ARTICLE pubs.acs.org/est
Dissipation and Sequestration of the Veterinary Antibiotic Sulfadiazine and Its Metabolites under Field Conditions Ingrid Rosendahl,*,† Jan Siemens,† Joost Groeneweg,‡ Elisabeth Linzbach,§,† Volker Laabs,||,† Christina Herrmann,^,† Harry Vereecken,‡ and Wulf Amelung† †
Institute of Crop Science and Resource Conservation, Soil Science and Soil Ecology, University of Bonn, Nussallee 13, D-53115 Bonn, Germany ‡ Institute of Bio- and Geosciences 3, Agrosphere, Forschungszentrum J€ulich GmbH, D-52425 J€ulich, Germany
bS Supporting Information ABSTRACT: Veterinary antibiotics introduced into the environment may change the composition and functioning of soil microbial communities and promote the spreading of antibiotic resistance. Actual risks depend on the antibiotic’s persistence and (bio)accessibility, which may differ between laboratory and field conditions. We examined the dissipation and sequestration of sulfadiazine (SDZ) and its main metabolites in soil under field conditions and how it was influenced by temperature, soil moisture, plant roots, and soil aggregation compared to controlled laboratory experiments. A sequential extraction accounted for easily extractable (CaCl2-extractable) and sequestered (microwave-extractable, residual) SDZ fractions. Dissipation from both fractions was largely temperature-dependent and could be well predicted from laboratory data recorded at different temperatures. Soil moisture additionally seemed to control sequestration, being accelerated in dry soil. Sequestration, as indicated by increasing apparent distribution coefficients and decreasing rates of kinetic release into CaCl2, governed the antibiotic’s long-term fate in soil. Besides, we observed spatial gradients of antibiotic concentrations across soil aggregates and in the vicinity of roots. The former were short-lived and equilibrated due to aggregate reorganization, while dissipation of the easily extractable fraction was accelerated near roots throughout the growth period. There was little if any impact of the plants on residual SDZ concentrations.
’ INTRODUCTION Veterinary antibiotics like the sulfonamide sulfadiazine (SDZ, see Supporting Information (SI), Table S1 for chemical properties), introduced into soils with manure, may change the structure of microbial communities, modify the biogeochemical cycling of carbon and nutrients, and promote the spreading and formation of antibiotic resistance.1,2 The biological effects and the mobility of veterinary antibiotics in the environment are to a large extent governed by their binding in soil3 and subsequent sequestration. In this context, “sequestration” summarizes all mechanisms decreasing the extractability and thus bioaccessibility of the compound in soil.4 In a laboratory experiment, the dissipation half-lives (DT50) of SDZ and its main metabolites 4-hydroxy-SDZ (4-OH-SDZ) and N-acetyl-SDZ (N-Ac-SDZ) extracted from soil in the first (0.01 M CaCl2) and second (methanol) step of a sequential extraction amounted to 632 d.5 The concentrations of SDZ and 4-OH-SDZ in a third microwave-extractable (“residual”) fraction increased concomitantly, hence this fraction represented a pool of sequestered antibiotic residues with DT50 exceeding several months.5 Nonextractable residues (NER) amounted to approximately 40% of applied radioactivity5 and mineralization of the parent compound to r 2011 American Chemical Society
14
CO2 was negligibly small.6 Altogether, these results suggested that the long-term fate and effects of SDZ and its metabolites in soil are controlled by their sequestration into hardly extractable and nonextractable forms. This may lead to an accumulation of antibiotic concentrations in soil when repeated applications of manure occur. Kinetic modeling, however, suggested that the underlying sequestration mechanisms are at least partly reversible, so that SDZ could be released back into accessible forms,7 thus conserving the biological effectiveness of SDZ in soil over extended periods of time. Any compound that is persistent under laboratory conditions is not necessarily persistent under field conditions, where e.g. variations of soil temperature8 and cycles of freezing and thawing9 or wetting and drying8,10 may affect dissipation or sequestration of organic contaminants. The impact of these variables on the environmental fate of antibiotics, however, has hardly been investigated so far. Besides, there is an additional Received: January 27, 2011 Accepted: April 28, 2011 Revised: April 18, 2011 Published: May 19, 2011 5216
dx.doi.org/10.1021/es200326t | Environ. Sci. Technol. 2011, 45, 5216–5222
Environmental Science & Technology contribution of distinctive field features to the fate of antibiotics which is rarely covered in lab studies. Preferential deposition of organic pollutants on the surface of soil aggregates and slow diffusion into their interior, for instance, contributed to spatial gradients of pollutant concentrations in soil.11,12 Yet, higher biological activity on aggregate surfaces13 might also accelerate the dissipation of organic contaminants. Another hotspot of microbial activity is located in the vicinity of plant roots,14 and there is evidence that the dissipation14 and the sequestration15 of organic pollutants are altered in the rhizosphere. To better understand the fate of SDZ under field conditions, we conducted an experiment with repeated applications of manure from SDZ-treated pigs on grass or maize plots. Relating our observations to results from controlled laboratory experiments, our aim was to explore the effects of field conditions (variable soil moisture and temperature, plants, and soil aggregation) on the dissipation and sequestration of SDZ and its main metabolites in soil. Soil concentrations of SDZ and its metabolites were analyzed in fractions of different binding strength by sequential extraction. The influence of plant roots and soil aggregation was investigated by selective sampling of rhizosphere soil, soil aggregate surfaces, and aggregate cores. Our hypotheses were that (i) dissipation and sequestration of SDZ are enhanced under field conditions compared to laboratory conditions, (ii) soil aggregation forms microsites of enhanced dissipation and sequestration of SDZ, and (iii) plant roots accelerate the dissipation of easily extractable SDZ.
’ EXPERIMENTAL SECTION Field Trial. The field trial was conducted from May 2009 until June 2010 near J€ulich (Western Germany, 50550 48,7700 N, 6170 20,0200 E) on a Luvisol16 described earlier.5 The site has a mean annual precipitation of 700 mm and a mean annual temperature of 9.9 C. A meteorological station recorded weather conditions during the trial period (see SI, Figure S1). Experimental grassland and maize plots (3 * 6 m, n = 4) were established in a randomized block design in March 2009. Pig slurry was produced separately for each of three applications by intramuscular administration of the prescribed dose of SDZ (30 mg kg1 bodyweight) on four consecutive days and collection of slurry for ten days. SDZ injection solution (200 mg mL1) was kindly provided by Vetoquinol Biowet, Gorzow Wielkopolski, Poland. An amount of 30 m3 ha1 of the collected manure was applied on 19 May (day 0, see SI, Table S2 for antibiotic concentrations in manure and the amount recovered from soil after application). It was incorporated to a depth of 12 cm on maize plots before the sowing of maize. The manure was not incorporated on the grassland plots. A second manure application (10 m3 ha1) between maize rows and on mown grass followed on 7 July (day 49). A third application (10 m3 ha1) on harvested maize plots and on mown grass took place on 29 September (day 133). Soil samples were collected on days 0, 2, 7, 14, 28, 42, 48, 49, 56, 63, 77, 105, 132, 133, 140, 147, 161, 175, 288, and 378 (maize only) after the first application of manure. Maize bulk soil was sampled between rows using a soil sampling cylinder (5.6 cm inner diameter * 12 cm height) by mixing three subsamples. Rhizosphere soil was defined as soil adhering to the roots after gentle shaking. It was sampled by digging out plants (maize: g 3 plants, grass: approximately 0.04 m2 * 5 cm deep), shaking, cutting roots
ARTICLE
plus remaining soil into small pieces, and thorough mixing. Soil falling from the grass plants during shaking was collected as grass bulk soil. The development of maize plants allowed the collection of sufficient amounts of rhizosphere soil from day 42 until day 132 only. Samples were stored on ice and immediately transported to the laboratory, where they were kept at 25 C until further processing. Roots were removed from rhizosphere soil before extraction. Effects of soil structure were assessed on a fallow plot in a separate trial, beginning on the same day. Manure was only applied twice (30 m3 ha1 on day 0 and 10 m3 ha1 on day 133) on one plot (3 * 6 m) and manually incorporated to 7 cm depth. The plot was kept free of plants by weeding. Four replicate samples were collected on days 0, 2, 14, 42, 133 (before manure application), 135, 147, 175, and 252 using a cylinder (10 cm inner diameter * 15 cm height). Soil columns were transported into the laboratory upright, where they were immediately shock-frozen using liquid nitrogen. For the sampling of aggregates, the columns were then thawed shortly to enable removal of the soil core from the cylinders without its disintegration. The soil core was mechanically broken into its constituting aggregates, which were kept at 25 C until aggregate fractionation (adapted from ref 17, see SI for further details). Laboratory Analyses. SDZ and its metabolites were sequentially extracted from soil to obtain fractions of differing binding strength (modified from ref 5). The methanol-extraction performed by F€orster et al.5 was replaced by a second extraction with CaCl2, which proved superior with regard to extracted antibiotic concentrations in preliminary tests. A subsequent exhaustive microwave extraction was used to assess the residual fraction (RES, 5). Extracts were stored at 25 C until LC-MS/MS analysis. Routine limits of quantitation (RLOQ) were 1.25 μg kg1 dry soil for all compounds in the CaCl2-extracts and 2.5 μg kg1 dry soil for all compounds in the residual fraction. Details on the analytical protocol are given in the SI. Data Evaluation. Concentrations below RLOQ were set to zero for mathematical operations and statistical tests. Concentrations in the two CaCl2-extracts were summed up to give the easily extractable (EAS) fraction. The fitting of first-order dissipation models (eq 1) was performed using nonlinear regression (Sigma Plot 11.0, Systat Software GmbH, Erkrath, Germany) CðtÞ ¼ C0 eðktÞ
ð1Þ
with C(t) defining the concentration still present in soil at time t, C0 defining the initial concentration, and k defining the dissipation rate constant. Dissipation half-lives (DT50), i.e. the time required for 50% of the initial concentration to dissipate from a given fraction were calculated from eq 1 with DT50 = ln(2) k1. Adding a second exponential term to eq 1 sometimes yielded a better fit and DT50 were then calculated stepwise from the model equation. Daily dissipation rates under field conditions were predicted from two laboratory experiments at constant temperature (283 and 294 K, details see SI) using an Arrhenius-type equation (eq 2, 18) γ 1= T 1=T
kðTÞ ¼ kref e
ref
ð2Þ
with k as the first-order dissipation rate constant [d1], T as the temperature [K], kref as the reference rate at 283 K, and γ as the empirical constant derived from comparing dissipation 5217
dx.doi.org/10.1021/es200326t |Environ. Sci. Technol. 2011, 45, 5216–5222
Environmental Science & Technology
ARTICLE
Figure 1. Maize bulk soil concentrations (mean of four replicates ( SE) of easily extractable (EAS) and residual (RES) SDZ, 4-OH-SDZ, and N-AcSDZ; continuous lines represent fitted dissipation curves, dashed lines represent temperature-adjusted dissipation predicted from laboratory experiments (eq 2); triangles mark concentrations
