Environ. Sci. Technol. 2008, 42, 3102–3108
Evidence for Bioavailable Copper-Dissolved Organic Matter Complexes and Transiently Increased Copper Bioavailability in Manure-Amended Soils as Determined by Bioluminescent Bacterial Biosensors K R I S T I A N K . B R A N D T , * ,† PETER E. HOLM,‡ AND OLE NYBROE† Department of Ecology, Department of Natural Sciences, Faculty of Life Sciences, University of Copenhagen, Thorvaldsensvej 40, DK-1871 Frederiksberg, Denmark
Received August 1, 2007. Revised manuscript received January 31, 2008. Accepted February 5, 2008.
The short-term (3 months) dynamics of bioavailable copper (Cu) species was determined in soils amended with various amounts of manure and Cu. Bioavailable Cu species were operationally defined as those species that were able to induce gene expression in a Cu-specific Pseudomonas fluorescens biosensor. Biosensor measurements were backed by analysis of total Cu in soil and of total Cu and free Cu2+ ion activity in solution. Cu bioavailability relative to the total Cu concentration increased dramatically with increasing Cu loading of manure and with increasing manure amendment to soil. In both cases, the immediate increase in bioavailability could be explained in part by increased Cu concentration in solution and in part by an increased bioavailability of dissolved Cu species. In contrast to Cu bioavailability, Cu2+ ion activity decreased progressively with increasing manure loading. Cu bioavailability declined rapidly during the weeks after manure amendment concomitant with a marked slow-down of C mineralization indicating a shift from initially bioavailable Cu-dissolved organic matter (Cu-DOM) complexes to nonavailable Cu-DOM complexes over time. Our data do not support the conventional view of metal bioavailability being primarily related to the free metal ion activity and strongly suggest differential bioavailability of Cu-DOM complexes in manure-amended soils.
Introduction Copper (Cu) is accumulating in many agricultural soils worldwide (1–3). Application of metal-contaminated manures is increasingly being recognized as a dominant source of Cu inputs to agricultural soils due to the increasing application of Cu as a growth promoter in animal production systems (2). For example, pig manure has been estimated to contribute approximately 50% of the total Cu emissions to the Danish * Corresponding author phone: + 45 3533 2612; fax: + 45 3533 2606; e-mail:
[email protected]. † Department of Ecology. ‡ Department of Natural Sciences. 3102
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 8, 2008
environment resulting in soil Cu accumulation in intensive pig production areas (4). Total Cu concentrations in soil are poor predictors of toxicity to soil biota, and consequently bioavailability-based risk assessment should be considered the ultimate goal for proper environmental protection of soil resources (5, 6). This aspiration represents a serious challenge to the scientific community and calls for detailed studies of bioavailability to different types of organisms. Soil bacterial communities should constitute a primary target group given their higher sensitivity to metal exposure than fungi, plants, and animals (7, 8). Several studies have indicated that Cu added to soil with manures or other biosolids are less available and toxic to plants and microorganisms than Cu added as pure salts (2, 9). Likewise, addition of manures with low metal contents to metal-contaminated soils has been shown to decrease metal availability and toxicity to plants and microorganisms as compared to unamended soils (2). This “biosolids effect” has been attributed to their high proportion of organic matter and amorphous oxide components, which have high capacity for specific sorption of metals and for Cu complexation (2, 10, 11). Both of these mechanisms reduce the chemical activity of the free metal ions in the soil solution, which is generally considered to be the toxic metal species responsible for inhibitory effects in soil biota (7, 9, 12–16). The aim of the present work was to challenge the validity of this conventional view by linking specific chemical descriptors of copper exposure to biological assessment of copper bioavailability using a pair of well-characterized Pseudomonas fluorescens biosensor strains (17–19). Specifically, we monitored the short-term dynamics (3 months) of bioavailable Cu, total dissolved Cu, and free Cu2+ ion activity in soil–water extracts following manure amendment to soil. Cu exposure, as determined by the above panel of descriptors, was further related to microbial soil respiration (i.e., organic C mineralization) following manure amendment.
Materials and Methods Soil and Manure Sampling. Agricultural soil (loamy sand) used for cultivation of cereals (barley and oats) without application of organic manures for several years was obtained from an experimental farm located in Taastrup, Denmark. Soil was sampled from experimental field plots amended with Cu in 1998 and corresponding control plots (20). Surface soil (0–20 cm depth) from triplicate field plots (12 soil cores per plot) were sampled and pooled to yield one composite soil sample with a low total Cu content of 9.5 mg kg-1 (referred to as low-Cu soil hereafter) and one corresponding moderately contaminated soil sample containing 82 mg kg-1 (high-Cu soil). Low-Cu pig manure (33 mg kg-1) was obtained from an organic pig farm in Gørløse (Zealand, Denmark). Soil and Manure Characterization. Elemental composition of used soils and manure was determined by inductively coupled plasma-mass spectrometry (ICP-MS). Total Cu concentrations determined by this method are in the following referred to as [Cu]total. Detailed description of analytical procedures as well as obtained characterization data (Table S1) can be found as part of the Supporting Information. Microcosm Setup. The first series of experiments was designed to monitor the bioavailability of Cu added to the soil as part of manure. These experiments were set up with low-Cu soil amended with pig manure (5% dry wt/dry wt) containing various realistic Cu contents. The second series of experiments addressed the impact of manure amendment 10.1021/es071916+ CCC: $40.75
2008 American Chemical Society
Published on Web 03/18/2008
on the bioavailability of soil Cu. In these experiments the high-Cu soil was amended with 1, 2, or 5% (dry wt/dry wt) of low-Cu manure. In both series of experiments, unamended soil served as control treatments. Both experimental series consisted of two experiments. The first experiment was made with triplicate soil microcosms that were harvested 1 day and 3 months after experimental setup, respectively. At these time points determinations of bioavailable Cu ([Cu]bio; biosensor assay), total solution Cu ([Cu]sol; GFAAS), and free Cu ion activity ([Cu2+]; Cu ion specific electrode) were carried out in soil–water extracts (see below for details). Cu-spiked manure used for the first experimental series was prepared by mixing one part of dry manure powder with 3 parts of Milli Q water or CuSO4 solution to increase the manure Cu content by 0, 200, 500, or 1000 µg of Cu g-1 of dry wt, respectively. The spiked manures were allowed to air-dry for 24 h and reconverted into a dry powder before use. For all experiments, dried manure powders were mixed into 5 g of soil kept in sterile 50 mL plastic tubes (Bibby Sterilin Ltd., Stone, Staffordhire, U.K.). Subsequently, Milli Q water was added to 60% of the water-holding capacity, and the resulting soil microcosms were incubated in the dark at 15 °C under aerobic conditions. Both experimental series included a second experiment to study the effect of time on bioavailability and organic C mineralization following manure amendment to soil. All microcosms (16 replicates per treatment) were prepared by adding 1 g of dry wt soil and various amounts of manure (see above) into 9 mL polyethylene centrifuge tubes (Ole Dich Instrument Makers, Hvidovre, Denmark). Subsequently, Milli Q water was added to 60% of the water-holding capacity and the centrifuge tubes were placed without lids in a 324 mL glass infusion flask closed with a gastight rubber septum. Finally, the headspace was outgassed with CO2-free air and microcosms were incubated in the dark at 15 °C under aerobic conditions. Carbon Mineralization. Microbial activity was determined nondestructively as accumulated CO2 from C mineralization during the soil microcosm incubations. CO2 was determined by gas chromatography (GC) using the same equipment and procedure as described previously (21). After each GC measurement, the headspace was outgassed with CO2-free air to ensure that headspace CO2 never exceeded concentrations of 2% (v/v). Amounts of mineralized manureorganic matter during microcosm incubations were estimated on the basis of the difference between accumulated CO2 in microcosms with or without added manure. Harvest of Soil Microcosms. Soil microcosms were destructively sampled by adding Milli Q water to each microcosm to give a soil-to-water ratio of 1-to-5 (w/w). The resulting soil slurries were subjected to shaking on a horizontal shaker for 2 h. Soil–water extracts (supernatants) were finally obtained by centrifugation (10000g, 10 min, 22 °C) and immediately used for determination of [Cu2+] and pH. Samples for determination of [Cu]bio and [Cu]sol were conserved by freezing until analysis. Biosensor Assay for Determination of Bioavailable Cu. [Cu]bio in soil–water extracts were estimated using a dual strain whole-cell bacterial biosensor assay and used to calculate bioavailable Cu per gram of dry weight soil ([Cu]biosoil) as described previously (19, 22). Cu bioavailability was operationally defined as Cu species that were able to induce expression of Cu-regulated luxAB genes in the employed Pseudomonas fluorescens (P. fluorescens) biosensor within a 90 min incubation period. In brief, the P. fluorescens DF57-Cu15 biosensor induced specifically by Cu (17) was used to quantify Cu bioavailability in soil–water extracts by comparing bioluminescence determined in soil-derived samples with an external standard curve obtained by
nonlinear regression of corresponding bioluminescence values determined in CuSO4 standard solutions. Hence, [Cu]bio was expressed with reference to the freely dissolved Cu2+ concentration in the external standard solutions. Another biosensor, P. fluorescens DF57-40E7 was used in parallel to correct for possible sample matrix interference with sensor bacteria due to the presence of substrates (stimulation), toxicants (inhibition), and sample quenching of luminescence (19). Biosensor cell suspensions were prepared in a minimal medium (MM) with a low capacity for Cu-complexation (22). MM consisted of 100 mM KCl, 20 mM Hepes (pH 7.2), 7.6 mM (NH4)2SO4, 4 mM glycerol-2-phosphate disodium salt, and 0.8% (w/v) glucose. Complete Cu speciation at a total Cu concentration of 0.04 µM was predicted using the Visual MINTEQ equilibrium model (23). Free Cu2+ ions constituted 27% of the total Cu species after biosensor addition, while the remaining Cu existed as part of low-affinity organic complexes (copper-glycerol-2-phosphate, 40%) or lowaffinity inorganic complexes (Cu-NH3, 17%; Cu-OH, 7%; Cu-CO3, 4%; Cu-SO4, 3%; Cu-Cl, 1%). Biosensor cell suspensions were mixed with Cu standard solutions or environmental samples, incubated for 90 min, after which bacterial bioluminescence was determined (17). All bioluminescence values are given as the mean of two analytical replicates. The quantification limit, defined as the amount of Cu necessary to cause a 2-fold induction, and the maximum induction of the biosensor were achieved with around 15 and 300 nM Cu in the standard CuSO4 solutions, respectively. Representative standard curves (Figure S1) and detailed procedures for calculation of [Cu]bio and [Cu]biosoil can be found as part of the Supporting Information. Chemical Analysis. [Cu]sol was determined by GFAAS (Perkin-Elmer 5100, Zeeman 5100, PE Applied Biosystems, Foster City, CA) as described previously (18) and subsequently used to calculate the total pool of water extractable Cu per dry weight of soil ([Cu]ext). A Cu ion specific electrode (Cu-ISE, Cu-electrode ISE25Cu, Radiometer, Copenhagen, Denmark), coupled with a doublejunction (KCl-saturated inner reservoir) reference electrode (Reference electrode 251, Radiometer) was used to determine the activity of Cu2+, [Cu2+], in soil–water extracts (24). A calibration curve was established allowing for direct comparison with biosensor-derived bioavailability estimates in parallel soil–water extracts (see above). Calibration buffers were made with 1 mM iminodiacetic acid (IDA), 0.1 mM Cu(NO3)2, 6.0 mM NaOH, and 2.5 mM potassium acid phthalate (KHC8H4O4). The Cu electrode was considered to be equilibrated when the potential stayed within the same (0.01 mV for 1 min. The electrode potential determined by Cu-ISE was then plotted against Cu2+ activity calculated by the computer equilibrium model Visual MINTEQ (23). A calibrated pH electrode (Metrohm 691, Metrohm AG, Herisau, Switzerland) was used to determine pH in soil–water extracts. Statistics. SigmaStat Version 3.5 (Systat Software, Point Richmond, CA) was used for statistical significance testing. If not stated otherwise, the effects of Cu or manure application rates and time on soil Cu descriptors and C mineralization rates were tested by two-way ANOVA for each sampling date and all pairwise multiple comparisons were performed by the Holm-Sidak method. When required, the raw data were log transformed as suggested by SigmaStat output.
Results Impacts of Amendment with Cu-Containing Manure on the Bioavailability and Speciation of Cu in Low-Cu Soil. Addition of low-Cu manure spiked with CuSO4 caused an up to 6-fold increase of the total soil Cu content from 9.5 to 60.7 mg of Cu kg-1 ([Cu]total, Table 1). The bioavailable Cu per dry VOL. 42, NO. 8, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3103
3104
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 8, 2008
293.3 ( 35.8 137.9 ( 26.7 495.6 ( 69.2 129.0 ( 6.7
6.48 ( 0.08 6.40 ( 0.03 7.03 ( 0.02 6.90 ( 0.04
492.2 ( 60.7 166.1 ( 6.9
7.00 ( 0.04 6.69 ( 0.02
250.3 ( 55.0 118.7 ( 29.2
230.4 ( 38.5 90.1 ( 7.4
7.12 ( 0.03 6.92 ( 0.04
6.11 ( 0.09 6.02 ( 0.09
124.1 ( 11.0 49.4 ( 2.2
7.13 ( 0.02 6.83 ( 0.04
152.3 ( 52.0 124.3 ( 28.6
41.5 ( 4.8 19.8 ( 0.3
7.14 ( 0.02 6.97 ( 0.04
5.87 ( 0.04 5.73 ( 0.06
10.5 ( 2.6 9.4 ( 0.2
[Cu]sol
6.04 ( 0.04 5.61 ( 0.02
pHb [Cu]bio
130.4 ( 26.6 13.6 ( 0.7
42.0 ( 14.8 5.4 ( 0.7
13.6 ( 1.3 1.2 ( 0.1
