuric chloride (HgCl2)
effectively destroys the microbial population without changing
soil physical and chemical properties (Wolf et al., 1989).
Batch equilibrium sorption data were fit with the linear
Freundlich (1909) sorption equation:
s = Kdc
[1]
where s is the concentration of hormone in the sorbed phase
(μg g−1), c is the concentration of hormone in solution (μg
mL−1), and Kd is the sorption coefficient (mL g−1). The sorption data were also fit with the nonlinear Freundlich (1909)
sorption equation:
s = Kfcn
[2]
where Kf is the nonlinear Freundlich sorption coefficient
(μg1-n mLn g−1) and n is the Freundlich exponent.
Desorption isotherms were also developed. Ten milliliters
of unlabeled estradiol or testosterone (1 μg mL−1) were added
to 1 g air-dried soil or litter in a 15-mL glass centrifuge tube.
For soil mixed with litter, 0.07 g of litter was added to each
g of soil. Test tubes were sealed by Teflon-lined caps. In each
test tube, 5 μL of labeled testosterone and estradiol (0.082
KBq 6,7-3H-estradiol and 0.019 KBq 4-14C-testosterone) were
added. Samples were shaken in a reciprocating shaker for 24
h, centrifuged for 15 min at 3500 rpm, and then 1 mL of
supernatant was taken for analysis of radioactivity using the
liquid scintillation counter. The supernatant was discarded and
10 mL of hormone-free 0.01 M CaCl2 solution was added to
the same test tube and it was shaken for 24 more hours. This
process was repeated four times at 24-h intervals.
Statistical Analysis
Results for sorption coefficients were analyzed statistically using
the General Linear Model (PROC GLM) procedure in SAS
(SAS Institute, Inc., Cary, NC). Differences between means
were tested using Fisher’s protected least significance difference
(LSD) with p = 0.05 or 0.01, except as noted.
1985
The sorption kinetics and equilibrium isotherm data were not
significantly different in nonsterile and sterile medium, indicating that biotic processes such as mineralization were not
important in the time frame of our study. Therefore, only the
sterile data are presented.
Sorption of both hormones varied with contact time in
our soil and soil+litter treatments, whereas sorption to litter
was equal within error across all reaction times (Fig. 1 and
2). 14C-testosterone sorption to soil and soil+litter increased
from 2 to 48 h before reaching pseudo-equilibrium (Fig. 1),
whereas 3H-estradiol exhibited a sorption maximum to soil
and soil+litter between 2 and 24 h followed by a subsequent
decrease in the amount retained on the solid after 24 h (Fig. 2).
In a prior study of 14C-testosterone and 3H-estradiol sorption
to surface samples from a similar Cecil series soil, Sangsupan et al.
(2006) found comparable time-dependent sorption trends with
a 14C-testosterone sorption maximum at 72 h and a decrease in
3
H-estradiol sorption after 48 h. However, Casey et al. (2004)
found that maximum sorption of testosterone occurred much
more rapidly (1 to 5 h) for an assortment of Mollisols. Rapid
14
C-testosterone sorption (<2 h to apparent equilibrium) was
also observed by Sangsupan et al. (2006) when examining subsurface samples of the Cecil series soil. Because each of these
studies—ours included—monitored sorption by loss of the
14
C-label from solution, this heterogeneity in the time to maximum sorption across different soils may result from variation in
the rates of testosterone transformation and subsequent transformation-product sorption. When Lee et al. (2003) used HPLC to
monitor testosterone transformation during sorption reactions,
they found the terminal hydroxyl group in testosterone to be
rapidly transformed into a ketone group forming the daughter
product androstenedione. Ketone groups are stronger hydrogen
acceptors than hydroxyl groups but cannot serve as hydrogen
donors (Neale et al., 2009). Thus, depending on the H-bonding
environment in the soil, the rate of testosterone transformation
could alter the apparent sorption kinetics (Lee et al., 2003).
Transformation of the parent hormone may also explain
the decrease in 3H-estradiol sorption with increasing reaction
time (Fig. 2), as again our methodology could not distinguish
between estradiol and its daughter transformation products
(Lai et al., 2000) if they retained the 3H label. Similar to the
transformation reaction described above for testosterone, Lee
et al. (2003) found that within 11 min after reacting estradiol
with soils, the phenolic hydroxyl group at position 17 on estradiol was oxidized to a ketone, forming the daughter product
estrone. This transformation reaction is found to occur even
in sterile soil, suggesting an abiotic mechanism (Colucci et
al., 2001). In addition to estrone, Casey et al. (2003) found
trace concentrations of estriol as a transformation product of
17β-estradiol. The presence of the ketone group on estrone
compared with the hydroxyl group on estradiol has important bearing on their relative affinity for soil sorbents. Ketones
evidently react much more strongly with phenolic hydrogen
donors than they do with carboxylic acid or alcohol hydrogen donors (Neale et al., 2009). Thus, estrone has a lower Kow
value than estradiol because the hydrogen donor in octanol is
an alcohol group, whereas estrone exhibits higher Kd values
than estradiol for sorbents dominated by phenolic hydrogen
donors such as humic acid, tannic acid, and toluene (Neale
et al., 2009; Qiao et al., 2011). This seems to be reflected in
Fig. 1. Effect of shaking time on sorption of 14C-testosterone in poultry
litter and in soil with and without poultry litter. The initial concentration was 0.5 μg mL−1. Standard error bars are shown for three
replicate samples per treatment.
Fig. 2. Effect of shaking time on sorption of 3H-estradiol in poultry
litter and in soil with and without poultry litter. The initial concentration was 0.5 μg mL−1. Standard error bars are shown for three
replicate samples per treatment.
Results and Discussion
Soil and Litter Properties
The sand, silt, and clay contents of the soil from the 0- to 30-cm
depth were 40, 28, and 32% by weight, respectively (sandy clay
loam textural class). Other properties were pH = 6.15, base
saturation = 87.3%, total C = 16.8 g kg−1, and CEC = 9.19
cmolc kg−1. The background concentrations of 17β-estradiol
and testosterone in the litter were 60 and 303 ng g−1, respectively, similar to the values reported by Lorenzen et al. (2004).
The background soil concentrations of hormones in the plot
where we took the samples for our study were measured on 14
May 2008 (we collected our study samples in mid-June). The
concentration in the top 0- to 2.5-cm depth increments were
52, 61, and 110 ng g−1 for 17β-estradiol, estrone, and testosterone, respectively.
Sorption Kinetics
1986
Journal of Environmental Quality • Volume 40 • November–December 2011
the spread of Kd values for sorption of these compounds to
soil across a range of studies. Most often, estradiol exhibits
higher Kd values than estrone for sorption to soils and sediments (Lee et al., 2003; Bonin and Simpson, 2007; Fan et al.,
2008; Sarmah et al., 2008), but this is less consistent when
sorption is particularly nonlinear (Yu et al., 2004), or involves
sandy soils with low organic matter content where Hildebrand
et al. (2006) measured a higher Kd for estrone than estradiol,
although Lee et al. (2003) did not.
Equilibrium Sorption Isotherms
All 14C-testosterone and 3H-estradiol equilibrium isotherms
were strongly linear across the concentration range examined
(0.001 to 1 μg mL−1) (Fig. 3 and 4), yielding Freundlich exponent values (n) between 0.92 to 0.98 for all treatments, with
the majority of n values greater than 0.95 (data not shown).
These values are comparable to values reported previously
(Casey et al., 2003), although nonlinear isotherms for sorption
of estradiol to soils are also common (Yu et al., 2004). Because
the linear model adequately fit the isotherm data, it will be
used in the remainder of our discussion. The background
concentrations of hormones in the soil (52, 61, and 110 μg
kg−1 for 17β-estradiol, estrone, and testosterone, respectively)
were small compared with the concentrations of unlabeled
17β-estradiol and testosterone added (10, 100, 1000, 5000,
and 10000 μg kg−1), except at the lowest levels. The fact that
the isotherms were linear indicated that the background concentrations did not have an effect.
The maximum linear sorption coefficients to soil were
similar for both hormones (Table 1), although Kd values for
3
H-estradiol were larger at early reaction time and smaller at
long reaction time than 14C-testosterone, reflecting our kinetic
sorption data (Fig. 1 and 2). Estradiol typically exhibits higher
affinity for soils, sediment (Lee et al., 2003; Casey et al., 2004),
and organic (Neale et al., 2009) sorbents than testosterone,
although Lee et al. (2003) reported higher testosterone linear
sorption coefficients compared with estradiol for the sandy,
low-organic matter soil, EPA-1. We used total C as a surrogate
for organic C in our soil because of its noncalcareous nature
and lack of liming history. Using total C, we calculated monosorbate log Koc values for 3H-estradiol and 14C-testosterone
from the 24-h isotherm to be 3.04 and 3.07, respectively. These
Fig. 3. Linear isotherms for noncompetitive 3H-estradiol sorption to
soil-only (filled circle), soil+litter (open circle), and litter-only (open
square) and competitive sorption in the presence of 14C-testosterone
to soil-only (open triangle). All reaction times were 24 h. All data
points shown without averaging.
Fig. 4. Linear isotherms for noncompetitive 14C-testosterone sorption
to soil-only (filled circle), soil+litter (open circle), and litter-only (open
square) and competitive sorption in the presence of 3H-estradiol to
soil-only (open triangle). All reaction times were 24 h. All data points
shown without averaging.
are within the range of log Koc values for similar Cecil topsoils
(Sangsupan et al., 2006; Kozarek et al., 2008) and comparable
to literature values reported for a wide range of soil series (Fig.
5 and 6).
Table 1. Mean linear sorption coefficients (Kd) for soil, soil with litter, and litter as a function of time for hormones applied separately. Values are also
shown for hormones applied together to soil-alone after 24 h. Coefficients of determination (r2) are shown for the fit of the linear isotherm to the
data.
3
2 hr
Soil
r2
Soil + litter
r2
Litter
r2
Soil
r2
H-Estradiol Kd
24 hr
14
48 hr
2 hr
C-Testosterone Kd
24 hr
48 hr
—————————————————————————— mL g−1 ——————————————————————————
Hormones applied separately
19.6
18.6
8.7
14.9
19.7
20.2
0.96
1.00
1.00
0.99
0.99
0.98
23.1
24.3
19.6
11.8
16.0
17.0
0.99
0.99
1.00
1.00
1.00
0.99
55.5
59.8
59.5
35.3
36.1
38.0
0.99
1.00
1.00
0.99
1.00
1.00
Hormones applied together
7.4
12.5
0.99
0.96
Bera et al.: Effect of Poultry Litter on Estradiol and Testosterone Sorption
1987
Fig. 5. Log organic carbon sorption coefficient (Koc) from previous
studies and our study of estradiol as a function of reaction time.
Some studies used radio-labeled estradiol (our study; Casey et al.,
2003, 2005; Fan et al., 2008; Sangsupan et al., 2006). Other studies
measured 17β-estradiol, estrone, or 17α-ethynilestradiol directly
(Bonin and Simpson, 2007; Hildebrand et al., 2006; Kozarek et al.,
2008; Lee et al., 2003; Sarmah et al., 2008; and Yu et al., 2004).
Fig. 6. Log organic carbon sorption coefficient (Koc) from previous
studies of testosterone as a function of reaction time. Our study and
Casey et al. (2004) used radio-labeled testosterone. Fan et al. (2007)
and Lee et al. (2003) measured testosterone and metabolites directly.
The Influence of Poultry Litter
Sorption coefficients were much higher in litter without soil
than in soil alone for both 3H-estradiol and 14C-testosterone
(Table 1); this was expected given the attraction of these
hormones to organic matter (Casey et al. 2003, 2004). We
measured higher Kd values for sorption of 3H-estradiol (max.
59.8 mL g−1) to litter compared with 14C-testosterone (max.
38.0 mL g−1). When poultry litter was mixed with soil, the
affinity of 3H-estradiol for the resulting solid phase materials
was higher than the soil-only treatment (Kd soil+litter > Kd
soil; Table 1) for all reaction times. Similar to the soil-only
treatment, we observed a trend of rapid initial sorption (Kd
= 23.1 mL g−1) followed by subsequent desorption leading
to a significant decrease in Kd values after 48 h of reaction
(19.6 mL g−1). This desorption step was not detected in the
litter-only treatment. If we interpret the desorption as trans1988
formation to a lower affinity sorbate, as discussed above,
then our data suggest either (i) the affinity of the transformation product for litter is equal to or within error of that
of estradiol, or (ii) that the transformation reactions were
suppressed in the litter-only treatment.
In contrast, sorption of 14C-testosterone in the soil+litter
treatment was significantly lower than in the soil-only treatment in both our sorption isotherm experiments after 2, 24,
and 48 h of reaction (Table 1 and Fig. 4) and in our single-concentration sorption experiments that extended to 140 h of reaction (Fig. 1). These results suggest that leaching of testosterone
from soil may be greater in the presence of poultry litter than
when testosterone is applied alone. Sorption of testosterone to
dissolved organic matter (DOM) released from the litter might
explain these results as the centrifugation rates used in our
experiments were unlikely to sediment most DOM. However,
we would expect similar artifacts to reduce our measurements
of estradiol sorption, but they did not.
Instead, we propose the presence of litter inhibited the oxidation of the phenolic hydroxyl on both hormones and delayed
the transformation of 17β-estradiol and testosterone to estrone
and androstenedione, respectively. Because 17β-estradiol has a
higher Kd than estrone (Lee et al., 2003; Bonin and Simpson,
2007; Fan et al., 2008; Sarmah et al., 2008), this would increase
apparent sorption of 3H-estradiol in the soil+litter treatment
(Table 1, Fig. 3). Delaying the transformation of testosterone
to androstenedione—which has a higher Kd than testosterone
(Lee et al., 2003)—would decrease sorption (Table 1, Fig. 4).
Supporting this hypothesis, equilibrium concentrations of
3
H-estradiol and 14C-testosterone were constant in our litteronly treatments (Fig. 1 and 2). In addition, the desorption of
3
H-estradiol, which became readily apparent between 24 and
48 h in our soil-only experiment, was much more gradual in
the soil+litter treatment, extending from 24 to 120 h (Fig. 2).
14
C-testosterone increased slowly from 2 to 48 h in both the
soil and soil+litter treatments (Fig. 1).
To explore the effect of delayed transformation, we modeled the sorption coefficient of the soil+litter treatments after
24 and 48 h of reaction using a mass-weighted average (1 g of
soil and 0.07 g of litter) of the soil-only and litter-only sorption coefficients after 2 h of reaction and compared them to
measured Kd values for 24 and 48 h of reaction (Table 1). For
14
C-testosterone, we calculated an estimated 14C-testosterone
Kd for soil+litter of 16.2 mL g−1, which compared well with the
measured Kd for soil+litter of 16.0 and 17.0 after 24 and 48 h
of reaction, respectively (Table 1). Thus, the reduction in sorption in the 24-h soil+litter treatment with 14C-testosterone is
consistent with our hypothesis that the addition of litter slowed
the transformation process. For 3H-estradiol, the calculated Kd
for the soil+litter treatment was 21.9 mL g−1, which was intermediate between the measured values of 24.3 and 19.6 mL g−1
after 24 and 48 h of reaction, respectively (Table 1). Further
investigation involving direct measurements of both the hormones and their transformation products would be necessary
to confirm this hypothesis.
As discussed above in regard to the soil samples, the background concentrations of 17β-estradiol and testosterone in
the litter (60 and 303 μg kg−1, respectively) were small compared with the concentrations of unlabeled 17β-estradiol and
Journal of Environmental Quality • Volume 40 • November–December 2011
testosterone added (10–10,000 μg kg−1), except at the lowest
level and had a negligible effect on the calculated Kds for litter
and soil+litter.
Combined Effect of Testosterone and Estradiol
Multisorbate experiments were also performed to evaluate the affinity of 14C-testosterone and 3H-estradiol for soil
when both hormones were present (as they always are in
poultry litter). After 24 h of reaction, both Kd values were
significantly lower in the multisorbate system (7.4 and
12.5 mL g−1 for 3H-estradiol and 14C-testosterone, respectively) than in the single-sorbate system (18.6 and 19.7 mL
g−1 for 3H-estradiol and 14C-testosterone, respectively) (Table
1). The affinity of estradiol for soils and sediment is commonly observed to decrease in the presence of competing
estrogen sorptives (Lai et al., 2000; Yu et al., 2004; Bonin
and Simpson, 2007). This effect has not been reported previously for testosterone but could be expected given its structural similarities to estradiol.
Unexpectedly, the testosterone Kd was significantly higher
(12.5 mL g−1) than the estradiol Kd (7.4 mL g−1) (Table 1).
Estradiol typically exhibits higher affinity for soil (Lee et al.,
2003; Casey et al., 2004), sediment (Lee et al., 2003), and
organic (Neale et al., 2009) sorbents than testosterone, although
there may be exceptions in sandy soils with low organic matter
(Lee et al., 2003). Indeed, we found much larger Kd values for
3
H-estradiol sorption to poultry litter and soil+litter than for
14
C-testosterone (Table 1). However, as we note above, there is
increasing evidence that the hydroxyl group at position 17 of
estradiol and testosterone undergoes rapid oxidation to ketone
moieties during sorption to soil (Lee et al., 2003), leading to the
transformation products estrone and androstenedione. After
confirming substantial hormone transformation in 17β-estradiol-soil and testosterone-soil experiments, Lee et al. (2003)
calculated higher soil Kd values for androstenedione than for
estrone. Furthermore, the observed desorption of 3H-estradiol
with increasing reaction time suggests estradiol or its transformation products are less tightly bound than testosterone or its
transformation products.
To investigate this, we tracked 3H or 14C desorption following sorption of 3H-estradiol or 14C-testosterone to soil in a
single-sorbate experiment. After four washings at 24-h intervals, more of the 14C-testosterone was retained on the soil than
3
H-estradiol (Fig. 7). The differences in coefficients of the fitted
polynomials were statistically significant at the p = 0.059 level.
This suggests testosterone or its transformation products bind
more strongly to the soil than estradiol or its transformation
products. A likely explanation for the higher affinity for soil of
14
C-testosterone than 3H-estradiol in our multisorbate experiments is the combined effects of a lower desorption potential
(Fig. 7) and greater affinity of its transformation product for
soil (Lee et al., 2003).
Conclusions
Our results suggest the environmental fate of estradiol and
testosterone in land-applied poultry litter is influenced by
the co-occurence of the hormones and the influence of the
litter on sorption to soil. We found when the hormones
Bera et al.: Effect of Poultry Litter on Estradiol and Testosterone Sorption
Fig. 7. Percentage of 3H-estradiol and testosterone sorbed as a function of four daily washings with hormone-free solution. All data
points shown without averaging. The lines through the data points
are fitted second-order polynomials.
were applied separately to soil, the maximum Kd values for
14
C-testosterone and 3H-estradiol were similar, with log Koc
values that were comparable to prior studies. Kinetic experiments showed that 14C-testosterone sorption to soil increased
steadily to 48 h before reaching pseudo-equilibrium, whereas
3
H-estradiol sorption peaked between 2 and 24 h followed
by desorption of the 3H label. These trends for both hormones are consistent with the oxidation of the phenolic
hydroxyl group to a ketone group forming the transformation products, estrone and androstenedione from estradiol
and testosterone, respectively. Estrone has a lower reported
affinity for soil than estradiol, whereas androstenedione has
a higher reported affinity for soil than testosterone. When
the hormones were applied together to soil, Kd for both hormones decreased and the 14C-testosterone Kd was significantly
greater than the value for 3H-estradiol. We suggest this is due
to competition for sorption sites between hormones and their
transformation products, with 14C-testosterone and its transformation products being better competitors than estradiol
and its transformation products. Desorption experiments
also showed more 14C-testosterone was retained on the soil
than 3H-estradiol. When poultry litter was mixed with soil,
sorption in soil increased for 3H-estradiol but decreased for
14
C-testosterone. This may have been due to poultry litter
slowing the transformation of parent hormones. These results
suggest that leaching of testosterone from soil may be greater
in the presence of poultry litter than when testosterone is
applied alone.
Acknowledgments
This work was supported by EPA STAR grant R833419.
References
Bishop, C.A., R.J. Brooks, J.H. Carey, P. Ng, R.J. Norstrom, and D.R.S. Lean.
1991. The case for a cause-effect linkage between environmental contamination and development in eggs of the common snapping turtle
(Chelydra-S Serpentina) from Ontario, Canada. J. Toxicol. Environ.
Health 33:521–547. doi:10.1080/15287399109531539
1989
Bonin, J.L., and M.J. Simpson. 2007. Sorption of steroid estrogens to soil and
soil constituents in single- and multi-sorbate systems. Environ. Toxicol.
Chem. 26:2604–2610. doi:10.1897/07-118.1
Casey, F.X.M., H. Hakk, J. Šimůnek, and G.L. Larsen. 2004. Fate and transport of testosterone in agricultural soils. Environ. Sci. Technol. 38:790–
798. doi:10.1021/es034822i
Casey, F.X.M., G.L. Larsen, H. Hakk, and J. Šimůnek. 2003. Fate and transport of 17 beta-estradiol in soil-water systems. Environ. Sci. Technol.
37:2400–2409. doi:10.1021/es026153z
Casey, F.X.M., J. Šimůnek, J. Lee, G.L. Larsen, and H. Hakk. 2005. Sorption,
mobility, and transformation of estrogenic hormones in natural soil. J.
Environ. Qual. 34:1372–1379. doi:10.2134/jeq2004.0290
Colucci, M.S., H. Bork, and E. Topp. 2001. Persistence of estrogenic hormones in agricultural soils: I. 17β-estradiol and estrone. J. Environ.
Qual. 30:2070–2076. doi:10.2134/jeq2001.2070
Fan, Z.S., F.X.M. Casey, H. Hakk, and G.L. Larsen. 2007. Discerning and
modeling the fate and transport of testosterone in undisturbed soil. J.
Environ. Qual. 36:864–873. doi:10.2134/jeq2006.0451
Fan, Z.S., F.X.M. Casey, H. Hakk, and G.L. Larsen. 2008. Modeling coupled
degradation, sorption, and transport of 17 beta-estradiol in undisturbed
soil. Water Resour. Res. 44:W08424. doi:10.1029/2007WR006407
Finlay-Moore, O., P.G. Hartel, and M.L. Cabrera. 2000. 17-β estradiol and testosterone in soil and runoff from grasslands amended
with broiler litter. J. Environ. Qual. 29:1604–1611. doi:10.2134/
jeq2000.00472425002900050030x
Freundlich, H. 1909. Kapillarchemie; eine Darstellung der Chemie der
Kolloide und verwandter Gebiete. Akademische Verlagsgellschaft,
Leipzig, Germany.
Gee, G.W., and D. Or. 2002. Particle-size analysis. p. 255–293. In J.H. Dane
and G.C. Topp (ed.) Methods of soil analysis. Part 4. SSSA, Madison, WI.
Hildebrand, C., K.L. Landry, and A. Farenhorst. 2006. Sorption and desorption of three endocrine disrupters in soils. J. Environ. Sci. Health B
41:907–921.
Holthaus, K.I.E., A.C. Johnson, M.D. Jurgens, R.J. Williams, J.J.L. Smith,
and J.E. Carter. 2002. The potential for estradiol and ethynylestradiol to
sorb to suspended and bed sediments in some English rivers. Environ.
Toxicol. Chem. 21:2526–2535. doi:10.1002/etc.5620211202
Jafari, A.J., R.P. Abasabad, and A. Salehzadeh. 2009. Endocrine disrupting
contaminants in water resources and sewage in Hamadan City of Iran.
Iran. J. Environ. Health Sci. Eng. 6:89–96.
Kozarek, J.L., M.L. Wolfe, N.G. Love, and K.F. Knowlton. 2008. Sorption
of estrogen to three agricultural soils from Virginia USA. Trans. ASAE
51:1591–1597.
Lai, K.M., K.L. Johnson, M.D. Scrimshaw, and J.N. Lester. 2000. Binding
of waterborne steroid estrogens to solid phases in river and estuarine
systems. Environ. Sci. Technol. 34:3890–3894. doi:10.1021/es9912729
Lee, L.S., T.J. Strock, A.K. Sarman, and P.S.C. Rao. 2003. Sorption and dissipation of testosterone, estrogens, and their primary transformation
products in soils and sediment. Environ. Sci. Technol. 37:4098–4105.
doi:10.1021/es020998t
Loffredo, E., and N. Senesi. 2002. Sorption and release of endocrine disruptor compounds onto/from surface and deep horizons of two sandy soils.
Dev. Soil Sci. 28(A):143–159. doi:10.1016/S0166-2481(02)80051-4
1990
Lorenzen, A., J.G. Hendel, K.L. Conn, S. Bittman, A.B. Kwabiah, G. Lazarovitz, D. Masse, T.A. McAllister, and E. Topp. 2004. Survey of hormone
activities in municipal biosolids and animal manures. Environ. Toxicol.
19:216–225. doi:10.1002/tox.20014
Martineau, D., A. Lagace, P. Beland, R. Higgins, D. Armstrong, and L.R.
Shugart. 1988. Pathology of stranded Beluga whales (DelphinapterusLeucas) from the St-Lawrence Estuary, Quebec, Canada. J. Comp.
Pathol. 98:287–311. doi:10.1016/0021-9975(88)90038-2
Mehlich, A. 1953. Determination of P, Ca, Mg, K, Na, and NH4. In North
Carolina Soil Test Division. [Mimeo.] Vol. 1–53. North Carolina Department of Agriculture, Raleigh.
Moccia, R.D., G.A. Fox, and A. Britton. 1986. A quantitative assessment
of thyroid histopathology of herring-gulls (Larus-Argentatus) from the
Great-Lakes and a hypothesis on the causal role of environmental contaminants. J. Wildl. Dis. 22:60–70.
Neale, P.A., B.I. Escher, and A.I. Schäfer. 2009. pH dependence of steroid
hormone–organic matter interactions at environmental concentrations.
Sci. Total Environ. 407:1164–1173 10.1016/j.scitotenv.2008.09.035.
doi:10.1016/j.scitotenv.2008.09.035
Nelson, D.W., and L.E. Sommers. 1982. Total carbon, organic carbon, and
organic matter. p. 539–579. In A.L. Page et al. (ed.) Methods of soil
analysis. Part 2. SSSA, Madison, WI.
Nichols, D.J., T.C. Daniel, P.A. Moore, Jr., D.R. Edwards, and D.H. Pote.
1997. Runoff of estrogen hormone 17β-estradiol from poultry litter applied to pasture. J. Environ. Qual. 26:1002–1006. doi:10.2134/
jeq1997.00472425002600040011x
Qiao, X., N. Carmosini, F. Li, and L.S. Lee. 2011. Probing the primary mechanisms affecting the environmental distribution of estrogen and androgen
isomers. Environ. Sci. Technol. 45:3989–3995 doi:10.1021/es200073h.
Routledge, E.J., D. Sheahan, C. Desbrow, G.C. Brighty, W. Waldock, and
J.P. Sumpter. 1998. Identification of estrogenic chemicals in STW effluent: 2. In vivo responses in trout and roach. Environ. Sci. Technol.
32:1559–1565. doi:10.1021/es970796a
Sangsupan, H.A., D.E. Radcliffe, P.G. Hartel, M.B. Jenkins, W.K. Vencill, and
M.L. Cabrera. 2006. Sorption and transport of 17 beta-estradiol and testosterone in undisturbed soil columns. J. Environ. Qual. 35:2261–2272.
doi:10.2134/jeq2005.0401
Sarmah, A.K., G.L. Northcut, and F.F. Scherr. 2008. Retention of estrogenic
steroid hormones by selected New Zealand soils. Environ. Int. 34:749–
755. doi:10.1016/j.envint.2007.12.017
USDA–NASS. 2011. Poultry: Production and value—2010 Summary. USDA
National Agricultural Statistics Service. Available at http://www.nass.
usda.gov (verified 27 Sept. 2011).
Wolf, D.C., T.H. Dao, H.D. Scott, and T.L. Lavy. 1989. Influence of sterilization methods on selected soil microbiological, physical, and
chemical properties. J. Environ. Qual. 18:39–44. doi:10.2134/
jeq1989.00472425001800010007x
Ying, G.G., R.S. Kookana, and Y.J. Ru. 2002. Occurrence and fate of hormone
steroid in the environment. Environ. Int. 28:545–551. doi:10.1016/
S0160-4120(02)00075-2
Yu, Z., B. Xiao, W. Huang, and P. Peng. 2004. Sorption of steroid estrogens to soils and sediments. Environ. Toxicol. Chem. 23:531–539.
doi:10.1897/03-192
Journal of Environmental Quality • Volume 40 • November–December 2011