TY - JOUR
AU - Nybroe,, Ole
AB - Abstract The impact of copper amendment on the bacterial community in agricultural soil was investigated by a 2-year field experiment complemented by short-term microcosm studies. In the field, the amendments led to total copper contents that were close to the safety limits laid down by European authorities. In parallel, bioavailable copper was determined with a copper-specific bioluminescent Pseudomonas reporter strain. The amounts of total Cu as well as of bioavailable Cu in the field declined throughout the experiment. Bacterial community structure was examined by terminal restriction fragment length polymorphism (T-RFLP) analysis of community DNA amplified with primers specific for 16S rDNA from the Bacteria domain, the Rhizobium-Agrobacterium group and the Cytophaga group. Similarity analysis of T-RFLP profiles from field samples demonstrated an impact of copper at the domain level and within the Rhizobium-Agrobacterium group. Comparable Cu effects were observed for microcosms, but in addition an impact on community structure within the Cytophaga group was observed. Field experiment, Pseudomonas reporter bacterium, Bacterial community structure 1 Introduction Copper (Cu) is an essential element for all living organisms yet high concentrations of Cu have detrimental effects on cell metabolism. Cu and Cu-containing compounds are widely used as bactericides and fungicides in agriculture and as growth promoters in pig production. Hence, accumulation of toxic levels of Cu in agricultural soil, e.g. through the application of pig manure, is of concern as the soil microbial community, and ultimately soil fertility, may be affected. Many studies on the effects of Cu and other heavy metals on soil microorganisms have focused on the functional properties of the entire community. For example, adverse effects on respiration and N2 fixation potential, as well as an increased tolerance to heavy metals, have been reported [1,2]. In parallel, a change in community composition could be an important and sensitive indicator of metal-induced effects. Culturing techniques have shown that functionally important bacterial groups, e.g. rhizobia, decrease in abundance in metal-contaminated soil. Consequently they have been used as indicators for metal toxicity [3–5]. The emergence of molecular techniques using polymerase chain reaction (PCR) amplification of phylogenetic markers such as small subunit (SSU) ribosomal RNA genes (rDNA) has provided new ways of studying community structures without the inherent biases of culture-dependent techniques. Hence, Smit and co-workers [6] used amplified rDNA restriction analysis (ARDRA) to demonstrate shifts in microbial community structure caused by copper contamination. However, due to the large number of fragments generated by ARDRA this technique has a limited resolution. The terminal restriction fragment length polymorphism (T-RFLP) method is a derivative of ARDRA and makes use of a fluorescence-labeled primer for PCR [7]. Subsequent analysis of restriction fragments by a DNA sequencing machine registers fluorescence-labeled terminal restriction fragments and provides a distinct fingerprint of a microbial community. Many studies have sought to establish the concentrations of Cu and other heavy metals in soil, at which the microbial community becomes affected. Bååth [8] found that metal concentrations that affected microbial communities in soil differed more than 100-fold between studies. Probably, these discrepancies between the studies are mainly caused by differences in bioavailability of metals as a consequence of different physical and chemical properties of investigated soils [9]. The chemistry of Cu in the soil environment is complicated and the forms of Cu that affect microorganisms are not known. Consequently it is difficult to foresee the toxic effects of Cu by traditional analytical chemistry. Bacterial Cu-specific reporter constructs have been used to detect and quantify bioavailable Cu species in environmental samples [10,11]. The reporters contain the luxAB reporter genes fused to indigenous Cu-induced promoters and they only respond to the bioavailable Cu fraction having impact on bacterial physiology. Relatively few field experiments have addressed Cu effects on microbial properties in agricultural soil, and most of these have been set up to monitor long-term effects of heavy metals applied to the soil with sewage sludge [9,12]. Often, the soils were contaminated with several metals making it difficult to interpret the effects of individual metals [4], or appropriate control soils were not included (as reviewed by McGrath et al. [12]). The objective of the present study was to examine the impact of Cu on soil bacterial community structure in the field during a 2-year period. The field work was complemented with microcosm studies performed under stable laboratory conditions. Community compositions were determined by T-RFLP analysis of SSU rRNA genes using PCR primers specific for the Bacteria domain, and for the Rhizobium-Agrobacterium and Cytophaga groups. Furthermore we examined the bioavailability of Cu in soil samples by means of a Pseudomonas fluorescens Cu reporter strain. 2 Materials and methods 2.1 Experimental setup A field trial was set up at an experimental field at the Royal Veterinary and Agricultural University, Taastrup, Denmark. The soil is a sandy loam containing 7.0 mg Cu in a kg of soil and 1.4% organic C. Soil characteristics are described in more detail by Jensen and Nybroe [13]. The field had been cultivated with spring barley and received inorganic NPK fertilizers during a period of at least 5 years preceding the field trial. The field was cropped with barley and followed by oats during the experimental period. In November 1998, CuSO4 was added to a series of 4×5-m plots arranged according to a randomized block design in an experimental area of 17×24 m. Three plots received a total of 21 g Cu m−2 (designated low-Cu plots), another three plots received 63 g Cu m−2 (designated high-Cu plots), while three control plots were not amended with Cu. The amendments were made to reach concentrations of about 50 mg Cu (kg soil)−1 (low-Cu plots) and 150 mg Cu (kg soil)−1 (high-Cu plots) assuming that all amended Cu remained in the plow layer (30 cm). In order to distribute the Cu more evenly in the soil, CuSO4 was sprayed onto the soil surface by a handheld syringe in two turns separated by plowing of the field. Soil samples were collected from plots 2, 9, 18 and 21 months after Cu amendments using a drill (2 cm wide and 25 cm long). From each plot nine soil cores were sampled evenly within the inner 2×3-m area. Soil cores from each plot were subsequently pooled, sieved (mesh size 2 mm) and kept at −20°C. Laboratory microcosms, 10 g each, were established using soil taken from a pool of control plot samples. Three microcosms were amended with 100 mg Cu (as CuSO4) per kg soil (equal to 1570 μmol (kg soil)−1), while three microcosms were unamended controls. Soil microcosms were kept in 50-ml polyethylene tubes at 25°C, and at ca. 80% water-holding capacity. The microcosms were harvested after 4 weeks and stored at −20°C. Parallel microcosm experiments were carried out with 1570 μmol SO4 (kg soil)−1 (as Na2SO4), to examine the effects of the SO42− anion. 2.2 Soil analysis of total Cu, bioavailable Cu and pH The pH of soil samples was measured with an Orion electrode (model 91-02, Boston, MA, USA) using a Radiometer pH meter (PHM28, Copenhagen, Denmark). Before measurement soil samples were diluted 1:5 with reverse osmosis-purified water (Milli-Q, Millipore, Bedford, MA, USA), shaken for 2 h at 200 rpm and subsequently equilibrated for 1 day at room temperature. The total Cu content in field plots was determined by Steins Laboratory (Brørup, Denmark) on duplicate samples of 1 g soil extracted with 20 ml 3.5 M HNO3 and heated to 120°C for 30 min. The Cu concentration in the filtered extract was measured by inductively coupled plasma spectrometry (Optima 4300 DV, Applied Biosystems, Foster City, CA, USA). Bioavailable Cu in the collected soil samples that had been stored at −20°C was measured using the Cu-specific reporter strain P. fluorescens DF57-Cu15, which carries a promoterless Tn5::luxAB cassette inserted on the chromosome under control of an indigenous Cu-induced promoter [11]. Water extracts were obtained from 1 g of soil (dry weight) shaken (in polyethylene tubes) for 2 h at 200 rpm at room temperature with 5 ml of Milli-Q water. Supernatants were recovered after centrifugation (10 000×g, 10 min, room temperature). Strain DF57-Cu15 was exposed to water extracts of the soil samples essentially as in Tom-Petersen et al. [11]. In brief, samples of 0.1 ml were mixed with 0.9 ml DF57-Cu15 cell suspension and incubated for 1.5 h at room temperature. Induction of bioluminescence from the luxAB reporter construct was measured by a luminometer (Bio Orbit 1253, Turku, Finland) as previously described [11]. A standard curve of bioluminescence obtained from DF57-Cu15 exposed to increasing concentrations of CuSO4 in Milli-Q water was used to convert relative luminescence unit values obtained from soil samples to bioavailable Cu expressed as mg Cu (kg soil)−1. Bioluminescence was measured in duplicate samples for each water extract. 2.3 DNA extractions and PCR amplifications Community DNA was extracted from 0.5-g soil samples (wet weight) with the Ultra Clean™ Soil DNA Isolation Kit (Mo Bio Laboratories, Solana Beach, CA, USA), using the heating procedure described by the manufacturer. The DNA yield was determined spectrometrically (Gene Quant II, Pharmacia Biotech, Cambridge, UK) and the DNA was further analyzed by agarose gel electrophoresis. The extracted DNA was used as target DNA in the PCR to amplify the community 16S rDNA. All PCR amplifications were carried out using the forward primer 8-27f [14], which targets the 5′ end of the 16S rDNA of prokaryotes from the Bacteria domain. This primer was used in combination with three different reverse primers (see Table 1). The forward primer was fluorescence-labeled with carboxyfluorescein-N-hydroxysuccinimide ester-dimethyl sulfoxide. All primers were delivered by DNA Technology, Århus, Denmark. 1 Primers used in this study Primer name Sequence Complement 16S rDNA at Escherichia coli position number Reference 8-27f 5′-AGAGTTTGATCCTGGCTCAG-3′ 8 [14] BACT 5′-CCGTCAATTCCTTTRAGTTT-3′ 926 [15] CYT 5′-GGATCATGGCTGATATCCGAT-3′ 1370 D. Buckley and T.L. Marsh RHIZ 5′-CTCGCTGCCCACTGTCAC-3′ 1244 This study Primer name Sequence Complement 16S rDNA at Escherichia coli position number Reference 8-27f 5′-AGAGTTTGATCCTGGCTCAG-3′ 8 [14] BACT 5′-CCGTCAATTCCTTTRAGTTT-3′ 926 [15] CYT 5′-GGATCATGGCTGATATCCGAT-3′ 1370 D. Buckley and T.L. Marsh RHIZ 5′-CTCGCTGCCCACTGTCAC-3′ 1244 This study The RHIZ primer was found using ARB software (rRNA database release No. 7.0), with the following probe design parameters: length of probe=18; Tm=45–85; GC content=45–75; E. coli position=0–100 000; max. non-group hits=15%; min. group hits=65%. Open in new tab 1 Primers used in this study Primer name Sequence Complement 16S rDNA at Escherichia coli position number Reference 8-27f 5′-AGAGTTTGATCCTGGCTCAG-3′ 8 [14] BACT 5′-CCGTCAATTCCTTTRAGTTT-3′ 926 [15] CYT 5′-GGATCATGGCTGATATCCGAT-3′ 1370 D. Buckley and T.L. Marsh RHIZ 5′-CTCGCTGCCCACTGTCAC-3′ 1244 This study Primer name Sequence Complement 16S rDNA at Escherichia coli position number Reference 8-27f 5′-AGAGTTTGATCCTGGCTCAG-3′ 8 [14] BACT 5′-CCGTCAATTCCTTTRAGTTT-3′ 926 [15] CYT 5′-GGATCATGGCTGATATCCGAT-3′ 1370 D. Buckley and T.L. Marsh RHIZ 5′-CTCGCTGCCCACTGTCAC-3′ 1244 This study The RHIZ primer was found using ARB software (rRNA database release No. 7.0), with the following probe design parameters: length of probe=18; Tm=45–85; GC content=45–75; E. coli position=0–100 000; max. non-group hits=15%; min. group hits=65%. Open in new tab For PCR amplification of community 16S rDNA of the Bacteria domain we used the forward primer in combination with the Bacteria-specific reverse primer 926r [15], hereafter referred to as BACT. The CYT reverse primer complements 16S rDNA of bacteria belonging to Cytophaga subgroups I and II within the Cytophaga-Flexibacter-Bacteroides division. The primer does not complement any other bacterial groups represented in the Ribosomal Database Project (RDP) [16]. The ability of this primer to target isolates within Cytophaga subgroups I and II has been experimentally verified from 100 sequenced clones derived from amplification with CYT that all belonged to Cytophaga. The RHIZ reverse primer targets 16S rDNA within subgroups of the Rhizobium-Agrobacterium group belonging to the α subdivision of Proteobacteria. The ability of the RHIZ primer to complement 16S rDNA from Rhizobium strains but not Bradyrhizobium strains (all belonging to the Rhizobium-Agrobacterium group) was experimentally verified on culture collection strains (data not shown). The Rhizobium primer also targets known sequences in the Azospirillum and Sphingomonas groups. However, sequences in these two groups do not have CfoI restriction sites and are therefore not detected in our T-RFLP analyses (see Section 2.4 for further details). PCR amplifications of each DNA extraction were performed in three separate reaction tubes each with a total volume of 50 μl, and the PCR products from these three reactions were subsequently pooled. Reaction mixtures for PCR contained full strength reaction buffer (Applied Biosystems), each dNTP at a concentration of 250 μM, 3 mM MgCl2, 400 ng μl−1 bovine serum albumin, 0.025 U μl−1 ampliTaq DNA polymerase (Applied Biosystems), each primer at 0.25 μM and 0.04 μl μl−1 template DNA. We used the following PCR program for assays including the BACT primer: initial denaturation at 94°C for 5 min; 25 cycles consisting of denaturation at 94°C for 30 s, primer annealing at 55°C for 30 s and elongation at 72°C for 90 s. The final elongation was extended to 10 min. PCR conditions used for reactions including the RHIZ or the CYT primer were similar, except that we used 30 thermal cycles and annealing temperatures of 60 and 62°C, respectively. 2.4 Restriction digests and T-RFLP gel electrophoresis The fluorescently labeled PCR products were purified on QIAquick PCR purification kit columns (Qiagen, Hilden Germany). 200 ng of purified products from PCR using primers targeting eubacterial 16S rDNA or 100 ng of purified products from PCR using Rhizobium-Agrobacterium and Cytophaga group-specific primers were digested for 4 h at 37°C with 20 U of CfoI (Boehringer, Mannheim, Germany) in 20-μl reaction mixtures. The fluorescently labeled terminal restriction fragments (T-RFs) were analyzed by electrophoresis on an automatic sequence analyzer (ABI Prism 373 DNA Sequencer, Applied Biosystems) in Genescan mode. Aliquots (2 μl) of T-RFs were mixed with 2 μl of deionized formamide, 0.4 μl of loading buffer (Applied Biosystems), and 0.6 μl of DNA size marker. The standard size marker was a 1:1 mixture of the size standards GS-500 and GS-1000 ROX (Applied Biosystems). The T-RF mixture was denatured at 94°C for 2 min and chilled on ice prior to electrophoresis. Samples were loaded on a 36-cm, 6% denaturing polyacrylamide gel. Electrophoreses were run for 10 or 13 h with the settings 2500 V and 40 mA, using the B filter set. 2.5 Analysis of T-RFLP profiles After electrophoresis, the sizes (bp) of fluorescently labelled T-RFs were determined by comparison with the internal ROX standard size markers using Genescan™ software (Applied Biosystems), and the Local Southern size calling method. Experimentally determined sizes of T-RFs were compared with 5′-terminal fragments of computer-simulated restriction enzyme digestions of 16S rDNA sequences in RDP using the web-based TAP software program [17]. Primer specificity was set to no mismatches in the last three nucleotides from the 3′ end, while a single mismatch was allowed at any other position. Normalized peak heights of T-RFs were used for comparisons of T-RFLP profiles. Normalization of peak heights was done as described by Dunbar et al. [18], and the DNA quantity of each profile was estimated by taking the sum of all T-RF peak heights between 30 and 674 bp. In the normalization procedure we excluded T-RFs with peak heights that were reduced to less than 25 relative fluorescence units, corresponding to the initial threshold value chosen for T-RF heights. 2.6 Analyses of community similarity and of individual T-RFs In some cases similarity analyses were carried out on ‘consensus profiles’ to eliminate the impact of seasonal variation, plot-to-plot variation and field-to-microcosm differences from the analyses. Consensus profiles included only peaks constituting a certain percentage of the sum of peak heights in at least one profile essentially as described by Lukow et al. [19]. Using this procedure and a threshold of 2%, a total number of 28 T-RFs were considered for profiles generated by the BACT reverse primer. For analyses of profiles obtained by the RHIZ or CYT primer T-RFs a 5% threshold was applied, giving a total number of nine and seven T-RFs, respectively. Subsequently, profiles were generated from T-RF data by expressing the height of a specific T-RF as a percentage of the average peak height of that particular T-RF in all profiles derived from a given sampling. These profiles were used to compare the height of individual T-RFs from Cu-treated soils with corresponding T-RFs from control soils within each sampling. Finally one consensus profile was made from each treatment at each sampling, generated as an average profile from triplicate plot profiles or triplicate microcosm profiles. Similarity analyses of T-RFLP table data, consensus or not, were carried out using Bionumerics software (Applied Maths, Kortrijk, Belgium). The Jaccard similarity coefficient (JC, expressed in percent values) for each pair of T-RFLP profiles was calculated by dividing the number of T-RFs that occur in both profiles by the total number of T-RFs (common and unique) in both T-RFLP profiles. The JCs were calculated using the area-sensitive option of the software. The area-sensitive option takes into account differences in area or height (this study) between matching T-RFs, and can thus differentiate T-RFLP profiles even if the same T-RFs are present in all of the compared samples. Dendrograms were constructed from the similarity matrix using the unweighted pair group method with arithmetic mean (UPGMA). 2.7 Statistics Significant differences of means were compared pairwise between control samples and Cu-treated samples by the t-test (two-tailed) after examination of normality and variance using probability plots and an F-test (two-tailed) respectively. Analyses were performed using SYSTAT 10 (SPSS, Chicago, IL, USA). 3 Results 3.1 Total Cu, bioavailable Cu and pH The average total and bioavailable Cu contents in the surface soil from field plots are presented in Fig. 1. At the first sampling 2 months after Cu amendments the total contents of low-Cu and high-Cu plots were 73 mg (kg soil)−1 and 213 mg (kg soil)−1, respectively (Fig. 1A). The total Cu concentrations decreased by about 36% in the low-Cu plots and by about 45% in the high-Cu plots during the experimental period. The measured Cu concentrations in the low-Cu and high-Cu plots were comparable to the current EU limits of 50 mg Cu (kg soil)−1 (advisable maximal load) and 140 mg Cu (kg soil)−1 (upper limit) [20], which are valid for soils receiving sewage sludge. The average total Cu content in untreated plots was 6.8 mg (kg soil)−1 at the first sampling and 8.7 mg (kg soil)−1 at the last sampling, and thus comparable to the 7.0 mg Cu (kg soil)−1 measured before the onset of the field experiment. 1 Open in new tabDownload slide Cu content in the topsoil of field plots during the 21-month experimental period. A: Total Cu in soil. B: Bioavailable Cu in water extracts. C: Bioavailability of Cu in water extracts (bioavailable Cu in percent of total Cu). Control plots (white bars), low-Cu plots (gray bars) and high-Cu plots (black bars). Error bars indicate S.D. values from triplicate plots. 1 Open in new tabDownload slide Cu content in the topsoil of field plots during the 21-month experimental period. A: Total Cu in soil. B: Bioavailable Cu in water extracts. C: Bioavailability of Cu in water extracts (bioavailable Cu in percent of total Cu). Control plots (white bars), low-Cu plots (gray bars) and high-Cu plots (black bars). Error bars indicate S.D. values from triplicate plots. Fig. 1B shows changes in bioavailable Cu measured with the bacterial Cu reporter DF57-Cu15 in water extracts of the same soil samples. The amount of bioavailable Cu in the plots generally reflected the level of Cu amendment but during the experiment the bioavailable Cu content was reduced. Cu bioavailability (bioavailable Cu as a fraction of total Cu) was approximately 0.4% after 2 months and declined with time as seen from Fig. 1C. Furthermore, the bioavailable Cu fraction tended to be smaller in the high-Cu plots than the low-Cu plots at all samplings. Soil pH was identical, 6.0–6.1, in control plots and Cu-treated plots. For the Cu-amended microcosms a total Cu content of 105 mg (kg soil)−1 was determined at the end of the experiment after 1 month. The bioavailable Cu fraction constituted 0.3% of the total Cu content. The pH declined slightly, but significantly (P<0.05) from 6.0 in the control microcosms to 5.8 in the Cu-amended microcosms. 3.2 Analysis of T-RFLP profiles The yield of the DNA extractions was estimated to be between 1.7 and 2.8 μg of DNA (g soil)−1. The recovered DNA migrated in agarose gels as fragments of approximately 20 kb, with no significant smearing (data not shown). The T-RFLP analysis included T-RFs in the range of 30–674 bp, as the largest fragment of the marker detected under the applied electrophoresis conditions was 674 bp. Examples of T-RFLP profiles obtained from a field sample using the BACT, CYT and RHIZ reverse primers are presented in Fig. 2. 2 Open in new tabDownload slide T-RFLP profiles obtained from DNA extracted from a control field plot. A: Profile generated by the BACT reverse primer amplifying 16S rDNA from the Bacteria domain. B: Profile generated by the CYT reverse primer amplifying 16S rDNA from the Cytophaga group. C: Profile generated by the RHIZ reverse primer amplifying 16S rDNA from the Rhizobium-Agrobacterium group. Sizes in bp of selected fragments are indicated. 2 Open in new tabDownload slide T-RFLP profiles obtained from DNA extracted from a control field plot. A: Profile generated by the BACT reverse primer amplifying 16S rDNA from the Bacteria domain. B: Profile generated by the CYT reverse primer amplifying 16S rDNA from the Cytophaga group. C: Profile generated by the RHIZ reverse primer amplifying 16S rDNA from the Rhizobium-Agrobacterium group. Sizes in bp of selected fragments are indicated. The total number of detected T-RFs with the BACT primer increased linearly to the sum of all fragment heights in each profile (r2=0.51; P<0.05), suggesting that the observed number of T-RFs in individual profiles depended on the amount of DNA loaded onto the gel. Therefore, the height of each T-RF was normalized (see Section 2.5 for details). Upon normalization the number of detected T-RFs ranged from 63 to 88 for the BACT primer, from 15 to 24 for the CYT primer, and from 19 to 30 for the RHIZ primer. All subsequent analyses were done on the normalized data. No significant difference was detected (P>0.05) in the total numbers of T-RFs between the Cu-amended soils and the control soils at any sampling with either of the three primer sets used. 3.3 Similarity analysis of the bacterial communities Initially, Cu effects on the bacterial community were monitored in laboratory microcosms. Fig. 3 shows the dendrograms obtained from T-RFLP profile analyses of the microcosms using the three primer sets. Profiles from untreated microcosms and from Cu-amended microcosms clustered into two groups with all primer sets. The mean JCs between profiles derived from the two types of microcosms were 74, 65 and 85, for the BACT, CYT, and RHIZ primers, respectively. 3 Open in new tabDownload slide Dendrograms based on T-RFLP profiles from triplicate control and triplicate Cu-amended microcosms using: (A) the BACT primer, (B) the CYT primer, and (C) the RHIZ primer. Dendrograms were constructed from the similarity matrix using the UPGMA method. 3 Open in new tabDownload slide Dendrograms based on T-RFLP profiles from triplicate control and triplicate Cu-amended microcosms using: (A) the BACT primer, (B) the CYT primer, and (C) the RHIZ primer. Dendrograms were constructed from the similarity matrix using the UPGMA method. T-RFLP profiles from microcosms amended with Na2SO4 were compared to the untreated and the Cu-amended microcosms to examine whether SO4, constituting the anion of the Cu salt used in this study, could be responsible for the observed effects. However, the Na2SO4-treated microcosms were quite similar to the control microcosms (mean JC=95), compared with the similarity to the Cu-amended microcosms (mean JC=77), indicating that SO4 did not account for the effects of CuSO4 presented here. T-RFLP profiles generated by the BACT primer from the winter/spring samples of field plots clustered distinctly from samples taken during summer. This indicates a seasonal variation of the bacterial community structure. In addition, variations of T-RFLP profiles between similarly treated field plots were seen (data not shown). This variability obscured any effects of the Cu treatment in the field experiment. To circumvent this problem, we constructed consensus T-RFLP profiles (see Section 2 for details), and performed all subsequent similarity analyses on these profiles. Fig. 4 shows the result of comparing all consensus T-RFLP profiles generated from field samples and soil microcosms with the BACT reverse primer. The profiles from field samples clustered according to the Cu treatment. Interestingly, the consensus profile from control microcosms clustered with those from control plots and Cu microcosms with high-Cu plot profiles. Table 2 shows the mean JCs obtained by comparison of control, low-Cu, high-Cu, control microcosm or Cu microcosm consensus profiles. Profiles from untreated plots were significantly more similar to profiles from low-Cu plots than from high-Cu plots (P<0.01), indicating a Cu dose-dependent effect on community similarity. Also untreated microcosms were significantly more similar (P<0.05) to the untreated field plot profiles than to low-Cu and high-Cu plot profiles. Correspondingly, consensus profiles from Cu-treated microcosms were significantly more similar to the high-Cu plot profiles (P<0.05) than to the remaining field plot profiles, and more similar (P<0.05) to low-Cu plot profiles than to untreated plot profiles (see Table 2). 4 Open in new tabDownload slide Dendrogram of consensus T-RFLP profiles generated by the BACT primer from field and microcosm samples. Dendrograms were constructed from the similarity matrix using the UPGMA method. 4 Open in new tabDownload slide Dendrogram of consensus T-RFLP profiles generated by the BACT primer from field and microcosm samples. Dendrograms were constructed from the similarity matrix using the UPGMA method. 2 Mean JCs of the four control, four low-Cu, four high-Cu, one control microcosm, and one Cu microcosm consensus profiles, obtained by the comparison of each consensus profile with all other consensus profiles from each treatment Control plots Low-Cu plots High-Cu plots Control plots 87±2.3 Low-Cu plots 81±2.4 86±1.5 High-Cu plots 75±3.2 80±3.2 90±2.1 Control microcosms 86±2.3 82±2.1 74±2.3 Cu microcosms 76±1.9 82±3.8 89±1.7 Control plots Low-Cu plots High-Cu plots Control plots 87±2.3 Low-Cu plots 81±2.4 86±1.5 High-Cu plots 75±3.2 80±3.2 90±2.1 Control microcosms 86±2.3 82±2.1 74±2.3 Cu microcosms 76±1.9 82±3.8 89±1.7 Mean values±S.D. are shown. Open in new tab 2 Mean JCs of the four control, four low-Cu, four high-Cu, one control microcosm, and one Cu microcosm consensus profiles, obtained by the comparison of each consensus profile with all other consensus profiles from each treatment Control plots Low-Cu plots High-Cu plots Control plots 87±2.3 Low-Cu plots 81±2.4 86±1.5 High-Cu plots 75±3.2 80±3.2 90±2.1 Control microcosms 86±2.3 82±2.1 74±2.3 Cu microcosms 76±1.9 82±3.8 89±1.7 Control plots Low-Cu plots High-Cu plots Control plots 87±2.3 Low-Cu plots 81±2.4 86±1.5 High-Cu plots 75±3.2 80±3.2 90±2.1 Control microcosms 86±2.3 82±2.1 74±2.3 Cu microcosms 76±1.9 82±3.8 89±1.7 Mean values±S.D. are shown. Open in new tab T-RFLP consensus profiles using the CYT and RHIZ primers were analyzed for field soil samples collected after 21 months. The similarity analysis of profiles using the RHIZ primer (nine T-RFs) showed that the field plot profiles clustered according to Cu treatment (Fig. 5). Furthermore, profiles from untreated microcosms clustered with profiles from untreated field plots, whereas profiles from Cu-treated microcosms and Cu-treated field plots clustered together. The field plot profiles generated with the CYT primers did not cluster according to the Cu treatment (data not shown). 5 Open in new tabDownload slide Dendrogram of consensus T-RFLP profiles generated by the RHIZ primer from field and microcosm samples. Dendrograms were constructed from the similarity matrix using the UPGMA method. 5 Open in new tabDownload slide Dendrogram of consensus T-RFLP profiles generated by the RHIZ primer from field and microcosm samples. Dendrograms were constructed from the similarity matrix using the UPGMA method. 3.4 Effects on abundance of individual T-RFs The abundance (peak height) of each of the 28 individual T-RFs from the BACT primer profiles was compared pairwise between Cu-treated soils and control soils, and the T-RFs from Cu-treated soils with different mean abundance are shown in Fig. 6. In Cu-amended microcosms 11 of 28 T-RFs had a significantly (P<0.05) altered abundance when compared to untreated microcosms. For comparison, only five of 28 T-RFs from the high-Cu field plots were significantly affected (P<0.05) after 2 and 9 months while a lower number was affected after 18 and 21 months. In general, fewer T-RFs were affected in the low-Cu-treated field plots. Fig. 6 shows that changes in abundance of a specific T-RF were always consistent (increasing or decreasing abundance, respectively) when comparing low-Cu and high-Cu plots, or when comparing field plots and laboratory microcosms. 6 Open in new tabDownload slide T-RFs obtained from Bacteria domain-specific T-RFLP profiles, which had a significantly (P<0.05) different mean abundance when Cu-treated soils were compared to corresponding control soils. The direction of the arrows shows whether the relative abundance of T-RFs in Cu-treated soil increased (↑) or decreased (↓) compared to control soils. 6 Open in new tabDownload slide T-RFs obtained from Bacteria domain-specific T-RFLP profiles, which had a significantly (P<0.05) different mean abundance when Cu-treated soils were compared to corresponding control soils. The direction of the arrows shows whether the relative abundance of T-RFs in Cu-treated soil increased (↑) or decreased (↓) compared to control soils. The number of T-RFs in the RDP database matching the size of each of the 28 T-RFs varied between 1 and 36. The average was 10 RDP sequences per determined T-RF, but eight of the T-RFs matched only one sequence in the database. Thirteen of the T-RFs originated from Gram-negative bacteria, while five represented Gram-positive bacteria. The remaining 10 T-RFs could potentially have originated from both Gram-negative and Gram-positive bacteria. Our analysis did not provide evidence that ‘Gram-positive’ and ‘Gram-negative’ T-RFs responded differently to the Cu amendments. For the corresponding analyses performed on T-RFs generated by the RHIZ or CYT primers, a 1-bp uncertainty in the size determination was allowed for fragments below 500 bp and a 3-bp uncertainty for larger fragments. Profiles generated with the RHIZ primers included nine T-RFs. A 407-bp fragment was more abundant in the Cu-treated field plots and microcosms than in the untreated field plots and microcosms, whereas a 511-bp fragment, matching the Bartonella subgroup, showed an increased abundance only in the Cu-treated microcosms. Comparison of the remaining experimentally determined T-RFs to the RDP database revealed that an about 61-bp fragment could originate from nine subgroups of the Rhizobium-Agrobacterium group, a 339-bp fragment from the Agrobacterium subgroup and a ca. 82-bp fragment from the Bartonella subgroup. Of the seven T-RFs in the profiles generated with the CYT primers, a 92-bp fragment was significantly less abundant in the Cu-treated microcosms than the untreated microcosms. This fragment matched 16S rDNA sequences affiliated with the Capnocytophaga ochracea subgroup, the Flavobacterium flevense subgroup, the Psychroserpens burtonensis subgroup, and the Empedobacter brevis subgroup, all within Cytophaga group I. A fragment of 98 bp matching the P. burtonensis subgroup was significantly more abundant in the high-Cu field plots than in the untreated plots. In addition several fragments matched known phylogenetic groups within Cytophaga group I. Hence an 88-bp T-RF matched the C. ochracea subgroup, while 90-bp fragments could be derived either from the C. ochracea subgroup or the F. flevense subgroup. Finally, a 96-bp T-RF matched the P. burtonensis subgroup, the Flexibacter maritimus subgroup, the Cytophaga uliginosa subgroup, and the Microsilla aggregans subgroup. 4 Discussion 4.1 Total and bioavailable Cu Total Cu concentration in the field plots decreased throughout the experimental period. Plowing of the soil might lead to a net movement of Cu to adjacent or underlying soil while leaching of Cu to lower soil horizons probably is less important as concluded from previous field studies [21,22]. We estimated the amount of bioavailable Cu in water extracts of the soil, normally considered to contain bioavailable Cu fraction for soil organisms residing in the soil soluble phase. Measured Cu bioavailability is expressed as mg bioavailable Cu per kg soil. This unit should be interpreted with the reservation that freezing of soil samples and presence of nutrients in the assay medium could modify Cu speciation [23]. Initially less than 0.5% of the amended Cu was biologically available to the Pseudomonas reporter strain, which is in agreement with our previous observations for soil slurry microcosms [11]. Most likely the low bioavailability is due primarily to retention of Cu by the soil solid phase. Additionally, soluble complexes between Cu and soil organic matter may not be readily available to living organisms [11]. We found a general correspondence between the total Cu content and bioavailable Cu in the field plots. However, the bioavailable Cu fraction tended to be higher in low-Cu field plots compared with high-Cu field plots. This could in part be explained by a relative decrease in water solubility of Cu with increasing levels of Cu contamination, which we have observed in other experiments using soil from the same location (unpublished results). Furthermore, during the experimental period the bioavailable Cu fraction decreased. We speculate that the mechanism leading to progressive reduction in water solubility, and hence in bioavailability, might be diffusion of Cu into inaccessible pores in the soil matrix [24,25]. Our study included measurements of soil pH as some effects on the soil microflora, which are observed when soil is amended with heavy metals, can be attributed to soil acidification rather than to direct metal toxicity [26]. However, we observed no significant difference in pH in the field experiment. The minor decrease in pH found for Cu-treated microcosms did not seem to significantly affect bacterial communities, as Cu-treated microcosms and field plots clustered together in our community analyses. 4.2 Target groups for T-RFLP analysis Our T-RFLP analysis of PCR-amplified community 16S rDNA used a forward primer specific for the Bacteria domain (8-27) for all amplifications, while three reverse primers were used. The BACT reverse primer has been selected in silico in combination with the 8-27f forward primer and the CfoI restriction enzyme to produce the largest number of different T-RFs and, hence, the highest resolution of the analysis [7]. A reverse primer targeting the Rhizobium-Agrobacterium group was designed since this ecologically important group, which includes nitrogen-fixing symbionts, has attracted considerable attention as a sensitive indicator of heavy metal contamination when culturing methods are employed [3–5]. The Cytophaga group belongs to the Cytophaga-Flexibacter-Bacteroides division, members of which have been found to constitute approximately 20% of the bacteria in soil by both culture-dependent and culture-independent surveys [27,28]. The Cytophaga group exhibits a high physiological diversity, and includes bacteria able to degrade complex organic compounds [29]. A reverse primer targeting the Cytophaga group was designed as this group may well be important to transformation of organic matter in soil and appears to be affected by chromium contamination (unpublished results). In this study we addressed Cu effects on soil microbial community at two levels: by comparing community structures using similarity analysis of T-RFLP profiles, and by comparing abundances of individual T-RFs. 4.3 Similarity analysis Seasonal and spatial variations in the field are known from the literature [20] and in our field experiment these variations had a larger effect on community structure than Cu amendment. However, by constructing consensus T-RFLP profiles for each treatment group, and by complementing field experiments by highly controllable microcosm setups we were able to detect an effect of the Cu treatment on the bacterial community structure. T-RFLP analyses using the BACT and RHIZ primers revealed a more significant impact of the high-Cu treatment than of the low-Cu treatment as displayed by the decreasing JCs. It is notable that Cu-induced effects on the bacterial community structure could be detected at concentrations close to the EU safety limits of 50–150 mg Cu (kg soil)−1 corresponding to a bioavailable Cu concentration as low as 0.3 mg Cu (kg soil)−1 as defined by the current assay. For comparison, the field study by Smit et al. [6] referred to above, which is possibly the only field experiments addressing the effects of Cu added as a single metal, revealed quite severe effects of Cu contamination as seen from altered community ARDRA profiles. Closer comparisons are made difficult by differences in Cu doses, soil properties (especially soil pH) and duration of the experiments. A Cu dose-dependent change in community composition has been reported by Frostegaard et al. [30]. These authors examined phospholipid fatty acid (PLFA) patterns of the microbial community in Cu-contaminated agricultural soil incubated in the laboratory. Their studies showed that PLFA profiles gradually changed from the control profiles, with significant changes appearing above approximately 130 mg Cu (kg soil)−1. 4.4 Effect on abundance of individual T-RFs Significant changes in the abundance of some of the T-RFs between treatments were found, whereas appearance of new T-RFs or disappearance of existing T-RFs was only seen in a few cases in the consensus profiles generated with any of the three primer sets. In the microcosms 39% of the T-RFs were significantly affected by the Cu treatment, whereas in the field experiment a maximum of 18% of the T-RFs were affected by the high-Cu treatment. The larger number of affected T-RFs in the microcosms was probably a consequence of the lower data variability in the microcosms producing more significant differences. The number of significantly affected T-RFs decreased throughout the experimental period suggesting either a reduced effect of Cu concomitant to the reduced concentration of bioavailable Cu, or an adaptation of bacterial community to Cu through increased metal tolerance [8,31]. Adaptation might also explain that Cu impact on the Cytophaga group was pronounced in short-term microcosm experiments but absent in the field samples obtained 21 months after Cu amendment. Within the limited resolution of the BACT primer, we did not obtain evidence for a specific impact on the Gram-positive population in Cu-treated soils. An increased abundance of culturable Gram-negative bacteria relative to Gram-positive bacteria has been found in heavy metal-contaminated soil microcosms [32], and in a field study [33]. However, other studies have failed to show such changes [6,34,35]. A better phylogenetic resolution, generally to groups or subgroups as defined in the RDP, was obtained with the RHIZ primers. The analyses showed a corresponding significant impact of Cu on a single T-RFs from the Rhizobium-Agrobacterium group in both Cu microcosms and in high-Cu plots after 21 months, indicating similar short-term and long-term effects on this group. In conclusion, we demonstrate that Cu amendment causes changes in the composition of the taxonomic groups examined here. These changes could well lead to changes in beneficial soil functions such as nitrogen fixation or hydrolytic activity. However, it remains to be solved how the impact on taxonomic groups can be related to effects on functional capabilities of soil microbial communities, which we plan to address in future experiments. Furthermore, the field plot soils will allow us to compare resistance and resilience of bacterial communities to different stressful conditions as affected by a previous exposure to Cu. Acknowledgements This work was supported by grants from the Danish Biotechnology Program and the Danish Ministry of Food, Agriculture and Fisheries, Project MIL-97. References [1] Bååth E. Diaz-Raviña M. Frostegård Å. Campbell C.D. ( 1998 ) Effects of metal-rich sludge amendment on the soil microbial community . Appl. Environ. Microbiol. 64 , 238 – 245 . Google Scholar PubMed OpenURL Placeholder Text WorldCat [2] Lorenz S.E. Brookes P.C. Giller K.E. ( 1992 ) Assessment of free-living nitrogen fixation activity as a biological indicator of heavy metal toxicity in soil . Soil Biol. Biochem. 24 , 601 – 606 . Google Scholar Crossref Search ADS WorldCat [3] Giller K.E. Nussbaum R. Chaudri A.M. McGrath S.P. ( 1993 ) Rhizobium meliloti is less sensitive to heavy metal contamination in soil than R. leguminosarum bv. Trifolii or R. loti . Soil Biol. Biochem. 25 , 273 – 278 . Google Scholar Crossref Search ADS WorldCat [4] Dahlin S. Witter E. Mårtensson A. Turner A. Bååth E. ( 1997 ) Where's the limit? Changes in the microbiological properties of agricultural soils at low levels of metal contamination . Soil Biol. Biochem. 29 , 1405 – 1415 . Google Scholar Crossref Search ADS WorldCat [5] Chaudri A.M. Allain C.M.G. Barbosa-Jefferson V.L. Nicholson F.A. Chambers B.J. McGrath S.P. ( 2000 ) A study of the impacts of Zn and Cu on two rhizobial species in soils of a long-term field experiment . Plant Soil 221 , 167 – 179 . Google Scholar Crossref Search ADS WorldCat [6] Smit E. Leeflang P. Wernars K. ( 1997 ) Detection of shifts in microbial community structure and diversity in soil caused by copper contamination using amplified ribosomal DNA restriction analysis . FEMS Microbiol. Ecol. 23 , 249 – 261 . Google Scholar Crossref Search ADS WorldCat [7] Liu W. Marsh T.L. Cheng H. Forney L.J. ( 1997 ) Characterization of microbial diversity by determining terminal restriction fragment length polymorphisms of genes encoding 16S rRNA . Appl. Environ. Microbiol. 63 , 4516 – 4522 . Google Scholar PubMed OpenURL Placeholder Text WorldCat [8] Bååth E. ( 1989 ) Effects of heavy metals in soil on microbial processes and populations: A review . Water Air Soil Poll. 47 , 335 – 379 . Google Scholar Crossref Search ADS WorldCat [9] Giller K.E. Witter E. McGrath S.P. ( 1998 ) Toxicity of heavy metals to microorganisms and microbial processes in agricultural soils: A review . Soil Biol. Biochem. 30 , 1389 – 1414 . Google Scholar Crossref Search ADS WorldCat [10] Corbisier P. Thiry E. Diels L. ( 1996 ) Bacterial biosensors for toxicity assessment of solid wastes . Environ. Tox. Water Qual. 11 , 171 – 177 . Google Scholar Crossref Search ADS WorldCat [11] Tom-Petersen A. Hosbond C. Nybroe O. ( 2001 ) Identification of copper-induced genes in Pseudomonas fluorescens and use of a reporter strain to monitor bioavailable copper in soil . FEMS Microbiol. Ecol. 38 , 59 – 67 . Google Scholar Crossref Search ADS WorldCat [12] McGrath S.P. Chaudri A.M. Giller K.E. ( 1995 ) Long-term effects of metals in sewage sludge on soils, microorganisms and plants . J. Ind. Microbiol. 14 , 94 – 104 . Google Scholar Crossref Search ADS PubMed WorldCat [13] Jensen L.E. Nybroe O. ( 1999 ) Nitrogen availability to Pseudomonas fluorescens DF57 is limited during decomposition of barley straw in bulk soil and in the barley rhizosphere . Appl. Environ. Microbiol. 65 , 4320 – 4328 . Google Scholar PubMed OpenURL Placeholder Text WorldCat [14] Edwards U. Rogall T. Blocker H. Emde M. Bottger E.C. ( 1989 ) Isolation and direct complete determination of entire genes . Nucleic Acids Res. 17 , 7843 – 7853 . Google Scholar Crossref Search ADS PubMed WorldCat [15] Muyzer G. Teske A. Wirsen C.O. Jannasch H.W. ( 1995 ) Phylogenetic relationships of Thiomicrospira species and their identification in deep-sea hydrothermal vent samples by denaturing gradient gel electrophoresis of 16S rDNA fragments . Arch. Microbiol. 164 , 165 – 172 . Google Scholar Crossref Search ADS PubMed WorldCat [16] Maidak B.L. Cole J.R. Lilburn T.G. Parker C.T. Saxman P.R. Stredwick J.M. Garrity G.M. Li G. Olsen G.J. Pramanik S. Schmidt T.M. Tiedje J.M. ( 2000 ) The RDP (Ribosomal Database Project) continues . Nucleic Acids Res. 28 , 173 – 174 . Google Scholar Crossref Search ADS PubMed WorldCat [17] Marsh T.L. Saxman P. Cole J. Tiedje J. ( 2000 ) Terminal restriction fragment length polymorphism analysis program, a web-based research tool for microbial community analysis . Appl. Environ. Microbiol. 66 , 3616 – 3620 . Google Scholar Crossref Search ADS PubMed WorldCat [18] Dunbar J. Ticknor L.O. Kuske C.R. ( 2001 ) Phylogenetic specificity and reproducibility and new method for analysis of terminal restriction fragment profiles of 16S rRNA genes from bacterial communities . Appl. Environ. Microbiol. 67 , 190 – 197 . Google Scholar Crossref Search ADS PubMed WorldCat [19] Lukow T. Dunfield P.F. Liesack W. ( 2000 ) Use of T-RFLP technique to assess spatial and temporal changes in the bacterial community structure within an agricultural soil planted with transgenic and non-transgenic potato plants . FEMS Microbiol. Ecol. 32 , 241 – 247 . Google Scholar Crossref Search ADS PubMed WorldCat [20] Commision of the European Communities ( 1986 ) Council directive of 12 June 1986 on the protection of the environment, and in particular of the soil, when sewage sludge is used in agriculture . Off. J. Eur. Commun. L181 , 6 – 12 . OpenURL Placeholder Text WorldCat Google Scholar [21] Mullins G.L. Martens D.C. Miller W.P. Korngay E.T. Hallock D.L. ( 1982 ) Copper availability, form and mobility in soils from three copper-enriched hog manure applications . J. Environ. Qual. 11 , 316 – 320 . Google Scholar Crossref Search ADS WorldCat [22] Römkens P. Hoenderboom G. Dolfing J. ( 1999 ) Copper solution geochemistry in arable soils: field observations and model application . J. Environ. Qual. 28 , 776 – 783 . Google Scholar Crossref Search ADS WorldCat [23] Bartlett R. James B. ( 1980 ) Studying dried soil samples – Some pitfalls . Soil Sci. Soc. Am. J. 44 , 721 – 724 . Google Scholar Crossref Search ADS WorldCat [24] Bruemmer G.W. Gerth J. Herms U. ( 1986 ) Heavy metal species, mobility and availability in soils . Z. Pflanzenernaehr. Bodenkd. 149 , 382 – 398 . Google Scholar Crossref Search ADS WorldCat [25] Swift R.S. McLaren R.G. ( 1991 ) Micronutrient adsorbtion by soils and soil colloids . In: Interactions at the Soil Colloid–Soil Solution Interphase ( Bolt G.H. Ed.), pp. 257 – 292 . Kluwer , Dordrecht . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [26] Speir T.W. Kettles H.A. Percival H.J. Parshotam A. ( 1999 ) Is soil acidification the cause of biochemical responses when soils are amended with heavy metal salts Soil Biol. Biochem. 31 , 1953 – 1961 . Google Scholar Crossref Search ADS WorldCat [27] Øvreås L. Torsvik V. ( 1998 ) Microbial diversity and community structure in two different agricultural soil communities . Microb. Ecol. 36 , 303 – 315 . Google Scholar Crossref Search ADS PubMed WorldCat [28] Borneman J. Skroch P.W. O'Sullivan K.M. Palus J.A. Rumjanek N.G. Jansen J.L. Nienhuis J. Triplett E.W. ( 1996 ) Molecular microbial diversity of an agricultural soil in Wiscounsin . Appl. Environ. Microbiol. 62 , 1935 – 1943 . Google Scholar PubMed OpenURL Placeholder Text WorldCat [29] Reichenbach H. ( 1992 ) The order Cytophagales . In: The Prokaryotes, Vol. 4 , 2nd edn. ( Barlows A. Truper H.G. Dworkin M. Harder W. Schleifer K.-H. Eds.), pp. 3631 – 3687 . Springer-Verlag , Berlin . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [30] Frostegård Å. Tunlid A. Bååth E. ( 1993 ) Phospholipid fatty acid composition, biomass and activity of microbial communities from two soil types experimentally exposed to different metals . Appl. Environ. Microbiol. 59 , 3605 – 3617 . Google Scholar PubMed OpenURL Placeholder Text WorldCat [31] Duxbury T. Bicknell B. ( 1983 ) Metal tolerant bacterial populations from natural and metal-polluted soils . Soil Biol. Biochem. 15 , 243 – 250 . Google Scholar Crossref Search ADS WorldCat [32] Maliszewska W. Dec S. Wierzbicka H. Wozniakowska A. ( 1985 ) The influence of various heavy metal compounds on the development and activity of soil micro-organisms . Environ. Pollut. 37 , 195 – 215 . Google Scholar Crossref Search ADS WorldCat [33] Doelman P. Jansen E. Michels M. ( 1994 ) Effects of heavy metals in soil on microbial diversity and activity as shown by the sensitivity-resistance index, an ecological relevant parameter . Biol. Fertil. Soils 17 , 177 – 184 . Google Scholar Crossref Search ADS WorldCat [34] Barkay T. Tripp S.C. Olson B.H. ( 1985 ) Effects of metal-rich sewage sludge application on the bacterial communities of grasslands . Appl. Environ. Microbiol. 49 , 333 – 337 . Google Scholar PubMed OpenURL Placeholder Text WorldCat [35] Roane T.M. Kellogg S.T. ( 1996 ) Characterization of bacterial communities in heavy metal contaminated soils . Can. J. Microbiol. 42 , 593 – 603 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes 1 Chr. Hansen A/S, Animal Health RD&A, Bøge Allé 10–12, DK-2970 Hørsholm, Denmark. © 2003 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
TI - Effects of copper amendment on the bacterial community in agricultural soil analyzed by the T-RFLP technique
JF - FEMS Microbiology Ecology
DO - 10.1016/S0168-6496(03)00192-2
DA - 2003-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/effects-of-copper-amendment-on-the-bacterial-community-in-agricultural-lSl66Q8PQc
SP - 53
EP - 62
VL - 46
IS - 1
DP - DeepDyve
ER -