Environ. Sci. Technol. 2003, 37, 1509-1514
Environmental Fate of Roxarsone in Poultry Litter. I. Degradation of Roxarsone during Composting J . R . G A R B A R I N O , * ,† A . J . B E D N A R , † D. W. RUTHERFORD,‡ R. S. BEYER,§ AND R. L. WERSHAW‡ National Water Quality Laboratory, U.S. Geological Survey, P. O. Box 25046 MS407, Denver Federal Center, Denver, Colorado 80225-0046, Branch of Regional Research Central Region, U.S. Geological Survey, P. O. Box 25046 MS408, Denver Federal Center, Denver, Colorado 80225-0046, and Department of Animal Science and Industry, Kansas State University, Manhattan, Kansas, 66506
Roxarsone, 3-nitro-4-hydroxyphenylarsonic acid, is an organoarsenic compound that is used extensively in the feed of broiler poultry to control coccidial intestinal parasites, improve feed efficiency, and promote rapid growth. Nearly all the roxarsone in the feed is excreted unchanged in the manure. Poultry litter composed of the manure and bedding material has a high nutrient content and is used routinely as a fertilizer on cropland and pasture. Investigations were conducted to determine the fate of poultrylitter roxarsone in the environment. Experiments indicated that roxarsone was stable in fresh dried litter; the primary arsenic species extracted with water from dried litter was roxarsone. However, when water was added to litter at about 50 wt % and the mixture was allowed to compost at 40 °C, the speciation of arsenic shifted from roxarsone to primarily arsenate in about 30 days. Increasing the amount of water increased the rate of degradation. Experiments also suggested that the degradation process most likely was biotic in nature. The rate of degradation was directly proportional to the incubation temperature; heat sterilization eliminated the degradation. Biotic degradation also was supported by results from enterobacteriaceae growth media that were inoculated with litter slurry to enhance the biotic processes and to reduce the concomitant abiotic effects from the complex litter solution. Samples collected from a variety of litter windrows in Arkansas, Oklahoma, and Maryland also showed that roxarsone originally present had been converted to arsenate.
Introduction Roxarsone, 3-nitro-4-hydroxyphenylarsonic acid, was added to broiler poultry feed to control coccidial intestinal parasites, improve feed efficiency, and promote rapid growth in about 70% of the broiler industry operations from 1999 to 2000 (1). The concentration of roxarsone in feed formulations ranges from 22.7 to 45.4 g/ton (2). Nearly all the roxarsone is excreted unchanged in the manure, however, 18% of 3-amino-4* Corresponding author phone: (303) 236-3945; fax: (303) 2363499; E-mail:
[email protected]. † National Water Quality Laboratory, USGS. ‡ Branch of Regional Research - Central Region, USGS. § Kansas State University. 10.1021/es026219q CCC: $25.00 Published on Web 03/18/2003
2003 American Chemical Society
hydroxyphenylarsonic acid (as a percentage of the roxarsone dosage) reportedly has been found in fresh manure (3, 4). The amount of roxarsone that is excreted by a single broiler when fed the 45.4 g/ton formulation is estimated to be 150 mg over the typical growth period of 42 days (2). Hence, if 70% of the 8.3 billion broilers grown in the United States during 2000 (5) were fed roxarsone-containing feed, the manure would contain about 9 × 105 kg of roxarsone or 2.5 × 105 kg of arsenic. Many of the top broiler-producing states have regions of concentrated broiler production. For example, the Delmarva Peninsula (DE, MD and VA) produced 620 million broilers in 2000 (6), thereby potentially producing manure with a total of about 2.6 × 104 kg of arsenic. Litter typically is applied as fertilizer at a rate of at least 5 metric tons per hectare to nearby farmland to minimize the cost of disposal. Repeated and intense localized disposal of manure can introduce a substantial amount of arsenic into the environment considering that about 60-250 g of arsenic per hectare can be introduced with every application. Furthermore, problems associated with litter disposal might increase if the U.S. broiler industry maintains the 1990s growth rate of about 6% a year (7). Many regions that are high-poultry-production areas (e.g., Ozark Plateau (AR, MO, and OK) and Delmarva Peninsula) also receive large amounts of rainfall. Wetting of litter during storage will promote growth of endemic microbial populations that could affect the stability of roxarsone, as well as increase its mobility through leaching. Several studies have shown that roxarsone can be easily extracted from poultry litter with water (4, 8, 9), but few studies have investigated the fate of roxarsone or its degradates in the environment. One study showed that the distribution of arsenic in soil fertilized for 20 years with litter was indistinguishable from unamended soil and concluded that the application of poultry litter did not increase the arsenic content of soil (4). Representative poultry-house litter samples were shown to have 30-60 mg/kg of roxarsone, accounting for only 36 to 88% of the total arsenic present (4). Neither the speciation of the arsenic present in the litter nor information on whether roxarsone was actually present in the litter at the time it was applied to the fields were provided. Furthermore, only total arsenic in the amended soil was measured so that the distribution of the arsenic species in the soil was not known. Knowing that a substantial amount of arsenic is being introduced into the environment through the application of poultry litter to cropland, and that 70 to 90% of the arsenic present in poultry litter is water-soluble (8, 9), raises questions on the ultimate fate of arsenic. Investigations were conducted to determine the fate of roxarsone in the environment from poultry litter use. The purposes of this report (Part I) were to identify the arsenic species in fresh litter from poultry fed roxarsone, to study the effects of composting on the distribution of arsenic species in litter, and to investigate the processes that cause changes in arsenic speciation. Part II of the report describes the mobility, concentration, and speciation of arsenic in different soils amended with litter from poultry that were fed roxarsone (10).
Experimental Section Poultry-Litter Samples. The poultry litter was obtained from a broiler diet study conducted in the Department of Animal Science and Industry at Kansas State University (KSU). The use of this litter was advantageous because the composition of the diet, the number of broilers raised, the mass of litter produced, and the handling of the litter after the trial were VOL. 37, NO. 8, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1509
known. The concentration of roxarsone in the corn- and soybean-based diet was 45.4 g/ton, which is within the concentration range recommended; a detailed list of other ingredients in the diet is published elsewhere (11). Forty-five broiler chicks were reared in each of sixteen 1.5 × 1.9 m experimental pens inside a positive-pressure ventilated house. The pen bedding material was composed of fresh wood shavings. A starter diet was fed to the broilers for 3 weeks followed by a grower diet for an additional 3 weeks; roxarsone was removed from the diet 1 week before the end of the trial (7 weeks). All the litter from one pen was immediately collected at the end of the trial, air-dried, disaggregated with a hand mill, and sieved to 2 mm to remove detritus material. Detritus material, composed primarily of feathers and twigs, was not analyzed. The concentration of arsenic in the dried litter after total dissolution was 28.7(0.5 mg/kg (n ) 3) as determined by inductively coupled plasma mass spectrometry (ICPMS). Poultry litter also was obtained from a poultry house in Hudson, Colorado that used feed without roxarsone. The litter sample was prepared in the same manner as the KSU litter. The concentration of arsenic in the dried litter was about 0.6 mg/kg. This poultry litter was used to test the effect of heat sterilization on the degradation of roxarsone. Additional litter samples were obtained from several poultry houses in the Delmarva Peninsula (MD) and the Ozark Plateau (AR, MO, and OK) to determine whether roxarsone was present in litter managed using routine procedures. The age and management practices of the litters were not known, but the litters came from poultry operations that used roxarsone-containing feed. Samples were collected either from inside the poultry houses or from uncovered or covered litter windrows. Whenever available, naturally occurring litter leachates were collected near litter windrows that were exposed to the weather. Analytical Methods. The distribution of arsenic species in test samples was determined using ion chromatography (IC)-ICPMS (9, 12). The arsenic species were separated by gradient elution using a nitric acid in 0.5% methanol mobile phase and a Dionex AS-7 analytical column with an AG-7 guard column. The gradient elution sequence at a flow rate of 1.00 mL/min was 2.5-mM nitric acid for the first 3 min, 50-mM nitric acid for the second 3 min, and 2.5-mM nitric acid for the final 3 min. The injection volume typically was 100 µL. The ICPMS was used to quantify the arsenic species at m/z 75, and the chloride interference on monoisotopic arsenic was corrected by using the standard procedure (13). Total arsenic also was determined using ICPMS; other elemental determinations were determined using inductively coupled plasma atomic emission spectrometry (14). All chemicals used in this study were reagent grade or higher purity and were used without further purification; the reagent water used had a resistivity of 18.3 MΩ-cm. Primary stock solutions of arsenate [As(V)], arsenite [As(III)], monomethylarsonate (MMA), dimethylarsinate (DMA), and roxarsone were prepared in 0.25 mM EDTA, filtered to 0.2 µm, and stored in the dark in designated fluoropolymer bottles at 4 °C. The concentration and species purity of the inorganic arsenic primary standards were verified using commercially available certified reference materials for As(III) and As(V) (Spex CertiPrep, Metuchen, NJ); reference materials were not available for the organoarsenic species. Six mixed-species calibration standards extending over the concentration range of 0 to 100 µg As/L were prepared in 1.25 mM EDTA (15). The arsenic species distribution in calibration standards prepared in this manner were stable for several days in airtight amber glass autosampler vials. The method separates the arsenic species in about 8 min with method detection limits ranging from 0.2 to 0.3 µg As/L (12). 1510
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 8, 2003
Litter Leachate Experiments. The distribution of arsenic species in poultry-litter samples was determined in water leachates. Water leachates of the KSU litter were prepared by mixing 2 g of litter with 20 mL of reagent water for 1 h. The aqueous phase was separated from the solid phase by centrifugation at 5000 relative centrifugal force (RCF) immediately after mixing. An aliquot of the leachate was filtered immediately to 0.45 µm, and analyzed for total arsenic by ICPMS and for arsenic species by IC-ICPMS to determine the initial arsenic species distribution. Water leachates of litter samples collected from the Delmarva Peninsula and the Ozark Plateau were prepared and analyzed using the same procedure. Analyses of these leachates were used only to determine the distribution of arsenic species in actual field samples. Litter Compost Experiments. The effects of composting on roxarsone-containing poultry litter were determined over a 38-d period. Compost was prepared in triplicate by mixing 100 g of dry KSU litter with 120 mL of reagent water (about 50% moisture by weight) in a glass beaker. The temperature of the compost was maintained at 40 °C in a humidified environmental chamber; moisture content was maintained at 50% by adding reagent water to the compost whenever needed. Each compost sample was sampled daily to monitor changes in the distribution of arsenic species. Multiple samples were taken at selected time-points to determine experimental variability. A 1-g sample of the compost was extracted with 20 mL of reagent water for about 1 h at room temperature. The aqueous fraction was separated from the mixture using centrifugation at 5000 RCF followed by filtration to 0.2 µm. The compost extracts were analyzed immediately for total arsenic and for arsenic species by ICPMS and ICICPMS, respectively. The solid compost material remaining after the extraction was not analyzed. An additional compost mixture was prepared to determine whether biotic processes were involved in the degradation of roxarsone. The compost was prepared and incubated as above except that the reagent water used to prepare the compost contained 0.5% sodium azide, a broad-spectrum microbial inhibitor (16), yielding about 0.25% sodium azide, by weight, in the compost. Daily samples of the compost were extracted and analyzed using the same procedures as described above. Litter Slurry Experiments. Litter slurries were different from the leachates described above because the solid litter material remained in contact with the aqueous phase throughout the entire experiment (14 d). The only difference between the slurry and compost samples was the amount of water used in the preparation. Two pairs of slurry samples were prepared by mixing 20 g of KSU litter with 200 mL of reagent water (the same proportion that was used for litter leachates) in a polyethylene bottle for about 1 h. Reagent water containing sodium azide (0.5 wt %) was used to prepare one pair of samples to inhibit biological processes; reagent water was used to prepare the other pair. The slurry samples were allowed to settle by gravity after the mixing period; however, the aqueous solution always remained in contact with the solid material. The capped bottles of slurry were incubated in an environmental chamber at 40 °C throughout the experiment. The aqueous solution of the slurry was sampled daily, filtered to 0.2 µm, and analyzed immediately for total arsenic and arsenic species. The KSU and Hudson litters also were used to investigate the effects of temperature on the degradation of roxarsone. Slurries of the KSU and Hudson litters were prepared by mixing 20 g of litter with 200 mL of reagent water for about 24 h at room temperature to allow enough time for microbial populations to grow. The aqueous supernatant was separated from the solid material using centrifugation at 5000 RCF. Because the Hudson litter did not contain roxarsone,
roxarsone was spiked into the supernatant at a concentration similar to that in the KSU supernatant (about 1700 µg As/L). Aliquots of the KSU and spiked Hudson supernatants were incubated in tightly capped sterilized culture tubes at 15, 20, and 40 °C. The incubation was considered to be anaerobic even though about 10% of the culture tube headspace was air. Another aliquot of the unspiked Hudson supernatant was heat-sterilized prior to spiking with roxarsone, capped in a culture tube, and stored at room temperature. Daily samples were taken from each aliquot, filtered to 0.45 µm, diluted into the calibration range with reagent water, and analyzed immediately for arsenic species distribution. Microbial-Growth Medium Experiments. Microbial processes were isolated from possible abiotic reactions in the complex litter leachates by inoculating a microbial growth medium with KSU slurry. The growth medium provided nutrients necessary to stimulate the growth of bacteria present in the KSU slurry. Enterobacteriaceae enrichment broth (Mossel Broth, 1-05394, EM Science, Gibbstown, NJ) was prepared by dissolving 45 g of broth salt in 1 L of reagent water. Two pairs of 20-mL aliquots of the broth were autoclaved at 120 °C and 117 kPa for 60 min in capped culture tubes. One pair of aliquots (the control samples) was spiked to 500 µg As/L with roxarsone, whereas the other pair was spiked with roxarsone and inoculated with 100 µL of KSU slurry. One sample of the control and the inoculated samples were analyzed to establish the initial distribution of arsenic species. The remaining samples were incubated at room temperature and analyzed after 7 d. No attempt was made to isolate, measure the growth rate, or identify the microbes present in the KSU slurry.
Results and Discussion The elemental compositions of the poultry litters used in this study and of those reported in the literature are listed in Table 1. The total arsenic concentration was similar for all the litter samples listed (a total of 14 samples). The most notable differences between the litters were the Cu and Fe concentrations. Elements such as Cu and Zn typically are added to the feed as fungicides and micronutrients. The total arsenic concentration in the KSU litter was about 29 mg/kg in comparison to the range of 12 to 30 mg/kg reported elsewhere (see Table 1) (4, 9, 17). Between 70 and 90% of the total arsenic present was extractable with water. The distribution of arsenic species in the KSU leachate was 91% roxarsone, 1.5% dimethylarsinate (DMA), 1.1% arsenate [As(V)], 0.8% arsenite [As(III)], and 5.6% of other unknown arsenic compounds (see KSU litter in Figure 1). Others have reported similar arsenic species distributions in poultry-litter leachate (9). Clearly, a high percentage of the total arsenic available in fresh dried poultry litter can be mobilized easily with water, and the major arsenic species is roxarsone. However, no studies have reported whether roxarsone in the litter is stable indefinitely or whether its stability is affected by certain litter management practices. The primary arsenic species in the litter leachates for samples collected from poultry operations in the Ozark Plateau and Delmarva Peninsula was not roxarsone but As(V) (for example, see the University of Maryland Agricultural Research Station (ARS) Delmarva Peninsula composted litter in Figure 1). The fact that roxarsone was the major arsenic species in the fresh KSU litter suggests that the degradation process might be controlled by poultry-litter management practices. During litter storage or after litter has been applied to fields, exposure to sunlight, elevated air temperature, and precipitation most likely promote photodegradation and microbial degradation reactions that affect the transformation of roxarsone to As(V). Other arsenic species identified in the litter leachates were As(III) and DMA. Furthermore, several
TABLE 1. Elemental Composition of Various Poultry Litters, in mg/kg
element As Ba Cd Co Cr Cu Fe Mn Ni Pb Se U V Zn
KSUa
ARSb
28.7 ( 0.5 29 ( 3 17.4 ( 0.6 12 ( 2 0.22 ( 0.06 0.26 ( 0.02 0.80 ( 0.08 1.1 ( 0.3 9.2 ( 0.7 11.24 ( 0.01 76.9 ( 0.9 359 ( 9 700 ( 100 1600 ( 400 310 ( 10 410 ( 20 9.7 ( 0.3 8.0 ( 0.1
