Arsenic solubility and distribution in poultry waste and long term amended soil

The Science of the Total Environment 320 (2004) 51–61
0048-9697/04/$ - see front matter ᮊ 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0048-9697(03)00441-8
Arsenic solubility and distribution in poultry waste and long-term
amended soil
F.X. Han *, W.L. Kingery , H.M. Selim , P.D. Gerard , M.S. Cox , J.L. Oldham
a,b, acda a
Department of Plant and Soil Sciences, Mississippi State University, Mississippi State, MS 39762, USA
a
Diagnostic Instrumentation and Analysis Laboratory, Mississippi State University, 205 Research Blvd., Starkville, MS 39759,
b
USA
Department of Agronomy, Louisiana State University, Baton Rouge, LA 70803, USA
c
Experimental Statistics Unit, Mississippi State University, Mississippi State, MS 39762, USA
d
Received 21 February 2003; accepted 15 July 2003
Abstract
The purpose of this study was to quantify the solubility and distribution of As among solid-phase components in
poultry wastes and soils receiving long-term poultry waste applications. Arsenic in the water-soluble, NaOCl-
extractable (organically bound),NHOHØHCl-extractable (oxide bound) and residual fractions were quantified in an
2
Upper Coastal Plain soil (Neshoba County, MS) that received annual waste applications. After 25 years, As in the
amended soil had a mean of 8.4 mg kg compared to 2.68 mg kg for a non-amended soil. Arsenic in the amended
y1 y1
soil was mainly in the residual fraction (72% of total), which is generally considered the least bioavailable fraction.
Arsenic in poultry waste samples was primarily water-soluble (5.3–25.1 mg kg ), representing 36–75% of the total
y1
As. To assess the extent of spatial heterogeneity, total As in a 0.5-ha area within the long-term waste-amended field
was quantified. Soil surface samples were taken on 10-m grid points and results for total As appeared negatively
skewed and approximated a bimodal distribution. Total As in the amended soil was strongly correlated with Fe

oxides, clay and hydroxy interlayered vermiculite concentrations, and negatively correlated with Mehlich III-P, mica
and quartz contents.
ᮊ 2003 Elsevier B.V. All rights reserved.
Keywords: Arsenic; Poultry waste; Fractionation; Solubility; Spatial distribution
1. Introduction
Arsenic (As) has become an increasingly impor-
tant environmental concern due to its potential
carcinogenic properties (Goyer et al., 1995).
Recently, the USEPA announced a decrease in the
allowable As level for drinking water from 50 to
*Corresponding author. Tel.: q1-662-325-2897; fax: q1-
662-325-8465.
E-mail address: (F.X. Han).
10 mgl (USEPA, 2001). Arsenic is added to
y1
poultry diets for the control of coccidial intestinal
parasites and to improve feed efficiency (Moore
et al., 1995; Wershaw et al., 1999). The organo-
arsenical compounds, p-arsanilic acid (4-amino-
phenylarsonic acid) and roxarsone 4
(3-nitro-hydroxyphenylarsonic acid), are typical
feed additives (Wershaw et al., 1999; Jackson and
Miller, 2000). Because these compounds are not
readily absorbed in tissues, they can occur in
52 F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
excreta. Therefore, the As in poultry wastes and
waste-amended soils may be present primarily in
organic forms (Morrison, 1969; Woolson, 1975).
Moore et al. (1998) found that As concentrations
in runoff from poultry-waste-amended soils

increased as application rates increased. In addi-
tion, because long-term applications to relatively
small areas of land have been shown to lead to
soil accumulation of nutrients and trace elements
(Kingery et al., 1994; Han et al., 2000), the study
of As behavior in waste-amended soil systems is
crucial due to potential contamination of surface
and groundwater via runoff and leaching.
Several studies have documented As speciation
and toxicity in soils (Onken and Hossner, 1995;
Cox et al., 1996a). The bioavailability, toxicity
and mobility of As in soil–water–plant systems
are largely determined by its speciation and distri-
bution, or partitioning between the solution and
soil matrix. Moreover, As mobility and possible
release into runoff from waste-amended fields may
be governed by its distribution among various soil
solid-phase components. Arsenic distribution
among solid-phase components in poultry waste
and in waste-amended soils is not well understood.
Sequential dissolutionyextraction techniques, as
opposed to a single extractant, have recently been
adopted as indicators for As binding, mobility and
bioavailability (Wenzel et al., 2001). Since As has
similar physicochemical properties to P in soils,
inorganic P fractionation techniques have been
adapted for soil As (Johnston and Barnard, 1979;
Onken and Adriano, 1997). Moore et al. (1988)
divided As in sediments into oxyhydroxide (Fe
and Mn)-bound, organically bound and sulfide-

bound fractions. Recently, Wenzel et al. (2001)
proposed a fractionation procedure that included
non-specifically adsorbed, specifically sorbed,
amorphous and poorly crystalline Fe and Al
oxides-bound, crystalline Fe and Al hydrous
oxides-bound and residual fractions. Since the
main species of As in poultry waste is as an
organic compound, about which little is know
concerning its binding by solid phases, it is appro-
priate to include an extractant for organic matter
(OM).
The primary objective of this study was to
determine the solubility and distribution of As
among solid-phase components in poultry waste
and waste-amended soils receiving long-term
applications. A second objective was to correlate
soil properties to the spatial distribution of As in
a long-term amended soil.
2. Materials and methods
2.1. Poultry-waste-amended soil and poultry
wastes
Six soil samples (0–20 cm) were taken from a
waste-amended pasture on a poultry farm located
in Neshoba County, Mississippi, where annual
applications had occurred for 25 years. Although
historical records of application rates are not com-
plete, recent measurements of typical application
management indicate that rates were approximately
10 Mg ha per application, one to three times
y1

each year (Curtis, 1998). The pasture consisted
predominantly of bermudagrass (Cynodon dacty-
lon) harvested for hay each summer and ryegrass
(Lolium multiflorum) sown each fall, after shallow
plowing. Cattle grazed during the winter months.
In order to assess the spatial variation of As
resulting from the long-term poultry waste amend-
ments, surface soil samples were collected on a
grid from the waste-amended field. Specifically,
66 surface samples (0- to 5-cm depth) from the
waste-amended field were sampled at 10-m inter-
vals on a 50=100 m grid, located in the center
2
of the field. In addition, surface (0–5 cm) soil
samples were collected from an adjacent, non-
amended forest soil where loblolly pine (Pinus
taeda) grew. Both soils were clayey, mixed, ther-
mic Typic Hapludults (Upper Coastal Plain) from
shale parent material. Properties of both soils were
reported earlier in Han et al. (2000). Soil pH
ranged from 4.7 to 6.3. The amended soil had
higher pH, OM and CEC than the non-amended
soil (Table 1)(Curtis, 1998).
Soil samples were air-dried and ground to pass
a 2-mm sieve. All soil samples were analyzed for
total As. Furthermore, we carried out analysis for
water-soluble As in 10 randomly selected surface
samples from our grid scheme.
Poultry waste samples were collected from two
locations (Marshall and Newton counties) in Mis-

53F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
Table 1
Selected properties of the non-amended and poultry-waste-amended surface soils
Property Non-amended soil Amended soil
pH 4.74–5.06 5.15–6.28
CEC (cmol kg )
y1
c
39.5"10.1
a
107.7"14.8
SSA (mg )
2 y1
179"64 187"41
OM (%)
b
3.4 4.45
Mineralogy of the clay fraction (%)
b
2:1 type 51.6 46.5
Mica 10.6 15.9
Kaolinite 25 19.3
Quartz 3.6 12.2
Fe O
23
0.7 1.0
Mean of five samples and standard error.
a
From Ref. Curtis (1998).
b

Table 2
Total and water-soluble As concentrations in poultry wastes on oven-dry basis
County Year Total As Water-soluble Water-soluble As
(mg kg )
y1
(mg kg )
y1
(% of total)
Newton 2000 32.6 (0.0) 22.3 (1.1) 68
31.1 (0.3) 20.7 (0.1) 67
33.3 (0.5) 25.1 (0.9) 75
31.8 (1.1) 21.2 (2.2) 67
31.3 (0.1) 22.5 (2.3) 72
30.9 (1.3) 23.2 (0.2) 75
Marshall 1997 32.0 (0.4) 19.8 (2.0) 62
1998 36.1 (0.1) 24.5 (1.4) 68
1999 26.7 (1.0) 9.7 (0.5) 36
11.1 (0.1) 6.0 (0.2) 54
34.2 (0.2) 15.6 (0.0) 46
12.4 (0.7) 5.3 (0.1) 43
28.0 (0.2) 12.5 (0.8) 45
2000 17.8 (0.4) 8.1 (0.4) 46
19.2 (1.1) 9.5 (1.2) 50
22.0 (0.5) 13.0 (0.1) 59
Summary
Average 26.9 (7.8) 15.0 (7.6) 58 (13)
Maximum 36.2 25.7 75
Minimum 11.1 5.3 36
Average followed by standard deviation in parentheses.
sissippi between 1997 and 2000 (Table 2). In 1999

and 2000, multiple, composite waste samples were
collected throughout the year. The samples were
air-dried and ground to pass a 1-mm sieve. Solid-
phase As fractionation was conducted on 10 sam-
ples. This study focused on overall As solubility
and its solid-phase distribution in both poultry
wastes and long-term waste-amended soils, and no
attempt was made to differentiate arsenate and
arsenite forms.
2.2. Analytical methods
Arsenic in soils and poultry wastes was meas-
ured in four operationally defined solid-phase frac-
54 F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
tions, which were obtained by selective sequential
dissolution. This method is based on both solubil-
ity of individual solid-phase components and the
selectivity and specificity of chemical reagents.
The procedure provides a gradient for the physi-
cochemical association between trace elements and
solid particles rather than actual chemical specia-
tion (Martin et al., 1987). But, it can nonetheless
provide an indication of their relative availability
to plants or to further migration with percolating
andyor runoff water. The terms of all fractions are
more appropriately considered to be operationally
rather than chemically defined (Han et al., 2001).
Each extractant in the sequential selective proce-
dures is assumed to effectively target one major
solid-phase component. It is recognized that no
extractant can remove all of a targeted solid-phase

component without attacking other components.
No selective dissolution scheme can be considered
completely accurate in distinguishing between dif-
ferent forms of an element, i.e. various organic
inorganic solid-phase components. Despite possi-
ble re-adsorption during sequential extraction,
common to any chemical extraction procedure,
sequential dissolution techniques still furnish use-
ful information on metal binding, mobility and
availability (Han et al., 2001) . The fractionation
procedures employed in this study were modified
from the sequential selective procedures developed
by MacLeod et al. (1998), Shuman (1983), Moore
et al. (1998) and Sposito et al. (1982). Arsenic in
both waste-amended soil and wastes was fraction-
ated into water-soluble, NaOCl-extractable,
NH OHØHCl-extractable and residual fractions (4
2
M HNO ).
3
(1) Water-soluble arsenic: Twenty milliliter of
distilled water was added to2gofsoil or waste
(oven-dry weight basis) in 50-ml Teflon centrifuge
tube and the mixture was shaken for 30 min at 25
8C. The sample was then centrifuged at 10 000=g
and the supernatant decanted and filtered through
a 0.45-mm filter. The soil residue was kept for the
subsequent extraction. The same centrifugation–
decantion steps were used after each of the follow-
ing extractions.

(2) NaOCl-extractable arsenic: Arsenic extract-
ed in this way may be mainly bound to OM
(Shuman, 1983). Twenty milliliter of 0.7 M NaOCl
solution at pH 8.5 (pH adjusted with NaOH and
HCl) was added to the soil residue. The mixture
was boiled in a water-bath at 95–100 8C for 30
min. During digestion, the mixture was continu-
ously stirred.
(3) NH OHØHCl-extractable arsenic: Arsenic
2
extracted in this step may be mostly bound to
oxides (Han and Banin, 1997): Twenty milliliter
of 0.04 M NH OHØHClq25% HOAc solution was
2
added to the soil residue and boiled in the water-
bath at 100 8C for 3 h.
(4) Arsenic in the residual fraction (RES):
Twenty milliliter of 4 M HNO solution was added
3
to the residue and the sample transferred to a glass
digestion tube. Digestion was conducted in a water-
bath at 80 8C for 16 h (Sposito et al., 1982; Han
and Banin, 1997). This fraction includes As that
was not extracted in the previous steps and repre-
sents the very stable fraction in soil and wastes.
Total As was extracted with heating with
HNO –H SO (Ganje and Rains, 1982). Amended
324
and non-amended soils were analyzed both by
Galbraith Laboratory (Knoxville, TN) and Missis-

sippi State Chemical Laboratory (Mississippi
State, MS) for total As concentrations. The results
from the two laboratories were very consistent. A
large number of samples were analyzed by Missis-
sippi State Chemical Laboratory. Arsenic concen-
trations in the extracts were determined using
graphite furnace atomic absorption spectroscopy
(GFAAS)(Perkin Elmer, Norwalk, CT) at 193.7
nm wavelength with background correction. A
mixed matrix modifier containing 0.015-mg Pd
and 0.01-mg Mg(NO ) was used for each 20-ml
32
standard or sample solution (Perkin Elmer, 1995).
Arsenic concentration in each soilywaste sample
was analyzed in duplicate.
Due to perceived relationship between P and As
mobility (Peryea, 1991), Mehlich III-extractable P
was also determined (Mehlich, 1982). Organic
carbon in soils was measured by wet combustion
(Nelson and Sommers, 1982). Iron oxides were
extracted by citrate-dithionate-bicarbonate (Dixon
and White, 1997) and the Fe was determined by
atomic absorption spectroscopy. Soil pH was meas-
ured in 1:1 soilywater ratio using a combination
glass pH electrode. Particle size distribution was
determined with a hydrometer (Day, 1965).
55F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
2.3. Mineralogical analyses
Quantitative analyses of minerals in clay frac-
tion of soils were determined following the meth-

ods of Karathanasis and Hajek (1982) and
Karathanasis and Harris (1994). Samples were
pretreated to remove salts, carbonates, OM and Fe
oxides, and the fine clay-sized fraction (-0.2 mm)
was collected by sieving and centrifugation (Jack-
son, 1956; Dixon and White, 1997). Clays were
saturated with Mg and K by washing with 1 M
MgCl or KCl, respectively. The Mg-clay was
2
solvated with glycerol, and the K-clay was sequen-
tially heated to 300 and 500 8C for 4 h before
analysis by X-ray diffraction (XRD). A Philips
X’Pert-MPD PW 3050 diffractometer (Philips
Electronics, Almelo, The Netherlands) equipped
with a ceramic long, fine focus copper anode tube
was used for XRD analysis. Samples were step-
scanned from 28 to 308 2u at 1 s per step with a
step size of 0.038 2u. Differential scanning calor-
imetric (DSC) analysis (DSC 910S, TA Instru-
ments Inc., New Castle, DE) was conducted on
the Mg-clays that had been equilibrated at 54%
relative humidity. Magnesium-clay was first cooled
to 5 8C, and then heated in covered aluminum
pans from 5 to 625 8Cat108C min in N
y1
2
atmosphere. An empty, covered aluminum pan was
used as the reference (Karathanasis and Harris,
1994).
2.4. Correlation analysis

Pearson Product Moment correlation coefficients
were calculated using available software (SAS
Institute Inc., 1989) for all pairs of variables: pH,
organic carbon, Fe oxide, particle size distribution,
Mehlich III-P concentrations, total As concentra-
tions and clay mineralogical composition. The
effects of soil properties on As accumulation in
the amended field were estimated.
3. Results and discussion
3.1. Arsenic in amended soil and poultry wastes
Total as well as water-soluble As in poultry
waste samples are presented in Table 2. Total As
in poultry wastes ranged from 11.1 to 36.2
mg kg with an average of 26.9 mg kg . The
y1 y1
As concentrations in more than 50% of the samples
analyzed were between 30 and 35 mg kg , which
y1
were similar to As ranges reported by Moore et
al. (1995). Total As varied with sampling location
and time (Table 2). We sampled poultry waste
from Marshall county, Mississippi, from 1997 to
2000 and found that As concentrations in 2000
were smaller than that in 1997 and 1998. As a
comparison, As concentrations in all samples of
waste were less than the permitted monthly aver-
age concentration of 41 mgAs kg for land appli-
y1
cation of sewage sludge (USEPA, 1994).
Solubility of As in poultry waste, indicated here

by water-soluble As, is linked to its mobility and
toxicity in soilywater systems. Water-soluble As
concentrations in wastes varied from 5.3 to 25.1
mg kg with an average of 15 mg kg over the
y1 y1
period samples were collected (Table 2). This As
accounted for 36–75% of measured total As (Table
2). Some 35% of the samples were in the range
of 20–25 mg kg of water-soluble As. There was
y1
also large variation in As concentrations among
sites and among years of sampling (Table 2).In
addition, water-soluble As concentrations in the
wastes were correlated (r s0.78, P-0.05) with
2
total As (Fig. 1). Jackson et al. (2000) reported
that most water-soluble As was in organo-arsenical
forms, such as p-arsanilic acid and roxarsone.
In the 10 poultry waste samples used for frac-
tionation analysis, As in the water-soluble fraction
was the largest fraction with an average of 47%,
followed by the NH OHØHCl-extractable fraction,
2
which represented 33% of the total (Fig. 2). The
NaOCl-extractable As made up 13% and the resid-
ual fraction accounted for 7% of the total As (Fig.
2). The high solubility of As in poultry waste may
be due to a large portion of it existing as organic
species, and lack of solid-phase components, such
as Fe oxides, with high binding affinity for As. It

has been shown that incorporation of alum into
poultry-house bedding materials significantly
decreases soluble As concentrations in poultry
wastes and in runoff from amended soils (Sims et
al., 2001; Moore et al., 1998).
Arsenic accumulation in surface soils (0–5 cm)
over approximately 25 years of poultry waste
56 F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
Fig. 1. Water-soluble vs. total As concentrations in poultry waste samples.
applications is indicated by the results given in
Table 3. Total As concentration in the amended
soil ranged from 1.7 to 15.2 mg kg with an
y1
average of 8.4 mg kg . By comparison, total As
y1
in the non-amended soils was from 0.59 to 4.5
mg kg and averaged 2.68 mg kg with a stan-
y1 y1
dard deviation of 1.35 mg kg . Thus, total As in
y1
the amended soil was four times greater than that
in the adjacent non-amended soils. If we assume
recent application rates of 10 Mg ha per appli-
y1
cation and two applications per year over the
history of the field (Curtis, 1998), As input is
estimated to be 5.3 mgAs kg in the surface soil
y1
(0–20 cm). In other words, the average As input
rate was approximately 0.54 kgAs ha yr ,

y1 y1
which is equivalent to 0.21 mgAs kg yr in the
y1 y1
top 20 cm of soil. This suggests that annual As
loading at current application rates is below the
annual ceiling rates (2.0 kgAs ha yr ) for safe
y1 y1
land application of sewage sludge (USEPA, 1994).
It should be noted, however, that this field was
plowed annually and therefore subject to a rela-
tively high degree of erosion. This practice is
typical for the region. Removal of As by eroded
soil particles is unknown.
Arsenic in the long-term waste-amended soils
was mostly present in the residual fraction (72%),
followed by NH OHØHCl-extractable fraction
2
(21%) and NaOCl-extractable fraction (6%)(Fig.
2). Compared to As distribution in poultry wastes,
As in the long-term amended soil appears to be in
more stable forms, resulting in decreased As bio-
availability and mobility. This indicates that pos-
sible quick leaching of water-soluble As into
surface water probably occurs shortly after wastes
are applied to fields. Jackson and Miller (2000)
have shown that aryl-organoarsenical compounds
are well adsorbed on amorphous Fe oxides and on
goethite. Arsenic is known to become rapidly
recalcitrant in soil with time after application
(Onken and Adriano, 1997). At present there are

no detailed studies on As distribution in poultry
wastes and poultry-waste-amended soils available
in the literature. However, studies on As solid-
phase fractionation in soils that received long-term
pesticide applications showed oxide-bound As to
be the dominant fraction. It was reported that
oxyhydroxides of Fe, Al, Mn are the primary solid
phases influencing soil As solubility (Woolson et
al., 1971; Johnston and Barnard, 1979; Livesey
and Huang, 1981). However, the present study
indicates that in addition to Fe-oxide-bound As,
the residual As is a major solid-phase fraction in
amended soils.
3.2. Spatial heterogeneity
Results of spatial measurements of total As in
the waste-amended field indicate extensive hetero-
57F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
Fig. 2. Comparisons of distributions of As among solid-phase components in poultry waste and long-term poultry-waste-amended
soil.
Table 3
Total As concentrations and As fractionation in poultry wastes and the surface layer of a long-term poultry-waste-amended and a
non-amended soil
Soilypoultry wastes Depth n
a
Mean Standard deviation Maximum Minimum
(cm)
(mg kg )
y1
Amended soil Total As 0–5 66 8.4 3.5 15.2 1.7
As fraction 0–20 6

Water-soluble As 0.12 0.04 0.18 0.08
Organically bound 0.99 0.644 2.5 0.28
Oxide bound 2.77 1.404 5.77 0.84
Residual fraction 9.52 2.42 12.77 5.45
Non-amended soil Total As 0–5 6 2.68 1.35 4.5 0.59
Poultry wastes Total As 16 26.9 7.8 36.1 11.1
As fraction 10
Water-soluble As 13.9 6.7 25.5 5.2
Organically bound 3.7 2.9 8.1 0.06
Oxide bound 10.7 6.7 20.1 1.6
Residual fraction 2.0 1.2 12.8 1.4
n, sample number.
a
geneity. This is illustrated in Fig. 3 where the
lowest values were in the northeast section and
the highest values of As tended to be in the
southwest section of the field. On the basis of the
coefficient of variation (CV), a high degree of
variability in As concentrations was observed.
Reasons for this variability are not obvious and
reflect non-uniformity of waste applications, soil
adsorption–desorption properties for As, plant
uptake, slopes and others. The CV is consistent
with that of soil Cd measured by Murray and
Baker (1992) of 91.5%. This indicates that trace
element concentrations in field soils are highly
variable. Test for normality using frequency distri-
butions and the Kologorov–Smirnov D-statistic
suggested that As concentrations were not normal-
ly distributed. Such finding has been reported by

others. For example, Murray and Baker (1992)
58 F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
Fig. 3. Spatial distribution of total As (mg kg ) in surface soils of a long-term poultry-waste-amended field. The water flows of
y1
two small streams were indicated as arrows.
showed that total Cd concentrations taken on a
15.2=15.2 m grid from a 2.1-ha site were nega-
2
tively skewed and approximated a lognormal dis-
tribution. A histogram of As concentrations from
our waste-amended field is shown in Fig. 4. It
suggests that As distribution is somewhat nega-
tively skewed with an apparent bimodal distribu-
tion. We are not aware of such distributions for
heavy metal observations at the field scale. If
spatial analysis of As data shown in Fig. 3 is
desired, such as in ordinary kriging, variogram
analysis of the data is necessary. Semi-variogram
analysis (not shown) exhibited a gradual increase
and then leveling off and reaching a sill after four
separation distances or lags.
3.3. Correlation with soil properties
Total As concentrations in poultry-waste-amend-
ed surface soils were positively correlated with
clay content, Fe oxide and hydroxy-interlayer ver-
miculite content, and negatively correlated with
mica, quartz, silt and Mehlich III-P concentrations
(Table 4). In the clay fraction, hydroxy interlay-
59F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
Fig. 4. Frequency distribution of total As in a 0.5-ha site within a long-term poultry-waste-amended field.

Table 4
Correlation analyses of total As concentrations with selected soil properties and clay mineral composition in the clay fraction (ns
60) of poultry-waste-amended field
Properties As pH Organic Fe O
23
Clay Silt Sand P
Mehlich
HIV Kaol Mica
(mg kg )
y1
carbon (%)(%)(%)(%)(%)(mg kg )
y1
(%)(%)(%)
pH 0.16
Organic carbon 0.18 0.28
*
Fe O
23
0.64
*
0.22 0.39
*
Clay 0.60
*
0.20 0.61
*
0.70
*
Silt y0.35
*

y0.09 y0.49
*
y0.40
*
y0.69
*
Sand y0.20 y0.10 y0.04 y0.25
*
y0.17 y0.59
*
P
a
Mehlich
y0.31
*
0.07 y0.04 y0.23 y0.15 0.21 y0.11
HIV 0.34
*
0.21 0.30
*
0.39
*
0.51
*
y0.37
*
y0.08 y0.02
Kaol 0.14 0.14 0.14 0.20 0.17 y0.16 0.03 y0.05 y0.03
Mica y0.52
*

y0.06 y0.35
*
y0.47
*
y0.61
*
0.72
*
y0.29
*
0.42
*
y0.52
*
y0.27
*
Quartz y0.52
*
y0.20 y0.42
*
y0.53
*
y0.58
*
0.62
*
y0.18 0.43
*
y0.45
*

0.09 0.70
*
P , HIV and Kaol represent Mehlich P in the soils, and HIV and kaolinite in the clay fraction, respectively.
a
Mehlich III
Correlation is significant at P-0.05 level.
*
ered vermiculite (HIV) was the major mineral,
followed by kaolinite and mica (Curtis, 1998, data
not shown). These correlations suggest that more
As may accumulate in soils with higher clay
contents.
Total As was negatively correlated with Mehlich
III-extractable P in the poultry-waste-amended soil
(Table 4). Enhanced As mobility, phytoavailability
and phytotoxicity were reported in lead arsenate-
contaminated soils amended with monoammonium
phosphate (Peryea, 1991). Arsenic is adsorbed on
soil mineral surfaces through ligand exchange with
surface hydroxide or hydrated metal-oxide miner-
als (Goldberg, 1986). Phosphate and AsO
3y
4
exhibit similar physicochemical behavior in soils
and compete directly for specific adsorption sites
in soil particles (Hingston et al., 1972; Woolson,
1983). Total P in the poultry-waste-amended sur-
face soil was 2000 mg kg and Mehlich III-P
y1
was 500 mg kg (Curtis, 1998). Thus, As solu-

y1
bility and mobility in both waste and waste-
amended soils may be enhanced by these high
60 F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
concentrations of P. Similarly, Cox et al. (1996b)
showed that addition of As increased solution P
concentrations in the soil.
4. Conclusions
At the current application rates, arsenic accu-
mulated over 25 years of poultry waste applica-
tions is estimated at 5.9 mgAs kg in the surface
y1
20 cm or with an input rate of approximately 0.54
kgAs ha yr . This annual As loading is much
y1 y1
lower than the rates established by USEPA for
land application of sewage sludge. Moreover,
arsenic in the amended soil was mainly in the
residual fraction (72% of total), which is the least
susceptible fraction to runoff losses as soluble As
or downward movement. However, since As in the
applied poultry waste was primarily water-soluble
(5.3–26 mg kg ), representing 36–75% of total
y1
As, an excessive application of poultry wastes per
time could release soluble As from amended fields.
Assessment of the extent of spatial heterogeneity
revealed that distribution of total As in the amend-
ed soil appeared negatively skewed and approxi-
mated a bimodal distribution. We also found that

total As in the amended soil was strongly corre-
lated with Fe oxide, clay and HIV concentrations,
and negatively correlated with Mehlich III-P, mica
and quartz contents.
References
Cox MS, Bell PF, Kovar JL. Differential tolerance of canola
to arsenic when grown hydroponically or in soil. J Plant
Nutr 1996;19:1599–1610.
Cox MS, Bell PF, Kovar JL. Arsenic supply characteristics of
four cotton-producing soils. Plant Soil 1996;180:11–17.
Curtis JL. The distribution and dynamics of phosphorus in an
agricultural watershed with a long-term history of poultry
waste application. M.S. Thesis. Mississippi State University,
Mississippi State, MS, 1998.
Day PR. Particle fractionation and particle size analysis. In:
Black CA, editor. Methods of soil analysis. Part I. Agron.
Monogr. 9. Madison, WI: ASA, 1965. p. 552–562.
Dixon JB, White GW. Soil mineralogy laboratory manual. Soil
and Crop Sci. Dept., Texas A&M Univ, College Station,
TX; 1997.
Ganje TJ, Rains DW. Arsenic. In: Page AL, Miller RH, Keeney
DR, editors. Methods of soil analysis. Part 2, 2nd ed.
Madison, WI: ASA, 1982. p. 385–402.
Goldberg S. Chemical modeling of arsenate on aluminum and
iron oxide minerals. Soil Sci Soc Am J 1986;50:1154–
1157.
Goyer RA, Klaassen CD, Waalkes MP. Metal toxicology. San
Diego, CA: Academic Press, 1995.
Han FX, Banin A. Long-term transformations and redistribu-
tion of potentially toxic heavy metals in arid-zone soils. I:

Under saturated conditions. Water Air Soil Pollut
1997;95:399–423.
Han FX, Kingery WL, Selim HM, Gerard P. Accumulation of
heavy metals in a long-term poultry waste-amended soil.
Soil Sci 2000;165:260–268.
Han FX, Banin A, Triplett GB. Redistribution of heavy metals
in arid-zone soils under a wetting–drying soil moisture
regime. Soil Sci 2001;166:18–28.
Hingston FJ, Posner AM, Quirk JP. Competitive adsorption of
negative charged ligands on oxide surfaces. Faraday Discuss
Chem Soc 1972;52:334–342.
Jackson BP, Miller WP. Effectiveness of phosphate and hydrox-
ide for desorption of arsenic and selenium species from iron
oxide. J Environ Qual 2000;64:1616–1622.
Jackson BP, Bertsch PM, Seaman JC. Fate of p-arsanilic acid
and roxarsone in soils. 2000 agronomy abstracts. Madison,
WI: ASA, 2000. p. 225.
Jackson ML. Soil chemical analysis—advanced course. Mad-
ison, WI: Department of Soils, University of Wisconsin,
1956.
Johnston SE, Barnard WM. Comparative effectiveness of
fourteen solutions for extracting arsenic from four western
New York soils. Soil Sci Soc Am J 1979;43:304–308.
Karathanasis AD, Harris WG. Quantitative thermal analysis of
soil materials. In: Amonette JE, Zelazny LW, editors. Quan-
titative methods in soil mineralogy. Madison, WI: SSSA,
1994. p. 360–411.
Karathanasis AD, Hajek BF. Revised methods for rapid quan-
titative determination of minerals in soil clays. Soil Sci Soc
Am J 1982;46:419–425.

Kingery WL, Wood CW, Delaney DP, Williams JC, Mullins
GL. Impact of long-term land application of broiler litter on
environmentally related soil properties. J Environ Qual
1994;23:139–147.
Livesey NT, Huang PM. Adsorption of arsenate by soils and
its relationship to selected chemical properties and anions.
Soil Sci 1981;131:88–94.
MacLeod F, McGaw BB, Shand CA. Sequential extraction of
selenium from four Scottish soils and a sewage sludge.
Commun Soil Sci Plant Anal 1998;29:523–534.
Martin JM, Nirel P, Thomas AJ. Sequential extraction tech-
niques: promises and problems. Mar Chem 1987;22:313–
341.
Mehlich A. Mehlich 3 soil test extraction: a modification of
Mehlich 2 extraction. Commun Soil Sci Plant Anal
1982;15:1409–1416.
Moore JN, Ficklin WH, Johns C. Partitioning of arsenic and
metals in reduced sulfidic sediments. Environ Sci Technol
1988;22:432–437.
61F.X. Han et al. / The Science of the Total Environment 320 (2004) 51–61
Moore PA Jr., Daniel TC, Sharpley AN, Wood CW. Poultry
manure management: environmentally sound options. J Soil
Water Conserv 1995;50:321–327.
Moore PA Jr., Daniel TC, Gilmour JT, Shreve BR, Edwards
DR, Wood BH. Decreasing metal runoff from poultry litter
with aluminum sulfate. J Environ Qual 1998;27:92–99.
Morrison JL. Distribution of arsenic from poultry litter in
broiler chickens, soil and crops. J Agri Food Chem
1969;17:1288–1290.
Murray MR, Baker DE. Use of probability spatial analysis in

assessing soil cadmium contamination. In: Iskandar IK,
Selim HM, editors. Engineering aspects of metal waste
management. Boca Raton, FL: CRCyLewis Publishers,
1992. p. 25–47.
Nelson DW, Sommers LE. Total carbon, organic carbon and
organic matter. In: Page AL, Miller RH, Keeney DR, editors.
Methods of soil analysis. Part 2, 2nd ed. Madison, WI:
ASA and SSSA, 1982. p. 539–580.
Onken BM, Hossner LR. Plant uptake and determination of
arsenic species in soil solution under flooded conditions. J
Environ Qual 1995;24:373–381.
Onken BM, Adriano DC. Arsenic availability in soil with time
under saturated and subsaturated conditions. Soil Sci Soc
Am J 1997;61:746–752.
Perkin Elmer. The THGA graphite furnace techniques and
recommended conditions. Norwalk, CT: Perkin Elmer Corp,
1995.
Peryea FJ. Phosphate-induced release of arsenic from soils
contaminated with lead arsenate. Soil Sci Soc Am J
1991;55:1301–1306.
SAS Institute Inc. SASySTAT User’s Guide. Release 6. 4th
ed. Cary, NC; 1989.
Sims JT, Luka-Mccafferty NJ, Syme L. Alum as a poultry
litter amendment: I. Farm-scale evaluation of alum effects
on litter properties. In: 2001 Agronomy abstract. Madison,
WI: ASA; 2001.
Shuman LM. Sodium hypochlorite methods for extracting
microelements associated with soil organic matter. Soil Sci
Soc Am J 1983;47:656–660.
Sposito G, Lund LJ, Chang AC. Trace metal chemistry in

arid-zone field soils amended with sewage sludge: I. Frac-
tionation of Ni, Cu, Zn, Cd, and Pb in solid phases. Soil
Sci Soc Am J 1982;46:260–264.
USEPA. EPA land application of sewage sludge: a guide for
land appliers on the requirements of the federal standards
for the use or disposal of sewage sludge, 40 CFR Part 503;
1994. (Available on-line with update at http:yy
www.epa.govyowmypubb.htm.)(Verified 10 Nov. 2001.).
USEPA. EPA announces arsenic standard for drinking water
of 10 parts per billion; 2001 (Available on-line with update
at http:yyyosemite1.epa.govy opayadmpress.nsf? Open-
Database&Starts1&Counts30&Expands2.)(Verified 22
Nov. 2001.).
Wenzel WW, Kirchbaumer N, Prohaska T, Stingeder G, Lombi
Z, Adriano DC. Arsenic fractionation in soils using an
improved sequential extraction procedure. Anal Chim Acta
2001;436:309–323.
Wershaw RL, Garbarino JR, Burkhardt MR. Roxarsone in
natural water systems. In: Wilde FD et al., editors. Proceed-
ings of the Conference on the Effects of animal feeding
operations on water resources and the environment. Fort
Collins, Colorado, 30 August–1 September. US Geological
Survey Open-File Report 00-204; 1999. p. 95–96.
Woolson EA, Axley JH, Kearney PC. The chemistry and
phytotoxicity if arsenic in soils: I. Contaminated field soils.
Soil Sci Soc Am J 1971;35:938–943.
Woolson EA. The persistence and distribution of arsanilic acid
in three soils. J Agro Food Chem 1975;23:677–681.
Woolson EA. Emissions, cycling and effects of arsenic in soil
ecosystems. In: Fowler BA, editor. Biological and environ-

mental effects of arsenic. New York: Elsevier, 1983. p. 51–
139.