Rhizodeposits of Trifolium pratense and Lolium perenne: their comparative effects on 2,4-D mineralization in two contrasting soils

Soil Biology & Biochemistry 37 (2005) 995–1002 www.elsevier.com/locate/soilbio

Rhizodeposits of Trifolium pratense and Lolium perenne: their comparative effects on 2,4-D mineralization in two contrasting soils Liz J Shawa, Richard G. Burnsb,* a b

Imperial College London, Wye Campus, Department of Agricultural Sciences, Wye, Kent TN25 5AH, UK School of Land and Food Sciences, The University of Queensland, Brisbane, Queensland 4072, Australia Received 6 April 2004; received in revised form 25 October 2004; accepted 29 October 2004

Abstract Rhizosphere enhanced biodegradation of organic pollutants has been reported frequently and a stimulatory role for specific components of rhizodeposits postulated. As rhizodeposit composition is a function of plant species and soil type, we compared the effect of Lolium perenne and Trifolium pratense grown in two different soils (a sandy silt loam: pH 4, 2.8% OC, no previous 2,4-D exposure and a silt loam: pH 6.5, 4.3% OC, previous 2,4-D exposure) on the mineralization of the herbicide 2,4-D (2,4-dichlorophenoxyacetic acid). We investigated the relationship of mineralization kinetics to dehydrogenase activity, most probable number of 2,4-D degraders (MPN2,4-D) and 2,4-D degrader composition (using sequence analysis of the gene encoding a-ketoglutarate/2,4-D dioxygenase (tfdA)). There were significant (P!0.01) plant–soil interaction effects on MPN2,4-D and 2,4-D mineralization kinetics (e.g. T. pratense rhizodeposits enhanced the maximum mineralization rate by 30% in the acid sandy silt loam soil, but not in the neutral silt loam soil). Differences in mineralization kinetics could not be ascribed to 2,4-D degrader composition as both soils had tfdA sequences which clustered with tfdAs representative of two distinct classes of 2,4-D degrader: canonical R. eutropha JMP134-like and oligotrophic a-proteobacterial-like. Other explanations for the differential rhizodeposit effect between soils and plants (e.g. nutrient competition effects) are discussed. Our findings stress that complexity of soil– plant–microbe interactions in the rhizosphere make the occurrence and extent of rhizosphere-enhanced xenobiotic degradation difficult to predict. q 2004 Elsevier Ltd. All rights reserved. Keywords: Biodegradation; 2,4-Dichlorophenoxyacetic acid; Rhizosphere; Rhizoremediation; Rhizodeposition; TfdA; Trifolium repens; Lolium perenne

1. Introduction Rhizoremediation, defined as the acceleration of organic pollutant breakdown in soil as a consequence of the enhanced biodegradative activity of rhizosphere microorganisms, is a property with potential for use in the clean up of contaminated soils (Siciliano and Germida, 1998). There are numerous studies (Shaw and Burns, 2003) and throughout) that present evidence for rhizosphere-enhanced biodegradation of a wide range of organic xenobiotics involving a large number of plant species. However, the mechanisms of rhizosphere-enhanced biodegradation are not fully understood although several have been postulated (Siciliano and Germida, 1998; Shaw and Burns, 2003). * Corresponding author. Tel.: C61 7 3365 2509; fax: C61 7 3365 1177. E-mail address: [email protected] (R.G. Burns). 0038-0717/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.soilbio.2004.10.020

One suggestion concerns the role of chemicals released by plant roots (rhizodeposits) as natural analogs of xenobiotic catabolic pathways since it has been noted (Gilbert and Crowley, 1997; Siciliano and Germida, 1998; Dunning Hotopp and Hausinger, 2001; Singer et al., 2003) that some rhizodeposit components are structurally similar to xenobiotics or their metabolites. The term rhizodeposit is used to define a spectrum of components ranging from simple exudate compounds to entire root fragments, released during sloughing of border cells and turnover of dead roots. It is known that both soil physicochemical properties and microbial community composition influence the quality and the quantity of rhizodeposition. For example, plant species growing in low nutrient environments may increase the concentration of nutrient-scavenging extracellular enzymes (e.g. phosphatases) or nutrient-solubilizing phenolics and organic

996

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

acids (Dakora and Phillips, 2002). Meharg and Killham (1995) have shown that Lolium perenne seedlings increase exudation of labelled C from 1% of that photosynthetically assimilated in the absence of rhizosphere microorganisms to up to 34% in the presence of microorganisms. In turn, soil type (Marschner et al., 2001) and the presence of plant roots (Marschner et al., 2001; Smalla et al., 2001) influence microbial community structure. Thus, the rhizosphere may be defined as a complex reciprocal relationship of soil, plant and microorganism with rhizodeposition both a driver and consequence of this interaction. It follows that the kinetics of xenobiotic biodegradation in the rhizosphere may be sensitive to such plant–soil–microbe interactions if specific components of rhizodeposition are key to rhizosphereenhanced breakdown. Shaw and Burns (2004, 2005) have shown that mineralization of the broad-leaved weed herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) is enhanced due to the presence of legume rhizodeposits in an acidic (pH 4) sandy silt loam soil from Sourhope, Scotland. The rhizodeposit enhancement had legume (e.g. Trifolium pratense, Lotus corniculatus, Medicago sativa) specificity; no enhancement was recorded for non-legume species (e.g. the monocotyledon Lolium perenne). Our aim was to compare the legume versus monocotyledon rhizodeposit effect on 2,4-D mineralization kinetics recorded for Sourhope soil with that recorded for a second, neutral (pH 6.5) sandy loam soil. In the resulting soil and planting combinations, we related kinetics of 2,4-D mineralization to soil chemical and microbial properties.

and 15 mg mK2 of 2,4-D ethyl hexyl ester, mecoprop-P ethyl hexyl ester and dicamba, respectively) applied 9 months prior to sampling. Grid reference details and some soil properties are given in Table 1. The soils were collected from 10–40 cm depth, sealed in sterile polythene bags and transported to the laboratory where they were were sieved (!2.8 mm) and stored at 4 8C until use. 2.2. Plant growth Plants were grown in boiling tubes (25 mm external diameter!150 mm length) as described in (Shaw and Burns, 2004). Briefly, replicate tubes containing 12 g (dry weight equivalent) field moist soil were planted with 20 seeds per tube of either Lolium perenne or Trifolium pratense (Herbiseed, Twyford, England). As controls, additional tubes were left non-planted but were otherwise treated identically to the planted tubes. Tubes were kept at 20 8C under a light-dark cycle of 16 h light (4 200 lux) and 8 h dark. Tubes were weighed every 2 days and sufficient distilled water added to bring the weight to the initial value; no correction was made for increasing biomass of the plant when adjusting the water content. After 25 d from sowing, tubes were destructively sampled: shoots were excised and weighed and the roots chopped finely with a sterile scalpel and homogenised with the soil (i.e. the entire below-soil surface contents of the tube were defined as the rhizosphere and included rhizoplane and endorhizosphere microbial colonizers). Sub-samples were taken for biochemical, microbiological and 2,4-D mineralisation assays.

2. Materials and methods

2.3. Soil biochemical and microbiological analysis

2.1. Soil and soil properties

Soil dehydrogenase activity was determined using an unbuffered iodonitrotetrazolium chloride (INT; SigmaAldrich Co. Ltd, Gillingham, Dorest, UK) substrate solution and an assay adapted from Trevors (1984) and von Mersi and Schinner (1991), see Shaw and Burns (2004). Soil sub-samples (0.5 g dry weight) were used as the basis for a 10-fold dilution series in the determination of most probable number of 2,4-D degraders (MPN2,4-D) by assay of 2,4-D disappearance from inoculated mineral broth

The two different soils used were a brown forest soil from Sourhope Research Station in Scotland (see Shaw and Burns, 2004 for further details) and a garden soil (Boughton Lees, Kent, England). Sourhope (SH) soil had no known history of 2,4-D application whereas Boughton Lees (BL) soil had SBK Brushwood Killer (Vitax Ltd, Coalville, Leics, UK) at a rate of 1 ml m K2 (equivalent to 71, 35

Table 1 Selected physico-chemical properties of Sourhope (SH) and Boughton Lees (BL) soil. Data are meanGstandard error of the mean (nR3) Soil (GB grid reference)

Texturea

pHb

Total organic carbon (%)c

Ammonium-N (mg gK1)d

Nitrate-N (mg gK1)e

Phosphate-P (mg gK1)f

BL (TR 025 471) SH (NT 850 205)

Sandy loam Sandy silt loam

6.53G0.023 4.03G0.01

4.33G0.08 2.84G0.05

208G14 168G20

42.2G3.2 26.0G0.9

104G5 32.4G1

a

Soil Survey of England and Wales classification. 1:2.5 (w/v) soil: 10 mM CaCl2. c According to the method of Nelson and Sommers (1982). d Ammonium was extracted using 2 M KCl and concentrations in extracts determined using a colorimetric method (Forster, 1995a). e Nitrate was extracted using distilled water and concentrations in extracts determined using a colorimetric method (Forster, 1995a). f Phosphate was extracted using 0.5 M NaHCO3 and concentration in extracts determined using the ammonium molybdate-ascorbic acid colorimetric method (Forster, 1995b). b

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

(50 mg 2,4-D mlK1) after incubation (9 weeks, 20 8C) (five replicate tubes per dilution). Further details are given in Shaw and Burns (2004). 2.4. Survey of putatively dominant 2,4-D degraders by PCR amplification from positive MPN2,4-D tubes and cloning amplified sequences At the conclusion of 2,4-D mineralization assays (d 36 and d 22 for SH and BL soil, respectively), soil sub-samples were taken for MPN2,4-D assays (see Section 2.3). After incubation, the bacteria in the positive tubes were harvested by centrifugation (10 min, 13,000!g), and the pellets were re-suspended in 20 ml dd H2O. DNA was extracted from the cell concentrate (3 ml) by Microlysise (Cambio, Cambridge, UK) according to the suppliers instructions. The resulting DNA was the template in PCR using primers (tfdAf (5 0 -AC(C/G)GAGTTC(G/T)(C/G)CGACATGCG3 0 ) and tfdAr (5 0 -GCGGTTGTCCCACATCAC-3 0 )) designed by Itoh et al. (2002) to target the tfdA gene which encodes a-ketoglutarate/2,4-D dioxygenase. Reaction mixtures contained PCR buffer (including 1.5 mM MgCl2), 1!Q solution, 1 U HotStarTaq DNA polymerase (Qiagen, UK), dNTP solution (200 mM dATP, dCTP, dGTP, dTTP), forward and reverse primers (0.4 mM of each), 2 ml template and sterile distilled water up to 50 ml. Reactions were held at 95 8C for 15 min to activate the HotStarTaq prior to the start of the temperature program (30 cycles of: denaturation (45 s at 94 8C); annealing (45 s at 53 8C); extension (1 min at 72 8C) and a final extension of 10 min at 72 8C). A total of 13 positive MPN2,4-D tubes were screened by PCR for SH soil and 12 for BL soil. Out of these, 9 PCR products (of the correct 357 bp size) were obtained for each soil. tfdA diversity was examined for each PCR product by single strand conformational polymorphism (SSCP) analysis and products producing unique patterns (1 pattern for BL, two for SH) were used as the basis for cloning (see Shaw and Burns, 2004). Briefly, PCR products were ligated into pGEM-T Easy vector (Promega, Southampton, UK) and transformed into E. coli JM109 competent cells (Promega). Plasmid DNA extracted (QIAprepw Spin Miniprep Kit, Qiagen, Crawley, UK) from 16 unique (as determined by SSCP) clones (10 from SH, six from BL) was used for sequencing (Comfort read; MWG Biotech, Ebersberg, Germany). Resulting sequences have been deposited in the Genbank database under accession numbers AY554184–AY554199. 2.5. Analysis of cloned tfdA sequences Cloned tfdA sequences were compared with selected reference tfdA sequences (trimmed to the same size as cloned sequences) in the Genbank database. Alignments were made using ClustalX1.81 and the neighbor-joining phylogenetic tree constructed (Phylip 3.6a3). Bootstrap analysis was conducted with 100 replicates.

997

2.6. 2,4-D mineralization assay The 2,4-D mineralization was assayed as described by Shaw and Burns (2004, 2005). In brief, soil samples (4 g dry weight basis) in 40 ml gas-tight vials were amended with 14 C–UL–2,4-D (50 mg gK1 and 250 Bq gK1). A test tube containing NaOH (1 ml, 1 M) was placed onto the surface of the soil to trap the 14CO2 released. Vials were incubated at 25 8C and 70% of the maximum moisture holding capacity. On sampling days, the NaOH solution was taken for quantification of radioactivity (using a Beckman LS6000TA Liquid Scintillation System with internal quench correction) and replaced with fresh NaOH solution. UltimaGolde (Packard Bioscience, Groningen, The Netherlands) was the scintillant. 14C mineralisation results were expressed as a percentage of the total initial activity of 14C-2,4-D added (determined by counting an aliquot of the spike). 2.7. Statistical analysis Analyses of variance (ANOVA; one-way and two-way) were performed using Minitab 13.1 (Minitab Inc.). MPN data was log10-transformed prior to ANOVA. Curve fitting was carried out using SigmaPlot 8.0 (SPSS Inc., Chicago).

3. Results 3.1. Soil type and plant species effects on dehydrogenase activity and 2,4-D degrader numbers The effects of plant species on soil microbiological and biochemical properties for the two soils collected from the rhizosphere after 25 d plant growth or from non-planted controls are shown in Fig. 1(a) and (b). For dehydrogenase activity (Fig. 1(a)), two-way ANOVA revealed a highly significant (P!0.001) effect of soil type; activities in BL soil were between 10- and 20-fold greater than those in SH, however there was no significant effect of plant treatment (PZ0.098). Most probable numbers of 2,4-D degraders were low (!20 gK1) in both soils with the exception of Trifoliumplanted BL soil where median MPN2,4-D were 315 gK1 (Fig. 1(b)). In one replicate from non-planted SH soil, MPN 2,4-D were below the detection limit (1.8 MPN2,4-D gK1). Two-way ANOVA (on log10-transformed data) revealed highly significant effects of soil, plant and soil * plant interaction (P!0.001 in all cases), presumably largely due to the increased MPN2,4-D in Trifolium-planted BL soil. 3.2. Diversity of dominant 2,4-D degraders in SH and BL soil To assess the diversity of dominant 2,4-D degraders in the two soils we extracted DNA from positive most probable number tubes from the highest dilution (10K4 and 10K5)

998

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

Fig. 1. Effect of plant species and soil type on: (a) dehydrogenase activity and (b) most probable number of 2,4-D degraders for soil collected after 25 d growth from the rhizosphere or from non-planted controls. Bars represent standard error (a) or maximum and minimum values (b) (nZ4). Columns superscripted by the same lowercase letter are not significantly different (PO0.05) within soils; columns superscripted by the same uppercase letter are not significantly different (PO0.05) within plant treatments. DL, detection limit (1.8 MPN gK1).

and used this as a template in PCR using primers targeting the a-ketoglutare/2,4-D dioxygenase gene (tfdA). The phylogenetic tree constructed from cloned tfdA sequences from MPN cultures and reference strains is shown in Fig. 2. The sequences separate into two distinct clusters: Cluster 1 containing tfdA sequences from reference strains R. eutropha JMP134 and Burkholderia sp. RASC and Cluster 2 containing tfdA sequences from reference strains HW12, HW13, RD2-C5, closely related to Bradyrhizobium spp. Both Clusters contained cloned sequences from both soil types. 3.3. Soil type and plant species effects on 2,4-D mineralization The effects of soil type and planting on mineralization kinetics of 14C-2,4-D (50 mg gK1) are shown in Fig. 3(a) and (b). Cumulative mineralization curves for all plant–soil combinations were sigmoidal. For statistical comparison of mineralization kinetics, data in Fig. 3 were fitted to the function YZa[1C(t/t0)b]K1 (Shaw et al., 2002). This function has an asymptote at a, an inflection point (t1, time at which the maximum rate occurs) at t0{[(bK1)/ (bC1)]1/b} and a maximum rate Kab/4t1. YZpercent 2,4-D ring 14C evolved as 14CO2 and tZtime (d). Parameter values and statistical comparisons for the individual effect of soil

Fig. 2. Neighbour-joining dendrogram (Jukes–Cantor distances) to show the phylogenetic position of tfdA sequences cloned from MPN tubes among common partial (357 bp) tfdA sequences from reference strains: Ralstonia eutrophus JMP134 (pJP4), Burkholderia sp. RASC, strains HW12, HW13 and RD2-C5 are members of the BANA cluster of the a proteobacteria most closely related to Bradyrhizobium spp. (Kamagata et al., 1997; Itoh et al., 2000). E. coli TauD (encodes taurine/a-KG dioxygenase; (Van der Ploeg et al., 1996)) was used as the outgroup. MPN clones: B1, BL soil; S1 and S2, SH soil MPN patterns 1 and 2, respectively. Bootstrap confidence limits (percentages) O50% are indicated. Scale bar represents a Jukes– Cantor distance of 0.1. Designations in parenthesis are Genbank accession numbers.

type and planting treatment are given in Table 2. Two-way ANOVA indicated that overall soil type had a highly significant (P!0.001) effect on all three mineralization parameters (a, t1 and maximum rate) whereas plant and soil*plant interaction terms were significant for t1 and maximum rate (P!0.001) but not for a (PR0.14). Examining the individual effect of soil type (within plant) revealed that in comparison to SH, BL soil mineralised 2,4D with an earlier t1, and at a faster maximum rate in nonplanted soil but with a lower a value for Lolium and Trifolium treatments (Table 2, Fig. 3). Separating the individual effect of plants (within soil) reveals that the response of 2,4-D mineralization kinetics to planting differed for the two soils; specifically, Trifolium planting enhanced mineralization in SH soil (significantly increased rate (PZ0.006) and decreased t1 (P!0.001)) whereas Lolium planting inhibited mineralization in BL soil (lower rate (PZ0.008) and later t1 (P!0.001)). Compared to nonplanted controls, Trifolium had no effect (PO0.05) on

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

999

contrast to the use of intact rhizosphere soil where the plant is present in the experimental system. In experiments using intact rhizospheres, the fate of the xenobiotic may be a product of both plant and microbial metabolism and the relative contributions of each may be difficult to disentangle. Accurate quantification of microbial mineralization is particularly problematic in intact systems due to the photosynthetic re-adsorption of 14 C-CO2 (Shaw and Burns, 2003). For these reasons, we chose to use harvested rhizosphere soil that was not subject to current rhizodeposition but had been impacted by the quality and quantity of rhizodeposition prior to harvest. We incorporated the harvested roots so as not to exclude possible microbial colonizers of the rhizoplane and endorhizosphere not easily separated from plant tissues. 4.1. Soil type and plant species effects on dehydrogenase activity and 2,4-D degrader numbers Using dehydrogenase activity as a measure of soil respiratory activity, we found that between-soil differences were pronounced and that activity for SH soil was an order of magnitude lower than in the BL soil. Respiratory activity may have been lower in the SH soil because of its acidic pH and smaller total organic carbon content. Camina et al. (1998) also report very low INT reduction activities for acid (pHw4) soils. vonMersi and Schinner (1991) have determined a pH optima for INT reduction of 7–7.5 whereas Trevors (1984) recorded very little activity below a pH 6.6 or above 9.5. On the other hand, Merlin et al. (1995) have pointed out that the bioavailability of the assay substrate, INT, depends on soil physico-chemical properties, therefore suggesting that it may be misleading to compare absolute dehydrogenase activities between soil types. Comparing within soil type, we found no significant (P!0.05) effect of plant treatment on dehydrogenase activity. This is in contrast to other workers who have reported significantly greater microbial numbers and activities in the rhizosphere in comparison to bulk soil (de Neergaard and Magid, 2001; Buyer et al., 2002). Most probable number assays revealed significantly elevated numbers of 2,4-D degraders in Trifolium planted

Fig. 3. Effect of plant species and soil type on mineralization of 50 mg 2,4-D gK1 in soil sampled from the rhizosphere of L. perenne (,), T. pratense (:) or non-planted controls (C). Bars represent standard error (nZ4).

mineralisation kinetics for BL, whereas Lolium had no effect (PO0.05) on mineralisation kinetics for SH soil (Table 2, Fig. 3).

4. Discussion The direct and indirect effects of plant species and soil properties on the composition and activity of microorganisms in the rhizosphere are numerous and difficult to resolve. We investigated the consequences of plant–soil interactions for the microbial mineralization of the xenobiotic 2,4-D. For the mineralization experiments, harvested (destructively sampled) rhizosphere soils were used. This is in Table 2 Effect of soil type and planting on 2,4-D mineralization parameters Soil

Plant

Parameter (GSEM) a

BL

SH

Non-planted Lolium Trifolium Non-planted Lolium Trifolium

a

61.5 (3.0)aA 62.3 (0.7)aA 63.7 (2.3)aA 68.7 (0.9)aA 67.0 (0.7)abB 72.1 (0.9)bB

t1

maximum rate

09.3 (0.1)aA 14.0 (0.4)bA 10.0 (0.5)aA 19.9 (0.5)aB 18.7 (0.2)aB 14.3 (0.4)bB

13.6 (1.0)aA 08.5 (0.6)bA 14.4 (1.5)aA 07.7 (0.5)aB 08.1 (0.2)aA 10.0 (0.5)bA

a Mean values superscripted by the same lowercase letter are not significantly different (P!0.05) for within-soil comparisons. Mean values superscripted by the same uppercase letter are not significantly different (P!0.05) for within-plant comparisons.

1000

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

BL soil, but not for any other plant–soil treatment combination. Trifolium rhizosphere-to-BL bulk soil ratios (R:S) were 1.63 and 0.90 for culturable heterotrophs (data not shown) and dehydrogenase activity, respectively, compared with a R:S of 27.0 for MPN2,4-D. Bearing in mind the caveats associated with interpretation on cultivation-based techniques, the MPN2,4-D R:S ratio provides evidence for selective enrichment of 2,4-D degraders (i.e. a shift in the microbial community composition such that a greater proportion of the population were 2,4-D degraders). This implies that a component of the rhizodeposits in this particular soil–plant combination acted as a substrate for growth that was selective for the degradative microorganisms. Sandmann and Loos (1984) estimated MPN2,4-D in rhizosphere and non-rhizosphere soil and found that soil sampled from the sugarcane rhizosphere with no previous exposure to phenoxyacetic acid herbicides had an enriched 2,4-D degrading community (R:S2,4-DZ105; R:SheterotrophZ6) and speculated that the stimulatory effect was due to the release of 2,4-D analogs (phenolic compounds) by the sugarcane roots. 4.2. Soil–plant interaction effects on 2,4-D mineralization Despite the evidence for increased abundance of 2,4-D degraders in Trifolium-planted BL soil, there was no difference in mineralization kinetics compared to the nonplanted control. Furthermore, mineralization was inhibited by Lolium in BL soil. Contrastingly, mineralization was enhanced in the Trifolium rhizosphere for SH soil which did not contain a rhizosphere-enriched 2,4-D degrading community. The lack of a Trifolium effect in BL soil may be accounted for in several ways. Previously, it has been hypothesized that legume flavonoids are the stimulatory rhizodeposit fraction responsible for enhanced 2,4-D biodegradation in SH soil, possibly by acting as inducers of the pathway (Shaw and Burns, 2004, 2005). Perhaps a combination of soil chemical and biological factors meant that even though 2,4-D degraders were enriched in BL Trifolium soil, the putative rhizodeposit inducer responsible for enhanced mineralization and produced by Trifolium in SH soil was not produced in sufficient quantity, or at all, in the BL soil. It is known that flavonoid production in plants is regulated by variables related to soil type, such as nutrient availability (Coronado et al., 1995; Dakora and Phillips, 1996) and the presence of specific microbes (Broughton et al., 2000). Another possible explanation for the different mineralization responses of the two soils may relate to the differential growth of the plants. Plants in the BL soil grew better than those in SH soil (shoot dry weights when soil was collected for the mineralization experiments were 15.2G0.4 and 43.2G1.3 mg (Lolium) and 6.6G1.5 and 14.7G1.1 mg (Trifolium) for SH and BL soil, respectively). As previously indicated, rhizodeposition may be a source of compounds

stimulatory to xenobiotic degradation through their roles as inducers or selective carbon sources. However, it should be remembered that in rhizosphere soil, any stimulatory compounds produced are present against a background of comparatively high concentrations of labile, organic carbon rich exudates. Therefore, microorganisms degrading organic xenobiotics as a C source in the rhizosphere will encounter competition for essential nutrient elements (e.g. N and P) from other microorganisms who are catabolising rhizodeposit C sources (Jensen and Nybroe, 1999). Thus, although the greater exudate production during growth and incorporation of the larger root biomass of Lolium to BL soil suggests otherwise, it may have had negative effects on 2,4-D mineralization through competition effects. 4.3. Diversity of 2,4-D degrading populations A third explanation may be that different communities of 2,4-D degraders were present in the SH and BL soil and those in the BL soil did not respond to the rhizodeposit stimulants. To answer this point, we assessed the composition of the community of 2,4-D degraders in the two soils using the following approach. As the template for PCR (using primers targeting the tfdA gene), we used DNA extracted from positive MPN tubes from the highest dilution (10K4 and 10K5). The highest dilution method avoids one bias inherent in the use of culture techniques to study microbial communities (Dunbar et al., 1997), namely the most abundant microbes are sampled, rather than the fastest growing (Ogram et al., 1995). Because the soil used to inoculate the MPN tubes was taken from the conclusion of mineralization experiments and therefore had already been exposed to, and mineralised, 2,4-D, only sequences from those 2,4-D degraders which grew in response to 2,4-D application would be represented in the tfdA clone library. It is interesting that for both SH and BL soil, approximately 25% of the positive MPN tubes (i.e where 2,4-D had been degraded) did not yield a tfdA PCR product. The design of PCR primers relies on depositions of relevant sequence information to databases. Therefore, the failure of the PCR may reflect the inability of the primers to amplify some of the tfdA-like genes present and may hint at the existence of further, as yet undiscovered, diversity in 2,4-D degradative genes. Most 2,4-D degrading bacteria studied to date are members of the b and g proteobacteria and were isolated from 2,4-D contaminated environments (referred to here as canonical degraders). Recently, 2,4-D degraders belonging to the a proteobacteria have been obtained from soils with no history of 2,4-D exposure (Kamagata et al., 1997; Itoh et al., 2000; Itoh et al., 2002). These strains are slowgrowing oligotrophs, are most closely related to the legume symbiont bradyrhizobia, possess tfdA genes (designated tfdAa), which have %60% sequence similarity to the canonical b- and g-proteobacterial tfdA genes, but also novel 2,4-D degradative genes, the cad genes with sequence

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

similarity to 2,4,5-T oxygenase (Kitagawa et al., 2002; Itoh et al., 2004). We did not screen for the presence of cad genes in our study, and indeed, the presence of these genes may have been responsible for 2,4-D degradation in the positive MPN tubes not producing a tfdA product. However, tfdA sequences of both the canonical and the tfdAa cluster were present in both soil types. Given the 2,4-D exposure history of the soils, we might have expected the canonical tfdA to have dominated in the previously-contaminated BL soil, and the pristine SH soil containing only those of the tfdAa type. Interestingly, this was not the case. Thus, differences in mineralization kinetics could not be ascribed to 2,4-D degrader class as both soils had tfdA sequences that clustered with tfdAs representative of the two classes of degrader. However, the detection of tfdAa-like gene sequences in both soils is significant and shows that these, until recently overlooked class of 2,4-D degraders, are widespread (Itoh et al., 2002). Few investigators have compared rhizosphere xenobiotic biodegradation kinetics between soil types. Knaebel and Vestal (1992) tested the effects of soybean (Glycine max) and maize (Zea mays) rhizosphere microbial communities on the mineralization of four surfactants in two soil types: a non-amended woodlot soil (pH 7.9, 10.0% TOC) and an agricultural soil (pH 4.9, 2.9% TOC) with previous exposure to sewage sludge. In contrast to our findings, they found that the rhizosphere significantly increased the first order mineralization rate by 1.1–1.9 times but that soil type had no overall effect on mineralization parameters. Boyle and Shann (1998) examined both the influence of plant species (Timothy grass (Phleum pratense), red clover (Trifolium pratense) and sunflower (Helianthus annuus)) and soil type on the mineralization of 2,4,5-T in rhizosphere soil. The three soil types used ranged in fertility (lowZpH 6.9, 2.2% OM; mediumZpH 7.4, 5.3% OM; and highZpH 7.2, 9.7% OM). They found that 2,4,5-T mineralization was most affected by soil type. However, planting also promoted mineralization with the greatest enhancement recorded for the low fertility soil. As in our study, the importance of the individual plant species depended on the soil type, as indicated by a highly significant (PZ0.0001) plant–soil interaction. Here we show that the response of microbial 2,4-D biodegradation kinetics to the presence of plant roots was not consistent for the two soils tested. These findings stress that the complex soil–plant–microbial interactions occuring in the rhizosphere result in different consequences for xenobiotic biodegradation. Thus, the successful use of a given plant species for rhizoremediation should be proceeded by a study of the biodegradation response in the target soil.

Acknowledgements Most of the research reported in this paper was carried out while the authors were associated with the Research

1001

School of Biosciences, University of Kent, Canterbury, Kent, UK. This research was funded by the European Union and the UK. Natural Environmental Research Council. We thank Sue Edwards for help with the analysis of soil physico-chemical properties.

References Boyle, J.J., Shann, J.R., 1998. The influence of planting and soil characteristics on mineralization of 2,4,5-T in rhizosphere soil. Journal of Environmental Quality 27, 704–709. Broughton, W.J., Jabbouri, S., Perret, X., 2000. Keys to symbiotic harmony. Journal of Bacteriology 182, 5641–5652. Buyer, J.S., Roberts, D.P., Russek-Cohen, E., 2002. Soil and plant effects on microbial community structure. Canadian Journal of Microbiology 48, 955–964. Camina, F., Trasar-Cepeda, C., Gil-Sotres, F., Leiros, C., 1998. Measurement of dehydrogenase activity in acid soils rich in organic matter. Soil Biology & Biochemistry 30, 1005–1011. Coronado, C., Zuanazzi, J.A.S., Sallaud, C., Quirion, J.C., Esnault, R., Husson, H.P., Kondorosi, A., Ratet, P., 1995. Alfalfa root flavonoid production is nitrogen regulated. Plant Physiology 108, 533–542. Dakora, F.D., Phillips, D.A., 1996. Diverse functions of isoflavonoids in legumes transcend anti-microbial definitions of phytoalexins. Physiological and Molecular Plant Pathology 49, 1–20. Dakora, F.D., Phillips, D.A., 2002. Root exudates as mediators of mineral acquisition in low-nutrient environments. Plant and Soil 245, 35–47. de Neergaard, A., Magid, J., 2001. Influence of the rhizosphere on microbial biomass and recently formed organic matter. European Journal of Soil Science 52, 377–384. Dunbar, J., White, S., Forney, L., 1997. Genetic diversity through the looking glass: effect of enrichment bias. Applied and Environmental Microbiology 63, 1326–1331. Dunning Hotopp, J.C., Hausinger, R.P., 2001. Alternative substrates of 2,4dichlorophenoxyacetate/alpa-ketoglutarate dioxygenase. Journal of Molecular Catalysis B: Enzymatic 15, 155–162. Forster, J.C., 1995a. Soil Nitrogen, in: Alef, K., Nannipieri, P. (Eds.), Methods in Applied Soil Microbiology and Biochemistry. Academic Press, New York, pp. 79–87. Forster, J.C., 1995b. Soil Phosphorus, in: Alef, K., Nannipieri, P. (Eds.), Methods in Applied Soil Microbiology and Biochemistry. Academic Press, New York, pp. 88–93. Gilbert, E.S., Crowley, D.E., 1997. Plant compounds that induce polychlorinated biphenyl biodegradation by Arthrobacter sp. strain B1B. Applied and Environmental Microbiology 63, 1933–1938. Itoh, K., Kanda, R., Momoda, Y., Sumita, Y., Kamagata, Y., Suyama, K., Yamamoto, H., 2000. Presence of a 2,4-D-catabolizing bacteria in a Japanese arable soil that belong to BANA (Bradyrhizobium-AgromonasNitrobacter-Afipia) cluster in alpha-Proteobacteria. Microbes and Environments 15, 113–117. Itoh, K., Kanda, R., Sumita, Y., Kim, H., Kamagata, Y., Suyama, K., Yamamoto, H., Hausinger, R.P., Tiedje, J.M., 2002. tfdA-like genes in 2,4-dichlorophenoxyacetic acid-degrading bacteria belonging to the Bradyrhizobium-Agromonas-Nitrobacter-Afipia cluster in a-Proteobacteria. Applied and Environmental Microbiology 68, 3449–3454. Itoh, K., Tashiro, Y., Uobe, K., Kamagata, Y., Suyama, K., Yamamoto, H., 2004. Root nodule Bradyrhizobium spp. Harbor tfdA alpha and cadA, homologous with genes encoding 2,4-dichlorophenoxyacetic aciddegrading proteins. Applied and Environmental Microbiology 70, 2110–2118. 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. Applied and Environmental Microbiology 65, 4320–4328.

1002

L.J. Shaw, R.G. Burns / Soil Biology & Biochemistry 37 (2005) 995–1002

Kamagata, Y., Fulthorpe, R.R., Tamura, K., Takami, H., Forney, L.J., Tiedje, J.M., 1997. Pristine environments harbour a new group of oligotrophic 2,4-dichlorophenoxyacetic acid-degrading bacteria. Applied and Environmental Microbiology 63, 2266–2272. Kitagawa, W., Takami, S., Miyauchi, K., Masai, E., Kamagata, Y., Tiedje, J.M., Fukuda, M., 2002. Novel 2,4-dichlorophenoxyacetic acid degradation genes from oligotrophic Bradyrhizobium sp. strain HW13 isolated from a pristine environment. Journal of Bacteriology 184, 509–-518. Knaebel, D.B., Vestal, J.R., 1992. Effects of intact rhizosphere microbial communities on the mineralization of surfactants in surface soils. Canadian Journal of Microbiology 38, 643–653. Marschner, P., Yang, C.H., Lieberei, R., Crowley, D.E., 2001. Soil and plant specific effects on bacterial community composition in the rhizosphere. Soil Biology & Biochemistry 33, 1437–1445. Meharg, A.A., Killham, K., 1995. Loss of exudates from the roots of perennial ryegrass inoculated with a range of microorganisms. Plant and Soil 170, 345–349. Merlin, G., Lissolo, T., Morel, V., Rossel, D., Tarradellas, J., 1995. Precautions for routine use of INT-reductase activity for measuring biological activities in soil and sediments. Environmental Toxicology and Water Quality 10, 185–192. Nelson, D.W., Sommers, L.E., 1982. Total carbon, organic carbon and organic matter, in: Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis Part 2 - Chemical and Microbiological Properties. Soil Science Society of America, Wisconsin, pp. 539–579. Ogram, A., Xia, X., Farrow, F., Bollinger, J., 1995. Ecology of 2,4-D degradation in three Palouse silt loam soils, in: Huang, P.M., Berthelin, J., Bollag, J.-M., McGill, W.B., Page, A.L. (Eds.), Environmental Impact of Soil Component Interactions. CRC Press, London, pp. 253–265. Sandmann, E.R.I.C., Loos, M.A., 1984. Enumeration of 2,4-D-degrading microorganisms in soils and crop plant rhizospheres using indicator media: high populations associated with sugarcane (Saccharum officinarum). Chemosphere 13, 1073–1084.

Shaw, L.J., Burns, R.G., 2003. Biodegradation of organic pollutants in the rhizosphere. Advances in Applied Microbiology 53, 1–60. Shaw, L.J., Burns, R.G., 2004. Enhanced mineralisation of [U-14C]2,4dichlorophenoxyacetic acid in soil from the rhizosphere of Trifolium pratense. Applied and Environmental Microbiology 70, 4766–4774. Shaw, L.J., Burns, R.G., 2005. Rhizodeposition and the enhanced mineralization of 2,4-dichlorophenoxyacetic acid in soil from the Trifolium pratense rhizosphere. Environmental Microbiology 2005; (in press). Shaw, L.J., Beaton, Y., Sousa, S., Glover, L.A., Killham, K., Meharg, A.A., 2002. Mineralisation of 2,4-dichlorophenol and glucose placed into the same or different hydrological domains as a bacterial inoculant. Soil Biology & Biochemistry 34, 531–539. Siciliano, S.D., Germida, J.J., 1998. Mechanisms of phytoremediation: biochemical and ecological interactions between plants and bacteria. Environmental Reviews 6, 65–79. Singer, A.C., Crowley, D.E., Thompson, I.P., 2003. Secondary plant metabolites in phytoremediation and biotransformation. Trends in Biotechnology 21, 123–130. Smalla, K., Wieland, G., Buchner, A., Zock, A., Parzy, J., Kaiser, S., Roskot, N., Heuer, H., Berg, G., 2001. Bulk and rhizosphere soil bacterial communities studied by gradient gel electrophoresis: plantdependent enrichment and seasonal shifts revealed. Applied and Environmental Microbiology 67, 4742–4751. Trevors, J.T., 1984. Effect of substrate concentration, inorganic nitrogen, O2 concentration, temperature and pH on dehydrogenase activity in soil. Plant and Soil 77, 285–293. Van der Ploeg, J.R., Weiss, M.A., Saller, E., Nashimoto, H., Saito, N., Kertesz, M.A., Leisinger, T., 1996. Identification of sulfate starvationregulated genes in Escherichia coli: a gene cluster involved in the utilization of taurine as a sulfur source. Journal of Bacteriology 178, 5438–5446. von Mersi, W., Schinner, F., 1991. An improved and accurate method for determining the dehydrogenase activity of soils with iodonitrotetrazolium chloride. Biology and Fertility of Soils 11, 216–220.