Introduction

The s-triazine herbicide, atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-s-triazine), is a common water and soil contaminant due to its intense agricultural application and high mobility (Jablonowski et al. 2011). At spill sites, atrazine can be present in soil and groundwater at very high concentrations (Strong et al. 2002; Silva et al. 2004). The potential serious ecological and health risks have increased the attention to atrazine reduction in contaminated sites. Atrazine can be degraded by both abiotic and biotic processes (Kruger et al. 1997); however, biodegradation has been suggested to be the major mechanism of atrazine dissipation in soil environments (Udiković-Kolić et al. 2010; Zhang et al. 2011a; Sebaï et al. 2011). Atrazine biodegradation in soils can be inhibited by nitrogen fertilization (Abdelhafid et al. 2000; Rhine et al. 2003; Zablotowicz et al. 2008). A large concentration of inorganic nitrogen can greatly decrease atrazine degradation in soils (Entry et al. 1993; Abdelhafid et al. 2000). However, mineralization of atrazine in soils can be inhibited even at nitrogen levels below typical field applications (Garces et al. 2007). Therefore, for soil bioremediation at spill sites, bioaugmentation with atrazine degraders insensitive to exogenous nitrogen sources is of practical importance.

A number of atrazine-degrading microorganisms from diverse bacterial genera have been isolated and characterized (Wackett et al. 2002; Li et al. 2008), and most of them can use atrazine as a nitrogen source (Yang et al. 2010). Therefore, nitrogen compounds in soils might be a major constraint in atrazine reduction (Yang et al. 2010; Wang and Xie 2012). This might partly account for poor survival of atrazine degraders inoculated in soils with high concentrations of atrazine (Struthers et al. 1998; Garcia-Gonzalez et al. 2003; Silva et al. 2004). Atrazine-degrading Arthrobacter species usually harbor three functional genes, trzN, atzB, and atzC, with a capacity to convert atrazine to cyanuric acid (Wang and Xie 2012). There have been numerous reports on atrazine-degrading Arthrobacter isolates from agricultural soils (Aislabie et al. 2005; Sebaï et al. 2011; Zhang et al. 2011a, Wang and Xie 2012). Some Arthrobacter strains were isolated from soils in agrochemical factory area, exposed to repeated spills of effluent from atrazine synthesis (Kolić et al. 2007). Arthrobacter-like microorganisms were dominant in bacterial communities enriched from soil collected within the same area (Udiković-Kolić et al. 2010). An Arthrobacter aurescens strain was also isolated from heavily contaminated soils (with 29,000 mg kg−1 atrazine) at a spill site (Strong et al. 2002). Moreover, several Arthrobacter strains have been obtained from industrial wastewater from atrazine production plants (Cai et al. 2003; Li et al. 2008; Wang et al. 2011a). Arthrobacter may have a competitive advantage at contaminated sites, where s-triazine compounds are present at high concentrations (Udiković-Kolić et al. 2010). Many Arthrobacter species have shown high degradation ability in liquid cultures (Cai et al. 2003; Wang et al. 2011a; Zhang et al. 2011a; Wang and Xie 2012). Moreover, atrazine degradation by Arthrobacter species may not be inhibited in liquid culture containing nitrogen compounds (Sajjaphan et al. 2010; Wang and Xie 2012). These previous reports suggest a strong potential of Arthrobacter species to remediate heavily contaminated soils. Unfortunately, little is known about bioaugmentation with Arthrobacter species to bioremediate-contaminated soils. Li et al. (2008) found that Arthrobacter strain AD26 could completely eliminate atrazine in soil (300 mg kg−1) in 20 days after inoculation. To the authors' knowledge, there is no information available on nitrogen impacts on survival of Arthrobacter strain and the strain's degradation efficiency in soils.

Traditional culture-dependent approaches have been usually applied to assess the biodegradation potential of s-triazine compounds in soils (Clausen et al. 2002; Morán et al. 2006; Dinamarca et al. 2007; Hernández et al. 2008; Morgante et al. 2010). However, culture-dependent approaches cannot adequately estimate the herbicide degradation potential (Clausen et al. 2002; Morán et al. 2006). The atrazine-degrading potential of soil microbiota can be more precisely estimated by quantifying atz and trz gene copies (Piutti et al. 2003; Martin-Laurent et al. 2004; Udiković-Kolić et al. 2010, 2011). Moreover, culture independent techniques (e.g., terminal restriction fragment length polymorphism (TRFLP)) are also useful for assessing the structure of soil microbial community and its change (Wang et al. 2011b; Zhang et al. 2011b, c). Although many previous research works have provided a substantial knowledge about the diversity of atrazine degraders, the change of microbial community structure associated with inoculation of atrazine degraders remains poorly known (Morgante et al. 2010). The aim of the current study was to investigate the impacts of exogenous nitrogen on a microbial community inoculated with the atrazine-degrading Arthrobacter sp. strain DAT1 in soil amended with a high concentration of atrazine. The strain DAT1 harbors atrazine-metabolic genes, trzN, atzB, and atzC and illustrates very high degradation ability in liquid culture regardless of high concentration of exogenous nitrogen compounds (Wang and Xie 2012). In this study, the relative abundance of the strain DAT1 and the bacterial community structure in soils were characterized using quantitative PCR (q-PCR) in combination with TRFLP.

Materials and methods

Microcosm set-up

The strain DAT1 was incubated in a basic salt medium (Wang and Xie 2012), with a supplement of 3 g L−1 sucrose and 50 mg L−1 atrazine, at 30 °C under shaking (150 rpm) for 48 h. Cells grown until turbidity600 nm = 1.0 (2.65 × 109 CFU mL−1), were harvested, and washed three times in sterile NaCl solution (8.5 g L−1).

Soil was collected from an agricultural area that has received typical atrazine application for more than 20 years. Before use, the soil was air dried and sieved through a 5-mm sieve. The soil contained 1,100 mg kg−1 organic nitrogen and 85 mg kg−1 available nitrogen (sum of NO −3 -N and NH +4 -N), including 30.8 mg kg−1 NO −3 -N. At the time of collection, no residual atrazine was detected. Microcosms were prepared in closed 250-mL jars with 40 g soil (dry weight) and incubated at 25 °C in the dark. Soil moisture was maintained at 10 % of the water-holding capacity. Four sets of treatments were performed as follows: (A) 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1; (B) 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 240 mg kg−1 NO −3 -N; (C) 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 1,000 mg/kg NO −3 -N; (D) 400 mg kg−1 atrazine + 1,000 mg kg−1 NO −3 -N. All microcosms were prepared in triplicate.

For measurement of the residual atrazine in soils, samples from treatments A,B, C, and D were collected at days 0, 1, 2, 3, 4, 5, 7, and 10. Each soil sample (1 g, dry weight) was extracted with 20 mL of methanol under ultrasonication (30 min; frequency 30 kHz), followed by a 5-min centrifugation (5,000 rpm). The extracts were subject to high performance liquid chromatography (HPLC) analysis using a HPLC apparatus equipped with a LC-10AT pump, a UV-detector (Shimadzu), and a Venusil XBP C18 column (Agela Technologies), using methanol–water (70:30) as the mobile phase at a flow rate of 1.0 mL min−1 (Behki and Khan 1994). Atrazine was detected by absorbance at 220 nm. The mean recovery rate of atrazine extraction was 96 %. Calibration was performed using the analytical standard of atrazine and the correlation was determined by linear regression with coefficient of determination, r 2 ≥ 0.99.

Quantitative PCR assays

For molecular analysis, genomic DNA samples from treatments A,B, C, and D were collected at days 0, 5, 7, and 10, using the UltraClean DNA extraction kit (Mobio Laboratories). SYBR Green q-PCR assays was performed in an ABI 7500 FAST (Applied Biosystems) in a reaction mixture of 25 μL. The mixture contained 12.5 μL of SYBR Green PCR master mix, 1 μmol L−1 of each primer and 1.5 μL template DNA (10 ng). The amplification conditions were as follows: 95 °C for 2 min; 40 cycles of 10 s at 95 °C, 20 s at 60 °C, 1 min at 72 °C followed by one melting cycle carried out from 60 °C to 95 °C by incrementing temperature of 0.2 °C s−1. The specific primers for the amplification of the 16S rRNA, atzB, and atzC genes were as previously described (Piutti et al. 2003; Devers et al. 2004; Udiković-Kolić et al. 2010, 2011). Calibration curves relating the log of the copy number of the target gene as a function of the C t (cycle threshold) were as follows: log (16S rDNA) = −0.300 C t +10.20 (R 2 = 0.998), log (atzB) = −0.288 C t +11.44 (R 2 = 0.997), and log (atzC) = −0.297 C t +11.18 (R 2 = 0.998).

TRFLP analysis

16S rRNA genes were amplified using bacterial primers 27 F-FAM (5′-AGAGTTTGATCMTGGCTCAG-3′, 5′ end labeled with carboxyfluorescine) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) under the PCR conditions as previously described (Zhang et al. 2011b, c). Purified PCR products (300 ng) were digested with HhaI. TRFLP fragment was analyzed using an ABI 3730 DNA Analyzer (Applied Biosystems) and Genemapper v3.5 software (Applied Biosystems). The relative abundance of each fragment equals the ratio of the area of each peak to the total area of all peaks in a given TRFLP profile (Zhang et al. 2011b, c). Similarities were calculated using the Bray–Curtis similarity index and clustering was performed by the unweighted pair group mean average method using the PRIMER 5.0 software (Clarke and Warwick 2001).

Results and discussion

Degradation of atrazine

Figure 1 showed a very limited reduction of atrazine in the microcosm without added bacteria (treatment D) during the 10-day incubation period. In the microcosms with treatments A and B, high removal rates (average 71.7 % and 81.2 %, respectively) were observed in 3 days after inoculation with the strain DAT1. Moreover, a nearly complete removal (>97 %) was achieved by day 5. However, only poor atrazine reduction (<5 %) occurred in the microcosm with treatment C in 10 days after inoculation.

Fig. 1
figure 1

Residual atrazine in microcosms with treatments A–D during the incubation period. Treatment A: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1; treatment B: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 240 mg kg−1 NO −3 -N; treatment C: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 1,000 mg/kg NO −3 -N; and treatment D: 400 mg kg−1 atrazine + 1,000 mg kg−1 NO −3 -N

Full size image

Rapid atrazine degradation has usually been found in agricultural soils that have been exposed to herbicide in normal farming practice (Krutz et al. 2010). Under field conditions, herbicides are usually applied annually at low dose. The concentration of atrazine can affect biodegradation and the microbial response (Mahía et al. 2011). Dzantor and Felsot (1991) found that in soil with high concentration of herbicides, degradation was negligible even after 1 year of incubation. Udiković-Kolić et al. (2011) indicated that increase of atrazine concentration may affect the growth of degraders. Our previous experiment indicated that no obvious reduction of atrazine was observed in an uninoculated soil microcosm containing 400 mg kg−1 atrazine (without exogenous nitrogen sources) during a 17-day incubation period (data not shown). In this study, although the raw soil sample used for microcosm studies was collected from an agricultural field receiving atrazine application in normal farming practice, the atrazine degradation rate in the uninoculated microcosm was very slow. The autochthonous degraders in raw soil were in low abundance and their growths were limited during the short incubation period (see below), which suggested that there was a lag period before the autochthonous degraders could degrade atrazine.

Bioaugmentation by introducing microorganisms with high degradation ability seems to be an ideal way to enhance the biodegradation of s-triazine herbicides in contaminated sites (Morgante et al. 2010). Exogenous nitrogen sources, especially at high concentrations, usually have negative impacts on atrazine biodegradation in soils (Entry et al. 1993; Abdelhafid et al. 2000; Rhine et al. 2003; Zablotowicz et al. 2008). However, little is known about the nitrogen impact on the added degraders for bioremediation of soils with high concentrations of s-triazine herbicides. The available nitrogen in soils is usually below 300 mg kg−1 (Garces et al. 2007). In this study, external addition of a medium level of nitrate (240 mg kg−1 NO −3 -N) had no negative effect on atrazine degradation. This result implied that bioaugmentation with the strain DAT1 could be a very efficient biotechnology for bioremediation of soils with high concentrations of atrazine. However, in some soils (e.g., overfertilized soil), the available nitrogen could exceed 1,000 mg kg−1 (Garces et al. 2007), which might significantly inhibit atrazine degradation by the strain DAT1. Moreover, in uninoculated soil microcosms (treatment D), addition of a high level of nitrate (1,000 mg kg−1 NO −3 -N) could further affect atrazine degradation by the indigenous degraders.

Organic nitrogen is the most prevalent nitrogen form in soils (Holst et al. 2012). Several recent works have investigated the effects of organic nitrogen on atrazine degradation (Sajjaphan et al. 2010; Yang et al. 2010; Tappin et al. 2012). Sajjaphan et al. (2010) found that urea had no effect on atrazine degradation by Arthrobacter sp strain KU001. However, Yang et al. (2010) found that urea as an additional nitrogen source could increase the biomass of an atrazine-degrading consortium. Tappin et al. (2012) showed that addition of a peptide co-substrate increased atrazine removal in the River Mersey incubation but did not increase the removal rate in the River Tamar incubation. These others' previous reports implied that organic nitrogen have no negative impact on atrazine biodegradation. However, our previous reports showed, under culture conditions, atrazine degradation activity of the strain DAT1 was only slightly lower at a high level of urea (1,000 mg L−1) than that at a medium level (100 mg L−1) (Wang and Xie 2012). In this study, the soil contained 1,100 mg kg−1 organic nitrogen, which may have certain effect on atrazine biodegradation in soils. However, more efforts will be necessary in order to clarify the effect of organic nitrogen on atrazine degradation by the strain DAT1.

Change of relative abundance of atzB and atzC genes

The abundance of atzB and atzC genes and 16S rRNA gene sequences was assessed by q-PCR. Ratios of atz genes to 16S rRNA were used to estimate the relative abundance of atrazine degraders among the bacterial community (Udiković-Kolić et al. 2010, 2011). On day 0, the relative abundance of atzB gene was 0.3–0.4 % in the inoculated microcosms, compared with that of 0.1 % in the uninoculated microcosm (Fig. 2). In the microcosms with treatment D, the relative abundance of atzB gene remained very low during the 10-day incubation period, although an increase (from 0.1 % to 2.5 %) could be observed (Fig. 2). However, in the microcosms with treatments A and B, high proportion of atzB gene (16 % and 18 %, respectively) was observed in 5 days after inoculation, followed by a slight variation in subsequent incubation. In the microcosm with treatments C, the relative abundance of atzB gene increased from 0.4 % on day 0 to 4 % on day 10.

Fig. 2
figure 2

Change of the relative abundance of atrB gene during the incubation period. Treatment A: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1; treatment B: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 240 mg kg−1 NO −3 -N; treatment C: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 1,000 mg/kg NO −3 -N; and treatment D: 400 mg kg−1 atrazine + 1,000 mg kg−1 NO −3 -N. The relative abundance of atzB gene equals the ratio of atzB gene copy numbers to 16S rRNA gene copy numbers

Full size image

On day 0, the relative abundance of atzC gene in each microcosm was low (below 0.3 %) (Fig. 3). Similar to the change pattern of atzB gene, in the microcosms with treatments C and D, the relative abundance of atzC gene remained low but increased during the 10-day incubation period. In the microcosms with treatments A and B, high proportion of atzC gene (15–18 % and 16–18 %, respectively) was observed in 5–10 days after inoculation.

Fig. 3
figure 3

Change of the relative abundance of atrC gene during the incubation period. Treatment A: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1; treatment B: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 240 mg kg−1 NO −3 -N; treatment C: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 1,000 mg/kg NO −3 -N; and treatment D: 400 mg kg−1 atrazine + 1,000 mg kg−1 NO −3 -N. The relative abundance of atzC gene equals the ratio of atzC gene copy numbers to 16S rRNA gene copy numbers

Full size image

Repeated application of atrazine promotes the abundance and activity of atrazine-degrading microorganisms in agricultural soils (Martin-Laurent et al. 2004). Presence of atzB and atzC genes in agricultural soils has been well-documented (Piutti et al. 2002; Martin-Laurent et al. 2004). The amount of atzC was founded to be transiently enhanced in agricultural soils following atrazine treatment at low concentration (1.5 mg kg−1 soil) (Piutti et al. 2002). In this study, the proportion of atzB and atzC genes in the uninoculated microcosm gradually increased following amendment of a high concentration of atrazine (400 mg kg−1). However, in a bacterial culture previously adapted to a low atrazine concentration, relative abundance of atzB and atzC genes decreased by a factor of 3–4 following an increase in atrazine concentration (Udiković-Kolić et al. 2011). Moreover, low abundance of atzB and atzC genes in the uninoculated microcosms suggests only a small proportion of the autochthonous bacterial community had the capacity to degrade atrazine. This may explain the low atrazine degradation rate in the uninoculated microcosm.

Several previous works have been carried out to investigate the feasibility of using atrazine-degrading microorganisms to remediate soils with high atrazine contamination (Struthers et al. 1998; Silva et al. 2004; Li et al. 2008). However, these previous works have not paid attention to the abundance of atrazine-metabolic genes or degraders. To the authors' knowledge, only Clausen et al. (2002) evaluated the potential of bioaugmentation using Pseudomonas sp. strain ADP in aquifer sediment by quantification of atzA gene. Moreover, although the effects of exogenous nitrogen sources on atrazine biodegradation have been intensively investigated, the information on nitrogen impact on the abundance of atrazine-metabolic genes is still lacking. In the microcosms with treatments A and B, high proportion of atzB and atzC genes was observed in 5 days after inoculation, which suggested a significant growth of the strain DAT1 and might account for the rapid atrazine biodegradation. Moreover, at each sampling date, the relative abundance of atzB and atzC genes with both treatments was almost similar. Therefore, addition of a medium level of nitrate did not decrease the relative abundance of atrazine-metabolic genes. However, addition of a high level of nitrate (1,000 mg kg−1 NO −3 -N) significantly decreased the relative abundance of atrazine-metabolic genes. This was also in agreement with the results of atrazine biodegradation.

Change of relative abundance of TRFLP fragment

The relative abundance of the strain DAT1was further assessed using TRFLP. The in silico HhaI cut site of the cloned 16S rRNA sequence of the strain DAT1 was 479 bp, while the observed HhaI terminal fragment was 477 bp (data not shown). The observed fragment lengths and those predicted using sequence data can often have a slight difference (2–3 bases) (Sun et al. 2010; Xie et al. 2011; Zhang et al. 2011d, 2012). On day 0, fragment 477 bp in each microcosm was below the detection limit (Fig. 4). In the microcosm with treatments C and D, fragment 477 bp was not detected during the 10-day incubation period. However, in the microcosms with treatments A and B, high relative abundance of fragment 477 bp (4.3 % and 7.1 %, respectively) was observed in 5 days after inoculation, followed by a decrease in subsequent incubation.

Fig. 4
figure 4

Change of the relative abundance of fragment 477 bp (HhaI). Treatment A: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1; treatment B: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 240 mg kg−1 NO −3 -N; treatment C: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 1,000 mg/kg NO −3 -N; and treatment D: 400 mg kg−1 atrazine + 1,000 mg kg−1 NO −3 -N. The relative abundance of fragment 477 bp equals the ratio of the area of its peak to the total area of all peaks in a given TRFLP profile

Full size image

TRFLP can be used to characterize the relative abundance of specific microorganisms in the microbial community (Xie et al. 2011; Zhang et al. 2011d, 2012). However, to the authors' knowledge, there has been no report available on the application of TRFLP to monitor the survival of inoculated atrazine-degrading microorganisms. Failure of bioaugmentation was usually due to low survival of inoculated microorganisms (Silva et al. 2004; Morán et al. 2006). In the microcosms with treatments A and B, high relative abundance of fragment 477 bp further confirmed the strong survival ability of the strain DAT1. Moreover, at each sampling date, the relative abundance of the strain DAT1 was higher with treatment B than that with treatment A. This further confirmed that addition of a medium level of nitrate did not have a negative effect on the survival of the strain DAT1. No detection of fragment 477 bp in the microcosm with treatment C also further confirmed that high nitrogen concentration had a negative effect on the survival of the strain DAT1.

Bacterial community structure change

Figure 5 showed the dendrogram constructed for community structures in soils with each treatment. For each treatment, the sample at each sampling date was not clustered with each other, indicating the shift in microbial community structure with time. Moreover, at days 5, 7, and 10, the sample with treatment A was always clustered with the sample with treatment B, while the sample with treatment C clustered with the sample with treatment D. This indicated that the samples in the microcosms with treatments A and B shared the similar microbial community structure but differed from that in the microcosms with treatments C and D.

Fig. 5
figure 5

Cluster diagram of similarity values for soil samples with each treatment. Similarity levels are indicated below the diagram. Upper case letters refer to treatment, and lower case letters indicate sampling date. Treatment A: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1; treatment B: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 240 mg kg−1 NO −3 -N; treatment C: 400 mg kg−1 atrazine + 6.5 × 106 CFU g−1 strain DAT1+ 1,000 mg/kg NO −3 -N; and treatment D: 400 mg kg−1 atrazine + 1,000 mg kg−1 NO −3 -N

Full size image

Microbial community usually changes with hydrocarbon biodegradation in soils (Mills et al. 2003; Muckian et al. 2009; Zhang et al. 2011b, c). However, little is known about the microbial community in response to biodegradation of s-triazine herbicide. Mahía et al. (2011) found that microbial community structure was significantly affected by both soil type and incubation time after atrazine addition. Bacterial communities can also be significantly impacted during simazine biodegradation (Morán et al. 2006). Moreover, Morgante et al. (2010) revealed that bioaugmentation with Pseudomonas sp strain MHP41 promotes simazine attenuation and bacterial community changes in agricultural soils. However, unfortunately, no information is available on the effect of nitrogen sources on bacterial community changes associated with bioaugmentation with the s-triazine herbicide degrader. In this study, shift in microbial community structures was observed in both the inoculated and uninoculated microcosms. Since the samples in the microcosms with treatments A and B shared the similar microbial community structure at each sampling date, addition of a medium level of nitrate did not show significant impact on the microbial community structure during the incubation. However, the distinct difference of microbial community structure in the microcosm with treatment C from that in the microcosm with treatment A or B illustrated a significant impact of high levels of nitrate.

Concluding remarks

The autochthonous bacterial community alone could not quickly dissipate the high concentration of atrazine in soil. However, bioaugmentation with the strain DAT1 could be a very efficient biotechnology for bioremediation of soils with high concentrations of atrazine. Atrazine degradation by the strain DAT1and the strain's atrazine-metabolic potential and survival were not affected by addition of a medium level of nitrate, but these processes were inhibited by the addition of a high level of nitrate. Change of microbial community structure occurred in both inoculated and uninoculated microcosms and was dependent on the level of added nitrate.